Merge branch 'drm-next-4.8' of git://people.freedesktop.org/~agd5f/linux into drm-next

A few more patches for 4.8.  Mostly bug fixes and some prep work
for iceland powerplay support.  I have a couple polaris patches and
Edward's misc cleanups that require a merge with Linus'.  I don't know
if you are planning a merge anytime soon.

[airlied: fixed up endian vs 32-bit change in ppatomctrl]

* 'drm-next-4.8' of git://people.freedesktop.org/~agd5f/linux: (26 commits)
  drm/amdgpu: comment out unused defaults_bonaire_pro static const structures to fix the build
  drm/amdgpu: temporary comment out unused static const structures to fix the build
  drm/amdgpu: S3 resume fail on Polaris10
  drm/amd/powerplay: add pp_tables_get_response_times function in process pptables
  drm/amd/powerplay: fix the incorrect return value
  drm/amd/powerplay: add atomctrl_get_voltage_evv function in ppatomctrl
  drm/amdgpu: add new definitions into ppsmc.h for iceland
  drm/amd/powerplay: add SMU register macro for future use
  drm/amdgpu: add ucode_start_address into cgs_firmware_info
  drm/amdgpu: no need load microcode at sdma if powerplay is enabled
  drm/amdgpu: rename smumgr to smum for dpm
  drm/amdgpu: disable GFX PG on CZ/BR/ST
  drivers: gpu: drm: amd: powerplay: hwmgr: Remove unused variable
  drm/amdgpu: return -ENOSPC when running out of UVD handles
  drm/amdgpu: trace need_flush in grab_vm as well
  drm/amdgpu: always signal all fences
  drm/amdgpu: check flush fence context instead of same ring v2
  drm/radeon: support backlight control for UNIPHY3
  drm/amdgpu: support backlight control for UNIPHY3
  drm/amdgpu: remove usec timeout loop from IB tests
  ...

30 files changed, 469 insertions, 360 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 5556ce979199..ee95e950a19b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -752,6 +752,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 
 		if (!adev->pm.fw) {
 			switch (adev->asic_type) {
+			case CHIP_TOPAZ:
+				strcpy(fw_name, "amdgpu/topaz_smc.bin");
+				break;
 			case CHIP_TONGA:
 				strcpy(fw_name, "amdgpu/tonga_smc.bin");
 				break;
@@ -800,6 +803,7 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 
 		info->version = adev->pm.fw_version;
 		info->image_size = ucode_size;
+		info->ucode_start_address = ucode_start_address;
 		info->kptr = (void *)src;
 	}
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index d1558768cfb7..0b109aebfec6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -204,16 +204,25 @@ void amdgpu_fence_process(struct amdgpu_ring *ring)
 	if (seq != ring->fence_drv.sync_seq)
 		amdgpu_fence_schedule_fallback(ring);
 
-	while (last_seq != seq) {
+	if (unlikely(seq == last_seq))
+		return;
+
+	last_seq &= drv->num_fences_mask;
+	seq &= drv->num_fences_mask;
+
+	do {
 		struct fence *fence, **ptr;
 
-		ptr = &drv->fences[++last_seq & drv->num_fences_mask];
+		++last_seq;
+		last_seq &= drv->num_fences_mask;
+		ptr = &drv->fences[last_seq];
 
 		/* There is always exactly one thread signaling this fence slot */
 		fence = rcu_dereference_protected(*ptr, 1);
 		RCU_INIT_POINTER(*ptr, NULL);
 
-		BUG_ON(!fence);
+		if (!fence)
+			continue;
 
 		r = fence_signal(fence);
 		if (!r)
@@ -222,7 +231,7 @@ void amdgpu_fence_process(struct amdgpu_ring *ring)
 			BUG();
 
 		fence_put(fence);
-	}
+	} while (last_seq != seq);
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index 46c3097c5224..428ebf3a4387 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -122,7 +122,6 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	bool skip_preamble, need_ctx_switch;
 	unsigned patch_offset = ~0;
 	struct amdgpu_vm *vm;
-	struct fence *hwf;
 	uint64_t ctx;
 
 	unsigned i;
@@ -190,7 +189,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	if (ring->funcs->emit_hdp_invalidate)
 		amdgpu_ring_emit_hdp_invalidate(ring);
 
-	r = amdgpu_fence_emit(ring, &hwf);
+	r = amdgpu_fence_emit(ring, f);
 	if (r) {
 		dev_err(adev->dev, "failed to emit fence (%d)\n", r);
 		if (job && job->vm_id)
@@ -205,9 +204,6 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 				       AMDGPU_FENCE_FLAG_64BIT);
 	}
 
-	if (f)
-		*f = fence_get(hwf);
-
 	if (patch_offset != ~0 && ring->funcs->patch_cond_exec)
 		amdgpu_ring_patch_cond_exec(ring, patch_offset);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
index aaee0c8f6731..6674d40eb3ab 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
@@ -172,15 +172,13 @@ static struct fence *amdgpu_job_run(struct amd_sched_job *sched_job)
 	trace_amdgpu_sched_run_job(job);
 	r = amdgpu_ib_schedule(job->ring, job->num_ibs, job->ibs,
 			       job->sync.last_vm_update, job, &fence);
-	if (r) {
+	if (r)
 		DRM_ERROR("Error scheduling IBs (%d)\n", r);
-		goto err;
-	}
 
-err:
 	/* if gpu reset, hw fence will be replaced here */
 	fence_put(job->fence);
-	job->fence = fence;
+	job->fence = fence_get(fence);
+	amdgpu_job_free_resources(job);
 	return fence;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
index 499803f3ce3a..0d8d65eb46cd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
@@ -149,24 +149,26 @@ TRACE_EVENT(amdgpu_sched_run_job,
 
 
 TRACE_EVENT(amdgpu_vm_grab_id,
-	    TP_PROTO(struct amdgpu_vm *vm, int ring, unsigned vmid,
-		     uint64_t pd_addr),
-	    TP_ARGS(vm, ring, vmid, pd_addr),
+	    TP_PROTO(struct amdgpu_vm *vm, int ring, struct amdgpu_job *job),
+	    TP_ARGS(vm, ring, job),
 	    TP_STRUCT__entry(
 			     __field(struct amdgpu_vm *, vm)
 			     __field(u32, ring)
 			     __field(u32, vmid)
 			     __field(u64, pd_addr)
+			     __field(u32, needs_flush)
 			     ),
 
 	    TP_fast_assign(
 			   __entry->vm = vm;
 			   __entry->ring = ring;
-			   __entry->vmid = vmid;
-			   __entry->pd_addr = pd_addr;
+			   __entry->vmid = job->vm_id;
+			   __entry->pd_addr = job->vm_pd_addr;
+			   __entry->needs_flush = job->vm_needs_flush;
 			   ),
-	    TP_printk("vm=%p, ring=%u, id=%u, pd_addr=%010Lx", __entry->vm,
-		      __entry->ring, __entry->vmid, __entry->pd_addr)
+	    TP_printk("vm=%p, ring=%u, id=%u, pd_addr=%010Lx needs_flush=%u",
+		      __entry->vm, __entry->ring, __entry->vmid,
+		      __entry->pd_addr, __entry->needs_flush)
 );
 
 TRACE_EVENT(amdgpu_vm_bo_map,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index d9c88d13f8db..a46a64c125d1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -40,7 +40,7 @@
 #include "uvd/uvd_4_2_d.h"
 
 /* 1 second timeout */
-#define UVD_IDLE_TIMEOUT_MS	1000
+#define UVD_IDLE_TIMEOUT	msecs_to_jiffies(1000)
 /* Polaris10/11 firmware version */
 #define FW_1_66_16 ((1 << 24) | (66 << 16) | (16 << 8))
 
@@ -662,7 +662,7 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 		}
 
 		DRM_ERROR("No more free UVD handles!\n");
-		return -EINVAL;
+		return -ENOSPC;
 
 	case 1:
 		/* it's a decode msg, calc buffer sizes */
@@ -968,7 +968,7 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 
 	if (direct) {
 		r = amdgpu_ib_schedule(ring, 1, ib, NULL, NULL, &f);
-		job->fence = f;
+		job->fence = fence_get(f);
 		if (r)
 			goto err_free;
 
@@ -1114,8 +1114,7 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 			amdgpu_asic_set_uvd_clocks(adev, 0, 0);
 		}
 	} else {
-		schedule_delayed_work(&adev->uvd.idle_work,
-				      msecs_to_jiffies(UVD_IDLE_TIMEOUT_MS));
+		schedule_delayed_work(&adev->uvd.idle_work, UVD_IDLE_TIMEOUT);
 	}
 }
 
@@ -1123,7 +1122,7 @@ static void amdgpu_uvd_note_usage(struct amdgpu_device *adev)
 {
 	bool set_clocks = !cancel_delayed_work_sync(&adev->uvd.idle_work);
 	set_clocks &= schedule_delayed_work(&adev->uvd.idle_work,
-					    msecs_to_jiffies(UVD_IDLE_TIMEOUT_MS));
+					    UVD_IDLE_TIMEOUT);
 
 	if (set_clocks) {
 		if (adev->pm.dpm_enabled) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index 875626a2eccb..aeeeb72ebbc4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -36,7 +36,7 @@
 #include "cikd.h"
 
 /* 1 second timeout */
-#define VCE_IDLE_TIMEOUT_MS	1000
+#define VCE_IDLE_TIMEOUT	msecs_to_jiffies(1000)
 
 /* Firmware Names */
 #ifdef CONFIG_DRM_AMDGPU_CIK
@@ -310,8 +310,7 @@ static void amdgpu_vce_idle_work_handler(struct work_struct *work)
 			amdgpu_asic_set_vce_clocks(adev, 0, 0);
 		}
 	} else {
-		schedule_delayed_work(&adev->vce.idle_work,
-				      msecs_to_jiffies(VCE_IDLE_TIMEOUT_MS));
+		schedule_delayed_work(&adev->vce.idle_work, VCE_IDLE_TIMEOUT);
 	}
 }
 
@@ -324,17 +323,12 @@ static void amdgpu_vce_idle_work_handler(struct work_struct *work)
  */
 static void amdgpu_vce_note_usage(struct amdgpu_device *adev)
 {
-	bool streams_changed = false;
 	bool set_clocks = !cancel_delayed_work_sync(&adev->vce.idle_work);
-	set_clocks &= schedule_delayed_work(&adev->vce.idle_work,
-					    msecs_to_jiffies(VCE_IDLE_TIMEOUT_MS));
 
-	if (adev->pm.dpm_enabled) {
-		/* XXX figure out if the streams changed */
-		streams_changed = false;
-	}
+	set_clocks &= schedule_delayed_work(&adev->vce.idle_work,
+					    VCE_IDLE_TIMEOUT);
 
-	if (set_clocks || streams_changed) {
+	if (set_clocks) {
 		if (adev->pm.dpm_enabled) {
 			amdgpu_dpm_enable_vce(adev, true);
 		} else {
@@ -357,6 +351,7 @@ void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
 	int i, r;
 	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i) {
 		uint32_t handle = atomic_read(&adev->vce.handles[i]);
+
 		if (!handle || adev->vce.filp[i] != filp)
 			continue;
 
