Merge tag 'drm-fixes-2018-12-14' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "While I hoped things would calm down, the world hasn't joined with me,
  but it's a few things scattered over a wide area. The i915 workarounds
  regression fix is probably the largest, the rest are more usual sized.
  We also get some new AMD PCI IDs.

  There is also a patch in here to MAINTAINERS to added Daniel as an
  official DRM toplevel co-maintainer, he's decided he wants to step up
  and share the glory, and he'll likely process next weeks fixes while
  I'm away on holidays.

  Summary:

  amdgpu:

   - some new PCI IDs

   - fixed firmware image updates

   - power management fixes

   - locking warning fix

  nouveau:

   - framebuffer flushing fix

   - memory leak fix

   - tegra device init regression fix

  vmwgfx:

   - OOM kernel memory fix

   - excess return in function fix

  i915:

   - the biggest fix is a regression fix where workarounds weren't
     getting reapplied after a gpu hang causing further crashing, this
     fixes the workaround application to make it happen again

   - GPU hang fixes for Braswell and some GEN3 GPUs

   - GVT fix for broadwell tiling

  rockchip:

   - revert to fix a regression causing a WARN on shutdown

  mediatek:

   - avoid crash attaching to non-existant bridges"

* tag 'drm-fixes-2018-12-14' of git://anongit.freedesktop.org/drm/drm: (23 commits)
  drm/vmwgfx: Protect from excessive execbuf kernel memory allocations v3
  MAINTAINERS: Daniel for drm co-maintainer
  drm/amdgpu: drop fclk/gfxclk ratio setting
  drm/vmwgfx: remove redundant return ret statement
  drm/i915: Flush GPU relocs harder for gen3
  drm/i915: Allocate a common scratch page
  drm/i915/execlists: Apply a full mb before execution for Braswell
  drm/nouveau/kms: Fix memory leak in nv50_mstm_del()
  drm/nouveau/kms/nv50-: also flush fb writes when rewinding push buffer
  drm/amdgpu: Fix DEBUG_LOCKS_WARN_ON(depth <= 0) in amdgpu_ctx.lock
  Revert "drm/rockchip: Allow driver to be shutdown on reboot/kexec"
  drm/nouveau/drm/nouveau: tegra: Call nouveau_drm_device_init()
  drm/amdgpu/powerplay: Apply avfs cks-off voltages on VI
  drm/amdgpu: update SMC firmware image for polaris10 variants
  drm/amdkfd: add new vega20 pci id
  drm/amdkfd: add new vega10 pci ids
  drm/amdgpu: add some additional vega20 pci ids
  drm/amdgpu: add some additional vega10 pci ids
  drm/amdgpu: update smu firmware images for VI variants (v2)
  drm/i915: Introduce per-engine workarounds
  ...

31 files changed, 718 insertions, 335 deletions

diff --git a/MAINTAINERS b/MAINTAINERS
index 0767f1d1b579..3318f30903b2 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -4847,6 +4847,7 @@ F:	include/uapi/drm/vmwgfx_drm.h
 
 DRM DRIVERS
 M:	David Airlie <airlied@linux.ie>
+M:	Daniel Vetter <daniel@ffwll.ch>
 L:	dri-devel@lists.freedesktop.org
 T:	git git://anongit.freedesktop.org/drm/drm
 B:	https://bugs.freedesktop.org/
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 8816c697b205..387f1cf1dc20 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -330,7 +330,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 			case CHIP_TOPAZ:
 				if (((adev->pdev->device == 0x6900) && (adev->pdev->revision == 0x81)) ||
 				    ((adev->pdev->device == 0x6900) && (adev->pdev->revision == 0x83)) ||
-				    ((adev->pdev->device == 0x6907) && (adev->pdev->revision == 0x87))) {
+				    ((adev->pdev->device == 0x6907) && (adev->pdev->revision == 0x87)) ||
+				    ((adev->pdev->device == 0x6900) && (adev->pdev->revision == 0xD1)) ||
+				    ((adev->pdev->device == 0x6900) && (adev->pdev->revision == 0xD3))) {
 					info->is_kicker = true;
 					strcpy(fw_name, "amdgpu/topaz_k_smc.bin");
 				} else
@@ -351,7 +353,6 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 				if (type == CGS_UCODE_ID_SMU) {
 					if (((adev->pdev->device == 0x67ef) &&
 					     ((adev->pdev->revision == 0xe0) ||
-					      (adev->pdev->revision == 0xe2) ||
 					      (adev->pdev->revision == 0xe5))) ||
 					    ((adev->pdev->device == 0x67ff) &&
 					     ((adev->pdev->revision == 0xcf) ||
@@ -359,8 +360,13 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 					      (adev->pdev->revision == 0xff)))) {
 						info->is_kicker = true;
 						strcpy(fw_name, "amdgpu/polaris11_k_smc.bin");
-					} else
+					} else if ((adev->pdev->device == 0x67ef) &&
+						   (adev->pdev->revision == 0xe2)) {
+						info->is_kicker = true;
+						strcpy(fw_name, "amdgpu/polaris11_k2_smc.bin");
+					} else {
 						strcpy(fw_name, "amdgpu/polaris11_smc.bin");
+					}
 				} else if (type == CGS_UCODE_ID_SMU_SK) {
 					strcpy(fw_name, "amdgpu/polaris11_smc_sk.bin");
 				}
@@ -375,17 +381,35 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 					      (adev->pdev->revision == 0xe7) ||
 					      (adev->pdev->revision == 0xef))) ||
 					    ((adev->pdev->device == 0x6fdf) &&
-					     (adev->pdev->revision == 0xef))) {
+					     ((adev->pdev->revision == 0xef) ||
+					      (adev->pdev->revision == 0xff)))) {
 						info->is_kicker = true;
 						strcpy(fw_name, "amdgpu/polaris10_k_smc.bin");
-					} else
+					} else if ((adev->pdev->device == 0x67df) &&
+						   ((adev->pdev->revision == 0xe1) ||
+						    (adev->pdev->revision == 0xf7))) {
+						info->is_kicker = true;
+						strcpy(fw_name, "amdgpu/polaris10_k2_smc.bin");
+					} else {
 						strcpy(fw_name, "amdgpu/polaris10_smc.bin");
+					}
 				} 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");
+				if (((adev->pdev->device == 0x6987) &&
+				     ((adev->pdev->revision == 0xc0) ||
+				      (adev->pdev->revision == 0xc3))) ||
+				    ((adev->pdev->device == 0x6981) &&
+				     ((adev->pdev->revision == 0x00) ||
+				      (adev->pdev->revision == 0x01) ||
+				      (adev->pdev->revision == 0x10)))) {
+					info->is_kicker = true;
+					strcpy(fw_name, "amdgpu/polaris12_k_smc.bin");
+				} else {
+					strcpy(fw_name, "amdgpu/polaris12_smc.bin");
+				}
 				break;
 			case CHIP_VEGAM:
 				strcpy(fw_name, "amdgpu/vegam_smc.bin");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 663043c8f0f5..0acc8dee2cb8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -124,14 +124,14 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs
 		goto free_chunk;
 	}
 
+	mutex_lock(&p->ctx->lock);
+
 	/* skip guilty context job */
 	if (atomic_read(&p->ctx->guilty) == 1) {
 		ret = -ECANCELED;
 		goto free_chunk;
 	}
 
-	mutex_lock(&p->ctx->lock);
-
 	/* get chunks */
 	chunk_array_user = u64_to_user_ptr(cs->in.chunks);
 	if (copy_from_user(chunk_array, chunk_array_user,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 8de55f7f1a3a..74b611e8a1b1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -872,7 +872,13 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x6864, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x6867, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x6868, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
+	{0x1002, 0x6869, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
+	{0x1002, 0x686a, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
+	{0x1002, 0x686b, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x686c, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
+	{0x1002, 0x686d, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
+	{0x1002, 0x686e, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
+	{0x1002, 0x686f, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x687f, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	/* Vega 12 */
 	{0x1002, 0x69A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
@@ -885,6 +891,7 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x66A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
 	{0x1002, 0x66A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
 	{0x1002, 0x66A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66A4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
 	{0x1002, 0x66A7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
 	{0x1002, 0x66AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
 	/* Raven */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index a9f18ea7e354..e4ded890b1cb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -337,12 +337,19 @@ static const struct kfd_deviceid supported_devices[] = {
 	{ 0x6864, &vega10_device_info },	/* Vega10 */
 	{ 0x6867, &vega10_device_info },	/* Vega10 */
 	{ 0x6868, &vega10_device_info },	/* Vega10 */
+	{ 0x6869, &vega10_device_info },	/* Vega10 */
+	{ 0x686A, &vega10_device_info },	/* Vega10 */
+	{ 0x686B, &vega10_device_info },	/* Vega10 */
 	{ 0x686C, &vega10_vf_device_info },	/* Vega10  vf*/
+	{ 0x686D, &vega10_device_info },	/* Vega10 */
+	{ 0x686E, &vega10_device_info },	/* Vega10 */
+	{ 0x686F, &vega10_device_info },	/* Vega10 */
 	{ 0x687F, &vega10_device_info },	/* Vega10 */
 	{ 0x66a0, &vega20_device_info },	/* Vega20 */
 	{ 0x66a1, &vega20_device_info },	/* Vega20 */
 	{ 0x66a2, &vega20_device_info },	/* Vega20 */
 	{ 0x66a3, &vega20_device_info },	/* Vega20 */
+	{ 0x66a4, &vega20_device_info },	/* Vega20 */
 	{ 0x66a7, &vega20_device_info },	/* Vega20 */
 	{ 0x66af, &vega20_device_info }		/* Vega20 */
 };
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index 3367dd30cdd0..3b7fce5d7258 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -130,7 +130,7 @@ static void vega20_set_default_registry_data(struct pp_hwmgr *hwmgr)
 	data->registry_data.disable_auto_wattman = 1;
 	data->registry_data.auto_wattman_debug = 0;
 	data->registry_data.auto_wattman_sample_period = 100;
-	data->registry_data.fclk_gfxclk_ratio = 0x3F6CCCCD;
+	data->registry_data.fclk_gfxclk_ratio = 0;
 	data->registry_data.auto_wattman_threshold = 50;
 	data->registry_data.gfxoff_controlled_by_driver = 1;
 	data->gfxoff_allowed = false;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu7_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/smu7_ppsmc.h
index 62f36ba2435b..c1a99dfe4913 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu7_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu7_ppsmc.h
@@ -386,6 +386,8 @@ typedef uint16_t PPSMC_Result;
 #define PPSMC_MSG_AgmResetPsm                 ((uint16_t) 0x403)
 #define PPSMC_MSG_ReadVftCell                 ((uint16_t) 0x404)
 
