Merge branch 'drm-next-4.4' of git://people.freedesktop.org/~agd5f/linux into drm-next

More amdgpu and radeon stuff for drm-next.  Stoney support is the big change.
The rest is just bug fixes and code cleanups.  The Stoney stuff is pretty
low impact with respect to existing chips.

* 'drm-next-4.4' of git://people.freedesktop.org/~agd5f/linux:
  drm/amdgpu: change VM size default to 64GB
  drm/amdgpu: add Stoney pci ids
  drm/amdgpu: update the core VI support for Stoney
  drm/amdgpu: add VCE support for Stoney (v2)
  drm/amdgpu: add UVD support for Stoney
  drm/amdgpu: add GFX support for Stoney (v2)
  drm/amdgpu: add SDMA support for Stoney (v2)
  drm/amdgpu: add DCE support for Stoney
  drm/amdgpu: Update SMC/DPM for Stoney
  drm/amdgpu: add GMC support for Stoney
  drm/amdgpu: add Stoney chip family
  drm/amdgpu: fix the broken vm->mutex V2
  drm/amdgpu: remove the unnecessary parameter adev for amdgpu_fence_wait_any()
  drm/amdgpu: remove the exclusive lock
  drm/amdgpu: remove old lockup detection infrastructure
  drm: fix trivial typos
  drm/amdgpu/dce: simplify suspend/resume
  drm/amdgpu/gfx8: set TC_WB_ACTION_EN in RELEASE_MEM packet
  drm/radeon: Use rdev->gem.mutex to protect hyperz/cmask owners

36 files changed, 616 insertions, 351 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 3fa1397fd7ae..5c400f4b87fd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -345,7 +345,6 @@ struct amdgpu_ring_funcs {
 	/* testing functions */
 	int (*test_ring)(struct amdgpu_ring *ring);
 	int (*test_ib)(struct amdgpu_ring *ring);
-	bool (*is_lockup)(struct amdgpu_ring *ring);
 	/* insert NOP packets */
 	void (*insert_nop)(struct amdgpu_ring *ring, uint32_t count);
 };
@@ -448,8 +447,7 @@ int amdgpu_fence_wait_next(struct amdgpu_ring *ring);
 int amdgpu_fence_wait_empty(struct amdgpu_ring *ring);
 unsigned amdgpu_fence_count_emitted(struct amdgpu_ring *ring);
 
-signed long amdgpu_fence_wait_any(struct amdgpu_device *adev,
-				  struct fence **array,
+signed long amdgpu_fence_wait_any(struct fence **array,
 				  uint32_t count,
 				  bool intr,
 				  signed long t);
@@ -907,8 +905,6 @@ struct amdgpu_ring {
 	unsigned		ring_size;
 	unsigned		ring_free_dw;
 	int			count_dw;
-	atomic_t		last_rptr;
-	atomic64_t		last_activity;
 	uint64_t		gpu_addr;
 	uint32_t		align_mask;
 	uint32_t		ptr_mask;
@@ -1230,8 +1226,6 @@ void amdgpu_ring_commit(struct amdgpu_ring *ring);
 void amdgpu_ring_unlock_commit(struct amdgpu_ring *ring);
 void amdgpu_ring_undo(struct amdgpu_ring *ring);
 void amdgpu_ring_unlock_undo(struct amdgpu_ring *ring);
-void amdgpu_ring_lockup_update(struct amdgpu_ring *ring);
-bool amdgpu_ring_test_lockup(struct amdgpu_ring *ring);
 unsigned amdgpu_ring_backup(struct amdgpu_ring *ring,
 			    uint32_t **data);
 int amdgpu_ring_restore(struct amdgpu_ring *ring,
@@ -1960,7 +1954,6 @@ struct amdgpu_device {
 	struct device			*dev;
 	struct drm_device		*ddev;
 	struct pci_dev			*pdev;
-	struct rw_semaphore		exclusive_lock;
 
 	/* ASIC */
 	enum amd_asic_type		asic_type;
@@ -1974,7 +1967,6 @@ struct amdgpu_device {
 	bool				suspend;
 	bool				need_dma32;
 	bool				accel_working;
-	bool				needs_reset;
 	struct work_struct 		reset_work;
 	struct notifier_block		acpi_nb;
 	struct amdgpu_i2c_chan		*i2c_bus[AMDGPU_MAX_I2C_BUS];
@@ -2253,7 +2245,6 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_ring_parse_cs(r, p, ib) ((r)->funcs->parse_cs((p), (ib)))
 #define amdgpu_ring_test_ring(r) (r)->funcs->test_ring((r))
 #define amdgpu_ring_test_ib(r) (r)->funcs->test_ib((r))
-#define amdgpu_ring_is_lockup(r) (r)->funcs->is_lockup((r))
 #define amdgpu_ring_get_rptr(r) (r)->funcs->get_rptr((r))
 #define amdgpu_ring_get_wptr(r) (r)->funcs->get_wptr((r))
 #define amdgpu_ring_set_wptr(r) (r)->funcs->set_wptr((r))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index baf00617fe90..27ef52847e6d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -609,7 +609,6 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
 		}
 	}
 
-	mutex_lock(&vm->mutex);
 	r = amdgpu_bo_vm_update_pte(parser, vm);
 	if (r) {
 		goto out;
@@ -620,7 +619,6 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
 				       parser->filp);
 
 out:
-	mutex_unlock(&vm->mutex);
 	return r;
 }
 
@@ -828,15 +826,14 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 {
 	struct amdgpu_device *adev = dev->dev_private;
 	union drm_amdgpu_cs *cs = data;
+	struct amdgpu_fpriv *fpriv = filp->driver_priv;
+	struct amdgpu_vm *vm = &fpriv->vm;
 	struct amdgpu_cs_parser *parser;
 	bool reserved_buffers = false;
 	int i, r;
 
-	down_read(&adev->exclusive_lock);
-	if (!adev->accel_working) {
-		up_read(&adev->exclusive_lock);
+	if (!adev->accel_working)
 		return -EBUSY;
-	}
 
 	parser = amdgpu_cs_parser_create(adev, filp, NULL, NULL, 0);
 	if (!parser)
@@ -844,12 +841,11 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 	r = amdgpu_cs_parser_init(parser, data);
 	if (r) {
 		DRM_ERROR("Failed to initialize parser !\n");
-		kfree(parser);
-		up_read(&adev->exclusive_lock);
+		amdgpu_cs_parser_fini(parser, r, false);
 		r = amdgpu_cs_handle_lockup(adev, r);
 		return r;
 	}
-
+	mutex_lock(&vm->mutex);
 	r = amdgpu_cs_parser_relocs(parser);
 	if (r == -ENOMEM)
 		DRM_ERROR("Not enough memory for command submission!\n");
@@ -916,14 +912,14 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 
 		mutex_unlock(&job->job_lock);
 		amdgpu_cs_parser_fini_late(parser);
-		up_read(&adev->exclusive_lock);
+		mutex_unlock(&vm->mutex);
 		return 0;
 	}
 
 	cs->out.handle = parser->ibs[parser->num_ibs - 1].sequence;
 out:
 	amdgpu_cs_parser_fini(parser, r, reserved_buffers);
-	up_read(&adev->exclusive_lock);
+	mutex_unlock(&vm->mutex);
 	r = amdgpu_cs_handle_lockup(adev, r);
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 901a460b2c55..d5b421330145 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -57,6 +57,7 @@ static const char *amdgpu_asic_name[] = {
 	"TONGA",
 	"FIJI",
 	"CARRIZO",
+	"STONEY",
 	"LAST",
 };
 
@@ -1165,7 +1166,8 @@ static int amdgpu_early_init(struct amdgpu_device *adev)
 	case CHIP_TONGA:
 	case CHIP_FIJI:
 	case CHIP_CARRIZO:
-		if (adev->asic_type == CHIP_CARRIZO)
+	case CHIP_STONEY:
+		if (adev->asic_type == CHIP_CARRIZO || adev->asic_type == CHIP_STONEY)
 			adev->family = AMDGPU_FAMILY_CZ;
 		else
 			adev->family = AMDGPU_FAMILY_VI;
@@ -1418,7 +1420,6 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	mutex_init(&adev->gfx.gpu_clock_mutex);
 	mutex_init(&adev->srbm_mutex);
 	mutex_init(&adev->grbm_idx_mutex);
-	init_rwsem(&adev->exclusive_lock);
 	mutex_init(&adev->mn_lock);
 	hash_init(adev->mn_hash);
 
@@ -1814,14 +1815,6 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev)
 	int i, r;
 	int resched;
 
-	down_write(&adev->exclusive_lock);
-
-	if (!adev->needs_reset) {
-		up_write(&adev->exclusive_lock);
-		return 0;
-	}
-
-	adev->needs_reset = false;
 	atomic_inc(&adev->gpu_reset_counter);
 
 	/* block TTM */
@@ -1885,7 +1878,6 @@ retry:
 		dev_info(adev->dev, "GPU reset failed\n");
 	}
 
-	up_write(&adev->exclusive_lock);
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
index de116398fa49..d747fca59669 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -47,11 +47,8 @@ static void amdgpu_flip_wait_fence(struct amdgpu_device *adev,
 	fence = to_amdgpu_fence(*f);
 	if (fence) {
 		r = fence_wait(&fence->base, false);
-		if (r == -EDEADLK) {
-			up_read(&adev->exclusive_lock);
+		if (r == -EDEADLK)
 			r = amdgpu_gpu_reset(adev);
-			down_read(&adev->exclusive_lock);
-		}
 	} else
 		r = fence_wait(*f, false);
 
@@ -77,7 +74,6 @@ static void amdgpu_flip_work_func(struct work_struct *__work)
 	unsigned long flags;
 	unsigned i;
 
-	down_read(&adev->exclusive_lock);
 	amdgpu_flip_wait_fence(adev, &work->excl);
 	for (i = 0; i < work->shared_count; ++i)
 		amdgpu_flip_wait_fence(adev, &work->shared[i]);
@@ -93,7 +89,6 @@ static void amdgpu_flip_work_func(struct work_struct *__work)
 	amdgpuCrtc->pflip_status = AMDGPU_FLIP_SUBMITTED;
 
