Merge drm/drm-next into drm-misc-next

Let's kickstart the v6.20 (7.0?) release cycle.

Signed-off-by: Maxime Ripard <mripard@kernel.org>

203 files changed, 6694 insertions, 2099 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig
index 1acfed2f92ef..7f515be5185d 100644
--- a/drivers/gpu/drm/amd/amdgpu/Kconfig
+++ b/drivers/gpu/drm/amd/amdgpu/Kconfig
@@ -43,14 +43,16 @@ config DRM_AMDGPU_SI
 	bool "Enable amdgpu support for SI parts"
 	depends on DRM_AMDGPU
 	help
-	  Choose this option if you want to enable experimental support
+	  Choose this option if you want to enable support
 	  for SI (Southern Islands) asics.
 
-	  SI is already supported in radeon. Experimental support for SI
-	  in amdgpu will be disabled by default and is still provided by
-	  radeon. Use module options to override this:
+	  SI (Southern Islands) are first generation GCN GPUs,
+	  supported by both drivers: radeon (old) and amdgpu (new).
+	  By default, SI dedicated GPUs are supported by amdgpu.
 
-	  radeon.si_support=0 amdgpu.si_support=1
+	  Use module options to override this:
+	  To use radeon for SI,
+	  radeon.si_support=1 amdgpu.si_support=0
 
 config DRM_AMDGPU_CIK
 	bool "Enable amdgpu support for CIK parts"
@@ -59,11 +61,17 @@ config DRM_AMDGPU_CIK
 	  Choose this option if you want to enable support for CIK (Sea
 	  Islands) asics.
 
-	  CIK is already supported in radeon. Support for CIK in amdgpu
-	  will be disabled by default and is still provided by radeon.
-	  Use module options to override this:
+	  CIK (Sea Islands) are second generation GCN GPUs,
+	  supported by both drivers: radeon (old) and amdgpu (new).
+	  By default,
+	  CIK dedicated GPUs are supported by amdgpu
+	  CIK APUs are supported by radeon
 
+	  Use module options to override this:
+	  To use amdgpu for CIK,
 	  radeon.cik_support=0 amdgpu.cik_support=1
+	  To use radeon for CIK,
+	  radeon.cik_support=1 amdgpu.cik_support=0
 
 config DRM_AMDGPU_USERPTR
 	bool "Always enable userptr write support"
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index ebe08947c5a3..c88760fb52ea 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -78,7 +78,7 @@ amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o \
 	dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o
 
 amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o \
-	uvd_v3_1.o
+	uvd_v3_1.o vce_v1_0.o
 
 amdgpu-y += \
 	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o \
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 50079209c472..9f9774f58ce1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1316,6 +1316,7 @@ struct amdgpu_device {
 	bool                            apu_prefer_gtt;
 
 	bool                            userq_halt_for_enforce_isolation;
+	struct work_struct              userq_reset_work;
 	struct amdgpu_uid *uid_info;
 
 	/* KFD
@@ -1539,11 +1540,6 @@ int emu_soc_asic_init(struct amdgpu_device *adev);
 #define amdgpu_asic_read_bios_from_rom(adev, b, l) (adev)->asic_funcs->read_bios_from_rom((adev), (b), (l))
 #define amdgpu_asic_read_register(adev, se, sh, offset, v)((adev)->asic_funcs->read_register((adev), (se), (sh), (offset), (v)))
 #define amdgpu_asic_get_config_memsize(adev) (adev)->asic_funcs->get_config_memsize((adev))
-#define amdgpu_asic_flush_hdp(adev, r) \
-	((adev)->asic_funcs->flush_hdp ? (adev)->asic_funcs->flush_hdp((adev), (r)) : (adev)->hdp.funcs->flush_hdp((adev), (r)))
-#define amdgpu_asic_invalidate_hdp(adev, r) \
-	((adev)->asic_funcs->invalidate_hdp ? (adev)->asic_funcs->invalidate_hdp((adev), (r)) : \
-	 ((adev)->hdp.funcs->invalidate_hdp ? (adev)->hdp.funcs->invalidate_hdp((adev), (r)) : (void)0))
 #define amdgpu_asic_need_full_reset(adev) (adev)->asic_funcs->need_full_reset((adev))
 #define amdgpu_asic_init_doorbell_index(adev) (adev)->asic_funcs->init_doorbell_index((adev))
 #define amdgpu_asic_get_pcie_usage(adev, cnt0, cnt1) ((adev)->asic_funcs->get_pcie_usage((adev), (cnt0), (cnt1)))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
index 4926996f94da..381ef205b0df 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
@@ -302,17 +302,19 @@ static int acp_hw_init(struct amdgpu_ip_block *ip_block)
 		adev->acp.acp_res[2].end = adev->acp.acp_res[2].start;
 
 		adev->acp.acp_cell[0].name = "acp_audio_dma";
+		adev->acp.acp_cell[0].id = 0;
 		adev->acp.acp_cell[0].num_resources = 3;
 		adev->acp.acp_cell[0].resources = &adev->acp.acp_res[0];
 		adev->acp.acp_cell[0].platform_data = &adev->asic_type;
 		adev->acp.acp_cell[0].pdata_size = sizeof(adev->asic_type);
 
 		adev->acp.acp_cell[1].name = "designware-i2s";
+		adev->acp.acp_cell[1].id = 1;
 		adev->acp.acp_cell[1].num_resources = 1;
 		adev->acp.acp_cell[1].resources = &adev->acp.acp_res[1];
 		adev->acp.acp_cell[1].platform_data = &i2s_pdata[0];
 		adev->acp.acp_cell[1].pdata_size = sizeof(struct i2s_platform_data);
-		r = mfd_add_hotplug_devices(adev->acp.parent, adev->acp.acp_cell, 2);
+		r = mfd_add_devices(adev->acp.parent, 0, adev->acp.acp_cell, 2, NULL, 0, NULL);
 		if (r)
 			goto failure;
 		r = device_for_each_child(adev->acp.parent, &adev->acp.acp_genpd->gpd,
@@ -410,30 +412,34 @@ static int acp_hw_init(struct amdgpu_ip_block *ip_block)
 		adev->acp.acp_res[4].end = adev->acp.acp_res[4].start;
 
 		adev->acp.acp_cell[0].name = "acp_audio_dma";
+		adev->acp.acp_cell[0].id = 0;
 		adev->acp.acp_cell[0].num_resources = 5;
 		adev->acp.acp_cell[0].resources = &adev->acp.acp_res[0];
 		adev->acp.acp_cell[0].platform_data = &adev->asic_type;
 		adev->acp.acp_cell[0].pdata_size = sizeof(adev->asic_type);
 
 		adev->acp.acp_cell[1].name = "designware-i2s";
+		adev->acp.acp_cell[1].id = 1;
 		adev->acp.acp_cell[1].num_resources = 1;
 		adev->acp.acp_cell[1].resources = &adev->acp.acp_res[1];
 		adev->acp.acp_cell[1].platform_data = &i2s_pdata[0];
 		adev->acp.acp_cell[1].pdata_size = sizeof(struct i2s_platform_data);
 
 		adev->acp.acp_cell[2].name = "designware-i2s";
+		adev->acp.acp_cell[2].id = 2;
 		adev->acp.acp_cell[2].num_resources = 1;
 		adev->acp.acp_cell[2].resources = &adev->acp.acp_res[2];
 		adev->acp.acp_cell[2].platform_data = &i2s_pdata[1];
 		adev->acp.acp_cell[2].pdata_size = sizeof(struct i2s_platform_data);
 
 		adev->acp.acp_cell[3].name = "designware-i2s";
+		adev->acp.acp_cell[3].id = 3;
 		adev->acp.acp_cell[3].num_resources = 1;
 		adev->acp.acp_cell[3].resources = &adev->acp.acp_res[3];
 		adev->acp.acp_cell[3].platform_data = &i2s_pdata[2];
 		adev->acp.acp_cell[3].pdata_size = sizeof(struct i2s_platform_data);
 
-		r = mfd_add_hotplug_devices(adev->acp.parent, adev->acp.acp_cell, ACP_DEVS);
+		r = mfd_add_devices(adev->acp.parent, 0, adev->acp.acp_cell, ACP_DEVS, NULL, 0, NULL);
 		if (r)
 			goto failure;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 644f79f3c9af..a2879d2b7c8e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -36,7 +36,6 @@
 #include "amdgpu_ras.h"
 #include "amdgpu_umc.h"
 #include "amdgpu_reset.h"
-#include "amdgpu_ras_mgr.h"
 
 /* Total memory size in system memory and all GPU VRAM. Used to
  * estimate worst case amount of memory to reserve for page tables
@@ -747,20 +746,6 @@ void amdgpu_amdkfd_ras_pasid_poison_consumption_handler(struct amdgpu_device *ad
 				enum amdgpu_ras_block block, uint16_t pasid,
 				pasid_notify pasid_fn, void *data, uint32_t reset)
 {
-
-	if (amdgpu_uniras_enabled(adev)) {
-		struct ras_ih_info ih_info;
-
-		memset(&ih_info, 0, sizeof(ih_info));
-		ih_info.block = block;
-		ih_info.pasid = pasid;
-		ih_info.reset = reset;
-		ih_info.pasid_fn = pasid_fn;
-		ih_info.data = data;
-		amdgpu_ras_mgr_handle_consumer_interrupt(adev, &ih_info);
-		return;
-	}
-
 	amdgpu_umc_pasid_poison_handler(adev, block, pasid, pasid_fn, data, reset);
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 96ccd5ade031..b1c24c8fa686 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -1274,6 +1274,10 @@ static int unmap_bo_from_gpuvm(struct kgd_mem *mem,
 
 	(void)amdgpu_vm_bo_unmap(adev, bo_va, entry->va);
 
+	/* VM entity stopped if process killed, don't clear freed pt bo */
+	if (!amdgpu_vm_ready(vm))
+		return 0;
+
 	(void)amdgpu_vm_clear_freed(adev, vm, &bo_va->last_pt_update);
 
 	(void)amdgpu_sync_fence(sync, bo_va->last_pt_update, GFP_KERNEL);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index f5d5c45ddc0d..afedea02188d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -236,7 +236,7 @@ static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, u32 hw_ip,
 		r = amdgpu_xcp_select_scheds(adev, hw_ip, hw_prio, fpriv,
 						&num_scheds, &scheds);
 		if (r)
-			goto cleanup_entity;
+			goto error_free_entity;
 	}
 
 	/* disable load balance if the hw engine retains context among dependent jobs */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c
index 8a026bc9ea44..4e2fe6674db8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c
@@ -217,8 +217,7 @@ amdgpu_devcoredump_read(char *buffer, loff_t offset, size_t count,
 	drm_printf(&p, "version: " AMDGPU_COREDUMP_VERSION "\n");
 	drm_printf(&p, "kernel: " UTS_RELEASE "\n");
 	drm_printf(&p, "module: " KBUILD_MODNAME "\n");
-	drm_printf(&p, "time: %lld.%09ld\n", coredump->reset_time.tv_sec,
-		   coredump->reset_time.tv_nsec);
+	drm_printf(&p, "time: %ptSp\n", &coredump->reset_time);
 
 	if (coredump->reset_task_info.task.pid)
 		drm_printf(&p, "process_name: %s PID: %d\n",
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 654f4844b7ad..58c3ffe707d1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -71,6 +71,7 @@
 
 #include "amdgpu_xgmi.h"
 #include "amdgpu_ras.h"
+#include "amdgpu_ras_mgr.h"
 #include "amdgpu_pmu.h"
 #include "amdgpu_fru_eeprom.h"
 #include "amdgpu_reset.h"
@@ -179,6 +180,10 @@ struct amdgpu_init_level amdgpu_init_minimal_xgmi = {
 		BIT(AMD_IP_BLOCK_TYPE_PSP)
 };
 
+static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev);
+static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev);
+static int amdgpu_device_ip_resume_phase3(struct amdgpu_device *adev);
+
 static void amdgpu_device_load_switch_state(struct amdgpu_device *adev);
 
 static inline bool amdgpu_ip_member_of_hwini(struct amdgpu_device *adev,
@@ -1673,9 +1678,9 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
 	int rbar_size = pci_rebar_bytes_to_size(adev->gmc.real_vram_size);
 	struct pci_bus *root;
 	struct resource *res;
+	int max_size, r;
 	unsigned int i;
 	u16 cmd;
-	int r;
 
 	if (!IS_ENABLED(CONFIG_PHYS_ADDR_T_64BIT))
 		return 0;
@@ -1721,30 +1726,28 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
 		return 0;
 
 	/* Limit the BAR size to what is available */
-	rbar_size = min(fls(pci_rebar_get_possible_sizes(adev->pdev, 0)) - 1,
-			rbar_size);
+	max_size = pci_rebar_get_max_size(adev->pdev, 0);
+	if (max_size < 0)
+		return 0;
+	rbar_size = min(max_size, rbar_size);
 
 	/* Disable memory decoding while we change the BAR addresses and size */
 	pci_read_config_word(adev->pdev, PCI_COMMAND, &cmd);
 	pci_write_config_word(adev->pdev, PCI_COMMAND,
 			      cmd & ~PCI_COMMAND_MEMORY);
 
-	/* Free the VRAM and doorbell BAR, we most likely need to move both. */
+	/* Tear down doorbell as resizing will release BARs */
 	amdgpu_doorbell_fini(adev);
-	if (adev->asic_type >= CHIP_BONAIRE)
-		pci_release_resource(adev->pdev, 2);
-
-	pci_release_resource(adev->pdev, 0);
 
-	r = pci_resize_resource(adev->pdev, 0, rbar_size);
+	r = pci_resize_resource(adev->pdev, 0, rbar_size,
+				(adev->asic_type >= CHIP_BONAIRE) ? 1 << 5
+								  : 1 << 2);
 	if (r == -ENOSPC)
 		dev_info(adev->dev,
 			 "Not enough PCI address space for a large BAR.");
 	else if (r && r != -ENOTSUPP)
 		dev_err(adev->dev, "Problem resizing BAR0 (%d).", r);
 
-	pci_assign_unassigned_bus_resources(adev->pdev->bus);
-
 	/* When the doorbell or fb BAR isn't available we have no chance of
 	 * using the device.
 	 */
@@ -2494,6 +2497,7 @@ static const char *ip_block_names[] = {
 	[AMD_IP_BLOCK_TYPE_VPE] = "vpe",
 	[AMD_IP_BLOCK_TYPE_UMSCH_MM] = "umsch_mm",
 	[AMD_IP_BLOCK_TYPE_ISP] = "isp",
+	[AMD_IP_BLOCK_TYPE_RAS] = "ras",
 };
 
 static const char *ip_block_name(struct amdgpu_device *adev, enum amd_ip_block_type type)
@@ -2659,6 +2663,8 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 		chip_name = "navi12";
 		break;
 	case CHIP_CYAN_SKILLFISH:
+		if (adev->discovery.bin)
+			return 0;
 		chip_name = "cyan_skillfish";
 		break;
 	}
@@ -3439,10 +3445,11 @@ int amdgpu_device_set_pg_state(struct amdgpu_device *adev,
 		    (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GFX ||
 		     adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SDMA))
 			continue;
-		/* skip CG for VCE/UVD, it's handled specially */
+		/* skip CG for VCE/UVD/VPE, it's handled specially */
 		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
 		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE &&
 		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN &&
+		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VPE &&
 		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_JPEG &&
 		    adev->ip_blocks[i].version->funcs->set_powergating_state) {
 			/* enable powergating to save power */
@@ -3674,6 +3681,20 @@ static int amdgpu_device_ip_fini_early(struct amdgpu_device *adev)
 				"failed to release exclusive mode on fini\n");
 	}
 
+	/*
+	 * Driver reload on the APU can fail due to firmware validation because
+	 * the PSP is always running, as it is shared across the whole SoC.
+	 * This same issue does not occur on dGPU because it has a mechanism
+	 * that checks whether the PSP is running. A solution for those issues
+	 * in the APU is to trigger a GPU reset, but this should be done during
+	 * the unload phase to avoid adding boot latency and screen flicker.
+	 */
+	if ((adev->flags & AMD_IS_APU) && !adev->gmc.is_app_apu) {
+		r = amdgpu_asic_reset(adev);
+		if (r)
+			dev_err(adev->dev, "asic reset on %s failed\n", __func__);
+	}
+
 	return 0;
 }
 
@@ -3784,7 +3805,7 @@ static void amdgpu_device_delay_enable_gfx_off(struct work_struct *work)
  */
 static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
 {
-	int i, r;
+	int i, r, rec;
 
 	amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE);
 	amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE);
@@ -3807,10 +3828,23 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
 
 		r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]);
 		if (r)
-			return r;
+			goto unwind;
 	}
 
 	return 0;
+unwind:
+	rec = amdgpu_device_ip_resume_phase3(adev);
+	if (rec)
+		dev_err(adev->dev,
+			"amdgpu_device_ip_resume_phase3 failed during unwind: %d\n",
+			rec);
+
+	amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_ALLOW);
+
+	amdgpu_device_set_pg_state(adev, AMD_PG_STATE_GATE);
+	amdgpu_device_set_cg_state(adev, AMD_CG_STATE_GATE);
+
+	return r;
 }
 
 /**
@@ -3826,7 +3860,7 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
  */
 static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 {
-	int i, r;
+	int i, r, rec;
 
 	if (adev->in_s0ix)
 		amdgpu_dpm_gfx_state_change(adev, sGpuChangeState_D3Entry);
@@ -3889,7 +3923,7 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 
 		r = amdgpu_ip_block_suspend(&adev->ip_blocks[i]);
 		if (r)
-			return r;
+			goto unwind;
 
 		/* handle putting the SMC in the appropriate state */
 		if (!amdgpu_sriov_vf(adev)) {
@@ -3899,13 +3933,40 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 					dev_err(adev->dev,
 						"SMC failed to set mp1 state %d, %d\n",
 						adev->mp1_state, r);
-					return r;
+					goto unwind;
 				}
 			}
 		}
 	}
 
 	return 0;
+unwind:
+	/* suspend phase 2 = resume phase 1 + resume phase 2 */
+	rec = amdgpu_device_ip_resume_phase1(adev);
+	if (rec) {
+		dev_err(adev->dev,
+			"amdgpu_device_ip_resume_phase1 failed during unwind: %d\n",
+			rec);
+		return r;
+	}
+
+	rec = amdgpu_device_fw_loading(adev);
+	if (rec) {
+		dev_err(adev->dev,
+			"amdgpu_device_fw_loading failed during unwind: %d\n",
+			rec);
+		return r;
+	}
+
+	rec = amdgpu_device_ip_resume_phase2(adev);
+	if (rec) {
+		dev_err(adev->dev,
+			"amdgpu_device_ip_resume_phase2 failed during unwind: %d\n",
+			rec);
+		return r;
+	}
+
+	return r;
 }
 
 /**
@@ -4203,24 +4264,13 @@ bool amdgpu_device_asic_has_dc_support(struct pci_dev *pdev,
 	case CHIP_PITCAIRN:
 	case CHIP_VERDE:
 	case CHIP_OLAND:
-		/*
-		 * We have systems in the wild with these ASICs that require
-		 * LVDS and VGA support which is not supported with DC.
-		 *
-		 * Fallback to the non-DC driver here by default so as not to
-		 * cause regressions.
-		 */
-#if defined(CONFIG_DRM_AMD_DC_SI)
-		return amdgpu_dc > 0;
-#else
-		return false;
-#endif
+		return amdgpu_dc != 0 && IS_ENABLED(CONFIG_DRM_AMD_DC_SI);
 	case CHIP_KAVERI:
 	case CHIP_KABINI:
 	case CHIP_MULLINS:
 		/*
 		 * We have systems in the wild with these ASICs that require
-		 * VGA support which is not supported with DC.
+		 * TRAVIS and NUTMEG support which is not supported with DC.
 		 *
 		 * Fallback to the non-DC driver here by default so as not to
 		 * cause regressions.
@@ -4607,6 +4657,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	}
 
 	INIT_WORK(&adev->xgmi_reset_work, amdgpu_device_xgmi_reset_func);
+	INIT_WORK(&adev->userq_reset_work, amdgpu_userq_reset_work);
 
 	adev->gfx.gfx_off_req_count = 1;
 	adev->gfx.gfx_off_residency = 0;
@@ -5229,7 +5280,7 @@ void amdgpu_device_complete(struct drm_device *dev)
 int amdgpu_device_suspend(struct drm_device *dev, bool notify_clients)
 {
 	struct amdgpu_device *adev = drm_to_adev(dev);
-	int r = 0;
+	int r, rec;
 
 	if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
@@ -5245,8 +5296,9 @@ int amdgpu_device_suspend(struct drm_device *dev, bool notify_clients)
 			return r;
 	}
 
-	if (amdgpu_acpi_smart_shift_update(adev, AMDGPU_SS_DEV_D3))
-		dev_warn(adev->dev, "smart shift update failed\n");
+	r = amdgpu_acpi_smart_shift_update(adev, AMDGPU_SS_DEV_D3);
+	if (r)
+		goto unwind_sriov;
 
 	if (notify_clients)
 		drm_client_dev_suspend(adev_to_drm(adev));
@@ -5257,16 +5309,16 @@ int amdgpu_device_suspend(struct drm_device *dev, bool notify_clients)
 
 	r = amdgpu_device_ip_suspend_phase1(adev);
 	if (r)
-		return r;
+		goto unwind_smartshift;
 
 	amdgpu_amdkfd_suspend(adev, !amdgpu_sriov_vf(adev) && !adev->in_runpm);
 	r = amdgpu_userq_suspend(adev);
 	if (r)
-		return r;
+		goto unwind_ip_phase1;
 
 	r = amdgpu_device_evict_resources(adev);
 	if (r)
-		return r;
+		goto unwind_userq;
 
 	amdgpu_ttm_set_buffer_funcs_status(adev, false);
 
@@ -5274,16 +5326,62 @@ int amdgpu_device_suspend(struct drm_device *dev, bool notify_clients)
 
 	r = amdgpu_device_ip_suspend_phase2(adev);
 	if (r)
-		return r;
+		goto unwind_evict;
 
 	if (amdgpu_sriov_vf(adev))
 		amdgpu_virt_release_full_gpu(adev, false);
 
-	r = amdgpu_dpm_notify_rlc_state(adev, false);
-	if (r)
+	return 0;
+
+unwind_evict:
+	if (adev->mman.buffer_funcs_ring->sched.ready)
+		amdgpu_ttm_set_buffer_funcs_status(adev, true);
+	amdgpu_fence_driver_hw_init(adev);
+
+unwind_userq:
+	rec = amdgpu_userq_resume(adev);
+	if (rec) {
+		dev_warn(adev->dev, "failed to re-initialize user queues: %d\n", rec);
 		return r;
+	}
+	rec = amdgpu_amdkfd_resume(adev, !amdgpu_sriov_vf(adev) && !adev->in_runpm);
+	if (rec) {
+		dev_warn(adev->dev, "failed to re-initialize kfd: %d\n", rec);
+		return r;
+	}
 
-	return 0;
+unwind_ip_phase1:
+	/* suspend phase 1 = resume phase 3 */
+	rec = amdgpu_device_ip_resume_phase3(adev);
+	if (rec) {
+		dev_warn(adev->dev, "failed to re-initialize IPs phase1: %d\n", rec);
+		return r;
+	}
+
+unwind_smartshift:
+	rec = amdgpu_acpi_smart_shift_update(adev, AMDGPU_SS_DEV_D0);
+	if (rec) {
+		dev_warn(adev->dev, "failed to re-update smart shift: %d\n", rec);
+		return r;
+	}
+
+	if (notify_clients)
+		drm_client_dev_resume(adev_to_drm(adev));
+
+	amdgpu_ras_resume(adev);
+
+unwind_sriov:
+	if (amdgpu_sriov_vf(adev)) {
+		rec = amdgpu_virt_request_full_gpu(adev, true);
+		if (rec) {
+			dev_warn(adev->dev, "failed to reinitialize sriov: %d\n", rec);
+			return r;
+		}
+	}
+
+	adev->in_suspend = adev->in_s0ix = adev->in_s3 = false;
+
+	return r;
 }
 
 static inline int amdgpu_virt_resume(struct amdgpu_device *adev)
@@ -5989,6 +6087,10 @@ int amdgpu_device_reinit_after_reset(struct amdgpu_reset_context *reset_context)
 				if (r)
 					goto out;
 
+				r = amdgpu_userq_post_reset(tmp_adev, vram_lost);
+				if (r)
+					goto out;
+
 				drm_client_dev_resume(adev_to_drm(tmp_adev));
 
 				/*
@@ -6211,6 +6313,7 @@ static inline void amdgpu_device_stop_pending_resets(struct amdgpu_device *adev)
 	if (!amdgpu_sriov_vf(adev))
 		cancel_work(&adev->reset_work);
 #endif
+	cancel_work(&adev->userq_reset_work);
 
 	if (adev->kfd.dev)
 		cancel_work(&adev->kfd.reset_work);
@@ -6331,6 +6434,8 @@ static void amdgpu_device_halt_activities(struct amdgpu_device *adev,
 		    amdgpu_device_ip_need_full_reset(tmp_adev))
 			amdgpu_ras_suspend(tmp_adev);
 
+		amdgpu_userq_pre_reset(tmp_adev);
+
 		for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 			struct amdgpu_ring *ring = tmp_adev->rings[i];
 
@@ -6560,6 +6665,9 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 			goto end_reset;
 	}
 
+	/* Cannot be called after locking reset domain */
+	amdgpu_ras_pre_reset(adev, &device_list);
+
 	/* We need to lock reset domain only once both for XGMI and single device */
 	amdgpu_device_recovery_get_reset_lock(adev, &device_list);
 
@@ -6590,6 +6698,7 @@ skip_sched_resume:
 	amdgpu_device_gpu_resume(adev, &device_list, need_emergency_restart);
 reset_unlock:
 	amdgpu_device_recovery_put_reset_lock(adev, &device_list);
+	amdgpu_ras_post_reset(adev, &device_list);
 end_reset:
 	if (hive) {
 		mutex_unlock(&hive->hive_lock);
@@ -7327,7 +7436,7 @@ void amdgpu_device_flush_hdp(struct amdgpu_device *adev,
 			return;
 	}
 
-	amdgpu_asic_flush_hdp(adev, ring);
+	amdgpu_hdp_flush(adev, ring);
 }
 
 void amdgpu_device_invalidate_hdp(struct amdgpu_device *adev,
@@ -7340,7 +7449,7 @@ void amdgpu_device_invalidate_hdp(struct amdgpu_device *adev,
 	if (adev->gmc.xgmi.connected_to_cpu)
 		return;
 
-	amdgpu_asic_invalidate_hdp(adev, ring);
+	amdgpu_hdp_invalidate(adev, ring);
 }
 
 int amdgpu_in_reset(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
index 8561ad7f6180..e22cfa7c6d32 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
@@ -81,13 +81,44 @@ static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf,
 	struct drm_gem_object *obj = dmabuf->priv;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	int r;
+
+	/*
+	 * Disable peer-to-peer access for DCC-enabled VRAM surfaces on GFX12+.
+	 * Such buffers cannot be safely accessed over P2P due to device-local
+	 * compression metadata. Fallback to system-memory path instead.
+	 * Device supports GFX12 (GC 12.x or newer)
+	 * BO was created with the AMDGPU_GEM_CREATE_GFX12_DCC flag
+	 *
+	 */
+	if (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(12, 0, 0) &&
+	    bo->flags & AMDGPU_GEM_CREATE_GFX12_DCC)
+		attach->peer2peer = false;
+
+	/*
+	 * Disable peer-to-peer access for DCC-enabled VRAM surfaces on GFX12+.
+	 * Such buffers cannot be safely accessed over P2P due to device-local
+	 * compression metadata. Fallback to system-memory path instead.
+	 * Device supports GFX12 (GC 12.x or newer)
+	 * BO was created with the AMDGPU_GEM_CREATE_GFX12_DCC flag
+	 *
+	 */
+	if (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(12, 0, 0) &&
+	    bo->flags & AMDGPU_GEM_CREATE_GFX12_DCC)
+		attach->peer2peer = false;
 
 	if (!amdgpu_dmabuf_is_xgmi_accessible(attach_adev, bo) &&
 	    pci_p2pdma_distance(adev->pdev, attach->dev, false) < 0)
 		attach->peer2peer = false;
 
+	r = dma_resv_lock(bo->tbo.base.resv, NULL);
+	if (r)
+		return r;
+
 	amdgpu_vm_bo_update_shared(bo);
 
+	dma_resv_unlock(bo->tbo.base.resv);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 3776901bbb1b..2dfbddcef9ab 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -312,7 +312,7 @@ module_param_named(moverate, amdgpu_moverate, int, 0600);
  * DOC: audio (int)
  * Set HDMI/DPAudio. Only affects non-DC display handling. The default is -1 (Enabled), set 0 to disabled it.
  */
-MODULE_PARM_DESC(audio, "Audio enable (-1 = auto, 0 = disable, 1 = enable)");
+MODULE_PARM_DESC(audio, "HDMI/DP Audio enable for non DC displays (-1 = auto, 0 = disable, 1 = enable)");
 module_param_named(audio, amdgpu_audio, int, 0444);
 
 /**
@@ -618,39 +618,39 @@ module_param_named(timeout_period, amdgpu_watchdog_timer.period, uint, 0644);
 
 /**
  * DOC: si_support (int)
- * Set SI support driver. This parameter works after set config CONFIG_DRM_AMDGPU_SI. For SI asic, when radeon driver is enabled,
- * set value 0 to use radeon driver, while set value 1 to use amdgpu driver. The default is using radeon driver when it available,
- * otherwise using amdgpu driver.
- */
+ * 1 = enabled, 0 = disabled, -1 = default
+ *
+ * SI (Southern Islands) are first generation GCN GPUs, supported by both
+ * drivers: radeon (old) and amdgpu (new). This parameter controls whether
+ * amdgpu should support SI.
+ * By default, SI dedicated GPUs are supported by amdgpu.
+ * Only relevant when CONFIG_DRM_AMDGPU_SI is enabled to build SI support in amdgpu.
+ * See also radeon.si_support which should be disabled when amdgpu.si_support is
+ * enabled, and vice versa.
+ */
+int amdgpu_si_support = -1;
 #ifdef CONFIG_DRM_AMDGPU_SI
-
-#if IS_ENABLED(CONFIG_DRM_RADEON) || IS_ENABLED(CONFIG_DRM_RADEON_MODULE)
-int amdgpu_si_support;
-MODULE_PARM_DESC(si_support, "SI support (1 = enabled, 0 = disabled (default))");
-#else
-int amdgpu_si_support = 1;
-MODULE_PARM_DESC(si_support, "SI support (1 = enabled (default), 0 = disabled)");
-#endif
-
+MODULE_PARM_DESC(si_support, "SI support (1 = enabled, 0 = disabled, -1 = default)");
 module_param_named(si_support, amdgpu_si_support, int, 0444);
 #endif
 
 /**
  * DOC: cik_support (int)
- * Set CIK support driver. This parameter works after set config CONFIG_DRM_AMDGPU_CIK. For CIK asic, when radeon driver is enabled,
- * set value 0 to use radeon driver, while set value 1 to use amdgpu driver. The default is using radeon driver when it available,
- * otherwise using amdgpu driver.
- */
+ * 1 = enabled, 0 = disabled, -1 = default
+ *
+ * CIK (Sea Islands) are second generation GCN GPUs, supported by both
+ * drivers: radeon (old) and amdgpu (new). This parameter controls whether
+ * amdgpu should support CIK.
+ * By default:
+ * - CIK dedicated GPUs are supported by amdgpu.
+ * - CIK APUs are supported by radeon (except when radeon is not built).
+ * Only relevant when CONFIG_DRM_AMDGPU_CIK is enabled to build CIK support in amdgpu.
+ * See also radeon.cik_support which should be disabled when amdgpu.cik_support is
+ * enabled, and vice versa.
+ */
+int amdgpu_cik_support = -1;
 #ifdef CONFIG_DRM_AMDGPU_CIK
-
-#if IS_ENABLED(CONFIG_DRM_RADEON) || IS_ENABLED(CONFIG_DRM_RADEON_MODULE)
-int amdgpu_cik_support;
-MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled, 0 = disabled (default))");
-#else
-int amdgpu_cik_support = 1;
-MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled (default), 0 = disabled)");
-#endif
-
+MODULE_PARM_DESC(cik_support, "CIK support  (1 = enabled, 0 = disabled, -1 = default)");
 module_param_named(cik_support, amdgpu_cik_support, int, 0444);
 #endif
 
@@ -2306,6 +2306,72 @@ static unsigned long amdgpu_fix_asic_type(struct pci_dev *pdev, unsigned long fl
 	return flags;
 }
 
+static bool amdgpu_support_enabled(struct device *dev,
+				   const enum amd_asic_type family)
+{
+	const char *gen;
+	const char *param;
+	int module_param = -1;
+	bool radeon_support_built = IS_ENABLED(CONFIG_DRM_RADEON);
+	bool amdgpu_support_built = false;
+	bool support_by_default = false;
+
+	switch (family) {
+	case CHIP_TAHITI:
+	case CHIP_PITCAIRN:
+	case CHIP_VERDE:
+	case CHIP_OLAND:
+	case CHIP_HAINAN:
+		gen = "SI";
+		param = "si_support";
+		module_param = amdgpu_si_support;
+		amdgpu_support_built = IS_ENABLED(CONFIG_DRM_AMDGPU_SI);
+		support_by_default = true;
+		break;
+
+	case CHIP_BONAIRE:
+	case CHIP_HAWAII:
+		support_by_default = true;
+		fallthrough;
+	case CHIP_KAVERI:
+	case CHIP_KABINI:
+	case CHIP_MULLINS:
+		gen = "CIK";
+		param = "cik_support";
+		module_param = amdgpu_cik_support;
+		amdgpu_support_built = IS_ENABLED(CONFIG_DRM_AMDGPU_CIK);
+		break;
+
+	default:
+		/* All other chips are supported by amdgpu only */
+		return true;
+	}
+
+	if (!amdgpu_support_built) {
+		dev_info(dev, "amdgpu built without %s support\n", gen);
+		return false;
+	}
+
+	if ((module_param == -1 && (support_by_default || !radeon_support_built)) ||
+	    module_param == 1) {
+		if (radeon_support_built)
+			dev_info(dev, "%s support provided by amdgpu.\n"
+				 "Use radeon.%s=1 amdgpu.%s=0 to override.\n",
+				 gen, param, param);
+
+		return true;
+	}
+
+	if (radeon_support_built)
+		dev_info(dev, "%s support provided by radeon.\n"
+			 "Use radeon.%s=0 amdgpu.%s=1 to override.\n",
+			 gen, param, param);
+	else if (module_param == 0)
+		dev_info(dev, "%s support disabled by module param\n", gen);
+
+	return false;
+}
+
 static int amdgpu_pci_probe(struct pci_dev *pdev,
 			    const struct pci_device_id *ent)
 {
@@ -2353,48 +2419,8 @@ static int amdgpu_pci_probe(struct pci_dev *pdev,
 		return -ENOTSUPP;
 	}
 
-	switch (flags & AMD_ASIC_MASK) {
-	case CHIP_TAHITI:
-	case CHIP_PITCAIRN:
-	case CHIP_VERDE:
-	case CHIP_OLAND:
-	case CHIP_HAINAN:
-#ifdef CONFIG_DRM_AMDGPU_SI
-		if (!amdgpu_si_support) {
-			dev_info(&pdev->dev,
-				 "SI support provided by radeon.\n");
-			dev_info(&pdev->dev,
-				 "Use radeon.si_support=0 amdgpu.si_support=1 to override.\n"
-				);
-			return -ENODEV;
-		}
-		break;
-#else
-		dev_info(&pdev->dev, "amdgpu is built without SI support.\n");
-		return -ENODEV;
-#endif
-	case CHIP_KAVERI:
-	case CHIP_BONAIRE:
-	case CHIP_HAWAII:
-	case CHIP_KABINI:
-	case CHIP_MULLINS:
-#ifdef CONFIG_DRM_AMDGPU_CIK
-		if (!amdgpu_cik_support) {
-			dev_info(&pdev->dev,
-				 "CIK support provided by radeon.\n");
-			dev_info(&pdev->dev,
-				 "Use radeon.cik_support=0 amdgpu.cik_support=1 to override.\n"
-				);
-			return -ENODEV;
-		}
-		break;
-#else
-		dev_info(&pdev->dev, "amdgpu is built without CIK support.\n");
+	if (!amdgpu_support_enabled(&pdev->dev, flags & AMD_ASIC_MASK))
 		return -ENODEV;
-#endif
-	default:
-		break;
-	}
 
 	adev = devm_drm_dev_alloc(&pdev->dev, &amdgpu_kms_driver, typeof(*adev), ddev);
 	if (IS_ERR(adev))
@@ -2625,9 +2651,14 @@ static int amdgpu_pmops_suspend_noirq(struct device *dev)
 {
 	struct drm_device *drm_dev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = drm_to_adev(drm_dev);
+	int r;
 
-	if (amdgpu_acpi_should_gpu_reset(adev))
-		return amdgpu_asic_reset(adev);
+	if (amdgpu_acpi_should_gpu_reset(adev)) {
+		amdgpu_device_lock_reset_domain(adev->reset_domain);
+		r = amdgpu_asic_reset(adev);
+		amdgpu_device_unlock_reset_domain(adev->reset_domain);
+		return r;
+	}
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
index 83f3b94ed975..d2237ce9da70 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
@@ -368,6 +368,42 @@ void amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset,
 }
 
 /**
+ * amdgpu_gart_map_vram_range - map VRAM pages into the GART page table
+ *
+ * @adev: amdgpu_device pointer
+ * @pa: physical address of the first page to be mapped
+ * @start_page: first page to map in the GART aperture
+ * @num_pages: number of pages to be mapped
+ * @flags: page table entry flags
+ * @dst: CPU address of the GART table
+ *
+ * Binds a BO that is allocated in VRAM to the GART page table
+ * (all ASICs).
+ *
+ * Useful when a kernel BO is located in VRAM but
+ * needs to be accessed from the GART address space.
+ */
+void amdgpu_gart_map_vram_range(struct amdgpu_device *adev, uint64_t pa,
+				uint64_t start_page, uint64_t num_pages,
+				uint64_t flags, void *dst)
+{
+	u32 i, idx;
+
+	/* The SYSTEM flag indicates the pages aren't in VRAM. */
+	WARN_ON_ONCE(flags & AMDGPU_PTE_SYSTEM);
+
+	if (!drm_dev_enter(adev_to_drm(adev), &idx))
+		return;
+
+	for (i = 0; i < num_pages; ++i) {
+		amdgpu_gmc_set_pte_pde(adev, adev->gart.ptr,
+			start_page + i, pa + AMDGPU_GPU_PAGE_SIZE * i, flags);
+	}
+
+	drm_dev_exit(idx);
+}
+
+/**
  * amdgpu_gart_bind - bind pages into the gart page table
  *
  * @adev: amdgpu_device pointer
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
index 7cc980bf4725..d3118275ddae 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
@@ -64,5 +64,8 @@ void amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset,
 		     void *dst);
 void amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset,
 		      int pages, dma_addr_t *dma_addr, uint64_t flags);
+void amdgpu_gart_map_vram_range(struct amdgpu_device *adev, uint64_t pa,
+				uint64_t start_page, uint64_t num_pages,
+				uint64_t flags, void *dst);
 void amdgpu_gart_invalidate_tlb(struct amdgpu_device *adev);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index 9dcf51991b5b..869bceb0fe2c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -597,6 +597,9 @@ int amdgpu_gmc_allocate_vm_inv_eng(struct amdgpu_device *adev)
 		/* reserve engine 5 for firmware */
 		if (adev->enable_mes)
 			vm_inv_engs[i] &= ~(1 << 5);
+		/* reserve engine 6 for uni mes */
+		if (adev->enable_uni_mes)
+			vm_inv_engs[i] &= ~(1 << 6);
 		/* reserve mmhub engine 3 for firmware */
 		if (adev->enable_umsch_mm)
 			vm_inv_engs[i] &= ~(1 << 3);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
index 55097ca10738..727342689d4b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
@@ -86,6 +86,11 @@ enum amdgpu_memory_partition {
 
 #define AMDGPU_MAX_MEM_RANGES 8
 
+#define AMDGPU_GMC9_FAULT_SOURCE_DATA_RETRY 0x80
+#define AMDGPU_GMC9_FAULT_SOURCE_DATA_READ  0x40
+#define AMDGPU_GMC9_FAULT_SOURCE_DATA_WRITE 0x20
+#define AMDGPU_GMC9_FAULT_SOURCE_DATA_EXE   0x10
+
 /*
  * GMC page fault information
  */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
index 0760e70402ec..895c1e4c6747 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
@@ -284,6 +284,7 @@ int amdgpu_gtt_mgr_init(struct amdgpu_device *adev, uint64_t gtt_size)
 	ttm_resource_manager_init(man, &adev->mman.bdev, gtt_size);
 
 	start = AMDGPU_GTT_MAX_TRANSFER_SIZE * AMDGPU_GTT_NUM_TRANSFER_WINDOWS;
+	start += amdgpu_vce_required_gart_pages(adev);
 	size = (adev->gmc.gart_size >> PAGE_SHIFT) - start;
 	drm_mm_init(&mgr->mm, start, size);
 	spin_lock_init(&mgr->lock);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.c
index 6e02fb9ac2f6..5a60d69a3e1f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.c
@@ -66,3 +66,19 @@ void amdgpu_hdp_generic_flush(struct amdgpu_device *adev,
 				      0);
 	}
 }
+
+void amdgpu_hdp_invalidate(struct amdgpu_device *adev, struct amdgpu_ring *ring)
+{
+	if (adev->asic_funcs && adev->asic_funcs->invalidate_hdp)
+		adev->asic_funcs->invalidate_hdp(adev, ring);
+	else if (adev->hdp.funcs && adev->hdp.funcs->invalidate_hdp)
+		adev->hdp.funcs->invalidate_hdp(adev, ring);
+}
+
+void amdgpu_hdp_flush(struct amdgpu_device *adev, struct amdgpu_ring *ring)
+{
+	if (adev->asic_funcs && adev->asic_funcs->flush_hdp)
+		adev->asic_funcs->flush_hdp(adev, ring);
+	else if (adev->hdp.funcs && adev->hdp.funcs->flush_hdp)
+		adev->hdp.funcs->flush_hdp(adev, ring);
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h
index 4cfd932b7e91..d9f488fa76b9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h
@@ -46,4 +46,8 @@ struct amdgpu_hdp {
 int amdgpu_hdp_ras_sw_init(struct amdgpu_device *adev);
 void amdgpu_hdp_generic_flush(struct amdgpu_device *adev,
 			      struct amdgpu_ring *ring);
+void amdgpu_hdp_invalidate(struct amdgpu_device *adev,
+			   struct amdgpu_ring *ring);
+void amdgpu_hdp_flush(struct amdgpu_device *adev,
+		      struct amdgpu_ring *ring);
 #endif /* __AMDGPU_HDP_H__ */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c
index 518ca3f4db2b..90d26d820bac 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_hmm.c
@@ -287,9 +287,7 @@ void amdgpu_hmm_range_free(struct amdgpu_hmm_range *range)
 	if (!range)
 		return;
 
-	if (range->hmm_range.hmm_pfns)
-		kvfree(range->hmm_range.hmm_pfns);
-
+	kvfree(range->hmm_range.hmm_pfns);
 	amdgpu_bo_unref(&range->bo);
 	kfree(range);
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index 39229ece83f8..586a58facca1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -176,18 +176,21 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs,
 
 	if (!ring->sched.ready) {
 		dev_err(adev->dev, "couldn't schedule ib on ring <%s>\n", ring->name);
-		return -EINVAL;
+		r = -EINVAL;
+		goto free_fence;
 	}
 
 	if (vm && !job->vmid) {
 		dev_err(adev->dev, "VM IB without ID\n");
-		return -EINVAL;
+		r = -EINVAL;
+		goto free_fence;
 	}
 
 	if ((ib->flags & AMDGPU_IB_FLAGS_SECURE) &&
 	    (!ring->funcs->secure_submission_supported)) {
 		dev_err(adev->dev, "secure submissions not supported on ring <%s>\n", ring->name);
-		return -EINVAL;
+		r = -EINVAL;
+		goto free_fence;
 	}
 
 	alloc_size = ring->funcs->emit_frame_size + num_ibs *
@@ -196,7 +199,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs,
 	r = amdgpu_ring_alloc(ring, alloc_size);
 	if (r) {
 		dev_err(adev->dev, "scheduling IB failed (%d).\n", r);
-		return r;
+		goto free_fence;
 	}
 
 	need_ctx_switch = ring->current_ctx != fence_ctx;
@@ -302,6 +305,9 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs,
 		return r;
 	}
 	*f = &af->base;
+	/* get a ref for the job */
+	if (job)
+		dma_fence_get(*f);
 
 	if (ring->funcs->insert_end)
 		ring->funcs->insert_end(ring);
@@ -328,6 +334,11 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs,
 	amdgpu_ring_commit(ring);
 
 	return 0;
+
+free_fence:
+	if (!job)
+		kfree(af);
+	return r;
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
index 3ef5bc95642c..9cab36322c16 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
@@ -201,58 +201,34 @@ static int amdgpu_vmid_grab_idle(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 	unsigned vmhub = ring->vm_hub;
 	struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub];
-	struct dma_fence **fences;
-	unsigned i;
 
+	/* If anybody is waiting for a VMID let everybody wait for fairness */
 	if (!dma_fence_is_signaled(ring->vmid_wait)) {
 		*fence = dma_fence_get(ring->vmid_wait);
 		return 0;
 	}
 
-	fences = kmalloc_array(id_mgr->num_ids, sizeof(void *), GFP_NOWAIT);
-	if (!fences)
-		return -ENOMEM;
-
 	/* Check if we have an idle VMID */
-	i = 0;
-	list_for_each_entry((*idle), &id_mgr->ids_lru, list) {
+	list_for_each_entry_reverse((*idle), &id_mgr->ids_lru, list) {
 		/* Don't use per engine and per process VMID at the same time */
 		struct amdgpu_ring *r = adev->vm_manager.concurrent_flush ?
 			NULL : ring;
 
-		fences[i] = amdgpu_sync_peek_fence(&(*idle)->active, r);
-		if (!fences[i])
-			break;
-		++i;
+		*fence = amdgpu_sync_peek_fence(&(*idle)->active, r);
+		if (!(*fence))
+			return 0;
 	}
 
-	/* If we can't find a idle VMID to use, wait till one becomes available */
-	if (&(*idle)->list == &id_mgr->ids_lru) {
-		u64 fence_context = adev->vm_manager.fence_context + ring->idx;
-		unsigned seqno = ++adev->vm_manager.seqno[ring->idx];
-		struct dma_fence_array *array;
-		unsigned j;
-
-		*idle = NULL;
-		for (j = 0; j < i; ++j)
-			dma_fence_get(fences[j]);
-
-		array = dma_fence_array_create(i, fences, fence_context,
-					       seqno, true);
-		if (!array) {
-			for (j = 0; j < i; ++j)
-				dma_fence_put(fences[j]);
-			kfree(fences);
-			return -ENOMEM;
-		}
-
-		*fence = dma_fence_get(&array->base);
-		dma_fence_put(ring->vmid_wait);
-		ring->vmid_wait = &array->base;
-		return 0;
-	}
-	kfree(fences);
+	/*
+	 * If we can't find a idle VMID to use, wait on a fence from the least
+	 * recently used in the hope that it will be available soon.
+	 */
+	*idle = NULL;
+	dma_fence_put(ring->vmid_wait);
+	ring->vmid_wait = dma_fence_get(*fence);
 
+	/* This is the reference we return */
+	dma_fence_get(*fence);
 	return 0;
 }
 
@@ -313,7 +289,7 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm,
 	* user of the VMID.
 	*/
 	r = amdgpu_sync_fence(&(*id)->active, &job->base.s_fence->finished,
-			      GFP_NOWAIT);
+			      GFP_ATOMIC);
 	if (r)
 		return r;
 
@@ -373,7 +349,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm,
 		 */
 		r = amdgpu_sync_fence(&(*id)->active,
 				      &job->base.s_fence->finished,
-				      GFP_NOWAIT);
+				      GFP_ATOMIC);
 		if (r)
 			return r;
 
@@ -426,7 +402,7 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
 			/* Remember this submission as user of the VMID */
 			r = amdgpu_sync_fence(&id->active,
 					      &job->base.s_fence->finished,
-					      GFP_NOWAIT);
+					      GFP_ATOMIC);
 			if (r)
 				goto error;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c
index 9cddbf50442a..37270c4dab8d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_isp.c
@@ -280,6 +280,8 @@ int isp_kernel_buffer_alloc(struct device *dev, u64 size,
 	if (ret)
 		return ret;
 
+	/* Ensure *bo is NULL so a new BO will be created */
+	*bo = NULL;
 	ret = amdgpu_bo_create_kernel(adev,
 				      size,
 				      ISP_MC_ADDR_ALIGN,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
index e08d837668f1..0a0dcbf0798d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
@@ -130,11 +130,9 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 	}
 
