Merge tag 'amd-drm-next-5.16-2021-10-29' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-5.16-2021-10-29: amdgpu: - RAS fixes - Fix a potential memory leak in device tear down - Add a stutter mode quirk - Misc display fixes - Further display FP refactoring - Display USB4 fixes - Display DP2.0 fixes - DCN 3.1 fixes - Display 8 ch audio fix - Fix DMA mask regression for SI parts - Aldebaran fixes amdkfd: - userptr fix - BO lifetime fix - Misc code cleanup UAPI: - Minor header cleanup (no functional change) Signed-off-by: Dave Airlie <airlied@redhat.com> From: Alex Deucher <alexander.deucher@amd.com> Link: https://patchwork.freedesktop.org/patch/msgid/20211029184338.4863-1-alexander.deucher@amd.com
author: Dave Airlie <airlied@redhat.com> 2021-11-02 05:40:57 +0300
committer: Dave Airlie <airlied@redhat.com> 2021-11-02 05:40:58 +0300
commit: d9bd054177fbd2c4762546aec40fc3071bfe4cc0 (patch)
tree: 70e34d106c87a7a7fefb810213786a0b8a19b69c
parent: 31fa8cbce4664946a1688898410fee41ad05364d (diff)
parent: 403475be6d8b122c3e6b8a47e075926d7299e5ef (diff)
download: linux-d9bd054177fbd2c4762546aec40fc3071bfe4cc0.tar.xz
89 files changed, 2054 insertions, 1334 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index d58e37fd01f4..b85b67a88a3d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -205,6 +205,7 @@ extern struct amdgpu_mgpu_info mgpu_info;
 extern int amdgpu_ras_enable;
 extern uint amdgpu_ras_mask;
 extern int amdgpu_bad_page_threshold;
+extern bool amdgpu_ignore_bad_page_threshold;
 extern struct amdgpu_watchdog_timer amdgpu_watchdog_timer;
 extern int amdgpu_async_gfx_ring;
 extern int amdgpu_mcbp;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index 69de31754907..751557af09bb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -279,6 +279,8 @@ int amdgpu_amdkfd_gpuvm_sync_memory(
 		struct kgd_dev *kgd, struct kgd_mem *mem, bool intr);
 int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd,
 		struct kgd_mem *mem, void **kptr, uint64_t *size);
+void amdgpu_amdkfd_gpuvm_unmap_gtt_bo_from_kernel(struct kgd_dev *kgd, struct kgd_mem *mem);
+
 int amdgpu_amdkfd_gpuvm_restore_process_bos(void *process_info,
 					    struct dma_fence **ef);
 int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct kgd_dev *kgd,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index cdf46bd0d8d5..0e9cfe99ae9e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -1871,6 +1871,16 @@ bo_reserve_failed:
 	return ret;
 }
 
+void amdgpu_amdkfd_gpuvm_unmap_gtt_bo_from_kernel(struct kgd_dev *kgd, struct kgd_mem *mem)
+{
+	struct amdgpu_bo *bo = mem->bo;
+
+	amdgpu_bo_reserve(bo, true);
+	amdgpu_bo_kunmap(bo);
+	amdgpu_bo_unpin(bo);
+	amdgpu_bo_unreserve(bo);
+}
+
 int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct kgd_dev *kgd,
 					      struct kfd_vm_fault_info *mem)
 {
@@ -2041,19 +2051,26 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
 		/* Get updated user pages */
 		ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages);
 		if (ret) {
-			pr_debug("%s: Failed to get user pages: %d\n",
-				__func__, ret);
+			pr_debug("Failed %d to get user pages\n", ret);
+
+			/* Return -EFAULT bad address error as success. It will
+			 * fail later with a VM fault if the GPU tries to access
+			 * it. Better than hanging indefinitely with stalled
+			 * user mode queues.
+			 *
+			 * Return other error -EBUSY or -ENOMEM to retry restore
+			 */
+			if (ret != -EFAULT)
+				return ret;
+		} else {
 
-			/* Return error -EBUSY or -ENOMEM, retry restore */
-			return ret;
+			/*
+			 * FIXME: Cannot ignore the return code, must hold
+			 * notifier_lock
+			 */
+			amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
 		}
 
-		/*
-		 * FIXME: Cannot ignore the return code, must hold
-		 * notifier_lock
-		 */
-		amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
-
 		/* Mark the BO as valid unless it was invalidated
 		 * again concurrently.
 		 */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index c218d53a031d..6e40cc1bc6dc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -3852,8 +3852,8 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev)
 
 void amdgpu_device_fini_sw(struct amdgpu_device *adev)
 {
-	amdgpu_device_ip_fini(adev);
 	amdgpu_fence_driver_sw_fini(adev);
+	amdgpu_device_ip_fini(adev);
 	release_firmware(adev->firmware.gpu_info_fw);
 	adev->firmware.gpu_info_fw = NULL;
 	adev->accel_working = false;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
index 208a784475bd..d7c8d9e3c203 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
@@ -931,6 +931,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[OSSSYS_HWIP][0] = IP_VERSION(4, 0, 0);
 		adev->ip_versions[HDP_HWIP][0] = IP_VERSION(4, 0, 0);
 		adev->ip_versions[SDMA0_HWIP][0] = IP_VERSION(4, 0, 0);
+		adev->ip_versions[SDMA1_HWIP][0] = IP_VERSION(4, 0, 0);
 		adev->ip_versions[DF_HWIP][0] = IP_VERSION(2, 1, 0);
 		adev->ip_versions[NBIO_HWIP][0] = IP_VERSION(6, 1, 0);
 		adev->ip_versions[UMC_HWIP][0] = IP_VERSION(6, 0, 0);
@@ -951,6 +952,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[OSSSYS_HWIP][0] = IP_VERSION(4, 0, 1);
 		adev->ip_versions[HDP_HWIP][0] = IP_VERSION(4, 0, 1);
 		adev->ip_versions[SDMA0_HWIP][0] = IP_VERSION(4, 0, 1);
+		adev->ip_versions[SDMA1_HWIP][0] = IP_VERSION(4, 0, 1);
 		adev->ip_versions[DF_HWIP][0] = IP_VERSION(2, 5, 0);
 		adev->ip_versions[NBIO_HWIP][0] = IP_VERSION(6, 2, 0);
 		adev->ip_versions[UMC_HWIP][0] = IP_VERSION(6, 1, 0);
@@ -1009,6 +1011,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[OSSSYS_HWIP][0] = IP_VERSION(4, 2, 0);
 		adev->ip_versions[HDP_HWIP][0] = IP_VERSION(4, 2, 0);
 		adev->ip_versions[SDMA0_HWIP][0] = IP_VERSION(4, 2, 0);
+		adev->ip_versions[SDMA1_HWIP][0] = IP_VERSION(4, 2, 0);
 		adev->ip_versions[DF_HWIP][0] = IP_VERSION(3, 6, 0);
 		adev->ip_versions[NBIO_HWIP][0] = IP_VERSION(7, 4, 0);
 		adev->ip_versions[UMC_HWIP][0] = IP_VERSION(6, 1, 1);
@@ -1018,6 +1021,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[SMUIO_HWIP][0] = IP_VERSION(11, 0, 2);
 		adev->ip_versions[GC_HWIP][0] = IP_VERSION(9, 4, 0);
 		adev->ip_versions[UVD_HWIP][0] = IP_VERSION(7, 2, 0);
+		adev->ip_versions[UVD_HWIP][1] = IP_VERSION(7, 2, 0);
 		adev->ip_versions[VCE_HWIP][0] = IP_VERSION(4, 1, 0);
 		adev->ip_versions[DCI_HWIP][0] = IP_VERSION(12, 1, 0);
 		break;
@@ -1030,6 +1034,13 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[OSSSYS_HWIP][0] = IP_VERSION(4, 2, 1);
 		adev->ip_versions[HDP_HWIP][0] = IP_VERSION(4, 2, 1);
 		adev->ip_versions[SDMA0_HWIP][0] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][0] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][1] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][2] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][3] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][4] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][5] = IP_VERSION(4, 2, 2);
+		adev->ip_versions[SDMA1_HWIP][6] = IP_VERSION(4, 2, 2);
 		adev->ip_versions[DF_HWIP][0] = IP_VERSION(3, 6, 1);
 		adev->ip_versions[NBIO_HWIP][0] = IP_VERSION(7, 4, 1);
 		adev->ip_versions[UMC_HWIP][0] = IP_VERSION(6, 1, 2);
@@ -1039,6 +1050,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[SMUIO_HWIP][0] = IP_VERSION(11, 0, 3);
 		adev->ip_versions[GC_HWIP][0] = IP_VERSION(9, 4, 1);
 		adev->ip_versions[UVD_HWIP][0] = IP_VERSION(2, 5, 0);
+		adev->ip_versions[UVD_HWIP][1] = IP_VERSION(2, 5, 0);
 		break;
 	case CHIP_ALDEBARAN:
 		aldebaran_reg_base_init(adev);
@@ -1049,6 +1061,10 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[OSSSYS_HWIP][0] = IP_VERSION(4, 4, 0);
 		adev->ip_versions[HDP_HWIP][0] = IP_VERSION(4, 4, 0);
 		adev->ip_versions[SDMA0_HWIP][0] = IP_VERSION(4, 4, 0);
+		adev->ip_versions[SDMA0_HWIP][1] = IP_VERSION(4, 4, 0);
+		adev->ip_versions[SDMA0_HWIP][2] = IP_VERSION(4, 4, 0);
+		adev->ip_versions[SDMA0_HWIP][3] = IP_VERSION(4, 4, 0);
+		adev->ip_versions[SDMA0_HWIP][4] = IP_VERSION(4, 4, 0);
 		adev->ip_versions[DF_HWIP][0] = IP_VERSION(3, 6, 2);
 		adev->ip_versions[NBIO_HWIP][0] = IP_VERSION(7, 4, 4);
 		adev->ip_versions[UMC_HWIP][0] = IP_VERSION(6, 7, 0);
@@ -1058,6 +1074,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[SMUIO_HWIP][0] = IP_VERSION(13, 0, 2);
 		adev->ip_versions[GC_HWIP][0] = IP_VERSION(9, 4, 2);
 		adev->ip_versions[UVD_HWIP][0] = IP_VERSION(2, 6, 0);
+		adev->ip_versions[UVD_HWIP][1] = IP_VERSION(2, 6, 0);
 		adev->ip_versions[XGMI_HWIP][0] = IP_VERSION(6, 1, 0);
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index c718fb5f3f8a..4cefe867ccf2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -877,7 +877,7 @@ 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)");
+MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = auto(default value), 0 = disable bad page retirement, -2 = ignore bad page 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+)");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index d2955ea4a62b..dfe667ea8b05 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -340,32 +340,32 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 	case AMDGPU_INFO_FW_TA:
 		switch (query_fw->index) {
 		case TA_FW_TYPE_PSP_XGMI:
-			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->ver = adev->psp.xgmi_context.context.bin_desc.fw_version;
 			fw_info->feature = adev->psp.xgmi_context.context
 						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_RAS:
-			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->ver = adev->psp.ras_context.context.bin_desc.fw_version;
 			fw_info->feature = adev->psp.ras_context.context
 						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_HDCP:
-			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->ver = adev->psp.hdcp_context.context.bin_desc.fw_version;
 			fw_info->feature = adev->psp.hdcp_context.context
 						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_DTM:
-			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->ver = adev->psp.dtm_context.context.bin_desc.fw_version;
 			fw_info->feature = adev->psp.dtm_context.context
 						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_RAP:
-			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->ver = adev->psp.rap_context.context.bin_desc.fw_version;
 			fw_info->feature = adev->psp.rap_context.context
 						   .bin_desc.feature_version;
 			break;
 		case TA_FW_TYPE_PSP_SECUREDISPLAY:
-			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->ver = adev->psp.securedisplay_context.context.bin_desc.fw_version;
 			fw_info->feature =
 				adev->psp.securedisplay_context.context.bin_desc
 					.feature_version;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index fd04e83031d6..c641f84649d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -1115,7 +1115,7 @@ int psp_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id)
 static bool psp_xgmi_peer_link_info_supported(struct psp_context *psp)
 {
 	return psp->adev->ip_versions[MP0_HWIP][0] == IP_VERSION(13, 0, 2) &&
-		psp->xgmi_context.context.bin_desc.feature_version >= 0x2000000b;
+		psp->xgmi_context.context.bin_desc.fw_version >= 0x2000000b;
 }
 
 /*
@@ -3108,32 +3108,32 @@ static int parse_ta_bin_descriptor(struct psp_context *psp,
 		psp->asd_context.bin_desc.start_addr        = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_XGMI:
-		psp->xgmi_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->xgmi_context.context.bin_desc.fw_version       = le32_to_cpu(desc->fw_version);
 		psp->xgmi_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
 		psp->xgmi_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_RAS:
-		psp->ras_context.context.bin_desc.feature_version   = le32_to_cpu(desc->fw_version);
+		psp->ras_context.context.bin_desc.fw_version        = le32_to_cpu(desc->fw_version);
 		psp->ras_context.context.bin_desc.size_bytes        = le32_to_cpu(desc->size_bytes);
 		psp->ras_context.context.bin_desc.start_addr        = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_HDCP:
-		psp->hdcp_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->hdcp_context.context.bin_desc.fw_version       = le32_to_cpu(desc->fw_version);
 		psp->hdcp_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
 		psp->hdcp_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_DTM:
-		psp->dtm_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->dtm_context.context.bin_desc.fw_version       = le32_to_cpu(desc->fw_version);
 		psp->dtm_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
 		psp->dtm_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_RAP:
-		psp->rap_context.context.bin_desc.feature_version  = le32_to_cpu(desc->fw_version);
+		psp->rap_context.context.bin_desc.fw_version       = le32_to_cpu(desc->fw_version);
 		psp->rap_context.context.bin_desc.size_bytes       = le32_to_cpu(desc->size_bytes);
 		psp->rap_context.context.bin_desc.start_addr       = ucode_start_addr;
 		break;
 	case TA_FW_TYPE_PSP_SECUREDISPLAY:
-		psp->securedisplay_context.context.bin_desc.feature_version =
+		psp->securedisplay_context.context.bin_desc.fw_version =
 			le32_to_cpu(desc->fw_version);
 		psp->securedisplay_context.context.bin_desc.size_bytes =
 			le32_to_cpu(desc->size_bytes);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index f4c05ff4b26c..05117eda105b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -1077,6 +1077,13 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
 		if (res)
 			DRM_ERROR("RAS table incorrect checksum or error:%d\n",
 				  res);
+
+		/* Warn if we are at 90% of the threshold or above
+		 */
+		if (10 * control->ras_num_recs >= 9 * ras->bad_page_cnt_threshold)
+			dev_warn(adev->dev, "RAS records:%u exceeds 90%% of threshold:%d",
+					control->ras_num_recs,
+					ras->bad_page_cnt_threshold);
 	} else if (hdr->header == RAS_TABLE_HDR_BAD &&
 		   amdgpu_bad_page_threshold != 0) {
 		res = __verify_ras_table_checksum(control);
@@ -1098,11 +1105,18 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
 			res = amdgpu_ras_eeprom_correct_header_tag(control,
 								   RAS_TABLE_HDR_VAL);
 		} else {
-			*exceed_err_limit = true;
-			dev_err(adev->dev,
-				"RAS records:%d exceed threshold:%d, "
-				"GPU will not be initialized. Replace this GPU or increase the threshold",
+			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.");
+				res = 0;
+			} else {
+				*exceed_err_limit = true;
+				dev_err(adev->dev,
+					"RAS records:%d exceed threshold:%d, "
+					"GPU will not be initialized. Replace this GPU or increase the threshold",
+					control->ras_num_recs, ras->bad_page_cnt_threshold);
+			}
 		}
 	} else {
 		DRM_INFO("Creating a new EEPROM table");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 590537b62a0a..17b5dd6adf8c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -696,6 +696,9 @@ int amdgpu_ttm_tt_get_user_pages(struct amdgpu_bo *bo, struct page **pages)
 				       true, NULL);
 out_unlock:
 	mmap_read_unlock(mm);
+	if (r)
+		pr_debug("failed %d to get user pages 0x%lx\n", r, start);
+
 	mmput(mm);
 
 	return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index 920c4f6ac914..ca3350502618 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -527,8 +527,8 @@ FW_VERSION_ATTR(mec_fw_version, 0444, gfx.mec_fw_version);
 FW_VERSION_ATTR(mec2_fw_version, 0444, gfx.mec2_fw_version);
 FW_VERSION_ATTR(sos_fw_version, 0444, psp.sos.fw_version);
 FW_VERSION_ATTR(asd_fw_version, 0444, psp.asd_context.bin_desc.fw_version);
-FW_VERSION_ATTR(ta_ras_fw_version, 0444, psp.ras_context.context.bin_desc.feature_version);
-FW_VERSION_ATTR(ta_xgmi_fw_version, 0444, psp.xgmi_context.context.bin_desc.feature_version);
+FW_VERSION_ATTR(ta_ras_fw_version, 0444, psp.ras_context.context.bin_desc.fw_version);
+FW_VERSION_ATTR(ta_xgmi_fw_version, 0444, psp.xgmi_context.context.bin_desc.fw_version);
 FW_VERSION_ATTR(smc_fw_version, 0444, pm.fw_version);
 FW_VERSION_ATTR(sdma_fw_version, 0444, sdma.instance[0].fw_version);
 FW_VERSION_ATTR(sdma2_fw_version, 0444, sdma.instance[1].fw_version);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
index 99c149397aae..04cf9b207e62 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -535,9 +535,9 @@ static void amdgpu_virt_populate_vf2pf_ucode_info(struct amdgpu_device *adev)
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_ASD,
 			    adev->psp.asd_context.bin_desc.fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_RAS,
-			    adev->psp.ras_context.context.bin_desc.feature_version);
+			    adev->psp.ras_context.context.bin_desc.fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_XGMI,
-			    adev->psp.xgmi_context.context.bin_desc.feature_version);
+			    adev->psp.xgmi_context.context.bin_desc.fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SMC,      adev->pm.fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SDMA,     adev->sdma.instance[0].fw_version);
 	POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SDMA2,    adev->sdma.instance[1].fw_version);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c
index 00a2b36a24b3..c4f37a161875 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c
@@ -706,6 +706,11 @@ int gfx_v9_4_2_do_edc_gpr_workarounds(struct amdgpu_device *adev)
 	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX))
 		return 0;
 
+	/* Workaround for ALDEBARAN, skip GPRs init in GPU reset.
+	   Will remove it once GPRs init algorithm works for all CU settings. */
+	if (amdgpu_in_reset(adev))
+		return 0;
+
 	gfx_v9_4_2_do_sgprs_init(adev);
 
 	gfx_v9_4_2_do_vgprs_init(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 0e81e03e9b49..0fe714f54cca 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -841,12 +841,12 @@ static int gmc_v6_0_sw_init(void *handle)
 
 	adev->gmc.mc_mask = 0xffffffffffULL;
 
-	r = dma_set_mask_and_coherent(adev->dev, DMA_BIT_MASK(44));
+	r = dma_set_mask_and_coherent(adev->dev, DMA_BIT_MASK(40));
 	if (r) {
 		dev_warn(adev->dev, "No suitable DMA available.\n");
 		return r;
 	}
-	adev->need_swiotlb = drm_need_swiotlb(44);
+	adev->need_swiotlb = drm_need_swiotlb(40);
 
 	r = gmc_v6_0_init_microcode(adev);
 	if (r) {
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
index 59644015dfc3..ed2293686f0d 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
@@ -84,7 +84,7 @@ static int psp_v10_0_init_microcode(struct psp_context *psp)
 
 		ta_hdr = (const struct ta_firmware_header_v1_0 *)
 				 adev->psp.ta_fw->data;
-		adev->psp.hdcp_context.context.bin_desc.feature_version =
+		adev->psp.hdcp_context.context.bin_desc.fw_version =
 			le32_to_cpu(ta_hdr->hdcp.fw_version);
 		adev->psp.hdcp_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->hdcp.size_bytes);
@@ -92,7 +92,7 @@ static int psp_v10_0_init_microcode(struct psp_context *psp)
 			(uint8_t *)ta_hdr +
 			le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
 
-		adev->psp.dtm_context.context.bin_desc.feature_version =
+		adev->psp.dtm_context.context.bin_desc.fw_version =
 			le32_to_cpu(ta_hdr->dtm.fw_version);
 		adev->psp.dtm_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->dtm.size_bytes);
@@ -100,7 +100,7 @@ static int psp_v10_0_init_microcode(struct psp_context *psp)
 			(uint8_t *)adev->psp.hdcp_context.context.bin_desc.start_addr +
 			le32_to_cpu(ta_hdr->dtm.offset_bytes);
 
-		adev->psp.securedisplay_context.context.bin_desc.feature_version =
+		adev->psp.securedisplay_context.context.bin_desc.fw_version =
 			le32_to_cpu(ta_hdr->securedisplay.fw_version);
 		adev->psp.securedisplay_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->securedisplay.size_bytes);
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
index 21a325ea49cb..2176ef85f137 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
@@ -151,7 +151,7 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 				goto out2;
 
 			ta_hdr = (const struct ta_firmware_header_v1_0 *)adev->psp.ta_fw->data;
-			adev->psp.xgmi_context.context.bin_desc.feature_version =
+			adev->psp.xgmi_context.context.bin_desc.fw_version =
 				le32_to_cpu(ta_hdr->xgmi.fw_version);
 			adev->psp.xgmi_context.context.bin_desc.size_bytes =
 				le32_to_cpu(ta_hdr->xgmi.size_bytes);
@@ -159,7 +159,7 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 				(uint8_t *)ta_hdr +
 				le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes);
 			adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
-			adev->psp.ras_context.context.bin_desc.feature_version =
+			adev->psp.ras_context.context.bin_desc.fw_version =
 				le32_to_cpu(ta_hdr->ras.fw_version);
 			adev->psp.ras_context.context.bin_desc.size_bytes =
 				le32_to_cpu(ta_hdr->ras.size_bytes);
@@ -192,7 +192,7 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 				goto out2;
 
 			ta_hdr = (const struct ta_firmware_header_v1_0 *)adev->psp.ta_fw->data;
-			adev->psp.hdcp_context.context.bin_desc.feature_version =
+			adev->psp.hdcp_context.context.bin_desc.fw_version =
 				le32_to_cpu(ta_hdr->hdcp.fw_version);
 			adev->psp.hdcp_context.context.bin_desc.size_bytes =
 				le32_to_cpu(ta_hdr->hdcp.size_bytes);
@@ -203,7 +203,7 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 
 			adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
 
-			adev->psp.dtm_context.context.bin_desc.feature_version =
+			adev->psp.dtm_context.context.bin_desc.fw_version =
 				le32_to_cpu(ta_hdr->dtm.fw_version);
 			adev->psp.dtm_context.context.bin_desc.size_bytes =
 				le32_to_cpu(ta_hdr->dtm.size_bytes);
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
index 281bc4d7f0a1..a2588200ea58 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c
@@ -84,7 +84,7 @@ static int psp_v12_0_init_microcode(struct psp_context *psp)
 
 		ta_hdr = (const struct ta_firmware_header_v1_0 *)
 				 adev->psp.ta_fw->data;
-		adev->psp.hdcp_context.context.bin_desc.feature_version =
+		adev->psp.hdcp_context.context.bin_desc.fw_version =
 			le32_to_cpu(ta_hdr->hdcp.fw_version);
 		adev->psp.hdcp_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->hdcp.size_bytes);
@@ -94,7 +94,7 @@ static int psp_v12_0_init_microcode(struct psp_context *psp)
 
 		adev->psp.ta_fw_version = le32_to_cpu(ta_hdr->header.ucode_version);
 
-		adev->psp.dtm_context.context.bin_desc.feature_version =
+		adev->psp.dtm_context.context.bin_desc.fw_version =
 			le32_to_cpu(ta_hdr->dtm.fw_version);
 		adev->psp.dtm_context.context.bin_desc.size_bytes =
 			le32_to_cpu(ta_hdr->dtm.size_bytes);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index f1e7edeb4e6b..24ebd61395d8 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -405,7 +405,7 @@ static int kfd_ioctl_update_queue(struct file *filp, struct kfd_process *p,
 
 	mutex_lock(&p->mutex);
 
-	retval = pqm_update_queue(&p->pqm, args->queue_id, &properties);
+	retval = pqm_update_queue_properties(&p->pqm, args->queue_id, &properties);
 
 	mutex_unlock(&p->mutex);
 
@@ -418,7 +418,7 @@ static int kfd_ioctl_set_cu_mask(struct file *filp, struct kfd_process *p,
 	int retval;
 	const int max_num_cus = 1024;
 	struct kfd_ioctl_set_cu_mask_args *args = data;
-	struct queue_properties properties;
+	struct mqd_update_info minfo = {0};
 	uint32_t __user *cu_mask_ptr = (uint32_t __user *)args->cu_mask_ptr;
 	size_t cu_mask_size = sizeof(uint32_t) * (args->num_cu_mask / 32);
 
@@ -428,8 +428,8 @@ static int kfd_ioctl_set_cu_mask(struct file *filp, struct kfd_process *p,
 		return -EINVAL;
 	}
 
-	properties.cu_mask_count = args->num_cu_mask;
-	if (properties.cu_mask_count == 0) {
+	minfo.cu_mask.count = args->num_cu_mask;
+	if (minfo.cu_mask.count == 0) {
 		pr_debug("CU mask cannot be 0");
 		return -EINVAL;
 	}
@@ -438,32 +438,33 @@ static int kfd_ioctl_set_cu_mask(struct file *filp, struct kfd_process *p,
 	 * limit of max_num_cus bits.  We can then just drop any CU mask bits
 	 * past max_num_cus bits and just use the first max_num_cus bits.
 	 */
-	if (properties.cu_mask_count > max_num_cus) {
+	if (minfo.cu_mask.count > max_num_cus) {
 		pr_debug("CU mask cannot be greater than 1024 bits");
-		properties.cu_mask_count = max_num_cus;
+		minfo.cu_mask.count = max_num_cus;
 		cu_mask_size = sizeof(uint32_t) * (max_num_cus/32);
 	}
 
-	properties.cu_mask = kzalloc(cu_mask_size, GFP_KERNEL);
-	if (!properties.cu_mask)
+	minfo.cu_mask.ptr = kzalloc(cu_mask_size, GFP_KERNEL);
+	if (!minfo.cu_mask.ptr)
 		return -ENOMEM;
 
-	retval = copy_from_user(properties.cu_mask, cu_mask_ptr, cu_mask_size);
+	retval = copy_from_user(minfo.cu_mask.ptr, cu_mask_ptr, cu_mask_size);
 	if (retval) {
 		pr_debug("Could not copy CU mask from userspace");
-		kfree(properties.cu_mask);
-		return -EFAULT;
+		retval = -EFAULT;
+		goto out;
 	}
 
+	minfo.update_flag = UPDATE_FLAG_CU_MASK;
+
 	mutex_lock(&p->mutex);
 
-	retval = pqm_set_cu_mask(&p->pqm, args->queue_id, &properties);
+	retval = pqm_update_mqd(&p->pqm, args->queue_id, &minfo);
 
 	mutex_unlock(&p->mutex);
 
-	if (retval)
-		kfree(properties.cu_mask);
-
+out:
+	kfree(minfo.cu_mask.ptr);
 	return retval;
 }
 
@@ -1011,11 +1012,6 @@ static int kfd_ioctl_create_event(struct file *filp, struct kfd_process *p,
 		void *mem, *kern_addr;
 		uint64_t size;
 
-		if (p->signal_page) {
-			pr_err("Event page is already set\n");
-			return -EINVAL;
-		}
-
 		kfd = kfd_device_by_id(GET_GPU_ID(args->event_page_offset));
 		if (!kfd) {
 			pr_err("Getting device by id failed in %s\n", __func__);
@@ -1023,6 +1019,13 @@ static int kfd_ioctl_create_event(struct file *filp, struct kfd_process *p,
 		}
 
 		mutex_lock(&p->mutex);
+
+		if (p->signal_page) {
+			pr_err("Event page is already set\n");
+			err = -EINVAL;
+			goto out_unlock;
+		}
+
 		pdd = kfd_bind_process_to_device(kfd, p);
 		if (IS_ERR(pdd)) {
 			err = PTR_ERR(pdd);
@@ -1037,20 +1040,24 @@ static int kfd_ioctl_create_event(struct file *filp, struct kfd_process *p,
 			err = -EINVAL;
 			goto out_unlock;
 		}
-		mutex_unlock(&p->mutex);
 
 		err = amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(kfd->kgd,
 						mem, &kern_addr, &size);
 		if (err) {
 			pr_err("Failed to map event page to kernel\n");
-			return err;
+			goto out_unlock;
 		}
 
 		err = kfd_event_page_set(p, kern_addr, size);
 		if (err) {
 			pr_err("Failed to set event page\n");
-			return err;
+			amdgpu_amdkfd_gpuvm_unmap_gtt_bo_from_kernel(kfd->kgd, mem);
+			goto out_unlock;
 		}
+
+		p->signal_handle = args->event_page_offset;
+
+		mutex_unlock(&p->mutex);
 	}
 
 	err = kfd_event_create(filp, p, args->event_type,
@@ -1368,6 +1375,15 @@ static int kfd_ioctl_free_memory_of_gpu(struct file *filep,
 		return -EINVAL;
 
 	mutex_lock(&p->mutex);
+	/*
+	 * Safeguard to prevent user space from freeing signal BO.
+	 * It will be freed at process termination.
+	 */
+	if (p->signal_handle && (p->signal_handle == args->handle)) {
+		pr_err("Free signal BO is not allowed\n");
+		ret = -EPERM;
+		goto err_unlock;
+	}
 
 	pdd = kfd_get_process_device_data(dev, p);
 	if (!pdd) {
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 f8fce9d05f50..533b27b35fc9 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -557,7 +557,8 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm,
 	return retval;
 }
 
-static int update_queue(struct device_queue_manager *dqm, struct queue *q)
+static int update_queue(struct device_queue_manager *dqm, struct queue *q,
+			struct mqd_update_info *minfo)
 {
 	int retval = 0;
 	struct mqd_manager *mqd_mgr;
@@ -605,7 +606,7 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q)
 		}
 	}
 
-	mqd_mgr->update_mqd(mqd_mgr, q->mqd, &q->properties);
+	mqd_mgr->update_mqd(mqd_mgr, q->mqd, &q->properties, minfo);
 
 	/*
 	 * check active state vs. the previous state and modify
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
index c8719682c4da..499fc0ea387f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -93,7 +93,7 @@ struct device_queue_manager_ops {
 				struct queue *q);
 
 	int	(*update_queue)(struct device_queue_manager *dqm,
-				struct queue *q);
+				struct queue *q, struct mqd_update_info *minfo);
 
 	int	(*register_process)(struct device_queue_manager *dqm,
 					struct qcm_process_device *qpd);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
index a2b77d1df854..64b4ac339904 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -136,7 +136,6 @@ static bool kq_initialize(struct kernel_queue *kq, struct kfd_dev *dev,
 	prop.write_ptr = (uint32_t *) kq->wptr_gpu_addr;
 	prop.eop_ring_buffer_address = kq->eop_gpu_addr;
 	prop.eop_ring_buffer_size = PAGE_SIZE;
-	prop.cu_mask = NULL;
 
 	if (init_queue(&kq->queue, &prop) != 0)
 		goto err_init_queue;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h
index 6e6918ccedfd..965e17c5dbb4 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h
@@ -80,7 +80,8 @@ struct mqd_manager {
 				struct mm_struct *mms);
 
 	void	(*update_mqd)(struct mqd_manager *mm, void *mqd,
-				struct queue_properties *q);
+				struct queue_properties *q,
+				struct mqd_update_info *minfo);
 
 	int	(*destroy_mqd)(struct mqd_manager *mm, void *mqd,
 				enum kfd_preempt_type type,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
index 064914e1e8d6..8128f4d312f1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
@@ -42,16 +42,17 @@ static inline struct cik_sdma_rlc_registers *get_sdma_mqd(void *mqd)
 }
 
 static void update_cu_mask(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct mqd_update_info *minfo)
 {
 	struct cik_mqd *m;
 	uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */
 
-	if (q->cu_mask_count == 0)
+	if (!minfo || (minfo->update_flag != UPDATE_FLAG_CU_MASK) ||
+	    !minfo->cu_mask.ptr)
 		return;
 
 	mqd_symmetrically_map_cu_mask(mm,
-		q->cu_mask, q->cu_mask_count, se_mask);
+		minfo->cu_mask.ptr, minfo->cu_mask.count, se_mask);
 
 	m = get_mqd(mqd);
 	m->compute_static_thread_mgmt_se0 = se_mask[0];
@@ -135,7 +136,7 @@ static void init_mqd(struct mqd_manager *mm, void **mqd,
 	*mqd = m;
 	if (gart_addr)
 		*gart_addr = addr;
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
@@ -152,7 +153,7 @@ static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
 	if (gart_addr)
 		*gart_addr = mqd_mem_obj->gpu_addr;
 
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static void free_mqd(struct mqd_manager *mm, void *mqd,
@@ -185,7 +186,8 @@ static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
 }
 
 static void __update_mqd(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q, unsigned int atc_bit)
+			struct queue_properties *q, struct mqd_update_info *minfo,
+			unsigned int atc_bit)
 {
 	struct cik_mqd *m;
 
