Merge branch 'drm-next-4.10' of git://people.freedesktop.org/~agd5f/linux into drm-next

Misc fixes and cleanups for 4.10. Highlights:
- Cursor fixes for radeon and amdgpu
- DPM fixes for some new SI variants
- Powerplay fixes
- Clock and powergating fixes

* 'drm-next-4.10' of git://people.freedesktop.org/~agd5f/linux: (39 commits)
  drm/amdgpu/dce6: Set MASTER_UPDATE_MODE to 0 in resume_mc_access as well
  drm/amdgpu: use pin rather than pin_restricted in a few cases
  drm/amd/powerplay: use pr_debug to print function not implemented message
  drm/amd/amdgpu: Add gpr reading for GFX v7
  drm/amd/amdgpu: Add gpr reading for GFX v6
  drm/amd/amdgpu: Add debugfs support for reading GPRs (v2)
  drm/amd/amdgpu: export vbios information (v2)
  drm/amd/amdgpu: stored bios_size
  drm/amdgpu: update golden setting of tonga
  drm/amdgpu/si: load the proper firmware on 0x87 oland boards
  drm/amdgpu: add additional pci revision to dpm workaround
  drm/radeon/si: load the proper firmware on 0x87 oland boards
  drm/radeon: add additional pci revision to dpm workaround
  drm/amd/powerplay: Fix potential NULL pointer issue
  drm/amdgpu: Don't touch GFX hw during HW fini
  drm/amd/powerplay: Adjust the position of data size initial
  drm/amd/powerplay: Ignore smu buffer usage
  drm/amd/powerplay: cut digest part
  drm/amdgpu: drop redundant vi_mqd define
  drm/amdgpu: fix gtt available page num accounting
  ...

41 files changed, 557 insertions, 436 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index f53e52f4d672..54ac8a845e9f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -842,6 +842,8 @@ struct amdgpu_gfx_funcs {
 	uint64_t (*get_gpu_clock_counter)(struct amdgpu_device *adev);
 	void (*select_se_sh)(struct amdgpu_device *adev, u32 se_num, u32 sh_num, u32 instance);
 	void (*read_wave_data)(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t *dst, int *no_fields);
+	void (*read_wave_vgprs)(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t thread, uint32_t start, uint32_t size, uint32_t *dst);
+	void (*read_wave_sgprs)(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t start, uint32_t size, uint32_t *dst);
 };
 
 struct amdgpu_gfx {
@@ -1330,6 +1332,7 @@ struct amdgpu_device {
 
 	/* BIOS */
 	uint8_t				*bios;
+	uint32_t			bios_size;
 	bool				is_atom_bios;
 	struct amdgpu_bo		*stollen_vga_memory;
 	uint32_t			bios_scratch[AMDGPU_BIOS_NUM_SCRATCH];
@@ -1679,8 +1682,6 @@ uint32_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 base);
 void amdgpu_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc);
 void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size);
-u64 amdgpu_ttm_get_gtt_mem_size(struct amdgpu_device *adev);
-int amdgpu_ttm_global_init(struct amdgpu_device *adev);
 int amdgpu_ttm_init(struct amdgpu_device *adev);
 void amdgpu_ttm_fini(struct amdgpu_device *adev);
 void amdgpu_program_register_sequence(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c
index b7e2762fcdd2..4f973a9c7b87 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c
@@ -74,6 +74,7 @@ static bool igp_read_bios_from_vram(struct amdgpu_device *adev)
 		iounmap(bios);
 		return false;
 	}
+	adev->bios_size = size;
 	memcpy_fromio(adev->bios, bios, size);
 	iounmap(bios);
 	return true;
@@ -103,6 +104,7 @@ bool amdgpu_read_bios(struct amdgpu_device *adev)
 		pci_unmap_rom(adev->pdev, bios);
 		return false;
 	}
+	adev->bios_size = size;
 	memcpy_fromio(adev->bios, bios, size);
 	pci_unmap_rom(adev->pdev, bios);
 	return true;
@@ -135,6 +137,7 @@ static bool amdgpu_read_bios_from_rom(struct amdgpu_device *adev)
 		DRM_ERROR("no memory to allocate for BIOS\n");
 		return false;
 	}
+	adev->bios_size = len;
 
 	/* read complete BIOS */
 	return amdgpu_asic_read_bios_from_rom(adev, adev->bios, len);
@@ -159,6 +162,7 @@ static bool amdgpu_read_platform_bios(struct amdgpu_device *adev)
 	if (adev->bios == NULL) {
 		return false;
 	}
+	adev->bios_size = size;
 
 	return true;
 }
@@ -273,6 +277,7 @@ static bool amdgpu_atrm_get_bios(struct amdgpu_device *adev)
 		kfree(adev->bios);
 		return false;
 	}
+	adev->bios_size = size;
 	return true;
 }
 #else
@@ -334,6 +339,7 @@ static bool amdgpu_acpi_vfct_bios(struct amdgpu_device *adev)
 	}
 
 	adev->bios = kmemdup(&vbios->VbiosContent, vhdr->ImageLength, GFP_KERNEL);
+	adev->bios_size = vhdr->ImageLength;
 	ret = !!adev->bios;
 
 out_unmap:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 7ded61e6dd81..9ada56c16a58 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -723,7 +723,7 @@ static uint16_t amdgpu_get_firmware_version(struct cgs_device *cgs_device,
 					enum cgs_ucode_id type)
 {
 	CGS_FUNC_ADEV;
-	uint16_t fw_version;
+	uint16_t fw_version = 0;
 
 	switch (type) {
 		case CGS_UCODE_ID_SDMA0:
@@ -753,9 +753,11 @@ static uint16_t amdgpu_get_firmware_version(struct cgs_device *cgs_device,
 		case CGS_UCODE_ID_RLC_G:
 			fw_version = adev->gfx.rlc_fw_version;
 			break;
+		case CGS_UCODE_ID_STORAGE:
+			break;
 		default:
 			DRM_ERROR("firmware type %d do not have version\n", type);
-			fw_version = 0;
+			break;
 	}
 	return fw_version;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 5a277495d6a3..29d6d84d1c28 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -451,7 +451,7 @@ static int amdgpu_cs_validate(void *param, struct amdgpu_bo *bo)
 		return r;
 
 	if (bo->shadow)
-		r = amdgpu_cs_bo_validate(p, bo);
+		r = amdgpu_cs_bo_validate(p, bo->shadow);
 
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index fc790e5c46fd..53ea782c2c47 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1470,20 +1470,26 @@ static int amdgpu_fini(struct amdgpu_device *adev)
 			amdgpu_wb_fini(adev);
 			amdgpu_vram_scratch_fini(adev);
 		}
-		/* ungate blocks before hw fini so that we can shutdown the blocks safely */
-		r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev,
-									     AMD_CG_STATE_UNGATE);
-		if (r) {
-			DRM_ERROR("set_clockgating_state(ungate) of IP block <%s> failed %d\n",
-				  adev->ip_blocks[i].version->funcs->name, r);
-			return r;
+
+		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
+			adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE) {
+			/* ungate blocks before hw fini so that we can shutdown the blocks safely */
+			r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev,
+										     AMD_CG_STATE_UNGATE);
+			if (r) {
+				DRM_ERROR("set_clockgating_state(ungate) of IP block <%s> failed %d\n",
+					  adev->ip_blocks[i].version->funcs->name, r);
+				return r;
+			}
 		}
+
 		r = adev->ip_blocks[i].version->funcs->hw_fini((void *)adev);
 		/* XXX handle errors */
 		if (r) {
 			DRM_DEBUG("hw_fini of IP block <%s> failed %d\n",
 				  adev->ip_blocks[i].version->funcs->name, r);
 		}
+
 		adev->ip_blocks[i].status.hw = false;
 	}
 
@@ -2979,6 +2985,66 @@ static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
 	return result;
 }
 
+static ssize_t amdgpu_debugfs_gpr_read(struct file *f, char __user *buf,
+					size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = f->f_inode->i_private;
+	int r;
+	ssize_t result = 0;
+	uint32_t offset, se, sh, cu, wave, simd, thread, bank, *data;
+
+	if (size & 3 || *pos & 3)
+		return -EINVAL;
+
+	/* decode offset */
+	offset = (*pos & 0xFFF);       /* in dwords */
+	se = ((*pos >> 12) & 0xFF);
+	sh = ((*pos >> 20) & 0xFF);
+	cu = ((*pos >> 28) & 0xFF);
+	wave = ((*pos >> 36) & 0xFF);
+	simd = ((*pos >> 44) & 0xFF);
+	thread = ((*pos >> 52) & 0xFF);
+	bank = ((*pos >> 60) & 1);
+
+	data = kmalloc_array(1024, sizeof(*data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	/* switch to the specific se/sh/cu */
+	mutex_lock(&adev->grbm_idx_mutex);
+	amdgpu_gfx_select_se_sh(adev, se, sh, cu);
+
+	if (bank == 0) {
+		if (adev->gfx.funcs->read_wave_vgprs)
+			adev->gfx.funcs->read_wave_vgprs(adev, simd, wave, thread, offset, size>>2, data);
+	} else {
+		if (adev->gfx.funcs->read_wave_sgprs)
+			adev->gfx.funcs->read_wave_sgprs(adev, simd, wave, offset, size>>2, data);
+	}
+
+	amdgpu_gfx_select_se_sh(adev, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF);
+	mutex_unlock(&adev->grbm_idx_mutex);
+
+	while (size) {
+		uint32_t value;
+
+		value = data[offset++];
+		r = put_user(value, (uint32_t *)buf);
+		if (r) {
+			result = r;
+			goto err;
+		}
+
+		result += 4;
+		buf += 4;
+		size -= 4;
+	}
+
+err:
+	kfree(data);
+	return result;
+}
+
 static const struct file_operations amdgpu_debugfs_regs_fops = {
 	.owner = THIS_MODULE,
 	.read = amdgpu_debugfs_regs_read,
@@ -3021,6 +3087,11 @@ static const struct file_operations amdgpu_debugfs_wave_fops = {
 	.read = amdgpu_debugfs_wave_read,
 	.llseek = default_llseek
 };
+static const struct file_operations amdgpu_debugfs_gpr_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_gpr_read,
+	.llseek = default_llseek
+};
 
 static const struct file_operations *debugfs_regs[] = {
 	&amdgpu_debugfs_regs_fops,
@@ -3030,6 +3101,7 @@ static const struct file_operations *debugfs_regs[] = {
 	&amdgpu_debugfs_gca_config_fops,
 	&amdgpu_debugfs_sensors_fops,
 	&amdgpu_debugfs_wave_fops,
+	&amdgpu_debugfs_gpr_fops,
 };
 
 static const char *debugfs_regs_names[] = {
@@ -3040,6 +3112,7 @@ static const char *debugfs_regs_names[] = {
 	"amdgpu_gca_config",
 	"amdgpu_sensors",
 	"amdgpu_wave",
+	"amdgpu_gpr",
 };
 
 static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
index 741144fcc7bc..581601ca6b89 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -187,7 +187,7 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc,
 		goto cleanup;
 	}
 
-	r = amdgpu_bo_pin_restricted(new_abo, AMDGPU_GEM_DOMAIN_VRAM, 0, 0, &base);
+	r = amdgpu_bo_pin(new_abo, AMDGPU_GEM_DOMAIN_VRAM, &base);
 	if (unlikely(r != 0)) {
 		r = -EINVAL;
 		DRM_ERROR("failed to pin new abo buffer before flip\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
index f1c9e59a7c87..24629bec181a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -171,7 +171,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
 	}
 