 	/* attempt a per ring reset */
-	if (unlikely(adev->debug_disable_gpu_ring_reset)) {
-		dev_err(adev->dev, "Ring reset disabled by debug mask\n");
-	} else if (amdgpu_gpu_recovery &&
-		   amdgpu_ring_is_reset_type_supported(ring, AMDGPU_RESET_TYPE_PER_QUEUE) &&
-		   ring->funcs->reset) {
+	if (amdgpu_gpu_recovery &&
+	    amdgpu_ring_is_reset_type_supported(ring, AMDGPU_RESET_TYPE_PER_QUEUE) &&
+	    ring->funcs->reset) {
 		dev_err(adev->dev, "Starting %s ring reset\n",
 			s_job->sched->name);
 		r = amdgpu_ring_reset(ring, job->vmid, job->hw_fence);
@@ -226,6 +224,7 @@ err_fence:
 	kfree((*job)->hw_fence);
 err_job:
 	kfree(*job);
+	*job = NULL;
 
 	return r;
 }
@@ -247,7 +246,10 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev,
 	if (r) {
 		if (entity)
 			drm_sched_job_cleanup(&(*job)->base);
+		kfree((*job)->hw_vm_fence);
+		kfree((*job)->hw_fence);
 		kfree(*job);
+		*job = NULL;
 	}
 
 	return r;
@@ -295,6 +297,15 @@ static void amdgpu_job_free_cb(struct drm_sched_job *s_job)
 
 	amdgpu_sync_free(&job->explicit_sync);
 
+	if (job->hw_fence->base.ops)
+		dma_fence_put(&job->hw_fence->base);
+	else
+		kfree(job->hw_fence);
+	if (job->hw_vm_fence->base.ops)
+		dma_fence_put(&job->hw_vm_fence->base);
+	else
+		kfree(job->hw_vm_fence);
+
 	kfree(job);
 }
 
@@ -324,6 +335,15 @@ void amdgpu_job_free(struct amdgpu_job *job)
 	if (job->gang_submit != &job->base.s_fence->scheduled)
 		dma_fence_put(job->gang_submit);
 
+	if (job->hw_fence->base.ops)
+		dma_fence_put(&job->hw_fence->base);
+	else
+		kfree(job->hw_fence);
+	if (job->hw_vm_fence->base.ops)
+		dma_fence_put(&job->hw_vm_fence->base);
+	else
+		kfree(job->hw_vm_fence);
+
 	kfree(job);
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index 8c0e5d03de50..0b10497d487c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -1539,6 +1539,7 @@ static void psp_xgmi_reflect_topology_info(struct psp_context *psp,
 	uint64_t src_node_id = psp->adev->gmc.xgmi.node_id;
 	uint64_t dst_node_id = node_info.node_id;
 	uint8_t dst_num_hops = node_info.num_hops;
+	uint8_t dst_is_sharing_enabled = node_info.is_sharing_enabled;
 	uint8_t dst_num_links = node_info.num_links;
 
 	hive = amdgpu_get_xgmi_hive(psp->adev);
@@ -1558,13 +1559,20 @@ static void psp_xgmi_reflect_topology_info(struct psp_context *psp,
 				continue;
 
 			mirror_top_info->nodes[j].num_hops = dst_num_hops;
-			/*
-			 * prevent 0 num_links value re-reflection since reflection
+			mirror_top_info->nodes[j].is_sharing_enabled = dst_is_sharing_enabled;
+			/* prevent 0 num_links value re-reflection since reflection
 			 * criteria is based on num_hops (direct or indirect).
-			 *
 			 */
-			if (dst_num_links)
+			if (dst_num_links) {
 				mirror_top_info->nodes[j].num_links = dst_num_links;
+				/* swap src and dst due to frame of reference */
+				for (int k = 0; k < dst_num_links; k++) {
+					mirror_top_info->nodes[j].port_num[k].src_xgmi_port_num =
+						node_info.port_num[k].dst_xgmi_port_num;
+					mirror_top_info->nodes[j].port_num[k].dst_xgmi_port_num =
+						node_info.port_num[k].src_xgmi_port_num;
+				}
+			}
 
 			break;
 		}
@@ -1639,9 +1647,10 @@ int psp_xgmi_get_topology_info(struct psp_context *psp,
 			amdgpu_ip_version(psp->adev, MP0_HWIP, 0) ==
 				IP_VERSION(13, 0, 6) ||
 			amdgpu_ip_version(psp->adev, MP0_HWIP, 0) ==
-				IP_VERSION(13, 0, 14);
-		bool ta_port_num_support = amdgpu_sriov_vf(psp->adev) ? 0 :
-				psp->xgmi_context.xgmi_ta_caps & EXTEND_PEER_LINK_INFO_CMD_FLAG;
+				IP_VERSION(13, 0, 14) ||
+			amdgpu_sriov_vf(psp->adev);
+		bool ta_port_num_support = psp->xgmi_context.xgmi_ta_caps & EXTEND_PEER_LINK_INFO_CMD_FLAG ||
+			amdgpu_sriov_xgmi_ta_ext_peer_link_en(psp->adev);
 
 		/* popluate the shared output buffer rather than the cmd input buffer
 		 * with node_ids as the input for GET_PEER_LINKS command execution.
@@ -2355,8 +2364,11 @@ static int psp_securedisplay_initialize(struct psp_context *psp)
 	if (!ret && !psp->securedisplay_context.context.resp_status) {
 		psp->securedisplay_context.context.initialized = true;
 		mutex_init(&psp->securedisplay_context.mutex);
-	} else
+	} else {
+		/* don't try again */
+		psp->securedisplay_context.context.bin_desc.size_bytes = 0;
 		return ret;
+	}
 
 	mutex_lock(&psp->securedisplay_context.mutex);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index c8b4dd3ea5c3..2a6cf7963dde 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -150,6 +150,8 @@ static void amdgpu_ras_critical_region_fini(struct amdgpu_device *adev);
 
 #ifdef CONFIG_X86_MCE_AMD
 static void amdgpu_register_bad_pages_mca_notifier(struct amdgpu_device *adev);
+static void
+amdgpu_unregister_bad_pages_mca_notifier(struct amdgpu_device *adev);
 struct mce_notifier_adev_list {
 	struct amdgpu_device *devs[MAX_GPU_INSTANCE];
 	int num_gpu;
@@ -1661,7 +1663,7 @@ static int amdgpu_uniras_error_inject(struct amdgpu_device *adev,
 	inject_req.address = info->address;
 	inject_req.error_type = info->head.type;
 	inject_req.instance_mask = info->instance_mask;
-	inject_req.value = info->value;
+	inject_req.method = info->value;
 
 	return amdgpu_ras_mgr_handle_ras_cmd(adev, RAS_CMD__INJECT_ERROR,
 			&inject_req, sizeof(inject_req), &rsp, sizeof(rsp));
@@ -2921,8 +2923,12 @@ static void amdgpu_ras_do_recovery(struct work_struct *work)
 		type = amdgpu_ras_get_fatal_error_event(adev);
 		list_for_each_entry(remote_adev,
 				device_list_handle, gmc.xgmi.head) {
-			amdgpu_ras_query_err_status(remote_adev);
-			amdgpu_ras_log_on_err_counter(remote_adev, type);
+			if (amdgpu_uniras_enabled(remote_adev)) {
+				amdgpu_ras_mgr_update_ras_ecc(remote_adev);
+			} else {
+				amdgpu_ras_query_err_status(remote_adev);
+				amdgpu_ras_log_on_err_counter(remote_adev, type);
+			}
 		}
 
 	}
@@ -3010,8 +3016,13 @@ static int amdgpu_ras_mca2pa_by_idx(struct amdgpu_device *adev,
 	addr_in.ma.err_addr = bps->address;
 	addr_in.ma.socket_id = socket;
 	addr_in.ma.ch_inst = bps->mem_channel;
-	/* tell RAS TA the node instance is not used */
-	addr_in.ma.node_inst = TA_RAS_INV_NODE;
+	if (!amdgpu_ras_smu_eeprom_supported(adev)) {
+		/* tell RAS TA the node instance is not used */
+		addr_in.ma.node_inst = TA_RAS_INV_NODE;
+	} else {
+		addr_in.ma.umc_inst = bps->mcumc_id;
+		addr_in.ma.node_inst = bps->cu;
+	}
 
 	if (adev->umc.ras && adev->umc.ras->convert_ras_err_addr)
 		ret = adev->umc.ras->convert_ras_err_addr(adev, err_data,
@@ -3154,8 +3165,16 @@ static int __amdgpu_ras_convert_rec_from_rom(struct amdgpu_device *adev,
 	int i = 0;
 	enum amdgpu_memory_partition save_nps;
 
-	save_nps = (bps->retired_page >> UMC_NPS_SHIFT) & UMC_NPS_MASK;
-	bps->retired_page &= ~(UMC_NPS_MASK << UMC_NPS_SHIFT);
+	if (!amdgpu_ras_smu_eeprom_supported(adev)) {
+		save_nps = (bps->retired_page >> UMC_NPS_SHIFT) & UMC_NPS_MASK;
+		bps->retired_page &= ~(UMC_NPS_MASK << UMC_NPS_SHIFT);
+	} else {
+		/* if pmfw manages eeprom, save_nps is not stored on eeprom,
+		 * we should always convert mca address into physical address,
+		 * make save_nps different from nps
+		 */
+		save_nps = nps + 1;
+	}
 
 	if (save_nps == nps) {
 		if (amdgpu_umc_pages_in_a_row(adev, err_data,
@@ -3221,7 +3240,8 @@ int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
 
 	if (from_rom) {
 		/* there is no pa recs in V3, so skip pa recs processing */
-		if (control->tbl_hdr.version < RAS_TABLE_VER_V3) {
+		if ((control->tbl_hdr.version < RAS_TABLE_VER_V3) &&
+		    !amdgpu_ras_smu_eeprom_supported(adev)) {
 			for (i = 0; i < pages; i++) {
 				if (control->ras_num_recs - i >= adev->umc.retire_unit) {
 					if ((bps[i].address == bps[i + 1].address) &&
@@ -3291,7 +3311,13 @@ int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev,
 	mutex_lock(&con->recovery_lock);
 	control = &con->eeprom_control;
 	data = con->eh_data;
-	unit_num = data->count / adev->umc.retire_unit - control->ras_num_recs;
+	if (amdgpu_ras_smu_eeprom_supported(adev))
+		unit_num = control->ras_num_recs -
+			control->ras_num_recs_old;
+	else
+		unit_num = data->count / adev->umc.retire_unit -
+			control->ras_num_recs;
+
 	save_count = con->bad_page_num - control->ras_num_bad_pages;
 	mutex_unlock(&con->recovery_lock);
 
@@ -3352,7 +3378,8 @@ static int amdgpu_ras_load_bad_pages(struct amdgpu_device *adev)
 			/*In V3, there is no pa recs, and some cases(when address==0) may be parsed
 			as pa recs, so add verion check to avoid it.
 			*/
-			if (control->tbl_hdr.version < RAS_TABLE_VER_V3) {
+			if ((control->tbl_hdr.version < RAS_TABLE_VER_V3) &&
+			    !amdgpu_ras_smu_eeprom_supported(adev)) {
 				for (i = 0; i < control->ras_num_recs; i++) {
 					if ((control->ras_num_recs - i) >= adev->umc.retire_unit) {
 						if ((bps[i].address == bps[i + 1].address) &&
@@ -3767,6 +3794,8 @@ int amdgpu_ras_init_badpage_info(struct amdgpu_device *adev)
 		return 0;
 
 	control = &con->eeprom_control;
+	con->ras_smu_drv = amdgpu_dpm_get_ras_smu_driver(adev);
+
 	ret = amdgpu_ras_eeprom_init(control);
 	control->is_eeprom_valid = !ret;
 
@@ -3927,7 +3956,9 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev)
 	mutex_unlock(&con->recovery_lock);
 
 	amdgpu_ras_critical_region_init(adev);
-
+#ifdef CONFIG_X86_MCE_AMD
+	amdgpu_unregister_bad_pages_mca_notifier(adev);
+#endif
 	return 0;
 }
 /* recovery end */
@@ -4961,6 +4992,28 @@ static void amdgpu_register_bad_pages_mca_notifier(struct amdgpu_device *adev)
 		notifier_registered = true;
 	}
 }
+static void amdgpu_unregister_bad_pages_mca_notifier(struct amdgpu_device *adev)
+{
+	int i, j;
+
+	if (!notifier_registered && !mce_adev_list.num_gpu)
+		return;
+	for (i = 0, j = 0; i < mce_adev_list.num_gpu; i++) {
+		if (mce_adev_list.devs[i] == adev)
+			mce_adev_list.devs[i] = NULL;
+		if (!mce_adev_list.devs[i])
+			++j;
+	}
+
+	if (j == mce_adev_list.num_gpu) {
+		mce_adev_list.num_gpu = 0;
+		/* Unregister x86 notifier with MCE subsystem. */
+		if (notifier_registered) {
+			mce_unregister_decode_chain(&amdgpu_bad_page_nb);
+			notifier_registered = false;
+		}
+	}
+}
 #endif
 
 struct amdgpu_ras *amdgpu_ras_get_context(struct amdgpu_device *adev)
@@ -5671,3 +5724,25 @@ bool amdgpu_ras_check_critical_address(struct amdgpu_device *adev, uint64_t addr
 
 	return ret;
 }
+
+void amdgpu_ras_pre_reset(struct amdgpu_device *adev,
+					  struct list_head *device_list)
+{
+	struct amdgpu_device *tmp_adev = NULL;
+
+	list_for_each_entry(tmp_adev, device_list, reset_list) {
+		if (amdgpu_uniras_enabled(tmp_adev))
+			amdgpu_ras_mgr_pre_reset(tmp_adev);
+	}
+}
+
+void amdgpu_ras_post_reset(struct amdgpu_device *adev,
+					  struct list_head *device_list)
+{
+	struct amdgpu_device *tmp_adev = NULL;
+
+	list_for_each_entry(tmp_adev, device_list, reset_list) {
+		if (amdgpu_uniras_enabled(tmp_adev))
+			amdgpu_ras_mgr_post_reset(tmp_adev);
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index 556cf4d7b5ef..ff44190d7d98 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -503,6 +503,32 @@ struct ras_critical_region {
 	uint64_t size;
 };
 
+struct ras_eeprom_table_version {
+	uint32_t minor    : 16;
+	uint32_t major    : 16;
+};
+
+struct ras_eeprom_smu_funcs {
+	int (*get_ras_table_version)(struct amdgpu_device *adev,
+							uint32_t *table_version);
+	int (*get_badpage_count)(struct amdgpu_device *adev, uint32_t *count, uint32_t timeout);
+	int (*get_badpage_mca_addr)(struct amdgpu_device *adev, uint16_t index, uint64_t *mca_addr);
+	int (*set_timestamp)(struct amdgpu_device *adev, uint64_t timestamp);
+	int (*get_timestamp)(struct amdgpu_device *adev,
+							uint16_t index, uint64_t *timestamp);
+	int (*get_badpage_ipid)(struct amdgpu_device *adev, uint16_t index, uint64_t *ipid);
+	int (*erase_ras_table)(struct amdgpu_device *adev, uint32_t *result);
+};
+
+enum ras_smu_feature_flags {
+	RAS_SMU_FEATURE_BIT__RAS_EEPROM = BIT_ULL(0),
+};
+
+struct ras_smu_drv {
+	const struct ras_eeprom_smu_funcs *smu_eeprom_funcs;
+	void (*ras_smu_feature_flags)(struct amdgpu_device *adev, uint64_t *flags);
+};
+
 struct amdgpu_ras {
 	void *ras_mgr;
 	/* ras infrastructure */
@@ -591,6 +617,10 @@ struct amdgpu_ras {
 
 	/* Protect poison injection */
 	struct mutex poison_lock;
+
+	/* Disable/Enable uniras switch */
+	bool uniras_enabled;
+	const struct ras_smu_drv *ras_smu_drv;
 };
 
 struct ras_fs_data {
@@ -1009,4 +1039,9 @@ void amdgpu_ras_event_log_print(struct amdgpu_device *adev, u64 event_id,
 				const char *fmt, ...);
 
 bool amdgpu_ras_is_rma(struct amdgpu_device *adev);
+
+void amdgpu_ras_pre_reset(struct amdgpu_device *adev,
+					  struct list_head *device_list);
+void amdgpu_ras_post_reset(struct amdgpu_device *adev,
+					  struct list_head *device_list);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 5a7bf0661dbf..64dd7a81bff5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -124,6 +124,8 @@
 					RAS_TABLE_V2_1_INFO_SIZE) \
 					/ RAS_TABLE_RECORD_SIZE)
 
+#define RAS_SMU_MESSAGE_TIMEOUT_MS 1000 /* 1s */
+
 /* Given a zero-based index of an EEPROM RAS record, yields the EEPROM
  * offset off of RAS_TABLE_START.  That is, this is something you can
  * add to control->i2c_address, and then tell I2C layer to read
@@ -444,40 +446,57 @@ int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control)
 	struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr;
 	struct amdgpu_ras_eeprom_table_ras_info *rai = &control->tbl_rai;
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	u32 erase_res = 0;
 	u8 csum;
 	int res;
 
 	mutex_lock(&control->ras_tbl_mutex);
 
-	hdr->header = RAS_TABLE_HDR_VAL;
-	amdgpu_ras_set_eeprom_table_version(control);
+	if (!amdgpu_ras_smu_eeprom_supported(adev)) {
+		hdr->header = RAS_TABLE_HDR_VAL;
+		amdgpu_ras_set_eeprom_table_version(control);
 
-	if (hdr->version >= RAS_TABLE_VER_V2_1) {
-		hdr->first_rec_offset = RAS_RECORD_START_V2_1;
-		hdr->tbl_size = RAS_TABLE_HEADER_SIZE +
-				RAS_TABLE_V2_1_INFO_SIZE;
-		rai->rma_status = GPU_HEALTH_USABLE;
-		/**
-		 * GPU health represented as a percentage.
-		 * 0 means worst health, 100 means fully health.
-		 */
-		rai->health_percent = 100;
-		/* ecc_page_threshold = 0 means disable bad page retirement */
-		rai->ecc_page_threshold = con->bad_page_cnt_threshold;
+		if (hdr->version >= RAS_TABLE_VER_V2_1) {
+			hdr->first_rec_offset = RAS_RECORD_START_V2_1;
+			hdr->tbl_size = RAS_TABLE_HEADER_SIZE +
+					RAS_TABLE_V2_1_INFO_SIZE;
+			rai->rma_status = GPU_HEALTH_USABLE;
+
+			control->ras_record_offset = RAS_RECORD_START_V2_1;
+			control->ras_max_record_count = RAS_MAX_RECORD_COUNT_V2_1;
+			/**
+			 * GPU health represented as a percentage.
+			 * 0 means worst health, 100 means fully health.
+			 */
+			rai->health_percent = 100;
+			/* ecc_page_threshold = 0 means disable bad page retirement */
+			rai->ecc_page_threshold = con->bad_page_cnt_threshold;
+		} else {
+			hdr->first_rec_offset = RAS_RECORD_START;
+			hdr->tbl_size = RAS_TABLE_HEADER_SIZE;
+
+			control->ras_record_offset = RAS_RECORD_START;
+			control->ras_max_record_count = RAS_MAX_RECORD_COUNT;
+		}
+
+		csum = __calc_hdr_byte_sum(control);
+		if (hdr->version >= RAS_TABLE_VER_V2_1)
+			csum += __calc_ras_info_byte_sum(control);
+		csum = -csum;
+		hdr->checksum = csum;
+		res = __write_table_header(control);
+		if (!res && hdr->version > RAS_TABLE_VER_V1)
+			res = __write_table_ras_info(control);
 	} else {
-		hdr->first_rec_offset = RAS_RECORD_START;
-		hdr->tbl_size = RAS_TABLE_HEADER_SIZE;
+		res = amdgpu_ras_smu_erase_ras_table(adev, &erase_res);
+		if (res || erase_res) {
+			dev_warn(adev->dev, "RAS EEPROM reset failed, res:%d result:%d",
+										res, erase_res);
+			if (!res)
+				res = -EIO;
+		}
 	}
 
-	csum = __calc_hdr_byte_sum(control);
-	if (hdr->version >= RAS_TABLE_VER_V2_1)
-		csum += __calc_ras_info_byte_sum(control);
-	csum = -csum;
-	hdr->checksum = csum;
-	res = __write_table_header(control);
-	if (!res && hdr->version > RAS_TABLE_VER_V1)
-		res = __write_table_ras_info(control);
-
 	control->ras_num_recs = 0;
 	control->ras_num_bad_pages = 0;
 	control->ras_num_mca_recs = 0;
@@ -770,7 +789,8 @@ amdgpu_ras_eeprom_update_header(struct amdgpu_ras_eeprom_control *control)
 			"Saved bad pages %d reaches threshold value %d\n",
 			control->ras_num_bad_pages, ras->bad_page_cnt_threshold);
 
-		if (adev->cper.enabled && amdgpu_cper_generate_bp_threshold_record(adev))
+		if (adev->cper.enabled && !amdgpu_uniras_enabled(adev) &&
+		    amdgpu_cper_generate_bp_threshold_record(adev))
 			dev_warn(adev->dev, "fail to generate bad page threshold cper records\n");
 
 		if ((amdgpu_bad_page_threshold != -1) &&
@@ -853,6 +873,71 @@ Out:
 	return res;
 }
 
+int amdgpu_ras_eeprom_update_record_num(struct amdgpu_ras_eeprom_control *control)
+{
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	int ret, retry = 20;
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return 0;
+
+	control->ras_num_recs_old = control->ras_num_recs;
+
+	do {
+		/* 1000ms timeout is long enough, smu_get_badpage_count won't
+		 * return -EBUSY before timeout.
+		 */
+		ret = amdgpu_ras_smu_get_badpage_count(adev,
+			&(control->ras_num_recs), RAS_SMU_MESSAGE_TIMEOUT_MS);
+		if (!ret &&
+		    (control->ras_num_recs_old == control->ras_num_recs)) {
+			/* record number update in PMFW needs some time,
+			 * smu_get_badpage_count may return immediately without
+			 * count update, sleep for a while and retry again.
+			 */
+			msleep(50);
+			retry--;
+		} else {
+			break;
+		}
+	} while (retry);
+
+	/* no update of record number is not a real failure,
+	 * don't print warning here
+	 */
+	if (!ret && (control->ras_num_recs_old == control->ras_num_recs))
+		ret = -EINVAL;
+
+	return ret;
+}
+
+static int amdgpu_ras_smu_eeprom_append(struct amdgpu_ras_eeprom_control *control)
+{
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev) || !con)
+		return 0;
+
+	control->ras_num_bad_pages = con->bad_page_num;
+
+	if (amdgpu_bad_page_threshold != 0 &&
+	    control->ras_num_bad_pages > con->bad_page_cnt_threshold) {
+		dev_warn(adev->dev,
+			"Saved bad pages %d reaches threshold value %d\n",
+			control->ras_num_bad_pages, con->bad_page_cnt_threshold);
+
+		if (adev->cper.enabled && amdgpu_cper_generate_bp_threshold_record(adev))
+			dev_warn(adev->dev, "fail to generate bad page threshold cper records\n");
+
+		if ((amdgpu_bad_page_threshold != -1) &&
+		    (amdgpu_bad_page_threshold != -2))
+			con->is_rma = true;
+	}
+
+	return 0;
+}
+
 /**
  * amdgpu_ras_eeprom_append -- append records to the EEPROM RAS table
  * @control: pointer to control structure
@@ -877,6 +962,9 @@ int amdgpu_ras_eeprom_append(struct amdgpu_ras_eeprom_control *control,
 	if (!__is_ras_eeprom_supported(adev))
 		return 0;
 
+	if (amdgpu_ras_smu_eeprom_supported(adev))
+		return amdgpu_ras_smu_eeprom_append(control);
+
 	if (num == 0) {
 		dev_err(adev->dev, "will not append 0 records\n");
 		return -EINVAL;
@@ -952,6 +1040,50 @@ static int __amdgpu_ras_eeprom_read(struct amdgpu_ras_eeprom_control *control,
 	return res;
 }
 
+int amdgpu_ras_eeprom_read_idx(struct amdgpu_ras_eeprom_control *control,
+			struct eeprom_table_record *record, u32 rec_idx,
+			const u32 num)
+{
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	uint64_t ts, end_idx;
+	int i, ret;
+	u64 mca, ipid;
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return 0;
+
+	if (!adev->umc.ras || !adev->umc.ras->mca_ipid_parse)
+		return -EOPNOTSUPP;
+
+	end_idx = rec_idx + num;
+	for (i = rec_idx; i < end_idx; i++) {
+		ret = amdgpu_ras_smu_get_badpage_mca_addr(adev, i, &mca);
+		if (ret)
+			return ret;
+
+		ret = amdgpu_ras_smu_get_badpage_ipid(adev, i, &ipid);
+		if (ret)
+			return ret;
+
+		ret = amdgpu_ras_smu_get_timestamp(adev, i, &ts);
+		if (ret)
+			return ret;
+
+		record[i - rec_idx].address = mca;
+		/* retired_page (pa) is unused now */
+		record[i - rec_idx].retired_page = 0x1ULL;
+		record[i - rec_idx].ts = ts;
+		record[i - rec_idx].err_type = AMDGPU_RAS_EEPROM_ERR_NON_RECOVERABLE;
+
+		adev->umc.ras->mca_ipid_parse(adev, ipid,
+			(uint32_t *)&(record[i - rec_idx].cu),
+			(uint32_t *)&(record[i - rec_idx].mem_channel),
+			(uint32_t *)&(record[i - rec_idx].mcumc_id), NULL);
+	}
+
+	return 0;
+}
+
 /**
  * amdgpu_ras_eeprom_read -- read EEPROM
  * @control: pointer to control structure
@@ -973,6 +1105,9 @@ int amdgpu_ras_eeprom_read(struct amdgpu_ras_eeprom_control *control,
 	u8 *buf, *pp;
 	u32 g0, g1;
 
+	if (amdgpu_ras_smu_eeprom_supported(adev))
+		return amdgpu_ras_eeprom_read_idx(control, record, 0, num);
+
 	if (!__is_ras_eeprom_supported(adev))
 		return 0;
 
@@ -1144,6 +1279,10 @@ static ssize_t amdgpu_ras_debugfs_table_read(struct file *f, char __user *buf,
 	int res = -EFAULT;
 	size_t data_len;
 
+	/* pmfw manages eeprom data by itself */
+	if (amdgpu_ras_smu_eeprom_supported(adev))
+		return 0;
+
 	mutex_lock(&control->ras_tbl_mutex);
 
 	/* We want *pos - data_len > 0, which means there's
@@ -1374,6 +1513,42 @@ Out:
 	return res == RAS_TABLE_V2_1_INFO_SIZE ? 0 : res;
 }
 
+static int amdgpu_ras_smu_eeprom_init(struct amdgpu_ras_eeprom_control *control)
+{
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr;
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+	uint64_t local_time;
+	int res;
+
+	ras->is_rma = false;
+
+	if (!__is_ras_eeprom_supported(adev))
+		return 0;
+	mutex_init(&control->ras_tbl_mutex);
+
+	res = amdgpu_ras_smu_get_table_version(adev, &(hdr->version));
+	if (res)
+		return res;
+
+	res = amdgpu_ras_smu_get_badpage_count(adev,
+								&(control->ras_num_recs), 100);
+	if (res)
+		return res;
+
+	local_time = (uint64_t)ktime_get_real_seconds();
+	res = amdgpu_ras_smu_set_timestamp(adev, local_time);
+	if (res)
+		return res;
+
+	control->ras_max_record_count = 4000;
+
+	control->ras_num_mca_recs = 0;
+	control->ras_num_pa_recs = 0;
+
+	return 0;
+}
+
 int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 {
 	struct amdgpu_device *adev = to_amdgpu_device(control);
@@ -1382,6 +1557,9 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
 	int res;
 
+	if (amdgpu_ras_smu_eeprom_supported(adev))
+		return amdgpu_ras_smu_eeprom_init(control);
+
 	ras->is_rma = false;
 
 	if (!__is_ras_eeprom_supported(adev))
@@ -1448,6 +1626,47 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 	return 0;
 }
 
+static int amdgpu_ras_smu_eeprom_check(struct amdgpu_ras_eeprom_control *control)
+{
+	struct amdgpu_device *adev = to_amdgpu_device(control);
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	if (!__is_ras_eeprom_supported(adev))
+		return 0;
+
+	control->ras_num_bad_pages = ras->bad_page_num;
+
+	if ((ras->bad_page_cnt_threshold < control->ras_num_bad_pages) &&
+	    amdgpu_bad_page_threshold != 0) {
+		dev_warn(adev->dev,
+			"RAS records:%d exceed threshold:%d\n",
+			control->ras_num_bad_pages, ras->bad_page_cnt_threshold);
+		if ((amdgpu_bad_page_threshold == -1) ||
+			(amdgpu_bad_page_threshold == -2)) {
+			dev_warn(adev->dev,
+				 "Please consult AMD Service Action Guide (SAG) for appropriate service procedures\n");
+		} else {
+			ras->is_rma = true;
+			dev_warn(adev->dev,
+				 "User defined threshold is set, runtime service will be halt when threshold is reached\n");
+		}
+
+		return 0;
+	}
+
+	dev_dbg(adev->dev,
+		"Found existing EEPROM table with %d records",
+		control->ras_num_bad_pages);
+
+	/* Warn if we are at 90% of the threshold or above
+	 */
+	if (10 * control->ras_num_bad_pages >= 9 * ras->bad_page_cnt_threshold)
+		dev_warn(adev->dev, "RAS records:%u exceeds 90%% of threshold:%d",
+				control->ras_num_bad_pages,
+				ras->bad_page_cnt_threshold);
+	return 0;
+}
+
 int amdgpu_ras_eeprom_check(struct amdgpu_ras_eeprom_control *control)
 {
 	struct amdgpu_device *adev = to_amdgpu_device(control);
@@ -1455,6 +1674,9 @@ int amdgpu_ras_eeprom_check(struct amdgpu_ras_eeprom_control *control)
 	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
 	int res = 0;
 
+	if (amdgpu_ras_smu_eeprom_supported(adev))
+		return amdgpu_ras_smu_eeprom_check(control);
+
 	if (!__is_ras_eeprom_supported(adev))
 		return 0;
 
@@ -1545,7 +1767,8 @@ void amdgpu_ras_eeprom_check_and_recover(struct amdgpu_device *adev)
 	struct amdgpu_ras_eeprom_control *control;
 	int res;
 
-	if (!__is_ras_eeprom_supported(adev) || !ras)
+	if (!__is_ras_eeprom_supported(adev) || !ras ||
+	    amdgpu_ras_smu_eeprom_supported(adev))
 		return;
 	control = &ras->eeprom_control;
 	if (!control->is_eeprom_valid)
@@ -1565,4 +1788,143 @@ void amdgpu_ras_eeprom_check_and_recover(struct amdgpu_device *adev)
 		control->is_eeprom_valid = false;
 	}
 	return;
-}
\ No newline at end of file
+}
+
+static const struct ras_smu_drv *amdgpu_ras_get_smu_ras_drv(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	if (!ras)
+		return NULL;
+
+	return ras->ras_smu_drv;
+}
+
+static uint64_t amdgpu_ras_smu_get_feature_flags(struct amdgpu_device *adev)
+{
+	const struct ras_smu_drv *ras_smu_drv = amdgpu_ras_get_smu_ras_drv(adev);
+	uint64_t flags = 0ULL;
+
+	if (!ras_smu_drv)
+		goto out;
+
+	if (ras_smu_drv->ras_smu_feature_flags)
+		ras_smu_drv->ras_smu_feature_flags(adev, &flags);
+
+out:
+	return flags;
+}
+
+bool amdgpu_ras_smu_eeprom_supported(struct amdgpu_device *adev)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+	uint64_t flags = 0ULL;
+
+	if (!__is_ras_eeprom_supported(adev) || !smu_ras_drv)
+		return false;
+
+	if (!smu_ras_drv->smu_eeprom_funcs)
+		return false;
+
+	flags = amdgpu_ras_smu_get_feature_flags(adev);
+
+	return !!(flags & RAS_SMU_FEATURE_BIT__RAS_EEPROM);
+}
+
+int amdgpu_ras_smu_get_table_version(struct amdgpu_device *adev,
+				     uint32_t *table_version)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->get_ras_table_version)
+		return smu_ras_drv->smu_eeprom_funcs->get_ras_table_version(adev,
+										 table_version);
+	return -EOPNOTSUPP;
+}
+
+int amdgpu_ras_smu_get_badpage_count(struct amdgpu_device *adev,
+				     uint32_t *count, uint32_t timeout)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->get_badpage_count)
+		return smu_ras_drv->smu_eeprom_funcs->get_badpage_count(adev,
+									     count, timeout);
+	return -EOPNOTSUPP;
+}
+
+int amdgpu_ras_smu_get_badpage_mca_addr(struct amdgpu_device *adev,
+					uint16_t index, uint64_t *mca_addr)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->get_badpage_mca_addr)
+		return smu_ras_drv->smu_eeprom_funcs->get_badpage_mca_addr(adev,
+										index, mca_addr);
+	return -EOPNOTSUPP;
+}
+
+int amdgpu_ras_smu_set_timestamp(struct amdgpu_device *adev,
+				 uint64_t timestamp)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->set_timestamp)
+		return smu_ras_drv->smu_eeprom_funcs->set_timestamp(adev,
+									 timestamp);
+	return -EOPNOTSUPP;
+}
+
+int amdgpu_ras_smu_get_timestamp(struct amdgpu_device *adev,
+				 uint16_t index, uint64_t *timestamp)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->get_timestamp)
+		return smu_ras_drv->smu_eeprom_funcs->get_timestamp(adev,
+									 index, timestamp);
+	return -EOPNOTSUPP;
+}
+
+int amdgpu_ras_smu_get_badpage_ipid(struct amdgpu_device *adev,
+				    uint16_t index, uint64_t *ipid)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->get_badpage_ipid)
+		return smu_ras_drv->smu_eeprom_funcs->get_badpage_ipid(adev,
+									    index, ipid);
+	return -EOPNOTSUPP;
+}
+
+int amdgpu_ras_smu_erase_ras_table(struct amdgpu_device *adev,
+				   uint32_t *result)
+{
+	const struct ras_smu_drv *smu_ras_drv = amdgpu_ras_get_smu_ras_drv(adev);
+
+	if (!amdgpu_ras_smu_eeprom_supported(adev))
+		return -EOPNOTSUPP;
+
+	if (smu_ras_drv->smu_eeprom_funcs->erase_ras_table)
+		return smu_ras_drv->smu_eeprom_funcs->erase_ras_table(adev,
+									   result);
+	return -EOPNOTSUPP;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
index ebfca4cb5688..2e5d63957e71 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
@@ -82,6 +82,7 @@ struct amdgpu_ras_eeprom_control {
 	/* Number of records in the table.
 	 */
 	u32 ras_num_recs;
+	u32 ras_num_recs_old;
 
 	/* the bad page number is ras_num_recs or
 	 * ras_num_recs * umc.retire_unit
@@ -163,6 +164,35 @@ int amdgpu_ras_eeprom_check(struct amdgpu_ras_eeprom_control *control);
 
 void amdgpu_ras_eeprom_check_and_recover(struct amdgpu_device *adev);
 
+bool amdgpu_ras_smu_eeprom_supported(struct amdgpu_device *adev);
+
+int amdgpu_ras_smu_get_table_version(struct amdgpu_device *adev,
+							uint32_t *table_version);
+
+int amdgpu_ras_smu_get_badpage_count(struct amdgpu_device *adev,
+								uint32_t *count, uint32_t timeout);
+
+int amdgpu_ras_smu_get_badpage_mca_addr(struct amdgpu_device *adev,
+								uint16_t index, uint64_t *mca_addr);
+
+int amdgpu_ras_smu_set_timestamp(struct amdgpu_device *adev,
+										uint64_t timestamp);
+
+int amdgpu_ras_smu_get_timestamp(struct amdgpu_device *adev,
+							uint16_t index, uint64_t *timestamp);
+
+int amdgpu_ras_smu_get_badpage_ipid(struct amdgpu_device *adev,
+								uint16_t index, uint64_t *ipid);
+
+int amdgpu_ras_smu_erase_ras_table(struct amdgpu_device *adev,
+									uint32_t *result);
+
+int amdgpu_ras_eeprom_read_idx(struct amdgpu_ras_eeprom_control *control,
+			struct eeprom_table_record *record, u32 rec_idx,
+			const u32 num);
+
+int amdgpu_ras_eeprom_update_record_num(struct amdgpu_ras_eeprom_control *control);
+
 extern const struct file_operations amdgpu_ras_debugfs_eeprom_size_ops;
 extern const struct file_operations amdgpu_ras_debugfs_eeprom_table_ops;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index 43f769fed810..c596b6df2e2d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -33,6 +33,7 @@
 
 #include <drm/amdgpu_drm.h>
 #include "amdgpu.h"
+#include "amdgpu_ras_mgr.h"
 #include "atom.h"
 
 /*
@@ -468,9 +469,6 @@ bool amdgpu_ring_soft_recovery(struct amdgpu_ring *ring, unsigned int vmid,
 	ktime_t deadline;
 	bool ret;
 
-	if (unlikely(ring->adev->debug_disable_soft_recovery))
-		return false;
-
 	deadline = ktime_add_us(ktime_get(), 10000);
 
 	if (amdgpu_sriov_vf(ring->adev) || !ring->funcs->soft_recovery || !fence)
@@ -498,6 +496,66 @@ bool amdgpu_ring_soft_recovery(struct amdgpu_ring *ring, unsigned int vmid,
  */
 #if defined(CONFIG_DEBUG_FS)
 
+static ssize_t amdgpu_ras_cper_debugfs_read(struct file *f, char __user *buf,
+					    size_t size, loff_t *offset)
+{
+	const uint8_t ring_header_size = 12;
+	struct amdgpu_ring *ring = file_inode(f)->i_private;
+	struct ras_cmd_cper_snapshot_req *snapshot_req __free(kfree) =
+		kzalloc(sizeof(struct ras_cmd_cper_snapshot_req), GFP_KERNEL);
+	struct ras_cmd_cper_snapshot_rsp *snapshot_rsp __free(kfree) =
+		kzalloc(sizeof(struct ras_cmd_cper_snapshot_rsp), GFP_KERNEL);
+	struct ras_cmd_cper_record_req *record_req __free(kfree) =
+		kzalloc(sizeof(struct ras_cmd_cper_record_req), GFP_KERNEL);
+	struct ras_cmd_cper_record_rsp *record_rsp __free(kfree) =
+		kzalloc(sizeof(struct ras_cmd_cper_record_rsp), GFP_KERNEL);
+	uint8_t *ring_header __free(kfree) =
+		kzalloc(ring_header_size, GFP_KERNEL);
+	uint32_t total_cper_num;
+	uint64_t start_cper_id;
+	int r;
+
+	if (!snapshot_req || !snapshot_rsp || !record_req || !record_rsp ||
+	    !ring_header)
+		return -ENOMEM;
+
+	if (!(*offset)) {
+		/* Need at least 12 bytes for the header on the first read */
+		if (size < ring_header_size)
+			return -EINVAL;
+
+		if (copy_to_user(buf, ring_header, ring_header_size))
+			return -EFAULT;
+		buf += ring_header_size;
+		size -= ring_header_size;
+	}
+
+	r = amdgpu_ras_mgr_handle_ras_cmd(ring->adev,
+					  RAS_CMD__GET_CPER_SNAPSHOT,
+					  snapshot_req, sizeof(struct ras_cmd_cper_snapshot_req),
+					  snapshot_rsp, sizeof(struct ras_cmd_cper_snapshot_rsp));
+	if (r || !snapshot_rsp->total_cper_num)
+		return r;
+
+	start_cper_id = snapshot_rsp->start_cper_id;
+	total_cper_num = snapshot_rsp->total_cper_num;
+
+	record_req->buf_ptr = (uint64_t)(uintptr_t)buf;
+	record_req->buf_size = size;
+	record_req->cper_start_id = start_cper_id + *offset;
+	record_req->cper_num = total_cper_num;
+	r = amdgpu_ras_mgr_handle_ras_cmd(ring->adev, RAS_CMD__GET_CPER_RECORD,
+					  record_req, sizeof(struct ras_cmd_cper_record_req),
+					  record_rsp, sizeof(struct ras_cmd_cper_record_rsp));
+	if (r)
+		return r;
+
+	r = *offset ? record_rsp->real_data_size : record_rsp->real_data_size + ring_header_size;
+	(*offset) += record_rsp->real_cper_num;
+
+	return r;
+}
+
 /* Layout of file is 12 bytes consisting of
  * - rptr
  * - wptr
@@ -514,6 +572,9 @@ static ssize_t amdgpu_debugfs_ring_read(struct file *f, char __user *buf,
 	loff_t i;
 	int r;
 
+	if (ring->funcs->type == AMDGPU_RING_TYPE_CPER && amdgpu_uniras_enabled(ring->adev))
+		return amdgpu_ras_cper_debugfs_read(f, buf, size, pos);
+
 	if (*pos & 3 || size & 3)
 		return -EINVAL;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index 87b962df5460..7a27c6c4bb44 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -83,6 +83,7 @@ enum amdgpu_ring_type {
 	AMDGPU_RING_TYPE_MES,
 	AMDGPU_RING_TYPE_UMSCH_MM,
 	AMDGPU_RING_TYPE_CPER,
+	AMDGPU_RING_TYPE_MAX,
 };
 
 enum amdgpu_ib_pool_type {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 77f1f0d1593b..2b931e855abd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -188,7 +188,6 @@ static int amdgpu_ttm_map_buffer(struct ttm_buffer_object *bo,
 	struct amdgpu_job *job;
 	void *cpu_addr;
 	uint64_t flags;
-	unsigned int i;
 	int r;
 
 	BUG_ON(adev->mman.buffer_funcs->copy_max_bytes <
@@ -255,16 +254,9 @@ static int amdgpu_ttm_map_buffer(struct ttm_buffer_object *bo,
 		dma_addr = &bo->ttm->dma_address[mm_cur->start >> PAGE_SHIFT];
 		amdgpu_gart_map(adev, 0, num_pages, dma_addr, flags, cpu_addr);
 	} else {
-		dma_addr_t dma_address;
-
-		dma_address = mm_cur->start;
-		dma_address += adev->vm_manager.vram_base_offset;
+		u64 pa = mm_cur->start + adev->vm_manager.vram_base_offset;
 
-		for (i = 0; i < num_pages; ++i) {
-			amdgpu_gart_map(adev, i << PAGE_SHIFT, 1, &dma_address,
-					flags, cpu_addr);
-			dma_address += PAGE_SIZE;
-		}
+		amdgpu_gart_map_vram_range(adev, pa, 0, num_pages, flags, cpu_addr);
 	}
 
 	dma_fence_put(amdgpu_job_submit(job));
@@ -1337,7 +1329,7 @@ uint64_t amdgpu_ttm_tt_pde_flags(struct ttm_tt *ttm, struct ttm_resource *mem)
 		    mem->mem_type == AMDGPU_PL_MMIO_REMAP)) {
 		flags |= AMDGPU_PTE_SYSTEM;
 
-		if (ttm->caching == ttm_cached)
+		if (ttm && ttm->caching == ttm_cached)
 			flags |= AMDGPU_PTE_SNOOPED;
 	}
 
@@ -1494,6 +1486,7 @@ static int amdgpu_ttm_access_memory_sdma(struct ttm_buffer_object *bo,
 	if (r)
 		goto out;
 
+	mutex_lock(&adev->mman.gtt_window_lock);
 	amdgpu_res_first(abo->tbo.resource, offset, len, &src_mm);
 	src_addr = amdgpu_ttm_domain_start(adev, bo->resource->mem_type) +
 		src_mm.start;
@@ -1508,6 +1501,7 @@ static int amdgpu_ttm_access_memory_sdma(struct ttm_buffer_object *bo,
 	WARN_ON(job->ibs[0].length_dw > num_dw);
 
 	fence = amdgpu_job_submit(job);
+	mutex_unlock(&adev->mman.gtt_window_lock);
 
 	if (!dma_fence_wait_timeout(fence, false, adev->sdma_timeout))
 		r = -ETIMEDOUT;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
index 2e039fb778ea..3f0b0e9af4f3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
@@ -24,6 +24,7 @@
 #include <linux/sort.h>
 #include "amdgpu.h"
 #include "umc_v6_7.h"
+#include "amdgpu_ras_mgr.h"
 #define MAX_UMC_POISON_POLLING_TIME_SYNC   20  //ms
 
 #define MAX_UMC_HASH_STRING_SIZE  256
@@ -96,67 +97,96 @@ void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev,
 {
 	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	struct amdgpu_ras_eeprom_control *control = &con->eeprom_control;
 	unsigned int error_query_mode;
 	int ret = 0;
 	unsigned long err_count;
 
 	amdgpu_ras_get_error_query_mode(adev, &error_query_mode);
 
+	err_data->err_addr =
+		kcalloc(adev->umc.max_ras_err_cnt_per_query,
+			sizeof(struct eeprom_table_record), GFP_KERNEL);
+
+	/* still call query_ras_error_address to clear error status
+	 * even NOMEM error is encountered
+	 */
+	if (!err_data->err_addr)
+		dev_warn(adev->dev,
+			"Failed to alloc memory for umc error address record!\n");
+	else
+		err_data->err_addr_len = adev->umc.max_ras_err_cnt_per_query;
+
 	mutex_lock(&con->page_retirement_lock);
-	ret = amdgpu_dpm_get_ecc_info(adev, (void *)&(con->umc_ecc));
-	if (ret == -EOPNOTSUPP &&
-	    error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY) {
-		if (adev->umc.ras && adev->umc.ras->ras_block.hw_ops &&
-		    adev->umc.ras->ras_block.hw_ops->query_ras_error_count)
-		    adev->umc.ras->ras_block.hw_ops->query_ras_error_count(adev, ras_error_status);
-
-		if (adev->umc.ras && adev->umc.ras->ras_block.hw_ops &&
-		    adev->umc.ras->ras_block.hw_ops->query_ras_error_address &&
-		    adev->umc.max_ras_err_cnt_per_query) {
-			err_data->err_addr =
-				kcalloc(adev->umc.max_ras_err_cnt_per_query,
-					sizeof(struct eeprom_table_record), GFP_KERNEL);
-
-			/* still call query_ras_error_address to clear error status
-			 * even NOMEM error is encountered
-			 */
-			if(!err_data->err_addr)
-				dev_warn(adev->dev, "Failed to alloc memory for "
-						"umc error address record!\n");
-			else
-				err_data->err_addr_len = adev->umc.max_ras_err_cnt_per_query;
-
-			/* umc query_ras_error_address is also responsible for clearing
-			 * error status
-			 */
-			adev->umc.ras->ras_block.hw_ops->query_ras_error_address(adev, ras_error_status);
+	if (!amdgpu_ras_smu_eeprom_supported(adev)) {
+		ret = amdgpu_dpm_get_ecc_info(adev, (void *)&(con->umc_ecc));
+		if (ret == -EOPNOTSUPP &&
+		    error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY) {
+			if (adev->umc.ras && adev->umc.ras->ras_block.hw_ops &&
+			    adev->umc.ras->ras_block.hw_ops->query_ras_error_count)
+				adev->umc.ras->ras_block.hw_ops->query_ras_error_count(adev,
+								ras_error_status);
+
+			if (adev->umc.ras && adev->umc.ras->ras_block.hw_ops &&
+			    adev->umc.ras->ras_block.hw_ops->query_ras_error_address &&
+			    adev->umc.max_ras_err_cnt_per_query) {
+				err_data->err_addr =
+					kcalloc(adev->umc.max_ras_err_cnt_per_query,
+						sizeof(struct eeprom_table_record), GFP_KERNEL);
+
+				/* still call query_ras_error_address to clear error status
+				 * even NOMEM error is encountered
+				 */
+				if (!err_data->err_addr)
+					dev_warn(adev->dev,
+						"Failed to alloc memory for umc error address record!\n");
+				else
+					err_data->err_addr_len =
+						adev->umc.max_ras_err_cnt_per_query;
+
+				/* umc query_ras_error_address is also responsible for clearing
+				 * error status
+				 */
+				adev->umc.ras->ras_block.hw_ops->query_ras_error_address(adev,
+								ras_error_status);
+			}
+		} else if (error_query_mode == AMDGPU_RAS_FIRMWARE_ERROR_QUERY ||
+		    (!ret && error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY)) {
+			if (adev->umc.ras &&
+			    adev->umc.ras->ecc_info_query_ras_error_count)
+				adev->umc.ras->ecc_info_query_ras_error_count(adev,
+								ras_error_status);
+
+			if (adev->umc.ras &&
+			    adev->umc.ras->ecc_info_query_ras_error_address &&
+			    adev->umc.max_ras_err_cnt_per_query) {
+				err_data->err_addr =
+					kcalloc(adev->umc.max_ras_err_cnt_per_query,
+						sizeof(struct eeprom_table_record), GFP_KERNEL);
+
+				/* still call query_ras_error_address to clear error status
+				 * even NOMEM error is encountered
+				 */
+				if (!err_data->err_addr)
+					dev_warn(adev->dev,
+						"Failed to alloc memory for umc error address record!\n");
+				else
+					err_data->err_addr_len =
+						adev->umc.max_ras_err_cnt_per_query;
+
+				/* umc query_ras_error_address is also responsible for clearing
+				 * error status
+				 */
+				adev->umc.ras->ecc_info_query_ras_error_address(adev,
+								ras_error_status);
+			}
 		}
-	} else if (error_query_mode == AMDGPU_RAS_FIRMWARE_ERROR_QUERY ||
-	    (!ret && error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY)) {
-		if (adev->umc.ras &&
-		    adev->umc.ras->ecc_info_query_ras_error_count)
-		    adev->umc.ras->ecc_info_query_ras_error_count(adev, ras_error_status);
-
-		if (adev->umc.ras &&
-		    adev->umc.ras->ecc_info_query_ras_error_address &&
-		    adev->umc.max_ras_err_cnt_per_query) {
-			err_data->err_addr =
-				kcalloc(adev->umc.max_ras_err_cnt_per_query,
-					sizeof(struct eeprom_table_record), GFP_KERNEL);
-
-			/* still call query_ras_error_address to clear error status
-			 * even NOMEM error is encountered
-			 */
-			if(!err_data->err_addr)
-				dev_warn(adev->dev, "Failed to alloc memory for "
-						"umc error address record!\n");
-			else
-				err_data->err_addr_len = adev->umc.max_ras_err_cnt_per_query;
-
-			/* umc query_ras_error_address is also responsible for clearing
-			 * error status
-			 */
-			adev->umc.ras->ecc_info_query_ras_error_address(adev, ras_error_status);
+	} else {
+		if (!amdgpu_ras_eeprom_update_record_num(control)) {
+			err_data->err_addr_cnt = err_data->de_count =
+				control->ras_num_recs -	control->ras_num_recs_old;
+			amdgpu_ras_eeprom_read_idx(control, err_data->err_addr,
+				control->ras_num_recs_old, err_data->de_count);
 		}
 	}
 
