Merge tag 'amd-drm-next-5.16-2021-09-27' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-5.16-2021-09-27:

amdgpu:
- RAS improvements
- BACO fixes
- Yellow Carp updates
- Misc code cleanups
- Initial DP 2.0 support
- VCN priority handling
- Cyan Skillfish updates
- Rework IB handling for multimedia engine tests
- Backlight fixes
- DCN 3.1 power saving improvements
- Runtime PM fixes
- Modifier support for DCC image stores for gfx 10.3
- Hotplug fixes
- Clean up stack related warnings in display code
- DP alt mode fixes
- Display rework for better handling FP code
- Debugfs fixes

amdkfd:
- SVM fixes
- DMA map fixes

radeon:
- AGP fix

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210927212653.4575-1-alexander.deucher@amd.com
Signed-off-by: Dave Airlie <airlied@redhat.com>

55 files changed, 1281 insertions, 1094 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index e7a010b7ca1f..468003583b2a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -43,14 +43,61 @@ const unsigned int amdgpu_ctx_num_entities[AMDGPU_HW_IP_NUM] = {
 	[AMDGPU_HW_IP_VCN_JPEG]	=	1,
 };
 
+bool amdgpu_ctx_priority_is_valid(int32_t ctx_prio)
+{
+	switch (ctx_prio) {
+	case AMDGPU_CTX_PRIORITY_UNSET:
+	case AMDGPU_CTX_PRIORITY_VERY_LOW:
+	case AMDGPU_CTX_PRIORITY_LOW:
+	case AMDGPU_CTX_PRIORITY_NORMAL:
+	case AMDGPU_CTX_PRIORITY_HIGH:
+	case AMDGPU_CTX_PRIORITY_VERY_HIGH:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static enum drm_sched_priority
+amdgpu_ctx_to_drm_sched_prio(int32_t ctx_prio)
+{
+	switch (ctx_prio) {
+	case AMDGPU_CTX_PRIORITY_UNSET:
+		return DRM_SCHED_PRIORITY_UNSET;
+
+	case AMDGPU_CTX_PRIORITY_VERY_LOW:
+		return DRM_SCHED_PRIORITY_MIN;
+
+	case AMDGPU_CTX_PRIORITY_LOW:
+		return DRM_SCHED_PRIORITY_MIN;
+
+	case AMDGPU_CTX_PRIORITY_NORMAL:
+		return DRM_SCHED_PRIORITY_NORMAL;
+
+	case AMDGPU_CTX_PRIORITY_HIGH:
+		return DRM_SCHED_PRIORITY_HIGH;
+
+	case AMDGPU_CTX_PRIORITY_VERY_HIGH:
+		return DRM_SCHED_PRIORITY_HIGH;
+
+	/* This should not happen as we sanitized userspace provided priority
+	 * already, WARN if this happens.
+	 */
+	default:
+		WARN(1, "Invalid context priority %d\n", ctx_prio);
+		return DRM_SCHED_PRIORITY_NORMAL;
+	}
+
+}
+
 static int amdgpu_ctx_priority_permit(struct drm_file *filp,
-				      enum drm_sched_priority priority)
+				      int32_t priority)
 {
-	if (priority < 0 || priority >= DRM_SCHED_PRIORITY_COUNT)
+	if (!amdgpu_ctx_priority_is_valid(priority))
 		return -EINVAL;
 
 	/* NORMAL and below are accessible by everyone */
-	if (priority <= DRM_SCHED_PRIORITY_NORMAL)
+	if (priority <= AMDGPU_CTX_PRIORITY_NORMAL)
 		return 0;
 
 	if (capable(CAP_SYS_NICE))
@@ -62,26 +109,51 @@ static int amdgpu_ctx_priority_permit(struct drm_file *filp,
 	return -EACCES;
 }
 
-static enum gfx_pipe_priority amdgpu_ctx_sched_prio_to_compute_prio(enum drm_sched_priority prio)
+static enum amdgpu_gfx_pipe_priority amdgpu_ctx_prio_to_compute_prio(int32_t prio)
 {
 	switch (prio) {
-	case DRM_SCHED_PRIORITY_HIGH:
-	case DRM_SCHED_PRIORITY_KERNEL:
+	case AMDGPU_CTX_PRIORITY_HIGH:
+	case AMDGPU_CTX_PRIORITY_VERY_HIGH:
 		return AMDGPU_GFX_PIPE_PRIO_HIGH;
 	default:
 		return AMDGPU_GFX_PIPE_PRIO_NORMAL;
 	}
 }
 
-static unsigned int amdgpu_ctx_prio_sched_to_hw(struct amdgpu_device *adev,
-						 enum drm_sched_priority prio,
-						 u32 hw_ip)
+static enum amdgpu_ring_priority_level amdgpu_ctx_sched_prio_to_ring_prio(int32_t prio)
 {
+	switch (prio) {
+	case AMDGPU_CTX_PRIORITY_HIGH:
+		return AMDGPU_RING_PRIO_1;
+	case AMDGPU_CTX_PRIORITY_VERY_HIGH:
+		return AMDGPU_RING_PRIO_2;
+	default:
+		return AMDGPU_RING_PRIO_0;
+	}
+}
+
+static unsigned int amdgpu_ctx_get_hw_prio(struct amdgpu_ctx *ctx, u32 hw_ip)
+{
+	struct amdgpu_device *adev = ctx->adev;
+	int32_t ctx_prio;
 	unsigned int hw_prio;
 
-	hw_prio = (hw_ip == AMDGPU_HW_IP_COMPUTE) ?
-			amdgpu_ctx_sched_prio_to_compute_prio(prio) :
-			AMDGPU_RING_PRIO_DEFAULT;
+	ctx_prio = (ctx->override_priority == AMDGPU_CTX_PRIORITY_UNSET) ?
+			ctx->init_priority : ctx->override_priority;
+
+	switch (hw_ip) {
+	case AMDGPU_HW_IP_COMPUTE:
+		hw_prio = amdgpu_ctx_prio_to_compute_prio(ctx_prio);
+		break;
+	case AMDGPU_HW_IP_VCE:
+	case AMDGPU_HW_IP_VCN_ENC:
+		hw_prio = amdgpu_ctx_sched_prio_to_ring_prio(ctx_prio);
+		break;
+	default:
+		hw_prio = AMDGPU_RING_PRIO_DEFAULT;
+		break;
+	}
+
 	hw_ip = array_index_nospec(hw_ip, AMDGPU_HW_IP_NUM);
 	if (adev->gpu_sched[hw_ip][hw_prio].num_scheds == 0)
 		hw_prio = AMDGPU_RING_PRIO_DEFAULT;
@@ -89,15 +161,17 @@ static unsigned int amdgpu_ctx_prio_sched_to_hw(struct amdgpu_device *adev,
 	return hw_prio;
 }
 
+
 static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, u32 hw_ip,
-				   const u32 ring)
+				  const u32 ring)
 {
 	struct amdgpu_device *adev = ctx->adev;
 	struct amdgpu_ctx_entity *entity;
 	struct drm_gpu_scheduler **scheds = NULL, *sched = NULL;
 	unsigned num_scheds = 0;
+	int32_t ctx_prio;
 	unsigned int hw_prio;
-	enum drm_sched_priority priority;
+	enum drm_sched_priority drm_prio;
 	int r;
 
 	entity = kzalloc(struct_size(entity, fences, amdgpu_sched_jobs),
@@ -105,10 +179,11 @@ static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, u32 hw_ip,
 	if (!entity)
 		return  -ENOMEM;
 
+	ctx_prio = (ctx->override_priority == AMDGPU_CTX_PRIORITY_UNSET) ?
+			ctx->init_priority : ctx->override_priority;
 	entity->sequence = 1;
-	priority = (ctx->override_priority == DRM_SCHED_PRIORITY_UNSET) ?
-				ctx->init_priority : ctx->override_priority;
-	hw_prio = amdgpu_ctx_prio_sched_to_hw(adev, priority, hw_ip);
+	hw_prio = amdgpu_ctx_get_hw_prio(ctx, hw_ip);
+	drm_prio = amdgpu_ctx_to_drm_sched_prio(ctx_prio);
 
 	hw_ip = array_index_nospec(hw_ip, AMDGPU_HW_IP_NUM);
 	scheds = adev->gpu_sched[hw_ip][hw_prio].sched;
@@ -124,7 +199,7 @@ static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, u32 hw_ip,
 		num_scheds = 1;
 	}
 
-	r = drm_sched_entity_init(&entity->entity, priority, scheds, num_scheds,
+	r = drm_sched_entity_init(&entity->entity, drm_prio, scheds, num_scheds,
 				  &ctx->guilty);
 	if (r)
 		goto error_free_entity;
@@ -139,7 +214,7 @@ error_free_entity:
 }
 
 static int amdgpu_ctx_init(struct amdgpu_device *adev,
-			   enum drm_sched_priority priority,
+			   int32_t priority,
 			   struct drm_file *filp,
 			   struct amdgpu_ctx *ctx)
 {
@@ -161,7 +236,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 	ctx->reset_counter_query = ctx->reset_counter;
 	ctx->vram_lost_counter = atomic_read(&adev->vram_lost_counter);
 	ctx->init_priority = priority;
-	ctx->override_priority = DRM_SCHED_PRIORITY_UNSET;
+	ctx->override_priority = AMDGPU_CTX_PRIORITY_UNSET;
 
 	return 0;
 }
@@ -234,7 +309,7 @@ int amdgpu_ctx_get_entity(struct amdgpu_ctx *ctx, u32 hw_ip, u32 instance,
 static int amdgpu_ctx_alloc(struct amdgpu_device *adev,
 			    struct amdgpu_fpriv *fpriv,
 			    struct drm_file *filp,
-			    enum drm_sched_priority priority,
+			    int32_t priority,
 			    uint32_t *id)
 {
 	struct amdgpu_ctx_mgr *mgr = &fpriv->ctx_mgr;
@@ -397,19 +472,19 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 {
 	int r;
 	uint32_t id;
-	enum drm_sched_priority priority;
+	int32_t priority;
 
 	union drm_amdgpu_ctx *args = data;
 	struct amdgpu_device *adev = drm_to_adev(dev);
 	struct amdgpu_fpriv *fpriv = filp->driver_priv;
 
 	id = args->in.ctx_id;
-	r = amdgpu_to_sched_priority(args->in.priority, &priority);
+	priority = args->in.priority;
 
 	/* For backwards compatibility reasons, we need to accept
 	 * ioctls with garbage in the priority field */
-	if (r == -EINVAL)
-		priority = DRM_SCHED_PRIORITY_NORMAL;
+	if (!amdgpu_ctx_priority_is_valid(priority))
+		priority = AMDGPU_CTX_PRIORITY_NORMAL;
 
 	switch (args->in.op) {
 	case AMDGPU_CTX_OP_ALLOC_CTX:
@@ -515,9 +590,9 @@ struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
 }
 
 static void amdgpu_ctx_set_entity_priority(struct amdgpu_ctx *ctx,
-					    struct amdgpu_ctx_entity *aentity,
-					    int hw_ip,
-					    enum drm_sched_priority priority)
+					   struct amdgpu_ctx_entity *aentity,
+					   int hw_ip,
+					   int32_t priority)
 {
 	struct amdgpu_device *adev = ctx->adev;
 	unsigned int hw_prio;
@@ -525,12 +600,12 @@ static void amdgpu_ctx_set_entity_priority(struct amdgpu_ctx *ctx,
 	unsigned num_scheds;
 
 	/* set sw priority */
-	drm_sched_entity_set_priority(&aentity->entity, priority);
+	drm_sched_entity_set_priority(&aentity->entity,
+				      amdgpu_ctx_to_drm_sched_prio(priority));
 
 	/* set hw priority */
 	if (hw_ip == AMDGPU_HW_IP_COMPUTE) {
-		hw_prio = amdgpu_ctx_prio_sched_to_hw(adev, priority,
-						      AMDGPU_HW_IP_COMPUTE);
+		hw_prio = amdgpu_ctx_get_hw_prio(ctx, hw_ip);
 		hw_prio = array_index_nospec(hw_prio, AMDGPU_RING_PRIO_MAX);
 		scheds = adev->gpu_sched[hw_ip][hw_prio].sched;
 		num_scheds = adev->gpu_sched[hw_ip][hw_prio].num_scheds;
@@ -540,14 +615,14 @@ static void amdgpu_ctx_set_entity_priority(struct amdgpu_ctx *ctx,
 }
 
 void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx,
-				  enum drm_sched_priority priority)
+				  int32_t priority)
 {
-	enum drm_sched_priority ctx_prio;
+	int32_t ctx_prio;
 	unsigned i, j;
 
 	ctx->override_priority = priority;
 
-	ctx_prio = (ctx->override_priority == DRM_SCHED_PRIORITY_UNSET) ?
+	ctx_prio = (ctx->override_priority == AMDGPU_CTX_PRIORITY_UNSET) ?
 			ctx->init_priority : ctx->override_priority;
 	for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) {
 		for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
index 14db16bc3322..a44b8b8ed39c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
@@ -47,8 +47,8 @@ struct amdgpu_ctx {
 	spinlock_t			ring_lock;
 	struct amdgpu_ctx_entity	*entities[AMDGPU_HW_IP_NUM][AMDGPU_MAX_ENTITY_NUM];
 	bool				preamble_presented;
-	enum drm_sched_priority		init_priority;
-	enum drm_sched_priority		override_priority;
+	int32_t				init_priority;
+	int32_t				override_priority;
 	struct mutex			lock;
 	atomic_t			guilty;
 	unsigned long			ras_counter_ce;
@@ -75,8 +75,8 @@ void amdgpu_ctx_add_fence(struct amdgpu_ctx *ctx,
 struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
 				       struct drm_sched_entity *entity,
 				       uint64_t seq);
-void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx,
-				  enum drm_sched_priority priority);
+bool amdgpu_ctx_priority_is_valid(int32_t ctx_prio);
+void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx, int32_t ctx_prio);
 
 int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 		     struct drm_file *filp);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
index 463b9c0283f7..074ffcf0dac2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -36,6 +36,7 @@
 #include "amdgpu_rap.h"
 #include "amdgpu_securedisplay.h"
 #include "amdgpu_fw_attestation.h"
+#include "amdgpu_umr.h"
 
 int amdgpu_debugfs_wait_dump(struct amdgpu_device *adev)
 {
@@ -279,6 +280,145 @@ static ssize_t amdgpu_debugfs_regs_write(struct file *f, const char __user *buf,
 	return amdgpu_debugfs_process_reg_op(false, f, (char __user *)buf, size, pos);
 }
 
+static int amdgpu_debugfs_regs2_open(struct inode *inode, struct file *file)
+{
+	struct amdgpu_debugfs_regs2_data *rd;
+
+	rd = kzalloc(sizeof *rd, GFP_KERNEL);
+	if (!rd)
+		return -ENOMEM;
+	rd->adev = file_inode(file)->i_private;
+	file->private_data = rd;
+	mutex_init(&rd->lock);
+
+	return 0;
+}
+
+static int amdgpu_debugfs_regs2_release(struct inode *inode, struct file *file)
+{
+	struct amdgpu_debugfs_regs2_data *rd = file->private_data;
+	mutex_destroy(&rd->lock);
+	kfree(file->private_data);
+	return 0;
+}
+
+static ssize_t amdgpu_debugfs_regs2_op(struct file *f, char __user *buf, u32 offset, size_t size, int write_en)
+{
+	struct amdgpu_debugfs_regs2_data *rd = f->private_data;
+	struct amdgpu_device *adev = rd->adev;
+	ssize_t result = 0;
+	int r;
+	uint32_t value;
+
+	if (size & 0x3 || offset & 0x3)
+		return -EINVAL;
+
+	r = pm_runtime_get_sync(adev_to_drm(adev)->dev);
+	if (r < 0) {
+		pm_runtime_put_autosuspend(adev_to_drm(adev)->dev);
+		return r;
+	}
+
+	r = amdgpu_virt_enable_access_debugfs(adev);
+	if (r < 0) {
+		pm_runtime_put_autosuspend(adev_to_drm(adev)->dev);
+		return r;
+	}
+
+	mutex_lock(&rd->lock);
+
+	if (rd->id.use_grbm) {
+		if ((rd->id.grbm.sh != 0xFFFFFFFF && rd->id.grbm.sh >= adev->gfx.config.max_sh_per_se) ||
+		    (rd->id.grbm.se != 0xFFFFFFFF && rd->id.grbm.se >= adev->gfx.config.max_shader_engines)) {
+			pm_runtime_mark_last_busy(adev_to_drm(adev)->dev);
+			pm_runtime_put_autosuspend(adev_to_drm(adev)->dev);
+			amdgpu_virt_disable_access_debugfs(adev);
+			mutex_unlock(&rd->lock);
+			return -EINVAL;
+		}
+		mutex_lock(&adev->grbm_idx_mutex);
+		amdgpu_gfx_select_se_sh(adev, rd->id.grbm.se,
+								rd->id.grbm.sh,
+								rd->id.grbm.instance);
+	}
+
+	if (rd->id.use_srbm) {
+		mutex_lock(&adev->srbm_mutex);
+		amdgpu_gfx_select_me_pipe_q(adev, rd->id.srbm.me, rd->id.srbm.pipe,
+									rd->id.srbm.queue, rd->id.srbm.vmid);
+	}
+
+	if (rd->id.pg_lock)
+		mutex_lock(&adev->pm.mutex);
+
+	while (size) {
+		if (!write_en) {
+			value = RREG32(offset >> 2);
+			r = put_user(value, (uint32_t *)buf);
+		} else {
+			r = get_user(value, (uint32_t *)buf);
+			if (!r)
+				amdgpu_mm_wreg_mmio_rlc(adev, offset >> 2, value);
+		}
+		if (r) {
+			result = r;
+			goto end;
+		}
+		offset += 4;
+		size -= 4;
+		result += 4;
+		buf += 4;
+	}
+end:
+	if (rd->id.use_grbm) {
+		amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+		mutex_unlock(&adev->grbm_idx_mutex);
+	}
+
+	if (rd->id.use_srbm) {
+		amdgpu_gfx_select_me_pipe_q(adev, 0, 0, 0, 0);
+		mutex_unlock(&adev->srbm_mutex);
+	}
+
+	if (rd->id.pg_lock)
+		mutex_unlock(&adev->pm.mutex);
+
+	mutex_unlock(&rd->lock);
+
+	pm_runtime_mark_last_busy(adev_to_drm(adev)->dev);
+	pm_runtime_put_autosuspend(adev_to_drm(adev)->dev);
+
+	amdgpu_virt_disable_access_debugfs(adev);
+	return result;
+}
+
+static long amdgpu_debugfs_regs2_ioctl(struct file *f, unsigned int cmd, unsigned long data)
+{
+	struct amdgpu_debugfs_regs2_data *rd = f->private_data;
+	int r;
+
+	switch (cmd) {
+	case AMDGPU_DEBUGFS_REGS2_IOC_SET_STATE:
+		mutex_lock(&rd->lock);
+		r = copy_from_user(&rd->id, (struct amdgpu_debugfs_regs2_iocdata *)data, sizeof rd->id);
+		mutex_unlock(&rd->lock);
+		return r ? -EINVAL : 0;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static ssize_t amdgpu_debugfs_regs2_read(struct file *f, char __user *buf, size_t size, loff_t *pos)
+{
+	return amdgpu_debugfs_regs2_op(f, buf, *pos, size, 0);
+}
+
+static ssize_t amdgpu_debugfs_regs2_write(struct file *f, const char __user *buf, size_t size, loff_t *pos)
+{
+	return amdgpu_debugfs_regs2_op(f, (char __user *)buf, *pos, size, 1);
+}
+
 
 /**
  * amdgpu_debugfs_regs_pcie_read - Read from a PCIE register
@@ -1091,6 +1231,16 @@ static ssize_t amdgpu_debugfs_gfxoff_read(struct file *f, char __user *buf,
 	return result;
 }
 
+static const struct file_operations amdgpu_debugfs_regs2_fops = {
+	.owner = THIS_MODULE,
+	.unlocked_ioctl = amdgpu_debugfs_regs2_ioctl,
+	.read = amdgpu_debugfs_regs2_read,
+	.write = amdgpu_debugfs_regs2_write,
+	.open = amdgpu_debugfs_regs2_open,
+	.release = amdgpu_debugfs_regs2_release,
+	.llseek = default_llseek
+};
+
 static const struct file_operations amdgpu_debugfs_regs_fops = {
 	.owner = THIS_MODULE,
 	.read = amdgpu_debugfs_regs_read,
@@ -1148,6 +1298,7 @@ static const struct file_operations amdgpu_debugfs_gfxoff_fops = {
 
 static const struct file_operations *debugfs_regs[] = {
 	&amdgpu_debugfs_regs_fops,
+	&amdgpu_debugfs_regs2_fops,
 	&amdgpu_debugfs_regs_didt_fops,
 	&amdgpu_debugfs_regs_pcie_fops,
 	&amdgpu_debugfs_regs_smc_fops,
@@ -1160,6 +1311,7 @@ static const struct file_operations *debugfs_regs[] = {
 
 static const char *debugfs_regs_names[] = {
 	"amdgpu_regs",
+	"amdgpu_regs2",
 	"amdgpu_regs_didt",
 	"amdgpu_regs_pcie",
 	"amdgpu_regs_smc",
@@ -1206,7 +1358,7 @@ static int amdgpu_debugfs_test_ib_show(struct seq_file *m, void *unused)
 	}
 