@@ -214,16 +216,17 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd,
 	if (q->format == KFD_QUEUE_FORMAT_AQL)
 		m->cp_hqd_pq_control |= NO_UPDATE_RPTR;
 
-	update_cu_mask(mm, mqd, q);
+	update_cu_mask(mm, mqd, minfo);
 	set_priority(m, q);
 
 	q->is_active = QUEUE_IS_ACTIVE(*q);
 }
 
 static void update_mqd(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
-	__update_mqd(mm, mqd, q, 1);
+	__update_mqd(mm, mqd, q, minfo, 1);
 }
 
 static uint32_t read_doorbell_id(void *mqd)
@@ -234,13 +237,15 @@ static uint32_t read_doorbell_id(void *mqd)
 }
 
 static void update_mqd_hawaii(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
-	__update_mqd(mm, mqd, q, 0);
+	__update_mqd(mm, mqd, q, minfo, 0);
 }
 
 static void update_mqd_sdma(struct mqd_manager *mm, void *mqd,
-				struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct cik_sdma_rlc_registers *m;
 
@@ -318,7 +323,8 @@ static void init_mqd_hiq(struct mqd_manager *mm, void **mqd,
 }
 
 static void update_mqd_hiq(struct mqd_manager *mm, void *mqd,
-				struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct cik_mqd *m;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
index c7fb59ca597f..270160fc401b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
@@ -42,16 +42,17 @@ static inline struct v10_sdma_mqd *get_sdma_mqd(void *mqd)
 }
 
 static void update_cu_mask(struct mqd_manager *mm, void *mqd,
-			   struct queue_properties *q)
+			struct mqd_update_info *minfo)
 {
 	struct v10_compute_mqd *m;
 	uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */
 
-	if (q->cu_mask_count == 0)
+	if (!minfo || (minfo->update_flag != UPDATE_FLAG_CU_MASK) ||
+	    !minfo->cu_mask.ptr)
 		return;
 
 	mqd_symmetrically_map_cu_mask(mm,
-		q->cu_mask, q->cu_mask_count, se_mask);
+		minfo->cu_mask.ptr, minfo->cu_mask.count, se_mask);
 
 	m = get_mqd(mqd);
 	m->compute_static_thread_mgmt_se0 = se_mask[0];
@@ -136,7 +137,7 @@ static void init_mqd(struct mqd_manager *mm, void **mqd,
 	*mqd = m;
 	if (gart_addr)
 		*gart_addr = addr;
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static int load_mqd(struct mqd_manager *mm, void *mqd,
@@ -162,7 +163,8 @@ static int hiq_load_mqd_kiq(struct mqd_manager *mm, void *mqd,
 }
 
 static void update_mqd(struct mqd_manager *mm, void *mqd,
-		      struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct v10_compute_mqd *m;
 
@@ -218,7 +220,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd,
 	if (mm->dev->cwsr_enabled)
 		m->cp_hqd_ctx_save_control = 0;
 
-	update_cu_mask(mm, mqd, q);
+	update_cu_mask(mm, mqd, minfo);
 	set_priority(m, q);
 
 	q->is_active = QUEUE_IS_ACTIVE(*q);
@@ -311,7 +313,7 @@ static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
 	if (gart_addr)
 		*gart_addr = mqd_mem_obj->gpu_addr;
 
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
@@ -326,7 +328,8 @@ static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
 #define SDMA_RLC_DUMMY_DEFAULT 0xf
 
 static void update_mqd_sdma(struct mqd_manager *mm, void *mqd,
-		struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct v10_sdma_mqd *m;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
index 7f4e102ff4bd..4e5932f54b5a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
@@ -43,16 +43,17 @@ static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd)
 }
 
 static void update_cu_mask(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct mqd_update_info *minfo)
 {
 	struct v9_mqd *m;
 	uint32_t se_mask[KFD_MAX_NUM_SE] = {0};
 
-	if (q->cu_mask_count == 0)
+	if (!minfo || (minfo->update_flag != UPDATE_FLAG_CU_MASK) ||
+	    !minfo->cu_mask.ptr)
 		return;
 
 	mqd_symmetrically_map_cu_mask(mm,
-		q->cu_mask, q->cu_mask_count, se_mask);
+		minfo->cu_mask.ptr, minfo->cu_mask.count, se_mask);
 
 	m = get_mqd(mqd);
 	m->compute_static_thread_mgmt_se0 = se_mask[0];
@@ -188,7 +189,7 @@ static void init_mqd(struct mqd_manager *mm, void **mqd,
 	*mqd = m;
 	if (gart_addr)
 		*gart_addr = addr;
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static int load_mqd(struct mqd_manager *mm, void *mqd,
@@ -212,7 +213,8 @@ static int hiq_load_mqd_kiq(struct mqd_manager *mm, void *mqd,
 }
 
 static void update_mqd(struct mqd_manager *mm, void *mqd,
-		      struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct v9_mqd *m;
 
@@ -269,7 +271,7 @@ static void update_mqd(struct mqd_manager *mm, void *mqd,
 	if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address)
 		m->cp_hqd_ctx_save_control = 0;
 
-	update_cu_mask(mm, mqd, q);
+	update_cu_mask(mm, mqd, minfo);
 	set_priority(m, q);
 
 	q->is_active = QUEUE_IS_ACTIVE(*q);
@@ -366,7 +368,7 @@ static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
 	if (gart_addr)
 		*gart_addr = mqd_mem_obj->gpu_addr;
 
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
@@ -381,7 +383,8 @@ static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
 #define SDMA_RLC_DUMMY_DEFAULT 0xf
 
 static void update_mqd_sdma(struct mqd_manager *mm, void *mqd,
-		struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct v9_sdma_mqd *m;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
index 33dbd22d290f..cd9220eb8a7a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
@@ -45,16 +45,17 @@ static inline struct vi_sdma_mqd *get_sdma_mqd(void *mqd)
 }
 
 static void update_cu_mask(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct mqd_update_info *minfo)
 {
 	struct vi_mqd *m;
 	uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */
 
-	if (q->cu_mask_count == 0)
+	if (!minfo || (minfo->update_flag != UPDATE_FLAG_CU_MASK) ||
+	    !minfo->cu_mask.ptr)
 		return;
 
 	mqd_symmetrically_map_cu_mask(mm,
-		q->cu_mask, q->cu_mask_count, se_mask);
+		minfo->cu_mask.ptr, minfo->cu_mask.count, se_mask);
 
 	m = get_mqd(mqd);
 	m->compute_static_thread_mgmt_se0 = se_mask[0];
@@ -150,7 +151,7 @@ static void init_mqd(struct mqd_manager *mm, void **mqd,
 	*mqd = m;
 	if (gart_addr)
 		*gart_addr = addr;
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static int load_mqd(struct mqd_manager *mm, void *mqd,
@@ -167,8 +168,8 @@ static int load_mqd(struct mqd_manager *mm, void *mqd,
 }
 
 static void __update_mqd(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q, unsigned int mtype,
-			unsigned int atc_bit)
+			struct queue_properties *q, struct mqd_update_info *minfo,
+			unsigned int mtype, unsigned int atc_bit)
 {
 	struct vi_mqd *m;
 
@@ -230,7 +231,7 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd,
 			atc_bit << CP_HQD_CTX_SAVE_CONTROL__ATC__SHIFT |
 			mtype << CP_HQD_CTX_SAVE_CONTROL__MTYPE__SHIFT;
 
-	update_cu_mask(mm, mqd, q);
+	update_cu_mask(mm, mqd, minfo);
 	set_priority(m, q);
 
 	q->is_active = QUEUE_IS_ACTIVE(*q);
@@ -238,9 +239,10 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd,
 
 
 static void update_mqd(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
-	__update_mqd(mm, mqd, q, MTYPE_CC, 1);
+	__update_mqd(mm, mqd, q, minfo, MTYPE_CC, 1);
 }
 
 static uint32_t read_doorbell_id(void *mqd)
@@ -251,9 +253,10 @@ static uint32_t read_doorbell_id(void *mqd)
 }
 
 static void update_mqd_tonga(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
-	__update_mqd(mm, mqd, q, MTYPE_UC, 0);
+	__update_mqd(mm, mqd, q, minfo, MTYPE_UC, 0);
 }
 
 static int destroy_mqd(struct mqd_manager *mm, void *mqd,
@@ -317,9 +320,10 @@ static void init_mqd_hiq(struct mqd_manager *mm, void **mqd,
 }
 
 static void update_mqd_hiq(struct mqd_manager *mm, void *mqd,
-			struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
-	__update_mqd(mm, mqd, q, MTYPE_UC, 0);
+	__update_mqd(mm, mqd, q, minfo, MTYPE_UC, 0);
 }
 
 static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
@@ -336,7 +340,7 @@ static void init_mqd_sdma(struct mqd_manager *mm, void **mqd,
 	if (gart_addr)
 		*gart_addr = mqd_mem_obj->gpu_addr;
 
-	mm->update_mqd(mm, m, q);
+	mm->update_mqd(mm, m, q, NULL);
 }
 
 static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
@@ -349,7 +353,8 @@ static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
 }
 
 static void update_mqd_sdma(struct mqd_manager *mm, void *mqd,
-		struct queue_properties *q)
+			struct queue_properties *q,
+			struct mqd_update_info *minfo)
 {
 	struct vi_sdma_mqd *m;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 6d8f9bb2d905..4104b167e721 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -472,9 +472,6 @@ struct queue_properties {
 	uint32_t ctl_stack_size;
 	uint64_t tba_addr;
 	uint64_t tma_addr;
-	/* Relevant for CU */
-	uint32_t cu_mask_count; /* Must be a multiple of 32 */
-	uint32_t *cu_mask;
 };
 
 #define QUEUE_IS_ACTIVE(q) ((q).queue_size > 0 &&	\
@@ -482,6 +479,20 @@ struct queue_properties {
 			    (q).queue_percent > 0 &&	\
 			    !(q).is_evicted)
 
+enum mqd_update_flag {
+	UPDATE_FLAG_CU_MASK = 0,
+};
+
+struct mqd_update_info {
+	union {
+		struct {
+			uint32_t count; /* Must be a multiple of 32 */
+			uint32_t *ptr;
+		} cu_mask;
+	};
+	enum mqd_update_flag update_flag;
+};
+
 /**
  * struct queue
  *
@@ -608,12 +619,14 @@ struct qcm_process_device {
 	uint32_t sh_hidden_private_base;
 
 	/* CWSR memory */
+	struct kgd_mem *cwsr_mem;
 	void *cwsr_kaddr;
 	uint64_t cwsr_base;
 	uint64_t tba_addr;
 	uint64_t tma_addr;
 
 	/* IB memory */
+	struct kgd_mem *ib_mem;
 	uint64_t ib_base;
 	void *ib_kaddr;
 
@@ -808,6 +821,7 @@ struct kfd_process {
 	/* Event ID allocator and lookup */
 	struct idr event_idr;
 	/* Event page */
+	u64 signal_handle;
 	struct kfd_signal_page *signal_page;
 	size_t signal_mapped_size;
 	size_t signal_event_count;
@@ -1031,10 +1045,10 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 			    unsigned int *qid,
 			    uint32_t *p_doorbell_offset_in_process);
 int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid);
-int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid,
-			struct queue_properties *p);
-int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid,
+int pqm_update_queue_properties(struct process_queue_manager *pqm, unsigned int qid,
 			struct queue_properties *p);
+int pqm_update_mqd(struct process_queue_manager *pqm, unsigned int qid,
+			struct mqd_update_info *minfo);
 int pqm_set_gws(struct process_queue_manager *pqm, unsigned int qid,
 			void *gws);
 struct kernel_queue *pqm_get_kernel_queue(struct process_queue_manager *pqm,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 21ec8a18cad2..457863861d6f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -72,6 +72,8 @@ static int kfd_process_init_cwsr_apu(struct kfd_process *p, struct file *filep);
 static void evict_process_worker(struct work_struct *work);
 static void restore_process_worker(struct work_struct *work);
 
+static void kfd_process_device_destroy_cwsr_dgpu(struct kfd_process_device *pdd);
+
 struct kfd_procfs_tree {
 	struct kobject *kobj;
 };
@@ -685,10 +687,15 @@ void kfd_process_destroy_wq(void)
 }
 
 static void kfd_process_free_gpuvm(struct kgd_mem *mem,
-			struct kfd_process_device *pdd)
+			struct kfd_process_device *pdd, void *kptr)
 {
 	struct kfd_dev *dev = pdd->dev;
 
+	if (kptr) {
+		amdgpu_amdkfd_gpuvm_unmap_gtt_bo_from_kernel(dev->kgd, mem);
+		kptr = NULL;
+	}
+
 	amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu(dev->kgd, mem, pdd->drm_priv);
 	amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, mem, pdd->drm_priv,
 					       NULL);
@@ -702,63 +709,46 @@ static void kfd_process_free_gpuvm(struct kgd_mem *mem,
  */
 static int kfd_process_alloc_gpuvm(struct kfd_process_device *pdd,
 				   uint64_t gpu_va, uint32_t size,
-				   uint32_t flags, void **kptr)
+				   uint32_t flags, struct kgd_mem **mem, void **kptr)
 {
 	struct kfd_dev *kdev = pdd->dev;
-	struct kgd_mem *mem = NULL;
-	int handle;
 	int err;
 
 	err = amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(kdev->kgd, gpu_va, size,
-						 pdd->drm_priv, &mem, NULL, flags);
+						 pdd->drm_priv, mem, NULL, flags);
 	if (err)
 		goto err_alloc_mem;
 
-	err = amdgpu_amdkfd_gpuvm_map_memory_to_gpu(kdev->kgd, mem,
+	err = amdgpu_amdkfd_gpuvm_map_memory_to_gpu(kdev->kgd, *mem,
 			pdd->drm_priv, NULL);
 	if (err)
 		goto err_map_mem;
 
-	err = amdgpu_amdkfd_gpuvm_sync_memory(kdev->kgd, mem, true);
+	err = amdgpu_amdkfd_gpuvm_sync_memory(kdev->kgd, *mem, true);
 	if (err) {
 		pr_debug("Sync memory failed, wait interrupted by user signal\n");
 		goto sync_memory_failed;
 	}
 
-	/* Create an obj handle so kfd_process_device_remove_obj_handle
-	 * will take care of the bo removal when the process finishes.
-	 * We do not need to take p->mutex, because the process is just
-	 * created and the ioctls have not had the chance to run.
-	 */
-	handle = kfd_process_device_create_obj_handle(pdd, mem);
-
-	if (handle < 0) {
-		err = handle;
-		goto free_gpuvm;
-	}
-
 	if (kptr) {
 		err = amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(kdev->kgd,
-				(struct kgd_mem *)mem, kptr, NULL);
+				(struct kgd_mem *)*mem, kptr, NULL);
 		if (err) {
 			pr_debug("Map GTT BO to kernel failed\n");
-			goto free_obj_handle;
+			goto sync_memory_failed;
 		}
 	}
 
 	return err;
 
-free_obj_handle:
-	kfd_process_device_remove_obj_handle(pdd, handle);
-free_gpuvm:
 sync_memory_failed:
-	kfd_process_free_gpuvm(mem, pdd);
-	return err;
+	amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu(kdev->kgd, *mem, pdd->drm_priv);
 
 err_map_mem:
-	amdgpu_amdkfd_gpuvm_free_memory_of_gpu(kdev->kgd, mem, pdd->drm_priv,
+	amdgpu_amdkfd_gpuvm_free_memory_of_gpu(kdev->kgd, *mem, pdd->drm_priv,
 					       NULL);
 err_alloc_mem:
+	*mem = NULL;
 	*kptr = NULL;
 	return err;
 }
@@ -776,6 +766,7 @@ static int kfd_process_device_reserve_ib_mem(struct kfd_process_device *pdd)
 			KFD_IOC_ALLOC_MEM_FLAGS_NO_SUBSTITUTE |
 			KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE |
 			KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE;
+	struct kgd_mem *mem;
 	void *kaddr;
 	int ret;
 
@@ -784,15 +775,26 @@ static int kfd_process_device_reserve_ib_mem(struct kfd_process_device *pdd)
 
 	/* ib_base is only set for dGPU */
 	ret = kfd_process_alloc_gpuvm(pdd, qpd->ib_base, PAGE_SIZE, flags,
-				      &kaddr);
+				      &mem, &kaddr);
 	if (ret)
 		return ret;
 
+	qpd->ib_mem = mem;
 	qpd->ib_kaddr = kaddr;
 
 	return 0;
 }
 
+static void kfd_process_device_destroy_ib_mem(struct kfd_process_device *pdd)
+{
+	struct qcm_process_device *qpd = &pdd->qpd;
+
+	if (!qpd->ib_kaddr || !qpd->ib_base)
+		return;
+
+	kfd_process_free_gpuvm(qpd->ib_mem, pdd, qpd->ib_kaddr);
+}
+
 struct kfd_process *kfd_create_process(struct file *filep)
 {
 	struct kfd_process *process;
@@ -947,6 +949,37 @@ static void kfd_process_device_free_bos(struct kfd_process_device *pdd)
 	}
 }
 
+/*
+ * Just kunmap and unpin signal BO here. It will be freed in
+ * kfd_process_free_outstanding_kfd_bos()
+ */
+static void kfd_process_kunmap_signal_bo(struct kfd_process *p)
+{
+	struct kfd_process_device *pdd;
+	struct kfd_dev *kdev;
+	void *mem;
+
+	kdev = kfd_device_by_id(GET_GPU_ID(p->signal_handle));
+	if (!kdev)
+		return;
+
+	mutex_lock(&p->mutex);
+
+	pdd = kfd_get_process_device_data(kdev, p);
+	if (!pdd)
+		goto out;
+
+	mem = kfd_process_device_translate_handle(
+		pdd, GET_IDR_HANDLE(p->signal_handle));
+	if (!mem)
+		goto out;
+
+	amdgpu_amdkfd_gpuvm_unmap_gtt_bo_from_kernel(kdev->kgd, mem);
+
+out:
+	mutex_unlock(&p->mutex);
+}
+
 static void kfd_process_free_outstanding_kfd_bos(struct kfd_process *p)
 {
 	int i;
@@ -965,6 +998,9 @@ static void kfd_process_destroy_pdds(struct kfd_process *p)
 		pr_debug("Releasing pdd (topology id %d) for process (pasid 0x%x)\n",
 				pdd->dev->id, p->pasid);
 
+		kfd_process_device_destroy_cwsr_dgpu(pdd);
+		kfd_process_device_destroy_ib_mem(pdd);
+
 		if (pdd->drm_file) {
 			amdgpu_amdkfd_gpuvm_release_process_vm(
 					pdd->dev->kgd, pdd->drm_priv);
@@ -1049,9 +1085,11 @@ static void kfd_process_wq_release(struct work_struct *work)
 {
 	struct kfd_process *p = container_of(work, struct kfd_process,
 					     release_work);
+
 	kfd_process_remove_sysfs(p);
 	kfd_iommu_unbind_process(p);
 
+	kfd_process_kunmap_signal_bo(p);
 	kfd_process_free_outstanding_kfd_bos(p);
 	svm_range_list_fini(p);
 
@@ -1198,6 +1236,7 @@ static int kfd_process_device_init_cwsr_dgpu(struct kfd_process_device *pdd)
 	uint32_t flags = KFD_IOC_ALLOC_MEM_FLAGS_GTT
 			| KFD_IOC_ALLOC_MEM_FLAGS_NO_SUBSTITUTE
 			| KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE;
+	struct kgd_mem *mem;
 	void *kaddr;
 	int ret;
 
@@ -1206,10 +1245,11 @@ static int kfd_process_device_init_cwsr_dgpu(struct kfd_process_device *pdd)
 
 	/* cwsr_base is only set for dGPU */
 	ret = kfd_process_alloc_gpuvm(pdd, qpd->cwsr_base,
-				      KFD_CWSR_TBA_TMA_SIZE, flags, &kaddr);
+				      KFD_CWSR_TBA_TMA_SIZE, flags, &mem, &kaddr);
 	if (ret)
 		return ret;
 
+	qpd->cwsr_mem = mem;
 	qpd->cwsr_kaddr = kaddr;
 	qpd->tba_addr = qpd->cwsr_base;
 
@@ -1222,6 +1262,17 @@ static int kfd_process_device_init_cwsr_dgpu(struct kfd_process_device *pdd)
 	return 0;
 }
 
+static void kfd_process_device_destroy_cwsr_dgpu(struct kfd_process_device *pdd)
+{
+	struct kfd_dev *dev = pdd->dev;
+	struct qcm_process_device *qpd = &pdd->qpd;
+
+	if (!dev->cwsr_enabled || !qpd->cwsr_kaddr || !qpd->cwsr_base)
+		return;
+
+	kfd_process_free_gpuvm(qpd->cwsr_mem, pdd, qpd->cwsr_kaddr);
+}
+
 void kfd_process_set_trap_handler(struct qcm_process_device *qpd,
 				  uint64_t tba_addr,
 				  uint64_t tma_addr)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 243dd1efcdbf..3627e7ac161b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -121,7 +121,7 @@ int pqm_set_gws(struct process_queue_manager *pqm, unsigned int qid,
 	pdd->qpd.num_gws = gws ? amdgpu_amdkfd_get_num_gws(dev->kgd) : 0;
 
 	return pqn->q->device->dqm->ops.update_queue(pqn->q->device->dqm,
-							pqn->q);
+							pqn->q, NULL);
 }
 
 void kfd_process_dequeue_from_all_devices(struct kfd_process *p)
@@ -394,8 +394,6 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid)
 			pdd->qpd.num_gws = 0;
 		}
 
-		kfree(pqn->q->properties.cu_mask);
-		pqn->q->properties.cu_mask = NULL;
 		uninit_queue(pqn->q);
 	}
 
@@ -411,8 +409,8 @@ err_destroy_queue:
 	return retval;
 }
 
-int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid,
-			struct queue_properties *p)
+int pqm_update_queue_properties(struct process_queue_manager *pqm,
+				unsigned int qid, struct queue_properties *p)
 {
 	int retval;
 	struct process_queue_node *pqn;
@@ -429,15 +427,15 @@ int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid,
 	pqn->q->properties.priority = p->priority;
 
 	retval = pqn->q->device->dqm->ops.update_queue(pqn->q->device->dqm,
-							pqn->q);
+							pqn->q, NULL);
 	if (retval != 0)
 		return retval;
 
 	return 0;
 }
 
-int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid,
-			struct queue_properties *p)
+int pqm_update_mqd(struct process_queue_manager *pqm,
+				unsigned int qid, struct mqd_update_info *minfo)
 {
 	int retval;
 	struct process_queue_node *pqn;
@@ -448,16 +446,8 @@ int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid,
 		return -EFAULT;
 	}
 
-	/* Free the old CU mask memory if it is already allocated, then
-	 * allocate memory for the new CU mask.
-	 */
-	kfree(pqn->q->properties.cu_mask);
-
-	pqn->q->properties.cu_mask_count = p->cu_mask_count;
-	pqn->q->properties.cu_mask = p->cu_mask;
-
 	retval = pqn->q->device->dqm->ops.update_queue(pqn->q->device->dqm,
-							pqn->q);
+							pqn->q, minfo);
 	if (retval != 0)
 		return retval;
 
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 ad571f71ee43..43e983e42c0f 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -730,6 +730,8 @@ static void dm_handle_hpd_work(struct work_struct *work)
 		dmub_hpd_wrk->adev->dm.dmub_callback[dmub_hpd_wrk->dmub_notify->type](dmub_hpd_wrk->adev,
 		dmub_hpd_wrk->dmub_notify);
 	}
+
+	kfree(dmub_hpd_wrk->dmub_notify);
 	kfree(dmub_hpd_wrk);
 
 }
@@ -755,12 +757,6 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
 
 	if (dc_enable_dmub_notifications(adev->dm.dc) &&
 		irq_params->irq_src == DC_IRQ_SOURCE_DMCUB_OUTBOX) {
-		dmub_hpd_wrk = kzalloc(sizeof(*dmub_hpd_wrk), GFP_ATOMIC);
-		if (!dmub_hpd_wrk) {
-			DRM_ERROR("Failed to allocate dmub_hpd_wrk");
-			return;
-		}
-		INIT_WORK(&dmub_hpd_wrk->handle_hpd_work, dm_handle_hpd_work);
 
 		do {
 			dc_stat_get_dmub_notification(adev->dm.dc, &notify);
@@ -769,7 +765,20 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
 				continue;
 			}
 			if (dm->dmub_thread_offload[notify.type] == true) {
-				dmub_hpd_wrk->dmub_notify = &notify;
+				dmub_hpd_wrk = kzalloc(sizeof(*dmub_hpd_wrk), GFP_ATOMIC);
+				if (!dmub_hpd_wrk) {
+					DRM_ERROR("Failed to allocate dmub_hpd_wrk");
+					return;
+				}
+				dmub_hpd_wrk->dmub_notify = kzalloc(sizeof(struct dmub_notification), GFP_ATOMIC);
+				if (!dmub_hpd_wrk->dmub_notify) {
+					kfree(dmub_hpd_wrk);
+					DRM_ERROR("Failed to allocate dmub_hpd_wrk->dmub_notify");
+					return;
+				}
+				INIT_WORK(&dmub_hpd_wrk->handle_hpd_work, dm_handle_hpd_work);
+				if (dmub_hpd_wrk->dmub_notify)
+					memcpy(dmub_hpd_wrk->dmub_notify, &notify, sizeof(struct dmub_notification));
 				dmub_hpd_wrk->adev = adev;
 				if (notify.type == DMUB_NOTIFICATION_HPD) {
 					plink = adev->dm.dc->links[notify.link_index];
@@ -1008,6 +1017,7 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
 	const unsigned char *fw_inst_const, *fw_bss_data;
 	uint32_t i, fw_inst_const_size, fw_bss_data_size;
 	bool has_hw_support;
+	struct dc *dc = adev->dm.dc;
 
 	if (!dmub_srv)
 		/* DMUB isn't supported on the ASIC. */
@@ -1094,6 +1104,19 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
 	for (i = 0; i < fb_info->num_fb; ++i)
 		hw_params.fb[i] = &fb_info->fb[i];
 
+	switch (adev->asic_type) {
+	case CHIP_YELLOW_CARP:
+		if (dc->ctx->asic_id.hw_internal_rev != YELLOW_CARP_A0) {
+			hw_params.dpia_supported = true;
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+			hw_params.disable_dpia = dc->debug.dpia_debug.bits.disable_dpia;
+#endif
+		}
+		break;
+	default:
+		break;
+	}
+
 	status = dmub_srv_hw_init(dmub_srv, &hw_params);
 	if (status != DMUB_STATUS_OK) {
 		DRM_ERROR("Error initializing DMUB HW: %d\n", status);
@@ -1295,6 +1318,37 @@ static struct hpd_rx_irq_offload_work_queue *hpd_rx_irq_create_workqueue(struct
 	return hpd_rx_offload_wq;
 }
 
+struct amdgpu_stutter_quirk {
+	u16 chip_vendor;
+	u16 chip_device;
+	u16 subsys_vendor;
+	u16 subsys_device;
+	u8 revision;
+};
+
+static const struct amdgpu_stutter_quirk amdgpu_stutter_quirk_list[] = {
+	/* https://bugzilla.kernel.org/show_bug.cgi?id=214417 */
+	{ 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc8 },
+	{ 0, 0, 0, 0, 0 },
+};
+
+static bool dm_should_disable_stutter(struct pci_dev *pdev)
+{
+	const struct amdgpu_stutter_quirk *p = amdgpu_stutter_quirk_list;
+
+	while (p && p->chip_device != 0) {
+		if (pdev->vendor == p->chip_vendor &&
+		    pdev->device == p->chip_device &&
+		    pdev->subsystem_vendor == p->subsys_vendor &&
+		    pdev->subsystem_device == p->subsys_device &&
+		    pdev->revision == p->revision) {
+			return true;
+		}
+		++p;
+	}
+	return false;
+}
+
 static int amdgpu_dm_init(struct amdgpu_device *adev)
 {
 	struct dc_init_data init_data;
@@ -1406,6 +1460,8 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 
 	if (adev->asic_type != CHIP_CARRIZO && adev->asic_type != CHIP_STONEY)
 		adev->dm.dc->debug.disable_stutter = amdgpu_pp_feature_mask & PP_STUTTER_MODE ? false : true;
+	if (dm_should_disable_stutter(adev->pdev))
+		adev->dm.dc->debug.disable_stutter = true;
 
 	if (amdgpu_dc_debug_mask & DC_DISABLE_STUTTER)
 		adev->dm.dc->debug.disable_stutter = true;
@@ -1792,7 +1848,7 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
 		break;
 	case IP_VERSION(3, 1, 2):
 	case IP_VERSION(3, 1, 3):
-		dmub_asic = DMUB_ASIC_DCN31;
+		dmub_asic = (adev->external_rev_id == YELLOW_CARP_B0) ? DMUB_ASIC_DCN31B : DMUB_ASIC_DCN31;
 		fw_name_dmub = FIRMWARE_YELLOW_CARP_DMUB;
 		break;
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index 1a68a674913c..3655663e079b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -491,7 +491,7 @@ static ssize_t dp_phy_settings_write(struct file *f, const char __user *buf,
 	if (!wr_buf)
 		return -ENOSPC;
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					   (long *)param, buf,
 					   max_param_num,
 					   &param_nums)) {
@@ -643,7 +643,7 @@ static ssize_t dp_phy_test_pattern_debugfs_write(struct file *f, const char __us
 	if (!wr_buf)
 		return -ENOSPC;
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					   (long *)param, buf,
 					   max_param_num,
 					   &param_nums)) {
@@ -918,7 +918,7 @@ static ssize_t dp_dsc_passthrough_set(struct file *f, const char __user *buf,
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					   &param, buf,
 					   max_param_num,
 					   &param_nums)) {
@@ -1215,7 +1215,7 @@ static ssize_t trigger_hotplug(struct file *f, const char __user *buf,
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 						(long *)param, buf,
 						max_param_num,
 						&param_nums)) {
@@ -1400,7 +1400,7 @@ static ssize_t dp_dsc_clock_en_write(struct file *f, const char __user *buf,
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					    (long *)param, buf,
 					    max_param_num,
 					    &param_nums)) {
@@ -1585,7 +1585,7 @@ static ssize_t dp_dsc_slice_width_write(struct file *f, const char __user *buf,
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					    (long *)param, buf,
 					    max_param_num,
 					    &param_nums)) {
@@ -1770,7 +1770,7 @@ static ssize_t dp_dsc_slice_height_write(struct file *f, const char __user *buf,
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					    (long *)param, buf,
 					    max_param_num,
 					    &param_nums)) {
@@ -1948,7 +1948,7 @@ static ssize_t dp_dsc_bits_per_pixel_write(struct file *f, const char __user *bu
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					    (long *)param, buf,
 					    max_param_num,
 					    &param_nums)) {
@@ -2386,7 +2386,7 @@ static ssize_t dp_max_bpc_write(struct file *f, const char __user *buf,
 		return -ENOSPC;
 	}
 
-	if (parse_write_buffer_into_params(wr_buf, size,
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
 					   (long *)param, buf,
 					   max_param_num,
 					   &param_nums)) {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 1a99fcc27078..874a49b605c7 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -64,6 +64,8 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 	payload.i2c_over_aux = (msg->request & DP_AUX_NATIVE_WRITE) == 0;
 	payload.write = (msg->request & DP_AUX_I2C_READ) == 0;
 	payload.mot = (msg->request & DP_AUX_I2C_MOT) != 0;
+	payload.write_status_update =
+			(msg->request & DP_AUX_I2C_WRITE_STATUS_UPDATE) != 0;
 	payload.defer_delay = 0;
 
 	result = dc_link_aux_transfer_raw(TO_DM_AUX(aux)->ddc_service, &payload,
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c
index 70a554f1e725..c022e56f9459 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c
@@ -107,6 +107,8 @@ bool amdgpu_dm_psr_enable(struct dc_stream_state *stream)
 	 */
 	// Init fail safe of 2 frames static
 	unsigned int num_frames_static = 2;
+	unsigned int power_opt = 0;
+	bool psr_enable = true;
 
 	DRM_DEBUG_DRIVER("Enabling psr...\n");
 
@@ -133,7 +135,9 @@ bool amdgpu_dm_psr_enable(struct dc_stream_state *stream)
 					   &stream, 1,
 					   &params);
 
-	return dc_link_set_psr_allow_active(link, true, false, false);
+	power_opt |= psr_power_opt_z10_static_screen;
+
+	return dc_link_set_psr_allow_active(link, &psr_enable, false, false, &power_opt);
 }
 
 /*
@@ -144,10 +148,12 @@ bool amdgpu_dm_psr_enable(struct dc_stream_state *stream)
  */
 bool amdgpu_dm_psr_disable(struct dc_stream_state *stream)
 {
+	unsigned int power_opt = 0;
+	bool psr_enable = false;
 
 	DRM_DEBUG_DRIVER("Disabling psr...\n");
 
-	return dc_link_set_psr_allow_active(stream->link, false, true, false);
+	return dc_link_set_psr_allow_active(stream->link, &psr_enable, true, false, &power_opt);
 }
 