@@ -166,7 +196,7 @@ void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev,
 		if ((amdgpu_bad_page_threshold != 0) &&
 			err_data->err_addr_cnt) {
 			amdgpu_ras_add_bad_pages(adev, err_data->err_addr,
-						err_data->err_addr_cnt, false);
+				err_data->err_addr_cnt, amdgpu_ras_smu_eeprom_supported(adev));
 			amdgpu_ras_save_bad_pages(adev, &err_count);
 
 			amdgpu_dpm_send_hbm_bad_pages_num(adev,
@@ -244,6 +274,15 @@ int amdgpu_umc_pasid_poison_handler(struct amdgpu_device *adev,
 			}
 
 			amdgpu_ras_error_data_fini(&err_data);
+		} else if (amdgpu_uniras_enabled(adev)) {
+			struct ras_ih_info ih_info = {0};
+
+			ih_info.block = block;
+			ih_info.pasid = pasid;
+			ih_info.reset = reset;
+			ih_info.pasid_fn = pasid_fn;
+			ih_info.data = data;
+			amdgpu_ras_mgr_handle_consumer_interrupt(adev, &ih_info);
 		} else {
 			struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 			int ret;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
index ec203f9e5ffa..28dff750c47e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h
@@ -113,6 +113,8 @@ struct amdgpu_umc_ras {
 	uint32_t (*get_die_id_from_pa)(struct amdgpu_device *adev,
 			uint64_t mca_addr, uint64_t retired_page);
 	void (*get_retire_flip_bits)(struct amdgpu_device *adev);
+	void (*mca_ipid_parse)(struct amdgpu_device *adev, uint64_t ipid,
+			uint32_t *did, uint32_t *ch, uint32_t *umc_inst, uint32_t *sid);
 };
 
 struct amdgpu_umc_funcs {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
index 13cc5a686dfd..9a969175900e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
@@ -25,12 +25,13 @@
 #include <drm/drm_auth.h>
 #include <drm/drm_exec.h>
 #include <linux/pm_runtime.h>
+#include <drm/drm_drv.h>
 
 #include "amdgpu.h"
+#include "amdgpu_reset.h"
 #include "amdgpu_vm.h"
 #include "amdgpu_userq.h"
 #include "amdgpu_hmm.h"
-#include "amdgpu_reset.h"
 #include "amdgpu_userq_fence.h"
 
 u32 amdgpu_userq_get_supported_ip_mask(struct amdgpu_device *adev)
@@ -46,6 +47,107 @@ u32 amdgpu_userq_get_supported_ip_mask(struct amdgpu_device *adev)
 	return userq_ip_mask;
 }
 
+static bool amdgpu_userq_is_reset_type_supported(struct amdgpu_device *adev,
+				enum amdgpu_ring_type ring_type, int reset_type)
+{
+
+	if (ring_type < 0 || ring_type >= AMDGPU_RING_TYPE_MAX)
+		return false;
+
+	switch (ring_type) {
+	case AMDGPU_RING_TYPE_GFX:
+		if (adev->gfx.gfx_supported_reset & reset_type)
+			return true;
+		break;
+	case AMDGPU_RING_TYPE_COMPUTE:
+		if (adev->gfx.compute_supported_reset & reset_type)
+			return true;
+		break;
+	case AMDGPU_RING_TYPE_SDMA:
+		if (adev->sdma.supported_reset & reset_type)
+			return true;
+		break;
+	case AMDGPU_RING_TYPE_VCN_DEC:
+	case AMDGPU_RING_TYPE_VCN_ENC:
+		if (adev->vcn.supported_reset & reset_type)
+			return true;
+		break;
+	case AMDGPU_RING_TYPE_VCN_JPEG:
+		if (adev->jpeg.supported_reset & reset_type)
+			return true;
+		break;
+	default:
+		break;
+	}
+	return false;
+}
+
+static void amdgpu_userq_gpu_reset(struct amdgpu_device *adev)
+{
+	if (amdgpu_device_should_recover_gpu(adev)) {
+		amdgpu_reset_domain_schedule(adev->reset_domain,
+					     &adev->userq_reset_work);
+		/* Wait for the reset job to complete */
+		flush_work(&adev->userq_reset_work);
+	}
+}
+
+static int
+amdgpu_userq_detect_and_reset_queues(struct amdgpu_userq_mgr *uq_mgr)
+{
+	struct amdgpu_device *adev = uq_mgr->adev;
+	const int queue_types[] = {
+		AMDGPU_RING_TYPE_COMPUTE,
+		AMDGPU_RING_TYPE_GFX,
+		AMDGPU_RING_TYPE_SDMA
+	};
+	const int num_queue_types = ARRAY_SIZE(queue_types);
+	bool gpu_reset = false;
+	int r = 0;
+	int i;
+
+	/* Warning if current process mutex is not held */
+	WARN_ON(!mutex_is_locked(&uq_mgr->userq_mutex));
+
+	if (unlikely(adev->debug_disable_gpu_ring_reset)) {
+		dev_err(adev->dev, "userq reset disabled by debug mask\n");
+		return 0;
+	}
+
+	/*
+	 * If GPU recovery feature is disabled system-wide,
+	 * skip all reset detection logic
+	 */
+	if (!amdgpu_gpu_recovery)
+		return 0;
+
+	/*
+	 * Iterate through all queue types to detect and reset problematic queues
+	 * Process each queue type in the defined order
+	 */
+	for (i = 0; i < num_queue_types; i++) {
+		int ring_type = queue_types[i];
+		const struct amdgpu_userq_funcs *funcs = adev->userq_funcs[ring_type];
+
+		if (!amdgpu_userq_is_reset_type_supported(adev, ring_type, AMDGPU_RESET_TYPE_PER_QUEUE))
+				continue;
+
+		if (atomic_read(&uq_mgr->userq_count[ring_type]) > 0 &&
+		    funcs && funcs->detect_and_reset) {
+			r = funcs->detect_and_reset(adev, ring_type);
+			if (r) {
+				gpu_reset = true;
+				break;
+			}
+		}
+	}
+
+	if (gpu_reset)
+		amdgpu_userq_gpu_reset(adev);
+
+	return r;
+}
+
 static int amdgpu_userq_buffer_va_list_add(struct amdgpu_usermode_queue *queue,
 					   struct amdgpu_bo_va_mapping *va_map, u64 addr)
 {
@@ -176,17 +278,22 @@ amdgpu_userq_preempt_helper(struct amdgpu_userq_mgr *uq_mgr,
 	struct amdgpu_device *adev = uq_mgr->adev;
 	const struct amdgpu_userq_funcs *userq_funcs =
 		adev->userq_funcs[queue->queue_type];
+	bool found_hung_queue = false;
 	int r = 0;
 
 	if (queue->state == AMDGPU_USERQ_STATE_MAPPED) {
 		r = userq_funcs->preempt(uq_mgr, queue);
 		if (r) {
 			queue->state = AMDGPU_USERQ_STATE_HUNG;
+			found_hung_queue = true;
 		} else {
 			queue->state = AMDGPU_USERQ_STATE_PREEMPTED;
 		}
 	}
 
+	if (found_hung_queue)
+		amdgpu_userq_detect_and_reset_queues(uq_mgr);
+
 	return r;
 }
 
@@ -218,16 +325,23 @@ amdgpu_userq_unmap_helper(struct amdgpu_userq_mgr *uq_mgr,
 	struct amdgpu_device *adev = uq_mgr->adev;
 	const struct amdgpu_userq_funcs *userq_funcs =
 		adev->userq_funcs[queue->queue_type];
+	bool found_hung_queue = false;
 	int r = 0;
 
 	if ((queue->state == AMDGPU_USERQ_STATE_MAPPED) ||
 		(queue->state == AMDGPU_USERQ_STATE_PREEMPTED)) {
 		r = userq_funcs->unmap(uq_mgr, queue);
-		if (r)
+		if (r) {
 			queue->state = AMDGPU_USERQ_STATE_HUNG;
-		else
+			found_hung_queue = true;
+		} else {
 			queue->state = AMDGPU_USERQ_STATE_UNMAPPED;
+		}
 	}
+
+	if (found_hung_queue)
+		amdgpu_userq_detect_and_reset_queues(uq_mgr);
+
 	return r;
 }
 
@@ -244,10 +358,12 @@ amdgpu_userq_map_helper(struct amdgpu_userq_mgr *uq_mgr,
 		r = userq_funcs->map(uq_mgr, queue);
 		if (r) {
 			queue->state = AMDGPU_USERQ_STATE_HUNG;
+			amdgpu_userq_detect_and_reset_queues(uq_mgr);
 		} else {
 			queue->state = AMDGPU_USERQ_STATE_MAPPED;
 		}
 	}
+
 	return r;
 }
 
@@ -475,10 +591,11 @@ amdgpu_userq_destroy(struct drm_file *filp, int queue_id)
 		amdgpu_bo_unreserve(queue->db_obj.obj);
 	}
 	amdgpu_bo_unref(&queue->db_obj.obj);
-
+	atomic_dec(&uq_mgr->userq_count[queue->queue_type]);
 #if defined(CONFIG_DEBUG_FS)
 	debugfs_remove_recursive(queue->debugfs_queue);
 #endif
+	amdgpu_userq_detect_and_reset_queues(uq_mgr);
 	r = amdgpu_userq_unmap_helper(uq_mgr, queue);
 	/*TODO: It requires a reset for userq hw unmap error*/
 	if (unlikely(r != AMDGPU_USERQ_STATE_UNMAPPED)) {
@@ -701,6 +818,7 @@ amdgpu_userq_create(struct drm_file *filp, union drm_amdgpu_userq *args)
 	kfree(queue_name);
 
 	args->out.queue_id = qid;
+	atomic_inc(&uq_mgr->userq_count[queue->queue_type]);
 
 unlock:
 	mutex_unlock(&uq_mgr->userq_mutex);
@@ -943,6 +1061,8 @@ retry_lock:
 	/* Validate User Ptr BOs */
 	list_for_each_entry(bo_va, &vm->done, base.vm_status) {
 		bo = bo_va->base.bo;
+		if (!bo)
+			continue;
 
 		if (!amdgpu_ttm_tt_is_userptr(bo->tbo.ttm))
 			continue;
@@ -1045,6 +1165,7 @@ amdgpu_userq_evict_all(struct amdgpu_userq_mgr *uq_mgr)
 	unsigned long queue_id;
 	int ret = 0, r;
 
+	amdgpu_userq_detect_and_reset_queues(uq_mgr);
 	/* Try to unmap all the queues in this process ctx */
 	xa_for_each(&uq_mgr->userq_mgr_xa, queue_id, queue) {
 		r = amdgpu_userq_preempt_helper(uq_mgr, queue);
@@ -1057,6 +1178,23 @@ amdgpu_userq_evict_all(struct amdgpu_userq_mgr *uq_mgr)
 	return ret;
 }
 
+void amdgpu_userq_reset_work(struct work_struct *work)
+{
+	struct amdgpu_device *adev = container_of(work, struct amdgpu_device,
+						  userq_reset_work);
+	struct amdgpu_reset_context reset_context;
+
+	memset(&reset_context, 0, sizeof(reset_context));
+
+	reset_context.method = AMD_RESET_METHOD_NONE;
+	reset_context.reset_req_dev = adev;
+	reset_context.src = AMDGPU_RESET_SRC_USERQ;
+	set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
+	/*set_bit(AMDGPU_SKIP_COREDUMP, &reset_context.flags);*/
+
+	amdgpu_device_gpu_recover(adev, NULL, &reset_context);
+}
+
 static int
 amdgpu_userq_wait_for_signal(struct amdgpu_userq_mgr *uq_mgr)
 {
@@ -1084,22 +1222,19 @@ void
 amdgpu_userq_evict(struct amdgpu_userq_mgr *uq_mgr,
 		   struct amdgpu_eviction_fence *ev_fence)
 {
-	int ret;
 	struct amdgpu_fpriv *fpriv = uq_mgr_to_fpriv(uq_mgr);
 	struct amdgpu_eviction_fence_mgr *evf_mgr = &fpriv->evf_mgr;
+	struct amdgpu_device *adev = uq_mgr->adev;
+	int ret;
 
 	/* Wait for any pending userqueue fence work to finish */
 	ret = amdgpu_userq_wait_for_signal(uq_mgr);
-	if (ret) {
-		drm_file_err(uq_mgr->file, "Not evicting userqueue, timeout waiting for work\n");
-		return;
-	}
+	if (ret)
+		dev_err(adev->dev, "Not evicting userqueue, timeout waiting for work\n");
 
 	ret = amdgpu_userq_evict_all(uq_mgr);
-	if (ret) {
-		drm_file_err(uq_mgr->file, "Failed to evict userqueue\n");
-		return;
-	}
+	if (ret)
+		dev_err(adev->dev, "Failed to evict userqueue\n");
 
 	/* Signal current eviction fence */
 	amdgpu_eviction_fence_signal(evf_mgr, ev_fence);
@@ -1133,6 +1268,7 @@ void amdgpu_userq_mgr_fini(struct amdgpu_userq_mgr *userq_mgr)
 	cancel_delayed_work_sync(&userq_mgr->resume_work);
 
 	mutex_lock(&userq_mgr->userq_mutex);
+	amdgpu_userq_detect_and_reset_queues(userq_mgr);
 	xa_for_each(&userq_mgr->userq_mgr_xa, queue_id, queue) {
 		amdgpu_userq_wait_for_last_fence(userq_mgr, queue);
 		amdgpu_userq_unmap_helper(userq_mgr, queue);
@@ -1159,6 +1295,7 @@ int amdgpu_userq_suspend(struct amdgpu_device *adev)
 		uqm = queue->userq_mgr;
 		cancel_delayed_work_sync(&uqm->resume_work);
 		guard(mutex)(&uqm->userq_mutex);
+		amdgpu_userq_detect_and_reset_queues(uqm);
 		if (adev->in_s0ix)
 			r = amdgpu_userq_preempt_helper(uqm, queue);
 		else
@@ -1217,6 +1354,7 @@ int amdgpu_userq_stop_sched_for_enforce_isolation(struct amdgpu_device *adev,
 		if (((queue->queue_type == AMDGPU_HW_IP_GFX) ||
 		     (queue->queue_type == AMDGPU_HW_IP_COMPUTE)) &&
 		    (queue->xcp_id == idx)) {
+			amdgpu_userq_detect_and_reset_queues(uqm);
 			r = amdgpu_userq_preempt_helper(uqm, queue);
 			if (r)
 				ret = r;
@@ -1289,3 +1427,56 @@ int amdgpu_userq_gem_va_unmap_validate(struct amdgpu_device *adev,
 
 	return 0;
 }
+
+void amdgpu_userq_pre_reset(struct amdgpu_device *adev)
+{
+	const struct amdgpu_userq_funcs *userq_funcs;
+	struct amdgpu_usermode_queue *queue;
+	struct amdgpu_userq_mgr *uqm;
+	unsigned long queue_id;
+
+	xa_for_each(&adev->userq_doorbell_xa, queue_id, queue) {
+		uqm = queue->userq_mgr;
+		cancel_delayed_work_sync(&uqm->resume_work);
+		if (queue->state == AMDGPU_USERQ_STATE_MAPPED) {
+			amdgpu_userq_wait_for_last_fence(uqm, queue);
+			userq_funcs = adev->userq_funcs[queue->queue_type];
+			userq_funcs->unmap(uqm, queue);
+			/* just mark all queues as hung at this point.
+			 * if unmap succeeds, we could map again
+			 * in amdgpu_userq_post_reset() if vram is not lost
+			 */
+			queue->state = AMDGPU_USERQ_STATE_HUNG;
+			amdgpu_userq_fence_driver_force_completion(queue);
+		}
+	}
+}
+
+int amdgpu_userq_post_reset(struct amdgpu_device *adev, bool vram_lost)
+{
+	/* if any queue state is AMDGPU_USERQ_STATE_UNMAPPED
+	 * at this point, we should be able to map it again
+	 * and continue if vram is not lost.
+	 */
+	struct amdgpu_userq_mgr *uqm;
+	struct amdgpu_usermode_queue *queue;
+	const struct amdgpu_userq_funcs *userq_funcs;
+	unsigned long queue_id;
+	int r = 0;
+
+	xa_for_each(&adev->userq_doorbell_xa, queue_id, queue) {
+		uqm = queue->userq_mgr;
+		if (queue->state == AMDGPU_USERQ_STATE_HUNG && !vram_lost) {
+			userq_funcs = adev->userq_funcs[queue->queue_type];
+			/* Re-map queue */
+			r = userq_funcs->map(uqm, queue);
+			if (r) {
+				dev_err(adev->dev, "Failed to remap queue %ld\n", queue_id);
+				continue;
+			}
+			queue->state = AMDGPU_USERQ_STATE_MAPPED;
+		}
+	}
+
+	return r;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h
index 09da0617bfa2..c37444427a14 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h
@@ -106,6 +106,7 @@ struct amdgpu_userq_mgr {
 	struct amdgpu_device		*adev;
 	struct delayed_work		resume_work;
 	struct drm_file			*file;
+	atomic_t                        userq_count[AMDGPU_RING_TYPE_MAX];
 };
 
 struct amdgpu_db_info {
@@ -148,6 +149,10 @@ int amdgpu_userq_stop_sched_for_enforce_isolation(struct amdgpu_device *adev,
 						  u32 idx);
 int amdgpu_userq_start_sched_for_enforce_isolation(struct amdgpu_device *adev,
 						   u32 idx);
+void amdgpu_userq_reset_work(struct work_struct *work);
+void amdgpu_userq_pre_reset(struct amdgpu_device *adev);
+int amdgpu_userq_post_reset(struct amdgpu_device *adev, bool vram_lost);
+
 int amdgpu_userq_input_va_validate(struct amdgpu_usermode_queue *queue,
 				   u64 addr, u64 expected_size);
 int amdgpu_userq_gem_va_unmap_validate(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
index 2aeeaa954882..eba9fb359047 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
@@ -151,15 +151,16 @@ void amdgpu_userq_fence_driver_process(struct amdgpu_userq_fence_driver *fence_d
 {
 	struct amdgpu_userq_fence *userq_fence, *tmp;
 	struct dma_fence *fence;
+	unsigned long flags;
 	u64 rptr;
 	int i;
 
 	if (!fence_drv)
 		return;
 
+	spin_lock_irqsave(&fence_drv->fence_list_lock, flags);
 	rptr = amdgpu_userq_fence_read(fence_drv);
 
-	spin_lock(&fence_drv->fence_list_lock);
 	list_for_each_entry_safe(userq_fence, tmp, &fence_drv->fences, link) {
 		fence = &userq_fence->base;
 
@@ -174,7 +175,7 @@ void amdgpu_userq_fence_driver_process(struct amdgpu_userq_fence_driver *fence_d
 		list_del(&userq_fence->link);
 		dma_fence_put(fence);
 	}
-	spin_unlock(&fence_drv->fence_list_lock);
+	spin_unlock_irqrestore(&fence_drv->fence_list_lock, flags);
 }
 
 void amdgpu_userq_fence_driver_destroy(struct kref *ref)
@@ -386,6 +387,7 @@ static int amdgpu_userq_fence_read_wptr(struct amdgpu_usermode_queue *queue,
 	amdgpu_bo_unreserve(queue->vm->root.bo);
 	r = amdgpu_bo_reserve(bo, true);
 	if (r) {
+		amdgpu_bo_unref(&bo);
 		DRM_ERROR("Failed to reserve userqueue wptr bo");
 		return r;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index ce318f5de047..a7d8f1ce6ac2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -41,6 +41,9 @@
 #define VCE_IDLE_TIMEOUT	msecs_to_jiffies(1000)
 
 /* Firmware Names */
+#ifdef CONFIG_DRM_AMDGPU_SI
+#define FIRMWARE_VCE_V1_0	"amdgpu/vce_1_0_0.bin"
+#endif
 #ifdef CONFIG_DRM_AMDGPU_CIK
 #define FIRMWARE_BONAIRE	"amdgpu/bonaire_vce.bin"
 #define FIRMWARE_KABINI	"amdgpu/kabini_vce.bin"
@@ -61,6 +64,9 @@
 #define FIRMWARE_VEGA12		"amdgpu/vega12_vce.bin"
 #define FIRMWARE_VEGA20		"amdgpu/vega20_vce.bin"
 
+#ifdef CONFIG_DRM_AMDGPU_SI
+MODULE_FIRMWARE(FIRMWARE_VCE_V1_0);
+#endif
 #ifdef CONFIG_DRM_AMDGPU_CIK
 MODULE_FIRMWARE(FIRMWARE_BONAIRE);
 MODULE_FIRMWARE(FIRMWARE_KABINI);
@@ -88,82 +94,93 @@ static int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 				      bool direct, struct dma_fence **fence);
 
 /**
- * amdgpu_vce_sw_init - allocate memory, load vce firmware
+ * amdgpu_vce_firmware_name() - determine the firmware file name for VCE
  *
  * @adev: amdgpu_device pointer
- * @size: size for the new BO
  *
- * First step to get VCE online, allocate memory and load the firmware
+ * Each chip that has VCE IP may need a different firmware.
+ * This function returns the name of the VCE firmware file
+ * appropriate for the current chip.
  */
-int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
+static const char *amdgpu_vce_firmware_name(struct amdgpu_device *adev)
 {
-	const char *fw_name;
-	const struct common_firmware_header *hdr;
-	unsigned int ucode_version, version_major, version_minor, binary_id;
-	int i, r;
-
 	switch (adev->asic_type) {
+#ifdef CONFIG_DRM_AMDGPU_SI
+	case CHIP_PITCAIRN:
+	case CHIP_TAHITI:
+	case CHIP_VERDE:
+		return FIRMWARE_VCE_V1_0;
+#endif
 #ifdef CONFIG_DRM_AMDGPU_CIK
 	case CHIP_BONAIRE:
-		fw_name = FIRMWARE_BONAIRE;
-		break;
+		return FIRMWARE_BONAIRE;
 	case CHIP_KAVERI:
-		fw_name = FIRMWARE_KAVERI;
-		break;
+		return FIRMWARE_KAVERI;
 	case CHIP_KABINI:
-		fw_name = FIRMWARE_KABINI;
-		break;
+		return FIRMWARE_KABINI;
 	case CHIP_HAWAII:
-		fw_name = FIRMWARE_HAWAII;
-		break;
+		return FIRMWARE_HAWAII;
 	case CHIP_MULLINS:
-		fw_name = FIRMWARE_MULLINS;
-		break;
+		return FIRMWARE_MULLINS;
 #endif
 	case CHIP_TONGA:
-		fw_name = FIRMWARE_TONGA;
-		break;
+		return  FIRMWARE_TONGA;
 	case CHIP_CARRIZO:
-		fw_name = FIRMWARE_CARRIZO;
-		break;
+		return  FIRMWARE_CARRIZO;
 	case CHIP_FIJI:
-		fw_name = FIRMWARE_FIJI;
-		break;
+		return  FIRMWARE_FIJI;
 	case CHIP_STONEY:
-		fw_name = FIRMWARE_STONEY;
-		break;
+		return  FIRMWARE_STONEY;
 	case CHIP_POLARIS10:
-		fw_name = FIRMWARE_POLARIS10;
-		break;
+		return  FIRMWARE_POLARIS10;
 	case CHIP_POLARIS11:
-		fw_name = FIRMWARE_POLARIS11;
-		break;
+		return  FIRMWARE_POLARIS11;
 	case CHIP_POLARIS12:
-		fw_name = FIRMWARE_POLARIS12;
-		break;
+		return  FIRMWARE_POLARIS12;
 	case CHIP_VEGAM:
-		fw_name = FIRMWARE_VEGAM;
-		break;
+		return  FIRMWARE_VEGAM;
 	case CHIP_VEGA10:
-		fw_name = FIRMWARE_VEGA10;
-		break;
+		return  FIRMWARE_VEGA10;
 	case CHIP_VEGA12:
-		fw_name = FIRMWARE_VEGA12;
-		break;
+		return  FIRMWARE_VEGA12;
 	case CHIP_VEGA20:
-		fw_name = FIRMWARE_VEGA20;
-		break;
+		return  FIRMWARE_VEGA20;
 
 	default:
-		return -EINVAL;
+		return NULL;
 	}
+}
+
+/**
+ * amdgpu_vce_early_init() - try to load VCE firmware
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Tries to load the VCE firmware.
+ *
+ * When not found, returns ENOENT so that the driver can
+ * still load and initialize the rest of the IP blocks.
+ * The GPU can function just fine without VCE, they will just
+ * not support video encoding.
+ */
+int amdgpu_vce_early_init(struct amdgpu_device *adev)
+{
+	const char *fw_name = amdgpu_vce_firmware_name(adev);
+	const struct common_firmware_header *hdr;
+	unsigned int ucode_version, version_major, version_minor, binary_id;
+	int r;
+
+	if (!fw_name)
+		return -ENOENT;
 
 	r = amdgpu_ucode_request(adev, &adev->vce.fw, AMDGPU_UCODE_REQUIRED, "%s", fw_name);
 	if (r) {
-		dev_err(adev->dev, "amdgpu_vce: Can't validate firmware \"%s\"\n",
-			fw_name);
+		dev_err(adev->dev,
+			"amdgpu_vce: Firmware \"%s\" not found or failed to validate (%d)\n",
+			fw_name, r);
+
 		amdgpu_ucode_release(&adev->vce.fw);
-		return r;
+		return -ENOENT;
 	}
 
 	hdr = (const struct common_firmware_header *)adev->vce.fw->data;
@@ -172,11 +189,35 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	version_major = (ucode_version >> 20) & 0xfff;
 	version_minor = (ucode_version >> 8) & 0xfff;
 	binary_id = ucode_version & 0xff;
-	DRM_INFO("Found VCE firmware Version: %d.%d Binary ID: %d\n",
+	dev_info(adev->dev, "Found VCE firmware Version: %d.%d Binary ID: %d\n",
 		version_major, version_minor, binary_id);
 	adev->vce.fw_version = ((version_major << 24) | (version_minor << 16) |
 				(binary_id << 8));
 
+	return 0;
+}
+
+/**
+ * amdgpu_vce_sw_init() - allocate memory for VCE BO
+ *
+ * @adev: amdgpu_device pointer
+ * @size: size for the new BO
+ *
+ * First step to get VCE online: allocate memory for VCE BO.
+ * The VCE firmware binary is copied into the VCE BO later,
+ * in amdgpu_vce_resume. The VCE executes its code from the
+ * VCE BO and also uses the space in this BO for its stack and data.
+ *
+ * Ideally this BO should be placed in VRAM for optimal performance,
+ * although technically it also runs from system RAM (albeit slowly).
+ */
+int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
+{
+	int i, r;
+
+	if (!adev->vce.fw)
+		return -ENOENT;
+
 	r = amdgpu_bo_create_kernel(adev, size, PAGE_SIZE,
 				    AMDGPU_GEM_DOMAIN_VRAM |
 				    AMDGPU_GEM_DOMAIN_GTT,
@@ -285,40 +326,23 @@ int amdgpu_vce_suspend(struct amdgpu_device *adev)
  */
 int amdgpu_vce_resume(struct amdgpu_device *adev)
 {
-	void *cpu_addr;
 	const struct common_firmware_header *hdr;
 	unsigned int offset;
-	int r, idx;
+	int idx;
 
 	if (adev->vce.vcpu_bo == NULL)
 		return -EINVAL;
 
-	r = amdgpu_bo_reserve(adev->vce.vcpu_bo, false);
-	if (r) {
-		dev_err(adev->dev, "(%d) failed to reserve VCE bo\n", r);
-		return r;
-	}
-
-	r = amdgpu_bo_kmap(adev->vce.vcpu_bo, &cpu_addr);
-	if (r) {
-		amdgpu_bo_unreserve(adev->vce.vcpu_bo);
-		dev_err(adev->dev, "(%d) VCE map failed\n", r);
-		return r;
-	}
-
 	hdr = (const struct common_firmware_header *)adev->vce.fw->data;
 	offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
 
 	if (drm_dev_enter(adev_to_drm(adev), &idx)) {
-		memcpy_toio(cpu_addr, adev->vce.fw->data + offset,
+		memset_io(adev->vce.cpu_addr, 0, amdgpu_bo_size(adev->vce.vcpu_bo));
+		memcpy_toio(adev->vce.cpu_addr, adev->vce.fw->data + offset,
 			    adev->vce.fw->size - offset);
 		drm_dev_exit(idx);
 	}
 
-	amdgpu_bo_kunmap(adev->vce.vcpu_bo);
-
-	amdgpu_bo_unreserve(adev->vce.vcpu_bo);
-
 	return 0;
 }
 
@@ -427,6 +451,24 @@ void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
 }
 
 /**
+ * amdgpu_vce_required_gart_pages() - gets number of GART pages required by VCE
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Returns how many GART pages we need before GTT for the VCE IP block.
+ * For VCE1, see vce_v1_0_ensure_vcpu_bo_32bit_addr for details.
+ * For VCE2+, this is not needed so return zero.
+ */
+u32 amdgpu_vce_required_gart_pages(struct amdgpu_device *adev)
+{
+	/* VCE IP block not added yet, so can't use amdgpu_ip_version */
+	if (adev->family == AMDGPU_FAMILY_SI)
+		return 512;
+
+	return 0;
+}
+
+/**
  * amdgpu_vce_get_create_msg - generate a VCE create msg
  *
  * @ring: ring we should submit the msg to
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
index 6e53f872d084..1c3464ce5037 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
@@ -51,14 +51,17 @@ struct amdgpu_vce {
 	struct drm_sched_entity	entity;
 	uint32_t                srbm_soft_reset;
 	unsigned		num_rings;
+	uint32_t		keyselect;
 };
 
+int amdgpu_vce_early_init(struct amdgpu_device *adev);
 int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size);
 int amdgpu_vce_sw_fini(struct amdgpu_device *adev);
 int amdgpu_vce_entity_init(struct amdgpu_device *adev, struct amdgpu_ring *ring);
 int amdgpu_vce_suspend(struct amdgpu_device *adev);
 int amdgpu_vce_resume(struct amdgpu_device *adev);
 void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp);
+u32 amdgpu_vce_required_gart_pages(struct amdgpu_device *adev);
 int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, struct amdgpu_job *job,
 			     struct amdgpu_ib *ib);
 int amdgpu_vce_ring_parse_cs_vm(struct amdgpu_cs_parser *p,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
index f2ce8f506aa8..47a6ce4fdc74 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -917,6 +917,7 @@ static void amdgpu_virt_init_ras(struct amdgpu_device *adev)
 			    RATELIMIT_MSG_ON_RELEASE);
 
 	mutex_init(&adev->virt.ras.ras_telemetry_mutex);
+	mutex_init(&adev->virt.access_req_mutex);
 
 	adev->virt.ras.cper_rptr = 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
index 14d864be5800..01d5bca2dee1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
@@ -150,6 +150,7 @@ enum AMDGIM_FEATURE_FLAG {
 	AMDGIM_FEATURE_RAS_CAPS = (1 << 9),
 	AMDGIM_FEATURE_RAS_TELEMETRY = (1 << 10),
 	AMDGIM_FEATURE_RAS_CPER = (1 << 11),
+	AMDGIM_FEATURE_XGMI_TA_EXT_PEER_LINK = (1 << 12),
 };
 
 enum AMDGIM_REG_ACCESS_FLAG {
@@ -324,6 +325,8 @@ struct amdgpu_virt {
 	/* Spinlock to protect access to the RLCG register interface */
 	spinlock_t rlcg_reg_lock;
 
+	struct mutex access_req_mutex;
+
 	union amd_sriov_ras_caps ras_en_caps;
 	union amd_sriov_ras_caps ras_telemetry_en_caps;
 	struct amdgpu_virt_ras ras;
@@ -395,6 +398,9 @@ struct amdgpu_video_codec_info;
 #define amdgpu_sriov_ras_cper_en(adev) \
 ((adev)->virt.gim_feature & AMDGIM_FEATURE_RAS_CPER)
 
+#define amdgpu_sriov_xgmi_ta_ext_peer_link_en(adev) \
+((adev)->virt.gim_feature & AMDGIM_FEATURE_XGMI_TA_EXT_PEER_LINK)
+
 static inline bool is_virtual_machine(void)
 {
 #if defined(CONFIG_X86)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 9309830821b7..c362d4dfb5bb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -484,15 +484,19 @@ int amdgpu_vm_lock_done_list(struct amdgpu_vm *vm, struct drm_exec *exec,
 	spin_lock(&vm->status_lock);
 	while (!list_is_head(prev->next, &vm->done)) {
 		bo_va = list_entry(prev->next, typeof(*bo_va), base.vm_status);
-		spin_unlock(&vm->status_lock);
 
 		bo = bo_va->base.bo;
 		if (bo) {
+			amdgpu_bo_ref(bo);
+			spin_unlock(&vm->status_lock);
+
 			ret = drm_exec_prepare_obj(exec, &bo->tbo.base, 1);
+			amdgpu_bo_unref(&bo);
 			if (unlikely(ret))
 				return ret;
+
+			spin_lock(&vm->status_lock);
 		}
-		spin_lock(&vm->status_lock);
 		prev = prev->next;
 	}
 	spin_unlock(&vm->status_lock);
@@ -845,6 +849,8 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
 		if (r)
 			return r;
 		fence = &job->hw_vm_fence->base;
+		/* get a ref for the job */
+		dma_fence_get(fence);
 	}
 
 	if (vm_flush_needed) {
@@ -1063,7 +1069,9 @@ amdgpu_vm_tlb_flush(struct amdgpu_vm_update_params *params,
 	}
 
 	/* Prepare a TLB flush fence to be attached to PTs */
-	if (!params->unlocked && vm->is_compute_context) {
+	if (!params->unlocked &&
+	    /* SI doesn't support pasid or KIQ/MES */
+	    params->adev->family > AMDGPU_FAMILY_SI) {
 		amdgpu_vm_tlb_fence_create(params->adev, vm, fence);
 
 		/* Makes sure no PD/PT is freed before the flush */
@@ -2087,7 +2095,7 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 		struct amdgpu_bo *bo = before->bo_va->base.bo;
 
 		amdgpu_vm_it_insert(before, &vm->va);
-		if (before->flags & AMDGPU_PTE_PRT_FLAG(adev))
+		if (before->flags & AMDGPU_VM_PAGE_PRT)
 			amdgpu_vm_prt_get(adev);
 
 		if (amdgpu_vm_is_bo_always_valid(vm, bo) &&
@@ -2102,7 +2110,7 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 		struct amdgpu_bo *bo = after->bo_va->base.bo;
 
 		amdgpu_vm_it_insert(after, &vm->va);
-		if (after->flags & AMDGPU_PTE_PRT_FLAG(adev))
+		if (after->flags & AMDGPU_VM_PAGE_PRT)
 			amdgpu_vm_prt_get(adev);
 
 		if (amdgpu_vm_is_bo_always_valid(vm, bo) &&
@@ -2837,8 +2845,6 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
  */
 void amdgpu_vm_manager_init(struct amdgpu_device *adev)
 {
-	unsigned i;
-
 	/* Concurrent flushes are only possible starting with Vega10 and
 	 * are broken on Navi10 and Navi14.
 	 */
@@ -2847,11 +2853,6 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev)
 					      adev->asic_type == CHIP_NAVI14);
 	amdgpu_vmid_mgr_init(adev);
 
-	adev->vm_manager.fence_context =
-		dma_fence_context_alloc(AMDGPU_MAX_RINGS);
-	for (i = 0; i < AMDGPU_MAX_RINGS; ++i)
-		adev->vm_manager.seqno[i] = 0;
-
 	spin_lock_init(&adev->vm_manager.prt_lock);
 	atomic_set(&adev->vm_manager.num_prt_users, 0);
 
@@ -2917,8 +2918,7 @@ int amdgpu_vm_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 	switch (args->in.op) {
 	case AMDGPU_VM_OP_RESERVE_VMID:
 		/* We only have requirement to reserve vmid from gfxhub */
-		amdgpu_vmid_alloc_reserved(adev, vm, AMDGPU_GFXHUB(0));
-		break;
+		return amdgpu_vmid_alloc_reserved(adev, vm, AMDGPU_GFXHUB(0));
 	case AMDGPU_VM_OP_UNRESERVE_VMID:
 		amdgpu_vmid_free_reserved(adev, vm, AMDGPU_GFXHUB(0));
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index cf0ec94e8a07..15d757c016cb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -453,10 +453,6 @@ struct amdgpu_vm_manager {
 	unsigned int				first_kfd_vmid;
 	bool					concurrent_flush;
 
-	/* Handling of VM fences */
-	u64					fence_context;
-	unsigned				seqno[AMDGPU_MAX_RINGS];
-
 	uint64_t				max_pfn;
 	uint32_t				num_level;
 	uint32_t				block_size;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
index 1ede308a7c67..aad530c46a9f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@@ -298,6 +298,9 @@ int amdgpu_xgmi_get_ext_link(struct amdgpu_device *adev, int link_num)
 {
 	int link_map_6_4_x[8] = { 0, 3, 1, 2, 7, 6, 4, 5 };
 
+	if (adev->gmc.xgmi.num_physical_nodes <= 1)
+		return -EINVAL;
+
 	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
 	case IP_VERSION(6, 4, 0):
 	case IP_VERSION(6, 4, 1):
@@ -333,6 +336,10 @@ static u32 xgmi_v6_4_get_link_status(struct amdgpu_device *adev, int global_link
 	}
 
 	i = global_link_num / n;
+
+	if (!(adev->aid_mask & BIT(i)))
+		return U32_MAX;
+
 	addr += adev->asic_funcs->encode_ext_smn_addressing(i);
 
 	return RREG32_PCIE_EXT(addr);
@@ -342,6 +349,9 @@ int amdgpu_get_xgmi_link_status(struct amdgpu_device *adev, int global_link_num)
 {
 	u32 xgmi_state_reg_val;
 
+	if (adev->gmc.xgmi.num_physical_nodes <= 1)
+		return -EINVAL;
+
 	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
 	case IP_VERSION(6, 4, 0):
 	case IP_VERSION(6, 4, 1):
@@ -958,28 +968,6 @@ static int amdgpu_xgmi_initialize_hive_get_data_partition(struct amdgpu_hive_inf
 	return 0;
 }
 
-static void amdgpu_xgmi_fill_topology_info(struct amdgpu_device *adev,
-	struct amdgpu_device *peer_adev)
-{
-	struct psp_xgmi_topology_info *top_info = &adev->psp.xgmi_context.top_info;
-	struct psp_xgmi_topology_info *peer_info = &peer_adev->psp.xgmi_context.top_info;
-
-	for (int i = 0; i < peer_info->num_nodes; i++) {
-		if (peer_info->nodes[i].node_id == adev->gmc.xgmi.node_id) {
-			for (int j = 0; j < top_info->num_nodes; j++) {
-				if (top_info->nodes[j].node_id == peer_adev->gmc.xgmi.node_id) {
-					peer_info->nodes[i].num_hops = top_info->nodes[j].num_hops;
-					peer_info->nodes[i].is_sharing_enabled =
-							top_info->nodes[j].is_sharing_enabled;
-					peer_info->nodes[i].num_links =
-							top_info->nodes[j].num_links;
-					return;
-				}
-			}
-		}
-	}
-}
-
 int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
 {
 	struct psp_xgmi_topology_info *top_info;
@@ -1065,11 +1053,6 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
 				/* To do: continue with some node failed or disable the whole hive*/
 				goto exit_unlock;
 			}
-
-			/* fill the topology info for peers instead of getting from PSP */
-			list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) {
-				amdgpu_xgmi_fill_topology_info(adev, tmp_adev);
-			}
 		} else {
 			/* get latest topology info for each device from psp */
 			list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h
index 1cee083fb6bd..3cdb1e0eca37 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h
@@ -160,7 +160,8 @@ union amd_sriov_msg_feature_flags {
 		uint32_t ras_caps		: 1;
 		uint32_t ras_telemetry		: 1;
 		uint32_t ras_cper		: 1;
-		uint32_t reserved		: 20;
+		uint32_t xgmi_ta_ext_peer_link	: 1;
+		uint32_t reserved		: 19;
 	} flags;
 	uint32_t all;
 };
@@ -198,8 +199,9 @@ union amd_sriov_ras_caps {
 		uint64_t block_jpeg			: 1;
 		uint64_t block_ih			: 1;
 		uint64_t block_mpio			: 1;
+		uint64_t block_mmsch			: 1;
 		uint64_t poison_propogation_mode	: 1;
-		uint64_t reserved			: 44;
+		uint64_t reserved			: 43;
 	} bits;
 	uint64_t all;
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c
index 811124ff88a8..f9e2edf5260b 100644
--- a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c
+++ b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c
@@ -407,7 +407,8 @@ static int aqua_vanjaram_switch_partition_mode(struct amdgpu_xcp_mgr *xcp_mgr,
 		return -EINVAL;
 	}
 
-	if (adev->kfd.init_complete && !amdgpu_in_reset(adev))
+	if (adev->kfd.init_complete && !amdgpu_in_reset(adev) &&
+		!adev->in_suspend)
 		flags |= AMDGPU_XCP_OPS_KFD;
 
 	if (flags & AMDGPU_XCP_OPS_KFD) {
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_ih.c b/drivers/gpu/drm/amd/amdgpu/cik_ih.c
index 41f4705bdbbd..876a3256dba4 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik_ih.c
@@ -156,6 +156,9 @@ static int cik_ih_irq_init(struct amdgpu_device *adev)
 	/* enable irqs */
 	cik_ih_enable_interrupts(adev);
 
+	if (adev->irq.ih_soft.ring_size)
+		adev->irq.ih_soft.enabled = true;
+
 	return 0;
 }
 
@@ -192,6 +195,9 @@ static u32 cik_ih_get_wptr(struct amdgpu_device *adev,
 
 	wptr = le32_to_cpu(*ih->wptr_cpu);
 
+	if (ih == &adev->irq.ih_soft)
+		goto out;
+
 	if (wptr & IH_RB_WPTR__RB_OVERFLOW_MASK) {
 		wptr &= ~IH_RB_WPTR__RB_OVERFLOW_MASK;
 		/* When a ring buffer overflow happen start parsing interrupt
@@ -211,6 +217,8 @@ static u32 cik_ih_get_wptr(struct amdgpu_device *adev,
 		tmp &= ~IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK;
 		WREG32(mmIH_RB_CNTL, tmp);
 	}
+
+out:
 	return (wptr & ih->ptr_mask);
 }
 
@@ -306,6 +314,10 @@ static int cik_ih_sw_init(struct amdgpu_ip_block *ip_block)
 	if (r)
 		return r;
 
+	r = amdgpu_ih_ring_init(adev, &adev->irq.ih_soft, IH_SW_RING_SIZE, true);
+	if (r)
+		return r;
+
 	r = amdgpu_irq_init(adev);
 
 	return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/cz_ih.c b/drivers/gpu/drm/amd/amdgpu/cz_ih.c
index 2f891fb846d5..bc7a2e06ab5f 100644
--- a/drivers/gpu/drm/amd/amdgpu/cz_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/cz_ih.c
@@ -157,6 +157,9 @@ static int cz_ih_irq_init(struct amdgpu_device *adev)
 	/* enable interrupts */
 	cz_ih_enable_interrupts(adev);
 
+	if (adev->irq.ih_soft.ring_size)
+		adev->irq.ih_soft.enabled = true;
+
 	return 0;
 }
 
@@ -194,6 +197,9 @@ static u32 cz_ih_get_wptr(struct amdgpu_device *adev,
 
 	wptr = le32_to_cpu(*ih->wptr_cpu);
 
+	if (ih == &adev->irq.ih_soft)
+		goto out;
+
 	if (!REG_GET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW))
 		goto out;
 
@@ -297,6 +303,10 @@ static int cz_ih_sw_init(struct amdgpu_ip_block *ip_block)
 	if (r)
 		return r;
 
+	r = amdgpu_ih_ring_init(adev, &adev->irq.ih_soft, IH_SW_RING_SIZE, true);
+	if (r)
+		return r;
+
 	r = amdgpu_irq_init(adev);
 
 	return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index 751732f3e883..d75b9940f248 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -4956,7 +4956,8 @@ static int gfx_v10_0_sw_init(struct amdgpu_ip_block *ip_block)
 		amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]);
 	adev->gfx.compute_supported_reset =
 		amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]);
-	if (!amdgpu_sriov_vf(adev)) {
+	if (!amdgpu_sriov_vf(adev) &&
+	    !adev->debug_disable_gpu_ring_reset) {
 		adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		adev->gfx.gfx_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
index 252517ce5d5a..8a2ee2de390f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c
@@ -1821,13 +1821,15 @@ static int gfx_v11_0_sw_init(struct amdgpu_ip_block *ip_block)
 	case IP_VERSION(11, 0, 3):
 		if ((adev->gfx.me_fw_version >= 2280) &&
 		    (adev->gfx.mec_fw_version >= 2410) &&
-		    !amdgpu_sriov_vf(adev)) {
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset) {
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 			adev->gfx.gfx_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		}
 		break;
 	default:
-		if (!amdgpu_sriov_vf(adev)) {
+		if (!amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset) {
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 			adev->gfx.gfx_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		}
@@ -5872,9 +5874,9 @@ static void gfx_v11_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 		if (flags & AMDGPU_IB_PREEMPTED)
 			control |= INDIRECT_BUFFER_PRE_RESUME(1);
 
-		if (vmid)
+		if (vmid && !ring->adev->gfx.rs64_enable)
 			gfx_v11_0_ring_emit_de_meta(ring,
-				    (!amdgpu_sriov_vf(ring->adev) && flags & AMDGPU_IB_PREEMPTED) ? true : false);
+				!amdgpu_sriov_vf(ring->adev) && (flags & AMDGPU_IB_PREEMPTED));
 	}
 
 	amdgpu_ring_write(ring, header);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
index 35d5a7e99a7c..d01d2712cf57 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
@@ -1548,7 +1548,8 @@ static int gfx_v12_0_sw_init(struct amdgpu_ip_block *ip_block)
 	case IP_VERSION(12, 0, 1):
 		if ((adev->gfx.me_fw_version >= 2660) &&
 		    (adev->gfx.mec_fw_version >= 2920) &&
-		    !amdgpu_sriov_vf(adev)) {
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset) {
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 			adev->gfx.gfx_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		}
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
index 7693b7953426..80565392313f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
@@ -3102,6 +3102,11 @@ static int gfx_v6_0_sw_init(struct amdgpu_ip_block *ip_block)
 			return r;
 	}
 
+	adev->gfx.gfx_supported_reset =
+		amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]);
+	adev->gfx.compute_supported_reset =
+		amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]);
+
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
index 5976ed55d9db..2b7aba22ecc1 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -4399,6 +4399,11 @@ static int gfx_v7_0_sw_init(struct amdgpu_ip_block *ip_block)
 
 	gfx_v7_0_gpu_early_init(adev);
 
+	adev->gfx.gfx_supported_reset =
+		amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]);
+	adev->gfx.compute_supported_reset =
+		amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]);
+
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index d3d0a4b0380c..1c87375e1dd5 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -2023,6 +2023,11 @@ static int gfx_v8_0_sw_init(struct amdgpu_ip_block *ip_block)
 	if (r)
 		return r;
 