 
-	ret = amdgpu_bo_pin_restricted(abo, AMDGPU_GEM_DOMAIN_VRAM, 0, 0, NULL);
+	ret = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, NULL);
 	if (ret) {
 		amdgpu_bo_unreserve(abo);
 		goto out_unref;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
index 3c634f02a3d5..00f46b0e076d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
@@ -164,8 +164,10 @@ static int amdgpu_gtt_mgr_new(struct ttm_mem_type_manager *man,
 	spin_unlock(&mgr->lock);
 
 	node = kzalloc(sizeof(*node), GFP_KERNEL);
-	if (!node)
-		return -ENOMEM;
+	if (!node) {
+		r = -ENOMEM;
+		goto err_out;
+	}
 
 	node->start = AMDGPU_BO_INVALID_OFFSET;
 	node->size = mem->num_pages;
@@ -176,12 +178,20 @@ static int amdgpu_gtt_mgr_new(struct ttm_mem_type_manager *man,
 		if (unlikely(r)) {
 			kfree(node);
 			mem->mm_node = NULL;
+			r = 0;
+			goto err_out;
 		}
 	} else {
 		mem->start = node->start;
 	}
 
 	return 0;
+err_out:
+	spin_lock(&mgr->lock);
+	mgr->available += mem->num_pages;
+	spin_unlock(&mgr->lock);
+
+	return r;
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index d1cf9ac0dff1..9af87eaf8ee3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -544,6 +544,32 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		return copy_to_user(out, &vce_clk_table,
 				    min((size_t)size, sizeof(vce_clk_table))) ? -EFAULT : 0;
 	}
+	case AMDGPU_INFO_VBIOS: {
+		uint32_t bios_size = adev->bios_size;
+
+		switch (info->vbios_info.type) {
+		case AMDGPU_INFO_VBIOS_SIZE:
+			return copy_to_user(out, &bios_size,
+					min((size_t)size, sizeof(bios_size)))
+					? -EFAULT : 0;
+		case AMDGPU_INFO_VBIOS_IMAGE: {
+			uint8_t *bios;
+			uint32_t bios_offset = info->vbios_info.offset;
+
+			if (bios_offset >= bios_size)
+				return -EINVAL;
+
+			bios = adev->bios + bios_offset;
+			return copy_to_user(out, bios,
+					    min((size_t)size, (size_t)(bios_size - bios_offset)))
+					? -EFAULT : 0;
+		}
+		default:
+			DRM_DEBUG_KMS("Invalid request %d\n",
+					info->vbios_info.type);
+			return -EINVAL;
+		}
+	}
 	default:
 		DRM_DEBUG_KMS("Invalid request %d\n", info->query);
 		return -EINVAL;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 8f18b8ed2b3a..bc70f80260d8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -34,7 +34,6 @@
 #include <ttm/ttm_placement.h>
 #include <ttm/ttm_module.h>
 #include <ttm/ttm_page_alloc.h>
-#include <ttm/ttm_memory.h>
 #include <drm/drmP.h>
 #include <drm/amdgpu_drm.h>
 #include <linux/seq_file.h>
@@ -65,7 +64,7 @@ static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref)
 	ttm_mem_global_release(ref->object);
 }
 
-int amdgpu_ttm_global_init(struct amdgpu_device *adev)
+static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 {
 	struct drm_global_reference *global_ref;
 	struct amdgpu_ring *ring;
@@ -1151,6 +1150,10 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	unsigned i, j;
 	int r;
 
+	r = amdgpu_ttm_global_init(adev);
+	if (r) {
+		return r;
+	}
 	/* No others user of address space so set it to 0 */
 	r = ttm_bo_device_init(&adev->mman.bdev,
 			       adev->mman.bo_global_ref.ref.object,
@@ -1650,8 +1653,3 @@ static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev)
 
 #endif
 }
-
-u64 amdgpu_ttm_get_gtt_mem_size(struct amdgpu_device *adev)
-{
-	return ttm_get_kernel_zone_memory_size(adev->mman.mem_global_ref.object);
-}
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index 1027f92de32b..bda9e3de191e 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -6083,7 +6083,7 @@ ci_dpm_debugfs_print_current_performance_level(struct amdgpu_device *adev,
 		activity_percent = activity_percent > 100 ? 100 : activity_percent;
 	}
 
-	seq_printf(m, "uvd %sabled\n", pi->uvd_enabled ? "en" : "dis");
+	seq_printf(m, "uvd %sabled\n", pi->uvd_power_gated ? "dis" : "en");
 	seq_printf(m, "vce %sabled\n", rps->vce_active ? "en" : "dis");
 	seq_printf(m, "power level avg    sclk: %u mclk: %u\n",
 		   sclk, mclk);
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 075aa0b1b075..9999dc71b998 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -2493,6 +2493,9 @@ static int dce_v10_0_cursor_move_locked(struct drm_crtc *crtc,
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	int xorigin = 0, yorigin = 0;
 
+	amdgpu_crtc->cursor_x = x;
+	amdgpu_crtc->cursor_y = y;
+
 	/* avivo cursor are offset into the total surface */
 	x += crtc->x;
 	y += crtc->y;
@@ -2509,11 +2512,6 @@ static int dce_v10_0_cursor_move_locked(struct drm_crtc *crtc,
 
 	WREG32(mmCUR_POSITION + amdgpu_crtc->crtc_offset, (x << 16) | y);
 	WREG32(mmCUR_HOT_SPOT + amdgpu_crtc->crtc_offset, (xorigin << 16) | yorigin);
-	WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
-	       ((amdgpu_crtc->cursor_width - 1) << 16) | (amdgpu_crtc->cursor_height - 1));
-
-	amdgpu_crtc->cursor_x = x;
-	amdgpu_crtc->cursor_y = y;
 
 	return 0;
 }
@@ -2539,6 +2537,7 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc,
 				      int32_t hot_y)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *aobj;
 	int ret;
@@ -2577,9 +2576,6 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	amdgpu_crtc->cursor_width = width;
-	amdgpu_crtc->cursor_height = height;
-
 	dce_v10_0_lock_cursor(crtc, true);
 
 	if (hot_x != amdgpu_crtc->cursor_hot_x ||
@@ -2595,6 +2591,14 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
+	if (width != amdgpu_crtc->cursor_width ||
+	    height != amdgpu_crtc->cursor_height) {
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (width - 1) << 16 | (height - 1));
+		amdgpu_crtc->cursor_width = width;
+		amdgpu_crtc->cursor_height = height;
+	}
+
 	dce_v10_0_show_cursor(crtc);
 	dce_v10_0_lock_cursor(crtc, false);
 
@@ -2616,6 +2620,7 @@ unpin:
 static void dce_v10_0_cursor_reset(struct drm_crtc *crtc)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 
 	if (amdgpu_crtc->cursor_bo) {
 		dce_v10_0_lock_cursor(crtc, true);
@@ -2623,6 +2628,10 @@ static void dce_v10_0_cursor_reset(struct drm_crtc *crtc)
 		dce_v10_0_cursor_move_locked(crtc, amdgpu_crtc->cursor_x,
 					     amdgpu_crtc->cursor_y);
 
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (amdgpu_crtc->cursor_width - 1) << 16 |
+		       (amdgpu_crtc->cursor_height - 1));
+
 		dce_v10_0_show_cursor(crtc);
 
 		dce_v10_0_lock_cursor(crtc, false);
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index a6717487ac78..b3d62b909f43 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -2509,6 +2509,9 @@ static int dce_v11_0_cursor_move_locked(struct drm_crtc *crtc,
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	int xorigin = 0, yorigin = 0;
 
+	amdgpu_crtc->cursor_x = x;
+	amdgpu_crtc->cursor_y = y;
+
 	/* avivo cursor are offset into the total surface */
 	x += crtc->x;
 	y += crtc->y;
@@ -2525,11 +2528,6 @@ static int dce_v11_0_cursor_move_locked(struct drm_crtc *crtc,
 
 	WREG32(mmCUR_POSITION + amdgpu_crtc->crtc_offset, (x << 16) | y);
 	WREG32(mmCUR_HOT_SPOT + amdgpu_crtc->crtc_offset, (xorigin << 16) | yorigin);
-	WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
-	       ((amdgpu_crtc->cursor_width - 1) << 16) | (amdgpu_crtc->cursor_height - 1));
-
-	amdgpu_crtc->cursor_x = x;
-	amdgpu_crtc->cursor_y = y;
 
 	return 0;
 }
@@ -2555,6 +2553,7 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc,
 				      int32_t hot_y)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *aobj;
 	int ret;
@@ -2593,9 +2592,6 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	amdgpu_crtc->cursor_width = width;
-	amdgpu_crtc->cursor_height = height;
-
 	dce_v11_0_lock_cursor(crtc, true);
 
 	if (hot_x != amdgpu_crtc->cursor_hot_x ||
@@ -2611,6 +2607,14 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
+	if (width != amdgpu_crtc->cursor_width ||
+	    height != amdgpu_crtc->cursor_height) {
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (width - 1) << 16 | (height - 1));
+		amdgpu_crtc->cursor_width = width;
+		amdgpu_crtc->cursor_height = height;
+	}
+
 	dce_v11_0_show_cursor(crtc);
 	dce_v11_0_lock_cursor(crtc, false);
 
@@ -2632,6 +2636,7 @@ unpin:
 static void dce_v11_0_cursor_reset(struct drm_crtc *crtc)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 
 	if (amdgpu_crtc->cursor_bo) {
 		dce_v11_0_lock_cursor(crtc, true);
@@ -2639,6 +2644,10 @@ static void dce_v11_0_cursor_reset(struct drm_crtc *crtc)
 		dce_v11_0_cursor_move_locked(crtc, amdgpu_crtc->cursor_x,
 					     amdgpu_crtc->cursor_y);
 
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (amdgpu_crtc->cursor_width - 1) << 16 |
+		       (amdgpu_crtc->cursor_height - 1));
+
 		dce_v11_0_show_cursor(crtc);
 
 		dce_v11_0_lock_cursor(crtc, false);
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index 15d98ef696a2..e564442b6393 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -460,9 +460,8 @@ static void dce_v6_0_resume_mc_access(struct amdgpu_device *adev,
 	for (i = 0; i < adev->mode_info.num_crtc; i++) {
 		if (save->crtc_enabled[i]) {
 			tmp = RREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i]);
-			if ((tmp & 0x7) != 3) {
+			if ((tmp & 0x7) != 0) {
 				tmp &= ~0x7;
-				tmp |= 0x3;
 				WREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i], tmp);
 			}
 			tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]);
@@ -1860,7 +1859,8 @@ static int dce_v6_0_cursor_move_locked(struct drm_crtc *crtc,
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	int xorigin = 0, yorigin = 0;
 
-	int w = amdgpu_crtc->cursor_width;
+	amdgpu_crtc->cursor_x = x;
+	amdgpu_crtc->cursor_y = y;
 
 	/* avivo cursor are offset into the total surface */
 	x += crtc->x;
@@ -1878,11 +1878,7 @@ static int dce_v6_0_cursor_move_locked(struct drm_crtc *crtc,
 
 	WREG32(mmCUR_POSITION + amdgpu_crtc->crtc_offset, (x << 16) | y);
 	WREG32(mmCUR_HOT_SPOT + amdgpu_crtc->crtc_offset, (xorigin << 16) | yorigin);
-	WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
-	       ((w - 1) << 16) | (amdgpu_crtc->cursor_height - 1));
 