 	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
-	up_read(&adev->exclusive_lock);
 }
 
 /*
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index ef58774b242c..0508c5cd103a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -73,7 +73,7 @@ int amdgpu_hard_reset = 0;
 unsigned amdgpu_ip_block_mask = 0xffffffff;
 int amdgpu_bapm = -1;
 int amdgpu_deep_color = 0;
-int amdgpu_vm_size = 8;
+int amdgpu_vm_size = 64;
 int amdgpu_vm_block_size = -1;
 int amdgpu_vm_fault_stop = 0;
 int amdgpu_vm_debug = 0;
@@ -137,7 +137,7 @@ module_param_named(bapm, amdgpu_bapm, int, 0444);
 MODULE_PARM_DESC(deep_color, "Deep Color support (1 = enable, 0 = disable (default))");
 module_param_named(deep_color, amdgpu_deep_color, int, 0444);
 
-MODULE_PARM_DESC(vm_size, "VM address space size in gigabytes (default 8GB)");
+MODULE_PARM_DESC(vm_size, "VM address space size in gigabytes (default 64GB)");
 module_param_named(vm_size, amdgpu_vm_size, int, 0444);
 
 MODULE_PARM_DESC(vm_block_size, "VM page table size in bits (default depending on vm_size)");
@@ -273,6 +273,8 @@ static struct pci_device_id pciidlist[] = {
 	{0x1002, 0x9875, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_CARRIZO|AMD_IS_APU},
 	{0x1002, 0x9876, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_CARRIZO|AMD_IS_APU},
 	{0x1002, 0x9877, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_CARRIZO|AMD_IS_APU},
+	/* stoney */
+	{0x1002, 0x98E4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_STONEY|AMD_IS_APU},
 
 	{0, 0, 0}
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index fcad7e060938..003a219943f1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -260,27 +260,8 @@ static void amdgpu_fence_check_lockup(struct work_struct *work)
 				lockup_work.work);
 	ring = fence_drv->ring;
 
-	if (!down_read_trylock(&ring->adev->exclusive_lock)) {
-		/* just reschedule the check if a reset is going on */
-		amdgpu_fence_schedule_check(ring);
-		return;
-	}
-
-	if (amdgpu_fence_activity(ring)) {
-		wake_up_all(&ring->fence_drv.fence_queue);
-	}
-	else if (amdgpu_ring_is_lockup(ring)) {
-		/* good news we believe it's a lockup */
-		dev_warn(ring->adev->dev, "GPU lockup (current fence id "
-			"0x%016llx last fence id 0x%016llx on ring %d)\n",
-			(uint64_t)atomic64_read(&fence_drv->last_seq),
-			fence_drv->sync_seq[ring->idx], ring->idx);
-
-		/* remember that we need an reset */
-		ring->adev->needs_reset = true;
+	if (amdgpu_fence_activity(ring))
 		wake_up_all(&ring->fence_drv.fence_queue);
-	}
-	up_read(&ring->adev->exclusive_lock);
 }
 
 /**
@@ -328,18 +309,15 @@ static bool amdgpu_fence_is_signaled(struct fence *f)
 {
 	struct amdgpu_fence *fence = to_amdgpu_fence(f);
 	struct amdgpu_ring *ring = fence->ring;
-	struct amdgpu_device *adev = ring->adev;
 
 	if (atomic64_read(&ring->fence_drv.last_seq) >= fence->seq)
 		return true;
 
-	if (down_read_trylock(&adev->exclusive_lock)) {
-		amdgpu_fence_process(ring);
-		up_read(&adev->exclusive_lock);
+	amdgpu_fence_process(ring);
+
+	if (atomic64_read(&ring->fence_drv.last_seq) >= fence->seq)
+		return true;
 
-		if (atomic64_read(&ring->fence_drv.last_seq) >= fence->seq)
-			return true;
-	}
 	return false;
 }
 
@@ -380,7 +358,6 @@ static bool amdgpu_fence_enable_signaling(struct fence *f)
  */
 static int amdgpu_fence_ring_wait_seq(struct amdgpu_ring *ring, uint64_t seq)
 {
-	struct amdgpu_device *adev = ring->adev;
 	bool signaled = false;
 
 	BUG_ON(!ring);
@@ -391,8 +368,7 @@ static int amdgpu_fence_ring_wait_seq(struct amdgpu_ring *ring, uint64_t seq)
 		return 0;
 
 	wait_event(ring->fence_drv.fence_queue, (
-		   (signaled = amdgpu_fence_seq_signaled(ring, seq))
-		   || adev->needs_reset));
+		   (signaled = amdgpu_fence_seq_signaled(ring, seq))));
 
 	if (signaled)
 		return 0;
@@ -881,16 +857,12 @@ static void amdgpu_fence_wait_cb(struct fence *fence, struct fence_cb *cb)
 static signed long amdgpu_fence_default_wait(struct fence *f, bool intr,
 					     signed long t)
 {
-	struct amdgpu_fence *fence = to_amdgpu_fence(f);
-	struct amdgpu_device *adev = fence->ring->adev;
-
-	return amdgpu_fence_wait_any(adev, &f, 1, intr, t);
+	return amdgpu_fence_wait_any(&f, 1, intr, t);
 }
 
 /**
  * Wait the fence array with timeout
  *
- * @adev:     amdgpu device
  * @array:    the fence array with amdgpu fence pointer
  * @count:    the number of the fence array
  * @intr:     when sleep, set the current task interruptable or not
@@ -898,8 +870,7 @@ static signed long amdgpu_fence_default_wait(struct fence *f, bool intr,
  *
  * It will return when any fence is signaled or timeout.
  */
-signed long amdgpu_fence_wait_any(struct amdgpu_device *adev,
-				  struct fence **array, uint32_t count,
+signed long amdgpu_fence_wait_any(struct fence **array, uint32_t count,
 				  bool intr, signed long t)
 {
 	struct amdgpu_wait_cb *cb;
@@ -939,11 +910,6 @@ signed long amdgpu_fence_wait_any(struct amdgpu_device *adev,
 		if (amdgpu_test_signaled_any(array, count))
 			break;
 
-		if (adev->needs_reset) {
-			t = -EDEADLK;
-			break;
-		}
-
 		t = schedule_timeout(t);
 
 		if (t > 0 && intr && signal_pending(current))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 7297ca3a0ba7..087332858853 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -115,9 +115,10 @@ int amdgpu_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_pri
 	struct amdgpu_vm *vm = &fpriv->vm;
 	struct amdgpu_bo_va *bo_va;
 	int r;
-
+	mutex_lock(&vm->mutex);
 	r = amdgpu_bo_reserve(rbo, false);
 	if (r) {
+		mutex_unlock(&vm->mutex);
 		return r;
 	}
 
@@ -128,7 +129,7 @@ int amdgpu_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_pri
 		++bo_va->ref_count;
 	}
 	amdgpu_bo_unreserve(rbo);
-
+	mutex_unlock(&vm->mutex);
 	return 0;
 }
 
@@ -141,9 +142,10 @@ void amdgpu_gem_object_close(struct drm_gem_object *obj,
 	struct amdgpu_vm *vm = &fpriv->vm;
 	struct amdgpu_bo_va *bo_va;
 	int r;
-
+	mutex_lock(&vm->mutex);
 	r = amdgpu_bo_reserve(rbo, true);
 	if (r) {
+		mutex_unlock(&vm->mutex);
 		dev_err(adev->dev, "leaking bo va because "
 			"we fail to reserve bo (%d)\n", r);
 		return;
@@ -155,6 +157,7 @@ void amdgpu_gem_object_close(struct drm_gem_object *obj,
 		}
 	}
 	amdgpu_bo_unreserve(rbo);
+	mutex_unlock(&vm->mutex);
 }
 
 static int amdgpu_gem_handle_lockup(struct amdgpu_device *adev, int r)
@@ -181,7 +184,6 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 	bool kernel = false;
 	int r;
 
-	down_read(&adev->exclusive_lock);
 	/* create a gem object to contain this object in */
 	if (args->in.domains & (AMDGPU_GEM_DOMAIN_GDS |
 	    AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA)) {
@@ -214,11 +216,9 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 
 	memset(args, 0, sizeof(*args));
 	args->out.handle = handle;
-	up_read(&adev->exclusive_lock);
 	return 0;
 
 error_unlock:
-	up_read(&adev->exclusive_lock);
 	r = amdgpu_gem_handle_lockup(adev, r);
 	return r;
 }
@@ -250,8 +250,6 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
 		return -EACCES;
 	}
 
-	down_read(&adev->exclusive_lock);
-
 	/* create a gem object to contain this object in */
 	r = amdgpu_gem_object_create(adev, args->size, 0,
 				     AMDGPU_GEM_DOMAIN_CPU, 0,
@@ -293,14 +291,12 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
 		goto handle_lockup;
 
 	args->handle = handle;
-	up_read(&adev->exclusive_lock);
 	return 0;
 
 release_object:
 	drm_gem_object_unreference_unlocked(gobj);
 
 handle_lockup:
-	up_read(&adev->exclusive_lock);
 	r = amdgpu_gem_handle_lockup(adev, r);
 
 	return r;
@@ -488,18 +484,13 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
 			goto error_unreserve;
 	}
 
-	mutex_lock(&bo_va->vm->mutex);
 	r = amdgpu_vm_clear_freed(adev, bo_va->vm);
 	if (r)
-		goto error_unlock;
-
+		goto error_unreserve;
 
 	if (operation == AMDGPU_VA_OP_MAP)
 		r = amdgpu_vm_bo_update(adev, bo_va, &bo_va->bo->tbo.mem);
 
-error_unlock:
-	mutex_unlock(&bo_va->vm->mutex);
-
 error_unreserve:
 	ttm_eu_backoff_reservation(&ticket, &list);
 