 /*
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index c17732fba039..a4bef4364afd 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -2321,6 +2321,8 @@ static enum bp_result get_integrated_info_v2_2(
 
 	info->ext_disp_conn_info.checksum =
 		info_v2_2->extdispconninfo.checksum;
+	info->ext_disp_conn_info.fixdpvoltageswing =
+		info_v2_2->extdispconninfo.fixdpvoltageswing;
 
 	info->edp1_info.edp_backlight_pwm_hz =
 	le16_to_cpu(info_v2_2->edp1_info.edp_backlight_pwm_hz);
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
index 1548b2a3fe03..26f96ee32472 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
@@ -100,11 +100,13 @@ void clk_mgr_exit_optimized_pwr_state(const struct dc *dc, struct clk_mgr *clk_m
 
 	if (edp_num) {
 		for (panel_inst = 0; panel_inst < edp_num; panel_inst++) {
+			bool allow_active = false;
+
 			edp_link = edp_links[panel_inst];
 			if (!edp_link->psr_settings.psr_feature_enabled)
 				continue;
 			clk_mgr->psr_allow_active_cache = edp_link->psr_settings.psr_allow_active;
-			dc_link_set_psr_allow_active(edp_link, false, false, false);
+			dc_link_set_psr_allow_active(edp_link, &allow_active, false, false, NULL);
 		}
 	}
 
@@ -124,7 +126,7 @@ void clk_mgr_optimize_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr)
 			if (!edp_link->psr_settings.psr_feature_enabled)
 				continue;
 			dc_link_set_psr_allow_active(edp_link,
-					clk_mgr->psr_allow_active_cache, false, false);
+					&clk_mgr->psr_allow_active_cache, false, false, NULL);
 		}
 	}
 
@@ -283,13 +285,8 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p
 			BREAK_TO_DEBUGGER();
 			return NULL;
 		}
-		if (ASICREV_IS_YELLOW_CARP(asic_id.hw_internal_rev)) {
-			/* TODO: to add DCN31 clk_mgr support, once CLK IP header files are available,
-			 * for now use DCN3.0 clk mgr.
-			 */
-			dcn31_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
-			return &clk_mgr->base.base;
-		}
+
+		dcn31_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
 		return &clk_mgr->base.base;
 	}
 #endif
@@ -326,7 +323,6 @@ void dc_destroy_clk_mgr(struct clk_mgr *clk_mgr_base)
 		break;
 
 	case FAMILY_YELLOW_CARP:
-		if (ASICREV_IS_YELLOW_CARP(clk_mgr_base->ctx->asic_id.hw_internal_rev))
 			dcn31_clk_mgr_destroy(clk_mgr);
 		break;
 
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c
index 0088dff441da..f4c9a458ace8 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c
@@ -523,14 +523,21 @@ static unsigned int find_clk_for_voltage(
 		unsigned int voltage)
 {
 	int i;
+	int max_voltage = 0;
+	int clock = 0;
 
 	for (i = 0; i < NUM_SOC_VOLTAGE_LEVELS; i++) {
-		if (clock_table->SocVoltage[i] == voltage)
+		if (clock_table->SocVoltage[i] == voltage) {
 			return clocks[i];
+		} else if (clock_table->SocVoltage[i] >= max_voltage &&
+				clock_table->SocVoltage[i] < voltage) {
+			max_voltage = clock_table->SocVoltage[i];
+			clock = clocks[i];
+		}
 	}
 
-	ASSERT(0);
-	return 0;
+	ASSERT(clock);
+	return clock;
 }
 
 void dcn31_clk_mgr_helper_populate_bw_params(
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 935a50d6e933..12e5470fa567 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1897,12 +1897,14 @@ static bool is_flip_pending_in_pipes(struct dc *dc, struct dc_state *context)
 static void process_deferred_updates(struct dc *dc)
 {
 #ifdef CONFIG_DRM_AMD_DC_DCN
-	int i;
+	int i = 0;
 
-	if (dc->debug.enable_mem_low_power.bits.cm)
+	if (dc->debug.enable_mem_low_power.bits.cm) {
+		ASSERT(dc->dcn_ip->max_num_dpp);
 		for (i = 0; i < dc->dcn_ip->max_num_dpp; i++)
 			if (dc->res_pool->dpps[i]->funcs->dpp_deferred_update)
 				dc->res_pool->dpps[i]->funcs->dpp_deferred_update(dc->res_pool->dpps[i]);
+	}
 #endif
 }
 
@@ -2283,6 +2285,9 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
 			update_flags->bits.gamma_change = 1;
 	}
 
+	if (u->lut3d_func || u->func_shaper)
+		update_flags->bits.lut_3d = 1;
+
 	if (u->hdr_mult.value)
 		if (u->hdr_mult.value != u->surface->hdr_mult.value) {
 			update_flags->bits.hdr_mult = 1;
@@ -2296,6 +2301,7 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
 
 	if (update_flags->bits.input_csc_change
 			|| update_flags->bits.coeff_reduction_change
+			|| update_flags->bits.lut_3d
 			|| update_flags->bits.gamma_change
 			|| update_flags->bits.gamut_remap_change) {
 		type = UPDATE_TYPE_FULL;
@@ -3493,6 +3499,7 @@ void dc_get_clock(struct dc *dc, enum dc_clock_type clock_type, struct dc_clock_
 bool dc_set_psr_allow_active(struct dc *dc, bool enable)
 {
 	int i;
+	bool allow_active;
 
 	for (i = 0; i < dc->current_state->stream_count ; i++) {
 		struct dc_link *link;
@@ -3504,10 +3511,12 @@ bool dc_set_psr_allow_active(struct dc *dc, bool enable)
 
 		if (link->psr_settings.psr_feature_enabled) {
 			if (enable && !link->psr_settings.psr_allow_active) {
-				if (!dc_link_set_psr_allow_active(link, true, false, false))
+				allow_active = true;
+				if (!dc_link_set_psr_allow_active(link, &allow_active, false, false, NULL))
 					return false;
 			} else if (!enable && link->psr_settings.psr_allow_active) {
-				if (!dc_link_set_psr_allow_active(link, false, true, false))
+				allow_active = false;
+				if (!dc_link_set_psr_allow_active(link, &allow_active, true, false, NULL))
 					return false;
 			}
 		}
@@ -3742,6 +3751,60 @@ bool dc_process_dmub_set_config_async(struct dc *dc,
 }
 
 /**
+ *****************************************************************************
+ *  Function: dc_process_dmub_set_mst_slots
+ *
+ *  @brief
+ *		Submits mst slot allocation command to dmub via inbox message
+ *
+ *  @param
+ *		[in] dc: dc structure
+ *		[in] link_index: link index
+ *		[in] mst_alloc_slots: mst slots to be allotted
+ *		[out] mst_slots_in_use: mst slots in use returned in failure case
+ *
+ *	@return
+ *		DC_OK if successful, DC_ERROR if failure
+ *****************************************************************************
+ */
+enum dc_status dc_process_dmub_set_mst_slots(const struct dc *dc,
+				uint32_t link_index,
+				uint8_t mst_alloc_slots,
+				uint8_t *mst_slots_in_use)
+{
+	union dmub_rb_cmd cmd = {0};
+	struct dc_dmub_srv *dmub_srv = dc->ctx->dmub_srv;
+
+	/* prepare MST_ALLOC_SLOTS command */
+	cmd.set_mst_alloc_slots.header.type = DMUB_CMD__DPIA;
+	cmd.set_mst_alloc_slots.header.sub_type = DMUB_CMD__DPIA_MST_ALLOC_SLOTS;
+
+	cmd.set_mst_alloc_slots.mst_slots_control.instance = dc->links[link_index]->ddc_hw_inst;
+	cmd.set_mst_alloc_slots.mst_slots_control.mst_alloc_slots = mst_alloc_slots;
+
+	if (!dc_dmub_srv_cmd_with_reply_data(dmub_srv, &cmd))
+		/* command is not processed by dmub */
+		return DC_ERROR_UNEXPECTED;
+
+	/* command processed by dmub, if ret_status is 1 */
+	if (cmd.set_config_access.header.ret_status != 1)
+		/* command processing error */
+		return DC_ERROR_UNEXPECTED;
+
+	/* command processed and we have a status of 2, mst not enabled in dpia */
+	if (cmd.set_mst_alloc_slots.mst_slots_control.immed_status == 2)
+		return DC_FAIL_UNSUPPORTED_1;
+
+	/* previously configured mst alloc and used slots did not match */
+	if (cmd.set_mst_alloc_slots.mst_slots_control.immed_status == 3) {
+		*mst_slots_in_use = cmd.set_mst_alloc_slots.mst_slots_control.mst_slots_in_use;
+		return DC_NOT_SUPPORTED;
+	}
+
+	return DC_OK;
+}
+
+/**
  * dc_disable_accelerated_mode - disable accelerated mode
  * @dc: dc structure
  */
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index ec5f107bc85a..2796bdd17de1 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -1660,6 +1660,14 @@ static bool dc_link_construct_legacy(struct dc_link *link,
 				DC_LOG_DC("BIOS object table - ddi_channel_mapping: 0x%04X", link->ddi_channel_mapping.raw);
 				DC_LOG_DC("BIOS object table - chip_caps: %d", link->chip_caps);
 			}
+
+			if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) {
+				link->bios_forced_drive_settings.VOLTAGE_SWING =
+						(info->ext_disp_conn_info.fixdpvoltageswing & 0x3);
+				link->bios_forced_drive_settings.PRE_EMPHASIS =
+						((info->ext_disp_conn_info.fixdpvoltageswing >> 2) & 0x3);
+			}
+
 			break;
 		}
 	}
@@ -1757,6 +1765,9 @@ static bool dc_link_construct_dpia(struct dc_link *link,
 
 	link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED;
 
+	/* Some docks seem to NAK I2C writes to segment pointer with mot=0. */
+	link->wa_flags.dp_mot_reset_segment = true;
+
 	return true;
 
 ddc_create_fail:
@@ -2916,8 +2927,8 @@ bool dc_link_set_backlight_level(const struct dc_link *link,
 	return true;
 }
 
-bool dc_link_set_psr_allow_active(struct dc_link *link, bool allow_active,
-		bool wait, bool force_static)
+bool dc_link_set_psr_allow_active(struct dc_link *link, const bool *allow_active,
+		bool wait, bool force_static, const unsigned int *power_opts)
 {
 	struct dc  *dc = link->ctx->dc;
 	struct dmcu *dmcu = dc->res_pool->dmcu;
@@ -2930,20 +2941,33 @@ bool dc_link_set_psr_allow_active(struct dc_link *link, bool allow_active,
 	if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst))
 		return false;
 
-	link->psr_settings.psr_allow_active = allow_active;
+	/* Set power optimization flag */
+	if (power_opts && link->psr_settings.psr_power_opt != *power_opts) {
+		link->psr_settings.psr_power_opt = *power_opts;
+
+		if (psr != NULL && link->psr_settings.psr_feature_enabled && psr->funcs->psr_set_power_opt)
+			psr->funcs->psr_set_power_opt(psr, link->psr_settings.psr_power_opt);
+	}
+
+	/* Enable or Disable PSR */
+	if (allow_active && link->psr_settings.psr_allow_active != *allow_active) {
+		link->psr_settings.psr_allow_active = *allow_active;
+
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-	if (!allow_active)
-		dc_z10_restore(dc);
+		if (!link->psr_settings.psr_allow_active)
+			dc_z10_restore(dc);
 #endif
 
-	if (psr != NULL && link->psr_settings.psr_feature_enabled) {
-		if (force_static && psr->funcs->psr_force_static)
-			psr->funcs->psr_force_static(psr, panel_inst);
-		psr->funcs->psr_enable(psr, allow_active, wait, panel_inst);
-	} else if ((dmcu != NULL && dmcu->funcs->is_dmcu_initialized(dmcu)) && link->psr_settings.psr_feature_enabled)
-		dmcu->funcs->set_psr_enable(dmcu, allow_active, wait);
-	else
-		return false;
+		if (psr != NULL && link->psr_settings.psr_feature_enabled) {
+			if (force_static && psr->funcs->psr_force_static)
+				psr->funcs->psr_force_static(psr, panel_inst);
+			psr->funcs->psr_enable(psr, link->psr_settings.psr_allow_active, wait, panel_inst);
+		} else if ((dmcu != NULL && dmcu->funcs->is_dmcu_initialized(dmcu)) &&
+			link->psr_settings.psr_feature_enabled)
+			dmcu->funcs->set_psr_enable(dmcu, link->psr_settings.psr_allow_active, wait);
+		else
+			return false;
+	}
 
 	return true;
 }
@@ -3475,6 +3499,20 @@ enum dc_status dc_link_allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 
 	ASSERT(proposed_table.stream_count > 0);
 
+	if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
+		static enum dc_status status;
+		uint8_t mst_alloc_slots = 0, prev_mst_slots_in_use = 0xFF;
+
+		for (i = 0; i < link->mst_stream_alloc_table.stream_count; i++)
+			mst_alloc_slots += link->mst_stream_alloc_table.stream_allocations[i].slot_count;
+
+		status = dc_process_dmub_set_mst_slots(link->dc, link->link_index,
+			mst_alloc_slots, &prev_mst_slots_in_use);
+		ASSERT(status == DC_OK);
+		DC_LOG_MST("dpia : status[%d]: alloc_slots[%d]: used_slots[%d]\n",
+				status, mst_alloc_slots, prev_mst_slots_in_use);
+	}
+
 	/* program DP source TX for payload */
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	switch (dp_get_link_encoding_format(&link->cur_link_settings)) {
@@ -3818,6 +3856,20 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 #endif
 	}
 
+	if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
+		enum dc_status status;
+		uint8_t mst_alloc_slots = 0, prev_mst_slots_in_use = 0xFF;
+
+		for (i = 0; i < link->mst_stream_alloc_table.stream_count; i++)
+			mst_alloc_slots += link->mst_stream_alloc_table.stream_allocations[i].slot_count;
+
+		status = dc_process_dmub_set_mst_slots(link->dc, link->link_index,
+			mst_alloc_slots, &prev_mst_slots_in_use);
+		ASSERT(status != DC_NOT_SUPPORTED);
+		DC_LOG_MST("dpia : status[%d]: alloc_slots[%d]: used_slots[%d]\n",
+				status, mst_alloc_slots, prev_mst_slots_in_use);
+	}
+
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	switch (dp_get_link_encoding_format(&link->cur_link_settings)) {
 	case DP_8b_10b_ENCODING:
@@ -3861,6 +3913,9 @@ static void update_psp_stream_config(struct pipe_ctx *pipe_ctx, bool dpms_off)
 	struct cp_psp *cp_psp = &pipe_ctx->stream->ctx->cp_psp;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	struct link_encoder *link_enc = NULL;
+	struct dc_state *state = pipe_ctx->stream->ctx->dc->current_state;
+	struct link_enc_assignment link_enc_assign;
+	int i;
 #endif
 
 	if (cp_psp && cp_psp->funcs.update_stream_config) {
@@ -3874,9 +3929,72 @@ static void update_psp_stream_config(struct pipe_ctx *pipe_ctx, bool dpms_off)
 		config.dig_be = pipe_ctx->stream->link->link_enc_hw_inst;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 		config.stream_enc_idx = pipe_ctx->stream_res.stream_enc->id - ENGINE_ID_DIGA;
-		if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_PHY) {
+		
+		if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_PHY ||
+				pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
 			link_enc = pipe_ctx->stream->link->link_enc;
+			config.dio_output_type = pipe_ctx->stream->link->ep_type;
+			config.dio_output_idx = link_enc->transmitter - TRANSMITTER_UNIPHY_A;
+			if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_PHY)
+				link_enc = pipe_ctx->stream->link->link_enc;
+			else if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA)
+				if (pipe_ctx->stream->link->dc->res_pool->funcs->link_encs_assign) {
+					link_enc = link_enc_cfg_get_link_enc_used_by_stream(
+							pipe_ctx->stream->ctx->dc,
+							pipe_ctx->stream);
+			}
+			// Initialize PHY ID with ABCDE - 01234 mapping except when it is B0
 			config.phy_idx = link_enc->transmitter - TRANSMITTER_UNIPHY_A;
+
+			//look up the link_enc_assignment for the current pipe_ctx
+			for (i = 0; i < state->stream_count; i++) {
+				if (pipe_ctx->stream == state->streams[i]) {
+					link_enc_assign = state->res_ctx.link_enc_cfg_ctx.link_enc_assignments[i];
+				}
+			}
+			// Add flag to guard new A0 DIG mapping
+			if (pipe_ctx->stream->ctx->dc->enable_c20_dtm_b0 == true) {
+				config.dig_be = link_enc_assign.eng_id;
+				config.dio_output_type = pipe_ctx->stream->link->ep_type;
+				config.dio_output_idx = link_enc->transmitter - TRANSMITTER_UNIPHY_A;
+			} else {
+				config.dio_output_type = 0;
+				config.dio_output_idx = 0;
+			}
+
+			// Add flag to guard B0 implementation
+			if (pipe_ctx->stream->ctx->dc->enable_c20_dtm_b0 == true &&
+					link_enc->ctx->asic_id.hw_internal_rev == YELLOW_CARP_B0) {
+				if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
+					link_enc = link_enc_assign.stream->link_enc;
+
+					// enum ID 1-4 maps to DPIA PHY ID 0-3
+					config.phy_idx = link_enc_assign.ep_id.link_id.enum_id - ENUM_ID_1;
+				} else {  // for non DPIA mode over B0, ABCDE maps to 01564
+
+					switch (link_enc->transmitter) {
+					case TRANSMITTER_UNIPHY_A:
+						config.phy_idx = 0;
+						break;
+					case TRANSMITTER_UNIPHY_B:
+						config.phy_idx = 1;
+						break;
+					case TRANSMITTER_UNIPHY_C:
+						config.phy_idx = 5;
+						break;
+					case TRANSMITTER_UNIPHY_D:
+						config.phy_idx = 6;
+						break;
+					case TRANSMITTER_UNIPHY_E:
+						config.phy_idx = 4;
+						break;
+					default:
+						config.phy_idx = 0;
+						break;
+					}
+
+				}
+			}
 		} else if (pipe_ctx->stream->link->dc->res_pool->funcs->link_encs_assign) {
 			link_enc = link_enc_cfg_get_link_enc_used_by_stream(
 					pipe_ctx->stream->ctx->dc,
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
index 471a67a64299..60539b1f2a80 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
@@ -554,6 +554,7 @@ bool dal_ddc_service_query_ddc_data(
 		payload.address = address;
 		payload.reply = NULL;
 		payload.defer_delay = get_defer_delay(ddc);
+		payload.write_status_update = false;
 
 		if (write_size != 0) {
 			payload.write = true;
@@ -625,24 +626,24 @@ bool dal_ddc_submit_aux_command(struct ddc_service *ddc,
 	do {
 		struct aux_payload current_payload;
 		bool is_end_of_payload = (retrieved + DEFAULT_AUX_MAX_DATA_SIZE) >=
-			payload->length;
+				payload->length ? true : false;
+		uint32_t payload_length = is_end_of_payload ?
+				payload->length - retrieved : DEFAULT_AUX_MAX_DATA_SIZE;
 
 		current_payload.address = payload->address;
 		current_payload.data = &payload->data[retrieved];
 		current_payload.defer_delay = payload->defer_delay;
 		current_payload.i2c_over_aux = payload->i2c_over_aux;
-		current_payload.length = is_end_of_payload ?
-			payload->length - retrieved : DEFAULT_AUX_MAX_DATA_SIZE;
-		/* set mot (middle of transaction) to false
-		 * if it is the last payload
-		 */
+		current_payload.length = payload_length;
+		/* set mot (middle of transaction) to false if it is the last payload */
 		current_payload.mot = is_end_of_payload ? payload->mot:true;
+		current_payload.write_status_update = false;
 		current_payload.reply = payload->reply;
 		current_payload.write = payload->write;
 
 		ret = dc_link_aux_transfer_with_retries(ddc, &current_payload);
 
-		retrieved += current_payload.length;
+		retrieved += payload_length;
 	} while (retrieved < payload->length && ret == true);
 
 	return ret;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index bb96e4e9ccfc..cc25ba0ec7db 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -106,6 +106,10 @@ static bool decide_fallback_link_setting(
 static struct dc_link_settings get_common_supported_link_settings(
 		struct dc_link_settings link_setting_a,
 		struct dc_link_settings link_setting_b);
+static void maximize_lane_settings(const struct link_training_settings *lt_settings,
+		struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX]);
+static void override_lane_settings(const struct link_training_settings *lt_settings,
+		struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX]);
 
 static uint32_t get_cr_training_aux_rd_interval(struct dc_link *link,
 		const struct dc_link_settings *link_settings)
@@ -515,8 +519,6 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 	enum dc_dp_training_pattern pattern,
 	uint32_t offset)
 {
-	union dpcd_training_lane dpcd_lane[LANE_COUNT_DP_MAX] = { { {0} } };
-
 	uint32_t dpcd_base_lt_offset;
 
 	uint8_t dpcd_lt_buffer[5] = {0};
@@ -554,16 +556,14 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 			dpcd_pattern.v1_4.TRAINING_PATTERN_SET);
 	}
 
-	dp_hw_to_dpcd_lane_settings(lt_settings, lt_settings->lane_settings, dpcd_lane);
-
 	/* concatenate everything into one buffer*/
-
-	size_in_bytes = lt_settings->link_settings.lane_count * sizeof(dpcd_lane[0]);
+	size_in_bytes = lt_settings->link_settings.lane_count *
+			sizeof(lt_settings->dpcd_lane_settings[0]);
 
 	 // 0x00103 - 0x00102
 	memmove(
 		&dpcd_lt_buffer[DP_TRAINING_LANE0_SET - DP_TRAINING_PATTERN_SET],
-		dpcd_lane,
+		lt_settings->dpcd_lane_settings,
 		size_in_bytes);
 
 	if (is_repeater(link, offset)) {
@@ -575,7 +575,7 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 					__func__,
 					offset,
 					dpcd_base_lt_offset,
-					dpcd_lane[0].tx_ffe.PRESET_VALUE);
+					lt_settings->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE);
 		else if (dp_get_link_encoding_format(&lt_settings->link_settings) ==
 				DP_8b_10b_ENCODING)
 #endif
@@ -584,10 +584,10 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 			__func__,
 			offset,
 			dpcd_base_lt_offset,
-			dpcd_lane[0].bits.VOLTAGE_SWING_SET,
-			dpcd_lane[0].bits.PRE_EMPHASIS_SET,
-			dpcd_lane[0].bits.MAX_SWING_REACHED,
-			dpcd_lane[0].bits.MAX_PRE_EMPHASIS_REACHED);
+			lt_settings->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET,
+			lt_settings->dpcd_lane_settings[0].bits.PRE_EMPHASIS_SET,
+			lt_settings->dpcd_lane_settings[0].bits.MAX_SWING_REACHED,
+			lt_settings->dpcd_lane_settings[0].bits.MAX_PRE_EMPHASIS_REACHED);
 	} else {
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 		if (dp_get_link_encoding_format(&lt_settings->link_settings) ==
@@ -595,17 +595,17 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 			DC_LOG_HW_LINK_TRAINING("%s:\n 0x%X TX_FFE_PRESET_VALUE = %x\n",
 					__func__,
 					dpcd_base_lt_offset,
-					dpcd_lane[0].tx_ffe.PRESET_VALUE);
+					lt_settings->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE);
 		else if (dp_get_link_encoding_format(&lt_settings->link_settings) ==
 				DP_8b_10b_ENCODING)
 #endif
 		DC_LOG_HW_LINK_TRAINING("%s:\n 0x%X VS set = %x  PE set = %x max VS Reached = %x  max PE Reached = %x\n",
 			__func__,
 			dpcd_base_lt_offset,
-			dpcd_lane[0].bits.VOLTAGE_SWING_SET,
-			dpcd_lane[0].bits.PRE_EMPHASIS_SET,
-			dpcd_lane[0].bits.MAX_SWING_REACHED,
-			dpcd_lane[0].bits.MAX_PRE_EMPHASIS_REACHED);
+			lt_settings->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET,
+			lt_settings->dpcd_lane_settings[0].bits.PRE_EMPHASIS_SET,
+			lt_settings->dpcd_lane_settings[0].bits.MAX_SWING_REACHED,
+			lt_settings->dpcd_lane_settings[0].bits.MAX_PRE_EMPHASIS_REACHED);
 	}
 	if (edp_workaround) {
 		/* for eDP write in 2 parts because the 5-byte burst is
@@ -620,7 +620,7 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 		core_link_write_dpcd(
 			link,
 			DP_TRAINING_LANE0_SET,
-			(uint8_t *)(dpcd_lane),
+			(uint8_t *)(lt_settings->dpcd_lane_settings),
 			size_in_bytes);
 
 #if defined(CONFIG_DRM_AMD_DC_DCN)
@@ -711,34 +711,44 @@ void dp_hw_to_dpcd_lane_settings(
 	}
 }
 
-void dp_update_drive_settings(
-		struct link_training_settings *dest,
-		struct link_training_settings src)
+void dp_decide_lane_settings(
+		const struct link_training_settings *lt_settings,
+		const union lane_adjust ln_adjust[LANE_COUNT_DP_MAX],
+		struct dc_lane_settings hw_lane_settings[LANE_COUNT_DP_MAX],
+		union dpcd_training_lane dpcd_lane_settings[LANE_COUNT_DP_MAX])
 {
 	uint32_t lane;
-	for (lane = 0; lane < src.link_settings.lane_count; lane++) {
-		if (dest->voltage_swing == NULL)
-			dest->lane_settings[lane].VOLTAGE_SWING = src.lane_settings[lane].VOLTAGE_SWING;
-		else
-			dest->lane_settings[lane].VOLTAGE_SWING = *dest->voltage_swing;
-
-		if (dest->pre_emphasis == NULL)
-			dest->lane_settings[lane].PRE_EMPHASIS = src.lane_settings[lane].PRE_EMPHASIS;
-		else
-			dest->lane_settings[lane].PRE_EMPHASIS = *dest->pre_emphasis;
-
-		if (dest->post_cursor2 == NULL)
-			dest->lane_settings[lane].POST_CURSOR2 = src.lane_settings[lane].POST_CURSOR2;
-		else
-			dest->lane_settings[lane].POST_CURSOR2 = *dest->post_cursor2;
 
+	for (lane = 0; lane < LANE_COUNT_DP_MAX; lane++) {
+		if (dp_get_link_encoding_format(&lt_settings->link_settings) ==
+				DP_8b_10b_ENCODING) {
+			hw_lane_settings[lane].VOLTAGE_SWING =
+					(enum dc_voltage_swing)(ln_adjust[lane].bits.
+							VOLTAGE_SWING_LANE);
+			hw_lane_settings[lane].PRE_EMPHASIS =
+					(enum dc_pre_emphasis)(ln_adjust[lane].bits.
+							PRE_EMPHASIS_LANE);
+		}
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-		if (dest->ffe_preset == NULL)
-			dest->lane_settings[lane].FFE_PRESET = src.lane_settings[lane].FFE_PRESET;
-		else
-			dest->lane_settings[lane].FFE_PRESET = *dest->ffe_preset;
+		else if (dp_get_link_encoding_format(&lt_settings->link_settings) ==
+				DP_128b_132b_ENCODING) {
+			hw_lane_settings[lane].FFE_PRESET.raw =
+					ln_adjust[lane].tx_ffe.PRESET_VALUE;
+		}
 #endif
 	}
+	dp_hw_to_dpcd_lane_settings(lt_settings, hw_lane_settings, dpcd_lane_settings);
+
+	if (lt_settings->disallow_per_lane_settings) {
+		/* we find the maximum of the requested settings across all lanes*/
+		/* and set this maximum for all lanes*/
+		maximize_lane_settings(lt_settings, hw_lane_settings);
+		override_lane_settings(lt_settings, hw_lane_settings);
+
+		if (lt_settings->always_match_dpcd_with_hw_lane_settings)
+			dp_hw_to_dpcd_lane_settings(lt_settings, hw_lane_settings, dpcd_lane_settings);
+	}
+
 }
 
 static uint8_t get_nibble_at_index(const uint8_t *buf,
@@ -768,55 +778,29 @@ static enum dc_pre_emphasis get_max_pre_emphasis_for_voltage_swing(
 
 }
 
-static void find_max_drive_settings(
-	const struct link_training_settings *link_training_setting,
-	struct link_training_settings *max_lt_setting)
+static void maximize_lane_settings(const struct link_training_settings *lt_settings,
+		struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX])
 {
 	uint32_t lane;
 	struct dc_lane_settings max_requested;
 
-	max_requested.VOLTAGE_SWING =
-		link_training_setting->
-		lane_settings[0].VOLTAGE_SWING;
-	max_requested.PRE_EMPHASIS =
-		link_training_setting->
-		lane_settings[0].PRE_EMPHASIS;
-	/*max_requested.postCursor2 =
-	 * link_training_setting->laneSettings[0].postCursor2;*/
+	max_requested.VOLTAGE_SWING = lane_settings[0].VOLTAGE_SWING;
+	max_requested.PRE_EMPHASIS = lane_settings[0].PRE_EMPHASIS;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-	max_requested.FFE_PRESET =
-		link_training_setting->lane_settings[0].FFE_PRESET;
+	max_requested.FFE_PRESET = lane_settings[0].FFE_PRESET;
 #endif
 
 	/* Determine what the maximum of the requested settings are*/
-	for (lane = 1; lane < link_training_setting->link_settings.lane_count;
-			lane++) {
-		if (link_training_setting->lane_settings[lane].VOLTAGE_SWING >
-			max_requested.VOLTAGE_SWING)
-
-			max_requested.VOLTAGE_SWING =
-			link_training_setting->
-			lane_settings[lane].VOLTAGE_SWING;
+	for (lane = 1; lane < lt_settings->link_settings.lane_count; lane++) {
+		if (lane_settings[lane].VOLTAGE_SWING > max_requested.VOLTAGE_SWING)
+			max_requested.VOLTAGE_SWING = lane_settings[lane].VOLTAGE_SWING;
 
-		if (link_training_setting->lane_settings[lane].PRE_EMPHASIS >
-				max_requested.PRE_EMPHASIS)
-			max_requested.PRE_EMPHASIS =
-			link_training_setting->
-			lane_settings[lane].PRE_EMPHASIS;
-
-		/*
-		if (link_training_setting->laneSettings[lane].postCursor2 >
-		 max_requested.postCursor2)
-		{
-		max_requested.postCursor2 =
-		link_training_setting->laneSettings[lane].postCursor2;
-		}
-		*/
+		if (lane_settings[lane].PRE_EMPHASIS > max_requested.PRE_EMPHASIS)
+			max_requested.PRE_EMPHASIS = lane_settings[lane].PRE_EMPHASIS;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-		if (link_training_setting->lane_settings[lane].FFE_PRESET.settings.level >
+		if (lane_settings[lane].FFE_PRESET.settings.level >
 				max_requested.FFE_PRESET.settings.level)
 			max_requested.FFE_PRESET.settings.level =
-					link_training_setting->
 					lane_settings[lane].FFE_PRESET.settings.level;
 #endif
 	}
@@ -828,10 +812,6 @@ static void find_max_drive_settings(
 
 	if (max_requested.PRE_EMPHASIS > PRE_EMPHASIS_MAX_LEVEL)
 		max_requested.PRE_EMPHASIS = PRE_EMPHASIS_MAX_LEVEL;
-	/*
-	if (max_requested.postCursor2 > PostCursor2_MaxLevel)
-	max_requested.postCursor2 = PostCursor2_MaxLevel;
-	*/
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	if (max_requested.FFE_PRESET.settings.level > DP_FFE_PRESET_MAX_LEVEL)
 		max_requested.FFE_PRESET.settings.level = DP_FFE_PRESET_MAX_LEVEL;
@@ -845,61 +825,58 @@ static void find_max_drive_settings(
 		get_max_pre_emphasis_for_voltage_swing(
 			max_requested.VOLTAGE_SWING);
 
-	/*
-	 * Post Cursor2 levels are completely independent from
-	 * pre-emphasis (Post Cursor1) levels. But Post Cursor2 levels
-	 * can only be applied to each allowable combination of voltage
-	 * swing and pre-emphasis levels */
-	 /* if ( max_requested.postCursor2 >
-	  *  getMaxPostCursor2ForVoltageSwing(max_requested.voltageSwing))
-	  *  max_requested.postCursor2 =
-	  *  getMaxPostCursor2ForVoltageSwing(max_requested.voltageSwing);
-	  */
-
-	max_lt_setting->link_settings.link_rate =
-		link_training_setting->link_settings.link_rate;
-	max_lt_setting->link_settings.lane_count =
-	link_training_setting->link_settings.lane_count;
-	max_lt_setting->link_settings.link_spread =
-		link_training_setting->link_settings.link_spread;
-
-	for (lane = 0; lane <
-		link_training_setting->link_settings.lane_count;
-		lane++) {
-		max_lt_setting->lane_settings[lane].VOLTAGE_SWING =
-			max_requested.VOLTAGE_SWING;
-		max_lt_setting->lane_settings[lane].PRE_EMPHASIS =
-			max_requested.PRE_EMPHASIS;
-		/*max_lt_setting->laneSettings[lane].postCursor2 =
-		 * max_requested.postCursor2;
-		 */
+	for (lane = 0; lane < LANE_COUNT_DP_MAX; lane++) {
+		lane_settings[lane].VOLTAGE_SWING = max_requested.VOLTAGE_SWING;
+		lane_settings[lane].PRE_EMPHASIS = max_requested.PRE_EMPHASIS;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-		max_lt_setting->lane_settings[lane].FFE_PRESET =
-			max_requested.FFE_PRESET;
+		lane_settings[lane].FFE_PRESET = max_requested.FFE_PRESET;
 #endif
 	}
+}
+
+static void override_lane_settings(const struct link_training_settings *lt_settings,
+		struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX])
+{
+	uint32_t lane;
+
+	if (lt_settings->voltage_swing == NULL &&
+	    lt_settings->pre_emphasis == NULL &&
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+	    lt_settings->ffe_preset == NULL &&
+#endif
+	    lt_settings->post_cursor2 == NULL)
+
+		return;
 