+#define PPSMC_MSG_ApplyAvfsCksOffVoltage      ((uint16_t) 0x415)
+
 #define PPSMC_MSG_GFX_CU_PG_ENABLE            ((uint16_t) 0x280)
 #define PPSMC_MSG_GFX_CU_PG_DISABLE           ((uint16_t) 0x281)
 #define PPSMC_MSG_GetCurrPkgPwr               ((uint16_t) 0x282)
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index 872d3824337b..a1e0ac9ae248 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -1985,6 +1985,12 @@ int polaris10_thermal_avfs_enable(struct pp_hwmgr *hwmgr)
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_EnableAvfs);
 
+	/* Apply avfs cks-off voltages to avoid the overshoot
+	 * when switching to the highest sclk frequency
+	 */
+	if (data->apply_avfs_cks_off_voltage)
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_ApplyAvfsCksOffVoltage);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
index 99d5e4f98f49..a6edd5df33b0 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
@@ -37,10 +37,13 @@ MODULE_FIRMWARE("amdgpu/fiji_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_smc_sk.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/polaris10_k2_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_smc_sk.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_k2_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_smc.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_k_smc.bin");
 MODULE_FIRMWARE("amdgpu/vegam_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_acg_smc.bin");
diff --git a/drivers/gpu/drm/i915/gvt/fb_decoder.c b/drivers/gpu/drm/i915/gvt/fb_decoder.c
index 481896fb712a..85e6736f0a32 100644
--- a/drivers/gpu/drm/i915/gvt/fb_decoder.c
+++ b/drivers/gpu/drm/i915/gvt/fb_decoder.c
@@ -235,7 +235,7 @@ int intel_vgpu_decode_primary_plane(struct intel_vgpu *vgpu,
 		plane->bpp = skl_pixel_formats[fmt].bpp;
 		plane->drm_format = skl_pixel_formats[fmt].drm_format;
 	} else {
-		plane->tiled = !!(val & DISPPLANE_TILED);
+		plane->tiled = val & DISPPLANE_TILED;
 		fmt = bdw_format_to_drm(val & DISPPLANE_PIXFORMAT_MASK);
 		plane->bpp = bdw_pixel_formats[fmt].bpp;
 		plane->drm_format = bdw_pixel_formats[fmt].drm_format;
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index ffdbbac4400e..47062ee979cf 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -1444,6 +1444,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv)
 
 	intel_uncore_sanitize(dev_priv);
 
+	intel_gt_init_workarounds(dev_priv);
 	i915_gem_load_init_fences(dev_priv);
 
 	/* On the 945G/GM, the chipset reports the MSI capability on the
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 9102571e9692..872a2e159a5f 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -67,6 +67,7 @@
 #include "intel_ringbuffer.h"
 #include "intel_uncore.h"
 #include "intel_wopcm.h"
+#include "intel_workarounds.h"
 #include "intel_uc.h"
 
 #include "i915_gem.h"
@@ -1805,6 +1806,7 @@ struct drm_i915_private {
 	int dpio_phy_iosf_port[I915_NUM_PHYS_VLV];
 
 	struct i915_workarounds workarounds;
+	struct i915_wa_list gt_wa_list;
 
 	struct i915_frontbuffer_tracking fb_tracking;
 
@@ -2148,6 +2150,8 @@ struct drm_i915_private {
 		struct delayed_work idle_work;
 
 		ktime_t last_init_time;
+
+		struct i915_vma *scratch;
 	} gt;
 
 	/* perform PHY state sanity checks? */
@@ -3870,4 +3874,9 @@ static inline int intel_hws_csb_write_index(struct drm_i915_private *i915)
 		return I915_HWS_CSB_WRITE_INDEX;
 }
 
+static inline u32 i915_scratch_offset(const struct drm_i915_private *i915)
+{
+	return i915_ggtt_offset(i915->gt.scratch);
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0c8aa57ce83b..6ae9a6080cc8 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -5305,7 +5305,7 @@ int i915_gem_init_hw(struct drm_i915_private *dev_priv)
 		}
 	}
 
-	intel_gt_workarounds_apply(dev_priv);
+	intel_gt_apply_workarounds(dev_priv);
 
 	i915_gem_init_swizzling(dev_priv);
 
@@ -5500,6 +5500,44 @@ err_active:
 	goto out_ctx;
 }
 
+static int
+i915_gem_init_scratch(struct drm_i915_private *i915, unsigned int size)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int ret;
+
+	obj = i915_gem_object_create_stolen(i915, size);
+	if (!obj)
+		obj = i915_gem_object_create_internal(i915, size);
+	if (IS_ERR(obj)) {
+		DRM_ERROR("Failed to allocate scratch page\n");
+		return PTR_ERR(obj);
+	}
+
+	vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL);
+	if (IS_ERR(vma)) {
+		ret = PTR_ERR(vma);
+		goto err_unref;
+	}
+
+	ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
+	if (ret)
+		goto err_unref;
+
+	i915->gt.scratch = vma;
+	return 0;
+
+err_unref:
+	i915_gem_object_put(obj);
+	return ret;
+}
+
+static void i915_gem_fini_scratch(struct drm_i915_private *i915)
+{
+	i915_vma_unpin_and_release(&i915->gt.scratch, 0);
+}
+
 int i915_gem_init(struct drm_i915_private *dev_priv)
 {
 	int ret;
@@ -5546,12 +5584,19 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 		goto err_unlock;
 	}
 
-	ret = i915_gem_contexts_init(dev_priv);
+	ret = i915_gem_init_scratch(dev_priv,
+				    IS_GEN2(dev_priv) ? SZ_256K : PAGE_SIZE);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
 		goto err_ggtt;
 	}
 
+	ret = i915_gem_contexts_init(dev_priv);
+	if (ret) {
+		GEM_BUG_ON(ret == -EIO);
+		goto err_scratch;
+	}
+
 	ret = intel_engines_init(dev_priv);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
@@ -5624,6 +5669,8 @@ err_pm:
 err_context:
 	if (ret != -EIO)
 		i915_gem_contexts_fini(dev_priv);
+err_scratch:
+	i915_gem_fini_scratch(dev_priv);
 err_ggtt:
 err_unlock:
 	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
@@ -5675,8 +5722,11 @@ void i915_gem_fini(struct drm_i915_private *dev_priv)
 	intel_uc_fini(dev_priv);
 	i915_gem_cleanup_engines(dev_priv);
 	i915_gem_contexts_fini(dev_priv);
+	i915_gem_fini_scratch(dev_priv);
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
+	intel_wa_list_free(&dev_priv->gt_wa_list);
+
 	intel_cleanup_gt_powersave(dev_priv);
 
 	intel_uc_fini_misc(dev_priv);
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index d4fac09095f8..1aaccbe7e1de 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1268,7 +1268,7 @@ relocate_entry(struct i915_vma *vma,
 		else if (gen >= 4)
 			len = 4;
 		else
-			len = 6;
+			len = 3;
 
 		batch = reloc_gpu(eb, vma, len);
 		if (IS_ERR(batch))
@@ -1309,11 +1309,6 @@ relocate_entry(struct i915_vma *vma,
 			*batch++ = MI_STORE_DWORD_IMM | MI_MEM_VIRTUAL;
 			*batch++ = addr;
 			*batch++ = target_offset;
-
-			/* And again for good measure (blb/pnv) */
-			*batch++ = MI_STORE_DWORD_IMM | MI_MEM_VIRTUAL;
-			*batch++ = addr;
-			*batch++ = target_offset;
 		}
 
 		goto out;
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 3eb33e000d6f..db4128d6c09b 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -1495,7 +1495,7 @@ static void gem_record_rings(struct i915_gpu_state *error)
 			if (HAS_BROKEN_CS_TLB(i915))
 				ee->wa_batchbuffer =
 					i915_error_object_create(i915,
-								 engine->scratch);
+								 i915->gt.scratch);
 			request_record_user_bo(request, ee);
 
 			ee->ctx =
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 217ed3ee1cab..76b5f94ea6cb 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -490,46 +490,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine)
 	intel_engine_init_cmd_parser(engine);
 }
 
