Merge tag 'v4.10-rc6' into devel

Linux 4.10-rc6

Resolved conflicts in:
	drivers/pinctrl/pinctrl-amd.c
	drivers/pinctrl/samsung/pinctrl-exynos.c

91 files changed, 1188 insertions, 831 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 9ada56c16a58..4c851fde1e82 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -840,6 +840,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 				else if (type == CGS_UCODE_ID_SMU_SK)
 					strcpy(fw_name, "amdgpu/polaris10_smc_sk.bin");
 				break;
+			case CHIP_POLARIS12:
+				strcpy(fw_name, "amdgpu/polaris12_smc.bin");
+				break;
 			default:
 				DRM_ERROR("SMC firmware not supported\n");
 				return -EINVAL;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 29d6d84d1c28..41e41f90265d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -83,6 +83,13 @@ int amdgpu_cs_get_ring(struct amdgpu_device *adev, u32 ip_type,
 		}
 		break;
 	}
+
+	if (!(*out_ring && (*out_ring)->adev)) {
+		DRM_ERROR("Ring %d is not initialized on IP %d\n",
+			  ring, ip_type);
+		return -EINVAL;
+	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 60bd4afe45c8..fe3bb94fe58d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -73,6 +73,7 @@ static const char *amdgpu_asic_name[] = {
 	"STONEY",
 	"POLARIS10",
 	"POLARIS11",
+	"POLARIS12",
 	"LAST",
 };
 
@@ -1277,6 +1278,7 @@ static int amdgpu_early_init(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS12:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
 		if (adev->asic_type == CHIP_CARRIZO || adev->asic_type == CHIP_STONEY)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 8cb937b2bfcc..2534adaebe30 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -418,6 +418,13 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x67CA, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS10},
 	{0x1002, 0x67CC, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS10},
 	{0x1002, 0x67CF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS10},
+	/* Polaris12 */
+	{0x1002, 0x6980, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	{0x1002, 0x6981, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	{0x1002, 0x6985, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	{0x1002, 0x6986, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	{0x1002, 0x6987, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	{0x1002, 0x699F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
 
 	{0, 0, 0}
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c
index fc592c2b0e16..95a568df8551 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c
@@ -98,6 +98,7 @@ static int amdgpu_pp_early_init(void *handle)
 	switch (adev->asic_type) {
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS12:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
 	case CHIP_TOPAZ:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index a81dfaeeb8c0..1d564beb0fde 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -65,6 +65,7 @@
 #define FIRMWARE_STONEY		"amdgpu/stoney_uvd.bin"
 #define FIRMWARE_POLARIS10	"amdgpu/polaris10_uvd.bin"
 #define FIRMWARE_POLARIS11	"amdgpu/polaris11_uvd.bin"
+#define FIRMWARE_POLARIS12	"amdgpu/polaris12_uvd.bin"
 
 /**
  * amdgpu_uvd_cs_ctx - Command submission parser context
@@ -98,6 +99,7 @@ MODULE_FIRMWARE(FIRMWARE_FIJI);
 MODULE_FIRMWARE(FIRMWARE_STONEY);
 MODULE_FIRMWARE(FIRMWARE_POLARIS10);
 MODULE_FIRMWARE(FIRMWARE_POLARIS11);
+MODULE_FIRMWARE(FIRMWARE_POLARIS12);
 
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
 
@@ -149,6 +151,9 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	case CHIP_POLARIS11:
 		fw_name = FIRMWARE_POLARIS11;
 		break;
+	case CHIP_POLARIS12:
+		fw_name = FIRMWARE_POLARIS12;
+		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 69b66b9e7f57..8fec802d3908 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -52,6 +52,7 @@
 #define FIRMWARE_STONEY		"amdgpu/stoney_vce.bin"
 #define FIRMWARE_POLARIS10	"amdgpu/polaris10_vce.bin"
 #define FIRMWARE_POLARIS11         "amdgpu/polaris11_vce.bin"
+#define FIRMWARE_POLARIS12         "amdgpu/polaris12_vce.bin"
 
 #ifdef CONFIG_DRM_AMDGPU_CIK
 MODULE_FIRMWARE(FIRMWARE_BONAIRE);
@@ -66,6 +67,7 @@ MODULE_FIRMWARE(FIRMWARE_FIJI);
 MODULE_FIRMWARE(FIRMWARE_STONEY);
 MODULE_FIRMWARE(FIRMWARE_POLARIS10);
 MODULE_FIRMWARE(FIRMWARE_POLARIS11);
+MODULE_FIRMWARE(FIRMWARE_POLARIS12);
 
 static void amdgpu_vce_idle_work_handler(struct work_struct *work);
 
@@ -121,6 +123,9 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	case CHIP_POLARIS11:
 		fw_name = FIRMWARE_POLARIS11;
 		break;
+	case CHIP_POLARIS12:
+		fw_name = FIRMWARE_POLARIS12;
+		break;
 
 	default:
 		return -EINVAL;
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 9999dc71b998..ccb5e02e7b20 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -2512,6 +2512,8 @@ 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));
 
 	return 0;
 }
@@ -2537,7 +2539,6 @@ 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;
@@ -2578,7 +2579,9 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 	dce_v10_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;
 
@@ -2587,16 +2590,10 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 		dce_v10_0_cursor_move_locked(crtc, x, y);
 
-		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;
+		amdgpu_crtc->cursor_hot_x = hot_x;
+		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
 	dce_v10_0_show_cursor(crtc);
@@ -2620,7 +2617,6 @@ 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);
@@ -2628,10 +2624,6 @@ 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 b3d62b909f43..a7af5b33a5e3 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -167,6 +167,7 @@ static void dce_v11_0_init_golden_registers(struct amdgpu_device *adev)
 						 (const u32)ARRAY_SIZE(stoney_golden_settings_a11));
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		amdgpu_program_register_sequence(adev,
 						 polaris11_golden_settings_a11,
 						 (const u32)ARRAY_SIZE(polaris11_golden_settings_a11));
@@ -608,6 +609,7 @@ static int dce_v11_0_get_num_crtc (struct amdgpu_device *adev)
 		num_crtc = 6;
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		num_crtc = 5;
 		break;
 	default:
@@ -1589,6 +1591,7 @@ static int dce_v11_0_audio_init(struct amdgpu_device *adev)
 		adev->mode_info.audio.num_pins = 8;
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		adev->mode_info.audio.num_pins = 6;
 		break;
 	default:
@@ -2388,7 +2391,8 @@ static u32 dce_v11_0_pick_pll(struct drm_crtc *crtc)
 	int pll;
 
 	if ((adev->asic_type == CHIP_POLARIS10) ||
-	    (adev->asic_type == CHIP_POLARIS11)) {
+	    (adev->asic_type == CHIP_POLARIS11) ||
+	    (adev->asic_type == CHIP_POLARIS12)) {
 		struct amdgpu_encoder *amdgpu_encoder =
 			to_amdgpu_encoder(amdgpu_crtc->encoder);
 		struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
@@ -2528,6 +2532,8 @@ 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));
 
 	return 0;
 }
@@ -2553,7 +2559,6 @@ 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;
@@ -2594,7 +2599,9 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 	dce_v11_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;
 
@@ -2603,16 +2610,10 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 		dce_v11_0_cursor_move_locked(crtc, x, y);
 
-		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;
+		amdgpu_crtc->cursor_hot_x = hot_x;
+		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
 	dce_v11_0_show_cursor(crtc);
@@ -2636,7 +2637,6 @@ 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);
@@ -2644,10 +2644,6 @@ 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);
@@ -2822,7 +2818,8 @@ static int dce_v11_0_crtc_mode_set(struct drm_crtc *crtc,
 		return -EINVAL;
 
 	if ((adev->asic_type == CHIP_POLARIS10) ||
-	    (adev->asic_type == CHIP_POLARIS11)) {
+	    (adev->asic_type == CHIP_POLARIS11) ||
+	    (adev->asic_type == CHIP_POLARIS12)) {
 		struct amdgpu_encoder *amdgpu_encoder =
 			to_amdgpu_encoder(amdgpu_crtc->encoder);
 		int encoder_mode =
@@ -2992,6 +2989,7 @@ static int dce_v11_0_early_init(void *handle)
 		adev->mode_info.num_dig = 6;
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		adev->mode_info.num_hpd = 5;
 		adev->mode_info.num_dig = 5;
 		break;
@@ -3101,7 +3099,8 @@ static int dce_v11_0_hw_init(void *handle)
 	amdgpu_atombios_crtc_powergate_init(adev);
 	amdgpu_atombios_encoder_init_dig(adev);
 	if ((adev->asic_type == CHIP_POLARIS10) ||
-	    (adev->asic_type == CHIP_POLARIS11)) {
+	    (adev->asic_type == CHIP_POLARIS11) ||
+	    (adev->asic_type == CHIP_POLARIS12)) {
 		amdgpu_atombios_crtc_set_dce_clock(adev, adev->clock.default_dispclk,
 						   DCE_CLOCK_TYPE_DISPCLK, ATOM_GCK_DFS);
 		amdgpu_atombios_crtc_set_dce_clock(adev, 0,
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index b4e4ec630e8c..39df6a50637f 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -1859,6 +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;
 
@@ -1878,6 +1880,8 @@ 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));
 
 	return 0;
 }
@@ -1903,7 +1907,6 @@ 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;
@@ -1944,7 +1947,9 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 	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;
 
@@ -1953,16 +1958,10 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 		dce_v6_0_cursor_move_locked(crtc, x, y);
 
-		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;
+		amdgpu_crtc->cursor_hot_x = hot_x;
+		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
 	dce_v6_0_show_cursor(crtc);
@@ -1986,7 +1985,6 @@ 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);
@@ -1994,10 +1992,6 @@ 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 584abe834a3c..28102bb1704d 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -2363,6 +2363,8 @@ 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));
 
 	return 0;
 }
@@ -2388,7 +2390,6 @@ 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;
@@ -2429,7 +2430,9 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 	dce_v8_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;
 
@@ -2438,16 +2441,10 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc,
 
 		dce_v8_0_cursor_move_locked(crtc, x, y);
 
-		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;
+		amdgpu_crtc->cursor_hot_x = hot_x;
+		amdgpu_crtc->cursor_hot_y = hot_y;
 	}
 
 	dce_v8_0_show_cursor(crtc);
@@ -2471,7 +2468,6 @@ 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);
@@ -2479,10 +2475,6 @@ 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/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index 762f8e82ceb7..e9a176891e13 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -627,11 +627,8 @@ static const struct drm_encoder_helper_funcs dce_virtual_encoder_helper_funcs =
 
 static void dce_virtual_encoder_destroy(struct drm_encoder *encoder)
 {
-	struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
-
-	kfree(amdgpu_encoder->enc_priv);
 	drm_encoder_cleanup(encoder);
-	kfree(amdgpu_encoder);
+	kfree(encoder);
 }
 
 static const struct drm_encoder_funcs dce_virtual_encoder_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index d0ec00986f38..373374164bd5 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -139,6 +139,13 @@ MODULE_FIRMWARE("amdgpu/polaris10_mec.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_mec2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/polaris12_ce.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_pfp.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_me.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_mec.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_mec2.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_rlc.bin");
+
 static const struct amdgpu_gds_reg_offset amdgpu_gds_reg_offset[] =
 {
 	{mmGDS_VMID0_BASE, mmGDS_VMID0_SIZE, mmGDS_GWS_VMID0, mmGDS_OA_VMID0},
@@ -689,6 +696,7 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev)
 						 (const u32)ARRAY_SIZE(tonga_golden_common_all));
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		amdgpu_program_register_sequence(adev,
 						 golden_settings_polaris11_a11,
 						 (const u32)ARRAY_SIZE(golden_settings_polaris11_a11));
@@ -903,6 +911,9 @@ static int gfx_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS12:
+		chip_name = "polaris12";
+		break;
 	case CHIP_STONEY:
 		chip_name = "stoney";
 		break;
@@ -1768,6 +1779,7 @@ static int gfx_v8_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config = TONGA_GB_ADDR_CONFIG_GOLDEN;
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		ret = amdgpu_atombios_get_gfx_info(adev);
 		if (ret)
 			return ret;
@@ -2682,6 +2694,7 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev)
 
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		modearray[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
 				PIPE_CONFIG(ADDR_SURF_P4_16x16) |
 				TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
@@ -3503,6 +3516,7 @@ gfx_v8_0_raster_config(struct amdgpu_device *adev, u32 *rconf, u32 *rconf1)
 		*rconf1 |= 0x0;
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		*rconf |= RB_MAP_PKR0(2) | RB_XSEL2(1) | SE_MAP(2) |
 			  SE_XSEL(1) | SE_YSEL(1);
 		*rconf1 |= 0x0;
@@ -4021,7 +4035,8 @@ static void gfx_v8_0_init_pg(struct amdgpu_device *adev)
 			cz_enable_cp_power_gating(adev, true);
 		else
 			cz_enable_cp_power_gating(adev, false);
-	} else if (adev->asic_type == CHIP_POLARIS11) {
+	} else if ((adev->asic_type == CHIP_POLARIS11) ||
+		   (adev->asic_type == CHIP_POLARIS12)) {
 		gfx_v8_0_init_csb(adev);
 		gfx_v8_0_init_save_restore_list(adev);
 		gfx_v8_0_enable_save_restore_machine(adev);
@@ -4095,7 +4110,8 @@ static int gfx_v8_0_rlc_resume(struct amdgpu_device *adev)
 		 RLC_CGCG_CGLS_CTRL__CGLS_EN_MASK);
 	WREG32(mmRLC_CGCG_CGLS_CTRL, tmp);
 	if (adev->asic_type == CHIP_POLARIS11 ||
-	    adev->asic_type == CHIP_POLARIS10) {
+	    adev->asic_type == CHIP_POLARIS10 ||
+	    adev->asic_type == CHIP_POLARIS12) {
 		tmp = RREG32(mmRLC_CGCG_CGLS_CTRL_3D);
 		tmp &= ~0x3;
 		WREG32(mmRLC_CGCG_CGLS_CTRL_3D, tmp);
@@ -4283,6 +4299,7 @@ static int gfx_v8_0_cp_gfx_start(struct amdgpu_device *adev)
 		amdgpu_ring_write(ring, 0x0000002A);
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		amdgpu_ring_write(ring, 0x16000012);
 		amdgpu_ring_write(ring, 0x00000000);
 		break;
@@ -4664,7 +4681,8 @@ static int gfx_v8_0_cp_compute_resume(struct amdgpu_device *adev)
 			    (adev->asic_type == CHIP_FIJI) ||
 			    (adev->asic_type == CHIP_STONEY) ||
 			    (adev->asic_type == CHIP_POLARIS11) ||
-			    (adev->asic_type == CHIP_POLARIS10)) {
+			    (adev->asic_type == CHIP_POLARIS10) ||
+			    (adev->asic_type == CHIP_POLARIS12)) {
 				WREG32(mmCP_MEC_DOORBELL_RANGE_LOWER,
 				       AMDGPU_DOORBELL_KIQ << 2);
 				WREG32(mmCP_MEC_DOORBELL_RANGE_UPPER,
@@ -4700,7 +4718,8 @@ static int gfx_v8_0_cp_compute_resume(struct amdgpu_device *adev)
 		mqd->cp_hqd_persistent_state = tmp;
 		if (adev->asic_type == CHIP_STONEY ||
 			adev->asic_type == CHIP_POLARIS11 ||
-			adev->asic_type == CHIP_POLARIS10) {
+			adev->asic_type == CHIP_POLARIS10 ||
+			adev->asic_type == CHIP_POLARIS12) {
 			tmp = RREG32(mmCP_ME1_PIPE3_INT_CNTL);
 			tmp = REG_SET_FIELD(tmp, CP_ME1_PIPE3_INT_CNTL, GENERIC2_INT_ENABLE, 1);
 			WREG32(mmCP_ME1_PIPE3_INT_CNTL, tmp);
@@ -5279,7 +5298,8 @@ static int gfx_v8_0_late_init(void *handle)
 static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev,
 						       bool enable)
 {
-	if (adev->asic_type == CHIP_POLARIS11)
+	if ((adev->asic_type == CHIP_POLARIS11) ||
+	    (adev->asic_type == CHIP_POLARIS12))
 		/* Send msg to SMU via Powerplay */
 		amdgpu_set_powergating_state(adev,
 					     AMD_IP_BLOCK_TYPE_SMC,
@@ -5353,6 +5373,7 @@ static int gfx_v8_0_set_powergating_state(void *handle,
 			gfx_v8_0_enable_gfx_dynamic_mg_power_gating(adev, false);
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG) && enable)
 			gfx_v8_0_enable_gfx_static_mg_power_gating(adev, true);
 		else
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 45a573e63d4a..e2b0b1646f99 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -44,6 +44,7 @@ MODULE_FIRMWARE("radeon/tahiti_mc.bin");
 MODULE_FIRMWARE("radeon/pitcairn_mc.bin");
 MODULE_FIRMWARE("radeon/verde_mc.bin");
 MODULE_FIRMWARE("radeon/oland_mc.bin");
+MODULE_FIRMWARE("radeon/si58_mc.bin");
 
 #define MC_SEQ_MISC0__MT__MASK   0xf0000000
 #define MC_SEQ_MISC0__MT__GDDR1  0x10000000
@@ -113,6 +114,7 @@ static int gmc_v6_0_init_microcode(struct amdgpu_device *adev)
 	const char *chip_name;
 	char fw_name[30];
 	int err;
+	bool is_58_fw = false;
 
 	DRM_DEBUG("\n");
 
@@ -135,7 +137,14 @@ static int gmc_v6_0_init_microcode(struct amdgpu_device *adev)
 	default: BUG();
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+	/* this memory configuration requires special firmware */
+	if (((RREG32(mmMC_SEQ_MISC0) & 0xff000000) >> 24) == 0x58)
+		is_58_fw = true;
+
+	if (is_58_fw)
+		snprintf(fw_name, sizeof(fw_name), "radeon/si58_mc.bin");
+	else
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
 	err = request_firmware(&adev->mc.fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -463,19 +472,11 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev)
 	WREG32(mmVM_CONTEXT1_CNTL,
 	       VM_CONTEXT1_CNTL__ENABLE_CONTEXT_MASK |
 	       (1UL << VM_CONTEXT1_CNTL__PAGE_TABLE_DEPTH__SHIFT) |
-	       ((amdgpu_vm_block_size - 9) << VM_CONTEXT1_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT) |
-	       VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
-	       VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK |
-	       VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
-	       VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK |
-	       VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
-	       VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK |
-	       VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
-	       VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK |
-	       VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
-	       VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_DEFAULT_MASK |
-	       VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
-	       VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK);
+	       ((amdgpu_vm_block_size - 9) << VM_CONTEXT1_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT));
+	if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS)
+		gmc_v6_0_set_fault_enable_default(adev, false);
+	else
+		gmc_v6_0_set_fault_enable_default(adev, true);
 
 	gmc_v6_0_gart_flush_gpu_tlb(adev, 0);
 	dev_info(adev->dev, "PCIE GART of %uM enabled (table at 0x%016llX).\n",
@@ -754,7 +755,10 @@ static int gmc_v6_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0);
+	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
+		return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0);
+	else
+		return 0;
 }
 
 static int gmc_v6_0_sw_init(void *handle)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index 0daac3a5be79..476bc9f1954b 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -46,6 +46,7 @@ static int gmc_v8_0_wait_for_idle(void *handle);
 MODULE_FIRMWARE("amdgpu/tonga_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_mc.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_mc.bin");
 
 static const u32 golden_settings_tonga_a11[] =
 {
@@ -130,6 +131,7 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev)
 						 (const u32)ARRAY_SIZE(golden_settings_tonga_a11));
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		amdgpu_program_register_sequence(adev,
 						 golden_settings_polaris11_a11,
 						 (const u32)ARRAY_SIZE(golden_settings_polaris11_a11));
@@ -225,6 +227,9 @@ static int gmc_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS12:
+		chip_name = "polaris12";
+		break;
 	case CHIP_FIJI:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
index 1170a64a3184..034ace79ed49 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -60,6 +60,8 @@ MODULE_FIRMWARE("amdgpu/polaris10_sdma.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_sdma.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_sdma.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_sdma1.bin");
 
 
 static const u32 sdma_offsets[SDMA_MAX_INSTANCE] =
@@ -206,6 +208,7 @@ static void sdma_v3_0_init_golden_registers(struct amdgpu_device *adev)
 						 (const u32)ARRAY_SIZE(golden_settings_tonga_a11));
 		break;
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		amdgpu_program_register_sequence(adev,
 						 golden_settings_polaris11_a11,
 						 (const u32)ARRAY_SIZE(golden_settings_polaris11_a11));
@@ -278,6 +281,9 @@ static int sdma_v3_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS12:
+		chip_name = "polaris12";
+		break;
 	case CHIP_CARRIZO:
 		chip_name = "carrizo";
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
index 6c65a1a2de79..6e150db8f380 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
@@ -56,7 +56,6 @@
 #define BIOS_SCRATCH_4                                    0x5cd
 