@@ -556,10 +547,11 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 	gobj = drm_gem_object_lookup(dev, filp, args->handle);
 	if (gobj == NULL)
 		return -ENOENT;
-
+	mutex_lock(&fpriv->vm.mutex);
 	rbo = gem_to_amdgpu_bo(gobj);
 	r = amdgpu_bo_reserve(rbo, false);
 	if (r) {
+		mutex_unlock(&fpriv->vm.mutex);
 		drm_gem_object_unreference_unlocked(gobj);
 		return r;
 	}
@@ -567,6 +559,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 	bo_va = amdgpu_vm_bo_find(&fpriv->vm, rbo);
 	if (!bo_va) {
 		amdgpu_bo_unreserve(rbo);
+		mutex_unlock(&fpriv->vm.mutex);
 		return -ENOENT;
 	}
 
@@ -591,7 +584,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 
 	if (!r && !(args->flags & AMDGPU_VM_DELAY_UPDATE))
 		amdgpu_gem_va_update_vm(adev, bo_va, args->operation);
-
+	mutex_unlock(&fpriv->vm.mutex);
 	drm_gem_object_unreference_unlocked(gobj);
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index c439735ee670..aad4c1c69448 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -298,7 +298,6 @@ int amdgpu_ib_ring_tests(struct amdgpu_device *adev)
 		r = amdgpu_ring_test_ib(ring);
 		if (r) {
 			ring->ready = false;
-			adev->needs_reset = false;
 
 			if (ring == &adev->gfx.gfx_ring[0]) {
 				/* oh, oh, that's really bad */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index b13a74b273a6..b2df348aa223 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -67,8 +67,6 @@ void amdgpu_ring_free_size(struct amdgpu_ring *ring)
 	if (!ring->ring_free_dw) {
 		/* this is an empty ring */
 		ring->ring_free_dw = ring->ring_size / 4;
-		/*  update lockup info to avoid false positive */
-		amdgpu_ring_lockup_update(ring);
 	}
 }
 
@@ -209,46 +207,6 @@ void amdgpu_ring_unlock_undo(struct amdgpu_ring *ring)
 }
 
 /**
- * amdgpu_ring_lockup_update - update lockup variables
- *
- * @ring: amdgpu_ring structure holding ring information
- *
- * Update the last rptr value and timestamp (all asics).
- */
-void amdgpu_ring_lockup_update(struct amdgpu_ring *ring)
-{
-	atomic_set(&ring->last_rptr, amdgpu_ring_get_rptr(ring));
-	atomic64_set(&ring->last_activity, jiffies_64);
-}
-
-/**
- * amdgpu_ring_test_lockup() - check if ring is lockedup by recording information
- * @ring:       amdgpu_ring structure holding ring information
- *
- */
-bool amdgpu_ring_test_lockup(struct amdgpu_ring *ring)
-{
-	uint32_t rptr = amdgpu_ring_get_rptr(ring);
-	uint64_t last = atomic64_read(&ring->last_activity);
-	uint64_t elapsed;
-
-	if (rptr != atomic_read(&ring->last_rptr)) {
-		/* ring is still working, no lockup */
-		amdgpu_ring_lockup_update(ring);
-		return false;
-	}
-
-	elapsed = jiffies_to_msecs(jiffies_64 - last);
-	if (amdgpu_lockup_timeout && elapsed >= amdgpu_lockup_timeout) {
-		dev_err(ring->adev->dev, "ring %d stalled for more than %llumsec\n",
-			ring->idx, elapsed);
-		return true;
-	}
-	/* give a chance to the GPU ... */
-	return false;
-}
-
-/**
  * amdgpu_ring_backup - Back up the content of a ring
  *
  * @ring: the ring we want to back up
@@ -436,7 +394,6 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 	if (amdgpu_debugfs_ring_init(adev, ring)) {
 		DRM_ERROR("Failed to register debugfs file for rings !\n");
 	}
-	amdgpu_ring_lockup_update(ring);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c
index e90712443fe9..5cb27d525e43 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c
@@ -372,7 +372,7 @@ int amdgpu_sa_bo_new(struct amdgpu_device *adev,
 		} while (amdgpu_sa_bo_next_hole(sa_manager, fences, tries));
 
 		spin_unlock(&sa_manager->wq.lock);
-		t = amdgpu_fence_wait_any(adev, fences, AMDGPU_MAX_RINGS,
+		t = amdgpu_fence_wait_any(fences, AMDGPU_MAX_RINGS,
 					  false, MAX_SCHEDULE_TIMEOUT);
 		r = (t > 0) ? 0 : t;
 		spin_lock(&sa_manager->wq.lock);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index d0312364d950..53f987aeeacf 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -53,6 +53,7 @@
 #define FIRMWARE_TONGA		"amdgpu/tonga_uvd.bin"
 #define FIRMWARE_CARRIZO	"amdgpu/carrizo_uvd.bin"
 #define FIRMWARE_FIJI		"amdgpu/fiji_uvd.bin"
+#define FIRMWARE_STONEY		"amdgpu/stoney_uvd.bin"
 
 /**
  * amdgpu_uvd_cs_ctx - Command submission parser context
@@ -83,6 +84,7 @@ MODULE_FIRMWARE(FIRMWARE_MULLINS);
 MODULE_FIRMWARE(FIRMWARE_TONGA);
 MODULE_FIRMWARE(FIRMWARE_CARRIZO);
 MODULE_FIRMWARE(FIRMWARE_FIJI);
+MODULE_FIRMWARE(FIRMWARE_STONEY);
 
 static void amdgpu_uvd_note_usage(struct amdgpu_device *adev);
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
@@ -124,6 +126,9 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	case CHIP_CARRIZO:
 		fw_name = FIRMWARE_CARRIZO;
 		break;
+	case CHIP_STONEY:
+		fw_name = FIRMWARE_STONEY;
+		break;
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index 74f2038ac747..03f0c3bae516 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -49,6 +49,7 @@
 #define FIRMWARE_TONGA		"amdgpu/tonga_vce.bin"
 #define FIRMWARE_CARRIZO	"amdgpu/carrizo_vce.bin"
 #define FIRMWARE_FIJI		"amdgpu/fiji_vce.bin"
+#define FIRMWARE_STONEY		"amdgpu/stoney_vce.bin"
 
 #ifdef CONFIG_DRM_AMDGPU_CIK
 MODULE_FIRMWARE(FIRMWARE_BONAIRE);
@@ -60,6 +61,7 @@ MODULE_FIRMWARE(FIRMWARE_MULLINS);
 MODULE_FIRMWARE(FIRMWARE_TONGA);
 MODULE_FIRMWARE(FIRMWARE_CARRIZO);
 MODULE_FIRMWARE(FIRMWARE_FIJI);
+MODULE_FIRMWARE(FIRMWARE_STONEY);
 
 static void amdgpu_vce_idle_work_handler(struct work_struct *work);
 
@@ -106,6 +108,9 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	case CHIP_FIJI:
 		fw_name = FIRMWARE_FIJI;
 		break;
+	case CHIP_STONEY:
+		fw_name = FIRMWARE_STONEY;
+		break;
 
 	default:
 		return -EINVAL;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 644fd9b8591f..ff26e330ccd6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -90,11 +90,9 @@ struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev,
 	struct amdgpu_bo_list_entry *list;
 	unsigned i, idx;
 
-	mutex_lock(&vm->mutex);
 	list = drm_malloc_ab(vm->max_pde_used + 2,
 			     sizeof(struct amdgpu_bo_list_entry));
 	if (!list) {
-		mutex_unlock(&vm->mutex);
 		return NULL;
 	}
 
@@ -119,7 +117,6 @@ struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev,
 		list[idx].tv.shared = true;
 		list_add(&list[idx++].tv.head, head);
 	}
-	mutex_unlock(&vm->mutex);
 
 	return list;
 }
@@ -972,9 +969,7 @@ struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
 	INIT_LIST_HEAD(&bo_va->invalids);
 	INIT_LIST_HEAD(&bo_va->vm_status);
 
-	mutex_lock(&vm->mutex);
 	list_add_tail(&bo_va->bo_list, &bo->va);
-	mutex_unlock(&vm->mutex);
 
 	return bo_va;
 }
@@ -1027,8 +1022,6 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
 		return -EINVAL;
 	}
 
-	mutex_lock(&vm->mutex);
-
 	saddr /= AMDGPU_GPU_PAGE_SIZE;
 	eaddr /= AMDGPU_GPU_PAGE_SIZE;
 
@@ -1042,14 +1035,14 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
 			tmp->it.start, tmp->it.last + 1);
 		amdgpu_bo_unreserve(bo_va->bo);
 		r = -EINVAL;
-		goto error_unlock;
+		goto error;
 	}
 
 	mapping = kmalloc(sizeof(*mapping), GFP_KERNEL);
 	if (!mapping) {
 		amdgpu_bo_unreserve(bo_va->bo);
 		r = -ENOMEM;
-		goto error_unlock;
+		goto error;
 	}
 
 	INIT_LIST_HEAD(&mapping->list);
@@ -1081,9 +1074,6 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
 		if (vm->page_tables[pt_idx].bo)
 			continue;
 
-		/* drop mutex to allocate and clear page table */
-		mutex_unlock(&vm->mutex);
-
 		ww_mutex_lock(&resv->lock, NULL);
 		r = amdgpu_bo_create(adev, AMDGPU_VM_PTE_COUNT * 8,
 				     AMDGPU_GPU_PAGE_SIZE, true,
@@ -1100,32 +1090,19 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
 			goto error_free;
 		}
 
-		/* aquire mutex again */
-		mutex_lock(&vm->mutex);
-		if (vm->page_tables[pt_idx].bo) {
-			/* someone else allocated the pt in the meantime */
-			mutex_unlock(&vm->mutex);
-			amdgpu_bo_unref(&pt);
-			mutex_lock(&vm->mutex);
-			continue;
-		}
-
 		vm->page_tables[pt_idx].addr = 0;
 		vm->page_tables[pt_idx].bo = pt;
 	}
 
-	mutex_unlock(&vm->mutex);
 	return 0;
 
 error_free:
-	mutex_lock(&vm->mutex);
 	list_del(&mapping->list);
 	interval_tree_remove(&mapping->it, &vm->va);
 	trace_amdgpu_vm_bo_unmap(bo_va, mapping);
 	kfree(mapping);
 