-	amdgpu_crtc->cursor_x = x;
-	amdgpu_crtc->cursor_y = y;
 	return 0;
 }
 
@@ -1907,6 +1903,7 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
 				     int32_t hot_y)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *aobj;
 	int ret;
@@ -1945,12 +1942,11 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	amdgpu_crtc->cursor_width = width;
-	amdgpu_crtc->cursor_height = height;
-
 	dce_v6_0_lock_cursor(crtc, true);
 
-	if (hot_x != amdgpu_crtc->cursor_hot_x ||
+	if (width != amdgpu_crtc->cursor_width ||
+	    height != amdgpu_crtc->cursor_height ||
+	    hot_x != amdgpu_crtc->cursor_hot_x ||
 	    hot_y != amdgpu_crtc->cursor_hot_y) {
 		int x, y;
 
@@ -1959,10 +1955,20 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 		dce_v6_0_cursor_move_locked(crtc, x, y);
 
+		amdgpu_crtc->cursor_width = width;
+		amdgpu_crtc->cursor_height = height;
 		amdgpu_crtc->cursor_hot_x = hot_x;
 		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
+	if (width != amdgpu_crtc->cursor_width ||
+	    height != amdgpu_crtc->cursor_height) {
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (width - 1) << 16 | (height - 1));
+		amdgpu_crtc->cursor_width = width;
+		amdgpu_crtc->cursor_height = height;
+	}
+
 	dce_v6_0_show_cursor(crtc);
 	dce_v6_0_lock_cursor(crtc, false);
 
@@ -1984,6 +1990,7 @@ unpin:
 static void dce_v6_0_cursor_reset(struct drm_crtc *crtc)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 
 	if (amdgpu_crtc->cursor_bo) {
 		dce_v6_0_lock_cursor(crtc, true);
@@ -1991,6 +1998,10 @@ static void dce_v6_0_cursor_reset(struct drm_crtc *crtc)
 		dce_v6_0_cursor_move_locked(crtc, amdgpu_crtc->cursor_x,
 					    amdgpu_crtc->cursor_y);
 
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (amdgpu_crtc->cursor_width - 1) << 16 |
+		       (amdgpu_crtc->cursor_height - 1));
+
 		dce_v6_0_show_cursor(crtc);
 		dce_v6_0_lock_cursor(crtc, false);
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index a699896eeabc..6ce7fb42dbef 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -2344,6 +2344,9 @@ static int dce_v8_0_cursor_move_locked(struct drm_crtc *crtc,
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	int xorigin = 0, yorigin = 0;
 
+	amdgpu_crtc->cursor_x = x;
+	amdgpu_crtc->cursor_y = y;
+
 	/* avivo cursor are offset into the total surface */
 	x += crtc->x;
 	y += crtc->y;
@@ -2360,11 +2363,6 @@ static int dce_v8_0_cursor_move_locked(struct drm_crtc *crtc,
 
 	WREG32(mmCUR_POSITION + amdgpu_crtc->crtc_offset, (x << 16) | y);
 	WREG32(mmCUR_HOT_SPOT + amdgpu_crtc->crtc_offset, (xorigin << 16) | yorigin);
-	WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
-	       ((amdgpu_crtc->cursor_width - 1) << 16) | (amdgpu_crtc->cursor_height - 1));
-
-	amdgpu_crtc->cursor_x = x;
-	amdgpu_crtc->cursor_y = y;
 
 	return 0;
 }
@@ -2390,6 +2388,7 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc,
 				     int32_t hot_y)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *aobj;
 	int ret;
@@ -2428,9 +2427,6 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	amdgpu_crtc->cursor_width = width;
-	amdgpu_crtc->cursor_height = height;
-
 	dce_v8_0_lock_cursor(crtc, true);
 
 	if (hot_x != amdgpu_crtc->cursor_hot_x ||
@@ -2442,10 +2438,20 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 		dce_v8_0_cursor_move_locked(crtc, x, y);
 
+		amdgpu_crtc->cursor_width = width;
+		amdgpu_crtc->cursor_height = height;
 		amdgpu_crtc->cursor_hot_x = hot_x;
 		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
+	if (width != amdgpu_crtc->cursor_width ||
+	    height != amdgpu_crtc->cursor_height) {
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (width - 1) << 16 | (height - 1));
+		amdgpu_crtc->cursor_width = width;
+		amdgpu_crtc->cursor_height = height;
+	}
+
 	dce_v8_0_show_cursor(crtc);
 	dce_v8_0_lock_cursor(crtc, false);
 
@@ -2467,6 +2473,7 @@ unpin:
 static void dce_v8_0_cursor_reset(struct drm_crtc *crtc)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 
 	if (amdgpu_crtc->cursor_bo) {
 		dce_v8_0_lock_cursor(crtc, true);
@@ -2474,6 +2481,10 @@ static void dce_v8_0_cursor_reset(struct drm_crtc *crtc)
 		dce_v8_0_cursor_move_locked(crtc, amdgpu_crtc->cursor_x,
 					    amdgpu_crtc->cursor_y);
 
+		WREG32(mmCUR_SIZE + amdgpu_crtc->crtc_offset,
+		       (amdgpu_crtc->cursor_width - 1) << 16 |
+		       (amdgpu_crtc->cursor_height - 1));
+
 		dce_v8_0_show_cursor(crtc);
 
 		dce_v8_0_lock_cursor(crtc, false);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
index 879a94bbfe12..558640aee15a 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
@@ -2827,6 +2827,21 @@ static uint32_t wave_read_ind(struct amdgpu_device *adev, uint32_t simd, uint32_
 	return RREG32(mmSQ_IND_DATA);
 }
 
+static void wave_read_regs(struct amdgpu_device *adev, uint32_t simd,
+			   uint32_t wave, uint32_t thread,
+			   uint32_t regno, uint32_t num, uint32_t *out)
+{
+	WREG32(mmSQ_IND_INDEX,
+		(wave << SQ_IND_INDEX__WAVE_ID__SHIFT) |
+		(simd << SQ_IND_INDEX__SIMD_ID__SHIFT) |
+		(regno << SQ_IND_INDEX__INDEX__SHIFT) |
+		(thread << SQ_IND_INDEX__THREAD_ID__SHIFT) |
+		(SQ_IND_INDEX__FORCE_READ_MASK) |
+		(SQ_IND_INDEX__AUTO_INCR_MASK));
+	while (num--)
+		*(out++) = RREG32(mmSQ_IND_DATA);
+}
+
 static void gfx_v6_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t *dst, int *no_fields)
 {
 	/* type 0 wave data */
@@ -2851,10 +2866,20 @@ static void gfx_v6_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, u
 	dst[(*no_fields)++] = wave_read_ind(adev, simd, wave, ixSQ_WAVE_M0);
 }
 
+static void gfx_v6_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd,
+				     uint32_t wave, uint32_t start,
+				     uint32_t size, uint32_t *dst)
+{
+	wave_read_regs(
+		adev, simd, wave, 0,
+		start + SQIND_WAVE_SGPRS_OFFSET, size, dst);
+}
+
 static const struct amdgpu_gfx_funcs gfx_v6_0_gfx_funcs = {
 	.get_gpu_clock_counter = &gfx_v6_0_get_gpu_clock_counter,
 	.select_se_sh = &gfx_v6_0_select_se_sh,
 	.read_wave_data = &gfx_v6_0_read_wave_data,
+	.read_wave_sgprs = &gfx_v6_0_read_wave_sgprs,
 };
 
 static int gfx_v6_0_early_init(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 1a745cf93f47..c4e14015ec5b 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -4380,6 +4380,21 @@ static uint32_t wave_read_ind(struct amdgpu_device *adev, uint32_t simd, uint32_
 	return RREG32(mmSQ_IND_DATA);
 }
 
+static void wave_read_regs(struct amdgpu_device *adev, uint32_t simd,
+			   uint32_t wave, uint32_t thread,
+			   uint32_t regno, uint32_t num, uint32_t *out)
+{
+	WREG32(mmSQ_IND_INDEX,
+		(wave << SQ_IND_INDEX__WAVE_ID__SHIFT) |
+		(simd << SQ_IND_INDEX__SIMD_ID__SHIFT) |
+		(regno << SQ_IND_INDEX__INDEX__SHIFT) |
+		(thread << SQ_IND_INDEX__THREAD_ID__SHIFT) |
+		(SQ_IND_INDEX__FORCE_READ_MASK) |
+		(SQ_IND_INDEX__AUTO_INCR_MASK));
+	while (num--)
+		*(out++) = RREG32(mmSQ_IND_DATA);
+}
+
 static void gfx_v7_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t *dst, int *no_fields)
 {
 	/* type 0 wave data */
@@ -4404,10 +4419,20 @@ static void gfx_v7_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, u
 	dst[(*no_fields)++] = wave_read_ind(adev, simd, wave, ixSQ_WAVE_M0);
 }
 
+static void gfx_v7_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd,
+				     uint32_t wave, uint32_t start,
+				     uint32_t size, uint32_t *dst)
+{
+	wave_read_regs(
+		adev, simd, wave, 0,
+		start + SQIND_WAVE_SGPRS_OFFSET, size, dst);
+}
+
 static const struct amdgpu_gfx_funcs gfx_v7_0_gfx_funcs = {
 	.get_gpu_clock_counter = &gfx_v7_0_get_gpu_clock_counter,
 	.select_se_sh = &gfx_v7_0_select_se_sh,
 	.read_wave_data = &gfx_v7_0_read_wave_data,
+	.read_wave_sgprs = &gfx_v7_0_read_wave_sgprs,
 };
 
 static const struct amdgpu_rlc_funcs gfx_v7_0_rlc_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index a3684891c6e1..6324f67bdb1f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -25,6 +25,7 @@
 #include "amdgpu.h"
 #include "amdgpu_gfx.h"
 #include "vi.h"
+#include "vi_structs.h"
 #include "vid.h"
 #include "amdgpu_ucode.h"
 #include "amdgpu_atombios.h"
@@ -167,6 +168,7 @@ static const u32 golden_settings_tonga_a11[] =
 	mmPA_SC_ENHANCE, 0xffffffff, 0x20000001,
 	mmPA_SC_FIFO_DEPTH_CNTL, 0x000003ff, 0x000000fc,
 	mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000,
+	mmRLC_CGCG_CGLS_CTRL, 0x00000003, 0x0000003c,
 	mmSQ_RANDOM_WAVE_PRI, 0x001fffff, 0x000006fd,
 	mmTA_CNTL_AUX, 0x000f000f, 0x000b0000,
 	mmTCC_CTRL, 0x00100000, 0xf31fff7f,
@@ -1371,7 +1373,7 @@ static int gfx_v8_0_mec_init(struct amdgpu_device *adev)
 
 	if (adev->gfx.mec.hpd_eop_obj == NULL) {
 		r = amdgpu_bo_create(adev,
-				     adev->gfx.mec.num_mec *adev->gfx.mec.num_pipe * MEC_HPD_SIZE * 2,
+				     adev->gfx.mec.num_queue * MEC_HPD_SIZE,
 				     PAGE_SIZE, true,
 				     AMDGPU_GEM_DOMAIN_GTT, 0, NULL, NULL,
 				     &adev->gfx.mec.hpd_eop_obj);
@@ -1400,7 +1402,7 @@ static int gfx_v8_0_mec_init(struct amdgpu_device *adev)
 		return r;
 	}
 