 MODULE_FIRMWARE("radeon/tahiti_smc.bin");
-MODULE_FIRMWARE("radeon/tahiti_k_smc.bin");
 MODULE_FIRMWARE("radeon/pitcairn_smc.bin");
 MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin");
 MODULE_FIRMWARE("radeon/verde_smc.bin");
@@ -65,6 +64,7 @@ MODULE_FIRMWARE("radeon/oland_smc.bin");
 MODULE_FIRMWARE("radeon/oland_k_smc.bin");
 MODULE_FIRMWARE("radeon/hainan_smc.bin");
 MODULE_FIRMWARE("radeon/hainan_k_smc.bin");
+MODULE_FIRMWARE("radeon/banks_k_2_smc.bin");
 
 union power_info {
 	struct _ATOM_POWERPLAY_INFO info;
@@ -3488,30 +3488,6 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev,
 		    (adev->pdev->device == 0x6817) ||
 		    (adev->pdev->device == 0x6806))
 			max_mclk = 120000;
-	} else if (adev->asic_type == CHIP_VERDE) {
-		if ((adev->pdev->revision == 0x81) ||
-		    (adev->pdev->revision == 0x83) ||
-		    (adev->pdev->revision == 0x87) ||
-		    (adev->pdev->device == 0x6820) ||
-		    (adev->pdev->device == 0x6821) ||
-		    (adev->pdev->device == 0x6822) ||
-		    (adev->pdev->device == 0x6823) ||
-		    (adev->pdev->device == 0x682A) ||
-		    (adev->pdev->device == 0x682B)) {
-			max_sclk = 75000;
-			max_mclk = 80000;
-		}
-	} else if (adev->asic_type == CHIP_OLAND) {
-		if ((adev->pdev->revision == 0xC7) ||
-		    (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;
-			max_mclk = 80000;
-		}
 	} else if (adev->asic_type == CHIP_HAINAN) {
 		if ((adev->pdev->revision == 0x81) ||
 		    (adev->pdev->revision == 0x83) ||
@@ -3520,7 +3496,6 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev,
 		    (adev->pdev->device == 0x6665) ||
 		    (adev->pdev->device == 0x6667)) {
 			max_sclk = 75000;
-			max_mclk = 80000;
 		}
 	}
 	/* Apply dpm quirks */
@@ -7687,50 +7662,51 @@ static int si_dpm_init_microcode(struct amdgpu_device *adev)
 		chip_name = "tahiti";
 		break;
 	case CHIP_PITCAIRN:
-		if ((adev->pdev->revision == 0x81) ||
-		    (adev->pdev->device == 0x6810) ||
-		    (adev->pdev->device == 0x6811) ||
-		    (adev->pdev->device == 0x6816) ||
-		    (adev->pdev->device == 0x6817) ||
-		    (adev->pdev->device == 0x6806))
+		if ((adev->pdev->revision == 0x81) &&
+		    ((adev->pdev->device == 0x6810) ||
+		    (adev->pdev->device == 0x6811)))
 			chip_name = "pitcairn_k";
 		else
 			chip_name = "pitcairn";
 		break;
 	case CHIP_VERDE:
-		if ((adev->pdev->revision == 0x81) ||
-		    (adev->pdev->revision == 0x83) ||
-		    (adev->pdev->revision == 0x87) ||
-		    (adev->pdev->device == 0x6820) ||
-		    (adev->pdev->device == 0x6821) ||
-		    (adev->pdev->device == 0x6822) ||
-		    (adev->pdev->device == 0x6823) ||
-		    (adev->pdev->device == 0x682A) ||
-		    (adev->pdev->device == 0x682B))
+		if (((adev->pdev->device == 0x6820) &&
+			((adev->pdev->revision == 0x81) ||
+			(adev->pdev->revision == 0x83))) ||
+		    ((adev->pdev->device == 0x6821) &&
+			((adev->pdev->revision == 0x83) ||
+			(adev->pdev->revision == 0x87))) ||
+		    ((adev->pdev->revision == 0x87) &&
+			((adev->pdev->device == 0x6823) ||
+			(adev->pdev->device == 0x682b))))
 			chip_name = "verde_k";
 		else
 			chip_name = "verde";
 		break;
 	case CHIP_OLAND:
-		if ((adev->pdev->revision == 0xC7) ||
-		    (adev->pdev->revision == 0x80) ||
-		    (adev->pdev->revision == 0x81) ||
-		    (adev->pdev->revision == 0x83) ||
-		    (adev->pdev->revision == 0x87) ||
-		    (adev->pdev->device == 0x6604) ||
-		    (adev->pdev->device == 0x6605))
+		if (((adev->pdev->revision == 0x81) &&
+			((adev->pdev->device == 0x6600) ||
+			(adev->pdev->device == 0x6604) ||
+			(adev->pdev->device == 0x6605) ||
+			(adev->pdev->device == 0x6610))) ||
+		    ((adev->pdev->revision == 0x83) &&
+			(adev->pdev->device == 0x6610)))
 			chip_name = "oland_k";
 		else
 			chip_name = "oland";
 		break;
 	case CHIP_HAINAN:
-		if ((adev->pdev->revision == 0x81) ||
-		    (adev->pdev->revision == 0x83) ||
-		    (adev->pdev->revision == 0xC3) ||
-		    (adev->pdev->device == 0x6664) ||
-		    (adev->pdev->device == 0x6665) ||
-		    (adev->pdev->device == 0x6667))
+		if (((adev->pdev->revision == 0x81) &&
+			(adev->pdev->device == 0x6660)) ||
+		    ((adev->pdev->revision == 0x83) &&
+			((adev->pdev->device == 0x6660) ||
+			(adev->pdev->device == 0x6663) ||
+			(adev->pdev->device == 0x6665) ||
+			 (adev->pdev->device == 0x6667))))
 			chip_name = "hainan_k";
+		else if ((adev->pdev->revision == 0xc3) &&
+			 (adev->pdev->device == 0x6665))
+			chip_name = "banks_k_2";
 		else
 			chip_name = "hainan";
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index 96444e4d862a..7fb9137dd89b 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -40,13 +40,14 @@
 #include "smu/smu_7_0_1_sh_mask.h"
 
 static void uvd_v4_2_mc_resume(struct amdgpu_device *adev);
-static void uvd_v4_2_init_cg(struct amdgpu_device *adev);
 static void uvd_v4_2_set_ring_funcs(struct amdgpu_device *adev);
 static void uvd_v4_2_set_irq_funcs(struct amdgpu_device *adev);
 static int uvd_v4_2_start(struct amdgpu_device *adev);
 static void uvd_v4_2_stop(struct amdgpu_device *adev);
 static int uvd_v4_2_set_clockgating_state(void *handle,
 				enum amd_clockgating_state state);
+static void uvd_v4_2_set_dcm(struct amdgpu_device *adev,
+			     bool sw_mode);
 /**
  * uvd_v4_2_ring_get_rptr - get read pointer
  *
@@ -140,7 +141,8 @@ static int uvd_v4_2_sw_fini(void *handle)
 
 	return r;
 }
-
+static void uvd_v4_2_enable_mgcg(struct amdgpu_device *adev,
+				 bool enable);
 /**
  * uvd_v4_2_hw_init - start and test UVD block
  *
@@ -155,8 +157,7 @@ static int uvd_v4_2_hw_init(void *handle)
 	uint32_t tmp;
 	int r;
 
-	uvd_v4_2_init_cg(adev);
-	uvd_v4_2_set_clockgating_state(adev, AMD_CG_STATE_GATE);
+	uvd_v4_2_enable_mgcg(adev, true);
 	amdgpu_asic_set_uvd_clocks(adev, 10000, 10000);
 	r = uvd_v4_2_start(adev);
 	if (r)
@@ -266,11 +267,13 @@ static int uvd_v4_2_start(struct amdgpu_device *adev)
 	struct amdgpu_ring *ring = &adev->uvd.ring;
 	uint32_t rb_bufsz;
 	int i, j, r;
-
 	/* disable byte swapping */
 	u32 lmi_swap_cntl = 0;
 	u32 mp_swap_cntl = 0;
 
+	WREG32(mmUVD_CGC_GATE, 0);
+	uvd_v4_2_set_dcm(adev, true);
+
 	uvd_v4_2_mc_resume(adev);
 
 	/* disable interupt */
@@ -406,6 +409,8 @@ static void uvd_v4_2_stop(struct amdgpu_device *adev)
 
 	/* Unstall UMC and register bus */
 	WREG32_P(mmUVD_LMI_CTRL2, 0, ~(1 << 8));
+
+	uvd_v4_2_set_dcm(adev, false);
 }
 
 /**
@@ -619,19 +624,6 @@ static void uvd_v4_2_set_dcm(struct amdgpu_device *adev,
 	WREG32_UVD_CTX(ixUVD_CGC_CTRL2, tmp2);
 }
 
-static void uvd_v4_2_init_cg(struct amdgpu_device *adev)
-{
-	bool hw_mode = true;
-
-	if (hw_mode) {
-		uvd_v4_2_set_dcm(adev, false);
-	} else {
-		u32 tmp = RREG32(mmUVD_CGC_CTRL);
-		tmp &= ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
-		WREG32(mmUVD_CGC_CTRL, tmp);
-	}
-}
-
 static bool uvd_v4_2_is_idle(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@@ -685,17 +677,6 @@ static int uvd_v4_2_process_interrupt(struct amdgpu_device *adev,
 static int uvd_v4_2_set_clockgating_state(void *handle,
 					  enum amd_clockgating_state state)
 {
-	bool gate = false;
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
-	if (!(adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG))
-		return 0;
-
-	if (state == AMD_CG_STATE_GATE)
-		gate = true;
-
-	uvd_v4_2_enable_mgcg(adev, gate);
-
 	return 0;
 }
 
@@ -711,9 +692,6 @@ static int uvd_v4_2_set_powergating_state(void *handle,
 	 */
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!(adev->pg_flags & AMD_PG_SUPPORT_UVD))
-		return 0;
-
 	if (state == AMD_PG_STATE_GATE) {
 		uvd_v4_2_stop(adev);
 		return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index a79e283590fb..6de6becce745 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -791,15 +791,10 @@ static int uvd_v5_0_set_clockgating_state(void *handle,
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
-	static int curstate = -1;
 
 	if (!(adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG))
 		return 0;
 
-	if (curstate == state)
-		return 0;
-
-	curstate = state;
 	if (enable) {
 		/* wait for STATUS to clear */
 		if (uvd_v5_0_wait_for_idle(handle))
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 6b3293a1c7b8..37ca685e5a9a 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -43,9 +43,13 @@
 
 #define GRBM_GFX_INDEX__VCE_INSTANCE__SHIFT	0x04
 #define GRBM_GFX_INDEX__VCE_INSTANCE_MASK	0x10
+#define GRBM_GFX_INDEX__VCE_ALL_PIPE		0x07
+
 #define mmVCE_LMI_VCPU_CACHE_40BIT_BAR0	0x8616
 #define mmVCE_LMI_VCPU_CACHE_40BIT_BAR1	0x8617
 #define mmVCE_LMI_VCPU_CACHE_40BIT_BAR2	0x8618
+#define mmGRBM_GFX_INDEX_DEFAULT 0xE0000000
+
 #define VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK	0x02
 
 #define VCE_V3_0_FW_SIZE	(384 * 1024)
@@ -54,6 +58,9 @@
 
 #define FW_52_8_3	((52 << 24) | (8 << 16) | (3 << 8))
 
+#define GET_VCE_INSTANCE(i)  ((i) << GRBM_GFX_INDEX__VCE_INSTANCE__SHIFT \
+					| GRBM_GFX_INDEX__VCE_ALL_PIPE)
+
 static void vce_v3_0_mc_resume(struct amdgpu_device *adev, int idx);
 static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev);
 static void vce_v3_0_set_irq_funcs(struct amdgpu_device *adev);
@@ -175,7 +182,7 @@ static void vce_v3_0_set_vce_sw_clock_gating(struct amdgpu_device *adev,
 		WREG32(mmVCE_UENC_CLOCK_GATING_2, data);
 
 		data = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
-		data &= ~0xffc00000;
+		data &= ~0x3ff;
 		WREG32(mmVCE_UENC_REG_CLOCK_GATING, data);
 
 		data = RREG32(mmVCE_UENC_DMA_DCLK_CTRL);
@@ -249,7 +256,7 @@ static int vce_v3_0_start(struct amdgpu_device *adev)
 		if (adev->vce.harvest_config & (1 << idx))
 			continue;
 
-		WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, idx);
+		WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(idx));
 		vce_v3_0_mc_resume(adev, idx);
 		WREG32_FIELD(VCE_STATUS, JOB_BUSY, 1);
 
@@ -273,7 +280,7 @@ static int vce_v3_0_start(struct amdgpu_device *adev)
 		}
 	}
 
-	WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0);
+	WREG32(mmGRBM_GFX_INDEX, mmGRBM_GFX_INDEX_DEFAULT);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	return 0;
@@ -288,7 +295,7 @@ static int vce_v3_0_stop(struct amdgpu_device *adev)
 		if (adev->vce.harvest_config & (1 << idx))
 			continue;
 
-		WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, idx);
+		WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(idx));
 
 		if (adev->asic_type >= CHIP_STONEY)
 			WREG32_P(mmVCE_VCPU_CNTL, 0, ~0x200001);
@@ -306,7 +313,7 @@ static int vce_v3_0_stop(struct amdgpu_device *adev)
 			vce_v3_0_set_vce_sw_clock_gating(adev, false);
 	}
 
-	WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0);
+	WREG32(mmGRBM_GFX_INDEX, mmGRBM_GFX_INDEX_DEFAULT);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	return 0;
@@ -320,11 +327,12 @@ static unsigned vce_v3_0_get_harvest_config(struct amdgpu_device *adev)
 {
 	u32 tmp;
 
-	/* Fiji, Stoney, Polaris10, Polaris11 are single pipe */
+	/* Fiji, Stoney, Polaris10, Polaris11, Polaris12 are single pipe */
 	if ((adev->asic_type == CHIP_FIJI) ||
 	    (adev->asic_type == CHIP_STONEY) ||
 	    (adev->asic_type == CHIP_POLARIS10) ||
-	    (adev->asic_type == CHIP_POLARIS11))
+	    (adev->asic_type == CHIP_POLARIS11) ||
+	    (adev->asic_type == CHIP_POLARIS12))
 		return AMDGPU_VCE_HARVEST_VCE1;
 
 	/* Tonga and CZ are dual or single pipe */
@@ -585,17 +593,17 @@ static bool vce_v3_0_check_soft_reset(void *handle)
 	 * VCE team suggest use bit 3--bit 6 for busy status check
 	 */
 	mutex_lock(&adev->grbm_idx_mutex);
-	WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0);
+	WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(0));
 	if (RREG32(mmVCE_STATUS) & AMDGPU_VCE_STATUS_BUSY_MASK) {
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE0, 1);
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE1, 1);
 	}
-	WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0x10);
+	WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1));
 	if (RREG32(mmVCE_STATUS) & AMDGPU_VCE_STATUS_BUSY_MASK) {
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE0, 1);
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE1, 1);
 	}
-	WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0);
+	WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(0));
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	if (srbm_soft_reset) {
@@ -733,7 +741,7 @@ static int vce_v3_0_set_clockgating_state(void *handle,
 		if (adev->vce.harvest_config & (1 << i))
 			continue;
 
-		WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, i);
+		WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(i));
 
 		if (enable) {
 			/* initialize VCE_CLOCK_GATING_A: Clock ON/OFF delay */
@@ -752,7 +760,7 @@ static int vce_v3_0_set_clockgating_state(void *handle,
 		vce_v3_0_set_vce_sw_clock_gating(adev, enable);
 	}
 
-	WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0);
+	WREG32(mmGRBM_GFX_INDEX, mmGRBM_GFX_INDEX_DEFAULT);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index bf088d6d9bf1..c2ac54f11341 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -88,6 +88,7 @@ MODULE_FIRMWARE("amdgpu/polaris10_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_smc_sk.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_smc_sk.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_smc.bin");
 
 /*
  * Indirect registers accessor
@@ -312,6 +313,7 @@ static void vi_init_golden_registers(struct amdgpu_device *adev)
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS12:
 	default:
 		break;
 	}
@@ -671,6 +673,7 @@ static int vi_read_register(struct amdgpu_device *adev, u32 se_num,
 	case CHIP_TONGA:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS12:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
 		asic_register_table = cz_allowed_read_registers;
@@ -994,6 +997,11 @@ static int vi_common_early_init(void *handle)
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x50;
 		break;
+	case CHIP_POLARIS12:
+		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG;
+		adev->pg_flags = 0;
+		adev->external_rev_id = adev->rev_id + 0x64;
+		break;
 	case CHIP_CARRIZO:
 		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG |
 			AMD_CG_SUPPORT_GFX_MGCG |
@@ -1346,6 +1354,7 @@ static int vi_common_set_clockgating_state(void *handle,
 	case CHIP_TONGA:
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
 		vi_common_set_clockgating_state_by_smu(adev, state);
 	default:
 		break;
@@ -1429,6 +1438,7 @@ int vi_set_ip_blocks(struct amdgpu_device *adev)
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS12:
 		amdgpu_ip_block_add(adev, &vi_common_ip_block);
 		amdgpu_ip_block_add(adev, &gmc_v8_1_ip_block);
 		amdgpu_ip_block_add(adev, &tonga_ih_ip_block);
diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index c02469ada9f1..85f358764bbc 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -23,7 +23,7 @@
 #ifndef __AMD_SHARED_H__
 #define __AMD_SHARED_H__
 
-#define AMD_MAX_USEC_TIMEOUT		100000  /* 100 ms */
+#define AMD_MAX_USEC_TIMEOUT		200000  /* 200 ms */
 
 /*
  * Supported ASIC types
@@ -46,6 +46,7 @@ enum amd_asic_type {
 	CHIP_STONEY,
 	CHIP_POLARIS10,
 	CHIP_POLARIS11,
+	CHIP_POLARIS12,
 	CHIP_LAST,
 };
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c
index b0c63c5f54c9..6bb79c94cb9f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c
@@ -200,7 +200,7 @@ int cz_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 				cgs_set_clockgating_state(
 							hwmgr->device,
 							AMD_IP_BLOCK_TYPE_VCE,
-							AMD_CG_STATE_UNGATE);
+							AMD_CG_STATE_GATE);
 				cgs_set_powergating_state(
 							hwmgr->device,
 							AMD_IP_BLOCK_TYPE_VCE,
@@ -218,7 +218,7 @@ int cz_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 				cgs_set_clockgating_state(
 							hwmgr->device,
 							AMD_IP_BLOCK_TYPE_VCE,
-							AMD_PG_STATE_GATE);
+							AMD_PG_STATE_UNGATE);
 				cz_dpm_update_vce_dpm(hwmgr);
 				cz_enable_disable_vce_dpm(hwmgr, true);
 				return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c
index 4b14f259a147..0fb4e8c8f5e1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c
@@ -1402,14 +1402,22 @@ int  cz_dpm_update_vce_dpm(struct pp_hwmgr *hwmgr)
 					     cz_hwmgr->vce_dpm.hard_min_clk,
 						PPSMC_MSG_SetEclkHardMin));
 	} else {
-		/*EPR# 419220 -HW limitation to to */
-		cz_hwmgr->vce_dpm.hard_min_clk = hwmgr->vce_arbiter.ecclk;
-		smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
-					    PPSMC_MSG_SetEclkHardMin,
-					    cz_get_eclk_level(hwmgr,
-				     cz_hwmgr->vce_dpm.hard_min_clk,
-					  PPSMC_MSG_SetEclkHardMin));
-
+		/*Program HardMin based on the vce_arbiter.ecclk */
+		if (hwmgr->vce_arbiter.ecclk == 0) {
+			smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+					    PPSMC_MSG_SetEclkHardMin, 0);
+		/* disable ECLK DPM 0. Otherwise VCE could hang if
+		 * switching SCLK from DPM 0 to 6/7 */
+			smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+					PPSMC_MSG_SetEclkSoftMin, 1);
+		} else {
+			cz_hwmgr->vce_dpm.hard_min_clk = hwmgr->vce_arbiter.ecclk;
+			smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+						PPSMC_MSG_SetEclkHardMin,
+						cz_get_eclk_level(hwmgr,
+						cz_hwmgr->vce_dpm.hard_min_clk,
+						PPSMC_MSG_SetEclkHardMin));
+		}
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index dc6700aee18f..b03606405a53 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -95,6 +95,7 @@ int hwmgr_init(struct amd_pp_init *pp_init, struct pp_instance *handle)
 			break;
 		case CHIP_POLARIS11:
 		case CHIP_POLARIS10:
+		case CHIP_POLARIS12:
 			polaris_set_asic_special_caps(hwmgr);
 			hwmgr->feature_mask &= ~(PP_UVD_HANDSHAKE_MASK);
 			break;
@@ -745,7 +746,7 @@ int polaris_set_asic_special_caps(struct pp_hwmgr *hwmgr)
 	phm_cap_set(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_TablelessHardwareInterface);
 
-	if (hwmgr->chip_id == CHIP_POLARIS11)
+	if ((hwmgr->chip_id == CHIP_POLARIS11) || (hwmgr->chip_id == CHIP_POLARIS12))
 		phm_cap_set(hwmgr->platform_descriptor.platformCaps,
 					PHM_PlatformCaps_SPLLShutdownSupport);
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index 26477f0f09dc..6cd1287a7a8f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -521,7 +521,7 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris10);
 				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
-			} else if (hwmgr->chip_id == CHIP_POLARIS11) {
+			} else if ((hwmgr->chip_id == CHIP_POLARIS11) || (hwmgr->chip_id == CHIP_POLARIS12)) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11);
 				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
index e5812aa456f3..6e618aa20719 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
@@ -65,6 +65,7 @@ int smum_init(struct amd_pp_init *pp_init, struct pp_instance *handle)
 			break;
 		case CHIP_POLARIS11:
 		case CHIP_POLARIS10:
+		case CHIP_POLARIS12:
 			polaris10_smum_init(smumgr);
 			break;
 		default:
diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h
index 908011d2c8f5..7abda94fc2cf 100644
--- a/drivers/gpu/drm/ast/ast_drv.h
+++ b/drivers/gpu/drm/ast/ast_drv.h
@@ -113,6 +113,7 @@ struct ast_private {
 	struct ttm_bo_kmap_obj cache_kmap;
 	int next_cursor;
 	bool support_wide_screen;
+	bool DisableP2A;
 
 	enum ast_tx_chip tx_chip_type;
 	u8 dp501_maxclk;
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index f75c6421db62..533e762d036d 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -124,6 +124,12 @@ static int ast_detect_chip(struct drm_device *dev, bool *need_post)
 	} else
 		*need_post = false;
 
+	/* Check P2A Access */
+	ast->DisableP2A = true;
+	data = ast_read32(ast, 0xf004);
+	if (data != 0xFFFFFFFF)
+		ast->DisableP2A = false;
+
 	/* Check if we support wide screen */
 	switch (ast->chip) {
 	case AST1180:
@@ -140,15 +146,17 @@ static int ast_detect_chip(struct drm_device *dev, bool *need_post)
 			ast->support_wide_screen = true;
 		else {
 			ast->support_wide_screen = false;
-			/* Read SCU7c (silicon revision register) */
-			ast_write32(ast, 0xf004, 0x1e6e0000);
-			ast_write32(ast, 0xf000, 0x1);
-			data = ast_read32(ast, 0x1207c);
-			data &= 0x300;
-			if (ast->chip == AST2300 && data == 0x0) /* ast1300 */
-				ast->support_wide_screen = true;
-			if (ast->chip == AST2400 && data == 0x100) /* ast1400 */
-				ast->support_wide_screen = true;
+			if (ast->DisableP2A == false) {
+				/* Read SCU7c (silicon revision register) */
+				ast_write32(ast, 0xf004, 0x1e6e0000);
+				ast_write32(ast, 0xf000, 0x1);
+				data = ast_read32(ast, 0x1207c);
+				data &= 0x300;
+				if (ast->chip == AST2300 && data == 0x0) /* ast1300 */
+					ast->support_wide_screen = true;
+				if (ast->chip == AST2400 && data == 0x100) /* ast1400 */
+					ast->support_wide_screen = true;
+			}
 		}
 		break;
 	}
@@ -216,80 +224,81 @@ static int ast_get_dram_info(struct drm_device *dev)
 	uint32_t data, data2;
 	uint32_t denum, num, div, ref_pll;
 