+	adev->gfx.gfx_supported_reset =
+		amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]);
+	adev->gfx.compute_supported_reset =
+		amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index f1a2efc2a8d0..0148d7ff34d9 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -2409,7 +2409,7 @@ static int gfx_v9_0_sw_init(struct amdgpu_ip_block *ip_block)
 		amdgpu_get_soft_full_reset_mask(&adev->gfx.gfx_ring[0]);
 	adev->gfx.compute_supported_reset =
 		amdgpu_get_soft_full_reset_mask(&adev->gfx.compute_ring[0]);
-	if (!amdgpu_sriov_vf(adev))
+	if (!amdgpu_sriov_vf(adev) && !adev->debug_disable_gpu_ring_reset)
 		adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 
 	r = amdgpu_gfx_kiq_init(adev, GFX9_MEC_HPD_SIZE, 0);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
index e0b50c690f8c..cbb74ffc4792 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
@@ -1149,14 +1149,16 @@ static int gfx_v9_4_3_sw_init(struct amdgpu_ip_block *ip_block)
 	case IP_VERSION(9, 4, 3):
 	case IP_VERSION(9, 4, 4):
 		if ((adev->gfx.mec_fw_version >= 155) &&
-		    !amdgpu_sriov_vf(adev)) {
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset) {
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_PIPE;
 		}
 		break;
 	case IP_VERSION(9, 5, 0):
 		if ((adev->gfx.mec_fw_version >= 21) &&
-		    !amdgpu_sriov_vf(adev)) {
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset) {
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 			adev->gfx.compute_supported_reset |= AMDGPU_RESET_TYPE_PER_PIPE;
 		}
@@ -2289,7 +2291,9 @@ static int gfx_v9_4_3_cp_resume(struct amdgpu_device *adev)
 		r = amdgpu_xcp_init(adev->xcp_mgr, num_xcp, mode);
 
 	} else {
-		if (amdgpu_xcp_query_partition_mode(adev->xcp_mgr,
+		if (adev->in_suspend)
+			amdgpu_xcp_restore_partition_mode(adev->xcp_mgr);
+		else if (amdgpu_xcp_query_partition_mode(adev->xcp_mgr,
 						    AMDGPU_XCP_FL_NONE) ==
 		    AMDGPU_UNKNOWN_COMPUTE_PARTITION_MODE)
 			r = amdgpu_xcp_switch_partition_mode(
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
index d7499be8c4bf..ce6e04242c52 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
@@ -103,8 +103,10 @@ static int gmc_v10_0_process_interrupt(struct amdgpu_device *adev,
 	uint32_t vmhub_index = entry->client_id == SOC15_IH_CLIENTID_VMC ?
 			       AMDGPU_MMHUB0(0) : AMDGPU_GFXHUB(0);
 	struct amdgpu_vmhub *hub = &adev->vmhub[vmhub_index];
-	bool retry_fault = !!(entry->src_data[1] & 0x80);
-	bool write_fault = !!(entry->src_data[1] & 0x20);
+	bool retry_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_RETRY);
+	bool write_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_WRITE);
 	struct amdgpu_task_info *task_info;
 	uint32_t status = 0;
 	u64 addr;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
index 7bc389d9f5c4..ba59ee8e398a 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
@@ -103,12 +103,41 @@ static int gmc_v11_0_process_interrupt(struct amdgpu_device *adev,
 	uint32_t vmhub_index = entry->client_id == SOC21_IH_CLIENTID_VMC ?
 			       AMDGPU_MMHUB0(0) : AMDGPU_GFXHUB(0);
 	struct amdgpu_vmhub *hub = &adev->vmhub[vmhub_index];
+	bool retry_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_RETRY);
+	bool write_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_WRITE);
 	uint32_t status = 0;
 	u64 addr;
 
 	addr = (u64)entry->src_data[0] << 12;
 	addr |= ((u64)entry->src_data[1] & 0xf) << 44;
 
+	if (retry_fault) {
+		/* Returning 1 here also prevents sending the IV to the KFD */
+
+		/* Process it only if it's the first fault for this address */
+		if (entry->ih != &adev->irq.ih_soft &&
+		    amdgpu_gmc_filter_faults(adev, entry->ih, addr, entry->pasid,
+					     entry->timestamp))
+			return 1;
+
+		/* Delegate it to a different ring if the hardware hasn't
+		 * already done it.
+		 */
+		if (entry->ih == &adev->irq.ih) {
+			amdgpu_irq_delegate(adev, entry, 8);
+			return 1;
+		}
+
+		/* Try to handle the recoverable page faults by filling page
+		 * tables
+		 */
+		if (amdgpu_vm_handle_fault(adev, entry->pasid, 0, 0, addr,
+					   entry->timestamp, write_fault))
+			return 1;
+	}
+
 	if (!amdgpu_sriov_vf(adev)) {
 		/*
 		 * Issue a dummy read to wait for the status register to
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
index cad2d19105c4..7a9d6894e321 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
@@ -91,6 +91,10 @@ static int gmc_v12_0_process_interrupt(struct amdgpu_device *adev,
 				       struct amdgpu_iv_entry *entry)
 {
 	struct amdgpu_vmhub *hub;
+	bool retry_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_RETRY);
+	bool write_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_WRITE);
 	uint32_t status = 0;
 	u64 addr;
 
@@ -102,6 +106,31 @@ static int gmc_v12_0_process_interrupt(struct amdgpu_device *adev,
 	else
 		hub = &adev->vmhub[AMDGPU_GFXHUB(0)];
 
+	if (retry_fault) {
+		/* Returning 1 here also prevents sending the IV to the KFD */
+
+		/* Process it only if it's the first fault for this address */
+		if (entry->ih != &adev->irq.ih_soft &&
+		    amdgpu_gmc_filter_faults(adev, entry->ih, addr, entry->pasid,
+					     entry->timestamp))
+			return 1;
+
+		/* Delegate it to a different ring if the hardware hasn't
+		 * already done it.
+		 */
+		if (entry->ih == &adev->irq.ih) {
+			amdgpu_irq_delegate(adev, entry, 8);
+			return 1;
+		}
+
+		/* Try to handle the recoverable page faults by filling page
+		 * tables
+		 */
+		if (amdgpu_vm_handle_fault(adev, entry->pasid, 0, 0, addr,
+					   entry->timestamp, write_fault))
+			return 1;
+	}
+
 	if (!amdgpu_sriov_vf(adev)) {
 		/*
 		 * Issue a dummy read to wait for the status register to
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index f6ad7911f1e6..a8ec95f42926 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -213,7 +213,7 @@ static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev,
 
 	amdgpu_gmc_set_agp_default(adev, mc);
 	amdgpu_gmc_vram_location(adev, mc, base);
-	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_BEST_FIT);
+	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_LOW);
 }
 
 static void gmc_v6_0_mc_program(struct amdgpu_device *adev)
@@ -610,23 +610,21 @@ static void gmc_v6_0_gart_disable(struct amdgpu_device *adev)
 }
 
 static void gmc_v6_0_vm_decode_fault(struct amdgpu_device *adev,
-				     u32 status, u32 addr, u32 mc_client)
+				     u32 status, u32 addr)
 {
 	u32 mc_id;
 	u32 vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, VMID);
 	u32 protections = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS,
 					PROTECTIONS);
-	char block[5] = { mc_client >> 24, (mc_client >> 16) & 0xff,
-		(mc_client >> 8) & 0xff, mc_client & 0xff, 0 };
 
 	mc_id = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS,
 			      MEMORY_CLIENT_ID);
 
-	dev_err(adev->dev, "VM fault (0x%02x, vmid %d) at page %u, %s from '%s' (0x%08x) (%d)\n",
+	dev_err(adev->dev, "VM fault (0x%02x, vmid %d) at page %u, %s from %d\n",
 	       protections, vmid, addr,
 	       REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS,
 			     MEMORY_CLIENT_RW) ?
-	       "write" : "read", block, mc_client, mc_id);
+	       "write" : "read", mc_id);
 }
 
 static const u32 mc_cg_registers[] = {
@@ -1072,6 +1070,12 @@ static int gmc_v6_0_process_interrupt(struct amdgpu_device *adev,
 {
 	u32 addr, status;
 
+	/* Delegate to the soft IRQ handler ring */
+	if (adev->irq.ih_soft.enabled && entry->ih != &adev->irq.ih_soft) {
+		amdgpu_irq_delegate(adev, entry, 4);
+		return 1;
+	}
+
 	addr = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_ADDR);
 	status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS);
 	WREG32_P(mmVM_CONTEXT1_CNTL2, 1, ~1);
@@ -1079,6 +1083,10 @@ static int gmc_v6_0_process_interrupt(struct amdgpu_device *adev,
 	if (!addr && !status)
 		return 0;
 
+	amdgpu_vm_update_fault_cache(adev, entry->pasid,
+				     ((u64)addr) << AMDGPU_GPU_PAGE_SHIFT,
+				     status, AMDGPU_GFXHUB(0));
+
 	if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_FIRST)
 		gmc_v6_0_set_fault_enable_default(adev, false);
 
@@ -1089,7 +1097,7 @@ static int gmc_v6_0_process_interrupt(struct amdgpu_device *adev,
 			addr);
 		dev_err(adev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n",
 			status);
-		gmc_v6_0_vm_decode_fault(adev, status, addr, 0);
+		gmc_v6_0_vm_decode_fault(adev, status, addr);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 0e5e54d0a9a5..fbd0bf147f50 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -1261,6 +1261,12 @@ static int gmc_v7_0_process_interrupt(struct amdgpu_device *adev,
 {
 	u32 addr, status, mc_client, vmid;
 
+	/* Delegate to the soft IRQ handler ring */
+	if (adev->irq.ih_soft.enabled && entry->ih != &adev->irq.ih_soft) {
+		amdgpu_irq_delegate(adev, entry, 4);
+		return 1;
+	}
+
 	addr = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_ADDR);
 	status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS);
 	mc_client = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_MCCLIENT);
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index e1509480dfc2..6551b60f2584 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -1439,6 +1439,12 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev,
 		return 0;
 	}
 
+	/* Delegate to the soft IRQ handler ring */
+	if (adev->irq.ih_soft.enabled && entry->ih != &adev->irq.ih_soft) {
+		amdgpu_irq_delegate(adev, entry, 4);
+		return 1;
+	}
+
 	addr = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_ADDR);
 	status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS);
 	mc_client = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_MCCLIENT);
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index e716097dfde4..8ad7519f7b58 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -544,8 +544,10 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_irq_src *source,
 				      struct amdgpu_iv_entry *entry)
 {
-	bool retry_fault = !!(entry->src_data[1] & 0x80);
-	bool write_fault = !!(entry->src_data[1] & 0x20);
+	bool retry_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_RETRY);
+	bool write_fault = !!(entry->src_data[1] &
+			      AMDGPU_GMC9_FAULT_SOURCE_DATA_WRITE);
 	uint32_t status = 0, cid = 0, rw = 0, fed = 0;
 	struct amdgpu_task_info *task_info;
 	struct amdgpu_vmhub *hub;
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
index 1317ede131b6..01cadf898c00 100644
--- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
@@ -157,6 +157,9 @@ static int iceland_ih_irq_init(struct amdgpu_device *adev)
 	/* enable interrupts */
 	iceland_ih_enable_interrupts(adev);
 
+	if (adev->irq.ih_soft.ring_size)
+		adev->irq.ih_soft.enabled = true;
+
 	return 0;
 }
 
@@ -194,6 +197,9 @@ static u32 iceland_ih_get_wptr(struct amdgpu_device *adev,
 
 	wptr = le32_to_cpu(*ih->wptr_cpu);
 
+	if (ih == &adev->irq.ih_soft)
+		goto out;
+
 	if (!REG_GET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW))
 		goto out;
 
@@ -296,6 +302,10 @@ static int iceland_ih_sw_init(struct amdgpu_ip_block *ip_block)
 	if (r)
 		return r;
 
+	r = amdgpu_ih_ring_init(adev, &adev->irq.ih_soft, IH_SW_RING_SIZE, true);
+	if (r)
+		return r;
+
 	r = amdgpu_irq_init(adev);
 
 	return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_1.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_1.c
index baf097d2e1ac..ab0bf880d3d8 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_1.c
@@ -878,6 +878,7 @@ static const struct amdgpu_ring_funcs jpeg_v5_0_1_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v5_0_1_dec_ring_get_rptr,
 	.get_wptr = jpeg_v5_0_1_dec_ring_get_wptr,
 	.set_wptr = jpeg_v5_0_1_dec_ring_set_wptr,
+	.parse_cs = amdgpu_jpeg_dec_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
diff --git a/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c b/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
index b1ee9473d628..64cae89357b6 100644
--- a/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
+++ b/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
@@ -208,6 +208,7 @@ static int mes_userq_detect_and_reset(struct amdgpu_device *adev,
 	unsigned int hung_db_num = 0;
 	unsigned long queue_id;
 	u32 db_array[8];
+	bool found_hung_queue = false;
 	int r, i;
 
 	if (db_array_size > 8) {
@@ -232,6 +233,7 @@ static int mes_userq_detect_and_reset(struct amdgpu_device *adev,
 				for (i = 0; i < hung_db_num; i++) {
 					if (queue->doorbell_index == db_array[i]) {
 						queue->state = AMDGPU_USERQ_STATE_HUNG;
+						found_hung_queue = true;
 						atomic_inc(&adev->gpu_reset_counter);
 						amdgpu_userq_fence_driver_force_completion(queue);
 						drm_dev_wedged_event(adev_to_drm(adev), DRM_WEDGE_RECOVERY_NONE, NULL);
@@ -241,6 +243,11 @@ static int mes_userq_detect_and_reset(struct amdgpu_device *adev,
 		}
 	}
 
+	if (found_hung_queue) {
+		/* Resume scheduling after hang recovery */
+		r = amdgpu_mes_resume(adev);
+	}
+
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
index cd5b2f07edb8..e7cd07383d56 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
@@ -173,13 +173,17 @@ static void xgpu_nv_mailbox_trans_msg (struct amdgpu_device *adev,
 static int xgpu_nv_send_access_requests_with_param(struct amdgpu_device *adev,
 			enum idh_request req, u32 data1, u32 data2, u32 data3)
 {
-	int r, retry = 1;
+	struct amdgpu_virt *virt = &adev->virt;
+	int r = 0, retry = 1;
 	enum idh_event event = -1;
 
+	mutex_lock(&virt->access_req_mutex);
 send_request:
 
-	if (amdgpu_ras_is_rma(adev))
-		return -ENODEV;
+	if (amdgpu_ras_is_rma(adev)) {
+		r = -ENODEV;
+		goto out;
+	}
 
 	xgpu_nv_mailbox_trans_msg(adev, req, data1, data2, data3);
 
@@ -217,7 +221,7 @@ send_request:
 
 			if (req != IDH_REQ_GPU_INIT_DATA) {
 				dev_err(adev->dev, "Doesn't get msg:%d from pf, error=%d\n", event, r);
-				return r;
+				goto out;
 			} else /* host doesn't support REQ_GPU_INIT_DATA handshake */
 				adev->virt.req_init_data_ver = 0;
 		} else {
@@ -246,7 +250,10 @@ send_request:
 		}
 	}
 
-	return 0;
+out:
+	mutex_unlock(&virt->access_req_mutex);
+
+	return r;
 }
 
 static int xgpu_nv_send_access_requests(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
index 64b240b51f1a..a9be7a505026 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c
@@ -142,13 +142,37 @@ static int psp_v11_0_init_microcode(struct psp_context *psp)
 	return err;
 }
 
-static int psp_v11_0_wait_for_bootloader(struct psp_context *psp)
+static int psp_v11_wait_for_tos_unload(struct psp_context *psp)
 {
 	struct amdgpu_device *adev = psp->adev;
+	uint32_t sol_reg1, sol_reg2;
+	int retry_loop;
 
+	/* Wait for the TOS to be unloaded */
+	for (retry_loop = 0; retry_loop < 20; retry_loop++) {
+		sol_reg1 = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81);
+		usleep_range(1000, 2000);
+		sol_reg2 = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81);
+		if (sol_reg1 == sol_reg2)
+			return 0;
+	}
+	dev_err(adev->dev, "TOS unload failed, C2PMSG_33: %x C2PMSG_81: %x",
+		RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_33),
+		RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81));
+
+	return -ETIME;
+}
+
+static int psp_v11_0_wait_for_bootloader(struct psp_context *psp)
+{
+	struct amdgpu_device *adev = psp->adev;
 	int ret;
 	int retry_loop;
 
+	/* For a reset done at the end of S3, only wait for TOS to be unloaded */
+	if (adev->in_s3 && !(adev->flags & AMD_IS_APU) && amdgpu_in_reset(adev))
+		return psp_v11_wait_for_tos_unload(psp);
+
 	for (retry_loop = 0; retry_loop < 20; retry_loop++) {
 		/* Wait for bootloader to signify that is
 		    ready having bit 31 of C2PMSG_35 set to 1 */
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
index 36b1ca73c2ed..a1443990d5c6 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
@@ -2361,11 +2361,15 @@ static void sdma_v4_4_2_update_reset_mask(struct amdgpu_device *adev)
 	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(9, 4, 3):
 	case IP_VERSION(9, 4, 4):
-		if ((adev->gfx.mec_fw_version >= 0xb0) && amdgpu_dpm_reset_sdma_is_supported(adev))
+		if ((adev->gfx.mec_fw_version >= 0xb0) &&
+		    amdgpu_dpm_reset_sdma_is_supported(adev) &&
+		    !adev->debug_disable_gpu_ring_reset)
 			adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	case IP_VERSION(9, 5, 0):
-		if ((adev->gfx.mec_fw_version >= 0xf) && amdgpu_dpm_reset_sdma_is_supported(adev))
+		if ((adev->gfx.mec_fw_version >= 0xf) &&
+		    amdgpu_dpm_reset_sdma_is_supported(adev) &&
+		    !adev->debug_disable_gpu_ring_reset)
 			adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
index 7dc67a22a7a0..8ddc4df06a1f 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
@@ -1429,7 +1429,8 @@ static int sdma_v5_0_sw_init(struct amdgpu_ip_block *ip_block)
 	case IP_VERSION(5, 0, 2):
 	case IP_VERSION(5, 0, 5):
 		if ((adev->sdma.instance[0].fw_version >= 35) &&
-		    !amdgpu_sriov_vf(adev))
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset)
 			adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c
index 3bd44c24f692..51101b0aa2fa 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c
@@ -342,7 +342,7 @@ static void sdma_v5_2_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg;
 
 	if (ring->me > 1) {
-		amdgpu_asic_flush_hdp(adev, ring);
+		amdgpu_hdp_flush(adev, ring);
 	} else {
 		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me;
 
@@ -1348,12 +1348,14 @@ static int sdma_v5_2_sw_init(struct amdgpu_ip_block *ip_block)
 	case IP_VERSION(5, 2, 3):
 	case IP_VERSION(5, 2, 4):
 		if ((adev->sdma.instance[0].fw_version >= 76) &&
-		    !amdgpu_sriov_vf(adev))
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset)
 			adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	case IP_VERSION(5, 2, 5):
 		if ((adev->sdma.instance[0].fw_version >= 34) &&
-		    !amdgpu_sriov_vf(adev))
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset)
 			adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
index db6e41967f12..217040044987 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c
@@ -1356,7 +1356,8 @@ static int sdma_v6_0_sw_init(struct amdgpu_ip_block *ip_block)
 	case IP_VERSION(6, 0, 2):
 	case IP_VERSION(6, 0, 3):
 		if ((adev->sdma.instance[0].fw_version >= 21) &&
-		    !amdgpu_sriov_vf(adev))
+		    !amdgpu_sriov_vf(adev) &&
+		    !adev->debug_disable_gpu_ring_reset)
 			adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	default:
@@ -1389,7 +1390,7 @@ static int sdma_v6_0_sw_init(struct amdgpu_ip_block *ip_block)
 			adev->userq_funcs[AMDGPU_HW_IP_DMA] = &userq_mes_funcs;
 		break;
 	case IP_VERSION(6, 0, 3):
-		if ((adev->sdma.instance[0].fw_version >= 27) && !adev->sdma.disable_uq)
+		if (adev->sdma.instance[0].fw_version >= 29 && !adev->sdma.disable_uq)
 			adev->userq_funcs[AMDGPU_HW_IP_DMA] = &userq_mes_funcs;
 		break;
 	case IP_VERSION(6, 1, 0):
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c
index 326ecc8d37d2..2b81344dcd66 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c
@@ -1337,7 +1337,8 @@ static int sdma_v7_0_sw_init(struct amdgpu_ip_block *ip_block)
 
 	adev->sdma.supported_reset =
 		amdgpu_get_soft_full_reset_mask(&adev->sdma.instance[0].ring);
-	if (!amdgpu_sriov_vf(adev))
+	if (!amdgpu_sriov_vf(adev) &&
+	    !adev->debug_disable_gpu_ring_reset)
 		adev->sdma.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 
 	r = amdgpu_sdma_sysfs_reset_mask_init(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c
index e0f139de7991..f7288372ee61 100644
--- a/drivers/gpu/drm/amd/amdgpu/si.c
+++ b/drivers/gpu/drm/amd/amdgpu/si.c
@@ -45,6 +45,7 @@
 #include "dce_v6_0.h"
 #include "si.h"
 #include "uvd_v3_1.h"
+#include "vce_v1_0.h"
 
 #include "uvd/uvd_4_0_d.h"
 
@@ -921,8 +922,6 @@ static const u32 hainan_mgcg_cgcg_init[] =
 	0x3630, 0xfffffff0, 0x00000100,
 };
 
-/* XXX: update when we support VCE */
-#if 0
 /* tahiti, pitcairn, verde */
 static const struct amdgpu_video_codec_info tahiti_video_codecs_encode_array[] =
 {
@@ -940,13 +939,7 @@ static const struct amdgpu_video_codecs tahiti_video_codecs_encode =
 	.codec_count = ARRAY_SIZE(tahiti_video_codecs_encode_array),
 	.codec_array = tahiti_video_codecs_encode_array,
 };
-#else
-static const struct amdgpu_video_codecs tahiti_video_codecs_encode =
-{
-	.codec_count = 0,
-	.codec_array = NULL,
-};
-#endif
+
 /* oland and hainan don't support encode */
 static const struct amdgpu_video_codecs hainan_video_codecs_encode =
 {
@@ -1925,6 +1918,14 @@ static int si_set_vce_clocks(struct amdgpu_device *adev, u32 evclk, u32 ecclk)
 		     ~VCEPLL_BYPASS_EN_MASK);
 
 	if (!evclk || !ecclk) {
+		/*
+		 * On some chips, the PLL takes way too long to get out of
+		 * sleep mode, causing a timeout waiting on CTLACK/CTLACK2.
+		 * Leave the PLL running in bypass mode.
+		 */
+		if (adev->pdev->device == 0x6780)
+			return 0;
+
 		/* Keep the Bypass mode, put PLL to sleep */
 		WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, VCEPLL_SLEEP_MASK,
 			     ~VCEPLL_SLEEP_MASK);
@@ -2717,7 +2718,7 @@ int si_set_ip_blocks(struct amdgpu_device *adev)
 		else
 			amdgpu_device_ip_block_add(adev, &dce_v6_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &uvd_v3_1_ip_block);
-		/* amdgpu_device_ip_block_add(adev, &vce_v1_0_ip_block); */
+		amdgpu_device_ip_block_add(adev, &vce_v1_0_ip_block);
 		break;
 	case CHIP_OLAND:
 		amdgpu_device_ip_block_add(adev, &si_common_ip_block);
@@ -2735,7 +2736,6 @@ int si_set_ip_blocks(struct amdgpu_device *adev)
 		else
 			amdgpu_device_ip_block_add(adev, &dce_v6_4_ip_block);
 		amdgpu_device_ip_block_add(adev, &uvd_v3_1_ip_block);
-		/* amdgpu_device_ip_block_add(adev, &vce_v1_0_ip_block); */
 		break;
 	case CHIP_HAINAN:
 		amdgpu_device_ip_block_add(adev, &si_common_ip_block);
diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c
index 1df00f8a2406..66f650f87243 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c
@@ -96,6 +96,9 @@ static int si_ih_irq_init(struct amdgpu_device *adev)
 	pci_set_master(adev->pdev);
 	si_ih_enable_interrupts(adev);
 
+	if (adev->irq.ih_soft.ring_size)
+		adev->irq.ih_soft.enabled = true;
+
 	return 0;
 }
 
@@ -112,6 +115,9 @@ static u32 si_ih_get_wptr(struct amdgpu_device *adev,
 
 	wptr = le32_to_cpu(*ih->wptr_cpu);
 
+	if (ih == &adev->irq.ih_soft)
+		goto out;
+
 	if (wptr & IH_RB_WPTR__RB_OVERFLOW_MASK) {
 		wptr &= ~IH_RB_WPTR__RB_OVERFLOW_MASK;
 		dev_warn(adev->dev, "IH ring buffer overflow (0x%08X, 0x%08X, 0x%08X)\n",
@@ -127,6 +133,8 @@ static u32 si_ih_get_wptr(struct amdgpu_device *adev,
 		tmp &= ~IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK;
 		WREG32(IH_RB_CNTL, tmp);
 	}
+
+out:
 	return (wptr & ih->ptr_mask);
 }
 
@@ -175,6 +183,10 @@ static int si_ih_sw_init(struct amdgpu_ip_block *ip_block)
 	if (r)
 		return r;
 
+	r = amdgpu_ih_ring_init(adev, &adev->irq.ih_soft, IH_SW_RING_SIZE, true);
+	if (r)
+		return r;
+
 	return amdgpu_irq_init(adev);
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/sid.h b/drivers/gpu/drm/amd/amdgpu/sid.h
index cbd4f8951cfa..561462a8332e 100644
--- a/drivers/gpu/drm/amd/amdgpu/sid.h
+++ b/drivers/gpu/drm/amd/amdgpu/sid.h
@@ -582,45 +582,6 @@
 #define	DMA_PACKET_NOP					  0xf
 
 /* VCE */
-#define VCE_STATUS					0x20004
-#define VCE_VCPU_CNTL					0x20014
-#define		VCE_CLK_EN				(1 << 0)
-#define VCE_VCPU_CACHE_OFFSET0				0x20024
-#define VCE_VCPU_CACHE_SIZE0				0x20028
-#define VCE_VCPU_CACHE_OFFSET1				0x2002c
-#define VCE_VCPU_CACHE_SIZE1				0x20030
-#define VCE_VCPU_CACHE_OFFSET2				0x20034
-#define VCE_VCPU_CACHE_SIZE2				0x20038
-#define VCE_SOFT_RESET					0x20120
-#define 	VCE_ECPU_SOFT_RESET			(1 << 0)
-#define 	VCE_FME_SOFT_RESET			(1 << 2)
-#define VCE_RB_BASE_LO2					0x2016c
-#define VCE_RB_BASE_HI2					0x20170
-#define VCE_RB_SIZE2					0x20174
-#define VCE_RB_RPTR2					0x20178
-#define VCE_RB_WPTR2					0x2017c
-#define VCE_RB_BASE_LO					0x20180
-#define VCE_RB_BASE_HI					0x20184
-#define VCE_RB_SIZE					0x20188
-#define VCE_RB_RPTR					0x2018c
-#define VCE_RB_WPTR					0x20190
-#define VCE_CLOCK_GATING_A				0x202f8
-#define VCE_CLOCK_GATING_B				0x202fc
-#define VCE_UENC_CLOCK_GATING				0x205bc
-#define VCE_UENC_REG_CLOCK_GATING			0x205c0
-#define VCE_FW_REG_STATUS				0x20e10
-#	define VCE_FW_REG_STATUS_BUSY			(1 << 0)
-#	define VCE_FW_REG_STATUS_PASS			(1 << 3)
-#	define VCE_FW_REG_STATUS_DONE			(1 << 11)
-#define VCE_LMI_FW_START_KEYSEL				0x20e18
-#define VCE_LMI_FW_PERIODIC_CTRL			0x20e20
-#define VCE_LMI_CTRL2					0x20e74
-#define VCE_LMI_CTRL					0x20e98
-#define VCE_LMI_VM_CTRL					0x20ea0
-#define VCE_LMI_SWAP_CNTL				0x20eb4
-#define VCE_LMI_SWAP_CNTL1				0x20eb8
-#define VCE_LMI_CACHE_CTRL				0x20ef4
-
 #define VCE_CMD_NO_OP					0x00000000
 #define VCE_CMD_END					0x00000001
 #define VCE_CMD_IB					0x00000002
@@ -629,7 +590,6 @@
 #define VCE_CMD_IB_AUTO					0x00000005
 #define VCE_CMD_SEMAPHORE				0x00000006
 
-
 //#dce stupp
 /* display controller offsets used for crtc/cur/lut/grph/viewport/etc. */
 #define CRTC0_REGISTER_OFFSET                 (0x1b7c - 0x1b7c) //(0x6df0 - 0x6df0)/4
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 9785fada4fa7..42f5d9c0e3af 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -853,10 +853,6 @@ static bool soc15_need_reset_on_init(struct amdgpu_device *adev)
 {
 	u32 sol_reg;
 
-	/* CP hangs in IGT reloading test on RN, reset to WA */
-	if (adev->asic_type == CHIP_RENOIR)
-		return true;
-
 	if (amdgpu_gmc_need_reset_on_init(adev))
 		return true;
 	if (amdgpu_psp_tos_reload_needed(adev))
diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
index 7d17ae56f901..ee8038df17e3 100644
--- a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
@@ -159,6 +159,9 @@ static int tonga_ih_irq_init(struct amdgpu_device *adev)
 	/* enable interrupts */
 	tonga_ih_enable_interrupts(adev);
 
+	if (adev->irq.ih_soft.ring_size)
+		adev->irq.ih_soft.enabled = true;
+
 	return 0;
 }
 
@@ -196,6 +199,9 @@ static u32 tonga_ih_get_wptr(struct amdgpu_device *adev,
 
 	wptr = le32_to_cpu(*ih->wptr_cpu);
 
+	if (ih == &adev->irq.ih_soft)
+		goto out;
+
 	if (!REG_GET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW))
 		goto out;
 
@@ -306,6 +312,10 @@ static int tonga_ih_sw_init(struct amdgpu_ip_block *ip_block)
 	if (r)
 		return r;
 
+	r = amdgpu_ih_ring_init(adev, &adev->irq.ih_soft, IH_SW_RING_SIZE, true);
+	if (r)
+		return r;
+
 	adev->irq.ih.use_doorbell = true;
 	adev->irq.ih.doorbell_index = adev->doorbell_index.ih;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c
index 8dc32787d625..0f5b1719fda5 100644
--- a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c
@@ -711,6 +711,19 @@ static uint32_t umc_v12_0_get_die_id(struct amdgpu_device *adev,
 	return die;
 }
 
+static void umc_v12_0_mca_ipid_parse(struct amdgpu_device *adev, uint64_t ipid,
+		uint32_t *did, uint32_t *ch, uint32_t *umc_inst, uint32_t *sid)
+{
+	if (did)
+		*did = MCA_IPID_2_DIE_ID(ipid);
+	if (ch)
+		*ch = MCA_IPID_2_UMC_CH(ipid);
+	if (umc_inst)
+		*umc_inst = MCA_IPID_2_UMC_INST(ipid);
+	if (sid)
+		*sid = MCA_IPID_2_SOCKET_ID(ipid);
+}
+
 struct amdgpu_umc_ras umc_v12_0_ras = {
 	.ras_block = {
 		.hw_ops = &umc_v12_0_ras_hw_ops,
@@ -724,5 +737,6 @@ struct amdgpu_umc_ras umc_v12_0_ras = {
 	.convert_ras_err_addr = umc_v12_0_convert_error_address,
 	.get_die_id_from_pa = umc_v12_0_get_die_id,
 	.get_retire_flip_bits = umc_v12_0_get_retire_flip_bits,
+	.mca_ipid_parse = umc_v12_0_mca_ipid_parse,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v1_0.c
new file mode 100644
index 000000000000..9ae424618556
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v1_0.c
@@ -0,0 +1,839 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * Copyright 2025 Valve Corporation
+ * Copyright 2025 Alexandre Demers
+ * All Rights Reserved.
+ *
+ * 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, sub license, 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 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 NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS 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.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ *          Timur Kristóf <timur.kristof@gmail.com>
+ *          Alexandre Demers <alexandre.f.demers@gmail.com>
+ */
+
+#include <linux/firmware.h>
+
+#include "amdgpu.h"
+#include "amdgpu_vce.h"
+#include "amdgpu_gart.h"
+#include "sid.h"
+#include "vce_v1_0.h"
+#include "vce/vce_1_0_d.h"
+#include "vce/vce_1_0_sh_mask.h"
+#include "oss/oss_1_0_d.h"
+#include "oss/oss_1_0_sh_mask.h"
+
+#define VCE_V1_0_FW_SIZE	(256 * 1024)
+#define VCE_V1_0_STACK_SIZE	(64 * 1024)
+#define VCE_V1_0_DATA_SIZE	(7808 * (AMDGPU_MAX_VCE_HANDLES + 1))
+#define VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK	0x02
+
+#define VCE_V1_0_GART_PAGE_START \
+	(AMDGPU_GTT_MAX_TRANSFER_SIZE * AMDGPU_GTT_NUM_TRANSFER_WINDOWS)
+#define VCE_V1_0_GART_ADDR_START \
+	(VCE_V1_0_GART_PAGE_START * AMDGPU_GPU_PAGE_SIZE)
+
+static void vce_v1_0_set_ring_funcs(struct amdgpu_device *adev);
+static void vce_v1_0_set_irq_funcs(struct amdgpu_device *adev);
+
+struct vce_v1_0_fw_signature {
+	int32_t offset;
+	uint32_t length;
+	int32_t number;
+	struct {
+		uint32_t chip_id;
+		uint32_t keyselect;
+		uint32_t nonce[4];
+		uint32_t sigval[4];
+	} val[8];
+};
+
+/**
+ * vce_v1_0_ring_get_rptr - get read pointer
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Returns the current hardware read pointer
+ */
+static uint64_t vce_v1_0_ring_get_rptr(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	if (ring->me == 0)
+		return RREG32(mmVCE_RB_RPTR);
+	else
+		return RREG32(mmVCE_RB_RPTR2);
+}
+
+/**
+ * vce_v1_0_ring_get_wptr - get write pointer
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Returns the current hardware write pointer
+ */
+static uint64_t vce_v1_0_ring_get_wptr(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	if (ring->me == 0)
+		return RREG32(mmVCE_RB_WPTR);
+	else
+		return RREG32(mmVCE_RB_WPTR2);
+}
+
+/**
+ * vce_v1_0_ring_set_wptr - set write pointer
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Commits the write pointer to the hardware
+ */
+static void vce_v1_0_ring_set_wptr(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	if (ring->me == 0)
+		WREG32(mmVCE_RB_WPTR, lower_32_bits(ring->wptr));
+	else
+		WREG32(mmVCE_RB_WPTR2, lower_32_bits(ring->wptr));
+}
+
+static int vce_v1_0_lmi_clean(struct amdgpu_device *adev)
+{
+	int i, j;
+
+	for (i = 0; i < 10; ++i) {
+		for (j = 0; j < 100; ++j) {
+			if (RREG32(mmVCE_LMI_STATUS) & 0x337f)
+				return 0;
+
+			mdelay(10);
+		}
+	}
+
+	return -ETIMEDOUT;
+}
+
+static int vce_v1_0_firmware_loaded(struct amdgpu_device *adev)
+{
+	int i, j;
+
+	for (i = 0; i < 10; ++i) {
+		for (j = 0; j < 100; ++j) {
+			if (RREG32(mmVCE_STATUS) & VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK)
+				return 0;
+			mdelay(10);
+		}
+
+		dev_err(adev->dev, "VCE not responding, trying to reset the ECPU\n");
+
+		WREG32_P(mmVCE_SOFT_RESET,
+			VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
+			~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+		mdelay(10);
+		WREG32_P(mmVCE_SOFT_RESET, 0,
+			~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+		mdelay(10);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static void vce_v1_0_init_cg(struct amdgpu_device *adev)
+{
+	u32 tmp;
+
+	tmp = RREG32(mmVCE_CLOCK_GATING_A);
+	tmp |= VCE_CLOCK_GATING_A__CGC_DYN_CLOCK_MODE_MASK;
+	WREG32(mmVCE_CLOCK_GATING_A, tmp);
+
+	tmp = RREG32(mmVCE_CLOCK_GATING_B);
+	tmp |= 0x1e;
+	tmp &= ~0xe100e1;
+	WREG32(mmVCE_CLOCK_GATING_B, tmp);
+
+	tmp = RREG32(mmVCE_UENC_CLOCK_GATING);
+	tmp &= ~0xff9ff000;
+	WREG32(mmVCE_UENC_CLOCK_GATING, tmp);
+
+	tmp = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
+	tmp &= ~0x3ff;
+	WREG32(mmVCE_UENC_REG_CLOCK_GATING, tmp);
+}
+
+/**
+ * vce_v1_0_load_fw_signature - load firmware signature into VCPU BO
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * The VCE1 firmware validation mechanism needs a firmware signature.
+ * This function finds the signature appropriate for the current
+ * ASIC and writes that into the VCPU BO.
+ */
+static int vce_v1_0_load_fw_signature(struct amdgpu_device *adev)
+{
+	const struct common_firmware_header *hdr;
+	struct vce_v1_0_fw_signature *sign;
+	unsigned int ucode_offset;
+	uint32_t chip_id;
+	u32 *cpu_addr;
+	int i;
+
+	hdr = (const struct common_firmware_header *)adev->vce.fw->data;
+	ucode_offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
+	cpu_addr = adev->vce.cpu_addr;
+
+	sign = (void *)adev->vce.fw->data + ucode_offset;
+
+	switch (adev->asic_type) {
+	case CHIP_TAHITI:
+		chip_id = 0x01000014;
+		break;
+	case CHIP_VERDE:
+		chip_id = 0x01000015;
+		break;
+	case CHIP_PITCAIRN:
+		chip_id = 0x01000016;
+		break;
+	default:
+		dev_err(adev->dev, "asic_type %#010x was not found!", adev->asic_type);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < le32_to_cpu(sign->number); ++i) {
+		if (le32_to_cpu(sign->val[i].chip_id) == chip_id)
+			break;
+	}
+
+	if (i == le32_to_cpu(sign->number)) {
+		dev_err(adev->dev, "chip_id 0x%x for %s was not found in VCE firmware",
+			chip_id, amdgpu_asic_name[adev->asic_type]);
+		return -EINVAL;
+	}
+
+	cpu_addr += (256 - 64) / 4;
+	memcpy_toio(&cpu_addr[0], &sign->val[i].nonce[0], 16);
+	cpu_addr[4] = cpu_to_le32(le32_to_cpu(sign->length) + 64);
+
+	memset_io(&cpu_addr[5], 0, 44);
+	memcpy_toio(&cpu_addr[16], &sign[1], hdr->ucode_size_bytes - sizeof(*sign));
+
+	cpu_addr += (le32_to_cpu(sign->length) + 64) / 4;
+	memcpy_toio(&cpu_addr[0], &sign->val[i].sigval[0], 16);
+
+	adev->vce.keyselect = le32_to_cpu(sign->val[i].keyselect);
+
+	return 0;
+}
+
+static int vce_v1_0_wait_for_fw_validation(struct amdgpu_device *adev)
+{
+	int i;
+
+	dev_dbg(adev->dev, "VCE keyselect: %d", adev->vce.keyselect);
+	WREG32(mmVCE_LMI_FW_START_KEYSEL, adev->vce.keyselect);
+
+	for (i = 0; i < 10; ++i) {
+		mdelay(10);
+		if (RREG32(mmVCE_FW_REG_STATUS) & VCE_FW_REG_STATUS__DONE_MASK)
+			break;
+	}
+
+	if (!(RREG32(mmVCE_FW_REG_STATUS) & VCE_FW_REG_STATUS__DONE_MASK)) {
+		dev_err(adev->dev, "VCE FW validation timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	if (!(RREG32(mmVCE_FW_REG_STATUS) & VCE_FW_REG_STATUS__PASS_MASK)) {
+		dev_err(adev->dev, "VCE FW validation failed\n");
+		return -EINVAL;
+	}
+
+	for (i = 0; i < 10; ++i) {
+		mdelay(10);
+		if (!(RREG32(mmVCE_FW_REG_STATUS) & VCE_FW_REG_STATUS__BUSY_MASK))
+			break;
+	}
+
+	if (RREG32(mmVCE_FW_REG_STATUS) & VCE_FW_REG_STATUS__BUSY_MASK) {
+		dev_err(adev->dev, "VCE FW busy timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static int vce_v1_0_mc_resume(struct amdgpu_device *adev)
+{
+	uint32_t offset;
+	uint32_t size;
+
+	/*
+	 * When the keyselect is already set, don't perturb VCE FW.
+	 * Validation seems to always fail the second time.
+	 */
+	if (RREG32(mmVCE_LMI_FW_START_KEYSEL)) {
+		dev_dbg(adev->dev, "keyselect already set: 0x%x (on CPU: 0x%x)\n",
+			RREG32(mmVCE_LMI_FW_START_KEYSEL), adev->vce.keyselect);
+
+		WREG32_P(mmVCE_LMI_CTRL2, 0x0, ~0x100);
+		return 0;
+	}
+
+	WREG32_P(mmVCE_CLOCK_GATING_A, 0, ~(1 << 16));
+	WREG32_P(mmVCE_UENC_CLOCK_GATING, 0x1FF000, ~0xFF9FF000);
+	WREG32_P(mmVCE_UENC_REG_CLOCK_GATING, 0x3F, ~0x3F);
+	WREG32(mmVCE_CLOCK_GATING_B, 0);
+
+	WREG32_P(mmVCE_LMI_FW_PERIODIC_CTRL, 0x4, ~0x4);
+
+	WREG32(mmVCE_LMI_CTRL, 0x00398000);
+
+	WREG32_P(mmVCE_LMI_CACHE_CTRL, 0x0, ~0x1);
+	WREG32(mmVCE_LMI_SWAP_CNTL, 0);
+	WREG32(mmVCE_LMI_SWAP_CNTL1, 0);
+	WREG32(mmVCE_LMI_VM_CTRL, 0);
+
+	WREG32(mmVCE_VCPU_SCRATCH7, AMDGPU_MAX_VCE_HANDLES);
+
+	offset =  adev->vce.gpu_addr + AMDGPU_VCE_FIRMWARE_OFFSET;
+	size = VCE_V1_0_FW_SIZE;
+	WREG32(mmVCE_VCPU_CACHE_OFFSET0, offset & 0x7fffffff);
+	WREG32(mmVCE_VCPU_CACHE_SIZE0, size);
+
+	offset += size;
+	size = VCE_V1_0_STACK_SIZE;
+	WREG32(mmVCE_VCPU_CACHE_OFFSET1, offset & 0x7fffffff);
+	WREG32(mmVCE_VCPU_CACHE_SIZE1, size);
+
+	offset += size;
+	size = VCE_V1_0_DATA_SIZE;
+	WREG32(mmVCE_VCPU_CACHE_OFFSET2, offset & 0x7fffffff);
+	WREG32(mmVCE_VCPU_CACHE_SIZE2, size);
+
+	WREG32_P(mmVCE_LMI_CTRL2, 0x0, ~0x100);
+
+	return vce_v1_0_wait_for_fw_validation(adev);
+}
+
+/**
+ * vce_v1_0_is_idle() - Check idle status of VCE1 IP block
+ *
+ * @ip_block: amdgpu_ip_block pointer
+ *
+ * Check whether VCE is busy according to VCE_STATUS.
+ * Also check whether the SRBM thinks VCE is busy, although
+ * SRBM_STATUS.VCE_BUSY seems to be bogus because it
+ * appears to mirror the VCE_STATUS.VCPU_REPORT_FW_LOADED bit.
+ */
+static bool vce_v1_0_is_idle(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	bool busy =
+		(RREG32(mmVCE_STATUS) & (VCE_STATUS__JOB_BUSY_MASK | VCE_STATUS__UENC_BUSY_MASK)) ||
+		(RREG32(mmSRBM_STATUS2) & SRBM_STATUS2__VCE_BUSY_MASK);
+
+	return !busy;
+}
+
+static int vce_v1_0_wait_for_idle(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	unsigned int i;
+
+	for (i = 0; i < adev->usec_timeout; i++) {
+		udelay(1);
+		if (vce_v1_0_is_idle(ip_block))
+			return 0;
+	}
+	return -ETIMEDOUT;
+}
+
+/**
+ * vce_v1_0_start - start VCE block
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Setup and start the VCE block
+ */
+static int vce_v1_0_start(struct amdgpu_device *adev)
+{
+	struct amdgpu_ring *ring;
+	int r;
+
+	WREG32_P(mmVCE_STATUS, 1, ~1);
+
+	r = vce_v1_0_mc_resume(adev);
+	if (r)
+		return r;
+
+	ring = &adev->vce.ring[0];
+	WREG32(mmVCE_RB_RPTR, lower_32_bits(ring->wptr));
+	WREG32(mmVCE_RB_WPTR, lower_32_bits(ring->wptr));
+	WREG32(mmVCE_RB_BASE_LO, lower_32_bits(ring->gpu_addr));
+	WREG32(mmVCE_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+	WREG32(mmVCE_RB_SIZE, ring->ring_size / 4);
+
+	ring = &adev->vce.ring[1];
+	WREG32(mmVCE_RB_RPTR2, lower_32_bits(ring->wptr));
+	WREG32(mmVCE_RB_WPTR2, lower_32_bits(ring->wptr));
+	WREG32(mmVCE_RB_BASE_LO2, lower_32_bits(ring->gpu_addr));
+	WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+	WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4);
+
+	WREG32_P(mmVCE_VCPU_CNTL, VCE_VCPU_CNTL__CLK_EN_MASK,
+		 ~VCE_VCPU_CNTL__CLK_EN_MASK);
+
+	WREG32_P(mmVCE_SOFT_RESET,
+		VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK |
+		VCE_SOFT_RESET__FME_SOFT_RESET_MASK,
+		~(VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK |
+		  VCE_SOFT_RESET__FME_SOFT_RESET_MASK));
+
+	mdelay(100);
+
+	WREG32_P(mmVCE_SOFT_RESET, 0,
+		~(VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK |
+		  VCE_SOFT_RESET__FME_SOFT_RESET_MASK));
+
+	r = vce_v1_0_firmware_loaded(adev);
+
+	/* Clear VCE_STATUS, otherwise SRBM thinks VCE1 is busy. */
+	WREG32(mmVCE_STATUS, 0);
+
+	if (r) {
+		dev_err(adev->dev, "VCE not responding, giving up\n");
+		return r;
+	}
+
+	return 0;
+}
+
+static int vce_v1_0_stop(struct amdgpu_device *adev)
+{
+	struct amdgpu_ip_block *ip_block;
+	int status;
+	int i;
+
+	ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_VCE);
+	if (!ip_block)
+		return -EINVAL;
+
+	if (vce_v1_0_lmi_clean(adev))
+		dev_warn(adev->dev, "VCE not idle\n");
+
+	if (vce_v1_0_wait_for_idle(ip_block))
+		dev_warn(adev->dev, "VCE busy: VCE_STATUS=0x%x, SRBM_STATUS2=0x%x\n",
+			RREG32(mmVCE_STATUS), RREG32(mmSRBM_STATUS2));
+
+	/* Stall UMC and register bus before resetting VCPU */
+	WREG32_P(mmVCE_LMI_CTRL2, 1 << 8, ~(1 << 8));
+
+	for (i = 0; i < 100; ++i) {
+		status = RREG32(mmVCE_LMI_STATUS);
+		if (status & 0x240)
+			break;
+		mdelay(1);
+	}
+
+	WREG32_P(mmVCE_VCPU_CNTL, 0, ~VCE_VCPU_CNTL__CLK_EN_MASK);
+
+	WREG32_P(mmVCE_SOFT_RESET,
+		VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK |
+		VCE_SOFT_RESET__FME_SOFT_RESET_MASK,
+		~(VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK |
+		  VCE_SOFT_RESET__FME_SOFT_RESET_MASK));
+
+	WREG32(mmVCE_STATUS, 0);
+
+	return 0;
+}
+
+static void vce_v1_0_enable_mgcg(struct amdgpu_device *adev, bool enable)
+{
+	u32 tmp;
+
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)) {
+		tmp = RREG32(mmVCE_CLOCK_GATING_A);
+		tmp |= VCE_CLOCK_GATING_A__CGC_DYN_CLOCK_MODE_MASK;
+		WREG32(mmVCE_CLOCK_GATING_A, tmp);
+
+		tmp = RREG32(mmVCE_UENC_CLOCK_GATING);
+		tmp &= ~0x1ff000;
+		tmp |= 0xff800000;
+		WREG32(mmVCE_UENC_CLOCK_GATING, tmp);
+
+		tmp = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
+		tmp &= ~0x3ff;
+		WREG32(mmVCE_UENC_REG_CLOCK_GATING, tmp);
+	} else {
+		tmp = RREG32(mmVCE_CLOCK_GATING_A);
+		tmp &= ~VCE_CLOCK_GATING_A__CGC_DYN_CLOCK_MODE_MASK;
+		WREG32(mmVCE_CLOCK_GATING_A, tmp);
+
+		tmp = RREG32(mmVCE_UENC_CLOCK_GATING);
+		tmp |= 0x1ff000;
+		tmp &= ~0xff800000;
+		WREG32(mmVCE_UENC_CLOCK_GATING, tmp);
+
+		tmp = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
+		tmp |= 0x3ff;
+		WREG32(mmVCE_UENC_REG_CLOCK_GATING, tmp);
+	}
+}
+
+static int vce_v1_0_early_init(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	int r;
+
+	r = amdgpu_vce_early_init(adev);
+	if (r)
+		return r;
+
+	adev->vce.num_rings = 2;
+
+	vce_v1_0_set_ring_funcs(adev);
+	vce_v1_0_set_irq_funcs(adev);
+
+	return 0;
+}
+
+/**
+ * vce_v1_0_ensure_vcpu_bo_32bit_addr() - ensure the VCPU BO has a 32-bit address
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Due to various hardware limitations, the VCE1 requires
+ * the VCPU BO to be in the low 32 bit address range.
+ * Ensure that the VCPU BO has a 32-bit GPU address,
+ * or return an error code when that isn't possible.
+ *
+ * To accomodate that, we put GART to the LOW address range
+ * and reserve some GART pages where we map the VCPU BO,
+ * so that it gets a 32-bit address.
+ */
+static int vce_v1_0_ensure_vcpu_bo_32bit_addr(struct amdgpu_device *adev)
+{
+	u64 gpu_addr = amdgpu_bo_gpu_offset(adev->vce.vcpu_bo);
+	u64 bo_size = amdgpu_bo_size(adev->vce.vcpu_bo);
+	u64 max_vcpu_bo_addr = 0xffffffff - bo_size;
+	u64 num_pages = ALIGN(bo_size, AMDGPU_GPU_PAGE_SIZE) / AMDGPU_GPU_PAGE_SIZE;
+	u64 pa = amdgpu_gmc_vram_pa(adev, adev->vce.vcpu_bo);
+	u64 flags = AMDGPU_PTE_READABLE | AMDGPU_PTE_WRITEABLE | AMDGPU_PTE_VALID;
+
+	/*
+	 * Check if the VCPU BO already has a 32-bit address.
+	 * Eg. if MC is configured to put VRAM in the low address range.
+	 */
+	if (gpu_addr <= max_vcpu_bo_addr)
+		return 0;
+
+	/* Check if we can map the VCPU BO in GART to a 32-bit address. */
+	if (adev->gmc.gart_start + VCE_V1_0_GART_ADDR_START > max_vcpu_bo_addr)
+		return -EINVAL;
+
+	amdgpu_gart_map_vram_range(adev, pa, VCE_V1_0_GART_PAGE_START,
+				   num_pages, flags, adev->gart.ptr);
+	adev->vce.gpu_addr = adev->gmc.gart_start + VCE_V1_0_GART_ADDR_START;
+	if (adev->vce.gpu_addr > max_vcpu_bo_addr)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int vce_v1_0_sw_init(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	struct amdgpu_ring *ring;
+	int r, i;
+
+	r = amdgpu_irq_add_id(adev, AMDGPU_IRQ_CLIENTID_LEGACY, 167, &adev->vce.irq);
+	if (r)
+		return r;
+
+	r = amdgpu_vce_sw_init(adev, VCE_V1_0_FW_SIZE +
+		VCE_V1_0_STACK_SIZE + VCE_V1_0_DATA_SIZE);
+	if (r)
+		return r;
+
+	r = amdgpu_vce_resume(adev);
+	if (r)
+		return r;
+	r = vce_v1_0_load_fw_signature(adev);
+	if (r)
+		return r;
+	r = vce_v1_0_ensure_vcpu_bo_32bit_addr(adev);
+	if (r)
+		return r;
+
+	for (i = 0; i < adev->vce.num_rings; i++) {
+		enum amdgpu_ring_priority_level hw_prio = amdgpu_vce_get_ring_prio(i);
+
+		ring = &adev->vce.ring[i];
+		sprintf(ring->name, "vce%d", i);
+		r = amdgpu_ring_init(adev, ring, 512, &adev->vce.irq, 0,
+				     hw_prio, NULL);
+		if (r)
+			return r;
+	}
+
+	return r;
+}
+
+static int vce_v1_0_sw_fini(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	int r;
+
+	r = amdgpu_vce_suspend(adev);
+	if (r)
+		return r;
+
+	return amdgpu_vce_sw_fini(adev);
+}
+
+/**
+ * vce_v1_0_hw_init - start and test VCE block
+ *
+ * @ip_block: Pointer to the amdgpu_ip_block for this hw instance.
+ *
+ * Initialize the hardware, boot up the VCPU and do some testing
+ */
+static int vce_v1_0_hw_init(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	int i, r;
+
+	if (adev->pm.dpm_enabled)
+		amdgpu_dpm_enable_vce(adev, true);
+	else
+		amdgpu_asic_set_vce_clocks(adev, 10000, 10000);
+
+	for (i = 0; i < adev->vce.num_rings; i++) {
+		r = amdgpu_ring_test_helper(&adev->vce.ring[i]);
+		if (r)
+			return r;
+	}
+
+	dev_info(adev->dev, "VCE initialized successfully.\n");
+
+	return 0;
+}
+
+static int vce_v1_0_hw_fini(struct amdgpu_ip_block *ip_block)
+{
+	int r;
+
+	r = vce_v1_0_stop(ip_block->adev);
+	if (r)
+		return r;
+
+	cancel_delayed_work_sync(&ip_block->adev->vce.idle_work);
+	return 0;
+}
+
+static int vce_v1_0_suspend(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	int r;
+
+	/*
+	 * Proper cleanups before halting the HW engine:
+	 *   - cancel the delayed idle work
+	 *   - enable powergating
+	 *   - enable clockgating
+	 *   - disable dpm
+	 *
+	 * TODO: to align with the VCN implementation, move the
+	 * jobs for clockgating/powergating/dpm setting to
+	 * ->set_powergating_state().
+	 */
+	cancel_delayed_work_sync(&adev->vce.idle_work);
+
+	if (adev->pm.dpm_enabled) {
+		amdgpu_dpm_enable_vce(adev, false);
+	} else {
+		amdgpu_asic_set_vce_clocks(adev, 0, 0);
+		amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+						       AMD_PG_STATE_GATE);
+		amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+						       AMD_CG_STATE_GATE);
+	}
+
+	r = vce_v1_0_hw_fini(ip_block);
+	if (r) {
+		dev_err(adev->dev, "vce_v1_0_hw_fini() failed with error %i", r);
+		return r;
+	}
+
+	return amdgpu_vce_suspend(adev);
+}
+
+static int vce_v1_0_resume(struct amdgpu_ip_block *ip_block)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+	int r;
+
+	r = amdgpu_vce_resume(adev);
+	if (r)
+		return r;
+	r = vce_v1_0_load_fw_signature(adev);
+	if (r)
+		return r;
+	r = vce_v1_0_ensure_vcpu_bo_32bit_addr(adev);
+	if (r)
+		return r;
+
+	return vce_v1_0_hw_init(ip_block);
+}
+
+static int vce_v1_0_set_interrupt_state(struct amdgpu_device *adev,
+					struct amdgpu_irq_src *source,
+					unsigned int type,
+					enum amdgpu_interrupt_state state)
+{
+	uint32_t val = 0;
+
+	if (state == AMDGPU_IRQ_STATE_ENABLE)
+		val |= VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK;
+
+	WREG32_P(mmVCE_SYS_INT_EN, val,
+		 ~VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK);
+	return 0;
+}
+
+static int vce_v1_0_process_interrupt(struct amdgpu_device *adev,
+				      struct amdgpu_irq_src *source,
+				      struct amdgpu_iv_entry *entry)
+{
+	dev_dbg(adev->dev, "IH: VCE\n");
+	switch (entry->src_data[0]) {
+	case 0:
+	case 1:
+		amdgpu_fence_process(&adev->vce.ring[entry->src_data[0]]);
+		break;
+	default:
+		dev_err(adev->dev, "Unhandled interrupt: %d %d\n",
+			  entry->src_id, entry->src_data[0]);
+		break;
+	}
+
+	return 0;
+}
+
+static int vce_v1_0_set_clockgating_state(struct amdgpu_ip_block *ip_block,
+					  enum amd_clockgating_state state)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+
+	vce_v1_0_init_cg(adev);
+	vce_v1_0_enable_mgcg(adev, state == AMD_CG_STATE_GATE);
+
+	return 0;
+}
+
+static int vce_v1_0_set_powergating_state(struct amdgpu_ip_block *ip_block,
+					  enum amd_powergating_state state)
+{
+	struct amdgpu_device *adev = ip_block->adev;
+
+	/*
+	 * This doesn't actually powergate the VCE block.
+	 * That's done in the dpm code via the SMC.  This
+	 * just re-inits the block as necessary.  The actual
+	 * gating still happens in the dpm code.  We should
+	 * revisit this when there is a cleaner line between
+	 * the smc and the hw blocks
+	 */
+	if (state == AMD_PG_STATE_GATE)
+		return vce_v1_0_stop(adev);
+	else
+		return vce_v1_0_start(adev);
+}
+
+static const struct amd_ip_funcs vce_v1_0_ip_funcs = {
+	.name = "vce_v1_0",
+	.early_init = vce_v1_0_early_init,
+	.sw_init = vce_v1_0_sw_init,
+	.sw_fini = vce_v1_0_sw_fini,
+	.hw_init = vce_v1_0_hw_init,
+	.hw_fini = vce_v1_0_hw_fini,
+	.suspend = vce_v1_0_suspend,
+	.resume = vce_v1_0_resume,
+	.is_idle = vce_v1_0_is_idle,
+	.wait_for_idle = vce_v1_0_wait_for_idle,
+	.set_clockgating_state = vce_v1_0_set_clockgating_state,
+	.set_powergating_state = vce_v1_0_set_powergating_state,
+};
+
+static const struct amdgpu_ring_funcs vce_v1_0_ring_funcs = {
+	.type = AMDGPU_RING_TYPE_VCE,
+	.align_mask = 0xf,
+	.nop = VCE_CMD_NO_OP,
+	.support_64bit_ptrs = false,
+	.no_user_fence = true,
+	.get_rptr = vce_v1_0_ring_get_rptr,
+	.get_wptr = vce_v1_0_ring_get_wptr,
+	.set_wptr = vce_v1_0_ring_set_wptr,
+	.parse_cs = amdgpu_vce_ring_parse_cs,
+	.emit_frame_size = 6, /* amdgpu_vce_ring_emit_fence  x1 no user fence */
+	.emit_ib_size = 4, /* amdgpu_vce_ring_emit_ib */
+	.emit_ib = amdgpu_vce_ring_emit_ib,
+	.emit_fence = amdgpu_vce_ring_emit_fence,
+	.test_ring = amdgpu_vce_ring_test_ring,
+	.test_ib = amdgpu_vce_ring_test_ib,
+	.insert_nop = amdgpu_ring_insert_nop,
+	.pad_ib = amdgpu_ring_generic_pad_ib,
+	.begin_use = amdgpu_vce_ring_begin_use,
+	.end_use = amdgpu_vce_ring_end_use,
+};
+
+static void vce_v1_0_set_ring_funcs(struct amdgpu_device *adev)
+{
+	int i;
+
+	for (i = 0; i < adev->vce.num_rings; i++) {
+		adev->vce.ring[i].funcs = &vce_v1_0_ring_funcs;
+		adev->vce.ring[i].me = i;
+	}
+};
+
+static const struct amdgpu_irq_src_funcs vce_v1_0_irq_funcs = {
+	.set = vce_v1_0_set_interrupt_state,
+	.process = vce_v1_0_process_interrupt,
+};
+
+static void vce_v1_0_set_irq_funcs(struct amdgpu_device *adev)
+{
+	adev->vce.irq.num_types = 1;
+	adev->vce.irq.funcs = &vce_v1_0_irq_funcs;
+};
+
+const struct amdgpu_ip_block_version vce_v1_0_ip_block = {
+	.type = AMD_IP_BLOCK_TYPE_VCE,
+	.major = 1,
+	.minor = 0,
+	.rev = 0,
+	.funcs = &vce_v1_0_ip_funcs,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v1_0.h b/drivers/gpu/drm/amd/amdgpu/vce_v1_0.h
new file mode 100644
index 000000000000..206e7bec897f
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v1_0.h
@@ -0,0 +1,32 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright 2025 Advanced Micro Devices, Inc.
+ * Copyright 2025 Valve Corporation
+ * Copyright 2025 Alexandre Demers
+ *
+ * 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.
+ *
+ */
+
+#ifndef __VCE_V1_0_H__
+#define __VCE_V1_0_H__
+
+extern const struct amdgpu_ip_block_version vce_v1_0_ip_block;
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
index bee3e904a6bc..8ea8a6193492 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
@@ -407,6 +407,11 @@ static void vce_v2_0_enable_mgcg(struct amdgpu_device *adev, bool enable,
 static int vce_v2_0_early_init(struct amdgpu_ip_block *ip_block)
 {
 	struct amdgpu_device *adev = ip_block->adev;
+	int r;
+
+	r = amdgpu_vce_early_init(adev);
+	if (r)
+		return r;
 