@@ -437,7 +432,7 @@ int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 		ib->ptr[i] = 0x0;
 
 	r = amdgpu_ib_schedule(ring, 1, ib, NULL, NULL, &f);
-	job->fence = f;
+	job->fence = fence_get(f);
 	if (r)
 		goto err;
 
@@ -499,7 +494,7 @@ int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 
 	if (direct) {
 		r = amdgpu_ib_schedule(ring, 1, ib, NULL, NULL, &f);
-		job->fence = f;
+		job->fence = fence_get(f);
 		if (r)
 			goto err;
 
@@ -580,12 +575,10 @@ static int amdgpu_vce_cs_reloc(struct amdgpu_cs_parser *p, uint32_t ib_idx,
  * we we don't have another free session index.
  */
 static int amdgpu_vce_validate_handle(struct amdgpu_cs_parser *p,
-				      uint32_t handle, bool *allocated)
+				      uint32_t handle, uint32_t *allocated)
 {
 	unsigned i;
 
-	*allocated = false;
-
 	/* validate the handle */
 	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i) {
 		if (atomic_read(&p->adev->vce.handles[i]) == handle) {
@@ -602,7 +595,7 @@ static int amdgpu_vce_validate_handle(struct amdgpu_cs_parser *p,
 		if (!atomic_cmpxchg(&p->adev->vce.handles[i], 0, handle)) {
 			p->adev->vce.filp[i] = p->filp;
 			p->adev->vce.img_size[i] = 0;
-			*allocated = true;
+			*allocated |= 1 << i;
 			return i;
 		}
 	}
@@ -622,9 +615,9 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 	struct amdgpu_ib *ib = &p->job->ibs[ib_idx];
 	unsigned fb_idx = 0, bs_idx = 0;
 	int session_idx = -1;
-	bool destroyed = false;
-	bool created = false;
-	bool allocated = false;
+	uint32_t destroyed = 0;
+	uint32_t created = 0;
+	uint32_t allocated = 0;
 	uint32_t tmp, handle = 0;
 	uint32_t *size = &tmp;
 	int i, r = 0, idx = 0;
@@ -641,30 +634,30 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 			goto out;
 		}
 
-		if (destroyed) {
-			DRM_ERROR("No other command allowed after destroy!\n");
-			r = -EINVAL;
-			goto out;
-		}
-
 		switch (cmd) {
-		case 0x00000001: // session
+		case 0x00000001: /* session */
 			handle = amdgpu_get_ib_value(p, ib_idx, idx + 2);
 			session_idx = amdgpu_vce_validate_handle(p, handle,
 								 &allocated);
-			if (session_idx < 0)
-				return session_idx;
+			if (session_idx < 0) {
+				r = session_idx;
+				goto out;
+			}
 			size = &p->adev->vce.img_size[session_idx];
 			break;
 
-		case 0x00000002: // task info
+		case 0x00000002: /* task info */
 			fb_idx = amdgpu_get_ib_value(p, ib_idx, idx + 6);
 			bs_idx = amdgpu_get_ib_value(p, ib_idx, idx + 7);
 			break;
 
-		case 0x01000001: // create
-			created = true;
-			if (!allocated) {
+		case 0x01000001: /* create */
+			created |= 1 << session_idx;
+			if (destroyed & (1 << session_idx)) {
+				destroyed &= ~(1 << session_idx);
+				allocated |= 1 << session_idx;
+
+			} else if (!(allocated & (1 << session_idx))) {
 				DRM_ERROR("Handle already in use!\n");
 				r = -EINVAL;
 				goto out;
@@ -675,16 +668,16 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 				8 * 3 / 2;
 			break;
 
-		case 0x04000001: // config extension
-		case 0x04000002: // pic control
-		case 0x04000005: // rate control
-		case 0x04000007: // motion estimation
-		case 0x04000008: // rdo
-		case 0x04000009: // vui
-		case 0x05000002: // auxiliary buffer
+		case 0x04000001: /* config extension */
+		case 0x04000002: /* pic control */
+		case 0x04000005: /* rate control */
+		case 0x04000007: /* motion estimation */
+		case 0x04000008: /* rdo */
+		case 0x04000009: /* vui */
+		case 0x05000002: /* auxiliary buffer */
 			break;
 
-		case 0x03000001: // encode
+		case 0x03000001: /* encode */
 			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 10, idx + 9,
 						*size, 0);
 			if (r)
@@ -696,18 +689,18 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 				goto out;
 			break;
 
-		case 0x02000001: // destroy
-			destroyed = true;
+		case 0x02000001: /* destroy */
+			destroyed |= 1 << session_idx;
 			break;
 
-		case 0x05000001: // context buffer
+		case 0x05000001: /* context buffer */
 			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3, idx + 2,
 						*size * 2, 0);
 			if (r)
 				goto out;
 			break;
 
-		case 0x05000004: // video bitstream buffer
+		case 0x05000004: /* video bitstream buffer */
 			tmp = amdgpu_get_ib_value(p, ib_idx, idx + 4);
 			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3, idx + 2,
 						tmp, bs_idx);
@@ -715,7 +708,7 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 				goto out;
 			break;
 
-		case 0x05000005: // feedback buffer
+		case 0x05000005: /* feedback buffer */
 			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3, idx + 2,
 						4096, fb_idx);
 			if (r)
@@ -737,21 +730,24 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 		idx += len / 4;
 	}
 
-	if (allocated && !created) {
+	if (allocated & ~created) {
 		DRM_ERROR("New session without create command!\n");
 		r = -ENOENT;
 	}
 
 out:
-	if ((!r && destroyed) || (r && allocated)) {
-		/*
-		 * IB contains a destroy msg or we have allocated an
-		 * handle and got an error, anyway free the handle
-		 */
-		for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i)
-			atomic_cmpxchg(&p->adev->vce.handles[i], handle, 0);
+	if (!r) {
+		/* No error, free all destroyed handle slots */
+		tmp = destroyed;
+	} else {
+		/* Error during parsing, free all allocated handle slots */
+		tmp = allocated;
 	}
 
+	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i)
+		if (tmp & (1 << i))
+			atomic_set(&p->adev->vce.handles[i], 0);
+
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 2f8496d48c94..8e642fc48df4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -195,6 +195,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 		      struct amdgpu_job *job)
 {
 	struct amdgpu_device *adev = ring->adev;
+	uint64_t fence_context = adev->fence_context + ring->idx;
 	struct fence *updates = sync->last_vm_update;
 	struct amdgpu_vm_id *id, *idle;
 	struct fence **fences;
@@ -254,7 +255,6 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 	i = ring->idx;
 	do {
 		struct fence *flushed;
-		bool same_ring = ring->idx == i;
 
 		id = vm->ids[i++];
 		if (i == AMDGPU_MAX_RINGS)
@@ -272,8 +272,11 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 		if (job->vm_pd_addr != id->pd_gpu_addr)
 			continue;
 
-		if (!same_ring &&
-		    (!id->last_flush || !fence_is_signaled(id->last_flush)))
+		if (!id->last_flush)
+			continue;
+
+		if (id->last_flush->context != fence_context &&
+		    !fence_is_signaled(id->last_flush))
 			continue;
 
 		flushed  = id->flushed_updates;
@@ -294,7 +297,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 
 		job->vm_id = id - adev->vm_manager.ids;
 		job->vm_needs_flush = false;
-		trace_amdgpu_vm_grab_id(vm, ring->idx, job->vm_id, job->vm_pd_addr);
+		trace_amdgpu_vm_grab_id(vm, ring->idx, job);
 
 		mutex_unlock(&adev->vm_manager.lock);
 		return 0;
@@ -325,7 +328,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 	vm->ids[ring->idx] = id;
 
 	job->vm_id = id - adev->vm_manager.ids;
-	trace_amdgpu_vm_grab_id(vm, ring->idx, job->vm_id, job->vm_pd_addr);
+	trace_amdgpu_vm_grab_id(vm, ring->idx, job);
 
 error:
 	mutex_unlock(&adev->vm_manager.lock);
diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c b/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c
index 48b6bd671cda..c32eca26155c 100644
--- a/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c
+++ b/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c
@@ -98,6 +98,7 @@ amdgpu_atombios_encoder_set_backlight_level(struct amdgpu_encoder *amdgpu_encode
 		case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
 		case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
 		case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+		case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
 			if (dig->backlight_level == 0)
 				amdgpu_atombios_encoder_setup_dig_transmitter(encoder,
 								       ATOM_TRANSMITTER_ACTION_LCD_BLOFF, 0, 0);
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index 5c33ed862695..573bc973d692 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -86,12 +86,14 @@ static const struct ci_pt_defaults defaults_bonaire_xt =
 	{ 0x17C, 0x172, 0x180, 0x1BC, 0x1B3, 0x1BD, 0x206, 0x200, 0x203, 0x25D, 0x25A, 0x255, 0x2C3, 0x2C5, 0x2B4 }
 };
 
+#if 0
 static const struct ci_pt_defaults defaults_bonaire_pro =
 {
 	1, 0xF, 0xFD, 0x19, 5, 45, 0, 0x65062,
 	{ 0x8C,  0x23F, 0x244, 0xA6,  0x83,  0x85,  0x86,  0x86,  0x83,  0xDB,  0xDB,  0xDA,  0x67,  0x60,  0x5F  },
 	{ 0x187, 0x193, 0x193, 0x1C7, 0x1D1, 0x1D1, 0x210, 0x219, 0x219, 0x266, 0x26C, 0x26C, 0x2C9, 0x2CB, 0x2CB }
 };
+#endif
 
 static const struct ci_pt_defaults defaults_saturn_xt =
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
index 46aca16a40aa..6507a7ee75e4 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
@@ -622,7 +622,6 @@ static int cik_sdma_ring_test_ib(struct amdgpu_ring *ring)
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_ib ib;
 	struct fence *f = NULL;
-	unsigned i;
 	unsigned index;
 	int r;
 	u32 tmp = 0;
@@ -644,7 +643,8 @@ static int cik_sdma_ring_test_ib(struct amdgpu_ring *ring)
 		goto err0;
 	}
 