 	/* Avoid accidently unparking the sched thread during GPU reset */
-	r = down_read_killable(&adev->reset_sem);
+	r = down_write_killable(&adev->reset_sem);
 	if (r)
 		return r;
 
@@ -1235,7 +1387,7 @@ static int amdgpu_debugfs_test_ib_show(struct seq_file *m, void *unused)
 		kthread_unpark(ring->sched.thread);
 	}
 
-	up_read(&adev->reset_sem);
+	up_write(&adev->reset_sem);
 
 	pm_runtime_mark_last_busy(dev->dev);
 	pm_runtime_put_autosuspend(dev->dev);
@@ -1582,9 +1734,7 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 		if (!ring)
 			continue;
 
-		if (amdgpu_debugfs_ring_init(adev, ring)) {
-			DRM_ERROR("Failed to register debugfs file for rings !\n");
-		}
+		amdgpu_debugfs_ring_init(adev, ring);
 	}
 
 	amdgpu_ras_debugfs_create_all(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.h
index 141a8474e24f..6d4965b2d01e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.h
@@ -22,7 +22,6 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-
 /*
  * Debugfs
  */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index ab3794c42d36..48089dc0180b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2745,6 +2745,11 @@ static int amdgpu_device_ip_fini_early(struct amdgpu_device *adev)
 		adev->ip_blocks[i].status.hw = false;
 	}
 
+	if (amdgpu_sriov_vf(adev)) {
+		if (amdgpu_virt_release_full_gpu(adev, false))
+			DRM_ERROR("failed to release exclusive mode on fini\n");
+	}
+
 	return 0;
 }
 
@@ -2805,10 +2810,6 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 
 	amdgpu_ras_fini(adev);
 
-	if (amdgpu_sriov_vf(adev))
-		if (amdgpu_virt_release_full_gpu(adev, false))
-			DRM_ERROR("failed to release exclusive mode on fini\n");
-
 	return 0;
 }
 
@@ -3538,17 +3539,6 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	DRM_INFO("register mmio base: 0x%08X\n", (uint32_t)adev->rmmio_base);
 	DRM_INFO("register mmio size: %u\n", (unsigned)adev->rmmio_size);
 
-	/* enable PCIE atomic ops */
-	r = pci_enable_atomic_ops_to_root(adev->pdev,
-					  PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
-					  PCI_EXP_DEVCAP2_ATOMIC_COMP64);
-	if (r) {
-		adev->have_atomics_support = false;
-		DRM_INFO("PCIE atomic ops is not supported\n");
-	} else {
-		adev->have_atomics_support = true;
-	}
-
 	amdgpu_device_get_pcie_info(adev);
 
 	if (amdgpu_mcbp)
@@ -3571,6 +3561,19 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (r)
 		return r;
 
+	/* enable PCIE atomic ops */
+	if (amdgpu_sriov_vf(adev))
+		adev->have_atomics_support = ((struct amd_sriov_msg_pf2vf_info *)
+			adev->virt.fw_reserve.p_pf2vf)->pcie_atomic_ops_enabled_flags ==
+			(PCI_EXP_DEVCAP2_ATOMIC_COMP32 | PCI_EXP_DEVCAP2_ATOMIC_COMP64);
+	else
+		adev->have_atomics_support =
+			!pci_enable_atomic_ops_to_root(adev->pdev,
+					  PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
+					  PCI_EXP_DEVCAP2_ATOMIC_COMP64);
+	if (!adev->have_atomics_support)
+		dev_info(adev->dev, "PCIE atomic ops is not supported\n");
+
 	/* doorbell bar mapping and doorbell index init*/
 	amdgpu_device_doorbell_init(adev);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index f18240f87387..15b27bcf5273 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -1508,6 +1508,10 @@ static int amdgpu_pmops_resume(struct device *dev)
 	struct amdgpu_device *adev = drm_to_adev(drm_dev);
 	int r;
 
+	/* Avoids registers access if device is physically gone */
+	if (!pci_device_is_present(adev->pdev))
+		adev->no_hw_access = true;
+
 	r = amdgpu_device_resume(drm_dev, true);
 	if (amdgpu_acpi_is_s0ix_active(adev))
 		adev->in_s0ix = false;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
index 76efd5f8950f..d7e4f4660acf 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
@@ -34,6 +34,7 @@
 #include <asm/set_memory.h>
 #endif
 #include "amdgpu.h"
+#include <drm/drm_drv.h>
 
 /*
  * GART
@@ -230,12 +231,16 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset,
 	u64 page_base;
 	/* Starting from VEGA10, system bit must be 0 to mean invalid. */
 	uint64_t flags = 0;
+	int idx;
 
 	if (!adev->gart.ready) {
 		WARN(1, "trying to unbind memory from uninitialized GART !\n");
 		return -EINVAL;
 	}
 
+	if (!drm_dev_enter(&adev->ddev, &idx))
+		return 0;
+
 	t = offset / AMDGPU_GPU_PAGE_SIZE;
 	p = t / AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 	for (i = 0; i < pages; i++, p++) {
@@ -254,6 +259,7 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset,
 	for (i = 0; i < adev->num_vmhubs; i++)
 		amdgpu_gmc_flush_gpu_tlb(adev, 0, i, 0);
 
+	drm_dev_exit(idx);
 	return 0;
 }
 
@@ -276,12 +282,16 @@ int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset,
 {
 	uint64_t page_base;
 	unsigned i, j, t;
+	int idx;
 
 	if (!adev->gart.ready) {
 		WARN(1, "trying to bind memory to uninitialized GART !\n");
 		return -EINVAL;
 	}
 
+	if (!drm_dev_enter(&adev->ddev, &idx))
+		return 0;
+
 	t = offset / AMDGPU_GPU_PAGE_SIZE;
 
 	for (i = 0; i < pages; i++) {
@@ -291,6 +301,7 @@ int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset,
 			page_base += AMDGPU_GPU_PAGE_SIZE;
 		}
 	}
+	drm_dev_exit(idx);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index d6aa032890ee..a573424a6e0b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -60,10 +60,9 @@ static vm_fault_t amdgpu_gem_fault(struct vm_fault *vmf)
 			goto unlock;
 		}
 
-		 ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
-						TTM_BO_VM_NUM_PREFAULT, 1);
-
-		 drm_dev_exit(idx);
+		ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+					       TTM_BO_VM_NUM_PREFAULT, 1);
+		drm_dev_exit(idx);
 	} else {
 		ret = ttm_bo_vm_dummy_page(vmf, vmf->vma->vm_page_prot);
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
index d43fe2ed8116..f851196c83a5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
@@ -42,10 +42,9 @@
 #define AMDGPU_MAX_GFX_QUEUES KGD_MAX_QUEUES
 #define AMDGPU_MAX_COMPUTE_QUEUES KGD_MAX_QUEUES
 
-enum gfx_pipe_priority {
-	AMDGPU_GFX_PIPE_PRIO_NORMAL = 1,
-	AMDGPU_GFX_PIPE_PRIO_HIGH,
-	AMDGPU_GFX_PIPE_PRIO_MAX
+enum amdgpu_gfx_pipe_priority {
+	AMDGPU_GFX_PIPE_PRIO_NORMAL = AMDGPU_RING_PRIO_1,
+	AMDGPU_GFX_PIPE_PRIO_HIGH = AMDGPU_RING_PRIO_2
 };
 
 /* Argument for PPSMC_MSG_GpuChangeState */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index 9ff600a38559..a0dec7f211f0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -153,10 +153,6 @@ int amdgpu_gmc_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
 {
 	void __iomem *ptr = (void *)cpu_pt_addr;
 	uint64_t value;
-	int idx;
-
-	if (!drm_dev_enter(&adev->ddev, &idx))
-		return 0;
 
 	/*
 	 * The following is for PTE only. GART does not have PDEs.
@@ -165,8 +161,6 @@ int amdgpu_gmc_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
 	value |= flags;
 	writeq(value, ptr + (gpu_page_idx * 8));
 
-	drm_dev_exit(idx);
-
 	return 0;
 }
 
@@ -749,6 +743,10 @@ void amdgpu_gmc_init_pdb0(struct amdgpu_device *adev)
 		adev->gmc.xgmi.physical_node_id * adev->gmc.xgmi.node_segment_size;
 	u64 vram_end = vram_addr + vram_size;
 	u64 gart_ptb_gpu_pa = amdgpu_gmc_vram_pa(adev, adev->gart.bo);
+	int idx;
+
+	if (!drm_dev_enter(&adev->ddev, &idx))
+		return;
 
 	flags |= AMDGPU_PTE_VALID | AMDGPU_PTE_READABLE;
 	flags |= AMDGPU_PTE_WRITEABLE;
@@ -770,6 +768,7 @@ void amdgpu_gmc_init_pdb0(struct amdgpu_device *adev)
 	flags |= AMDGPU_PDE_BFS(0) | AMDGPU_PTE_SNOOPED;
 	/* Requires gart_ptb_gpu_pa to be 4K aligned */
 	amdgpu_gmc_set_pte_pde(adev, adev->gmc.ptr_pdb0, i, gart_ptb_gpu_pa, flags);
+	drm_dev_exit(idx);
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index c076a6b9a5a2..bc1297dcdf97 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -300,20 +300,15 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
  */
 int amdgpu_ib_pool_init(struct amdgpu_device *adev)
 {
-	unsigned size;
 	int r, i;
 
 	if (adev->ib_pool_ready)
 		return 0;
 
 	for (i = 0; i < AMDGPU_IB_POOL_MAX; i++) {
-		if (i == AMDGPU_IB_POOL_DIRECT)
-			size = PAGE_SIZE * 6;
-		else
-			size = AMDGPU_IB_POOL_SIZE;
-
 		r = amdgpu_sa_bo_manager_init(adev, &adev->ib_pools[i],
-					      size, AMDGPU_GPU_PAGE_SIZE,
+					      AMDGPU_IB_POOL_SIZE,
+					      AMDGPU_GPU_PAGE_SIZE,
 					      AMDGPU_GEM_DOMAIN_GTT);
 		if (r)
 			goto error;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 7e45640fbee0..d2955ea4a62b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -341,27 +341,34 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 		switch (query_fw->index) {
 		case TA_FW_TYPE_PSP_XGMI:
 			fw_info->ver = adev->psp.ta_fw_version;
-			fw_info->feature = adev->psp.xgmi.feature_version;
+			fw_info->feature = adev->psp.xgmi_context.context
+						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_RAS:
 			fw_info->ver = adev->psp.ta_fw_version;
-			fw_info->feature = adev->psp.ras.feature_version;
+			fw_info->feature = adev->psp.ras_context.context
+						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_HDCP:
 			fw_info->ver = adev->psp.ta_fw_version;
-			fw_info->feature = adev->psp.hdcp.feature_version;
+			fw_info->feature = adev->psp.hdcp_context.context
+						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_DTM:
 			fw_info->ver = adev->psp.ta_fw_version;
-			fw_info->feature = adev->psp.dtm.feature_version;
+			fw_info->feature = adev->psp.dtm_context.context
+						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_RAP:
 			fw_info->ver = adev->psp.ta_fw_version;
-			fw_info->feature = adev->psp.rap.feature_version;
+			fw_info->feature = adev->psp.rap_context.context
+						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_SECUREDISPLAY:
 			fw_info->ver = adev->psp.ta_fw_version;
-			fw_info->feature = adev->psp.securedisplay.feature_version;
+			fw_info->feature =
+				adev->psp.securedisplay_context.context.bin_desc
+					.feature_version;
 			break;
 		default:
 			return -EINVAL;
@@ -378,8 +385,8 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 		fw_info->feature = adev->psp.sos.feature_version;
 		break;
 	case AMDGPU_INFO_FW_ASD:
-		fw_info->ver = adev->psp.asd.fw_version;
-		fw_info->feature = adev->psp.asd.feature_version;
+		fw_info->ver = adev->psp.asd_context.bin_desc.fw_version;
+		fw_info->feature = adev->psp.asd_context.bin_desc.feature_version;
 		break;
 	case AMDGPU_INFO_FW_DMCU:
 		fw_info->ver = adev->dm.dmcu_fw_version;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c
index a2d3dbbf7d25..ce538f4819f9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c
@@ -31,7 +31,7 @@ void amdgpu_mca_query_correctable_error_count(struct amdgpu_device *adev,
 					      uint64_t mc_status_addr,
 					      unsigned long *error_count)
 {
-	uint64_t mc_status = RREG64_PCIE(mc_status_addr * 4);
+	uint64_t mc_status = RREG64_PCIE(mc_status_addr);
 
 	if (REG_GET_FIELD(mc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 &&
 	    REG_GET_FIELD(mc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1)
@@ -42,7 +42,7 @@ void amdgpu_mca_query_uncorrectable_error_count(struct amdgpu_device *adev,
 						uint64_t mc_status_addr,
 						unsigned long *error_count)
 {
-	uint64_t mc_status = RREG64_PCIE(mc_status_addr * 4);
+	uint64_t mc_status = RREG64_PCIE(mc_status_addr);
 
 	if ((REG_GET_FIELD(mc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1) &&
 	    (REG_GET_FIELD(mc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Deferred) == 1 ||
@@ -56,7 +56,7 @@ void amdgpu_mca_query_uncorrectable_error_count(struct amdgpu_device *adev,
 void amdgpu_mca_reset_error_count(struct amdgpu_device *adev,
 				  uint64_t mc_status_addr)
 {
-	WREG64_PCIE(mc_status_addr * 4, 0x0ULL);
+	WREG64_PCIE(mc_status_addr, 0x0ULL);
 }
 
 void amdgpu_mca_query_ras_error_count(struct amdgpu_device *adev,
@@ -87,8 +87,8 @@ int amdgpu_mca_ras_late_init(struct amdgpu_device *adev,
 		if (!mca_dev->ras_if)
 			return -ENOMEM;
 		mca_dev->ras_if->block = mca_dev->ras_funcs->ras_block;
+		mca_dev->ras_if->sub_block_index = mca_dev->ras_funcs->ras_sub_block;
 		mca_dev->ras_if->type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE;
-		mca_dev->ras_if->sub_block_index = 0;
 	}
 	ih_info.head = fs_info.head = *mca_dev->ras_if;
 	r = amdgpu_ras_late_init(adev, mca_dev->ras_if,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h
index f860f2f0e296..c74bc7177066 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h
@@ -29,6 +29,7 @@ struct amdgpu_mca_ras_funcs {
 	void (*query_ras_error_address)(struct amdgpu_device *adev,
 					void *ras_error_status);
 	uint32_t ras_block;
+	uint32_t ras_sub_block;
 	const char* sysfs_name;
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 01a78c786536..32f36992291e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -695,40 +695,6 @@ int amdgpu_bo_create_vm(struct amdgpu_device *adev,
 }
 
 /**
- * amdgpu_bo_validate - validate an &amdgpu_bo buffer object
- * @bo: pointer to the buffer object
- *
- * Sets placement according to domain; and changes placement and caching
- * policy of the buffer object according to the placement.
- * This is used for validating shadow bos.  It calls ttm_bo_validate() to
- * make sure the buffer is resident where it needs to be.
- *
- * Returns:
- * 0 for success or a negative error code on failure.
- */
-int amdgpu_bo_validate(struct amdgpu_bo *bo)
-{
-	struct ttm_operation_ctx ctx = { false, false };
-	uint32_t domain;
-	int r;
-
-	if (bo->tbo.pin_count)
-		return 0;
-
-	domain = bo->preferred_domains;
-
-retry:
-	amdgpu_bo_placement_from_domain(bo, domain);
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
-		domain = bo->allowed_domains;
-		goto retry;
-	}
-
-	return r;
-}
-
-/**
  * amdgpu_bo_add_to_shadow_list - add a BO to the shadow list
  *
  * @vmbo: BO that will be inserted into the shadow list
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 9d6c001c15f8..d2cbabaf2a46 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -327,7 +327,6 @@ int amdgpu_bo_sync_wait_resv(struct amdgpu_device *adev, struct dma_resv *resv,
 int amdgpu_bo_sync_wait(struct amdgpu_bo *bo, void *owner, bool intr);
 u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo);
 u64 amdgpu_bo_gpu_offset_no_check(struct amdgpu_bo *bo);
-int amdgpu_bo_validate(struct amdgpu_bo *bo);
 void amdgpu_bo_get_memory(struct amdgpu_bo *bo, uint64_t *vram_mem,
 				uint64_t *gtt_mem, uint64_t *cpu_mem);
 void amdgpu_bo_add_to_shadow_list(struct amdgpu_bo_vm *vmbo);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index 9b41cb8c3de5..55ffc3da89ce 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -46,6 +46,10 @@ static int psp_sysfs_init(struct amdgpu_device *adev);
 static void psp_sysfs_fini(struct amdgpu_device *adev);
 
 static int psp_load_smu_fw(struct psp_context *psp);
+static int psp_ta_unload(struct psp_context *psp, struct ta_context *context);
+static int psp_ta_load(struct psp_context *psp, struct ta_context *context);
+static int psp_rap_terminate(struct psp_context *psp);
+static int psp_securedisplay_terminate(struct psp_context *psp);
 
 /*
  * Due to DF Cstate management centralized to PMFW, the firmware
@@ -778,46 +782,29 @@ static int psp_rl_load(struct amdgpu_device *adev)
 	return ret;
 }
 
-static void psp_prep_asd_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-				uint64_t asd_mc, uint32_t size)
+static int psp_asd_load(struct psp_context *psp)
 {
-	cmd->cmd_id = GFX_CMD_ID_LOAD_ASD;
-	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(asd_mc);
-	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(asd_mc);
-	cmd->cmd.cmd_load_ta.app_len = size;
-
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = 0;
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = 0;
-	cmd->cmd.cmd_load_ta.cmd_buf_len = 0;
+	return psp_ta_load(psp, &psp->asd_context);
 }
 
-static int psp_asd_load(struct psp_context *psp)
+static int psp_asd_initialize(struct psp_context *psp)
 {
 	int ret;
-	struct psp_gfx_cmd_resp *cmd;
 
 	/* If PSP version doesn't match ASD version, asd loading will be failed.
 	 * add workaround to bypass it for sriov now.
 	 * TODO: add version check to make it common
 	 */
-	if (amdgpu_sriov_vf(psp->adev) || !psp->asd.size_bytes)
+	if (amdgpu_sriov_vf(psp->adev) || !psp->asd_context.bin_desc.size_bytes)
 		return 0;
 
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_copy_fw(psp, psp->asd.start_addr, psp->asd.size_bytes);
+	psp->asd_context.mem_context.shared_mc_addr  = 0;
+	psp->asd_context.mem_context.shared_mem_size = PSP_ASD_SHARED_MEM_SIZE;
+	psp->asd_context.ta_load_type                = GFX_CMD_ID_LOAD_ASD;
 
-	psp_prep_asd_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
-				  psp->asd.size_bytes);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-				 psp->fence_buf_mc_addr);
-	if (!ret) {
-		psp->asd_context.asd_initialized = true;
-		psp->asd_context.session_id = cmd->resp.session_id;
-	}
-
-	release_psp_cmd_buf(psp);
+	ret = psp_asd_load(psp);
+	if (!ret)
+		psp->asd_context.initialized = true;
 
 	return ret;
 }
@@ -829,27 +816,39 @@ static void psp_prep_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
 	cmd->cmd.cmd_unload_ta.session_id = session_id;
 }
 
+static int psp_ta_unload(struct psp_context *psp, struct ta_context *context)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd = acquire_psp_cmd_buf(psp);
+
+	psp_prep_ta_unload_cmd_buf(cmd, context->session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	release_psp_cmd_buf(psp);
+
+	return ret;
+}
+
 static int psp_asd_unload(struct psp_context *psp)
 {
+	return psp_ta_unload(psp, &psp->asd_context);
+}
+
+static int psp_asd_terminate(struct psp_context *psp)
+{
 	int ret;
-	struct psp_gfx_cmd_resp *cmd;
 
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	if (!psp->asd_context.asd_initialized)
+	if (!psp->asd_context.initialized)
 		return 0;
 
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->asd_context.session_id);
+	ret = psp_asd_unload(psp);
 
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-				 psp->fence_buf_mc_addr);
 	if (!ret)
-		psp->asd_context.asd_initialized = false;
-
-	release_psp_cmd_buf(psp);
+		psp->asd_context.initialized = false;
 
 	return ret;
 }
@@ -885,23 +884,22 @@ int psp_reg_program(struct psp_context *psp, enum psp_reg_prog_id reg,
 
 static void psp_prep_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
 				     uint64_t ta_bin_mc,
-				     uint32_t ta_bin_size,
-				     uint64_t ta_shared_mc,
-				     uint32_t ta_shared_size)
+				     struct ta_context *context)
 {
-	cmd->cmd_id				= GFX_CMD_ID_LOAD_TA;
+	cmd->cmd_id				= context->ta_load_type;
 	cmd->cmd.cmd_load_ta.app_phy_addr_lo 	= lower_32_bits(ta_bin_mc);
 	cmd->cmd.cmd_load_ta.app_phy_addr_hi	= upper_32_bits(ta_bin_mc);
-	cmd->cmd.cmd_load_ta.app_len		= ta_bin_size;
+	cmd->cmd.cmd_load_ta.app_len		= context->bin_desc.size_bytes;
 