-	memset(hpd, 0, adev->gfx.mec.num_mec *adev->gfx.mec.num_pipe * MEC_HPD_SIZE * 2);
+	memset(hpd, 0, adev->gfx.mec.num_queue * MEC_HPD_SIZE);
 
 	amdgpu_bo_kunmap(adev->gfx.mec.hpd_eop_obj);
 	amdgpu_bo_unreserve(adev->gfx.mec.hpd_eop_obj);
@@ -4469,267 +4471,6 @@ static int gfx_v8_0_cp_compute_load_microcode(struct amdgpu_device *adev)
 	return 0;
 }
 
-struct vi_mqd {
-	uint32_t header;  /* ordinal0 */
-	uint32_t compute_dispatch_initiator;  /* ordinal1 */
-	uint32_t compute_dim_x;  /* ordinal2 */
-	uint32_t compute_dim_y;  /* ordinal3 */
-	uint32_t compute_dim_z;  /* ordinal4 */
-	uint32_t compute_start_x;  /* ordinal5 */
-	uint32_t compute_start_y;  /* ordinal6 */
-	uint32_t compute_start_z;  /* ordinal7 */
-	uint32_t compute_num_thread_x;  /* ordinal8 */
-	uint32_t compute_num_thread_y;  /* ordinal9 */
-	uint32_t compute_num_thread_z;  /* ordinal10 */
-	uint32_t compute_pipelinestat_enable;  /* ordinal11 */
-	uint32_t compute_perfcount_enable;  /* ordinal12 */
-	uint32_t compute_pgm_lo;  /* ordinal13 */
-	uint32_t compute_pgm_hi;  /* ordinal14 */
-	uint32_t compute_tba_lo;  /* ordinal15 */
-	uint32_t compute_tba_hi;  /* ordinal16 */
-	uint32_t compute_tma_lo;  /* ordinal17 */
-	uint32_t compute_tma_hi;  /* ordinal18 */
-	uint32_t compute_pgm_rsrc1;  /* ordinal19 */
-	uint32_t compute_pgm_rsrc2;  /* ordinal20 */
-	uint32_t compute_vmid;  /* ordinal21 */
-	uint32_t compute_resource_limits;  /* ordinal22 */
-	uint32_t compute_static_thread_mgmt_se0;  /* ordinal23 */
-	uint32_t compute_static_thread_mgmt_se1;  /* ordinal24 */
-	uint32_t compute_tmpring_size;  /* ordinal25 */
-	uint32_t compute_static_thread_mgmt_se2;  /* ordinal26 */
-	uint32_t compute_static_thread_mgmt_se3;  /* ordinal27 */
-	uint32_t compute_restart_x;  /* ordinal28 */
-	uint32_t compute_restart_y;  /* ordinal29 */
-	uint32_t compute_restart_z;  /* ordinal30 */
-	uint32_t compute_thread_trace_enable;  /* ordinal31 */
-	uint32_t compute_misc_reserved;  /* ordinal32 */
-	uint32_t compute_dispatch_id;  /* ordinal33 */
-	uint32_t compute_threadgroup_id;  /* ordinal34 */
-	uint32_t compute_relaunch;  /* ordinal35 */
-	uint32_t compute_wave_restore_addr_lo;  /* ordinal36 */
-	uint32_t compute_wave_restore_addr_hi;  /* ordinal37 */
-	uint32_t compute_wave_restore_control;  /* ordinal38 */
-	uint32_t reserved9;  /* ordinal39 */
-	uint32_t reserved10;  /* ordinal40 */
-	uint32_t reserved11;  /* ordinal41 */
-	uint32_t reserved12;  /* ordinal42 */
-	uint32_t reserved13;  /* ordinal43 */
-	uint32_t reserved14;  /* ordinal44 */
-	uint32_t reserved15;  /* ordinal45 */
-	uint32_t reserved16;  /* ordinal46 */
-	uint32_t reserved17;  /* ordinal47 */
-	uint32_t reserved18;  /* ordinal48 */
-	uint32_t reserved19;  /* ordinal49 */
-	uint32_t reserved20;  /* ordinal50 */
-	uint32_t reserved21;  /* ordinal51 */
-	uint32_t reserved22;  /* ordinal52 */
-	uint32_t reserved23;  /* ordinal53 */
-	uint32_t reserved24;  /* ordinal54 */
-	uint32_t reserved25;  /* ordinal55 */
-	uint32_t reserved26;  /* ordinal56 */
-	uint32_t reserved27;  /* ordinal57 */
-	uint32_t reserved28;  /* ordinal58 */
-	uint32_t reserved29;  /* ordinal59 */
-	uint32_t reserved30;  /* ordinal60 */
-	uint32_t reserved31;  /* ordinal61 */
-	uint32_t reserved32;  /* ordinal62 */
-	uint32_t reserved33;  /* ordinal63 */
-	uint32_t reserved34;  /* ordinal64 */
-	uint32_t compute_user_data_0;  /* ordinal65 */
-	uint32_t compute_user_data_1;  /* ordinal66 */
-	uint32_t compute_user_data_2;  /* ordinal67 */
-	uint32_t compute_user_data_3;  /* ordinal68 */
-	uint32_t compute_user_data_4;  /* ordinal69 */
-	uint32_t compute_user_data_5;  /* ordinal70 */
-	uint32_t compute_user_data_6;  /* ordinal71 */
-	uint32_t compute_user_data_7;  /* ordinal72 */
-	uint32_t compute_user_data_8;  /* ordinal73 */
-	uint32_t compute_user_data_9;  /* ordinal74 */
-	uint32_t compute_user_data_10;  /* ordinal75 */
-	uint32_t compute_user_data_11;  /* ordinal76 */
-	uint32_t compute_user_data_12;  /* ordinal77 */
-	uint32_t compute_user_data_13;  /* ordinal78 */
-	uint32_t compute_user_data_14;  /* ordinal79 */
-	uint32_t compute_user_data_15;  /* ordinal80 */
-	uint32_t cp_compute_csinvoc_count_lo;  /* ordinal81 */
-	uint32_t cp_compute_csinvoc_count_hi;  /* ordinal82 */
-	uint32_t reserved35;  /* ordinal83 */
-	uint32_t reserved36;  /* ordinal84 */
-	uint32_t reserved37;  /* ordinal85 */
-	uint32_t cp_mqd_query_time_lo;  /* ordinal86 */
-	uint32_t cp_mqd_query_time_hi;  /* ordinal87 */
-	uint32_t cp_mqd_connect_start_time_lo;  /* ordinal88 */
-	uint32_t cp_mqd_connect_start_time_hi;  /* ordinal89 */
-	uint32_t cp_mqd_connect_end_time_lo;  /* ordinal90 */
-	uint32_t cp_mqd_connect_end_time_hi;  /* ordinal91 */
-	uint32_t cp_mqd_connect_end_wf_count;  /* ordinal92 */
-	uint32_t cp_mqd_connect_end_pq_rptr;  /* ordinal93 */
-	uint32_t cp_mqd_connect_end_pq_wptr;  /* ordinal94 */
-	uint32_t cp_mqd_connect_end_ib_rptr;  /* ordinal95 */
-	uint32_t reserved38;  /* ordinal96 */
-	uint32_t reserved39;  /* ordinal97 */
-	uint32_t cp_mqd_save_start_time_lo;  /* ordinal98 */
-	uint32_t cp_mqd_save_start_time_hi;  /* ordinal99 */
-	uint32_t cp_mqd_save_end_time_lo;  /* ordinal100 */
-	uint32_t cp_mqd_save_end_time_hi;  /* ordinal101 */
-	uint32_t cp_mqd_restore_start_time_lo;  /* ordinal102 */
-	uint32_t cp_mqd_restore_start_time_hi;  /* ordinal103 */
-	uint32_t cp_mqd_restore_end_time_lo;  /* ordinal104 */
-	uint32_t cp_mqd_restore_end_time_hi;  /* ordinal105 */
-	uint32_t reserved40;  /* ordinal106 */
-	uint32_t reserved41;  /* ordinal107 */
-	uint32_t gds_cs_ctxsw_cnt0;  /* ordinal108 */
-	uint32_t gds_cs_ctxsw_cnt1;  /* ordinal109 */
-	uint32_t gds_cs_ctxsw_cnt2;  /* ordinal110 */
-	uint32_t gds_cs_ctxsw_cnt3;  /* ordinal111 */
-	uint32_t reserved42;  /* ordinal112 */
-	uint32_t reserved43;  /* ordinal113 */
-	uint32_t cp_pq_exe_status_lo;  /* ordinal114 */
-	uint32_t cp_pq_exe_status_hi;  /* ordinal115 */
-	uint32_t cp_packet_id_lo;  /* ordinal116 */
-	uint32_t cp_packet_id_hi;  /* ordinal117 */
-	uint32_t cp_packet_exe_status_lo;  /* ordinal118 */
-	uint32_t cp_packet_exe_status_hi;  /* ordinal119 */
-	uint32_t gds_save_base_addr_lo;  /* ordinal120 */
-	uint32_t gds_save_base_addr_hi;  /* ordinal121 */
-	uint32_t gds_save_mask_lo;  /* ordinal122 */
-	uint32_t gds_save_mask_hi;  /* ordinal123 */
-	uint32_t ctx_save_base_addr_lo;  /* ordinal124 */
-	uint32_t ctx_save_base_addr_hi;  /* ordinal125 */
-	uint32_t reserved44;  /* ordinal126 */
-	uint32_t reserved45;  /* ordinal127 */
-	uint32_t cp_mqd_base_addr_lo;  /* ordinal128 */
-	uint32_t cp_mqd_base_addr_hi;  /* ordinal129 */
-	uint32_t cp_hqd_active;  /* ordinal130 */
-	uint32_t cp_hqd_vmid;  /* ordinal131 */
-	uint32_t cp_hqd_persistent_state;  /* ordinal132 */
-	uint32_t cp_hqd_pipe_priority;  /* ordinal133 */
-	uint32_t cp_hqd_queue_priority;  /* ordinal134 */
-	uint32_t cp_hqd_quantum;  /* ordinal135 */
-	uint32_t cp_hqd_pq_base_lo;  /* ordinal136 */
-	uint32_t cp_hqd_pq_base_hi;  /* ordinal137 */
-	uint32_t cp_hqd_pq_rptr;  /* ordinal138 */
-	uint32_t cp_hqd_pq_rptr_report_addr_lo;  /* ordinal139 */
-	uint32_t cp_hqd_pq_rptr_report_addr_hi;  /* ordinal140 */
-	uint32_t cp_hqd_pq_wptr_poll_addr;  /* ordinal141 */
-	uint32_t cp_hqd_pq_wptr_poll_addr_hi;  /* ordinal142 */
-	uint32_t cp_hqd_pq_doorbell_control;  /* ordinal143 */
-	uint32_t cp_hqd_pq_wptr;  /* ordinal144 */
-	uint32_t cp_hqd_pq_control;  /* ordinal145 */
-	uint32_t cp_hqd_ib_base_addr_lo;  /* ordinal146 */
-	uint32_t cp_hqd_ib_base_addr_hi;  /* ordinal147 */
-	uint32_t cp_hqd_ib_rptr;  /* ordinal148 */
-	uint32_t cp_hqd_ib_control;  /* ordinal149 */
-	uint32_t cp_hqd_iq_timer;  /* ordinal150 */
-	uint32_t cp_hqd_iq_rptr;  /* ordinal151 */
-	uint32_t cp_hqd_dequeue_request;  /* ordinal152 */
-	uint32_t cp_hqd_dma_offload;  /* ordinal153 */
-	uint32_t cp_hqd_sema_cmd;  /* ordinal154 */
-	uint32_t cp_hqd_msg_type;  /* ordinal155 */
-	uint32_t cp_hqd_atomic0_preop_lo;  /* ordinal156 */
-	uint32_t cp_hqd_atomic0_preop_hi;  /* ordinal157 */
-	uint32_t cp_hqd_atomic1_preop_lo;  /* ordinal158 */
-	uint32_t cp_hqd_atomic1_preop_hi;  /* ordinal159 */
-	uint32_t cp_hqd_hq_status0;  /* ordinal160 */
-	uint32_t cp_hqd_hq_control0;  /* ordinal161 */
-	uint32_t cp_mqd_control;  /* ordinal162 */
-	uint32_t cp_hqd_hq_status1;  /* ordinal163 */
-	uint32_t cp_hqd_hq_control1;  /* ordinal164 */
-	uint32_t cp_hqd_eop_base_addr_lo;  /* ordinal165 */
-	uint32_t cp_hqd_eop_base_addr_hi;  /* ordinal166 */
-	uint32_t cp_hqd_eop_control;  /* ordinal167 */
-	uint32_t cp_hqd_eop_rptr;  /* ordinal168 */
-	uint32_t cp_hqd_eop_wptr;  /* ordinal169 */
-	uint32_t cp_hqd_eop_done_events;  /* ordinal170 */
-	uint32_t cp_hqd_ctx_save_base_addr_lo;  /* ordinal171 */
-	uint32_t cp_hqd_ctx_save_base_addr_hi;  /* ordinal172 */
-	uint32_t cp_hqd_ctx_save_control;  /* ordinal173 */
-	uint32_t cp_hqd_cntl_stack_offset;  /* ordinal174 */
-	uint32_t cp_hqd_cntl_stack_size;  /* ordinal175 */
-	uint32_t cp_hqd_wg_state_offset;  /* ordinal176 */
-	uint32_t cp_hqd_ctx_save_size;  /* ordinal177 */
-	uint32_t cp_hqd_gds_resource_state;  /* ordinal178 */
-	uint32_t cp_hqd_error;  /* ordinal179 */
-	uint32_t cp_hqd_eop_wptr_mem;  /* ordinal180 */
-	uint32_t cp_hqd_eop_dones;  /* ordinal181 */
-	uint32_t reserved46;  /* ordinal182 */
-	uint32_t reserved47;  /* ordinal183 */
-	uint32_t reserved48;  /* ordinal184 */
-	uint32_t reserved49;  /* ordinal185 */
-	uint32_t reserved50;  /* ordinal186 */
-	uint32_t reserved51;  /* ordinal187 */
-	uint32_t reserved52;  /* ordinal188 */
-	uint32_t reserved53;  /* ordinal189 */
-	uint32_t reserved54;  /* ordinal190 */
-	uint32_t reserved55;  /* ordinal191 */
-	uint32_t iqtimer_pkt_header;  /* ordinal192 */
-	uint32_t iqtimer_pkt_dw0;  /* ordinal193 */
-	uint32_t iqtimer_pkt_dw1;  /* ordinal194 */
-	uint32_t iqtimer_pkt_dw2;  /* ordinal195 */
-	uint32_t iqtimer_pkt_dw3;  /* ordinal196 */
-	uint32_t iqtimer_pkt_dw4;  /* ordinal197 */
-	uint32_t iqtimer_pkt_dw5;  /* ordinal198 */
-	uint32_t iqtimer_pkt_dw6;  /* ordinal199 */
-	uint32_t iqtimer_pkt_dw7;  /* ordinal200 */
-	uint32_t iqtimer_pkt_dw8;  /* ordinal201 */
-	uint32_t iqtimer_pkt_dw9;  /* ordinal202 */
-	uint32_t iqtimer_pkt_dw10;  /* ordinal203 */
-	uint32_t iqtimer_pkt_dw11;  /* ordinal204 */
-	uint32_t iqtimer_pkt_dw12;  /* ordinal205 */
-	uint32_t iqtimer_pkt_dw13;  /* ordinal206 */
-	uint32_t iqtimer_pkt_dw14;  /* ordinal207 */
-	uint32_t iqtimer_pkt_dw15;  /* ordinal208 */
-	uint32_t iqtimer_pkt_dw16;  /* ordinal209 */
-	uint32_t iqtimer_pkt_dw17;  /* ordinal210 */
-	uint32_t iqtimer_pkt_dw18;  /* ordinal211 */
-	uint32_t iqtimer_pkt_dw19;  /* ordinal212 */
-	uint32_t iqtimer_pkt_dw20;  /* ordinal213 */
-	uint32_t iqtimer_pkt_dw21;  /* ordinal214 */
-	uint32_t iqtimer_pkt_dw22;  /* ordinal215 */
-	uint32_t iqtimer_pkt_dw23;  /* ordinal216 */
-	uint32_t iqtimer_pkt_dw24;  /* ordinal217 */
-	uint32_t iqtimer_pkt_dw25;  /* ordinal218 */
-	uint32_t iqtimer_pkt_dw26;  /* ordinal219 */
-	uint32_t iqtimer_pkt_dw27;  /* ordinal220 */
-	uint32_t iqtimer_pkt_dw28;  /* ordinal221 */
-	uint32_t iqtimer_pkt_dw29;  /* ordinal222 */
-	uint32_t iqtimer_pkt_dw30;  /* ordinal223 */
-	uint32_t iqtimer_pkt_dw31;  /* ordinal224 */
-	uint32_t reserved56;  /* ordinal225 */
-	uint32_t reserved57;  /* ordinal226 */
-	uint32_t reserved58;  /* ordinal227 */
-	uint32_t set_resources_header;  /* ordinal228 */
-	uint32_t set_resources_dw1;  /* ordinal229 */
-	uint32_t set_resources_dw2;  /* ordinal230 */
-	uint32_t set_resources_dw3;  /* ordinal231 */
-	uint32_t set_resources_dw4;  /* ordinal232 */
-	uint32_t set_resources_dw5;  /* ordinal233 */
-	uint32_t set_resources_dw6;  /* ordinal234 */
-	uint32_t set_resources_dw7;  /* ordinal235 */
-	uint32_t reserved59;  /* ordinal236 */
-	uint32_t reserved60;  /* ordinal237 */
-	uint32_t reserved61;  /* ordinal238 */
-	uint32_t reserved62;  /* ordinal239 */
-	uint32_t reserved63;  /* ordinal240 */
-	uint32_t reserved64;  /* ordinal241 */
-	uint32_t reserved65;  /* ordinal242 */
-	uint32_t reserved66;  /* ordinal243 */
-	uint32_t reserved67;  /* ordinal244 */
-	uint32_t reserved68;  /* ordinal245 */
-	uint32_t reserved69;  /* ordinal246 */
-	uint32_t reserved70;  /* ordinal247 */
-	uint32_t reserved71;  /* ordinal248 */
-	uint32_t reserved72;  /* ordinal249 */
-	uint32_t reserved73;  /* ordinal250 */
-	uint32_t reserved74;  /* ordinal251 */
-	uint32_t reserved75;  /* ordinal252 */
-	uint32_t reserved76;  /* ordinal253 */
-	uint32_t reserved77;  /* ordinal254 */
-	uint32_t reserved78;  /* ordinal255 */
-
-	uint32_t reserved_t[256]; /* Reserve 256 dword buffer used by ucode */
-};
-
 static void gfx_v8_0_cp_compute_fini(struct amdgpu_device *adev)
 {
 	int i, r;
@@ -4763,34 +4504,7 @@ static int gfx_v8_0_cp_compute_resume(struct amdgpu_device *adev)
 	u32 *buf;
 	struct vi_mqd *mqd;
 