-error_unlock:
-	mutex_unlock(&vm->mutex);
+error:
 	return r;
 }
 
@@ -1170,7 +1147,6 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
 		}
 	}
 
-	mutex_lock(&vm->mutex);
 	list_del(&mapping->list);
 	interval_tree_remove(&mapping->it, &vm->va);
 	trace_amdgpu_vm_bo_unmap(bo_va, mapping);
@@ -1179,7 +1155,6 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
 		list_add(&mapping->list, &vm->freed);
 	else
 		kfree(mapping);
-	mutex_unlock(&vm->mutex);
 	amdgpu_bo_unreserve(bo_va->bo);
 
 	return 0;
@@ -1203,8 +1178,6 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
 
 	list_del(&bo_va->bo_list);
 
-	mutex_lock(&vm->mutex);
-
 	spin_lock(&vm->status_lock);
 	list_del(&bo_va->vm_status);
 	spin_unlock(&vm->status_lock);
@@ -1223,8 +1196,6 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
 
 	fence_put(bo_va->last_pt_update);
 	kfree(bo_va);
-
-	mutex_unlock(&vm->mutex);
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
index 814598e76c98..5f712ceddf08 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
@@ -1290,24 +1290,6 @@ const struct amd_ip_funcs cik_sdma_ip_funcs = {
 	.set_powergating_state = cik_sdma_set_powergating_state,
 };
 
-/**
- * cik_sdma_ring_is_lockup - Check if the DMA engine is locked up
- *
- * @ring: amdgpu_ring structure holding ring information
- *
- * Check if the async DMA engine is locked up (CIK).
- * Returns true if the engine appears to be locked up, false if not.
- */
-static bool cik_sdma_ring_is_lockup(struct amdgpu_ring *ring)
-{
-
-	if (cik_sdma_is_idle(ring->adev)) {
-		amdgpu_ring_lockup_update(ring);
-		return false;
-	}
-	return amdgpu_ring_test_lockup(ring);
-}
-
 static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = {
 	.get_rptr = cik_sdma_ring_get_rptr,
 	.get_wptr = cik_sdma_ring_get_wptr,
@@ -1320,7 +1302,6 @@ static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = {
 	.emit_hdp_flush = cik_sdma_ring_emit_hdp_flush,
 	.test_ring = cik_sdma_ring_test_ring,
 	.test_ib = cik_sdma_ring_test_ib,
-	.is_lockup = cik_sdma_ring_is_lockup,
 	.insert_nop = cik_sdma_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/cz_dpm.c b/drivers/gpu/drm/amd/amdgpu/cz_dpm.c
index 2e3373ed4c94..8035d4d6a4f5 100644
--- a/drivers/gpu/drm/amd/amdgpu/cz_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/cz_dpm.c
@@ -1264,6 +1264,7 @@ static void cz_apply_state_adjust_rules(struct amdgpu_device *adev,
 
 static int cz_dpm_enable(struct amdgpu_device *adev)
 {
+	const char *chip_name;
 	int ret = 0;
 
 	/* renable will hang up SMU, so check first */
@@ -1272,21 +1273,33 @@ static int cz_dpm_enable(struct amdgpu_device *adev)
 
 	cz_program_voting_clients(adev);
 
+	switch (adev->asic_type) {
+	case CHIP_CARRIZO:
+		chip_name = "carrizo";
+		break;
+	case CHIP_STONEY:
+		chip_name = "stoney";
+		break;
+	default:
+		BUG();
+	}
+
+
 	ret = cz_start_dpm(adev);
 	if (ret) {
-		DRM_ERROR("Carrizo DPM enable failed\n");
+		DRM_ERROR("%s DPM enable failed\n", chip_name);
 		return -EINVAL;
 	}
 
 	ret = cz_program_bootup_state(adev);
 	if (ret) {
-		DRM_ERROR("Carrizo bootup state program failed\n");
+		DRM_ERROR("%s bootup state program failed\n", chip_name);
 		return -EINVAL;
 	}
 
 	ret = cz_enable_didt(adev, true);
 	if (ret) {
-		DRM_ERROR("Carrizo enable di/dt failed\n");
+		DRM_ERROR("%s enable di/dt failed\n", chip_name);
 		return -EINVAL;
 	}
 
@@ -1353,7 +1366,7 @@ static int cz_dpm_disable(struct amdgpu_device *adev)
 
 	ret = cz_enable_didt(adev, false);
 	if (ret) {
-		DRM_ERROR("Carrizo disable di/dt failed\n");
+		DRM_ERROR("disable di/dt failed\n");
 		return -EINVAL;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/cz_smc.c b/drivers/gpu/drm/amd/amdgpu/cz_smc.c
index e33180d3314a..ac7fee7b7eca 100644
--- a/drivers/gpu/drm/amd/amdgpu/cz_smc.c
+++ b/drivers/gpu/drm/amd/amdgpu/cz_smc.c
@@ -312,13 +312,16 @@ int cz_smu_start(struct amdgpu_device *adev)
 				UCODE_ID_CP_MEC_JT1_MASK |
 				UCODE_ID_CP_MEC_JT2_MASK;
 
+	if (adev->asic_type == CHIP_STONEY)
+		fw_to_check &= ~(UCODE_ID_SDMA1_MASK | UCODE_ID_CP_MEC_JT2_MASK);
+
 	cz_smu_request_load_fw(adev);
 	ret = cz_smu_check_fw_load_finish(adev, fw_to_check);
 	if (ret)
 		return ret;
 
 	/* manually load MEC firmware for CZ */
-	if (adev->asic_type == CHIP_CARRIZO) {
+	if (adev->asic_type == CHIP_CARRIZO || adev->asic_type == CHIP_STONEY) {
 		ret = cz_load_mec_firmware(adev);
 		if (ret) {
 			dev_err(adev->dev, "(%d) Mec Firmware load failed\n", ret);
@@ -336,6 +339,9 @@ int cz_smu_start(struct amdgpu_device *adev)
 				AMDGPU_CPMEC2_UCODE_LOADED |
 				AMDGPU_CPRLC_UCODE_LOADED;
 
+	if (adev->asic_type == CHIP_STONEY)
+		adev->smu.fw_flags &= ~(AMDGPU_SDMA1_UCODE_LOADED | AMDGPU_CPMEC2_UCODE_LOADED);
+
 	return ret;
 }
 
@@ -601,8 +607,13 @@ static int cz_smu_construct_toc_for_vddgfx_exit(struct amdgpu_device *adev)
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_ME, false);
 		cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false);
-		cz_smu_populate_single_ucode_load_task(adev,
+		if (adev->asic_type == CHIP_STONEY) {
+			cz_smu_populate_single_ucode_load_task(adev,
+				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false);
+		} else {
+			cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT2, false);
+		}
 		cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_RLC_G, false);
 	}
@@ -642,8 +653,13 @@ static int cz_smu_construct_toc_for_bootup(struct amdgpu_device *adev)
 	if (adev->firmware.smu_load) {
 		cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0, false);
-		cz_smu_populate_single_ucode_load_task(adev,
+		if (adev->asic_type == CHIP_STONEY) {
+			cz_smu_populate_single_ucode_load_task(adev,
+				CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0, false);
+		} else {
+			cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_SDMA1, false);
+		}
 		cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_CE, false);
 		cz_smu_populate_single_ucode_load_task(adev,
@@ -652,8 +668,13 @@ static int cz_smu_construct_toc_for_bootup(struct amdgpu_device *adev)
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_ME, false);
 		cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false);
-		cz_smu_populate_single_ucode_load_task(adev,
+		if (adev->asic_type == CHIP_STONEY) {
+			cz_smu_populate_single_ucode_load_task(adev,
+				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false);
+		} else {
+			cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT2, false);
+		}
 		cz_smu_populate_single_ucode_load_task(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_RLC_G, true);
 	}
@@ -888,10 +909,18 @@ int cz_smu_init(struct amdgpu_device *adev)
 				CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0,
 				&priv->driver_buffer[priv->driver_buffer_length++]))
 			goto smu_init_failed;
-		if (cz_smu_populate_single_firmware_entry(adev,
-				CZ_SCRATCH_ENTRY_UCODE_ID_SDMA1,
-				&priv->driver_buffer[priv->driver_buffer_length++]))
-			goto smu_init_failed;
+
+		if (adev->asic_type == CHIP_STONEY) {
+			if (cz_smu_populate_single_firmware_entry(adev,
+					CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0,
+					&priv->driver_buffer[priv->driver_buffer_length++]))
+				goto smu_init_failed;
+		} else {
+			if (cz_smu_populate_single_firmware_entry(adev,
+					CZ_SCRATCH_ENTRY_UCODE_ID_SDMA1,
+					&priv->driver_buffer[priv->driver_buffer_length++]))
+				goto smu_init_failed;
+		}
 		if (cz_smu_populate_single_firmware_entry(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_CE,
 				&priv->driver_buffer[priv->driver_buffer_length++]))
@@ -908,10 +937,17 @@ int cz_smu_init(struct amdgpu_device *adev)
 				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1,
 				&priv->driver_buffer[priv->driver_buffer_length++]))
 			goto smu_init_failed;
-		if (cz_smu_populate_single_firmware_entry(adev,
-				CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT2,
-				&priv->driver_buffer[priv->driver_buffer_length++]))
-			goto smu_init_failed;
+		if (adev->asic_type == CHIP_STONEY) {
+			if (cz_smu_populate_single_firmware_entry(adev,
+					CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1,
+					&priv->driver_buffer[priv->driver_buffer_length++]))
+				goto smu_init_failed;
+		} else {
+			if (cz_smu_populate_single_firmware_entry(adev,
+					CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT2,
+					&priv->driver_buffer[priv->driver_buffer_length++]))
+				goto smu_init_failed;
+		}
 		if (cz_smu_populate_single_firmware_entry(adev,
 				CZ_SCRATCH_ENTRY_UCODE_ID_RLC_G,
 				&priv->driver_buffer[priv->driver_buffer_length++]))
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 37073930e2c9..a6ea2d8e85df 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -3086,22 +3086,18 @@ static int dce_v10_0_suspend(void *handle)
 
 	amdgpu_atombios_scratch_regs_save(adev);
 