-int intel_engine_create_scratch(struct intel_engine_cs *engine,
-				unsigned int size)
-{
-	struct drm_i915_gem_object *obj;
-	struct i915_vma *vma;
-	int ret;
-
-	WARN_ON(engine->scratch);
-
-	obj = i915_gem_object_create_stolen(engine->i915, size);
-	if (!obj)
-		obj = i915_gem_object_create_internal(engine->i915, size);
-	if (IS_ERR(obj)) {
-		DRM_ERROR("Failed to allocate scratch page\n");
-		return PTR_ERR(obj);
-	}
-
-	vma = i915_vma_instance(obj, &engine->i915->ggtt.vm, NULL);
-	if (IS_ERR(vma)) {
-		ret = PTR_ERR(vma);
-		goto err_unref;
-	}
-
-	ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
-	if (ret)
-		goto err_unref;
-
-	engine->scratch = vma;
-	return 0;
-
-err_unref:
-	i915_gem_object_put(obj);
-	return ret;
-}
-
-void intel_engine_cleanup_scratch(struct intel_engine_cs *engine)
-{
-	i915_vma_unpin_and_release(&engine->scratch, 0);
-}
-
 static void cleanup_status_page(struct intel_engine_cs *engine)
 {
 	if (HWS_NEEDS_PHYSICAL(engine->i915)) {
@@ -704,8 +664,6 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *i915 = engine->i915;
 
-	intel_engine_cleanup_scratch(engine);
-
 	cleanup_status_page(engine);
 
 	intel_engine_fini_breadcrumbs(engine);
@@ -720,6 +678,8 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 	__intel_context_unpin(i915->kernel_context, engine);
 
 	i915_timeline_fini(&engine->timeline);
+
+	intel_wa_list_free(&engine->wa_list);
 }
 
 u64 intel_engine_get_active_head(const struct intel_engine_cs *engine)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 37c94a54efcb..58d1d3d47dd3 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -442,8 +442,13 @@ static u64 execlists_update_context(struct i915_request *rq)
 	 * may not be visible to the HW prior to the completion of the UC
 	 * register write and that we may begin execution from the context
 	 * before its image is complete leading to invalid PD chasing.
+	 *
+	 * Furthermore, Braswell, at least, wants a full mb to be sure that
+	 * the writes are coherent in memory (visible to the GPU) prior to
+	 * execution, and not just visible to other CPUs (as is the result of
+	 * wmb).
 	 */
-	wmb();
+	mb();
 	return ce->lrc_desc;
 }
 
@@ -1443,9 +1448,10 @@ static int execlists_request_alloc(struct i915_request *request)
 static u32 *
 gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch)
 {
+	/* NB no one else is allowed to scribble over scratch + 256! */
 	*batch++ = MI_STORE_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT;
 	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
-	*batch++ = i915_ggtt_offset(engine->scratch) + 256;
+	*batch++ = i915_scratch_offset(engine->i915) + 256;
 	*batch++ = 0;
 
 	*batch++ = MI_LOAD_REGISTER_IMM(1);
@@ -1459,7 +1465,7 @@ gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch)
 
 	*batch++ = MI_LOAD_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT;
 	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
-	*batch++ = i915_ggtt_offset(engine->scratch) + 256;
+	*batch++ = i915_scratch_offset(engine->i915) + 256;
 	*batch++ = 0;
 
 	return batch;
@@ -1496,7 +1502,7 @@ static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
 				       PIPE_CONTROL_GLOBAL_GTT_IVB |
 				       PIPE_CONTROL_CS_STALL |
 				       PIPE_CONTROL_QW_WRITE,
-				       i915_ggtt_offset(engine->scratch) +
+				       i915_scratch_offset(engine->i915) +
 				       2 * CACHELINE_BYTES);
 
 	*batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
@@ -1573,7 +1579,7 @@ static u32 *gen9_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
 					       PIPE_CONTROL_GLOBAL_GTT_IVB |
 					       PIPE_CONTROL_CS_STALL |
 					       PIPE_CONTROL_QW_WRITE,
-					       i915_ggtt_offset(engine->scratch)
+					       i915_scratch_offset(engine->i915)
 					       + 2 * CACHELINE_BYTES);
 	}
 
@@ -1793,6 +1799,8 @@ static bool unexpected_starting_state(struct intel_engine_cs *engine)
 
 static int gen8_init_common_ring(struct intel_engine_cs *engine)
 {
+	intel_engine_apply_workarounds(engine);
+
 	intel_mocs_init_engine(engine);
 
 	intel_engine_reset_breadcrumbs(engine);
@@ -2139,7 +2147,7 @@ static int gen8_emit_flush_render(struct i915_request *request,
 {
 	struct intel_engine_cs *engine = request->engine;
 	u32 scratch_addr =
-		i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES;
+		i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES;
 	bool vf_flush_wa = false, dc_flush_wa = false;
 	u32 *cs, flags = 0;
 	int len;
@@ -2476,10 +2484,6 @@ int logical_render_ring_init(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
-	ret = intel_engine_create_scratch(engine, PAGE_SIZE);
-	if (ret)
-		goto err_cleanup_common;
-
 	ret = intel_init_workaround_bb(engine);
 	if (ret) {
 		/*
@@ -2491,11 +2495,9 @@ int logical_render_ring_init(struct intel_engine_cs *engine)
 			  ret);
 	}
 
-	return 0;
+	intel_engine_init_workarounds(engine);
 
-err_cleanup_common:
-	intel_engine_cleanup_common(engine);
-	return ret;
+	return 0;
 }
 
 int logical_xcs_ring_init(struct intel_engine_cs *engine)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 187bb0ceb4ac..1f8d2a66c791 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -69,19 +69,28 @@ unsigned int intel_ring_update_space(struct intel_ring *ring)
 static int
 gen2_render_ring_flush(struct i915_request *rq, u32 mode)
 {
+	unsigned int num_store_dw;
 	u32 cmd, *cs;
 
 	cmd = MI_FLUSH;
-
+	num_store_dw = 0;
 	if (mode & EMIT_INVALIDATE)
 		cmd |= MI_READ_FLUSH;
+	if (mode & EMIT_FLUSH)
+		num_store_dw = 4;
 
-	cs = intel_ring_begin(rq, 2);
+	cs = intel_ring_begin(rq, 2 + 3 * num_store_dw);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
 	*cs++ = cmd;
-	*cs++ = MI_NOOP;
+	while (num_store_dw--) {
+		*cs++ = MI_STORE_DWORD_IMM | MI_MEM_VIRTUAL;
+		*cs++ = i915_scratch_offset(rq->i915);
+		*cs++ = 0;
+	}
+	*cs++ = MI_FLUSH | MI_NO_WRITE_FLUSH;
+
 	intel_ring_advance(rq, cs);
 
 	return 0;
@@ -150,8 +159,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode)
 	 */
 	if (mode & EMIT_INVALIDATE) {
 		*cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE;
-		*cs++ = i915_ggtt_offset(rq->engine->scratch) |
-			PIPE_CONTROL_GLOBAL_GTT;
+		*cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT;
 		*cs++ = 0;
 		*cs++ = 0;
 
@@ -159,8 +167,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode)
 			*cs++ = MI_FLUSH;
 
 		*cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE;
-		*cs++ = i915_ggtt_offset(rq->engine->scratch) |
-			PIPE_CONTROL_GLOBAL_GTT;
+		*cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT;
 		*cs++ = 0;
 		*cs++ = 0;
 	}
@@ -212,8 +219,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode)
 static int
 intel_emit_post_sync_nonzero_flush(struct i915_request *rq)
 {
-	u32 scratch_addr =
-		i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES;
+	u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES;
 	u32 *cs;
 
 	cs = intel_ring_begin(rq, 6);
@@ -246,8 +252,7 @@ intel_emit_post_sync_nonzero_flush(struct i915_request *rq)
 static int
 gen6_render_ring_flush(struct i915_request *rq, u32 mode)
 {
-	u32 scratch_addr =
-		i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES;
+	u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES;
 	u32 *cs, flags = 0;
 	int ret;
 
@@ -316,8 +321,7 @@ gen7_render_ring_cs_stall_wa(struct i915_request *rq)
 static int
 gen7_render_ring_flush(struct i915_request *rq, u32 mode)
 {
-	u32 scratch_addr =
-		i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES;
+	u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES;
 	u32 *cs, flags = 0;
 
