1 files changed, 154 insertions, 22 deletions

diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
index b235d8da0602..5c1e178fc5f9 100644
--- a/drivers/mmc/host/mtk-sd.c
+++ b/drivers/mmc/host/mtk-sd.c
@@ -76,6 +76,7 @@
 #define MSDC_PATCH_BIT1  0xb4
 #define MSDC_PAD_TUNE    0xec
 #define PAD_DS_TUNE      0x188
+#define PAD_CMD_TUNE     0x18c
 #define EMMC50_CFG0      0x208
 
 /*--------------------------------------------------------------------------*/
@@ -211,13 +212,18 @@
 #define MSDC_PATCH_BIT_SPCPUSH    (0x1 << 29)	/* RW */
 #define MSDC_PATCH_BIT_DECRCTMO   (0x1 << 30)	/* RW */
 
+#define MSDC_PAD_TUNE_DATWRDLY	  (0x1f <<  0)	/* RW */
 #define MSDC_PAD_TUNE_DATRRDLY	  (0x1f <<  8)	/* RW */
 #define MSDC_PAD_TUNE_CMDRDLY	  (0x1f << 16)  /* RW */
+#define MSDC_PAD_TUNE_CMDRRDLY	  (0x1f << 22)	/* RW */
+#define MSDC_PAD_TUNE_CLKTDLY	  (0x1f << 27)  /* RW */
 
 #define PAD_DS_TUNE_DLY1	  (0x1f << 2)   /* RW */
 #define PAD_DS_TUNE_DLY2	  (0x1f << 7)   /* RW */
 #define PAD_DS_TUNE_DLY3	  (0x1f << 12)  /* RW */
 
+#define PAD_CMD_TUNE_RX_DLY3	  (0x1f << 1)  /* RW */
+
 #define EMMC50_CFG_PADCMD_LATCHCK (0x1 << 0)   /* RW */
 #define EMMC50_CFG_CRCSTS_EDGE    (0x1 << 3)   /* RW */
 #define EMMC50_CFG_CFCSTS_SEL     (0x1 << 4)   /* RW */
@@ -285,12 +291,14 @@ struct msdc_save_para {
 	u32 patch_bit0;
 	u32 patch_bit1;
 	u32 pad_ds_tune;
+	u32 pad_cmd_tune;
 	u32 emmc50_cfg0;
 };
 
 struct msdc_tune_para {
 	u32 iocon;
 	u32 pad_tune;
+	u32 pad_cmd_tune;
 };
 
 struct msdc_delay_phase {
@@ -332,6 +340,10 @@ struct msdc_host {
 	unsigned char timing;
 	bool vqmmc_enabled;
 	u32 hs400_ds_delay;
+	u32 hs200_cmd_int_delay; /* cmd internal delay for HS200/SDR104 */
+	u32 hs400_cmd_int_delay; /* cmd internal delay for HS400 */
+	bool hs400_cmd_resp_sel_rising;
+				 /* cmd response sample selection for HS400 */
 	bool hs400_mode;	/* current eMMC will run at hs400 mode */
 	struct msdc_save_para save_para; /* used when gate HCLK */
 	struct msdc_tune_para def_tune_para; /* default tune setting */
@@ -462,11 +474,9 @@ static void msdc_prepare_data(struct msdc_host *host, struct mmc_request *mrq)
 	struct mmc_data *data = mrq->data;
 
 	if (!(data->host_cookie & MSDC_PREPARE_FLAG)) {
-		bool read = (data->flags & MMC_DATA_READ) != 0;
-
 		data->host_cookie |= MSDC_PREPARE_FLAG;
 		data->sg_count = dma_map_sg(host->dev, data->sg, data->sg_len,
-					   read ? DMA_FROM_DEVICE : DMA_TO_DEVICE);
+					    mmc_get_dma_dir(data));
 	}
 }
 