-	dce_v10_0_hpd_fini(adev);
-
-	return 0;
+	return dce_v10_0_hw_fini(handle);
 }
 
 static int dce_v10_0_resume(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int ret;
 
-	dce_v10_0_init_golden_registers(adev);
+	ret = dce_v10_0_hw_init(handle);
 
 	amdgpu_atombios_scratch_regs_restore(adev);
 
-	/* init dig PHYs, disp eng pll */
-	amdgpu_atombios_encoder_init_dig(adev);
-	amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk);
 	/* turn on the BL */
 	if (adev->mode_info.bl_encoder) {
 		u8 bl_level = amdgpu_display_backlight_get_level(adev,
@@ -3110,10 +3106,7 @@ static int dce_v10_0_resume(void *handle)
 						    bl_level);
 	}
 
-	/* initialize hpd */
-	dce_v10_0_hpd_init(adev);
-
-	return 0;
+	return ret;
 }
 
 static bool dce_v10_0_is_idle(void *handle)
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index 5be83a3f0192..784afb5978ac 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -126,6 +126,13 @@ static const u32 cz_mgcg_cgcg_init[] =
 	mmXDMA_MEM_POWER_CNTL, 0x00000101, 0x00000000,
 };
 
+static const u32 stoney_golden_settings_a11[] =
+{
+	mmCRTC_DOUBLE_BUFFER_CONTROL, 0x00010101, 0x00010000,
+	mmFBC_MISC, 0x1f311fff, 0x14302000,
+};
+
+
 static void dce_v11_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
@@ -137,6 +144,11 @@ static void dce_v11_0_init_golden_registers(struct amdgpu_device *adev)
 						 cz_golden_settings_a11,
 						 (const u32)ARRAY_SIZE(cz_golden_settings_a11));
 		break;
+	case CHIP_STONEY:
+		amdgpu_program_register_sequence(adev,
+						 stoney_golden_settings_a11,
+						 (const u32)ARRAY_SIZE(stoney_golden_settings_a11));
+		break;
 	default:
 		break;
 	}
@@ -2425,7 +2437,7 @@ static u32 dce_v11_0_pick_pll(struct drm_crtc *crtc)
 
 	/* XXX need to determine what plls are available on each DCE11 part */
 	pll_in_use = amdgpu_pll_get_use_mask(crtc);
-	if (adev->asic_type == CHIP_CARRIZO) {
+	if (adev->asic_type == CHIP_CARRIZO || adev->asic_type == CHIP_STONEY) {
 		if (!(pll_in_use & (1 << ATOM_PPLL1)))
 			return ATOM_PPLL1;
 		if (!(pll_in_use & (1 << ATOM_PPLL0)))
@@ -2930,6 +2942,11 @@ static int dce_v11_0_early_init(void *handle)
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 9;
 		break;
+	case CHIP_STONEY:
+		adev->mode_info.num_crtc = 2;
+		adev->mode_info.num_hpd = 6;
+		adev->mode_info.num_dig = 9;
+		break;
 	default:
 		/* FIXME: not supported yet */
 		return -EINVAL;
@@ -3028,6 +3045,7 @@ static int dce_v11_0_hw_init(void *handle)
 	dce_v11_0_init_golden_registers(adev);
 
 	/* init dig PHYs, disp eng pll */
+	amdgpu_atombios_crtc_powergate_init(adev);
 	amdgpu_atombios_encoder_init_dig(adev);
 	amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk);
 
@@ -3061,23 +3079,18 @@ static int dce_v11_0_suspend(void *handle)
 
 	amdgpu_atombios_scratch_regs_save(adev);
 
-	dce_v11_0_hpd_fini(adev);
-
-	return 0;
+	return dce_v11_0_hw_fini(handle);
 }
 
 static int dce_v11_0_resume(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int ret;
 
-	dce_v11_0_init_golden_registers(adev);
+	ret = dce_v11_0_hw_init(handle);
 
 	amdgpu_atombios_scratch_regs_restore(adev);
 
-	/* init dig PHYs, disp eng pll */
-	amdgpu_atombios_crtc_powergate_init(adev);
-	amdgpu_atombios_encoder_init_dig(adev);
-	amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk);
 	/* turn on the BL */
 	if (adev->mode_info.bl_encoder) {
 		u8 bl_level = amdgpu_display_backlight_get_level(adev,
@@ -3086,10 +3099,7 @@ static int dce_v11_0_resume(void *handle)
 						    bl_level);
 	}
 
-	/* initialize hpd */
-	dce_v11_0_hpd_init(adev);
-
-	return 0;
+	return ret;
 }
 
 static bool dce_v11_0_is_idle(void *handle)
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index d784fb43efc2..00c34f87ac20 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -2994,20 +2994,18 @@ static int dce_v8_0_suspend(void *handle)
 
 	amdgpu_atombios_scratch_regs_save(adev);
 
-	dce_v8_0_hpd_fini(adev);
-
-	return 0;
+	return dce_v8_0_hw_fini(handle);
 }
 
 static int dce_v8_0_resume(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int ret;
+
+	ret = dce_v8_0_hw_init(handle);
 
 	amdgpu_atombios_scratch_regs_restore(adev);
 
-	/* init dig PHYs, disp eng pll */
-	amdgpu_atombios_encoder_init_dig(adev);
-	amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk);
 	/* turn on the BL */
 	if (adev->mode_info.bl_encoder) {
 		u8 bl_level = amdgpu_display_backlight_get_level(adev,
@@ -3016,10 +3014,7 @@ static int dce_v8_0_resume(void *handle)
 						    bl_level);
 	}
 
-	/* initialize hpd */
-	dce_v8_0_hpd_init(adev);
-
-	return 0;
+	return ret;
 }
 
 static bool dce_v8_0_is_idle(void *handle)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
index e992bf2ff66c..72793f93e2fc 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -5542,24 +5542,6 @@ const struct amd_ip_funcs gfx_v7_0_ip_funcs = {
 	.set_powergating_state = gfx_v7_0_set_powergating_state,
 };
 
-/**
- * gfx_v7_0_ring_is_lockup - check if the 3D engine is locked up
- *
- * @adev: amdgpu_device pointer
- * @ring: amdgpu_ring structure holding ring information
- *
- * Check if the 3D engine is locked up (CIK).
- * Returns true if the engine is locked, false if not.
- */
-static bool gfx_v7_0_ring_is_lockup(struct amdgpu_ring *ring)
-{
-	if (gfx_v7_0_is_idle(ring->adev)) {
-		amdgpu_ring_lockup_update(ring);
-		return false;
-	}
-	return amdgpu_ring_test_lockup(ring);
-}
-
 static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = {
 	.get_rptr = gfx_v7_0_ring_get_rptr_gfx,
 	.get_wptr = gfx_v7_0_ring_get_wptr_gfx,
@@ -5573,7 +5555,6 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = {
 	.emit_hdp_flush = gfx_v7_0_ring_emit_hdp_flush,
 	.test_ring = gfx_v7_0_ring_test_ring,
 	.test_ib = gfx_v7_0_ring_test_ib,
-	.is_lockup = gfx_v7_0_ring_is_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
@@ -5590,7 +5571,6 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = {
 	.emit_hdp_flush = gfx_v7_0_ring_emit_hdp_flush,
 	.test_ring = gfx_v7_0_ring_test_ring,
 	.test_ib = gfx_v7_0_ring_test_ib,
-	.is_lockup = gfx_v7_0_ring_is_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 718250ae9856..cbc46a34987c 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -73,6 +73,12 @@ MODULE_FIRMWARE("amdgpu/carrizo_mec.bin");
 MODULE_FIRMWARE("amdgpu/carrizo_mec2.bin");
 MODULE_FIRMWARE("amdgpu/carrizo_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/stoney_ce.bin");
+MODULE_FIRMWARE("amdgpu/stoney_pfp.bin");
+MODULE_FIRMWARE("amdgpu/stoney_me.bin");
+MODULE_FIRMWARE("amdgpu/stoney_mec.bin");
+MODULE_FIRMWARE("amdgpu/stoney_rlc.bin");
+
 MODULE_FIRMWARE("amdgpu/tonga_ce.bin");
 MODULE_FIRMWARE("amdgpu/tonga_pfp.bin");
 MODULE_FIRMWARE("amdgpu/tonga_me.bin");
@@ -493,6 +499,42 @@ static const u32 cz_mgcg_cgcg_init[] =
 	mmCP_MEM_SLP_CNTL, 0x00000001, 0x00000001,
 };
 