-	ib.ptr[0] = SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
+	ib.ptr[0] = SDMA_PACKET(SDMA_OPCODE_WRITE,
+				SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
 	ib.ptr[1] = lower_32_bits(gpu_addr);
 	ib.ptr[2] = upper_32_bits(gpu_addr);
 	ib.ptr[3] = 1;
@@ -659,23 +659,15 @@ static int cik_sdma_ring_test_ib(struct amdgpu_ring *ring)
 		DRM_ERROR("amdgpu: fence wait failed (%d).\n", r);
 		goto err1;
 	}
-	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = le32_to_cpu(adev->wb.wb[index]);
-		if (tmp == 0xDEADBEEF)
-			break;
-		DRM_UDELAY(1);
-	}
-	if (i < adev->usec_timeout) {
-		DRM_INFO("ib test on ring %d succeeded in %u usecs\n",
-			 ring->idx, i);
-		goto err1;
+	tmp = le32_to_cpu(adev->wb.wb[index]);
+	if (tmp == 0xDEADBEEF) {
+		DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
 	} else {
 		DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp);
 		r = -EINVAL;
 	}
 
 err1:
-	fence_put(f);
 	amdgpu_ib_free(adev, &ib, NULL);
 	fence_put(f);
 err0:
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
index f6bd9465dbdc..1ac5ad01bf58 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -2112,7 +2112,6 @@ static int gfx_v7_0_ring_test_ib(struct amdgpu_ring *ring)
 	struct fence *f = NULL;
 	uint32_t scratch;
 	uint32_t tmp = 0;
-	unsigned i;
 	int r;
 
 	r = amdgpu_gfx_scratch_get(adev, &scratch);
@@ -2141,16 +2140,9 @@ static int gfx_v7_0_ring_test_ib(struct amdgpu_ring *ring)
 		DRM_ERROR("amdgpu: fence wait failed (%d).\n", r);
 		goto err2;
 	}
-	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = RREG32(scratch);
-		if (tmp == 0xDEADBEEF)
-			break;
-		DRM_UDELAY(1);
-	}
-	if (i < adev->usec_timeout) {
-		DRM_INFO("ib test on ring %d succeeded in %u usecs\n",
-			 ring->idx, i);
-		goto err2;
+	tmp = RREG32(scratch);
+	if (tmp == 0xDEADBEEF) {
+		DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
 	} else {
 		DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n",
 			  scratch, tmp);
@@ -2158,7 +2150,6 @@ static int gfx_v7_0_ring_test_ib(struct amdgpu_ring *ring)
 	}
 
 err2:
-	fence_put(f);
 	amdgpu_ib_free(adev, &ib, NULL);
 	fence_put(f);
 err1:
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index aabaf2f92575..d97b962bc3de 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -794,7 +794,6 @@ static int gfx_v8_0_ring_test_ib(struct amdgpu_ring *ring)
 	struct fence *f = NULL;
 	uint32_t scratch;
 	uint32_t tmp = 0;
-	unsigned i;
 	int r;
 
 	r = amdgpu_gfx_scratch_get(adev, &scratch);
@@ -823,23 +822,15 @@ static int gfx_v8_0_ring_test_ib(struct amdgpu_ring *ring)
 		DRM_ERROR("amdgpu: fence wait failed (%d).\n", r);
 		goto err2;
 	}
-	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = RREG32(scratch);
-		if (tmp == 0xDEADBEEF)
-			break;
-		DRM_UDELAY(1);
-	}
-	if (i < adev->usec_timeout) {
-		DRM_INFO("ib test on ring %d succeeded in %u usecs\n",
-			 ring->idx, i);
-		goto err2;
+	tmp = RREG32(scratch);
+	if (tmp == 0xDEADBEEF) {
+		DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
 	} else {
 		DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n",
 			  scratch, tmp);
 		r = -EINVAL;
 	}
 err2:
-	fence_put(f);
 	amdgpu_ib_free(adev, &ib, NULL);
 	fence_put(f);
 err1:
@@ -1729,7 +1720,6 @@ static int gfx_v8_0_do_edc_gpr_workarounds(struct amdgpu_device *adev)
 		RREG32(sec_ded_counter_registers[i]);
 
 fail:
-	fence_put(f);
 	amdgpu_ib_free(adev, &ib, NULL);
 	fence_put(f);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c b/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c
index 825ccd63f2dc..2f078ad6095c 100644
--- a/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c
@@ -24,7 +24,7 @@
 #include <linux/firmware.h>
 #include "drmP.h"
 #include "amdgpu.h"
-#include "iceland_smumgr.h"
+#include "iceland_smum.h"
 
 MODULE_FIRMWARE("amdgpu/topaz_smc.bin");
 
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_smc.c b/drivers/gpu/drm/amd/amdgpu/iceland_smc.c
index 52ee08193295..528571285e69 100644
--- a/drivers/gpu/drm/amd/amdgpu/iceland_smc.c
+++ b/drivers/gpu/drm/amd/amdgpu/iceland_smc.c
@@ -25,7 +25,7 @@
 #include "drmP.h"
 #include "amdgpu.h"
 #include "ppsmc.h"
-#include "iceland_smumgr.h"
+#include "iceland_smum.h"
 #include "smu_ucode_xfer_vi.h"
 #include "amdgpu_ucode.h"
 
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_smumgr.h b/drivers/gpu/drm/amd/amdgpu/iceland_smum.h
index 1e0769e110fa..1e0769e110fa 100644
--- a/drivers/gpu/drm/amd/amdgpu/iceland_smumgr.h
+++ b/drivers/gpu/drm/amd/amdgpu/iceland_smum.h
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
index 5a0e245771ce..a845e883f5fa 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -191,6 +191,7 @@ static void sumo_construct_vid_mapping_table(struct amdgpu_device *adev,
 	vid_mapping_table->num_entries = i;
 }
 
+#if 0
 static const struct kv_lcac_config_values sx_local_cac_cfg_kv[] =
 {
 	{  0,       4,        1    },
@@ -289,6 +290,7 @@ static const struct kv_lcac_config_reg cpl_cac_config_reg[] =
 {
 	{ 0xc0400d80, 0x003e0000, 17, 0x3fc00000, 22, 0x0001fffe, 1, 0x00000001, 0 }
 };
+#endif
 
 static const struct kv_pt_config_reg didt_config_kv[] =
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/ppsmc.h b/drivers/gpu/drm/amd/amdgpu/ppsmc.h
index 7837f2ecc357..8463245f424f 100644
--- a/drivers/gpu/drm/amd/amdgpu/ppsmc.h
+++ b/drivers/gpu/drm/amd/amdgpu/ppsmc.h
@@ -90,7 +90,9 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_StartFanControl               ((uint8_t)0x5B)
 #define PPSMC_StopFanControl                ((uint8_t)0x5C)
 #define PPSMC_MSG_NoDisplay                 ((uint8_t)0x5D)
+#define PPSMC_NoDisplay                     ((uint8_t)0x5D)
 #define PPSMC_MSG_HasDisplay                ((uint8_t)0x5E)
+#define PPSMC_HasDisplay                    ((uint8_t)0x5E)
 #define PPSMC_MSG_UVDPowerOFF               ((uint8_t)0x60)
 #define PPSMC_MSG_UVDPowerON                ((uint8_t)0x61)
 #define PPSMC_MSG_EnableULV                 ((uint8_t)0x62)
@@ -108,6 +110,7 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_MSG_DisableDTE                ((uint8_t)0x88)
 #define PPSMC_MSG_ThrottleOVRDSCLKDS        ((uint8_t)0x96)
 #define PPSMC_MSG_CancelThrottleOVRDSCLKDS  ((uint8_t)0x97)
+#define PPSMC_MSG_EnableACDCGPIOInterrupt   ((uint16_t) 0x149)
 
 /* CI/KV/KB */
 #define PPSMC_MSG_UVDDPM_SetEnabledMask       ((uint16_t) 0x12D)
@@ -161,6 +164,7 @@ typedef uint8_t PPSMC_Result;
 #define PPSMC_MSG_MASTER_DeepSleep_OFF        ((uint16_t) 0x190)
 #define PPSMC_MSG_Remove_DC_Clamp             ((uint16_t) 0x191)
 #define PPSMC_MSG_SetFanPwmMax                ((uint16_t) 0x19A)
+#define PPSMC_MSG_SetFanRpmMax                ((uint16_t) 0x205)
 
 #define PPSMC_MSG_ENABLE_THERMAL_DPM          ((uint16_t) 0x19C)
 #define PPSMC_MSG_DISABLE_THERMAL_DPM         ((uint16_t) 0x19D)
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
index ac3730a6e49f..0111d153411b 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
@@ -567,19 +567,21 @@ static int sdma_v2_4_start(struct amdgpu_device *adev)
 {
 	int r;
 
-	if (!adev->firmware.smu_load) {
-		r = sdma_v2_4_load_microcode(adev);
-		if (r)
-			return r;
-	} else {
-		r = adev->smu.smumgr_funcs->check_fw_load_finish(adev,
-						AMDGPU_UCODE_ID_SDMA0);
-		if (r)
-			return -EINVAL;
-		r = adev->smu.smumgr_funcs->check_fw_load_finish(adev,
-						AMDGPU_UCODE_ID_SDMA1);
-		if (r)
-			return -EINVAL;
+	if (!adev->pp_enabled) {
+		if (!adev->firmware.smu_load) {
+			r = sdma_v2_4_load_microcode(adev);
+			if (r)
+				return r;
+		} else {
+			r = adev->smu.smumgr_funcs->check_fw_load_finish(adev,
+							AMDGPU_UCODE_ID_SDMA0);
+			if (r)
+				return -EINVAL;
+			r = adev->smu.smumgr_funcs->check_fw_load_finish(adev,
+							AMDGPU_UCODE_ID_SDMA1);
+			if (r)
+				return -EINVAL;
+		}
 	}
 
 	/* halt the engine before programing */
@@ -671,7 +673,6 @@ static int sdma_v2_4_ring_test_ib(struct amdgpu_ring *ring)
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_ib ib;
 	struct fence *f = NULL;
-	unsigned i;
 	unsigned index;
 	int r;
 	u32 tmp = 0;
@@ -713,23 +714,15 @@ static int sdma_v2_4_ring_test_ib(struct amdgpu_ring *ring)
 		DRM_ERROR("amdgpu: fence wait failed (%d).\n", r);
 		goto err1;
 	}
-	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = le32_to_cpu(adev->wb.wb[index]);
-		if (tmp == 0xDEADBEEF)
-			break;
-		DRM_UDELAY(1);
-	}
-	if (i < adev->usec_timeout) {
-		DRM_INFO("ib test on ring %d succeeded in %u usecs\n",
-			 ring->idx, i);
-		goto err1;
+	tmp = le32_to_cpu(adev->wb.wb[index]);
+	if (tmp == 0xDEADBEEF) {
+		DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
 	} else {
 		DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp);
 		r = -EINVAL;
 	}
 
 err1:
-	fence_put(f);
 	amdgpu_ib_free(adev, &ib, NULL);
 	fence_put(f);
 err0:
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
index f00db6f4c04c..2b8ce5808a14 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -901,7 +901,6 @@ static int sdma_v3_0_ring_test_ib(struct amdgpu_ring *ring)
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_ib ib;
 	struct fence *f = NULL;
-	unsigned i;
 	unsigned index;
 	int r;
 	u32 tmp = 0;
@@ -943,22 +942,14 @@ static int sdma_v3_0_ring_test_ib(struct amdgpu_ring *ring)
 		DRM_ERROR("amdgpu: fence wait failed (%d).\n", r);
 		goto err1;
 	}
-	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = le32_to_cpu(adev->wb.wb[index]);
-		if (tmp == 0xDEADBEEF)
-			break;
-		DRM_UDELAY(1);
-	}
-	if (i < adev->usec_timeout) {
-		DRM_INFO("ib test on ring %d succeeded in %u usecs\n",
-			 ring->idx, i);
-		goto err1;
+	tmp = le32_to_cpu(adev->wb.wb[index]);
+	if (tmp == 0xDEADBEEF) {
+		DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
 	} else {
 		DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp);
 		r = -EINVAL;
 	}
 err1:
-	fence_put(f);
 	amdgpu_ib_free(adev, &ib, NULL);
 	fence_put(f);
 err0:
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 30e8099e94c5..d7b8da433fe2 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -43,6 +43,7 @@
 #define mmVCE_LMI_VCPU_CACHE_40BIT_BAR0	0x8616
 #define mmVCE_LMI_VCPU_CACHE_40BIT_BAR1	0x8617
 #define mmVCE_LMI_VCPU_CACHE_40BIT_BAR2	0x8618
+#define VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK	0x02
 