-	ast_write32(ast, 0xf004, 0x1e6e0000);
-	ast_write32(ast, 0xf000, 0x1);
-
-
-	ast_write32(ast, 0x10000, 0xfc600309);
-
-	do {
-		if (pci_channel_offline(dev->pdev))
-			return -EIO;
-	} while (ast_read32(ast, 0x10000) != 0x01);
-	data = ast_read32(ast, 0x10004);
-
-	if (data & 0x40)
+	if (ast->DisableP2A)
+	{
 		ast->dram_bus_width = 16;
+		ast->dram_type = AST_DRAM_1Gx16;
+		ast->mclk = 396;
+	}
 	else
-		ast->dram_bus_width = 32;
+	{
+		ast_write32(ast, 0xf004, 0x1e6e0000);
+		ast_write32(ast, 0xf000, 0x1);
+		data = ast_read32(ast, 0x10004);
+
+		if (data & 0x40)
+			ast->dram_bus_width = 16;
+		else
+			ast->dram_bus_width = 32;
+
+		if (ast->chip == AST2300 || ast->chip == AST2400) {
+			switch (data & 0x03) {
+			case 0:
+				ast->dram_type = AST_DRAM_512Mx16;
+				break;
+			default:
+			case 1:
+				ast->dram_type = AST_DRAM_1Gx16;
+				break;
+			case 2:
+				ast->dram_type = AST_DRAM_2Gx16;
+				break;
+			case 3:
+				ast->dram_type = AST_DRAM_4Gx16;
+				break;
+			}
+		} else {
+			switch (data & 0x0c) {
+			case 0:
+			case 4:
+				ast->dram_type = AST_DRAM_512Mx16;
+				break;
+			case 8:
+				if (data & 0x40)
+					ast->dram_type = AST_DRAM_1Gx16;
+				else
+					ast->dram_type = AST_DRAM_512Mx32;
+				break;
+			case 0xc:
+				ast->dram_type = AST_DRAM_1Gx32;
+				break;
+			}
+		}
 
-	if (ast->chip == AST2300 || ast->chip == AST2400) {
-		switch (data & 0x03) {
-		case 0:
-			ast->dram_type = AST_DRAM_512Mx16;
-			break;
-		default:
-		case 1:
-			ast->dram_type = AST_DRAM_1Gx16;
-			break;
-		case 2:
-			ast->dram_type = AST_DRAM_2Gx16;
-			break;
+		data = ast_read32(ast, 0x10120);
+		data2 = ast_read32(ast, 0x10170);
+		if (data2 & 0x2000)
+			ref_pll = 14318;
+		else
+			ref_pll = 12000;
+
+		denum = data & 0x1f;
+		num = (data & 0x3fe0) >> 5;
+		data = (data & 0xc000) >> 14;
+		switch (data) {
 		case 3:
-			ast->dram_type = AST_DRAM_4Gx16;
-			break;
-		}
-	} else {
-		switch (data & 0x0c) {
-		case 0:
-		case 4:
-			ast->dram_type = AST_DRAM_512Mx16;
+			div = 0x4;
 			break;
-		case 8:
-			if (data & 0x40)
-				ast->dram_type = AST_DRAM_1Gx16;
-			else
-				ast->dram_type = AST_DRAM_512Mx32;
+		case 2:
+		case 1:
+			div = 0x2;
 			break;
-		case 0xc:
-			ast->dram_type = AST_DRAM_1Gx32;
+		default:
+			div = 0x1;
 			break;
 		}
+		ast->mclk = ref_pll * (num + 2) / (denum + 2) * (div * 1000);
 	}
-
-	data = ast_read32(ast, 0x10120);
-	data2 = ast_read32(ast, 0x10170);
-	if (data2 & 0x2000)
-		ref_pll = 14318;
-	else
-		ref_pll = 12000;
-
-	denum = data & 0x1f;
-	num = (data & 0x3fe0) >> 5;
-	data = (data & 0xc000) >> 14;
-	switch (data) {
-	case 3:
-		div = 0x4;
-		break;
-	case 2:
-	case 1:
-		div = 0x2;
-		break;
-	default:
-		div = 0x1;
-		break;
-	}
-	ast->mclk = ref_pll * (num + 2) / (denum + 2) * (div * 1000);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/ast/ast_post.c b/drivers/gpu/drm/ast/ast_post.c
index 810c51d92b99..5331ee1df086 100644
--- a/drivers/gpu/drm/ast/ast_post.c
+++ b/drivers/gpu/drm/ast/ast_post.c
@@ -379,12 +379,20 @@ void ast_post_gpu(struct drm_device *dev)
 	ast_open_key(ast);
 	ast_set_def_ext_reg(dev);
 
-	if (ast->chip == AST2300 || ast->chip == AST2400)
-		ast_init_dram_2300(dev);
-	else
-		ast_init_dram_reg(dev);
+	if (ast->DisableP2A == false)
+	{
+		if (ast->chip == AST2300 || ast->chip == AST2400)
+			ast_init_dram_2300(dev);
+		else
+			ast_init_dram_reg(dev);
 
-	ast_init_3rdtx(dev);
+		ast_init_3rdtx(dev);
+	}
+	else
+	{
+		if (ast->tx_chip_type != AST_TX_NONE)
+			ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xcf, 0x80);	/* Enable DVO */
+	}
 }
 
 /* AST 2300 DRAM settings */
diff --git a/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c b/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
index eb9bf8786c24..18eefdcbf1ba 100644
--- a/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
+++ b/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
@@ -1382,6 +1382,7 @@ int analogix_dp_bind(struct device *dev, struct drm_device *drm_dev,
 
 	pm_runtime_enable(dev);
 
+	pm_runtime_get_sync(dev);
 	phy_power_on(dp->phy);
 
 	analogix_dp_init_dp(dp);
@@ -1414,9 +1415,15 @@ int analogix_dp_bind(struct device *dev, struct drm_device *drm_dev,
 		goto err_disable_pm_runtime;
 	}
 
+	phy_power_off(dp->phy);
+	pm_runtime_put(dev);
+
 	return 0;
 
 err_disable_pm_runtime:
+
+	phy_power_off(dp->phy);
+	pm_runtime_put(dev);
 	pm_runtime_disable(dev);
 
 	return ret;
diff --git a/drivers/gpu/drm/cirrus/Kconfig b/drivers/gpu/drm/cirrus/Kconfig
index 04b3c161dfae..7f4cc6e172ab 100644
--- a/drivers/gpu/drm/cirrus/Kconfig
+++ b/drivers/gpu/drm/cirrus/Kconfig
@@ -7,3 +7,12 @@ config DRM_CIRRUS_QEMU
 	 This is a KMS driver for emulated cirrus device in qemu.
 	 It is *NOT* intended for real cirrus devices. This requires
 	 the modesetting userspace X.org driver.
+
+	 Cirrus is obsolete, the hardware was designed in the 90ies
+	 and can't keep up with todays needs.  More background:
+	 https://www.kraxel.org/blog/2014/10/qemu-using-cirrus-considered-harmful/
+
+	 Better alternatives are:
+	   - stdvga (DRM_BOCHS, qemu -vga std, default in qemu 2.2+)
+	   - qxl (DRM_QXL, qemu -vga qxl, works best with spice)
+	   - virtio (DRM_VIRTIO_GPU), qemu -vga virtio)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 60697482b94c..50f5cf7b69d1 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -291,15 +291,15 @@ drm_atomic_get_crtc_state(struct drm_atomic_state *state,
 EXPORT_SYMBOL(drm_atomic_get_crtc_state);
 
 static void set_out_fence_for_crtc(struct drm_atomic_state *state,
-				   struct drm_crtc *crtc, s64 __user *fence_ptr)
+				   struct drm_crtc *crtc, s32 __user *fence_ptr)
 {
 	state->crtcs[drm_crtc_index(crtc)].out_fence_ptr = fence_ptr;
 }
 
-static s64 __user *get_out_fence_for_crtc(struct drm_atomic_state *state,
+static s32 __user *get_out_fence_for_crtc(struct drm_atomic_state *state,
 					  struct drm_crtc *crtc)
 {
-	s64 __user *fence_ptr;
+	s32 __user *fence_ptr;
 
 	fence_ptr = state->crtcs[drm_crtc_index(crtc)].out_fence_ptr;
 	state->crtcs[drm_crtc_index(crtc)].out_fence_ptr = NULL;
@@ -512,7 +512,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc,
 		state->color_mgmt_changed |= replaced;
 		return ret;
 	} else if (property == config->prop_out_fence_ptr) {
-		s64 __user *fence_ptr = u64_to_user_ptr(val);
+		s32 __user *fence_ptr = u64_to_user_ptr(val);
 
 		if (!fence_ptr)
 			return 0;
@@ -1915,7 +1915,7 @@ EXPORT_SYMBOL(drm_atomic_clean_old_fb);
  */
 
 struct drm_out_fence_state {
-	s64 __user *out_fence_ptr;
+	s32 __user *out_fence_ptr;
 	struct sync_file *sync_file;
 	int fd;
 };
@@ -1952,7 +1952,7 @@ static int prepare_crtc_signaling(struct drm_device *dev,
 		return 0;
 
 	for_each_crtc_in_state(state, crtc, crtc_state, i) {
-		u64 __user *fence_ptr;
+		s32 __user *fence_ptr;
 
 		fence_ptr = get_out_fence_for_crtc(crtc_state->state, crtc);
 
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
index 583f47f27b36..34f757bcabae 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -1259,8 +1259,10 @@ int drm_atomic_helper_commit(struct drm_device *dev,
 
 	if (!nonblock) {
 		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
-		if (ret)
+		if (ret) {
+			drm_atomic_helper_cleanup_planes(dev, state);
 			return ret;
+		}
 	}
 
 	/*
diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c
index ac6a35212501..e6b19bc9021a 100644
--- a/drivers/gpu/drm/drm_modes.c
+++ b/drivers/gpu/drm/drm_modes.c
@@ -1460,6 +1460,13 @@ drm_mode_create_from_cmdline_mode(struct drm_device *dev,
 		return NULL;
 
 	mode->type |= DRM_MODE_TYPE_USERDEF;
+	/* fix up 1368x768: GFT/CVT can't express 1366 width due to alignment */
+	if (cmd->xres == 1366 && mode->hdisplay == 1368) {
+		mode->hdisplay = 1366;
+		mode->hsync_start--;
+		mode->hsync_end--;
+		drm_mode_set_name(mode);
+	}
 	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V);
 	return mode;
 }
diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
index ac953f037be7..cf8f0128c161 100644
--- a/drivers/gpu/drm/drm_probe_helper.c
+++ b/drivers/gpu/drm/drm_probe_helper.c
@@ -143,8 +143,18 @@ void drm_kms_helper_poll_enable_locked(struct drm_device *dev)
 	}
 
 	if (dev->mode_config.delayed_event) {
+		/*
+		 * FIXME:
+		 *
+		 * Use short (1s) delay to handle the initial delayed event.
+		 * This delay should not be needed, but Optimus/nouveau will
+		 * fail in a mysterious way if the delayed event is handled as
+		 * soon as possible like it is done in
+		 * drm_helper_probe_single_connector_modes() in case the poll
+		 * was enabled before.
+		 */
 		poll = true;
-		delay = 0;
+		delay = HZ;
 	}
 
 	if (poll)
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
index 169ac96e8f08..fe0e85b41310 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
@@ -116,9 +116,14 @@ static int etnaviv_iommu_find_iova(struct etnaviv_iommu *mmu,
 		struct list_head list;
 		bool found;
 
+		/*
+		 * XXX: The DRM_MM_SEARCH_BELOW is really a hack to trick
+		 * drm_mm into giving out a low IOVA after address space
+		 * rollover. This needs a proper fix.
+		 */
 		ret = drm_mm_insert_node_in_range(&mmu->mm, node,
 			size, 0, mmu->last_iova, ~0UL,
-			DRM_MM_SEARCH_DEFAULT);
+			mmu->last_iova ? DRM_MM_SEARCH_DEFAULT : DRM_MM_SEARCH_BELOW);
 
 		if (ret != -ENOSPC)
 			break;
diff --git a/drivers/gpu/drm/exynos/exynos5433_drm_decon.c b/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
index 6ca1f3117fe8..75eeb831ed6a 100644
--- a/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
+++ b/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
@@ -46,7 +46,8 @@ enum decon_flag_bits {
 	BIT_CLKS_ENABLED,
 	BIT_IRQS_ENABLED,
 	BIT_WIN_UPDATED,
-	BIT_SUSPENDED
+	BIT_SUSPENDED,
+	BIT_REQUEST_UPDATE
 };
 
 struct decon_context {
@@ -141,12 +142,6 @@ static void decon_commit(struct exynos_drm_crtc *crtc)
 		m->crtc_vsync_end = m->crtc_vsync_start + 1;
 	}
 
-	decon_set_bits(ctx, DECON_VIDCON0, VIDCON0_ENVID, 0);
-
-	/* enable clock gate */
-	val = CMU_CLKGAGE_MODE_SFR_F | CMU_CLKGAGE_MODE_MEM_F;
-	writel(val, ctx->addr + DECON_CMU);
-
 	if (ctx->out_type & (IFTYPE_I80 | I80_HW_TRG))
 		decon_setup_trigger(ctx);
 
@@ -315,6 +310,7 @@ static void decon_update_plane(struct exynos_drm_crtc *crtc,
 
 	/* window enable */
 	decon_set_bits(ctx, DECON_WINCONx(win), WINCONx_ENWIN_F, ~0);
+	set_bit(BIT_REQUEST_UPDATE, &ctx->flags);
 }
 
 static void decon_disable_plane(struct exynos_drm_crtc *crtc,
@@ -327,6 +323,7 @@ static void decon_disable_plane(struct exynos_drm_crtc *crtc,
 		return;
 
 	decon_set_bits(ctx, DECON_WINCONx(win), WINCONx_ENWIN_F, 0);
+	set_bit(BIT_REQUEST_UPDATE, &ctx->flags);
 }
 
 static void decon_atomic_flush(struct exynos_drm_crtc *crtc)
@@ -340,8 +337,8 @@ static void decon_atomic_flush(struct exynos_drm_crtc *crtc)
 	for (i = ctx->first_win; i < WINDOWS_NR; i++)
 		decon_shadow_protect_win(ctx, i, false);
 
-	/* standalone update */
-	decon_set_bits(ctx, DECON_UPDATE, STANDALONE_UPDATE_F, ~0);
+	if (test_and_clear_bit(BIT_REQUEST_UPDATE, &ctx->flags))
+		decon_set_bits(ctx, DECON_UPDATE, STANDALONE_UPDATE_F, ~0);
 
 	if (ctx->out_type & IFTYPE_I80)
 		set_bit(BIT_WIN_UPDATED, &ctx->flags);
diff --git a/drivers/gpu/drm/i915/gvt/aperture_gm.c b/drivers/gpu/drm/i915/gvt/aperture_gm.c
index 0d41ebc4aea6..f7bce8603958 100644
--- a/drivers/gpu/drm/i915/gvt/aperture_gm.c
+++ b/drivers/gpu/drm/i915/gvt/aperture_gm.c
@@ -37,13 +37,6 @@
 #include "i915_drv.h"
 #include "gvt.h"
 
-#define MB_TO_BYTES(mb) ((mb) << 20ULL)
-#define BYTES_TO_MB(b) ((b) >> 20ULL)
-
-#define HOST_LOW_GM_SIZE MB_TO_BYTES(128)
-#define HOST_HIGH_GM_SIZE MB_TO_BYTES(384)
-#define HOST_FENCE 4
-
 static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
@@ -165,6 +158,14 @@ void intel_vgpu_write_fence(struct intel_vgpu *vgpu,
 	POSTING_READ(fence_reg_lo);
 }
 
+static void _clear_vgpu_fence(struct intel_vgpu *vgpu)
+{
+	int i;
+
+	for (i = 0; i < vgpu_fence_sz(vgpu); i++)
+		intel_vgpu_write_fence(vgpu, i, 0);
+}
+
 static void free_vgpu_fence(struct intel_vgpu *vgpu)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
@@ -178,9 +179,9 @@ static void free_vgpu_fence(struct intel_vgpu *vgpu)
 	intel_runtime_pm_get(dev_priv);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
+	_clear_vgpu_fence(vgpu);
 	for (i = 0; i < vgpu_fence_sz(vgpu); i++) {
 		reg = vgpu->fence.regs[i];
-		intel_vgpu_write_fence(vgpu, i, 0);
 		list_add_tail(&reg->link,
 			      &dev_priv->mm.fence_list);
 	}
@@ -208,13 +209,14 @@ static int alloc_vgpu_fence(struct intel_vgpu *vgpu)
 			continue;
 		list_del(pos);
 		vgpu->fence.regs[i] = reg;
-		intel_vgpu_write_fence(vgpu, i, 0);
 		if (++i == vgpu_fence_sz(vgpu))
 			break;
 	}
 	if (i != vgpu_fence_sz(vgpu))
 		goto out_free_fence;
 
+	_clear_vgpu_fence(vgpu);
+
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 	intel_runtime_pm_put(dev_priv);
 	return 0;
@@ -314,6 +316,22 @@ void intel_vgpu_free_resource(struct intel_vgpu *vgpu)
 }
 
 /**
+ * intel_vgpu_reset_resource - reset resource state owned by a vGPU
+ * @vgpu: a vGPU
+ *
+ * This function is used to reset resource state owned by a vGPU.
+ *
+ */
+void intel_vgpu_reset_resource(struct intel_vgpu *vgpu)
+{
+	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+
+	intel_runtime_pm_get(dev_priv);
+	_clear_vgpu_fence(vgpu);
+	intel_runtime_pm_put(dev_priv);
+}
+
+/**
  * intel_alloc_vgpu_resource - allocate HW resource for a vGPU
  * @vgpu: vGPU
  * @param: vGPU creation params
diff --git a/drivers/gpu/drm/i915/gvt/cfg_space.c b/drivers/gpu/drm/i915/gvt/cfg_space.c
index db516382a4d4..4a6a2ed65732 100644
--- a/drivers/gpu/drm/i915/gvt/cfg_space.c
+++ b/drivers/gpu/drm/i915/gvt/cfg_space.c
@@ -123,6 +123,7 @@ static int emulate_pci_command_write(struct intel_vgpu *vgpu,
 	u8 changed = old ^ new;
 	int ret;
 
+	memcpy(vgpu_cfg_space(vgpu) + offset, p_data, bytes);
 	if (!(changed & PCI_COMMAND_MEMORY))
 		return 0;
 
@@ -142,7 +143,6 @@ static int emulate_pci_command_write(struct intel_vgpu *vgpu,
 			return ret;
 	}
 
-	memcpy(vgpu_cfg_space(vgpu) + offset, p_data, bytes);
 	return 0;
 }
 
@@ -240,7 +240,7 @@ int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset,
 	if (WARN_ON(bytes > 4))
 		return -EINVAL;
 
-	if (WARN_ON(offset + bytes >= INTEL_GVT_MAX_CFG_SPACE_SZ))
+	if (WARN_ON(offset + bytes > INTEL_GVT_MAX_CFG_SPACE_SZ))
 		return -EINVAL;
 
 	/* First check if it's PCI_COMMAND */
@@ -282,3 +282,77 @@ int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset,
 	}
 	return 0;
 }
+
+/**
+ * intel_vgpu_init_cfg_space - init vGPU configuration space when create vGPU
+ *
+ * @vgpu: a vGPU
+ * @primary: is the vGPU presented as primary
+ *
+ */
+void intel_vgpu_init_cfg_space(struct intel_vgpu *vgpu,
+			       bool primary)
+{
+	struct intel_gvt *gvt = vgpu->gvt;
+	const struct intel_gvt_device_info *info = &gvt->device_info;
+	u16 *gmch_ctl;
+	int i;
+
+	memcpy(vgpu_cfg_space(vgpu), gvt->firmware.cfg_space,
+	       info->cfg_space_size);
+
+	if (!primary) {
+		vgpu_cfg_space(vgpu)[PCI_CLASS_DEVICE] =
+			INTEL_GVT_PCI_CLASS_VGA_OTHER;
+		vgpu_cfg_space(vgpu)[PCI_CLASS_PROG] =
+			INTEL_GVT_PCI_CLASS_VGA_OTHER;
+	}
+
+	/* Show guest that there isn't any stolen memory.*/
+	gmch_ctl = (u16 *)(vgpu_cfg_space(vgpu) + INTEL_GVT_PCI_GMCH_CONTROL);
+	*gmch_ctl &= ~(BDW_GMCH_GMS_MASK << BDW_GMCH_GMS_SHIFT);
+
+	intel_vgpu_write_pci_bar(vgpu, PCI_BASE_ADDRESS_2,
+				 gvt_aperture_pa_base(gvt), true);
+
+	vgpu_cfg_space(vgpu)[PCI_COMMAND] &= ~(PCI_COMMAND_IO
+					     | PCI_COMMAND_MEMORY
+					     | PCI_COMMAND_MASTER);
+	/*
+	 * Clear the bar upper 32bit and let guest to assign the new value
+	 */
+	memset(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_1, 0, 4);
+	memset(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_3, 0, 4);
+	memset(vgpu_cfg_space(vgpu) + INTEL_GVT_PCI_OPREGION, 0, 4);
+
+	for (i = 0; i < INTEL_GVT_MAX_BAR_NUM; i++) {
+		vgpu->cfg_space.bar[i].size = pci_resource_len(
+					      gvt->dev_priv->drm.pdev, i * 2);
+		vgpu->cfg_space.bar[i].tracked = false;
+	}
+}
+
+/**
+ * intel_vgpu_reset_cfg_space - reset vGPU configuration space
+ *
+ * @vgpu: a vGPU
+ *
+ */
+void intel_vgpu_reset_cfg_space(struct intel_vgpu *vgpu)
+{
+	u8 cmd = vgpu_cfg_space(vgpu)[PCI_COMMAND];
+	bool primary = vgpu_cfg_space(vgpu)[PCI_CLASS_DEVICE] !=
+				INTEL_GVT_PCI_CLASS_VGA_OTHER;
+
+	if (cmd & PCI_COMMAND_MEMORY) {
+		trap_gttmmio(vgpu, false);
+		map_aperture(vgpu, false);
+	}
+
+	/**
+	 * Currently we only do such reset when vGPU is not
+	 * owned by any VM, so we simply restore entire cfg
+	 * space to default value.
+	 */
+	intel_vgpu_init_cfg_space(vgpu, primary);
+}
diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index d26a092c70e8..e4563984cb1e 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -481,7 +481,6 @@ struct parser_exec_state {
 	(s->vgpu->gvt->device_info.gmadr_bytes_in_cmd >> 2)
 
 static unsigned long bypass_scan_mask = 0;
-static bool bypass_batch_buffer_scan = true;
 