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(ta_shared_mc);
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(ta_shared_mc);
-	cmd->cmd.cmd_load_ta.cmd_buf_len	 = ta_shared_size;
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo =
+		lower_32_bits(context->mem_context.shared_mc_addr);
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi =
+		upper_32_bits(context->mem_context.shared_mc_addr);
+	cmd->cmd.cmd_load_ta.cmd_buf_len = context->mem_context.shared_mem_size;
 }
 
 static int psp_ta_init_shared_buf(struct psp_context *psp,
-				  struct ta_mem_context *mem_ctx,
-				  uint32_t shared_mem_size)
+				  struct ta_mem_context *mem_ctx)
 {
 	int ret;
 
@@ -909,8 +907,8 @@ static int psp_ta_init_shared_buf(struct psp_context *psp,
 	* Allocate 16k memory aligned to 4k from Frame Buffer (local
 	* physical) for ta to host memory
 	*/
-	ret = amdgpu_bo_create_kernel(psp->adev, shared_mem_size, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
+	ret = amdgpu_bo_create_kernel(psp->adev, mem_ctx->shared_mem_size,
+				      PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
 				      &mem_ctx->shared_bo,
 				      &mem_ctx->shared_mc_addr,
 				      &mem_ctx->shared_buf);
@@ -926,8 +924,7 @@ static void psp_ta_free_shared_buf(struct ta_mem_context *mem_ctx)
 
 static int psp_xgmi_init_shared_buf(struct psp_context *psp)
 {
-	return psp_ta_init_shared_buf(psp, &psp->xgmi_context.context.mem_context,
-				      PSP_XGMI_SHARED_MEM_SIZE);
+	return psp_ta_init_shared_buf(psp, &psp->xgmi_context.context.mem_context);
 }
 
 static void psp_prep_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
@@ -956,31 +953,23 @@ static int psp_ta_invoke(struct psp_context *psp,
 	return ret;
 }
 
-static int psp_xgmi_load(struct psp_context *psp)
+static int psp_ta_load(struct psp_context *psp, struct ta_context *context)
 {
 	int ret;
 	struct psp_gfx_cmd_resp *cmd;
 
-	/*
-	 * TODO: bypass the loading in sriov for now
-	 */
-
 	cmd = acquire_psp_cmd_buf(psp);
 
-	psp_copy_fw(psp, psp->xgmi.start_addr, psp->xgmi.size_bytes);
+	psp_copy_fw(psp, context->bin_desc.start_addr,
+		    context->bin_desc.size_bytes);
 
-	psp_prep_ta_load_cmd_buf(cmd,
-				 psp->fw_pri_mc_addr,
-				 psp->xgmi.size_bytes,
-				 psp->xgmi_context.context.mem_context.shared_mc_addr,
-				 PSP_XGMI_SHARED_MEM_SIZE);
+	psp_prep_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr, context);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 				 psp->fence_buf_mc_addr);
 
 	if (!ret) {
-		psp->xgmi_context.context.initialized = true;
-		psp->xgmi_context.context.session_id = cmd->resp.session_id;
+		context->session_id = cmd->resp.session_id;
 	}
 
 	release_psp_cmd_buf(psp);
@@ -988,31 +977,14 @@ static int psp_xgmi_load(struct psp_context *psp)
 	return ret;
 }
 
-static int psp_xgmi_unload(struct psp_context *psp)
+static int psp_xgmi_load(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-	struct amdgpu_device *adev = psp->adev;
-
-	/* XGMI TA unload currently is not supported on Arcturus/Aldebaran A+A */
-	if (adev->asic_type == CHIP_ARCTURUS ||
-		(adev->asic_type == CHIP_ALDEBARAN && adev->gmc.xgmi.connected_to_cpu))
-		return 0;
-
-	/*
-	 * TODO: bypass the unloading in sriov for now
-	 */
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->xgmi_context.context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-				 psp->fence_buf_mc_addr);
-
-	release_psp_cmd_buf(psp);
+	return psp_ta_load(psp, &psp->xgmi_context.context);
+}
 
-	return ret;
+static int psp_xgmi_unload(struct psp_context *psp)
+{
+	return psp_ta_unload(psp, &psp->xgmi_context.context);
 }
 
 int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
@@ -1023,6 +995,12 @@ int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 int psp_xgmi_terminate(struct psp_context *psp)
 {
 	int ret;
+	struct amdgpu_device *adev = psp->adev;
+
+	/* XGMI TA unload currently is not supported on Arcturus/Aldebaran A+A */
+	if (adev->asic_type == CHIP_ARCTURUS ||
+		(adev->asic_type == CHIP_ALDEBARAN && adev->gmc.xgmi.connected_to_cpu))
+		return 0;
 
 	if (!psp->xgmi_context.context.initialized)
 		return 0;
@@ -1045,13 +1023,16 @@ int psp_xgmi_initialize(struct psp_context *psp, bool set_extended_data, bool lo
 	int ret;
 
 	if (!psp->ta_fw ||
-	    !psp->xgmi.size_bytes ||
-	    !psp->xgmi.start_addr)
+	    !psp->xgmi_context.context.bin_desc.size_bytes ||
+	    !psp->xgmi_context.context.bin_desc.start_addr)
 		return -ENOENT;
 
 	if (!load_ta)
 		goto invoke;
 
+	psp->xgmi_context.context.mem_context.shared_mem_size = PSP_XGMI_SHARED_MEM_SIZE;
+	psp->xgmi_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
+
 	if (!psp->xgmi_context.context.initialized) {
 		ret = psp_xgmi_init_shared_buf(psp);
 		if (ret)
@@ -1060,7 +1041,9 @@ int psp_xgmi_initialize(struct psp_context *psp, bool set_extended_data, bool lo
 
 	/* Load XGMI TA */
 	ret = psp_xgmi_load(psp);
-	if (ret)
+	if (!ret)
+		psp->xgmi_context.context.initialized = true;
+	else
 		return ret;
 
 invoke:
@@ -1118,7 +1101,7 @@ int psp_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id)
 static bool psp_xgmi_peer_link_info_supported(struct psp_context *psp)
 {
 	return psp->adev->asic_type == CHIP_ALDEBARAN &&
-				psp->xgmi.feature_version >= 0x2000000b;
+				psp->xgmi_context.context.bin_desc.feature_version >= 0x2000000b;
 }
 
 /*
@@ -1282,80 +1265,17 @@ int psp_xgmi_set_topology_info(struct psp_context *psp,
 // ras begin
 static int psp_ras_init_shared_buf(struct psp_context *psp)
 {
-	return psp_ta_init_shared_buf(psp, &psp->ras_context.context.mem_context,
-				      PSP_RAS_SHARED_MEM_SIZE);
+	return psp_ta_init_shared_buf(psp, &psp->ras_context.context.mem_context);
 }
 
 static int psp_ras_load(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-	struct ta_ras_shared_memory *ras_cmd;
-
-	/*
-	 * TODO: bypass the loading in sriov for now
-	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	psp_copy_fw(psp, psp->ras.start_addr, psp->ras.size_bytes);
-
-	ras_cmd = (struct ta_ras_shared_memory *)psp->ras_context.context.mem_context.shared_buf;
-
-	if (psp->adev->gmc.xgmi.connected_to_cpu)
-		ras_cmd->ras_in_message.init_flags.poison_mode_en = 1;
-	else
-		ras_cmd->ras_in_message.init_flags.dgpu_mode = 1;
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_load_cmd_buf(cmd,
-				 psp->fw_pri_mc_addr,
-				 psp->ras.size_bytes,
-				 psp->ras_context.context.mem_context.shared_mc_addr,
-				 PSP_RAS_SHARED_MEM_SIZE);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-			psp->fence_buf_mc_addr);
-
-	if (!ret) {
-		psp->ras_context.context.session_id = cmd->resp.session_id;
-
-		if (!ras_cmd->ras_status)
-			psp->ras_context.context.initialized = true;
-		else
-			dev_warn(psp->adev->dev, "RAS Init Status: 0x%X\n", ras_cmd->ras_status);
-	}
-
-	release_psp_cmd_buf(psp);
-
-	if (ret || ras_cmd->ras_status)
-		amdgpu_ras_fini(psp->adev);
-
-	return ret;
+	return psp_ta_load(psp, &psp->ras_context.context);
 }
 
 static int psp_ras_unload(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	/*
-	 * TODO: bypass the unloading in sriov for now
-	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->ras_context.context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-			psp->fence_buf_mc_addr);
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_unload(psp, &psp->ras_context.context);
 }
 
 int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
@@ -1391,31 +1311,11 @@ int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 		else if (ras_cmd->ras_out_message.flags.reg_access_failure_flag)
 			dev_warn(psp->adev->dev,
 				 "RAS internal register access blocked\n");
-	}
 
-	return ret;
-}
-
-static int psp_ras_status_to_errno(struct amdgpu_device *adev,
-					 enum ta_ras_status ras_status)
-{
-	int ret = -EINVAL;
-
-	switch (ras_status) {
-	case TA_RAS_STATUS__SUCCESS:
-		ret = 0;
-		break;
-	case TA_RAS_STATUS__RESET_NEEDED:
-		ret = -EAGAIN;
-		break;
-	case TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE:
-		dev_warn(adev->dev, "RAS WARN: ras function unavailable\n");
-		break;
-	case TA_RAS_STATUS__ERROR_ASD_READ_WRITE:
-		dev_warn(adev->dev, "RAS WARN: asd read or write failed\n");
-		break;
-	default:
-		dev_err(adev->dev, "RAS ERROR: ras function failed ret 0x%X\n", ret);
+		if (ras_cmd->ras_status == TA_RAS_STATUS__ERROR_UNSUPPORTED_IP)
+		    dev_warn(psp->adev->dev, "RAS WARNING: cmd failed due to unsupported ip\n");
+		else if (ras_cmd->ras_status)
+		    dev_warn(psp->adev->dev, "RAS WARNING: ras status = 0x%X\n", ras_cmd->ras_status);
 	}
 
 	return ret;
@@ -1444,7 +1344,7 @@ int psp_ras_enable_features(struct psp_context *psp,
 	if (ret)
 		return -EINVAL;
 
-	return psp_ras_status_to_errno(psp->adev, ras_cmd->ras_status);
+	return 0;
 }
 
 static int psp_ras_terminate(struct psp_context *psp)
@@ -1477,6 +1377,7 @@ static int psp_ras_initialize(struct psp_context *psp)
 	int ret;
 	uint32_t boot_cfg = 0xFF;
 	struct amdgpu_device *adev = psp->adev;
+	struct ta_ras_shared_memory *ras_cmd;
 
 	/*
 	 * TODO: bypass the initialize in sriov for now
@@ -1484,8 +1385,8 @@ static int psp_ras_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(adev))
 		return 0;
 
-	if (!adev->psp.ras.size_bytes ||
-	    !adev->psp.ras.start_addr) {
+	if (!adev->psp.ras_context.context.bin_desc.size_bytes ||
+	    !adev->psp.ras_context.context.bin_desc.start_addr) {
 		dev_info(adev->dev, "RAS: optional ras ta ucode is not available\n");
 		return 0;
 	}
@@ -1531,17 +1432,34 @@ static int psp_ras_initialize(struct psp_context *psp)
 		}
 	}
 
+	psp->ras_context.context.mem_context.shared_mem_size = PSP_RAS_SHARED_MEM_SIZE;
+	psp->ras_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
+
 	if (!psp->ras_context.context.initialized) {
 		ret = psp_ras_init_shared_buf(psp);
 		if (ret)
 			return ret;
 	}
 
+	ras_cmd = (struct ta_ras_shared_memory *)psp->ras_context.context.mem_context.shared_buf;
+	memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory));
+
+	if (psp->adev->gmc.xgmi.connected_to_cpu)
+		ras_cmd->ras_in_message.init_flags.poison_mode_en = 1;
+	else
+		ras_cmd->ras_in_message.init_flags.dgpu_mode = 1;
+
 	ret = psp_ras_load(psp);
-	if (ret)
-		return ret;
 
-	return 0;
+	if (!ret && !ras_cmd->ras_status)
+		psp->ras_context.context.initialized = true;
+	else {
+		if (ras_cmd->ras_status)
+			dev_warn(psp->adev->dev, "RAS Init Status: 0x%X\n", ras_cmd->ras_status);
+		amdgpu_ras_fini(psp->adev);
+	}
+
+	return ret;
 }
 
 int psp_ras_trigger_error(struct psp_context *psp,
@@ -1568,51 +1486,24 @@ int psp_ras_trigger_error(struct psp_context *psp,
 	if (amdgpu_ras_intr_triggered())
 		return 0;
 
-	return psp_ras_status_to_errno(psp->adev, ras_cmd->ras_status);
+	if (ras_cmd->ras_status)
+		return -EINVAL;
+
+	return 0;
 }
 // ras end
 
 // HDCP start
 static int psp_hdcp_init_shared_buf(struct psp_context *psp)
 {
-	return psp_ta_init_shared_buf(psp, &psp->hdcp_context.context.mem_context,
-				      PSP_HDCP_SHARED_MEM_SIZE);
+	return psp_ta_init_shared_buf(psp, &psp->hdcp_context.context.mem_context);
 }
 
 static int psp_hdcp_load(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	/*
-	 * TODO: bypass the loading in sriov for now
-	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	psp_copy_fw(psp, psp->hdcp.start_addr,
-		    psp->hdcp.size_bytes);
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_load_cmd_buf(cmd,
-				 psp->fw_pri_mc_addr,
-				 psp->hdcp.size_bytes,
-				 psp->hdcp_context.context.mem_context.shared_mc_addr,
-				 PSP_HDCP_SHARED_MEM_SIZE);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	if (!ret) {
-		psp->hdcp_context.context.initialized = true;
-		psp->hdcp_context.context.session_id = cmd->resp.session_id;
-		mutex_init(&psp->hdcp_context.mutex);
-	}
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_load(psp, &psp->hdcp_context.context);
 }
+
 static int psp_hdcp_initialize(struct psp_context *psp)
 {
 	int ret;
@@ -1623,12 +1514,15 @@ static int psp_hdcp_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	if (!psp->hdcp.size_bytes ||
-	    !psp->hdcp.start_addr) {
+	if (!psp->hdcp_context.context.bin_desc.size_bytes ||
+	    !psp->hdcp_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "HDCP: optional hdcp ta ucode is not available\n");
 		return 0;
 	}
 
+	psp->hdcp_context.context.mem_context.shared_mem_size = PSP_HDCP_SHARED_MEM_SIZE;
+	psp->hdcp_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
+
 	if (!psp->hdcp_context.context.initialized) {
 		ret = psp_hdcp_init_shared_buf(psp);
 		if (ret)
@@ -1636,32 +1530,17 @@ static int psp_hdcp_initialize(struct psp_context *psp)
 	}
 
 	ret = psp_hdcp_load(psp);
-	if (ret)
-		return ret;
+	if (!ret) {
+		psp->hdcp_context.context.initialized = true;
+		mutex_init(&psp->hdcp_context.mutex);
+	}
 
-	return 0;
+	return ret;
 }
 
 static int psp_hdcp_unload(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	/*
-	 * TODO: bypass the unloading in sriov for now
-	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->hdcp_context.context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_unload(psp, &psp->hdcp_context.context);
 }
 
 int psp_hdcp_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
@@ -1709,42 +1588,12 @@ out:
 // DTM start
 static int psp_dtm_init_shared_buf(struct psp_context *psp)
 {
-	return psp_ta_init_shared_buf(psp, &psp->dtm_context.context.mem_context,
-				      PSP_DTM_SHARED_MEM_SIZE);
+	return psp_ta_init_shared_buf(psp, &psp->dtm_context.context.mem_context);
 }
 
 static int psp_dtm_load(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	/*
-	 * TODO: bypass the loading in sriov for now
-	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	psp_copy_fw(psp, psp->dtm.start_addr, psp->dtm.size_bytes);
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_load_cmd_buf(cmd,
-				 psp->fw_pri_mc_addr,
-				 psp->dtm.size_bytes,
-				 psp->dtm_context.context.mem_context.shared_mc_addr,
-				 PSP_DTM_SHARED_MEM_SIZE);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	if (!ret) {
-		psp->dtm_context.context.initialized = true;
-		psp->dtm_context.context.session_id = cmd->resp.session_id;
-		mutex_init(&psp->dtm_context.mutex);
-	}
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_load(psp, &psp->dtm_context.context);
 }
 
 static int psp_dtm_initialize(struct psp_context *psp)
@@ -1757,12 +1606,15 @@ static int psp_dtm_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	if (!psp->dtm.size_bytes ||
-	    !psp->dtm.start_addr) {
+	if (!psp->dtm_context.context.bin_desc.size_bytes ||
+	    !psp->dtm_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "DTM: optional dtm ta ucode is not available\n");
 		return 0;
 	}
 
+	psp->dtm_context.context.mem_context.shared_mem_size = PSP_DTM_SHARED_MEM_SIZE;
+	psp->dtm_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
+
 	if (!psp->dtm_context.context.initialized) {
 		ret = psp_dtm_init_shared_buf(psp);
 		if (ret)
@@ -1770,32 +1622,17 @@ static int psp_dtm_initialize(struct psp_context *psp)
 	}
 
 	ret = psp_dtm_load(psp);
-	if (ret)
-		return ret;
+	if (!ret) {
+		psp->dtm_context.context.initialized = true;
+		mutex_init(&psp->dtm_context.mutex);
+	}
 
-	return 0;
+	return ret;
 }
 
 static int psp_dtm_unload(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	/*
-	 * TODO: bypass the unloading in sriov for now
-	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->dtm_context.context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_unload(psp, &psp->dtm_context.context);
 }
 
 int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
@@ -1843,50 +1680,17 @@ out:
 // RAP start
 static int psp_rap_init_shared_buf(struct psp_context *psp)
 {
-	return psp_ta_init_shared_buf(psp, &psp->rap_context.context.mem_context,
-				      PSP_RAP_SHARED_MEM_SIZE);
+	return psp_ta_init_shared_buf(psp, &psp->rap_context.context.mem_context);
 }
 
 static int psp_rap_load(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	psp_copy_fw(psp, psp->rap.start_addr, psp->rap.size_bytes);
-
-	cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_load_cmd_buf(cmd,
-				 psp->fw_pri_mc_addr,
-				 psp->rap.size_bytes,
-				 psp->rap_context.context.mem_context.shared_mc_addr,
-				 PSP_RAP_SHARED_MEM_SIZE);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	if (!ret) {
-		psp->rap_context.context.initialized = true;
-		psp->rap_context.context.session_id = cmd->resp.session_id;
-		mutex_init(&psp->rap_context.mutex);
-	}
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_load(psp, &psp->rap_context.context);
 }
 
 static int psp_rap_unload(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->rap_context.context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_unload(psp, &psp->rap_context.context);
 }
 
 static int psp_rap_initialize(struct psp_context *psp)
@@ -1900,12 +1704,15 @@ static int psp_rap_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	if (!psp->rap.size_bytes ||
-	    !psp->rap.start_addr) {
+	if (!psp->rap_context.context.bin_desc.size_bytes ||
+	    !psp->rap_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "RAP: optional rap ta ucode is not available\n");
 		return 0;
 	}
 
+	psp->rap_context.context.mem_context.shared_mem_size = PSP_RAP_SHARED_MEM_SIZE;
+	psp->rap_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
+
 	if (!psp->rap_context.context.initialized) {
 		ret = psp_rap_init_shared_buf(psp);
 		if (ret)
@@ -1913,16 +1720,15 @@ static int psp_rap_initialize(struct psp_context *psp)
 	}
 
 	ret = psp_rap_load(psp);
-	if (ret)
+	if (!ret) {
+		psp->rap_context.context.initialized = true;
+		mutex_init(&psp->rap_context.mutex);
+	} else
 		return ret;
 
 	ret = psp_rap_invoke(psp, TA_CMD_RAP__INITIALIZE, &status);
 	if (ret || status != TA_RAP_STATUS__SUCCESS) {
-		psp_rap_unload(psp);
-
-		psp_ta_free_shared_buf(&psp->rap_context.context.mem_context);
-
-		psp->rap_context.context.initialized = false;
+		psp_rap_terminate(psp);
 
 		dev_warn(psp->adev->dev, "RAP TA initialize fail (%d) status %d.\n",
 			 ret, status);
@@ -1989,49 +1795,17 @@ out_unlock:
 static int psp_securedisplay_init_shared_buf(struct psp_context *psp)
 {
 	return psp_ta_init_shared_buf(
-		psp, &psp->securedisplay_context.context.mem_context,
-		PSP_SECUREDISPLAY_SHARED_MEM_SIZE);
+		psp, &psp->securedisplay_context.context.mem_context);
 }
 
 static int psp_securedisplay_load(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd = acquire_psp_cmd_buf(psp);
-
-	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
-	memcpy(psp->fw_pri_buf, psp->securedisplay.start_addr, psp->securedisplay.size_bytes);
-
-	psp_prep_ta_load_cmd_buf(cmd,
-				 psp->fw_pri_mc_addr,
-				 psp->securedisplay.size_bytes,
-				 psp->securedisplay_context.context.mem_context.shared_mc_addr,
-				 PSP_SECUREDISPLAY_SHARED_MEM_SIZE);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	if (!ret) {
-		psp->securedisplay_context.context.initialized = true;
-		psp->securedisplay_context.context.session_id = cmd->resp.session_id;
-		mutex_init(&psp->securedisplay_context.mutex);
-	}
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_load(psp, &psp->securedisplay_context.context);
 }
 
 static int psp_securedisplay_unload(struct psp_context *psp)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd = acquire_psp_cmd_buf(psp);
-
-	psp_prep_ta_unload_cmd_buf(cmd, psp->securedisplay_context.context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	release_psp_cmd_buf(psp);
-
-	return ret;
+	return psp_ta_unload(psp, &psp->securedisplay_context.context);
 }
 
 static int psp_securedisplay_initialize(struct psp_context *psp)
@@ -2045,12 +1819,16 @@ static int psp_securedisplay_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	if (!psp->securedisplay.size_bytes ||
-	    !psp->securedisplay.start_addr) {
+	if (!psp->securedisplay_context.context.bin_desc.size_bytes ||
+	    !psp->securedisplay_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "SECUREDISPLAY: securedisplay ta ucode is not available\n");
 		return 0;
 	}
 