+	for (lane = 1; lane < LANE_COUNT_DP_MAX; lane++) {
+		if (lt_settings->voltage_swing)
+			lane_settings[lane].VOLTAGE_SWING = *lt_settings->voltage_swing;
+		if (lt_settings->pre_emphasis)
+			lane_settings[lane].PRE_EMPHASIS = *lt_settings->pre_emphasis;
+		if (lt_settings->post_cursor2)
+			lane_settings[lane].POST_CURSOR2 = *lt_settings->post_cursor2;
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+		if (lt_settings->ffe_preset)
+			lane_settings[lane].FFE_PRESET = *lt_settings->ffe_preset;
+#endif
+	}
 }
 
-enum dc_status dp_get_lane_status_and_drive_settings(
+enum dc_status dp_get_lane_status_and_lane_adjust(
 	struct dc_link *link,
 	const struct link_training_settings *link_training_setting,
-	union lane_status *ln_status,
-	union lane_align_status_updated *ln_status_updated,
-	struct link_training_settings *req_settings,
+	union lane_status ln_status[LANE_COUNT_DP_MAX],
+	union lane_align_status_updated *ln_align,
+	union lane_adjust ln_adjust[LANE_COUNT_DP_MAX],
 	uint32_t offset)
 {
 	unsigned int lane01_status_address = DP_LANE0_1_STATUS;
 	uint8_t lane_adjust_offset = 4;
 	unsigned int lane01_adjust_address;
 	uint8_t dpcd_buf[6] = {0};
-	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
-	struct link_training_settings request_settings = { {0} };
 	uint32_t lane;
 	enum dc_status status;
 
-	memset(req_settings, '\0', sizeof(struct link_training_settings));
-
 	if (is_repeater(link, offset)) {
 		lane01_status_address =
 				DP_LANE0_1_STATUS_PHY_REPEATER1 +
@@ -919,11 +896,11 @@ enum dc_status dp_get_lane_status_and_drive_settings(
 
 		ln_status[lane].raw =
 			get_nibble_at_index(&dpcd_buf[0], lane);
-		dpcd_lane_adjust[lane].raw =
+		ln_adjust[lane].raw =
 			get_nibble_at_index(&dpcd_buf[lane_adjust_offset], lane);
 	}
 
-	ln_status_updated->raw = dpcd_buf[2];
+	ln_align->raw = dpcd_buf[2];
 
 	if (is_repeater(link, offset)) {
 		DC_LOG_HW_LINK_TRAINING("%s:\n LTTPR Repeater ID: %d\n"
@@ -962,55 +939,6 @@ enum dc_status dp_get_lane_status_and_drive_settings(
 			dpcd_buf[lane_adjust_offset + 1]);
 	}
 
-	/*copy to req_settings*/
-	request_settings.link_settings.lane_count =
-		link_training_setting->link_settings.lane_count;
-	request_settings.link_settings.link_rate =
-		link_training_setting->link_settings.link_rate;
-	request_settings.link_settings.link_spread =
-		link_training_setting->link_settings.link_spread;
-
-	for (lane = 0; lane <
-		(uint32_t)(link_training_setting->link_settings.lane_count);
-		lane++) {
-
-#if defined(CONFIG_DRM_AMD_DC_DCN)
-		if (dp_get_link_encoding_format(&link_training_setting->link_settings) ==
-				DP_128b_132b_ENCODING) {
-			request_settings.lane_settings[lane].FFE_PRESET.raw =
-					dpcd_lane_adjust[lane].tx_ffe.PRESET_VALUE;
-		} else if (dp_get_link_encoding_format(&link_training_setting->link_settings) ==
-				DP_8b_10b_ENCODING) {
-			request_settings.lane_settings[lane].VOLTAGE_SWING =
-					(enum dc_voltage_swing)(dpcd_lane_adjust[lane].bits.
-							VOLTAGE_SWING_LANE);
-			request_settings.lane_settings[lane].PRE_EMPHASIS =
-					(enum dc_pre_emphasis)(dpcd_lane_adjust[lane].bits.
-							PRE_EMPHASIS_LANE);
-		}
-#else
-		request_settings.lane_settings[lane].VOLTAGE_SWING =
-			(enum dc_voltage_swing)(dpcd_lane_adjust[lane].bits.
-				VOLTAGE_SWING_LANE);
-		request_settings.lane_settings[lane].PRE_EMPHASIS =
-			(enum dc_pre_emphasis)(dpcd_lane_adjust[lane].bits.
-				PRE_EMPHASIS_LANE);
-#endif
-	}
-
-	/*Note: for postcursor2, read adjusted
-	 * postcursor2 settings from*/
-	/*DpcdAddress_AdjustRequestPostCursor2 =
-	 *0x020C (not implemented yet)*/
-
-	/* we find the maximum of the requested settings across all lanes*/
-	/* and set this maximum for all lanes*/
-	find_max_drive_settings(&request_settings, req_settings);
-
-	/* if post cursor 2 is needed in the future,
-	 * read DpcdAddress_AdjustRequestPostCursor2 = 0x020C
-	 */
-
 	return status;
 }
 
@@ -1019,7 +947,6 @@ enum dc_status dpcd_set_lane_settings(
 	const struct link_training_settings *link_training_setting,
 	uint32_t offset)
 {
-	union dpcd_training_lane dpcd_lane[LANE_COUNT_DP_MAX] = {{{0}}};
 	unsigned int lane0_set_address;
 	enum dc_status status;
 
@@ -1029,34 +956,11 @@ enum dc_status dpcd_set_lane_settings(
 		lane0_set_address = DP_TRAINING_LANE0_SET_PHY_REPEATER1 +
 		((DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE) * (offset - 1));
 
-	dp_hw_to_dpcd_lane_settings(link_training_setting,
-			link_training_setting->lane_settings,
-			dpcd_lane);
-
 	status = core_link_write_dpcd(link,
 		lane0_set_address,
-		(uint8_t *)(dpcd_lane),
+		(uint8_t *)(link_training_setting->dpcd_lane_settings),
 		link_training_setting->link_settings.lane_count);
 
-	/*
-	if (LTSettings.link.rate == LinkRate_High2)
-	{
-		DpcdTrainingLaneSet2 dpcd_lane2[lane_count_DPMax] = {0};
-		for ( uint32_t lane = 0;
-		lane < lane_count_DPMax; lane++)
-		{
-			dpcd_lane2[lane].bits.post_cursor2_set =
-			static_cast<unsigned char>(
-			LTSettings.laneSettings[lane].postCursor2);
-			dpcd_lane2[lane].bits.max_post_cursor2_reached = 0;
-		}
-		m_pDpcdAccessSrv->WriteDpcdData(
-		DpcdAddress_Lane0Set2,
-		reinterpret_cast<unsigned char*>(dpcd_lane2),
-		LTSettings.link.lanes);
-	}
-	*/
-
 	if (is_repeater(link, offset)) {
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 		if (dp_get_link_encoding_format(&link_training_setting->link_settings) ==
@@ -1066,7 +970,7 @@ enum dc_status dpcd_set_lane_settings(
 					__func__,
 					offset,
 					lane0_set_address,
-					dpcd_lane[0].tx_ffe.PRESET_VALUE);
+					link_training_setting->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE);
 		else if (dp_get_link_encoding_format(&link_training_setting->link_settings) ==
 				DP_8b_10b_ENCODING)
 #endif
@@ -1075,10 +979,10 @@ enum dc_status dpcd_set_lane_settings(
 			__func__,
 			offset,
 			lane0_set_address,
-			dpcd_lane[0].bits.VOLTAGE_SWING_SET,
-			dpcd_lane[0].bits.PRE_EMPHASIS_SET,
-			dpcd_lane[0].bits.MAX_SWING_REACHED,
-			dpcd_lane[0].bits.MAX_PRE_EMPHASIS_REACHED);
+			link_training_setting->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET,
+			link_training_setting->dpcd_lane_settings[0].bits.PRE_EMPHASIS_SET,
+			link_training_setting->dpcd_lane_settings[0].bits.MAX_SWING_REACHED,
+			link_training_setting->dpcd_lane_settings[0].bits.MAX_PRE_EMPHASIS_REACHED);
 
 	} else {
 #if defined(CONFIG_DRM_AMD_DC_DCN)
@@ -1087,17 +991,17 @@ enum dc_status dpcd_set_lane_settings(
 			DC_LOG_HW_LINK_TRAINING("%s:\n 0x%X TX_FFE_PRESET_VALUE = %x\n",
 					__func__,
 					lane0_set_address,
-					dpcd_lane[0].tx_ffe.PRESET_VALUE);
+					link_training_setting->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE);
 		else if (dp_get_link_encoding_format(&link_training_setting->link_settings) ==
 				DP_8b_10b_ENCODING)
 #endif
 		DC_LOG_HW_LINK_TRAINING("%s\n 0x%X VS set = %x  PE set = %x max VS Reached = %x  max PE Reached = %x\n",
 			__func__,
 			lane0_set_address,
-			dpcd_lane[0].bits.VOLTAGE_SWING_SET,
-			dpcd_lane[0].bits.PRE_EMPHASIS_SET,
-			dpcd_lane[0].bits.MAX_SWING_REACHED,
-			dpcd_lane[0].bits.MAX_PRE_EMPHASIS_REACHED);
+			link_training_setting->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET,
+			link_training_setting->dpcd_lane_settings[0].bits.PRE_EMPHASIS_SET,
+			link_training_setting->dpcd_lane_settings[0].bits.MAX_SWING_REACHED,
+			link_training_setting->dpcd_lane_settings[0].bits.MAX_PRE_EMPHASIS_REACHED);
 	}
 
 	return status;
@@ -1110,7 +1014,7 @@ bool dp_is_max_vs_reached(
 	for (lane = 0; lane <
 		(uint32_t)(lt_settings->link_settings.lane_count);
 		lane++) {
-		if (lt_settings->lane_settings[lane].VOLTAGE_SWING
+		if (lt_settings->dpcd_lane_settings[lane].bits.VOLTAGE_SWING_SET
 			== VOLTAGE_SWING_MAX_LEVEL)
 			return true;
 	}
@@ -1140,17 +1044,17 @@ static bool perform_post_lt_adj_req_sequence(
 			adj_req_timer < POST_LT_ADJ_REQ_TIMEOUT;
 			adj_req_timer++) {
 
-			struct link_training_settings req_settings;
 			union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX];
 			union lane_align_status_updated
 				dpcd_lane_status_updated;
+			union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 
-			dp_get_lane_status_and_drive_settings(
+			dp_get_lane_status_and_lane_adjust(
 				link,
 				lt_settings,
 				dpcd_lane_status,
 				&dpcd_lane_status_updated,
-				&req_settings,
+				dpcd_lane_adjust,
 				DPRX);
 
 			if (dpcd_lane_status_updated.bits.
@@ -1168,11 +1072,10 @@ static bool perform_post_lt_adj_req_sequence(
 			for (lane = 0; lane < (uint32_t)(lane_count); lane++) {
 
 				if (lt_settings->
-				lane_settings[lane].VOLTAGE_SWING !=
-				req_settings.lane_settings[lane].
-				VOLTAGE_SWING ||
-				lt_settings->lane_settings[lane].PRE_EMPHASIS !=
-				req_settings.lane_settings[lane].PRE_EMPHASIS) {
+				dpcd_lane_settings[lane].bits.VOLTAGE_SWING_SET !=
+				dpcd_lane_adjust[lane].bits.VOLTAGE_SWING_LANE ||
+				lt_settings->dpcd_lane_settings[lane].bits.PRE_EMPHASIS_SET !=
+				dpcd_lane_adjust[lane].bits.PRE_EMPHASIS_LANE) {
 
 					req_drv_setting_changed = true;
 					break;
@@ -1180,8 +1083,8 @@ static bool perform_post_lt_adj_req_sequence(
 			}
 
 			if (req_drv_setting_changed) {
-				dp_update_drive_settings(
-					lt_settings, req_settings);
+				dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+						lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 
 				dc_link_dp_set_drive_settings(link,
 						lt_settings);
@@ -1261,16 +1164,15 @@ static enum link_training_result perform_channel_equalization_sequence(
 	struct link_training_settings *lt_settings,
 	uint32_t offset)
 {
-	struct link_training_settings req_settings;
 	enum dc_dp_training_pattern tr_pattern;
 	uint32_t retries_ch_eq;
 	uint32_t wait_time_microsec;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
 	union lane_align_status_updated dpcd_lane_status_updated = {0};
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {0};
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = {0};
 
 	/* Note: also check that TPS4 is a supported feature*/
-
 	tr_pattern = lt_settings->pattern_for_eq;
 
 #if defined(CONFIG_DRM_AMD_DC_DCN)
@@ -1316,12 +1218,12 @@ static enum link_training_result perform_channel_equalization_sequence(
 		/* 4. Read lane status and requested
 		 * drive settings as set by the sink*/
 
-		dp_get_lane_status_and_drive_settings(
+		dp_get_lane_status_and_lane_adjust(
 			link,
 			lt_settings,
 			dpcd_lane_status,
 			&dpcd_lane_status_updated,
-			&req_settings,
+			dpcd_lane_adjust,
 			offset);
 
 		/* 5. check CR done*/
@@ -1335,7 +1237,8 @@ static enum link_training_result perform_channel_equalization_sequence(
 			return LINK_TRAINING_SUCCESS;
 
 		/* 7. update VS/PE/PC2 in lt_settings*/
-		dp_update_drive_settings(lt_settings, req_settings);
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+				lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 	}
 
 	return LINK_TRAINING_EQ_FAIL_EQ;
@@ -1361,10 +1264,10 @@ static enum link_training_result perform_clock_recovery_sequence(
 	uint32_t retries_cr;
 	uint32_t retry_count;
 	uint32_t wait_time_microsec;
-	struct link_training_settings req_settings;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX];
 	union lane_align_status_updated dpcd_lane_status_updated;
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 
 	retries_cr = 0;
 	retry_count = 0;
@@ -1418,12 +1321,12 @@ static enum link_training_result perform_clock_recovery_sequence(
 		/* 4. Read lane status and requested drive
 		* settings as set by the sink
 		*/
-		dp_get_lane_status_and_drive_settings(
+		dp_get_lane_status_and_lane_adjust(
 				link,
 				lt_settings,
 				dpcd_lane_status,
 				&dpcd_lane_status_updated,
-				&req_settings,
+				dpcd_lane_adjust,
 				offset);
 
 		/* 5. check CR done*/
@@ -1441,33 +1344,25 @@ static enum link_training_result perform_clock_recovery_sequence(
 			break;
 #endif
 
-#if defined(CONFIG_DRM_AMD_DC_DCN)
-		if ((dp_get_link_encoding_format(&lt_settings->link_settings) == DP_128b_132b_ENCODING) &&
-				lt_settings->lane_settings[0].FFE_PRESET.settings.level ==
-						req_settings.lane_settings[0].FFE_PRESET.settings.level)
-			retries_cr++;
-		else if ((dp_get_link_encoding_format(&lt_settings->link_settings) == DP_8b_10b_ENCODING) &&
-				lt_settings->lane_settings[0].VOLTAGE_SWING ==
-						req_settings.lane_settings[0].VOLTAGE_SWING)
-			retries_cr++;
-		else
-			retries_cr = 0;
-#else
 		/* 7. same lane settings*/
 		/* Note: settings are the same for all lanes,
 		 * so comparing first lane is sufficient*/
-		if ((lt_settings->lane_settings[0].VOLTAGE_SWING ==
-			req_settings.lane_settings[0].VOLTAGE_SWING)
-			&& (lt_settings->lane_settings[0].PRE_EMPHASIS ==
-				req_settings.lane_settings[0].PRE_EMPHASIS))
+		if ((dp_get_link_encoding_format(&lt_settings->link_settings) == DP_8b_10b_ENCODING) &&
+				lt_settings->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET ==
+						dpcd_lane_adjust[0].bits.VOLTAGE_SWING_LANE)
+			retries_cr++;
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+		else if ((dp_get_link_encoding_format(&lt_settings->link_settings) == DP_128b_132b_ENCODING) &&
+				lt_settings->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE ==
+						dpcd_lane_adjust[0].tx_ffe.PRESET_VALUE)
 			retries_cr++;
+#endif
 		else
 			retries_cr = 0;
-#endif
 
 		/* 8. update VS/PE/PC2 in lt_settings*/
-		dp_update_drive_settings(lt_settings, req_settings);
-
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+				lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 		retry_count++;
 	}
 
@@ -1596,6 +1491,9 @@ static inline void decide_8b_10b_training_settings(
 	lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_setting);
 	lt_settings->enhanced_framing = 1;
 	lt_settings->should_set_fec_ready = true;
+	lt_settings->disallow_per_lane_settings = true;
+	lt_settings->always_match_dpcd_with_hw_lane_settings = true;
+	dp_hw_to_dpcd_lane_settings(lt_settings, lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 }
 
 #if defined(CONFIG_DRM_AMD_DC_DCN)
@@ -1621,6 +1519,9 @@ static inline void decide_128b_132b_training_settings(struct dc_link *link,
 			link->dpcd_caps.lttpr_caps.phy_repeater_cnt) + 1) * 20000;
 	lt_settings->lttpr_mode = dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) ?
 			LTTPR_MODE_NON_TRANSPARENT : LTTPR_MODE_TRANSPARENT;
+	lt_settings->disallow_per_lane_settings = true;
+	dp_hw_to_dpcd_lane_settings(lt_settings,
+			lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 }
 #endif
 
@@ -1661,7 +1562,13 @@ static void override_training_settings(
 	if (overrides->ffe_preset != NULL)
 		lt_settings->ffe_preset = overrides->ffe_preset;
 #endif
-
+	/* Override HW lane settings with BIOS forced values if present */
+	if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN &&
+			link->lttpr_mode == LTTPR_MODE_TRANSPARENT) {
+		lt_settings->voltage_swing = &link->bios_forced_drive_settings.VOLTAGE_SWING;
+		lt_settings->pre_emphasis = &link->bios_forced_drive_settings.PRE_EMPHASIS;
+		lt_settings->always_match_dpcd_with_hw_lane_settings = false;
+	}
 	for (lane = 0; lane < LANE_COUNT_DP_MAX; lane++) {
 		lt_settings->lane_settings[lane].VOLTAGE_SWING =
 			lt_settings->voltage_swing != NULL ?
@@ -1677,6 +1584,9 @@ static void override_training_settings(
 			: POST_CURSOR2_DISABLED;
 	}
 
+	dp_hw_to_dpcd_lane_settings(lt_settings,
+			lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
+
 	/* Initialize training timings */
 	if (overrides->cr_pattern_time != NULL)
 		lt_settings->cr_pattern_time = *overrides->cr_pattern_time;
@@ -1947,6 +1857,9 @@ void dc_link_dp_set_drive_settings(
 	/* program ASIC PHY settings*/
 	dp_set_hw_lane_settings(link, lt_settings, DPRX);
 
+	dp_hw_to_dpcd_lane_settings(lt_settings,
+			lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
+
 	/* Notify DP sink the PHY settings from source */
 	dpcd_set_lane_settings(link, lt_settings, DPRX);
 }
@@ -2074,38 +1987,43 @@ static enum link_training_result dp_perform_128b_132b_channel_eq_done_sequence(
 		struct dc_link *link,
 		struct link_training_settings *lt_settings)
 {
-	uint8_t loop_count = 0;
+	uint8_t loop_count;
 	uint32_t aux_rd_interval = 0;
 	uint32_t wait_time = 0;
-	struct link_training_settings req_settings;
 	union lane_align_status_updated dpcd_lane_status_updated = {0};
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {0};
 	enum link_training_result status = LINK_TRAINING_SUCCESS;
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = {0};
 
-	/* Transmit 128b/132b_TPS1 over Main-Link and Set TRAINING_PATTERN_SET to 01h */
+	/* Transmit 128b/132b_TPS1 over Main-Link */
 	dp_set_hw_training_pattern(link, lt_settings->pattern_for_cr, DPRX);
+	/* Set TRAINING_PATTERN_SET to 01h */
 	dpcd_set_training_pattern(link, lt_settings->pattern_for_cr);
 
-	/* Adjust TX_FFE_PRESET_VALUE as requested */
-	dp_get_lane_status_and_drive_settings(link, lt_settings, dpcd_lane_status,
-			&dpcd_lane_status_updated, &req_settings, DPRX);
-	dp_update_drive_settings(lt_settings, req_settings);
+	/* Adjust TX_FFE_PRESET_VALUE and Transmit 128b/132b_TPS2 over Main-Link */
 	dpcd_128b_132b_get_aux_rd_interval(link, &aux_rd_interval);
+	dp_get_lane_status_and_lane_adjust(link, lt_settings, dpcd_lane_status,
+			&dpcd_lane_status_updated, dpcd_lane_adjust, DPRX);
+	dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+			lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 	dp_set_hw_lane_settings(link, lt_settings, DPRX);
-	dpcd_set_lane_settings(link, lt_settings, DPRX);
-
-	/* Transmit 128b/132b_TPS2 over Main-Link and Set TRAINING_PATTERN_SET to 02h */
 	dp_set_hw_training_pattern(link, lt_settings->pattern_for_eq, DPRX);
-	dpcd_set_training_pattern(link, lt_settings->pattern_for_eq);
+
+	/* Set loop counter to start from 1 */
+	loop_count = 1;
+
+	/* Set TRAINING_PATTERN_SET to 02h and TX_FFE_PRESET_VALUE in one AUX transaction */
+	dpcd_set_lt_pattern_and_lane_settings(link, lt_settings,
+			lt_settings->pattern_for_eq, DPRX);
 
 	/* poll for channel EQ done */
 	while (status == LINK_TRAINING_SUCCESS) {
-		loop_count++;
 		dp_wait_for_training_aux_rd_interval(link, aux_rd_interval);
 		wait_time += aux_rd_interval;
-		dp_get_lane_status_and_drive_settings(link, lt_settings, dpcd_lane_status,
-				&dpcd_lane_status_updated, &req_settings, DPRX);
-		dp_update_drive_settings(lt_settings, req_settings);
+		dp_get_lane_status_and_lane_adjust(link, lt_settings, dpcd_lane_status,
+				&dpcd_lane_status_updated, dpcd_lane_adjust, DPRX);
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+			lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 		dpcd_128b_132b_get_aux_rd_interval(link, &aux_rd_interval);
 		if (dp_is_ch_eq_done(lt_settings->link_settings.lane_count,
 				dpcd_lane_status)) {
@@ -2119,6 +2037,7 @@ static enum link_training_result dp_perform_128b_132b_channel_eq_done_sequence(
 			dp_set_hw_lane_settings(link, lt_settings, DPRX);
 			dpcd_set_lane_settings(link, lt_settings, DPRX);
 		}
+		loop_count++;
 	}
 
 	/* poll for EQ interlane align done */
@@ -2134,8 +2053,8 @@ static enum link_training_result dp_perform_128b_132b_channel_eq_done_sequence(
 			dp_wait_for_training_aux_rd_interval(link,
 					lt_settings->eq_pattern_time);
 			wait_time += lt_settings->eq_pattern_time;
-			dp_get_lane_status_and_drive_settings(link, lt_settings, dpcd_lane_status,
-					&dpcd_lane_status_updated, &req_settings, DPRX);
+			dp_get_lane_status_and_lane_adjust(link, lt_settings, dpcd_lane_status,
+					&dpcd_lane_status_updated, dpcd_lane_adjust, DPRX);
 		}
 	}
 
@@ -2148,9 +2067,9 @@ static enum link_training_result dp_perform_128b_132b_cds_done_sequence(
 {
 	/* Assumption: assume hardware has transmitted eq pattern */
 	enum link_training_result status = LINK_TRAINING_SUCCESS;
-	struct link_training_settings req_settings;
 	union lane_align_status_updated dpcd_lane_status_updated = {0};
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {0};
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 	uint32_t wait_time = 0;
 
 	/* initiate CDS done sequence */
@@ -2161,8 +2080,8 @@ static enum link_training_result dp_perform_128b_132b_cds_done_sequence(
 		dp_wait_for_training_aux_rd_interval(link,
 				lt_settings->cds_pattern_time);
 		wait_time += lt_settings->cds_pattern_time;
-		dp_get_lane_status_and_drive_settings(link, lt_settings, dpcd_lane_status,
-						&dpcd_lane_status_updated, &req_settings, DPRX);
+		dp_get_lane_status_and_lane_adjust(link, lt_settings, dpcd_lane_status,
+						&dpcd_lane_status_updated, dpcd_lane_adjust, DPRX);
 		if (dp_is_symbol_locked(lt_settings->link_settings.lane_count, dpcd_lane_status) &&
 				dpcd_lane_status_updated.bits.CDS_INTERLANE_ALIGN_DONE_128b_132b) {
 			/* pass */
@@ -2219,7 +2138,7 @@ static enum link_training_result dp_perform_8b_10b_link_training(
 		}
 
 		for (lane = 0; lane < (uint8_t)lt_settings->link_settings.lane_count; lane++)
-			lt_settings->lane_settings[lane].VOLTAGE_SWING = VOLTAGE_SWING_LEVEL0;
+			lt_settings->dpcd_lane_settings[lane].bits.VOLTAGE_SWING_SET = VOLTAGE_SWING_LEVEL0;
 	}
 
 	if (status == LINK_TRAINING_SUCCESS) {
@@ -3523,6 +3442,8 @@ static bool handle_hpd_irq_psr_sink(struct dc_link *link)
 		if (psr_error_status.bits.LINK_CRC_ERROR ||
 				psr_error_status.bits.RFB_STORAGE_ERROR ||
 				psr_error_status.bits.VSC_SDP_ERROR) {
+			bool allow_active;
+
 			/* Acknowledge and clear error bits */
 			dm_helpers_dp_write_dpcd(
 				link->ctx,
@@ -3532,8 +3453,10 @@ static bool handle_hpd_irq_psr_sink(struct dc_link *link)
 				sizeof(psr_error_status.raw));
 
 			/* PSR error, disable and re-enable PSR */
-			dc_link_set_psr_allow_active(link, false, true, false);
-			dc_link_set_psr_allow_active(link, true, true, false);
+			allow_active = false;
+			dc_link_set_psr_allow_active(link, &allow_active, true, false, NULL);
+			allow_active = true;
+			dc_link_set_psr_allow_active(link, &allow_active, true, false, NULL);
 
 			return true;
 		} else if (psr_sink_psr_status.bits.SINK_SELF_REFRESH_STATUS ==
@@ -3591,15 +3514,13 @@ static void dp_test_send_phy_test_pattern(struct dc_link *link)
 #endif
 	unsigned int test_pattern_size = 0;
 	enum dp_test_pattern test_pattern;
-	struct dc_link_training_settings link_settings;
 	union lane_adjust dpcd_lane_adjust;
 	unsigned int lane;
 	struct link_training_settings link_training_settings;
-	int i = 0;
 
 	dpcd_test_pattern.raw = 0;
 	memset(dpcd_lane_adjustment, 0, sizeof(dpcd_lane_adjustment));
-	memset(&link_settings, 0, sizeof(link_settings));
+	memset(&link_training_settings, 0, sizeof(link_training_settings));
 
 	/* get phy test pattern and pattern parameters from DP receiver */
 	core_link_read_dpcd(
@@ -3720,48 +3641,37 @@ static void dp_test_send_phy_test_pattern(struct dc_link *link)
 #endif
 
 	/* prepare link training settings */
-	link_settings.link = link->cur_link_settings;
+	link_training_settings.link_settings = link->cur_link_settings;
 
 	for (lane = 0; lane <
 		(unsigned int)(link->cur_link_settings.lane_count);
 		lane++) {
 		dpcd_lane_adjust.raw =
 			get_nibble_at_index(&dpcd_lane_adjustment[0].raw, lane);
-#if defined(CONFIG_DRM_AMD_DC_DCN)
 		if (dp_get_link_encoding_format(&link->cur_link_settings) ==
-				DP_128b_132b_ENCODING) {
-			link_settings.lane_settings[lane].FFE_PRESET.raw =
-					dpcd_lane_adjust.tx_ffe.PRESET_VALUE;
-		} else if (dp_get_link_encoding_format(&link->cur_link_settings) ==
 				DP_8b_10b_ENCODING) {
-			link_settings.lane_settings[lane].VOLTAGE_SWING =
+			link_training_settings.hw_lane_settings[lane].VOLTAGE_SWING =
 				(enum dc_voltage_swing)
 				(dpcd_lane_adjust.bits.VOLTAGE_SWING_LANE);
-			link_settings.lane_settings[lane].PRE_EMPHASIS =
+			link_training_settings.hw_lane_settings[lane].PRE_EMPHASIS =
 				(enum dc_pre_emphasis)
 				(dpcd_lane_adjust.bits.PRE_EMPHASIS_LANE);
-			link_settings.lane_settings[lane].POST_CURSOR2 =
+			link_training_settings.hw_lane_settings[lane].POST_CURSOR2 =
 				(enum dc_post_cursor2)
 				((dpcd_post_cursor_2_adjustment >> (lane * 2)) & 0x03);
 		}
-#else
-		link_settings.lane_settings[lane].VOLTAGE_SWING =
-			(enum dc_voltage_swing)
-			(dpcd_lane_adjust.bits.VOLTAGE_SWING_LANE);
-		link_settings.lane_settings[lane].PRE_EMPHASIS =
-			(enum dc_pre_emphasis)
-			(dpcd_lane_adjust.bits.PRE_EMPHASIS_LANE);
-		link_settings.lane_settings[lane].POST_CURSOR2 =
-			(enum dc_post_cursor2)
-			((dpcd_post_cursor_2_adjustment >> (lane * 2)) & 0x03);
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+		else if (dp_get_link_encoding_format(&link->cur_link_settings) ==
+				DP_128b_132b_ENCODING) {
+			link_training_settings.hw_lane_settings[lane].FFE_PRESET.raw =
+					dpcd_lane_adjust.tx_ffe.PRESET_VALUE;
+		}
 #endif
 	}
 
-	for (i = 0; i < 4; i++)
-		link_training_settings.lane_settings[i] =
-				link_settings.lane_settings[i];
-	link_training_settings.link_settings = link_settings.link;
-	link_training_settings.allow_invalid_msa_timing_param = false;
+	dp_hw_to_dpcd_lane_settings(&link_training_settings,
+			link_training_settings.hw_lane_settings,
+			link_training_settings.dpcd_lane_settings);
 	/*Usage: Measure DP physical lane signal
 	 * by DP SI test equipment automatically.
 	 * PHY test pattern request is generated by equipment via HPD interrupt.
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c
index e9006d099393..b1c9f77d6bf4 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c
@@ -263,10 +263,10 @@ static enum link_training_result dpia_training_cr_non_transparent(struct dc_link
 	uint32_t retry_count = 0;
 	/* From DP spec, CR read interval is always 100us. */
 	uint32_t wait_time_microsec = TRAINING_AUX_RD_INTERVAL;
-	struct link_training_settings req_settings;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = { { {0} } };
 	union lane_align_status_updated dpcd_lane_status_updated = { {0} };
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 	uint8_t set_cfg_data;
 	enum dpia_set_config_ts ts;
 
@@ -345,11 +345,12 @@ static enum link_training_result dpia_training_cr_non_transparent(struct dc_link
 		dp_wait_for_training_aux_rd_interval(link, wait_time_microsec);
 
 		/* Read status and adjustment requests from DPCD. */
-		status = dp_get_lane_status_and_drive_settings(link,
+		status = dp_get_lane_status_and_lane_adjust(
+				link,
 				lt_settings,
 				dpcd_lane_status,
 				&dpcd_lane_status_updated,
-				&req_settings,
+				dpcd_lane_adjust,
 				hop);
 		if (status != DC_OK) {
 			result = LINK_TRAINING_ABORT;
@@ -371,16 +372,18 @@ static enum link_training_result dpia_training_cr_non_transparent(struct dc_link
 		 * Note: settings are the same for all lanes,
 		 * so comparing first lane is sufficient.
 		 */
-		if (lt_settings->lane_settings[0].VOLTAGE_SWING ==
-			req_settings.lane_settings[0].VOLTAGE_SWING &&
-			lt_settings->lane_settings[0].PRE_EMPHASIS ==
-			req_settings.lane_settings[0].PRE_EMPHASIS)
+		if ((lt_settings->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET ==
+				dpcd_lane_adjust[0].bits.VOLTAGE_SWING_LANE)
+				&& (lt_settings->dpcd_lane_settings[0].bits.PRE_EMPHASIS_SET ==
+						dpcd_lane_adjust[0].bits.PRE_EMPHASIS_LANE))
 			retries_cr++;
 		else
 			retries_cr = 0;
 