 /* ring ALL, type = 0 */
 static struct sub_op_bits sub_op_mi[] = {
@@ -1525,9 +1524,6 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s)
 {
 	struct intel_gvt *gvt = s->vgpu->gvt;
 
-	if (bypass_batch_buffer_scan)
-		return 0;
-
 	if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) {
 		/* BDW decides privilege based on address space */
 		if (cmd_val(s, 0) & (1 << 8))
diff --git a/drivers/gpu/drm/i915/gvt/execlist.c b/drivers/gpu/drm/i915/gvt/execlist.c
index f32bb6f6495c..34083731669d 100644
--- a/drivers/gpu/drm/i915/gvt/execlist.c
+++ b/drivers/gpu/drm/i915/gvt/execlist.c
@@ -364,58 +364,30 @@ static void free_workload(struct intel_vgpu_workload *workload)
 #define get_desc_from_elsp_dwords(ed, i) \
 	((struct execlist_ctx_descriptor_format *)&((ed)->data[i * 2]))
 
-
-#define BATCH_BUFFER_ADDR_MASK ((1UL << 32) - (1U << 2))
-#define BATCH_BUFFER_ADDR_HIGH_MASK ((1UL << 16) - (1U))
-static int set_gma_to_bb_cmd(struct intel_shadow_bb_entry *entry_obj,
-			     unsigned long add, int gmadr_bytes)
-{
-	if (WARN_ON(gmadr_bytes != 4 && gmadr_bytes != 8))
-		return -1;
-
-	*((u32 *)(entry_obj->bb_start_cmd_va + (1 << 2))) = add &
-		BATCH_BUFFER_ADDR_MASK;
-	if (gmadr_bytes == 8) {
-		*((u32 *)(entry_obj->bb_start_cmd_va + (2 << 2))) =
-			add & BATCH_BUFFER_ADDR_HIGH_MASK;
-	}
-
-	return 0;
-}
-
 static void prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 {
-	int gmadr_bytes = workload->vgpu->gvt->device_info.gmadr_bytes_in_cmd;
+	const int gmadr_bytes = workload->vgpu->gvt->device_info.gmadr_bytes_in_cmd;
+	struct intel_shadow_bb_entry *entry_obj;
 
 	/* pin the gem object to ggtt */
-	if (!list_empty(&workload->shadow_bb)) {
-		struct intel_shadow_bb_entry *entry_obj =
-			list_first_entry(&workload->shadow_bb,
-					 struct intel_shadow_bb_entry,
-					 list);
-		struct intel_shadow_bb_entry *temp;
+	list_for_each_entry(entry_obj, &workload->shadow_bb, list) {
+		struct i915_vma *vma;
 
-		list_for_each_entry_safe(entry_obj, temp, &workload->shadow_bb,
-				list) {
-			struct i915_vma *vma;
-
-			vma = i915_gem_object_ggtt_pin(entry_obj->obj, NULL, 0,
-						       4, 0);
-			if (IS_ERR(vma)) {
-				gvt_err("Cannot pin\n");
-				return;
-			}
-
-			/* FIXME: we are not tracking our pinned VMA leaving it
-			 * up to the core to fix up the stray pin_count upon
-			 * free.
-			 */
-
-			/* update the relocate gma with shadow batch buffer*/
-			set_gma_to_bb_cmd(entry_obj,
-					  i915_ggtt_offset(vma),
-					  gmadr_bytes);
+		vma = i915_gem_object_ggtt_pin(entry_obj->obj, NULL, 0, 4, 0);
+		if (IS_ERR(vma)) {
+			gvt_err("Cannot pin\n");
+			return;
 		}
+
+		/* FIXME: we are not tracking our pinned VMA leaving it
+		 * up to the core to fix up the stray pin_count upon
+		 * free.
+		 */
+
+		/* update the relocate gma with shadow batch buffer*/
+		entry_obj->bb_start_cmd_va[1] = i915_ggtt_offset(vma);
+		if (gmadr_bytes == 8)
+			entry_obj->bb_start_cmd_va[2] = 0;
 	}
 }
 
@@ -826,7 +798,7 @@ int intel_vgpu_init_execlist(struct intel_vgpu *vgpu)
 		INIT_LIST_HEAD(&vgpu->workload_q_head[i]);
 	}
 
-	vgpu->workloads = kmem_cache_create("gvt-g vgpu workload",
+	vgpu->workloads = kmem_cache_create("gvt-g_vgpu_workload",
 			sizeof(struct intel_vgpu_workload), 0,
 			SLAB_HWCACHE_ALIGN,
 			NULL);
diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
index 7eaaf1c9ed2b..47dec4acf7ff 100644
--- a/drivers/gpu/drm/i915/gvt/gtt.c
+++ b/drivers/gpu/drm/i915/gvt/gtt.c
@@ -240,15 +240,8 @@ static inline int get_pse_type(int type)
 static u64 read_pte64(struct drm_i915_private *dev_priv, unsigned long index)
 {
 	void __iomem *addr = (gen8_pte_t __iomem *)dev_priv->ggtt.gsm + index;
-	u64 pte;
 
-#ifdef readq
-	pte = readq(addr);
-#else
-	pte = ioread32(addr);
-	pte |= (u64)ioread32(addr + 4) << 32;
-#endif
-	return pte;
+	return readq(addr);
 }
 
 static void write_pte64(struct drm_i915_private *dev_priv,
@@ -256,12 +249,8 @@ static void write_pte64(struct drm_i915_private *dev_priv,
 {
 	void __iomem *addr = (gen8_pte_t __iomem *)dev_priv->ggtt.gsm + index;
 
-#ifdef writeq
 	writeq(pte, addr);
-#else
-	iowrite32((u32)pte, addr);
-	iowrite32(pte >> 32, addr + 4);
-#endif
+
 	I915_WRITE(GFX_FLSH_CNTL_GEN6, GFX_FLSH_CNTL_EN);
 	POSTING_READ(GFX_FLSH_CNTL_GEN6);
 }
@@ -1380,8 +1369,7 @@ static int gen8_mm_alloc_page_table(struct intel_vgpu_mm *mm)
 			info->gtt_entry_size;
 		mem = kzalloc(mm->has_shadow_page_table ?
 			mm->page_table_entry_size * 2
-				: mm->page_table_entry_size,
-			GFP_ATOMIC);
+				: mm->page_table_entry_size, GFP_KERNEL);
 		if (!mem)
 			return -ENOMEM;
 		mm->virtual_page_table = mem;
@@ -1532,7 +1520,7 @@ struct intel_vgpu_mm *intel_vgpu_create_mm(struct intel_vgpu *vgpu,
 	struct intel_vgpu_mm *mm;
 	int ret;
 
-	mm = kzalloc(sizeof(*mm), GFP_ATOMIC);
+	mm = kzalloc(sizeof(*mm), GFP_KERNEL);
 	if (!mm) {
 		ret = -ENOMEM;
 		goto fail;
@@ -1886,30 +1874,27 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 	struct intel_gvt_gtt_pte_ops *ops = vgpu->gvt->gtt.pte_ops;
 	int page_entry_num = GTT_PAGE_SIZE >>
 				vgpu->gvt->device_info.gtt_entry_size_shift;
-	struct page *scratch_pt;
+	void *scratch_pt;
 	unsigned long mfn;
 	int i;
-	void *p;
 
 	if (WARN_ON(type < GTT_TYPE_PPGTT_PTE_PT || type >= GTT_TYPE_MAX))
 		return -EINVAL;
 
-	scratch_pt = alloc_page(GFP_KERNEL | GFP_ATOMIC | __GFP_ZERO);
+	scratch_pt = (void *)get_zeroed_page(GFP_KERNEL);
 	if (!scratch_pt) {
 		gvt_err("fail to allocate scratch page\n");
 		return -ENOMEM;
 	}
 
-	p = kmap_atomic(scratch_pt);
-	mfn = intel_gvt_hypervisor_virt_to_mfn(p);
+	mfn = intel_gvt_hypervisor_virt_to_mfn(scratch_pt);
 	if (mfn == INTEL_GVT_INVALID_ADDR) {
-		gvt_err("fail to translate vaddr:0x%llx\n", (u64)p);
-		kunmap_atomic(p);
-		__free_page(scratch_pt);
+		gvt_err("fail to translate vaddr:0x%lx\n", (unsigned long)scratch_pt);
+		free_page((unsigned long)scratch_pt);
 		return -EFAULT;
 	}
 	gtt->scratch_pt[type].page_mfn = mfn;
-	gtt->scratch_pt[type].page = scratch_pt;
+	gtt->scratch_pt[type].page = virt_to_page(scratch_pt);
 	gvt_dbg_mm("vgpu%d create scratch_pt: type %d mfn=0x%lx\n",
 			vgpu->id, type, mfn);
 
@@ -1918,7 +1903,7 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 	 * scratch_pt[type] indicate the scratch pt/scratch page used by the
 	 * 'type' pt.
 	 * e.g. scratch_pt[GTT_TYPE_PPGTT_PDE_PT] is used by
-	 * GTT_TYPE_PPGTT_PDE_PT level pt, that means this scatch_pt it self
+	 * GTT_TYPE_PPGTT_PDE_PT level pt, that means this scratch_pt it self
 	 * is GTT_TYPE_PPGTT_PTE_PT, and full filled by scratch page mfn.
 	 */
 	if (type > GTT_TYPE_PPGTT_PTE_PT && type < GTT_TYPE_MAX) {
@@ -1936,11 +1921,9 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 			se.val64 |= PPAT_CACHED_INDEX;
 
 		for (i = 0; i < page_entry_num; i++)
-			ops->set_entry(p, &se, i, false, 0, vgpu);
+			ops->set_entry(scratch_pt, &se, i, false, 0, vgpu);
 	}
 
-	kunmap_atomic(p);
-
 	return 0;
 }
 
@@ -1998,6 +1981,8 @@ int intel_vgpu_init_gtt(struct intel_vgpu *vgpu)
 	INIT_LIST_HEAD(&gtt->oos_page_list_head);
 	INIT_LIST_HEAD(&gtt->post_shadow_list_head);
 
+	intel_vgpu_reset_ggtt(vgpu);
+
 	ggtt_mm = intel_vgpu_create_mm(vgpu, INTEL_GVT_MM_GGTT,
 			NULL, 1, 0);
 	if (IS_ERR(ggtt_mm)) {
@@ -2206,6 +2191,7 @@ int intel_vgpu_g2v_destroy_ppgtt_mm(struct intel_vgpu *vgpu,
 int intel_gvt_init_gtt(struct intel_gvt *gvt)
 {
 	int ret;
+	void *page;
 
 	gvt_dbg_core("init gtt\n");
 
@@ -2218,6 +2204,20 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt)
 		return -ENODEV;
 	}
 
+	page = (void *)get_zeroed_page(GFP_KERNEL);
+	if (!page) {
+		gvt_err("fail to allocate scratch ggtt page\n");
+		return -ENOMEM;
+	}
+	gvt->gtt.scratch_ggtt_page = virt_to_page(page);
+
+	gvt->gtt.scratch_ggtt_mfn = intel_gvt_hypervisor_virt_to_mfn(page);
+	if (gvt->gtt.scratch_ggtt_mfn == INTEL_GVT_INVALID_ADDR) {
+		gvt_err("fail to translate scratch ggtt page\n");
+		__free_page(gvt->gtt.scratch_ggtt_page);
+		return -EFAULT;
+	}
+
 	if (enable_out_of_sync) {
 		ret = setup_spt_oos(gvt);
 		if (ret) {
@@ -2239,6 +2239,68 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt)
  */
 void intel_gvt_clean_gtt(struct intel_gvt *gvt)
 {
+	__free_page(gvt->gtt.scratch_ggtt_page);
+
 	if (enable_out_of_sync)
 		clean_spt_oos(gvt);
 }
+
+/**
+ * intel_vgpu_reset_ggtt - reset the GGTT entry
+ * @vgpu: a vGPU
+ *
+ * This function is called at the vGPU create stage
+ * to reset all the GGTT entries.
+ *
+ */
+void intel_vgpu_reset_ggtt(struct intel_vgpu *vgpu)
+{
+	struct intel_gvt *gvt = vgpu->gvt;
+	struct intel_gvt_gtt_pte_ops *ops = vgpu->gvt->gtt.pte_ops;
+	u32 index;
+	u32 offset;
+	u32 num_entries;
+	struct intel_gvt_gtt_entry e;
+
+	memset(&e, 0, sizeof(struct intel_gvt_gtt_entry));
+	e.type = GTT_TYPE_GGTT_PTE;
+	ops->set_pfn(&e, gvt->gtt.scratch_ggtt_mfn);
+	e.val64 |= _PAGE_PRESENT;
+
+	index = vgpu_aperture_gmadr_base(vgpu) >> PAGE_SHIFT;
+	num_entries = vgpu_aperture_sz(vgpu) >> PAGE_SHIFT;
+	for (offset = 0; offset < num_entries; offset++)
+		ops->set_entry(NULL, &e, index + offset, false, 0, vgpu);
+
+	index = vgpu_hidden_gmadr_base(vgpu) >> PAGE_SHIFT;
+	num_entries = vgpu_hidden_sz(vgpu) >> PAGE_SHIFT;
+	for (offset = 0; offset < num_entries; offset++)
+		ops->set_entry(NULL, &e, index + offset, false, 0, vgpu);
+}
+
+/**
+ * intel_vgpu_reset_gtt - reset the all GTT related status
+ * @vgpu: a vGPU
+ * @dmlr: true for vGPU Device Model Level Reset, false for GT Reset
+ *
+ * This function is called from vfio core to reset reset all
+ * GTT related status, including GGTT, PPGTT, scratch page.
+ *
+ */
+void intel_vgpu_reset_gtt(struct intel_vgpu *vgpu, bool dmlr)
+{
+	int i;
+
+	ppgtt_free_all_shadow_page(vgpu);
+	if (!dmlr)
+		return;
+
+	intel_vgpu_reset_ggtt(vgpu);
+
+	/* clear scratch page for security */
+	for (i = GTT_TYPE_PPGTT_PTE_PT; i < GTT_TYPE_MAX; i++) {
+		if (vgpu->gtt.scratch_pt[i].page != NULL)
+			memset(page_address(vgpu->gtt.scratch_pt[i].page),
+				0, PAGE_SIZE);
+	}
+}
diff --git a/drivers/gpu/drm/i915/gvt/gtt.h b/drivers/gpu/drm/i915/gvt/gtt.h
index d250013bc37b..f88eb5e89bea 100644
--- a/drivers/gpu/drm/i915/gvt/gtt.h
+++ b/drivers/gpu/drm/i915/gvt/gtt.h
@@ -81,6 +81,9 @@ struct intel_gvt_gtt {
 	struct list_head oos_page_use_list_head;
 	struct list_head oos_page_free_list_head;
 	struct list_head mm_lru_list_head;
+
+	struct page *scratch_ggtt_page;
+	unsigned long scratch_ggtt_mfn;
 };
 
 enum {
@@ -202,8 +205,10 @@ struct intel_vgpu_gtt {
 
 extern int intel_vgpu_init_gtt(struct intel_vgpu *vgpu);
 extern void intel_vgpu_clean_gtt(struct intel_vgpu *vgpu);
+void intel_vgpu_reset_ggtt(struct intel_vgpu *vgpu);
 
 extern int intel_gvt_init_gtt(struct intel_gvt *gvt);
+extern void intel_vgpu_reset_gtt(struct intel_vgpu *vgpu, bool dmlr);
 extern void intel_gvt_clean_gtt(struct intel_gvt *gvt);
 
 extern struct intel_vgpu_mm *intel_gvt_find_ppgtt_mm(struct intel_vgpu *vgpu,
diff --git a/drivers/gpu/drm/i915/gvt/gvt.c b/drivers/gpu/drm/i915/gvt/gvt.c
index 398877c3d2fd..e6bf5c533fbe 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.c
+++ b/drivers/gpu/drm/i915/gvt/gvt.c
@@ -201,6 +201,8 @@ void intel_gvt_clean_device(struct drm_i915_private *dev_priv)
 	intel_gvt_hypervisor_host_exit(&dev_priv->drm.pdev->dev, gvt);
 	intel_gvt_clean_vgpu_types(gvt);
 
+	idr_destroy(&gvt->vgpu_idr);
+
 	kfree(dev_priv->gvt);
 	dev_priv->gvt = NULL;
 }
@@ -237,6 +239,8 @@ int intel_gvt_init_device(struct drm_i915_private *dev_priv)
 
 	gvt_dbg_core("init gvt device\n");
 
+	idr_init(&gvt->vgpu_idr);
+
 	mutex_init(&gvt->lock);
 	gvt->dev_priv = dev_priv;
 
@@ -244,7 +248,7 @@ int intel_gvt_init_device(struct drm_i915_private *dev_priv)
 
 	ret = intel_gvt_setup_mmio_info(gvt);
 	if (ret)
-		return ret;
+		goto out_clean_idr;
 
 	ret = intel_gvt_load_firmware(gvt);
 	if (ret)
@@ -313,6 +317,8 @@ out_free_firmware:
 	intel_gvt_free_firmware(gvt);
 out_clean_mmio_info:
 	intel_gvt_clean_mmio_info(gvt);
+out_clean_idr:
+	idr_destroy(&gvt->vgpu_idr);
 	kfree(gvt);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h
index ad0e9364ee70..e227caf5859e 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.h
+++ b/drivers/gpu/drm/i915/gvt/gvt.h
@@ -175,6 +175,7 @@ struct intel_vgpu {
 		struct notifier_block group_notifier;
 		struct kvm *kvm;
 		struct work_struct release_work;
+		atomic_t released;
 	} vdev;
 #endif
 };
@@ -322,6 +323,7 @@ struct intel_vgpu_creation_params {
 
 int intel_vgpu_alloc_resource(struct intel_vgpu *vgpu,
 			      struct intel_vgpu_creation_params *param);
+void intel_vgpu_reset_resource(struct intel_vgpu *vgpu);
 void intel_vgpu_free_resource(struct intel_vgpu *vgpu);
 void intel_vgpu_write_fence(struct intel_vgpu *vgpu,
 	u32 fence, u64 value);
@@ -374,6 +376,8 @@ void intel_gvt_clean_vgpu_types(struct intel_gvt *gvt);
 struct intel_vgpu *intel_gvt_create_vgpu(struct intel_gvt *gvt,
 					 struct intel_vgpu_type *type);
 void intel_gvt_destroy_vgpu(struct intel_vgpu *vgpu);
+void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr,
+				 unsigned int engine_mask);
 void intel_gvt_reset_vgpu(struct intel_vgpu *vgpu);
 
 
@@ -410,6 +414,10 @@ int intel_gvt_ggtt_index_g2h(struct intel_vgpu *vgpu, unsigned long g_index,
 int intel_gvt_ggtt_h2g_index(struct intel_vgpu *vgpu, unsigned long h_index,
 			     unsigned long *g_index);
 
+void intel_vgpu_init_cfg_space(struct intel_vgpu *vgpu,
+		bool primary);
+void intel_vgpu_reset_cfg_space(struct intel_vgpu *vgpu);
+
 int intel_vgpu_emulate_cfg_read(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes);
 
@@ -423,7 +431,6 @@ void intel_vgpu_clean_opregion(struct intel_vgpu *vgpu);
 int intel_vgpu_init_opregion(struct intel_vgpu *vgpu, u32 gpa);
 
 int intel_vgpu_emulate_opregion_request(struct intel_vgpu *vgpu, u32 swsci);
-int setup_vgpu_mmio(struct intel_vgpu *vgpu);
 void populate_pvinfo_page(struct intel_vgpu *vgpu);
 
 struct intel_gvt_ops {
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index 522809710312..ab2ea157da4c 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -93,7 +93,8 @@ static void write_vreg(struct intel_vgpu *vgpu, unsigned int offset,
 static int new_mmio_info(struct intel_gvt *gvt,
 		u32 offset, u32 flags, u32 size,
 		u32 addr_mask, u32 ro_mask, u32 device,
-		void *read, void *write)
+		int (*read)(struct intel_vgpu *, unsigned int, void *, unsigned int),
+		int (*write)(struct intel_vgpu *, unsigned int, void *, unsigned int))
 {
 	struct intel_gvt_mmio_info *info, *p;
 	u32 start, end, i;
@@ -219,7 +220,7 @@ static int mul_force_wake_write(struct intel_vgpu *vgpu,
 		default:
 			/*should not hit here*/
 			gvt_err("invalid forcewake offset 0x%x\n", offset);
-			return 1;
+			return -EINVAL;
 		}
 	} else {
 		ack_reg_offset = FORCEWAKE_ACK_HSW_REG;
@@ -230,77 +231,45 @@ static int mul_force_wake_write(struct intel_vgpu *vgpu,
 	return 0;
 }
 