@@ -478,10 +488,8 @@ static void msdc_unprepare_data(struct msdc_host *host, struct mmc_request *mrq)
 		return;
 
 	if (data->host_cookie & MSDC_PREPARE_FLAG) {
-		bool read = (data->flags & MMC_DATA_READ) != 0;
-
 		dma_unmap_sg(host->dev, data->sg, data->sg_len,
-			     read ? DMA_FROM_DEVICE : DMA_TO_DEVICE);
+			     mmc_get_dma_dir(data));
 		data->host_cookie &= ~MSDC_PREPARE_FLAG;
 	}
 }
@@ -601,8 +609,14 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 	} else {
 		writel(host->saved_tune_para.iocon, host->base + MSDC_IOCON);
 		writel(host->saved_tune_para.pad_tune, host->base + MSDC_PAD_TUNE);
+		writel(host->saved_tune_para.pad_cmd_tune,
+		       host->base + PAD_CMD_TUNE);
 	}
 
+	if (timing == MMC_TIMING_MMC_HS400)
+		sdr_set_field(host->base + PAD_CMD_TUNE,
+			      MSDC_PAD_TUNE_CMDRRDLY,
+			      host->hs400_cmd_int_delay);
 	dev_dbg(host->dev, "sclk: %d, timing: %d\n", host->sclk, timing);
 }
 
@@ -1303,7 +1317,7 @@ static struct msdc_delay_phase get_best_delay(struct msdc_host *host, u32 delay)
 			len_final = len;
 		}
 		start += len ? len : 1;
-		if (len >= 8 && start_final < 4)
+		if (len >= 12 && start_final < 4)
 			break;
 	}
 
@@ -1326,36 +1340,67 @@ static int msdc_tune_response(struct mmc_host *mmc, u32 opcode)
 	struct msdc_host *host = mmc_priv(mmc);
 	u32 rise_delay = 0, fall_delay = 0;
 	struct msdc_delay_phase final_rise_delay, final_fall_delay = { 0,};
+	struct msdc_delay_phase internal_delay_phase;
 	u8 final_delay, final_maxlen;
+	u32 internal_delay = 0;
 	int cmd_err;
-	int i;
+	int i, j;
+
+	if (mmc->ios.timing == MMC_TIMING_MMC_HS200 ||
+	    mmc->ios.timing == MMC_TIMING_UHS_SDR104)
+		sdr_set_field(host->base + MSDC_PAD_TUNE,
+			      MSDC_PAD_TUNE_CMDRRDLY,
+			      host->hs200_cmd_int_delay);
 
 	sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
 	for (i = 0 ; i < PAD_DELAY_MAX; i++) {
 		sdr_set_field(host->base + MSDC_PAD_TUNE,
 			      MSDC_PAD_TUNE_CMDRDLY, i);
-		mmc_send_tuning(mmc, opcode, &cmd_err);
-		if (!cmd_err)
-			rise_delay |= (1 << i);
+		/*
+		 * Using the same parameters, it may sometimes pass the test,
+		 * but sometimes it may fail. To make sure the parameters are
+		 * more stable, we test each set of parameters 3 times.
+		 */
+		for (j = 0; j < 3; j++) {
+			mmc_send_tuning(mmc, opcode, &cmd_err);
+			if (!cmd_err) {
+				rise_delay |= (1 << i);
+			} else {
+				rise_delay &= ~(1 << i);
+				break;
+			}
+		}
 	}
 	final_rise_delay = get_best_delay(host, rise_delay);
 	/* if rising edge has enough margin, then do not scan falling edge */
-	if (final_rise_delay.maxlen >= 10 ||
-	    (final_rise_delay.start == 0 && final_rise_delay.maxlen >= 4))
+	if (final_rise_delay.maxlen >= 12 && final_rise_delay.start < 4)
 		goto skip_fall;
 
 	sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
 	for (i = 0; i < PAD_DELAY_MAX; i++) {
 		sdr_set_field(host->base + MSDC_PAD_TUNE,
 			      MSDC_PAD_TUNE_CMDRDLY, i);
-		mmc_send_tuning(mmc, opcode, &cmd_err);
-		if (!cmd_err)
-			fall_delay |= (1 << i);
+		/*
+		 * Using the same parameters, it may sometimes pass the test,
+		 * but sometimes it may fail. To make sure the parameters are
+		 * more stable, we test each set of parameters 3 times.
+		 */
+		for (j = 0; j < 3; j++) {
+			mmc_send_tuning(mmc, opcode, &cmd_err);
+			if (!cmd_err) {
+				fall_delay |= (1 << i);
+			} else {
+				fall_delay &= ~(1 << i);
+				break;
+			}
+		}
 	}
 	final_fall_delay = get_best_delay(host, fall_delay);
 