+static const u32 stoney_golden_settings_a11[] =
+{
+	mmDB_DEBUG2, 0xf00fffff, 0x00000400,
+	mmGB_GPU_ID, 0x0000000f, 0x00000000,
+	mmPA_SC_ENHANCE, 0xffffffff, 0x20000001,
+	mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000,
+	mmRLC_CGCG_CGLS_CTRL, 0x00000003, 0x0001003c,
+	mmTA_CNTL_AUX, 0x000f000f, 0x000b0000,
+  	mmTCC_CTRL, 0x00100000, 0xf31fff7f,
+	mmTCC_EXE_DISABLE, 0x00000002, 0x00000002,
+	mmTCP_ADDR_CONFIG, 0x0000000f, 0x000000f1,
+	mmTCP_CHAN_STEER_LO, 0xffffffff, 0x10101010,
+};
+
+static const u32 stoney_golden_common_all[] =
+{
+	mmGRBM_GFX_INDEX, 0xffffffff, 0xe0000000,
+	mmPA_SC_RASTER_CONFIG, 0xffffffff, 0x00000000,
+	mmPA_SC_RASTER_CONFIG_1, 0xffffffff, 0x00000000,
+	mmGB_ADDR_CONFIG, 0xffffffff, 0x12010001,
+	mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800,
+	mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800,
+	mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF,
+	mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF,
+};
+
+static const u32 stoney_mgcg_cgcg_init[] =
+{
+	mmGRBM_GFX_INDEX, 0xffffffff, 0xe0000000,
+	mmRLC_CGCG_CGLS_CTRL, 0xffffffff, 0x0020003f,
+	mmCP_MEM_SLP_CNTL, 0xffffffff, 0x00020201,
+	mmRLC_MEM_SLP_CNTL, 0xffffffff, 0x00020201,
+	mmCGTS_SM_CTRL_REG, 0xffffffff, 0x96940200,
+	mmATC_MISC_CG, 0xffffffff, 0x000c0200,
+};
+
 static void gfx_v8_0_set_ring_funcs(struct amdgpu_device *adev);
 static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev);
 static void gfx_v8_0_set_gds_init(struct amdgpu_device *adev);
@@ -545,6 +587,17 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev)
 						 cz_golden_common_all,
 						 (const u32)ARRAY_SIZE(cz_golden_common_all));
 		break;
+	case CHIP_STONEY:
+		amdgpu_program_register_sequence(adev,
+						 stoney_mgcg_cgcg_init,
+						 (const u32)ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		amdgpu_program_register_sequence(adev,
+						 stoney_golden_settings_a11,
+						 (const u32)ARRAY_SIZE(stoney_golden_settings_a11));
+		amdgpu_program_register_sequence(adev,
+						 stoney_golden_common_all,
+						 (const u32)ARRAY_SIZE(stoney_golden_common_all));
+		break;
 	default:
 		break;
 	}
@@ -691,6 +744,9 @@ static int gfx_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 		chip_name = "fiji";
 		break;
+	case CHIP_STONEY:
+		chip_name = "stoney";
+		break;
 	default:
 		BUG();
 	}
@@ -748,21 +804,23 @@ static int gfx_v8_0_init_microcode(struct amdgpu_device *adev)
 	adev->gfx.mec_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
 	adev->gfx.mec_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
 
-	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec2.bin", chip_name);
-	err = request_firmware(&adev->gfx.mec2_fw, fw_name, adev->dev);
-	if (!err) {
-		err = amdgpu_ucode_validate(adev->gfx.mec2_fw);
-		if (err)
-			goto out;
-		cp_hdr = (const struct gfx_firmware_header_v1_0 *)
-						adev->gfx.mec2_fw->data;
-		adev->gfx.mec2_fw_version = le32_to_cpu(
-						cp_hdr->header.ucode_version);
-		adev->gfx.mec2_feature_version = le32_to_cpu(
-						cp_hdr->ucode_feature_version);
-	} else {
-		err = 0;
-		adev->gfx.mec2_fw = NULL;
+	if (adev->asic_type != CHIP_STONEY) {
+		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec2.bin", chip_name);
+		err = request_firmware(&adev->gfx.mec2_fw, fw_name, adev->dev);
+		if (!err) {
+			err = amdgpu_ucode_validate(adev->gfx.mec2_fw);
+			if (err)
+				goto out;
+			cp_hdr = (const struct gfx_firmware_header_v1_0 *)
+				adev->gfx.mec2_fw->data;
+			adev->gfx.mec2_fw_version =
+				le32_to_cpu(cp_hdr->header.ucode_version);
+			adev->gfx.mec2_feature_version =
+				le32_to_cpu(cp_hdr->ucode_feature_version);
+		} else {
+			err = 0;
+			adev->gfx.mec2_fw = NULL;
+		}
 	}
 
 	if (adev->firmware.smu_load) {
@@ -1010,6 +1068,40 @@ static void gfx_v8_0_gpu_early_init(struct amdgpu_device *adev)
 		adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
 		gb_addr_config = CARRIZO_GB_ADDR_CONFIG_GOLDEN;
 		break;
+	case CHIP_STONEY:
+		adev->gfx.config.max_shader_engines = 1;
+		adev->gfx.config.max_tile_pipes = 2;
+		adev->gfx.config.max_sh_per_se = 1;
+		adev->gfx.config.max_backends_per_se = 1;
+
+		switch (adev->pdev->revision) {
+		case 0xc0:
+		case 0xc1:
+		case 0xc2:
+		case 0xc4:
+		case 0xc8:
+		case 0xc9:
+			adev->gfx.config.max_cu_per_sh = 3;
+			break;
+		case 0xd0:
+		case 0xd1:
+		case 0xd2:
+		default:
+			adev->gfx.config.max_cu_per_sh = 2;
+			break;
+		}
+
+		adev->gfx.config.max_texture_channel_caches = 2;
+		adev->gfx.config.max_gprs = 256;
+		adev->gfx.config.max_gs_threads = 16;
+		adev->gfx.config.max_hw_contexts = 8;
+
+		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+		adev->gfx.config.sc_prim_fifo_size_backend = 0x100;
+		adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+		adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
+		gb_addr_config = CARRIZO_GB_ADDR_CONFIG_GOLDEN;
+		break;
 	default:
 		adev->gfx.config.max_shader_engines = 2;
 		adev->gfx.config.max_tile_pipes = 4;
@@ -1797,6 +1889,273 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev)
 			WREG32(mmGB_MACROTILE_MODE0 + reg_offset, gb_tile_moden);
 		}
 		break;
+	case CHIP_STONEY:
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
+			switch (reg_offset) {
+			case 0:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 1:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 2:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 3:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 4:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 5:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 6:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+				break;
+			case 8:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+						PIPE_CONFIG(ADDR_SURF_P2));
+				break;
+			case 9:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 10:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 11:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
+				break;
+			case 13:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 14:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 15:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 16:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
+				break;
+			case 18:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 19:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 20:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 21:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_THICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 22:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 24:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 25:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_XTHICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 26:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_XTHICK) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
+				break;
+			case 27:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 28:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+				break;
+			case 29:
+				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+						PIPE_CONFIG(ADDR_SURF_P2) |
+						MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+						SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
+				break;
+			case 7:
+			case 12:
+			case 17:
+			case 23:
+				/* unused idx */
+				continue;
+			default:
+				gb_tile_moden = 0;
+				break;
+			};
+			adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden;
+			WREG32(mmGB_TILE_MODE0 + reg_offset, gb_tile_moden);
+		}
+		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) {
+			switch (reg_offset) {
+			case 0:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 1:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 2:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 3:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 4:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 5:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 6:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 8:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_4) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_16_BANK));
+				break;
+			case 9:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_4) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_16_BANK));
+				break;
+			case 10:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_16_BANK));
+				break;
+			case 11:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_16_BANK));
+				break;
+			case 12:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_16_BANK));
+				break;
+			case 13:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+						NUM_BANKS(ADDR_SURF_16_BANK));
+				break;
+			case 14:
+				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+						BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+						MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+						NUM_BANKS(ADDR_SURF_8_BANK));
+				break;
+			case 7:
+				/* unused idx */
+				continue;
+			default:
+				gb_tile_moden = 0;
+				break;
+			};
+			adev->gfx.config.macrotile_mode_array[reg_offset] = gb_tile_moden;
+			WREG32(mmGB_MACROTILE_MODE0 + reg_offset, gb_tile_moden);
+		}
+		break;
 	case CHIP_CARRIZO:
 	default:
 		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
@@ -2384,7 +2743,7 @@ static void gfx_v8_0_rlc_start(struct amdgpu_device *adev)
 	WREG32(mmRLC_CNTL, tmp);
 
 	/* carrizo do enable cp interrupt after cp inited */
-	if (adev->asic_type != CHIP_CARRIZO)
+	if (!(adev->flags & AMD_IS_APU))
 		gfx_v8_0_enable_gui_idle_interrupt(adev, true);
 
 	udelay(50);
@@ -2606,6 +2965,10 @@ static int gfx_v8_0_cp_gfx_start(struct amdgpu_device *adev)
 		amdgpu_ring_write(ring, 0x00000002);
 		amdgpu_ring_write(ring, 0x00000000);
 		break;
+	case CHIP_STONEY:
+		amdgpu_ring_write(ring, 0x00000000);
+		amdgpu_ring_write(ring, 0x00000000);
+		break;
 	default:
 		BUG();
 	}
@@ -3240,7 +3603,8 @@ static int gfx_v8_0_cp_compute_resume(struct amdgpu_device *adev)
 		/* enable the doorbell if requested */
 		if (use_doorbell) {
 			if ((adev->asic_type == CHIP_CARRIZO) ||
-			    (adev->asic_type == CHIP_FIJI)) {
+			    (adev->asic_type == CHIP_FIJI) ||
+			    (adev->asic_type == CHIP_STONEY)) {
 				WREG32(mmCP_MEC_DOORBELL_RANGE_LOWER,
 				       AMDGPU_DOORBELL_KIQ << 2);
 				WREG32(mmCP_MEC_DOORBELL_RANGE_UPPER,
@@ -3312,7 +3676,7 @@ static int gfx_v8_0_cp_resume(struct amdgpu_device *adev)
 {
 	int r;
 
-	if (adev->asic_type != CHIP_CARRIZO)
+	if (!(adev->flags & AMD_IS_APU))
 		gfx_v8_0_enable_gui_idle_interrupt(adev, false);
 
 	if (!adev->firmware.smu_load) {
@@ -4075,15 +4439,6 @@ static void gfx_v8_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 	}
 }
 
-static bool gfx_v8_0_ring_is_lockup(struct amdgpu_ring *ring)
-{
-	if (gfx_v8_0_is_idle(ring->adev)) {
-		amdgpu_ring_lockup_update(ring);
-		return false;
-	}
-	return amdgpu_ring_test_lockup(ring);
-}
-
 static u32 gfx_v8_0_ring_get_rptr_compute(struct amdgpu_ring *ring)
 {
 	return ring->adev->wb.wb[ring->rptr_offs];
@@ -4114,6 +4469,7 @@ static void gfx_v8_0_ring_emit_fence_compute(struct amdgpu_ring *ring,
 	amdgpu_ring_write(ring, PACKET3(PACKET3_RELEASE_MEM, 5));
 	amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN |
 				 EOP_TC_ACTION_EN |
+				 EOP_TC_WB_ACTION_EN |
 				 EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) |
 				 EVENT_INDEX(5)));
 	amdgpu_ring_write(ring, DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0));
@@ -4364,7 +4720,6 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = {
 	.emit_hdp_flush = gfx_v8_0_ring_emit_hdp_flush,
 	.test_ring = gfx_v8_0_ring_test_ring,
 	.test_ib = gfx_v8_0_ring_test_ib,
-	.is_lockup = gfx_v8_0_ring_is_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
@@ -4381,7 +4736,6 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = {
 	.emit_hdp_flush = gfx_v8_0_ring_emit_hdp_flush,
 	.test_ring = gfx_v8_0_ring_test_ring,
 	.test_ib = gfx_v8_0_ring_test_ib,
-	.is_lockup = gfx_v8_0_ring_is_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index 42b5ff827055..72e977b1685d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -93,6 +93,12 @@ static const u32 cz_mgcg_cgcg_init[] =
 	mmMC_MEM_POWER_LS, 0xffffffff, 0x00000104
 };
 