+	psp->securedisplay_context.context.mem_context.shared_mem_size =
+		PSP_SECUREDISPLAY_SHARED_MEM_SIZE;
+	psp->securedisplay_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
+
 	if (!psp->securedisplay_context.context.initialized) {
 		ret = psp_securedisplay_init_shared_buf(psp);
 		if (ret)
@@ -2058,7 +1836,10 @@ static int psp_securedisplay_initialize(struct psp_context *psp)
 	}
 
 	ret = psp_securedisplay_load(psp);
-	if (ret)
+	if (!ret) {
+		psp->securedisplay_context.context.initialized = true;
+		mutex_init(&psp->securedisplay_context.mutex);
+	} else
 		return ret;
 
 	psp_prep_securedisplay_cmd_buf(psp, &securedisplay_cmd,
@@ -2066,12 +1847,7 @@ static int psp_securedisplay_initialize(struct psp_context *psp)
 
 	ret = psp_securedisplay_invoke(psp, TA_SECUREDISPLAY_COMMAND__QUERY_TA);
 	if (ret) {
-		psp_securedisplay_unload(psp);
-
-		psp_ta_free_shared_buf(&psp->securedisplay_context.context.mem_context);
-
-		psp->securedisplay_context.context.initialized = false;
-
+		psp_securedisplay_terminate(psp);
 		dev_err(psp->adev->dev, "SECUREDISPLAY TA initialize fail.\n");
 		return -EINVAL;
 	}
@@ -2629,7 +2405,7 @@ skip_memalloc:
 	if (ret)
 		goto failed;
 
-	ret = psp_asd_load(psp);
+	ret = psp_asd_initialize(psp);
 	if (ret) {
 		DRM_ERROR("PSP load asd failed!\n");
 		return ret;
@@ -2721,7 +2497,7 @@ static int psp_hw_fini(void *handle)
 		psp_hdcp_terminate(psp);
 	}
 
-	psp_asd_unload(psp);
+	psp_asd_terminate(psp);
 
 	psp_tmr_terminate(psp);
 	psp_ring_destroy(psp, PSP_RING_TYPE__KM);
@@ -2779,9 +2555,9 @@ static int psp_suspend(void *handle)
 		}
 	}
 
-	ret = psp_asd_unload(psp);
+	ret = psp_asd_terminate(psp);
 	if (ret) {
-		DRM_ERROR("Failed to unload asd\n");
+		DRM_ERROR("Failed to terminate asd\n");
 		return ret;
 	}
 
@@ -2826,7 +2602,7 @@ static int psp_resume(void *handle)
 	if (ret)
 		goto failed;
 
-	ret = psp_asd_load(psp);
+	ret = psp_asd_initialize(psp);
 	if (ret) {
 		DRM_ERROR("PSP load asd failed!\n");
 		goto failed;
@@ -2994,10 +2770,10 @@ int psp_init_asd_microcode(struct psp_context *psp,
 		goto out;
 
 	asd_hdr = (const struct psp_firmware_header_v1_0 *)adev->psp.asd_fw->data;
-	adev->psp.asd.fw_version = le32_to_cpu(asd_hdr->header.ucode_version);
-	adev->psp.asd.feature_version = le32_to_cpu(asd_hdr->sos.fw_version);
-	adev->psp.asd.size_bytes = le32_to_cpu(asd_hdr->header.ucode_size_bytes);
-	adev->psp.asd.start_addr = (uint8_t *)asd_hdr +
+	adev->psp.asd_context.bin_desc.fw_version = le32_to_cpu(asd_hdr->header.ucode_version);
+	adev->psp.asd_context.bin_desc.feature_version = le32_to_cpu(asd_hdr->sos.fw_version);
+	adev->psp.asd_context.bin_desc.size_bytes = le32_to_cpu(asd_hdr->header.ucode_size_bytes);
+	adev->psp.asd_context.bin_desc.start_addr = (uint8_t *)asd_hdr +
 				le32_to_cpu(asd_hdr->header.ucode_array_offset_bytes);
 	return 0;
 out:
@@ -3284,40 +3060,43 @@ static int parse_ta_bin_descriptor(struct psp_context *psp,
 
 	switch (desc->fw_type) {
 	case TA_FW_TYPE_PSP_ASD:
-		psp->asd.fw_version        = le32_to_cpu(desc->fw_version);
-		psp->asd.feature_version   = le32_to_cpu(desc->fw_version);
-		psp->asd.size_bytes        = le32_to_cpu(desc->size_bytes);
-		psp->asd.start_addr 	   = ucode_start_addr;
+		psp->asd_context.bin_desc.fw_version        = le32_to_cpu(desc->fw_version);
+		psp->asd_context.bin_desc.feature_version   = le32_to_cpu(desc->fw_version);
+		psp->asd_context.bin_desc.size_bytes        = le32_to_cpu(desc->size_bytes);
+		psp->asd_context.bin_desc.start_addr        = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_XGMI:
-		psp->xgmi.feature_version  = le32_to_cpu(desc->fw_version);
-		psp->xgmi.size_bytes       = le32_to_cpu(desc->size_bytes);
-		psp->xgmi.start_addr       = ucode_start_addr;
+		psp->xgmi_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->xgmi_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
+		psp->xgmi_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_RAS:
-		psp->ras.feature_version   = le32_to_cpu(desc->fw_version);
-		psp->ras.size_bytes        = le32_to_cpu(desc->size_bytes);
-		psp->ras.start_addr        = ucode_start_addr;
+		psp->ras_context.context.bin_desc.feature_version   = le32_to_cpu(desc->fw_version);
+		psp->ras_context.context.bin_desc.size_bytes        = le32_to_cpu(desc->size_bytes);
+		psp->ras_context.context.bin_desc.start_addr        = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_HDCP:
-		psp->hdcp.feature_version  = le32_to_cpu(desc->fw_version);
-		psp->hdcp.size_bytes       = le32_to_cpu(desc->size_bytes);
-		psp->hdcp.start_addr       = ucode_start_addr;
+		psp->hdcp_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->hdcp_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
+		psp->hdcp_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_DTM:
-		psp->dtm.feature_version  = le32_to_cpu(desc->fw_version);
-		psp->dtm.size_bytes       = le32_to_cpu(desc->size_bytes);
-		psp->dtm.start_addr       = ucode_start_addr;
+		psp->dtm_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->dtm_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
+		psp->dtm_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_RAP:
-		psp->rap.feature_version  = le32_to_cpu(desc->fw_version);
-		psp->rap.size_bytes       = le32_to_cpu(desc->size_bytes);
-		psp->rap.start_addr       = ucode_start_addr;
+		psp->rap_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->rap_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
+		psp->rap_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_SECUREDISPLAY:
-		psp->securedisplay.feature_version  = le32_to_cpu(desc->fw_version);
-		psp->securedisplay.size_bytes       = le32_to_cpu(desc->size_bytes);
-		psp->securedisplay.start_addr       = ucode_start_addr;
+		psp->securedisplay_context.context.bin_desc.feature_version =
+			le32_to_cpu(desc->fw_version);
+		psp->securedisplay_context.context.bin_desc.size_bytes =
+			le32_to_cpu(desc->size_bytes);
+		psp->securedisplay_context.context.bin_desc.start_addr =
+			ucode_start_addr;
 		break;
 	default:
 		dev_warn(psp->adev->dev, "Unsupported TA type: %d\n", desc->fw_type);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
index 8ef2d28af92a..f29afabbff1f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
@@ -34,17 +34,20 @@
 
 #define PSP_FENCE_BUFFER_SIZE	0x1000
 #define PSP_CMD_BUFFER_SIZE	0x1000
-#define PSP_XGMI_SHARED_MEM_SIZE 0x4000
-#define PSP_RAS_SHARED_MEM_SIZE 0x4000
 #define PSP_1_MEG		0x100000
 #define PSP_TMR_SIZE(adev)	((adev)->asic_type == CHIP_ALDEBARAN ? 0x800000 : 0x400000)
-#define PSP_HDCP_SHARED_MEM_SIZE	0x4000
-#define PSP_DTM_SHARED_MEM_SIZE	0x4000
-#define PSP_RAP_SHARED_MEM_SIZE	0x4000
-#define PSP_SECUREDISPLAY_SHARED_MEM_SIZE	0x4000
-#define PSP_SHARED_MEM_SIZE		0x4000
 #define PSP_FW_NAME_LEN		0x24
 
+enum psp_shared_mem_size {
+	PSP_ASD_SHARED_MEM_SIZE				= 0x0,
+	PSP_XGMI_SHARED_MEM_SIZE			= 0x4000,
+	PSP_RAS_SHARED_MEM_SIZE				= 0x4000,
+	PSP_HDCP_SHARED_MEM_SIZE			= 0x4000,
+	PSP_DTM_SHARED_MEM_SIZE				= 0x4000,
+	PSP_RAP_SHARED_MEM_SIZE				= 0x4000,
+	PSP_SECUREDISPLAY_SHARED_MEM_SIZE	= 0x4000,
+};
+
 struct psp_context;
 struct psp_xgmi_node_info;
 struct psp_xgmi_topology_info;
@@ -131,21 +134,26 @@ struct psp_xgmi_topology_info {
 	struct psp_xgmi_node_info	nodes[AMDGPU_XGMI_MAX_CONNECTED_NODES];
 };
 
-struct psp_asd_context {
-	bool			asd_initialized;
-	uint32_t		session_id;
+struct psp_bin_desc {
+	uint32_t fw_version;
+	uint32_t feature_version;
+	uint32_t size_bytes;
+	uint8_t *start_addr;
 };
 
 struct ta_mem_context {
 	struct amdgpu_bo		*shared_bo;
 	uint64_t		shared_mc_addr;
 	void			*shared_buf;
+	enum psp_shared_mem_size	shared_mem_size;
 };
 
 struct ta_context {
 	bool			initialized;
 	uint32_t		session_id;
 	struct ta_mem_context	mem_context;
+	struct psp_bin_desc		bin_desc;
+	enum psp_gfx_cmd_id		ta_load_type;
 };
 
 struct ta_cp_context {
@@ -263,13 +271,6 @@ struct psp_runtime_boot_cfg_entry {
 	uint32_t reserved;
 };
 
-struct psp_bin_desc {
-	uint32_t fw_version;
-	uint32_t feature_version;
-	uint32_t size_bytes;
-	uint8_t *start_addr;
-};
-
 struct psp_context
 {
 	struct amdgpu_device            *adev;
@@ -301,7 +302,6 @@ struct psp_context
 
 	/* asd firmware */
 	const struct firmware	*asd_fw;
-	struct psp_bin_desc		asd;
 
 	/* toc firmware */
 	const struct firmware		*toc_fw;
@@ -326,14 +326,8 @@ struct psp_context
 	/* xgmi ta firmware and buffer */
 	const struct firmware		*ta_fw;
 	uint32_t			ta_fw_version;
-	struct psp_bin_desc		xgmi;
-	struct psp_bin_desc		ras;
-	struct psp_bin_desc		hdcp;
-	struct psp_bin_desc		dtm;
-	struct psp_bin_desc		rap;
-	struct psp_bin_desc		securedisplay;
-
-	struct psp_asd_context		asd_context;
+
+	struct ta_context		asd_context;
 	struct psp_xgmi_context		xgmi_context;
 	struct psp_ras_context		ras_context;
 	struct ta_cp_context		hdcp_context;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 96a8fd0ca1df..e1c34eef76b7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -61,8 +61,30 @@ const char *ras_block_string[] = {
 	"mp0",
 	"mp1",
 	"fuse",
+	"mca",
 };
 
+const char *ras_mca_block_string[] = {
+	"mca_mp0",
+	"mca_mp1",
+	"mca_mpio",
+	"mca_iohc",
+};
+
+const char *get_ras_block_str(struct ras_common_if *ras_block)
+{
+	if (!ras_block)
+		return "NULL";
+
+	if (ras_block->block >= AMDGPU_RAS_BLOCK_COUNT)
+		return "OUT OF RANGE";
+
+	if (ras_block->block == AMDGPU_RAS_BLOCK__MCA)
+		return ras_mca_block_string[ras_block->sub_block_index];
+
+	return ras_block_string[ras_block->block];
+}
+
 #define ras_err_str(i) (ras_error_string[ffs(i)])
 
 #define RAS_DEFAULT_FLAGS (AMDGPU_RAS_FLAG_INIT_BY_VBIOS)
@@ -187,7 +209,7 @@ static int amdgpu_ras_find_block_id_by_name(const char *name, int *block_id)
 
 	for (i = 0; i < ARRAY_SIZE(ras_block_string); i++) {
 		*block_id = i;
-		if (strcmp(name, ras_block_str(i)) == 0)
+		if (strcmp(name, ras_block_string[i]) == 0)
 			return 0;
 	}
 	return -EINVAL;
@@ -509,7 +531,6 @@ static ssize_t amdgpu_ras_sysfs_read(struct device *dev,
 	if (amdgpu_ras_query_error_status(obj->adev, &info))
 		return -EINVAL;
 
-
 	if (obj->adev->asic_type == CHIP_ALDEBARAN) {
 		if (amdgpu_ras_reset_error_status(obj->adev, info.head.block))
 			DRM_WARN("Failed to reset error counter and error status");
@@ -529,7 +550,7 @@ static inline void put_obj(struct ras_manager *obj)
 	if (obj && (--obj->use == 0))
 		list_del(&obj->node);
 	if (obj && (obj->use < 0))
-		DRM_ERROR("RAS ERROR: Unbalance obj(%s) use\n", ras_block_str(obj->head.block));
+		DRM_ERROR("RAS ERROR: Unbalance obj(%s) use\n", get_ras_block_str(&obj->head));
 }
 
 /* make one obj and return it. */
@@ -545,7 +566,14 @@ static struct ras_manager *amdgpu_ras_create_obj(struct amdgpu_device *adev,
 	if (head->block >= AMDGPU_RAS_BLOCK_COUNT)
 		return NULL;
 
-	obj = &con->objs[head->block];
+	if (head->block == AMDGPU_RAS_BLOCK__MCA) {
+		if (head->sub_block_index >= AMDGPU_RAS_MCA_BLOCK__LAST)
+			return NULL;
+
+		obj = &con->objs[AMDGPU_RAS_BLOCK__LAST + head->sub_block_index];
+	} else
+		obj = &con->objs[head->block];
+
 	/* already exist. return obj? */
 	if (alive_obj(obj))
 		return NULL;
@@ -573,19 +601,21 @@ struct ras_manager *amdgpu_ras_find_obj(struct amdgpu_device *adev,
 		if (head->block >= AMDGPU_RAS_BLOCK_COUNT)
 			return NULL;
 
-		obj = &con->objs[head->block];
+		if (head->block == AMDGPU_RAS_BLOCK__MCA) {
+			if (head->sub_block_index >= AMDGPU_RAS_MCA_BLOCK__LAST)
+				return NULL;
+
+			obj = &con->objs[AMDGPU_RAS_BLOCK__LAST + head->sub_block_index];
+		} else
+			obj = &con->objs[head->block];
 
-		if (alive_obj(obj)) {
-			WARN_ON(head->block != obj->head.block);
+		if (alive_obj(obj))
 			return obj;
-		}
 	} else {
-		for (i = 0; i < AMDGPU_RAS_BLOCK_COUNT; i++) {
+		for (i = 0; i < AMDGPU_RAS_BLOCK_COUNT + AMDGPU_RAS_MCA_BLOCK_COUNT; i++) {
 			obj = &con->objs[i];
-			if (alive_obj(obj)) {
-				WARN_ON(i != obj->head.block);
+			if (alive_obj(obj))
 				return obj;
-			}
 		}
 	}
 
@@ -626,8 +656,6 @@ static int __amdgpu_ras_feature_enable(struct amdgpu_device *adev,
 	 */
 	if (!amdgpu_ras_is_feature_allowed(adev, head))
 		return 0;
-	if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head)))
-		return 0;
 
 	if (enable) {
 		if (!obj) {
@@ -678,18 +706,13 @@ int amdgpu_ras_feature_enable(struct amdgpu_device *adev,
 
 	/* Do not enable if it is not allowed. */
 	WARN_ON(enable && !amdgpu_ras_is_feature_allowed(adev, head));
-	/* Are we alerady in that state we are going to set? */
-	if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head))) {
-		ret = 0;
-		goto out;
-	}
 
 	if (!amdgpu_ras_intr_triggered()) {
 		ret = psp_ras_enable_features(&adev->psp, info, enable);
 		if (ret) {
 			dev_err(adev->dev, "ras %s %s failed %d\n",
 				enable ? "enable":"disable",
-				ras_block_str(head->block),
+				get_ras_block_str(head),
 				ret);
 			goto out;
 		}
@@ -731,7 +754,7 @@ int amdgpu_ras_feature_enable_on_boot(struct amdgpu_device *adev,
 				if (!ret)
 					dev_info(adev->dev,
 						"RAS INFO: %s setup object\n",
-						ras_block_str(head->block));
+						get_ras_block_str(head));
 			}
 		} else {
 			/* setup the object then issue a ras TA disable cmd.*/
@@ -781,17 +804,39 @@ static int amdgpu_ras_enable_all_features(struct amdgpu_device *adev,
 		bool bypass)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
-	int ras_block_count = AMDGPU_RAS_BLOCK_COUNT;
 	int i;
-	const enum amdgpu_ras_error_type default_ras_type =
-		AMDGPU_RAS_ERROR__NONE;
+	const enum amdgpu_ras_error_type default_ras_type = AMDGPU_RAS_ERROR__NONE;
 
-	for (i = 0; i < ras_block_count; i++) {
+	for (i = 0; i < AMDGPU_RAS_BLOCK_COUNT; i++) {
 		struct ras_common_if head = {
 			.block = i,
 			.type = default_ras_type,
 			.sub_block_index = 0,
 		};
+
+		if (i == AMDGPU_RAS_BLOCK__MCA)
+			continue;
+
+		if (bypass) {
+			/*
+			 * bypass psp. vbios enable ras for us.
+			 * so just create the obj
+			 */
+			if (__amdgpu_ras_feature_enable(adev, &head, 1))
+				break;
+		} else {
+			if (amdgpu_ras_feature_enable(adev, &head, 1))
+				break;
+		}
+	}
+
+	for (i = 0; i < AMDGPU_RAS_MCA_BLOCK_COUNT; i++) {
+		struct ras_common_if head = {
+			.block = AMDGPU_RAS_BLOCK__MCA,
+			.type = default_ras_type,
+			.sub_block_index = i,
+		};
+
 		if (bypass) {
 			/*
 			 * bypass psp. vbios enable ras for us.
@@ -809,6 +854,32 @@ static int amdgpu_ras_enable_all_features(struct amdgpu_device *adev,
 }
 /* feature ctl end */
 
+
+void amdgpu_ras_mca_query_error_status(struct amdgpu_device *adev,
+				       struct ras_common_if *ras_block,
+				       struct ras_err_data  *err_data)
+{
+	switch (ras_block->sub_block_index) {
+	case AMDGPU_RAS_MCA_BLOCK__MP0:
+		if (adev->mca.mp0.ras_funcs &&
+		    adev->mca.mp0.ras_funcs->query_ras_error_count)
+			adev->mca.mp0.ras_funcs->query_ras_error_count(adev, &err_data);
+		break;
+	case AMDGPU_RAS_MCA_BLOCK__MP1:
+		if (adev->mca.mp1.ras_funcs &&
+		    adev->mca.mp1.ras_funcs->query_ras_error_count)
+			adev->mca.mp1.ras_funcs->query_ras_error_count(adev, &err_data);
+		break;
+	case AMDGPU_RAS_MCA_BLOCK__MPIO:
+		if (adev->mca.mpio.ras_funcs &&
+		    adev->mca.mpio.ras_funcs->query_ras_error_count)
+			adev->mca.mpio.ras_funcs->query_ras_error_count(adev, &err_data);
+		break;
+	default:
+		break;
+	}
+}
+
 /* query/inject/cure begin */
 int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
 				  struct ras_query_if *info)
@@ -872,6 +943,9 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
 		    adev->hdp.ras_funcs->query_ras_error_count)
 			adev->hdp.ras_funcs->query_ras_error_count(adev, &err_data);
 		break;
+	case AMDGPU_RAS_BLOCK__MCA:
+		amdgpu_ras_mca_query_error_status(adev, &info->head, &err_data);
+		break;
 	default:
 		break;
 	}
@@ -893,13 +967,13 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
 					adev->smuio.funcs->get_socket_id(adev),
 					adev->smuio.funcs->get_die_id(adev),
 					obj->err_data.ce_count,
-					ras_block_str(info->head.block));
+					get_ras_block_str(&info->head));
 		} else {
 			dev_info(adev->dev, "%ld correctable hardware errors "
 					"detected in %s block, no user "
 					"action is needed.\n",
 					obj->err_data.ce_count,
-					ras_block_str(info->head.block));
+					get_ras_block_str(&info->head));
 		}
 	}
 	if (err_data.ue_count) {
@@ -912,12 +986,12 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
 					adev->smuio.funcs->get_socket_id(adev),
 					adev->smuio.funcs->get_die_id(adev),
 					obj->err_data.ue_count,
-					ras_block_str(info->head.block));
+					get_ras_block_str(&info->head));
 		} else {
 			dev_info(adev->dev, "%ld uncorrectable hardware errors "
 					"detected in %s block\n",
 					obj->err_data.ue_count,
-					ras_block_str(info->head.block));
+					get_ras_block_str(&info->head));
 		}
 	}
 