 	adev->vce.num_rings = 2;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 708123899c41..719e9643c43d 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -399,6 +399,7 @@ static unsigned vce_v3_0_get_harvest_config(struct amdgpu_device *adev)
 static int vce_v3_0_early_init(struct amdgpu_ip_block *ip_block)
 {
 	struct amdgpu_device *adev = ip_block->adev;
+	int r;
 
 	adev->vce.harvest_config = vce_v3_0_get_harvest_config(adev);
 
@@ -407,6 +408,10 @@ static int vce_v3_0_early_init(struct amdgpu_ip_block *ip_block)
 	    (AMDGPU_VCE_HARVEST_VCE0 | AMDGPU_VCE_HARVEST_VCE1))
 		return -ENOENT;
 
+	r = amdgpu_vce_early_init(adev);
+	if (r)
+		return r;
+
 	adev->vce.num_rings = 3;
 
 	vce_v3_0_set_ring_funcs(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
index 335bda64ff5b..2d64002bed61 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
@@ -410,6 +410,11 @@ static int vce_v4_0_stop(struct amdgpu_device *adev)
 static int vce_v4_0_early_init(struct amdgpu_ip_block *ip_block)
 {
 	struct amdgpu_device *adev = ip_block->adev;
+	int r;
+
+	r = amdgpu_vce_early_init(adev);
+	if (r)
+		return r;
 
 	if (amdgpu_sriov_vf(adev)) /* currently only VCN0 support SRIOV */
 		adev->vce.num_rings = 1;
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c
index eacf4e93ba2f..cb7123ec1a5d 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c
@@ -141,7 +141,7 @@ static int vcn_v4_0_3_late_init(struct amdgpu_ip_block *ip_block)
 	adev->vcn.supported_reset =
 		amdgpu_get_soft_full_reset_mask(&adev->vcn.inst[0].ring_enc[0]);
 
-	if (amdgpu_dpm_reset_vcn_is_supported(adev))
+	if (amdgpu_dpm_reset_vcn_is_supported(adev) && !amdgpu_sriov_vf(adev))
 		adev->vcn.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c
index b107ee80e472..1f6a22983c0d 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c
@@ -265,6 +265,8 @@ static int vcn_v4_0_5_sw_fini(struct amdgpu_ip_block *ip_block)
 	if (amdgpu_sriov_vf(adev))
 		amdgpu_virt_free_mm_table(adev);
 
+	amdgpu_vcn_sysfs_reset_mask_fini(adev);
+
 	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
 		r = amdgpu_vcn_suspend(adev, i);
 		if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c
index 714350cabf2f..8bd457dea4cf 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c
@@ -122,7 +122,9 @@ static int vcn_v5_0_1_late_init(struct amdgpu_ip_block *ip_block)
 
 	switch (amdgpu_ip_version(adev, MP0_HWIP, 0)) {
 	case IP_VERSION(13, 0, 12):
-		if ((adev->psp.sos.fw_version >= 0x00450025) && amdgpu_dpm_reset_vcn_is_supported(adev))
+		if ((adev->psp.sos.fw_version >= 0x00450025) &&
+			amdgpu_dpm_reset_vcn_is_supported(adev) &&
+			!amdgpu_sriov_vf(adev))
 			adev->vcn.supported_reset |= AMDGPU_RESET_TYPE_PER_QUEUE;
 		break;
 	default:
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
index 0320163b6e74..f98c735b2905 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
@@ -3644,14 +3644,18 @@ static const uint32_t cwsr_trap_gfx9_4_3_hex[] = {
 };
 
 static const uint32_t cwsr_trap_gfx12_hex[] = {
-	0xbfa00001, 0xbfa002a2,
-	0xb0804009, 0xb8f8f804,
+	0xbfa00001, 0xbfa002b2,
+	0xb0804009, 0xb8eef81a,
+	0xbf880000, 0xb980081a,
+	0x00000000, 0xb8f8f804,
+	0x9177ff77, 0x0c000000,
+	0x846e9a6e, 0x8c776e77,
 	0x9178ff78, 0x00008c00,
 	0xb8fbf811, 0x8b6eff78,
 	0x00004000, 0xbfa10008,
 	0x8b6eff7b, 0x00000080,
 	0xbfa20018, 0x8b6ea07b,
-	0xbfa20042, 0xbf830010,
+	0xbfa2004a, 0xbf830010,
 	0xb8fbf811, 0xbfa0fffb,
 	0x8b6eff7b, 0x00000bd0,
 	0xbfa20010, 0xb8eef812,
@@ -3662,28 +3666,32 @@ static const uint32_t cwsr_trap_gfx12_hex[] = {
 	0xf0000000, 0xbfa20005,
 	0x8b6fff6f, 0x00000200,
 	0xbfa20002, 0x8b6ea07b,
-	0xbfa2002c, 0xbefa4d82,
+	0xbfa20034, 0xbefa4d82,
 	0xbf8a0000, 0x84fa887a,
 	0xbf0d8f7b, 0xbfa10002,
 	0x8c7bff7b, 0xffff0000,
-	0xf4601bbd, 0xf8000010,
-	0xbf8a0000, 0x846e976e,
-	0x9177ff77, 0x00800000,
-	0x8c776e77, 0xf4603bbd,
-	0xf8000000, 0xbf8a0000,
-	0xf4603ebd, 0xf8000008,
-	0xbf8a0000, 0x8bee6e6e,
-	0xbfa10001, 0xbe80486e,
-	0x8b6eff6d, 0xf0000000,
-	0xbfa20009, 0xb8eef811,
-	0x8b6eff6e, 0x00000080,
-	0xbfa20007, 0x8c78ff78,
-	0x00004000, 0x80ec886c,
-	0x82ed806d, 0xbfa00002,
-	0x806c846c, 0x826d806d,
-	0x8b6dff6d, 0x0000ffff,
-	0x8bfe7e7e, 0x8bea6a6a,
-	0x85788978, 0xb9783244,
+	0x8b6eff77, 0x0c000000,
+	0x916dff6d, 0x0c000000,
+	0x8c6d6e6d, 0xf4601bbd,
+	0xf8000010, 0xbf8a0000,
+	0x846e976e, 0x9177ff77,
+	0x00800000, 0x8c776e77,
+	0xf4603bbd, 0xf8000000,
+	0xbf8a0000, 0xf4603ebd,
+	0xf8000008, 0xbf8a0000,
+	0x8bee6e6e, 0xbfa10001,
+	0xbe80486e, 0x8b6eff6d,
+	0xf0000000, 0xbfa20009,
+	0xb8eef811, 0x8b6eff6e,
+	0x00000080, 0xbfa20007,
+	0x8c78ff78, 0x00004000,
+	0x80ec886c, 0x82ed806d,
+	0xbfa00002, 0x806c846c,
+	0x826d806d, 0x8b6dff6d,
+	0x0000ffff, 0x8bfe7e7e,
+	0x8bea6a6a, 0x85788978,
+	0x936eff77, 0x0002001a,
+	0xb96ef81a, 0xb9783244,
 	0xbe804a6c, 0xb8faf802,
 	0xbf0d987a, 0xbfa10001,
 	0xbfb00000, 0x8b6dff6d,
@@ -3981,7 +3989,7 @@ static const uint32_t cwsr_trap_gfx12_hex[] = {
 	0x008ce800, 0x00000000,
 	0x807d817d, 0x8070ff70,
 	0x00000080, 0xbf0a7b7d,
-	0xbfa2fff7, 0xbfa0016e,
+	0xbfa2fff7, 0xbfa00171,
 	0xbef4007e, 0x8b75ff7f,
 	0x0000ffff, 0x8c75ff75,
 	0x00040000, 0xbef60080,
@@ -4163,12 +4171,14 @@ static const uint32_t cwsr_trap_gfx12_hex[] = {
 	0xf8000074, 0xbf8a0000,
 	0x8b6dff6d, 0x0000ffff,
 	0x8bfe7e7e, 0x8bea6a6a,
-	0xb97af804, 0xbe804ec2,
-	0xbf94fffe, 0xbe804a6c,
+	0x936eff77, 0x0002001a,
+	0xb96ef81a, 0xb97af804,
 	0xbe804ec2, 0xbf94fffe,
-	0xbfb10000, 0xbf9f0000,
+	0xbe804a6c, 0xbe804ec2,
+	0xbf94fffe, 0xbfb10000,
 	0xbf9f0000, 0xbf9f0000,
 	0xbf9f0000, 0xbf9f0000,
+	0xbf9f0000, 0x00000000,
 };
 
 static const uint32_t cwsr_trap_gfx9_5_0_hex[] = {
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx12.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx12.asm
index 5a1a1b1f897f..07999b4649de 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx12.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx12.asm
@@ -78,9 +78,16 @@ var SQ_WAVE_EXCP_FLAG_PRIV_RESTORE_PART_2_SHIFT	= SQ_WAVE_EXCP_FLAG_PRIV_ILLEGAL
 var SQ_WAVE_EXCP_FLAG_PRIV_RESTORE_PART_2_SIZE	= SQ_WAVE_EXCP_FLAG_PRIV_HOST_TRAP_SHIFT - SQ_WAVE_EXCP_FLAG_PRIV_ILLEGAL_INST_SHIFT
 var SQ_WAVE_EXCP_FLAG_PRIV_RESTORE_PART_3_SHIFT	= SQ_WAVE_EXCP_FLAG_PRIV_WAVE_START_SHIFT
 var SQ_WAVE_EXCP_FLAG_PRIV_RESTORE_PART_3_SIZE	= 32 - SQ_WAVE_EXCP_FLAG_PRIV_RESTORE_PART_3_SHIFT
+
+var SQ_WAVE_SCHED_MODE_DEP_MODE_SHIFT		= 0
+var SQ_WAVE_SCHED_MODE_DEP_MODE_SIZE		= 2
+
 var BARRIER_STATE_SIGNAL_OFFSET			= 16
 var BARRIER_STATE_VALID_OFFSET			= 0
 
+var TTMP11_SCHED_MODE_SHIFT			= 26
+var TTMP11_SCHED_MODE_SIZE			= 2
+var TTMP11_SCHED_MODE_MASK			= 0xC000000
 var TTMP11_DEBUG_TRAP_ENABLED_SHIFT		= 23
 var TTMP11_DEBUG_TRAP_ENABLED_MASK		= 0x800000
 
@@ -160,8 +167,19 @@ L_JUMP_TO_RESTORE:
 	s_branch	L_RESTORE
 
 L_SKIP_RESTORE:
+	// Assume most relaxed scheduling mode is set. Save and revert to normal mode.
+	s_getreg_b32	ttmp2, hwreg(HW_REG_WAVE_SCHED_MODE)
+	s_wait_alu	0
+	s_setreg_imm32_b32	hwreg(HW_REG_WAVE_SCHED_MODE, \
+		SQ_WAVE_SCHED_MODE_DEP_MODE_SHIFT, SQ_WAVE_SCHED_MODE_DEP_MODE_SIZE), 0
+
 	s_getreg_b32	s_save_state_priv, hwreg(HW_REG_WAVE_STATE_PRIV)	//save STATUS since we will change SCC
 
+	// Save SCHED_MODE[1:0] into ttmp11[27:26].
+	s_andn2_b32	ttmp11, ttmp11, TTMP11_SCHED_MODE_MASK
+	s_lshl_b32	ttmp2, ttmp2, TTMP11_SCHED_MODE_SHIFT
+	s_or_b32	ttmp11, ttmp11, ttmp2
+
 	// Clear SPI_PRIO: do not save with elevated priority.
 	// Clear ECC_ERR: prevents SQC store and triggers FATAL_HALT if setreg'd.
 	s_andn2_b32	s_save_state_priv, s_save_state_priv, SQ_WAVE_STATE_PRIV_ALWAYS_CLEAR_MASK
@@ -238,6 +256,13 @@ L_FETCH_2ND_TRAP:
 	s_cbranch_scc0	L_NO_SIGN_EXTEND_TMA
 	s_or_b32	ttmp15, ttmp15, 0xFFFF0000
 L_NO_SIGN_EXTEND_TMA:
+#if ASIC_FAMILY == CHIP_GFX12
+	// Move SCHED_MODE[1:0] from ttmp11 to unused bits in ttmp1[27:26] (return PC_HI).
+	// The second-level trap will restore from ttmp1 for backwards compatibility.
+	s_and_b32	ttmp2, ttmp11, TTMP11_SCHED_MODE_MASK
+	s_andn2_b32	ttmp1, ttmp1, TTMP11_SCHED_MODE_MASK
+	s_or_b32	ttmp1, ttmp1, ttmp2
+#endif
 
 	s_load_dword    ttmp2, [ttmp14, ttmp15], 0x10 scope:SCOPE_SYS		// debug trap enabled flag
 	s_wait_idle
@@ -287,6 +312,10 @@ L_EXIT_TRAP:
 	// STATE_PRIV.BARRIER_COMPLETE may have changed since we read it.
 	// Only restore fields which the trap handler changes.
 	s_lshr_b32	s_save_state_priv, s_save_state_priv, SQ_WAVE_STATE_PRIV_SCC_SHIFT
+
+	// Assume relaxed scheduling mode after this point.
+	restore_sched_mode(ttmp2)
+
 	s_setreg_b32	hwreg(HW_REG_WAVE_STATE_PRIV, SQ_WAVE_STATE_PRIV_SCC_SHIFT, \
 		SQ_WAVE_STATE_PRIV_POISON_ERR_SHIFT - SQ_WAVE_STATE_PRIV_SCC_SHIFT + 1), s_save_state_priv
 
@@ -1043,6 +1072,9 @@ L_SKIP_BARRIER_RESTORE:
 	s_and_b64	exec, exec, exec					// Restore STATUS.EXECZ, not writable by s_setreg_b32
 	s_and_b64	vcc, vcc, vcc						// Restore STATUS.VCCZ, not writable by s_setreg_b32
 
+	// Assume relaxed scheduling mode after this point.
+	restore_sched_mode(s_restore_tmp)
+
 	s_setreg_b32	hwreg(HW_REG_WAVE_STATE_PRIV), s_restore_state_priv	// SCC is included, which is changed by previous salu
 
 	// Make barrier and LDS state visible to all waves in the group.
@@ -1134,3 +1166,8 @@ function valu_sgpr_hazard
 	end
 #endif
 end
+
+function restore_sched_mode(s_tmp)
+	s_bfe_u32	s_tmp, ttmp11, (TTMP11_SCHED_MODE_SHIFT | (TTMP11_SCHED_MODE_SIZE << 0x10))
+	s_setreg_b32	hwreg(HW_REG_WAVE_SCHED_MODE), s_tmp
+end
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 0f0719528bcc..22925df6a791 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -2826,7 +2826,7 @@ retry:
 
 static int runtime_disable(struct kfd_process *p)
 {
-	int i = 0, ret;
+	int i = 0, ret = 0;
 	bool was_enabled = p->runtime_info.runtime_state == DEBUG_RUNTIME_STATE_ENABLED;
 
 	p->runtime_info.runtime_state = DEBUG_RUNTIME_STATE_DISABLED;
@@ -2863,6 +2863,7 @@ static int runtime_disable(struct kfd_process *p)
 	/* disable ttmp setup */
 	for (i = 0; i < p->n_pdds; i++) {
 		struct kfd_process_device *pdd = p->pdds[i];
+		int last_err = 0;
 
 		if (kfd_dbg_is_per_vmid_supported(pdd->dev)) {
 			pdd->spi_dbg_override =
@@ -2872,14 +2873,17 @@ static int runtime_disable(struct kfd_process *p)
 					pdd->dev->vm_info.last_vmid_kfd);
 
 			if (!pdd->dev->kfd->shared_resources.enable_mes)
-				debug_refresh_runlist(pdd->dev->dqm);
+				last_err = debug_refresh_runlist(pdd->dev->dqm);
 			else
-				kfd_dbg_set_mes_debug_mode(pdd,
+				last_err = kfd_dbg_set_mes_debug_mode(pdd,
 							   !kfd_dbg_has_cwsr_workaround(pdd->dev));
+
+			if (last_err)
+				ret = last_err;
 		}
 	}
 
-	return 0;
+	return ret;
 }
 
 static int kfd_ioctl_runtime_enable(struct file *filep, struct kfd_process *p, void *data)
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 4fbe865ff279..d7a2e7178ea9 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -2094,7 +2094,8 @@ int amdkfd_fence_wait_timeout(struct device_queue_manager *dqm,
 
 	while (*fence_addr != fence_value) {
 		/* Fatal err detected, this response won't come */
-		if (amdgpu_amdkfd_is_fed(dqm->dev->adev))
+		if (amdgpu_amdkfd_is_fed(dqm->dev->adev) ||
+		    amdgpu_in_reset(dqm->dev->adev))
 			return -EIO;
 
 		if (time_after(jiffies, end_jiffies)) {
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index 82905f3e54dd..5a190dd6be4e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -748,16 +748,6 @@ void kfd_signal_event_interrupt(u32 pasid, uint32_t partial_id,
 		uint64_t *slots = page_slots(p->signal_page);
 		uint32_t id;
 
-		/*
-		 * If id is valid but slot is not signaled, GPU may signal the same event twice
-		 * before driver have chance to process the first interrupt, then signal slot is
-		 * auto-reset after set_event wakeup the user space, just drop the second event as
-		 * the application only need wakeup once.
-		 */
-		if ((valid_id_bits > 31 || (1U << valid_id_bits) >= KFD_SIGNAL_EVENT_LIMIT) &&
-		    partial_id < KFD_SIGNAL_EVENT_LIMIT && slots[partial_id] == UNSIGNALED_EVENT_SLOT)
-			goto out_unlock;
-
 		if (valid_id_bits)
 			pr_debug_ratelimited("Partial ID invalid: %u (%u valid bits)\n",
 					     partial_id, valid_id_bits);
@@ -786,7 +776,6 @@ void kfd_signal_event_interrupt(u32 pasid, uint32_t partial_id,
 		}
 	}
 
-out_unlock:
 	rcu_read_unlock();
 	kfd_unref_process(p);
 }
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
index 46c84fc60af1..af53e796ea1b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
@@ -217,7 +217,7 @@ svm_migrate_get_vram_page(struct svm_range *prange, unsigned long pfn)
 	page = pfn_to_page(pfn);
 	svm_range_bo_ref(prange->svm_bo);
 	page->zone_device_data = prange->svm_bo;
-	zone_device_page_init(page);
+	zone_device_page_init(page, 0);
 }
 
 static void
@@ -567,8 +567,9 @@ out:
 	return r < 0 ? r : 0;
 }
 
-static void svm_migrate_page_free(struct page *page)
+static void svm_migrate_folio_free(struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct svm_range_bo *svm_bo = page->zone_device_data;
 
 	if (svm_bo) {
@@ -1008,7 +1009,7 @@ out_mmput:
 }
 
 static const struct dev_pagemap_ops svm_migrate_pgmap_ops = {
-	.page_free		= svm_migrate_page_free,
+	.folio_free		= svm_migrate_folio_free,
 	.migrate_to_ram		= svm_migrate_to_ram,
 };
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
index a65c67cf56ff..80c4fa2b0975 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
@@ -297,16 +297,16 @@ int kfd_queue_acquire_buffers(struct kfd_process_device *pdd, struct queue_prope
 		goto out_err_unreserve;
 	}
 
-	if (properties->ctx_save_restore_area_size != topo_dev->node_props.cwsr_size) {
-		pr_debug("queue cwsr size 0x%x not equal to node cwsr size 0x%x\n",
+	if (properties->ctx_save_restore_area_size < topo_dev->node_props.cwsr_size) {
+		pr_debug("queue cwsr size 0x%x not sufficient for node cwsr size 0x%x\n",
 			properties->ctx_save_restore_area_size,
 			topo_dev->node_props.cwsr_size);
 		err = -EINVAL;
 		goto out_err_unreserve;
 	}
 
-	total_cwsr_size = (topo_dev->node_props.cwsr_size + topo_dev->node_props.debug_memory_size)
-			  * NUM_XCC(pdd->dev->xcc_mask);
+	total_cwsr_size = (properties->ctx_save_restore_area_size +
+			   topo_dev->node_props.debug_memory_size) * NUM_XCC(pdd->dev->xcc_mask);
 	total_cwsr_size = ALIGN(total_cwsr_size, PAGE_SIZE);
 
 	err = kfd_queue_buffer_get(vm, (void *)properties->ctx_save_restore_area_address,
@@ -352,8 +352,8 @@ int kfd_queue_release_buffers(struct kfd_process_device *pdd, struct queue_prope
 	topo_dev = kfd_topology_device_by_id(pdd->dev->id);
 	if (!topo_dev)
 		return -EINVAL;
-	total_cwsr_size = (topo_dev->node_props.cwsr_size + topo_dev->node_props.debug_memory_size)
-			  * NUM_XCC(pdd->dev->xcc_mask);
+	total_cwsr_size = (properties->ctx_save_restore_area_size +
+			   topo_dev->node_props.debug_memory_size) * NUM_XCC(pdd->dev->xcc_mask);
 	total_cwsr_size = ALIGN(total_cwsr_size, PAGE_SIZE);
 
 	kfd_queue_buffer_svm_put(pdd, properties->ctx_save_restore_area_address, total_cwsr_size);
@@ -409,6 +409,7 @@ static u32 kfd_get_vgpr_size_per_cu(u32 gfxv)
 		vgpr_size = 0x80000;
 	else if (gfxv == 110000 ||		/* GFX_VERSION_PLUM_BONITO */
 		 gfxv == 110001 ||		/* GFX_VERSION_WHEAT_NAS */
+		 gfxv == 110501 ||		/* GFX_VERSION_GFX1151 */
 		 gfxv == 120000 ||		/* GFX_VERSION_GFX1200 */
 		 gfxv == 120001)		/* GFX_VERSION_GFX1201 */
 		vgpr_size = 0x60000;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
index ffb7b36e577c..79ea138897fc 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
@@ -1144,30 +1144,48 @@ static int
 svm_range_split_tail(struct svm_range *prange, uint64_t new_last,
 		     struct list_head *insert_list, struct list_head *remap_list)
 {
+	unsigned long last_align_down = ALIGN_DOWN(prange->last, 512);
+	unsigned long start_align = ALIGN(prange->start, 512);
+	bool huge_page_mapping = last_align_down > start_align;
 	struct svm_range *tail = NULL;
-	int r = svm_range_split(prange, prange->start, new_last, &tail);
+	int r;
 
-	if (!r) {
-		list_add(&tail->list, insert_list);
-		if (!IS_ALIGNED(new_last + 1, 1UL << prange->granularity))
-			list_add(&tail->update_list, remap_list);
-	}
-	return r;
+	r = svm_range_split(prange, prange->start, new_last, &tail);
+
+	if (r)
+		return r;
+
+	list_add(&tail->list, insert_list);
+
+	if (huge_page_mapping && tail->start > start_align &&
+	    tail->start < last_align_down && (!IS_ALIGNED(tail->start, 512)))
+		list_add(&tail->update_list, remap_list);
+
+	return 0;
 }
 
 static int
 svm_range_split_head(struct svm_range *prange, uint64_t new_start,
 		     struct list_head *insert_list, struct list_head *remap_list)
 {
+	unsigned long last_align_down = ALIGN_DOWN(prange->last, 512);
+	unsigned long start_align = ALIGN(prange->start, 512);
+	bool huge_page_mapping = last_align_down > start_align;
 	struct svm_range *head = NULL;
-	int r = svm_range_split(prange, new_start, prange->last, &head);
+	int r;
 
-	if (!r) {
-		list_add(&head->list, insert_list);
-		if (!IS_ALIGNED(new_start, 1UL << prange->granularity))
-			list_add(&head->update_list, remap_list);
-	}
-	return r;
+	r = svm_range_split(prange, new_start, prange->last, &head);
+
+	if (r)
+		return r;
+
+	list_add(&head->list, insert_list);
+
+	if (huge_page_mapping && head->last + 1 > start_align &&
+	    head->last + 1 < last_align_down && (!IS_ALIGNED(head->last, 512)))
+		list_add(&head->update_list, remap_list);
+
+	return 0;
 }
 
 static void
@@ -1744,11 +1762,8 @@ static int svm_range_validate_and_map(struct mm_struct *mm,
 			else
 				r = -ENOMEM;
 			WRITE_ONCE(p->svms.faulting_task, NULL);
-			if (r) {
-				amdgpu_hmm_range_free(range);
-				range = NULL;
+			if (r)
 				pr_debug("failed %d to get svm range pages\n", r);
-			}
 		} else {
 			r = -EFAULT;
 		}
@@ -1771,10 +1786,9 @@ static int svm_range_validate_and_map(struct mm_struct *mm,
 			pr_debug("hmm update the range, need validate again\n");
 			r = -EAGAIN;
 		}
-		/* Free the hmm range */
-		if (range)
-			amdgpu_hmm_range_free(range);
 
+		/* Free the hmm range */
+		amdgpu_hmm_range_free(range);
 
 		if (!r && !list_empty(&prange->child_list)) {
 			pr_debug("range split by unmap in parallel, validate again\n");
@@ -3697,6 +3711,8 @@ svm_range_set_attr(struct kfd_process *p, struct mm_struct *mm,
 		svm_range_apply_attrs(p, prange, nattr, attrs, &update_mapping);
 		/* TODO: unmap ranges from GPU that lost access */
 	}
+	update_mapping |= !p->xnack_enabled && !list_empty(&remap_list);
+
 	list_for_each_entry_safe(prange, next, &remove_list, update_list) {
 		pr_debug("unlink old 0x%p prange 0x%p [0x%lx 0x%lx]\n",
 			 prange->svms, prange, prange->start,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index 5c98746eb72d..3eb32d58a120 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -491,6 +491,10 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 			      dev->node_props.num_sdma_queues_per_engine);
 	sysfs_show_32bit_prop(buffer, offs, "num_cp_queues",
 			      dev->node_props.num_cp_queues);
+	sysfs_show_32bit_prop(buffer, offs, "cwsr_size",
+			      dev->node_props.cwsr_size);
+	sysfs_show_32bit_prop(buffer, offs, "ctl_stack_size",
+			      dev->node_props.ctl_stack_size);
 
 	if (dev->gpu) {
 		log_max_watch_addr =
@@ -530,7 +534,9 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 		sysfs_show_32bit_prop(buffer, offs, "sdma_fw_version",
 				      dev->gpu->kfd->sdma_fw_version);
 		sysfs_show_64bit_prop(buffer, offs, "unique_id",
-				      dev->gpu->xcp ?
+				      dev->gpu->xcp &&
+				      (dev->gpu->xcp->xcp_mgr->mode !=
+				       AMDGPU_SPX_PARTITION_MODE) ?
 				      dev->gpu->xcp->unique_id :
 				      dev->gpu->adev->unique_id);
 		sysfs_show_32bit_prop(buffer, offs, "num_xcc",
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 6e5c3e7afb09..740711ac1037 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3624,6 +3624,7 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
 	/* Do mst topology probing after resuming cached state*/
 	drm_connector_list_iter_begin(ddev, &iter);
 	drm_for_each_connector_iter(connector, &iter) {
+		bool init = false;
 
 		if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK)
 			continue;
@@ -3633,7 +3634,14 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
 		    aconnector->mst_root)
 			continue;
 
-		drm_dp_mst_topology_queue_probe(&aconnector->mst_mgr);
+		scoped_guard(mutex, &aconnector->mst_mgr.lock) {
+			init = !aconnector->mst_mgr.mst_primary;
+		}
+		if (init)
+			dm_helpers_dp_mst_start_top_mgr(aconnector->dc_link->ctx,
+				aconnector->dc_link, false);
+		else
+			drm_dp_mst_topology_queue_probe(&aconnector->mst_mgr);
 	}
 	drm_connector_list_iter_end(&iter);
 
@@ -3924,6 +3932,97 @@ void amdgpu_dm_update_connector_after_detect(
 		mutex_unlock(&dev->mode_config.mutex);
 }
 
+static bool are_sinks_equal(const struct dc_sink *sink1, const struct dc_sink *sink2)
+{
+	if (!sink1 || !sink2)
+		return false;
+	if (sink1->sink_signal != sink2->sink_signal)
+		return false;
+
+	if (sink1->dc_edid.length != sink2->dc_edid.length)
+		return false;
+
+	if (memcmp(sink1->dc_edid.raw_edid, sink2->dc_edid.raw_edid,
+		   sink1->dc_edid.length) != 0)
+		return false;
+	return true;
+}
+
+
+/**
+ * DOC: hdmi_hpd_debounce_work
+ *
+ * HDMI HPD debounce delay in milliseconds. When an HDMI display toggles HPD
+ * (such as during power save transitions), this delay determines how long to
+ * wait before processing the HPD event. This allows distinguishing between a
+ * physical unplug (>hdmi_hpd_debounce_delay)
+ * and a spontaneous RX HPD toggle (<hdmi_hpd_debounce_delay).
+ *
+ * If the toggle is less than this delay, the driver compares sink capabilities
+ * and permits a hotplug event if they changed.
+ *
+ * The default value of 1500ms was chosen based on experimental testing with
+ * various monitors that exhibit spontaneous HPD toggling behavior.
+ */
+static void hdmi_hpd_debounce_work(struct work_struct *work)
+{
+	struct amdgpu_dm_connector *aconnector =
+		container_of(to_delayed_work(work), struct amdgpu_dm_connector,
+			     hdmi_hpd_debounce_work);
+	struct drm_connector *connector = &aconnector->base;
+	struct drm_device *dev = connector->dev;
+	struct amdgpu_device *adev = drm_to_adev(dev);
+	struct dc *dc = aconnector->dc_link->ctx->dc;
+	bool fake_reconnect = false;
+	bool reallow_idle = false;
+	bool ret = false;
+	guard(mutex)(&aconnector->hpd_lock);
+
+	/* Re-detect the display */
+	scoped_guard(mutex, &adev->dm.dc_lock) {
+		if (dc->caps.ips_support && dc->ctx->dmub_srv->idle_allowed) {
+			dc_allow_idle_optimizations(dc, false);
+			reallow_idle = true;
+		}
+		ret = dc_link_detect(aconnector->dc_link, DETECT_REASON_HPD);
+	}
+
+	if (ret) {
+		/* Apply workaround delay for certain panels */
+		apply_delay_after_dpcd_poweroff(adev, aconnector->dc_sink);
+		/* Compare sinks to determine if this was a spontaneous HPD toggle */
+		if (are_sinks_equal(aconnector->dc_link->local_sink, aconnector->hdmi_prev_sink)) {
+			/*
+			* Sinks match - this was a spontaneous HDMI HPD toggle.
+			*/
+			drm_dbg_kms(dev, "HDMI HPD: Sink unchanged after debounce, internal re-enable\n");
+			fake_reconnect = true;
+		}
+
+		/* Update connector state */
+		amdgpu_dm_update_connector_after_detect(aconnector);
+
+		drm_modeset_lock_all(dev);
+		dm_restore_drm_connector_state(dev, connector);
+		drm_modeset_unlock_all(dev);
+
+		/* Only notify OS if sink actually changed */
+		if (!fake_reconnect && aconnector->base.force == DRM_FORCE_UNSPECIFIED)
+			drm_kms_helper_hotplug_event(dev);
+	}
+
+	/* Release the cached sink reference */
+	if (aconnector->hdmi_prev_sink) {
+		dc_sink_release(aconnector->hdmi_prev_sink);
+		aconnector->hdmi_prev_sink = NULL;
+	}
+
+	scoped_guard(mutex, &adev->dm.dc_lock) {
+		if (reallow_idle && dc->caps.ips_support)
+			dc_allow_idle_optimizations(dc, true);
+	}
+}
+
 static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
 {
 	struct drm_connector *connector = &aconnector->base;
@@ -3933,6 +4032,7 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
 	struct dm_connector_state *dm_con_state = to_dm_connector_state(connector->state);
 	struct dc *dc = aconnector->dc_link->ctx->dc;
 	bool ret = false;
+	bool debounce_required = false;
 
 	if (adev->dm.disable_hpd_irq)
 		return;
@@ -3955,6 +4055,14 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
 	if (!dc_link_detect_connection_type(aconnector->dc_link, &new_connection_type))
 		drm_err(adev_to_drm(adev), "KMS: Failed to detect connector\n");
 
+	/*
+	 * Check for HDMI disconnect with debounce enabled.
+	 */
+	debounce_required = (aconnector->hdmi_hpd_debounce_delay_ms > 0 &&
+			      dc_is_hdmi_signal(aconnector->dc_link->connector_signal) &&
+			      new_connection_type == dc_connection_none &&
+			      aconnector->dc_link->local_sink != NULL);
+
 	if (aconnector->base.force && new_connection_type == dc_connection_none) {
 		emulated_link_detect(aconnector->dc_link);
 
@@ -3964,7 +4072,34 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
 
 		if (aconnector->base.force == DRM_FORCE_UNSPECIFIED)
 			drm_kms_helper_connector_hotplug_event(connector);
+	} else if (debounce_required) {
+		/*
+		 * HDMI disconnect detected - schedule delayed work instead of
+		 * processing immediately. This allows us to coalesce spurious
+		 * HDMI signals from physical unplugs.
+		 */
+		drm_dbg_kms(dev, "HDMI HPD: Disconnect detected, scheduling debounce work (%u ms)\n",
+			    aconnector->hdmi_hpd_debounce_delay_ms);
+
+		/* Cache the current sink for later comparison */
+		if (aconnector->hdmi_prev_sink)
+			dc_sink_release(aconnector->hdmi_prev_sink);
+		aconnector->hdmi_prev_sink = aconnector->dc_link->local_sink;
+		if (aconnector->hdmi_prev_sink)
+			dc_sink_retain(aconnector->hdmi_prev_sink);
+
+		/* Schedule delayed detection. */
+		if (mod_delayed_work(system_wq,
+				 &aconnector->hdmi_hpd_debounce_work,
+				 msecs_to_jiffies(aconnector->hdmi_hpd_debounce_delay_ms)))
+			drm_dbg_kms(dev, "HDMI HPD: Re-scheduled debounce work\n");
+
 	} else {
+
+		/* If the aconnector->hdmi_hpd_debounce_work is scheduled, exit early */
+		if (delayed_work_pending(&aconnector->hdmi_hpd_debounce_work))
+			return;
+
 		scoped_guard(mutex, &adev->dm.dc_lock) {
 			dc_exit_ips_for_hw_access(dc);
 			ret = dc_link_detect(aconnector->dc_link, DETECT_REASON_HPD);
@@ -4990,6 +5125,21 @@ static void amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm,
 	struct dc_link *link;
 	u32 brightness;
 	bool rc, reallow_idle = false;
+	struct drm_connector *connector;
+
+	list_for_each_entry(connector, &dm->ddev->mode_config.connector_list, head) {
+		struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+
+		if (aconnector->bl_idx != bl_idx)
+			continue;
+
+		/* if connector is off, save the brightness for next time it's on */
+		if (!aconnector->base.encoder) {
+			dm->brightness[bl_idx] = user_brightness;
+			dm->actual_brightness[bl_idx] = 0;
+			return;
+		}
+	}
 
 	amdgpu_dm_update_backlight_caps(dm, bl_idx);
 	caps = &dm->backlight_caps[bl_idx];
@@ -7237,10 +7387,18 @@ finish:
 }
 
 /**
- * amdgpu_dm_connector_poll() - Poll a connector to see if it's connected to a display
+ * amdgpu_dm_connector_poll - Poll a connector to see if it's connected to a display
+ * @aconnector: DM connector to poll (owns @base drm_connector and @dc_link)
+ * @force: if true, force polling even when DAC load detection was used
+ *
+ * Used for connectors that don't support HPD (hotplug detection) to
+ * periodically check whether the connector is connected to a display.
+ *
+ * When connection was determined via DAC load detection, we avoid
+ * re-running it on normal polls to prevent visible glitches, unless
+ * @force is set.
  *
- * Used for connectors that don't support HPD (hotplug detection)
- * to periodically checked whether the connector is connected to a display.
+ * Return: The probed connector status (connected/disconnected/unknown).
  */
 static enum drm_connector_status
 amdgpu_dm_connector_poll(struct amdgpu_dm_connector *aconnector, bool force)
@@ -7308,6 +7466,14 @@ amdgpu_dm_connector_poll(struct amdgpu_dm_connector *aconnector, bool force)
  * 1. This interface is NOT called in context of HPD irq.
  * 2. This interface *is called* in context of user-mode ioctl. Which
  *    makes it a bad place for *any* MST-related activity.
+ *
+ * @connector: The DRM connector we are checking. We convert it to
+ *             amdgpu_dm_connector so we can read the DC link and state.
+ * @force:     If true, do a full detect again. This is used even when
+ *             a lighter check would normally be used to avoid flicker.
+ *
+ * Return: The connector status (connected, disconnected, or unknown).
+ *
  */
 static enum drm_connector_status
 amdgpu_dm_connector_detect(struct drm_connector *connector, bool force)
@@ -7393,7 +7559,7 @@ int amdgpu_dm_connector_atomic_set_property(struct drm_connector *connector,
 		default:
 			dm_new_state->abm_sysfs_forbidden = true;
 			dm_new_state->abm_level = val;
-		};
+		}
 		ret = 0;
 	}
 
