Merge tag 'amd-drm-fixes-6.3-2023-03-02' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-fixes-6.3-2023-03-02:

amdgpu:
- SMU 13 fixes
- Enable TMZ for GC 10.3.6
- Misc display fixes
- Buddy allocator fixes
- GC 11 fixes
- S0ix fix
- INFO IOCTL queries for GC 11
- VCN harvest fixes for SR-IOV
- UMC 8.10 RAS fixes
- Don't restrict bpc to 8
- NBIO 7.5 fix
- Allow freesync on PCon for more devices

amdkfd:
- SDMA fix
- Illegal memory access fix

radeon:
- Display fix for iMac11,2

UAPI:
- Add some additional INFO IOCTL queries for GC 11 fixes
  Mesa MR: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21403

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230302051843.7793-1-alexander.deucher@amd.com

38 files changed, 409 insertions, 128 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig
index 5341b6b242c3..a82d36ea88e2 100644
--- a/drivers/gpu/drm/amd/amdgpu/Kconfig
+++ b/drivers/gpu/drm/amd/amdgpu/Kconfig
@@ -6,6 +6,7 @@ config DRM_AMDGPU
 	select FW_LOADER
 	select DRM_DISPLAY_DP_HELPER
 	select DRM_DISPLAY_HDMI_HELPER
+	select DRM_DISPLAY_HDCP_HELPER
 	select DRM_DISPLAY_HELPER
 	select DRM_KMS_HELPER
 	select DRM_SCHED
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
index 458362e4ea01..d4196fcb85a0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
@@ -1073,6 +1073,9 @@ bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev)
 	    (pm_suspend_target_state != PM_SUSPEND_TO_IDLE))
 		return false;
 
+	if (adev->asic_type < CHIP_RAVEN)
+		return false;
+
 	/*
 	 * If ACPI_FADT_LOW_POWER_S0 is not set in the FADT, it is generally
 	 * risky to do any special firmware-related preparations for entering
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 86fbb4138285..f5ffca24def4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -107,9 +107,12 @@
  * - 3.50.0 - Update AMDGPU_INFO_DEV_INFO IOCTL for minimum engine and memory clock
  *            Update AMDGPU_INFO_SENSOR IOCTL for PEAK_PSTATE engine and memory clock
  *   3.51.0 - Return the PCIe gen and lanes from the INFO ioctl
+ *   3.52.0 - Add AMDGPU_IDS_FLAGS_CONFORMANT_TRUNC_COORD, add device_info fields:
+ *            tcp_cache_size, num_sqc_per_wgp, sqc_data_cache_size, sqc_inst_cache_size,
+ *            gl1c_cache_size, gl2c_cache_size, mall_size, enabled_rb_pipes_mask_hi
  */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	51
+#define KMS_DRIVER_MINOR	52
 #define KMS_DRIVER_PATCHLEVEL	0
 
 unsigned int amdgpu_vram_limit = UINT_MAX;
@@ -921,7 +924,7 @@ module_param_named(reset_method, amdgpu_reset_method, int, 0444);
  * result in the GPU entering bad status when the number of total
  * faulty pages by ECC exceeds the threshold value.
  */
-MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = auto(default value), 0 = disable bad page retirement, -2 = ignore bad page threshold)");
+MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = ignore threshold (default value), 0 = disable bad page retirement, -2 = driver sets threshold)");
 module_param_named(bad_page_threshold, amdgpu_bad_page_threshold, int, 0444);
 
 MODULE_PARM_DESC(num_kcq, "number of kernel compute queue user want to setup (8 if set to greater than 8 or less than 0, only affect gfx 8+)");
@@ -2414,8 +2417,10 @@ static int amdgpu_pmops_suspend(struct device *dev)
 
 	if (amdgpu_acpi_is_s0ix_active(adev))
 		adev->in_s0ix = true;
-	else
+	else if (amdgpu_acpi_is_s3_active(adev))
 		adev->in_s3 = true;
+	if (!adev->in_s0ix && !adev->in_s3)
+		return 0;
 	return amdgpu_device_suspend(drm_dev, true);
 }
 
@@ -2436,6 +2441,9 @@ static int amdgpu_pmops_resume(struct device *dev)
 	struct amdgpu_device *adev = drm_to_adev(drm_dev);
 	int r;
 
+	if (!adev->in_s0ix && !adev->in_s3)
+		return 0;
+
 	/* Avoids registers access if device is physically gone */
 	if (!pci_device_is_present(adev->pdev))
 		adev->no_hw_access = true;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
index 86ec9d0d12c8..de9e7a00bb15 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
@@ -178,6 +178,8 @@ struct amdgpu_gfx_config {
 	uint32_t num_sc_per_sh;
 	uint32_t num_packer_per_sc;
 	uint32_t pa_sc_tile_steering_override;
+	/* Whether texture coordinate truncation is conformant. */
+	bool ta_cntl2_truncate_coord_mode;
 	uint64_t tcc_disabled_mask;
 	uint32_t gc_num_tcp_per_sa;
 	uint32_t gc_num_sdp_interface;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index 94f10ac0eef7..12a6826caef4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -552,6 +552,7 @@ void amdgpu_gmc_tmz_set(struct amdgpu_device *adev)
 	case IP_VERSION(10, 3, 2):
 	case IP_VERSION(10, 3, 4):
 	case IP_VERSION(10, 3, 5):
+	case IP_VERSION(10, 3, 6):
 	/* VANGOGH */
 	case IP_VERSION(10, 3, 1):
 	/* YELLOW_CARP*/
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index ca945055e683..0efb38539d70 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -808,6 +808,8 @@ int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 			dev_info->ids_flags |= AMDGPU_IDS_FLAGS_PREEMPTION;
 		if (amdgpu_is_tmz(adev))
 			dev_info->ids_flags |= AMDGPU_IDS_FLAGS_TMZ;
+		if (adev->gfx.config.ta_cntl2_truncate_coord_mode)
+			dev_info->ids_flags |= AMDGPU_IDS_FLAGS_CONFORMANT_TRUNC_COORD;
 
 		vm_size = adev->vm_manager.max_pfn * AMDGPU_GPU_PAGE_SIZE;
 		vm_size -= AMDGPU_VA_RESERVED_SIZE;
@@ -865,6 +867,15 @@ int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 			adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 ? 4 :
 			adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 ? 2 : 1;
 