-	/* init the pipes */
-	mutex_lock(&adev->srbm_mutex);
-	for (i = 0; i < (adev->gfx.mec.num_pipe * adev->gfx.mec.num_mec); i++) {
-		int me = (i < 4) ? 1 : 2;
-		int pipe = (i < 4) ? i : (i - 4);
-
-		eop_gpu_addr = adev->gfx.mec.hpd_eop_gpu_addr + (i * MEC_HPD_SIZE);
-		eop_gpu_addr >>= 8;
-
-		vi_srbm_select(adev, me, pipe, 0, 0);
-
-		/* write the EOP addr */
-		WREG32(mmCP_HQD_EOP_BASE_ADDR, eop_gpu_addr);
-		WREG32(mmCP_HQD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr));
-
-		/* set the VMID assigned */
-		WREG32(mmCP_HQD_VMID, 0);
-
-		/* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */
-		tmp = RREG32(mmCP_HQD_EOP_CONTROL);
-		tmp = REG_SET_FIELD(tmp, CP_HQD_EOP_CONTROL, EOP_SIZE,
-				    (order_base_2(MEC_HPD_SIZE / 4) - 1));
-		WREG32(mmCP_HQD_EOP_CONTROL, tmp);
-	}
-	vi_srbm_select(adev, 0, 0, 0, 0);
-	mutex_unlock(&adev->srbm_mutex);
-
-	/* init the queues.  Just two for now. */
+	/* init the queues.  */
 	for (i = 0; i < adev->gfx.num_compute_rings; i++) {
 		struct amdgpu_ring *ring = &adev->gfx.compute_ring[i];
 
@@ -4842,6 +4556,22 @@ static int gfx_v8_0_cp_compute_resume(struct amdgpu_device *adev)
 			       ring->pipe,
 			       ring->queue, 0);
 
+		eop_gpu_addr = adev->gfx.mec.hpd_eop_gpu_addr + (i * MEC_HPD_SIZE);
+		eop_gpu_addr >>= 8;
+
+		/* write the EOP addr */
+		WREG32(mmCP_HQD_EOP_BASE_ADDR, eop_gpu_addr);
+		WREG32(mmCP_HQD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr));
+
+		/* set the VMID assigned */
+		WREG32(mmCP_HQD_VMID, 0);
+
+		/* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */
+		tmp = RREG32(mmCP_HQD_EOP_CONTROL);
+		tmp = REG_SET_FIELD(tmp, CP_HQD_EOP_CONTROL, EOP_SIZE,
+				    (order_base_2(MEC_HPD_SIZE / 4) - 1));
+		WREG32(mmCP_HQD_EOP_CONTROL, tmp);
+
 		/* disable wptr polling */
 		tmp = RREG32(mmCP_PQ_WPTR_POLL_CNTL);
 		tmp = REG_SET_FIELD(tmp, CP_PQ_WPTR_POLL_CNTL, EN, 0);
@@ -4925,9 +4655,9 @@ static int gfx_v8_0_cp_compute_resume(struct amdgpu_device *adev)
 
 		/* only used if CP_PQ_WPTR_POLL_CNTL.CP_PQ_WPTR_POLL_CNTL__EN_MASK=1 */
 		wb_gpu_addr = adev->wb.gpu_addr + (ring->wptr_offs * 4);
-		mqd->cp_hqd_pq_wptr_poll_addr = wb_gpu_addr & 0xfffffffc;
+		mqd->cp_hqd_pq_wptr_poll_addr_lo = wb_gpu_addr & 0xfffffffc;
 		mqd->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits(wb_gpu_addr) & 0xffff;
-		WREG32(mmCP_HQD_PQ_WPTR_POLL_ADDR, mqd->cp_hqd_pq_wptr_poll_addr);
+		WREG32(mmCP_HQD_PQ_WPTR_POLL_ADDR, mqd->cp_hqd_pq_wptr_poll_addr_lo);
 		WREG32(mmCP_HQD_PQ_WPTR_POLL_ADDR_HI,
 		       mqd->cp_hqd_pq_wptr_poll_addr_hi);
 