@@ -7570,6 +7736,13 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector)
 	if (aconnector->mst_mgr.dev)
 		drm_dp_mst_topology_mgr_destroy(&aconnector->mst_mgr);
 
+	/* Cancel and flush any pending HDMI HPD debounce work */
+	cancel_delayed_work_sync(&aconnector->hdmi_hpd_debounce_work);
+	if (aconnector->hdmi_prev_sink) {
+		dc_sink_release(aconnector->hdmi_prev_sink);
+		aconnector->hdmi_prev_sink = NULL;
+	}
+
 	if (aconnector->bl_idx != -1) {
 		backlight_device_unregister(dm->backlight_dev[aconnector->bl_idx]);
 		dm->backlight_dev[aconnector->bl_idx] = NULL;
@@ -8220,7 +8393,7 @@ static int dm_encoder_helper_atomic_check(struct drm_encoder *encoder,
 				       "mode %dx%d@%dHz is not native, enabling scaling\n",
 				       adjusted_mode->hdisplay, adjusted_mode->vdisplay,
 				       drm_mode_vrefresh(adjusted_mode));
-			dm_new_connector_state->scaling = RMX_FULL;
+			dm_new_connector_state->scaling = RMX_ASPECT;
 		}
 		return 0;
 	}
@@ -8666,8 +8839,14 @@ static void amdgpu_dm_connector_add_freesync_modes(struct drm_connector *connect
 	if (!(amdgpu_freesync_vid_mode && drm_edid))
 		return;
 
-	if (!amdgpu_dm_connector->dc_sink || amdgpu_dm_connector->dc_sink->edid_caps.analog ||
-		!dc_supports_vrr(amdgpu_dm_connector->dc_sink->ctx->dce_version))
+	if (!amdgpu_dm_connector->dc_sink || !amdgpu_dm_connector->dc_link)
+		return;
+
+	if (!dc_supports_vrr(amdgpu_dm_connector->dc_sink->ctx->dce_version))
+		return;
+
+	if (dc_connector_supports_analog(amdgpu_dm_connector->dc_link->link_id.id) &&
+	    amdgpu_dm_connector->dc_sink->edid_caps.analog)
 		return;
 
 	if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10)
@@ -8679,11 +8858,11 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector)
 {
 	struct amdgpu_dm_connector *amdgpu_dm_connector =
 			to_amdgpu_dm_connector(connector);
+	struct dc_link *dc_link = amdgpu_dm_connector->dc_link;
 	struct drm_encoder *encoder;
 	const struct drm_edid *drm_edid = amdgpu_dm_connector->drm_edid;
-	struct dc_link_settings *verified_link_cap =
-			&amdgpu_dm_connector->dc_link->verified_link_cap;
-	const struct dc *dc = amdgpu_dm_connector->dc_link->dc;
+	struct dc_link_settings *verified_link_cap = &dc_link->verified_link_cap;
+	const struct dc *dc = dc_link->dc;
 
 	encoder = amdgpu_dm_connector_to_encoder(connector);
 
@@ -8694,7 +8873,9 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector)
 			amdgpu_dm_connector->num_modes +=
 				drm_add_modes_noedid(connector, 1920, 1080);
 
-		if (amdgpu_dm_connector->dc_sink->edid_caps.analog) {
+		if (amdgpu_dm_connector->dc_sink &&
+		    amdgpu_dm_connector->dc_sink->edid_caps.analog &&
+		    dc_connector_supports_analog(dc_link->link_id.id)) {
 			/* Analog monitor connected by DAC load detection.
 			 * Add common modes. It will be up to the user to select one that works.
 			 */
@@ -8749,6 +8930,10 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 	mutex_init(&aconnector->hpd_lock);
 	mutex_init(&aconnector->handle_mst_msg_ready);
 
+	aconnector->hdmi_hpd_debounce_delay_ms = AMDGPU_DM_HDMI_HPD_DEBOUNCE_MS;
+	INIT_DELAYED_WORK(&aconnector->hdmi_hpd_debounce_work, hdmi_hpd_debounce_work);
+	aconnector->hdmi_prev_sink = NULL;
+
 	/*
 	 * configure support HPD hot plug connector_>polled default value is 0
 	 * which means HPD hot plug not supported
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index 5a7aa903bd3c..bd0403005f37 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -59,6 +59,7 @@
 
 #define AMDGPU_HDR_MULT_DEFAULT (0x100000000LL)
 
+#define AMDGPU_DM_HDMI_HPD_DEBOUNCE_MS 1500
 /*
 #include "include/amdgpu_dal_power_if.h"
 #include "amdgpu_dm_irq.h"
@@ -819,6 +820,11 @@ struct amdgpu_dm_connector {
 	bool pack_sdp_v1_3;
 	enum adaptive_sync_type as_type;
 	struct amdgpu_hdmi_vsdb_info vsdb_info;
+
+	/* HDMI HPD debounce support */
+	unsigned int hdmi_hpd_debounce_delay_ms;
+	struct delayed_work hdmi_hpd_debounce_work;
+	struct dc_sink *hdmi_prev_sink;
 };
 
 static inline void amdgpu_dm_set_mst_status(uint8_t *status,
@@ -1057,6 +1063,9 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc,
 void amdgpu_dm_update_connector_after_detect(
 		struct amdgpu_dm_connector *aconnector);
 
+void populate_hdmi_info_from_connector(struct drm_hdmi_info *info,
+				      struct dc_edid_caps *edid_caps);
+
 extern const struct drm_encoder_helper_funcs amdgpu_dm_encoder_helper_funcs;
 
 int amdgpu_dm_process_dmub_aux_transfer_sync(struct dc_context *ctx, unsigned int link_index,
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 cb4bb67289a4..a9839485f2a2 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
@@ -1303,7 +1303,8 @@ static int odm_combine_segments_show(struct seq_file *m, void *unused)
 	if (connector->status != connector_status_connected)
 		return -ENODEV;
 
-	if (pipe_ctx != NULL && pipe_ctx->stream_res.tg->funcs->get_odm_combine_segments)
+	if (pipe_ctx && pipe_ctx->stream_res.tg &&
+	    pipe_ctx->stream_res.tg->funcs->get_odm_combine_segments)
 		pipe_ctx->stream_res.tg->funcs->get_odm_combine_segments(pipe_ctx->stream_res.tg, &segments);
 
 	seq_printf(m, "%d\n", segments);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
index 19038f336155..85ce558cefc5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
@@ -201,6 +201,7 @@ void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
 	struct mod_hdcp_link_adjustment link_adjust;
 	struct mod_hdcp_display_adjustment display_adjust;
 	unsigned int conn_index = aconnector->base.index;
+	const struct dc *dc = aconnector->dc_link->dc;
 
 	guard(mutex)(&hdcp_w->mutex);
 	drm_connector_get(&aconnector->base);
@@ -231,6 +232,9 @@ void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
 			link_adjust.hdcp1.disable = 1;
 			link_adjust.hdcp2.force_type = MOD_HDCP_FORCE_TYPE_1;
 		}
+		link_adjust.hdcp2.use_fw_locality_check =
+				(dc->caps.fused_io_supported || dc->debug.hdcp_lc_force_fw_enable);
+		link_adjust.hdcp2.use_sw_locality_fallback = dc->debug.hdcp_lc_enable_sw_fallback;
 
 		schedule_delayed_work(&hdcp_w->property_validate_dwork,
 				      msecs_to_jiffies(DRM_HDCP_CHECK_PERIOD_MS));
@@ -534,6 +538,7 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
 	struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
 	struct dc_sink *sink = NULL;
 	bool link_is_hdcp14 = false;
+	const struct dc *dc = aconnector->dc_link->dc;
 
 	if (config->dpms_off) {
 		hdcp_remove_display(hdcp_work, link_index, aconnector);
@@ -575,6 +580,8 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
 	link->adjust.auth_delay = 2;
 	link->adjust.retry_limit = MAX_NUM_OF_ATTEMPTS;
 	link->adjust.hdcp1.disable = 0;
+	link->adjust.hdcp2.use_fw_locality_check = (dc->caps.fused_io_supported || dc->debug.hdcp_lc_force_fw_enable);
+	link->adjust.hdcp2.use_sw_locality_fallback = dc->debug.hdcp_lc_enable_sw_fallback;
 	hdcp_w->encryption_status[display->index] = MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
 
 	DRM_DEBUG_DRIVER("[HDCP_DM] display %d, CP %d, type %d\n", aconnector->base.index,
@@ -786,15 +793,8 @@ struct hdcp_workqueue *hdcp_create_workqueue(struct amdgpu_device *adev,
 		ddc_funcs->read_i2c = lp_read_i2c;
 		ddc_funcs->write_dpcd = lp_write_dpcd;
 		ddc_funcs->read_dpcd = lp_read_dpcd;
-
-		config->debug.lc_enable_sw_fallback = dc->debug.hdcp_lc_enable_sw_fallback;
-		if (dc->caps.fused_io_supported || dc->debug.hdcp_lc_force_fw_enable) {
-			ddc_funcs->atomic_write_poll_read_i2c = lp_atomic_write_poll_read_i2c;
-			ddc_funcs->atomic_write_poll_read_aux = lp_atomic_write_poll_read_aux;
-		} else {
-			ddc_funcs->atomic_write_poll_read_i2c = NULL;
-			ddc_funcs->atomic_write_poll_read_aux = NULL;
-		}
+		ddc_funcs->atomic_write_poll_read_i2c = lp_atomic_write_poll_read_i2c;
+		ddc_funcs->atomic_write_poll_read_aux = lp_atomic_write_poll_read_aux;
 
 		memset(hdcp_work[i].aconnector, 0,
 		       sizeof(struct amdgpu_dm_connector *) *
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index eb2c587b0b9b..e5e993d3ef74 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -139,6 +139,9 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
 
 	edid_caps->edid_hdmi = connector->display_info.is_hdmi;
 
+	if (edid_caps->edid_hdmi)
+		populate_hdmi_info_from_connector(&connector->display_info.hdmi, edid_caps);
+
 	apply_edid_quirks(dev, edid_buf, edid_caps);
 
 	sad_count = drm_edid_to_sad((struct edid *) edid->raw_edid, &sads);
@@ -990,6 +993,11 @@ dm_helpers_read_acpi_edid(struct amdgpu_dm_connector *aconnector)
 	return drm_edid_read_custom(connector, dm_helpers_probe_acpi_edid, connector);
 }
 
+void populate_hdmi_info_from_connector(struct drm_hdmi_info *hdmi, struct dc_edid_caps *edid_caps)
+{
+	edid_caps->scdc_present = hdmi->scdc.supported;
+}
+
 enum dc_edid_status dm_helpers_read_local_edid(
 		struct dc_context *ctx,
 		struct dc_link *link,
@@ -998,8 +1006,8 @@ enum dc_edid_status dm_helpers_read_local_edid(
 	struct amdgpu_dm_connector *aconnector = link->priv;
 	struct drm_connector *connector = &aconnector->base;
 	struct i2c_adapter *ddc;
-	int retry = 3;
-	enum dc_edid_status edid_status;
+	int retry = 25;
+	enum dc_edid_status edid_status = EDID_NO_RESPONSE;
 	const struct drm_edid *drm_edid;
 	const struct edid *edid;
 
@@ -1029,7 +1037,7 @@ enum dc_edid_status dm_helpers_read_local_edid(
 		}
 
 		if (!drm_edid)
-			return EDID_NO_RESPONSE;
+			continue;
 
 		edid = drm_edid_raw(drm_edid); // FIXME: Get rid of drm_edid_raw()
 		if (!edid ||
@@ -1047,7 +1055,7 @@ enum dc_edid_status dm_helpers_read_local_edid(
 						&sink->dc_edid,
 						&sink->edid_caps);
 
-	} while (edid_status == EDID_BAD_CHECKSUM && --retry > 0);
+	} while ((edid_status == EDID_BAD_CHECKSUM || edid_status == EDID_NO_RESPONSE) && --retry > 0);
 
 	if (edid_status != EDID_OK)
 		DRM_ERROR("EDID err: %d, on connector: %s",
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_replay.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_replay.c
index 80704d709e44..da94e3544b65 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_replay.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_replay.c
@@ -162,7 +162,7 @@ bool amdgpu_dm_replay_enable(struct dc_stream_state *stream, bool wait)
 
 	if (link) {
 		link->dc->link_srv->edp_setup_replay(link, stream);
-		link->dc->link_srv->edp_set_coasting_vtotal(link, stream->timing.v_total);
+		link->dc->link_srv->edp_set_coasting_vtotal(link, stream->timing.v_total, 0);
 		DRM_DEBUG_DRIVER("Enabling replay...\n");
 		link->dc->link_srv->edp_set_replay_allow_active(link, &replay_active, wait, false, NULL);
 		return true;
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
index 4120d6c4c5e4..d1471f34e419 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
@@ -805,7 +805,7 @@ static enum bp_result bios_parser_dac_load_detection(
 
 	if (bp_params.device_id == ATOM_DEVICE_CRT1_SUPPORT)
 		device_id_mask = ATOM_S0_CRT1_MASK;
-	else if (bp_params.device_id == ATOM_DEVICE_CRT1_SUPPORT)
+	else if (bp_params.device_id == ATOM_DEVICE_CRT2_SUPPORT)
 		device_id_mask = ATOM_S0_CRT2_MASK;
 	else
 		return BP_RESULT_UNSUPPORTED;
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 04eb647acc4e..550a9f1d03f8 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -1480,10 +1480,10 @@ static enum bp_result get_embedded_panel_info_v2_1(
 	/* not provided by VBIOS */
 	info->lcd_timing.misc_info.HORIZONTAL_CUT_OFF = 0;
 
-	info->lcd_timing.misc_info.H_SYNC_POLARITY = ~(uint32_t) (lvds->lcd_timing.miscinfo
-			& ATOM_HSYNC_POLARITY);
-	info->lcd_timing.misc_info.V_SYNC_POLARITY = ~(uint32_t) (lvds->lcd_timing.miscinfo
-			& ATOM_VSYNC_POLARITY);
+	info->lcd_timing.misc_info.H_SYNC_POLARITY = !(lvds->lcd_timing.miscinfo &
+						       ATOM_HSYNC_POLARITY);
+	info->lcd_timing.misc_info.V_SYNC_POLARITY = !(lvds->lcd_timing.miscinfo &
+						       ATOM_VSYNC_POLARITY);
 
 	/* not provided by VBIOS */
 	info->lcd_timing.misc_info.VERTICAL_CUT_OFF = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
index 35d20a663d67..dfd0c9505af0 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
@@ -1295,6 +1295,35 @@ static void dcn35_update_clocks_fpga(struct clk_mgr *clk_mgr,
 	dcn35_update_clocks_update_dtb_dto(clk_mgr_int, context, clk_mgr->clks.ref_dtbclk_khz);
 }
 
+static unsigned int dcn35_get_max_clock_khz(struct clk_mgr *clk_mgr_base, enum clk_type clk_type)
+{
+	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+	unsigned int num_clk_levels;
+
+	switch (clk_type) {
+	case CLK_TYPE_DISPCLK:
+		num_clk_levels = clk_mgr->base.bw_params->clk_table.num_entries_per_clk.num_dispclk_levels;
+		return num_clk_levels ?
+				clk_mgr->base.bw_params->clk_table.entries[num_clk_levels - 1].dispclk_mhz * 1000 :
+				clk_mgr->base.boot_snapshot.dispclk;
+	case CLK_TYPE_DPPCLK:
+		num_clk_levels = clk_mgr->base.bw_params->clk_table.num_entries_per_clk.num_dppclk_levels;
+		return num_clk_levels ?
+				clk_mgr->base.bw_params->clk_table.entries[num_clk_levels - 1].dppclk_mhz * 1000 :
+				clk_mgr->base.boot_snapshot.dppclk;
+	case CLK_TYPE_DSCCLK:
+		num_clk_levels = clk_mgr->base.bw_params->clk_table.num_entries_per_clk.num_dispclk_levels;
+		return num_clk_levels ?
+				clk_mgr->base.bw_params->clk_table.entries[num_clk_levels - 1].dispclk_mhz * 1000 / 3 :
+				clk_mgr->base.boot_snapshot.dispclk / 3;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
 static struct clk_mgr_funcs dcn35_funcs = {
 	.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,
 	.get_dtb_ref_clk_frequency = dcn31_get_dtb_ref_freq_khz,
@@ -1306,6 +1335,7 @@ static struct clk_mgr_funcs dcn35_funcs = {
 	.set_low_power_state = dcn35_set_low_power_state,
 	.exit_low_power_state = dcn35_exit_low_power_state,
 	.is_ips_supported = dcn35_is_ips_supported,
+	.get_max_clock_khz = dcn35_get_max_clock_khz,
 };
 
 struct clk_mgr_funcs dcn35_fpga_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index b720e007c654..8be9cbd43e18 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -303,6 +303,7 @@ static bool create_links(
 		link->link_id.id = CONNECTOR_ID_VIRTUAL;
 		link->link_id.enum_id = ENUM_ID_1;
 		link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED;
+		link->replay_settings.config.replay_version = DC_REPLAY_VERSION_UNSUPPORTED;
 		link->link_enc = kzalloc(sizeof(*link->link_enc), GFP_KERNEL);
 
 		if (!link->link_enc) {
@@ -2146,6 +2147,14 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 	if (!dcb->funcs->is_accelerated_mode(dcb)) {
 		disable_vbios_mode_if_required(dc, context);
 		dc->hwss.enable_accelerated_mode(dc, context);
+	} else if (get_seamless_boot_stream_count(dc->current_state) > 0) {
+		/* If the previous Stream still retains the apply seamless boot flag,
+		 * it means the OS has not actually performed a flip yet.
+		 * At this point, if we receive dc_commit_streams again, we should
+		 * once more check whether the actual HW timing matches what the OS
+		 * has provided
+		 */
+		disable_vbios_mode_if_required(dc, context);
 	}
 
 	if (dc->hwseq->funcs.wait_for_pipe_update_if_needed) {
@@ -2670,41 +2679,42 @@ static struct surface_update_descriptor get_plane_info_update_type(const struct
 	if (!u->plane_info)
 		return update_type;
 
-	elevate_update_type(&update_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_PLANE);
+	// `plane_info` present means at least `STREAM` lock is required
+	elevate_update_type(&update_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
 
 	if (u->plane_info->color_space != u->surface->color_space) {
 		update_flags->bits.color_space_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 	}
 
 	if (u->plane_info->horizontal_mirror != u->surface->horizontal_mirror) {
 		update_flags->bits.horizontal_mirror_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 	}
 
 	if (u->plane_info->rotation != u->surface->rotation) {
 		update_flags->bits.rotation_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (u->plane_info->format != u->surface->format) {
 		update_flags->bits.pixel_format_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (u->plane_info->stereo_format != u->surface->stereo_format) {
 		update_flags->bits.stereo_format_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (u->plane_info->per_pixel_alpha != u->surface->per_pixel_alpha) {
 		update_flags->bits.per_pixel_alpha_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 	}
 
 	if (u->plane_info->global_alpha_value != u->surface->global_alpha_value) {
 		update_flags->bits.global_alpha_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 	}
 
 	if (u->plane_info->dcc.enable != u->surface->dcc.enable
@@ -2716,7 +2726,7 @@ static struct surface_update_descriptor get_plane_info_update_type(const struct
 		 * recalculate stutter period.
 		 */
 		update_flags->bits.dcc_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (resource_pixel_format_to_bpp(u->plane_info->format) !=
@@ -2725,34 +2735,34 @@ static struct surface_update_descriptor get_plane_info_update_type(const struct
 		 * and DML calculation
 		 */
 		update_flags->bits.bpp_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (u->plane_info->plane_size.surface_pitch != u->surface->plane_size.surface_pitch
 			|| u->plane_info->plane_size.chroma_pitch != u->surface->plane_size.chroma_pitch) {
 		update_flags->bits.plane_size_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 	}
 
 	const struct dc_tiling_info *tiling = &u->plane_info->tiling_info;
 
 	if (memcmp(tiling, &u->surface->tiling_info, sizeof(*tiling)) != 0) {
 		update_flags->bits.swizzle_change = 1;
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 
 		switch (tiling->gfxversion) {
 		case DcGfxVersion9:
 		case DcGfxVersion10:
 		case DcGfxVersion11:
 			if (tiling->gfx9.swizzle != DC_SW_LINEAR) {
-				elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
 				update_flags->bits.bandwidth_change = 1;
+				elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 			}
 			break;
 		case DcGfxAddr3:
 			if (tiling->gfx_addr3.swizzle != DC_ADDR3_SW_LINEAR) {
-				elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
 				update_flags->bits.bandwidth_change = 1;
+				elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 			}
 			break;
 		case DcGfxVersion7:
@@ -2777,7 +2787,8 @@ static struct surface_update_descriptor get_scaling_info_update_type(
 	if (!u->scaling_info)
 		return update_type;
 
-	elevate_update_type(&update_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_PLANE);
+	// `scaling_info` present means at least `STREAM` lock is required
+	elevate_update_type(&update_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
 
 	if (u->scaling_info->src_rect.width != u->surface->src_rect.width
 			|| u->scaling_info->src_rect.height != u->surface->src_rect.height
@@ -2788,6 +2799,7 @@ static struct surface_update_descriptor get_scaling_info_update_type(
 			|| u->scaling_info->scaling_quality.integer_scaling !=
 					u->surface->scaling_quality.integer_scaling) {
 		update_flags->bits.scaling_change = 1;
+		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 
 		if (u->scaling_info->src_rect.width > u->surface->src_rect.width
 				|| u->scaling_info->src_rect.height > u->surface->src_rect.height)
@@ -2813,17 +2825,10 @@ static struct surface_update_descriptor get_scaling_info_update_type(
 			|| u->scaling_info->clip_rect.x != u->surface->clip_rect.x
 			|| u->scaling_info->clip_rect.y != u->surface->clip_rect.y
 			|| u->scaling_info->dst_rect.x != u->surface->dst_rect.x
-			|| u->scaling_info->dst_rect.y != u->surface->dst_rect.y)
+			|| u->scaling_info->dst_rect.y != u->surface->dst_rect.y) {
+		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 		update_flags->bits.position_change = 1;
-
-	/* process every update flag before returning */
-	if (update_flags->bits.clock_change
-			|| update_flags->bits.bandwidth_change
-			|| update_flags->bits.scaling_change)
-		elevate_update_type(&update_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
-
-	if (update_flags->bits.position_change)
-		elevate_update_type(&update_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+	}
 
 	return update_type;
 }
@@ -2837,7 +2842,7 @@ static struct surface_update_descriptor det_surface_update(
 
 	if (u->surface->force_full_update) {
 		update_flags->raw = 0xFFFFFFFF;
-		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 		return overall_type;
 	}
 
@@ -2852,76 +2857,69 @@ static struct surface_update_descriptor det_surface_update(
 
 	if (u->flip_addr) {
 		update_flags->bits.addr_update = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+
 		if (u->flip_addr->address.tmz_surface != u->surface->address.tmz_surface) {
 			update_flags->bits.tmz_changed = 1;
-			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 		}
 	}
-	if (u->in_transfer_func)
+	if (u->in_transfer_func) {
 		update_flags->bits.in_transfer_func_change = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
+	}
 
-	if (u->input_csc_color_matrix)
+	if (u->input_csc_color_matrix) {
 		update_flags->bits.input_csc_change = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+	}
 
-	if (u->coeff_reduction_factor)
+	if (u->coeff_reduction_factor) {
 		update_flags->bits.coeff_reduction_change = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+	}
 
-	if (u->gamut_remap_matrix)
+	if (u->gamut_remap_matrix) {
 		update_flags->bits.gamut_remap_change = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+	}
 
-	if (u->blend_tf)
+	if (u->blend_tf || (u->gamma && dce_use_lut(u->plane_info ? u->plane_info->format : u->surface->format))) {
 		update_flags->bits.gamma_change = 1;
-
-	if (u->gamma) {
-		enum surface_pixel_format format = SURFACE_PIXEL_FORMAT_GRPH_BEGIN;
-
-		if (u->plane_info)
-			format = u->plane_info->format;
-		else
-			format = u->surface->format;
-
-		if (dce_use_lut(format))
-			update_flags->bits.gamma_change = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
 	}
 
-	if (u->lut3d_func || u->func_shaper)
+	if (u->lut3d_func || u->func_shaper) {
 		update_flags->bits.lut_3d = 1;
+		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+	}
 
 	if (u->hdr_mult.value)
 		if (u->hdr_mult.value != u->surface->hdr_mult.value) {
-			update_flags->bits.hdr_mult = 1;
 			// TODO: Should be fast?
-			elevate_update_type(&overall_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+			update_flags->bits.hdr_mult = 1;
+			elevate_update_type(&overall_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STREAM);
 		}
 
 	if (u->sdr_white_level_nits)
 		if (u->sdr_white_level_nits != u->surface->sdr_white_level_nits) {
-			update_flags->bits.sdr_white_level_nits = 1;
 			// TODO: Should be fast?
-			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+			update_flags->bits.sdr_white_level_nits = 1;
+			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 		}
 
 	if (u->cm2_params) {
-		if ((u->cm2_params->component_settings.shaper_3dlut_setting
-					!= u->surface->mcm_shaper_3dlut_setting)
-				|| (u->cm2_params->component_settings.lut1d_enable
-					!= u->surface->mcm_lut1d_enable))
-			update_flags->bits.mcm_transfer_function_enable_change = 1;
-		if (u->cm2_params->cm2_luts.lut3d_data.lut3d_src
-				!= u->surface->mcm_luts.lut3d_data.lut3d_src)
+		if (u->cm2_params->component_settings.shaper_3dlut_setting != u->surface->mcm_shaper_3dlut_setting
+				|| u->cm2_params->component_settings.lut1d_enable != u->surface->mcm_lut1d_enable
+				|| u->cm2_params->cm2_luts.lut3d_data.lut3d_src != u->surface->mcm_luts.lut3d_data.lut3d_src) {
 			update_flags->bits.mcm_transfer_function_enable_change = 1;
-	}
-	if (update_flags->bits.in_transfer_func_change) {
-		// TODO: Fast?
-		elevate_update_type(&overall_type, UPDATE_TYPE_MED, LOCK_DESCRIPTOR_STATE);
+			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
+		}
 	}
 
 	if (update_flags->bits.lut_3d &&
 			u->surface->mcm_luts.lut3d_data.lut3d_src != DC_CM2_TRANSFER_FUNC_SOURCE_VIDMEM) {
-		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
-	}
-	if (update_flags->bits.mcm_transfer_function_enable_change) {
-		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (check_config->enable_legacy_fast_update &&
@@ -2929,7 +2927,7 @@ static struct surface_update_descriptor det_surface_update(
 			update_flags->bits.gamut_remap_change ||
 			update_flags->bits.input_csc_change ||
 			update_flags->bits.coeff_reduction_change)) {
-		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 	return overall_type;
 }
@@ -2966,19 +2964,19 @@ static struct surface_update_descriptor check_update_surfaces_for_stream(
 	struct surface_update_descriptor overall_type = { UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_NONE };
 
 	if (stream_update && stream_update->pending_test_pattern) {
-		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	if (stream_update && stream_update->hw_cursor_req) {
-		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 	}
 
 	/* some stream updates require passive update */
 	if (stream_update) {
-		union stream_update_flags *su_flags = &stream_update->stream->update_flags;
-
 		elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
 
+		union stream_update_flags *su_flags = &stream_update->stream->update_flags;
+
 		if ((stream_update->src.height != 0 && stream_update->src.width != 0) ||
 			(stream_update->dst.height != 0 && stream_update->dst.width != 0) ||
 			stream_update->integer_scaling_update)
@@ -2990,8 +2988,10 @@ static struct surface_update_descriptor check_update_surfaces_for_stream(
 		if (stream_update->abm_level)
 			su_flags->bits.abm_level = 1;
 
-		if (stream_update->dpms_off)
+		if (stream_update->dpms_off) {
 			su_flags->bits.dpms_off = 1;
+			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL | LOCK_DESCRIPTOR_LINK);
+		}
 
 		if (stream_update->gamut_remap)
 			su_flags->bits.gamut_remap = 1;
@@ -3019,17 +3019,20 @@ static struct surface_update_descriptor check_update_surfaces_for_stream(
 		if (stream_update->output_color_space)
 			su_flags->bits.out_csc = 1;
 
-		if (su_flags->raw != 0)
-			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_STATE);
+		// TODO: Make each elevation explicit, as to not override fast stream in crct_timing_adjust
+		if (su_flags->raw)
+			elevate_update_type(&overall_type, UPDATE_TYPE_FULL, LOCK_DESCRIPTOR_GLOBAL);
 
-		if (stream_update->output_csc_transform)
+		// Non-global cases
+		if (stream_update->output_csc_transform) {
 			su_flags->bits.out_csc = 1;
+			elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+		}
 
-		/* Output transfer function changes do not require bandwidth recalculation,
-		 * so don't trigger a full update
-		 */
-		if (!check_config->enable_legacy_fast_update && stream_update->out_transfer_func)
+		if (!check_config->enable_legacy_fast_update && stream_update->out_transfer_func) {
 			su_flags->bits.out_tf = 1;
+			elevate_update_type(&overall_type, UPDATE_TYPE_FAST, LOCK_DESCRIPTOR_STREAM);
+		}
 	}
 
 	for (int i = 0 ; i < surface_count; i++) {
@@ -3389,7 +3392,11 @@ static void restore_planes_and_stream_state(
 	for (i = 0; i < status->plane_count; i++) {
 		dc_plane_copy_config(status->plane_states[i], &scratch->plane_states[i]);
 	}
+
+	// refcount is persistent
+	struct kref temp_refcount = stream->refcount;
 	*stream = scratch->stream_state;
+	stream->refcount = temp_refcount;
 }
 
 /**
@@ -5973,6 +5980,101 @@ bool dc_process_dmub_aux_transfer_async(struct dc *dc,
 	return true;
 }
 
+bool dc_smart_power_oled_enable(const struct dc_link *link, bool enable, uint16_t peak_nits,
+			uint8_t debug_control, uint16_t fixed_CLL, uint32_t triggerline)
+{
+	bool status = false;
+	struct dc *dc = link->ctx->dc;
+	union dmub_rb_cmd cmd;
+	uint8_t otg_inst = 0;
+	unsigned int panel_inst = 0;
+	struct pipe_ctx *pipe_ctx = NULL;
+	struct resource_context *res_ctx = &link->ctx->dc->current_state->res_ctx;
+	int i = 0;
+
+	// get panel_inst
+	if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst))
+		return status;
+
+	// get otg_inst
+	for (i = 0; i < MAX_PIPES; i++) {
+		if (res_ctx &&
+			res_ctx->pipe_ctx[i].stream &&
+			res_ctx->pipe_ctx[i].stream->link &&
+			res_ctx->pipe_ctx[i].stream->link == link &&
+			res_ctx->pipe_ctx[i].stream->link->connector_signal == SIGNAL_TYPE_EDP) {
+			pipe_ctx = &res_ctx->pipe_ctx[i];
+			//TODO: refactor for multi edp support
+			break;
+		}
+	}
+
+	if (pipe_ctx)
+		otg_inst = pipe_ctx->stream_res.tg->inst;
+
+	// before enable smart power OLED, we need to call set pipe for DMUB to set ABM config
+	if (enable) {
+		if (dc->hwss.set_pipe && pipe_ctx)
+			dc->hwss.set_pipe(pipe_ctx);
+	}
+
+	// fill in cmd
+	memset(&cmd, 0, sizeof(cmd));
+
+	cmd.smart_power_oled_enable.header.type = DMUB_CMD__SMART_POWER_OLED;
+	cmd.smart_power_oled_enable.header.sub_type = DMUB_CMD__SMART_POWER_OLED_ENABLE;
+	cmd.smart_power_oled_enable.header.payload_bytes =
+		sizeof(struct dmub_rb_cmd_smart_power_oled_enable_data) - sizeof(struct dmub_cmd_header);
+	cmd.smart_power_oled_enable.header.ret_status = 1;
+	cmd.smart_power_oled_enable.data.enable = enable;
+	cmd.smart_power_oled_enable.data.panel_inst = panel_inst;
+	cmd.smart_power_oled_enable.data.peak_nits = peak_nits;
+	cmd.smart_power_oled_enable.data.otg_inst = otg_inst;
+	cmd.smart_power_oled_enable.data.digfe_inst = link->link_enc->preferred_engine;
+	cmd.smart_power_oled_enable.data.digbe_inst = link->link_enc->transmitter;
+
+	cmd.smart_power_oled_enable.data.debugcontrol = debug_control;
+	cmd.smart_power_oled_enable.data.triggerline = triggerline;
+	cmd.smart_power_oled_enable.data.fixed_max_cll = fixed_CLL;
+
+	// send cmd
+	status = dc_wake_and_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
+
+	return status;
+}
+
+bool dc_smart_power_oled_get_max_cll(const struct dc_link *link, unsigned int *pCurrent_MaxCLL)
+{
+	struct dc *dc = link->ctx->dc;
+	union dmub_rb_cmd cmd;
+	bool status = false;
+	unsigned int panel_inst = 0;
+
+	// get panel_inst
+	if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst))
+		return status;
+
+	// fill in cmd
+	memset(&cmd, 0, sizeof(cmd));
+
+	cmd.smart_power_oled_getmaxcll.header.type = DMUB_CMD__SMART_POWER_OLED;
+	cmd.smart_power_oled_getmaxcll.header.sub_type = DMUB_CMD__SMART_POWER_OLED_GETMAXCLL;
+	cmd.smart_power_oled_getmaxcll.header.payload_bytes = sizeof(cmd.smart_power_oled_getmaxcll.data);
+	cmd.smart_power_oled_getmaxcll.header.ret_status = 1;
+
+	cmd.smart_power_oled_getmaxcll.data.input.panel_inst = panel_inst;
+
+	// send cmd and wait for reply
+	status = dc_wake_and_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY);
+
+	if (status)
+		*pCurrent_MaxCLL = cmd.smart_power_oled_getmaxcll.data.output.current_max_cll;
+	else
+		*pCurrent_MaxCLL = 0;
+
+	return status;
+}
+
 uint8_t get_link_index_from_dpia_port_index(const struct dc *dc,
 					    uint8_t dpia_port_index)
 {
@@ -6417,6 +6519,574 @@ void dc_get_underflow_debug_data_for_otg(struct dc *dc, int primary_otg_inst,
 		dc->hwss.get_underflow_debug_data(dc, tg, out_data);
 }
 