+		dev_info->tcp_cache_size = adev->gfx.config.gc_tcp_l1_size;
+		dev_info->num_sqc_per_wgp = adev->gfx.config.gc_num_sqc_per_wgp;
+		dev_info->sqc_data_cache_size = adev->gfx.config.gc_l1_data_cache_size_per_sqc;
+		dev_info->sqc_inst_cache_size = adev->gfx.config.gc_l1_instruction_cache_size_per_sqc;
+		dev_info->gl1c_cache_size = adev->gfx.config.gc_gl1c_size_per_instance *
+					    adev->gfx.config.gc_gl1c_per_sa;
+		dev_info->gl2c_cache_size = adev->gfx.config.gc_gl2c_per_gpu;
+		dev_info->mall_size = adev->gmc.mall_size;
+
 		ret = copy_to_user(out, dev_info,
 				   min((size_t)size, sizeof(*dev_info))) ? -EFAULT : 0;
 		kfree(dev_info);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 981010de0a28..e3e1ed4314dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -139,7 +139,7 @@ void amdgpu_bo_placement_from_domain(struct amdgpu_bo *abo, u32 domain)
 
 		if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)
 			places[c].lpfn = visible_pfn;
-		else
+		else if (adev->gmc.real_vram_size != adev->gmc.visible_vram_size)
 			places[c].flags |= TTM_PL_FLAG_TOPDOWN;
 
 		if (flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index 15e601f09648..28fe6d941054 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -1683,7 +1683,7 @@ static int psp_hdcp_initialize(struct psp_context *psp)
 	psp->hdcp_context.context.mem_context.shared_mem_size = PSP_HDCP_SHARED_MEM_SIZE;
 	psp->hdcp_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
 
-	if (!psp->hdcp_context.context.initialized) {
+	if (!psp->hdcp_context.context.mem_context.shared_buf) {
 		ret = psp_ta_init_shared_buf(psp, &psp->hdcp_context.context.mem_context);
 		if (ret)
 			return ret;
@@ -1750,7 +1750,7 @@ static int psp_dtm_initialize(struct psp_context *psp)
 	psp->dtm_context.context.mem_context.shared_mem_size = PSP_DTM_SHARED_MEM_SIZE;
 	psp->dtm_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
 
-	if (!psp->dtm_context.context.initialized) {
+	if (!psp->dtm_context.context.mem_context.shared_buf) {
 		ret = psp_ta_init_shared_buf(psp, &psp->dtm_context.context.mem_context);
 		if (ret)
 			return ret;
@@ -1818,7 +1818,7 @@ static int psp_rap_initialize(struct psp_context *psp)
 	psp->rap_context.context.mem_context.shared_mem_size = PSP_RAP_SHARED_MEM_SIZE;
 	psp->rap_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA;
 
-	if (!psp->rap_context.context.initialized) {
+	if (!psp->rap_context.context.mem_context.shared_buf) {
 		ret = psp_ta_init_shared_buf(psp, &psp->rap_context.context.mem_context);
 		if (ret)
 			return ret;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 6e543558386d..63dfcc98152d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -176,7 +176,7 @@ static int amdgpu_reserve_page_direct(struct amdgpu_device *adev, uint64_t addre
 	if (amdgpu_bad_page_threshold != 0) {
 		amdgpu_ras_add_bad_pages(adev, err_data.err_addr,
 					 err_data.err_addr_cnt);
-		amdgpu_ras_save_bad_pages(adev);
+		amdgpu_ras_save_bad_pages(adev, NULL);
 	}
 
 	dev_warn(adev->dev, "WARNING: THIS IS ONLY FOR TEST PURPOSES AND WILL CORRUPT RAS EEPROM\n");
@@ -2084,22 +2084,32 @@ out:
 /*
  * write error record array to eeprom, the function should be
  * protected by recovery_lock
+ * new_cnt: new added UE count, excluding reserved bad pages, can be NULL
  */
-int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev)
+int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev,
+		unsigned long *new_cnt)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct ras_err_handler_data *data;
 	struct amdgpu_ras_eeprom_control *control;
 	int save_count;
 
-	if (!con || !con->eh_data)
+	if (!con || !con->eh_data) {
+		if (new_cnt)
+			*new_cnt = 0;
+
 		return 0;
+	}
 
 	mutex_lock(&con->recovery_lock);
 	control = &con->eeprom_control;
 	data = con->eh_data;
 	save_count = data->count - control->ras_num_recs;
 	mutex_unlock(&con->recovery_lock);
+
+	if (new_cnt)
+		*new_cnt = save_count / adev->umc.retire_unit;
+
 	/* only new entries are saved */
 	if (save_count > 0) {
 		if (amdgpu_ras_eeprom_append(control,
@@ -2186,11 +2196,12 @@ static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev,
 	/*
 	 * Justification of value bad_page_cnt_threshold in ras structure
 	 *
-	 * Generally, -1 <= amdgpu_bad_page_threshold <= max record length
-	 * in eeprom, and introduce two scenarios accordingly.
+	 * Generally, 0 <= amdgpu_bad_page_threshold <= max record length
+	 * in eeprom or amdgpu_bad_page_threshold == -2, introduce two
+	 * scenarios accordingly.
 	 *
 	 * Bad page retirement enablement:
-	 *    - If amdgpu_bad_page_threshold = -1,
+	 *    - If amdgpu_bad_page_threshold = -2,
 	 *      bad_page_cnt_threshold = typical value by formula.
 	 *
 	 *    - When the value from user is 0 < amdgpu_bad_page_threshold <
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index f2ad999993f6..ef38f4c93df0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -547,7 +547,8 @@ int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
 int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
 		struct eeprom_table_record *bps, int pages);
 
-int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev);
+int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev,
+		unsigned long *new_cnt);
 
 static inline enum ta_ras_block
 amdgpu_ras_block_to_ta(enum amdgpu_ras_block block) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 2d9f3f4cd79e..2e08fce87521 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -417,7 +417,8 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 
-	if (!__is_ras_eeprom_supported(adev))
+	if (!__is_ras_eeprom_supported(adev) ||
+	    !amdgpu_bad_page_threshold)
 		return false;
 