 #define VCE_V3_0_FW_SIZE	(384 * 1024)
 #define VCE_V3_0_STACK_SIZE	(64 * 1024)
@@ -51,6 +52,7 @@
 static void vce_v3_0_mc_resume(struct amdgpu_device *adev, int idx);
 static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev);
 static void vce_v3_0_set_irq_funcs(struct amdgpu_device *adev);
+static int vce_v3_0_wait_for_idle(void *handle);
 
 /**
  * vce_v3_0_ring_get_rptr - get read pointer
@@ -205,6 +207,32 @@ static void vce_v3_0_set_vce_sw_clock_gating(struct amdgpu_device *adev,
 	vce_v3_0_override_vce_clock_gating(adev, false);
 }
 
+static int vce_v3_0_firmware_loaded(struct amdgpu_device *adev)
+{
+	int i, j;
+	uint32_t status = 0;
+
+	for (i = 0; i < 10; ++i) {
+		for (j = 0; j < 100; ++j) {
+			status = RREG32(mmVCE_STATUS);
+			if (status & VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK)
+				return 0;
+			mdelay(10);
+		}
+
+		DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
+		WREG32_P(mmVCE_SOFT_RESET,
+			VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
+			~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+		mdelay(10);
+		WREG32_P(mmVCE_SOFT_RESET, 0,
+			~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+		mdelay(10);
+	}
+
+	return -ETIMEDOUT;
+}
+
 /**
  * vce_v3_0_start - start VCE block
  *
@@ -215,11 +243,24 @@ static void vce_v3_0_set_vce_sw_clock_gating(struct amdgpu_device *adev,
 static int vce_v3_0_start(struct amdgpu_device *adev)
 {
 	struct amdgpu_ring *ring;
-	int idx, i, j, r;
+	int idx, r;
+
+	ring = &adev->vce.ring[0];
+	WREG32(mmVCE_RB_RPTR, ring->wptr);
+	WREG32(mmVCE_RB_WPTR, ring->wptr);
+	WREG32(mmVCE_RB_BASE_LO, ring->gpu_addr);
+	WREG32(mmVCE_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+	WREG32(mmVCE_RB_SIZE, ring->ring_size / 4);
+
+	ring = &adev->vce.ring[1];
+	WREG32(mmVCE_RB_RPTR2, ring->wptr);
+	WREG32(mmVCE_RB_WPTR2, ring->wptr);
+	WREG32(mmVCE_RB_BASE_LO2, ring->gpu_addr);
+	WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+	WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4);
 
 	mutex_lock(&adev->grbm_idx_mutex);
 	for (idx = 0; idx < 2; ++idx) {
-
 		if (adev->vce.harvest_config & (1 << idx))
 			continue;
 
@@ -233,48 +274,24 @@ static int vce_v3_0_start(struct amdgpu_device *adev)
 
 		vce_v3_0_mc_resume(adev, idx);
 
-		/* set BUSY flag */
-		WREG32_P(mmVCE_STATUS, 1, ~1);
+		WREG32_P(mmVCE_STATUS, VCE_STATUS__JOB_BUSY_MASK,
+		         ~VCE_STATUS__JOB_BUSY_MASK);
+
 		if (adev->asic_type >= CHIP_STONEY)
 			WREG32_P(mmVCE_VCPU_CNTL, 1, ~0x200001);
 		else
 			WREG32_P(mmVCE_VCPU_CNTL, VCE_VCPU_CNTL__CLK_EN_MASK,
 				~VCE_VCPU_CNTL__CLK_EN_MASK);
 
-		WREG32_P(mmVCE_SOFT_RESET,
-			 VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
-			 ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
-
-		mdelay(100);
-
 		WREG32_P(mmVCE_SOFT_RESET, 0,
 			~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
 
-		for (i = 0; i < 10; ++i) {
-			uint32_t status;
-			for (j = 0; j < 100; ++j) {
-				status = RREG32(mmVCE_STATUS);
-				if (status & 2)
-					break;
-				mdelay(10);
-			}
-			r = 0;
-			if (status & 2)
-				break;
-
-			DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
-			WREG32_P(mmVCE_SOFT_RESET,
-				VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
-				~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
-			mdelay(10);
-			WREG32_P(mmVCE_SOFT_RESET, 0,
-				~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
-			mdelay(10);
-			r = -1;
-		}
+		mdelay(100);
+
+		r = vce_v3_0_firmware_loaded(adev);
 
 		/* clear BUSY flag */
-		WREG32_P(mmVCE_STATUS, 0, ~1);
+		WREG32_P(mmVCE_STATUS, 0, ~VCE_STATUS__JOB_BUSY_MASK);
 
 		/* Set Clock-Gating off */
 		if (adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)
@@ -290,19 +307,46 @@ static int vce_v3_0_start(struct amdgpu_device *adev)
 	WREG32_P(mmGRBM_GFX_INDEX, 0, ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
-	ring = &adev->vce.ring[0];
-	WREG32(mmVCE_RB_RPTR, ring->wptr);
-	WREG32(mmVCE_RB_WPTR, ring->wptr);
-	WREG32(mmVCE_RB_BASE_LO, ring->gpu_addr);
-	WREG32(mmVCE_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
-	WREG32(mmVCE_RB_SIZE, ring->ring_size / 4);
+	return 0;
+}
 
-	ring = &adev->vce.ring[1];
-	WREG32(mmVCE_RB_RPTR2, ring->wptr);
-	WREG32(mmVCE_RB_WPTR2, ring->wptr);
-	WREG32(mmVCE_RB_BASE_LO2, ring->gpu_addr);
-	WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
-	WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4);
+static int vce_v3_0_stop(struct amdgpu_device *adev)
+{
+	int idx;
+
+	mutex_lock(&adev->grbm_idx_mutex);
+	for (idx = 0; idx < 2; ++idx) {
+		if (adev->vce.harvest_config & (1 << idx))
+			continue;
+
+		if (idx == 0)
+			WREG32_P(mmGRBM_GFX_INDEX, 0,
+				~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+		else
+			WREG32_P(mmGRBM_GFX_INDEX,
+				GRBM_GFX_INDEX__VCE_INSTANCE_MASK,
+				~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+
+		if (adev->asic_type >= CHIP_STONEY)
+			WREG32_P(mmVCE_VCPU_CNTL, 0, ~0x200001);
+		else
+			WREG32_P(mmVCE_VCPU_CNTL, 0,
+				~VCE_VCPU_CNTL__CLK_EN_MASK);
+		/* hold on ECPU */
+		WREG32_P(mmVCE_SOFT_RESET,
+			 VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
+			 ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+
+		/* clear BUSY flag */
+		WREG32_P(mmVCE_STATUS, 0, ~VCE_STATUS__JOB_BUSY_MASK);
+
+		/* Set Clock-Gating off */
+		if (adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)
+			vce_v3_0_set_vce_sw_clock_gating(adev, false);
+	}
+
+	WREG32_P(mmGRBM_GFX_INDEX, 0, ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	return 0;
 }
@@ -441,7 +485,14 @@ static int vce_v3_0_hw_init(void *handle)
 
 static int vce_v3_0_hw_fini(void *handle)
 {
-	return 0;
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	r = vce_v3_0_wait_for_idle(handle);
+	if (r)
+		return r;
+
+	return vce_v3_0_stop(adev);
 }
 
 static int vce_v3_0_suspend(void *handle)
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index cda7def9dc2c..03a31c53aec3 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -1249,15 +1249,7 @@ static int vi_common_early_init(void *handle)
 			AMD_CG_SUPPORT_HDP_LS |
 			AMD_CG_SUPPORT_SDMA_MGCG |
 			AMD_CG_SUPPORT_SDMA_LS;
-		/* rev0 hardware doesn't support PG */
 		adev->pg_flags = 0;
-		if (adev->rev_id != 0x00)
-			adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
-				AMD_PG_SUPPORT_GFX_SMG |
-				AMD_PG_SUPPORT_GFX_DMG |
-				AMD_PG_SUPPORT_CP |
-				AMD_PG_SUPPORT_RLC_SMU_HS |
-				AMD_PG_SUPPORT_GFX_PIPELINE;
 		adev->external_rev_id = adev->rev_id + 0x1;
 		break;
 	case CHIP_STONEY:
@@ -1276,12 +1268,6 @@ static int vi_common_early_init(void *handle)
 			AMD_CG_SUPPORT_HDP_LS |
 			AMD_CG_SUPPORT_SDMA_MGCG |
 			AMD_CG_SUPPORT_SDMA_LS;
-		adev->pg_flags = AMD_PG_SUPPORT_GFX_PG |
-			AMD_PG_SUPPORT_GFX_SMG |
-			AMD_PG_SUPPORT_GFX_DMG |
-			AMD_PG_SUPPORT_GFX_PIPELINE |
-			AMD_PG_SUPPORT_CP |
-			AMD_PG_SUPPORT_RLC_SMU_HS;
 		adev->external_rev_id = adev->rev_id + 0x1;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h
index 0c8c85d2a2a5..f32af2f2091a 100644
--- a/drivers/gpu/drm/amd/include/cgs_common.h
+++ b/drivers/gpu/drm/amd/include/cgs_common.h
@@ -160,6 +160,10 @@ struct cgs_firmware_info {
 	uint16_t		feature_version;
 	uint32_t		image_size;
 	uint64_t		mc_addr;
+
+	/* only for smc firmware */
+	uint32_t		ucode_start_address;
+
 	void			*kptr;
 };
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c
index a6c9b4201e25..d1b528b401e9 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c
@@ -1828,7 +1828,7 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr)
 {
 	uint32_t ro, efuse, volt_without_cks, volt_with_cks, value, max, min;
 	struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
-	uint8_t i, stretch_amount, stretch_amount2, volt_offset = 0;
+	uint8_t i, stretch_amount, volt_offset = 0;
 	struct phm_ppt_v1_information *table_info =
 			(struct phm_ppt_v1_information *)(hwmgr->pptable);
 	struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table =
@@ -1879,11 +1879,8 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr)
 
 	data->smc_state_table.LdoRefSel = (table_info->cac_dtp_table->ucCKS_LDO_REFSEL != 0) ? table_info->cac_dtp_table->ucCKS_LDO_REFSEL : 6;
 	/* Populate CKS Lookup Table */
-	if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5)
-		stretch_amount2 = 0;
-	else if (stretch_amount == 3 || stretch_amount == 4)
-		stretch_amount2 = 1;
-	else {
+	if (stretch_amount != 1 && stretch_amount != 2 && stretch_amount != 3 &&
+			stretch_amount != 4 && stretch_amount != 5) {
 		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_ClockStretcher);
 		PP_ASSERT_WITH_CODE(false,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index a64db7ba4e0b..e2aece361eb0 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -179,13 +179,12 @@ int atomctrl_set_engine_dram_timings_rv770(
 