+void dc_get_power_feature_status(struct dc *dc, int primary_otg_inst,
+				struct power_features *out_data)
+{
+	out_data->uclk_p_state = dc->current_state->clk_mgr->clks.p_state_change_support;
+	out_data->fams = dc->current_state->bw_ctx.bw.dcn.clk.fw_based_mclk_switching;
+}
+
+bool dc_capture_register_software_state(struct dc *dc, struct dc_register_software_state *state)
+{
+	struct dc_state *context;
+	struct resource_context *res_ctx;
+	int i;
+
+	if (!dc || !dc->current_state || !state) {
+		if (state)
+			state->state_valid = false;
+		return false;
+	}
+
+	/* Initialize the state structure */
+	memset(state, 0, sizeof(struct dc_register_software_state));
+
+	context = dc->current_state;
+	res_ctx = &context->res_ctx;
+
+	/* Count active pipes and streams */
+	state->active_pipe_count = 0;
+	state->active_stream_count = context->stream_count;
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		if (res_ctx->pipe_ctx[i].stream)
+			state->active_pipe_count++;
+	}
+
+	/* Capture HUBP programming state for each pipe */
+	for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+
+		state->hubp[i].valid_stream = false;
+		if (!pipe_ctx->stream)
+			continue;
+
+		state->hubp[i].valid_stream = true;
+
+		/* HUBP register programming variables */
+		if (pipe_ctx->stream_res.tg)
+			state->hubp[i].vtg_sel = pipe_ctx->stream_res.tg->inst;
+
+		state->hubp[i].hubp_clock_enable = (pipe_ctx->plane_res.hubp != NULL) ? 1 : 0;
+
+		state->hubp[i].valid_plane_state = false;
+		if (pipe_ctx->plane_state) {
+			state->hubp[i].valid_plane_state = true;
+			state->hubp[i].surface_pixel_format = pipe_ctx->plane_state->format;
+			state->hubp[i].rotation_angle = pipe_ctx->plane_state->rotation;
+			state->hubp[i].h_mirror_en = pipe_ctx->plane_state->horizontal_mirror ? 1 : 0;
+
+			/* Surface size */
+			if (pipe_ctx->plane_state->plane_size.surface_size.width > 0) {
+				state->hubp[i].surface_size_width = pipe_ctx->plane_state->plane_size.surface_size.width;
+				state->hubp[i].surface_size_height = pipe_ctx->plane_state->plane_size.surface_size.height;
+			}
+
+			/* Viewport dimensions from scaler data */
+			if (pipe_ctx->plane_state->src_rect.width > 0) {
+				state->hubp[i].pri_viewport_width = pipe_ctx->plane_state->src_rect.width;
+				state->hubp[i].pri_viewport_height = pipe_ctx->plane_state->src_rect.height;
+				state->hubp[i].pri_viewport_x_start = pipe_ctx->plane_state->src_rect.x;
+				state->hubp[i].pri_viewport_y_start = pipe_ctx->plane_state->src_rect.y;
+			}
+
+			/* DCC settings */
+			state->hubp[i].surface_dcc_en = (pipe_ctx->plane_state->dcc.enable) ? 1 : 0;
+			state->hubp[i].surface_dcc_ind_64b_blk = pipe_ctx->plane_state->dcc.independent_64b_blks;
+			state->hubp[i].surface_dcc_ind_128b_blk = pipe_ctx->plane_state->dcc.dcc_ind_blk;
+
+			/* Surface pitch */
+			state->hubp[i].surface_pitch = pipe_ctx->plane_state->plane_size.surface_pitch;
+			state->hubp[i].meta_pitch = pipe_ctx->plane_state->dcc.meta_pitch;
+			state->hubp[i].chroma_pitch = pipe_ctx->plane_state->plane_size.chroma_pitch;
+			state->hubp[i].meta_pitch_c = pipe_ctx->plane_state->dcc.meta_pitch_c;
+
+			/* Surface addresses - primary */
+			state->hubp[i].primary_surface_address_low = pipe_ctx->plane_state->address.grph.addr.low_part;
+			state->hubp[i].primary_surface_address_high = pipe_ctx->plane_state->address.grph.addr.high_part;
+			state->hubp[i].primary_meta_surface_address_low = pipe_ctx->plane_state->address.grph.meta_addr.low_part;
+			state->hubp[i].primary_meta_surface_address_high = pipe_ctx->plane_state->address.grph.meta_addr.high_part;
+
+			/* TMZ settings */
+			state->hubp[i].primary_surface_tmz = pipe_ctx->plane_state->address.tmz_surface;
+			state->hubp[i].primary_meta_surface_tmz = pipe_ctx->plane_state->address.tmz_surface;
+
+			/* Tiling configuration */
+			state->hubp[i].min_dc_gfx_version9 = false;
+			if (pipe_ctx->plane_state->tiling_info.gfxversion >= DcGfxVersion9) {
+				state->hubp[i].min_dc_gfx_version9 = true;
+				state->hubp[i].sw_mode = pipe_ctx->plane_state->tiling_info.gfx9.swizzle;
+				state->hubp[i].num_pipes = pipe_ctx->plane_state->tiling_info.gfx9.num_pipes;
+				state->hubp[i].num_banks = pipe_ctx->plane_state->tiling_info.gfx9.num_banks;
+				state->hubp[i].pipe_interleave = pipe_ctx->plane_state->tiling_info.gfx9.pipe_interleave;
+				state->hubp[i].num_shader_engines = pipe_ctx->plane_state->tiling_info.gfx9.num_shader_engines;
+				state->hubp[i].num_rb_per_se = pipe_ctx->plane_state->tiling_info.gfx9.num_rb_per_se;
+				state->hubp[i].num_pkrs = pipe_ctx->plane_state->tiling_info.gfx9.num_pkrs;
+			}
+		}
+
+		/* DML Request Size Configuration */
+		if (pipe_ctx->rq_regs.rq_regs_l.chunk_size > 0) {
+			state->hubp[i].rq_chunk_size = pipe_ctx->rq_regs.rq_regs_l.chunk_size;
+			state->hubp[i].rq_min_chunk_size = pipe_ctx->rq_regs.rq_regs_l.min_chunk_size;
+			state->hubp[i].rq_meta_chunk_size = pipe_ctx->rq_regs.rq_regs_l.meta_chunk_size;
+			state->hubp[i].rq_min_meta_chunk_size = pipe_ctx->rq_regs.rq_regs_l.min_meta_chunk_size;
+			state->hubp[i].rq_dpte_group_size = pipe_ctx->rq_regs.rq_regs_l.dpte_group_size;
+			state->hubp[i].rq_mpte_group_size = pipe_ctx->rq_regs.rq_regs_l.mpte_group_size;
+			state->hubp[i].rq_swath_height_l = pipe_ctx->rq_regs.rq_regs_l.swath_height;
+			state->hubp[i].rq_pte_row_height_l = pipe_ctx->rq_regs.rq_regs_l.pte_row_height_linear;
+		}
+
+		/* Chroma request size configuration */
+		if (pipe_ctx->rq_regs.rq_regs_c.chunk_size > 0) {
+			state->hubp[i].rq_chunk_size_c = pipe_ctx->rq_regs.rq_regs_c.chunk_size;
+			state->hubp[i].rq_min_chunk_size_c = pipe_ctx->rq_regs.rq_regs_c.min_chunk_size;
+			state->hubp[i].rq_meta_chunk_size_c = pipe_ctx->rq_regs.rq_regs_c.meta_chunk_size;
+			state->hubp[i].rq_min_meta_chunk_size_c = pipe_ctx->rq_regs.rq_regs_c.min_meta_chunk_size;
+			state->hubp[i].rq_dpte_group_size_c = pipe_ctx->rq_regs.rq_regs_c.dpte_group_size;
+			state->hubp[i].rq_mpte_group_size_c = pipe_ctx->rq_regs.rq_regs_c.mpte_group_size;
+			state->hubp[i].rq_swath_height_c = pipe_ctx->rq_regs.rq_regs_c.swath_height;
+			state->hubp[i].rq_pte_row_height_c = pipe_ctx->rq_regs.rq_regs_c.pte_row_height_linear;
+		}
+
+		/* DML expansion modes */
+		state->hubp[i].drq_expansion_mode = pipe_ctx->rq_regs.drq_expansion_mode;
+		state->hubp[i].prq_expansion_mode = pipe_ctx->rq_regs.prq_expansion_mode;
+		state->hubp[i].mrq_expansion_mode = pipe_ctx->rq_regs.mrq_expansion_mode;
+		state->hubp[i].crq_expansion_mode = pipe_ctx->rq_regs.crq_expansion_mode;
+
+		/* DML DLG parameters - nominal */
+		state->hubp[i].dst_y_per_vm_vblank = pipe_ctx->dlg_regs.dst_y_per_vm_vblank;
+		state->hubp[i].dst_y_per_row_vblank = pipe_ctx->dlg_regs.dst_y_per_row_vblank;
+		state->hubp[i].dst_y_per_vm_flip = pipe_ctx->dlg_regs.dst_y_per_vm_flip;
+		state->hubp[i].dst_y_per_row_flip = pipe_ctx->dlg_regs.dst_y_per_row_flip;
+
+		/* DML prefetch settings */
+		state->hubp[i].dst_y_prefetch = pipe_ctx->dlg_regs.dst_y_prefetch;
+		state->hubp[i].vratio_prefetch = pipe_ctx->dlg_regs.vratio_prefetch;
+		state->hubp[i].vratio_prefetch_c = pipe_ctx->dlg_regs.vratio_prefetch_c;
+
+		/* TTU parameters */
+		state->hubp[i].qos_level_low_wm = pipe_ctx->ttu_regs.qos_level_low_wm;
+		state->hubp[i].qos_level_high_wm = pipe_ctx->ttu_regs.qos_level_high_wm;
+		state->hubp[i].qos_level_flip = pipe_ctx->ttu_regs.qos_level_flip;
+		state->hubp[i].min_ttu_vblank = pipe_ctx->ttu_regs.min_ttu_vblank;
+	}
+
+	/* Capture HUBBUB programming state */
+	if (dc->res_pool->hubbub) {
+		/* Individual DET buffer sizes - software state variables that program DET registers */
+		for (i = 0; i < 4 && i < dc->res_pool->pipe_count; i++) {
+			uint32_t det_size = res_ctx->pipe_ctx[i].det_buffer_size_kb;
+			switch (i) {
+			case 0:
+				state->hubbub.det0_size = det_size;
+				break;
+			case 1:
+				state->hubbub.det1_size = det_size;
+				break;
+			case 2:
+				state->hubbub.det2_size = det_size;
+				break;
+			case 3:
+				state->hubbub.det3_size = det_size;
+				break;
+			}
+		}
+
+		/* Compression buffer configuration - software state that programs COMPBUF_SIZE register */
+		// TODO: Handle logic for legacy DCN pre-DCN401
+		state->hubbub.compbuf_size = context->bw_ctx.bw.dcn.arb_regs.compbuf_size;
+	}
+
+	/* Capture DPP programming state for each pipe */
+	for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+
+		if (!pipe_ctx->stream)
+			continue;
+
+		state->dpp[i].dpp_clock_enable = (pipe_ctx->plane_res.dpp != NULL) ? 1 : 0;
+
+		if (pipe_ctx->plane_state && pipe_ctx->plane_res.scl_data.recout.width > 0) {
+			/* Access dscl_prog_data directly - this contains the actual software state used for register programming */
+			struct dscl_prog_data *dscl_data = &pipe_ctx->plane_res.scl_data.dscl_prog_data;
+
+			/* Recout (Rectangle of Interest) configuration - software state that programs RECOUT registers */
+			state->dpp[i].recout_start_x = dscl_data->recout.x;
+			state->dpp[i].recout_start_y = dscl_data->recout.y;
+			state->dpp[i].recout_width = dscl_data->recout.width;
+			state->dpp[i].recout_height = dscl_data->recout.height;
+
+			/* MPC (Multiple Pipe/Plane Combiner) size - software state that programs MPC_SIZE registers */
+			state->dpp[i].mpc_width = dscl_data->mpc_size.width;
+			state->dpp[i].mpc_height = dscl_data->mpc_size.height;
+
+			/* DSCL mode - software state that programs SCL_MODE registers */
+			state->dpp[i].dscl_mode = dscl_data->dscl_mode;
+
+			/* Scaler ratios - software state that programs scale ratio registers (use actual programmed ratios) */
+			state->dpp[i].horz_ratio_int = dscl_data->ratios.h_scale_ratio >> 19; // Extract integer part from programmed ratio
+			state->dpp[i].vert_ratio_int = dscl_data->ratios.v_scale_ratio >> 19; // Extract integer part from programmed ratio
+
+			/* Basic scaler taps - software state that programs tap control registers (use actual programmed taps) */
+			state->dpp[i].h_taps = dscl_data->taps.h_taps + 1; // dscl_prog_data.taps stores (taps - 1), so add 1 back
+			state->dpp[i].v_taps = dscl_data->taps.v_taps + 1; // dscl_prog_data.taps stores (taps - 1), so add 1 back
+		}
+	}
+
+	/* Capture essential clock state for underflow analysis */
+	if (dc->clk_mgr && dc->clk_mgr->clks.dispclk_khz > 0) {
+		/* Core display clocks affecting bandwidth and timing */
+		state->dccg.dispclk_khz = dc->clk_mgr->clks.dispclk_khz;
+
+		/* Per-pipe clock configuration - only capture what's essential */
+		for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+			struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+			if (pipe_ctx->stream) {
+				/* Essential clocks that directly affect underflow risk */
+				state->dccg.dppclk_khz[i] = dc->clk_mgr->clks.dppclk_khz;
+				state->dccg.pixclk_khz[i] = pipe_ctx->stream->timing.pix_clk_100hz / 10;
+				state->dccg.dppclk_enable[i] = 1;
+
+				/* DP stream clock only for DP signals */
+				if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT ||
+						pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
+					state->dccg.dpstreamclk_enable[i] = 1;
+				} else {
+					state->dccg.dpstreamclk_enable[i] = 0;
+				}
+			} else {
+				/* Inactive pipe - no clocks */
+				state->dccg.dppclk_khz[i] = 0;
+				state->dccg.pixclk_khz[i] = 0;
+				state->dccg.dppclk_enable[i] = 0;
+				if (i < 4) {
+					state->dccg.dpstreamclk_enable[i] = 0;
+				}
+			}
+		}
+
+		/* DSC clock state - only when actually using DSC */
+		for (i = 0; i < MAX_PIPES; i++) {
+			struct pipe_ctx *pipe_ctx = (i < dc->res_pool->pipe_count) ? &res_ctx->pipe_ctx[i] : NULL;
+			if (pipe_ctx && pipe_ctx->stream && pipe_ctx->stream->timing.dsc_cfg.num_slices_h > 0) {
+				state->dccg.dscclk_khz[i] = 400000; /* Typical DSC clock frequency */
+			} else {
+				state->dccg.dscclk_khz[i] = 0;
+			}
+		}
+
+		/* SYMCLK32 LE Control - only the essential HPO state for underflow analysis */
+		for (i = 0; i < 2; i++) {
+			state->dccg.symclk32_le_enable[i] = 0; /* Default: disabled */
+		}
+
+	}
+
+	/* Capture essential DSC configuration for underflow analysis */
+	for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+
+		if (pipe_ctx->stream && pipe_ctx->stream->timing.dsc_cfg.num_slices_h > 0) {
+			/* DSC is enabled - capture essential configuration */
+			state->dsc[i].dsc_clock_enable = 1;
+
+			/* DSC configuration affecting bandwidth and timing */
+			struct dc_dsc_config *dsc_cfg = &pipe_ctx->stream->timing.dsc_cfg;
+			state->dsc[i].dsc_num_slices_h = dsc_cfg->num_slices_h;
+			state->dsc[i].dsc_num_slices_v = dsc_cfg->num_slices_v;
+			state->dsc[i].dsc_bits_per_pixel = dsc_cfg->bits_per_pixel;
+
+			/* OPP pipe source for DSC forwarding */
+			if (pipe_ctx->stream_res.opp) {
+				state->dsc[i].dscrm_dsc_forward_enable = 1;
+				state->dsc[i].dscrm_dsc_opp_pipe_source = pipe_ctx->stream_res.opp->inst;
+			} else {
+				state->dsc[i].dscrm_dsc_forward_enable = 0;
+				state->dsc[i].dscrm_dsc_opp_pipe_source = 0;
+			}
+		} else {
+			/* DSC not enabled - clear all fields */
+			memset(&state->dsc[i], 0, sizeof(state->dsc[i]));
+		}
+	}
+
+	/* Capture MPC programming state - comprehensive register field coverage */
+	for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+
+		if (pipe_ctx->plane_state && pipe_ctx->stream) {
+			struct dc_plane_state *plane_state = pipe_ctx->plane_state;
+
+			/* MPCC blending tree and mode control - capture actual blend configuration */
+			state->mpc.mpcc_mode[i] = (plane_state->blend_tf.type != TF_TYPE_BYPASS) ? 1 : 0;
+			state->mpc.mpcc_alpha_blend_mode[i] = plane_state->per_pixel_alpha ? 1 : 0;
+			state->mpc.mpcc_alpha_multiplied_mode[i] = plane_state->pre_multiplied_alpha ? 1 : 0;
+			state->mpc.mpcc_blnd_active_overlap_only[i] = 0; /* Default - no overlap restriction */
+			state->mpc.mpcc_global_alpha[i] = plane_state->global_alpha_value;
+			state->mpc.mpcc_global_gain[i] = plane_state->global_alpha ? 255 : 0;
+			state->mpc.mpcc_bg_bpc[i] = 8; /* Standard 8-bit background */
+			state->mpc.mpcc_bot_gain_mode[i] = 0; /* Standard gain mode */
+
+			/* MPCC blending tree connections - capture tree topology */
+			if (pipe_ctx->bottom_pipe) {
+				state->mpc.mpcc_bot_sel[i] = pipe_ctx->bottom_pipe->pipe_idx;
+			} else {
+				state->mpc.mpcc_bot_sel[i] = 0xF; /* No bottom connection */
+			}
+			state->mpc.mpcc_top_sel[i] = pipe_ctx->pipe_idx; /* This pipe's DPP ID */
+
+			/* MPCC output gamma control - capture gamma programming */
+			if (plane_state->gamma_correction.type != GAMMA_CS_TFM_1D && plane_state->gamma_correction.num_entries > 0) {
+				state->mpc.mpcc_ogam_mode[i] = 1; /* Gamma enabled */
+				state->mpc.mpcc_ogam_select[i] = 0; /* Bank A selection */
+				state->mpc.mpcc_ogam_pwl_disable[i] = 0; /* PWL enabled */
+			} else {
+				state->mpc.mpcc_ogam_mode[i] = 0; /* Bypass mode */
+				state->mpc.mpcc_ogam_select[i] = 0;
+				state->mpc.mpcc_ogam_pwl_disable[i] = 1; /* PWL disabled */
+			}
+
+			/* MPCC pipe assignment and operational status */
+			if (pipe_ctx->stream_res.opp) {
+				state->mpc.mpcc_opp_id[i] = pipe_ctx->stream_res.opp->inst;
+			} else {
+				state->mpc.mpcc_opp_id[i] = 0xF; /* No OPP assignment */
+			}
+
+			/* MPCC status indicators - active pipe state */
+			state->mpc.mpcc_idle[i] = 0; /* Active pipe - not idle */
+			state->mpc.mpcc_busy[i] = 1; /* Active pipe - busy processing */
+
+		} else {
+			/* Pipe not active - set disabled/idle state for all fields */
+			state->mpc.mpcc_mode[i] = 0;
+			state->mpc.mpcc_alpha_blend_mode[i] = 0;
+			state->mpc.mpcc_alpha_multiplied_mode[i] = 0;
+			state->mpc.mpcc_blnd_active_overlap_only[i] = 0;
+			state->mpc.mpcc_global_alpha[i] = 0;
+			state->mpc.mpcc_global_gain[i] = 0;
+			state->mpc.mpcc_bg_bpc[i] = 0;
+			state->mpc.mpcc_bot_gain_mode[i] = 0;
+			state->mpc.mpcc_bot_sel[i] = 0xF; /* No bottom connection */
+			state->mpc.mpcc_top_sel[i] = 0xF; /* No top connection */
+			state->mpc.mpcc_ogam_mode[i] = 0; /* Bypass */
+			state->mpc.mpcc_ogam_select[i] = 0;
+			state->mpc.mpcc_ogam_pwl_disable[i] = 1; /* PWL disabled */
+			state->mpc.mpcc_opp_id[i] = 0xF; /* No OPP assignment */
+			state->mpc.mpcc_idle[i] = 1; /* Idle */
+			state->mpc.mpcc_busy[i] = 0; /* Not busy */
+		}
+	}
+
+	/* Capture OPP programming state for each pipe - comprehensive register field coverage */
+	for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+
+		if (!pipe_ctx->stream)
+			continue;
+
+		if (pipe_ctx->stream_res.opp) {
+			struct dc_crtc_timing *timing = &pipe_ctx->stream->timing;
+
+			/* OPP Pipe Control */
+			state->opp[i].opp_pipe_clock_enable = 1; /* Active pipe has clock enabled */
+
+			/* Display Pattern Generator (DPG) Control - 19 fields */
+			if (pipe_ctx->stream->test_pattern.type != DP_TEST_PATTERN_VIDEO_MODE) {
+				state->opp[i].dpg_enable = 1;
+			} else {
+				/* Video mode - DPG disabled */
+				state->opp[i].dpg_enable = 0;
+			}
+
+			/* Format Control (FMT) - 18 fields */
+			state->opp[i].fmt_pixel_encoding = timing->pixel_encoding;
+
+			/* Chroma subsampling mode based on pixel encoding */
+			if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) {
+				state->opp[i].fmt_subsampling_mode = 1; /* 4:2:0 subsampling */
+			} else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) {
+				state->opp[i].fmt_subsampling_mode = 2; /* 4:2:2 subsampling */
+			} else {
+				state->opp[i].fmt_subsampling_mode = 0; /* No subsampling (4:4:4) */
+			}
+
+			state->opp[i].fmt_cbcr_bit_reduction_bypass = (timing->pixel_encoding == PIXEL_ENCODING_RGB) ? 1 : 0;
+			state->opp[i].fmt_stereosync_override = (timing->timing_3d_format != TIMING_3D_FORMAT_NONE) ? 1 : 0;
+
+			/* Dithering control based on bit depth */
+			if (timing->display_color_depth < COLOR_DEPTH_121212) {
+				state->opp[i].fmt_spatial_dither_frame_counter_max = 15; /* Typical frame counter max */
+				state->opp[i].fmt_spatial_dither_frame_counter_bit_swap = 0; /* No bit swapping */
+				state->opp[i].fmt_spatial_dither_enable = 1;
+				state->opp[i].fmt_spatial_dither_mode = 0; /* Spatial dithering mode */
+				state->opp[i].fmt_spatial_dither_depth = timing->display_color_depth;
+				state->opp[i].fmt_temporal_dither_enable = 0; /* Spatial dithering preferred */
+			} else {
+				state->opp[i].fmt_spatial_dither_frame_counter_max = 0;
+				state->opp[i].fmt_spatial_dither_frame_counter_bit_swap = 0;
+				state->opp[i].fmt_spatial_dither_enable = 0;
+				state->opp[i].fmt_spatial_dither_mode = 0;
+				state->opp[i].fmt_spatial_dither_depth = 0;
+				state->opp[i].fmt_temporal_dither_enable = 0;
+			}
+
+			/* Truncation control for bit depth reduction */
+			if (timing->display_color_depth < COLOR_DEPTH_121212) {
+				state->opp[i].fmt_truncate_enable = 1;
+				state->opp[i].fmt_truncate_depth = timing->display_color_depth;
+				state->opp[i].fmt_truncate_mode = 0; /* Round mode */
+			} else {
+				state->opp[i].fmt_truncate_enable = 0;
+				state->opp[i].fmt_truncate_depth = 0;
+				state->opp[i].fmt_truncate_mode = 0;
+			}
+
+			/* Data clamping control */
+			state->opp[i].fmt_clamp_data_enable = 1; /* Clamping typically enabled */
+			state->opp[i].fmt_clamp_color_format = timing->pixel_encoding;
+
+			/* Dynamic expansion for limited range content */
+			if (timing->pixel_encoding != PIXEL_ENCODING_RGB) {
+				state->opp[i].fmt_dynamic_exp_enable = 1; /* YCbCr typically needs expansion */
+				state->opp[i].fmt_dynamic_exp_mode = 0; /* Standard expansion */
+			} else {
+				state->opp[i].fmt_dynamic_exp_enable = 0; /* RGB typically full range */
+				state->opp[i].fmt_dynamic_exp_mode = 0;
+			}
+
+			/* Legacy field for compatibility */
+			state->opp[i].fmt_bit_depth_control = timing->display_color_depth;
+
+			/* Output Buffer (OPPBUF) Control - 6 fields */
+			state->opp[i].oppbuf_active_width = timing->h_addressable;
+			state->opp[i].oppbuf_pixel_repetition = 0; /* No pixel repetition by default */
+
+			/* Multi-Stream Output (MSO) / ODM segmentation */
+			if (pipe_ctx->next_odm_pipe) {
+				state->opp[i].oppbuf_display_segmentation = 1; /* Segmented display */
+				state->opp[i].oppbuf_overlap_pixel_num = 0; /* ODM overlap pixels */
+			} else {
+				state->opp[i].oppbuf_display_segmentation = 0; /* Single segment */
+				state->opp[i].oppbuf_overlap_pixel_num = 0;
+			}
+
+			/* 3D/Stereo control */
+			if (timing->timing_3d_format != TIMING_3D_FORMAT_NONE) {
+				state->opp[i].oppbuf_3d_vact_space1_size = 30; /* Typical stereo blanking */
+				state->opp[i].oppbuf_3d_vact_space2_size = 30;
+			} else {
+				state->opp[i].oppbuf_3d_vact_space1_size = 0;
+				state->opp[i].oppbuf_3d_vact_space2_size = 0;
+			}
+
+			/* DSC Forward Config - 3 fields */
+			if (timing->dsc_cfg.num_slices_h > 0) {
+				state->opp[i].dscrm_dsc_forward_enable = 1;
+				state->opp[i].dscrm_dsc_opp_pipe_source = pipe_ctx->stream_res.opp->inst;
+				state->opp[i].dscrm_dsc_forward_enable_status = 1; /* Status follows enable */
+			} else {
+				state->opp[i].dscrm_dsc_forward_enable = 0;
+				state->opp[i].dscrm_dsc_opp_pipe_source = 0;
+				state->opp[i].dscrm_dsc_forward_enable_status = 0;
+			}
+		} else {
+			/* No OPP resource - set all fields to disabled state */
+			memset(&state->opp[i], 0, sizeof(state->opp[i]));
+		}
+	}
+
+	/* Capture OPTC programming state for each pipe - comprehensive register field coverage */
+	for (i = 0; i < MAX_PIPES && i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[i];
+
+		if (!pipe_ctx->stream)
+			continue;
+
+		if (pipe_ctx->stream_res.tg) {
+			struct dc_crtc_timing *timing = &pipe_ctx->stream->timing;
+
+			state->optc[i].otg_master_inst = pipe_ctx->stream_res.tg->inst;
+
+			/* OTG_CONTROL register - 5 fields */
+			state->optc[i].otg_master_enable = 1; /* Active stream */
+			state->optc[i].otg_disable_point_cntl = 0; /* Normal operation */
+			state->optc[i].otg_start_point_cntl = 0; /* Normal start */
+			state->optc[i].otg_field_number_cntl = (timing->flags.INTERLACE) ? 1 : 0;
+			state->optc[i].otg_out_mux = 0; /* Direct output */
+
+			/* OTG Horizontal Timing - 7 fields */
+			state->optc[i].otg_h_total = timing->h_total;
+			state->optc[i].otg_h_blank_start = timing->h_addressable;
+			state->optc[i].otg_h_blank_end = timing->h_total - timing->h_front_porch;
+			state->optc[i].otg_h_sync_start = timing->h_addressable + timing->h_front_porch;
+			state->optc[i].otg_h_sync_end = timing->h_addressable + timing->h_front_porch + timing->h_sync_width;
+			state->optc[i].otg_h_sync_polarity = timing->flags.HSYNC_POSITIVE_POLARITY ? 0 : 1;
+			state->optc[i].otg_h_timing_div_mode = (pipe_ctx->next_odm_pipe) ? 1 : 0; /* ODM divide mode */
+
+			/* OTG Vertical Timing - 7 fields */
+			state->optc[i].otg_v_total = timing->v_total;
+			state->optc[i].otg_v_blank_start = timing->v_addressable;
+			state->optc[i].otg_v_blank_end = timing->v_total - timing->v_front_porch;
+			state->optc[i].otg_v_sync_start = timing->v_addressable + timing->v_front_porch;
+			state->optc[i].otg_v_sync_end = timing->v_addressable + timing->v_front_porch + timing->v_sync_width;
+			state->optc[i].otg_v_sync_polarity = timing->flags.VSYNC_POSITIVE_POLARITY ? 0 : 1;
+			state->optc[i].otg_v_sync_mode = 0; /* Normal sync mode */
+
+			/* Initialize remaining core fields with appropriate defaults */
+			// TODO: Update logic for accurate vtotal min/max
+			state->optc[i].otg_v_total_max = timing->v_total + 100; /* Typical DRR range */
+			state->optc[i].otg_v_total_min = timing->v_total - 50;
+			state->optc[i].otg_v_total_mid = timing->v_total;
+
+			/* ODM configuration */
+			// TODO: Update logic to have complete ODM mappings (e.g. 3:1 and 4:1) stored in single pipe
+			if (pipe_ctx->next_odm_pipe) {
+				state->optc[i].optc_seg0_src_sel = pipe_ctx->stream_res.opp ? pipe_ctx->stream_res.opp->inst : 0;
+				state->optc[i].optc_seg1_src_sel = pipe_ctx->next_odm_pipe->stream_res.opp ? pipe_ctx->next_odm_pipe->stream_res.opp->inst : 0;
+				state->optc[i].optc_num_of_input_segment = 1; /* 2 segments - 1 */
+			} else {
+				state->optc[i].optc_seg0_src_sel = pipe_ctx->stream_res.opp ? pipe_ctx->stream_res.opp->inst : 0;
+				state->optc[i].optc_seg1_src_sel = 0;
+				state->optc[i].optc_num_of_input_segment = 0; /* Single segment */
+			}
+
+			/* DSC configuration */
+			if (timing->dsc_cfg.num_slices_h > 0) {
+				state->optc[i].optc_dsc_mode = 1; /* DSC enabled */
+				state->optc[i].optc_dsc_bytes_per_pixel = timing->dsc_cfg.bits_per_pixel / 16; /* Convert to bytes */
+				state->optc[i].optc_dsc_slice_width = timing->h_addressable / timing->dsc_cfg.num_slices_h;
+			} else {
+				state->optc[i].optc_dsc_mode = 0;
+				state->optc[i].optc_dsc_bytes_per_pixel = 0;
+				state->optc[i].optc_dsc_slice_width = 0;
+			}
+
+			/* Essential control fields */
+			state->optc[i].otg_stereo_enable = (timing->timing_3d_format != TIMING_3D_FORMAT_NONE) ? 1 : 0;
+			state->optc[i].otg_interlace_enable = timing->flags.INTERLACE ? 1 : 0;
+			state->optc[i].otg_clock_enable = 1; /* OTG clock enabled */
+			state->optc[i].vtg0_enable = 1; /* VTG enabled for timing generation */
+
+			/* Initialize other key fields to defaults */
+			state->optc[i].optc_input_pix_clk_en = 1;
+			state->optc[i].optc_segment_width = (pipe_ctx->next_odm_pipe) ? (timing->h_addressable / 2) : timing->h_addressable;
+			state->optc[i].otg_vready_offset = 1;
+			state->optc[i].otg_vstartup_start = timing->v_addressable + 10;
+			state->optc[i].otg_vupdate_offset = 0;
+			state->optc[i].otg_vupdate_width = 5;
+		} else {
+			/* No timing generator resource - initialize all fields to 0 */
+			memset(&state->optc[i], 0, sizeof(state->optc[i]));
+		}
+	}
+
+	state->state_valid = true;
+	return true;
+}
+
 void dc_log_preos_dmcub_info(const struct dc *dc)
 {
 	dc_dmub_srv_log_preos_dmcub_info(dc->ctx->dmub_srv);
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
index f95cb0cf4b8a..e2763b60482a 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
@@ -38,6 +38,7 @@
 #include "dccg.h"
 #include "abm.h"
 #include "dcn10/dcn10_hubbub.h"
+#include "dce/dmub_hw_lock_mgr.h"
 
 #define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0]))
 #define MAX_NUM_MCACHE 8
@@ -264,7 +265,7 @@ void color_space_to_black_color(
 			black_color_format[BLACK_COLOR_FORMAT_RGB_LIMITED];
 		break;
 
-	/**
+	/*
 	 * Remove default and add case for all color space
 	 * so when we forget to add new color space
 	 * compiler will give a warning
@@ -764,7 +765,9 @@ void hwss_build_fast_sequence(struct dc *dc,
 	if (dc->hwss.dmub_hw_control_lock_fast) {
 		block_sequence[*num_steps].params.dmub_hw_control_lock_fast_params.dc = dc;
 		block_sequence[*num_steps].params.dmub_hw_control_lock_fast_params.lock = true;
-		block_sequence[*num_steps].params.dmub_hw_control_lock_fast_params.is_required = dc_state_is_fams2_in_use(dc, context);
+		block_sequence[*num_steps].params.dmub_hw_control_lock_fast_params.is_required =
+			dc_state_is_fams2_in_use(dc, context) ||
+			dmub_hw_lock_mgr_does_link_require_lock(dc, stream->link);
 		block_sequence[*num_steps].func = DMUB_HW_CONTROL_LOCK_FAST;
 		(*num_steps)++;
 	}
@@ -1290,6 +1293,9 @@ void hwss_execute_sequence(struct dc *dc,
 		case HUBP_MEM_PROGRAM_VIEWPORT:
 			hwss_hubp_mem_program_viewport(params);
 			break;
+		case ABORT_CURSOR_OFFLOAD_UPDATE:
+			hwss_abort_cursor_offload_update(params);
+			break;
 		case SET_CURSOR_ATTRIBUTE:
 			hwss_set_cursor_attribute(params);
 			break;
@@ -1315,7 +1321,7 @@ void hwss_execute_sequence(struct dc *dc,
 	}
 }
 
-/**
+/*
  * Helper function to add OPTC pipe control lock to block sequence
  */
 void hwss_add_optc_pipe_control_lock(struct block_sequence_state *seq_state,
@@ -1332,7 +1338,7 @@ void hwss_add_optc_pipe_control_lock(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP set flip control GSL to block sequence
  */
 void hwss_add_hubp_set_flip_control_gsl(struct block_sequence_state *seq_state,
@@ -1347,7 +1353,7 @@ void hwss_add_hubp_set_flip_control_gsl(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP program triplebuffer to block sequence
  */
 void hwss_add_hubp_program_triplebuffer(struct block_sequence_state *seq_state,
@@ -1364,7 +1370,7 @@ void hwss_add_hubp_program_triplebuffer(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP update plane address to block sequence
  */
 void hwss_add_hubp_update_plane_addr(struct block_sequence_state *seq_state,
@@ -1379,7 +1385,7 @@ void hwss_add_hubp_update_plane_addr(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add DPP set input transfer function to block sequence
  */
 void hwss_add_dpp_set_input_transfer_func(struct block_sequence_state *seq_state,
@@ -1396,7 +1402,7 @@ void hwss_add_dpp_set_input_transfer_func(struct block_sequence_state *seq_state
 	}
 }
 
-/**
+/*
  * Helper function to add DPP program gamut remap to block sequence
  */
 void hwss_add_dpp_program_gamut_remap(struct block_sequence_state *seq_state,
@@ -1409,7 +1415,7 @@ void hwss_add_dpp_program_gamut_remap(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add DPP program bias and scale to block sequence
  */
 void hwss_add_dpp_program_bias_and_scale(struct block_sequence_state *seq_state, struct pipe_ctx *pipe_ctx)
@@ -1421,7 +1427,7 @@ void hwss_add_dpp_program_bias_and_scale(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add OPTC program manual trigger to block sequence
  */
 void hwss_add_optc_program_manual_trigger(struct block_sequence_state *seq_state,
@@ -1434,7 +1440,7 @@ void hwss_add_optc_program_manual_trigger(struct block_sequence_state *seq_state
 	}
 }
 
-/**
+/*
  * Helper function to add DPP set output transfer function to block sequence
  */
 void hwss_add_dpp_set_output_transfer_func(struct block_sequence_state *seq_state,
@@ -1451,7 +1457,7 @@ void hwss_add_dpp_set_output_transfer_func(struct block_sequence_state *seq_stat
 	}
 }
 
-/**
+/*
  * Helper function to add MPC update visual confirm to block sequence
  */
 void hwss_add_mpc_update_visual_confirm(struct block_sequence_state *seq_state,
@@ -1468,7 +1474,7 @@ void hwss_add_mpc_update_visual_confirm(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add MPC power on MPC mem PWR to block sequence
  */
 void hwss_add_mpc_power_on_mpc_mem_pwr(struct block_sequence_state *seq_state,
@@ -1485,7 +1491,7 @@ void hwss_add_mpc_power_on_mpc_mem_pwr(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add MPC set output CSC to block sequence
  */
 void hwss_add_mpc_set_output_csc(struct block_sequence_state *seq_state,
@@ -1504,7 +1510,7 @@ void hwss_add_mpc_set_output_csc(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add MPC set OCSC default to block sequence
  */
 void hwss_add_mpc_set_ocsc_default(struct block_sequence_state *seq_state,
@@ -1523,7 +1529,7 @@ void hwss_add_mpc_set_ocsc_default(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add DMUB send DMCUB command to block sequence
  */
 void hwss_add_dmub_send_dmcub_cmd(struct block_sequence_state *seq_state,
@@ -1540,7 +1546,7 @@ void hwss_add_dmub_send_dmcub_cmd(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add DMUB SubVP save surface address to block sequence
  */
 void hwss_add_dmub_subvp_save_surf_addr(struct block_sequence_state *seq_state,
@@ -1557,7 +1563,7 @@ void hwss_add_dmub_subvp_save_surf_addr(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP wait for DCC meta propagation to block sequence
  */
 void hwss_add_hubp_wait_for_dcc_meta_prop(struct block_sequence_state *seq_state,
@@ -1572,7 +1578,7 @@ void hwss_add_hubp_wait_for_dcc_meta_prop(struct block_sequence_state *seq_state
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP wait pipe read start to block sequence
  */
 void hwss_add_hubp_wait_pipe_read_start(struct block_sequence_state *seq_state,
@@ -1585,7 +1591,7 @@ void hwss_add_hubp_wait_pipe_read_start(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HWS apply update flags for phantom to block sequence
  */
 void hwss_add_hws_apply_update_flags_for_phantom(struct block_sequence_state *seq_state,
@@ -1598,7 +1604,7 @@ void hwss_add_hws_apply_update_flags_for_phantom(struct block_sequence_state *se
 	}
 }
 
-/**
+/*
  * Helper function to add HWS update phantom VP position to block sequence
  */
 void hwss_add_hws_update_phantom_vp_position(struct block_sequence_state *seq_state,
@@ -1615,7 +1621,7 @@ void hwss_add_hws_update_phantom_vp_position(struct block_sequence_state *seq_st
 	}
 }
 
-/**
+/*
  * Helper function to add OPTC set ODM combine to block sequence
  */
 void hwss_add_optc_set_odm_combine(struct block_sequence_state *seq_state,
@@ -1633,7 +1639,7 @@ void hwss_add_optc_set_odm_combine(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add OPTC set ODM bypass to block sequence
  */
 void hwss_add_optc_set_odm_bypass(struct block_sequence_state *seq_state,
@@ -1656,7 +1662,7 @@ void hwss_send_dmcub_cmd(union block_sequence_params *params)
 	dc_wake_and_execute_dmub_cmd(ctx, cmd, wait_type);
 }
 
-/**
+/*
  * Helper function to add TG program global sync to block sequence
  */
 void hwss_add_tg_program_global_sync(struct block_sequence_state *seq_state,
@@ -1679,7 +1685,7 @@ void hwss_add_tg_program_global_sync(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add TG wait for state to block sequence
  */
 void hwss_add_tg_wait_for_state(struct block_sequence_state *seq_state,
@@ -1694,7 +1700,7 @@ void hwss_add_tg_wait_for_state(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add TG set VTG params to block sequence
  */
 void hwss_add_tg_set_vtg_params(struct block_sequence_state *seq_state,
@@ -1711,7 +1717,7 @@ void hwss_add_tg_set_vtg_params(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add TG setup vertical interrupt2 to block sequence
  */
 void hwss_add_tg_setup_vertical_interrupt2(struct block_sequence_state *seq_state,
@@ -1725,7 +1731,7 @@ void hwss_add_tg_setup_vertical_interrupt2(struct block_sequence_state *seq_stat
 	}
 }
 
-/**
+/*
  * Helper function to add DPP set HDR multiplier to block sequence
  */
 void hwss_add_dpp_set_hdr_multiplier(struct block_sequence_state *seq_state,
@@ -1739,7 +1745,7 @@ void hwss_add_dpp_set_hdr_multiplier(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP program DET size to block sequence
  */
 void hwss_add_hubp_program_det_size(struct block_sequence_state *seq_state,
@@ -1782,7 +1788,7 @@ void hwss_add_hubbub_force_pstate_change_control(struct block_sequence_state *se
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP program DET segments to block sequence
  */
 void hwss_add_hubp_program_det_segments(struct block_sequence_state *seq_state,
@@ -1799,7 +1805,7 @@ void hwss_add_hubp_program_det_segments(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add OPP set dynamic expansion to block sequence
  */
 void hwss_add_opp_set_dyn_expansion(struct block_sequence_state *seq_state,
@@ -1818,7 +1824,7 @@ void hwss_add_opp_set_dyn_expansion(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add OPP program FMT to block sequence
  */
 void hwss_add_opp_program_fmt(struct block_sequence_state *seq_state,
@@ -1849,7 +1855,7 @@ void hwss_add_opp_program_left_edge_extra_pixel(struct block_sequence_state *seq
 	}
 }
 
-/**
+/*
  * Helper function to add ABM set pipe to block sequence
  */
 void hwss_add_abm_set_pipe(struct block_sequence_state *seq_state,
@@ -1864,7 +1870,7 @@ void hwss_add_abm_set_pipe(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add ABM set level to block sequence
  */
 void hwss_add_abm_set_level(struct block_sequence_state *seq_state,
@@ -1879,7 +1885,7 @@ void hwss_add_abm_set_level(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add TG enable CRTC to block sequence
  */
 void hwss_add_tg_enable_crtc(struct block_sequence_state *seq_state,
@@ -1892,7 +1898,7 @@ void hwss_add_tg_enable_crtc(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP wait flip pending to block sequence
  */
 void hwss_add_hubp_wait_flip_pending(struct block_sequence_state *seq_state,
@@ -1909,7 +1915,7 @@ void hwss_add_hubp_wait_flip_pending(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add TG wait double buffer pending to block sequence
  */
 void hwss_add_tg_wait_double_buffer_pending(struct block_sequence_state *seq_state,
@@ -3073,6 +3079,15 @@ void hwss_hubp_mem_program_viewport(union block_sequence_params *params)
 		hubp->funcs->mem_program_viewport(hubp, viewport, viewport_c);
 }
 
+void hwss_abort_cursor_offload_update(union block_sequence_params *params)
+{
+	struct dc *dc = params->abort_cursor_offload_update_params.dc;
+	struct pipe_ctx *pipe_ctx = params->abort_cursor_offload_update_params.pipe_ctx;
+
+	if (dc && dc->hwss.abort_cursor_offload_update)
+		dc->hwss.abort_cursor_offload_update(dc, pipe_ctx);
+}
+
 void hwss_set_cursor_attribute(union block_sequence_params *params)
 {
 	struct dc *dc = params->set_cursor_attribute_params.dc;
@@ -3267,7 +3282,7 @@ void hwss_add_opp_set_disp_pattern_generator(struct block_sequence_state *seq_st
 	}
 }
 
-/**
+/*
  * Helper function to add MPC update blending to block sequence
  */
 void hwss_add_mpc_update_blending(struct block_sequence_state *seq_state,
@@ -3284,7 +3299,7 @@ void hwss_add_mpc_update_blending(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add MPC insert plane to block sequence
  */
 void hwss_add_mpc_insert_plane(struct block_sequence_state *seq_state,
@@ -3309,7 +3324,7 @@ void hwss_add_mpc_insert_plane(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add MPC assert idle MPCC to block sequence
  */
 void hwss_add_mpc_assert_idle_mpcc(struct block_sequence_state *seq_state,
@@ -3324,7 +3339,7 @@ void hwss_add_mpc_assert_idle_mpcc(struct block_sequence_state *seq_state,
 	}
 }
 
-/**
+/*
  * Helper function to add HUBP set blank to block sequence
  */
 void hwss_add_hubp_set_blank(struct block_sequence_state *seq_state,
@@ -3931,6 +3946,18 @@ void hwss_add_hubp_mem_program_viewport(struct block_sequence_state *seq_state,
 	}
 }
 
+void hwss_add_abort_cursor_offload_update(struct block_sequence_state *seq_state,
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx)
+{
+	if (*seq_state->num_steps < MAX_HWSS_BLOCK_SEQUENCE_SIZE) {
+		seq_state->steps[*seq_state->num_steps].func = ABORT_CURSOR_OFFLOAD_UPDATE;
+		seq_state->steps[*seq_state->num_steps].params.abort_cursor_offload_update_params.dc = dc;
+		seq_state->steps[*seq_state->num_steps].params.abort_cursor_offload_update_params.pipe_ctx = pipe_ctx;
+		(*seq_state->num_steps)++;
+	}
+}
+
 void hwss_add_set_cursor_attribute(struct block_sequence_state *seq_state,
 		struct dc *dc,
 		struct pipe_ctx *pipe_ctx)
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 dc0c4065a92c..848c267ef11e 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -99,6 +99,40 @@
 
 #define UNABLE_TO_SPLIT -1
 
+static void capture_pipe_topology_data(struct dc *dc, int plane_idx, int slice_idx, int stream_idx,
+									   int dpp_inst, int opp_inst, int tg_inst, bool is_phantom_pipe)
+{
+	struct pipe_topology_snapshot *current_snapshot = &dc->debug_data.topology_history.snapshots[dc->debug_data.topology_history.current_snapshot_index];
+
+	if (current_snapshot->line_count >= MAX_PIPES)
+		return;
+
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].is_phantom_pipe = is_phantom_pipe;
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].plane_idx = plane_idx;
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].slice_idx = slice_idx;
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].stream_idx = stream_idx;
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].dpp_inst = dpp_inst;
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].opp_inst = opp_inst;
+	current_snapshot->pipe_log_lines[current_snapshot->line_count].tg_inst = tg_inst;
+
+	current_snapshot->line_count++;
+}
+
+static void start_new_topology_snapshot(struct dc *dc, struct dc_state *state)
+{
+	// Move to next snapshot slot (circular buffer)
+	dc->debug_data.topology_history.current_snapshot_index = (dc->debug_data.topology_history.current_snapshot_index + 1) % MAX_TOPOLOGY_SNAPSHOTS;
+
+	// Clear the new snapshot
+	struct pipe_topology_snapshot *current_snapshot = &dc->debug_data.topology_history.snapshots[dc->debug_data.topology_history.current_snapshot_index];
+	memset(current_snapshot, 0, sizeof(*current_snapshot));
+
+	// Set metadata
+	current_snapshot->timestamp_us = dm_get_timestamp(dc->ctx);
+	current_snapshot->stream_count = state->stream_count;
+	current_snapshot->phantom_stream_count = state->phantom_stream_count;
+}
+
 enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
 {
 	enum dce_version dc_version = DCE_VERSION_UNKNOWN;
@@ -2311,10 +2345,11 @@ bool resource_is_odm_topology_changed(const struct pipe_ctx *otg_master_a,
 
 static void resource_log_pipe(struct dc *dc, struct pipe_ctx *pipe,
 		int stream_idx, int slice_idx, int plane_idx, int slice_count,
-		bool is_primary)
+		bool is_primary, bool is_phantom_pipe)
 {
 	DC_LOGGER_INIT(dc->ctx->logger);
 
+	// new format for logging: bit storing code
 	if (slice_idx == 0 && plane_idx == 0 && is_primary) {
 		/* case 0 (OTG master pipe with plane) */
 		DC_LOG_DC(" | plane%d  slice%d  stream%d|",
@@ -2323,6 +2358,10 @@ static void resource_log_pipe(struct dc *dc, struct pipe_ctx *pipe,
 				pipe->plane_res.dpp->inst,
 				pipe->stream_res.opp->inst,
 				pipe->stream_res.tg->inst);
+		capture_pipe_topology_data(dc, plane_idx, slice_idx, stream_idx,
+				pipe->plane_res.dpp->inst,
+				pipe->stream_res.opp->inst,
+				pipe->stream_res.tg->inst, is_phantom_pipe);
 	} else if (slice_idx == 0 && plane_idx == -1) {
 		/* case 1 (OTG master pipe without plane) */
 		DC_LOG_DC(" |         slice%d  stream%d|",
@@ -2331,6 +2370,10 @@ static void resource_log_pipe(struct dc *dc, struct pipe_ctx *pipe,
 				pipe->stream_res.opp->inst,
 				pipe->stream_res.opp->inst,
 				pipe->stream_res.tg->inst);
+		capture_pipe_topology_data(dc, 0xF, slice_idx, stream_idx,
+				pipe->plane_res.dpp->inst,
+				pipe->stream_res.opp->inst,
+				pipe->stream_res.tg->inst, is_phantom_pipe);
 	} else if (slice_idx != 0 && plane_idx == 0 && is_primary) {
 		/* case 2 (OPP head pipe with plane) */
 		DC_LOG_DC(" | plane%d  slice%d |       |",
@@ -2338,27 +2381,43 @@ static void resource_log_pipe(struct dc *dc, struct pipe_ctx *pipe,
 		DC_LOG_DC(" |DPP%d----OPP%d----|       |",
 				pipe->plane_res.dpp->inst,
 				pipe->stream_res.opp->inst);
+		capture_pipe_topology_data(dc, plane_idx, slice_idx, stream_idx,
+				pipe->plane_res.dpp->inst,
+				pipe->stream_res.opp->inst,
+				pipe->stream_res.tg->inst, is_phantom_pipe);
 	} else if (slice_idx != 0 && plane_idx == -1) {
 		/* case 3 (OPP head pipe without plane) */
 		DC_LOG_DC(" |         slice%d |       |", slice_idx);
 		DC_LOG_DC(" |DPG%d----OPP%d----|       |",
 				pipe->plane_res.dpp->inst,
 				pipe->stream_res.opp->inst);
+		capture_pipe_topology_data(dc, 0xF, slice_idx, stream_idx,
+				pipe->plane_res.dpp->inst,
+				pipe->stream_res.opp->inst,
+				pipe->stream_res.tg->inst, is_phantom_pipe);
 	} else if (slice_idx == slice_count - 1) {
 		/* case 4 (DPP pipe in last slice) */
 		DC_LOG_DC(" | plane%d |               |", plane_idx);
 		DC_LOG_DC(" |DPP%d----|               |",
 				pipe->plane_res.dpp->inst);
+		capture_pipe_topology_data(dc, plane_idx, slice_idx, stream_idx,
+				pipe->plane_res.dpp->inst,
+				pipe->stream_res.opp->inst,
+				pipe->stream_res.tg->inst, is_phantom_pipe);
 	} else {
 		/* case 5 (DPP pipe not in last slice) */
 		DC_LOG_DC(" | plane%d |       |       |", plane_idx);
 		DC_LOG_DC(" |DPP%d----|       |       |",
 				pipe->plane_res.dpp->inst);
+		capture_pipe_topology_data(dc, plane_idx, slice_idx, stream_idx,
+				pipe->plane_res.dpp->inst,
+				pipe->stream_res.opp->inst,
+				pipe->stream_res.tg->inst, is_phantom_pipe);
 	}
 }
 
 static void resource_log_pipe_for_stream(struct dc *dc, struct dc_state *state,
-		struct pipe_ctx *otg_master, int stream_idx)
+		struct pipe_ctx *otg_master, int stream_idx, bool is_phantom_pipe)
 {
 	struct pipe_ctx *opp_heads[MAX_PIPES];
 	struct pipe_ctx *dpp_pipes[MAX_PIPES];
@@ -2384,12 +2443,12 @@ static void resource_log_pipe_for_stream(struct dc *dc, struct dc_state *state,
 				resource_log_pipe(dc, dpp_pipes[dpp_idx],
 						stream_idx, slice_idx,
 						plane_idx, slice_count,
-						is_primary);
+						is_primary, is_phantom_pipe);
 			}
 		} else {
 			resource_log_pipe(dc, opp_heads[slice_idx],
 					stream_idx, slice_idx, plane_idx,
-					slice_count, true);
+					slice_count, true, is_phantom_pipe);
 		}
 
 	}
@@ -2420,6 +2479,10 @@ void resource_log_pipe_topology_update(struct dc *dc, struct dc_state *state)
 	struct pipe_ctx *otg_master;
 	int stream_idx, phantom_stream_idx;
 	DC_LOGGER_INIT(dc->ctx->logger);
+	bool is_phantom_pipe = false;
+
+	// Start a new snapshot for this topology update
+	start_new_topology_snapshot(dc, state);
 
 	DC_LOG_DC("    pipe topology update");
 	DC_LOG_DC("  ________________________");
@@ -2433,9 +2496,10 @@ void resource_log_pipe_topology_update(struct dc *dc, struct dc_state *state)
 		if (!otg_master)
 			continue;
 
-		resource_log_pipe_for_stream(dc, state, otg_master, stream_idx);
+		resource_log_pipe_for_stream(dc, state, otg_master, stream_idx, is_phantom_pipe);
 	}
 	if (state->phantom_stream_count > 0) {
+		is_phantom_pipe = true;
 		DC_LOG_DC(" |    (phantom pipes)     |");
 		for (stream_idx = 0; stream_idx < state->stream_count; stream_idx++) {
 			if (state->stream_status[stream_idx].mall_stream_config.type != SUBVP_MAIN)
@@ -2448,7 +2512,7 @@ void resource_log_pipe_topology_update(struct dc *dc, struct dc_state *state)
 			if (!otg_master)
 				continue;
 
-			resource_log_pipe_for_stream(dc, state, otg_master, stream_idx);
+			resource_log_pipe_for_stream(dc, state, otg_master, stream_idx, is_phantom_pipe);
 		}
 	}
 	DC_LOG_DC(" |________________________|\n");
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index 6d309c320253..129cd5f84983 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -737,9 +737,14 @@ bool dc_stream_get_scanoutpos(const struct dc_stream_state *stream,
 {
 	uint8_t i;
 	bool ret = false;
-	struct dc  *dc = stream->ctx->dc;
-	struct resource_context *res_ctx =
-		&dc->current_state->res_ctx;
+	struct dc  *dc;
+	struct resource_context *res_ctx;
+
+	if (!stream->ctx)
+		return false;
+
+	dc = stream->ctx->dc;
+	res_ctx = &dc->current_state->res_ctx;
 
 	dc_exit_ips_for_hw_access(dc);
 
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
index 922f23557f5d..0971dfa25845 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -86,7 +86,7 @@ uint8_t  dc_plane_get_pipe_mask(struct dc_state *dc_state, const struct dc_plane
 struct dc_plane_state *dc_create_plane_state(const struct dc *dc)
 {
 	struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state),
-							GFP_KERNEL);
+							GFP_ATOMIC);
 
 	if (NULL == plane_state)
 		return NULL;
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 010d9315b96b..29edfa51ea2c 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -63,7 +63,7 @@ struct dcn_dsc_reg_state;
 struct dcn_optc_reg_state;
 struct dcn_dccg_reg_state;
 
-#define DC_VER "3.2.356"
+#define DC_VER "3.2.359"
 
 /**
  * MAX_SURFACES - representative of the upper bound of surfaces that can be piped to a single CRTC
@@ -469,10 +469,9 @@ enum surface_update_type {
 
 enum dc_lock_descriptor {
 	LOCK_DESCRIPTOR_NONE = 0x0,
-	LOCK_DESCRIPTOR_STATE = 0x1,
+	LOCK_DESCRIPTOR_STREAM = 0x1,
 	LOCK_DESCRIPTOR_LINK = 0x2,
-	LOCK_DESCRIPTOR_STREAM = 0x4,
-	LOCK_DESCRIPTOR_PLANE = 0x8,
+	LOCK_DESCRIPTOR_GLOBAL = 0x4,
 };
 
 struct surface_update_descriptor {
@@ -900,6 +899,7 @@ struct dc_debug_data {
 	uint32_t ltFailCount;
 	uint32_t i2cErrorCount;
 	uint32_t auxErrorCount;
+	struct pipe_topology_history topology_history;
 };
 
 struct dc_phy_addr_space_config {
@@ -1865,6 +1865,18 @@ struct dc_underflow_debug_data {
 	struct dcn_dccg_reg_state *dccg_reg_state[MAX_PIPES];
 };
 
+struct power_features {
+	bool ips;
+	bool rcg;
+	bool replay;
+	bool dds;
+	bool sprs;
+	bool psr;
+	bool fams;
+	bool mpo;
+	bool uclk_p_state;
+};
+
 /*
  * Create a new surface with default parameters;
  */
@@ -2707,6 +2719,13 @@ bool dc_process_dmub_aux_transfer_async(struct dc *dc,
 				uint32_t link_index,
 				struct aux_payload *payload);
 
+/*
+ * smart power OLED Interfaces
+ */
+bool dc_smart_power_oled_enable(const struct dc_link *link, bool enable, uint16_t peak_nits,
+	uint8_t debug_control, uint16_t fixed_CLL, uint32_t triggerline);
+bool dc_smart_power_oled_get_max_cll(const struct dc_link *link, unsigned int *pCurrent_MaxCLL);
+
 /* Get dc link index from dpia port index */
 uint8_t get_link_index_from_dpia_port_index(const struct dc *dc,
 				uint8_t dpia_port_index);
@@ -2772,4 +2791,493 @@ bool dc_can_clear_cursor_limit(const struct dc *dc);
  */
 void dc_get_underflow_debug_data_for_otg(struct dc *dc, int primary_otg_inst, struct dc_underflow_debug_data *out_data);
 