 	/*
@@ -971,7 +975,7 @@ i965_emit_bb_start(struct i915_request *rq,
 }
 
 /* Just userspace ABI convention to limit the wa batch bo to a resonable size */
-#define I830_BATCH_LIMIT (256*1024)
+#define I830_BATCH_LIMIT SZ_256K
 #define I830_TLB_ENTRIES (2)
 #define I830_WA_SIZE max(I830_TLB_ENTRIES*4096, I830_BATCH_LIMIT)
 static int
@@ -979,7 +983,9 @@ i830_emit_bb_start(struct i915_request *rq,
 		   u64 offset, u32 len,
 		   unsigned int dispatch_flags)
 {
-	u32 *cs, cs_offset = i915_ggtt_offset(rq->engine->scratch);
+	u32 *cs, cs_offset = i915_scratch_offset(rq->i915);
+
+	GEM_BUG_ON(rq->i915->gt.scratch->size < I830_WA_SIZE);
 
 	cs = intel_ring_begin(rq, 6);
 	if (IS_ERR(cs))
@@ -1437,7 +1443,6 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 {
 	struct i915_timeline *timeline;
 	struct intel_ring *ring;
-	unsigned int size;
 	int err;
 
 	intel_engine_setup_common(engine);
@@ -1462,21 +1467,12 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 	GEM_BUG_ON(engine->buffer);
 	engine->buffer = ring;
 
-	size = PAGE_SIZE;
-	if (HAS_BROKEN_CS_TLB(engine->i915))
-		size = I830_WA_SIZE;
-	err = intel_engine_create_scratch(engine, size);
-	if (err)
-		goto err_unpin;
-
 	err = intel_engine_init_common(engine);
 	if (err)
-		goto err_scratch;
+		goto err_unpin;
 
 	return 0;
 
-err_scratch:
-	intel_engine_cleanup_scratch(engine);
 err_unpin:
 	intel_ring_unpin(ring);
 err_ring:
@@ -1550,7 +1546,7 @@ static int flush_pd_dir(struct i915_request *rq)
 	/* Stall until the page table load is complete */
 	*cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
 	*cs++ = i915_mmio_reg_offset(RING_PP_DIR_BASE(engine));
-	*cs++ = i915_ggtt_offset(engine->scratch);
+	*cs++ = i915_scratch_offset(rq->i915);
 	*cs++ = MI_NOOP;
 
 	intel_ring_advance(rq, cs);
@@ -1659,7 +1655,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 			/* Insert a delay before the next switch! */
 			*cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
 			*cs++ = i915_mmio_reg_offset(last_reg);
-			*cs++ = i915_ggtt_offset(engine->scratch);
+			*cs++ = i915_scratch_offset(rq->i915);
 			*cs++ = MI_NOOP;
 		}
 		*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 2dfa585712c2..767a7192c969 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -15,6 +15,7 @@
 #include "i915_selftest.h"
 #include "i915_timeline.h"
 #include "intel_gpu_commands.h"
+#include "intel_workarounds.h"
 
 struct drm_printer;
 struct i915_sched_attr;
@@ -440,7 +441,7 @@ struct intel_engine_cs {
 
 	struct intel_hw_status_page status_page;
 	struct i915_ctx_workarounds wa_ctx;
-	struct i915_vma *scratch;
+	struct i915_wa_list wa_list;
 
 	u32             irq_keep_mask; /* always keep these interrupts */
 	u32		irq_enable_mask; /* bitmask to enable ring interrupt */
@@ -898,10 +899,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine);
 int intel_engine_init_common(struct intel_engine_cs *engine);
 void intel_engine_cleanup_common(struct intel_engine_cs *engine);
 
-int intel_engine_create_scratch(struct intel_engine_cs *engine,
-				unsigned int size);
-void intel_engine_cleanup_scratch(struct intel_engine_cs *engine);
-
 int intel_init_render_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine);
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 4bcdeaf8d98f..6e580891db96 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -48,6 +48,20 @@
  * - Public functions to init or apply the given workaround type.
  */
 
+static void wa_init_start(struct i915_wa_list *wal, const char *name)
+{
+	wal->name = name;
+}
+
+static void wa_init_finish(struct i915_wa_list *wal)
+{
+	if (!wal->count)
+		return;
+
+	DRM_DEBUG_DRIVER("Initialized %u %s workarounds\n",
+			 wal->count, wal->name);
+}
+
 static void wa_add(struct drm_i915_private *i915,
 		   i915_reg_t reg, const u32 mask, const u32 val)
 {
@@ -580,160 +594,175 @@ int intel_ctx_workarounds_emit(struct i915_request *rq)
 	return 0;
 }
 
-static void bdw_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void
+wal_add(struct i915_wa_list *wal, const struct i915_wa *wa)
+{
+	const unsigned int grow = 1 << 4;
+
+	GEM_BUG_ON(!is_power_of_2(grow));
+
+	if (IS_ALIGNED(wal->count, grow)) { /* Either uninitialized or full. */
+		struct i915_wa *list;
+
+		list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*wa),
+				     GFP_KERNEL);
+		if (!list) {
+			DRM_ERROR("No space for workaround init!\n");
+			return;
+		}
+
+		if (wal->list)
+			memcpy(list, wal->list, sizeof(*wa) * wal->count);
+
+		wal->list = list;
+	}
+
+	wal->list[wal->count++] = *wa;
+}
+
+static void
+wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val)
+{
+	struct i915_wa wa = {
+		.reg = reg,
+		.mask = val,
+		.val = _MASKED_BIT_ENABLE(val)
+	};
+
+	wal_add(wal, &wa);
+}
+
+static void
+wa_write_masked_or(struct i915_wa_list *wal, i915_reg_t reg, u32 mask,
+		   u32 val)
 {
+	struct i915_wa wa = {
+		.reg = reg,
+		.mask = mask,
+		.val = val
+	};
+
+	wal_add(wal, &wa);
 }
 
-static void chv_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void
+wa_write(struct i915_wa_list *wal, i915_reg_t reg, u32 val)
 {
+	wa_write_masked_or(wal, reg, ~0, val);
 }
 
-static void gen9_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void
+wa_write_or(struct i915_wa_list *wal, i915_reg_t reg, u32 val)
 {
-	/* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */
-	I915_WRITE(GEN9_CSFE_CHICKEN1_RCS,
-		   _MASKED_BIT_ENABLE(GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE));
+	wa_write_masked_or(wal, reg, val, val);
+}
 
-	/* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */
-	I915_WRITE(BDW_SCRATCH1, I915_READ(BDW_SCRATCH1) |
-		   GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE);
+static void gen9_gt_workarounds_init(struct drm_i915_private *i915)
+{
+	struct i915_wa_list *wal = &i915->gt_wa_list;
 
 	/* WaDisableKillLogic:bxt,skl,kbl */
-	if (!IS_COFFEELAKE(dev_priv))
-		I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) |
-			   ECOCHK_DIS_TLB);
+	if (!IS_COFFEELAKE(i915))
+		wa_write_or(wal,
+			    GAM_ECOCHK,
+			    ECOCHK_DIS_TLB);
 
-	if (HAS_LLC(dev_priv)) {
+	if (HAS_LLC(i915)) {
 		/* WaCompressedResourceSamplerPbeMediaNewHashMode:skl,kbl
 		 *
 		 * Must match Display Engine. See
 		 * WaCompressedResourceDisplayNewHashMode.
 		 */
-		I915_WRITE(MMCD_MISC_CTRL,
-			   I915_READ(MMCD_MISC_CTRL) |
-			   MMCD_PCLA |
-			   MMCD_HOTSPOT_EN);
+		wa_write_or(wal,
+			    MMCD_MISC_CTRL,
+			    MMCD_PCLA | MMCD_HOTSPOT_EN);
 	}
 
 	/* WaDisableHDCInvalidation:skl,bxt,kbl,cfl */
-	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) |
-		   BDW_DISABLE_HDC_INVALIDATION);
-
-	/* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */
-	if (IS_GEN9_LP(dev_priv)) {
-		u32 val = I915_READ(GEN8_L3SQCREG1);
-
-		val &= ~L3_PRIO_CREDITS_MASK;
-		val |= L3_GENERAL_PRIO_CREDITS(62) | L3_HIGH_PRIO_CREDITS(2);
-		I915_WRITE(GEN8_L3SQCREG1, val);
-	}
-
-	/* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */
-	I915_WRITE(GEN8_L3SQCREG4,
-		   I915_READ(GEN8_L3SQCREG4) | GEN8_LQSC_FLUSH_COHERENT_LINES);
-
-	/* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,[cnl] */
-	I915_WRITE(GEN7_FF_SLICE_CS_CHICKEN1,
-		   _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL));
+	wa_write_or(wal,
+		    GAM_ECOCHK,
+		    BDW_DISABLE_HDC_INVALIDATION);
 }
 
-static void skl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void skl_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	gen9_gt_workarounds_apply(dev_priv);
+	struct i915_wa_list *wal = &i915->gt_wa_list;
 
-	/* WaEnableGapsTsvCreditFix:skl */
-	I915_WRITE(GEN8_GARBCNTL,
-		   I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE);
+	gen9_gt_workarounds_init(i915);
 