+static const u32 stoney_mgcg_cgcg_init[] =
+{
+	mmMC_MEM_POWER_LS, 0xffffffff, 0x00000104
+};
+
+
 static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
@@ -125,6 +131,11 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev)
 						 cz_mgcg_cgcg_init,
 						 (const u32)ARRAY_SIZE(cz_mgcg_cgcg_init));
 		break;
+	case CHIP_STONEY:
+		amdgpu_program_register_sequence(adev,
+						 stoney_mgcg_cgcg_init,
+						 (const u32)ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		break;
 	default:
 		break;
 	}
@@ -228,6 +239,7 @@ static int gmc_v8_0_init_microcode(struct amdgpu_device *adev)
 		chip_name = "fiji";
 		break;
 	case CHIP_CARRIZO:
+	case CHIP_STONEY:
 		return 0;
 	default: BUG();
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
index f8b868c7c496..2cf50180cc51 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
@@ -1295,24 +1295,6 @@ const struct amd_ip_funcs sdma_v2_4_ip_funcs = {
 	.set_powergating_state = sdma_v2_4_set_powergating_state,
 };
 
-/**
- * sdma_v2_4_ring_is_lockup - Check if the DMA engine is locked up
- *
- * @ring: amdgpu_ring structure holding ring information
- *
- * Check if the async DMA engine is locked up (VI).
- * Returns true if the engine appears to be locked up, false if not.
- */
-static bool sdma_v2_4_ring_is_lockup(struct amdgpu_ring *ring)
-{
-
-	if (sdma_v2_4_is_idle(ring->adev)) {
-		amdgpu_ring_lockup_update(ring);
-		return false;
-	}
-	return amdgpu_ring_test_lockup(ring);
-}
-
 static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = {
 	.get_rptr = sdma_v2_4_ring_get_rptr,
 	.get_wptr = sdma_v2_4_ring_get_wptr,
@@ -1325,7 +1307,6 @@ static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = {
 	.emit_hdp_flush = sdma_v2_4_ring_emit_hdp_flush,
 	.test_ring = sdma_v2_4_ring_test_ring,
 	.test_ib = sdma_v2_4_ring_test_ib,
-	.is_lockup = sdma_v2_4_ring_is_lockup,
 	.insert_nop = sdma_v2_4_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
index 670555a45da9..7253132f04b8 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -55,6 +55,7 @@ MODULE_FIRMWARE("amdgpu/carrizo_sdma.bin");
 MODULE_FIRMWARE("amdgpu/carrizo_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/fiji_sdma.bin");
 MODULE_FIRMWARE("amdgpu/fiji_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/stoney_sdma.bin");
 
 static const u32 sdma_offsets[SDMA_MAX_INSTANCE] =
 {
@@ -122,6 +123,19 @@ static const u32 cz_mgcg_cgcg_init[] =
 	mmSDMA1_CLK_CTRL, 0xff000ff0, 0x00000100
 };
 
+static const u32 stoney_golden_settings_a11[] =
+{
+	mmSDMA0_GFX_IB_CNTL, 0x00000100, 0x00000100,
+	mmSDMA0_POWER_CNTL, 0x00000800, 0x0003c800,
+	mmSDMA0_RLC0_IB_CNTL, 0x00000100, 0x00000100,
+	mmSDMA0_RLC1_IB_CNTL, 0x00000100, 0x00000100,
+};
+
+static const u32 stoney_mgcg_cgcg_init[] =
+{
+	mmSDMA0_CLK_CTRL, 0xffffffff, 0x00000100,
+};
+
 /*
  * sDMA - System DMA
  * Starting with CIK, the GPU has new asynchronous
@@ -166,6 +180,14 @@ static void sdma_v3_0_init_golden_registers(struct amdgpu_device *adev)
 						 cz_golden_settings_a11,
 						 (const u32)ARRAY_SIZE(cz_golden_settings_a11));
 		break;
+	case CHIP_STONEY:
+		amdgpu_program_register_sequence(adev,
+						 stoney_mgcg_cgcg_init,
+						 (const u32)ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		amdgpu_program_register_sequence(adev,
+						 stoney_golden_settings_a11,
+						 (const u32)ARRAY_SIZE(stoney_golden_settings_a11));
+		break;
 	default:
 		break;
 	}
@@ -201,6 +223,9 @@ static int sdma_v3_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_CARRIZO:
 		chip_name = "carrizo";
 		break;
+	case CHIP_STONEY:
+		chip_name = "stoney";
+		break;
 	default: BUG();
 	}
 
@@ -1071,6 +1096,9 @@ static int sdma_v3_0_early_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	switch (adev->asic_type) {
+	case CHIP_STONEY:
+		adev->sdma.num_instances = 1;
+		break;
 	default:
 		adev->sdma.num_instances = SDMA_MAX_INSTANCE;
 		break;
@@ -1428,24 +1456,6 @@ const struct amd_ip_funcs sdma_v3_0_ip_funcs = {
 	.set_powergating_state = sdma_v3_0_set_powergating_state,
 };
 
-/**
- * sdma_v3_0_ring_is_lockup - Check if the DMA engine is locked up
- *
- * @ring: amdgpu_ring structure holding ring information
- *
- * Check if the async DMA engine is locked up (VI).
- * Returns true if the engine appears to be locked up, false if not.
- */
-static bool sdma_v3_0_ring_is_lockup(struct amdgpu_ring *ring)
-{
-
-	if (sdma_v3_0_is_idle(ring->adev)) {
-		amdgpu_ring_lockup_update(ring);
-		return false;
-	}
-	return amdgpu_ring_test_lockup(ring);
-}
-
 static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = {
 	.get_rptr = sdma_v3_0_ring_get_rptr,
 	.get_wptr = sdma_v3_0_ring_get_wptr,
@@ -1458,7 +1468,6 @@ static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = {
 	.emit_hdp_flush = sdma_v3_0_ring_emit_hdp_flush,
 	.test_ring = sdma_v3_0_ring_test_ring,
 	.test_ib = sdma_v3_0_ring_test_ib,
-	.is_lockup = sdma_v3_0_ring_is_lockup,
 	.insert_nop = sdma_v3_0_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index ed50dd725788..5e9f73af83a8 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -885,7 +885,6 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 	.emit_semaphore = uvd_v4_2_ring_emit_semaphore,
 	.test_ring = uvd_v4_2_ring_test_ring,
 	.test_ib = uvd_v4_2_ring_test_ib,
-	.is_lockup = amdgpu_ring_test_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index 9ad8b9906c0b..38864f562981 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -824,7 +824,6 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 	.emit_semaphore = uvd_v5_0_ring_emit_semaphore,
 	.test_ring = uvd_v5_0_ring_test_ring,
 	.test_ib = uvd_v5_0_ring_test_ib,
-	.is_lockup = amdgpu_ring_test_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index 7e9934fa4193..121915bbc3b6 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -808,7 +808,6 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_funcs = {
 	.emit_semaphore = uvd_v6_0_ring_emit_semaphore,
 	.test_ring = uvd_v6_0_ring_test_ring,
 	.test_ib = uvd_v6_0_ring_test_ib,
-	.is_lockup = amdgpu_ring_test_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
index cd16df543f64..52ac7a8f1e58 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
@@ -642,7 +642,6 @@ static const struct amdgpu_ring_funcs vce_v2_0_ring_funcs = {
 	.emit_semaphore = amdgpu_vce_ring_emit_semaphore,
 	.test_ring = amdgpu_vce_ring_test_ring,
 	.test_ib = amdgpu_vce_ring_test_ib,
-	.is_lockup = amdgpu_ring_test_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index f0656dfb53f3..6a52db6ad8d7 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -205,8 +205,9 @@ static unsigned vce_v3_0_get_harvest_config(struct amdgpu_device *adev)
 	u32 tmp;
 	unsigned ret;
 
-	/* Fiji is single pipe */
-	if (adev->asic_type == CHIP_FIJI) {
+	/* Fiji, Stoney are single pipe */
+	if ((adev->asic_type == CHIP_FIJI) ||
+	    (adev->asic_type == CHIP_STONEY)){
 		ret = AMDGPU_VCE_HARVEST_VCE1;
 		return ret;
 	}
@@ -643,7 +644,6 @@ static const struct amdgpu_ring_funcs vce_v3_0_ring_funcs = {
 	.emit_semaphore = amdgpu_vce_ring_emit_semaphore,
 	.test_ring = amdgpu_vce_ring_test_ring,
 	.test_ib = amdgpu_vce_ring_test_ib,
-	.is_lockup = amdgpu_ring_test_lockup,
 	.insert_nop = amdgpu_ring_insert_nop,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 0bac8702e934..d8c93f14c3b5 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -232,6 +232,13 @@ static const u32 cz_mgcg_cgcg_init[] =
 	mmHDP_XDP_CGTT_BLK_CTRL, 0xc0000fff, 0x00000104,
 };
 