@@ -1027,6 +1101,7 @@ int amdgpu_ras_error_inject(struct amdgpu_device *adev,
 	case AMDGPU_RAS_BLOCK__SDMA:
 	case AMDGPU_RAS_BLOCK__MMHUB:
 	case AMDGPU_RAS_BLOCK__PCIE_BIF:
+	case AMDGPU_RAS_BLOCK__MCA:
 		ret = psp_ras_trigger_error(&adev->psp, &block_info);
 		break;
 	case AMDGPU_RAS_BLOCK__XGMI_WAFL:
@@ -1034,13 +1109,13 @@ int amdgpu_ras_error_inject(struct amdgpu_device *adev,
 		break;
 	default:
 		dev_info(adev->dev, "%s error injection is not supported yet\n",
-			 ras_block_str(info->head.block));
+			 get_ras_block_str(&info->head));
 		ret = -EINVAL;
 	}
 
 	if (ret)
 		dev_err(adev->dev, "ras inject %s failed %d\n",
-			ras_block_str(info->head.block), ret);
+			get_ras_block_str(&info->head), ret);
 
 	return ret;
 }
@@ -1383,7 +1458,7 @@ void amdgpu_ras_debugfs_create_all(struct amdgpu_device *adev)
 		if (amdgpu_ras_is_supported(adev, obj->head.block) &&
 			(obj->attr_inuse == 1)) {
 			sprintf(fs_info.debugfs_name, "%s_err_inject",
-					ras_block_str(obj->head.block));
+					get_ras_block_str(&obj->head));
 			fs_info.head = obj->head;
 			amdgpu_ras_debugfs_create(adev, &fs_info, dir);
 		}
@@ -2056,19 +2131,6 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev)
 }
 /* recovery end */
 
-/* return 0 if ras will reset gpu and repost.*/
-int amdgpu_ras_request_reset_on_boot(struct amdgpu_device *adev,
-		unsigned int block)
-{
-	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
-
-	if (!ras)
-		return -EINVAL;
-
-	ras->flags |= AMDGPU_RAS_FLAG_INIT_NEED_RESET;
-	return 0;
-}
-
 static bool amdgpu_ras_asic_supported(struct amdgpu_device *adev)
 {
 	return adev->asic_type == CHIP_VEGA10 ||
@@ -2181,7 +2243,8 @@ int amdgpu_ras_init(struct amdgpu_device *adev)
 		return 0;
 
 	con = kmalloc(sizeof(struct amdgpu_ras) +
-			sizeof(struct ras_manager) * AMDGPU_RAS_BLOCK_COUNT,
+			sizeof(struct ras_manager) * AMDGPU_RAS_BLOCK_COUNT +
+			sizeof(struct ras_manager) * AMDGPU_RAS_MCA_BLOCK_COUNT,
 			GFP_KERNEL|__GFP_ZERO);
 	if (!con)
 		return -ENOMEM;
@@ -2306,12 +2369,7 @@ int amdgpu_ras_late_init(struct amdgpu_device *adev,
 
 	r = amdgpu_ras_feature_enable_on_boot(adev, ras_block, 1);
 	if (r) {
-		if (r == -EAGAIN) {
-			/* request gpu reset. will run again */
-			amdgpu_ras_request_reset_on_boot(adev,
-					ras_block->block);
-			return 0;
-		} else if (adev->in_suspend || amdgpu_in_reset(adev)) {
+		if (adev->in_suspend || amdgpu_in_reset(adev)) {
 			/* in resume phase, if fail to enable ras,
 			 * clean up all ras fs nodes, and disable ras */
 			goto cleanup;
@@ -2403,19 +2461,6 @@ void amdgpu_ras_resume(struct amdgpu_device *adev)
 			}
 		}
 	}
-
-	if (con->flags & AMDGPU_RAS_FLAG_INIT_NEED_RESET) {
-		con->flags &= ~AMDGPU_RAS_FLAG_INIT_NEED_RESET;
-		/* setup ras obj state as disabled.
-		 * for init_by_vbios case.
-		 * if we want to enable ras, just enable it in a normal way.
-		 * If we want do disable it, need setup ras obj as enabled,
-		 * then issue another TA disable cmd.
-		 * See feature_enable_on_boot
-		 */
-		amdgpu_ras_disable_all_features(adev, 1);
-		amdgpu_ras_reset_gpu(adev);
-	}
 }
 
 void amdgpu_ras_suspend(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index eae604fd90b8..37b3c40272b4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -32,7 +32,6 @@
 #include "amdgpu_ras_eeprom.h"
 
 #define AMDGPU_RAS_FLAG_INIT_BY_VBIOS		(0x1 << 0)
-#define AMDGPU_RAS_FLAG_INIT_NEED_RESET		(0x1 << 1)
 
 enum amdgpu_ras_block {
 	AMDGPU_RAS_BLOCK__UMC = 0,
@@ -49,15 +48,22 @@ enum amdgpu_ras_block {
 	AMDGPU_RAS_BLOCK__MP0,
 	AMDGPU_RAS_BLOCK__MP1,
 	AMDGPU_RAS_BLOCK__FUSE,
-	AMDGPU_RAS_BLOCK__MPIO,
+	AMDGPU_RAS_BLOCK__MCA,
 
 	AMDGPU_RAS_BLOCK__LAST
 };
 
-extern const char *ras_block_string[];
+enum amdgpu_ras_mca_block {
+	AMDGPU_RAS_MCA_BLOCK__MP0 = 0,
+	AMDGPU_RAS_MCA_BLOCK__MP1,
+	AMDGPU_RAS_MCA_BLOCK__MPIO,
+	AMDGPU_RAS_MCA_BLOCK__IOHC,
+
+	AMDGPU_RAS_MCA_BLOCK__LAST
+};
 
-#define ras_block_str(i) (ras_block_string[i])
 #define AMDGPU_RAS_BLOCK_COUNT	AMDGPU_RAS_BLOCK__LAST
+#define AMDGPU_RAS_MCA_BLOCK_COUNT	AMDGPU_RAS_MCA_BLOCK__LAST
 #define AMDGPU_RAS_BLOCK_MASK	((1ULL << AMDGPU_RAS_BLOCK_COUNT) - 1)
 
 enum amdgpu_ras_gfx_subblock {
@@ -488,8 +494,6 @@ static inline int amdgpu_ras_is_supported(struct amdgpu_device *adev,
 }
 
 int amdgpu_ras_recovery_init(struct amdgpu_device *adev);
-int amdgpu_ras_request_reset_on_boot(struct amdgpu_device *adev,
-		unsigned int block);
 
 void amdgpu_ras_resume(struct amdgpu_device *adev);
 void amdgpu_ras_suspend(struct amdgpu_device *adev);
@@ -544,6 +548,8 @@ amdgpu_ras_block_to_ta(enum amdgpu_ras_block block) {
 		return TA_RAS_BLOCK__MP1;
 	case AMDGPU_RAS_BLOCK__FUSE:
 		return TA_RAS_BLOCK__FUSE;
+	case AMDGPU_RAS_BLOCK__MCA:
+		return TA_RAS_BLOCK__MCA;
 	default:
 		WARN_ONCE(1, "RAS ERROR: unexpected block id %d\n", block);
 		return TA_RAS_BLOCK__UMC;
@@ -638,4 +644,6 @@ void amdgpu_release_ras_context(struct amdgpu_device *adev);
 
 int amdgpu_persistent_edc_harvesting_supported(struct amdgpu_device *adev);
 
+const char *get_ras_block_str(struct ras_common_if *ras_block);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index 0554576d3695..ab2351ba9574 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -415,26 +415,20 @@ static const struct file_operations amdgpu_debugfs_ring_fops = {
 
 #endif
 
-int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
-			     struct amdgpu_ring *ring)
+void amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
+			      struct amdgpu_ring *ring)
 {
 #if defined(CONFIG_DEBUG_FS)
 	struct drm_minor *minor = adev_to_drm(adev)->primary;
-	struct dentry *ent, *root = minor->debugfs_root;
+	struct dentry *root = minor->debugfs_root;
 	char name[32];
 
 	sprintf(name, "amdgpu_ring_%s", ring->name);
+	debugfs_create_file_size(name, S_IFREG | S_IRUGO, root, ring,
+				 &amdgpu_debugfs_ring_fops,
+				 ring->ring_size + 12);
 
-	ent = debugfs_create_file(name,
-				  S_IFREG | S_IRUGO, root,
-				  ring, &amdgpu_debugfs_ring_fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
-
-	i_size_write(ent->d_inode, ring->ring_size + 12);
-	ring->ent = ent;
 #endif
-	return 0;
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index e713d31619fe..4d380e79752c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -36,8 +36,13 @@
 #define AMDGPU_MAX_VCE_RINGS		3
 #define AMDGPU_MAX_UVD_ENC_RINGS	2
 
-#define AMDGPU_RING_PRIO_DEFAULT	1
-#define AMDGPU_RING_PRIO_MAX		AMDGPU_GFX_PIPE_PRIO_MAX
+enum amdgpu_ring_priority_level {
+	AMDGPU_RING_PRIO_0,
+	AMDGPU_RING_PRIO_1,
+	AMDGPU_RING_PRIO_DEFAULT = 1,
+	AMDGPU_RING_PRIO_2,
+	AMDGPU_RING_PRIO_MAX
+};
 
 /* some special values for the owner field */
 #define AMDGPU_FENCE_OWNER_UNDEFINED	((void *)0ul)
@@ -248,10 +253,6 @@ struct amdgpu_ring {
 	bool			has_compute_vm_bug;
 	bool			no_scheduler;
 	int			hw_prio;
-
-#if defined(CONFIG_DEBUG_FS)
-	struct dentry *ent;
-#endif
 };
 
 #define amdgpu_ring_parse_cs(r, p, ib) ((r)->funcs->parse_cs((p), (ib)))
@@ -351,8 +352,6 @@ static inline void amdgpu_ring_write_multiple(struct amdgpu_ring *ring,
 
 int amdgpu_ring_test_helper(struct amdgpu_ring *ring);
 
-int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
-			     struct amdgpu_ring *ring);
-void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring);
-
+void amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
+			      struct amdgpu_ring *ring);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
index b7d861ed5284..e9b45089a28a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
@@ -32,37 +32,9 @@
 #include "amdgpu_sched.h"
 #include "amdgpu_vm.h"
 
-int amdgpu_to_sched_priority(int amdgpu_priority,
-			     enum drm_sched_priority *prio)
-{
-	switch (amdgpu_priority) {
-	case AMDGPU_CTX_PRIORITY_VERY_HIGH:
-		*prio = DRM_SCHED_PRIORITY_HIGH;
-		break;
-	case AMDGPU_CTX_PRIORITY_HIGH:
-		*prio = DRM_SCHED_PRIORITY_HIGH;
-		break;
-	case AMDGPU_CTX_PRIORITY_NORMAL:
-		*prio = DRM_SCHED_PRIORITY_NORMAL;
-		break;
-	case AMDGPU_CTX_PRIORITY_LOW:
-	case AMDGPU_CTX_PRIORITY_VERY_LOW:
-		*prio = DRM_SCHED_PRIORITY_MIN;
-		break;
-	case AMDGPU_CTX_PRIORITY_UNSET:
-		*prio = DRM_SCHED_PRIORITY_UNSET;
-		break;
-	default:
-		WARN(1, "Invalid context priority %d\n", amdgpu_priority);
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
 static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
 						  int fd,
-						  enum drm_sched_priority priority)
+						  int32_t priority)
 {
 	struct fd f = fdget(fd);
 	struct amdgpu_fpriv *fpriv;
@@ -89,7 +61,7 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
 static int amdgpu_sched_context_priority_override(struct amdgpu_device *adev,
 						  int fd,
 						  unsigned ctx_id,
-						  enum drm_sched_priority priority)
+						  int32_t priority)
 {
 	struct fd f = fdget(fd);
 	struct amdgpu_fpriv *fpriv;
@@ -124,7 +96,6 @@ int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
 {
 	union drm_amdgpu_sched *args = data;
 	struct amdgpu_device *adev = drm_to_adev(dev);
-	enum drm_sched_priority priority;
 	int r;
 
 	/* First check the op, then the op's argument.
@@ -138,21 +109,22 @@ int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	r = amdgpu_to_sched_priority(args->in.priority, &priority);
-	if (r)
-		return r;
+	if (!amdgpu_ctx_priority_is_valid(args->in.priority)) {
+		WARN(1, "Invalid context priority %d\n", args->in.priority);
+		return -EINVAL;
+	}
 
 	switch (args->in.op) {
 	case AMDGPU_SCHED_OP_PROCESS_PRIORITY_OVERRIDE:
 		r = amdgpu_sched_process_priority_override(adev,
 							   args->in.fd,
-							   priority);
+							   args->in.priority);
 		break;
 	case AMDGPU_SCHED_OP_CONTEXT_PRIORITY_OVERRIDE:
 		r = amdgpu_sched_context_priority_override(adev,
 							   args->in.fd,
 							   args->in.ctx_id,
-							   priority);
+							   args->in.priority);
 		break;
 	default:
 		/* Impossible.
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index abd8469380e5..527d67ded8a0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -525,9 +525,9 @@ FW_VERSION_ATTR(rlc_srls_fw_version, 0444, gfx.rlc_srls_fw_version);
 FW_VERSION_ATTR(mec_fw_version, 0444, gfx.mec_fw_version);
 FW_VERSION_ATTR(mec2_fw_version, 0444, gfx.mec2_fw_version);
 FW_VERSION_ATTR(sos_fw_version, 0444, psp.sos.fw_version);
-FW_VERSION_ATTR(asd_fw_version, 0444, psp.asd.fw_version);
-FW_VERSION_ATTR(ta_ras_fw_version, 0444, psp.ras.feature_version);
-FW_VERSION_ATTR(ta_xgmi_fw_version, 0444, psp.xgmi.feature_version);
+FW_VERSION_ATTR(asd_fw_version, 0444, psp.asd_context.bin_desc.fw_version);
+FW_VERSION_ATTR(ta_ras_fw_version, 0444, psp.ras_context.context.bin_desc.feature_version);
+FW_VERSION_ATTR(ta_xgmi_fw_version, 0444, psp.xgmi_context.context.bin_desc.feature_version);
 FW_VERSION_ATTR(smc_fw_version, 0444, pm.fw_version);
 FW_VERSION_ATTR(sdma_fw_version, 0444, sdma.instance[0].fw_version);
 FW_VERSION_ATTR(sdma2_fw_version, 0444, sdma.instance[1].fw_version);
@@ -572,6 +572,7 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev,
 	const struct dmcu_firmware_header_v1_0 *dmcu_hdr = NULL;
 	const struct dmcub_firmware_header_v1_0 *dmcub_hdr = NULL;
 	const struct mes_firmware_header_v1_0 *mes_hdr = NULL;
+	u8 *ucode_addr;
 
 	if (NULL == ucode->fw)
 		return 0;
@@ -588,94 +589,83 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev,
 	dmcub_hdr = (const struct dmcub_firmware_header_v1_0 *)ucode->fw->data;
 	mes_hdr = (const struct mes_firmware_header_v1_0 *)ucode->fw->data;
 
-	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP ||
-	    (ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1 &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2 &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1_JT &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MES &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MES_DATA &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_IRAM &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_DRAM &&
-		 ucode->ucode_id != AMDGPU_UCODE_ID_DMCU_ERAM &&
-		 ucode->ucode_id != AMDGPU_UCODE_ID_DMCU_INTV &&
-		 ucode->ucode_id != AMDGPU_UCODE_ID_DMCUB)) {
-		ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes);
-
-		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
-					      le32_to_cpu(header->ucode_array_offset_bytes)),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_CP_MEC1 ||
-		   ucode->ucode_id == AMDGPU_UCODE_ID_CP_MEC2) {
-		ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes) -
-			le32_to_cpu(cp_hdr->jt_size) * 4;
-
-		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
-					      le32_to_cpu(header->ucode_array_offset_bytes)),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_CP_MEC1_JT ||
-		   ucode->ucode_id == AMDGPU_UCODE_ID_CP_MEC2_JT) {
-		ucode->ucode_size = le32_to_cpu(cp_hdr->jt_size) * 4;
-
-		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
-					      le32_to_cpu(header->ucode_array_offset_bytes) +
-					      le32_to_cpu(cp_hdr->jt_offset) * 4),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_DMCU_ERAM) {
-		ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes) -
+	if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
+		switch (ucode->ucode_id) {
+		case AMDGPU_UCODE_ID_CP_MEC1:
+		case AMDGPU_UCODE_ID_CP_MEC2:
+			ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes) -
+				le32_to_cpu(cp_hdr->jt_size) * 4;
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(header->ucode_array_offset_bytes);
+			break;
+		case AMDGPU_UCODE_ID_CP_MEC1_JT:
+		case AMDGPU_UCODE_ID_CP_MEC2_JT:
+			ucode->ucode_size = le32_to_cpu(cp_hdr->jt_size) * 4;
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(header->ucode_array_offset_bytes) +
+				le32_to_cpu(cp_hdr->jt_offset) * 4;
+			break;
+		case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL:
+			ucode->ucode_size = adev->gfx.rlc.save_restore_list_cntl_size_bytes;
+			ucode_addr = adev->gfx.rlc.save_restore_list_cntl;
+			break;
+		case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM:
+			ucode->ucode_size = adev->gfx.rlc.save_restore_list_gpm_size_bytes;
+			ucode_addr = adev->gfx.rlc.save_restore_list_gpm;
+			break;
+		case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM:
+			ucode->ucode_size = adev->gfx.rlc.save_restore_list_srm_size_bytes;
+			ucode_addr = adev->gfx.rlc.save_restore_list_srm;
+			break;
+		case AMDGPU_UCODE_ID_RLC_IRAM:
+			ucode->ucode_size = adev->gfx.rlc.rlc_iram_ucode_size_bytes;
+			ucode_addr = adev->gfx.rlc.rlc_iram_ucode;
+			break;
+		case AMDGPU_UCODE_ID_RLC_DRAM:
+			ucode->ucode_size = adev->gfx.rlc.rlc_dram_ucode_size_bytes;
+			ucode_addr = adev->gfx.rlc.rlc_dram_ucode;
+			break;
+		case AMDGPU_UCODE_ID_CP_MES:
+			ucode->ucode_size = le32_to_cpu(mes_hdr->mes_ucode_size_bytes);
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(mes_hdr->mes_ucode_offset_bytes);
+			break;
+		case AMDGPU_UCODE_ID_CP_MES_DATA:
+			ucode->ucode_size = le32_to_cpu(mes_hdr->mes_ucode_data_size_bytes);
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(mes_hdr->mes_ucode_data_offset_bytes);
+			break;
+		case AMDGPU_UCODE_ID_DMCU_ERAM:
+			ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes) -
 				le32_to_cpu(dmcu_hdr->intv_size_bytes);
-
-		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
-					      le32_to_cpu(header->ucode_array_offset_bytes)),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_DMCU_INTV) {
-		ucode->ucode_size = le32_to_cpu(dmcu_hdr->intv_size_bytes);
-
-		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
-					      le32_to_cpu(header->ucode_array_offset_bytes) +
-					      le32_to_cpu(dmcu_hdr->intv_offset_bytes)),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_DMCUB) {
-		ucode->ucode_size = le32_to_cpu(dmcub_hdr->inst_const_bytes);
-		memcpy(ucode->kaddr,
-		       (void *)((uint8_t *)ucode->fw->data +
-				le32_to_cpu(header->ucode_array_offset_bytes)),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL) {
-		ucode->ucode_size = adev->gfx.rlc.save_restore_list_cntl_size_bytes;
-		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_cntl,
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM) {
-		ucode->ucode_size = adev->gfx.rlc.save_restore_list_gpm_size_bytes;
-		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_gpm,
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM) {
-		ucode->ucode_size = adev->gfx.rlc.save_restore_list_srm_size_bytes;
-		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_srm,
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_IRAM) {
-		ucode->ucode_size = adev->gfx.rlc.rlc_iram_ucode_size_bytes;
-		memcpy(ucode->kaddr, adev->gfx.rlc.rlc_iram_ucode,
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_DRAM) {
-		ucode->ucode_size = adev->gfx.rlc.rlc_dram_ucode_size_bytes;
-		memcpy(ucode->kaddr, adev->gfx.rlc.rlc_dram_ucode,
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_CP_MES) {
-		ucode->ucode_size = le32_to_cpu(mes_hdr->mes_ucode_size_bytes);
-		memcpy(ucode->kaddr, (void *)((uint8_t *)adev->mes.fw->data +
-			      le32_to_cpu(mes_hdr->mes_ucode_offset_bytes)),
-		       ucode->ucode_size);
-	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_CP_MES_DATA) {
-		ucode->ucode_size = le32_to_cpu(mes_hdr->mes_ucode_data_size_bytes);
-		memcpy(ucode->kaddr, (void *)((uint8_t *)adev->mes.fw->data +
-			      le32_to_cpu(mes_hdr->mes_ucode_data_offset_bytes)),
-		       ucode->ucode_size);
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(header->ucode_array_offset_bytes);
+			break;
+		case AMDGPU_UCODE_ID_DMCU_INTV:
+			ucode->ucode_size = le32_to_cpu(dmcu_hdr->intv_size_bytes);
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(header->ucode_array_offset_bytes) +
+				le32_to_cpu(dmcu_hdr->intv_offset_bytes);
+			break;
+		case AMDGPU_UCODE_ID_DMCUB:
+			ucode->ucode_size = le32_to_cpu(dmcub_hdr->inst_const_bytes);
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(header->ucode_array_offset_bytes);
+			break;
+		default:
+			ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes);
+			ucode_addr = (u8 *)ucode->fw->data +
+				le32_to_cpu(header->ucode_array_offset_bytes);
+			break;
+		}
+	} else {
+		ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes);
+		ucode_addr = (u8 *)ucode->fw->data +
+			le32_to_cpu(header->ucode_array_offset_bytes);
 	}
 