 skip_fall:
 	final_maxlen = max(final_rise_delay.maxlen, final_fall_delay.maxlen);
+	if (final_fall_delay.maxlen >= 12 && final_fall_delay.start < 4)
+		final_maxlen = final_fall_delay.maxlen;
 	if (final_maxlen == final_rise_delay.maxlen) {
 		sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
 		sdr_set_field(host->base + MSDC_PAD_TUNE, MSDC_PAD_TUNE_CMDRDLY,
@@ -1367,7 +1412,71 @@ skip_fall:
 			      final_fall_delay.final_phase);
 		final_delay = final_fall_delay.final_phase;
 	}
+	if (host->hs200_cmd_int_delay)
+		goto skip_internal;
+
+	for (i = 0; i < PAD_DELAY_MAX; i++) {
+		sdr_set_field(host->base + MSDC_PAD_TUNE,
+			      MSDC_PAD_TUNE_CMDRRDLY, i);
+		mmc_send_tuning(mmc, opcode, &cmd_err);
+		if (!cmd_err)
+			internal_delay |= (1 << i);
+	}
+	dev_dbg(host->dev, "Final internal delay: 0x%x\n", internal_delay);
+	internal_delay_phase = get_best_delay(host, internal_delay);
+	sdr_set_field(host->base + MSDC_PAD_TUNE, MSDC_PAD_TUNE_CMDRRDLY,
+		      internal_delay_phase.final_phase);
+skip_internal:
+	dev_dbg(host->dev, "Final cmd pad delay: %x\n", final_delay);
+	return final_delay == 0xff ? -EIO : 0;
+}
+
+static int hs400_tune_response(struct mmc_host *mmc, u32 opcode)
+{
+	struct msdc_host *host = mmc_priv(mmc);
+	u32 cmd_delay = 0;
+	struct msdc_delay_phase final_cmd_delay = { 0,};
+	u8 final_delay;
+	int cmd_err;
+	int i, j;
+
+	/* select EMMC50 PAD CMD tune */
+	sdr_set_bits(host->base + PAD_CMD_TUNE, BIT(0));
+
+	if (mmc->ios.timing == MMC_TIMING_MMC_HS200 ||
+	    mmc->ios.timing == MMC_TIMING_UHS_SDR104)
+		sdr_set_field(host->base + MSDC_PAD_TUNE,
+			      MSDC_PAD_TUNE_CMDRRDLY,
+			      host->hs200_cmd_int_delay);
+
+	if (host->hs400_cmd_resp_sel_rising)
+		sdr_clr_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
+	else
+		sdr_set_bits(host->base + MSDC_IOCON, MSDC_IOCON_RSPL);
+	for (i = 0 ; i < PAD_DELAY_MAX; i++) {
+		sdr_set_field(host->base + PAD_CMD_TUNE,
+			      PAD_CMD_TUNE_RX_DLY3, i);
+		/*
+		 * Using the same parameters, it may sometimes pass the test,
+		 * but sometimes it may fail. To make sure the parameters are
+		 * more stable, we test each set of parameters 3 times.
+		 */
+		for (j = 0; j < 3; j++) {
+			mmc_send_tuning(mmc, opcode, &cmd_err);
+			if (!cmd_err) {
+				cmd_delay |= (1 << i);
+			} else {
+				cmd_delay &= ~(1 << i);
+				break;
+			}
+		}
+	}
+	final_cmd_delay = get_best_delay(host, cmd_delay);
+	sdr_set_field(host->base + PAD_CMD_TUNE, PAD_CMD_TUNE_RX_DLY3,
+		      final_cmd_delay.final_phase);
+	final_delay = final_cmd_delay.final_phase;
 
+	dev_dbg(host->dev, "Final cmd pad delay: %x\n", final_delay);
 	return final_delay == 0xff ? -EIO : 0;
 }
 
@@ -1390,7 +1499,7 @@ static int msdc_tune_data(struct mmc_host *mmc, u32 opcode)
 	}
 	final_rise_delay = get_best_delay(host, rise_delay);
 	/* if rising edge has enough margin, then do not scan falling edge */
-	if (final_rise_delay.maxlen >= 10 ||
+	if (final_rise_delay.maxlen >= 12 ||
 	    (final_rise_delay.start == 0 && final_rise_delay.maxlen >= 4))
 		goto skip_fall;
 