 		/* Update VS/PE. */
-		dp_update_drive_settings(lt_settings, req_settings);
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+				lt_settings->lane_settings,
+				lt_settings->dpcd_lane_settings);
 		retry_count++;
 	}
 
@@ -416,10 +419,10 @@ static enum link_training_result dpia_training_cr_transparent(struct dc_link *li
 	uint32_t retries_cr = 0; /* Number of consecutive attempts with same VS or PE. */
 	uint32_t retry_count = 0;
 	uint32_t wait_time_microsec = lt_settings->cr_pattern_time;
-	struct link_training_settings req_settings;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = { { {0} } };
 	union lane_align_status_updated dpcd_lane_status_updated = { {0} };
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 
 	/* Cap of LINK_TRAINING_MAX_CR_RETRY attempts at clock recovery.
 	 * Fix inherited from perform_clock_recovery_sequence() -
@@ -445,11 +448,12 @@ static enum link_training_result dpia_training_cr_transparent(struct dc_link *li
 		dp_wait_for_training_aux_rd_interval(link, wait_time_microsec);
 
 		/* Read status and adjustment requests from DPCD. */
-		status = dp_get_lane_status_and_drive_settings(link,
+		status = dp_get_lane_status_and_lane_adjust(
+				link,
 				lt_settings,
 				dpcd_lane_status,
 				&dpcd_lane_status_updated,
-				&req_settings,
+				dpcd_lane_adjust,
 				DPRX);
 		if (status != DC_OK) {
 			result = LINK_TRAINING_ABORT;
@@ -471,16 +475,17 @@ static enum link_training_result dpia_training_cr_transparent(struct dc_link *li
 		 * Note: settings are the same for all lanes,
 		 * so comparing first lane is sufficient.
 		 */
-		if (lt_settings->lane_settings[0].VOLTAGE_SWING ==
-			req_settings.lane_settings[0].VOLTAGE_SWING &&
-			lt_settings->lane_settings[0].PRE_EMPHASIS ==
-			req_settings.lane_settings[0].PRE_EMPHASIS)
+		if ((lt_settings->dpcd_lane_settings[0].bits.VOLTAGE_SWING_SET ==
+				dpcd_lane_adjust[0].bits.VOLTAGE_SWING_LANE)
+				&& (lt_settings->dpcd_lane_settings[0].bits.PRE_EMPHASIS_SET ==
+						dpcd_lane_adjust[0].bits.PRE_EMPHASIS_LANE))
 			retries_cr++;
 		else
 			retries_cr = 0;
 
 		/* Update VS/PE. */
-		dp_update_drive_settings(lt_settings, req_settings);
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+				lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 		retry_count++;
 	}
 
@@ -566,10 +571,10 @@ static enum link_training_result dpia_training_eq_non_transparent(struct dc_link
 	enum dc_status status;
 	enum dc_dp_training_pattern tr_pattern;
 	uint32_t wait_time_microsec;
-	struct link_training_settings req_settings;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
 	union lane_align_status_updated dpcd_lane_status_updated = { {0} };
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = { { {0} } };
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 	uint8_t set_cfg_data;
 	enum dpia_set_config_ts ts;
 
@@ -639,11 +644,12 @@ static enum link_training_result dpia_training_eq_non_transparent(struct dc_link
 		dp_wait_for_training_aux_rd_interval(link, wait_time_microsec);
 
 		/* Read status and adjustment requests from DPCD. */
-		status = dp_get_lane_status_and_drive_settings(link,
+		status = dp_get_lane_status_and_lane_adjust(
+				link,
 				lt_settings,
 				dpcd_lane_status,
 				&dpcd_lane_status_updated,
-				&req_settings,
+				dpcd_lane_adjust,
 				hop);
 		if (status != DC_OK) {
 			result = LINK_TRAINING_ABORT;
@@ -664,7 +670,8 @@ static enum link_training_result dpia_training_eq_non_transparent(struct dc_link
 		}
 
 		/* Update VS/PE. */
-		dp_update_drive_settings(lt_settings, req_settings);
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+				lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 	}
 
 	/* Abort link training if equalization failed due to HPD unplug. */
@@ -701,10 +708,10 @@ static enum link_training_result dpia_training_eq_transparent(struct dc_link *li
 	enum dc_status status;
 	enum dc_dp_training_pattern tr_pattern = lt_settings->pattern_for_eq;
 	uint32_t wait_time_microsec;
-	struct link_training_settings req_settings;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
 	union lane_align_status_updated dpcd_lane_status_updated = { {0} };
 	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = { { {0} } };
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
 
 	wait_time_microsec = dpia_get_eq_aux_rd_interval(link, lt_settings, DPRX);
 
@@ -720,11 +727,12 @@ static enum link_training_result dpia_training_eq_transparent(struct dc_link *li
 		dp_wait_for_training_aux_rd_interval(link, wait_time_microsec);
 
 		/* Read status and adjustment requests from DPCD. */
-		status = dp_get_lane_status_and_drive_settings(link,
+		status = dp_get_lane_status_and_lane_adjust(
+				link,
 				lt_settings,
 				dpcd_lane_status,
 				&dpcd_lane_status_updated,
-				&req_settings,
+				dpcd_lane_adjust,
 				DPRX);
 		if (status != DC_OK) {
 			result = LINK_TRAINING_ABORT;
@@ -745,7 +753,8 @@ static enum link_training_result dpia_training_eq_transparent(struct dc_link *li
 		}
 
 		/* Update VS/PE. */
-		dp_update_drive_settings(lt_settings, req_settings);
+		dp_decide_lane_settings(lt_settings, dpcd_lane_adjust,
+				lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
 	}
 
 	/* Abort link training if equalization failed due to HPD unplug. */
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 9e83fd54e2ca..c32fdccd4d92 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1155,9 +1155,17 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 			pipe_ctx->plane_res.scl_data.recout.x += pipe_ctx->plane_res.scl_data.recout.width;
 	}
 
-	if (pipe_ctx->plane_res.scl_data.viewport.height < MIN_VIEWPORT_SIZE ||
-			pipe_ctx->plane_res.scl_data.viewport.width < MIN_VIEWPORT_SIZE)
-		res = false;
+	if (!pipe_ctx->stream->ctx->dc->config.enable_windowed_mpo_odm) {
+		if (pipe_ctx->plane_res.scl_data.viewport.height < MIN_VIEWPORT_SIZE ||
+				pipe_ctx->plane_res.scl_data.viewport.width < MIN_VIEWPORT_SIZE)
+			res = false;
+	} else {
+		/* Clamp minimum viewport size */
+		if (pipe_ctx->plane_res.scl_data.viewport.height < MIN_VIEWPORT_SIZE)
+			pipe_ctx->plane_res.scl_data.viewport.height = MIN_VIEWPORT_SIZE;
+		if (pipe_ctx->plane_res.scl_data.viewport.width < MIN_VIEWPORT_SIZE)
+			pipe_ctx->plane_res.scl_data.viewport.width = MIN_VIEWPORT_SIZE;
+	}
 
 	DC_LOG_SCALER("%s pipe %d:\nViewport: height:%d width:%d x:%d y:%d  Recout: height:%d width:%d x:%d y:%d  HACTIVE:%d VACTIVE:%d\n"
 			"src_rect: height:%d width:%d x:%d y:%d  dst_rect: height:%d width:%d x:%d y:%d  clip_rect: height:%d width:%d x:%d y:%d\n",
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index fc3f0fd1f068..a5339796902a 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -47,7 +47,7 @@ struct aux_payload;
 struct set_config_cmd_payload;
 struct dmub_notification;
 
-#define DC_VER "3.2.157"
+#define DC_VER "3.2.159"
 
 #define MAX_SURFACES 3
 #define MAX_PLANES 6
@@ -211,12 +211,12 @@ struct dc_dcc_setting {
 	unsigned int max_uncompressed_blk_size;
 	bool independent_64b_blks;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-	//These bitfields to be used starting with DCN 3.0
+	//These bitfields to be used starting with DCN
 	struct {
-		uint32_t dcc_256_64_64 : 1;//available in ASICs before DCN 3.0 (the worst compression case)
-		uint32_t dcc_128_128_uncontrained : 1;  //available in ASICs before DCN 3.0
-		uint32_t dcc_256_128_128 : 1;		//available starting with DCN 3.0
-		uint32_t dcc_256_256_unconstrained : 1;  //available in ASICs before DCN 3.0 (the best compression case)
+		uint32_t dcc_256_64_64 : 1;//available in ASICs before DCN (the worst compression case)
+		uint32_t dcc_128_128_uncontrained : 1;  //available in ASICs before DCN
+		uint32_t dcc_256_128_128 : 1;		//available starting with DCN
+		uint32_t dcc_256_256_unconstrained : 1;  //available in ASICs before DCN (the best compression case)
 	} dcc_controls;
 #endif
 };
@@ -323,6 +323,7 @@ struct dc_config {
 	bool multi_mon_pp_mclk_switch;
 	bool disable_dmcu;
 	bool enable_4to1MPC;
+	bool enable_windowed_mpo_odm;
 	bool allow_edp_hotplug_detection;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	bool clamp_min_dcfclk;
@@ -342,6 +343,12 @@ enum visual_confirm {
 	VISUAL_CONFIRM_SWIZZLE = 9,
 };
 
+enum dc_psr_power_opts {
+	psr_power_opt_invalid = 0x0,
+	psr_power_opt_smu_opt_static_screen = 0x1,
+	psr_power_opt_z10_static_screen = 0x10,
+};
+
 enum dcc_option {
 	DCC_ENABLE = 0,
 	DCC_DISABLE = 1,
@@ -725,6 +732,9 @@ struct dc {
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	bool idle_optimizations_allowed;
 #endif
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+	bool enable_c20_dtm_b0;
+#endif
 
 	/* Require to maintain clocks and bandwidth for UEFI enabled HW */
 
@@ -931,6 +941,7 @@ union surface_update_flags {
 		uint32_t bandwidth_change:1;
 		uint32_t clock_change:1;
 		uint32_t stereo_format_change:1;
+		uint32_t lut_3d:1;
 		uint32_t full_update:1;
 	} bits;
 
@@ -1417,6 +1428,12 @@ bool dc_process_dmub_set_config_async(struct dc *dc,
 				uint32_t link_index,
 				struct set_config_cmd_payload *payload,
 				struct dmub_notification *notify);
+
+enum dc_status dc_process_dmub_set_mst_slots(const struct dc *dc,
+				uint32_t link_index,
+				uint8_t mst_alloc_slots,
+				uint8_t *mst_slots_in_use);
+
 /*******************************************************************************
  * DSC Interfaces
  ******************************************************************************/
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
index a23937e1dc5c..bc87ea0adf94 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
@@ -174,11 +174,6 @@ struct dc_lane_settings {
 #endif
 };
 
-struct dc_link_training_settings {
-	struct dc_link_settings link;
-	struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX];
-};
-
 struct dc_link_training_overrides {
 	enum dc_voltage_swing *voltage_swing;
 	enum dc_pre_emphasis *pre_emphasis;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h
index 08815310d85b..180ecd860296 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_link.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_link.h
@@ -85,6 +85,7 @@ struct psr_settings {
 	 */
 	bool psr_frame_capture_indication_req;
 	unsigned int psr_sdp_transmit_line_num_deadline;
+	unsigned int psr_power_opt;
 };
 
 /*
@@ -123,6 +124,10 @@ struct dc_link {
 	struct dc_link_settings cur_link_settings;
 	struct dc_lane_settings cur_lane_setting[LANE_COUNT_DP_MAX];
 	struct dc_link_settings preferred_link_setting;
+	/* preferred_training_settings are override values that
+	 * come from DM. DM is responsible for the memory
+	 * management of the override pointers.
+	 */
 	struct dc_link_training_overrides preferred_training_settings;
 	struct dp_audio_test_data audio_test_data;
 
@@ -177,11 +182,15 @@ struct dc_link {
 
 	struct psr_settings psr_settings;
 
+	/* Drive settings read from integrated info table */
+	struct dc_lane_settings bios_forced_drive_settings;
+
 	/* MST record stream using this link */
 	struct link_flags {
 		bool dp_keep_receiver_powered;
 		bool dp_skip_DID2;
 		bool dp_skip_reset_segment;
+		bool dp_mot_reset_segment;
 	} wa_flags;
 	struct link_mst_stream_allocation_table mst_stream_alloc_table;
 
@@ -267,8 +276,8 @@ int dc_link_get_backlight_level(const struct dc_link *dc_link);
 
 int dc_link_get_target_backlight_pwm(const struct dc_link *link);
 
-bool dc_link_set_psr_allow_active(struct dc_link *dc_link, bool enable,
-		bool wait, bool force_static);
+bool dc_link_set_psr_allow_active(struct dc_link *dc_link, const bool *enable,
+		bool wait, bool force_static, const unsigned int *power_opts);
 
 bool dc_link_get_psr_state(const struct dc_link *dc_link, enum dc_psr_state *state);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
index 7866cf2a668f..27218ede150a 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
@@ -514,13 +514,15 @@ void dce_aud_az_configure(
 			union audio_sample_rates sample_rates =
 					audio_mode->sample_rates;
 			uint8_t byte2 = audio_mode->max_bit_rate;
+			uint8_t channel_count = audio_mode->channel_count;
 
 			/* adjust specific properties */
 			switch (audio_format_code) {
 			case AUDIO_FORMAT_CODE_LINEARPCM: {
+
 				check_audio_bandwidth(
 					crtc_info,
-					audio_mode->channel_count,
+					channel_count,
 					signal,
 					&sample_rates);
 
@@ -548,7 +550,7 @@ void dce_aud_az_configure(
 
 			/* fill audio format data */
 			set_reg_field_value(value,
-					audio_mode->channel_count - 1,
+					channel_count - 1,
 					AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0,
 					MAX_CHANNELS);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c
index 95cb4d7cc76a..6d42a9cc9916 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c
@@ -534,17 +534,26 @@ struct dce_aux *dce110_aux_engine_construct(struct aux_engine_dce110 *aux_engine
 static enum i2caux_transaction_action i2caux_action_from_payload(struct aux_payload *payload)
 {
 	if (payload->i2c_over_aux) {
+		if (payload->write_status_update) {
+			if (payload->mot)
+				return I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT;
+			else
+				return I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST;
+		}
 		if (payload->write) {
 			if (payload->mot)
 				return I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT;
-			return I2CAUX_TRANSACTION_ACTION_I2C_WRITE;
+			else
+				return I2CAUX_TRANSACTION_ACTION_I2C_WRITE;
 		}
 		if (payload->mot)
 			return I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT;
+
 		return I2CAUX_TRANSACTION_ACTION_I2C_READ;
 	}
 	if (payload->write)
 		return I2CAUX_TRANSACTION_ACTION_DP_WRITE;
+
 	return I2CAUX_TRANSACTION_ACTION_DP_READ;
 }
 
@@ -698,7 +707,8 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc,
 		aux_defer_retries = 0,
 		aux_i2c_defer_retries = 0,
 		aux_timeout_retries = 0,
-		aux_invalid_reply_retries = 0;
+		aux_invalid_reply_retries = 0,
+		aux_ack_m_retries = 0;
 
 	if (ddc_pin) {
 		aux_engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en];
@@ -758,9 +768,27 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc,
 									aux_defer_retries,
 									AUX_MAX_RETRIES);
 						goto fail;
-					} else {
+					} else 
 						udelay(300);
+				} else if (payload->write && ret > 0) {
+					/* sink requested more time to complete the write via AUX_ACKM */
+					if (++aux_ack_m_retries >= AUX_MAX_RETRIES) {
+						DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_ERROR,
+								LOG_FLAG_Error_I2cAux,
+								"dce_aux_transfer_with_retries: FAILURE: aux_ack_m_retries=%d >= AUX_MAX_RETRIES=%d",
+								aux_ack_m_retries,
+								AUX_MAX_RETRIES);
+						goto fail;
 					}
+
+					/* retry reading the write status until complete
+					 * NOTE: payload is modified here
+					 */
+					payload->write = false;
+					payload->write_status_update = true;
+					payload->length = 0;
+					udelay(300);
+
 				} else
 					return true;
 			break;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
index 296b2f80a1ec..989f5b6907e2 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
@@ -1151,7 +1151,8 @@ struct dce_hwseq_registers {
 	type DOMAIN_POWER_GATE;\
 	type DOMAIN_PGFSM_PWR_STATUS;\
 	type HPO_HDMISTREAMCLK_G_GATE_DIS;\
-	type DISABLE_HOSTVM_FORCE_ALLOW_PSTATE;
+	type DISABLE_HOSTVM_FORCE_ALLOW_PSTATE;\
+	type I2C_LIGHT_SLEEP_FORCE;
 
 struct dce_hwseq_shift {
 	HWSEQ_REG_FIELD_LIST(uint8_t)
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
index 05d96ca80512..90eb8eedacf2 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
@@ -227,6 +227,25 @@ static void dmub_psr_set_level(struct dmub_psr *dmub, uint16_t psr_level, uint8_
 	dc_dmub_srv_wait_idle(dc->dmub_srv);
 }
 
+/**
+ * Set PSR power optimization flags.
+ */
+static void dmub_psr_set_power_opt(struct dmub_psr *dmub, unsigned int power_opt)
+{
+	union dmub_rb_cmd cmd;
+	struct dc_context *dc = dmub->ctx;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.psr_set_power_opt.header.type = DMUB_CMD__PSR;
+	cmd.psr_set_power_opt.header.sub_type = DMUB_CMD__SET_PSR_POWER_OPT;
+	cmd.psr_set_power_opt.header.payload_bytes = sizeof(struct dmub_cmd_psr_set_power_opt_data);
+	cmd.psr_set_power_opt.psr_set_power_opt_data.power_opt = power_opt;
+
+	dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd);
+	dc_dmub_srv_cmd_execute(dc->dmub_srv);
+	dc_dmub_srv_wait_idle(dc->dmub_srv);
+}
+
 /*
  * Setup PSR by programming phy registers and sending psr hw context values to firmware.
  */
@@ -358,6 +377,7 @@ static const struct dmub_psr_funcs psr_funcs = {
 	.psr_set_level			= dmub_psr_set_level,
 	.psr_force_static		= dmub_psr_force_static,
 	.psr_get_residency		= dmub_psr_get_residency,
+	.psr_set_power_opt		= dmub_psr_set_power_opt,
 };
 
 /*
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.h b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.h
index 9675c269e649..5dbd479660f1 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.h
@@ -46,6 +46,7 @@ struct dmub_psr_funcs {
 	void (*psr_force_static)(struct dmub_psr *dmub, uint8_t panel_inst);
 	void (*psr_get_residency)(struct dmub_psr *dmub, uint32_t *residency,
 	uint8_t panel_inst);
+	void (*psr_set_power_opt)(struct dmub_psr *dmub, unsigned int power_opt);
 };
 
 struct dmub_psr *dmub_psr_create(struct dc_context *ctx);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
index cb9767ddf93d..44293d66b46b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
@@ -205,9 +205,17 @@ static void dpp1_power_on_dscl(
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
 
 	if (dpp->tf_regs->DSCL_MEM_PWR_CTRL) {
-		REG_UPDATE(DSCL_MEM_PWR_CTRL, LUT_MEM_PWR_FORCE, power_on ? 0 : 3);
-		if (power_on)
+		if (power_on) {
+			REG_UPDATE(DSCL_MEM_PWR_CTRL, LUT_MEM_PWR_FORCE, 0);
 			REG_WAIT(DSCL_MEM_PWR_STATUS, LUT_MEM_PWR_STATE, 0, 1, 5);
+		} else {
+			if (dpp->base.ctx->dc->debug.enable_mem_low_power.bits.dscl) {
+				dpp->base.ctx->dc->optimized_required = true;
+				dpp->base.deferred_reg_writes.bits.disable_dscl = true;
+			} else {
+				REG_UPDATE(DSCL_MEM_PWR_CTRL, LUT_MEM_PWR_FORCE, 3);
+			}
+		}
 	}
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 44d27579d898..a25732d07222 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -1378,6 +1378,12 @@ void dcn10_init_hw(struct dc *dc)
 	if (dc->clk_mgr && dc->clk_mgr->funcs->init_clocks)
 		dc->clk_mgr->funcs->init_clocks(dc->clk_mgr);
 
+	/* Align bw context with hw config when system resume. */
+	if (dc->clk_mgr->clks.dispclk_khz != 0 && dc->clk_mgr->clks.dppclk_khz != 0) {
+		dc->current_state->bw_ctx.bw.dcn.clk.dispclk_khz = dc->clk_mgr->clks.dispclk_khz;
+		dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz = dc->clk_mgr->clks.dppclk_khz;
+	}
+
 	// Initialize the dccg
 	if (dc->res_pool->dccg && dc->res_pool->dccg->funcs->dccg_init)
 		dc->res_pool->dccg->funcs->dccg_init(res_pool->dccg);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
index 756f5d411d9a..3883f918b3bb 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
@@ -3660,9 +3660,6 @@ static enum dml_project get_dml_project_version(uint32_t hw_internal_rev)
 	return DML_PROJECT_NAVI10v2;
 }
 
-#define fixed16_to_double(x) (((double) x) / ((double) (1 << 16)))
-#define fixed16_to_double_to_cpu(x) fixed16_to_double(le32_to_cpu(x))
-
 static bool init_soc_bounding_box(struct dc *dc,
 				  struct dcn20_resource_pool *pool)
 {
@@ -3698,16 +3695,22 @@ static bool init_soc_bounding_box(struct dc *dc,
 			clock_limits_available = (status == PP_SMU_RESULT_OK);
 		}
 
-		if (clock_limits_available && uclk_states_available && num_states)
+		if (clock_limits_available && uclk_states_available && num_states) {
+			DC_FP_START();
 			dcn20_update_bounding_box(dc, loaded_bb, &max_clocks, uclk_states, num_states);
-		else if (clock_limits_available)
+			DC_FP_END();
+		} else if (clock_limits_available) {
+			DC_FP_START();
 			dcn20_cap_soc_clocks(loaded_bb, max_clocks);
+			DC_FP_END();
+		}
 	}
 
 	loaded_ip->max_num_otg = pool->base.res_cap->num_timing_generator;
 	loaded_ip->max_num_dpp = pool->base.pipe_count;
+	DC_FP_START();
 	dcn20_patch_bounding_box(dc, loaded_bb);
-
+	DC_FP_END();
 	return true;
 }
 
@@ -3727,8 +3730,6 @@ static bool dcn20_resource_construct(
 	enum dml_project dml_project_version =
 			get_dml_project_version(ctx->asic_id.hw_internal_rev);
 
-	DC_FP_START();
-
 	ctx->dc_bios->regs = &bios_regs;
 	pool->base.funcs = &dcn20_res_pool_funcs;
 
@@ -4077,12 +4078,10 @@ static bool dcn20_resource_construct(
 		pool->base.oem_device = NULL;
 	}
 
-	DC_FP_END();
 	return true;
 
 create_fail:
 
-	DC_FP_END();
 	dcn20_resource_destruct(pool);
 
 	return false;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c
index ef5f6da5248a..c1d967ed6551 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c
@@ -494,6 +494,20 @@ void dpp3_deferred_update(
 	int bypass_state;
 	struct dcn3_dpp *dpp = TO_DCN30_DPP(dpp_base);
 
+	if (dpp_base->deferred_reg_writes.bits.disable_dscl) {
+		REG_UPDATE(DSCL_MEM_PWR_CTRL, LUT_MEM_PWR_FORCE, 3);
+		dpp_base->deferred_reg_writes.bits.disable_dscl = false;
+	}
+
+	if (dpp_base->deferred_reg_writes.bits.disable_gamcor) {
+		REG_GET(CM_GAMCOR_CONTROL, CM_GAMCOR_MODE_CURRENT, &bypass_state);
+		if (bypass_state == 0) {	// only program if bypass was latched
+			REG_UPDATE(CM_MEM_PWR_CTRL, GAMCOR_MEM_PWR_FORCE, 3);
+		} else
+			ASSERT(0); // LUT select was updated again before vupdate
+		dpp_base->deferred_reg_writes.bits.disable_gamcor = false;
+	}
+
 	if (dpp_base->deferred_reg_writes.bits.disable_blnd_lut) {
 		REG_GET(CM_BLNDGAM_CONTROL, CM_BLNDGAM_MODE_CURRENT, &bypass_state);
 		if (bypass_state == 0) {	// only program if bypass was latched
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp_cm.c
index 72c5687adc68..387eec616162 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp_cm.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp_cm.c
@@ -136,9 +136,13 @@ static void dpp3_power_on_gamcor_lut(
 	struct dcn3_dpp *dpp = TO_DCN30_DPP(dpp_base);
 
 	if (dpp_base->ctx->dc->debug.enable_mem_low_power.bits.cm) {
-		REG_UPDATE(CM_MEM_PWR_CTRL, GAMCOR_MEM_PWR_FORCE, power_on ? 0 : 3);
-		if (power_on)
+		if (power_on) {
+			REG_UPDATE(CM_MEM_PWR_CTRL, GAMCOR_MEM_PWR_FORCE, 0);
 			REG_WAIT(CM_MEM_PWR_STATUS, GAMCOR_MEM_PWR_STATE, 0, 1, 5);
+		} else {
+			dpp_base->ctx->dc->optimized_required = true;
+			dpp_base->deferred_reg_writes.bits.disable_gamcor = true;
+		}
 	} else
 		REG_SET(CM_MEM_PWR_CTRL, 0,
 				GAMCOR_MEM_PWR_DIS, power_on == true ? 0:1);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c
index 3a8a3214f770..e50c695e3c96 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c
@@ -1707,9 +1707,6 @@ bool dcn30_release_post_bldn_3dlut(
 	return ret;
 }
 
-#define fixed16_to_double(x) (((double) x) / ((double) (1 << 16)))
-#define fixed16_to_double_to_cpu(x) fixed16_to_double(le32_to_cpu(x))
-
 static bool is_soc_bounding_box_valid(struct dc *dc)
 {
 	uint32_t hw_internal_rev = dc->ctx->asic_id.hw_internal_rev;
@@ -1929,23 +1926,25 @@ noinline bool dcn30_internal_validate_bw(
 	if (vlevel == context->bw_ctx.dml.soc.num_states)
 		goto validate_fail;
 
-	for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *mpo_pipe = pipe->bottom_pipe;
+	if (!dc->config.enable_windowed_mpo_odm) {
+		for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
+			struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+			struct pipe_ctx *mpo_pipe = pipe->bottom_pipe;
 
-		if (!pipe->stream)
-			continue;
+			if (!pipe->stream)
+				continue;
 
-		/* We only support full screen mpo with ODM */
-		if (vba->ODMCombineEnabled[vba->pipe_plane[pipe_idx]] != dm_odm_combine_mode_disabled
-				&& pipe->plane_state && mpo_pipe
-				&& memcmp(&mpo_pipe->plane_res.scl_data.recout,
-						&pipe->plane_res.scl_data.recout,
-						sizeof(struct rect)) != 0) {
-			ASSERT(mpo_pipe->plane_state != pipe->plane_state);
-			goto validate_fail;
+			/* We only support full screen mpo with ODM */
+			if (vba->ODMCombineEnabled[vba->pipe_plane[pipe_idx]] != dm_odm_combine_mode_disabled
+					&& pipe->plane_state && mpo_pipe
+					&& memcmp(&mpo_pipe->plane_res.scl_data.recout,
+							&pipe->plane_res.scl_data.recout,
+							sizeof(struct rect)) != 0) {
+				ASSERT(mpo_pipe->plane_state != pipe->plane_state);
+				goto validate_fail;
+			}
+			pipe_idx++;
 		}
-		pipe_idx++;
 	}
 
 	/* merge pipes if necessary */
@@ -2323,7 +2322,9 @@ bool dcn30_validate_bandwidth(struct dc *dc,
 		goto validate_out;
 	}
 
+	DC_FP_START();
 	dc->res_pool->funcs->calculate_wm_and_dlg(dc, context, pipes, pipe_cnt, vlevel);
+	DC_FP_END();
 
 	BW_VAL_TRACE_END_WATERMARKS();
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/Makefile b/drivers/gpu/drm/amd/display/dc/dcn301/Makefile
index 09264716d1dc..7aa628c21973 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn301/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dcn301/Makefile
@@ -13,32 +13,6 @@
 DCN301 = dcn301_init.o dcn301_resource.o dcn301_dccg.o \
 		dcn301_dio_link_encoder.o dcn301_hwseq.o dcn301_panel_cntl.o dcn301_hubbub.o
 
-ifdef CONFIG_X86
-CFLAGS_$(AMDDALPATH)/dc/dcn301/dcn301_resource.o := -msse
-endif
-
-ifdef CONFIG_PPC64
-CFLAGS_$(AMDDALPATH)/dc/dcn301/dcn301_resource.o := -mhard-float -maltivec
-endif
-
-ifdef CONFIG_CC_IS_GCC
-ifeq ($(call cc-ifversion, -lt, 0701, y), y)
-IS_OLD_GCC = 1
-endif
-CFLAGS_$(AMDDALPATH)/dc/dcn301/dcn301_resource.o += -mhard-float
-endif
-
-ifdef CONFIG_X86
-ifdef IS_OLD_GCC
-# Stack alignment mismatch, proceed with caution.
-# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
-# (8B stack alignment).
-CFLAGS_$(AMDDALPATH)/dc/dcn301/dcn301_resource.o += -mpreferred-stack-boundary=4
-else
-CFLAGS_$(AMDDALPATH)/dc/dcn301/dcn301_resource.o += -msse2
-endif
-endif
-
 AMD_DAL_DCN301 = $(addprefix $(AMDDALPATH)/dc/dcn301/,$(DCN301))
 
 AMD_DISPLAY_FILES += $(AMD_DAL_DCN301)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c
index 5350c93d7772..fbaa03f26d8b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c
@@ -82,6 +82,7 @@
 #include "dce/dce_i2c.h"
 
 #include "dml/dcn30/display_mode_vba_30.h"
+#include "dml/dcn301/dcn301_fpu.h"
 #include "vm_helper.h"
 #include "dcn20/dcn20_vmid.h"
 #include "amdgpu_socbb.h"
@@ -91,184 +92,6 @@
 
 #define DC_LOGGER_INIT(logger)
 