 	/* They are both in 10KHz Units. */
 	engine_clock_parameters.ulTargetEngineClock =
-		(uint32_t) engine_clock & SET_CLOCK_FREQ_MASK;
-	engine_clock_parameters.ulTargetEngineClock |=
-		(COMPUTE_ENGINE_PLL_PARAM << 24);
+		cpu_to_le32((engine_clock & SET_CLOCK_FREQ_MASK) |
+			    ((COMPUTE_ENGINE_PLL_PARAM << 24)));
 
 	/* in 10 khz units.*/
 	engine_clock_parameters.sReserved.ulClock =
-		(uint32_t) memory_clock & SET_CLOCK_FREQ_MASK;
+		cpu_to_le32(memory_clock & SET_CLOCK_FREQ_MASK);
 	return cgs_atom_exec_cmd_table(hwmgr->device,
 			GetIndexIntoMasterTable(COMMAND, DynamicMemorySettings),
 			&engine_clock_parameters);
@@ -252,7 +251,7 @@ int atomctrl_get_memory_pll_dividers_si(
 	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_1 mpll_parameters;
 	int result;
 
-	mpll_parameters.ulClock = (uint32_t) clock_value;
+	mpll_parameters.ulClock = cpu_to_le32(clock_value);
 	mpll_parameters.ucInputFlag = (uint8_t)((strobe_mode) ? 1 : 0);
 
 	result = cgs_atom_exec_cmd_table
@@ -262,9 +261,9 @@ int atomctrl_get_memory_pll_dividers_si(
 
 	if (0 == result) {
 		mpll_param->mpll_fb_divider.clk_frac =
-			mpll_parameters.ulFbDiv.usFbDivFrac;
+			le16_to_cpu(mpll_parameters.ulFbDiv.usFbDivFrac);
 		mpll_param->mpll_fb_divider.cl_kf =
-			mpll_parameters.ulFbDiv.usFbDiv;
+			le16_to_cpu(mpll_parameters.ulFbDiv.usFbDiv);
 		mpll_param->mpll_post_divider =
 			(uint32_t)mpll_parameters.ucPostDiv;
 		mpll_param->vco_mode =
@@ -300,7 +299,7 @@ int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2 mpll_parameters;
 	int result;
 
-	mpll_parameters.ulClock.ulClock = (uint32_t)clock_value;
+	mpll_parameters.ulClock.ulClock = cpu_to_le32(clock_value);
 
 	result = cgs_atom_exec_cmd_table(hwmgr->device,
 			GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
@@ -320,7 +319,7 @@ int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 	COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V4 pll_parameters;
 	int result;
 
-	pll_parameters.ulClock = clock_value;
+	pll_parameters.ulClock = cpu_to_le32(clock_value);
 
 	result = cgs_atom_exec_cmd_table
 		(hwmgr->device,
@@ -329,7 +328,7 @@ int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 
 	if (0 == result) {
 		dividers->pll_post_divider = pll_parameters.ucPostDiv;
-		dividers->real_clock = pll_parameters.ulClock;
+		dividers->real_clock = le32_to_cpu(pll_parameters.ulClock);
 	}
 
 	return result;
@@ -343,7 +342,7 @@ int atomctrl_get_engine_pll_dividers_vi(
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_6 pll_patameters;
 	int result;
 
-	pll_patameters.ulClock.ulClock = clock_value;
+	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv = COMPUTE_GPUCLK_INPUT_FLAG_SCLK;
 
 	result = cgs_atom_exec_cmd_table
@@ -355,12 +354,12 @@ int atomctrl_get_engine_pll_dividers_vi(
 		dividers->pll_post_divider =
 			pll_patameters.ulClock.ucPostDiv;
 		dividers->real_clock =
-			pll_patameters.ulClock.ulClock;
+			le32_to_cpu(pll_patameters.ulClock.ulClock);
 
 		dividers->ul_fb_div.ul_fb_div_frac =
-			pll_patameters.ulFbDiv.usFbDivFrac;
+			le16_to_cpu(pll_patameters.ulFbDiv.usFbDivFrac);
 		dividers->ul_fb_div.ul_fb_div =
-			pll_patameters.ulFbDiv.usFbDiv;
+			le16_to_cpu(pll_patameters.ulFbDiv.usFbDiv);
 
 		dividers->uc_pll_ref_div =
 			pll_patameters.ucPllRefDiv;
@@ -380,7 +379,7 @@ int atomctrl_get_engine_pll_dividers_ai(struct pp_hwmgr *hwmgr,
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_7 pll_patameters;
 	int result;
 
-	pll_patameters.ulClock.ulClock = clock_value;
+	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv = COMPUTE_GPUCLK_INPUT_FLAG_SCLK;
 
 	result = cgs_atom_exec_cmd_table
@@ -412,7 +411,7 @@ int atomctrl_get_dfs_pll_dividers_vi(
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_6 pll_patameters;
 	int result;
 
-	pll_patameters.ulClock.ulClock = clock_value;
+	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv =
 		COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK;
 
@@ -425,12 +424,12 @@ int atomctrl_get_dfs_pll_dividers_vi(
 		dividers->pll_post_divider =
 			pll_patameters.ulClock.ucPostDiv;
 		dividers->real_clock =
-			pll_patameters.ulClock.ulClock;
+			le32_to_cpu(pll_patameters.ulClock.ulClock);
 
 		dividers->ul_fb_div.ul_fb_div_frac =
-			pll_patameters.ulFbDiv.usFbDivFrac;
+			le16_to_cpu(pll_patameters.ulFbDiv.usFbDivFrac);
 		dividers->ul_fb_div.ul_fb_div =
-			pll_patameters.ulFbDiv.usFbDiv;
+			le16_to_cpu(pll_patameters.ulFbDiv.usFbDiv);
 
 		dividers->uc_pll_ref_div =
 			pll_patameters.ucPllRefDiv;
@@ -519,13 +518,13 @@ int atomctrl_get_voltage_table_v3(
 
 	for (i = 0; i < voltage_object->asGpioVoltageObj.ucGpioEntryNum; i++) {
 		voltage_table->entries[i].value =
-			voltage_object->asGpioVoltageObj.asVolGpioLut[i].usVoltageValue;
+			le16_to_cpu(voltage_object->asGpioVoltageObj.asVolGpioLut[i].usVoltageValue);
 		voltage_table->entries[i].smio_low =
-			voltage_object->asGpioVoltageObj.asVolGpioLut[i].ulVoltageId;
+			le32_to_cpu(voltage_object->asGpioVoltageObj.asVolGpioLut[i].ulVoltageId);
 	}
 
 	voltage_table->mask_low    =
-		voltage_object->asGpioVoltageObj.ulGpioMaskVal;
+		le32_to_cpu(voltage_object->asGpioVoltageObj.ulGpioMaskVal);
 	voltage_table->count      =
 		voltage_object->asGpioVoltageObj.ucGpioEntryNum;
 	voltage_table->phase_delay =
@@ -592,12 +591,12 @@ bool atomctrl_get_pp_assign_pin(
 		const uint32_t pinId,
 		pp_atomctrl_gpio_pin_assignment *gpio_pin_assignment)
 {
-	bool bRet = 0;
+	bool bRet = false;
 	ATOM_GPIO_PIN_LUT *gpio_lookup_table =
 		get_gpio_lookup_table(hwmgr->device);
 
 	PP_ASSERT_WITH_CODE((NULL != gpio_lookup_table),
-			"Could not find GPIO lookup Table in BIOS.", return -1);
+			"Could not find GPIO lookup Table in BIOS.", return false);
 
 	bRet = atomctrl_lookup_gpio_pin(gpio_lookup_table, pinId,
 		gpio_pin_assignment);
@@ -650,8 +649,8 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 		return -1;
 
 	if (getASICProfilingInfo->asHeader.ucTableFormatRevision < 3 ||
-			(getASICProfilingInfo->asHeader.ucTableFormatRevision == 3 &&
-			getASICProfilingInfo->asHeader.ucTableContentRevision < 4))
+	    (getASICProfilingInfo->asHeader.ucTableFormatRevision == 3 &&
+	     getASICProfilingInfo->asHeader.ucTableContentRevision < 4))
 		return -1;
 
 	/*-----------------------------------------------------------
@@ -662,37 +661,37 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	switch (dpm_level) {
 	case 1:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm1);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM1, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm1));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM1), 1000);
 		break;
 	case 2:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm2);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM2, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm2));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM2), 1000);
 		break;
 	case 3:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm3);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM3, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm3));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM3), 1000);
 		break;
 	case 4:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm4);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM4, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm4));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM4), 1000);
 		break;
 	case 5:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm5);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM5, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm5));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM5), 1000);
 		break;
 	case 6:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm6);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM6, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm6));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM6), 1000);
 		break;
 	case 7:
-		fPowerDPMx = Convert_ULONG_ToFraction(getASICProfilingInfo->usPowerDpm7);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM7, 1000);
+		fPowerDPMx = Convert_ULONG_ToFraction(le16_to_cpu(getASICProfilingInfo->usPowerDpm7));
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM7), 1000);
 		break;
 	default:
 		printk(KERN_ERR "DPM Level not supported\n");
 		fPowerDPMx = Convert_ULONG_ToFraction(1);
-		fDerateTDP = GetScaledFraction(getASICProfilingInfo->ulTdpDerateDPM0, 1000);
+		fDerateTDP = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulTdpDerateDPM0), 1000);
 	}
 
 	/*-------------------------
@@ -716,9 +715,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 		return result;
 
 	/* Finally, the actual fuse value */
-	ul_RO_fused = sOutput_FuseValues.ulEfuseValue;
-	fMin = GetScaledFraction(sRO_fuse.ulEfuseMin, 1);
-	fRange = GetScaledFraction(sRO_fuse.ulEfuseEncodeRange, 1);
+	ul_RO_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fMin = GetScaledFraction(le32_to_cpu(sRO_fuse.ulEfuseMin), 1);
+	fRange = GetScaledFraction(le32_to_cpu(sRO_fuse.ulEfuseEncodeRange), 1);
 	fRO_fused = fDecodeLinearFuse(ul_RO_fused, fMin, fRange, sRO_fuse.ucEfuseLength);
 
 	sCACm_fuse = getASICProfilingInfo->sCACm;
@@ -736,9 +735,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	if (result)
 		return result;
 