 	/* WaDisableGafsUnitClkGating:skl */
-	I915_WRITE(GEN7_UCGCTL4,
-		   I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE);
+	wa_write_or(wal,
+		    GEN7_UCGCTL4,
+		    GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE);
 
 	/* WaInPlaceDecompressionHang:skl */
-	if (IS_SKL_REVID(dev_priv, SKL_REVID_H0, REVID_FOREVER))
-		I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA,
-			   I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
-			   GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
+	if (IS_SKL_REVID(i915, SKL_REVID_H0, REVID_FOREVER))
+		wa_write_or(wal,
+			    GEN9_GAMT_ECO_REG_RW_IA,
+			    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 }
 
-static void bxt_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void bxt_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	gen9_gt_workarounds_apply(dev_priv);
+	struct i915_wa_list *wal = &i915->gt_wa_list;
 
-	/* WaDisablePooledEuLoadBalancingFix:bxt */
-	I915_WRITE(FF_SLICE_CS_CHICKEN2,
-		   _MASKED_BIT_ENABLE(GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE));
+	gen9_gt_workarounds_init(i915);
 
 	/* WaInPlaceDecompressionHang:bxt */
-	I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA,
-		   I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
-		   GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
+	wa_write_or(wal,
+		    GEN9_GAMT_ECO_REG_RW_IA,
+		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 }
 
-static void kbl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void kbl_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	gen9_gt_workarounds_apply(dev_priv);
+	struct i915_wa_list *wal = &i915->gt_wa_list;
 
-	/* WaEnableGapsTsvCreditFix:kbl */
-	I915_WRITE(GEN8_GARBCNTL,
-		   I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE);
+	gen9_gt_workarounds_init(i915);
 
 	/* WaDisableDynamicCreditSharing:kbl */
-	if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_B0))
-		I915_WRITE(GAMT_CHKN_BIT_REG,
-			   I915_READ(GAMT_CHKN_BIT_REG) |
-			   GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING);
+	if (IS_KBL_REVID(i915, 0, KBL_REVID_B0))
+		wa_write_or(wal,
+			    GAMT_CHKN_BIT_REG,
+			    GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING);
 
 	/* WaDisableGafsUnitClkGating:kbl */
-	I915_WRITE(GEN7_UCGCTL4,
-		   I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE);
+	wa_write_or(wal,
+		    GEN7_UCGCTL4,
+		    GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE);
 
 	/* WaInPlaceDecompressionHang:kbl */
-	I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA,
-		   I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
-		   GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
-
-	/* WaKBLVECSSemaphoreWaitPoll:kbl */
-	if (IS_KBL_REVID(dev_priv, KBL_REVID_A0, KBL_REVID_E0)) {
-		struct intel_engine_cs *engine;
-		unsigned int tmp;
-
-		for_each_engine(engine, dev_priv, tmp) {
-			if (engine->id == RCS)
-				continue;
-
-			I915_WRITE(RING_SEMA_WAIT_POLL(engine->mmio_base), 1);
-		}
-	}
+	wa_write_or(wal,
+		    GEN9_GAMT_ECO_REG_RW_IA,
+		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 }
 
-static void glk_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void glk_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	gen9_gt_workarounds_apply(dev_priv);
+	gen9_gt_workarounds_init(i915);
 }
 
-static void cfl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void cfl_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	gen9_gt_workarounds_apply(dev_priv);
+	struct i915_wa_list *wal = &i915->gt_wa_list;
 
-	/* WaEnableGapsTsvCreditFix:cfl */
-	I915_WRITE(GEN8_GARBCNTL,
-		   I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE);
+	gen9_gt_workarounds_init(i915);
 
 	/* WaDisableGafsUnitClkGating:cfl */
-	I915_WRITE(GEN7_UCGCTL4,
-		   I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE);
+	wa_write_or(wal,
+		    GEN7_UCGCTL4,
+		    GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE);
 
 	/* WaInPlaceDecompressionHang:cfl */
-	I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA,
-		   I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
-		   GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
+	wa_write_or(wal,
+		    GEN9_GAMT_ECO_REG_RW_IA,
+		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 }
 
 static void wa_init_mcr(struct drm_i915_private *dev_priv)
 {
 	const struct sseu_dev_info *sseu = &(INTEL_INFO(dev_priv)->sseu);
-	u32 mcr;
+	struct i915_wa_list *wal = &dev_priv->gt_wa_list;
 	u32 mcr_slice_subslice_mask;
 
 	/*
@@ -770,8 +799,6 @@ static void wa_init_mcr(struct drm_i915_private *dev_priv)
 		WARN_ON((enabled_mask & disabled_mask) != enabled_mask);
 	}
 
-	mcr = I915_READ(GEN8_MCR_SELECTOR);
-
 	if (INTEL_GEN(dev_priv) >= 11)
 		mcr_slice_subslice_mask = GEN11_MCR_SLICE_MASK |
 					  GEN11_MCR_SUBSLICE_MASK;
@@ -789,148 +816,170 @@ static void wa_init_mcr(struct drm_i915_private *dev_priv)
 	 * occasions, such as INSTDONE, where this value is dependent
 	 * on s/ss combo, the read should be done with read_subslice_reg.
 	 */
-	mcr &= ~mcr_slice_subslice_mask;
-	mcr |= intel_calculate_mcr_s_ss_select(dev_priv);
-	I915_WRITE(GEN8_MCR_SELECTOR, mcr);
+	wa_write_masked_or(wal,
+			   GEN8_MCR_SELECTOR,
+			   mcr_slice_subslice_mask,
+			   intel_calculate_mcr_s_ss_select(dev_priv));
 }
 
-static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void cnl_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	wa_init_mcr(dev_priv);
+	struct i915_wa_list *wal = &i915->gt_wa_list;
+
+	wa_init_mcr(i915);
 
 	/* WaDisableI2mCycleOnWRPort:cnl (pre-prod) */
-	if (IS_CNL_REVID(dev_priv, CNL_REVID_B0, CNL_REVID_B0))
-		I915_WRITE(GAMT_CHKN_BIT_REG,
-			   I915_READ(GAMT_CHKN_BIT_REG) |
-			   GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT);
+	if (IS_CNL_REVID(i915, CNL_REVID_B0, CNL_REVID_B0))
+		wa_write_or(wal,
+			    GAMT_CHKN_BIT_REG,
+			    GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT);
 
 	/* WaInPlaceDecompressionHang:cnl */
-	I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA,
-		   I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
-		   GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
-
-	/* WaEnablePreemptionGranularityControlByUMD:cnl */
-	I915_WRITE(GEN7_FF_SLICE_CS_CHICKEN1,
-		   _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL));
+	wa_write_or(wal,
+		    GEN9_GAMT_ECO_REG_RW_IA,
+		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 }
 
-static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+static void icl_gt_workarounds_init(struct drm_i915_private *i915)
 {
-	wa_init_mcr(dev_priv);
+	struct i915_wa_list *wal = &i915->gt_wa_list;
 
-	/* This is not an Wa. Enable for better image quality */
-	I915_WRITE(_3D_CHICKEN3,
-		   _MASKED_BIT_ENABLE(_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE));
+	wa_init_mcr(i915);
 
 	/* WaInPlaceDecompressionHang:icl */
-	I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
-					    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
-
-	/* WaPipelineFlushCoherentLines:icl */
-	I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
-				   GEN8_LQSC_FLUSH_COHERENT_LINES);
-
-	/* Wa_1405543622:icl
-	 * Formerly known as WaGAPZPriorityScheme
-	 */
-	I915_WRITE(GEN8_GARBCNTL, I915_READ(GEN8_GARBCNTL) |
-				  GEN11_ARBITRATION_PRIO_ORDER_MASK);
-
-	/* Wa_1604223664:icl
-	 * Formerly known as WaL3BankAddressHashing
-	 */
-	I915_WRITE(GEN8_GARBCNTL,
-		   (I915_READ(GEN8_GARBCNTL) & ~GEN11_HASH_CTRL_EXCL_MASK) |
-		   GEN11_HASH_CTRL_EXCL_BIT0);
-	I915_WRITE(GEN11_GLBLINVL,
-		   (I915_READ(GEN11_GLBLINVL) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK) |
-		   GEN11_BANK_HASH_ADDR_EXCL_BIT0);
+	wa_write_or(wal,
+		    GEN9_GAMT_ECO_REG_RW_IA,
+		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 
 	/* WaModifyGamTlbPartitioning:icl */
-	I915_WRITE(GEN11_GACB_PERF_CTRL,
-		   (I915_READ(GEN11_GACB_PERF_CTRL) & ~GEN11_HASH_CTRL_MASK) |
-		   GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4);
-
-	/* Wa_1405733216:icl
-	 * Formerly known as WaDisableCleanEvicts
-	 */
-	I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
-				   GEN11_LQSC_CLEAN_EVICT_DISABLE);
+	wa_write_masked_or(wal,
+			   GEN11_GACB_PERF_CTRL,
+			   GEN11_HASH_CTRL_MASK,
+			   GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4);
 