 	/* skip check eeprom table for VEGA20 Gaming */
@@ -428,10 +429,18 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)
 			return false;
 
 	if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) {
-		dev_warn(adev->dev, "This GPU is in BAD status.");
-		dev_warn(adev->dev, "Please retire it or set a larger "
-			 "threshold value when reloading driver.\n");
-		return true;
+		if (amdgpu_bad_page_threshold == -1) {
+			dev_warn(adev->dev, "RAS records:%d exceed threshold:%d",
+				con->eeprom_control.ras_num_recs, con->bad_page_cnt_threshold);
+			dev_warn(adev->dev,
+				"But GPU can be operated due to bad_page_threshold = -1.\n");
+			return false;
+		} else {
+			dev_warn(adev->dev, "This GPU is in BAD status.");
+			dev_warn(adev->dev, "Please retire it or set a larger "
+				 "threshold value when reloading driver.\n");
+			return true;
+		}
 	}
 
 	return false;
@@ -1191,8 +1200,8 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
 		} else {
 			dev_err(adev->dev, "RAS records:%d exceed threshold:%d",
 				control->ras_num_recs, ras->bad_page_cnt_threshold);
-			if (amdgpu_bad_page_threshold == -2) {
-				dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -2.");
+			if (amdgpu_bad_page_threshold == -1) {
+				dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -1.");
 				res = 0;
 			} else {
 				*exceed_err_limit = true;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
index 1c7fcb4f2380..1b8574bc4463 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
@@ -68,7 +68,7 @@ int amdgpu_umc_page_retirement_mca(struct amdgpu_device *adev,
 	if (amdgpu_bad_page_threshold != 0) {
 		amdgpu_ras_add_bad_pages(adev, err_data.err_addr,
 						err_data.err_addr_cnt);
-		amdgpu_ras_save_bad_pages(adev);
+		amdgpu_ras_save_bad_pages(adev, NULL);
 	}
 
 out:
@@ -147,7 +147,7 @@ static int amdgpu_umc_do_page_retirement(struct amdgpu_device *adev,
 			err_data->err_addr_cnt) {
 			amdgpu_ras_add_bad_pages(adev, err_data->err_addr,
 						err_data->err_addr_cnt);
-			amdgpu_ras_save_bad_pages(adev);
+			amdgpu_ras_save_bad_pages(adev, &(err_data->ue_count));
 
 			amdgpu_dpm_send_hbm_bad_pages_num(adev, con->eeprom_control.ras_num_recs);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
index a6951160f13a..f2bf979af588 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
@@ -74,6 +74,8 @@ struct amdgpu_umc {
 
 	/* UMC regiser per channel offset */
 	uint32_t channel_offs;
+	/* how many pages are retired in one UE */
+	uint32_t retire_unit;
 	/* channel index table of interleaved memory */
 	const uint32_t *channel_idx_tbl;
 	struct ras_common_if *ras_if;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
index 9fa1d814508a..43d6a9d6a538 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -453,7 +453,8 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man,
 		/* Limit maximum size to 2GiB due to SG table limitations */
 		size = min(remaining_size, 2ULL << 30);
 
-		if (size >= (u64)pages_per_block << PAGE_SHIFT)
+		if ((size >= (u64)pages_per_block << PAGE_SHIFT) &&
+				!(size & (((u64)pages_per_block << PAGE_SHIFT) - 1)))
 			min_block_size = (u64)pages_per_block << PAGE_SHIFT;
 
 		cur_size = size;
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
index 8ad8a0bffcac..3bf697a80cf2 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
@@ -1503,44 +1503,70 @@ static void gfx_v11_0_select_se_sh(struct amdgpu_device *adev, u32 se_num,
 	WREG32_SOC15(GC, 0, regGRBM_GFX_INDEX, data);
 }
 
-static u32 gfx_v11_0_get_rb_active_bitmap(struct amdgpu_device *adev)
+static u32 gfx_v11_0_get_sa_active_bitmap(struct amdgpu_device *adev)
 {
-	u32 data, mask;
+	u32 gc_disabled_sa_mask, gc_user_disabled_sa_mask, sa_mask;
+
+	gc_disabled_sa_mask = RREG32_SOC15(GC, 0, regCC_GC_SA_UNIT_DISABLE);
+	gc_disabled_sa_mask = REG_GET_FIELD(gc_disabled_sa_mask,
+					   CC_GC_SA_UNIT_DISABLE,
+					   SA_DISABLE);
+	gc_user_disabled_sa_mask = RREG32_SOC15(GC, 0, regGC_USER_SA_UNIT_DISABLE);
+	gc_user_disabled_sa_mask = REG_GET_FIELD(gc_user_disabled_sa_mask,
+						 GC_USER_SA_UNIT_DISABLE,
+						 SA_DISABLE);
+	sa_mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_sh_per_se *
+					    adev->gfx.config.max_shader_engines);
 
-	data = RREG32_SOC15(GC, 0, regCC_RB_BACKEND_DISABLE);
-	data |= RREG32_SOC15(GC, 0, regGC_USER_RB_BACKEND_DISABLE);
+	return sa_mask & (~(gc_disabled_sa_mask | gc_user_disabled_sa_mask));
+}
 
-	data &= CC_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK;
-	data >>= GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT;
+static u32 gfx_v11_0_get_rb_active_bitmap(struct amdgpu_device *adev)
+{
+	u32 gc_disabled_rb_mask, gc_user_disabled_rb_mask;
+	u32 rb_mask;
 
-	mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_backends_per_se /
-					 adev->gfx.config.max_sh_per_se);
+	gc_disabled_rb_mask = RREG32_SOC15(GC, 0, regCC_RB_BACKEND_DISABLE);
+	gc_disabled_rb_mask = REG_GET_FIELD(gc_disabled_rb_mask,
+					    CC_RB_BACKEND_DISABLE,
+					    BACKEND_DISABLE);
+	gc_user_disabled_rb_mask = RREG32_SOC15(GC, 0, regGC_USER_RB_BACKEND_DISABLE);
+	gc_user_disabled_rb_mask = REG_GET_FIELD(gc_user_disabled_rb_mask,
+						 GC_USER_RB_BACKEND_DISABLE,
+						 BACKEND_DISABLE);
+	rb_mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_backends_per_se *
+					    adev->gfx.config.max_shader_engines);
 
-	return (~data) & mask;
+	return rb_mask & (~(gc_disabled_rb_mask | gc_user_disabled_rb_mask));
 }
 
 static void gfx_v11_0_setup_rb(struct amdgpu_device *adev)
 {
-	int i, j;
-	u32 data;
-	u32 active_rbs = 0;
-	u32 rb_bitmap_width_per_sh = adev->gfx.config.max_backends_per_se /
-					adev->gfx.config.max_sh_per_se;
+	u32 rb_bitmap_width_per_sa;
+	u32 max_sa;
+	u32 active_sa_bitmap;
+	u32 global_active_rb_bitmap;
+	u32 active_rb_bitmap = 0;
+	u32 i;
 