-	ul_CACm_fused = sOutput_FuseValues.ulEfuseValue;
-	fMin = GetScaledFraction(sCACm_fuse.ulEfuseMin, 1000);
-	fRange = GetScaledFraction(sCACm_fuse.ulEfuseEncodeRange, 1000);
+	ul_CACm_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fMin = GetScaledFraction(le32_to_cpu(sCACm_fuse.ulEfuseMin), 1000);
+	fRange = GetScaledFraction(le32_to_cpu(sCACm_fuse.ulEfuseEncodeRange), 1000);
 
 	fCACm_fused = fDecodeLinearFuse(ul_CACm_fused, fMin, fRange, sCACm_fuse.ucEfuseLength);
 
@@ -756,9 +755,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	if (result)
 		return result;
 
-	ul_CACb_fused = sOutput_FuseValues.ulEfuseValue;
-	fMin = GetScaledFraction(sCACb_fuse.ulEfuseMin, 1000);
-	fRange = GetScaledFraction(sCACb_fuse.ulEfuseEncodeRange, 1000);
+	ul_CACb_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fMin = GetScaledFraction(le32_to_cpu(sCACb_fuse.ulEfuseMin), 1000);
+	fRange = GetScaledFraction(le32_to_cpu(sCACb_fuse.ulEfuseEncodeRange), 1000);
 
 	fCACb_fused = fDecodeLinearFuse(ul_CACb_fused, fMin, fRange, sCACb_fuse.ucEfuseLength);
 
@@ -777,9 +776,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	if (result)
 		return result;
 
-	ul_Kt_Beta_fused = sOutput_FuseValues.ulEfuseValue;
-	fAverage = GetScaledFraction(sKt_Beta_fuse.ulEfuseEncodeAverage, 1000);
-	fRange = GetScaledFraction(sKt_Beta_fuse.ulEfuseEncodeRange, 1000);
+	ul_Kt_Beta_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fAverage = GetScaledFraction(le32_to_cpu(sKt_Beta_fuse.ulEfuseEncodeAverage), 1000);
+	fRange = GetScaledFraction(le32_to_cpu(sKt_Beta_fuse.ulEfuseEncodeRange), 1000);
 
 	fKt_Beta_fused = fDecodeLogisticFuse(ul_Kt_Beta_fused,
 			fAverage, fRange, sKt_Beta_fuse.ucEfuseLength);
@@ -798,9 +797,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	if (result)
 		return result;
 
-	ul_Kv_m_fused = sOutput_FuseValues.ulEfuseValue;
-	fAverage = GetScaledFraction(sKv_m_fuse.ulEfuseEncodeAverage, 1000);
-	fRange = GetScaledFraction((sKv_m_fuse.ulEfuseEncodeRange & 0x7fffffff), 1000);
+	ul_Kv_m_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fAverage = GetScaledFraction(le32_to_cpu(sKv_m_fuse.ulEfuseEncodeAverage), 1000);
+	fRange = GetScaledFraction((le32_to_cpu(sKv_m_fuse.ulEfuseEncodeRange) & 0x7fffffff), 1000);
 	fRange = fMultiply(fRange, ConvertToFraction(-1));
 
 	fKv_m_fused = fDecodeLogisticFuse(ul_Kv_m_fused,
@@ -820,9 +819,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	if (result)
 		return result;
 
-	ul_Kv_b_fused = sOutput_FuseValues.ulEfuseValue;
-	fAverage = GetScaledFraction(sKv_b_fuse.ulEfuseEncodeAverage, 1000);
-	fRange = GetScaledFraction(sKv_b_fuse.ulEfuseEncodeRange, 1000);
+	ul_Kv_b_fused = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fAverage = GetScaledFraction(le32_to_cpu(sKv_b_fuse.ulEfuseEncodeAverage), 1000);
+	fRange = GetScaledFraction(le32_to_cpu(sKv_b_fuse.ulEfuseEncodeRange), 1000);
 
 	fKv_b_fused = fDecodeLogisticFuse(ul_Kv_b_fused,
 			fAverage, fRange, sKv_b_fuse.ucEfuseLength);
@@ -851,9 +850,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	if (result)
 		return result;
 
-	ul_FT_Lkg_V0NORM = sOutput_FuseValues.ulEfuseValue;
-	fLn_MaxDivMin = GetScaledFraction(getASICProfilingInfo->ulLkgEncodeLn_MaxDivMin, 10000);
-	fMin = GetScaledFraction(getASICProfilingInfo->ulLkgEncodeMin, 10000);
+	ul_FT_Lkg_V0NORM = le32_to_cpu(sOutput_FuseValues.ulEfuseValue);
+	fLn_MaxDivMin = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulLkgEncodeLn_MaxDivMin), 10000);
+	fMin = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulLkgEncodeMin), 10000);
 
 	fFT_Lkg_V0NORM = fDecodeLeakageID(ul_FT_Lkg_V0NORM,
 			fLn_MaxDivMin, fMin, getASICProfilingInfo->ucLkgEfuseLength);
@@ -863,40 +862,40 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	 * PART 2 - Grabbing all required values
 	 *-------------------------------------------
 	 */
-	fSM_A0 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A0, 1000000),
+	fSM_A0 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A0), 1000000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A0_sign)));
-	fSM_A1 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A1, 1000000),
+	fSM_A1 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A1), 1000000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A1_sign)));
-	fSM_A2 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A2, 100000),
+	fSM_A2 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A2), 100000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A2_sign)));
-	fSM_A3 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A3, 1000000),
+	fSM_A3 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A3), 1000000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A3_sign)));
-	fSM_A4 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A4, 1000000),
+	fSM_A4 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A4), 1000000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A4_sign)));
-	fSM_A5 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A5, 1000),
+	fSM_A5 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A5), 1000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A5_sign)));
-	fSM_A6 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A6, 1000),
+	fSM_A6 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A6), 1000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A6_sign)));
-	fSM_A7 = fMultiply(GetScaledFraction(getASICProfilingInfo->ulSM_A7, 1000),
+	fSM_A7 = fMultiply(GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulSM_A7), 1000),
 			ConvertToFraction(uPow(-1, getASICProfilingInfo->ucSM_A7_sign)));
 
-	fMargin_RO_a = ConvertToFraction(getASICProfilingInfo->ulMargin_RO_a);
-	fMargin_RO_b = ConvertToFraction(getASICProfilingInfo->ulMargin_RO_b);
-	fMargin_RO_c = ConvertToFraction(getASICProfilingInfo->ulMargin_RO_c);
+	fMargin_RO_a = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_RO_a));
+	fMargin_RO_b = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_RO_b));
+	fMargin_RO_c = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_RO_c));
 
-	fMargin_fixed = ConvertToFraction(getASICProfilingInfo->ulMargin_fixed);
+	fMargin_fixed = ConvertToFraction(le32_to_cpu(getASICProfilingInfo->ulMargin_fixed));
 
 	fMargin_FMAX_mean = GetScaledFraction(
-			getASICProfilingInfo->ulMargin_Fmax_mean, 10000);
+		le32_to_cpu(getASICProfilingInfo->ulMargin_Fmax_mean), 10000);
 	fMargin_Plat_mean = GetScaledFraction(
-			getASICProfilingInfo->ulMargin_plat_mean, 10000);
+		le32_to_cpu(getASICProfilingInfo->ulMargin_plat_mean), 10000);
 	fMargin_FMAX_sigma = GetScaledFraction(
-			getASICProfilingInfo->ulMargin_Fmax_sigma, 10000);
+		le32_to_cpu(getASICProfilingInfo->ulMargin_Fmax_sigma), 10000);
 	fMargin_Plat_sigma = GetScaledFraction(
-			getASICProfilingInfo->ulMargin_plat_sigma, 10000);
+		le32_to_cpu(getASICProfilingInfo->ulMargin_plat_sigma), 10000);
 
 	fMargin_DC_sigma = GetScaledFraction(
-			getASICProfilingInfo->ulMargin_DC_sigma, 100);
+		le32_to_cpu(getASICProfilingInfo->ulMargin_DC_sigma), 100);
 	fMargin_DC_sigma = fDivide(fMargin_DC_sigma, ConvertToFraction(1000));
 
 	fCACm_fused = fDivide(fCACm_fused, ConvertToFraction(100));
@@ -908,14 +907,14 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	fSclk = GetScaledFraction(sclk, 100);
 
 	fV_max = fDivide(GetScaledFraction(
-			getASICProfilingInfo->ulMaxVddc, 1000), ConvertToFraction(4));
-	fT_prod = GetScaledFraction(getASICProfilingInfo->ulBoardCoreTemp, 10);
-	fLKG_Factor = GetScaledFraction(getASICProfilingInfo->ulEvvLkgFactor, 100);
-	fT_FT = GetScaledFraction(getASICProfilingInfo->ulLeakageTemp, 10);
+				 le32_to_cpu(getASICProfilingInfo->ulMaxVddc), 1000), ConvertToFraction(4));
+	fT_prod = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulBoardCoreTemp), 10);
+	fLKG_Factor = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulEvvLkgFactor), 100);
+	fT_FT = GetScaledFraction(le32_to_cpu(getASICProfilingInfo->ulLeakageTemp), 10);
 	fV_FT = fDivide(GetScaledFraction(
-			getASICProfilingInfo->ulLeakageVoltage, 1000), ConvertToFraction(4));
+				le32_to_cpu(getASICProfilingInfo->ulLeakageVoltage), 1000), ConvertToFraction(4));
 	fV_min = fDivide(GetScaledFraction(
-			getASICProfilingInfo->ulMinVddc, 1000), ConvertToFraction(4));
+				 le32_to_cpu(getASICProfilingInfo->ulMinVddc), 1000), ConvertToFraction(4));
 