 	/* Wa_1405766107:icl
 	 * Formerly known as WaCL2SFHalfMaxAlloc
 	 */
-	I915_WRITE(GEN11_LSN_UNSLCVC, I915_READ(GEN11_LSN_UNSLCVC) |
-				      GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC |
-				      GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC);
+	wa_write_or(wal,
+		    GEN11_LSN_UNSLCVC,
+		    GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC |
+		    GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC);
 
 	/* Wa_220166154:icl
 	 * Formerly known as WaDisCtxReload
 	 */
-	I915_WRITE(GAMW_ECO_DEV_RW_IA_REG, I915_READ(GAMW_ECO_DEV_RW_IA_REG) |
-					   GAMW_ECO_DEV_CTX_RELOAD_DISABLE);
+	wa_write_or(wal,
+		    GEN8_GAMW_ECO_DEV_RW_IA,
+		    GAMW_ECO_DEV_CTX_RELOAD_DISABLE);
 
 	/* Wa_1405779004:icl (pre-prod) */
-	if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_A0))
-		I915_WRITE(SLICE_UNIT_LEVEL_CLKGATE,
-			   I915_READ(SLICE_UNIT_LEVEL_CLKGATE) |
-			   MSCUNIT_CLKGATE_DIS);
+	if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_A0))
+		wa_write_or(wal,
+			    SLICE_UNIT_LEVEL_CLKGATE,
+			    MSCUNIT_CLKGATE_DIS);
 
 	/* Wa_1406680159:icl */
-	I915_WRITE(SUBSLICE_UNIT_LEVEL_CLKGATE,
-		   I915_READ(SUBSLICE_UNIT_LEVEL_CLKGATE) |
-		   GWUNIT_CLKGATE_DIS);
-
-	/* Wa_1604302699:icl */
-	I915_WRITE(GEN10_L3_CHICKEN_MODE_REGISTER,
-		   I915_READ(GEN10_L3_CHICKEN_MODE_REGISTER) |
-		   GEN11_I2M_WRITE_DISABLE);
+	wa_write_or(wal,
+		    SUBSLICE_UNIT_LEVEL_CLKGATE,
+		    GWUNIT_CLKGATE_DIS);
 
 	/* Wa_1406838659:icl (pre-prod) */
-	if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0))
-		I915_WRITE(INF_UNIT_LEVEL_CLKGATE,
-			   I915_READ(INF_UNIT_LEVEL_CLKGATE) |
-			   CGPSF_CLKGATE_DIS);
-
-	/* WaForwardProgressSoftReset:icl */
-	I915_WRITE(GEN10_SCRATCH_LNCF2,
-		   I915_READ(GEN10_SCRATCH_LNCF2) |
-		   PMFLUSHDONE_LNICRSDROP |
-		   PMFLUSH_GAPL3UNBLOCK |
-		   PMFLUSHDONE_LNEBLK);
+	if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0))
+		wa_write_or(wal,
+			    INF_UNIT_LEVEL_CLKGATE,
+			    CGPSF_CLKGATE_DIS);
 
 	/* Wa_1406463099:icl
 	 * Formerly known as WaGamTlbPendError
 	 */
-	I915_WRITE(GAMT_CHKN_BIT_REG,
-		   I915_READ(GAMT_CHKN_BIT_REG) |
-		   GAMT_CHKN_DISABLE_L3_COH_PIPE);
+	wa_write_or(wal,
+		    GAMT_CHKN_BIT_REG,
+		    GAMT_CHKN_DISABLE_L3_COH_PIPE);
 }
 
-void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+void intel_gt_init_workarounds(struct drm_i915_private *i915)
 {
-	if (INTEL_GEN(dev_priv) < 8)
+	struct i915_wa_list *wal = &i915->gt_wa_list;
+
+	wa_init_start(wal, "GT");
+
+	if (INTEL_GEN(i915) < 8)
 		return;
-	else if (IS_BROADWELL(dev_priv))
-		bdw_gt_workarounds_apply(dev_priv);
-	else if (IS_CHERRYVIEW(dev_priv))
-		chv_gt_workarounds_apply(dev_priv);
-	else if (IS_SKYLAKE(dev_priv))
-		skl_gt_workarounds_apply(dev_priv);
-	else if (IS_BROXTON(dev_priv))
-		bxt_gt_workarounds_apply(dev_priv);
-	else if (IS_KABYLAKE(dev_priv))
-		kbl_gt_workarounds_apply(dev_priv);
-	else if (IS_GEMINILAKE(dev_priv))
-		glk_gt_workarounds_apply(dev_priv);
-	else if (IS_COFFEELAKE(dev_priv))
-		cfl_gt_workarounds_apply(dev_priv);
-	else if (IS_CANNONLAKE(dev_priv))
-		cnl_gt_workarounds_apply(dev_priv);
-	else if (IS_ICELAKE(dev_priv))
-		icl_gt_workarounds_apply(dev_priv);
+	else if (IS_BROADWELL(i915))
+		return;
+	else if (IS_CHERRYVIEW(i915))
+		return;
+	else if (IS_SKYLAKE(i915))
+		skl_gt_workarounds_init(i915);
+	else if (IS_BROXTON(i915))
+		bxt_gt_workarounds_init(i915);
+	else if (IS_KABYLAKE(i915))
+		kbl_gt_workarounds_init(i915);
+	else if (IS_GEMINILAKE(i915))
+		glk_gt_workarounds_init(i915);
+	else if (IS_COFFEELAKE(i915))
+		cfl_gt_workarounds_init(i915);
+	else if (IS_CANNONLAKE(i915))
+		cnl_gt_workarounds_init(i915);
+	else if (IS_ICELAKE(i915))
+		icl_gt_workarounds_init(i915);
 	else
-		MISSING_CASE(INTEL_GEN(dev_priv));
+		MISSING_CASE(INTEL_GEN(i915));
+
+	wa_init_finish(wal);
+}
+
+static enum forcewake_domains
+wal_get_fw_for_rmw(struct drm_i915_private *dev_priv,
+		   const struct i915_wa_list *wal)
+{
+	enum forcewake_domains fw = 0;
+	struct i915_wa *wa;
+	unsigned int i;
+
+	for (i = 0, wa = wal->list; i < wal->count; i++, wa++)
+		fw |= intel_uncore_forcewake_for_reg(dev_priv,
+						     wa->reg,
+						     FW_REG_READ |
+						     FW_REG_WRITE);
+
+	return fw;
+}
+
+static void
+wa_list_apply(struct drm_i915_private *dev_priv, const struct i915_wa_list *wal)
+{
+	enum forcewake_domains fw;
+	unsigned long flags;
+	struct i915_wa *wa;
+	unsigned int i;
+
+	if (!wal->count)
+		return;
+
+	fw = wal_get_fw_for_rmw(dev_priv, wal);
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
+	intel_uncore_forcewake_get__locked(dev_priv, fw);
+
+	for (i = 0, wa = wal->list; i < wal->count; i++, wa++) {
+		u32 val = I915_READ_FW(wa->reg);
+
+		val &= ~wa->mask;
+		val |= wa->val;
+
+		I915_WRITE_FW(wa->reg, val);
+	}
+
+	intel_uncore_forcewake_put__locked(dev_priv, fw);
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+
+	DRM_DEBUG_DRIVER("Applied %u %s workarounds\n", wal->count, wal->name);
+}
+
+void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv)
+{
+	wa_list_apply(dev_priv, &dev_priv->gt_wa_list);
 }
 
 struct whitelist {
@@ -1077,6 +1126,146 @@ void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine)
 	whitelist_apply(engine, whitelist_build(engine, &w));
 }
 