@@ -5098,6 +4828,10 @@ static int gfx_v8_0_hw_fini(void *handle)
 
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);
+	if (amdgpu_sriov_vf(adev)) {
+		pr_debug("For SRIOV client, shouldn't do anything.\n");
+		return 0;
+	}
 	gfx_v8_0_cp_enable(adev, false);
 	gfx_v8_0_rlc_stop(adev);
 	gfx_v8_0_cp_compute_fini(adev);
@@ -5450,6 +5184,21 @@ static uint32_t wave_read_ind(struct amdgpu_device *adev, uint32_t simd, uint32_
 	return RREG32(mmSQ_IND_DATA);
 }
 
+static void wave_read_regs(struct amdgpu_device *adev, uint32_t simd,
+			   uint32_t wave, uint32_t thread,
+			   uint32_t regno, uint32_t num, uint32_t *out)
+{
+	WREG32(mmSQ_IND_INDEX,
+		(wave << SQ_IND_INDEX__WAVE_ID__SHIFT) |
+		(simd << SQ_IND_INDEX__SIMD_ID__SHIFT) |
+		(regno << SQ_IND_INDEX__INDEX__SHIFT) |
+		(thread << SQ_IND_INDEX__THREAD_ID__SHIFT) |
+		(SQ_IND_INDEX__FORCE_READ_MASK) |
+		(SQ_IND_INDEX__AUTO_INCR_MASK));
+	while (num--)
+		*(out++) = RREG32(mmSQ_IND_DATA);
+}
+
 static void gfx_v8_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t *dst, int *no_fields)
 {
 	/* type 0 wave data */
@@ -5474,11 +5223,21 @@ static void gfx_v8_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, u
 	dst[(*no_fields)++] = wave_read_ind(adev, simd, wave, ixSQ_WAVE_M0);
 }
 
+static void gfx_v8_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd,
+				     uint32_t wave, uint32_t start,
+				     uint32_t size, uint32_t *dst)
+{
+	wave_read_regs(
+		adev, simd, wave, 0,
+		start + SQIND_WAVE_SGPRS_OFFSET, size, dst);
+}
+
 
 static const struct amdgpu_gfx_funcs gfx_v8_0_gfx_funcs = {
 	.get_gpu_clock_counter = &gfx_v8_0_get_gpu_clock_counter,
 	.select_se_sh = &gfx_v8_0_select_se_sh,
 	.read_wave_data = &gfx_v8_0_read_wave_data,
+	.read_wave_sgprs = &gfx_v8_0_read_wave_sgprs,
 };
 
 static int gfx_v8_0_early_init(void *handle)
@@ -5930,29 +5689,24 @@ static void gfx_v8_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev
 	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGCG)) {
-		/* 1 enable cntx_empty_int_enable/cntx_busy_int_enable/
-		 * Cmp_busy/GFX_Idle interrupts
-		 */
-		gfx_v8_0_enable_gui_idle_interrupt(adev, true);
-
 		temp1 = data1 =	RREG32(mmRLC_CGTT_MGCG_OVERRIDE);
 		data1 &= ~RLC_CGTT_MGCG_OVERRIDE__CGCG_MASK;
 		if (temp1 != data1)
 			WREG32(mmRLC_CGTT_MGCG_OVERRIDE, data1);
 
-		/* 2 wait for RLC_SERDES_CU_MASTER & RLC_SERDES_NONCU_MASTER idle */
+		/* : wait for RLC_SERDES_CU_MASTER & RLC_SERDES_NONCU_MASTER idle */
 		gfx_v8_0_wait_for_rlc_serdes(adev);
 
-		/* 3 - clear cgcg override */
+		/* 2 - clear cgcg override */
 		gfx_v8_0_send_serdes_cmd(adev, BPM_REG_CGCG_OVERRIDE, CLE_BPM_SERDES_CMD);
 
 		/* wait for RLC_SERDES_CU_MASTER & RLC_SERDES_NONCU_MASTER idle */
 		gfx_v8_0_wait_for_rlc_serdes(adev);
 
-		/* 4 - write cmd to set CGLS */
+		/* 3 - write cmd to set CGLS */
 		gfx_v8_0_send_serdes_cmd(adev, BPM_REG_CGLS_EN, SET_BPM_SERDES_CMD);
 
-		/* 5 - enable cgcg */
+		/* 4 - enable cgcg */
 		data |= RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK;
 
 		if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGLS) {
@@ -5970,6 +5724,11 @@ static void gfx_v8_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev
 
 		if (temp != data)
 			WREG32(mmRLC_CGCG_CGLS_CTRL, data);
+
+		/* 5 enable cntx_empty_int_enable/cntx_busy_int_enable/
+		 * Cmp_busy/GFX_Idle interrupts
+		 */
+		gfx_v8_0_enable_gui_idle_interrupt(adev, true);
 	} else {
 		/* disable cntx_empty_int_enable & GFX Idle interrupt */
 		gfx_v8_0_enable_gui_idle_interrupt(adev, false);
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 64d3c1e6014c..45a573e63d4a 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -335,7 +335,7 @@ static int gmc_v6_0_mc_init(struct amdgpu_device *adev)
 	 * size equal to the 1024 or vram, whichever is larger.
 	 */
 	if (amdgpu_gart_size == -1)
-		adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev);
+		adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size);
 	else
 		adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20;
 
@@ -795,11 +795,6 @@ static int gmc_v6_0_sw_init(void *handle)
 		return r;
 	}
 
-	r = amdgpu_ttm_global_init(adev);
-	if (r) {
-		return r;
-	}
-
 	r = gmc_v6_0_mc_init(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index fbe1d9ac500a..273b16fb9459 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -385,7 +385,7 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev)
 	 * size equal to the 1024 or vram, whichever is larger.
 	 */
 	if (amdgpu_gart_size == -1)
-		adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev);
+		adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size);
 	else
 		adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20;
 
@@ -945,11 +945,6 @@ static int gmc_v7_0_sw_init(void *handle)
 		return r;
 	}
 
-	r = amdgpu_ttm_global_init(adev);
-	if (r) {
-		return r;
-	}
-
 	r = gmc_v7_0_mc_init(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index 12ea3404dd65..0daac3a5be79 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -472,7 +472,7 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev)
 	 * size equal to the 1024 or vram, whichever is larger.
 	 */
 	if (amdgpu_gart_size == -1)
-		adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev);
+		adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size);
 	else
 		adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20;
 
@@ -952,11 +952,6 @@ static int gmc_v8_0_sw_init(void *handle)
 		return r;
 	}
 
-	r = amdgpu_ttm_global_init(adev);
-	if (r) {
-		return r;
-	}
-
 	r = gmc_v8_0_mc_init(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
index f0f2f6c9718e..6c65a1a2de79 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
@@ -3506,6 +3506,7 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev,
 		    (adev->pdev->revision == 0x80) ||
 		    (adev->pdev->revision == 0x81) ||
 		    (adev->pdev->revision == 0x83) ||
+		    (adev->pdev->revision == 0x87) ||
 		    (adev->pdev->device == 0x6604) ||
 		    (adev->pdev->device == 0x6605)) {
 			max_sclk = 75000;
@@ -7715,6 +7716,7 @@ static int si_dpm_init_microcode(struct amdgpu_device *adev)
 		    (adev->pdev->revision == 0x80) ||
 		    (adev->pdev->revision == 0x81) ||
 		    (adev->pdev->revision == 0x83) ||
+		    (adev->pdev->revision == 0x87) ||
 		    (adev->pdev->device == 0x6604) ||
 		    (adev->pdev->device == 0x6605))
 			chip_name = "oland_k";
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index 95cabeafc18e..a79e283590fb 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -640,7 +640,7 @@ static void uvd_v5_0_enable_clock_gating(struct amdgpu_device *adev, bool enable
 		     UVD_SUVD_CGC_GATE__SDB_MASK;
 
 	if (enable) {
-		data3 |= (UVD_CGC_GATE__SYS_MASK       |
+		data3 |= (UVD_CGC_GATE__SYS_MASK     |
 			UVD_CGC_GATE__UDEC_MASK      |
 			UVD_CGC_GATE__MPEG2_MASK     |
 			UVD_CGC_GATE__RBC_MASK       |
@@ -656,9 +656,11 @@ static void uvd_v5_0_enable_clock_gating(struct amdgpu_device *adev, bool enable
 			UVD_CGC_GATE__UDEC_DB_MASK   |
 			UVD_CGC_GATE__UDEC_MP_MASK   |
 			UVD_CGC_GATE__WCB_MASK       |
-			UVD_CGC_GATE__VCPU_MASK      |
 			UVD_CGC_GATE__JPEG_MASK      |
 			UVD_CGC_GATE__SCPU_MASK);
+		/* only in pg enabled, we can gate clock to vcpu*/
+		if (adev->pg_flags & AMD_PG_SUPPORT_UVD)
+			data3 |= UVD_CGC_GATE__VCPU_MASK;
 		data3 &= ~UVD_CGC_GATE__REGS_MASK;
 		data1 |= suvd_flags;
 	} else {
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index 00fad6951d82..ba0bbf7138dc 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -42,6 +42,10 @@ static void uvd_v6_0_set_irq_funcs(struct amdgpu_device *adev);
 static int uvd_v6_0_start(struct amdgpu_device *adev);
 static void uvd_v6_0_stop(struct amdgpu_device *adev);
 static void uvd_v6_0_set_sw_clock_gating(struct amdgpu_device *adev);
+static int uvd_v6_0_set_clockgating_state(void *handle,
+					  enum amd_clockgating_state state);
+static void uvd_v6_0_enable_mgcg(struct amdgpu_device *adev,
+				 bool enable);
 
 /**
  * uvd_v6_0_ring_get_rptr - get read pointer
@@ -151,8 +155,6 @@ static int uvd_v6_0_hw_init(void *handle)
 	uint32_t tmp;
 	int r;
 
-	amdgpu_asic_set_uvd_clocks(adev, 10000, 10000);
-
 	r = uvd_v6_0_start(adev);
 	if (r)
 		goto done;
@@ -395,11 +397,11 @@ static int uvd_v6_0_start(struct amdgpu_device *adev)
 	lmi_swap_cntl = 0;
 	mp_swap_cntl = 0;
 
+	amdgpu_asic_set_uvd_clocks(adev, 10000, 10000);
+	uvd_v6_0_set_clockgating_state(adev, AMD_CG_STATE_UNGATE);
+	uvd_v6_0_enable_mgcg(adev, true);
 	uvd_v6_0_mc_resume(adev);
 
-	/* disable clock gating */
-	WREG32_FIELD(UVD_CGC_CTRL, DYN_CLOCK_MODE, 0);
-
 	/* disable interupt */
 	WREG32_FIELD(UVD_MASTINT_EN, VCPU_EN, 0);
 
@@ -838,22 +840,72 @@ static int uvd_v6_0_process_interrupt(struct amdgpu_device *adev,
 	return 0;
 }
 