 	/*-----------------------
 	 * PART 3
@@ -925,7 +924,7 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	fA_Term = fAdd(fMargin_RO_a, fAdd(fMultiply(fSM_A4, fSclk), fSM_A5));
 	fB_Term = fAdd(fAdd(fMultiply(fSM_A2, fSclk), fSM_A6), fMargin_RO_b);
 	fC_Term = fAdd(fMargin_RO_c,
-			fAdd(fMultiply(fSM_A0,fLkg_FT),
+			fAdd(fMultiply(fSM_A0, fLkg_FT),
 			fAdd(fMultiply(fSM_A1, fMultiply(fLkg_FT, fSclk)),
 			fAdd(fMultiply(fSM_A3, fSclk),
 			fSubtract(fSM_A7, fRO_fused)))));
@@ -1063,9 +1062,55 @@ int atomctrl_get_voltage_evv_on_sclk(
 	get_voltage_info_param_space.ucVoltageMode   =
 		ATOM_GET_VOLTAGE_EVV_VOLTAGE;
 	get_voltage_info_param_space.usVoltageLevel  =
-		virtual_voltage_Id;
+		cpu_to_le16(virtual_voltage_Id);
 	get_voltage_info_param_space.ulSCLKFreq      =
-		sclk;
+		cpu_to_le32(sclk);
+
+	result = cgs_atom_exec_cmd_table(hwmgr->device,
+			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
+			&get_voltage_info_param_space);
+
+	if (0 != result)
+		return result;
+
+	*voltage = le16_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
+				(&get_voltage_info_param_space))->usVoltageLevel);
+
+	return result;
+}
+
+/**
+ * atomctrl_get_voltage_evv gets voltage via call to ATOM COMMAND table.
+ * @param hwmgr	input: pointer to hwManager
+ * @param virtual_voltage_id      input: voltage id which match per voltage DPM state: 0xff01, 0xff02.. 0xff08
+ * @param voltage		       output: real voltage level in unit of mv
+ */
+int atomctrl_get_voltage_evv(struct pp_hwmgr *hwmgr,
+			     uint16_t virtual_voltage_id,
+			     uint16_t *voltage)
+{
+	int result;
+	int entry_id;
+	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
+
+	/* search for leakage voltage ID 0xff01 ~ 0xff08 and sckl */
+	for (entry_id = 0; entry_id < hwmgr->dyn_state.vddc_dependency_on_sclk->count; entry_id++) {
+		if (hwmgr->dyn_state.vddc_dependency_on_sclk->entries[entry_id].v == virtual_voltage_id) {
+			/* found */
+			break;
+		}
+	}
+
+	PP_ASSERT_WITH_CODE(entry_id < hwmgr->dyn_state.vddc_dependency_on_sclk->count,
+	        "Can't find requested voltage id in vddc_dependency_on_sclk table!",
+	        return -EINVAL;
+	);
+
+	get_voltage_info_param_space.ucVoltageType = VOLTAGE_TYPE_VDDC;
+	get_voltage_info_param_space.ucVoltageMode = ATOM_GET_VOLTAGE_EVV_VOLTAGE;
+	get_voltage_info_param_space.usVoltageLevel = virtual_voltage_id;
+	get_voltage_info_param_space.ulSCLKFreq =
+		cpu_to_le32(hwmgr->dyn_state.vddc_dependency_on_sclk->entries[entry_id].clk);
 
 	result = cgs_atom_exec_cmd_table(hwmgr->device,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
@@ -1074,8 +1119,8 @@ int atomctrl_get_voltage_evv_on_sclk(
 	if (0 != result)
 		return result;
 
-	*voltage = ((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
-			(&get_voltage_info_param_space))->usVoltageLevel;
+	*voltage = le16_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
+				(&get_voltage_info_param_space))->usVoltageLevel);
 
 	return result;
 }
@@ -1165,8 +1210,8 @@ static int asic_internal_ss_get_ss_asignment(struct pp_hwmgr *hwmgr,
 
 	if (entry_found) {
 		ssEntry->speed_spectrum_percentage =
-			ssInfo->usSpreadSpectrumPercentage;
-		ssEntry->speed_spectrum_rate = ssInfo->usSpreadRateInKhz;
+			le16_to_cpu(ssInfo->usSpreadSpectrumPercentage);
+		ssEntry->speed_spectrum_rate = le16_to_cpu(ssInfo->usSpreadRateInKhz);
 
 		if (((GET_DATA_TABLE_MAJOR_REVISION(table) == 2) &&
 			(GET_DATA_TABLE_MINOR_REVISION(table) >= 2)) ||
@@ -1222,7 +1267,7 @@ int atomctrl_read_efuse(void *device, uint16_t start_index,
 	int result;
 	READ_EFUSE_VALUE_PARAMETER efuse_param;
 
-	efuse_param.sEfuse.usEfuseIndex = (start_index / 32) * 4;
+	efuse_param.sEfuse.usEfuseIndex = cpu_to_le16((start_index / 32) * 4);
 	efuse_param.sEfuse.ucBitShift = (uint8_t)
 			(start_index - ((start_index / 32) * 32));
 	efuse_param.sEfuse.ucBitLength  = (uint8_t)
@@ -1232,19 +1277,21 @@ int atomctrl_read_efuse(void *device, uint16_t start_index,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
 			&efuse_param);
 	if (!result)
-		*efuse = efuse_param.ulEfuseValue & mask;
+		*efuse = le32_to_cpu(efuse_param.ulEfuseValue) & mask;
 
 	return result;
 }
 
 int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
-								uint8_t level)
+			      uint8_t level)
 {
 	DYNAMICE_MEMORY_SETTINGS_PARAMETER_V2_1 memory_clock_parameters;
 	int result;
 
-	memory_clock_parameters.asDPMMCReg.ulClock.ulClockFreq = memory_clock & SET_CLOCK_FREQ_MASK;
-	memory_clock_parameters.asDPMMCReg.ulClock.ulComputeClockFlag = ADJUST_MC_SETTING_PARAM;
+	memory_clock_parameters.asDPMMCReg.ulClock.ulClockFreq =
+		cpu_to_le32(memory_clock & SET_CLOCK_FREQ_MASK);
+	memory_clock_parameters.asDPMMCReg.ulClock.ulComputeClockFlag =
+		cpu_to_le32(ADJUST_MC_SETTING_PARAM);
 	memory_clock_parameters.asDPMMCReg.ucMclkDPMState = level;
 
 	result = cgs_atom_exec_cmd_table
@@ -1264,8 +1311,8 @@ int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_
 
 	get_voltage_info_param_space.ucVoltageType = voltage_type;
 	get_voltage_info_param_space.ucVoltageMode = ATOM_GET_VOLTAGE_EVV_VOLTAGE;
-	get_voltage_info_param_space.usVoltageLevel = virtual_voltage_Id;
-	get_voltage_info_param_space.ulSCLKFreq = sclk;
+	get_voltage_info_param_space.usVoltageLevel = cpu_to_le16(virtual_voltage_Id);
+	get_voltage_info_param_space.ulSCLKFreq = cpu_to_le32(sclk);
 
 	result = cgs_atom_exec_cmd_table(hwmgr->device,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
@@ -1274,7 +1321,7 @@ int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_
 	if (0 != result)
 		return result;
 
-	*voltage = ((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)(&get_voltage_info_param_space))->ulVoltageLevel;
+	*voltage = le32_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)(&get_voltage_info_param_space))->ulVoltageLevel);
 
 	return result;
 }
@@ -1295,15 +1342,19 @@ int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_atom_ctr
 	for (i = 0; i < psmu_info->ucSclkEntryNum; i++) {
 		table->entry[i].ucVco_setting = psmu_info->asSclkFcwRangeEntry[i].ucVco_setting;
 		table->entry[i].ucPostdiv = psmu_info->asSclkFcwRangeEntry[i].ucPostdiv;
-		table->entry[i].usFcw_pcc = psmu_info->asSclkFcwRangeEntry[i].ucFcw_pcc;
-		table->entry[i].usFcw_trans_upper = psmu_info->asSclkFcwRangeEntry[i].ucFcw_trans_upper;
-		table->entry[i].usRcw_trans_lower = psmu_info->asSclkFcwRangeEntry[i].ucRcw_trans_lower;
+		table->entry[i].usFcw_pcc =
+			le16_to_cpu(psmu_info->asSclkFcwRangeEntry[i].ucFcw_pcc);
+		table->entry[i].usFcw_trans_upper =
+			le16_to_cpu(psmu_info->asSclkFcwRangeEntry[i].ucFcw_trans_upper);
+		table->entry[i].usRcw_trans_lower =
+			le16_to_cpu(psmu_info->asSclkFcwRangeEntry[i].ucRcw_trans_lower);
 	}
 
 	return 0;
 }
 
-int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl__avfs_parameters *param)
+int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
+				  struct pp_atom_ctrl__avfs_parameters *param)
 {
 	ATOM_ASIC_PROFILING_INFO_V3_6 *profile = NULL;
 
@@ -1317,30 +1368,30 @@ int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl__a
 	if (!profile)
 		return -1;
 