+static void rcs_engine_wa_init(struct intel_engine_cs *engine)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	struct i915_wa_list *wal = &engine->wa_list;
+
+	if (IS_ICELAKE(i915)) {
+		/* This is not an Wa. Enable for better image quality */
+		wa_masked_en(wal,
+			     _3D_CHICKEN3,
+			     _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE);
+
+		/* WaPipelineFlushCoherentLines:icl */
+		wa_write_or(wal,
+			    GEN8_L3SQCREG4,
+			    GEN8_LQSC_FLUSH_COHERENT_LINES);
+
+		/*
+		 * Wa_1405543622:icl
+		 * Formerly known as WaGAPZPriorityScheme
+		 */
+		wa_write_or(wal,
+			    GEN8_GARBCNTL,
+			    GEN11_ARBITRATION_PRIO_ORDER_MASK);
+
+		/*
+		 * Wa_1604223664:icl
+		 * Formerly known as WaL3BankAddressHashing
+		 */
+		wa_write_masked_or(wal,
+				   GEN8_GARBCNTL,
+				   GEN11_HASH_CTRL_EXCL_MASK,
+				   GEN11_HASH_CTRL_EXCL_BIT0);
+		wa_write_masked_or(wal,
+				   GEN11_GLBLINVL,
+				   GEN11_BANK_HASH_ADDR_EXCL_MASK,
+				   GEN11_BANK_HASH_ADDR_EXCL_BIT0);
+
+		/*
+		 * Wa_1405733216:icl
+		 * Formerly known as WaDisableCleanEvicts
+		 */
+		wa_write_or(wal,
+			    GEN8_L3SQCREG4,
+			    GEN11_LQSC_CLEAN_EVICT_DISABLE);
+
+		/* Wa_1604302699:icl */
+		wa_write_or(wal,
+			    GEN10_L3_CHICKEN_MODE_REGISTER,
+			    GEN11_I2M_WRITE_DISABLE);
+
+		/* WaForwardProgressSoftReset:icl */
+		wa_write_or(wal,
+			    GEN10_SCRATCH_LNCF2,
+			    PMFLUSHDONE_LNICRSDROP |
+			    PMFLUSH_GAPL3UNBLOCK |
+			    PMFLUSHDONE_LNEBLK);
+	}
+
+	if (IS_GEN9(i915) || IS_CANNONLAKE(i915)) {
+		/* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,cnl */
+		wa_masked_en(wal,
+			     GEN7_FF_SLICE_CS_CHICKEN1,
+			     GEN9_FFSC_PERCTX_PREEMPT_CTRL);
+	}
+
+	if (IS_SKYLAKE(i915) || IS_KABYLAKE(i915) || IS_COFFEELAKE(i915)) {
+		/* WaEnableGapsTsvCreditFix:skl,kbl,cfl */
+		wa_write_or(wal,
+			    GEN8_GARBCNTL,
+			    GEN9_GAPS_TSV_CREDIT_DISABLE);
+	}
+
+	if (IS_BROXTON(i915)) {
+		/* WaDisablePooledEuLoadBalancingFix:bxt */
+		wa_masked_en(wal,
+			     FF_SLICE_CS_CHICKEN2,
+			     GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE);
+	}
+
+	if (IS_GEN9(i915)) {
+		/* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */
+		wa_masked_en(wal,
+			     GEN9_CSFE_CHICKEN1_RCS,
+			     GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE);
+
+		/* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */
+		wa_write_or(wal,
+			    BDW_SCRATCH1,
+			    GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE);
+
+		/* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */
+		if (IS_GEN9_LP(i915))
+			wa_write_masked_or(wal,
+					   GEN8_L3SQCREG1,
+					   L3_PRIO_CREDITS_MASK,
+					   L3_GENERAL_PRIO_CREDITS(62) |
+					   L3_HIGH_PRIO_CREDITS(2));
+
+		/* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */
+		wa_write_or(wal,
+			    GEN8_L3SQCREG4,
+			    GEN8_LQSC_FLUSH_COHERENT_LINES);
+	}
+}
+
+static void xcs_engine_wa_init(struct intel_engine_cs *engine)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	struct i915_wa_list *wal = &engine->wa_list;
+
+	/* WaKBLVECSSemaphoreWaitPoll:kbl */
+	if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) {
+		wa_write(wal,
+			 RING_SEMA_WAIT_POLL(engine->mmio_base),
+			 1);
+	}
+}
+
+void intel_engine_init_workarounds(struct intel_engine_cs *engine)
+{
+	struct i915_wa_list *wal = &engine->wa_list;
+
+	if (GEM_WARN_ON(INTEL_GEN(engine->i915) < 8))
+		return;
+
+	wa_init_start(wal, engine->name);
+
+	if (engine->id == RCS)
+		rcs_engine_wa_init(engine);
+	else
+		xcs_engine_wa_init(engine);
+
+	wa_init_finish(wal);
+}
+
+void intel_engine_apply_workarounds(struct intel_engine_cs *engine)
+{
+	wa_list_apply(engine->i915, &engine->wa_list);
+}
+
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftests/intel_workarounds.c"
 #endif
diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h
index b11d0623e626..979695a53964 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.h
+++ b/drivers/gpu/drm/i915/intel_workarounds.h
@@ -7,11 +7,35 @@
 #ifndef _I915_WORKAROUNDS_H_
 #define _I915_WORKAROUNDS_H_
 
+#include <linux/slab.h>
+
+struct i915_wa {
+	i915_reg_t	  reg;
+	u32		  mask;
+	u32		  val;
+};
+
+struct i915_wa_list {
+	const char	*name;
+	struct i915_wa	*list;
+	unsigned int	count;
+};
+
+static inline void intel_wa_list_free(struct i915_wa_list *wal)
+{
+	kfree(wal->list);
+	memset(wal, 0, sizeof(*wal));
+}
+
 int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv);
 int intel_ctx_workarounds_emit(struct i915_request *rq);
 
-void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv);
+void intel_gt_init_workarounds(struct drm_i915_private *dev_priv);
+void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv);
 
 void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine);
 
+void intel_engine_init_workarounds(struct intel_engine_cs *engine);
+void intel_engine_apply_workarounds(struct intel_engine_cs *engine);
+
 #endif
diff --git a/drivers/gpu/drm/mediatek/mtk_dsi.c b/drivers/gpu/drm/mediatek/mtk_dsi.c
index 66df1b177959..27b507eb4a99 100644
--- a/drivers/gpu/drm/mediatek/mtk_dsi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dsi.c
@@ -818,10 +818,13 @@ static int mtk_dsi_create_conn_enc(struct drm_device *drm, struct mtk_dsi *dsi)
 	dsi->encoder.possible_crtcs = 1;
 
 	/* If there's a bridge, attach to it and let it create the connector */
-	ret = drm_bridge_attach(&dsi->encoder, dsi->bridge, NULL);
-	if (ret) {
-		DRM_ERROR("Failed to attach bridge to drm\n");
-
+	if (dsi->bridge) {
+		ret = drm_bridge_attach(&dsi->encoder, dsi->bridge, NULL);
+		if (ret) {
+			DRM_ERROR("Failed to attach bridge to drm\n");
+			goto err_encoder_cleanup;
+		}
+	} else {
 		/* Otherwise create our own connector and attach to a panel */
 		ret = mtk_dsi_create_connector(drm, dsi);
 		if (ret)
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 6cbbae3f438b..db1bf7f88c1f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -198,6 +198,22 @@ nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
 /******************************************************************************
  * EVO channel helpers
  *****************************************************************************/
+static void
+evo_flush(struct nv50_dmac *dmac)
+{
+	/* Push buffer fetches are not coherent with BAR1, we need to ensure
+	 * writes have been flushed right through to VRAM before writing PUT.
+	 */
+	if (dmac->push.type & NVIF_MEM_VRAM) {
+		struct nvif_device *device = dmac->base.device;
+		nvif_wr32(&device->object, 0x070000, 0x00000001);
+		nvif_msec(device, 2000,
+			if (!(nvif_rd32(&device->object, 0x070000) & 0x00000002))
+				break;
+		);
+	}
+}
+
 u32 *
 evo_wait(struct nv50_dmac *evoc, int nr)
 {
@@ -208,6 +224,7 @@ evo_wait(struct nv50_dmac *evoc, int nr)
 	mutex_lock(&dmac->lock);
 	if (put + nr >= (PAGE_SIZE / 4) - 8) {
 		dmac->ptr[put] = 0x20000000;
+		evo_flush(dmac);
 
 		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
 		if (nvif_msec(device, 2000,
@@ -230,17 +247,7 @@ evo_kick(u32 *push, struct nv50_dmac *evoc)
 {
 	struct nv50_dmac *dmac = evoc;
 
-	/* Push buffer fetches are not coherent with BAR1, we need to ensure
-	 * writes have been flushed right through to VRAM before writing PUT.
-	 */
-	if (dmac->push.type & NVIF_MEM_VRAM) {
-		struct nvif_device *device = dmac->base.device;
-		nvif_wr32(&device->object, 0x070000, 0x00000001);
-		nvif_msec(device, 2000,
-			if (!(nvif_rd32(&device->object, 0x070000) & 0x00000002))
-				break;
-		);
-	}
+	evo_flush(dmac);
 
 	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
 	mutex_unlock(&dmac->lock);
@@ -1264,6 +1271,7 @@ nv50_mstm_del(struct nv50_mstm **pmstm)
 {
 	struct nv50_mstm *mstm = *pmstm;
 	if (mstm) {
+		drm_dp_mst_topology_mgr_destroy(&mstm->mgr);
 		kfree(*pmstm);
 		*pmstm = NULL;
 	}
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 2b2baf6e0e0d..d2928d43f29a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -1171,10 +1171,16 @@ nouveau_platform_device_create(const struct nvkm_device_tegra_func *func,
 		goto err_free;
 	}
 
+	err = nouveau_drm_device_init(drm);
+	if (err)
+		goto err_put;
+
 	platform_set_drvdata(pdev, drm);
 
 	return drm;
 
+err_put:
+	drm_dev_put(drm);
 err_free:
 	nvkm_device_del(pdevice);
 