+	memcpy(ucode->kaddr, ucode_addr, ucode->ucode_size);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umr.h
new file mode 100644
index 000000000000..919d9d401750
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umr.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include <linux/ioctl.h>
+
+/*
+ * MMIO debugfs IOCTL structure
+ */
+struct amdgpu_debugfs_regs2_iocdata {
+	__u32 use_srbm, use_grbm, pg_lock;
+	struct {
+		__u32 se, sh, instance;
+	} grbm;
+	struct {
+		__u32 me, pipe, queue, vmid;
+	} srbm;
+};
+
+/*
+ * MMIO debugfs state data (per file* handle)
+ */
+struct amdgpu_debugfs_regs2_data {
+	struct amdgpu_device *adev;
+	struct mutex lock;
+	struct amdgpu_debugfs_regs2_iocdata id;
+};
+
+enum AMDGPU_DEBUGFS_REGS2_CMDS {
+	AMDGPU_DEBUGFS_REGS2_CMD_SET_STATE=0,
+};
+
+#define AMDGPU_DEBUGFS_REGS2_IOC_SET_STATE _IOWR(0x20, AMDGPU_DEBUGFS_REGS2_CMD_SET_STATE, struct amdgpu_debugfs_regs2_iocdata)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index d451c359606a..8a26459bd80b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -134,6 +134,51 @@ MODULE_FIRMWARE(FIRMWARE_VEGA12);
 MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
+static void amdgpu_uvd_force_into_uvd_segment(struct amdgpu_bo *abo);
+
+static int amdgpu_uvd_create_msg_bo_helper(struct amdgpu_device *adev,
+					   uint32_t size,
+					   struct amdgpu_bo **bo_ptr)
+{
+	struct ttm_operation_ctx ctx = { true, false };
+	struct amdgpu_bo *bo = NULL;
+	void *addr;
+	int r;
+
+	r = amdgpu_bo_create_reserved(adev, size, PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_GTT,
+				      &bo, NULL, &addr);
+	if (r)
+		return r;
+
+	if (adev->uvd.address_64_bit)
+		goto succ;
+
+	amdgpu_bo_kunmap(bo);
+	amdgpu_bo_unpin(bo);
+	amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
+	amdgpu_uvd_force_into_uvd_segment(bo);
+	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+	if (r)
+		goto err;
+	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_VRAM);
+	if (r)
+		goto err_pin;
+	r = amdgpu_bo_kmap(bo, &addr);
+	if (r)
+		goto err_kmap;
+succ:
+	amdgpu_bo_unreserve(bo);
+	*bo_ptr = bo;
+	return 0;
+err_kmap:
+	amdgpu_bo_unpin(bo);
+err_pin:
+err:
+	amdgpu_bo_unreserve(bo);
+	amdgpu_bo_unref(&bo);
+	return r;
+}
 
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 {
@@ -302,6 +347,10 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	if (!amdgpu_device_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0))
 		adev->uvd.address_64_bit = true;
 
+	r = amdgpu_uvd_create_msg_bo_helper(adev, 128 << 10, &adev->uvd.ib_bo);
+	if (r)
+		return r;
+
 	switch (adev->asic_type) {
 	case CHIP_TONGA:
 		adev->uvd.use_ctx_buf = adev->uvd.fw_version >= FW_1_65_10;
@@ -324,6 +373,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 
 int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 {
+	void *addr = amdgpu_bo_kptr(adev->uvd.ib_bo);
 	int i, j;
 
 	drm_sched_entity_destroy(&adev->uvd.entity);
@@ -342,6 +392,7 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 		for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
 			amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
 	}
+	amdgpu_bo_free_kernel(&adev->uvd.ib_bo, NULL, &addr);
 	release_firmware(adev->uvd.fw);
 
 	return 0;
@@ -1080,23 +1131,10 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 	unsigned offset_idx = 0;
 	unsigned offset[3] = { UVD_BASE_SI, 0, 0 };
 
-	amdgpu_bo_kunmap(bo);
-	amdgpu_bo_unpin(bo);
-
-	if (!ring->adev->uvd.address_64_bit) {
-		struct ttm_operation_ctx ctx = { true, false };
-
-		amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
-		amdgpu_uvd_force_into_uvd_segment(bo);
-		r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-		if (r)
-			goto err;
-	}
-
 	r = amdgpu_job_alloc_with_ib(adev, 64, direct ? AMDGPU_IB_POOL_DIRECT :
 				     AMDGPU_IB_POOL_DELAYED, &job);
 	if (r)
-		goto err;
+		return r;
 
 	if (adev->asic_type >= CHIP_VEGA10) {
 		offset_idx = 1 + ring->me;
@@ -1147,9 +1185,9 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 			goto err_free;
 	}
 
+	amdgpu_bo_reserve(bo, true);
 	amdgpu_bo_fence(bo, f, false);
 	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_unref(&bo);
 
 	if (fence)
 		*fence = dma_fence_get(f);
@@ -1159,10 +1197,6 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 
 err_free:
 	amdgpu_job_free(job);
-
-err:
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_unref(&bo);
 	return r;
 }
 
@@ -1173,16 +1207,11 @@ int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 			      struct dma_fence **fence)
 {
 	struct amdgpu_device *adev = ring->adev;
-	struct amdgpu_bo *bo = NULL;
+	struct amdgpu_bo *bo = adev->uvd.ib_bo;
 	uint32_t *msg;
-	int r, i;
-
-	r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_GTT,
-				      &bo, NULL, (void **)&msg);
-	if (r)
-		return r;
+	int i;
 
+	msg = amdgpu_bo_kptr(bo);
 	/* stitch together an UVD create msg */
 	msg[0] = cpu_to_le32(0x00000de4);
 	msg[1] = cpu_to_le32(0x00000000);
@@ -1199,6 +1228,7 @@ int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 		msg[i] = cpu_to_le32(0x0);
 
 	return amdgpu_uvd_send_msg(ring, bo, true, fence);
+
 }
 
 int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
@@ -1209,12 +1239,15 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 	uint32_t *msg;
 	int r, i;
 
-	r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_GTT,
-				      &bo, NULL, (void **)&msg);
-	if (r)
-		return r;
+	if (direct) {
+		bo = adev->uvd.ib_bo;
+	} else {
+		r = amdgpu_uvd_create_msg_bo_helper(adev, 4096, &bo);
+		if (r)
+			return r;
+	}
 
+	msg = amdgpu_bo_kptr(bo);
 	/* stitch together an UVD destroy msg */
 	msg[0] = cpu_to_le32(0x00000de4);
 	msg[1] = cpu_to_le32(0x00000002);
@@ -1223,7 +1256,12 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 	for (i = 4; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	return amdgpu_uvd_send_msg(ring, bo, direct, fence);
+	r = amdgpu_uvd_send_msg(ring, bo, direct, fence);
+
+	if (!direct)
+		amdgpu_bo_free_kernel(&bo, NULL, (void **)&msg);
+
+	return r;
 }
 
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
@@ -1298,10 +1336,17 @@ int amdgpu_uvd_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	struct dma_fence *fence;
 	long r;
 
-	r = amdgpu_uvd_get_create_msg(ring, 1, NULL);
+	r = amdgpu_uvd_get_create_msg(ring, 1, &fence);
 	if (r)
 		goto error;
 
+	r = dma_fence_wait_timeout(fence, false, timeout);
+	dma_fence_put(fence);
+	if (r == 0)
+		r = -ETIMEDOUT;
+	if (r < 0)
+		goto error;
+
 	r = amdgpu_uvd_get_destroy_msg(ring, 1, true, &fence);
 	if (r)
 		goto error;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
index edbb8194ee81..76ac9699885d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
@@ -68,6 +68,7 @@ struct amdgpu_uvd {
 	/* store image width to adjust nb memory state */
 	unsigned		decode_image_width;
 	uint32_t                keyselect;
+	struct amdgpu_bo	*ib_bo;
 };
 
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index 8e8dee9fac9f..caa4d3420e00 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -82,7 +82,6 @@ MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_vce_idle_work_handler(struct work_struct *work);
 static int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-				     struct amdgpu_bo *bo,
 				     struct dma_fence **fence);
 static int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 				      bool direct, struct dma_fence **fence);
@@ -441,12 +440,12 @@ void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
  * Open up a stream for HW test
  */
 static int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-				     struct amdgpu_bo *bo,
 				     struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 1024;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
+	struct amdgpu_ib ib_msg;
 	struct dma_fence *f = NULL;
 	uint64_t addr;
 	int i, r;
@@ -456,9 +455,17 @@ static int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 	if (r)
 		return r;
 
-	ib = &job->ibs[0];
+	memset(&ib_msg, 0, sizeof(ib_msg));
+	/* only one gpu page is needed, alloc +1 page to make addr aligned. */
+	r = amdgpu_ib_get(ring->adev, NULL, AMDGPU_GPU_PAGE_SIZE * 2,
+			  AMDGPU_IB_POOL_DIRECT,
+			  &ib_msg);
+	if (r)
+		goto err;
 
-	addr = amdgpu_bo_gpu_offset(bo);
+	ib = &job->ibs[0];
+	/* let addr point to page boundary */
+	addr = AMDGPU_GPU_PAGE_ALIGN(ib_msg.gpu_addr);
 
 	/* stitch together an VCE create msg */
 	ib->length_dw = 0;
@@ -498,6 +505,7 @@ static int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 		ib->ptr[i] = 0x0;
 
 	r = amdgpu_job_submit_direct(job, ring, &f);
+	amdgpu_ib_free(ring->adev, &ib_msg, f);
 	if (r)
 		goto err;
 
@@ -1134,20 +1142,13 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring)
 int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
-	struct amdgpu_bo *bo = NULL;
 	long r;
 
 	/* skip vce ring1/2 ib test for now, since it's not reliable */
 	if (ring != &ring->adev->vce.ring[0])
 		return 0;
 
-	r = amdgpu_bo_create_reserved(ring->adev, 512, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      &bo, NULL, NULL);
-	if (r)
-		return r;
-
-	r = amdgpu_vce_get_create_msg(ring, 1, bo, NULL);
+	r = amdgpu_vce_get_create_msg(ring, 1, NULL);
 	if (r)
 		goto error;
 
@@ -1163,7 +1164,19 @@ int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_free_kernel(&bo, NULL, NULL);
 	return r;
 }
+
+enum amdgpu_ring_priority_level amdgpu_vce_get_ring_prio(int ring)
+{
+	switch(ring) {
+	case 0:
+		return AMDGPU_RING_PRIO_0;
+	case 1:
+		return AMDGPU_RING_PRIO_1;
+	case 2:
+		return AMDGPU_RING_PRIO_2;
+	default:
+		return AMDGPU_RING_PRIO_0;
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
index d6d83a3ec803..be4a6e773c5b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
@@ -71,5 +71,6 @@ void amdgpu_vce_ring_begin_use(struct amdgpu_ring *ring);
 void amdgpu_vce_ring_end_use(struct amdgpu_ring *ring);
 unsigned amdgpu_vce_ring_get_emit_ib_size(struct amdgpu_ring *ring);
 unsigned amdgpu_vce_ring_get_dma_frame_size(struct amdgpu_ring *ring);
+enum amdgpu_ring_priority_level amdgpu_vce_get_ring_prio(int ring);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 008a308a4eca..b60b8fe5bf67 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -541,15 +541,14 @@ int amdgpu_vcn_dec_sw_ring_test_ring(struct amdgpu_ring *ring)
 }
 
 static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
-				   struct amdgpu_bo *bo,
+				   struct amdgpu_ib *ib_msg,
 				   struct dma_fence **fence)
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct dma_fence *f = NULL;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
-	uint64_t addr;
-	void *msg = NULL;
+	uint64_t addr = AMDGPU_GPU_PAGE_ALIGN(ib_msg->gpu_addr);
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(adev, 64,
@@ -558,8 +557,6 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
 		goto err;
 
 	ib = &job->ibs[0];
-	addr = amdgpu_bo_gpu_offset(bo);
-	msg = amdgpu_bo_kptr(bo);
 	ib->ptr[0] = PACKET0(adev->vcn.internal.data0, 0);
 	ib->ptr[1] = addr;
 	ib->ptr[2] = PACKET0(adev->vcn.internal.data1, 0);
@@ -576,9 +573,7 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
 	if (r)
 		goto err_free;
 
-	amdgpu_bo_fence(bo, f, false);
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_free_kernel(&bo, NULL, (void **)&msg);
+	amdgpu_ib_free(adev, ib_msg, f);
 
 	if (fence)
 		*fence = dma_fence_get(f);
@@ -588,27 +583,26 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
 
 err_free:
 	amdgpu_job_free(job);
-
 err:
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_free_kernel(&bo, NULL, (void **)&msg);
+	amdgpu_ib_free(adev, ib_msg, f);
 	return r;
 }
 
 static int amdgpu_vcn_dec_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-					 struct amdgpu_bo **bo)
+		struct amdgpu_ib *ib)
 {
 	struct amdgpu_device *adev = ring->adev;
 	uint32_t *msg;
 	int r, i;
 
-	*bo = NULL;
-	r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      bo, NULL, (void **)&msg);
+	memset(ib, 0, sizeof(*ib));
+	r = amdgpu_ib_get(adev, NULL, AMDGPU_GPU_PAGE_SIZE * 2,
+			AMDGPU_IB_POOL_DIRECT,
+			ib);
 	if (r)
 		return r;
 
+	msg = (uint32_t *)AMDGPU_GPU_PAGE_ALIGN((unsigned long)ib->ptr);
 	msg[0] = cpu_to_le32(0x00000028);
 	msg[1] = cpu_to_le32(0x00000038);
 	msg[2] = cpu_to_le32(0x00000001);
@@ -630,19 +624,20 @@ static int amdgpu_vcn_dec_get_create_msg(struct amdgpu_ring *ring, uint32_t hand
 }
 
 static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-					  struct amdgpu_bo **bo)
+					  struct amdgpu_ib *ib)
 {
 	struct amdgpu_device *adev = ring->adev;
 	uint32_t *msg;
 	int r, i;
 
-	*bo = NULL;
-	r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      bo, NULL, (void **)&msg);
+	memset(ib, 0, sizeof(*ib));
+	r = amdgpu_ib_get(adev, NULL, AMDGPU_GPU_PAGE_SIZE * 2,
+			AMDGPU_IB_POOL_DIRECT,
+			ib);
 	if (r)
 		return r;
 
+	msg = (uint32_t *)AMDGPU_GPU_PAGE_ALIGN((unsigned long)ib->ptr);
 	msg[0] = cpu_to_le32(0x00000028);
 	msg[1] = cpu_to_le32(0x00000018);
 	msg[2] = cpu_to_le32(0x00000000);
@@ -658,21 +653,21 @@ static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han
 int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
-	struct amdgpu_bo *bo;
+	struct amdgpu_ib ib;
 	long r;
 
-	r = amdgpu_vcn_dec_get_create_msg(ring, 1, &bo);
+	r = amdgpu_vcn_dec_get_create_msg(ring, 1, &ib);
 	if (r)
 		goto error;
 
-	r = amdgpu_vcn_dec_send_msg(ring, bo, NULL);
+	r = amdgpu_vcn_dec_send_msg(ring, &ib, NULL);
 	if (r)
 		goto error;
-	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, &bo);
+	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, &ib);
 	if (r)
 		goto error;
 
-	r = amdgpu_vcn_dec_send_msg(ring, bo, &fence);
+	r = amdgpu_vcn_dec_send_msg(ring, &ib, &fence);
 	if (r)
 		goto error;
 
@@ -688,8 +683,8 @@ error:
 }
 
 static int amdgpu_vcn_dec_sw_send_msg(struct amdgpu_ring *ring,
-				   struct amdgpu_bo *bo,
-				   struct dma_fence **fence)
+				      struct amdgpu_ib *ib_msg,
+				      struct dma_fence **fence)
 {
 	struct amdgpu_vcn_decode_buffer *decode_buffer = NULL;
 	const unsigned int ib_size_dw = 64;
@@ -697,7 +692,7 @@ static int amdgpu_vcn_dec_sw_send_msg(struct amdgpu_ring *ring,
 	struct dma_fence *f = NULL;
 	struct amdgpu_job *job;
 	struct amdgpu_ib *ib;
-	uint64_t addr;
+	uint64_t addr = AMDGPU_GPU_PAGE_ALIGN(ib_msg->gpu_addr);
 	int i, r;
 
 	r = amdgpu_job_alloc_with_ib(adev, ib_size_dw * 4,
@@ -706,7 +701,6 @@ static int amdgpu_vcn_dec_sw_send_msg(struct amdgpu_ring *ring,
 		goto err;
 
 	ib = &job->ibs[0];
-	addr = amdgpu_bo_gpu_offset(bo);
 	ib->length_dw = 0;
 
 	ib->ptr[ib->length_dw++] = sizeof(struct amdgpu_vcn_decode_buffer) + 8;
@@ -726,9 +720,7 @@ static int amdgpu_vcn_dec_sw_send_msg(struct amdgpu_ring *ring,
 	if (r)
 		goto err_free;
 
-	amdgpu_bo_fence(bo, f, false);
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_unref(&bo);
+	amdgpu_ib_free(adev, ib_msg, f);
 
 	if (fence)
 		*fence = dma_fence_get(f);
@@ -738,31 +730,29 @@ static int amdgpu_vcn_dec_sw_send_msg(struct amdgpu_ring *ring,
 
 err_free:
 	amdgpu_job_free(job);
-
 err:
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_unref(&bo);
+	amdgpu_ib_free(adev, ib_msg, f);
 	return r;
 }
 
 int amdgpu_vcn_dec_sw_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
-	struct amdgpu_bo *bo;
+	struct amdgpu_ib ib;
 	long r;
 
-	r = amdgpu_vcn_dec_get_create_msg(ring, 1, &bo);
+	r = amdgpu_vcn_dec_get_create_msg(ring, 1, &ib);
 	if (r)
 		goto error;
 
-	r = amdgpu_vcn_dec_sw_send_msg(ring, bo, NULL);
+	r = amdgpu_vcn_dec_sw_send_msg(ring, &ib, NULL);
 	if (r)
 		goto error;
-	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, &bo);
+	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, &ib);
 	if (r)
 		goto error;
 
-	r = amdgpu_vcn_dec_sw_send_msg(ring, bo, &fence);
+	r = amdgpu_vcn_dec_sw_send_msg(ring, &ib, &fence);
 	if (r)
 		goto error;
 
@@ -809,7 +799,7 @@ int amdgpu_vcn_enc_ring_test_ring(struct amdgpu_ring *ring)
 }
 
 static int amdgpu_vcn_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
-					 struct amdgpu_bo *bo,
+					 struct amdgpu_ib *ib_msg,
 					 struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
@@ -825,7 +815,7 @@ static int amdgpu_vcn_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t hand
 		return r;
 
 	ib = &job->ibs[0];
-	addr = amdgpu_bo_gpu_offset(bo);
+	addr = AMDGPU_GPU_PAGE_ALIGN(ib_msg->gpu_addr);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
@@ -863,7 +853,7 @@ err:
 }
 
 static int amdgpu_vcn_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-					  struct amdgpu_bo *bo,
+					  struct amdgpu_ib *ib_msg,
 					  struct dma_fence **fence)
 {
 	const unsigned ib_size_dw = 16;
@@ -879,7 +869,7 @@ static int amdgpu_vcn_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han
 		return r;
 
 	ib = &job->ibs[0];
-	addr = amdgpu_bo_gpu_offset(bo);
+	addr = AMDGPU_GPU_PAGE_ALIGN(ib_msg->gpu_addr);
 
 	ib->length_dw = 0;
 	ib->ptr[ib->length_dw++] = 0x00000018;
@@ -918,21 +908,23 @@ err:
 
 int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
+	struct amdgpu_device *adev = ring->adev;
 	struct dma_fence *fence = NULL;
-	struct amdgpu_bo *bo = NULL;
+	struct amdgpu_ib ib;
 	long r;
 