@@ -1423,6 +1532,7 @@ skip_fall:
 		final_delay = final_fall_delay.final_phase;
 	}
 
+	dev_dbg(host->dev, "Final data pad delay: %x\n", final_delay);
 	return final_delay == 0xff ? -EIO : 0;
 }
 
@@ -1431,7 +1541,10 @@ static int msdc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 	struct msdc_host *host = mmc_priv(mmc);
 	int ret;
 
-	ret = msdc_tune_response(mmc, opcode);
+	if (host->hs400_mode)
+		ret = hs400_tune_response(mmc, opcode);
+	else
+		ret = msdc_tune_response(mmc, opcode);
 	if (ret == -EIO) {
 		dev_err(host->dev, "Tune response fail!\n");
 		return ret;
@@ -1444,6 +1557,7 @@ static int msdc_execute_tuning(struct mmc_host *mmc, u32 opcode)
 
 	host->saved_tune_para.iocon = readl(host->base + MSDC_IOCON);
 	host->saved_tune_para.pad_tune = readl(host->base + MSDC_PAD_TUNE);
+	host->saved_tune_para.pad_cmd_tune = readl(host->base + PAD_CMD_TUNE);
 	return ret;
 }
 
@@ -1478,6 +1592,25 @@ static struct mmc_host_ops mt_msdc_ops = {
 	.hw_reset = msdc_hw_reset,
 };
 
+static void msdc_of_property_parse(struct platform_device *pdev,
+				   struct msdc_host *host)
+{
+	of_property_read_u32(pdev->dev.of_node, "hs400-ds-delay",
+			     &host->hs400_ds_delay);
+
+	of_property_read_u32(pdev->dev.of_node, "mediatek,hs200-cmd-int-delay",
+			     &host->hs200_cmd_int_delay);
+
+	of_property_read_u32(pdev->dev.of_node, "mediatek,hs400-cmd-int-delay",
+			     &host->hs400_cmd_int_delay);
+
+	if (of_property_read_bool(pdev->dev.of_node,
+				  "mediatek,hs400-cmd-resp-sel-rising"))
+		host->hs400_cmd_resp_sel_rising = true;
+	else
+		host->hs400_cmd_resp_sel_rising = false;
+}
+
 static int msdc_drv_probe(struct platform_device *pdev)
 {
 	struct mmc_host *mmc;
@@ -1549,10 +1682,7 @@ static int msdc_drv_probe(struct platform_device *pdev)
 		goto host_free;
 	}
 
-	if (!of_property_read_u32(pdev->dev.of_node, "hs400-ds-delay",
-				  &host->hs400_ds_delay))
-		dev_dbg(&pdev->dev, "hs400-ds-delay: %x\n",
-			host->hs400_ds_delay);
+	msdc_of_property_parse(pdev, host);
 
 	host->dev = &pdev->dev;
 	host->mmc = mmc;
@@ -1664,6 +1794,7 @@ static void msdc_save_reg(struct msdc_host *host)
 	host->save_para.patch_bit0 = readl(host->base + MSDC_PATCH_BIT);
 	host->save_para.patch_bit1 = readl(host->base + MSDC_PATCH_BIT1);
 	host->save_para.pad_ds_tune = readl(host->base + PAD_DS_TUNE);
+	host->save_para.pad_cmd_tune = readl(host->base + PAD_CMD_TUNE);
 	host->save_para.emmc50_cfg0 = readl(host->base + EMMC50_CFG0);
 }
 
@@ -1676,6 +1807,7 @@ static void msdc_restore_reg(struct msdc_host *host)
 	writel(host->save_para.patch_bit0, host->base + MSDC_PATCH_BIT);
 	writel(host->save_para.patch_bit1, host->base + MSDC_PATCH_BIT1);
 	writel(host->save_para.pad_ds_tune, host->base + PAD_DS_TUNE);
+	writel(host->save_para.pad_cmd_tune, host->base + PAD_CMD_TUNE);
 	writel(host->save_para.emmc50_cfg0, host->base + EMMC50_CFG0);
 }