-static int handle_device_reset(struct intel_vgpu *vgpu, unsigned int offset,
-		void *p_data, unsigned int bytes, unsigned long bitmap)
-{
-	struct intel_gvt_workload_scheduler *scheduler =
-		&vgpu->gvt->scheduler;
-
-	vgpu->resetting = true;
-
-	intel_vgpu_stop_schedule(vgpu);
-	/*
-	 * The current_vgpu will set to NULL after stopping the
-	 * scheduler when the reset is triggered by current vgpu.
-	 */
-	if (scheduler->current_vgpu == NULL) {
-		mutex_unlock(&vgpu->gvt->lock);
-		intel_gvt_wait_vgpu_idle(vgpu);
-		mutex_lock(&vgpu->gvt->lock);
-	}
-
-	intel_vgpu_reset_execlist(vgpu, bitmap);
-
-	/* full GPU reset */
-	if (bitmap == 0xff) {
-		mutex_unlock(&vgpu->gvt->lock);
-		intel_vgpu_clean_gtt(vgpu);
-		mutex_lock(&vgpu->gvt->lock);
-		setup_vgpu_mmio(vgpu);
-		populate_pvinfo_page(vgpu);
-		intel_vgpu_init_gtt(vgpu);
-	}
-
-	vgpu->resetting = false;
-
-	return 0;
-}
-
 static int gdrst_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
-		void *p_data, unsigned int bytes)
+			    void *p_data, unsigned int bytes)
 {
+	unsigned int engine_mask = 0;
 	u32 data;
-	u64 bitmap = 0;
 
 	write_vreg(vgpu, offset, p_data, bytes);
 	data = vgpu_vreg(vgpu, offset);
 
 	if (data & GEN6_GRDOM_FULL) {
 		gvt_dbg_mmio("vgpu%d: request full GPU reset\n", vgpu->id);
-		bitmap = 0xff;
-	}
-	if (data & GEN6_GRDOM_RENDER) {
-		gvt_dbg_mmio("vgpu%d: request RCS reset\n", vgpu->id);
-		bitmap |= (1 << RCS);
-	}
-	if (data & GEN6_GRDOM_MEDIA) {
-		gvt_dbg_mmio("vgpu%d: request VCS reset\n", vgpu->id);
-		bitmap |= (1 << VCS);
-	}
-	if (data & GEN6_GRDOM_BLT) {
-		gvt_dbg_mmio("vgpu%d: request BCS Reset\n", vgpu->id);
-		bitmap |= (1 << BCS);
-	}
-	if (data & GEN6_GRDOM_VECS) {
-		gvt_dbg_mmio("vgpu%d: request VECS Reset\n", vgpu->id);
-		bitmap |= (1 << VECS);
-	}
-	if (data & GEN8_GRDOM_MEDIA2) {
-		gvt_dbg_mmio("vgpu%d: request VCS2 Reset\n", vgpu->id);
-		if (HAS_BSD2(vgpu->gvt->dev_priv))
-			bitmap |= (1 << VCS2);
+		engine_mask = ALL_ENGINES;
+	} else {
+		if (data & GEN6_GRDOM_RENDER) {
+			gvt_dbg_mmio("vgpu%d: request RCS reset\n", vgpu->id);
+			engine_mask |= (1 << RCS);
+		}
+		if (data & GEN6_GRDOM_MEDIA) {
+			gvt_dbg_mmio("vgpu%d: request VCS reset\n", vgpu->id);
+			engine_mask |= (1 << VCS);
+		}
+		if (data & GEN6_GRDOM_BLT) {
+			gvt_dbg_mmio("vgpu%d: request BCS Reset\n", vgpu->id);
+			engine_mask |= (1 << BCS);
+		}
+		if (data & GEN6_GRDOM_VECS) {
+			gvt_dbg_mmio("vgpu%d: request VECS Reset\n", vgpu->id);
+			engine_mask |= (1 << VECS);
+		}
+		if (data & GEN8_GRDOM_MEDIA2) {
+			gvt_dbg_mmio("vgpu%d: request VCS2 Reset\n", vgpu->id);
+			if (HAS_BSD2(vgpu->gvt->dev_priv))
+				engine_mask |= (1 << VCS2);
+		}
 	}
-	return handle_device_reset(vgpu, offset, p_data, bytes, bitmap);
+
+	intel_gvt_reset_vgpu_locked(vgpu, false, engine_mask);
+
+	return 0;
 }
 
 static int gmbus_mmio_read(struct intel_vgpu *vgpu, unsigned int offset,
@@ -974,7 +943,7 @@ static int sbi_data_mmio_read(struct intel_vgpu *vgpu, unsigned int offset,
 	return 0;
 }
 
-static bool sbi_ctl_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
+static int sbi_ctl_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
 	u32 data;
@@ -1366,7 +1335,6 @@ static int ring_mode_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 static int gvt_reg_tlb_control_handler(struct intel_vgpu *vgpu,
 		unsigned int offset, void *p_data, unsigned int bytes)
 {
-	int rc = 0;
 	unsigned int id = 0;
 
 	write_vreg(vgpu, offset, p_data, bytes);
@@ -1389,12 +1357,11 @@ static int gvt_reg_tlb_control_handler(struct intel_vgpu *vgpu,
 		id = VECS;
 		break;
 	default:
-		rc = -EINVAL;
-		break;
+		return -EINVAL;
 	}
 	set_bit(id, (void *)vgpu->tlb_handle_pending);
 
-	return rc;
+	return 0;
 }
 
 static int ring_reset_ctl_write(struct intel_vgpu *vgpu,
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 4dd6722a7339..3f656e3a6e5a 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -114,12 +114,15 @@ out:
 static kvm_pfn_t gvt_cache_find(struct intel_vgpu *vgpu, gfn_t gfn)
 {
 	struct gvt_dma *entry;
+	kvm_pfn_t pfn;
 
 	mutex_lock(&vgpu->vdev.cache_lock);
+
 	entry = __gvt_cache_find(vgpu, gfn);
-	mutex_unlock(&vgpu->vdev.cache_lock);
+	pfn = (entry == NULL) ? 0 : entry->pfn;
 
-	return entry == NULL ? 0 : entry->pfn;
+	mutex_unlock(&vgpu->vdev.cache_lock);
+	return pfn;
 }
 
 static void gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn, kvm_pfn_t pfn)
@@ -166,7 +169,7 @@ static void __gvt_cache_remove_entry(struct intel_vgpu *vgpu,
 
 static void gvt_cache_remove(struct intel_vgpu *vgpu, gfn_t gfn)
 {
-	struct device *dev = &vgpu->vdev.mdev->dev;
+	struct device *dev = mdev_dev(vgpu->vdev.mdev);
 	struct gvt_dma *this;
 	unsigned long g1;
 	int rc;
@@ -195,7 +198,7 @@ static void gvt_cache_destroy(struct intel_vgpu *vgpu)
 {
 	struct gvt_dma *dma;
 	struct rb_node *node = NULL;
-	struct device *dev = &vgpu->vdev.mdev->dev;
+	struct device *dev = mdev_dev(vgpu->vdev.mdev);
 	unsigned long gfn;
 
 	mutex_lock(&vgpu->vdev.cache_lock);
@@ -227,8 +230,8 @@ static struct intel_vgpu_type *intel_gvt_find_vgpu_type(struct intel_gvt *gvt,
 	return NULL;
 }
 
-static ssize_t available_instance_show(struct kobject *kobj, struct device *dev,
-		char *buf)
+static ssize_t available_instances_show(struct kobject *kobj,
+					struct device *dev, char *buf)
 {
 	struct intel_vgpu_type *type;
 	unsigned int num = 0;
@@ -266,12 +269,12 @@ static ssize_t description_show(struct kobject *kobj, struct device *dev,
 				type->fence);
 }
 
-static MDEV_TYPE_ATTR_RO(available_instance);
+static MDEV_TYPE_ATTR_RO(available_instances);
 static MDEV_TYPE_ATTR_RO(device_api);
 static MDEV_TYPE_ATTR_RO(description);
 
 static struct attribute *type_attrs[] = {
-	&mdev_type_attr_available_instance.attr,
+	&mdev_type_attr_available_instances.attr,
 	&mdev_type_attr_device_api.attr,
 	&mdev_type_attr_description.attr,
 	NULL,
@@ -395,21 +398,24 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 	struct intel_vgpu_type *type;
 	struct device *pdev;
 	void *gvt;
+	int ret;
 
-	pdev = mdev->parent->dev;
+	pdev = mdev_parent_dev(mdev);
 	gvt = kdev_to_i915(pdev)->gvt;
 
 	type = intel_gvt_find_vgpu_type(gvt, kobject_name(kobj));
 	if (!type) {
 		gvt_err("failed to find type %s to create\n",
 						kobject_name(kobj));
-		return -EINVAL;
+		ret = -EINVAL;
+		goto out;
 	}
 
 	vgpu = intel_gvt_ops->vgpu_create(gvt, type);
 	if (IS_ERR_OR_NULL(vgpu)) {
-		gvt_err("create intel vgpu failed\n");
-		return -EINVAL;
+		ret = vgpu == NULL ? -EFAULT : PTR_ERR(vgpu);
+		gvt_err("failed to create intel vgpu: %d\n", ret);
+		goto out;
 	}
 
 	INIT_WORK(&vgpu->vdev.release_work, intel_vgpu_release_work);
@@ -418,8 +424,11 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 	mdev_set_drvdata(mdev, vgpu);
 
 	gvt_dbg_core("intel_vgpu_create succeeded for mdev: %s\n",
-		     dev_name(&mdev->dev));
-	return 0;
+		     dev_name(mdev_dev(mdev)));
+	ret = 0;
+
+out:
+	return ret;
 }
 
 static int intel_vgpu_remove(struct mdev_device *mdev)
@@ -482,7 +491,7 @@ static int intel_vgpu_open(struct mdev_device *mdev)
 	vgpu->vdev.group_notifier.notifier_call = intel_vgpu_group_notifier;
 
 	events = VFIO_IOMMU_NOTIFY_DMA_UNMAP;
-	ret = vfio_register_notifier(&mdev->dev, VFIO_IOMMU_NOTIFY, &events,
+	ret = vfio_register_notifier(mdev_dev(mdev), VFIO_IOMMU_NOTIFY, &events,
 				&vgpu->vdev.iommu_notifier);
 	if (ret != 0) {
 		gvt_err("vfio_register_notifier for iommu failed: %d\n", ret);
@@ -490,17 +499,26 @@ static int intel_vgpu_open(struct mdev_device *mdev)
 	}
 
 	events = VFIO_GROUP_NOTIFY_SET_KVM;
-	ret = vfio_register_notifier(&mdev->dev, VFIO_GROUP_NOTIFY, &events,
+	ret = vfio_register_notifier(mdev_dev(mdev), VFIO_GROUP_NOTIFY, &events,
 				&vgpu->vdev.group_notifier);
 	if (ret != 0) {
 		gvt_err("vfio_register_notifier for group failed: %d\n", ret);
 		goto undo_iommu;
 	}
 
-	return kvmgt_guest_init(mdev);
+	ret = kvmgt_guest_init(mdev);
+	if (ret)
+		goto undo_group;
+
+	atomic_set(&vgpu->vdev.released, 0);
+	return ret;
+
+undo_group:
+	vfio_unregister_notifier(mdev_dev(mdev), VFIO_GROUP_NOTIFY,
+					&vgpu->vdev.group_notifier);
 
 undo_iommu:
-	vfio_unregister_notifier(&mdev->dev, VFIO_IOMMU_NOTIFY,
+	vfio_unregister_notifier(mdev_dev(mdev), VFIO_IOMMU_NOTIFY,
 					&vgpu->vdev.iommu_notifier);
 out:
 	return ret;
@@ -509,17 +527,26 @@ out:
 static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 {
 	struct kvmgt_guest_info *info;
+	int ret;
 
 	if (!handle_valid(vgpu->handle))
 		return;
 
-	vfio_unregister_notifier(&vgpu->vdev.mdev->dev, VFIO_IOMMU_NOTIFY,
+	if (atomic_cmpxchg(&vgpu->vdev.released, 0, 1))
+		return;
+
+	ret = vfio_unregister_notifier(mdev_dev(vgpu->vdev.mdev), VFIO_IOMMU_NOTIFY,
 					&vgpu->vdev.iommu_notifier);
-	vfio_unregister_notifier(&vgpu->vdev.mdev->dev, VFIO_GROUP_NOTIFY,
+	WARN(ret, "vfio_unregister_notifier for iommu failed: %d\n", ret);
+
+	ret = vfio_unregister_notifier(mdev_dev(vgpu->vdev.mdev), VFIO_GROUP_NOTIFY,
 					&vgpu->vdev.group_notifier);
+	WARN(ret, "vfio_unregister_notifier for group failed: %d\n", ret);
 
 	info = (struct kvmgt_guest_info *)vgpu->handle;
 	kvmgt_guest_exit(info);
+
+	vgpu->vdev.kvm = NULL;
 	vgpu->handle = 0;
 }
 
@@ -534,6 +561,7 @@ static void intel_vgpu_release_work(struct work_struct *work)
 {
 	struct intel_vgpu *vgpu = container_of(work, struct intel_vgpu,
 					vdev.release_work);
+
 	__intel_vgpu_release(vgpu);
 }
 
@@ -1089,7 +1117,7 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd,
 	return 0;
 }
 
-static const struct parent_ops intel_vgpu_ops = {
+static const struct mdev_parent_ops intel_vgpu_ops = {
 	.supported_type_groups	= intel_vgpu_type_groups,
 	.create			= intel_vgpu_create,
 	.remove			= intel_vgpu_remove,
@@ -1134,6 +1162,10 @@ static int kvmgt_write_protect_add(unsigned long handle, u64 gfn)
 
 	idx = srcu_read_lock(&kvm->srcu);
 	slot = gfn_to_memslot(kvm, gfn);
+	if (!slot) {
+		srcu_read_unlock(&kvm->srcu, idx);
+		return -EINVAL;
+	}
 
 	spin_lock(&kvm->mmu_lock);
 
@@ -1164,6 +1196,10 @@ static int kvmgt_write_protect_remove(unsigned long handle, u64 gfn)
 
 	idx = srcu_read_lock(&kvm->srcu);
 	slot = gfn_to_memslot(kvm, gfn);
+	if (!slot) {
+		srcu_read_unlock(&kvm->srcu, idx);
+		return -EINVAL;
+	}
 
 	spin_lock(&kvm->mmu_lock);
 
@@ -1311,18 +1347,14 @@ static int kvmgt_guest_init(struct mdev_device *mdev)
 
 static bool kvmgt_guest_exit(struct kvmgt_guest_info *info)
 {
-	struct intel_vgpu *vgpu;
-
 	if (!info) {
 		gvt_err("kvmgt_guest_info invalid\n");
 		return false;
 	}
 
-	vgpu = info->vgpu;
-
 	kvm_page_track_unregister_notifier(info->kvm, &info->track_node);
 	kvmgt_protect_table_destroy(info);
-	gvt_cache_destroy(vgpu);
+	gvt_cache_destroy(info->vgpu);
 	vfree(info);
 
 	return true;
@@ -1372,7 +1404,7 @@ static unsigned long kvmgt_gfn_to_pfn(unsigned long handle, unsigned long gfn)
 		return pfn;
 
 	pfn = INTEL_GVT_INVALID_ADDR;
-	dev = &info->vgpu->vdev.mdev->dev;
+	dev = mdev_dev(info->vgpu->vdev.mdev);
 	rc = vfio_pin_pages(dev, &gfn, 1, IOMMU_READ | IOMMU_WRITE, &pfn);
 	if (rc != 1) {
 		gvt_err("vfio_pin_pages failed for gfn 0x%lx: %d\n", gfn, rc);
diff --git a/drivers/gpu/drm/i915/gvt/mmio.c b/drivers/gpu/drm/i915/gvt/mmio.c
index 09c9450a1946..4df078bc5d04 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.c
+++ b/drivers/gpu/drm/i915/gvt/mmio.c
@@ -125,25 +125,12 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, uint64_t pa,
 	if (WARN_ON(!reg_is_mmio(gvt, offset + bytes - 1)))
 		goto err;
 
-	mmio = intel_gvt_find_mmio_info(gvt, rounddown(offset, 4));
-	if (!mmio && !vgpu->mmio.disable_warn_untrack) {
-		gvt_err("vgpu%d: read untracked MMIO %x len %d val %x\n",
-				vgpu->id, offset, bytes, *(u32 *)p_data);
-
-		if (offset == 0x206c) {
-			gvt_err("------------------------------------------\n");
-			gvt_err("vgpu%d: likely triggers a gfx reset\n",
-			vgpu->id);
-			gvt_err("------------------------------------------\n");
-			vgpu->mmio.disable_warn_untrack = true;
-		}
-	}
-
 	if (!intel_gvt_mmio_is_unalign(gvt, offset)) {
 		if (WARN_ON(!IS_ALIGNED(offset, bytes)))
 			goto err;
 	}
 
+	mmio = intel_gvt_find_mmio_info(gvt, rounddown(offset, 4));
 	if (mmio) {
 		if (!intel_gvt_mmio_is_unalign(gvt, mmio->offset)) {
 			if (WARN_ON(offset + bytes > mmio->offset + mmio->size))
@@ -152,9 +139,23 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, uint64_t pa,
 				goto err;
 		}
 		ret = mmio->read(vgpu, offset, p_data, bytes);
-	} else
+	} else {
 		ret = intel_vgpu_default_mmio_read(vgpu, offset, p_data, bytes);
 
+		if (!vgpu->mmio.disable_warn_untrack) {
+			gvt_err("vgpu%d: read untracked MMIO %x(%dB) val %x\n",
+				vgpu->id, offset, bytes, *(u32 *)p_data);
+
+			if (offset == 0x206c) {
+				gvt_err("------------------------------------------\n");
+				gvt_err("vgpu%d: likely triggers a gfx reset\n",
+					vgpu->id);
+				gvt_err("------------------------------------------\n");
+				vgpu->mmio.disable_warn_untrack = true;
+			}
+		}
+	}
+
 	if (ret)
 		goto err;
 
@@ -302,3 +303,56 @@ err:
 	mutex_unlock(&gvt->lock);
 	return ret;
 }
+
+
+/**
+ * intel_vgpu_reset_mmio - reset virtual MMIO space
+ * @vgpu: a vGPU
+ *
+ */
+void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu)
+{
+	struct intel_gvt *gvt = vgpu->gvt;
+	const struct intel_gvt_device_info *info = &gvt->device_info;
+
+	memcpy(vgpu->mmio.vreg, gvt->firmware.mmio, info->mmio_size);
+	memcpy(vgpu->mmio.sreg, gvt->firmware.mmio, info->mmio_size);
+
+	vgpu_vreg(vgpu, GEN6_GT_THREAD_STATUS_REG) = 0;
+
+	/* set the bit 0:2(Core C-State ) to C0 */
+	vgpu_vreg(vgpu, GEN6_GT_CORE_STATUS) = 0;
+}
+
+/**
+ * intel_vgpu_init_mmio - init MMIO  space
+ * @vgpu: a vGPU
+ *
+ * Returns:
+ * Zero on success, negative error code if failed
+ */
+int intel_vgpu_init_mmio(struct intel_vgpu *vgpu)
+{
+	const struct intel_gvt_device_info *info = &vgpu->gvt->device_info;
+
+	vgpu->mmio.vreg = vzalloc(info->mmio_size * 2);
+	if (!vgpu->mmio.vreg)
+		return -ENOMEM;
+
+	vgpu->mmio.sreg = vgpu->mmio.vreg + info->mmio_size;
+
+	intel_vgpu_reset_mmio(vgpu);
+
+	return 0;
+}
+
+/**
+ * intel_vgpu_clean_mmio - clean MMIO space
+ * @vgpu: a vGPU
+ *
+ */
+void intel_vgpu_clean_mmio(struct intel_vgpu *vgpu)
+{
+	vfree(vgpu->mmio.vreg);
+	vgpu->mmio.vreg = vgpu->mmio.sreg = NULL;
+}
diff --git a/drivers/gpu/drm/i915/gvt/mmio.h b/drivers/gpu/drm/i915/gvt/mmio.h
index 87d5b5e366a3..3bc620f56f35 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.h
+++ b/drivers/gpu/drm/i915/gvt/mmio.h
@@ -86,6 +86,10 @@ struct intel_gvt_mmio_info *intel_gvt_find_mmio_info(struct intel_gvt *gvt,
 	*offset; \
 })
 
+int intel_vgpu_init_mmio(struct intel_vgpu *vgpu);
+void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu);
+void intel_vgpu_clean_mmio(struct intel_vgpu *vgpu);
+
 int intel_vgpu_gpa_to_mmio_offset(struct intel_vgpu *vgpu, u64 gpa);
 
 int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, u64 pa,
diff --git a/drivers/gpu/drm/i915/gvt/opregion.c b/drivers/gpu/drm/i915/gvt/opregion.c
index d2a0fbc896c3..d9fb41ab7119 100644
--- a/drivers/gpu/drm/i915/gvt/opregion.c
+++ b/drivers/gpu/drm/i915/gvt/opregion.c
@@ -36,9 +36,9 @@ static int init_vgpu_opregion(struct intel_vgpu *vgpu, u32 gpa)
 			vgpu->id))
 		return -EINVAL;
 
-	vgpu_opregion(vgpu)->va = (void *)__get_free_pages(GFP_ATOMIC |
-			GFP_DMA32 | __GFP_ZERO,
-			INTEL_GVT_OPREGION_PORDER);
+	vgpu_opregion(vgpu)->va = (void *)__get_free_pages(GFP_KERNEL |
+			__GFP_ZERO,
+			get_order(INTEL_GVT_OPREGION_SIZE));
 
 	if (!vgpu_opregion(vgpu)->va)
 		return -ENOMEM;
@@ -65,7 +65,7 @@ static int map_vgpu_opregion(struct intel_vgpu *vgpu, bool map)
 	int i, ret;
 
 	for (i = 0; i < INTEL_GVT_OPREGION_PAGES; i++) {
-		mfn = intel_gvt_hypervisor_virt_to_mfn(vgpu_opregion(vgpu)
+		mfn = intel_gvt_hypervisor_virt_to_mfn(vgpu_opregion(vgpu)->va
 			+ i * PAGE_SIZE);
 		if (mfn == INTEL_GVT_INVALID_ADDR) {
 			gvt_err("fail to get MFN from VA\n");
@@ -97,7 +97,7 @@ void intel_vgpu_clean_opregion(struct intel_vgpu *vgpu)
 	if (intel_gvt_host.hypervisor_type == INTEL_GVT_HYPERVISOR_XEN) {
 		map_vgpu_opregion(vgpu, false);
 		free_pages((unsigned long)vgpu_opregion(vgpu)->va,
-				INTEL_GVT_OPREGION_PORDER);
+				get_order(INTEL_GVT_OPREGION_SIZE));
 
 		vgpu_opregion(vgpu)->va = NULL;
 	}
diff --git a/drivers/gpu/drm/i915/gvt/reg.h b/drivers/gpu/drm/i915/gvt/reg.h
index 0dfe789d8f02..fbd023a16f18 100644
--- a/drivers/gpu/drm/i915/gvt/reg.h
+++ b/drivers/gpu/drm/i915/gvt/reg.h
@@ -50,8 +50,7 @@
 #define INTEL_GVT_OPREGION_PARM                   0x204
 