-	mutex_lock(&adev->grbm_idx_mutex);
-	for (i = 0; i < adev->gfx.config.max_shader_engines; i++) {
-		for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) {
-			gfx_v11_0_select_se_sh(adev, i, j, 0xffffffff);
-			data = gfx_v11_0_get_rb_active_bitmap(adev);
-			active_rbs |= data << ((i * adev->gfx.config.max_sh_per_se + j) *
-					       rb_bitmap_width_per_sh);
-		}
+	/* query sa bitmap from SA_UNIT_DISABLE registers */
+	active_sa_bitmap = gfx_v11_0_get_sa_active_bitmap(adev);
+	/* query rb bitmap from RB_BACKEND_DISABLE registers */
+	global_active_rb_bitmap = gfx_v11_0_get_rb_active_bitmap(adev);
+
+	/* generate active rb bitmap according to active sa bitmap */
+	max_sa = adev->gfx.config.max_shader_engines *
+		 adev->gfx.config.max_sh_per_se;
+	rb_bitmap_width_per_sa = adev->gfx.config.max_backends_per_se /
+				 adev->gfx.config.max_sh_per_se;
+	for (i = 0; i < max_sa; i++) {
+		if (active_sa_bitmap & (1 << i))
+			active_rb_bitmap |= (0x3 << (i * rb_bitmap_width_per_sa));
 	}
-	gfx_v11_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
-	mutex_unlock(&adev->grbm_idx_mutex);
 
-	adev->gfx.config.backend_enable_mask = active_rbs;
-	adev->gfx.config.num_rbs = hweight32(active_rbs);
+	active_rb_bitmap |= global_active_rb_bitmap;
+	adev->gfx.config.backend_enable_mask = active_rb_bitmap;
+	adev->gfx.config.num_rbs = hweight32(active_rb_bitmap);
 }
 
 #define DEFAULT_SH_MEM_BASES	(0x6000)
@@ -1633,6 +1659,11 @@ static void gfx_v11_0_constants_init(struct amdgpu_device *adev)
 	gfx_v11_0_get_tcc_info(adev);
 	adev->gfx.config.pa_sc_tile_steering_override = 0;
 
+	/* Set whether texture coordinate truncation is conformant. */
+	tmp = RREG32_SOC15(GC, 0, regTA_CNTL2);
+	adev->gfx.config.ta_cntl2_truncate_coord_mode =
+		REG_GET_FIELD(tmp, TA_CNTL2, TRUNCATE_COORD_MODE);
+
 	/* XXX SH_MEM regs */
 	/* where to put LDS, scratch, GPUVM in FSA64 space */
 	mutex_lock(&adev->srbm_mutex);
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
index 7db1f1a7e33c..ab2556ca984e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
@@ -692,6 +692,7 @@ static void gmc_v10_0_set_umc_funcs(struct amdgpu_device *adev)
 		adev->umc.channel_inst_num = UMC_V8_7_CHANNEL_INSTANCE_NUM;
 		adev->umc.umc_inst_num = UMC_V8_7_UMC_INSTANCE_NUM;
 		adev->umc.channel_offs = UMC_V8_7_PER_CHANNEL_OFFSET_SIENNA;
+		adev->umc.retire_unit = 1;
 		adev->umc.channel_idx_tbl = &umc_v8_7_channel_idx_tbl[0][0];
 		adev->umc.ras = &umc_v8_7_ras;
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
index 0a31a341aa43..85e0afc3d4f7 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
@@ -570,6 +570,7 @@ static void gmc_v11_0_set_umc_funcs(struct amdgpu_device *adev)
 		adev->umc.node_inst_num = adev->gmc.num_umc;
 		adev->umc.max_ras_err_cnt_per_query = UMC_V8_10_TOTAL_CHANNEL_NUM(adev);
 		adev->umc.channel_offs = UMC_V8_10_PER_CHANNEL_OFFSET;
+		adev->umc.retire_unit = UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM;
 		if (adev->umc.node_inst_num == 4)
 			adev->umc.channel_idx_tbl = &umc_v8_10_channel_idx_tbl_ext0[0][0][0];
 		else
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index d65c6cea3445..b06170c00dfc 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -1288,6 +1288,7 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev)
 		adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM;
 		adev->umc.umc_inst_num = UMC_V6_1_UMC_INSTANCE_NUM;
 		adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET_VG20;
+		adev->umc.retire_unit = 1;
 		adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0];
 		adev->umc.ras = &umc_v6_1_ras;
 		break;
@@ -1296,6 +1297,7 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev)
 		adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM;
 		adev->umc.umc_inst_num = UMC_V6_1_UMC_INSTANCE_NUM;
 		adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET_ARCT;
+		adev->umc.retire_unit = 1;
 		adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0];
 		adev->umc.ras = &umc_v6_1_ras;
 		break;
@@ -1305,6 +1307,7 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev)
 		adev->umc.channel_inst_num = UMC_V6_7_CHANNEL_INSTANCE_NUM;
 		adev->umc.umc_inst_num = UMC_V6_7_UMC_INSTANCE_NUM;
 		adev->umc.channel_offs = UMC_V6_7_PER_CHANNEL_OFFSET;
+		adev->umc.retire_unit = (UMC_V6_7_NA_MAP_PA_NUM * 2);
 		if (!adev->gmc.xgmi.connected_to_cpu)
 			adev->umc.ras = &umc_v6_7_ras;
 		if (1 & adev->smuio.funcs->get_die_id(adev))
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c
index 31776b12e4c4..4b0d563c6522 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c
@@ -382,6 +382,11 @@ static void nbio_v7_2_init_registers(struct amdgpu_device *adev)
 		if (def != data)
 			WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regBIF1_PCIE_MST_CTRL_3), data);
 		break;
+	case IP_VERSION(7, 5, 1):
+		data = RREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2);
+		data &= ~RCC_DEV2_EPF0_STRAP2__STRAP_NO_SOFT_RESET_DEV2_F0_MASK;
+		WREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2, data);
+		fallthrough;
 	default:
 		def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CONFIG_CNTL));
 		data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL,
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c
index da394bc06bba..fb55e8cb9967 100644
--- a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c
@@ -209,6 +209,45 @@ static int umc_v8_10_swizzle_mode_na_to_pa(struct amdgpu_device *adev,
 	return 0;
 }
 