+static void uvd_v6_0_enable_clock_gating(struct amdgpu_device *adev, bool enable)
+{
+	uint32_t data1, data3;
+
+	data1 = RREG32(mmUVD_SUVD_CGC_GATE);
+	data3 = RREG32(mmUVD_CGC_GATE);
+
+	data1 |= UVD_SUVD_CGC_GATE__SRE_MASK |
+		     UVD_SUVD_CGC_GATE__SIT_MASK |
+		     UVD_SUVD_CGC_GATE__SMP_MASK |
+		     UVD_SUVD_CGC_GATE__SCM_MASK |
+		     UVD_SUVD_CGC_GATE__SDB_MASK |
+		     UVD_SUVD_CGC_GATE__SRE_H264_MASK |
+		     UVD_SUVD_CGC_GATE__SRE_HEVC_MASK |
+		     UVD_SUVD_CGC_GATE__SIT_H264_MASK |
+		     UVD_SUVD_CGC_GATE__SIT_HEVC_MASK |
+		     UVD_SUVD_CGC_GATE__SCM_H264_MASK |
+		     UVD_SUVD_CGC_GATE__SCM_HEVC_MASK |
+		     UVD_SUVD_CGC_GATE__SDB_H264_MASK |
+		     UVD_SUVD_CGC_GATE__SDB_HEVC_MASK;
+
+	if (enable) {
+		data3 |= (UVD_CGC_GATE__SYS_MASK       |
+			UVD_CGC_GATE__UDEC_MASK      |
+			UVD_CGC_GATE__MPEG2_MASK     |
+			UVD_CGC_GATE__RBC_MASK       |
+			UVD_CGC_GATE__LMI_MC_MASK    |
+			UVD_CGC_GATE__LMI_UMC_MASK   |
+			UVD_CGC_GATE__IDCT_MASK      |
+			UVD_CGC_GATE__MPRD_MASK      |
+			UVD_CGC_GATE__MPC_MASK       |
+			UVD_CGC_GATE__LBSI_MASK      |
+			UVD_CGC_GATE__LRBBM_MASK     |
+			UVD_CGC_GATE__UDEC_RE_MASK   |
+			UVD_CGC_GATE__UDEC_CM_MASK   |
+			UVD_CGC_GATE__UDEC_IT_MASK   |
+			UVD_CGC_GATE__UDEC_DB_MASK   |
+			UVD_CGC_GATE__UDEC_MP_MASK   |
+			UVD_CGC_GATE__WCB_MASK       |
+			UVD_CGC_GATE__JPEG_MASK      |
+			UVD_CGC_GATE__SCPU_MASK      |
+			UVD_CGC_GATE__JPEG2_MASK);
+		/* only in pg enabled, we can gate clock to vcpu*/
+		if (adev->pg_flags & AMD_PG_SUPPORT_UVD)
+			data3 |= UVD_CGC_GATE__VCPU_MASK;
+
+		data3 &= ~UVD_CGC_GATE__REGS_MASK;
+	} else {
+		data3 = 0;
+	}
+
+	WREG32(mmUVD_SUVD_CGC_GATE, data1);
+	WREG32(mmUVD_CGC_GATE, data3);
+}
+
 static void uvd_v6_0_set_sw_clock_gating(struct amdgpu_device *adev)
 {
-	uint32_t data, data1, data2, suvd_flags;
+	uint32_t data, data2;
 
 	data = RREG32(mmUVD_CGC_CTRL);
-	data1 = RREG32(mmUVD_SUVD_CGC_GATE);
 	data2 = RREG32(mmUVD_SUVD_CGC_CTRL);
 
+
 	data &= ~(UVD_CGC_CTRL__CLK_OFF_DELAY_MASK |
 		  UVD_CGC_CTRL__CLK_GATE_DLY_TIMER_MASK);
 
-	suvd_flags = UVD_SUVD_CGC_GATE__SRE_MASK |
-		     UVD_SUVD_CGC_GATE__SIT_MASK |
-		     UVD_SUVD_CGC_GATE__SMP_MASK |
-		     UVD_SUVD_CGC_GATE__SCM_MASK |
-		     UVD_SUVD_CGC_GATE__SDB_MASK;
 
 	data |= UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK |
 		(1 << REG_FIELD_SHIFT(UVD_CGC_CTRL, CLK_GATE_DLY_TIMER)) |
@@ -886,11 +938,8 @@ static void uvd_v6_0_set_sw_clock_gating(struct amdgpu_device *adev)
 			UVD_SUVD_CGC_CTRL__SMP_MODE_MASK |
 			UVD_SUVD_CGC_CTRL__SCM_MODE_MASK |
 			UVD_SUVD_CGC_CTRL__SDB_MODE_MASK);
-	data1 |= suvd_flags;
 
 	WREG32(mmUVD_CGC_CTRL, data);
-	WREG32(mmUVD_CGC_GATE, 0);
-	WREG32(mmUVD_SUVD_CGC_GATE, data1);
 	WREG32(mmUVD_SUVD_CGC_CTRL, data2);
 }
 
@@ -937,6 +986,32 @@ static void uvd_v6_0_set_hw_clock_gating(struct amdgpu_device *adev)
 }
 #endif
 
+static void uvd_v6_0_enable_mgcg(struct amdgpu_device *adev,
+				 bool enable)
+{
+	u32 orig, data;
+
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG)) {
+		data = RREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL);
+		data |= 0xfff;
+		WREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL, data);
+
+		orig = data = RREG32(mmUVD_CGC_CTRL);
+		data |= UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
+		if (orig != data)
+			WREG32(mmUVD_CGC_CTRL, data);
+	} else {
+		data = RREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL);
+		data &= ~0xfff;
+		WREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL, data);
+
+		orig = data = RREG32(mmUVD_CGC_CTRL);
+		data &= ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
+		if (orig != data)
+			WREG32(mmUVD_CGC_CTRL, data);
+	}
+}
+
 static int uvd_v6_0_set_clockgating_state(void *handle,
 					  enum amd_clockgating_state state)
 {
@@ -947,17 +1022,17 @@ static int uvd_v6_0_set_clockgating_state(void *handle,
 		return 0;
 
 	if (enable) {
-		/* disable HW gating and enable Sw gating */
-		uvd_v6_0_set_sw_clock_gating(adev);
-	} else {
 		/* wait for STATUS to clear */
 		if (uvd_v6_0_wait_for_idle(handle))
 			return -EBUSY;
-
+		uvd_v6_0_enable_clock_gating(adev, true);
 		/* enable HW gates because UVD is idle */
 /*		uvd_v6_0_set_hw_clock_gating(adev); */
+	} else {
+		/* disable HW gating and enable Sw gating */
+		uvd_v6_0_enable_clock_gating(adev, false);
 	}
-
+	uvd_v6_0_set_sw_clock_gating(adev);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 39f03f137a56..6b3293a1c7b8 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -134,7 +134,7 @@ static void vce_v3_0_set_vce_sw_clock_gating(struct amdgpu_device *adev,
 	   accessible but the firmware will throttle the clocks on the
 	   fly as necessary.
 	*/
-	if (gated) {
+	if (!gated) {
 		data = RREG32(mmVCE_CLOCK_GATING_B);
 		data |= 0x1ff;
 		data &= ~0xef0000;
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 243dcf7bae47..9f771f4ffcb7 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -937,12 +937,14 @@ static int vi_common_early_init(void *handle)
 		adev->external_rev_id = adev->rev_id + 0x14;
 		break;
 	case CHIP_POLARIS11:
-		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG;
+		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG |
+			AMD_CG_SUPPORT_VCE_MGCG;
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x5A;
 		break;
 	case CHIP_POLARIS10:
-		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG;
+		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG |
+			AMD_CG_SUPPORT_VCE_MGCG;
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x50;
 		break;
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 51a36077b993..c81cf1412728 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -1004,12 +1004,12 @@ int amd_powerplay_reset(void *handle)
 	if (ret)
 		return ret;
 
-	hw_init_power_state_table(instance->hwmgr);
-
 	if ((amdgpu_dpm == 0)
 		|| cgs_is_virtualization_enabled(instance->smu_mgr->device))
 		return 0;
 
+	hw_init_power_state_table(instance->hwmgr);
+
 	if (eventmgr == NULL || eventmgr->pp_eventmgr_init == NULL)
 		return -EINVAL;
 
diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c
index b6f45fd01fa6..ec36c0e28388 100644
--- a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c
+++ b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c
@@ -154,7 +154,7 @@ int pem_task_powerdown_vce_tasks(struct pp_eventmgr *eventmgr, struct pem_event_
 
 int pem_task_disable_clock_power_gatings_tasks(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data)
 {
-	/* TODO */
+	phm_disable_clock_power_gatings(eventmgr->hwmgr);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/functiontables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/functiontables.c
index a6abe81bc843..71822ae73a12 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/functiontables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/functiontables.c
@@ -35,7 +35,7 @@ static int phm_run_table(struct pp_hwmgr *hwmgr,
 	phm_table_function *function;
 
 	if (rt_table->function_list == NULL) {
-		printk(KERN_INFO "[ powerplay ] this function not implement!\n");
+		pr_debug("[ powerplay ] this function not implement!\n");
 		return 0;
 	}
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index 0723758ed065..c355a0f51663 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -209,6 +209,19 @@ int phm_enable_clock_power_gatings(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+int phm_disable_clock_power_gatings(struct pp_hwmgr *hwmgr)
+{
+	PHM_FUNC_CHECK(hwmgr);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+		PHM_PlatformCaps_TablelessHardwareInterface)) {
+		if (NULL != hwmgr->hwmgr_func->disable_clock_power_gating)
+			return hwmgr->hwmgr_func->disable_clock_power_gating(hwmgr);
+	}
+	return 0;
+}
+
+
 int phm_display_configuration_changed(struct pp_hwmgr *hwmgr)
 {
 	PHM_FUNC_CHECK(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
index d4495839c64c..26129972f686 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
@@ -334,6 +334,7 @@ struct phm_clocks {
 	uint32_t clock[MAX_NUM_CLOCKS];
 };
 
+extern int phm_disable_clock_power_gatings(struct pp_hwmgr *hwmgr);
 extern int phm_enable_clock_power_gatings(struct pp_hwmgr *hwmgr);
 extern int phm_powergate_uvd(struct pp_hwmgr *hwmgr, bool gate);
 extern int phm_powergate_vce(struct pp_hwmgr *hwmgr, bool gate);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c
index 34523fe6ed6f..6aeb1d20cc3b 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c
@@ -1958,6 +1958,12 @@ int fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr)
 	int res;
 	uint64_t tmp64;
 
+	if (hwmgr->thermal_controller.fanInfo.bNoFan) {
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_MicrocodeFanControl);
+		return 0;
+	}
+
 	if (smu_data->smu7_data.fan_table_start == 0) {
 		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_MicrocodeFanControl);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smc.c
index b579f0c175e6..a24971a33bfd 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smc.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smc.c
@@ -2006,6 +2006,12 @@ int iceland_thermal_setup_fan_table(struct pp_hwmgr *hwmgr)
 	if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl))
 		return 0;
 
+	if (hwmgr->thermal_controller.fanInfo.bNoFan) {
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_MicrocodeFanControl);
+		return 0;
+	}
+
 	if (0 == smu7_data->fan_table_start) {
 		phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl);
 		return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c
index 8db8e209d915..5190e821200c 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c
@@ -1885,6 +1885,12 @@ int polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr)
 	int res;
 	uint64_t tmp64;
 