 #define INTEL_GVT_OPREGION_PAGES	2
-#define INTEL_GVT_OPREGION_PORDER	1
-#define INTEL_GVT_OPREGION_SIZE		(2 * 4096)
+#define INTEL_GVT_OPREGION_SIZE		(INTEL_GVT_OPREGION_PAGES * PAGE_SIZE)
 
 #define VGT_SPRSTRIDE(pipe)	_PIPE(pipe, _SPRA_STRIDE, _PLANE_STRIDE_2_B)
 
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 4db242250235..e91885dffeff 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -350,13 +350,15 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 {
 	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
 	struct intel_vgpu_workload *workload;
+	struct intel_vgpu *vgpu;
 	int event;
 
 	mutex_lock(&gvt->lock);
 
 	workload = scheduler->current_workload[ring_id];
+	vgpu = workload->vgpu;
 
-	if (!workload->status && !workload->vgpu->resetting) {
+	if (!workload->status && !vgpu->resetting) {
 		wait_event(workload->shadow_ctx_status_wq,
 			   !atomic_read(&workload->shadow_ctx_active));
 
@@ -364,8 +366,7 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 
 		for_each_set_bit(event, workload->pending_events,
 				 INTEL_GVT_EVENT_MAX)
-			intel_vgpu_trigger_virtual_event(workload->vgpu,
-					event);
+			intel_vgpu_trigger_virtual_event(vgpu, event);
 	}
 
 	gvt_dbg_sched("ring id %d complete workload %p status %d\n",
@@ -373,11 +374,10 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 
 	scheduler->current_workload[ring_id] = NULL;
 
-	atomic_dec(&workload->vgpu->running_workload_num);
-
 	list_del_init(&workload->list);
 	workload->complete(workload);
 
+	atomic_dec(&vgpu->running_workload_num);
 	wake_up(&scheduler->workload_complete_wq);
 	mutex_unlock(&gvt->lock);
 }
@@ -459,11 +459,11 @@ complete:
 		gvt_dbg_sched("will complete workload %p\n, status: %d\n",
 				workload, workload->status);
 
-		complete_current_workload(gvt, ring_id);
-
 		if (workload->req)
 			i915_gem_request_put(fetch_and_zero(&workload->req));
 
+		complete_current_workload(gvt, ring_id);
+
 		if (need_force_wake)
 			intel_uncore_forcewake_put(gvt->dev_priv,
 					FORCEWAKE_ALL);
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.h b/drivers/gpu/drm/i915/gvt/scheduler.h
index 3b30c28bff51..2833dfa8c9ae 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.h
+++ b/drivers/gpu/drm/i915/gvt/scheduler.h
@@ -113,7 +113,7 @@ struct intel_shadow_bb_entry {
 	struct drm_i915_gem_object *obj;
 	void *va;
 	unsigned long len;
-	void *bb_start_cmd_va;
+	u32 *bb_start_cmd_va;
 };
 
 #define workload_q_head(vgpu, ring_id) \
diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c
index 536d2b9d5777..7295bc8e12fb 100644
--- a/drivers/gpu/drm/i915/gvt/vgpu.c
+++ b/drivers/gpu/drm/i915/gvt/vgpu.c
@@ -35,79 +35,6 @@
 #include "gvt.h"
 #include "i915_pvinfo.h"
 
-static void clean_vgpu_mmio(struct intel_vgpu *vgpu)
-{
-	vfree(vgpu->mmio.vreg);
-	vgpu->mmio.vreg = vgpu->mmio.sreg = NULL;
-}
-
-int setup_vgpu_mmio(struct intel_vgpu *vgpu)
-{
-	struct intel_gvt *gvt = vgpu->gvt;
-	const struct intel_gvt_device_info *info = &gvt->device_info;
-
-	if (vgpu->mmio.vreg)
-		memset(vgpu->mmio.vreg, 0, info->mmio_size * 2);
-	else {
-		vgpu->mmio.vreg = vzalloc(info->mmio_size * 2);
-		if (!vgpu->mmio.vreg)
-			return -ENOMEM;
-	}
-
-	vgpu->mmio.sreg = vgpu->mmio.vreg + info->mmio_size;
-
-	memcpy(vgpu->mmio.vreg, gvt->firmware.mmio, info->mmio_size);
-	memcpy(vgpu->mmio.sreg, gvt->firmware.mmio, info->mmio_size);
-
-	vgpu_vreg(vgpu, GEN6_GT_THREAD_STATUS_REG) = 0;
-
-	/* set the bit 0:2(Core C-State ) to C0 */
-	vgpu_vreg(vgpu, GEN6_GT_CORE_STATUS) = 0;
-	return 0;
-}
-
-static void setup_vgpu_cfg_space(struct intel_vgpu *vgpu,
-	struct intel_vgpu_creation_params *param)
-{
-	struct intel_gvt *gvt = vgpu->gvt;
-	const struct intel_gvt_device_info *info = &gvt->device_info;
-	u16 *gmch_ctl;
-	int i;
-
-	memcpy(vgpu_cfg_space(vgpu), gvt->firmware.cfg_space,
-	       info->cfg_space_size);
-
-	if (!param->primary) {
-		vgpu_cfg_space(vgpu)[PCI_CLASS_DEVICE] =
-			INTEL_GVT_PCI_CLASS_VGA_OTHER;
-		vgpu_cfg_space(vgpu)[PCI_CLASS_PROG] =
-			INTEL_GVT_PCI_CLASS_VGA_OTHER;
-	}
-
-	/* Show guest that there isn't any stolen memory.*/
-	gmch_ctl = (u16 *)(vgpu_cfg_space(vgpu) + INTEL_GVT_PCI_GMCH_CONTROL);
-	*gmch_ctl &= ~(BDW_GMCH_GMS_MASK << BDW_GMCH_GMS_SHIFT);
-
-	intel_vgpu_write_pci_bar(vgpu, PCI_BASE_ADDRESS_2,
-				 gvt_aperture_pa_base(gvt), true);
-
-	vgpu_cfg_space(vgpu)[PCI_COMMAND] &= ~(PCI_COMMAND_IO
-					     | PCI_COMMAND_MEMORY
-					     | PCI_COMMAND_MASTER);
-	/*
-	 * Clear the bar upper 32bit and let guest to assign the new value
-	 */
-	memset(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_1, 0, 4);
-	memset(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_3, 0, 4);
-	memset(vgpu_cfg_space(vgpu) + INTEL_GVT_PCI_OPREGION, 0, 4);
-
-	for (i = 0; i < INTEL_GVT_MAX_BAR_NUM; i++) {
-		vgpu->cfg_space.bar[i].size = pci_resource_len(
-					      gvt->dev_priv->drm.pdev, i * 2);
-		vgpu->cfg_space.bar[i].tracked = false;
-	}
-}
-
 void populate_pvinfo_page(struct intel_vgpu *vgpu)
 {
 	/* setup the ballooning information */
@@ -177,7 +104,7 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt)
 		if (low_avail / min_low == 0)
 			break;
 		gvt->types[i].low_gm_size = min_low;
-		gvt->types[i].high_gm_size = 3 * gvt->types[i].low_gm_size;
+		gvt->types[i].high_gm_size = max((min_low<<3), MB_TO_BYTES(384U));
 		gvt->types[i].fence = 4;
 		gvt->types[i].max_instance = low_avail / min_low;
 		gvt->types[i].avail_instance = gvt->types[i].max_instance;
@@ -217,7 +144,7 @@ static void intel_gvt_update_vgpu_types(struct intel_gvt *gvt)
 	 */
 	low_gm_avail = MB_TO_BYTES(256) - HOST_LOW_GM_SIZE -
 		gvt->gm.vgpu_allocated_low_gm_size;
-	high_gm_avail = MB_TO_BYTES(256) * 3 - HOST_HIGH_GM_SIZE -
+	high_gm_avail = MB_TO_BYTES(256) * 8UL - HOST_HIGH_GM_SIZE -
 		gvt->gm.vgpu_allocated_high_gm_size;
 	fence_avail = gvt_fence_sz(gvt) - HOST_FENCE -
 		gvt->fence.vgpu_allocated_fence_num;
@@ -268,7 +195,7 @@ void intel_gvt_destroy_vgpu(struct intel_vgpu *vgpu)
 	intel_vgpu_clean_gtt(vgpu);
 	intel_gvt_hypervisor_detach_vgpu(vgpu);
 	intel_vgpu_free_resource(vgpu);
-	clean_vgpu_mmio(vgpu);
+	intel_vgpu_clean_mmio(vgpu);
 	vfree(vgpu);
 
 	intel_gvt_update_vgpu_types(gvt);
@@ -300,11 +227,11 @@ static struct intel_vgpu *__intel_gvt_create_vgpu(struct intel_gvt *gvt,
 	vgpu->gvt = gvt;
 	bitmap_zero(vgpu->tlb_handle_pending, I915_NUM_ENGINES);
 
-	setup_vgpu_cfg_space(vgpu, param);
+	intel_vgpu_init_cfg_space(vgpu, param->primary);
 
-	ret = setup_vgpu_mmio(vgpu);
+	ret = intel_vgpu_init_mmio(vgpu);
 	if (ret)
-		goto out_free_vgpu;
+		goto out_clean_idr;
 
 	ret = intel_vgpu_alloc_resource(vgpu, param);
 	if (ret)
@@ -354,7 +281,9 @@ out_detach_hypervisor_vgpu:
 out_clean_vgpu_resource:
 	intel_vgpu_free_resource(vgpu);
 out_clean_vgpu_mmio:
-	clean_vgpu_mmio(vgpu);
+	intel_vgpu_clean_mmio(vgpu);
+out_clean_idr:
+	idr_remove(&gvt->vgpu_idr, vgpu->id);
 out_free_vgpu:
 	vfree(vgpu);
 	mutex_unlock(&gvt->lock);
@@ -398,7 +327,75 @@ struct intel_vgpu *intel_gvt_create_vgpu(struct intel_gvt *gvt,
 }
 
 /**
- * intel_gvt_reset_vgpu - reset a virtual GPU
+ * intel_gvt_reset_vgpu_locked - reset a virtual GPU by DMLR or GT reset
+ * @vgpu: virtual GPU
+ * @dmlr: vGPU Device Model Level Reset or GT Reset
+ * @engine_mask: engines to reset for GT reset
+ *
+ * This function is called when user wants to reset a virtual GPU through
+ * device model reset or GT reset. The caller should hold the gvt lock.
+ *
+ * vGPU Device Model Level Reset (DMLR) simulates the PCI level reset to reset
+ * the whole vGPU to default state as when it is created. This vGPU function
+ * is required both for functionary and security concerns.The ultimate goal
+ * of vGPU FLR is that reuse a vGPU instance by virtual machines. When we
+ * assign a vGPU to a virtual machine we must isse such reset first.
+ *
+ * Full GT Reset and Per-Engine GT Reset are soft reset flow for GPU engines
+ * (Render, Blitter, Video, Video Enhancement). It is defined by GPU Spec.
+ * Unlike the FLR, GT reset only reset particular resource of a vGPU per
+ * the reset request. Guest driver can issue a GT reset by programming the
+ * virtual GDRST register to reset specific virtual GPU engine or all
+ * engines.
+ *
+ * The parameter dev_level is to identify if we will do DMLR or GT reset.
+ * The parameter engine_mask is to specific the engines that need to be
+ * resetted. If value ALL_ENGINES is given for engine_mask, it means
+ * the caller requests a full GT reset that we will reset all virtual
+ * GPU engines. For FLR, engine_mask is ignored.
+ */
+void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr,
+				 unsigned int engine_mask)
+{
+	struct intel_gvt *gvt = vgpu->gvt;
+	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
+
+	gvt_dbg_core("------------------------------------------\n");
+	gvt_dbg_core("resseting vgpu%d, dmlr %d, engine_mask %08x\n",
+		     vgpu->id, dmlr, engine_mask);
+	vgpu->resetting = true;
+
+	intel_vgpu_stop_schedule(vgpu);
+	/*
+	 * The current_vgpu will set to NULL after stopping the
+	 * scheduler when the reset is triggered by current vgpu.
+	 */
+	if (scheduler->current_vgpu == NULL) {
+		mutex_unlock(&gvt->lock);
+		intel_gvt_wait_vgpu_idle(vgpu);
+		mutex_lock(&gvt->lock);
+	}
+
+	intel_vgpu_reset_execlist(vgpu, dmlr ? ALL_ENGINES : engine_mask);
+
+	/* full GPU reset or device model level reset */
+	if (engine_mask == ALL_ENGINES || dmlr) {
+		intel_vgpu_reset_gtt(vgpu, dmlr);
+		intel_vgpu_reset_resource(vgpu);
+		intel_vgpu_reset_mmio(vgpu);
+		populate_pvinfo_page(vgpu);
+
+		if (dmlr)
+			intel_vgpu_reset_cfg_space(vgpu);
+	}
+
+	vgpu->resetting = false;
+	gvt_dbg_core("reset vgpu%d done\n", vgpu->id);
+	gvt_dbg_core("------------------------------------------\n");
+}
+
+/**
+ * intel_gvt_reset_vgpu - reset a virtual GPU (Function Level)
  * @vgpu: virtual GPU
  *
  * This function is called when user wants to reset a virtual GPU.
@@ -406,4 +403,7 @@ struct intel_vgpu *intel_gvt_create_vgpu(struct intel_gvt *gvt,
  */
 void intel_gvt_reset_vgpu(struct intel_vgpu *vgpu)
 {
+	mutex_lock(&vgpu->gvt->lock);
+	intel_gvt_reset_vgpu_locked(vgpu, true, 0);
+	mutex_unlock(&vgpu->gvt->lock);
 }
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 445fec9c2841..b2c4a0b8a627 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -2378,7 +2378,7 @@ static int intel_runtime_suspend(struct device *kdev)
 
 	assert_forcewakes_inactive(dev_priv);
 
-	if (!IS_VALLEYVIEW(dev_priv) || !IS_CHERRYVIEW(dev_priv))
+	if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv))
 		intel_hpd_poll_init(dev_priv);
 
 	DRM_DEBUG_KMS("Device suspended\n");
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 243224aeabf8..69bc3b0c4390 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1977,6 +1977,11 @@ struct drm_i915_private {
 
 	struct i915_frontbuffer_tracking fb_tracking;
 
+	struct intel_atomic_helper {
+		struct llist_head free_list;
+		struct work_struct free_work;
+	} atomic_helper;
+
 	u16 orig_clock;
 
 	bool mchbar_need_disable;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 4a31b7a891ec..4b23a7814713 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -244,14 +244,16 @@ err_phys:
 
 static void
 __i915_gem_object_release_shmem(struct drm_i915_gem_object *obj,
-				struct sg_table *pages)
+				struct sg_table *pages,
+				bool needs_clflush)
 {
 	GEM_BUG_ON(obj->mm.madv == __I915_MADV_PURGED);
 
 	if (obj->mm.madv == I915_MADV_DONTNEED)
 		obj->mm.dirty = false;
 
-	if ((obj->base.read_domains & I915_GEM_DOMAIN_CPU) == 0 &&
+	if (needs_clflush &&
+	    (obj->base.read_domains & I915_GEM_DOMAIN_CPU) == 0 &&
 	    !cpu_cache_is_coherent(obj->base.dev, obj->cache_level))
 		drm_clflush_sg(pages);
 
@@ -263,7 +265,7 @@ static void
 i915_gem_object_put_pages_phys(struct drm_i915_gem_object *obj,
 			       struct sg_table *pages)
 {
-	__i915_gem_object_release_shmem(obj, pages);
+	__i915_gem_object_release_shmem(obj, pages, false);
 
 	if (obj->mm.dirty) {
 		struct address_space *mapping = obj->base.filp->f_mapping;
@@ -593,47 +595,21 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
 		     struct drm_i915_gem_pwrite *args,
 		     struct drm_file *file)
 {
-	struct drm_device *dev = obj->base.dev;
 	void *vaddr = obj->phys_handle->vaddr + args->offset;
 	char __user *user_data = u64_to_user_ptr(args->data_ptr);
-	int ret;
 
 	/* We manually control the domain here and pretend that it
 	 * remains coherent i.e. in the GTT domain, like shmem_pwrite.
 	 */
-	lockdep_assert_held(&obj->base.dev->struct_mutex);
-	ret = i915_gem_object_wait(obj,
-				   I915_WAIT_INTERRUPTIBLE |
-				   I915_WAIT_LOCKED |
-				   I915_WAIT_ALL,
-				   MAX_SCHEDULE_TIMEOUT,
-				   to_rps_client(file));
-	if (ret)
-		return ret;
-
 	intel_fb_obj_invalidate(obj, ORIGIN_CPU);
-	if (__copy_from_user_inatomic_nocache(vaddr, user_data, args->size)) {
-		unsigned long unwritten;
-
-		/* The physical object once assigned is fixed for the lifetime
-		 * of the obj, so we can safely drop the lock and continue
-		 * to access vaddr.
-		 */
-		mutex_unlock(&dev->struct_mutex);
-		unwritten = copy_from_user(vaddr, user_data, args->size);
-		mutex_lock(&dev->struct_mutex);
-		if (unwritten) {
-			ret = -EFAULT;
-			goto out;
-		}
-	}
+	if (copy_from_user(vaddr, user_data, args->size))
+		return -EFAULT;
 
 	drm_clflush_virt_range(vaddr, args->size);
-	i915_gem_chipset_flush(to_i915(dev));
+	i915_gem_chipset_flush(to_i915(obj->base.dev));
 
-out:
 	intel_fb_obj_flush(obj, false, ORIGIN_CPU);
-	return ret;
+	return 0;
 }
 
 void *i915_gem_object_alloc(struct drm_device *dev)
@@ -2231,7 +2207,7 @@ i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj,
 	struct sgt_iter sgt_iter;
 	struct page *page;
 
-	__i915_gem_object_release_shmem(obj, pages);
+	__i915_gem_object_release_shmem(obj, pages, true);
 
 	i915_gem_gtt_finish_pages(obj, pages);
 
@@ -2304,15 +2280,6 @@ unlock:
 	mutex_unlock(&obj->mm.lock);
 }
 
-static unsigned int swiotlb_max_size(void)
-{
-#if IS_ENABLED(CONFIG_SWIOTLB)
-	return rounddown(swiotlb_nr_tbl() << IO_TLB_SHIFT, PAGE_SIZE);
-#else
-	return 0;
-#endif
-}
-
 static void i915_sg_trim(struct sg_table *orig_st)
 {
 	struct sg_table new_st;
@@ -2322,7 +2289,7 @@ static void i915_sg_trim(struct sg_table *orig_st)
 	if (orig_st->nents == orig_st->orig_nents)
 		return;
 
-	if (sg_alloc_table(&new_st, orig_st->nents, GFP_KERNEL))
+	if (sg_alloc_table(&new_st, orig_st->nents, GFP_KERNEL | __GFP_NOWARN))
 		return;
 
 	new_sg = new_st.sgl;
@@ -2360,7 +2327,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	GEM_BUG_ON(obj->base.read_domains & I915_GEM_GPU_DOMAINS);
 	GEM_BUG_ON(obj->base.write_domain & I915_GEM_GPU_DOMAINS);
 
-	max_segment = swiotlb_max_size();
+	max_segment = swiotlb_max_segment();
 	if (!max_segment)
 		max_segment = rounddown(UINT_MAX, PAGE_SIZE);
 
@@ -2728,6 +2695,7 @@ static void i915_gem_reset_engine(struct intel_engine_cs *engine)
 	struct drm_i915_gem_request *request;
 	struct i915_gem_context *incomplete_ctx;
 	struct intel_timeline *timeline;
+	unsigned long flags;
 	bool ring_hung;
 
 	if (engine->irq_seqno_barrier)
@@ -2763,13 +2731,20 @@ static void i915_gem_reset_engine(struct intel_engine_cs *engine)
 	if (i915_gem_context_is_default(incomplete_ctx))
 		return;
 
+	timeline = i915_gem_context_lookup_timeline(incomplete_ctx, engine);
+
+	spin_lock_irqsave(&engine->timeline->lock, flags);
+	spin_lock(&timeline->lock);
+
 	list_for_each_entry_continue(request, &engine->timeline->requests, link)
 		if (request->ctx == incomplete_ctx)
 			reset_request(request);
 
-	timeline = i915_gem_context_lookup_timeline(incomplete_ctx, engine);
 	list_for_each_entry(request, &timeline->requests, link)
 		reset_request(request);
+
+	spin_unlock(&timeline->lock);
+	spin_unlock_irqrestore(&engine->timeline->lock, flags);
 }
 
 void i915_gem_reset(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index bd08814b015c..d534a316a16e 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -199,6 +199,7 @@ found:
 	}
 
 	/* Unbinding will emit any required flushes */
+	ret = 0;
 	while (!list_empty(&eviction_list)) {
 		vma = list_first_entry(&eviction_list,
 				       struct i915_vma,
diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h
index e2b077df2da0..d229f47d1028 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.h
+++ b/drivers/gpu/drm/i915/i915_gem_request.h
@@ -413,6 +413,25 @@ i915_gem_active_set(struct i915_gem_active *active,
 	rcu_assign_pointer(active->request, request);
 }
 
+/**
+ * i915_gem_active_set_retire_fn - updates the retirement callback
+ * @active - the active tracker
+ * @fn - the routine called when the request is retired
+ * @mutex - struct_mutex used to guard retirements
+ *
+ * i915_gem_active_set_retire_fn() updates the function pointer that
+ * is called when the final request associated with the @active tracker
+ * is retired.
+ */
+static inline void
+i915_gem_active_set_retire_fn(struct i915_gem_active *active,
+			      i915_gem_retire_fn fn,
+			      struct mutex *mutex)
+{
+	lockdep_assert_held(mutex);
+	active->retire = fn ?: i915_gem_retire_noop;
+}
+
 static inline struct drm_i915_gem_request *
 __i915_gem_active_peek(const struct i915_gem_active *active)
 {
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index a792dcb902b5..e924a9516079 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -185,6 +185,7 @@ int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level,
 			return ret;
 	}
 
+	trace_i915_vma_bind(vma, bind_flags);
 	ret = vma->vm->bind_vma(vma, cache_level, bind_flags);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index 86ecec5601d4..588470eb8d39 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -499,6 +499,7 @@ static bool intel_crt_detect_ddc(struct drm_connector *connector)
 	struct drm_i915_private *dev_priv = to_i915(crt->base.base.dev);
 	struct edid *edid;
 	struct i2c_adapter *i2c;
+	bool ret = false;
 