+static void umc_v8_10_convert_error_address(struct amdgpu_device *adev,
+					    struct ras_err_data *err_data, uint64_t err_addr,
+					    uint32_t ch_inst, uint32_t umc_inst,
+					    uint32_t node_inst, uint64_t mc_umc_status)
+{
+	uint64_t na_err_addr_base;
+	uint64_t na_err_addr, retired_page_addr;
+	uint32_t channel_index, addr_lsb, col = 0;
+	int ret = 0;
+
+	channel_index =
+		adev->umc.channel_idx_tbl[node_inst * adev->umc.umc_inst_num *
+					adev->umc.channel_inst_num +
+					umc_inst * adev->umc.channel_inst_num +
+					ch_inst];
+
+	/* the lowest lsb bits should be ignored */
+	addr_lsb = REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrLsb);
+	err_addr &= ~((0x1ULL << addr_lsb) - 1);
+	na_err_addr_base = err_addr & ~(0x3ULL << UMC_V8_10_NA_C5_BIT);
+
+	/* loop for all possibilities of [C6 C5] in normal address. */
+	for (col = 0; col < UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM; col++) {
+		na_err_addr = na_err_addr_base | (col << UMC_V8_10_NA_C5_BIT);
+
+		/* Mapping normal error address to retired soc physical address. */
+		ret = umc_v8_10_swizzle_mode_na_to_pa(adev, channel_index,
+						na_err_addr, &retired_page_addr);
+		if (ret) {
+			dev_err(adev->dev, "Failed to map pa from umc na.\n");
+			break;
+		}
+		dev_info(adev->dev, "Error Address(PA): 0x%llx\n",
+			retired_page_addr);
+		amdgpu_umc_fill_error_record(err_data, na_err_addr,
+				retired_page_addr, channel_index, umc_inst);
+	}
+}
+
 static void umc_v8_10_query_error_address(struct amdgpu_device *adev,
 					 struct ras_err_data *err_data,
 					 uint32_t umc_reg_offset,
@@ -218,10 +257,7 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev,
 {
 	uint64_t mc_umc_status_addr;
 	uint64_t mc_umc_status, err_addr;
-	uint64_t mc_umc_addrt0, na_err_addr_base;
-	uint64_t na_err_addr, retired_page_addr;
-	uint32_t channel_index, addr_lsb, col = 0;
-	int ret = 0;
+	uint64_t mc_umc_addrt0;
 
 	mc_umc_status_addr =
 		SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0);
@@ -236,12 +272,6 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev,
 		return;
 	}
 
-	channel_index =
-		adev->umc.channel_idx_tbl[node_inst * adev->umc.umc_inst_num *
-					adev->umc.channel_inst_num +
-					umc_inst * adev->umc.channel_inst_num +
-					ch_inst];
-
 	/* calculate error address if ue error is detected */
 	if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 &&
 	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrV) == 1 &&
@@ -251,27 +281,8 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev,
 		err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4);
 		err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr);
 
-		/* the lowest lsb bits should be ignored */
-		addr_lsb = REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrLsb);
-		err_addr &= ~((0x1ULL << addr_lsb) - 1);
-		na_err_addr_base = err_addr & ~(0x3ULL << UMC_V8_10_NA_C5_BIT);
-
-		/* loop for all possibilities of [C6 C5] in normal address. */
-		for (col = 0; col < UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM; col++) {
-			na_err_addr = na_err_addr_base | (col << UMC_V8_10_NA_C5_BIT);
-
-			/* Mapping normal error address to retired soc physical address. */
-			ret = umc_v8_10_swizzle_mode_na_to_pa(adev, channel_index,
-							na_err_addr, &retired_page_addr);
-			if (ret) {
-				dev_err(adev->dev, "Failed to map pa from umc na.\n");
-				break;
-			}
-			dev_info(adev->dev, "Error Address(PA): 0x%llx\n",
-				retired_page_addr);
-			amdgpu_umc_fill_error_record(err_data, na_err_addr,
-					retired_page_addr, channel_index, umc_inst);
-		}
+		umc_v8_10_convert_error_address(adev, err_data, err_addr,
+					ch_inst, umc_inst, node_inst, mc_umc_status);
 	}
 
 	/* clear umc status */
@@ -349,6 +360,133 @@ static bool umc_v8_10_query_ras_poison_mode(struct amdgpu_device *adev)
 	return true;
 }
 
+static void umc_v8_10_ecc_info_query_correctable_error_count(struct amdgpu_device *adev,
+				      uint32_t node_inst, uint32_t umc_inst, uint32_t ch_inst,
+				      unsigned long *error_count)
+{
+	uint64_t mc_umc_status;
+	uint32_t eccinfo_table_idx;
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	eccinfo_table_idx = node_inst * adev->umc.umc_inst_num *
+				  adev->umc.channel_inst_num +
+				  umc_inst * adev->umc.channel_inst_num +
+				  ch_inst;
+
+	/* check the MCUMC_STATUS */
+	mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status;
+	if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 &&
+	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1) {
+		*error_count += 1;
+	}
+}
+
+static void umc_v8_10_ecc_info_query_uncorrectable_error_count(struct amdgpu_device *adev,
+				      uint32_t node_inst, uint32_t umc_inst, uint32_t ch_inst,
+				      unsigned long *error_count)
+{
+	uint64_t mc_umc_status;
+	uint32_t eccinfo_table_idx;
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	eccinfo_table_idx = node_inst * adev->umc.umc_inst_num *
+				  adev->umc.channel_inst_num +
+				  umc_inst * adev->umc.channel_inst_num +
+				  ch_inst;
+
+	/* check the MCUMC_STATUS */
+	mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status;
+	if ((REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1) &&
+	    (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Deferred) == 1 ||
+	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1 ||
+	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, PCC) == 1 ||
+	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UC) == 1 ||
+	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, TCC) == 1)) {
+		*error_count += 1;
+	}
+}
+
+static void umc_v8_10_ecc_info_query_ras_error_count(struct amdgpu_device *adev,
+					void *ras_error_status)
+{
+	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
+
+	uint32_t node_inst       = 0;
+	uint32_t umc_inst        = 0;
+	uint32_t ch_inst         = 0;
+
+	/* TODO: driver needs to toggle DF Cstate to ensure
+	 * safe access of UMC registers. Will add the protection
+	 */
+	LOOP_UMC_EACH_NODE_INST_AND_CH(node_inst, umc_inst, ch_inst) {
+		umc_v8_10_ecc_info_query_correctable_error_count(adev,
+							node_inst, umc_inst, ch_inst,
+							&(err_data->ce_count));
+		umc_v8_10_ecc_info_query_uncorrectable_error_count(adev,
+							node_inst, umc_inst, ch_inst,
+							&(err_data->ue_count));
+	}
+}
+
+static void umc_v8_10_ecc_info_query_error_address(struct amdgpu_device *adev,
+					struct ras_err_data *err_data,
+					uint32_t ch_inst,
+					uint32_t umc_inst,
+					uint32_t node_inst)
+{
+	uint32_t eccinfo_table_idx;
+	uint64_t mc_umc_status, err_addr;
+
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	eccinfo_table_idx = node_inst * adev->umc.umc_inst_num *
+				  adev->umc.channel_inst_num +
+				  umc_inst * adev->umc.channel_inst_num +
+				  ch_inst;
+
+	mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status;
+
+	if (mc_umc_status == 0)
+		return;
+
+	if (!err_data->err_addr)
+		return;
+
+	/* calculate error address if ue error is detected */
+	if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 &&
+	    REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrV) == 1 &&
+	    (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1)) {
+
+		err_addr = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_addr;
+		err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr);
+
+		umc_v8_10_convert_error_address(adev, err_data, err_addr,
+					ch_inst, umc_inst, node_inst, mc_umc_status);
+	}
+}
+
+static void umc_v8_10_ecc_info_query_ras_error_address(struct amdgpu_device *adev,
+					void *ras_error_status)
+{
+	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
+
+	uint32_t node_inst       = 0;
+	uint32_t umc_inst        = 0;
+	uint32_t ch_inst         = 0;
+
+	/* TODO: driver needs to toggle DF Cstate to ensure
+	 * safe access of UMC resgisters. Will add the protection
+	 * when firmware interface is ready
+	 */
+	LOOP_UMC_EACH_NODE_INST_AND_CH(node_inst, umc_inst, ch_inst) {
+		umc_v8_10_ecc_info_query_error_address(adev,
+						err_data,
+						ch_inst,
+						umc_inst,
+						node_inst);
+	}
+}
+
 const struct amdgpu_ras_block_hw_ops umc_v8_10_ras_hw_ops = {
 	.query_ras_error_count = umc_v8_10_query_ras_error_count,
 	.query_ras_error_address = umc_v8_10_query_ras_error_address,
@@ -360,4 +498,6 @@ struct amdgpu_umc_ras umc_v8_10_ras = {
 	},
 	.err_cnt_init = umc_v8_10_err_cnt_init,
 	.query_ras_poison_mode = umc_v8_10_query_ras_poison_mode,
+	.ecc_info_query_ras_error_count = umc_v8_10_ecc_info_query_ras_error_count,
+	.ecc_info_query_ras_error_address = umc_v8_10_ecc_info_query_ras_error_address,
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c
index 22a41766a8c7..43d587404c3e 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c
@@ -78,9 +78,17 @@ static void vcn_v4_0_set_ras_funcs(struct amdgpu_device *adev);
 static int vcn_v4_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int i;
 