-struct _vcs_dpi_ip_params_st dcn3_01_ip = {
-	.odm_capable = 1,
-	.gpuvm_enable = 1,
-	.hostvm_enable = 1,
-	.gpuvm_max_page_table_levels = 1,
-	.hostvm_max_page_table_levels = 2,
-	.hostvm_cached_page_table_levels = 0,
-	.pte_group_size_bytes = 2048,
-	.num_dsc = 3,
-	.rob_buffer_size_kbytes = 184,
-	.det_buffer_size_kbytes = 184,
-	.dpte_buffer_size_in_pte_reqs_luma = 64,
-	.dpte_buffer_size_in_pte_reqs_chroma = 32,
-	.pde_proc_buffer_size_64k_reqs = 48,
-	.dpp_output_buffer_pixels = 2560,
-	.opp_output_buffer_lines = 1,
-	.pixel_chunk_size_kbytes = 8,
-	.meta_chunk_size_kbytes = 2,
-	.writeback_chunk_size_kbytes = 8,
-	.line_buffer_size_bits = 789504,
-	.is_line_buffer_bpp_fixed = 0,  // ?
-	.line_buffer_fixed_bpp = 48,     // ?
-	.dcc_supported = true,
-	.writeback_interface_buffer_size_kbytes = 90,
-	.writeback_line_buffer_buffer_size = 656640,
-	.max_line_buffer_lines = 12,
-	.writeback_luma_buffer_size_kbytes = 12,  // writeback_line_buffer_buffer_size = 656640
-	.writeback_chroma_buffer_size_kbytes = 8,
-	.writeback_chroma_line_buffer_width_pixels = 4,
-	.writeback_max_hscl_ratio = 1,
-	.writeback_max_vscl_ratio = 1,
-	.writeback_min_hscl_ratio = 1,
-	.writeback_min_vscl_ratio = 1,
-	.writeback_max_hscl_taps = 1,
-	.writeback_max_vscl_taps = 1,
-	.writeback_line_buffer_luma_buffer_size = 0,
-	.writeback_line_buffer_chroma_buffer_size = 14643,
-	.cursor_buffer_size = 8,
-	.cursor_chunk_size = 2,
-	.max_num_otg = 4,
-	.max_num_dpp = 4,
-	.max_num_wb = 1,
-	.max_dchub_pscl_bw_pix_per_clk = 4,
-	.max_pscl_lb_bw_pix_per_clk = 2,
-	.max_lb_vscl_bw_pix_per_clk = 4,
-	.max_vscl_hscl_bw_pix_per_clk = 4,
-	.max_hscl_ratio = 6,
-	.max_vscl_ratio = 6,
-	.hscl_mults = 4,
-	.vscl_mults = 4,
-	.max_hscl_taps = 8,
-	.max_vscl_taps = 8,
-	.dispclk_ramp_margin_percent = 1,
-	.underscan_factor = 1.11,
-	.min_vblank_lines = 32,
-	.dppclk_delay_subtotal = 46,
-	.dynamic_metadata_vm_enabled = true,
-	.dppclk_delay_scl_lb_only = 16,
-	.dppclk_delay_scl = 50,
-	.dppclk_delay_cnvc_formatter = 27,
-	.dppclk_delay_cnvc_cursor = 6,
-	.dispclk_delay_subtotal = 119,
-	.dcfclk_cstate_latency = 5.2, // SRExitTime
-	.max_inter_dcn_tile_repeaters = 8,
-	.max_num_hdmi_frl_outputs = 0,
-	.odm_combine_4to1_supported = true,
-
-	.xfc_supported = false,
-	.xfc_fill_bw_overhead_percent = 10.0,
-	.xfc_fill_constant_bytes = 0,
-	.gfx7_compat_tiling_supported = 0,
-	.number_of_cursors = 1,
-};
-
-struct _vcs_dpi_soc_bounding_box_st dcn3_01_soc = {
-	.clock_limits = {
-			{
-				.state = 0,
-				.dram_speed_mts = 2400.0,
-				.fabricclk_mhz = 600,
-				.socclk_mhz = 278.0,
-				.dcfclk_mhz = 400.0,
-				.dscclk_mhz = 206.0,
-				.dppclk_mhz = 1015.0,
-				.dispclk_mhz = 1015.0,
-				.phyclk_mhz = 600.0,
-			},
-			{
-				.state = 1,
-				.dram_speed_mts = 2400.0,
-				.fabricclk_mhz = 688,
-				.socclk_mhz = 278.0,
-				.dcfclk_mhz = 400.0,
-				.dscclk_mhz = 206.0,
-				.dppclk_mhz = 1015.0,
-				.dispclk_mhz = 1015.0,
-				.phyclk_mhz = 600.0,
-			},
-			{
-				.state = 2,
-				.dram_speed_mts = 4267.0,
-				.fabricclk_mhz = 1067,
-				.socclk_mhz = 278.0,
-				.dcfclk_mhz = 608.0,
-				.dscclk_mhz = 296.0,
-				.dppclk_mhz = 1015.0,
-				.dispclk_mhz = 1015.0,
-				.phyclk_mhz = 810.0,
-			},
-
-			{
-				.state = 3,
-				.dram_speed_mts = 4267.0,
-				.fabricclk_mhz = 1067,
-				.socclk_mhz = 715.0,
-				.dcfclk_mhz = 676.0,
-				.dscclk_mhz = 338.0,
-				.dppclk_mhz = 1015.0,
-				.dispclk_mhz = 1015.0,
-				.phyclk_mhz = 810.0,
-			},
-
-			{
-				.state = 4,
-				.dram_speed_mts = 4267.0,
-				.fabricclk_mhz = 1067,
-				.socclk_mhz = 953.0,
-				.dcfclk_mhz = 810.0,
-				.dscclk_mhz = 338.0,
-				.dppclk_mhz = 1015.0,
-				.dispclk_mhz = 1015.0,
-				.phyclk_mhz = 810.0,
-			},
-		},
-
-	.sr_exit_time_us = 9.0,
-	.sr_enter_plus_exit_time_us = 11.0,
-	.urgent_latency_us = 4.0,
-	.urgent_latency_pixel_data_only_us = 4.0,
-	.urgent_latency_pixel_mixed_with_vm_data_us = 4.0,
-	.urgent_latency_vm_data_only_us = 4.0,
-	.urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096,
-	.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096,
-	.urgent_out_of_order_return_per_channel_vm_only_bytes = 4096,
-	.pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 80.0,
-	.pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 75.0,
-	.pct_ideal_dram_sdp_bw_after_urgent_vm_only = 40.0,
-	.max_avg_sdp_bw_use_normal_percent = 60.0,
-	.max_avg_dram_bw_use_normal_percent = 60.0,
-	.writeback_latency_us = 12.0,
-	.max_request_size_bytes = 256,
-	.dram_channel_width_bytes = 4,
-	.fabric_datapath_to_dcn_data_return_bytes = 32,
-	.dcn_downspread_percent = 0.5,
-	.downspread_percent = 0.38,
-	.dram_page_open_time_ns = 50.0,
-	.dram_rw_turnaround_time_ns = 17.5,
-	.dram_return_buffer_per_channel_bytes = 8192,
-	.round_trip_ping_latency_dcfclk_cycles = 191,
-	.urgent_out_of_order_return_per_channel_bytes = 4096,
-	.channel_interleave_bytes = 256,
-	.num_banks = 8,
-	.num_chans = 4,
-	.gpuvm_min_page_size_bytes = 4096,
-	.hostvm_min_page_size_bytes = 4096,
-	.dram_clock_change_latency_us = 23.84,
-	.writeback_dram_clock_change_latency_us = 23.0,
-	.return_bus_width_bytes = 64,
-	.dispclk_dppclk_vco_speed_mhz = 3550,
-	.xfc_bus_transport_time_us = 20,      // ?
-	.xfc_xbuf_latency_tolerance_us = 4,  // ?
-	.use_urgent_burst_bw = 1,            // ?
-	.num_states = 5,
-	.do_urgent_latency_adjustment = false,
-	.urgent_latency_adjustment_fabric_clock_component_us = 0,
-	.urgent_latency_adjustment_fabric_clock_reference_mhz = 0,
-};
-
 enum dcn301_clk_src_array_id {
 	DCN301_CLK_SRC_PLL0,
 	DCN301_CLK_SRC_PLL1,
@@ -1480,8 +1303,6 @@ static struct dc_cap_funcs cap_funcs = {
 	.get_dcc_compression_cap = dcn20_get_dcc_compression_cap
 };
 
-#define fixed16_to_double(x) (((double) x) / ((double) (1 << 16)))
-#define fixed16_to_double_to_cpu(x) fixed16_to_double(le32_to_cpu(x))
 
 static bool is_soc_bounding_box_valid(struct dc *dc)
 {
@@ -1508,26 +1329,24 @@ static bool init_soc_bounding_box(struct dc *dc,
 
 	loaded_ip->max_num_otg = pool->base.res_cap->num_timing_generator;
 	loaded_ip->max_num_dpp = pool->base.pipe_count;
+	DC_FP_START();
 	dcn20_patch_bounding_box(dc, loaded_bb);
+	DC_FP_END();
 
 	if (dc->ctx->dc_bios->funcs->get_soc_bb_info) {
 		struct bp_soc_bb_info bb_info = {0};
 
 		if (dc->ctx->dc_bios->funcs->get_soc_bb_info(dc->ctx->dc_bios, &bb_info) == BP_RESULT_OK) {
-			if (bb_info.dram_clock_change_latency_100ns > 0)
-				dcn3_01_soc.dram_clock_change_latency_us = bb_info.dram_clock_change_latency_100ns * 10;
-
-			if (bb_info.dram_sr_enter_exit_latency_100ns > 0)
-				dcn3_01_soc.sr_enter_plus_exit_time_us = bb_info.dram_sr_enter_exit_latency_100ns * 10;
-
-			if (bb_info.dram_sr_exit_latency_100ns > 0)
-				dcn3_01_soc.sr_exit_time_us = bb_info.dram_sr_exit_latency_100ns * 10;
+			DC_FP_START();
+			dcn301_fpu_init_soc_bounding_box(bb_info);
+			DC_FP_END();
 		}
 	}
 
 	return true;
 }
 
+
 static void set_wm_ranges(
 		struct pp_smu_funcs *pp_smu,
 		struct _vcs_dpi_soc_bounding_box_st *loaded_bb)
@@ -1550,9 +1369,9 @@ static void set_wm_ranges(
 			ranges.reader_wm_sets[i].wm_inst = i;
 			ranges.reader_wm_sets[i].min_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MIN;
 			ranges.reader_wm_sets[i].max_drain_clk_mhz = PP_SMU_WM_SET_RANGE_CLK_UNCONSTRAINED_MAX;
-			ranges.reader_wm_sets[i].min_fill_clk_mhz = (i > 0) ? (loaded_bb->clock_limits[i - 1].dram_speed_mts / 16) + 1 : 0;
-			ranges.reader_wm_sets[i].max_fill_clk_mhz = loaded_bb->clock_limits[i].dram_speed_mts / 16;
-
+			DC_FP_START();
+			dcn301_fpu_set_wm_ranges(i, &ranges, loaded_bb);
+			DC_FP_END();
 			ranges.num_reader_wm_sets = i + 1;
 		}
 
@@ -1572,154 +1391,6 @@ static void set_wm_ranges(
 	pp_smu->nv_funcs.set_wm_ranges(&pp_smu->nv_funcs.pp_smu, &ranges);
 }
 
-static void dcn301_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
-{
-	struct dcn301_resource_pool *pool = TO_DCN301_RES_POOL(dc->res_pool);
-	struct clk_limit_table *clk_table = &bw_params->clk_table;
-	struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES];
-	unsigned int i, closest_clk_lvl;
-	int j;
-
-	// Default clock levels are used for diags, which may lead to overclocking.
-	if (!IS_DIAG_DC(dc->ctx->dce_environment)) {
-		dcn3_01_ip.max_num_otg = pool->base.res_cap->num_timing_generator;
-		dcn3_01_ip.max_num_dpp = pool->base.pipe_count;
-		dcn3_01_soc.num_chans = bw_params->num_channels;
-
-		ASSERT(clk_table->num_entries);
-		for (i = 0; i < clk_table->num_entries; i++) {
-			/* loop backwards*/
-			for (closest_clk_lvl = 0, j = dcn3_01_soc.num_states - 1; j >= 0; j--) {
-				if ((unsigned int) dcn3_01_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) {
-					closest_clk_lvl = j;
-					break;
-				}
-			}
-
-			clock_limits[i].state = i;
-			clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz;
-			clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz;
-			clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz;
-			clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2;
-
-			clock_limits[i].dispclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dispclk_mhz;
-			clock_limits[i].dppclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dppclk_mhz;
-			clock_limits[i].dram_bw_per_chan_gbps = dcn3_01_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps;
-			clock_limits[i].dscclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dscclk_mhz;
-			clock_limits[i].dtbclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dtbclk_mhz;
-			clock_limits[i].phyclk_d18_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz;
-			clock_limits[i].phyclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].phyclk_mhz;
-		}
-		for (i = 0; i < clk_table->num_entries; i++)
-			dcn3_01_soc.clock_limits[i] = clock_limits[i];
-		if (clk_table->num_entries) {
-			dcn3_01_soc.num_states = clk_table->num_entries;
-			/* duplicate last level */
-			dcn3_01_soc.clock_limits[dcn3_01_soc.num_states] = dcn3_01_soc.clock_limits[dcn3_01_soc.num_states - 1];
-			dcn3_01_soc.clock_limits[dcn3_01_soc.num_states].state = dcn3_01_soc.num_states;
-		}
-	}
-
-	dcn3_01_soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
-	dc->dml.soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
-
-	dml_init_instance(&dc->dml, &dcn3_01_soc, &dcn3_01_ip, DML_PROJECT_DCN30);
-}
-
-static void calculate_wm_set_for_vlevel(
-		int vlevel,
-		struct wm_range_table_entry *table_entry,
-		struct dcn_watermarks *wm_set,
-		struct display_mode_lib *dml,
-		display_e2e_pipe_params_st *pipes,
-		int pipe_cnt)
-{
-	double dram_clock_change_latency_cached = dml->soc.dram_clock_change_latency_us;
-
-	ASSERT(vlevel < dml->soc.num_states);
-	/* only pipe 0 is read for voltage and dcf/soc clocks */
-	pipes[0].clks_cfg.voltage = vlevel;
-	pipes[0].clks_cfg.dcfclk_mhz = dml->soc.clock_limits[vlevel].dcfclk_mhz;
-	pipes[0].clks_cfg.socclk_mhz = dml->soc.clock_limits[vlevel].socclk_mhz;
-
-	dml->soc.dram_clock_change_latency_us = table_entry->pstate_latency_us;
-	dml->soc.sr_exit_time_us = table_entry->sr_exit_time_us;
-	dml->soc.sr_enter_plus_exit_time_us = table_entry->sr_enter_plus_exit_time_us;
-
-	wm_set->urgent_ns = get_wm_urgent(dml, pipes, pipe_cnt) * 1000;
-	wm_set->cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(dml, pipes, pipe_cnt) * 1000;
-	wm_set->cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(dml, pipes, pipe_cnt) * 1000;
-	wm_set->cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(dml, pipes, pipe_cnt) * 1000;
-	wm_set->pte_meta_urgent_ns = get_wm_memory_trip(dml, pipes, pipe_cnt) * 1000;
-	wm_set->frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(dml, pipes, pipe_cnt) * 1000;
-	wm_set->frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(dml, pipes, pipe_cnt) * 1000;
-	wm_set->urgent_latency_ns = get_urgent_latency(dml, pipes, pipe_cnt) * 1000;
-	dml->soc.dram_clock_change_latency_us = dram_clock_change_latency_cached;
-
-}
-
-static void dcn301_calculate_wm_and_dlg(
-		struct dc *dc, struct dc_state *context,
-		display_e2e_pipe_params_st *pipes,
-		int pipe_cnt,
-		int vlevel_req)
-{
-	int i, pipe_idx;
-	int vlevel, vlevel_max;
-	struct wm_range_table_entry *table_entry;
-	struct clk_bw_params *bw_params = dc->clk_mgr->bw_params;
-
-	ASSERT(bw_params);
-
-	vlevel_max = bw_params->clk_table.num_entries - 1;
-
-	/* WM Set D */
-	table_entry = &bw_params->wm_table.entries[WM_D];
-	if (table_entry->wm_type == WM_TYPE_RETRAINING)
-		vlevel = 0;
-	else
-		vlevel = vlevel_max;
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.d,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-	/* WM Set C */
-	table_entry = &bw_params->wm_table.entries[WM_C];
-	vlevel = min(max(vlevel_req, 2), vlevel_max);
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.c,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-	/* WM Set B */
-	table_entry = &bw_params->wm_table.entries[WM_B];
-	vlevel = min(max(vlevel_req, 1), vlevel_max);
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.b,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-
-	/* WM Set A */
-	table_entry = &bw_params->wm_table.entries[WM_A];
-	vlevel = min(vlevel_req, vlevel_max);
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.a,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-
-	for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
-		if (!context->res_ctx.pipe_ctx[i].stream)
-			continue;
-
-		pipes[pipe_idx].clks_cfg.dispclk_mhz = get_dispclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt);
-		pipes[pipe_idx].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
-
-		if (dc->config.forced_clocks) {
-			pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dispclk_mhz;
-			pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz;
-		}
-		if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz * 1000)
-			pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz / 1000.0;
-		if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000)
-			pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz / 1000.0;
-
-		pipe_idx++;
-	}
-
-	dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
-}
-
 static struct resource_funcs dcn301_res_pool_funcs = {
 	.destroy = dcn301_destroy_resource_pool,
 	.link_enc_create = dcn301_link_encoder_create,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.h b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.h
index 17e4e91ff4b8..ae8672680cdd 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.h
@@ -32,6 +32,9 @@ struct dc;
 struct resource_pool;
 struct _vcs_dpi_display_pipe_params_st;
 
+extern struct _vcs_dpi_ip_params_st dcn3_01_ip;
+extern struct _vcs_dpi_soc_bounding_box_st dcn3_01_soc;
+
 struct dcn301_resource_pool {
 	struct resource_pool base;
 };
diff --git a/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c b/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c
index 2ce6eae7535d..4a9b64023675 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c
@@ -1344,6 +1344,20 @@ void dcn303_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param
 			dcn3_03_soc.clock_limits[i].phyclk_d18_mhz = dcn3_03_soc.clock_limits[0].phyclk_d18_mhz;
 			dcn3_03_soc.clock_limits[i].dscclk_mhz = dcn3_03_soc.clock_limits[0].dscclk_mhz;
 		}
+
+		// WA: patch strobe modes to compensate for DCN303 BW issue
+		if (dcn3_03_soc.num_chans <= 4) {
+			for (i = 0; i < dcn3_03_soc.num_states; i++) {
+				if (dcn3_03_soc.clock_limits[i].dram_speed_mts > 1700)
+					break;
+
+				if (dcn3_03_soc.clock_limits[i].dram_speed_mts >= 1500) {
+					dcn3_03_soc.clock_limits[i].dcfclk_mhz = 100;
+					dcn3_03_soc.clock_limits[i].fabricclk_mhz = 100;
+				}
+			}
+		}
+
 		/* re-init DML with updated bb */
 		dml_init_instance(&dc->dml, &dcn3_03_soc, &dcn3_03_ip, DML_PROJECT_DCN30);
 		if (dc->current_state)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c
index 8f8eee475144..ee6f13bef377 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c
@@ -431,7 +431,7 @@ void dcn31_link_encoder_enable_dp_output(
 
 		if (link) {
 			dpia_control.dpia_id = link->ddc_hw_inst;
-			dpia_control.fec_rdy = link->fec_state == dc_link_fec_ready ? 1 : 0;
+			dpia_control.fec_rdy = dc_link_should_enable_fec(link);
 		} else {
 			DC_LOG_ERROR("%s: Failed to execute DPIA enable DMUB command.\n", __func__);
 			BREAK_TO_DEBUGGER();
@@ -476,7 +476,7 @@ void dcn31_link_encoder_enable_dp_mst_output(
 
 		if (link) {
 			dpia_control.dpia_id = link->ddc_hw_inst;
-			dpia_control.fec_rdy = link->fec_state == dc_link_fec_ready ? 1 : 0;
+			dpia_control.fec_rdy = dc_link_should_enable_fec(link);
 		} else {
 			DC_LOG_ERROR("%s: Failed to execute DPIA enable DMUB command.\n", __func__);
 			BREAK_TO_DEBUGGER();
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
index 9a6ad1cebc85..d24ad7754d71 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
@@ -50,6 +50,7 @@
 #include "dcn10/dcn10_hw_sequencer.h"
 #include "inc/link_enc_cfg.h"
 #include "dcn30/dcn30_vpg.h"
+#include "dce/dce_i2c_hw.h"
 
 #define DC_LOGGER_INIT(logger)
 
@@ -259,6 +260,10 @@ void dcn31_init_hw(struct dc *dc)
 	/* power AFMT HDMI memory TODO: may move to dis/en output save power*/
 	REG_WRITE(DIO_MEM_PWR_CTRL, 0);
 
+	// Set i2c to light sleep until engine is setup
+	if (dc->debug.enable_mem_low_power.bits.i2c)
+		REG_UPDATE(DIO_MEM_PWR_CTRL, I2C_LIGHT_SLEEP_FORCE, 1);
+
 	if (!dc->debug.disable_clock_gate) {
 		/* enable all DCN clock gating */
 		REG_WRITE(DCCG_GATE_DISABLE_CNTL, 0);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c
index 2b4459909648..87b2c2428842 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c
@@ -101,8 +101,6 @@
 #include "link_enc_cfg.h"
 
 #define DC_LOGGER_INIT(logger)
-#define fixed16_to_double(x) (((double) x) / ((double) (1 << 16)))
-#define fixed16_to_double_to_cpu(x) fixed16_to_double(le32_to_cpu(x))
 
 #define DCN3_1_DEFAULT_DET_SIZE 384
 
@@ -899,7 +897,8 @@ static const struct dce_hwseq_registers hwseq_reg = {
 	HWS_SF(, DMU_MEM_PWR_CNTL, DMCU_ERAM_MEM_PWR_FORCE, mask_sh), \
 	HWS_SF(, ODM_MEM_PWR_CTRL3, ODM_MEM_UNASSIGNED_PWR_MODE, mask_sh), \
 	HWS_SF(, ODM_MEM_PWR_CTRL3, ODM_MEM_VBLANK_PWR_MODE, mask_sh), \
-	HWS_SF(, MMHUBBUB_MEM_PWR_CNTL, VGA_MEM_PWR_FORCE, mask_sh)
+	HWS_SF(, MMHUBBUB_MEM_PWR_CNTL, VGA_MEM_PWR_FORCE, mask_sh), \
+	HWS_SF(, DIO_MEM_PWR_CTRL, I2C_LIGHT_SLEEP_FORCE, mask_sh)
 
 static const struct dce_hwseq_shift hwseq_shift = {
 		HWSEQ_DCN31_MASK_SH_LIST(__SHIFT)
@@ -1833,7 +1832,7 @@ static int dcn31_populate_dml_pipes_from_context(
 	return pipe_cnt;
 }
 
-static void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context)
+void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context)
 {
 	if (dc->clk_mgr->bw_params->wm_table.entries[WM_A].valid) {
 		context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.entries[WM_A].pstate_latency_us;
@@ -1977,7 +1976,7 @@ static void dcn31_calculate_wm_and_dlg_fp(
 	dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
 }
 
-static void dcn31_calculate_wm_and_dlg(
+void dcn31_calculate_wm_and_dlg(
 		struct dc *dc, struct dc_state *context,
 		display_e2e_pipe_params_st *pipes,
 		int pipe_cnt,
@@ -2459,6 +2458,8 @@ static bool dcn31_resource_construct(
 
 	dc->cap_funcs = cap_funcs;
 
+	dc->dcn_ip->max_num_dpp = dcn3_1_ip.max_num_dpp;
+
 	DC_FP_END();
 
 	return true;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h
index 93571c976996..416fe7a721d8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h
@@ -35,6 +35,16 @@ struct dcn31_resource_pool {
 	struct resource_pool base;
 };
 
+bool dcn31_validate_bandwidth(struct dc *dc,
+		struct dc_state *context,
+		bool fast_validate);
+void dcn31_calculate_wm_and_dlg(
+		struct dc *dc, struct dc_state *context,
+		display_e2e_pipe_params_st *pipes,
+		int pipe_cnt,
+		int vlevel);
+void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context);
+
 struct resource_pool *dcn31_create_resource_pool(
 		const struct dc_init_data *init_data,
 		struct dc *dc);
diff --git a/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h b/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h
index 43f33e186088..511f9e1159c7 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_cp_psp.h
@@ -35,6 +35,8 @@ struct cp_psp_stream_config {
 	uint8_t link_enc_idx;
 	uint8_t stream_enc_idx;
 	uint8_t phy_idx;
+	uint8_t dio_output_idx;
+	uint8_t dio_output_type;
 	uint8_t assr_enabled;
 	uint8_t mst_enabled;
 	uint8_t dp2_enabled;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile
index 169a4e68f86e..eee6672bd32d 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
@@ -70,6 +70,8 @@ CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_mode_vba_30.o := $(dml_ccflags) $(fram
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_rq_dlg_calc_30.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_mode_vba_31.o := $(dml_ccflags) $(frame_warn_flag)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_rq_dlg_calc_31.o := $(dml_ccflags)
+CFLAGS_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_ccflags)
+CFLAGS_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags)
 CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_rcflags)
 CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn2x/dcn2x.o := $(dml_rcflags)
@@ -83,7 +85,9 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/display_mode_vba_30.o := $(dml_rcflags)
 CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/display_rq_dlg_calc_30.o := $(dml_rcflags)
 CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/display_mode_vba_31.o := $(dml_rcflags)
 CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/display_rq_dlg_calc_31.o := $(dml_rcflags)
+CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_rcflags)
 CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_rcflags)
+CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o  := $(dml_rcflags)
 endif
 CFLAGS_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/display_rq_dlg_helpers.o := $(dml_ccflags)
@@ -99,6 +103,8 @@ DML += dcn20/display_rq_dlg_calc_20v2.o dcn20/display_mode_vba_20v2.o
 DML += dcn21/display_rq_dlg_calc_21.o dcn21/display_mode_vba_21.o
 DML += dcn30/display_mode_vba_30.o dcn30/display_rq_dlg_calc_30.o
 DML += dcn31/display_mode_vba_31.o dcn31/display_rq_dlg_calc_31.o
+DML += dcn301/dcn301_fpu.o
+DML += dsc/rc_calc_fpu.o
 endif
 