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
index 941f35233b1f..5864cb452c5c 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
@@ -448,11 +448,6 @@ static int rockchip_drm_platform_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static void rockchip_drm_platform_shutdown(struct platform_device *pdev)
-{
-	rockchip_drm_platform_remove(pdev);
-}
-
 static const struct of_device_id rockchip_drm_dt_ids[] = {
 	{ .compatible = "rockchip,display-subsystem", },
 	{ /* sentinel */ },
@@ -462,7 +457,6 @@ MODULE_DEVICE_TABLE(of, rockchip_drm_dt_ids);
 static struct platform_driver rockchip_drm_platform_driver = {
 	.probe = rockchip_drm_platform_probe,
 	.remove = rockchip_drm_platform_remove,
-	.shutdown = rockchip_drm_platform_shutdown,
 	.driver = {
 		.name = "rockchip-drm",
 		.of_match_table = rockchip_drm_dt_ids,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index 61a84b958d67..d7a2dfb8ee9b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -49,6 +49,8 @@
 
 #define VMWGFX_REPO "In Tree"
 
+#define VMWGFX_VALIDATION_MEM_GRAN (16*PAGE_SIZE)
+
 
 /**
  * Fully encoded drm commands. Might move to vmw_drm.h
@@ -918,7 +920,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 		spin_unlock(&dev_priv->cap_lock);
 	}
 
-
+	vmw_validation_mem_init_ttm(dev_priv, VMWGFX_VALIDATION_MEM_GRAN);
 	ret = vmw_kms_init(dev_priv);
 	if (unlikely(ret != 0))
 		goto out_no_kms;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 59f614225bcd..aca974b14b55 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -606,6 +606,9 @@ struct vmw_private {
 
 	struct vmw_cmdbuf_man *cman;
 	DECLARE_BITMAP(irqthread_pending, VMW_IRQTHREAD_MAX);
+
+	/* Validation memory reservation */
+	struct vmw_validation_mem vvm;
 };
 
 static inline struct vmw_surface *vmw_res_to_srf(struct vmw_resource *res)
@@ -846,6 +849,8 @@ extern int vmw_ttm_global_init(struct vmw_private *dev_priv);
 extern void vmw_ttm_global_release(struct vmw_private *dev_priv);
 extern int vmw_mmap(struct file *filp, struct vm_area_struct *vma);
 
+extern void vmw_validation_mem_init_ttm(struct vmw_private *dev_priv,
+					size_t gran);
 /**
  * TTM buffer object driver - vmwgfx_ttm_buffer.c
  */
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index 5a6b70ba137a..f2d13a72c05d 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -1738,7 +1738,6 @@ static int vmw_cmd_check_define_gmrfb(struct vmw_private *dev_priv,
 				      void *buf)
 {
 	struct vmw_buffer_object *vmw_bo;
-	int ret;
 
 	struct {
 		uint32_t header;
@@ -1748,7 +1747,6 @@ static int vmw_cmd_check_define_gmrfb(struct vmw_private *dev_priv,
 	return vmw_translate_guest_ptr(dev_priv, sw_context,
 				       &cmd->body.ptr,
 				       &vmw_bo);
-	return ret;
 }
 
 
@@ -3837,6 +3835,8 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	struct sync_file *sync_file = NULL;
 	DECLARE_VAL_CONTEXT(val_ctx, &sw_context->res_ht, 1);
 
+	vmw_validation_set_val_mem(&val_ctx, &dev_priv->vvm);
+
 	if (flags & DRM_VMW_EXECBUF_FLAG_EXPORT_FENCE_FD) {
 		out_fence_fd = get_unused_fd_flags(O_CLOEXEC);
 		if (out_fence_fd < 0) {
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
index 7b1e5a5cbd2c..f88247046721 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
@@ -96,3 +96,39 @@ void vmw_ttm_global_release(struct vmw_private *dev_priv)
 	drm_global_item_unref(&dev_priv->bo_global_ref.ref);
 	drm_global_item_unref(&dev_priv->mem_global_ref);
 }
+
+/* struct vmw_validation_mem callback */
+static int vmw_vmt_reserve(struct vmw_validation_mem *m, size_t size)
+{
+	static struct ttm_operation_ctx ctx = {.interruptible = false,
+					       .no_wait_gpu = false};
+	struct vmw_private *dev_priv = container_of(m, struct vmw_private, vvm);
+
+	return ttm_mem_global_alloc(vmw_mem_glob(dev_priv), size, &ctx);
+}
+
+/* struct vmw_validation_mem callback */
+static void vmw_vmt_unreserve(struct vmw_validation_mem *m, size_t size)
+{
+	struct vmw_private *dev_priv = container_of(m, struct vmw_private, vvm);
+
+	return ttm_mem_global_free(vmw_mem_glob(dev_priv), size);
+}
+
+/**
+ * vmw_validation_mem_init_ttm - Interface the validation memory tracker
+ * to ttm.
+ * @dev_priv: Pointer to struct vmw_private. The reason we choose a vmw private
+ * rather than a struct vmw_validation_mem is to make sure assumption in the
+ * callbacks that struct vmw_private derives from struct vmw_validation_mem
+ * holds true.
+ * @gran: The recommended allocation granularity
+ */
+void vmw_validation_mem_init_ttm(struct vmw_private *dev_priv, size_t gran)
+{
+	struct vmw_validation_mem *vvm = &dev_priv->vvm;
+
+	vvm->reserve_mem = vmw_vmt_reserve;
+	vvm->unreserve_mem = vmw_vmt_unreserve;
+	vvm->gran = gran;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
index 184025fa938e..f116f092e00b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
@@ -104,11 +104,25 @@ void *vmw_validation_mem_alloc(struct vmw_validation_context *ctx,
 		return NULL;
 
 	if (ctx->mem_size_left < size) {
-		struct page *page = alloc_page(GFP_KERNEL | __GFP_ZERO);
+		struct page *page;
 
+		if (ctx->vm && ctx->vm_size_left < PAGE_SIZE) {
+			int ret = ctx->vm->reserve_mem(ctx->vm, ctx->vm->gran);
+
+			if (ret)
+				return NULL;
+
+			ctx->vm_size_left += ctx->vm->gran;
+			ctx->total_mem += ctx->vm->gran;
+		}
+
+		page = alloc_page(GFP_KERNEL | __GFP_ZERO);
 		if (!page)
 			return NULL;
 
+		if (ctx->vm)
+			ctx->vm_size_left -= PAGE_SIZE;
+
 		list_add_tail(&page->lru, &ctx->page_list);
 		ctx->page_address = page_address(page);
 		ctx->mem_size_left = PAGE_SIZE;
@@ -138,6 +152,11 @@ static void vmw_validation_mem_free(struct vmw_validation_context *ctx)
 	}
 
 	ctx->mem_size_left = 0;
+	if (ctx->vm && ctx->total_mem) {
+		ctx->vm->unreserve_mem(ctx->vm, ctx->total_mem);
+		ctx->total_mem = 0;
+		ctx->vm_size_left = 0;
+	}
 }
 
 /**
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h
index b57e3292c386..3b396fea40d7 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.h
@@ -34,6 +34,21 @@
 #include <drm/ttm/ttm_execbuf_util.h>
 
 /**
+ * struct vmw_validation_mem - Custom interface to provide memory reservations
+ * for the validation code.
+ * @reserve_mem: Callback to reserve memory
+ * @unreserve_mem: Callback to unreserve memory
+ * @gran: Reservation granularity. Contains a hint how much memory should
+ * be reserved in each call to @reserve_mem(). A slow implementation may want
+ * reservation to be done in large batches.
+ */
+struct vmw_validation_mem {
+	int (*reserve_mem)(struct vmw_validation_mem *m, size_t size);
+	void (*unreserve_mem)(struct vmw_validation_mem *m, size_t size);
+	size_t gran;
+};
+
+/**
  * struct vmw_validation_context - Per command submission validation context
  * @ht: Hash table used to find resource- or buffer object duplicates
  * @resource_list: List head for resource validation metadata
@@ -47,6 +62,10 @@
  * buffer objects
  * @mem_size_left: Free memory left in the last page in @page_list
  * @page_address: Kernel virtual address of the last page in @page_list
+ * @vm: A pointer to the memory reservation interface or NULL if no
+ * memory reservation is needed.
+ * @vm_size_left: Amount of reserved memory that so far has not been allocated.
+ * @total_mem: Amount of reserved memory.
  */
 struct vmw_validation_context {
 	struct drm_open_hash *ht;
@@ -59,6 +78,9 @@ struct vmw_validation_context {
 	unsigned int merge_dups;
 	unsigned int mem_size_left;
 	u8 *page_address;
+	struct vmw_validation_mem *vm;
+	size_t vm_size_left;
+	size_t total_mem;
 };
 
 struct vmw_buffer_object;
@@ -102,6 +124,21 @@ vmw_validation_has_bos(struct vmw_validation_context *ctx)
 }
 
 /**
+ * vmw_validation_set_val_mem - Register a validation mem object for
+ * validation memory reservation
+ * @ctx: The validation context
+ * @vm: Pointer to a struct vmw_validation_mem
+ *
+ * Must be set before the first attempt to allocate validation memory.
+ */
+static inline void
+vmw_validation_set_val_mem(struct vmw_validation_context *ctx,
+			   struct vmw_validation_mem *vm)
+{
+	ctx->vm = vm;
+}
+
+/**
  * vmw_validation_set_ht - Register a hash table for duplicate finding
  * @ctx: The validation context
  * @ht: Pointer to a hash table to use for duplicate finding