-	if (amdgpu_sriov_vf(adev))
+	if (amdgpu_sriov_vf(adev)) {
 		adev->vcn.harvest_config = VCN_HARVEST_MMSCH;
+		for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
+			if (amdgpu_vcn_is_disabled_vcn(adev, VCN_ENCODE_RING, i)) {
+				adev->vcn.harvest_config |= 1 << i;
+				dev_info(adev->dev, "VCN%d is disabled by hypervisor\n", i);
+			}
+		}
+	}
 
 	/* re-use enc ring as unified ring */
 	adev->vcn.num_enc_rings = 1;
@@ -238,16 +246,11 @@ static int vcn_v4_0_hw_init(void *handle)
 				continue;
 
 			ring = &adev->vcn.inst[i].ring_enc[0];
-			if (amdgpu_vcn_is_disabled_vcn(adev, VCN_ENCODE_RING, i)) {
-				ring->sched.ready = false;
-				ring->no_scheduler = true;
-				dev_info(adev->dev, "ring %s is disabled by hypervisor\n", ring->name);
-			} else {
-				ring->wptr = 0;
-				ring->wptr_old = 0;
-				vcn_v4_0_unified_ring_set_wptr(ring);
-				ring->sched.ready = true;
-			}
+			ring->wptr = 0;
+			ring->wptr_old = 0;
+			vcn_v4_0_unified_ring_set_wptr(ring);
+			ring->sched.ready = true;
+
 		}
 	} else {
 		for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index c06ada0844ba..7a95698d83f7 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -2373,7 +2373,7 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev)
 	if (init_mqd_managers(dqm))
 		goto out_free;
 
-	if (allocate_hiq_sdma_mqd(dqm)) {
+	if (!dev->shared_resources.enable_mes && allocate_hiq_sdma_mqd(dqm)) {
 		pr_err("Failed to allocate hiq sdma mqd trunk buffer\n");
 		goto out_free;
 	}
@@ -2397,7 +2397,8 @@ static void deallocate_hiq_sdma_mqd(struct kfd_dev *dev,
 void device_queue_manager_uninit(struct device_queue_manager *dqm)
 {
 	dqm->ops.uninitialize(dqm);
-	deallocate_hiq_sdma_mqd(dqm->dev, &dqm->hiq_sdma_mqd);
+	if (!dqm->dev->shared_resources.enable_mes)
+		deallocate_hiq_sdma_mqd(dqm->dev, &dqm->hiq_sdma_mqd);
 	kfree(dqm);
 }
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index 729d26d648af..2880ed96ac2e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -778,16 +778,13 @@ static struct kfd_event_waiter *alloc_event_waiters(uint32_t num_events)
 	struct kfd_event_waiter *event_waiters;
 	uint32_t i;
 
-	event_waiters = kmalloc_array(num_events,
-					sizeof(struct kfd_event_waiter),
-					GFP_KERNEL);
+	event_waiters = kcalloc(num_events, sizeof(struct kfd_event_waiter),
+				GFP_KERNEL);
 	if (!event_waiters)
 		return NULL;
 
-	for (i = 0; (event_waiters) && (i < num_events) ; i++) {
+	for (i = 0; i < num_events; i++)
 		init_wait(&event_waiters[i].wait);
-		event_waiters[i].activated = false;
-	}
 
 	return event_waiters;
 }
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c
index 4f6390f3236e..4a9af800b1f1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c
@@ -308,11 +308,16 @@ static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
 		struct queue_properties *q)
 {
 	struct v11_sdma_mqd *m;
+	int size;
 
 	m = (struct v11_sdma_mqd *) mqd_mem_obj->cpu_ptr;
 
-	memset(m, 0, sizeof(struct v11_sdma_mqd));
+	if (mm->dev->shared_resources.enable_mes)
+		size = PAGE_SIZE;
+	else
+		size = sizeof(struct v11_sdma_mqd);
 
+	memset(m, 0, size);
 	*mqd = m;
 	if (gart_addr)
 		*gart_addr = mqd_mem_obj->gpu_addr;
@@ -443,6 +448,14 @@ struct mqd_manager *mqd_manager_init_v11(enum KFD_MQD_TYPE type,
 #if defined(CONFIG_DEBUG_FS)
 		mqd->debugfs_show_mqd = debugfs_show_mqd_sdma;
 #endif
+		/*
+		 * To allocate SDMA MQDs by generic functions
+		 * when MES is enabled.
+		 */
+		if (dev->shared_resources.enable_mes) {
+			mqd->allocate_mqd = allocate_mqd;
+			mqd->free_mqd = kfd_free_mqd_cp;
+		}
 		pr_debug("%s@%i\n", __func__, __LINE__);
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index 2efe93f74f84..0c9bd0a53e60 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -28,7 +28,6 @@ config DRM_AMD_DC_DCN
 config DRM_AMD_DC_HDCP
 	bool "Enable HDCP support in DC"
 	depends on DRM_AMD_DC
-	select DRM_DISPLAY_HDCP_HELPER
 	help
 	  Choose this option if you want to support HDCP authentication.
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index c420bce47acb..009ef917dad4 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -41,6 +41,8 @@
 #include "dpcd_defs.h"
 #include "link/protocols/link_dpcd.h"
 #include "link_service_types.h"
+#include "link/protocols/link_dp_capability.h"
+#include "link/protocols/link_ddc.h"
 