 AMD_DAL_DML = $(addprefix $(AMDDALPATH)/dc/dml/,$(DML))
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.c
new file mode 100644
index 000000000000..94c32832a0e7
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.c
@@ -0,0 +1,390 @@
+/*
+ * Copyright 2019-2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+#include "resource.h"
+#include "clk_mgr.h"
+#include "dcn20/dcn20_resource.h"
+#include "dcn301/dcn301_resource.h"
+
+#include "dml/dcn20/dcn20_fpu.h"
+#include "dcn301_fpu.h"
+
+#define TO_DCN301_RES_POOL(pool)\
+	container_of(pool, struct dcn301_resource_pool, base)
+
+/* Based on: //vidip/dc/dcn3/doc/architecture/DCN3x_Display_Mode.xlsm#83 */
+struct _vcs_dpi_ip_params_st dcn3_01_ip = {
+	.odm_capable = 1,
+	.gpuvm_enable = 1,
+	.hostvm_enable = 1,
+	.gpuvm_max_page_table_levels = 1,
+	.hostvm_max_page_table_levels = 2,
+	.hostvm_cached_page_table_levels = 0,
+	.pte_group_size_bytes = 2048,
+	.num_dsc = 3,
+	.rob_buffer_size_kbytes = 184,
+	.det_buffer_size_kbytes = 184,
+	.dpte_buffer_size_in_pte_reqs_luma = 64,
+	.dpte_buffer_size_in_pte_reqs_chroma = 32,
+	.pde_proc_buffer_size_64k_reqs = 48,
+	.dpp_output_buffer_pixels = 2560,
+	.opp_output_buffer_lines = 1,
+	.pixel_chunk_size_kbytes = 8,
+	.meta_chunk_size_kbytes = 2,
+	.writeback_chunk_size_kbytes = 8,
+	.line_buffer_size_bits = 789504,
+	.is_line_buffer_bpp_fixed = 0,  // ?
+	.line_buffer_fixed_bpp = 48,     // ?
+	.dcc_supported = true,
+	.writeback_interface_buffer_size_kbytes = 90,
+	.writeback_line_buffer_buffer_size = 656640,
+	.max_line_buffer_lines = 12,
+	.writeback_luma_buffer_size_kbytes = 12,  // writeback_line_buffer_buffer_size = 656640
+	.writeback_chroma_buffer_size_kbytes = 8,
+	.writeback_chroma_line_buffer_width_pixels = 4,
+	.writeback_max_hscl_ratio = 1,
+	.writeback_max_vscl_ratio = 1,
+	.writeback_min_hscl_ratio = 1,
+	.writeback_min_vscl_ratio = 1,
+	.writeback_max_hscl_taps = 1,
+	.writeback_max_vscl_taps = 1,
+	.writeback_line_buffer_luma_buffer_size = 0,
+	.writeback_line_buffer_chroma_buffer_size = 14643,
+	.cursor_buffer_size = 8,
+	.cursor_chunk_size = 2,
+	.max_num_otg = 4,
+	.max_num_dpp = 4,
+	.max_num_wb = 1,
+	.max_dchub_pscl_bw_pix_per_clk = 4,
+	.max_pscl_lb_bw_pix_per_clk = 2,
+	.max_lb_vscl_bw_pix_per_clk = 4,
+	.max_vscl_hscl_bw_pix_per_clk = 4,
+	.max_hscl_ratio = 6,
+	.max_vscl_ratio = 6,
+	.hscl_mults = 4,
+	.vscl_mults = 4,
+	.max_hscl_taps = 8,
+	.max_vscl_taps = 8,
+	.dispclk_ramp_margin_percent = 1,
+	.underscan_factor = 1.11,
+	.min_vblank_lines = 32,
+	.dppclk_delay_subtotal = 46,
+	.dynamic_metadata_vm_enabled = true,
+	.dppclk_delay_scl_lb_only = 16,
+	.dppclk_delay_scl = 50,
+	.dppclk_delay_cnvc_formatter = 27,
+	.dppclk_delay_cnvc_cursor = 6,
+	.dispclk_delay_subtotal = 119,
+	.dcfclk_cstate_latency = 5.2, // SRExitTime
+	.max_inter_dcn_tile_repeaters = 8,
+	.max_num_hdmi_frl_outputs = 0,
+	.odm_combine_4to1_supported = true,
+
+	.xfc_supported = false,
+	.xfc_fill_bw_overhead_percent = 10.0,
+	.xfc_fill_constant_bytes = 0,
+	.gfx7_compat_tiling_supported = 0,
+	.number_of_cursors = 1,
+};
+
+struct _vcs_dpi_soc_bounding_box_st dcn3_01_soc = {
+	.clock_limits = {
+		{
+			.state = 0,
+			.dram_speed_mts = 2400.0,
+			.fabricclk_mhz = 600,
+			.socclk_mhz = 278.0,
+			.dcfclk_mhz = 400.0,
+			.dscclk_mhz = 206.0,
+			.dppclk_mhz = 1015.0,
+			.dispclk_mhz = 1015.0,
+			.phyclk_mhz = 600.0,
+		},
+
+		{
+			.state = 1,
+			.dram_speed_mts = 2400.0,
+			.fabricclk_mhz = 688,
+			.socclk_mhz = 278.0,
+			.dcfclk_mhz = 400.0,
+			.dscclk_mhz = 206.0,
+			.dppclk_mhz = 1015.0,
+			.dispclk_mhz = 1015.0,
+			.phyclk_mhz = 600.0,
+		},
+
+		{
+			.state = 2,
+			.dram_speed_mts = 4267.0,
+			.fabricclk_mhz = 1067,
+			.socclk_mhz = 278.0,
+			.dcfclk_mhz = 608.0,
+			.dscclk_mhz = 296.0,
+			.dppclk_mhz = 1015.0,
+			.dispclk_mhz = 1015.0,
+			.phyclk_mhz = 810.0,
+		},
+
+		{
+			.state = 3,
+			.dram_speed_mts = 4267.0,
+			.fabricclk_mhz = 1067,
+			.socclk_mhz = 715.0,
+			.dcfclk_mhz = 676.0,
+			.dscclk_mhz = 338.0,
+			.dppclk_mhz = 1015.0,
+			.dispclk_mhz = 1015.0,
+			.phyclk_mhz = 810.0,
+		},
+
+		{
+			.state = 4,
+			.dram_speed_mts = 4267.0,
+			.fabricclk_mhz = 1067,
+			.socclk_mhz = 953.0,
+			.dcfclk_mhz = 810.0,
+			.dscclk_mhz = 338.0,
+			.dppclk_mhz = 1015.0,
+			.dispclk_mhz = 1015.0,
+			.phyclk_mhz = 810.0,
+		},
+	},
+
+	.sr_exit_time_us = 9.0,
+	.sr_enter_plus_exit_time_us = 11.0,
+	.urgent_latency_us = 4.0,
+	.urgent_latency_pixel_data_only_us = 4.0,
+	.urgent_latency_pixel_mixed_with_vm_data_us = 4.0,
+	.urgent_latency_vm_data_only_us = 4.0,
+	.urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096,
+	.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096,
+	.urgent_out_of_order_return_per_channel_vm_only_bytes = 4096,
+	.pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 80.0,
+	.pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 75.0,
+	.pct_ideal_dram_sdp_bw_after_urgent_vm_only = 40.0,
+	.max_avg_sdp_bw_use_normal_percent = 60.0,
+	.max_avg_dram_bw_use_normal_percent = 60.0,
+	.writeback_latency_us = 12.0,
+	.max_request_size_bytes = 256,
+	.dram_channel_width_bytes = 4,
+	.fabric_datapath_to_dcn_data_return_bytes = 32,
+	.dcn_downspread_percent = 0.5,
+	.downspread_percent = 0.38,
+	.dram_page_open_time_ns = 50.0,
+	.dram_rw_turnaround_time_ns = 17.5,
+	.dram_return_buffer_per_channel_bytes = 8192,
+	.round_trip_ping_latency_dcfclk_cycles = 191,
+	.urgent_out_of_order_return_per_channel_bytes = 4096,
+	.channel_interleave_bytes = 256,
+	.num_banks = 8,
+	.num_chans = 4,
+	.gpuvm_min_page_size_bytes = 4096,
+	.hostvm_min_page_size_bytes = 4096,
+	.dram_clock_change_latency_us = 23.84,
+	.writeback_dram_clock_change_latency_us = 23.0,
+	.return_bus_width_bytes = 64,
+	.dispclk_dppclk_vco_speed_mhz = 3550,
+	.xfc_bus_transport_time_us = 20,      // ?
+	.xfc_xbuf_latency_tolerance_us = 4,  // ?
+	.use_urgent_burst_bw = 1,            // ?
+	.num_states = 5,
+	.do_urgent_latency_adjustment = false,
+	.urgent_latency_adjustment_fabric_clock_component_us = 0,
+	.urgent_latency_adjustment_fabric_clock_reference_mhz = 0,
+};
+
+static void calculate_wm_set_for_vlevel(int vlevel,
+		struct wm_range_table_entry *table_entry,
+		struct dcn_watermarks *wm_set,
+		struct display_mode_lib *dml,
+		display_e2e_pipe_params_st *pipes,
+		int pipe_cnt)
+{
+	double dram_clock_change_latency_cached = dml->soc.dram_clock_change_latency_us;
+
+	ASSERT(vlevel < dml->soc.num_states);
+	/* only pipe 0 is read for voltage and dcf/soc clocks */
+	pipes[0].clks_cfg.voltage = vlevel;
+	pipes[0].clks_cfg.dcfclk_mhz = dml->soc.clock_limits[vlevel].dcfclk_mhz;
+	pipes[0].clks_cfg.socclk_mhz = dml->soc.clock_limits[vlevel].socclk_mhz;
+
+	dml->soc.dram_clock_change_latency_us = table_entry->pstate_latency_us;
+	dml->soc.sr_exit_time_us = table_entry->sr_exit_time_us;
+	dml->soc.sr_enter_plus_exit_time_us = table_entry->sr_enter_plus_exit_time_us;
+
+	wm_set->urgent_ns = get_wm_urgent(dml, pipes, pipe_cnt) * 1000;
+	wm_set->cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(dml, pipes, pipe_cnt) * 1000;
+	wm_set->cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(dml, pipes, pipe_cnt) * 1000;
+	wm_set->cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(dml, pipes, pipe_cnt) * 1000;
+	wm_set->pte_meta_urgent_ns = get_wm_memory_trip(dml, pipes, pipe_cnt) * 1000;
+	wm_set->frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(dml, pipes, pipe_cnt) * 1000;
+	wm_set->frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(dml, pipes, pipe_cnt) * 1000;
+	wm_set->urgent_latency_ns = get_urgent_latency(dml, pipes, pipe_cnt) * 1000;
+	dml->soc.dram_clock_change_latency_us = dram_clock_change_latency_cached;
+
+}
+
+void dcn301_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
+{
+	struct dcn301_resource_pool *pool = TO_DCN301_RES_POOL(dc->res_pool);
+	struct clk_limit_table *clk_table = &bw_params->clk_table;
+	struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES];
+	unsigned int i, closest_clk_lvl;
+	int j;
+
+	dc_assert_fp_enabled();
+
+	/* Default clock levels are used for diags, which may lead to overclocking. */
+	if (!IS_DIAG_DC(dc->ctx->dce_environment)) {
+		dcn3_01_ip.max_num_otg = pool->base.res_cap->num_timing_generator;
+		dcn3_01_ip.max_num_dpp = pool->base.pipe_count;
+		dcn3_01_soc.num_chans = bw_params->num_channels;
+
+		ASSERT(clk_table->num_entries);
+		for (i = 0; i < clk_table->num_entries; i++) {
+			/* loop backwards*/
+			for (closest_clk_lvl = 0, j = dcn3_01_soc.num_states - 1; j >= 0; j--) {
+				if ((unsigned int) dcn3_01_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) {
+					closest_clk_lvl = j;
+					break;
+				}
+			}
+
+			clock_limits[i].state = i;
+			clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz;
+			clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz;
+			clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz;
+			clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2;
+
+			clock_limits[i].dispclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dispclk_mhz;
+			clock_limits[i].dppclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dppclk_mhz;
+			clock_limits[i].dram_bw_per_chan_gbps = dcn3_01_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps;
+			clock_limits[i].dscclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dscclk_mhz;
+			clock_limits[i].dtbclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].dtbclk_mhz;
+			clock_limits[i].phyclk_d18_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz;
+			clock_limits[i].phyclk_mhz = dcn3_01_soc.clock_limits[closest_clk_lvl].phyclk_mhz;
+		}
+
+		for (i = 0; i < clk_table->num_entries; i++)
+			dcn3_01_soc.clock_limits[i] = clock_limits[i];
+
+		if (clk_table->num_entries) {
+			dcn3_01_soc.num_states = clk_table->num_entries;
+			/* duplicate last level */
+			dcn3_01_soc.clock_limits[dcn3_01_soc.num_states] = dcn3_01_soc.clock_limits[dcn3_01_soc.num_states - 1];
+			dcn3_01_soc.clock_limits[dcn3_01_soc.num_states].state = dcn3_01_soc.num_states;
+		}
+	}
+
+	dcn3_01_soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
+	dc->dml.soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
+
+	dml_init_instance(&dc->dml, &dcn3_01_soc, &dcn3_01_ip, DML_PROJECT_DCN30);
+}
+
+void dcn301_fpu_set_wm_ranges(int i,
+	struct pp_smu_wm_range_sets *ranges,
+	struct _vcs_dpi_soc_bounding_box_st *loaded_bb)
+{
+	dc_assert_fp_enabled();
+
+	ranges->reader_wm_sets[i].min_fill_clk_mhz = (i > 0) ? (loaded_bb->clock_limits[i - 1].dram_speed_mts / 16) + 1 : 0;
+	ranges->reader_wm_sets[i].max_fill_clk_mhz = loaded_bb->clock_limits[i].dram_speed_mts / 16;
+}
+
+void dcn301_fpu_init_soc_bounding_box(struct bp_soc_bb_info bb_info)
+{
+	dc_assert_fp_enabled();
+
+	if (bb_info.dram_clock_change_latency_100ns > 0)
+		dcn3_01_soc.dram_clock_change_latency_us = bb_info.dram_clock_change_latency_100ns * 10;
+
+	if (bb_info.dram_sr_enter_exit_latency_100ns > 0)
+		dcn3_01_soc.sr_enter_plus_exit_time_us = bb_info.dram_sr_enter_exit_latency_100ns * 10;
+
+	if (bb_info.dram_sr_exit_latency_100ns > 0)
+		dcn3_01_soc.sr_exit_time_us = bb_info.dram_sr_exit_latency_100ns * 10;
+}
+
+void dcn301_calculate_wm_and_dlg(struct dc *dc,
+		struct dc_state *context,
+		display_e2e_pipe_params_st *pipes,
+		int pipe_cnt,
+		int vlevel_req)
+{
+	int i, pipe_idx;
+	int vlevel, vlevel_max;
+	struct wm_range_table_entry *table_entry;
+	struct clk_bw_params *bw_params = dc->clk_mgr->bw_params;
+
+	ASSERT(bw_params);
+	dc_assert_fp_enabled();
+
+	vlevel_max = bw_params->clk_table.num_entries - 1;
+
+	/* WM Set D */
+	table_entry = &bw_params->wm_table.entries[WM_D];
+	if (table_entry->wm_type == WM_TYPE_RETRAINING)
+		vlevel = 0;
+	else
+		vlevel = vlevel_max;
+	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.d,
+						&context->bw_ctx.dml, pipes, pipe_cnt);
+	/* WM Set C */
+	table_entry = &bw_params->wm_table.entries[WM_C];
+	vlevel = min(max(vlevel_req, 2), vlevel_max);
+	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.c,
+						&context->bw_ctx.dml, pipes, pipe_cnt);
+	/* WM Set B */
+	table_entry = &bw_params->wm_table.entries[WM_B];
+	vlevel = min(max(vlevel_req, 1), vlevel_max);
+	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.b,
+						&context->bw_ctx.dml, pipes, pipe_cnt);
+
+	/* WM Set A */
+	table_entry = &bw_params->wm_table.entries[WM_A];
+	vlevel = min(vlevel_req, vlevel_max);
+	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.a,
+						&context->bw_ctx.dml, pipes, pipe_cnt);
+
+	for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
+		if (!context->res_ctx.pipe_ctx[i].stream)
+			continue;
+
+		pipes[pipe_idx].clks_cfg.dispclk_mhz = get_dispclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt);
+		pipes[pipe_idx].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
+
+		if (dc->config.forced_clocks) {
+			pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dispclk_mhz;
+			pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz;
+		}
+		if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz * 1000)
+			pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz / 1000.0;
+		if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000)
+			pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz / 1000.0;
+		pipe_idx++;
+	}
+
+	dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.h
new file mode 100644
index 000000000000..fc7065d17842
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn301/dcn301_fpu.h
@@ -0,0 +1,42 @@
+/*
+ * Copyright 2019-2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DCN301_FPU_H__
+#define __DCN301_FPU_H__
+
+void dcn301_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params);
+
+void dcn301_fpu_set_wm_ranges(int i,
+	struct pp_smu_wm_range_sets *ranges,
+	struct _vcs_dpi_soc_bounding_box_st *loaded_bb);
+
+void dcn301_fpu_init_soc_bounding_box(struct bp_soc_bb_info bb_info);
+
+void dcn301_calculate_wm_and_dlg(struct dc *dc,
+		struct dc_state *context,
+		display_e2e_pipe_params_st *pipes,
+		int pipe_cnt,
+		int vlevel_req);
+#endif /* __DCN301_FPU_H__*/
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/qp_tables.h b/drivers/gpu/drm/amd/display/dc/dml/dsc/qp_tables.h
index e5fac9f4181d..e5fac9f4181d 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/qp_tables.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dsc/qp_tables.h
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dsc/rc_calc_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dsc/rc_calc_fpu.c
new file mode 100644
index 000000000000..3ee858f311d1
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml/dsc/rc_calc_fpu.c
@@ -0,0 +1,291 @@
+/*
+ * Copyright 2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "rc_calc_fpu.h"
+
+#include "qp_tables.h"
+#include "amdgpu_dm/dc_fpu.h"
+
+#define table_hash(mode, bpc, max_min) ((mode << 16) | (bpc << 8) | max_min)
+
+#define MODE_SELECT(val444, val422, val420) \
+	(cm == CM_444 || cm == CM_RGB) ? (val444) : (cm == CM_422 ? (val422) : (val420))
+
+
+#define TABLE_CASE(mode, bpc, max)   case (table_hash(mode, BPC_##bpc, max)): \
+	table = qp_table_##mode##_##bpc##bpc_##max; \
+	table_size = sizeof(qp_table_##mode##_##bpc##bpc_##max)/sizeof(*qp_table_##mode##_##bpc##bpc_##max); \
+	break
+
+static int median3(int a, int b, int c)
+{
+	if (a > b)
+		swap(a, b);
+	if (b > c)
+		swap(b, c);
+	if (a > b)
+		swap(b, c);
+
+	return b;
+}
+
+static double dsc_roundf(double num)
+{
+	if (num < 0.0)
+		num = num - 0.5;
+	else
+		num = num + 0.5;
+
+	return (int)(num);
+}
+
+static double dsc_ceil(double num)
+{
+	double retval = (int)num;
+
+	if (retval != num && num > 0)
+		retval = num + 1;
+
+	return (int)retval;
+}
+
+static void get_qp_set(qp_set qps, enum colour_mode cm, enum bits_per_comp bpc,
+		       enum max_min max_min, float bpp)
+{
+	int mode = MODE_SELECT(444, 422, 420);
+	int sel = table_hash(mode, bpc, max_min);
+	int table_size = 0;
+	int index;
+	const struct qp_entry *table = 0L;
+
+	// alias enum
+	enum { min = DAL_MM_MIN, max = DAL_MM_MAX };
+	switch (sel) {
+		TABLE_CASE(444,  8, max);
+		TABLE_CASE(444,  8, min);
+		TABLE_CASE(444, 10, max);
+		TABLE_CASE(444, 10, min);
+		TABLE_CASE(444, 12, max);
+		TABLE_CASE(444, 12, min);
+		TABLE_CASE(422,  8, max);
+		TABLE_CASE(422,  8, min);
+		TABLE_CASE(422, 10, max);
+		TABLE_CASE(422, 10, min);
+		TABLE_CASE(422, 12, max);
+		TABLE_CASE(422, 12, min);
+		TABLE_CASE(420,  8, max);
+		TABLE_CASE(420,  8, min);
+		TABLE_CASE(420, 10, max);
+		TABLE_CASE(420, 10, min);
+		TABLE_CASE(420, 12, max);
+		TABLE_CASE(420, 12, min);
+	}
+
+	if (table == 0)
+		return;
+
+	index = (bpp - table[0].bpp) * 2;
+
+	/* requested size is bigger than the table */
+	if (index >= table_size) {
+		dm_error("ERROR: Requested rc_calc to find a bpp entry that exceeds the table size\n");
+		return;
+	}
+
+	memcpy(qps, table[index].qps, sizeof(qp_set));
+}
+
+static void get_ofs_set(qp_set ofs, enum colour_mode mode, float bpp)
+{
+	int   *p = ofs;
+
+	if (mode == CM_444 || mode == CM_RGB) {
+		*p++ = (bpp <=  6) ? (0) : ((((bpp >=  8) && (bpp <= 12))) ? (2) : ((bpp >= 15) ? (10) : ((((bpp > 6) && (bpp < 8))) ? (0 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (2 + dsc_roundf((bpp - 12) * (8 / 3.0))))));
+		*p++ = (bpp <=  6) ? (-2) : ((((bpp >=  8) && (bpp <= 12))) ? (0) : ((bpp >= 15) ? (8) : ((((bpp > 6) && (bpp < 8))) ? (-2 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (0 + dsc_roundf((bpp - 12) * (8 / 3.0))))));
+		*p++ = (bpp <=  6) ? (-2) : ((((bpp >=  8) && (bpp <= 12))) ? (0) : ((bpp >= 15) ? (6) : ((((bpp > 6) && (bpp < 8))) ? (-2 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (0 + dsc_roundf((bpp - 12) * (6 / 3.0))))));
+		*p++ = (bpp <=  6) ? (-4) : ((((bpp >=  8) && (bpp <= 12))) ? (-2) : ((bpp >= 15) ? (4) : ((((bpp > 6) && (bpp < 8))) ? (-4 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (-2 + dsc_roundf((bpp - 12) * (6 / 3.0))))));
+		*p++ = (bpp <=  6) ? (-6) : ((((bpp >=  8) && (bpp <= 12))) ? (-4) : ((bpp >= 15) ? (2) : ((((bpp > 6) && (bpp < 8))) ? (-6 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (-4 + dsc_roundf((bpp - 12) * (6 / 3.0))))));
+		*p++ = (bpp <= 12) ? (-6) : ((bpp >= 15) ? (0) : (-6 + dsc_roundf((bpp - 12) * (6 / 3.0))));
+		*p++ = (bpp <= 12) ? (-8) : ((bpp >= 15) ? (-2) : (-8 + dsc_roundf((bpp - 12) * (6 / 3.0))));
+		*p++ = (bpp <= 12) ? (-8) : ((bpp >= 15) ? (-4) : (-8 + dsc_roundf((bpp - 12) * (4 / 3.0))));
+		*p++ = (bpp <= 12) ? (-8) : ((bpp >= 15) ? (-6) : (-8 + dsc_roundf((bpp - 12) * (2 / 3.0))));
+		*p++ = (bpp <= 12) ? (-10) : ((bpp >= 15) ? (-8) : (-10 + dsc_roundf((bpp - 12) * (2 / 3.0))));
+		*p++ = -10;
+		*p++ = (bpp <=  6) ? (-12) : ((bpp >=  8) ? (-10) : (-12 + dsc_roundf((bpp -  6) * (2 / 2.0))));
+		*p++ = -12;
+		*p++ = -12;
+		*p++ = -12;
+	} else if (mode == CM_422) {
+		*p++ = (bpp <=  8) ? (2) : ((bpp >= 10) ? (10) : (2 + dsc_roundf((bpp -  8) * (8 / 2.0))));
+		*p++ = (bpp <=  8) ? (0) : ((bpp >= 10) ? (8) : (0 + dsc_roundf((bpp -  8) * (8 / 2.0))));
+		*p++ = (bpp <=  8) ? (0) : ((bpp >= 10) ? (6) : (0 + dsc_roundf((bpp -  8) * (6 / 2.0))));
+		*p++ = (bpp <=  8) ? (-2) : ((bpp >= 10) ? (4) : (-2 + dsc_roundf((bpp -  8) * (6 / 2.0))));
+		*p++ = (bpp <=  8) ? (-4) : ((bpp >= 10) ? (2) : (-4 + dsc_roundf((bpp -  8) * (6 / 2.0))));
+		*p++ = (bpp <=  8) ? (-6) : ((bpp >= 10) ? (0) : (-6 + dsc_roundf((bpp -  8) * (6 / 2.0))));
+		*p++ = (bpp <=  8) ? (-8) : ((bpp >= 10) ? (-2) : (-8 + dsc_roundf((bpp -  8) * (6 / 2.0))));
+		*p++ = (bpp <=  8) ? (-8) : ((bpp >= 10) ? (-4) : (-8 + dsc_roundf((bpp -  8) * (4 / 2.0))));
+		*p++ = (bpp <=  8) ? (-8) : ((bpp >= 10) ? (-6) : (-8 + dsc_roundf((bpp -  8) * (2 / 2.0))));
+		*p++ = (bpp <=  8) ? (-10) : ((bpp >= 10) ? (-8) : (-10 + dsc_roundf((bpp -  8) * (2 / 2.0))));
+		*p++ = -10;
+		*p++ = (bpp <=  6) ? (-12) : ((bpp >= 7) ? (-10) : (-12 + dsc_roundf((bpp -  6) * (2.0 / 1))));
+		*p++ = -12;
+		*p++ = -12;
+		*p++ = -12;
+	} else {
+		*p++ = (bpp <=  6) ? (2) : ((bpp >=  8) ? (10) : (2 + dsc_roundf((bpp -  6) * (8 / 2.0))));
+		*p++ = (bpp <=  6) ? (0) : ((bpp >=  8) ? (8) : (0 + dsc_roundf((bpp -  6) * (8 / 2.0))));
+		*p++ = (bpp <=  6) ? (0) : ((bpp >=  8) ? (6) : (0 + dsc_roundf((bpp -  6) * (6 / 2.0))));
+		*p++ = (bpp <=  6) ? (-2) : ((bpp >=  8) ? (4) : (-2 + dsc_roundf((bpp -  6) * (6 / 2.0))));
+		*p++ = (bpp <=  6) ? (-4) : ((bpp >=  8) ? (2) : (-4 + dsc_roundf((bpp -  6) * (6 / 2.0))));
+		*p++ = (bpp <=  6) ? (-6) : ((bpp >=  8) ? (0) : (-6 + dsc_roundf((bpp -  6) * (6 / 2.0))));
+		*p++ = (bpp <=  6) ? (-8) : ((bpp >=  8) ? (-2) : (-8 + dsc_roundf((bpp -  6) * (6 / 2.0))));
+		*p++ = (bpp <=  6) ? (-8) : ((bpp >=  8) ? (-4) : (-8 + dsc_roundf((bpp -  6) * (4 / 2.0))));
+		*p++ = (bpp <=  6) ? (-8) : ((bpp >=  8) ? (-6) : (-8 + dsc_roundf((bpp -  6) * (2 / 2.0))));
+		*p++ = (bpp <=  6) ? (-10) : ((bpp >=  8) ? (-8) : (-10 + dsc_roundf((bpp -  6) * (2 / 2.0))));
+		*p++ = -10;
+		*p++ = (bpp <=  4) ? (-12) : ((bpp >=  5) ? (-10) : (-12 + dsc_roundf((bpp -  4) * (2 / 1.0))));
+		*p++ = -12;
+		*p++ = -12;
+		*p++ = -12;
+	}
+}
+
+void _do_calc_rc_params(struct rc_params *rc,
+		enum colour_mode cm,
+		enum bits_per_comp bpc,
+		u16 drm_bpp,
+		bool is_navite_422_or_420,
+		int slice_width,
+		int slice_height,
+		int minor_version)
+{
+	float bpp;
+	float bpp_group;
+	float initial_xmit_delay_factor;
+	int padding_pixels;
+	int i;
+
+	dc_assert_fp_enabled();
+
+	bpp = ((float)drm_bpp / 16.0);
+	/* in native_422 or native_420 modes, the bits_per_pixel is double the
+	 * target bpp (the latter is what calc_rc_params expects)
+	 */
+	if (is_navite_422_or_420)
+		bpp /= 2.0;
+
+	rc->rc_quant_incr_limit0 = ((bpc == BPC_8) ? 11 : (bpc == BPC_10 ? 15 : 19)) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
+	rc->rc_quant_incr_limit1 = ((bpc == BPC_8) ? 11 : (bpc == BPC_10 ? 15 : 19)) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
+
+	bpp_group = MODE_SELECT(bpp, bpp * 2.0, bpp * 2.0);
+
+	switch (cm) {
+	case CM_420:
+		rc->initial_fullness_offset = (bpp >=  6) ? (2048) : ((bpp <=  4) ? (6144) : ((((bpp >  4) && (bpp <=  5))) ? (6144 - dsc_roundf((bpp - 4) * (512))) : (5632 - dsc_roundf((bpp -  5) * (3584)))));
+		rc->first_line_bpg_offset   = median3(0, (12 + (int) (0.09 *  min(34, slice_height - 8))), (int)((3 * bpc * 3) - (3 * bpp_group)));
+		rc->second_line_bpg_offset  = median3(0, 12, (int)((3 * bpc * 3) - (3 * bpp_group)));
+		break;
+	case CM_422:
+		rc->initial_fullness_offset = (bpp >=  8) ? (2048) : ((bpp <=  7) ? (5632) : (5632 - dsc_roundf((bpp - 7) * (3584))));
+		rc->first_line_bpg_offset   = median3(0, (12 + (int) (0.09 *  min(34, slice_height - 8))), (int)((3 * bpc * 4) - (3 * bpp_group)));
+		rc->second_line_bpg_offset  = 0;
+		break;
+	case CM_444:
+	case CM_RGB:
+		rc->initial_fullness_offset = (bpp >= 12) ? (2048) : ((bpp <=  8) ? (6144) : ((((bpp >  8) && (bpp <= 10))) ? (6144 - dsc_roundf((bpp - 8) * (512 / 2))) : (5632 - dsc_roundf((bpp - 10) * (3584 / 2)))));
+		rc->first_line_bpg_offset   = median3(0, (12 + (int) (0.09 *  min(34, slice_height - 8))), (int)(((3 * bpc + (cm == CM_444 ? 0 : 2)) * 3) - (3 * bpp_group)));
+		rc->second_line_bpg_offset  = 0;
+		break;
+	}
+
+	initial_xmit_delay_factor = (cm == CM_444 || cm == CM_RGB) ? 1.0 : 2.0;
+	rc->initial_xmit_delay = dsc_roundf(8192.0/2.0/bpp/initial_xmit_delay_factor);
+
+	if (cm == CM_422 || cm == CM_420)
+		slice_width /= 2;
+
+	padding_pixels = ((slice_width % 3) != 0) ? (3 - (slice_width % 3)) * (rc->initial_xmit_delay / slice_width) : 0;
+	if (3 * bpp_group >= (((rc->initial_xmit_delay + 2) / 3) * (3 + (cm == CM_422)))) {
+		if ((rc->initial_xmit_delay + padding_pixels) % 3 == 1)
+			rc->initial_xmit_delay++;
+	}
+
+	rc->flatness_min_qp     = ((bpc == BPC_8) ?  (3) : ((bpc == BPC_10) ? (7)  : (11))) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
+	rc->flatness_max_qp     = ((bpc == BPC_8) ? (12) : ((bpc == BPC_10) ? (16) : (20))) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
+	rc->flatness_det_thresh = 2 << (bpc - 8);
+
+	get_qp_set(rc->qp_min, cm, bpc, DAL_MM_MIN, bpp);
+	get_qp_set(rc->qp_max, cm, bpc, DAL_MM_MAX, bpp);
+	if (cm == CM_444 && minor_version == 1) {
+		for (i = 0; i < QP_SET_SIZE; ++i) {
+			rc->qp_min[i] = rc->qp_min[i] > 0 ? rc->qp_min[i] - 1 : 0;
+			rc->qp_max[i] = rc->qp_max[i] > 0 ? rc->qp_max[i] - 1 : 0;
+		}
+	}
+	get_ofs_set(rc->ofs, cm, bpp);
+
+	/* fixed parameters */
+	rc->rc_model_size    = 8192;
+	rc->rc_edge_factor   = 6;
+	rc->rc_tgt_offset_hi = 3;
+	rc->rc_tgt_offset_lo = 3;
+
+	rc->rc_buf_thresh[0] = 896;
+	rc->rc_buf_thresh[1] = 1792;
+	rc->rc_buf_thresh[2] = 2688;
+	rc->rc_buf_thresh[3] = 3584;
+	rc->rc_buf_thresh[4] = 4480;
+	rc->rc_buf_thresh[5] = 5376;
+	rc->rc_buf_thresh[6] = 6272;
+	rc->rc_buf_thresh[7] = 6720;
+	rc->rc_buf_thresh[8] = 7168;
+	rc->rc_buf_thresh[9] = 7616;
+	rc->rc_buf_thresh[10] = 7744;
+	rc->rc_buf_thresh[11] = 7872;
+	rc->rc_buf_thresh[12] = 8000;
+	rc->rc_buf_thresh[13] = 8064;
+}
+
+u32 _do_bytes_per_pixel_calc(int slice_width,
+		u16 drm_bpp,
+		bool is_navite_422_or_420)
+{
+	float bpp;
+	u32 bytes_per_pixel;
+	double d_bytes_per_pixel;
+
+	dc_assert_fp_enabled();
+
+	bpp = ((float)drm_bpp / 16.0);
+	d_bytes_per_pixel = dsc_ceil(bpp * slice_width / 8.0) / slice_width;
+	// TODO: Make sure the formula for calculating this is precise (ceiling
+	// vs. floor, and at what point they should be applied)
+	if (is_navite_422_or_420)
+		d_bytes_per_pixel /= 2;
+
+	bytes_per_pixel = (u32)dsc_ceil(d_bytes_per_pixel * 0x10000000);
+
+	return bytes_per_pixel;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dsc/rc_calc_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dsc/rc_calc_fpu.h
new file mode 100644
index 000000000000..b93b95409fbe
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml/dsc/rc_calc_fpu.h
@@ -0,0 +1,94 @@
+/*
+ * Copyright 2021 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __RC_CALC_FPU_H__
+#define __RC_CALC_FPU_H__
+
+#include "os_types.h"
+#include <drm/drm_dsc.h>
+
+#define QP_SET_SIZE 15
+
+typedef int qp_set[QP_SET_SIZE];
+
+struct rc_params {
+	int      rc_quant_incr_limit0;
+	int      rc_quant_incr_limit1;
+	int      initial_fullness_offset;
+	int      initial_xmit_delay;
+	int      first_line_bpg_offset;
+	int      second_line_bpg_offset;
+	int      flatness_min_qp;
+	int      flatness_max_qp;
+	int      flatness_det_thresh;
+	qp_set   qp_min;
+	qp_set   qp_max;
+	qp_set   ofs;
+	int      rc_model_size;
+	int      rc_edge_factor;
+	int      rc_tgt_offset_hi;
+	int      rc_tgt_offset_lo;
+	int      rc_buf_thresh[QP_SET_SIZE - 1];
+};
+
+enum colour_mode {
+	CM_RGB,   /* 444 RGB */
+	CM_444,   /* 444 YUV or simple 422 */
+	CM_422,   /* native 422 */
+	CM_420    /* native 420 */
+};
+
+enum bits_per_comp {
+	BPC_8  =  8,
+	BPC_10 = 10,
+	BPC_12 = 12
+};
+
+enum max_min {
+	DAL_MM_MIN = 0,
+	DAL_MM_MAX = 1
+};
+
+struct qp_entry {
+	float         bpp;
+	const qp_set  qps;
+};
+
+typedef struct qp_entry qp_table[];
+
+u32 _do_bytes_per_pixel_calc(int slice_width,
+		u16 drm_bpp,
+		bool is_navite_422_or_420);
+
+void _do_calc_rc_params(struct rc_params *rc,
+		enum colour_mode cm,
+		enum bits_per_comp bpc,
+		u16 drm_bpp,
+		bool is_navite_422_or_420,
+		int slice_width,
+		int slice_height,
+		int minor_version);
+
+#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/Makefile b/drivers/gpu/drm/amd/display/dc/dsc/Makefile
index 8d31eb75c6a6..a2537229ee88 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dsc/Makefile
@@ -1,35 +1,6 @@
 # SPDX-License-Identifier: MIT
 #
 # Makefile for the 'dsc' sub-component of DAL.
-
-ifdef CONFIG_X86
-dsc_ccflags := -mhard-float -msse
-endif
-
-ifdef CONFIG_PPC64
-dsc_ccflags := -mhard-float -maltivec
-endif
-
-ifdef CONFIG_CC_IS_GCC
-ifeq ($(call cc-ifversion, -lt, 0701, y), y)
-IS_OLD_GCC = 1
-endif
-endif
-
-ifdef CONFIG_X86
-ifdef IS_OLD_GCC
-# Stack alignment mismatch, proceed with caution.
-# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3
-# (8B stack alignment).
-dsc_ccflags += -mpreferred-stack-boundary=4
-else
-dsc_ccflags += -msse2
-endif
-endif
-
-CFLAGS_$(AMDDALPATH)/dc/dsc/rc_calc.o := $(dsc_ccflags)
-CFLAGS_REMOVE_$(AMDDALPATH)/dc/dsc/rc_calc.o := $(dsc_rcflags)
-
 DSC = dc_dsc.o rc_calc.o rc_calc_dpi.o
 