-	r = amdgpu_bo_create_reserved(ring->adev, 128 * 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      &bo, NULL, NULL);
+	memset(&ib, 0, sizeof(ib));
+	r = amdgpu_ib_get(adev, NULL, (128 << 10) + AMDGPU_GPU_PAGE_SIZE,
+			AMDGPU_IB_POOL_DIRECT,
+			&ib);
 	if (r)
 		return r;
 
-	r = amdgpu_vcn_enc_get_create_msg(ring, 1, bo, NULL);
+	r = amdgpu_vcn_enc_get_create_msg(ring, 1, &ib, NULL);
 	if (r)
 		goto error;
 
-	r = amdgpu_vcn_enc_get_destroy_msg(ring, 1, bo, &fence);
+	r = amdgpu_vcn_enc_get_destroy_msg(ring, 1, &ib, &fence);
 	if (r)
 		goto error;
 
@@ -943,9 +935,22 @@ int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 		r = 0;
 
 error:
+	amdgpu_ib_free(adev, &ib, fence);
 	dma_fence_put(fence);
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_free_kernel(&bo, NULL, NULL);
 
 	return r;
 }
+
+enum amdgpu_ring_priority_level amdgpu_vcn_get_enc_ring_prio(int ring)
+{
+	switch(ring) {
+	case 0:
+		return AMDGPU_RING_PRIO_0;
+	case 1:
+		return AMDGPU_RING_PRIO_1;
+	case 2:
+		return AMDGPU_RING_PRIO_2;
+	default:
+		return AMDGPU_RING_PRIO_0;
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
index d74c62b49795..795cbaa02ff8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
@@ -308,4 +308,6 @@ int amdgpu_vcn_dec_sw_ring_test_ib(struct amdgpu_ring *ring, long timeout);
 int amdgpu_vcn_enc_ring_test_ring(struct amdgpu_ring *ring);
 int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout);
 
+enum amdgpu_ring_priority_level amdgpu_vcn_get_enc_ring_prio(int ring);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
index ca058fbcccd4..88c4177b708a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -532,9 +532,12 @@ static void amdgpu_virt_populate_vf2pf_ucode_info(struct amdgpu_device *adev)
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_MEC,      adev->gfx.mec_fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_MEC2,     adev->gfx.mec2_fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SOS,      adev->psp.sos.fw_version);
-	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_ASD,      adev->psp.asd.fw_version);
-	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_RAS,   adev->psp.ras.feature_version);
-	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_XGMI,  adev->psp.xgmi.feature_version);
+	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_ASD,
+			    adev->psp.asd_context.bin_desc.fw_version);
+	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_RAS,
+			    adev->psp.ras_context.context.bin_desc.feature_version);
+	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_XGMI,
+			    adev->psp.xgmi_context.context.bin_desc.feature_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SMC,      adev->pm.fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SDMA,     adev->sdma.instance[0].fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SDMA2,    adev->sdma.instance[1].fw_version);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 6b15cad78de9..a1ddf74bbdba 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -800,7 +800,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 	struct amdgpu_bo *bo = &vmbo->bo;
 	unsigned entries, ats_entries;
 	uint64_t addr;
-	int r;
+	int r, idx;
 
 	/* Figure out our place in the hierarchy */
 	if (ancestor->parent) {
@@ -845,9 +845,12 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 			return r;
 	}
 
+	if (!drm_dev_enter(&adev->ddev, &idx))
+		return -ENODEV;
+
 	r = vm->update_funcs->map_table(vmbo);
 	if (r)
-		return r;
+		goto exit;
 
 	memset(&params, 0, sizeof(params));
 	params.adev = adev;
@@ -856,7 +859,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 
 	r = vm->update_funcs->prepare(&params, NULL, AMDGPU_SYNC_EXPLICIT);
 	if (r)
-		return r;
+		goto exit;
 
 	addr = 0;
 	if (ats_entries) {
@@ -872,7 +875,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 		r = vm->update_funcs->update(&params, vmbo, addr, 0, ats_entries,
 					     value, flags);
 		if (r)
-			return r;
+			goto exit;
 
 		addr += ats_entries * 8;
 	}
@@ -895,10 +898,13 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 		r = vm->update_funcs->update(&params, vmbo, addr, 0, entries,
 					     value, flags);
 		if (r)
-			return r;
+			goto exit;
 	}
 
-	return vm->update_funcs->commit(&params, NULL);
+	r = vm->update_funcs->commit(&params, NULL);
+exit:
+	drm_dev_exit(idx);
+	return r;
 }
 
 /**
@@ -1384,11 +1390,14 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
 			  struct amdgpu_vm *vm, bool immediate)
 {
 	struct amdgpu_vm_update_params params;
-	int r;
+	int r, idx;
 
 	if (list_empty(&vm->relocated))
 		return 0;
 
+	if (!drm_dev_enter(&adev->ddev, &idx))
+		return -ENODEV;
+
 	memset(&params, 0, sizeof(params));
 	params.adev = adev;
 	params.vm = vm;
@@ -1396,7 +1405,7 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
 
 	r = vm->update_funcs->prepare(&params, NULL, AMDGPU_SYNC_EXPLICIT);
 	if (r)
-		return r;
+		goto exit;
 
 	while (!list_empty(&vm->relocated)) {
 		struct amdgpu_vm_bo_base *entry;
@@ -1414,10 +1423,13 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
 	r = vm->update_funcs->commit(&params, &vm->last_update);
 	if (r)
 		goto error;
+	drm_dev_exit(idx);
 	return 0;
 
 error:
 	amdgpu_vm_invalidate_pds(adev, vm);
+exit:
+	drm_dev_exit(idx);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h
index a434c71fde8e..995899191288 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h
@@ -204,8 +204,10 @@ struct amd_sriov_msg_pf2vf_info {
 	} mm_bw_management[AMD_SRIOV_MSG_RESERVE_VCN_INST];
 	/* UUID info */
 	struct amd_sriov_msg_uuid_info uuid_info;
+	/* pcie atomic Ops info */
+	uint32_t pcie_atomic_ops_enabled_flags;
 	/* reserved */
-	uint32_t reserved[256 - 47];
+	uint32_t reserved[256 - 48];
 };
 
 struct amd_sriov_msg_vf2pf_info_header {
diff --git a/drivers/gpu/drm/amd/amdgpu/mca_v3_0.c b/drivers/gpu/drm/amd/amdgpu/mca_v3_0.c
index 058b65730a84..8f7107d392af 100644
--- a/drivers/gpu/drm/amd/amdgpu/mca_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mca_v3_0.c
@@ -52,7 +52,8 @@ const struct amdgpu_mca_ras_funcs mca_v3_0_mp0_ras_funcs = {
 	.ras_fini = mca_v3_0_mp0_ras_fini,
 	.query_ras_error_count = mca_v3_0_mp0_query_ras_error_count,
 	.query_ras_error_address = NULL,
-	.ras_block = AMDGPU_RAS_BLOCK__MP0,
+	.ras_block = AMDGPU_RAS_BLOCK__MCA,
+	.ras_sub_block = AMDGPU_RAS_MCA_BLOCK__MP0,
 	.sysfs_name = "mp0_err_count",
 };
 
@@ -79,7 +80,8 @@ const struct amdgpu_mca_ras_funcs mca_v3_0_mp1_ras_funcs = {
 	.ras_fini = mca_v3_0_mp1_ras_fini,
 	.query_ras_error_count = mca_v3_0_mp1_query_ras_error_count,
 	.query_ras_error_address = NULL,
-	.ras_block = AMDGPU_RAS_BLOCK__MP1,
+	.ras_block = AMDGPU_RAS_BLOCK__MCA,
+	.ras_sub_block = AMDGPU_RAS_MCA_BLOCK__MP1,
 	.sysfs_name = "mp1_err_count",
 };
 
@@ -106,7 +108,8 @@ const struct amdgpu_mca_ras_funcs mca_v3_0_mpio_ras_funcs = {
 	.ras_fini = mca_v3_0_mpio_ras_fini,
 	.query_ras_error_count = mca_v3_0_mpio_query_ras_error_count,
 	.query_ras_error_address = NULL,
-	.ras_block = AMDGPU_RAS_BLOCK__MPIO,
+	.ras_block = AMDGPU_RAS_BLOCK__MCA,
+	.ras_sub_block = AMDGPU_RAS_MCA_BLOCK__MPIO,
 	.sysfs_name = "mpio_err_count",
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
index f50045cebd44..91b3afa946f5 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
@@ -387,13 +387,13 @@ static void nbio_v7_4_handle_ras_controller_intr_no_bifring(struct amdgpu_device
 						"errors detected in %s block, "
 						"no user action is needed.\n",
 						obj->err_data.ce_count,
-						ras_block_str(adev->nbio.ras_if->block));
+						get_ras_block_str(adev->nbio.ras_if));
 
 			if (err_data.ue_count)
 				dev_info(adev->dev, "%ld uncorrectable hardware "
 						"errors detected in %s block\n",
 						obj->err_data.ue_count,
-						ras_block_str(adev->nbio.ras_if->block));
+						get_ras_block_str(adev->nbio.ras_if));
 		}
 
 		dev_info(adev->dev, "RAS controller interrupt triggered "
@@ -566,7 +566,9 @@ static int nbio_v7_4_init_ras_err_event_athub_interrupt (struct amdgpu_device *a
 	return r;
 }
 
-#define smnPARITY_ERROR_STATUS_UNCORR_GRP2	0x13a20030
+#define smnPARITY_ERROR_STATUS_UNCORR_GRP2	    0x13a20030
+#define smnPARITY_ERROR_STATUS_UNCORR_GRP2_ALDE	0x13b20030
+#define smnRAS_GLOBAL_STATUS_LO_ALDE            0x13b20020
 
 static void nbio_v7_4_query_ras_error_count(struct amdgpu_device *adev,
 					void *ras_error_status)
@@ -575,12 +577,20 @@ static void nbio_v7_4_query_ras_error_count(struct amdgpu_device *adev,
 	uint32_t corr, fatal, non_fatal;
 	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
 
-	global_sts = RREG32_PCIE(smnRAS_GLOBAL_STATUS_LO);
+	if (adev->asic_type == CHIP_ALDEBARAN)
+		global_sts = RREG32_PCIE(smnRAS_GLOBAL_STATUS_LO_ALDE);
+	else
+		global_sts = RREG32_PCIE(smnRAS_GLOBAL_STATUS_LO);
+
 	corr = REG_GET_FIELD(global_sts, RAS_GLOBAL_STATUS_LO, ParityErrCorr);
 	fatal = REG_GET_FIELD(global_sts, RAS_GLOBAL_STATUS_LO, ParityErrFatal);
 	non_fatal = REG_GET_FIELD(global_sts, RAS_GLOBAL_STATUS_LO,
 				ParityErrNonFatal);
-	parity_sts = RREG32_PCIE(smnPARITY_ERROR_STATUS_UNCORR_GRP2);
+
+	if (adev->asic_type == CHIP_ALDEBARAN)
+		parity_sts = RREG32_PCIE(smnPARITY_ERROR_STATUS_UNCORR_GRP2_ALDE);
+	else
+		parity_sts = RREG32_PCIE(smnPARITY_ERROR_STATUS_UNCORR_GRP2);
 
 	if (corr)
 		err_data->ce_count++;
@@ -589,13 +599,21 @@ static void nbio_v7_4_query_ras_error_count(struct amdgpu_device *adev,
 
 	if (corr || fatal || non_fatal) {
 		central_sts = RREG32_PCIE(smnBIFL_RAS_CENTRAL_STATUS);
+
 		/* clear error status register */
-		WREG32_PCIE(smnRAS_GLOBAL_STATUS_LO, global_sts);
+		if (adev->asic_type == CHIP_ALDEBARAN)
+			WREG32_PCIE(smnRAS_GLOBAL_STATUS_LO_ALDE, global_sts);
+		else
+			WREG32_PCIE(smnRAS_GLOBAL_STATUS_LO, global_sts);
 
 		if (fatal)
+		{
 			/* clear parity fatal error indication field */
-			WREG32_PCIE(smnPARITY_ERROR_STATUS_UNCORR_GRP2,
-				    parity_sts);
+			if (adev->asic_type == CHIP_ALDEBARAN)
+				WREG32_PCIE(smnPARITY_ERROR_STATUS_UNCORR_GRP2_ALDE, parity_sts);
+			else
+				WREG32_PCIE(smnPARITY_ERROR_STATUS_UNCORR_GRP2, parity_sts);
+		}
 
 		if (REG_GET_FIELD(central_sts, BIFL_RAS_CENTRAL_STATUS,
 				BIFL_RasContller_Intr_Recv)) {
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
index 5872d68ed13d..59644015dfc3 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
@@ -84,28 +84,28 @@ static int psp_v10_0_init_microcode(struct psp_context *psp)
 
 		ta_hdr = (const struct ta_firmware_header_v1_0 *)
 				 adev->psp.ta_fw->data;
-		adev->psp.hdcp.feature_version =
+		adev->psp.hdcp_context.context.bin_desc.feature_version =
 			le32_to_cpu(ta_hdr->hdcp.fw_version);
-		adev->psp.hdcp.size_bytes =
+		adev->psp.hdcp_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->hdcp.size_bytes);
-		adev->psp.hdcp.start_addr =
+		adev->psp.hdcp_context.context.bin_desc.start_addr =
 			(uint8_t *)ta_hdr +
 			le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
 
-		adev->psp.dtm.feature_version =
+		adev->psp.dtm_context.context.bin_desc.feature_version =
 			le32_to_cpu(ta_hdr->dtm.fw_version);
-		adev->psp.dtm.size_bytes =
+		adev->psp.dtm_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->dtm.size_bytes);
-		adev->psp.dtm.start_addr =
-			(uint8_t *)adev->psp.hdcp.start_addr +
+		adev->psp.dtm_context.context.bin_desc.start_addr =
+			(uint8_t *)adev->psp.hdcp_context.context.bin_desc.start_addr +
 			le32_to_cpu(ta_hdr->dtm.offset_bytes);
 
-		adev->psp.securedisplay.feature_version =
+		adev->psp.securedisplay_context.context.bin_desc.feature_version =
 			le32_to_cpu(ta_hdr->securedisplay.fw_version);
-		adev->psp.securedisplay.size_bytes =
+		adev->psp.securedisplay_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->securedisplay.size_bytes);
-		adev->psp.securedisplay.start_addr =
-			(uint8_t *)adev->psp.hdcp.start_addr +
+		adev->psp.securedisplay_context.context.bin_desc.start_addr =
+			(uint8_t *)adev->psp.hdcp_context.context.bin_desc.start_addr +
 			le32_to_cpu(ta_hdr->securedisplay.offset_bytes);
 
 		adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
index 29bf9f09944b..3c02e75fd366 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
@@ -151,14 +151,20 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 				goto out2;
 
 			ta_hdr = (const struct ta_firmware_header_v1_0 *)adev->psp.ta_fw->data;
-			adev->psp.xgmi.feature_version = le32_to_cpu(ta_hdr->xgmi.fw_version);
-			adev->psp.xgmi.size_bytes = le32_to_cpu(ta_hdr->xgmi.size_bytes);
-			adev->psp.xgmi.start_addr = (uint8_t *)ta_hdr +
+			adev->psp.xgmi_context.context.bin_desc.feature_version =
+				le32_to_cpu(ta_hdr->xgmi.fw_version);
+			adev->psp.xgmi_context.context.bin_desc.size_bytes =
+				le32_to_cpu(ta_hdr->xgmi.size_bytes);
+			adev->psp.xgmi_context.context.bin_desc.start_addr =
+				(uint8_t *)ta_hdr +
 				le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
 			adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
-			adev->psp.ras.feature_version = le32_to_cpu(ta_hdr->ras.fw_version);
-			adev->psp.ras.size_bytes = le32_to_cpu(ta_hdr->ras.size_bytes);
-			adev->psp.ras.start_addr = (uint8_t *)adev->psp.xgmi.start_addr +
+			adev->psp.ras_context.context.bin_desc.feature_version =
+				le32_to_cpu(ta_hdr->ras.fw_version);
+			adev->psp.ras_context.context.bin_desc.size_bytes =
+				le32_to_cpu(ta_hdr->ras.size_bytes);
+			adev->psp.ras_context.context.bin_desc.start_addr =
+				(uint8_t *)adev->psp.xgmi_context.context.bin_desc.start_addr +
 				le32_to_cpu(ta_hdr->ras.offset_bytes);
 		}
 		break;
@@ -186,16 +192,24 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 				goto out2;
 
 			ta_hdr = (const struct ta_firmware_header_v1_0 *)adev->psp.ta_fw->data;
-			adev->psp.hdcp.feature_version = le32_to_cpu(ta_hdr->hdcp.fw_version);
-			adev->psp.hdcp.size_bytes = le32_to_cpu(ta_hdr->hdcp.size_bytes);
-			adev->psp.hdcp.start_addr = (uint8_t *)ta_hdr +
-				le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
+			adev->psp.hdcp_context.context.bin_desc.feature_version =
+				le32_to_cpu(ta_hdr->hdcp.fw_version);
+			adev->psp.hdcp_context.context.bin_desc.size_bytes =
+				le32_to_cpu(ta_hdr->hdcp.size_bytes);
+			adev->psp.hdcp_context.context.bin_desc.start_addr =
+				(uint8_t *)ta_hdr +
+				le32_to_cpu(
+					ta_hdr->header.ucode_array_offset_bytes);
 
 			adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
 
-			adev->psp.dtm.feature_version = le32_to_cpu(ta_hdr->dtm.fw_version);
-			adev->psp.dtm.size_bytes = le32_to_cpu(ta_hdr->dtm.size_bytes);
-			adev->psp.dtm.start_addr = (uint8_t *)adev->psp.hdcp.start_addr +
+			adev->psp.dtm_context.context.bin_desc.feature_version =
+				le32_to_cpu(ta_hdr->dtm.fw_version);
+			adev->psp.dtm_context.context.bin_desc.size_bytes =
+				le32_to_cpu(ta_hdr->dtm.size_bytes);
+			adev->psp.dtm_context.context.bin_desc.start_addr =
+				(uint8_t *)adev->psp.hdcp_context.context
+					.bin_desc.start_addr +
 				le32_to_cpu(ta_hdr->dtm.offset_bytes);
 		}
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
index cc649406234b..281bc4d7f0a1 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
@@ -84,22 +84,22 @@ static int psp_v12_0_init_microcode(struct psp_context *psp)
 
 		ta_hdr = (const struct ta_firmware_header_v1_0 *)
 				 adev->psp.ta_fw->data;
-		adev->psp.hdcp.feature_version =
+		adev->psp.hdcp_context.context.bin_desc.feature_version =
 			le32_to_cpu(ta_hdr->hdcp.fw_version);
-		adev->psp.hdcp.size_bytes =
+		adev->psp.hdcp_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->hdcp.size_bytes);
-		adev->psp.hdcp.start_addr =
+		adev->psp.hdcp_context.context.bin_desc.start_addr =
 			(uint8_t *)ta_hdr +
 			le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
 
 		adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
 
-		adev->psp.dtm.feature_version =
+		adev->psp.dtm_context.context.bin_desc.feature_version =
 			le32_to_cpu(ta_hdr->dtm.fw_version);
-		adev->psp.dtm.size_bytes =
+		adev->psp.dtm_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->dtm.size_bytes);
-		adev->psp.dtm.start_addr =
-			(uint8_t *)adev->psp.hdcp.start_addr +
+		adev->psp.dtm_context.context.bin_desc.start_addr =
+			(uint8_t *)adev->psp.hdcp_context.context.bin_desc.start_addr +
 			le32_to_cpu(ta_hdr->dtm.offset_bytes);
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c
index 779f5c911e11..e4a96e7e386d 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c
@@ -375,10 +375,10 @@ static void sdma_v5_2_ring_emit_ib(struct amdgpu_ring *ring,
  */
 static void sdma_v5_2_ring_emit_mem_sync(struct amdgpu_ring *ring)
 {
-    uint32_t gcr_cntl =
-		    SDMA_GCR_GL2_INV | SDMA_GCR_GL2_WB | SDMA_GCR_GLM_INV |
-			SDMA_GCR_GL1_INV | SDMA_GCR_GLV_INV | SDMA_GCR_GLK_INV |
-			SDMA_GCR_GLI_INV(1);
+	uint32_t gcr_cntl = SDMA_GCR_GL2_INV | SDMA_GCR_GL2_WB |
+			    SDMA_GCR_GLM_INV | SDMA_GCR_GL1_INV |
+			    SDMA_GCR_GLV_INV | SDMA_GCR_GLK_INV |
+			    SDMA_GCR_GLI_INV(1);
 
 	/* flush entire cache L0/L1/L2, this can be optimized by performance requirement */
 	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_GCR_REQ));
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 0fc97c364fd7..15e295a1412c 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -85,6 +85,8 @@
 #define mmMP0_MISC_LIGHT_SLEEP_CTRL                                                             0x01ba
 #define mmMP0_MISC_LIGHT_SLEEP_CTRL_BASE_IDX                                                    0
 
+static const struct amd_ip_funcs soc15_common_ip_funcs;
+
 /* Vega, Raven, Arcturus */
 static const struct amdgpu_video_codec_info vega_video_codecs_encode_array[] =
 {
@@ -1645,7 +1647,7 @@ static int soc15_common_set_powergating_state(void *handle,
 	return 0;
 }
 