 #include "vid.h"
 #include "amdgpu.h"
@@ -2302,6 +2304,14 @@ static void s3_handle_mst(struct drm_device *dev, bool suspend)
 		if (suspend) {
 			drm_dp_mst_topology_mgr_suspend(mgr);
 		} else {
+			/* if extended timeout is supported in hardware,
+			 * default to LTTPR timeout (3.2ms) first as a W/A for DP link layer
+			 * CTS 4.2.1.1 regression introduced by CTS specs requirement update.
+			 */
+			try_to_configure_aux_timeout(aconnector->dc_link->ddc, LINK_AUX_DEFAULT_LTTPR_TIMEOUT_PERIOD);
+			if (!dp_is_lttpr_present(aconnector->dc_link))
+				try_to_configure_aux_timeout(aconnector->dc_link->ddc, LINK_AUX_DEFAULT_TIMEOUT_PERIOD);
+
 			ret = drm_dp_mst_topology_mgr_resume(mgr, true);
 			if (ret < 0) {
 				dm_helpers_dp_mst_stop_top_mgr(aconnector->dc_link->ctx,
@@ -4265,6 +4275,8 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 	/* Update the actual used number of crtc */
 	adev->mode_info.num_crtc = adev->dm.display_indexes_num;
 
+	amdgpu_dm_set_irq_funcs(adev);
+
 	link_cnt = dm->dc->caps.max_links;
 	if (amdgpu_dm_mode_config_init(dm->adev)) {
 		DRM_ERROR("DM: Failed to initialize mode config\n");
@@ -4757,8 +4769,6 @@ static int dm_early_init(void *handle)
 		break;
 	}
 
-	amdgpu_dm_set_irq_funcs(adev);
-
 	if (adev->mode_info.funcs == NULL)
 		adev->mode_info.funcs = &dm_display_funcs;
 
@@ -7235,7 +7245,7 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 		drm_connector_attach_max_bpc_property(&aconnector->base, 8, 16);
 
 	/* This defaults to the max in the range, but we want 8bpc for non-edp. */
-	aconnector->base.state->max_bpc = (connector_type == DRM_MODE_CONNECTOR_eDP) ? 16 : 8;
+	aconnector->base.state->max_bpc = 16;
 	aconnector->base.state->max_requested_bpc = aconnector->base.state->max_bpc;
 
 	if (connector_type == DRM_MODE_CONNECTOR_eDP &&
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index 6fdc2027c2b4..1583157da355 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -1149,6 +1149,8 @@ static bool dm_is_freesync_pcon_whitelist(const uint32_t branch_dev_id)
 
 	switch (branch_dev_id) {
 	case DP_BRANCH_DEVICE_ID_0060AD:
+	case DP_BRANCH_DEVICE_ID_00E04C:
+	case DP_BRANCH_DEVICE_ID_90CC24:
 		ret_val = true;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index fb3fd5b7c78b..0d4d3d586166 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -779,10 +779,8 @@ void dce110_edp_wait_for_hpd_ready(
 
 	dal_gpio_destroy_irq(&hpd);
 
-	if (false == edp_hpd_high) {
-		DC_LOG_WARNING(
-				"%s: wait timed out!\n", __func__);
-	}
+	/* ensure that the panel is detected */
+	ASSERT(edp_hpd_high);
 }
 
 void dce110_edp_power_control(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
index df787fcf8e86..3b4d4d68359b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c
@@ -998,5 +998,8 @@ void dcn30_prepare_bandwidth(struct dc *dc,
 			dc->clk_mgr->funcs->set_max_memclk(dc->clk_mgr, dc->clk_mgr->bw_params->clk_table.entries[dc->clk_mgr->bw_params->clk_table.num_entries - 1].memclk_mhz);
 
 	dcn20_prepare_bandwidth(dc, context);
+
+	dc_dmub_srv_p_state_delegate(dc,
+		context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching, context);
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c
index 379729b02847..c3d75e56410c 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c
@@ -1802,7 +1802,10 @@ static unsigned int CalculateVMAndRowBytes(
 	}
 
 	if (SurfaceTiling == dm_sw_linear) {
-		*dpte_row_height = dml_min(128, 1 << (unsigned int) dml_floor(dml_log2(PTEBufferSizeInRequests * *PixelPTEReqWidth / Pitch), 1));
+		if (PTEBufferSizeInRequests == 0)
+			*dpte_row_height = 1;
+		else
+			*dpte_row_height = dml_min(128, 1 << (unsigned int) dml_floor(dml_log2(PTEBufferSizeInRequests * *PixelPTEReqWidth / Pitch), 1));
 		*dpte_row_width_ub = (dml_ceil(((double) SwathWidth - 1) / *PixelPTEReqWidth, 1) + 1) * *PixelPTEReqWidth;
 		*PixelPTEBytesPerRow = *dpte_row_width_ub / *PixelPTEReqWidth * *PTERequestSize;
 	} else if (ScanDirection != dm_vert) {
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.h
index 86e9d2e886d6..aaa5064408ba 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.h
@@ -33,6 +33,7 @@
 #define DPVGA_DONGLE_AUX_DEFER_WA_DELAY 40
 #define I2C_OVER_AUX_DEFER_WA_DELAY_1MS 1
 #define LINK_AUX_DEFAULT_LTTPR_TIMEOUT_PERIOD 3200 /*us*/
+#define LINK_AUX_DEFAULT_TIMEOUT_PERIOD 552 /*us*/
 
 #define EDID_SEGMENT_SIZE 256
 
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
index 4874d1bf1dcb..d4370856f164 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
@@ -60,8 +60,6 @@
 #define MIN(X, Y) ((X) < (Y) ? (X) : (Y))
 #endif
 
-#define LINK_AUX_DEFAULT_TIMEOUT_PERIOD 552 /*us*/
-
 struct dp_lt_fallback_entry {
 	enum dc_lane_count lane_count;
 	enum dc_link_rate link_rate;
diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
index 834d146c4991..0652b001ad54 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
@@ -1202,10 +1202,17 @@ static int smu_smc_hw_setup(struct smu_context *smu)
 		return ret;
 	}
 