 	BUG_ON(crt->base.type != INTEL_OUTPUT_ANALOG);
 
@@ -515,17 +516,17 @@ static bool intel_crt_detect_ddc(struct drm_connector *connector)
 		 */
 		if (!is_digital) {
 			DRM_DEBUG_KMS("CRT detected via DDC:0x50 [EDID]\n");
-			return true;
+			ret = true;
+		} else {
+			DRM_DEBUG_KMS("CRT not detected via DDC:0x50 [EDID reports a digital panel]\n");
 		}
-
-		DRM_DEBUG_KMS("CRT not detected via DDC:0x50 [EDID reports a digital panel]\n");
 	} else {
 		DRM_DEBUG_KMS("CRT not detected via DDC:0x50 [no valid EDID found]\n");
 	}
 
 	kfree(edid);
 
-	return false;
+	return ret;
 }
 
 static enum drm_connector_status
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 6daad8613760..f0b9aa7a0483 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -2251,6 +2251,9 @@ void intel_unpin_fb_obj(struct drm_framebuffer *fb, unsigned int rotation)
 	intel_fill_fb_ggtt_view(&view, fb, rotation);
 	vma = i915_gem_object_to_ggtt(obj, &view);
 
+	if (WARN_ON_ONCE(!vma))
+		return;
+
 	i915_vma_unpin_fence(vma);
 	i915_gem_object_unpin_from_display_plane(vma);
 }
@@ -2585,8 +2588,9 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 			 * We only keep the x/y offsets, so push all of the
 			 * gtt offset into the x/y offsets.
 			 */
-			_intel_adjust_tile_offset(&x, &y, tile_size,
-						  tile_width, tile_height, pitch_tiles,
+			_intel_adjust_tile_offset(&x, &y,
+						  tile_width, tile_height,
+						  tile_size, pitch_tiles,
 						  gtt_offset_rotated * tile_size, 0);
 
 			gtt_offset_rotated += rot_info->plane[i].width * rot_info->plane[i].height;
@@ -2967,6 +2971,9 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state)
 	unsigned int rotation = plane_state->base.rotation;
 	int ret;
 
+	if (!plane_state->base.visible)
+		return 0;
+
 	/* Rotate src coordinates to match rotated GTT view */
 	if (drm_rotation_90_or_270(rotation))
 		drm_rect_rotate(&plane_state->base.src,
@@ -6846,6 +6853,12 @@ static void intel_crtc_disable_noatomic(struct drm_crtc *crtc)
 	}
 
 	state = drm_atomic_state_alloc(crtc->dev);
+	if (!state) {
+		DRM_DEBUG_KMS("failed to disable [CRTC:%d:%s], out of memory",
+			      crtc->base.id, crtc->name);
+		return;
+	}
+
 	state->acquire_ctx = crtc->dev->mode_config.acquire_ctx;
 
 	/* Everything's already locked, -EDEADLK can't happen. */
@@ -11243,6 +11256,7 @@ found:
 	}
 
 	old->restore_state = restore_state;
+	drm_atomic_state_put(state);
 
 	/* let the connector get through one full cycle before testing */
 	intel_wait_for_vblank(dev_priv, intel_crtc->pipe);
@@ -14512,8 +14526,14 @@ intel_atomic_commit_ready(struct i915_sw_fence *fence,
 		break;
 
 	case FENCE_FREE:
-		drm_atomic_state_put(&state->base);
-		break;
+		{
+			struct intel_atomic_helper *helper =
+				&to_i915(state->base.dev)->atomic_helper;
+
+			if (llist_add(&state->freed, &helper->free_list))
+				schedule_work(&helper->free_work);
+			break;
+		}
 	}
 
 	return NOTIFY_DONE;
@@ -16392,6 +16412,18 @@ fail:
 	drm_modeset_acquire_fini(&ctx);
 }
 
+static void intel_atomic_helper_free_state(struct work_struct *work)
+{
+	struct drm_i915_private *dev_priv =
+		container_of(work, typeof(*dev_priv), atomic_helper.free_work);
+	struct intel_atomic_state *state, *next;
+	struct llist_node *freed;
+
+	freed = llist_del_all(&dev_priv->atomic_helper.free_list);
+	llist_for_each_entry_safe(state, next, freed, freed)
+		drm_atomic_state_put(&state->base);
+}
+
 int intel_modeset_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
@@ -16411,6 +16443,9 @@ int intel_modeset_init(struct drm_device *dev)
 
 	dev->mode_config.funcs = &intel_mode_funcs;
 
+	INIT_WORK(&dev_priv->atomic_helper.free_work,
+		  intel_atomic_helper_free_state);
+
 	intel_init_quirks(dev);
 
 	intel_init_pm(dev_priv);
@@ -16791,7 +16826,6 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 
 	for_each_intel_crtc(dev, crtc) {
 		struct intel_crtc_state *crtc_state = crtc->config;
-		int pixclk = 0;
 
 		__drm_atomic_helper_crtc_destroy_state(&crtc_state->base);
 		memset(crtc_state, 0, sizeof(*crtc_state));
@@ -16803,23 +16837,9 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		crtc->base.enabled = crtc_state->base.enable;
 		crtc->active = crtc_state->base.active;
 
-		if (crtc_state->base.active) {
+		if (crtc_state->base.active)
 			dev_priv->active_crtcs |= 1 << crtc->pipe;
 
-			if (INTEL_GEN(dev_priv) >= 9 || IS_BROADWELL(dev_priv))
-				pixclk = ilk_pipe_pixel_rate(crtc_state);
-			else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-				pixclk = crtc_state->base.adjusted_mode.crtc_clock;
-			else
-				WARN_ON(dev_priv->display.modeset_calc_cdclk);
-
-			/* pixel rate mustn't exceed 95% of cdclk with IPS on BDW */
-			if (IS_BROADWELL(dev_priv) && crtc_state->ips_enabled)
-				pixclk = DIV_ROUND_UP(pixclk * 100, 95);
-		}
-
-		dev_priv->min_pixclk[crtc->pipe] = pixclk;
-
 		readout_plane_state(crtc);
 
 		DRM_DEBUG_KMS("[CRTC:%d:%s] hw state readout: %s\n",
@@ -16892,6 +16912,8 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 	}
 
 	for_each_intel_crtc(dev, crtc) {
+		int pixclk = 0;
+
 		crtc->base.hwmode = crtc->config->base.adjusted_mode;
 
 		memset(&crtc->base.mode, 0, sizeof(crtc->base.mode));
@@ -16919,10 +16941,23 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			 */
 			crtc->base.state->mode.private_flags = I915_MODE_FLAG_INHERITED;
 
+			if (INTEL_GEN(dev_priv) >= 9 || IS_BROADWELL(dev_priv))
+				pixclk = ilk_pipe_pixel_rate(crtc->config);
+			else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+				pixclk = crtc->config->base.adjusted_mode.crtc_clock;
+			else
+				WARN_ON(dev_priv->display.modeset_calc_cdclk);
+
+			/* pixel rate mustn't exceed 95% of cdclk with IPS on BDW */
+			if (IS_BROADWELL(dev_priv) && crtc->config->ips_enabled)
+				pixclk = DIV_ROUND_UP(pixclk * 100, 95);
+
 			drm_calc_timestamping_constants(&crtc->base, &crtc->base.hwmode);
 			update_scanline_offset(crtc);
 		}
 
+		dev_priv->min_pixclk[crtc->pipe] = pixclk;
+
 		intel_pipe_config_sanity_check(dev_priv, crtc->config);
 	}
 }
@@ -17024,7 +17059,8 @@ void intel_display_resume(struct drm_device *dev)
 
 	if (ret)
 		DRM_ERROR("Restoring old state failed with %i\n", ret);
-	drm_atomic_state_put(state);
+	if (state)
+		drm_atomic_state_put(state);
 }
 
 void intel_modeset_gem_init(struct drm_device *dev)
@@ -17094,6 +17130,9 @@ void intel_modeset_cleanup(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
+	flush_work(&dev_priv->atomic_helper.free_work);
+	WARN_ON(!llist_empty(&dev_priv->atomic_helper.free_list));
+
 	intel_disable_gt_powersave(dev_priv);
 
 	/*
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index d9bc19be855e..0b8e8eb85c19 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -355,7 +355,8 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 				    struct intel_dp *intel_dp);
 static void
 intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
-					      struct intel_dp *intel_dp);
+					      struct intel_dp *intel_dp,
+					      bool force_disable_vdd);
 static void
 intel_dp_pps_init(struct drm_device *dev, struct intel_dp *intel_dp);
 
@@ -516,7 +517,7 @@ vlv_power_sequencer_pipe(struct intel_dp *intel_dp)
 
 	/* init power sequencer on this pipe and port */
 	intel_dp_init_panel_power_sequencer(dev, intel_dp);
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, true);
 
 	/*
 	 * Even vdd force doesn't work until we've made
@@ -553,7 +554,7 @@ bxt_power_sequencer_idx(struct intel_dp *intel_dp)
 	 * Only the HW needs to be reprogrammed, the SW state is fixed and
 	 * has been setup during connector init.
 	 */
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, false);
 
 	return 0;
 }
@@ -636,7 +637,7 @@ vlv_initial_power_sequencer_setup(struct intel_dp *intel_dp)
 		      port_name(port), pipe_name(intel_dp->pps_pipe));
 
 	intel_dp_init_panel_power_sequencer(dev, intel_dp);
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, false);
 }
 
 void intel_power_sequencer_reset(struct drm_i915_private *dev_priv)
@@ -2912,7 +2913,7 @@ static void vlv_init_panel_power_sequencer(struct intel_dp *intel_dp)
 
 	/* init power sequencer on this pipe and port */
 	intel_dp_init_panel_power_sequencer(dev, intel_dp);
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, true);
 }
 
 static void vlv_pre_enable_dp(struct intel_encoder *encoder,
@@ -5055,7 +5056,8 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 
 static void
 intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
-					      struct intel_dp *intel_dp)
+					      struct intel_dp *intel_dp,
+					      bool force_disable_vdd)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	u32 pp_on, pp_off, pp_div, port_sel = 0;
@@ -5068,6 +5070,31 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 
 	intel_pps_get_registers(dev_priv, intel_dp, &regs);
 
+	/*
+	 * On some VLV machines the BIOS can leave the VDD
+	 * enabled even on power seqeuencers which aren't
+	 * hooked up to any port. This would mess up the
+	 * power domain tracking the first time we pick
+	 * one of these power sequencers for use since
+	 * edp_panel_vdd_on() would notice that the VDD was
+	 * already on and therefore wouldn't grab the power
+	 * domain reference. Disable VDD first to avoid this.
+	 * This also avoids spuriously turning the VDD on as
+	 * soon as the new power seqeuencer gets initialized.
+	 */
+	if (force_disable_vdd) {
+		u32 pp = ironlake_get_pp_control(intel_dp);
+
+		WARN(pp & PANEL_POWER_ON, "Panel power already on\n");
+
+		if (pp & EDP_FORCE_VDD)
+			DRM_DEBUG_KMS("VDD already on, disabling first\n");
+
+		pp &= ~EDP_FORCE_VDD;
+
+		I915_WRITE(regs.pp_ctrl, pp);
+	}
+
 	pp_on = (seq->t1_t3 << PANEL_POWER_UP_DELAY_SHIFT) |
 		(seq->t8 << PANEL_LIGHT_ON_DELAY_SHIFT);
 	pp_off = (seq->t9 << PANEL_LIGHT_OFF_DELAY_SHIFT) |
@@ -5122,7 +5149,7 @@ static void intel_dp_pps_init(struct drm_device *dev,
 		vlv_initial_power_sequencer_setup(intel_dp);
 	} else {
 		intel_dp_init_panel_power_sequencer(dev, intel_dp);
-		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, false);
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index cd132c216a67..cd72ae171eeb 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -370,6 +370,8 @@ struct intel_atomic_state {
 	struct skl_wm_values wm_results;
 
 	struct i915_sw_fence commit_ready;
+
+	struct llist_node freed;
 };
 
 struct intel_plane_state {
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index beb08982dc0b..8cf2d80f2254 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -742,6 +742,9 @@ void intel_fbdev_initial_config_async(struct drm_device *dev)
 {
 	struct intel_fbdev *ifbdev = to_i915(dev)->fbdev;
 
+	if (!ifbdev)
+		return;
+
 	ifbdev->cookie = async_schedule(intel_fbdev_initial_config, ifbdev);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index d4961fa20c73..beabc17e7c8a 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -979,18 +979,8 @@ static inline int gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine,
 						uint32_t *batch,
 						uint32_t index)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
 	uint32_t l3sqc4_flush = (0x40400000 | GEN8_LQSC_FLUSH_COHERENT_LINES);
 
-	/*
-	 * WaDisableLSQCROPERFforOCL:kbl
-	 * This WA is implemented in skl_init_clock_gating() but since
-	 * this batch updates GEN8_L3SQCREG4 with default value we need to
-	 * set this bit here to retain the WA during flush.
-	 */
-	if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_E0))
-		l3sqc4_flush |= GEN8_LQSC_RO_PERF_DIS;
-
 	wa_ctx_emit(batch, index, (MI_STORE_REGISTER_MEM_GEN8 |
 				   MI_SRM_LRM_GLOBAL_GTT));
 	wa_ctx_emit_reg(batch, index, GEN8_L3SQCREG4);
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index fd0e4dac7cc1..e589e17876dc 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -216,7 +216,8 @@ static void intel_overlay_submit_request(struct intel_overlay *overlay,
 {
 	GEM_BUG_ON(i915_gem_active_peek(&overlay->last_flip,
 					&overlay->i915->drm.struct_mutex));
-	overlay->last_flip.retire = retire;
+	i915_gem_active_set_retire_fn(&overlay->last_flip, retire,
+				      &overlay->i915->drm.struct_mutex);
 	i915_gem_active_set(&overlay->last_flip, req);
 	i915_add_request(req);
 }
@@ -839,8 +840,8 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay,
 	if (ret)
 		goto out_unpin;
 
-	i915_gem_track_fb(overlay->vma->obj, new_bo,
-			  INTEL_FRONTBUFFER_OVERLAY(pipe));
+	i915_gem_track_fb(overlay->vma ? overlay->vma->obj : NULL,
+			  vma->obj, INTEL_FRONTBUFFER_OVERLAY(pipe));
 
 	overlay->old_vma = overlay->vma;
 	overlay->vma = vma;
@@ -1430,6 +1431,8 @@ void intel_setup_overlay(struct drm_i915_private *dev_priv)
 	overlay->contrast = 75;
 	overlay->saturation = 146;
 
+	init_request_active(&overlay->last_flip, NULL);
+
 	regs = intel_overlay_map_regs(overlay);
 	if (!regs)
 		goto out_unpin_bo;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index aeb637dc1fdf..91cb4c422ad5 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1095,14 +1095,6 @@ static int kbl_init_workarounds(struct intel_engine_cs *engine)
 		WA_SET_BIT_MASKED(HDC_CHICKEN0,
 				  HDC_FENCE_DEST_SLM_DISABLE);
 
-	/* GEN8_L3SQCREG4 has a dependency with WA batch so any new changes
-	 * involving this register should also be added to WA batch as required.
-	 */
-	if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_E0))
-		/* WaDisableLSQCROPERFforOCL:kbl */
-		I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
-			   GEN8_LQSC_RO_PERF_DIS);
-
 	/* WaToEnableHwFixForPushConstHWBug:kbl */
 	if (IS_KBL_REVID(dev_priv, KBL_REVID_C0, REVID_FOREVER))
 		WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2,
diff --git a/drivers/gpu/drm/meson/meson_plane.c b/drivers/gpu/drm/meson/meson_plane.c
index 4942ca090b46..7890e30eb584 100644
--- a/drivers/gpu/drm/meson/meson_plane.c
+++ b/drivers/gpu/drm/meson/meson_plane.c
@@ -51,6 +51,9 @@ static int meson_plane_atomic_check(struct drm_plane *plane,
 	struct drm_crtc_state *crtc_state;
 	struct drm_rect clip = { 0, };
 
+	if (!state->crtc)
+		return 0;
+
 	crtc_state = drm_atomic_get_crtc_state(state->state, state->crtc);
 	if (IS_ERR(crtc_state))
 		return PTR_ERR(crtc_state);
diff --git a/drivers/gpu/drm/meson/meson_venc.c b/drivers/gpu/drm/meson/meson_venc.c
index d836b2274531..f7c870172220 100644
--- a/drivers/gpu/drm/meson/meson_venc.c
+++ b/drivers/gpu/drm/meson/meson_venc.c
@@ -38,6 +38,11 @@
  * - TV Panel encoding via ENCT
  */
 
+/* HHI Registers */
+#define HHI_VDAC_CNTL0		0x2F4 /* 0xbd offset in data sheet */
+#define HHI_VDAC_CNTL1		0x2F8 /* 0xbe offset in data sheet */
+#define HHI_HDMI_PHY_CNTL0	0x3a0 /* 0xe8 offset in data sheet */
+
 struct meson_cvbs_enci_mode meson_cvbs_enci_pal = {
 	.mode_tag = MESON_VENC_MODE_CVBS_PAL,
 	.hso_begin = 3,
@@ -242,6 +247,20 @@ void meson_venc_disable_vsync(struct meson_drm *priv)
 
 void meson_venc_init(struct meson_drm *priv)
 {
+	/* Disable CVBS VDAC */
+	regmap_write(priv->hhi, HHI_VDAC_CNTL0, 0);
+	regmap_write(priv->hhi, HHI_VDAC_CNTL1, 8);
+
+	/* Power Down Dacs */
+	writel_relaxed(0xff, priv->io_base + _REG(VENC_VDAC_SETTING));
+
+	/* Disable HDMI PHY */
+	regmap_write(priv->hhi, HHI_HDMI_PHY_CNTL0, 0);
+
+	/* Disable HDMI */
+	writel_bits_relaxed(0x3, 0,
+			    priv->io_base + _REG(VPU_HDMI_SETTING));
+
 	/* Disable all encoders */
 	writel_relaxed(0, priv->io_base + _REG(ENCI_VIDEO_EN));
 	writel_relaxed(0, priv->io_base + _REG(ENCP_VIDEO_EN));
diff --git a/drivers/gpu/drm/meson/meson_venc_cvbs.c b/drivers/gpu/drm/meson/meson_venc_cvbs.c
index c809c085fd78..a2bcc70a03ef 100644
--- a/drivers/gpu/drm/meson/meson_venc_cvbs.c
+++ b/drivers/gpu/drm/meson/meson_venc_cvbs.c
@@ -167,7 +167,7 @@ static void meson_venc_cvbs_encoder_disable(struct drm_encoder *encoder)
 
 	/* Disable CVBS VDAC */
 	regmap_write(priv->hhi, HHI_VDAC_CNTL0, 0);
-	regmap_write(priv->hhi, HHI_VDAC_CNTL1, 0);
+	regmap_write(priv->hhi, HHI_VDAC_CNTL1, 8);
 }
 
 static void meson_venc_cvbs_encoder_enable(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index a18126150e11..686a580c711a 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -213,7 +213,14 @@ void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 void adreno_flush(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
-	uint32_t wptr = get_wptr(gpu->rb);
+	uint32_t wptr;
+
+	/*
+	 * Mask wptr value that we calculate to fit in the HW range. This is
+	 * to account for the possibility that the last command fit exactly into
+	 * the ringbuffer and rb->next hasn't wrapped to zero yet
+	 */
+	wptr = get_wptr(gpu->rb) & ((gpu->rb->size / 4) - 1);
 
 	/* ensure writes to ringbuffer have hit system memory: */
 	mb();
@@ -338,7 +345,6 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 {
 	struct adreno_platform_config *config = pdev->dev.platform_data;
 	struct msm_gpu *gpu = &adreno_gpu->base;
-	struct msm_mmu *mmu;
 	int ret;
 
 	adreno_gpu->funcs = funcs;
@@ -378,8 +384,8 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		return ret;
 	}
 
-	mmu = gpu->aspace->mmu;
-	if (mmu) {
+	if (gpu->aspace && gpu->aspace->mmu) {
+		struct msm_mmu *mmu = gpu->aspace->mmu;
 		ret = mmu->funcs->attach(mmu, iommu_ports,
 				ARRAY_SIZE(iommu_ports));
 		if (ret)
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
index 5f6cd8745dbc..c396d459a9d0 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
@@ -119,13 +119,7 @@ static void mdp5_prepare_commit(struct msm_kms *kms, struct drm_atomic_state *st
 
 static void mdp5_complete_commit(struct msm_kms *kms, struct drm_atomic_state *state)
 {
-	int i;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
-	struct drm_plane *plane;
-	struct drm_plane_state *plane_state;
-
-	for_each_plane_in_state(state, plane, plane_state, i)
-		mdp5_plane_complete_commit(plane, plane_state);
 
 	if (mdp5_kms->smp)
 		mdp5_smp_complete_commit(mdp5_kms->smp, &mdp5_kms->state->smp);
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
index 17b0cc101171..cdfc63d90c7b 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
@@ -104,8 +104,6 @@ struct mdp5_plane_state {
 
 	/* assigned by crtc blender */
 	enum mdp_mixer_stage_id stage;
-
-	bool pending : 1;
 };
 #define to_mdp5_plane_state(x) \
 		container_of(x, struct mdp5_plane_state, base)
@@ -232,8 +230,6 @@ int mdp5_irq_domain_init(struct mdp5_kms *mdp5_kms);
 void mdp5_irq_domain_fini(struct mdp5_kms *mdp5_kms);
 
 uint32_t mdp5_plane_get_flush(struct drm_plane *plane);
-void mdp5_plane_complete_commit(struct drm_plane *plane,
-	struct drm_plane_state *state);
 enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane);
 struct drm_plane *mdp5_plane_init(struct drm_device *dev, bool primary);
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
index c099da7bc212..25d9d0a97156 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
@@ -179,7 +179,6 @@ mdp5_plane_atomic_print_state(struct drm_printer *p,
 	drm_printf(p, "\tzpos=%u\n", pstate->zpos);
 	drm_printf(p, "\talpha=%u\n", pstate->alpha);
 	drm_printf(p, "\tstage=%s\n", stage2name(pstate->stage));
-	drm_printf(p, "\tpending=%u\n", pstate->pending);
 }
 
 static void mdp5_plane_reset(struct drm_plane *plane)
@@ -220,8 +219,6 @@ mdp5_plane_duplicate_state(struct drm_plane *plane)
 	if (mdp5_state && mdp5_state->base.fb)
 		drm_framebuffer_reference(mdp5_state->base.fb);
 