-	param->ulAVFS_meanNsigma_Acontant0 = profile->ulAVFS_meanNsigma_Acontant0;
-	param->ulAVFS_meanNsigma_Acontant1 = profile->ulAVFS_meanNsigma_Acontant1;
-	param->ulAVFS_meanNsigma_Acontant2 = profile->ulAVFS_meanNsigma_Acontant2;
-	param->usAVFS_meanNsigma_DC_tol_sigma = profile->usAVFS_meanNsigma_DC_tol_sigma;
-	param->usAVFS_meanNsigma_Platform_mean = profile->usAVFS_meanNsigma_Platform_mean;
-	param->usAVFS_meanNsigma_Platform_sigma = profile->usAVFS_meanNsigma_Platform_sigma;
-	param->ulGB_VDROOP_TABLE_CKSOFF_a0 = profile->ulGB_VDROOP_TABLE_CKSOFF_a0;
-	param->ulGB_VDROOP_TABLE_CKSOFF_a1 = profile->ulGB_VDROOP_TABLE_CKSOFF_a1;
-	param->ulGB_VDROOP_TABLE_CKSOFF_a2 = profile->ulGB_VDROOP_TABLE_CKSOFF_a2;
-	param->ulGB_VDROOP_TABLE_CKSON_a0 = profile->ulGB_VDROOP_TABLE_CKSON_a0;
-	param->ulGB_VDROOP_TABLE_CKSON_a1 = profile->ulGB_VDROOP_TABLE_CKSON_a1;
-	param->ulGB_VDROOP_TABLE_CKSON_a2 = profile->ulGB_VDROOP_TABLE_CKSON_a2;
-	param->ulAVFSGB_FUSE_TABLE_CKSOFF_m1 = profile->ulAVFSGB_FUSE_TABLE_CKSOFF_m1;
-	param->usAVFSGB_FUSE_TABLE_CKSOFF_m2 = profile->usAVFSGB_FUSE_TABLE_CKSOFF_m2;
-	param->ulAVFSGB_FUSE_TABLE_CKSOFF_b = profile->ulAVFSGB_FUSE_TABLE_CKSOFF_b;
-	param->ulAVFSGB_FUSE_TABLE_CKSON_m1 = profile->ulAVFSGB_FUSE_TABLE_CKSON_m1;
-	param->usAVFSGB_FUSE_TABLE_CKSON_m2 = profile->usAVFSGB_FUSE_TABLE_CKSON_m2;
-	param->ulAVFSGB_FUSE_TABLE_CKSON_b = profile->ulAVFSGB_FUSE_TABLE_CKSON_b;
-	param->usMaxVoltage_0_25mv = profile->usMaxVoltage_0_25mv;
+	param->ulAVFS_meanNsigma_Acontant0 = le32_to_cpu(profile->ulAVFS_meanNsigma_Acontant0);
+	param->ulAVFS_meanNsigma_Acontant1 = le32_to_cpu(profile->ulAVFS_meanNsigma_Acontant1);
+	param->ulAVFS_meanNsigma_Acontant2 = le32_to_cpu(profile->ulAVFS_meanNsigma_Acontant2);
+	param->usAVFS_meanNsigma_DC_tol_sigma = le16_to_cpu(profile->usAVFS_meanNsigma_DC_tol_sigma);
+	param->usAVFS_meanNsigma_Platform_mean = le16_to_cpu(profile->usAVFS_meanNsigma_Platform_mean);
+	param->usAVFS_meanNsigma_Platform_sigma = le16_to_cpu(profile->usAVFS_meanNsigma_Platform_sigma);
+	param->ulGB_VDROOP_TABLE_CKSOFF_a0 = le32_to_cpu(profile->ulGB_VDROOP_TABLE_CKSOFF_a0);
+	param->ulGB_VDROOP_TABLE_CKSOFF_a1 = le32_to_cpu(profile->ulGB_VDROOP_TABLE_CKSOFF_a1);
+	param->ulGB_VDROOP_TABLE_CKSOFF_a2 = le32_to_cpu(profile->ulGB_VDROOP_TABLE_CKSOFF_a2);
+	param->ulGB_VDROOP_TABLE_CKSON_a0 = le32_to_cpu(profile->ulGB_VDROOP_TABLE_CKSON_a0);
+	param->ulGB_VDROOP_TABLE_CKSON_a1 = le32_to_cpu(profile->ulGB_VDROOP_TABLE_CKSON_a1);
+	param->ulGB_VDROOP_TABLE_CKSON_a2 = le32_to_cpu(profile->ulGB_VDROOP_TABLE_CKSON_a2);
+	param->ulAVFSGB_FUSE_TABLE_CKSOFF_m1 = le32_to_cpu(profile->ulAVFSGB_FUSE_TABLE_CKSOFF_m1);
+	param->usAVFSGB_FUSE_TABLE_CKSOFF_m2 = le16_to_cpu(profile->usAVFSGB_FUSE_TABLE_CKSOFF_m2);
+	param->ulAVFSGB_FUSE_TABLE_CKSOFF_b = le32_to_cpu(profile->ulAVFSGB_FUSE_TABLE_CKSOFF_b);
+	param->ulAVFSGB_FUSE_TABLE_CKSON_m1 = le32_to_cpu(profile->ulAVFSGB_FUSE_TABLE_CKSON_m1);
+	param->usAVFSGB_FUSE_TABLE_CKSON_m2 = le16_to_cpu(profile->usAVFSGB_FUSE_TABLE_CKSON_m2);
+	param->ulAVFSGB_FUSE_TABLE_CKSON_b = le32_to_cpu(profile->ulAVFSGB_FUSE_TABLE_CKSON_b);
+	param->usMaxVoltage_0_25mv = le16_to_cpu(profile->usMaxVoltage_0_25mv);
 	param->ucEnableGB_VDROOP_TABLE_CKSOFF = profile->ucEnableGB_VDROOP_TABLE_CKSOFF;
 	param->ucEnableGB_VDROOP_TABLE_CKSON = profile->ucEnableGB_VDROOP_TABLE_CKSON;
 	param->ucEnableGB_FUSE_TABLE_CKSOFF = profile->ucEnableGB_FUSE_TABLE_CKSOFF;
 	param->ucEnableGB_FUSE_TABLE_CKSON = profile->ucEnableGB_FUSE_TABLE_CKSON;
-	param->usPSM_Age_ComFactor = profile->usPSM_Age_ComFactor;
+	param->usPSM_Age_ComFactor = le16_to_cpu(profile->usPSM_Age_ComFactor);
 	param->ucEnableApplyAVFS_CKS_OFF_Voltage = profile->ucEnableApplyAVFS_CKS_OFF_Voltage;
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
index 1e35a9625baf..fc898afce002 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
@@ -281,6 +281,7 @@ struct pp_atom_ctrl__avfs_parameters {
 
 extern bool atomctrl_get_pp_assign_pin(struct pp_hwmgr *hwmgr, const uint32_t pinId, pp_atomctrl_gpio_pin_assignment *gpio_pin_assignment);
 extern int atomctrl_get_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type, uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage);
+extern int atomctrl_get_voltage_evv(struct pp_hwmgr *hwmgr, uint16_t virtual_voltage_id, uint16_t *voltage);
 extern uint32_t atomctrl_get_mpll_reference_clock(struct pp_hwmgr *hwmgr);
 extern int atomctrl_get_memory_clock_spread_spectrum(struct pp_hwmgr *hwmgr, const uint32_t memory_clock, pp_atomctrl_internal_ss_info *ssInfo);
 extern int atomctrl_get_engine_clock_spread_spectrum(struct pp_hwmgr *hwmgr, const uint32_t engine_clock, pp_atomctrl_internal_ss_info *ssInfo);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
index 35bc8a29b773..6c321b0d8a1e 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
@@ -810,6 +810,19 @@ static const ATOM_PPLIB_POWERPLAYTABLE *get_powerplay_table(
 	return (const ATOM_PPLIB_POWERPLAYTABLE *)table_addr;
 }
 
+int pp_tables_get_response_times(struct pp_hwmgr *hwmgr,
+				uint32_t *vol_rep_time, uint32_t *bb_rep_time)
+{
+	const ATOM_PPLIB_POWERPLAYTABLE *powerplay_tab = get_powerplay_table(hwmgr);
+
+	PP_ASSERT_WITH_CODE(NULL != powerplay_tab,
+			    "Missing PowerPlay Table!", return -EINVAL);
+
+	*vol_rep_time = (uint32_t)le16_to_cpu(powerplay_tab->usVoltageTime);
+	*bb_rep_time = (uint32_t)le16_to_cpu(powerplay_tab->usBackbiasTime);
+
+	return 0;
+}
 
 int pp_tables_get_num_of_entries(struct pp_hwmgr *hwmgr,
 				     unsigned long *num_of_entries)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.h b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.h
index 30434802417e..baddaa75693b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.h
@@ -32,16 +32,19 @@ struct pp_hw_power_state;
 extern const struct pp_table_func pptable_funcs;
 
 typedef int (*pp_tables_hw_clock_info_callback)(struct pp_hwmgr *hwmgr,
-					struct pp_hw_power_state *hw_ps,
-							unsigned int index,
-						 const void *clock_info);
+						struct pp_hw_power_state *hw_ps,
+						unsigned int index,
+						const void *clock_info);
 
 int pp_tables_get_num_of_entries(struct pp_hwmgr *hwmgr,
-				     unsigned long *num_of_entries);
+				 unsigned long *num_of_entries);
 
 int pp_tables_get_entry(struct pp_hwmgr *hwmgr,
-						unsigned long entry_index,
-						struct pp_power_state *ps,
-				pp_tables_hw_clock_info_callback func);
+			unsigned long entry_index,
+			struct pp_power_state *ps,
+			pp_tables_hw_clock_info_callback func);
+
+int pp_tables_get_response_times(struct pp_hwmgr *hwmgr,
+				 uint32_t *vol_rep_time, uint32_t *bb_rep_time);
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h
index fc9e3d1dd409..3c235f0177cd 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h
@@ -131,6 +131,12 @@ extern int smu_free_memory(void *device, void *handle);
 	smum_wait_on_indirect_register(smumgr,				\
 				mm##port##_INDEX, index, value, mask)
 
+#define SMUM_WAIT_INDIRECT_REGISTER(smumgr, port, reg, value, mask)    \
+	    SMUM_WAIT_INDIRECT_REGISTER_GIVEN_INDEX(smumgr, port, ix##reg, value, mask)
+
+#define SMUM_WAIT_INDIRECT_FIELD(smumgr, port, reg, field, fieldval)                          \
+	    SMUM_WAIT_INDIRECT_REGISTER(smumgr, port, reg, (fieldval) << SMUM_FIELD_SHIFT(reg, field), \
+			            SMUM_FIELD_MASK(reg, field) )
 
 #define SMUM_WAIT_REGISTER_UNEQUAL_GIVEN_INDEX(smumgr,         \
 							index, value, mask) \
@@ -158,6 +164,10 @@ extern int smu_free_memory(void *device, void *handle);
 		(SMUM_FIELD_MASK(reg, field) & ((field_val) <<                 \
 			SMUM_FIELD_SHIFT(reg, field))))
 
+#define SMUM_READ_INDIRECT_FIELD(device, port, reg, field) \
+	    SMUM_GET_FIELD(cgs_read_ind_register(device, port, ix##reg), \
+			   reg, field)
+
 #define SMUM_WAIT_VFPF_INDIRECT_REGISTER_GIVEN_INDEX(smumgr,		\
 				port, index, value, mask)		\
 	smum_wait_on_indirect_register(smumgr,				\
@@ -191,6 +201,13 @@ extern int smu_free_memory(void *device, void *handle);
 			SMUM_SET_FIELD(cgs_read_ind_register(device, port, ix##reg), \
 			reg, field, fieldval))
 
+
+#define SMUM_WRITE_INDIRECT_FIELD(device, port, reg, field, fieldval)    		\
+		cgs_write_ind_register(device, port, ix##reg, 				\
+			SMUM_SET_FIELD(cgs_read_ind_register(device, port, ix##reg), 	\
+				       reg, field, fieldval))
+
+
 #define SMUM_WAIT_VFPF_INDIRECT_FIELD(smumgr, port, reg, field, fieldval) \
 	SMUM_WAIT_VFPF_INDIRECT_REGISTER(smumgr, port, reg,		\
 		(fieldval) << SMUM_FIELD_SHIFT(reg, field),		\
@@ -200,4 +217,16 @@ extern int smu_free_memory(void *device, void *handle);
 	SMUM_WAIT_VFPF_INDIRECT_REGISTER_UNEQUAL(smumgr, port, reg,	\
 		(fieldval) << SMUM_FIELD_SHIFT(reg, field),		\
 		SMUM_FIELD_MASK(reg, field))
+
+#define SMUM_WAIT_INDIRECT_REGISTER_UNEQUAL_GIVEN_INDEX(smumgr, port, index, value, mask)    \
+	smum_wait_for_indirect_register_unequal(smumgr,			\
+		mm##port##_INDEX, index, value, mask)
+
+#define SMUM_WAIT_INDIRECT_REGISTER_UNEQUAL(smumgr, port, reg, value, mask)    \
+	    SMUM_WAIT_INDIRECT_REGISTER_UNEQUAL_GIVEN_INDEX(smumgr, port, ix##reg, value, mask)
+
+#define SMUM_WAIT_INDIRECT_FIELD_UNEQUAL(smumgr, port, reg, field, fieldval)                          \
+	    SMUM_WAIT_INDIRECT_REGISTER_UNEQUAL(smumgr, port, reg, (fieldval) << SMUM_FIELD_SHIFT(reg, field), \
+			            SMUM_FIELD_MASK(reg, field) )
+
 #endif
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c
index 587cae4e73c9..56bb758f4e33 100644
--- a/drivers/gpu/drm/radeon/atombios_encoders.c
+++ b/drivers/gpu/drm/radeon/atombios_encoders.c
@@ -120,6 +120,7 @@ atombios_set_backlight_level(struct radeon_encoder *radeon_encoder, u8 level)
 		case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
 		case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
 		case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+		case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
 			if (dig->backlight_level == 0)
 				atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_LCD_BLOFF, 0, 0);
 			else {