+	if (hwmgr->thermal_controller.fanInfo.bNoFan) {
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_MicrocodeFanControl);
+		return 0;
+	}
+
 	if (smu_data->smu7_data.fan_table_start == 0) {
 		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_MicrocodeFanControl);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 6df0d6edfdd1..f49b5487b951 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -366,12 +366,16 @@ static int smu7_populate_single_firmware_entry(struct pp_smumgr *smumgr,
 				&info);
 
 	if (!result) {
-		entry->version = info.version;
+		entry->version = info.fw_version;
 		entry->id = (uint16_t)fw_type;
 		entry->image_addr_high = smu_upper_32_bits(info.mc_addr);
 		entry->image_addr_low = smu_lower_32_bits(info.mc_addr);
 		entry->meta_data_addr_high = 0;
 		entry->meta_data_addr_low = 0;
+
+		/* digest need be excluded out */
+		if (cgs_is_virtualization_enabled(smumgr->device))
+			info.image_size -= 20;
 		entry->data_size_byte = info.image_size;
 		entry->num_register_entries = 0;
 	}
@@ -403,8 +407,14 @@ int smu7_request_smu_load_fw(struct pp_smumgr *smumgr)
 					0x0);
 
 	if (smumgr->chip_id > CHIP_TOPAZ) { /* add support for Topaz */
-		smu7_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_SMU_DRAM_ADDR_HI, smu_data->smu_buffer.mc_addr_high);
-		smu7_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_SMU_DRAM_ADDR_LO, smu_data->smu_buffer.mc_addr_low);
+		if (!cgs_is_virtualization_enabled(smumgr->device)) {
+			smu7_send_msg_to_smc_with_parameter(smumgr,
+						PPSMC_MSG_SMU_DRAM_ADDR_HI,
+						smu_data->smu_buffer.mc_addr_high);
+			smu7_send_msg_to_smc_with_parameter(smumgr,
+						PPSMC_MSG_SMU_DRAM_ADDR_LO,
+						smu_data->smu_buffer.mc_addr_low);
+		}
 		fw_to_load = UCODE_ID_RLC_G_MASK
 			   + UCODE_ID_SDMA0_MASK
 			   + UCODE_ID_SDMA1_MASK
@@ -539,7 +549,6 @@ int smu7_init(struct pp_smumgr *smumgr)
 	smu_data = (struct smu7_smumgr *)(smumgr->backend);
 	smu_data->header_buffer.data_size =
 			((sizeof(struct SMU_DRAMData_TOC) / 4096) + 1) * 4096;
-	smu_data->smu_buffer.data_size = 200*4096;
 
 /* Allocate FW image data structure and header buffer and
  * send the header buffer address to SMU */
@@ -562,6 +571,10 @@ int smu7_init(struct pp_smumgr *smumgr)
 		(cgs_handle_t)smu_data->header_buffer.handle);
 		return -EINVAL);
 
+	if (cgs_is_virtualization_enabled(smumgr->device))
+		return 0;
+
+	smu_data->smu_buffer.data_size = 200*4096;
 	smu_allocate_memory(smumgr->device,
 		smu_data->smu_buffer.data_size,
 		CGS_GPU_MEM_TYPE__VISIBLE_CONTIG_FB,
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smc.c
index d08f6f19b454..2e1493ce1bb5 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smc.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smc.c
@@ -2496,6 +2496,12 @@ int tonga_thermal_setup_fan_table(struct pp_hwmgr *hwmgr)
 					PHM_PlatformCaps_MicrocodeFanControl))
 		return 0;
 
+	if (hwmgr->thermal_controller.fanInfo.bNoFan) {
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_MicrocodeFanControl);
+		return 0;
+	}
+
 	if (0 == smu_data->smu7_data.fan_table_start) {
 		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 					PHM_PlatformCaps_MicrocodeFanControl);
diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c
index 2a10e24b34b1..fb16070b266e 100644
--- a/drivers/gpu/drm/radeon/radeon_cursor.c
+++ b/drivers/gpu/drm/radeon/radeon_cursor.c
@@ -90,6 +90,9 @@ static void radeon_show_cursor(struct drm_crtc *crtc)
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct radeon_device *rdev = crtc->dev->dev_private;
 
+	if (radeon_crtc->cursor_out_of_bounds)
+		return;
+
 	if (ASIC_IS_DCE4(rdev)) {
 		WREG32(EVERGREEN_CUR_SURFACE_ADDRESS_HIGH + radeon_crtc->crtc_offset,
 		       upper_32_bits(radeon_crtc->cursor_addr));
@@ -143,21 +146,25 @@ static int radeon_cursor_move_locked(struct drm_crtc *crtc, int x, int y)
 	int xorigin = 0, yorigin = 0;
 	int w = radeon_crtc->cursor_width;
 
+	radeon_crtc->cursor_x = x;
+	radeon_crtc->cursor_y = y;
+
 	if (ASIC_IS_AVIVO(rdev)) {
 		/* avivo cursor are offset into the total surface */
 		x += crtc->x;
 		y += crtc->y;
 	}
-	DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y);
 
-	if (x < 0) {
+	if (x < 0)
 		xorigin = min(-x, radeon_crtc->max_cursor_width - 1);
-		x = 0;
-	}
-	if (y < 0) {
+	if (y < 0)
 		yorigin = min(-y, radeon_crtc->max_cursor_height - 1);
-		y = 0;
+
+	if (!ASIC_IS_AVIVO(rdev)) {
+		x += crtc->x;
+		y += crtc->y;
 	}
+	DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y);
 
 	/* fixed on DCE6 and newer */
 	if (ASIC_IS_AVIVO(rdev) && !ASIC_IS_DCE6(rdev)) {
@@ -180,27 +187,31 @@ static int radeon_cursor_move_locked(struct drm_crtc *crtc, int x, int y)
 		if (i > 1) {
 			int cursor_end, frame_end;
 
-			cursor_end = x - xorigin + w;
+			cursor_end = x + w;
 			frame_end = crtc->x + crtc->mode.crtc_hdisplay;
 			if (cursor_end >= frame_end) {
 				w = w - (cursor_end - frame_end);
 				if (!(frame_end & 0x7f))
 					w--;
-			} else {
-				if (!(cursor_end & 0x7f))
-					w--;
+			} else if (cursor_end <= 0) {
+				goto out_of_bounds;
+			} else if (!(cursor_end & 0x7f)) {
+				w--;
 			}
 			if (w <= 0) {
-				w = 1;
-				cursor_end = x - xorigin + w;
-				if (!(cursor_end & 0x7f)) {
-					x--;
-					WARN_ON_ONCE(x < 0);
-				}
+				goto out_of_bounds;
 			}
 		}
 	}
 
+	if (x <= (crtc->x - w) || y <= (crtc->y - radeon_crtc->cursor_height) ||
+	    x >= (crtc->x + crtc->mode.crtc_hdisplay) ||
+	    y >= (crtc->y + crtc->mode.crtc_vdisplay))
+		goto out_of_bounds;
+
+	x += xorigin;
+	y += yorigin;
+
 	if (ASIC_IS_DCE4(rdev)) {
 		WREG32(EVERGREEN_CUR_POSITION + radeon_crtc->crtc_offset, (x << 16) | y);
 		WREG32(EVERGREEN_CUR_HOT_SPOT + radeon_crtc->crtc_offset, (xorigin << 16) | yorigin);
@@ -212,6 +223,9 @@ static int radeon_cursor_move_locked(struct drm_crtc *crtc, int x, int y)
 		WREG32(AVIVO_D1CUR_SIZE + radeon_crtc->crtc_offset,
 		       ((w - 1) << 16) | (radeon_crtc->cursor_height - 1));
 	} else {
+		x -= crtc->x;
+		y -= crtc->y;
+
 		if (crtc->mode.flags & DRM_MODE_FLAG_DBLSCAN)
 			y *= 2;
 
@@ -229,10 +243,20 @@ static int radeon_cursor_move_locked(struct drm_crtc *crtc, int x, int y)
 		       yorigin * 256);
 	}
 
-	radeon_crtc->cursor_x = x;
-	radeon_crtc->cursor_y = y;
+	if (radeon_crtc->cursor_out_of_bounds) {
+		radeon_crtc->cursor_out_of_bounds = false;
+		if (radeon_crtc->cursor_bo)
+			radeon_show_cursor(crtc);
+	}
 
 	return 0;
+
+ out_of_bounds:
+	if (!radeon_crtc->cursor_out_of_bounds) {
+		radeon_hide_cursor(crtc);
+		radeon_crtc->cursor_out_of_bounds = true;
+	}
+	return 0;
 }
 
 int radeon_crtc_cursor_move(struct drm_crtc *crtc,
@@ -297,22 +321,23 @@ int radeon_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	radeon_crtc->cursor_width = width;
-	radeon_crtc->cursor_height = height;
-
 	radeon_lock_cursor(crtc, true);
 
-	if (hot_x != radeon_crtc->cursor_hot_x ||
+	if (width != radeon_crtc->cursor_width ||
+	    height != radeon_crtc->cursor_height ||
+	    hot_x != radeon_crtc->cursor_hot_x ||
 	    hot_y != radeon_crtc->cursor_hot_y) {
 		int x, y;
 
 		x = radeon_crtc->cursor_x + radeon_crtc->cursor_hot_x - hot_x;
 		y = radeon_crtc->cursor_y + radeon_crtc->cursor_hot_y - hot_y;
 
-		radeon_cursor_move_locked(crtc, x, y);
-
+		radeon_crtc->cursor_width = width;
+		radeon_crtc->cursor_height = height;
 		radeon_crtc->cursor_hot_x = hot_x;
 		radeon_crtc->cursor_hot_y = hot_y;
+
+		radeon_cursor_move_locked(crtc, x, y);
 	}
 
 	radeon_show_cursor(crtc);
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index bb75201a24ba..f1da484864a9 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -330,6 +330,7 @@ struct radeon_crtc {
 	u16 lut_r[256], lut_g[256], lut_b[256];
 	bool enabled;
 	bool can_tile;
+	bool cursor_out_of_bounds;
 	uint32_t crtc_offset;
 	struct drm_gem_object *cursor_bo;
 	uint64_t cursor_addr;
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 143280dc0851..ad4d7b8b8322 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -1714,6 +1714,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		    (rdev->pdev->revision == 0x80) ||
 		    (rdev->pdev->revision == 0x81) ||
 		    (rdev->pdev->revision == 0x83) ||
+		    (rdev->pdev->revision == 0x87) ||
 		    (rdev->pdev->device == 0x6604) ||
 		    (rdev->pdev->device == 0x6605))
 			new_smc = true;
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index c49934527a87..8b5e697f2549 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -3026,6 +3026,7 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 		    (rdev->pdev->revision == 0x80) ||
 		    (rdev->pdev->revision == 0x81) ||
 		    (rdev->pdev->revision == 0x83) ||
+		    (rdev->pdev->revision == 0x87) ||
 		    (rdev->pdev->device == 0x6604) ||
 		    (rdev->pdev->device == 0x6605)) {
 			max_sclk = 75000;
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 2191a9e4f3db..396183628f3c 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -522,6 +522,12 @@ struct drm_amdgpu_cs_chunk_data {
 #define AMDGPU_INFO_MEMORY			0x19
 /* Query vce clock table */
 #define AMDGPU_INFO_VCE_CLOCK_TABLE		0x1A
+/* Query vbios related information */
+#define AMDGPU_INFO_VBIOS			0x1B
+	/* Subquery id: Query vbios size */
+	#define AMDGPU_INFO_VBIOS_SIZE		0x1
+	/* Subquery id: Query vbios image */
+	#define AMDGPU_INFO_VBIOS_IMAGE		0x2
 
 #define AMDGPU_INFO_MMR_SE_INDEX_SHIFT	0
 #define AMDGPU_INFO_MMR_SE_INDEX_MASK	0xff
@@ -580,6 +586,11 @@ struct drm_amdgpu_info {
 		} read_mmr_reg;
 
 		struct drm_amdgpu_query_fw query_fw;
+
+		struct {
+			__u32 type;
+			__u32 offset;
+		} vbios_info;
 	};
 };