+static const u32 stoney_mgcg_cgcg_init[] =
+{
+	mmCGTT_DRM_CLK_CTRL0, 0xffffffff, 0x00000100,
+	mmHDP_XDP_CGTT_BLK_CTRL, 0xffffffff, 0x00000104,
+	mmHDP_HOST_PATH_CNTL, 0xffffffff, 0x0f000027,
+};
+
 static void vi_init_golden_registers(struct amdgpu_device *adev)
 {
 	/* Some of the registers might be dependent on GRBM_GFX_INDEX */
@@ -258,6 +265,11 @@ static void vi_init_golden_registers(struct amdgpu_device *adev)
 						 cz_mgcg_cgcg_init,
 						 (const u32)ARRAY_SIZE(cz_mgcg_cgcg_init));
 		break;
+	case CHIP_STONEY:
+		amdgpu_program_register_sequence(adev,
+						 stoney_mgcg_cgcg_init,
+						 (const u32)ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		break;
 	default:
 		break;
 	}
@@ -488,6 +500,7 @@ static int vi_read_register(struct amdgpu_device *adev, u32 se_num,
 	case CHIP_FIJI:
 	case CHIP_TONGA:
 	case CHIP_CARRIZO:
+	case CHIP_STONEY:
 		asic_register_table = cz_allowed_read_registers;
 		size = ARRAY_SIZE(cz_allowed_read_registers);
 		break;
@@ -543,8 +556,10 @@ static void vi_print_gpu_status_regs(struct amdgpu_device *adev)
 		RREG32(mmSRBM_STATUS2));
 	dev_info(adev->dev, "  SDMA0_STATUS_REG   = 0x%08X\n",
 		RREG32(mmSDMA0_STATUS_REG + SDMA0_REGISTER_OFFSET));
-	dev_info(adev->dev, "  SDMA1_STATUS_REG   = 0x%08X\n",
-		 RREG32(mmSDMA0_STATUS_REG + SDMA1_REGISTER_OFFSET));
+	if (adev->sdma.num_instances > 1) {
+		dev_info(adev->dev, "  SDMA1_STATUS_REG   = 0x%08X\n",
+			RREG32(mmSDMA0_STATUS_REG + SDMA1_REGISTER_OFFSET));
+	}
 	dev_info(adev->dev, "  CP_STAT = 0x%08x\n", RREG32(mmCP_STAT));
 	dev_info(adev->dev, "  CP_STALLED_STAT1 = 0x%08x\n",
 		 RREG32(mmCP_STALLED_STAT1));
@@ -639,9 +654,11 @@ u32 vi_gpu_check_soft_reset(struct amdgpu_device *adev)
 		reset_mask |= AMDGPU_RESET_DMA;
 
 	/* SDMA1_STATUS_REG */
-	tmp = RREG32(mmSDMA0_STATUS_REG + SDMA1_REGISTER_OFFSET);
-	if (!(tmp & SDMA0_STATUS_REG__IDLE_MASK))
-		reset_mask |= AMDGPU_RESET_DMA1;
+	if (adev->sdma.num_instances > 1) {
+		tmp = RREG32(mmSDMA0_STATUS_REG + SDMA1_REGISTER_OFFSET);
+		if (!(tmp & SDMA0_STATUS_REG__IDLE_MASK))
+			reset_mask |= AMDGPU_RESET_DMA1;
+	}
 #if 0
 	/* VCE_STATUS */
 	if (adev->asic_type != CHIP_TOPAZ) {
@@ -1319,6 +1336,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev)
 		adev->num_ip_blocks = ARRAY_SIZE(tonga_ip_blocks);
 		break;
 	case CHIP_CARRIZO:
+	case CHIP_STONEY:
 		adev->ip_blocks = cz_ip_blocks;
 		adev->num_ip_blocks = ARRAY_SIZE(cz_ip_blocks);
 		break;
@@ -1330,11 +1348,18 @@ int vi_set_ip_blocks(struct amdgpu_device *adev)
 	return 0;
 }
 
+#define ATI_REV_ID_FUSE_MACRO__ADDRESS      0xC0014044
+#define ATI_REV_ID_FUSE_MACRO__SHIFT        9
+#define ATI_REV_ID_FUSE_MACRO__MASK         0x00001E00
+
 static uint32_t vi_get_rev_id(struct amdgpu_device *adev)
 {
 	if (adev->asic_type == CHIP_TOPAZ)
 		return (RREG32(mmPCIE_EFUSE4) & PCIE_EFUSE4__STRAP_BIF_ATI_REV_ID_MASK)
 			>> PCIE_EFUSE4__STRAP_BIF_ATI_REV_ID__SHIFT;
+	else if (adev->flags & AMD_IS_APU)
+		return (RREG32_SMC(ATI_REV_ID_FUSE_MACRO__ADDRESS) & ATI_REV_ID_FUSE_MACRO__MASK)
+			>> ATI_REV_ID_FUSE_MACRO__SHIFT;
 	else
 		return (RREG32(mmCC_DRM_ID_STRAPS) & CC_DRM_ID_STRAPS__ATI_REV_ID_MASK)
 			>> CC_DRM_ID_STRAPS__ATI_REV_ID__SHIFT;
@@ -1401,6 +1426,7 @@ static int vi_common_early_init(void *handle)
 			adev->firmware.smu_load = true;
 		break;
 	case CHIP_CARRIZO:
+	case CHIP_STONEY:
 		adev->has_uvd = true;
 		adev->cg_flags = 0;
 		/* Disable UVD pg */
diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index 68a8eaa1b7d0..fe28fb353fab 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -47,6 +47,7 @@ enum amd_asic_type {
 	CHIP_TONGA,
 	CHIP_FIJI,
 	CHIP_CARRIZO,
+	CHIP_STONEY,
 	CHIP_LAST,
 };
 
diff --git a/drivers/gpu/drm/amd/include/atombios.h b/drivers/gpu/drm/amd/include/atombios.h
index 44c5d4a4d1bf..552622675ace 100644
--- a/drivers/gpu/drm/amd/include/atombios.h
+++ b/drivers/gpu/drm/amd/include/atombios.h
@@ -6784,7 +6784,7 @@ typedef struct _ATOM_MC_INIT_PARAM_TABLE_V2_1
   ULONG                         ulMCUcodeRomStartAddr;
   ULONG                         ulMCUcodeLength;
   USHORT                        usMcRegInitTableOffset;     // offset of ATOM_REG_INIT_SETTING array for MC core register settings.
-  USHORT                        usReserved;                 // offset of ATOM_INIT_REG_BLOCK for MC SEQ/PHY regsiter setting
+  USHORT                        usReserved;                 // offset of ATOM_INIT_REG_BLOCK for MC SEQ/PHY register setting
 }ATOM_MC_INIT_PARAM_TABLE_V2_1;
 
 
diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.c b/drivers/gpu/drm/radeon/cayman_blit_shaders.c
index 98d009e154bf..9fec4d09f383 100644
--- a/drivers/gpu/drm/radeon/cayman_blit_shaders.c
+++ b/drivers/gpu/drm/radeon/cayman_blit_shaders.c
@@ -32,7 +32,7 @@
  * evergreen cards need to use the 3D engine to blit data which requires
  * quite a bit of hw state setup.  Rather than pull the whole 3D driver
  * (which normally generates the 3D state) into the DRM, we opt to use
- * statically generated state tables.  The regsiter state and shaders
+ * statically generated state tables.  The register state and shaders
  * were hand generated to support blitting functionality.  See the 3D
  * driver or documentation for descriptions of the registers and
  * shader instructions.
diff --git a/drivers/gpu/drm/radeon/evergreen_blit_shaders.c b/drivers/gpu/drm/radeon/evergreen_blit_shaders.c
index d43383470cdf..1a96ddb3e5ed 100644
--- a/drivers/gpu/drm/radeon/evergreen_blit_shaders.c
+++ b/drivers/gpu/drm/radeon/evergreen_blit_shaders.c
@@ -32,7 +32,7 @@
  * evergreen cards need to use the 3D engine to blit data which requires
  * quite a bit of hw state setup.  Rather than pull the whole 3D driver
  * (which normally generates the 3D state) into the DRM, we opt to use
- * statically generated state tables.  The regsiter state and shaders
+ * statically generated state tables.  The register state and shaders
  * were hand generated to support blitting functionality.  See the 3D
  * driver or documentation for descriptions of the registers and
  * shader instructions.
diff --git a/drivers/gpu/drm/radeon/r600_blit_shaders.c b/drivers/gpu/drm/radeon/r600_blit_shaders.c
index 34c8b2340f33..443cbe59b274 100644
--- a/drivers/gpu/drm/radeon/r600_blit_shaders.c
+++ b/drivers/gpu/drm/radeon/r600_blit_shaders.c
@@ -32,7 +32,7 @@
  * R6xx+ cards need to use the 3D engine to blit data which requires
  * quite a bit of hw state setup.  Rather than pull the whole 3D driver
  * (which normally generates the 3D state) into the DRM, we opt to use
- * statically generated state tables.  The regsiter state and shaders
+ * statically generated state tables.  The register state and shaders
  * were hand generated to support blitting functionality.  See the 3D
  * driver or documentation for descriptions of the registers and
  * shader instructions.
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 5f687c95a1e1..0ec6fcca16d3 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -181,7 +181,9 @@ static void radeon_set_filp_rights(struct drm_device *dev,
 				   struct drm_file *applier,
 				   uint32_t *value)
 {
-	mutex_lock(&dev->struct_mutex);
+	struct radeon_device *rdev = dev->dev_private;
+
+	mutex_lock(&rdev->gem.mutex);
 	if (*value == 1) {
 		/* wants rights */
 		if (!*owner)
@@ -192,7 +194,7 @@ static void radeon_set_filp_rights(struct drm_device *dev,
 			*owner = NULL;
 	}
 	*value = *owner == applier ? 1 : 0;
-	mutex_unlock(&dev->struct_mutex);
+	mutex_unlock(&rdev->gem.mutex);
 }
 
 /*
@@ -727,10 +729,14 @@ void radeon_driver_preclose_kms(struct drm_device *dev,
 				struct drm_file *file_priv)
 {
 	struct radeon_device *rdev = dev->dev_private;
+
+	mutex_lock(&rdev->gem.mutex);
 	if (rdev->hyperz_filp == file_priv)
 		rdev->hyperz_filp = NULL;
 	if (rdev->cmask_filp == file_priv)
 		rdev->cmask_filp = NULL;
+	mutex_unlock(&rdev->gem.mutex);
+
 	radeon_uvd_free_handles(rdev, file_priv);
 	radeon_vce_free_handles(rdev, file_priv);
 }