-	mdp5_state->pending = false;
-
 	return &mdp5_state->base;
 }
 
@@ -288,13 +285,6 @@ static int mdp5_plane_atomic_check(struct drm_plane *plane,
 	DBG("%s: check (%d -> %d)", plane->name,
 			plane_enabled(old_state), plane_enabled(state));
 
-	/* We don't allow faster-than-vblank updates.. if we did add this
-	 * some day, we would need to disallow in cases where hwpipe
-	 * changes
-	 */
-	if (WARN_ON(to_mdp5_plane_state(old_state)->pending))
-		return -EBUSY;
-
 	max_width = config->hw->lm.max_width << 16;
 	max_height = config->hw->lm.max_height << 16;
 
@@ -370,12 +360,9 @@ static void mdp5_plane_atomic_update(struct drm_plane *plane,
 				     struct drm_plane_state *old_state)
 {
 	struct drm_plane_state *state = plane->state;
-	struct mdp5_plane_state *mdp5_state = to_mdp5_plane_state(state);
 
 	DBG("%s: update", plane->name);
 
-	mdp5_state->pending = true;
-
 	if (plane_enabled(state)) {
 		int ret;
 
@@ -851,15 +838,6 @@ uint32_t mdp5_plane_get_flush(struct drm_plane *plane)
 	return pstate->hwpipe->flush_mask;
 }
 
-/* called after vsync in thread context */
-void mdp5_plane_complete_commit(struct drm_plane *plane,
-	struct drm_plane_state *state)
-{
-	struct mdp5_plane_state *pstate = to_mdp5_plane_state(plane->state);
-
-	pstate->pending = false;
-}
-
 /* initialize plane */
 struct drm_plane *mdp5_plane_init(struct drm_device *dev, bool primary)
 {
diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index d8bc59c7e261..8098677a3916 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -294,6 +294,8 @@ put_iova(struct drm_gem_object *obj)
 	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
 
 	for (id = 0; id < ARRAY_SIZE(msm_obj->domain); id++) {
+		if (!priv->aspace[id])
+			continue;
 		msm_gem_unmap_vma(priv->aspace[id],
 				&msm_obj->domain[id], msm_obj->sgt);
 	}
diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c
index 166e84e4f0d4..489676568a10 100644
--- a/drivers/gpu/drm/msm/msm_gem_submit.c
+++ b/drivers/gpu/drm/msm/msm_gem_submit.c
@@ -106,7 +106,8 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
 			pagefault_disable();
 		}
 
-		if (submit_bo.flags & ~MSM_SUBMIT_BO_FLAGS) {
+		if ((submit_bo.flags & ~MSM_SUBMIT_BO_FLAGS) ||
+			!(submit_bo.flags & MSM_SUBMIT_BO_FLAGS)) {
 			DRM_ERROR("invalid flags: %x\n", submit_bo.flags);
 			ret = -EINVAL;
 			goto out_unlock;
@@ -290,7 +291,7 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
 {
 	uint32_t i, last_offset = 0;
 	uint32_t *ptr;
-	int ret;
+	int ret = 0;
 
 	if (offset % 4) {
 		DRM_ERROR("non-aligned cmdstream buffer: %u\n", offset);
@@ -318,12 +319,13 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
 
 		ret = copy_from_user(&submit_reloc, userptr, sizeof(submit_reloc));
 		if (ret)
-			return -EFAULT;
+			goto out;
 
 		if (submit_reloc.submit_offset % 4) {
 			DRM_ERROR("non-aligned reloc offset: %u\n",
 					submit_reloc.submit_offset);
-			return -EINVAL;
+			ret = -EINVAL;
+			goto out;
 		}
 
 		/* offset in dwords: */
@@ -332,12 +334,13 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
 		if ((off >= (obj->base.size / 4)) ||
 				(off < last_offset)) {
 			DRM_ERROR("invalid offset %u at reloc %u\n", off, i);
-			return -EINVAL;
+			ret = -EINVAL;
+			goto out;
 		}
 
 		ret = submit_bo(submit, submit_reloc.reloc_idx, NULL, &iova, &valid);
 		if (ret)
-			return ret;
+			goto out;
 
 		if (valid)
 			continue;
@@ -354,9 +357,10 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
 		last_offset = off;
 	}
 
+out:
 	msm_gem_put_vaddr_locked(&obj->base);
 
-	return 0;
+	return ret;
 }
 
 static void submit_cleanup(struct msm_gem_submit *submit)
diff --git a/drivers/gpu/drm/msm/msm_ringbuffer.c b/drivers/gpu/drm/msm/msm_ringbuffer.c
index f326cf6a32e6..67b34e069abf 100644
--- a/drivers/gpu/drm/msm/msm_ringbuffer.c
+++ b/drivers/gpu/drm/msm/msm_ringbuffer.c
@@ -23,7 +23,8 @@ struct msm_ringbuffer *msm_ringbuffer_new(struct msm_gpu *gpu, int size)
 	struct msm_ringbuffer *ring;
 	int ret;
 
-	size = ALIGN(size, 4);   /* size should be dword aligned */
+	if (WARN_ON(!is_power_of_2(size)))
+		return ERR_PTR(-EINVAL);
 
 	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
 	if (!ring) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index cef08da1da4e..6a157763dfc3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -411,7 +411,8 @@ nouveau_display_init(struct drm_device *dev)
 		return ret;
 
 	/* enable polling for external displays */
-	drm_kms_helper_poll_enable(dev);
+	if (!dev->mode_config.poll_enabled)
+		drm_kms_helper_poll_enable(dev);
 
 	/* enable hotplug interrupts */
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 59348fc41c77..bc85a45f91cd 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -773,7 +773,10 @@ nouveau_pmops_runtime_resume(struct device *dev)
 	pci_set_master(pdev);
 
 	ret = nouveau_do_resume(drm_dev, true);
-	drm_kms_helper_poll_enable(drm_dev);
+
+	if (!drm_dev->mode_config.poll_enabled)
+		drm_kms_helper_poll_enable(drm_dev);
+
 	/* do magic */
 	nvif_mask(&device->object, 0x088488, (1 << 25), (1 << 25));
 	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON);
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index 8d5ed5bfdacb..42c1fa53d431 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -165,6 +165,8 @@ struct nouveau_drm {
 	struct backlight_device *backlight;
 	struct list_head bl_connectors;
 	struct work_struct hpd_work;
+	struct work_struct fbcon_work;
+	int fbcon_new_state;
 #ifdef CONFIG_ACPI
 	struct notifier_block acpi_nb;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
index 2f2a3dcd4ad7..fa2d0a978ccc 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
@@ -470,19 +470,43 @@ static const struct drm_fb_helper_funcs nouveau_fbcon_helper_funcs = {
 	.fb_probe = nouveau_fbcon_create,
 };
 
+static void
+nouveau_fbcon_set_suspend_work(struct work_struct *work)
+{
+	struct nouveau_drm *drm = container_of(work, typeof(*drm), fbcon_work);
+	int state = READ_ONCE(drm->fbcon_new_state);
+
+	if (state == FBINFO_STATE_RUNNING)
+		pm_runtime_get_sync(drm->dev->dev);
+
+	console_lock();
+	if (state == FBINFO_STATE_RUNNING)
+		nouveau_fbcon_accel_restore(drm->dev);
+	drm_fb_helper_set_suspend(&drm->fbcon->helper, state);
+	if (state != FBINFO_STATE_RUNNING)
+		nouveau_fbcon_accel_save_disable(drm->dev);
+	console_unlock();
+
+	if (state == FBINFO_STATE_RUNNING) {
+		pm_runtime_mark_last_busy(drm->dev->dev);
+		pm_runtime_put_sync(drm->dev->dev);
+	}
+}
+
 void
 nouveau_fbcon_set_suspend(struct drm_device *dev, int state)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
-	if (drm->fbcon) {
-		console_lock();
-		if (state == FBINFO_STATE_RUNNING)
-			nouveau_fbcon_accel_restore(dev);
-		drm_fb_helper_set_suspend(&drm->fbcon->helper, state);
-		if (state != FBINFO_STATE_RUNNING)
-			nouveau_fbcon_accel_save_disable(dev);
-		console_unlock();
-	}
+
+	if (!drm->fbcon)
+		return;
+
+	drm->fbcon_new_state = state;
+	/* Since runtime resume can happen as a result of a sysfs operation,
+	 * it's possible we already have the console locked. So handle fbcon
+	 * init/deinit from a seperate work thread
+	 */
+	schedule_work(&drm->fbcon_work);
 }
 
 int
@@ -502,6 +526,7 @@ nouveau_fbcon_init(struct drm_device *dev)
 		return -ENOMEM;
 
 	drm->fbcon = fbcon;
+	INIT_WORK(&drm->fbcon_work, nouveau_fbcon_set_suspend_work);
 
 	drm_fb_helper_prepare(dev, &fbcon->helper, &nouveau_fbcon_helper_funcs);
 
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index 00ea0002b539..e0c143b865f3 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -366,11 +366,10 @@ static void
 radeon_pci_shutdown(struct pci_dev *pdev)
 {
 	/* if we are running in a VM, make sure the device
-	 * torn down properly on reboot/shutdown.
-	 * unfortunately we can't detect certain
-	 * hypervisors so just do this all the time.
+	 * torn down properly on reboot/shutdown
 	 */
-	radeon_pci_remove(pdev);
+	if (radeon_device_is_virtual())
+		radeon_pci_remove(pdev);
 }
 
 static int radeon_pmops_suspend(struct device *dev)
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index ad4d7b8b8322..414776811e71 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -50,7 +50,6 @@ MODULE_FIRMWARE("radeon/tahiti_ce.bin");
 MODULE_FIRMWARE("radeon/tahiti_mc.bin");
 MODULE_FIRMWARE("radeon/tahiti_rlc.bin");
 MODULE_FIRMWARE("radeon/tahiti_smc.bin");
-MODULE_FIRMWARE("radeon/tahiti_k_smc.bin");
 
 MODULE_FIRMWARE("radeon/PITCAIRN_pfp.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_me.bin");
@@ -115,6 +114,9 @@ MODULE_FIRMWARE("radeon/hainan_mc.bin");
 MODULE_FIRMWARE("radeon/hainan_rlc.bin");
 MODULE_FIRMWARE("radeon/hainan_smc.bin");
 MODULE_FIRMWARE("radeon/hainan_k_smc.bin");
+MODULE_FIRMWARE("radeon/banks_k_2_smc.bin");
+
+MODULE_FIRMWARE("radeon/si58_mc.bin");
 
 static u32 si_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh);
 static void si_pcie_gen3_enable(struct radeon_device *rdev);
@@ -1651,15 +1653,14 @@ static int si_init_microcode(struct radeon_device *rdev)
 	int err;
 	int new_fw = 0;
 	bool new_smc = false;
+	bool si58_fw = false;
+	bool banks2_fw = false;
 
 	DRM_DEBUG("\n");
 
 	switch (rdev->family) {
 	case CHIP_TAHITI:
 		chip_name = "TAHITI";
-		/* XXX: figure out which Tahitis need the new ucode */
-		if (0)
-			new_smc = true;
 		new_chip_name = "tahiti";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
@@ -1671,12 +1672,9 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_PITCAIRN:
 		chip_name = "PITCAIRN";
-		if ((rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->device == 0x6810) ||
-		    (rdev->pdev->device == 0x6811) ||
-		    (rdev->pdev->device == 0x6816) ||
-		    (rdev->pdev->device == 0x6817) ||
-		    (rdev->pdev->device == 0x6806))
+		if ((rdev->pdev->revision == 0x81) &&
+		    ((rdev->pdev->device == 0x6810) ||
+		     (rdev->pdev->device == 0x6811)))
 			new_smc = true;
 		new_chip_name = "pitcairn";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
@@ -1689,15 +1687,15 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_VERDE:
 		chip_name = "VERDE";
-		if ((rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->revision == 0x83) ||
-		    (rdev->pdev->revision == 0x87) ||
-		    (rdev->pdev->device == 0x6820) ||
-		    (rdev->pdev->device == 0x6821) ||
-		    (rdev->pdev->device == 0x6822) ||
-		    (rdev->pdev->device == 0x6823) ||
-		    (rdev->pdev->device == 0x682A) ||
-		    (rdev->pdev->device == 0x682B))
+		if (((rdev->pdev->device == 0x6820) &&
+		     ((rdev->pdev->revision == 0x81) ||
+		      (rdev->pdev->revision == 0x83))) ||
+		    ((rdev->pdev->device == 0x6821) &&
+		     ((rdev->pdev->revision == 0x83) ||
+		      (rdev->pdev->revision == 0x87))) ||
+		    ((rdev->pdev->revision == 0x87) &&
+		     ((rdev->pdev->device == 0x6823) ||
+		      (rdev->pdev->device == 0x682b))))
 			new_smc = true;
 		new_chip_name = "verde";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
@@ -1710,13 +1708,13 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_OLAND:
 		chip_name = "OLAND";
-		if ((rdev->pdev->revision == 0xC7) ||
-		    (rdev->pdev->revision == 0x80) ||
-		    (rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->revision == 0x83) ||
-		    (rdev->pdev->revision == 0x87) ||
-		    (rdev->pdev->device == 0x6604) ||
-		    (rdev->pdev->device == 0x6605))
+		if (((rdev->pdev->revision == 0x81) &&
+		     ((rdev->pdev->device == 0x6600) ||
+		      (rdev->pdev->device == 0x6604) ||
+		      (rdev->pdev->device == 0x6605) ||
+		      (rdev->pdev->device == 0x6610))) ||
+		    ((rdev->pdev->revision == 0x83) &&
+		     (rdev->pdev->device == 0x6610)))
 			new_smc = true;
 		new_chip_name = "oland";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
@@ -1728,13 +1726,17 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_HAINAN:
 		chip_name = "HAINAN";
-		if ((rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->revision == 0x83) ||
-		    (rdev->pdev->revision == 0xC3) ||
-		    (rdev->pdev->device == 0x6664) ||
-		    (rdev->pdev->device == 0x6665) ||
-		    (rdev->pdev->device == 0x6667))
+		if (((rdev->pdev->revision == 0x81) &&
+		     (rdev->pdev->device == 0x6660)) ||
+		    ((rdev->pdev->revision == 0x83) &&
+		     ((rdev->pdev->device == 0x6660) ||
+		      (rdev->pdev->device == 0x6663) ||
+		      (rdev->pdev->device == 0x6665) ||
+		      (rdev->pdev->device == 0x6667))))
 			new_smc = true;
+		else if ((rdev->pdev->revision == 0xc3) &&
+			 (rdev->pdev->device == 0x6665))
+			banks2_fw = true;
 		new_chip_name = "hainan";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
@@ -1746,6 +1748,10 @@ static int si_init_microcode(struct radeon_device *rdev)
 	default: BUG();
 	}
 
+	/* this memory configuration requires special firmware */
+	if (((RREG32(MC_SEQ_MISC0) & 0xff000000) >> 24) == 0x58)
+		si58_fw = true;
+
 	DRM_INFO("Loading %s Microcode\n", new_chip_name);
 
 	snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", new_chip_name);
@@ -1849,7 +1855,10 @@ static int si_init_microcode(struct radeon_device *rdev)
 		}
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", new_chip_name);
+	if (si58_fw)
+		snprintf(fw_name, sizeof(fw_name), "radeon/si58_mc.bin");
+	else
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", new_chip_name);
 	err = request_firmware(&rdev->mc_fw, fw_name, rdev->dev);
 	if (err) {
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc2.bin", chip_name);
@@ -1880,7 +1889,9 @@ static int si_init_microcode(struct radeon_device *rdev)
 		}
 	}
 
-	if (new_smc)
+	if (banks2_fw)
+		snprintf(fw_name, sizeof(fw_name), "radeon/banks_k_2_smc.bin");
+	else if (new_smc)
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_k_smc.bin", new_chip_name);
 	else
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name);
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 8b5e697f2549..2944916f7102 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -3008,30 +3008,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 		    (rdev->pdev->device == 0x6817) ||
 		    (rdev->pdev->device == 0x6806))
 			max_mclk = 120000;
-	} else if (rdev->family == CHIP_VERDE) {
-		if ((rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->revision == 0x83) ||
-		    (rdev->pdev->revision == 0x87) ||
-		    (rdev->pdev->device == 0x6820) ||
-		    (rdev->pdev->device == 0x6821) ||
-		    (rdev->pdev->device == 0x6822) ||
-		    (rdev->pdev->device == 0x6823) ||
-		    (rdev->pdev->device == 0x682A) ||
-		    (rdev->pdev->device == 0x682B)) {
-			max_sclk = 75000;
-			max_mclk = 80000;
-		}
-	} else if (rdev->family == CHIP_OLAND) {
-		if ((rdev->pdev->revision == 0xC7) ||
-		    (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;
-			max_mclk = 80000;
-		}
 	} else if (rdev->family == CHIP_HAINAN) {
 		if ((rdev->pdev->revision == 0x81) ||
 		    (rdev->pdev->revision == 0x83) ||
@@ -3040,7 +3016,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 		    (rdev->pdev->device == 0x6665) ||
 		    (rdev->pdev->device == 0x6667)) {
 			max_sclk = 75000;
-			max_mclk = 80000;
 		}
 	}
 	/* Apply dpm quirks */
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
index 725dffad5640..6dfdb145f3bb 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
@@ -856,7 +856,7 @@ irqreturn_t tilcdc_crtc_irq(struct drm_crtc *crtc)
 	struct tilcdc_crtc *tilcdc_crtc = to_tilcdc_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct tilcdc_drm_private *priv = dev->dev_private;
-	uint32_t stat;
+	uint32_t stat, reg;
 
 	stat = tilcdc_read_irqstatus(dev);
 	tilcdc_clear_irqstatus(dev, stat);
@@ -921,17 +921,26 @@ irqreturn_t tilcdc_crtc_irq(struct drm_crtc *crtc)
 		dev_err_ratelimited(dev->dev, "%s(0x%08x): Sync lost",
 				    __func__, stat);
 		tilcdc_crtc->frame_intact = false;
-		if (tilcdc_crtc->sync_lost_count++ >
-		    SYNC_LOST_COUNT_LIMIT) {
-			dev_err(dev->dev, "%s(0x%08x): Sync lost flood detected, recovering", __func__, stat);
-			queue_work(system_wq, &tilcdc_crtc->recover_work);
-			if (priv->rev == 1)
+		if (priv->rev == 1) {
+			reg = tilcdc_read(dev, LCDC_RASTER_CTRL_REG);
+			if (reg & LCDC_RASTER_ENABLE) {
 				tilcdc_clear(dev, LCDC_RASTER_CTRL_REG,
-					     LCDC_V1_SYNC_LOST_INT_ENA);
-			else
+					     LCDC_RASTER_ENABLE);
+				tilcdc_set(dev, LCDC_RASTER_CTRL_REG,
+					   LCDC_RASTER_ENABLE);
+			}
+		} else {
+			if (tilcdc_crtc->sync_lost_count++ >
+			    SYNC_LOST_COUNT_LIMIT) {
+				dev_err(dev->dev,
+					"%s(0x%08x): Sync lost flood detected, recovering",
+					__func__, stat);
+				queue_work(system_wq,
+					   &tilcdc_crtc->recover_work);
 				tilcdc_write(dev, LCDC_INT_ENABLE_CLR_REG,
 					     LCDC_SYNC_LOST);
-			tilcdc_crtc->sync_lost_count = 0;
+				tilcdc_crtc->sync_lost_count = 0;
+			}
 		}
 	}
 
diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c
index a0fd3e66bc4b..7aadce1f7e7a 100644
--- a/drivers/gpu/drm/vc4/vc4_crtc.c
+++ b/drivers/gpu/drm/vc4/vc4_crtc.c
@@ -839,7 +839,7 @@ static void vc4_crtc_destroy_state(struct drm_crtc *crtc,
 
 	}
 
-	__drm_atomic_helper_crtc_destroy_state(state);
+	drm_atomic_helper_crtc_destroy_state(crtc, state);
 }
 
 static const struct drm_crtc_funcs vc4_crtc_funcs = {
diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c
index db920771bfb5..ab3016982466 100644
--- a/drivers/gpu/drm/vc4/vc4_gem.c
+++ b/drivers/gpu/drm/vc4/vc4_gem.c
@@ -594,12 +594,14 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec)
 					  args->shader_rec_count);
 	struct vc4_bo *bo;
 
-	if (uniforms_offset < shader_rec_offset ||
+	if (shader_rec_offset < args->bin_cl_size ||
+	    uniforms_offset < shader_rec_offset ||
 	    exec_size < uniforms_offset ||
 	    args->shader_rec_count >= (UINT_MAX /
 					  sizeof(struct vc4_shader_state)) ||
 	    temp_size < exec_size) {
 		DRM_ERROR("overflow in exec arguments\n");
+		ret = -EINVAL;
 		goto fail;
 	}
 
diff --git a/drivers/gpu/drm/vc4/vc4_render_cl.c b/drivers/gpu/drm/vc4/vc4_render_cl.c
index 08886a309757..5cdd003605f5 100644
--- a/drivers/gpu/drm/vc4/vc4_render_cl.c
+++ b/drivers/gpu/drm/vc4/vc4_render_cl.c
@@ -461,7 +461,7 @@ static int vc4_rcl_surface_setup(struct vc4_exec_info *exec,
 		}
 
 		ret = vc4_full_res_bounds_check(exec, *obj, surf);
-		if (!ret)
+		if (ret)
 			return ret;
 
 		return 0;
diff --git a/drivers/gpu/drm/virtio/virtgpu_fb.c b/drivers/gpu/drm/virtio/virtgpu_fb.c
index dd21f950e129..cde9f3758106 100644
--- a/drivers/gpu/drm/virtio/virtgpu_fb.c
+++ b/drivers/gpu/drm/virtio/virtgpu_fb.c
@@ -331,7 +331,7 @@ static int virtio_gpufb_create(struct drm_fb_helper *helper,
 	info->fbops = &virtio_gpufb_ops;
 	info->pixmap.flags = FB_PIXMAP_SYSTEM;
 
-	info->screen_base = obj->vmap;
+	info->screen_buffer = obj->vmap;
 	info->screen_size = obj->gem_base.size;
 	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->depth);
 	drm_fb_helper_fill_var(info, &vfbdev->helper,