-	ret = smu_setup_pptable(smu);
-	if (ret) {
-		dev_err(adev->dev, "Failed to setup pptable!\n");
-		return ret;
+	/*
+	 * It is assumed the pptable used before runpm is same as
+	 * the one used afterwards. Thus, we can reuse the stored
+	 * copy and do not need to resetup the pptable again.
+	 */
+	if (!adev->in_runpm) {
+		ret = smu_setup_pptable(smu);
+		if (ret) {
+			dev_err(adev->dev, "Failed to setup pptable!\n");
+			return ret;
+		}
 	}
 
 	/* smu_dump_pptable(smu); */
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c
index 6492d69e2e60..e1ef88ee1ed3 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c
@@ -256,7 +256,7 @@ int smu_v11_0_check_fw_version(struct smu_context *smu)
 	 * to be backward compatible.
 	 * 2. New fw usually brings some optimizations. But that's visible
 	 * only on the paired driver.
-	 * Considering above, we just leave user a warning message instead
+	 * Considering above, we just leave user a verbal message instead
 	 * of halt driver loading.
 	 */
 	if (if_version != smu->smc_driver_if_version) {
@@ -264,7 +264,7 @@ int smu_v11_0_check_fw_version(struct smu_context *smu)
 			"smu fw program = %d, version = 0x%08x (%d.%d.%d)\n",
 			smu->smc_driver_if_version, if_version,
 			smu_program, smu_version, smu_major, smu_minor, smu_debug);
-		dev_warn(smu->adev->dev, "SMU driver if version not matched\n");
+		dev_info(smu->adev->dev, "SMU driver if version not matched\n");
 	}
 
 	return ret;
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c
index 56a02bc60cee..c788aa7a99a9 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu12/smu_v12_0.c
@@ -93,7 +93,7 @@ int smu_v12_0_check_fw_version(struct smu_context *smu)
 	 * to be backward compatible.
 	 * 2. New fw usually brings some optimizations. But that's visible
 	 * only on the paired driver.
-	 * Considering above, we just leave user a warning message instead
+	 * Considering above, we just leave user a verbal message instead
 	 * of halt driver loading.
 	 */
 	if (if_version != smu->smc_driver_if_version) {
@@ -101,7 +101,7 @@ int smu_v12_0_check_fw_version(struct smu_context *smu)
 			"smu fw program = %d, smu fw version = 0x%08x (%d.%d.%d)\n",
 			smu->smc_driver_if_version, if_version,
 			smu_program, smu_version, smu_major, smu_minor, smu_debug);
-		dev_warn(smu->adev->dev, "SMU driver if version not matched\n");
+		dev_info(smu->adev->dev, "SMU driver if version not matched\n");
 	}
 
 	return ret;
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c
index 78945e79dbee..a52ed0580fd7 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c
@@ -311,7 +311,7 @@ int smu_v13_0_check_fw_version(struct smu_context *smu)
 	 * to be backward compatible.
 	 * 2. New fw usually brings some optimizations. But that's visible
 	 * only on the paired driver.
-	 * Considering above, we just leave user a warning message instead
+	 * Considering above, we just leave user a verbal message instead
 	 * of halt driver loading.
 	 */
 	if (if_version != smu->smc_driver_if_version) {
@@ -319,7 +319,7 @@ int smu_v13_0_check_fw_version(struct smu_context *smu)
 			 "smu fw program = %d, smu fw version = 0x%08x (%d.%d.%d)\n",
 			 smu->smc_driver_if_version, if_version,
 			 smu_program, smu_version, smu_major, smu_minor, smu_debug);
-		dev_warn(adev->dev, "SMU driver if version not matched\n");
+		dev_info(adev->dev, "SMU driver if version not matched\n");
 	}
 
 	return ret;
@@ -2229,10 +2229,23 @@ int smu_v13_0_gfx_ulv_control(struct smu_context *smu,
 int smu_v13_0_baco_set_armd3_sequence(struct smu_context *smu,
 				      enum smu_baco_seq baco_seq)
 {
-	return smu_cmn_send_smc_msg_with_param(smu,
-					       SMU_MSG_ArmD3,
-					       baco_seq,
-					       NULL);
+	struct smu_baco_context *smu_baco = &smu->smu_baco;
+	int ret;
+
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+					      SMU_MSG_ArmD3,
+					      baco_seq,
+					      NULL);
+	if (ret)
+		return ret;
+
+	if (baco_seq == BACO_SEQ_BAMACO ||
+	    baco_seq == BACO_SEQ_BACO)
+		smu_baco->state = SMU_BACO_STATE_ENTER;
+	else
+		smu_baco->state = SMU_BACO_STATE_EXIT;
+
+	return 0;
 }
 
 bool smu_v13_0_baco_is_support(struct smu_context *smu)
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
index 7c906ab3ddd2..923a9fb3c887 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
@@ -147,6 +147,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_0_message_map[SMU_MSG_MAX_COUNT] =
 			    PPSMC_MSG_SetBadMemoryPagesRetiredFlagsPerChannel,   0),
 	MSG_MAP(AllowGpo,			PPSMC_MSG_SetGpoAllow,           0),
 	MSG_MAP(AllowIHHostInterrupt,		PPSMC_MSG_AllowIHHostInterrupt,       0),
+	MSG_MAP(ReenableAcDcInterrupt,		PPSMC_MSG_ReenableAcDcInterrupt,       0),
 };
 
 static struct cmn2asic_mapping smu_v13_0_0_clk_map[SMU_CLK_COUNT] = {
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c
index 1471c3a96602..4aca09cab4b8 100644
--- a/drivers/gpu/drm/radeon/atombios_encoders.c
+++ b/drivers/gpu/drm/radeon/atombios_encoders.c
@@ -2123,11 +2123,12 @@ int radeon_atom_pick_dig_encoder(struct drm_encoder *encoder, int fe_idx)
 
 	/*
 	 * On DCE32 any encoder can drive any block so usually just use crtc id,
-	 * but Apple thinks different at least on iMac10,1, so there use linkb,
+	 * but Apple thinks different at least on iMac10,1 and iMac11,2, so there use linkb,
 	 * otherwise the internal eDP panel will stay dark.
 	 */
 	if (ASIC_IS_DCE32(rdev)) {
-		if (dmi_match(DMI_PRODUCT_NAME, "iMac10,1"))
+		if (dmi_match(DMI_PRODUCT_NAME, "iMac10,1") ||
+		    dmi_match(DMI_PRODUCT_NAME, "iMac11,2"))
 			enc_idx = (dig->linkb) ? 1 : 0;
 		else
 			enc_idx = radeon_crtc->crtc_id;