-const struct amd_ip_funcs soc15_common_ip_funcs = {
+static const struct amd_ip_funcs soc15_common_ip_funcs = {
 	.name = "soc15_common",
 	.early_init = soc15_common_early_init,
 	.late_init = soc15_common_late_init,
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.h b/drivers/gpu/drm/amd/amdgpu/soc15.h
index 034cfdfc4dbe..a025339ac5e9 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.h
@@ -31,8 +31,6 @@
 #define SOC15_FLUSH_GPU_TLB_NUM_WREG		6
 #define SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT	3
 
-extern const struct amd_ip_funcs soc15_common_ip_funcs;
-
 struct soc15_reg_golden {
 	u32	hwip;
 	u32	instance;
diff --git a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h
index 0f214a398dd8..de24a0a97d5e 100644
--- a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h
+++ b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h
@@ -38,9 +38,8 @@ enum ras_command {
 	TA_RAS_COMMAND__TRIGGER_ERROR,
 };
 
-enum ta_ras_status
-{
-	TA_RAS_STATUS__SUCCESS                          = 0x00,
+enum ta_ras_status {
+	TA_RAS_STATUS__SUCCESS                          = 0x0000,
 	TA_RAS_STATUS__RESET_NEEDED                     = 0xA001,
 	TA_RAS_STATUS__ERROR_INVALID_PARAMETER          = 0xA002,
 	TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE          = 0xA003,
@@ -55,7 +54,12 @@ enum ta_ras_status
 	TA_RAS_STATUS__ERROR_GET_DEV_INFO               = 0xA00C,
 	TA_RAS_STATUS__ERROR_UNSUPPORTED_DEV            = 0xA00D,
 	TA_RAS_STATUS__ERROR_NOT_INITIALIZED            = 0xA00E,
-	TA_RAS_STATUS__ERROR_TEE_INTERNAL               = 0xA00F
+	TA_RAS_STATUS__ERROR_TEE_INTERNAL               = 0xA00F,
+	TA_RAS_STATUS__ERROR_UNSUPPORTED_FUNCTION       = 0xA010,
+	TA_RAS_STATUS__ERROR_SYS_DRV_REG_ACCESS         = 0xA011,
+	TA_RAS_STATUS__ERROR_RAS_READ_WRITE             = 0xA012,
+	TA_RAS_STATUS__ERROR_NULL_PTR                   = 0xA013,
+	TA_RAS_STATUS__ERROR_UNSUPPORTED_IP             = 0xA014
 };
 
 enum ta_ras_block {
@@ -73,9 +77,18 @@ enum ta_ras_block {
 	TA_RAS_BLOCK__MP0,
 	TA_RAS_BLOCK__MP1,
 	TA_RAS_BLOCK__FUSE,
+	TA_RAS_BLOCK__MCA,
 	TA_NUM_BLOCK_MAX
 };
 
+enum ta_ras_mca_block {
+	TA_RAS_MCA_BLOCK__MP0   = 0,
+	TA_RAS_MCA_BLOCK__MP1   = 1,
+	TA_RAS_MCA_BLOCK__MPIO  = 2,
+	TA_RAS_MCA_BLOCK__IOHC  = 3,
+	TA_MCA_NUM_BLOCK_MAX
+};
+
 enum ta_ras_error_type {
 	TA_RAS_ERROR__NONE			= 0,
 	TA_RAS_ERROR__PARITY			= 1,
@@ -105,17 +118,15 @@ struct ta_ras_trigger_error_input {
 	uint64_t		value;			// method if error injection. i.e persistent, coherent etc.
 };
 
-struct ta_ras_init_flags
-{
-    uint8_t     poison_mode_en;
-    uint8_t     dgpu_mode;
+struct ta_ras_init_flags {
+	uint8_t poison_mode_en;
+	uint8_t dgpu_mode;
 };
 
-struct ta_ras_output_flags
-{
-	uint8_t    ras_init_success_flag;
-	uint8_t    err_inject_switch_disable_flag;
-	uint8_t    reg_access_failure_flag;
+struct ta_ras_output_flags {
+	uint8_t ras_init_success_flag;
+	uint8_t err_inject_switch_disable_flag;
+	uint8_t reg_access_failure_flag;
 };
 
 /* Common input structure for RAS callbacks */
@@ -126,14 +137,13 @@ union ta_ras_cmd_input {
 	struct ta_ras_disable_features_input	disable_features;
 	struct ta_ras_trigger_error_input	trigger_error;
 
-	uint32_t	reserve_pad[256];
+	uint32_t reserve_pad[256];
 };
 
-union ta_ras_cmd_output
-{
-	struct ta_ras_output_flags  flags;
+union ta_ras_cmd_output {
+	struct ta_ras_output_flags flags;
 
-	uint32_t	reserve_pad[256];
+	uint32_t reserve_pad[256];
 };
 
 /* Shared Memory structures */
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c
index 7232241e3bfb..0fef925b6602 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c
@@ -698,6 +698,19 @@ static int uvd_v3_1_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cancel_delayed_work_sync(&adev->uvd.idle_work);
+
+	if (RREG32(mmUVD_STATUS) != 0)
+		uvd_v3_1_stop(adev);
+
+	return 0;
+}
+
+static int uvd_v3_1_suspend(void *handle)
+{
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
 	/*
 	 * Proper cleanups before halting the HW engine:
 	 *   - cancel the delayed idle work
@@ -722,17 +735,6 @@ static int uvd_v3_1_hw_fini(void *handle)
 						       AMD_CG_STATE_GATE);
 	}
 
-	if (RREG32(mmUVD_STATUS) != 0)
-		uvd_v3_1_stop(adev);
-
-	return 0;
-}
-
-static int uvd_v3_1_suspend(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
 	r = uvd_v3_1_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index 52d6de969f46..c108b8381795 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -212,6 +212,19 @@ static int uvd_v4_2_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cancel_delayed_work_sync(&adev->uvd.idle_work);
+
+	if (RREG32(mmUVD_STATUS) != 0)
+		uvd_v4_2_stop(adev);
+
+	return 0;
+}
+
+static int uvd_v4_2_suspend(void *handle)
+{
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
 	/*
 	 * Proper cleanups before halting the HW engine:
 	 *   - cancel the delayed idle work
@@ -236,17 +249,6 @@ static int uvd_v4_2_hw_fini(void *handle)
 						       AMD_CG_STATE_GATE);
 	}
 
-	if (RREG32(mmUVD_STATUS) != 0)
-		uvd_v4_2_stop(adev);
-
-	return 0;
-}
-
-static int uvd_v4_2_suspend(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
 	r = uvd_v4_2_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index db6d06758e4d..563493d1f830 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -210,6 +210,19 @@ static int uvd_v5_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cancel_delayed_work_sync(&adev->uvd.idle_work);
+
+	if (RREG32(mmUVD_STATUS) != 0)
+		uvd_v5_0_stop(adev);
+
+	return 0;
+}
+
+static int uvd_v5_0_suspend(void *handle)
+{
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
 	/*
 	 * Proper cleanups before halting the HW engine:
 	 *   - cancel the delayed idle work
@@ -234,17 +247,6 @@ static int uvd_v5_0_hw_fini(void *handle)
 						       AMD_CG_STATE_GATE);
 	}
 
-	if (RREG32(mmUVD_STATUS) != 0)
-		uvd_v5_0_stop(adev);
-
-	return 0;
-}
-
-static int uvd_v5_0_suspend(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
 	r = uvd_v5_0_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index bc571833632e..d5d023a24269 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -332,15 +332,9 @@ err:
 static int uvd_v6_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
-	struct amdgpu_bo *bo = NULL;
+	struct amdgpu_bo *bo = ring->adev->uvd.ib_bo;
 	long r;
 
-	r = amdgpu_bo_create_reserved(ring->adev, 128 * 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      &bo, NULL, NULL);
-	if (r)
-		return r;
-
 	r = uvd_v6_0_enc_get_create_msg(ring, 1, bo, NULL);
 	if (r)
 		goto error;
@@ -357,9 +351,6 @@ static int uvd_v6_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
-	amdgpu_bo_unpin(bo);
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_unref(&bo);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index b6e82d75561f..b483f03b4591 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -338,15 +338,9 @@ err:
 static int uvd_v7_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence = NULL;
-	struct amdgpu_bo *bo = NULL;
+	struct amdgpu_bo *bo = ring->adev->uvd.ib_bo;
 	long r;
 
-	r = amdgpu_bo_create_reserved(ring->adev, 128 * 1024, PAGE_SIZE,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      &bo, NULL, NULL);
-	if (r)
-		return r;
-
 	r = uvd_v7_0_enc_get_create_msg(ring, 1, bo, NULL);
 	if (r)
 		goto error;
@@ -363,9 +357,6 @@ static int uvd_v7_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 error:
 	dma_fence_put(fence);
-	amdgpu_bo_unpin(bo);
-	amdgpu_bo_unreserve(bo);
-	amdgpu_bo_unref(&bo);
 	return r;
 }
 
@@ -606,6 +597,23 @@ static int uvd_v7_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cancel_delayed_work_sync(&adev->uvd.idle_work);
+
+	if (!amdgpu_sriov_vf(adev))
+		uvd_v7_0_stop(adev);
+	else {
+		/* full access mode, so don't touch any UVD register */
+		DRM_DEBUG("For SRIOV client, shouldn't do anything.\n");
+	}
+
+	return 0;
+}
+
+static int uvd_v7_0_suspend(void *handle)
+{
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
 	/*
 	 * Proper cleanups before halting the HW engine:
 	 *   - cancel the delayed idle work
@@ -630,21 +638,6 @@ static int uvd_v7_0_hw_fini(void *handle)
 						       AMD_CG_STATE_GATE);
 	}
 
-	if (!amdgpu_sriov_vf(adev))
-		uvd_v7_0_stop(adev);
-	else {
-		/* full access mode, so don't touch any UVD register */
-		DRM_DEBUG("For SRIOV client, shouldn't do anything.\n");
-	}
-
-	return 0;
-}
-
-static int uvd_v7_0_suspend(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
 	r = uvd_v7_0_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
index b70c17f0c52e..67eb01fef789 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
@@ -431,10 +431,12 @@ static int vce_v2_0_sw_init(void *handle)
 		return r;
 
 	for (i = 0; i < adev->vce.num_rings; i++) {
+		enum amdgpu_ring_priority_level hw_prio = amdgpu_vce_get_ring_prio(i);
+
 		ring = &adev->vce.ring[i];
 		sprintf(ring->name, "vce%d", i);
 		r = amdgpu_ring_init(adev, ring, 512, &adev->vce.irq, 0,
-				     AMDGPU_RING_PRIO_DEFAULT, NULL);
+				     hw_prio, NULL);
 		if (r)
 			return r;
 	}
@@ -479,6 +481,17 @@ static int vce_v2_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cancel_delayed_work_sync(&adev->vce.idle_work);
+
+	return 0;
+}
+
+static int vce_v2_0_suspend(void *handle)
+{
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+
 	/*
 	 * Proper cleanups before halting the HW engine:
 	 *   - cancel the delayed idle work
@@ -502,14 +515,6 @@ static int vce_v2_0_hw_fini(void *handle)
 						       AMD_CG_STATE_GATE);
 	}
 
-	return 0;
-}
-
-static int vce_v2_0_suspend(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
 	r = vce_v2_0_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 9de66893ccd6..142e291983b4 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -440,10 +440,12 @@ static int vce_v3_0_sw_init(void *handle)
 		return r;
 
 	for (i = 0; i < adev->vce.num_rings; i++) {
+		enum amdgpu_ring_priority_level hw_prio = amdgpu_vce_get_ring_prio(i);
+
 		ring = &adev->vce.ring[i];
 		sprintf(ring->name, "vce%d", i);
 		r = amdgpu_ring_init(adev, ring, 512, &adev->vce.irq, 0,
-				     AMDGPU_RING_PRIO_DEFAULT, NULL);
+				     hw_prio, NULL);
 		if (r)
 			return r;
 	}
@@ -490,6 +492,21 @@ static int vce_v3_0_hw_fini(void *handle)
 	int r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cancel_delayed_work_sync(&adev->vce.idle_work);
+
+	r = vce_v3_0_wait_for_idle(handle);
+	if (r)
+		return r;
+
+	vce_v3_0_stop(adev);
+	return vce_v3_0_set_clockgating_state(adev, AMD_CG_STATE_GATE);
+}
+
+static int vce_v3_0_suspend(void *handle)
+{
+	int r;
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
 	/*
 	 * Proper cleanups before halting the HW engine:
 	 *   - cancel the delayed idle work
@@ -513,19 +530,6 @@ static int vce_v3_0_hw_fini(void *handle)
 						       AMD_CG_STATE_GATE);
 	}
 
-	r = vce_v3_0_wait_for_idle(handle);
-	if (r)
-		return r;
-
-	vce_v3_0_stop(adev);
-	return vce_v3_0_set_clockgating_state(adev, AMD_CG_STATE_GATE);
-}
-
-static int vce_v3_0_suspend(void *handle)
-{
-	int r;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
 	r = vce_v3_0_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
index fec902b800c2..226b79254db8 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
@@ -463,6 +463,8 @@ static int vce_v4_0_sw_init(void *handle)
 	}
 
 	for (i = 0; i < adev->vce.num_rings; i++) {
+		enum amdgpu_ring_priority_level hw_prio = amdgpu_vce_get_ring_prio(i);
+
 		ring = &adev->vce.ring[i];
 		sprintf(ring->name, "vce%d", i);
 		if (amdgpu_sriov_vf(adev)) {
@@ -478,7 +480,7 @@ static int vce_v4_0_sw_init(void *handle)
 				ring->doorbell_index = adev->doorbell_index.uvd_vce.vce_ring2_3 * 2 + 1;
 		}
 		r = amdgpu_ring_init(adev, ring, 512, &adev->vce.irq, 0,
-				     AMDGPU_RING_PRIO_DEFAULT, NULL);
+				     hw_prio, NULL);
 		if (r)
 			return r;
 	}
@@ -542,29 +544,8 @@ static int vce_v4_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	/*
-	 * Proper cleanups before halting the HW engine:
-	 *   - cancel the delayed idle work
-	 *   - enable powergating
-	 *   - enable clockgating
-	 *   - disable dpm
-	 *
-	 * TODO: to align with the VCN implementation, move the
-	 * jobs for clockgating/powergating/dpm setting to
-	 * ->set_powergating_state().
-	 */
 	cancel_delayed_work_sync(&adev->vce.idle_work);
 
-	if (adev->pm.dpm_enabled) {
-		amdgpu_dpm_enable_vce(adev, false);
-	} else {
-		amdgpu_asic_set_vce_clocks(adev, 0, 0);
-		amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-						       AMD_PG_STATE_GATE);
-		amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-						       AMD_CG_STATE_GATE);
-	}
-
 	if (!amdgpu_sriov_vf(adev)) {
 		/* vce_v4_0_wait_for_idle(handle); */
 		vce_v4_0_stop(adev);
@@ -594,6 +575,29 @@ static int vce_v4_0_suspend(void *handle)
 		drm_dev_exit(idx);
 	}
 
+	/*
+	 * Proper cleanups before halting the HW engine:
+	 *   - cancel the delayed idle work
+	 *   - enable powergating
+	 *   - enable clockgating
+	 *   - disable dpm
+	 *
+	 * TODO: to align with the VCN implementation, move the
+	 * jobs for clockgating/powergating/dpm setting to
+	 * ->set_powergating_state().
+	 */
+	cancel_delayed_work_sync(&adev->vce.idle_work);
+
+	if (adev->pm.dpm_enabled) {
+		amdgpu_dpm_enable_vce(adev, false);
+	} else {
+		amdgpu_asic_set_vce_clocks(adev, 0, 0);
+		amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+						       AMD_PG_STATE_GATE);
+		amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+						       AMD_CG_STATE_GATE);
+	}
+
 	r = vce_v4_0_hw_fini(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 121ee9f2b8d1..6c11739270c1 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -145,10 +145,12 @@ static int vcn_v1_0_sw_init(void *handle)
 		SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP);
 
 	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
+		enum amdgpu_ring_priority_level hw_prio = amdgpu_vcn_get_enc_ring_prio(i);
+
 		ring = &adev->vcn.inst->ring_enc[i];
 		sprintf(ring->name, "vcn_enc%d", i);
 		r = amdgpu_ring_init(adev, ring, 512, &adev->vcn.inst->irq, 0,
-				     AMDGPU_RING_PRIO_DEFAULT, NULL);
+				     hw_prio, NULL);
 		if (r)
 			return r;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
index f4686e918e0d..a03c0fc8338f 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
@@ -22,6 +22,7 @@
  */
 
 #include <linux/firmware.h>
+#include <drm/drm_drv.h>
 
 #include "amdgpu.h"
 #include "amdgpu_vcn.h"
@@ -159,6 +160,8 @@ static int vcn_v2_0_sw_init(void *handle)
 	adev->vcn.inst->external.nop = SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP);
 
 	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
+		enum amdgpu_ring_priority_level hw_prio = amdgpu_vcn_get_enc_ring_prio(i);
+
 		ring = &adev->vcn.inst->ring_enc[i];
 		ring->use_doorbell = true;
 		if (!amdgpu_sriov_vf(adev))
@@ -167,7 +170,7 @@ static int vcn_v2_0_sw_init(void *handle)
 			ring->doorbell_index = (adev->doorbell_index.vcn.vcn_ring0_1 << 1) + 1 + i;
 		sprintf(ring->name, "vcn_enc%d", i);
 		r = amdgpu_ring_init(adev, ring, 512, &adev->vcn.inst->irq, 0,
-				     AMDGPU_RING_PRIO_DEFAULT, NULL);
+				     hw_prio, NULL);
 		if (r)
 			return r;
 	}
@@ -192,11 +195,14 @@ static int vcn_v2_0_sw_init(void *handle)
  */
 static int vcn_v2_0_sw_fini(void *handle)
 {
-	int r;
+	int r, idx;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	volatile struct amdgpu_fw_shared *fw_shared = adev->vcn.inst->fw_shared_cpu_addr;
 
-	fw_shared->present_flag_0 = 0;
+	if (drm_dev_enter(&adev->ddev, &idx)) {
+		fw_shared->present_flag_0 = 0;
+		drm_dev_exit(idx);
+	}
 
 	amdgpu_virt_free_mm_table(adev);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
index e0c0c3734432..1780ad1eacd6 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
@@ -22,6 +22,7 @@
  */
 
 #include <linux/firmware.h>
+#include <drm/drm_drv.h>
 
 #include "amdgpu.h"
 #include "amdgpu_vcn.h"
@@ -194,6 +195,8 @@ static int vcn_v2_5_sw_init(void *handle)
 			return r;
 
 		for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
+			enum amdgpu_ring_priority_level hw_prio = amdgpu_vcn_get_enc_ring_prio(i);
+
 			ring = &adev->vcn.inst[j].ring_enc[i];
 			ring->use_doorbell = true;
 
@@ -203,7 +206,7 @@ static int vcn_v2_5_sw_init(void *handle)
 			sprintf(ring->name, "vcn_enc_%d.%d", j, i);
 			r = amdgpu_ring_init(adev, ring, 512,
 					     &adev->vcn.inst[j].irq, 0,
-					     AMDGPU_RING_PRIO_DEFAULT, NULL);
+					     hw_prio, NULL);
 			if (r)
 				return r;
 		}
@@ -233,17 +236,21 @@ static int vcn_v2_5_sw_init(void *handle)
  */
 static int vcn_v2_5_sw_fini(void *handle)
 {
-	int i, r;
+	int i, r, idx;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	volatile struct amdgpu_fw_shared *fw_shared;
 
-	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
-		if (adev->vcn.harvest_config & (1 << i))
-			continue;
-		fw_shared = adev->vcn.inst[i].fw_shared_cpu_addr;
-		fw_shared->present_flag_0 = 0;
+	if (drm_dev_enter(&adev->ddev, &idx)) {
+		for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
+			if (adev->vcn.harvest_config & (1 << i))
+				continue;
+			fw_shared = adev->vcn.inst[i].fw_shared_cpu_addr;
+			fw_shared->present_flag_0 = 0;
+		}
+		drm_dev_exit(idx);
 	}
 
+
 	if (amdgpu_sriov_vf(adev))
 		amdgpu_virt_free_mm_table(adev);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c
index 3d18aab88b4e..b1af70d49c6e 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c
@@ -224,6 +224,8 @@ static int vcn_v3_0_sw_init(void *handle)
 			return r;
 
 		for (j = 0; j < adev->vcn.num_enc_rings; ++j) {
+			enum amdgpu_ring_priority_level hw_prio = amdgpu_vcn_get_enc_ring_prio(j);
+
 			/* VCN ENC TRAP */
 			r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_vcns[i],
 				j + VCN_2_0__SRCID__UVD_ENC_GENERAL_PURPOSE, &adev->vcn.inst[i].irq);
@@ -239,8 +241,7 @@ static int vcn_v3_0_sw_init(void *handle)
 			}
 			sprintf(ring->name, "vcn_enc_%d.%d", i, j);
 			r = amdgpu_ring_init(adev, ring, 512, &adev->vcn.inst[i].irq, 0,
-					     AMDGPU_RING_PRIO_DEFAULT,
-					     &adev->vcn.inst[i].sched_score);
+					     hw_prio, &adev->vcn.inst[i].sched_score);
 			if (r)
 				return r;
 		}