 AMD_DAL_DSC = $(addprefix $(AMDDALPATH)/dc/dsc/,$(DSC))
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c
index 7b294f637881..b19d3aeb5962 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.c
@@ -23,266 +23,7 @@
  * Authors: AMD
  *
  */
-#include <drm/drm_dsc.h>
-
-#include "os_types.h"
 #include "rc_calc.h"
-#include "qp_tables.h"
-
-#define table_hash(mode, bpc, max_min) ((mode << 16) | (bpc << 8) | max_min)
-
-#define MODE_SELECT(val444, val422, val420) \
-	(cm == CM_444 || cm == CM_RGB) ? (val444) : (cm == CM_422 ? (val422) : (val420))
-
-
-#define TABLE_CASE(mode, bpc, max)   case (table_hash(mode, BPC_##bpc, max)): \
-	table = qp_table_##mode##_##bpc##bpc_##max; \
-	table_size = sizeof(qp_table_##mode##_##bpc##bpc_##max)/sizeof(*qp_table_##mode##_##bpc##bpc_##max); \
-	break
-
-
-static void get_qp_set(qp_set qps, enum colour_mode cm, enum bits_per_comp bpc,
-		       enum max_min max_min, float bpp)
-{
-	int mode = MODE_SELECT(444, 422, 420);
-	int sel = table_hash(mode, bpc, max_min);
-	int table_size = 0;
-	int index;
-	const struct qp_entry *table = 0L;
-
-	// alias enum
-	enum { min = DAL_MM_MIN, max = DAL_MM_MAX };
-	switch (sel) {
-		TABLE_CASE(444,  8, max);
-		TABLE_CASE(444,  8, min);
-		TABLE_CASE(444, 10, max);
-		TABLE_CASE(444, 10, min);
-		TABLE_CASE(444, 12, max);
-		TABLE_CASE(444, 12, min);
-		TABLE_CASE(422,  8, max);
-		TABLE_CASE(422,  8, min);
-		TABLE_CASE(422, 10, max);
-		TABLE_CASE(422, 10, min);
-		TABLE_CASE(422, 12, max);
-		TABLE_CASE(422, 12, min);
-		TABLE_CASE(420,  8, max);
-		TABLE_CASE(420,  8, min);
-		TABLE_CASE(420, 10, max);
-		TABLE_CASE(420, 10, min);
-		TABLE_CASE(420, 12, max);
-		TABLE_CASE(420, 12, min);
-	}
-
-	if (table == 0)
-		return;
-
-	index = (bpp - table[0].bpp) * 2;
-
-	/* requested size is bigger than the table */
-	if (index >= table_size) {
-		dm_error("ERROR: Requested rc_calc to find a bpp entry that exceeds the table size\n");
-		return;
-	}
-
-	memcpy(qps, table[index].qps, sizeof(qp_set));
-}
-
-static double dsc_roundf(double num)
-{
-	if (num < 0.0)
-		num = num - 0.5;
-	else
-		num = num + 0.5;
-
-	return (int)(num);
-}
-
-static double dsc_ceil(double num)
-{
-	double retval = (int)num;
-
-	if (retval != num && num > 0)
-		retval = num + 1;
-
-	return (int)retval;
-}
-
-static void get_ofs_set(qp_set ofs, enum colour_mode mode, float bpp)
-{
-	int   *p = ofs;
-
-	if (mode == CM_444 || mode == CM_RGB) {
-		*p++ = (bpp <=  6) ? (0) : ((((bpp >=  8) && (bpp <= 12))) ? (2) : ((bpp >= 15) ? (10) : ((((bpp > 6) && (bpp < 8))) ? (0 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (2 + dsc_roundf((bpp - 12) * (8 / 3.0))))));
-		*p++ = (bpp <=  6) ? (-2) : ((((bpp >=  8) && (bpp <= 12))) ? (0) : ((bpp >= 15) ? (8) : ((((bpp > 6) && (bpp < 8))) ? (-2 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (0 + dsc_roundf((bpp - 12) * (8 / 3.0))))));
-		*p++ = (bpp <=  6) ? (-2) : ((((bpp >=  8) && (bpp <= 12))) ? (0) : ((bpp >= 15) ? (6) : ((((bpp > 6) && (bpp < 8))) ? (-2 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (0 + dsc_roundf((bpp - 12) * (6 / 3.0))))));
-		*p++ = (bpp <=  6) ? (-4) : ((((bpp >=  8) && (bpp <= 12))) ? (-2) : ((bpp >= 15) ? (4) : ((((bpp > 6) && (bpp < 8))) ? (-4 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (-2 + dsc_roundf((bpp - 12) * (6 / 3.0))))));
-		*p++ = (bpp <=  6) ? (-6) : ((((bpp >=  8) && (bpp <= 12))) ? (-4) : ((bpp >= 15) ? (2) : ((((bpp > 6) && (bpp < 8))) ? (-6 + dsc_roundf((bpp -  6) * (2 / 2.0))) : (-4 + dsc_roundf((bpp - 12) * (6 / 3.0))))));
-		*p++ = (bpp <= 12) ? (-6) : ((bpp >= 15) ? (0) : (-6 + dsc_roundf((bpp - 12) * (6 / 3.0))));
-		*p++ = (bpp <= 12) ? (-8) : ((bpp >= 15) ? (-2) : (-8 + dsc_roundf((bpp - 12) * (6 / 3.0))));
-		*p++ = (bpp <= 12) ? (-8) : ((bpp >= 15) ? (-4) : (-8 + dsc_roundf((bpp - 12) * (4 / 3.0))));
-		*p++ = (bpp <= 12) ? (-8) : ((bpp >= 15) ? (-6) : (-8 + dsc_roundf((bpp - 12) * (2 / 3.0))));
-		*p++ = (bpp <= 12) ? (-10) : ((bpp >= 15) ? (-8) : (-10 + dsc_roundf((bpp - 12) * (2 / 3.0))));
-		*p++ = -10;
-		*p++ = (bpp <=  6) ? (-12) : ((bpp >=  8) ? (-10) : (-12 + dsc_roundf((bpp -  6) * (2 / 2.0))));
-		*p++ = -12;
-		*p++ = -12;
-		*p++ = -12;
-	} else if (mode == CM_422) {
-		*p++ = (bpp <=  8) ? (2) : ((bpp >= 10) ? (10) : (2 + dsc_roundf((bpp -  8) * (8 / 2.0))));
-		*p++ = (bpp <=  8) ? (0) : ((bpp >= 10) ? (8) : (0 + dsc_roundf((bpp -  8) * (8 / 2.0))));
-		*p++ = (bpp <=  8) ? (0) : ((bpp >= 10) ? (6) : (0 + dsc_roundf((bpp -  8) * (6 / 2.0))));
-		*p++ = (bpp <=  8) ? (-2) : ((bpp >= 10) ? (4) : (-2 + dsc_roundf((bpp -  8) * (6 / 2.0))));
-		*p++ = (bpp <=  8) ? (-4) : ((bpp >= 10) ? (2) : (-4 + dsc_roundf((bpp -  8) * (6 / 2.0))));
-		*p++ = (bpp <=  8) ? (-6) : ((bpp >= 10) ? (0) : (-6 + dsc_roundf((bpp -  8) * (6 / 2.0))));
-		*p++ = (bpp <=  8) ? (-8) : ((bpp >= 10) ? (-2) : (-8 + dsc_roundf((bpp -  8) * (6 / 2.0))));
-		*p++ = (bpp <=  8) ? (-8) : ((bpp >= 10) ? (-4) : (-8 + dsc_roundf((bpp -  8) * (4 / 2.0))));
-		*p++ = (bpp <=  8) ? (-8) : ((bpp >= 10) ? (-6) : (-8 + dsc_roundf((bpp -  8) * (2 / 2.0))));
-		*p++ = (bpp <=  8) ? (-10) : ((bpp >= 10) ? (-8) : (-10 + dsc_roundf((bpp -  8) * (2 / 2.0))));
-		*p++ = -10;
-		*p++ = (bpp <=  6) ? (-12) : ((bpp >= 7) ? (-10) : (-12 + dsc_roundf((bpp -  6) * (2.0 / 1))));
-		*p++ = -12;
-		*p++ = -12;
-		*p++ = -12;
-	} else {
-		*p++ = (bpp <=  6) ? (2) : ((bpp >=  8) ? (10) : (2 + dsc_roundf((bpp -  6) * (8 / 2.0))));
-		*p++ = (bpp <=  6) ? (0) : ((bpp >=  8) ? (8) : (0 + dsc_roundf((bpp -  6) * (8 / 2.0))));
-		*p++ = (bpp <=  6) ? (0) : ((bpp >=  8) ? (6) : (0 + dsc_roundf((bpp -  6) * (6 / 2.0))));
-		*p++ = (bpp <=  6) ? (-2) : ((bpp >=  8) ? (4) : (-2 + dsc_roundf((bpp -  6) * (6 / 2.0))));
-		*p++ = (bpp <=  6) ? (-4) : ((bpp >=  8) ? (2) : (-4 + dsc_roundf((bpp -  6) * (6 / 2.0))));
-		*p++ = (bpp <=  6) ? (-6) : ((bpp >=  8) ? (0) : (-6 + dsc_roundf((bpp -  6) * (6 / 2.0))));
-		*p++ = (bpp <=  6) ? (-8) : ((bpp >=  8) ? (-2) : (-8 + dsc_roundf((bpp -  6) * (6 / 2.0))));
-		*p++ = (bpp <=  6) ? (-8) : ((bpp >=  8) ? (-4) : (-8 + dsc_roundf((bpp -  6) * (4 / 2.0))));
-		*p++ = (bpp <=  6) ? (-8) : ((bpp >=  8) ? (-6) : (-8 + dsc_roundf((bpp -  6) * (2 / 2.0))));
-		*p++ = (bpp <=  6) ? (-10) : ((bpp >=  8) ? (-8) : (-10 + dsc_roundf((bpp -  6) * (2 / 2.0))));
-		*p++ = -10;
-		*p++ = (bpp <=  4) ? (-12) : ((bpp >=  5) ? (-10) : (-12 + dsc_roundf((bpp -  4) * (2 / 1.0))));
-		*p++ = -12;
-		*p++ = -12;
-		*p++ = -12;
-	}
-}
-
-static int median3(int a, int b, int c)
-{
-	if (a > b)
-		swap(a, b);
-	if (b > c)
-		swap(b, c);
-	if (a > b)
-		swap(b, c);
-
-	return b;
-}
-
-static void _do_calc_rc_params(struct rc_params *rc, enum colour_mode cm,
-			       enum bits_per_comp bpc, u16 drm_bpp,
-			       bool is_navite_422_or_420,
-			       int slice_width, int slice_height,
-			       int minor_version)
-{
-	float bpp;
-	float bpp_group;
-	float initial_xmit_delay_factor;
-	int padding_pixels;
-	int i;
-
-	bpp = ((float)drm_bpp / 16.0);
-	/* in native_422 or native_420 modes, the bits_per_pixel is double the
-	 * target bpp (the latter is what calc_rc_params expects)
-	 */
-	if (is_navite_422_or_420)
-		bpp /= 2.0;
-
-	rc->rc_quant_incr_limit0 = ((bpc == BPC_8) ? 11 : (bpc == BPC_10 ? 15 : 19)) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
-	rc->rc_quant_incr_limit1 = ((bpc == BPC_8) ? 11 : (bpc == BPC_10 ? 15 : 19)) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
-
-	bpp_group = MODE_SELECT(bpp, bpp * 2.0, bpp * 2.0);
-
-	switch (cm) {
-	case CM_420:
-		rc->initial_fullness_offset = (bpp >=  6) ? (2048) : ((bpp <=  4) ? (6144) : ((((bpp >  4) && (bpp <=  5))) ? (6144 - dsc_roundf((bpp - 4) * (512))) : (5632 - dsc_roundf((bpp -  5) * (3584)))));
-		rc->first_line_bpg_offset   = median3(0, (12 + (int) (0.09 *  min(34, slice_height - 8))), (int)((3 * bpc * 3) - (3 * bpp_group)));
-		rc->second_line_bpg_offset  = median3(0, 12, (int)((3 * bpc * 3) - (3 * bpp_group)));
-		break;
-	case CM_422:
-		rc->initial_fullness_offset = (bpp >=  8) ? (2048) : ((bpp <=  7) ? (5632) : (5632 - dsc_roundf((bpp - 7) * (3584))));
-		rc->first_line_bpg_offset   = median3(0, (12 + (int) (0.09 *  min(34, slice_height - 8))), (int)((3 * bpc * 4) - (3 * bpp_group)));
-		rc->second_line_bpg_offset  = 0;
-		break;
-	case CM_444:
-	case CM_RGB:
-		rc->initial_fullness_offset = (bpp >= 12) ? (2048) : ((bpp <=  8) ? (6144) : ((((bpp >  8) && (bpp <= 10))) ? (6144 - dsc_roundf((bpp - 8) * (512 / 2))) : (5632 - dsc_roundf((bpp - 10) * (3584 / 2)))));
-		rc->first_line_bpg_offset   = median3(0, (12 + (int) (0.09 *  min(34, slice_height - 8))), (int)(((3 * bpc + (cm == CM_444 ? 0 : 2)) * 3) - (3 * bpp_group)));
-		rc->second_line_bpg_offset  = 0;
-		break;
-	}
-
-	initial_xmit_delay_factor = (cm == CM_444 || cm == CM_RGB) ? 1.0 : 2.0;
-	rc->initial_xmit_delay = dsc_roundf(8192.0/2.0/bpp/initial_xmit_delay_factor);
-
-	if (cm == CM_422 || cm == CM_420)
-		slice_width /= 2;
-
-	padding_pixels = ((slice_width % 3) != 0) ? (3 - (slice_width % 3)) * (rc->initial_xmit_delay / slice_width) : 0;
-	if (3 * bpp_group >= (((rc->initial_xmit_delay + 2) / 3) * (3 + (cm == CM_422)))) {
-		if ((rc->initial_xmit_delay + padding_pixels) % 3 == 1)
-			rc->initial_xmit_delay++;
-	}
-
-	rc->flatness_min_qp     = ((bpc == BPC_8) ?  (3) : ((bpc == BPC_10) ? (7)  : (11))) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
-	rc->flatness_max_qp     = ((bpc == BPC_8) ? (12) : ((bpc == BPC_10) ? (16) : (20))) - ((minor_version == 1 && cm == CM_444) ? 1 : 0);
-	rc->flatness_det_thresh = 2 << (bpc - 8);
-
-	get_qp_set(rc->qp_min, cm, bpc, DAL_MM_MIN, bpp);
-	get_qp_set(rc->qp_max, cm, bpc, DAL_MM_MAX, bpp);
-	if (cm == CM_444 && minor_version == 1) {
-		for (i = 0; i < QP_SET_SIZE; ++i) {
-			rc->qp_min[i] = rc->qp_min[i] > 0 ? rc->qp_min[i] - 1 : 0;
-			rc->qp_max[i] = rc->qp_max[i] > 0 ? rc->qp_max[i] - 1 : 0;
-		}
-	}
-	get_ofs_set(rc->ofs, cm, bpp);
-
-	/* fixed parameters */
-	rc->rc_model_size    = 8192;
-	rc->rc_edge_factor   = 6;
-	rc->rc_tgt_offset_hi = 3;
-	rc->rc_tgt_offset_lo = 3;
-
-	rc->rc_buf_thresh[0] = 896;
-	rc->rc_buf_thresh[1] = 1792;
-	rc->rc_buf_thresh[2] = 2688;
-	rc->rc_buf_thresh[3] = 3584;
-	rc->rc_buf_thresh[4] = 4480;
-	rc->rc_buf_thresh[5] = 5376;
-	rc->rc_buf_thresh[6] = 6272;
-	rc->rc_buf_thresh[7] = 6720;
-	rc->rc_buf_thresh[8] = 7168;
-	rc->rc_buf_thresh[9] = 7616;
-	rc->rc_buf_thresh[10] = 7744;
-	rc->rc_buf_thresh[11] = 7872;
-	rc->rc_buf_thresh[12] = 8000;
-	rc->rc_buf_thresh[13] = 8064;
-}
-
-static u32 _do_bytes_per_pixel_calc(int slice_width, u16 drm_bpp,
-				    bool is_navite_422_or_420)
-{
-	float bpp;
-	u32 bytes_per_pixel;
-	double d_bytes_per_pixel;
-
-	bpp = ((float)drm_bpp / 16.0);
-	d_bytes_per_pixel = dsc_ceil(bpp * slice_width / 8.0) / slice_width;
-	// TODO: Make sure the formula for calculating this is precise (ceiling
-	// vs. floor, and at what point they should be applied)
-	if (is_navite_422_or_420)
-		d_bytes_per_pixel /= 2;
-
-	bytes_per_pixel = (u32)dsc_ceil(d_bytes_per_pixel * 0x10000000);
-
-	return bytes_per_pixel;
-}
 
 /**
  * calc_rc_params - reads the user's cmdline mode
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.h b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.h
index 262f06afcbf9..c2340e001b57 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.h
+++ b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc.h
@@ -27,55 +27,7 @@
 #ifndef __RC_CALC_H__
 #define __RC_CALC_H__
 
-
-#define QP_SET_SIZE 15
-
-typedef int qp_set[QP_SET_SIZE];
-
-struct rc_params {
-	int      rc_quant_incr_limit0;
-	int      rc_quant_incr_limit1;
-	int      initial_fullness_offset;
-	int      initial_xmit_delay;
-	int      first_line_bpg_offset;
-	int      second_line_bpg_offset;
-	int      flatness_min_qp;
-	int      flatness_max_qp;
-	int      flatness_det_thresh;
-	qp_set   qp_min;
-	qp_set   qp_max;
-	qp_set   ofs;
-	int      rc_model_size;
-	int      rc_edge_factor;
-	int      rc_tgt_offset_hi;
-	int      rc_tgt_offset_lo;
-	int      rc_buf_thresh[QP_SET_SIZE - 1];
-};
-
-enum colour_mode {
-	CM_RGB,   /* 444 RGB */
-	CM_444,   /* 444 YUV or simple 422 */
-	CM_422,   /* native 422 */
-	CM_420    /* native 420 */
-};
-
-enum bits_per_comp {
-	BPC_8  =  8,
-	BPC_10 = 10,
-	BPC_12 = 12
-};
-
-enum max_min {
-	DAL_MM_MIN = 0,
-	DAL_MM_MAX = 1
-};
-
-struct qp_entry {
-	float         bpp;
-	const qp_set  qps;
-};
-
-typedef struct qp_entry qp_table[];
+#include "dml/dsc/rc_calc_fpu.h"
 
 void calc_rc_params(struct rc_params *rc, const struct drm_dsc_config *pps);
 u32 calc_dsc_bytes_per_pixel(const struct drm_dsc_config *pps);
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c
index ef830aded5b1..1e19dd674e5a 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c
@@ -22,7 +22,6 @@
  * Authors: AMD
  *
  */
-#include "os_types.h"
 #include <drm/drm_dsc.h>
 #include "dscc_types.h"
 #include "rc_calc.h"
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
index c70375117519..a6d3d859754a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h
@@ -121,12 +121,12 @@ enum dc_status dpcd_set_lane_settings(
 	const struct link_training_settings *link_training_setting,
 	uint32_t offset);
 /* Read training status and adjustment requests from DPCD. */
-enum dc_status dp_get_lane_status_and_drive_settings(
+enum dc_status dp_get_lane_status_and_lane_adjust(
 	struct dc_link *link,
 	const struct link_training_settings *link_training_setting,
-	union lane_status *ln_status,
-	union lane_align_status_updated *ln_status_updated,
-	struct link_training_settings *req_settings,
+	union lane_status ln_status[LANE_COUNT_DP_MAX],
+	union lane_align_status_updated *ln_align,
+	union lane_adjust ln_adjust[LANE_COUNT_DP_MAX],
 	uint32_t offset);
 
 void dp_wait_for_training_aux_rd_interval(
@@ -151,9 +151,11 @@ void dp_hw_to_dpcd_lane_settings(
 	const struct link_training_settings *lt_settings,
 	const struct dc_lane_settings hw_lane_settings[LANE_COUNT_DP_MAX],
 	union dpcd_training_lane dpcd_lane_settings[LANE_COUNT_DP_MAX]);
-void dp_update_drive_settings(
-	struct link_training_settings *dest,
-	struct link_training_settings src);
+void dp_decide_lane_settings(
+	const struct link_training_settings *lt_settings,
+	const union lane_adjust ln_adjust[LANE_COUNT_DP_MAX],
+	struct dc_lane_settings hw_lane_settings[LANE_COUNT_DP_MAX],
+	union dpcd_training_lane dpcd_lane_settings[LANE_COUNT_DP_MAX]);
 
 uint32_t dp_translate_training_aux_read_interval(uint32_t dpcd_aux_read_interval);
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
index 9f12792b7e59..3ef7faa92052 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -34,6 +34,8 @@ union defer_reg_writes {
 		bool disable_blnd_lut:1;
 		bool disable_3dlut:1;
 		bool disable_shaper:1;
+		bool disable_gamcor:1;
+		bool disable_dscl:1;
 	} bits;
 	uint32_t raw;
 };
diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h
index 365fbc4758e1..717c0e572d2f 100644
--- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h
+++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h
@@ -97,6 +97,7 @@ enum dmub_asic {
 	DMUB_ASIC_DCN302,
 	DMUB_ASIC_DCN303,
 	DMUB_ASIC_DCN31,
+	DMUB_ASIC_DCN31B,
 	DMUB_ASIC_MAX,
 };
 
@@ -237,6 +238,7 @@ struct dmub_srv_hw_params {
 	bool load_inst_const;
 	bool skip_panel_power_sequence;
 	bool disable_z10;
+	bool dpia_supported;
 	bool disable_dpia;
 };
 
diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
index 4c436fb4624b..0293c58f0701 100644
--- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
+++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
@@ -46,10 +46,10 @@
 
 /* Firmware versioning. */
 #ifdef DMUB_EXPOSE_VERSION
-#define DMUB_FW_VERSION_GIT_HASH 0xd146258f
+#define DMUB_FW_VERSION_GIT_HASH 0x9525efb5
 #define DMUB_FW_VERSION_MAJOR 0
 #define DMUB_FW_VERSION_MINOR 0
-#define DMUB_FW_VERSION_REVISION 88
+#define DMUB_FW_VERSION_REVISION 90
 #define DMUB_FW_VERSION_TEST 0
 #define DMUB_FW_VERSION_VBIOS 0
 #define DMUB_FW_VERSION_HOTFIX 0
@@ -370,8 +370,13 @@ union dmub_fw_boot_options {
 		uint32_t z10_disable: 1; /**< 1 to disable z10 */
 		uint32_t enable_dpia: 1; /**< 1 if DPIA should be enabled */
 		uint32_t invalid_vbios_data: 1; /**< 1 if VBIOS data table is invalid */
-		uint32_t reserved_unreleased2: 1; /**< reserved for an unreleased feature */
-		uint32_t reserved : 22; /**< reserved */
+		uint32_t dpia_supported: 1; /**< 1 if DPIA is supported on this platform */
+		uint32_t sel_mux_phy_c_d_phy_f_g: 1; /**< 1 if PHYF/PHYG should be enabled */
+		/**< 1 if all root clock gating is enabled and low power memory is enabled*/
+		uint32_t power_optimization: 1;
+		uint32_t diag_env: 1; /* 1 if diagnostic environment */
+
+		uint32_t reserved : 19; /**< reserved */
 	} bits; /**< boot bits */
 	uint32_t all; /**< 32-bit access to bits */
 };
@@ -692,6 +697,7 @@ enum dmub_out_cmd_type {
 enum dmub_cmd_dpia_type {
 	DMUB_CMD__DPIA_DIG1_DPIA_CONTROL = 0,
 	DMUB_CMD__DPIA_SET_CONFIG_ACCESS = 1,
+	DMUB_CMD__DPIA_MST_ALLOC_SLOTS = 2,
 };
 
 #pragma pack(push, 1)
@@ -1067,6 +1073,24 @@ struct dmub_rb_cmd_set_config_access {
 };
 
 /**
+ * Data passed from driver to FW in a DMUB_CMD__DPIA_MST_ALLOC_SLOTS command.
+ */
+struct dmub_cmd_mst_alloc_slots_control_data {
+	uint8_t mst_alloc_slots; /* mst slots to be allotted */
+	uint8_t instance; /* DPIA instance */
+	uint8_t immed_status; /* Immediate status returned as there is no outbox msg posted */
+	uint8_t mst_slots_in_use; /* returns slots in use for error cases */
+};
+
+/**
+ * DMUB command structure for SET_ command.
+ */
+struct dmub_rb_cmd_set_mst_alloc_slots {
+	struct dmub_cmd_header header; /* header */
+	struct dmub_cmd_mst_alloc_slots_control_data mst_slots_control; /* mst slots control */
+};
+
+/**
  * struct dmub_rb_cmd_dpphy_init - DPPHY init.
  */
 struct dmub_rb_cmd_dpphy_init {
@@ -1378,6 +1402,10 @@ enum dmub_cmd_psr_type {
 	 * Forces PSR enabled until an explicit PSR disable call.
 	 */
 	DMUB_CMD__PSR_FORCE_STATIC		= 5,
+	/**
+	 * Set PSR power option
+	 */
+	DMUB_CMD__SET_PSR_POWER_OPT = 7,
 };
 
 /**
@@ -1676,6 +1704,44 @@ struct dmub_rb_cmd_psr_force_static {
 };
 
 /**
+ * Data passed from driver to FW in a DMUB_CMD__SET_PSR_POWER_OPT command.
+ */
+struct dmub_cmd_psr_set_power_opt_data {
+	/**
+	 * PSR control version.
+	 */
+	uint8_t cmd_version;
+	/**
+	 * Panel Instance.
+	 * Panel isntance to identify which psr_state to use
+	 * Currently the support is only for 0 or 1
+	 */
+	uint8_t panel_inst;
+	/**
+	 * Explicit padding to 4 byte boundary.
+	 */
+	uint8_t pad[2];
+	/**
+	 * PSR power option
+	 */
+	uint32_t power_opt;
+};
+
+/**
+ * Definition of a DMUB_CMD__SET_PSR_POWER_OPT command.
+ */
+struct dmub_rb_cmd_psr_set_power_opt {
+	/**
+	 * Command header.
+	 */
+	struct dmub_cmd_header header;
+	/**
+	 * Definition of a DMUB_CMD__SET_PSR_POWER_OPT command.
+	 */
+	struct dmub_cmd_psr_set_power_opt_data psr_set_power_opt_data;
+};
+
+/**
  * Set of HW components that can be locked.
  *
  * Note: If updating with more HW components, fields
@@ -2459,6 +2525,10 @@ union dmub_rb_cmd {
 	 */
 	struct dmub_rb_cmd_psr_force_static psr_force_static;
 	/**
+	 * Definition of a DMUB_CMD__SET_PSR_POWER_OPT command.
+	 */
+	struct dmub_rb_cmd_psr_set_power_opt psr_set_power_opt;
+	/**
 	 * Definition of a DMUB_CMD__PLAT_54186_WA command.
 	 */
 	struct dmub_rb_cmd_PLAT_54186_wa PLAT_54186_wa;
@@ -2543,6 +2613,10 @@ union dmub_rb_cmd {
 	 */
 	struct dmub_rb_cmd_set_config_access set_config_access;
 	/**
+	 * Definition of a DMUB_CMD__DPIA_MST_ALLOC_SLOTS command.
+	 */
+	struct dmub_rb_cmd_set_mst_alloc_slots set_mst_alloc_slots;
+	/**
 	 * Definition of a DMUB_CMD__EDID_CEA command.
 	 */
 	struct dmub_rb_cmd_edid_cea edid_cea;
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c
index 3988f65f1ea4..10ebf20eaa41 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c
@@ -338,8 +338,11 @@ void dmub_dcn31_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmu
 	union dmub_fw_boot_options boot_options = {0};
 
 	boot_options.bits.z10_disable = params->disable_z10;
+	boot_options.bits.dpia_supported = params->dpia_supported;
 	boot_options.bits.enable_dpia = params->disable_dpia ? 0 : 1;
 
+	boot_options.bits.sel_mux_phy_c_d_phy_f_g = (dmub->asic == DMUB_ASIC_DCN31B) ? 1 : 0;
+
 	REG_WRITE(DMCUB_SCRATCH14, boot_options.all);
 }
 
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
index 77c67222cabd..56d400ffa7ac 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
@@ -208,6 +208,7 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic)
 		break;
 
 	case DMUB_ASIC_DCN31:
+	case DMUB_ASIC_DCN31B:
 		dmub->regs_dcn31 = &dmub_srv_dcn31_regs;
 		funcs->reset = dmub_dcn31_reset;
 		funcs->reset_release = dmub_dcn31_reset_release;
diff --git a/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h b/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h
index 792652236c61..dd974c428d23 100644
--- a/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h
+++ b/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h
@@ -328,6 +328,7 @@ struct integrated_info {
 
 		uint8_t gu_id[NUMBER_OF_UCHAR_FOR_GUID];
 		uint8_t checksum;
+		uint8_t fixdpvoltageswing;
 	} ext_disp_conn_info; /* exiting long long time */
 
 	struct available_s_clk_list {
diff --git a/drivers/gpu/drm/amd/display/include/i2caux_interface.h b/drivers/gpu/drm/amd/display/include/i2caux_interface.h
index c7fbb9c3ad6b..418fbf8c5c3a 100644
--- a/drivers/gpu/drm/amd/display/include/i2caux_interface.h
+++ b/drivers/gpu/drm/amd/display/include/i2caux_interface.h
@@ -41,6 +41,8 @@ struct aux_payload {
 	 * reset it to read data */
 	bool write;
 	bool mot;
+	bool write_status_update;
+
 	uint32_t address;
 	uint32_t length;
 	uint8_t *data;
@@ -53,6 +55,7 @@ struct aux_payload {
 	 * zero means "use default value"
 	 */
 	uint32_t defer_delay;
+
 };
 
 struct aux_command {
diff --git a/drivers/gpu/drm/amd/display/include/link_service_types.h b/drivers/gpu/drm/amd/display/include/link_service_types.h
index 9ffea7b40545..424bccd36434 100644
--- a/drivers/gpu/drm/amd/display/include/link_service_types.h
+++ b/drivers/gpu/drm/amd/display/include/link_service_types.h
@@ -90,8 +90,11 @@ enum lttpr_mode {
 
 struct link_training_settings {
 	struct dc_link_settings link_settings;
-	struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX];
 
+	/* TODO: turn lane settings below into mandatory fields
+	 * as initial lane configuration
+	 */
+	struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX];
 	enum dc_voltage_swing *voltage_swing;
 	enum dc_pre_emphasis *pre_emphasis;
 	enum dc_post_cursor2 *post_cursor2;
@@ -115,8 +118,30 @@ struct link_training_settings {
 #endif
 
 	bool enhanced_framing;
-	bool allow_invalid_msa_timing_param;
 	enum lttpr_mode lttpr_mode;
+
+	/* disallow different lanes to have different lane settings */
+	bool disallow_per_lane_settings;
+	/* dpcd lane settings will always use the same hw lane settings
+	 * even if it doesn't match requested lane adjust */
+	bool always_match_dpcd_with_hw_lane_settings;
+
+	/*****************************************************************
+	* training states - parameters that can change in link training
+	*****************************************************************/
+	/* TODO: Move hw_lane_settings and dpcd_lane_settings
+	 * along with lane adjust, lane align, offset and all
+	 * other training states into a new structure called
+	 * training states, so link_training_settings becomes
+	 * a constant input pre-decided prior to link training.
+	 *
+	 * The goal is to strictly decouple link training settings
+	 * decision making process from link training states to
+	 * prevent it from messy code practice of changing training
+	 * decision on the fly.
+	 */
+	struct dc_lane_settings hw_lane_settings[LANE_COUNT_DP_MAX];
+	union dpcd_training_lane dpcd_lane_settings[LANE_COUNT_DP_MAX];
 };
 
 /*TODO: Move this enum test harness*/
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index b99aa232bd8b..bd1d1dc93629 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -155,9 +155,18 @@ static unsigned int calc_v_total_from_duration(
 	if (duration_in_us > vrr->max_duration_in_us)
 		duration_in_us = vrr->max_duration_in_us;
 
-	v_total = div64_u64(div64_u64(((unsigned long long)(
-				duration_in_us) * (stream->timing.pix_clk_100hz / 10)),
-				stream->timing.h_total), 1000);
+	if (dc_is_hdmi_signal(stream->signal)) {
+		uint32_t h_total_up_scaled;
+
+		h_total_up_scaled = stream->timing.h_total * 10000;
+		v_total = div_u64((unsigned long long)duration_in_us
+					* stream->timing.pix_clk_100hz + (h_total_up_scaled - 1),
+					h_total_up_scaled);
+	} else {
+		v_total = div64_u64(div64_u64(((unsigned long long)(
+					duration_in_us) * (stream->timing.pix_clk_100hz / 10)),
+					stream->timing.h_total), 1000);
+	}
 
 	/* v_total cannot be less than nominal */
 	if (v_total < stream->timing.v_total) {
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
index e9bd84ec027d..be61975f1470 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
@@ -105,6 +105,7 @@ static enum mod_hdcp_status remove_display_from_topology_v3(
 	dtm_cmd->dtm_status = TA_DTM_STATUS__GENERIC_FAILURE;
 
 	psp_dtm_invoke(psp, dtm_cmd->cmd_id);
+	mutex_unlock(&psp->dtm_context.mutex);
 
 	if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS) {
 		status = remove_display_from_topology_v2(hdcp, index);
@@ -115,8 +116,6 @@ static enum mod_hdcp_status remove_display_from_topology_v3(
 		HDCP_TOP_REMOVE_DISPLAY_TRACE(hdcp, display->index);
 	}
 
-	mutex_unlock(&psp->dtm_context.mutex);
-
 	return status;
 }
 
@@ -205,6 +204,7 @@ static enum mod_hdcp_status add_display_to_topology_v3(
 	dtm_cmd->dtm_in_message.topology_update_v3.link_hdcp_cap = link->hdcp_supported_informational;
 
 	psp_dtm_invoke(psp, dtm_cmd->cmd_id);
+	mutex_unlock(&psp->dtm_context.mutex);
 
 	if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS) {
 		status = add_display_to_topology_v2(hdcp, display);
@@ -214,8 +214,6 @@ static enum mod_hdcp_status add_display_to_topology_v3(
 		HDCP_TOP_ADD_DISPLAY_TRACE(hdcp, display->index);
 	}
 
-	mutex_unlock(&psp->dtm_context.mutex);
-
 	return status;
 }
 
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
index f37101f5a777..6d648c889866 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
@@ -249,6 +249,8 @@ struct mod_hdcp_link {
 	uint8_t ddc_line;
 	uint8_t link_enc_idx;
 	uint8_t phy_idx;
+	uint8_t dio_output_type;
+	uint8_t dio_output_id;
 	uint8_t hdcp_supported_informational;
 	union {
 		struct mod_hdcp_displayport dp;
diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
index 01cca08a774f..49fe4155c374 100644
--- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
@@ -310,7 +310,7 @@ static ssize_t amdgpu_set_power_dpm_force_performance_level(struct device *dev,
 	struct amdgpu_device *adev = drm_to_adev(ddev);
 	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
 	enum amd_dpm_forced_level level;
-	enum amd_dpm_forced_level current_level = 0xff;
+	enum amd_dpm_forced_level current_level;
 	int ret = 0;
 
 	if (amdgpu_in_reset(adev))
@@ -350,6 +350,8 @@ static ssize_t amdgpu_set_power_dpm_force_performance_level(struct device *dev,
 
 	if (pp_funcs->get_performance_level)
 		current_level = amdgpu_dpm_get_performance_level(adev);
+	else
+		current_level = adev->pm.dpm.forced_level;
 
 	if (current_level == level) {
 		pm_runtime_mark_last_busy(ddev->dev);
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 0cbd1540aeac..26e45fc5eb1a 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -786,13 +786,6 @@ struct drm_amdgpu_cs_chunk_data {
 #define AMDGPU_INFO_VRAM_LOST_COUNTER		0x1F
 /* query ras mask of enabled features*/
 #define AMDGPU_INFO_RAS_ENABLED_FEATURES	0x20
-/* query video encode/decode caps */
-#define AMDGPU_INFO_VIDEO_CAPS			0x21
-	/* Subquery id: Decode */
-	#define AMDGPU_INFO_VIDEO_CAPS_DECODE		0
-	/* Subquery id: Encode */
-	#define AMDGPU_INFO_VIDEO_CAPS_ENCODE		1
-
 /* RAS MASK: UMC (VRAM) */
 #define AMDGPU_INFO_RAS_ENABLED_UMC			(1 << 0)
 /* RAS MASK: SDMA */
@@ -821,6 +814,12 @@ struct drm_amdgpu_cs_chunk_data {
 #define AMDGPU_INFO_RAS_ENABLED_MP1			(1 << 12)
 /* RAS MASK: FUSE */
 #define AMDGPU_INFO_RAS_ENABLED_FUSE			(1 << 13)
+/* query video encode/decode caps */
+#define AMDGPU_INFO_VIDEO_CAPS			0x21
+	/* Subquery id: Decode */
+	#define AMDGPU_INFO_VIDEO_CAPS_DECODE		0
+	/* Subquery id: Encode */
+	#define AMDGPU_INFO_VIDEO_CAPS_ENCODE		1
 
 #define AMDGPU_INFO_MMR_SE_INDEX_SHIFT	0
 #define AMDGPU_INFO_MMR_SE_INDEX_MASK	0xff
author	Dave Airlie <airlied@redhat.com>	2021-11-02 05:40:57 +0300
committer	Dave Airlie <airlied@redhat.com>	2021-11-02 05:40:58 +0300
commit	d9bd054177fbd2c4762546aec40fc3071bfe4cc0 (patch)
tree	70e34d106c87a7a7fefb810213786a0b8a19b69c
parent	31fa8cbce4664946a1688898410fee41ad05364d (diff)
parent	403475be6d8b122c3e6b8a47e075926d7299e5ef (diff)
download	linux-d9bd054177fbd2c4762546aec40fc3071bfe4cc0.tar.xz