+void dc_get_power_feature_status(struct dc *dc, int primary_otg_inst, struct power_features *out_data);
+
+/**
+ * Software state variables used to program register fields across the display pipeline
+ */
+struct dc_register_software_state {
+	/* HUBP register programming variables for each pipe */
+	struct {
+		bool valid_plane_state;
+		bool valid_stream;
+		bool min_dc_gfx_version9;
+		uint32_t vtg_sel;                        /* DCHUBP_CNTL->HUBP_VTG_SEL from pipe_ctx->stream_res.tg->inst */
+		uint32_t hubp_clock_enable;              /* HUBP_CLK_CNTL->HUBP_CLOCK_ENABLE from power management */
+		uint32_t surface_pixel_format;           /* DCSURF_SURFACE_CONFIG->SURFACE_PIXEL_FORMAT from plane_state->format */
+		uint32_t rotation_angle;                 /* DCSURF_SURFACE_CONFIG->ROTATION_ANGLE from plane_state->rotation */
+		uint32_t h_mirror_en;                    /* DCSURF_SURFACE_CONFIG->H_MIRROR_EN from plane_state->horizontal_mirror */
+		uint32_t surface_dcc_en;                 /* DCSURF_SURFACE_CONTROL->PRIMARY_SURFACE_DCC_EN from dcc->enable */
+		uint32_t surface_size_width;             /* HUBP_SIZE->SURFACE_SIZE_WIDTH from plane_size.surface_size.width */
+		uint32_t surface_size_height;            /* HUBP_SIZE->SURFACE_SIZE_HEIGHT from plane_size.surface_size.height */
+		uint32_t pri_viewport_width;             /* DCSURF_PRI_VIEWPORT_DIMENSION->PRI_VIEWPORT_WIDTH from scaler_data.viewport.width */
+		uint32_t pri_viewport_height;            /* DCSURF_PRI_VIEWPORT_DIMENSION->PRI_VIEWPORT_HEIGHT from scaler_data.viewport.height */
+		uint32_t pri_viewport_x_start;           /* DCSURF_PRI_VIEWPORT_START->PRI_VIEWPORT_X_START from scaler_data.viewport.x */
+		uint32_t pri_viewport_y_start;           /* DCSURF_PRI_VIEWPORT_START->PRI_VIEWPORT_Y_START from scaler_data.viewport.y */
+		uint32_t cursor_enable;                  /* CURSOR_CONTROL->CURSOR_ENABLE from cursor_attributes.enable */
+		uint32_t cursor_width;                   /* CURSOR_SETTINGS->CURSOR_WIDTH from cursor_position.width */
+		uint32_t cursor_height;                  /* CURSOR_SETTINGS->CURSOR_HEIGHT from cursor_position.height */
+
+		/* Additional DCC configuration */
+		uint32_t surface_dcc_ind_64b_blk;        /* DCSURF_SURFACE_CONTROL->PRIMARY_SURFACE_DCC_IND_64B_BLK from dcc.independent_64b_blks */
+		uint32_t surface_dcc_ind_128b_blk;       /* DCSURF_SURFACE_CONTROL->PRIMARY_SURFACE_DCC_IND_128B_BLK from dcc.independent_128b_blks */
+
+		/* Surface pitch configuration */
+		uint32_t surface_pitch;                  /* DCSURF_SURFACE_PITCH->PITCH from plane_size.surface_pitch */
+		uint32_t meta_pitch;                     /* DCSURF_SURFACE_PITCH->META_PITCH from dcc.meta_pitch */
+		uint32_t chroma_pitch;                   /* DCSURF_SURFACE_PITCH_C->PITCH_C from plane_size.chroma_pitch */
+		uint32_t meta_pitch_c;                   /* DCSURF_SURFACE_PITCH_C->META_PITCH_C from dcc.meta_pitch_c */
+
+		/* Surface addresses */
+		uint32_t primary_surface_address_low;    /* DCSURF_PRIMARY_SURFACE_ADDRESS->PRIMARY_SURFACE_ADDRESS from address.grph.addr.low_part */
+		uint32_t primary_surface_address_high;   /* DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH->PRIMARY_SURFACE_ADDRESS_HIGH from address.grph.addr.high_part */
+		uint32_t primary_meta_surface_address_low;  /* DCSURF_PRIMARY_META_SURFACE_ADDRESS->PRIMARY_META_SURFACE_ADDRESS from address.grph.meta_addr.low_part */
+		uint32_t primary_meta_surface_address_high; /* DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH->PRIMARY_META_SURFACE_ADDRESS_HIGH from address.grph.meta_addr.high_part */
+
+		/* TMZ configuration */
+		uint32_t primary_surface_tmz;            /* DCSURF_SURFACE_CONTROL->PRIMARY_SURFACE_TMZ from address.tmz_surface */
+		uint32_t primary_meta_surface_tmz;       /* DCSURF_SURFACE_CONTROL->PRIMARY_META_SURFACE_TMZ from address.tmz_surface */
+
+		/* Tiling configuration */
+		uint32_t sw_mode;                        /* DCSURF_TILING_CONFIG->SW_MODE from tiling_info.gfx9.swizzle */
+		uint32_t num_pipes;                      /* DCSURF_ADDR_CONFIG->NUM_PIPES from tiling_info.gfx9.num_pipes */
+		uint32_t num_banks;                      /* DCSURF_ADDR_CONFIG->NUM_BANKS from tiling_info.gfx9.num_banks */
+		uint32_t pipe_interleave;                /* DCSURF_ADDR_CONFIG->PIPE_INTERLEAVE from tiling_info.gfx9.pipe_interleave */
+		uint32_t num_shader_engines;             /* DCSURF_ADDR_CONFIG->NUM_SE from tiling_info.gfx9.num_shader_engines */
+		uint32_t num_rb_per_se;                  /* DCSURF_ADDR_CONFIG->NUM_RB_PER_SE from tiling_info.gfx9.num_rb_per_se */
+		uint32_t num_pkrs;                       /* DCSURF_ADDR_CONFIG->NUM_PKRS from tiling_info.gfx9.num_pkrs */
+
+		/* DML Request Size Configuration - Luma */
+		uint32_t rq_chunk_size;                  /* DCHUBP_REQ_SIZE_CONFIG->CHUNK_SIZE from rq_regs.rq_regs_l.chunk_size */
+		uint32_t rq_min_chunk_size;              /* DCHUBP_REQ_SIZE_CONFIG->MIN_CHUNK_SIZE from rq_regs.rq_regs_l.min_chunk_size */
+		uint32_t rq_meta_chunk_size;             /* DCHUBP_REQ_SIZE_CONFIG->META_CHUNK_SIZE from rq_regs.rq_regs_l.meta_chunk_size */
+		uint32_t rq_min_meta_chunk_size;         /* DCHUBP_REQ_SIZE_CONFIG->MIN_META_CHUNK_SIZE from rq_regs.rq_regs_l.min_meta_chunk_size */
+		uint32_t rq_dpte_group_size;             /* DCHUBP_REQ_SIZE_CONFIG->DPTE_GROUP_SIZE from rq_regs.rq_regs_l.dpte_group_size */
+		uint32_t rq_mpte_group_size;             /* DCHUBP_REQ_SIZE_CONFIG->MPTE_GROUP_SIZE from rq_regs.rq_regs_l.mpte_group_size */
+		uint32_t rq_swath_height_l;              /* DCHUBP_REQ_SIZE_CONFIG->SWATH_HEIGHT_L from rq_regs.rq_regs_l.swath_height */
+		uint32_t rq_pte_row_height_l;            /* DCHUBP_REQ_SIZE_CONFIG->PTE_ROW_HEIGHT_L from rq_regs.rq_regs_l.pte_row_height */
+
+		/* DML Request Size Configuration - Chroma */
+		uint32_t rq_chunk_size_c;                /* DCHUBP_REQ_SIZE_CONFIG_C->CHUNK_SIZE_C from rq_regs.rq_regs_c.chunk_size */
+		uint32_t rq_min_chunk_size_c;            /* DCHUBP_REQ_SIZE_CONFIG_C->MIN_CHUNK_SIZE_C from rq_regs.rq_regs_c.min_chunk_size */
+		uint32_t rq_meta_chunk_size_c;           /* DCHUBP_REQ_SIZE_CONFIG_C->META_CHUNK_SIZE_C from rq_regs.rq_regs_c.meta_chunk_size */
+		uint32_t rq_min_meta_chunk_size_c;       /* DCHUBP_REQ_SIZE_CONFIG_C->MIN_META_CHUNK_SIZE_C from rq_regs.rq_regs_c.min_meta_chunk_size */
+		uint32_t rq_dpte_group_size_c;           /* DCHUBP_REQ_SIZE_CONFIG_C->DPTE_GROUP_SIZE_C from rq_regs.rq_regs_c.dpte_group_size */
+		uint32_t rq_mpte_group_size_c;           /* DCHUBP_REQ_SIZE_CONFIG_C->MPTE_GROUP_SIZE_C from rq_regs.rq_regs_c.mpte_group_size */
+		uint32_t rq_swath_height_c;              /* DCHUBP_REQ_SIZE_CONFIG_C->SWATH_HEIGHT_C from rq_regs.rq_regs_c.swath_height */
+		uint32_t rq_pte_row_height_c;            /* DCHUBP_REQ_SIZE_CONFIG_C->PTE_ROW_HEIGHT_C from rq_regs.rq_regs_c.pte_row_height */
+
+		/* DML Expansion Modes */
+		uint32_t drq_expansion_mode;             /* DCN_EXPANSION_MODE->DRQ_EXPANSION_MODE from rq_regs.drq_expansion_mode */
+		uint32_t prq_expansion_mode;             /* DCN_EXPANSION_MODE->PRQ_EXPANSION_MODE from rq_regs.prq_expansion_mode */
+		uint32_t mrq_expansion_mode;             /* DCN_EXPANSION_MODE->MRQ_EXPANSION_MODE from rq_regs.mrq_expansion_mode */
+		uint32_t crq_expansion_mode;             /* DCN_EXPANSION_MODE->CRQ_EXPANSION_MODE from rq_regs.crq_expansion_mode */
+
+		/* DML DLG parameters - nominal */
+		uint32_t dst_y_per_vm_vblank;            /* NOM_PARAMETERS_0->DST_Y_PER_VM_VBLANK from dlg_regs.dst_y_per_vm_vblank */
+		uint32_t dst_y_per_row_vblank;           /* NOM_PARAMETERS_0->DST_Y_PER_ROW_VBLANK from dlg_regs.dst_y_per_row_vblank */
+		uint32_t dst_y_per_vm_flip;              /* NOM_PARAMETERS_1->DST_Y_PER_VM_FLIP from dlg_regs.dst_y_per_vm_flip */
+		uint32_t dst_y_per_row_flip;             /* NOM_PARAMETERS_1->DST_Y_PER_ROW_FLIP from dlg_regs.dst_y_per_row_flip */
+
+		/* DML prefetch settings */
+		uint32_t dst_y_prefetch;                 /* PREFETCH_SETTINS->DST_Y_PREFETCH from dlg_regs.dst_y_prefetch */
+		uint32_t vratio_prefetch;                /* PREFETCH_SETTINS->VRATIO_PREFETCH from dlg_regs.vratio_prefetch */
+		uint32_t vratio_prefetch_c;              /* PREFETCH_SETTINS_C->VRATIO_PREFETCH_C from dlg_regs.vratio_prefetch_c */
+
+		/* TTU parameters */
+		uint32_t qos_level_low_wm;               /* TTU_CNTL1->QoSLevelLowWaterMark from ttu_regs.qos_level_low_wm */
+		uint32_t qos_level_high_wm;              /* TTU_CNTL1->QoSLevelHighWaterMark from ttu_regs.qos_level_high_wm */
+		uint32_t qos_level_flip;                 /* TTU_CNTL2->QoS_LEVEL_FLIP_L from ttu_regs.qos_level_flip */
+		uint32_t min_ttu_vblank;                 /* DCN_GLOBAL_TTU_CNTL->MIN_TTU_VBLANK from ttu_regs.min_ttu_vblank */
+	} hubp[MAX_PIPES];
+
+	/* HUBBUB register programming variables */
+	struct {
+		/* Individual DET buffer control per pipe - software state that programs DET registers */
+		uint32_t det0_size;                      /* DCHUBBUB_DET0_CTRL->DET0_SIZE from hubbub->funcs->program_det_size(hubbub, 0, det_buffer_size_kb) */
+		uint32_t det1_size;                      /* DCHUBBUB_DET1_CTRL->DET1_SIZE from hubbub->funcs->program_det_size(hubbub, 1, det_buffer_size_kb) */
+		uint32_t det2_size;                      /* DCHUBBUB_DET2_CTRL->DET2_SIZE from hubbub->funcs->program_det_size(hubbub, 2, det_buffer_size_kb) */
+		uint32_t det3_size;                      /* DCHUBBUB_DET3_CTRL->DET3_SIZE from hubbub->funcs->program_det_size(hubbub, 3, det_buffer_size_kb) */
+
+		/* Compression buffer control - software state that programs COMPBUF registers */
+		uint32_t compbuf_size;                   /* DCHUBBUB_COMPBUF_CTRL->COMPBUF_SIZE from hubbub->funcs->program_compbuf_size(hubbub, compbuf_size_kb, safe_to_increase) */
+		uint32_t compbuf_reserved_space_64b;     /* COMPBUF_RESERVED_SPACE->COMPBUF_RESERVED_SPACE_64B from hubbub2->pixel_chunk_size / 32 */
+		uint32_t compbuf_reserved_space_zs;      /* COMPBUF_RESERVED_SPACE->COMPBUF_RESERVED_SPACE_ZS from hubbub2->pixel_chunk_size / 128 */
+	} hubbub;
+
+	/* DPP register programming variables for each pipe (simplified for available fields) */
+	struct {
+		uint32_t dpp_clock_enable;               /* DPP_CONTROL->DPP_CLOCK_ENABLE from dppclk_enable */
+
+		/* Recout (Rectangle of Interest) configuration */
+		uint32_t recout_start_x;                 /* RECOUT_START->RECOUT_START_X from pipe_ctx->plane_res.scl_data.recout.x */
+		uint32_t recout_start_y;                 /* RECOUT_START->RECOUT_START_Y from pipe_ctx->plane_res.scl_data.recout.y */
+		uint32_t recout_width;                   /* RECOUT_SIZE->RECOUT_WIDTH from pipe_ctx->plane_res.scl_data.recout.width */
+		uint32_t recout_height;                  /* RECOUT_SIZE->RECOUT_HEIGHT from pipe_ctx->plane_res.scl_data.recout.height */
+
+		/* MPC (Multiple Pipe/Plane Combiner) size configuration */
+		uint32_t mpc_width;                      /* MPC_SIZE->MPC_WIDTH from pipe_ctx->plane_res.scl_data.h_active */
+		uint32_t mpc_height;                     /* MPC_SIZE->MPC_HEIGHT from pipe_ctx->plane_res.scl_data.v_active */
+
+		/* DSCL mode configuration */
+		uint32_t dscl_mode;                      /* SCL_MODE->DSCL_MODE from pipe_ctx->plane_res.scl_data.dscl_prog_data.dscl_mode */
+
+		/* Scaler ratios (simplified to integer parts) */
+		uint32_t horz_ratio_int;                 /* SCL_HORZ_FILTER_SCALE_RATIO->SCL_H_SCALE_RATIO integer part from ratios.horz */
+		uint32_t vert_ratio_int;                 /* SCL_VERT_FILTER_SCALE_RATIO->SCL_V_SCALE_RATIO integer part from ratios.vert */
+
+		/* Basic scaler taps */
+		uint32_t h_taps;                         /* SCL_TAP_CONTROL->SCL_H_NUM_TAPS from taps.h_taps */
+		uint32_t v_taps;                         /* SCL_TAP_CONTROL->SCL_V_NUM_TAPS from taps.v_taps */
+	} dpp[MAX_PIPES];
+
+	/* DCCG register programming variables */
+	struct {
+		/* Core Display Clock Control */
+		uint32_t dispclk_khz;                    /* DENTIST_DISPCLK_CNTL->DENTIST_DISPCLK_WDIVIDER from clk_mgr.dispclk_khz */
+		uint32_t dc_mem_global_pwr_req_dis;      /* DC_MEM_GLOBAL_PWR_REQ_CNTL->DC_MEM_GLOBAL_PWR_REQ_DIS from memory power management settings */
+
+		/* DPP Clock Control - 4 fields per pipe */
+		uint32_t dppclk_khz[MAX_PIPES];          /* DPPCLK_CTRL->DPPCLK_R_GATE_DISABLE from dpp_clocks[pipe] */
+		uint32_t dppclk_enable[MAX_PIPES];       /* DPPCLK_CTRL->DPPCLK0_EN,DPPCLK1_EN,DPPCLK2_EN,DPPCLK3_EN from dccg31_update_dpp_dto() */
+		uint32_t dppclk_dto_enable[MAX_PIPES];   /* DPPCLK_DTO_CTRL->DPPCLK_DTO_ENABLE from dccg->dpp_clock_gated[dpp_inst] state */
+		uint32_t dppclk_dto_phase[MAX_PIPES];    /* DPPCLK0_DTO_PARAM->DPPCLK0_DTO_PHASE from phase calculation req_dppclk/ref_dppclk */
+		uint32_t dppclk_dto_modulo[MAX_PIPES];   /* DPPCLK0_DTO_PARAM->DPPCLK0_DTO_MODULO from modulo = 0xff */
+
+		/* DSC Clock Control - 4 fields per DSC resource */
+		uint32_t dscclk_khz[MAX_PIPES]; /* DSCCLK_DTO_CTRL->DSCCLK_DTO_ENABLE from dsc_clocks */
+		uint32_t dscclk_dto_enable[MAX_PIPES]; /* DSCCLK_DTO_CTRL->DSCCLK0_DTO_ENABLE,DSCCLK1_DTO_ENABLE,DSCCLK2_DTO_ENABLE,DSCCLK3_DTO_ENABLE */
+		uint32_t dscclk_dto_phase[MAX_PIPES];  /* DSCCLK0_DTO_PARAM->DSCCLK0_DTO_PHASE from dccg31_enable_dscclk() */
+		uint32_t dscclk_dto_modulo[MAX_PIPES]; /* DSCCLK0_DTO_PARAM->DSCCLK0_DTO_MODULO from dccg31_enable_dscclk() */
+
+		/* Pixel Clock Control - per pipe */
+		uint32_t pixclk_khz[MAX_PIPES];          /* PIXCLK_RESYNC_CNTL->PIXCLK_RESYNC_ENABLE from stream.timing.pix_clk_100hz */
+		uint32_t otg_pixel_rate_div[MAX_PIPES];  /* OTG_PIXEL_RATE_DIV->OTG_PIXEL_RATE_DIV from OTG pixel rate divider control */
+		uint32_t dtbclk_dto_enable[MAX_PIPES];   /* OTG0_PIXEL_RATE_CNTL->DTBCLK_DTO_ENABLE from dccg31_set_dtbclk_dto() */
+		uint32_t pipe_dto_src_sel[MAX_PIPES];    /* OTG0_PIXEL_RATE_CNTL->PIPE_DTO_SRC_SEL from dccg31_set_dtbclk_dto() source selection */
+		uint32_t dtbclk_dto_div[MAX_PIPES];      /* OTG0_PIXEL_RATE_CNTL->DTBCLK_DTO_DIV from dtbdto_div calculation */
+		uint32_t otg_add_pixel[MAX_PIPES];       /* OTG0_PIXEL_RATE_CNTL->OTG_ADD_PIXEL from dccg31_otg_add_pixel() */
+		uint32_t otg_drop_pixel[MAX_PIPES];      /* OTG0_PIXEL_RATE_CNTL->OTG_DROP_PIXEL from dccg31_otg_drop_pixel() */
+
+		/* DTBCLK DTO Control - 4 DTOs */
+		uint32_t dtbclk_dto_modulo[4];           /* DTBCLK_DTO0_MODULO->DTBCLK_DTO0_MODULO from dccg31_set_dtbclk_dto() modulo calculation */
+		uint32_t dtbclk_dto_phase[4];            /* DTBCLK_DTO0_PHASE->DTBCLK_DTO0_PHASE from phase calculation pixclk_khz/ref_dtbclk_khz */
+		uint32_t dtbclk_dto_dbuf_en;             /* DTBCLK_DTO_DBUF_EN->DTBCLK DTO data buffer enable */
+
+		/* DP Stream Clock Control - 4 pipes */
+		uint32_t dpstreamclk_enable[MAX_PIPES];          /* DPSTREAMCLK_CNTL->DPSTREAMCLK_PIPE0_EN,DPSTREAMCLK_PIPE1_EN,DPSTREAMCLK_PIPE2_EN,DPSTREAMCLK_PIPE3_EN */
+		uint32_t dp_dto_modulo[4];               /* DP_DTO0_MODULO->DP_DTO0_MODULO from DP stream DTO programming */
+		uint32_t dp_dto_phase[4];                /* DP_DTO0_PHASE->DP_DTO0_PHASE from DP stream DTO programming */
+		uint32_t dp_dto_dbuf_en;                 /* DP_DTO_DBUF_EN->DP DTO data buffer enable */
+
+		/* PHY Symbol Clock Control - 5 PHYs (A,B,C,D,E) */
+		uint32_t phy_symclk_force_en[5];         /* PHYASYMCLK_CLOCK_CNTL->PHYASYMCLK_FORCE_EN from dccg31_set_physymclk() force_enable */
+		uint32_t phy_symclk_force_src_sel[5];    /* PHYASYMCLK_CLOCK_CNTL->PHYASYMCLK_FORCE_SRC_SEL from dccg31_set_physymclk() clk_src */
+		uint32_t phy_symclk_gate_disable[5];     /* DCCG_GATE_DISABLE_CNTL2->PHYASYMCLK_GATE_DISABLE from debug.root_clock_optimization.bits.physymclk */
+
+		/* SYMCLK32 SE Control - 4 instances */
+		uint32_t symclk32_se_src_sel[4];         /* SYMCLK32_SE_CNTL->SYMCLK32_SE0_SRC_SEL from dccg31_enable_symclk32_se() with get_phy_mux_symclk() mapping */
+		uint32_t symclk32_se_enable[4];          /* SYMCLK32_SE_CNTL->SYMCLK32_SE0_EN from dccg31_enable_symclk32_se() enable */
+		uint32_t symclk32_se_gate_disable[4];    /* DCCG_GATE_DISABLE_CNTL3->SYMCLK32_SE0_GATE_DISABLE from debug.root_clock_optimization.bits.symclk32_se */
+
+		/* SYMCLK32 LE Control - 2 instances */
+		uint32_t symclk32_le_src_sel[2];         /* SYMCLK32_LE_CNTL->SYMCLK32_LE0_SRC_SEL from dccg31_enable_symclk32_le() phyd32clk source */
+		uint32_t symclk32_le_enable[2];          /* SYMCLK32_LE_CNTL->SYMCLK32_LE0_EN from dccg31_enable_symclk32_le() enable */
+		uint32_t symclk32_le_gate_disable[2];    /* DCCG_GATE_DISABLE_CNTL3->SYMCLK32_LE0_GATE_DISABLE from debug.root_clock_optimization.bits.symclk32_le */
+
+		/* DPIA Clock Control */
+		uint32_t dpiaclk_540m_dto_modulo;        /* DPIACLK_540M_DTO_MODULO->DPIA 540MHz DTO modulo */
+		uint32_t dpiaclk_540m_dto_phase;         /* DPIACLK_540M_DTO_PHASE->DPIA 540MHz DTO phase */
+		uint32_t dpiaclk_810m_dto_modulo;        /* DPIACLK_810M_DTO_MODULO->DPIA 810MHz DTO modulo */
+		uint32_t dpiaclk_810m_dto_phase;         /* DPIACLK_810M_DTO_PHASE->DPIA 810MHz DTO phase */
+		uint32_t dpiaclk_dto_cntl;               /* DPIACLK_DTO_CNTL->DPIA clock DTO control */
+		uint32_t dpiasymclk_cntl;                /* DPIASYMCLK_CNTL->DPIA symbol clock control */
+
+		/* Clock Gating Control */
+		uint32_t dccg_gate_disable_cntl;         /* DCCG_GATE_DISABLE_CNTL->Clock gate disable control from dccg31_init() */
+		uint32_t dpstreamclk_gate_disable;       /* DCCG_GATE_DISABLE_CNTL3->DPSTREAMCLK_GATE_DISABLE from debug.root_clock_optimization.bits.dpstream */
+		uint32_t dpstreamclk_root_gate_disable;  /* DCCG_GATE_DISABLE_CNTL3->DPSTREAMCLK_ROOT_GATE_DISABLE from debug.root_clock_optimization.bits.dpstream */
+
+		/* VSync Control */
+		uint32_t vsync_cnt_ctrl;                 /* DCCG_VSYNC_CNT_CTRL->VSync counter control */
+		uint32_t vsync_cnt_int_ctrl;             /* DCCG_VSYNC_CNT_INT_CTRL->VSync counter interrupt control */
+		uint32_t vsync_otg_latch_value[6];       /* DCCG_VSYNC_OTG0_LATCH_VALUE->OTG0 VSync latch value (for OTG0-5) */
+
+		/* Time Base Control */
+		uint32_t microsecond_time_base_div;      /* MICROSECOND_TIME_BASE_DIV->Microsecond time base divider */
+		uint32_t millisecond_time_base_div;      /* MILLISECOND_TIME_BASE_DIV->Millisecond time base divider */
+	} dccg;
+
+	/* DSC essential configuration for underflow analysis */
+	struct {
+		/* DSC active state - critical for bandwidth analysis */
+		uint32_t dsc_clock_enable;               /* DSC enabled - affects bandwidth requirements */
+
+		/* DSC configuration affecting bandwidth and timing */
+		uint32_t dsc_num_slices_h;              /* Horizontal slice count - affects throughput */
+		uint32_t dsc_num_slices_v;              /* Vertical slice count - affects throughput */
+		uint32_t dsc_bits_per_pixel;            /* Compression ratio - affects bandwidth */
+
+		/* OPP integration - affects pipeline flow */
+		uint32_t dscrm_dsc_forward_enable;      /* DSC forwarding to OPP enabled */
+		uint32_t dscrm_dsc_opp_pipe_source;    /* Which OPP receives DSC output */
+	} dsc[MAX_PIPES];
+
+	/* MPC register programming variables */
+	struct {
+		/* MPCC blending tree and mode control */
+		uint32_t mpcc_mode[MAX_PIPES];           /* MPCC_CONTROL->MPCC_MODE from blend_cfg.blend_mode */
+		uint32_t mpcc_alpha_blend_mode[MAX_PIPES]; /* MPCC_CONTROL->MPCC_ALPHA_BLND_MODE from blend_cfg.alpha_mode */
+		uint32_t mpcc_alpha_multiplied_mode[MAX_PIPES]; /* MPCC_CONTROL->MPCC_ALPHA_MULTIPLIED_MODE from blend_cfg.pre_multiplied_alpha */
+		uint32_t mpcc_blnd_active_overlap_only[MAX_PIPES]; /* MPCC_CONTROL->MPCC_BLND_ACTIVE_OVERLAP_ONLY from blend_cfg.overlap_only */
+		uint32_t mpcc_global_alpha[MAX_PIPES];   /* MPCC_CONTROL->MPCC_GLOBAL_ALPHA from blend_cfg.global_alpha */
+		uint32_t mpcc_global_gain[MAX_PIPES];    /* MPCC_CONTROL->MPCC_GLOBAL_GAIN from blend_cfg.global_gain */
+		uint32_t mpcc_bg_bpc[MAX_PIPES];         /* MPCC_CONTROL->MPCC_BG_BPC from background color depth */
+		uint32_t mpcc_bot_gain_mode[MAX_PIPES];  /* MPCC_CONTROL->MPCC_BOT_GAIN_MODE from bottom layer gain control */
+
+		/* MPCC blending tree connections */
+		uint32_t mpcc_bot_sel[MAX_PIPES];        /* MPCC_BOT_SEL->MPCC_BOT_SEL from mpcc_state->bot_sel */
+		uint32_t mpcc_top_sel[MAX_PIPES];        /* MPCC_TOP_SEL->MPCC_TOP_SEL from mpcc_state->dpp_id */
+
+		/* MPCC output gamma control */
+		uint32_t mpcc_ogam_mode[MAX_PIPES];      /* MPCC_OGAM_CONTROL->MPCC_OGAM_MODE from output gamma mode */
+		uint32_t mpcc_ogam_select[MAX_PIPES];    /* MPCC_OGAM_CONTROL->MPCC_OGAM_SELECT from gamma LUT bank selection */
+		uint32_t mpcc_ogam_pwl_disable[MAX_PIPES]; /* MPCC_OGAM_CONTROL->MPCC_OGAM_PWL_DISABLE from PWL control */
+
+		/* MPCC pipe assignment and status */
+		uint32_t mpcc_opp_id[MAX_PIPES];         /* MPCC_OPP_ID->MPCC_OPP_ID from mpcc_state->opp_id */
+		uint32_t mpcc_idle[MAX_PIPES];           /* MPCC_STATUS->MPCC_IDLE from mpcc idle status */
+		uint32_t mpcc_busy[MAX_PIPES];           /* MPCC_STATUS->MPCC_BUSY from mpcc busy status */
+
+		/* MPC output processing */
+		uint32_t mpc_out_csc_mode;               /* MPC_OUT_CSC_COEF->MPC_OUT_CSC_MODE from output_csc */
+		uint32_t mpc_out_gamma_mode;             /* MPC_OUT_GAMMA_LUT->MPC_OUT_GAMMA_MODE from output_gamma */
+	} mpc;
+
+	/* OPP register programming variables for each pipe */
+	struct {
+		/* Display Pattern Generator (DPG) Control - 19 fields from DPG_CONTROL register */
+		uint32_t dpg_enable;                     /* DPG_CONTROL->DPG_EN from test_pattern parameter (enable/disable) */
+
+		/* Format Control (FMT) - 18 fields from FMT_CONTROL register */
+		uint32_t fmt_pixel_encoding;             /* FMT_CONTROL->FMT_PIXEL_ENCODING from clamping->pixel_encoding */
+		uint32_t fmt_subsampling_mode;           /* FMT_CONTROL->FMT_SUBSAMPLING_MODE from force_chroma_subsampling_1tap */
+		uint32_t fmt_cbcr_bit_reduction_bypass;  /* FMT_CONTROL->FMT_CBCR_BIT_REDUCTION_BYPASS from pixel_encoding bypass control */
+		uint32_t fmt_stereosync_override;        /* FMT_CONTROL->FMT_STEREOSYNC_OVERRIDE from stereo timing override */
+		uint32_t fmt_spatial_dither_frame_counter_max; /* FMT_CONTROL->FMT_SPATIAL_DITHER_FRAME_COUNTER_MAX from fmt_bit_depth->flags */
+		uint32_t fmt_spatial_dither_frame_counter_bit_swap; /* FMT_CONTROL->FMT_SPATIAL_DITHER_FRAME_COUNTER_BIT_SWAP from dither control */
+		uint32_t fmt_truncate_enable;            /* FMT_CONTROL->FMT_TRUNCATE_EN from fmt_bit_depth->flags.TRUNCATE_ENABLED */
+		uint32_t fmt_truncate_depth;             /* FMT_CONTROL->FMT_TRUNCATE_DEPTH from fmt_bit_depth->flags.TRUNCATE_DEPTH */
+		uint32_t fmt_truncate_mode;              /* FMT_CONTROL->FMT_TRUNCATE_MODE from fmt_bit_depth->flags.TRUNCATE_MODE */
+		uint32_t fmt_spatial_dither_enable;      /* FMT_CONTROL->FMT_SPATIAL_DITHER_EN from fmt_bit_depth->flags.SPATIAL_DITHER_ENABLED */
+		uint32_t fmt_spatial_dither_mode;        /* FMT_CONTROL->FMT_SPATIAL_DITHER_MODE from fmt_bit_depth->flags.SPATIAL_DITHER_MODE */
+		uint32_t fmt_spatial_dither_depth;       /* FMT_CONTROL->FMT_SPATIAL_DITHER_DEPTH from fmt_bit_depth->flags.SPATIAL_DITHER_DEPTH */
+		uint32_t fmt_temporal_dither_enable;     /* FMT_CONTROL->FMT_TEMPORAL_DITHER_EN from fmt_bit_depth->flags.TEMPORAL_DITHER_ENABLED */
+		uint32_t fmt_clamp_data_enable;          /* FMT_CONTROL->FMT_CLAMP_DATA_EN from clamping->clamping_range enable */
+		uint32_t fmt_clamp_color_format;         /* FMT_CONTROL->FMT_CLAMP_COLOR_FORMAT from clamping->color_format */
+		uint32_t fmt_dynamic_exp_enable;         /* FMT_CONTROL->FMT_DYNAMIC_EXP_EN from color_sp/color_dpth/signal */
+		uint32_t fmt_dynamic_exp_mode;           /* FMT_CONTROL->FMT_DYNAMIC_EXP_MODE from color space mode mapping */
+		uint32_t fmt_bit_depth_control;          /* Legacy field - kept for compatibility */
+
+		/* OPP Pipe Control - 1 field from OPP_PIPE_CONTROL register */
+		uint32_t opp_pipe_clock_enable;          /* OPP_PIPE_CONTROL->OPP_PIPE_CLOCK_EN from enable parameter (bool) */
+
+		/* OPP CRC Control - 3 fields from OPP_PIPE_CRC_CONTROL register */
+		uint32_t opp_crc_enable;                 /* OPP_PIPE_CRC_CONTROL->CRC_EN from CRC enable control */
+		uint32_t opp_crc_select_source;          /* OPP_PIPE_CRC_CONTROL->CRC_SELECT_SOURCE from CRC source selection */
+		uint32_t opp_crc_stereo_cont;            /* OPP_PIPE_CRC_CONTROL->CRC_STEREO_CONT from stereo continuous CRC */
+
+		/* Output Buffer (OPPBUF) Control - 6 fields from OPPBUF_CONTROL register */
+		uint32_t oppbuf_active_width;            /* OPPBUF_CONTROL->OPPBUF_ACTIVE_WIDTH from oppbuf_params->active_width */
+		uint32_t oppbuf_pixel_repetition;        /* OPPBUF_CONTROL->OPPBUF_PIXEL_REPETITION from oppbuf_params->pixel_repetition */
+		uint32_t oppbuf_display_segmentation;    /* OPPBUF_CONTROL->OPPBUF_DISPLAY_SEGMENTATION from oppbuf_params->mso_segmentation */
+		uint32_t oppbuf_overlap_pixel_num;       /* OPPBUF_CONTROL->OPPBUF_OVERLAP_PIXEL_NUM from oppbuf_params->mso_overlap_pixel_num */
+		uint32_t oppbuf_3d_vact_space1_size;     /* OPPBUF_CONTROL->OPPBUF_3D_VACT_SPACE1_SIZE from 3D timing space1_size */
+		uint32_t oppbuf_3d_vact_space2_size;     /* OPPBUF_CONTROL->OPPBUF_3D_VACT_SPACE2_SIZE from 3D timing space2_size */
+
+		/* DSC Forward Config - 3 fields from DSCRM_DSC_FORWARD_CONFIG register */
+		uint32_t dscrm_dsc_forward_enable;       /* DSCRM_DSC_FORWARD_CONFIG->DSCRM_DSC_FORWARD_EN from DSC forward enable control */
+		uint32_t dscrm_dsc_opp_pipe_source;      /* DSCRM_DSC_FORWARD_CONFIG->DSCRM_DSC_OPP_PIPE_SOURCE from opp_pipe parameter */
+		uint32_t dscrm_dsc_forward_enable_status; /* DSCRM_DSC_FORWARD_CONFIG->DSCRM_DSC_FORWARD_EN_STATUS from DSC forward status (read-only) */
+	} opp[MAX_PIPES];
+
+	/* OPTC register programming variables for each pipe */
+	struct {
+		uint32_t otg_master_inst;
+
+		/* OTG_CONTROL register - 5 fields for OTG control */
+		uint32_t otg_master_enable;              /* OTG_CONTROL->OTG_MASTER_EN from timing enable/disable control */
+		uint32_t otg_disable_point_cntl;         /* OTG_CONTROL->OTG_DISABLE_POINT_CNTL from disable timing control */
+		uint32_t otg_start_point_cntl;           /* OTG_CONTROL->OTG_START_POINT_CNTL from start timing control */
+		uint32_t otg_field_number_cntl;          /* OTG_CONTROL->OTG_FIELD_NUMBER_CNTL from interlace field control */
+		uint32_t otg_out_mux;                    /* OTG_CONTROL->OTG_OUT_MUX from output mux selection */
+
+		/* OTG Horizontal Timing - 7 fields */
+		uint32_t otg_h_total;                    /* OTG_H_TOTAL->OTG_H_TOTAL from dc_crtc_timing->h_total */
+		uint32_t otg_h_blank_start;              /* OTG_H_BLANK_START_END->OTG_H_BLANK_START from dc_crtc_timing->h_front_porch */
+		uint32_t otg_h_blank_end;                /* OTG_H_BLANK_START_END->OTG_H_BLANK_END from dc_crtc_timing->h_addressable_video_pixel_width */
+		uint32_t otg_h_sync_start;               /* OTG_H_SYNC_A->OTG_H_SYNC_A_START from dc_crtc_timing->h_sync_width */
+		uint32_t otg_h_sync_end;                 /* OTG_H_SYNC_A->OTG_H_SYNC_A_END from calculated sync end position */
+		uint32_t otg_h_sync_polarity;            /* OTG_H_SYNC_A_CNTL->OTG_H_SYNC_A_POL from dc_crtc_timing->flags.HSYNC_POSITIVE_POLARITY */
+		uint32_t otg_h_timing_div_mode;          /* OTG_H_TIMING_CNTL->OTG_H_TIMING_DIV_MODE from horizontal timing division mode */
+
+		/* OTG Vertical Timing - 7 fields */
+		uint32_t otg_v_total;                    /* OTG_V_TOTAL->OTG_V_TOTAL from dc_crtc_timing->v_total */
+		uint32_t otg_v_blank_start;              /* OTG_V_BLANK_START_END->OTG_V_BLANK_START from dc_crtc_timing->v_front_porch */
+		uint32_t otg_v_blank_end;                /* OTG_V_BLANK_START_END->OTG_V_BLANK_END from dc_crtc_timing->v_addressable_video_line_width */
+		uint32_t otg_v_sync_start;               /* OTG_V_SYNC_A->OTG_V_SYNC_A_START from dc_crtc_timing->v_sync_width */
+		uint32_t otg_v_sync_end;                 /* OTG_V_SYNC_A->OTG_V_SYNC_A_END from calculated sync end position */
+		uint32_t otg_v_sync_polarity;            /* OTG_V_SYNC_A_CNTL->OTG_V_SYNC_A_POL from dc_crtc_timing->flags.VSYNC_POSITIVE_POLARITY */
+		uint32_t otg_v_sync_mode;                /* OTG_V_SYNC_A_CNTL->OTG_V_SYNC_MODE from sync mode selection */
+
+		/* OTG DRR (Dynamic Refresh Rate) Control - 8 fields */
+		uint32_t otg_v_total_max;                /* OTG_V_TOTAL_MAX->OTG_V_TOTAL_MAX from drr_params->vertical_total_max */
+		uint32_t otg_v_total_min;                /* OTG_V_TOTAL_MIN->OTG_V_TOTAL_MIN from drr_params->vertical_total_min */
+		uint32_t otg_v_total_mid;                /* OTG_V_TOTAL_MID->OTG_V_TOTAL_MID from drr_params->vertical_total_mid */
+		uint32_t otg_v_total_max_sel;            /* OTG_V_TOTAL_CONTROL->OTG_V_TOTAL_MAX_SEL from DRR max selection enable */
+		uint32_t otg_v_total_min_sel;            /* OTG_V_TOTAL_CONTROL->OTG_V_TOTAL_MIN_SEL from DRR min selection enable */
+		uint32_t otg_vtotal_mid_replacing_max_en; /* OTG_V_TOTAL_CONTROL->OTG_VTOTAL_MID_REPLACING_MAX_EN from DRR mid-frame enable */
+		uint32_t otg_vtotal_mid_frame_num;       /* OTG_V_TOTAL_CONTROL->OTG_VTOTAL_MID_FRAME_NUM from drr_params->vertical_total_mid_frame_num */
+		uint32_t otg_set_v_total_min_mask;       /* OTG_V_TOTAL_CONTROL->OTG_SET_V_TOTAL_MIN_MASK from DRR trigger mask */
+		uint32_t otg_force_lock_on_event;        /* OTG_V_TOTAL_CONTROL->OTG_FORCE_LOCK_ON_EVENT from DRR force lock control */
+
+		/* OPTC Data Source and ODM - 6 fields */
+		uint32_t optc_seg0_src_sel;              /* OPTC_DATA_SOURCE_SELECT->OPTC_SEG0_SRC_SEL from opp_id[0] ODM segment 0 source */
+		uint32_t optc_seg1_src_sel;              /* OPTC_DATA_SOURCE_SELECT->OPTC_SEG1_SRC_SEL from opp_id[1] ODM segment 1 source */
+		uint32_t optc_seg2_src_sel;              /* OPTC_DATA_SOURCE_SELECT->OPTC_SEG2_SRC_SEL from opp_id[2] ODM segment 2 source */
+		uint32_t optc_seg3_src_sel;              /* OPTC_DATA_SOURCE_SELECT->OPTC_SEG3_SRC_SEL from opp_id[3] ODM segment 3 source */
+		uint32_t optc_num_of_input_segment;      /* OPTC_DATA_SOURCE_SELECT->OPTC_NUM_OF_INPUT_SEGMENT from opp_cnt-1 number of input segments */
+		uint32_t optc_mem_sel;                   /* OPTC_MEMORY_CONFIG->OPTC_MEM_SEL from memory_mask ODM memory selection */
+
+		/* OPTC Data Format and DSC - 4 fields */
+		uint32_t optc_data_format;               /* OPTC_DATA_FORMAT_CONTROL->OPTC_DATA_FORMAT from data format selection */
+		uint32_t optc_dsc_mode;                  /* OPTC_DATA_FORMAT_CONTROL->OPTC_DSC_MODE from dsc_mode parameter */
+		uint32_t optc_dsc_bytes_per_pixel;       /* OPTC_BYTES_PER_PIXEL->OPTC_DSC_BYTES_PER_PIXEL from dsc_bytes_per_pixel parameter */
+		uint32_t optc_segment_width;             /* OPTC_WIDTH_CONTROL->OPTC_SEGMENT_WIDTH from segment_width parameter */
+		uint32_t optc_dsc_slice_width;           /* OPTC_WIDTH_CONTROL->OPTC_DSC_SLICE_WIDTH from dsc_slice_width parameter */
+
+		/* OPTC Clock and Underflow Control - 4 fields */
+		uint32_t optc_input_pix_clk_en;          /* OPTC_INPUT_CLOCK_CONTROL->OPTC_INPUT_PIX_CLK_EN from pixel clock enable */
+		uint32_t optc_underflow_occurred_status; /* OPTC_INPUT_GLOBAL_CONTROL->OPTC_UNDERFLOW_OCCURRED_STATUS from underflow status (read-only) */
+		uint32_t optc_underflow_clear;           /* OPTC_INPUT_GLOBAL_CONTROL->OPTC_UNDERFLOW_CLEAR from underflow clear control */
+		uint32_t otg_clock_enable;               /* OTG_CLOCK_CONTROL->OTG_CLOCK_EN from OTG clock enable */
+		uint32_t otg_clock_gate_dis;             /* OTG_CLOCK_CONTROL->OTG_CLOCK_GATE_DIS from clock gate disable */
+
+		/* OTG Stereo and 3D Control - 6 fields */
+		uint32_t otg_stereo_enable;              /* OTG_STEREO_CONTROL->OTG_STEREO_EN from stereo enable control */
+		uint32_t otg_stereo_sync_output_line_num; /* OTG_STEREO_CONTROL->OTG_STEREO_SYNC_OUTPUT_LINE_NUM from timing->stereo_3d_format line num */
+		uint32_t otg_stereo_sync_output_polarity; /* OTG_STEREO_CONTROL->OTG_STEREO_SYNC_OUTPUT_POLARITY from stereo polarity control */
+		uint32_t otg_3d_structure_en;            /* OTG_3D_STRUCTURE_CONTROL->OTG_3D_STRUCTURE_EN from 3D structure enable */
+		uint32_t otg_3d_structure_v_update_mode; /* OTG_3D_STRUCTURE_CONTROL->OTG_3D_STRUCTURE_V_UPDATE_MODE from 3D vertical update mode */
+		uint32_t otg_3d_structure_stereo_sel_ovr; /* OTG_3D_STRUCTURE_CONTROL->OTG_3D_STRUCTURE_STEREO_SEL_OVR from 3D stereo selection override */
+		uint32_t otg_interlace_enable;           /* OTG_INTERLACE_CONTROL->OTG_INTERLACE_ENABLE from dc_crtc_timing->flags.INTERLACE */
+
+		/* OTG GSL (Global Sync Lock) Control - 5 fields */
+		uint32_t otg_gsl0_en;                    /* OTG_GSL_CONTROL->OTG_GSL0_EN from GSL group 0 enable */
+		uint32_t otg_gsl1_en;                    /* OTG_GSL_CONTROL->OTG_GSL1_EN from GSL group 1 enable */
+		uint32_t otg_gsl2_en;                    /* OTG_GSL_CONTROL->OTG_GSL2_EN from GSL group 2 enable */
+		uint32_t otg_gsl_master_en;              /* OTG_GSL_CONTROL->OTG_GSL_MASTER_EN from GSL master enable */
+		uint32_t otg_gsl_master_mode;            /* OTG_GSL_CONTROL->OTG_GSL_MASTER_MODE from gsl_params->gsl_master mode */
+
+		/* OTG DRR Advanced Control - 4 fields */
+		uint32_t otg_v_total_last_used_by_drr;   /* OTG_DRR_CONTROL->OTG_V_TOTAL_LAST_USED_BY_DRR from last used DRR V_TOTAL (read-only) */
+		uint32_t otg_drr_trigger_window_start_x; /* OTG_DRR_TRIGGER_WINDOW->OTG_DRR_TRIGGER_WINDOW_START_X from window_start parameter */
+		uint32_t otg_drr_trigger_window_end_x;   /* OTG_DRR_TRIGGER_WINDOW->OTG_DRR_TRIGGER_WINDOW_END_X from window_end parameter */
+		uint32_t otg_drr_v_total_change_limit;   /* OTG_DRR_V_TOTAL_CHANGE->OTG_DRR_V_TOTAL_CHANGE_LIMIT from limit parameter */
+
+		/* OTG DSC Position Control - 2 fields */
+		uint32_t otg_dsc_start_position_x;       /* OTG_DSC_START_POSITION->OTG_DSC_START_POSITION_X from DSC start X position */
+		uint32_t otg_dsc_start_position_line_num; /* OTG_DSC_START_POSITION->OTG_DSC_START_POSITION_LINE_NUM from DSC start line number */
+
+		/* OTG Double Buffer Control - 2 fields */
+		uint32_t otg_drr_timing_dbuf_update_mode; /* OTG_DOUBLE_BUFFER_CONTROL->OTG_DRR_TIMING_DBUF_UPDATE_MODE from DRR double buffer mode */
+		uint32_t otg_blank_data_double_buffer_en; /* OTG_DOUBLE_BUFFER_CONTROL->OTG_BLANK_DATA_DOUBLE_BUFFER_EN from blank data double buffer enable */
+
+		/* OTG Vertical Interrupts - 6 fields */
+		uint32_t otg_vertical_interrupt0_int_enable; /* OTG_VERTICAL_INTERRUPT0_CONTROL->OTG_VERTICAL_INTERRUPT0_INT_ENABLE from interrupt 0 enable */
+		uint32_t otg_vertical_interrupt0_line_start; /* OTG_VERTICAL_INTERRUPT0_POSITION->OTG_VERTICAL_INTERRUPT0_LINE_START from start_line parameter */
+		uint32_t otg_vertical_interrupt1_int_enable; /* OTG_VERTICAL_INTERRUPT1_CONTROL->OTG_VERTICAL_INTERRUPT1_INT_ENABLE from interrupt 1 enable */
+		uint32_t otg_vertical_interrupt1_line_start; /* OTG_VERTICAL_INTERRUPT1_POSITION->OTG_VERTICAL_INTERRUPT1_LINE_START from start_line parameter */
+		uint32_t otg_vertical_interrupt2_int_enable; /* OTG_VERTICAL_INTERRUPT2_CONTROL->OTG_VERTICAL_INTERRUPT2_INT_ENABLE from interrupt 2 enable */
+		uint32_t otg_vertical_interrupt2_line_start; /* OTG_VERTICAL_INTERRUPT2_POSITION->OTG_VERTICAL_INTERRUPT2_LINE_START from start_line parameter */
+
+		/* OTG Global Sync Parameters - 6 fields */
+		uint32_t otg_vready_offset;              /* OTG_VREADY_PARAM->OTG_VREADY_OFFSET from vready_offset parameter */
+		uint32_t otg_vstartup_start;             /* OTG_VSTARTUP_PARAM->OTG_VSTARTUP_START from vstartup_start parameter */
+		uint32_t otg_vupdate_offset;             /* OTG_VUPDATE_PARAM->OTG_VUPDATE_OFFSET from vupdate_offset parameter */
+		uint32_t otg_vupdate_width;              /* OTG_VUPDATE_PARAM->OTG_VUPDATE_WIDTH from vupdate_width parameter */
+		uint32_t master_update_lock_vupdate_keepout_start_offset; /* OTG_VUPDATE_KEEPOUT->MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_START_OFFSET from pstate_keepout start */
+		uint32_t master_update_lock_vupdate_keepout_end_offset;   /* OTG_VUPDATE_KEEPOUT->MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_END_OFFSET from pstate_keepout end */
+
+		/* OTG Manual Trigger Control - 11 fields */
+		uint32_t otg_triga_source_select;        /* OTG_TRIGA_CNTL->OTG_TRIGA_SOURCE_SELECT from trigger A source selection */
+		uint32_t otg_triga_source_pipe_select;   /* OTG_TRIGA_CNTL->OTG_TRIGA_SOURCE_PIPE_SELECT from trigger A pipe selection */
+		uint32_t otg_triga_rising_edge_detect_cntl; /* OTG_TRIGA_CNTL->OTG_TRIGA_RISING_EDGE_DETECT_CNTL from trigger A rising edge detect */
+		uint32_t otg_triga_falling_edge_detect_cntl; /* OTG_TRIGA_CNTL->OTG_TRIGA_FALLING_EDGE_DETECT_CNTL from trigger A falling edge detect */
+		uint32_t otg_triga_polarity_select;      /* OTG_TRIGA_CNTL->OTG_TRIGA_POLARITY_SELECT from trigger A polarity selection */
+		uint32_t otg_triga_frequency_select;     /* OTG_TRIGA_CNTL->OTG_TRIGA_FREQUENCY_SELECT from trigger A frequency selection */
+		uint32_t otg_triga_delay;                /* OTG_TRIGA_CNTL->OTG_TRIGA_DELAY from trigger A delay */
+		uint32_t otg_triga_clear;                /* OTG_TRIGA_CNTL->OTG_TRIGA_CLEAR from trigger A clear */
+		uint32_t otg_triga_manual_trig;          /* OTG_TRIGA_MANUAL_TRIG->OTG_TRIGA_MANUAL_TRIG from manual trigger A */
+		uint32_t otg_trigb_source_select;        /* OTG_TRIGB_CNTL->OTG_TRIGB_SOURCE_SELECT from trigger B source selection */
+		uint32_t otg_trigb_polarity_select;      /* OTG_TRIGB_CNTL->OTG_TRIGB_POLARITY_SELECT from trigger B polarity selection */
+		uint32_t otg_trigb_manual_trig;          /* OTG_TRIGB_MANUAL_TRIG->OTG_TRIGB_MANUAL_TRIG from manual trigger B */
+
+		/* OTG Static Screen and Update Control - 6 fields */
+		uint32_t otg_static_screen_event_mask;   /* OTG_STATIC_SCREEN_CONTROL->OTG_STATIC_SCREEN_EVENT_MASK from event_triggers parameter */
+		uint32_t otg_static_screen_frame_count;  /* OTG_STATIC_SCREEN_CONTROL->OTG_STATIC_SCREEN_FRAME_COUNT from num_frames parameter */
+		uint32_t master_update_lock;             /* OTG_MASTER_UPDATE_LOCK->MASTER_UPDATE_LOCK from update lock control */
+		uint32_t master_update_mode;             /* OTG_MASTER_UPDATE_MODE->MASTER_UPDATE_MODE from update mode selection */
+		uint32_t otg_force_count_now_mode;       /* OTG_FORCE_COUNT_NOW_CNTL->OTG_FORCE_COUNT_NOW_MODE from force count mode */
+		uint32_t otg_force_count_now_clear;      /* OTG_FORCE_COUNT_NOW_CNTL->OTG_FORCE_COUNT_NOW_CLEAR from force count clear */
+
+		/* VTG Control - 3 fields */
+		uint32_t vtg0_enable;                    /* CONTROL->VTG0_ENABLE from VTG enable control */
+		uint32_t vtg0_fp2;                       /* CONTROL->VTG0_FP2 from VTG front porch 2 */
+		uint32_t vtg0_vcount_init;               /* CONTROL->VTG0_VCOUNT_INIT from VTG vertical count init */
+
+		/* OTG Status (Read-Only) - 12 fields */
+		uint32_t otg_v_blank;                    /* OTG_STATUS->OTG_V_BLANK from vertical blank status (read-only) */
+		uint32_t otg_v_active_disp;              /* OTG_STATUS->OTG_V_ACTIVE_DISP from vertical active display (read-only) */
+		uint32_t otg_frame_count;                /* OTG_STATUS_FRAME_COUNT->OTG_FRAME_COUNT from frame count (read-only) */
+		uint32_t otg_horz_count;                 /* OTG_STATUS_POSITION->OTG_HORZ_COUNT from horizontal position (read-only) */
+		uint32_t otg_vert_count;                 /* OTG_STATUS_POSITION->OTG_VERT_COUNT from vertical position (read-only) */
+		uint32_t otg_horz_count_hv;              /* OTG_STATUS_HV_COUNT->OTG_HORZ_COUNT from horizontal count (read-only) */
+		uint32_t otg_vert_count_nom;             /* OTG_STATUS_HV_COUNT->OTG_VERT_COUNT_NOM from vertical count nominal (read-only) */
+		uint32_t otg_flip_pending;               /* OTG_PIPE_UPDATE_STATUS->OTG_FLIP_PENDING from flip pending status (read-only) */
+		uint32_t otg_dc_reg_update_pending;      /* OTG_PIPE_UPDATE_STATUS->OTG_DC_REG_UPDATE_PENDING from DC register update pending (read-only) */
+		uint32_t otg_cursor_update_pending;      /* OTG_PIPE_UPDATE_STATUS->OTG_CURSOR_UPDATE_PENDING from cursor update pending (read-only) */
+		uint32_t otg_vupdate_keepout_status;     /* OTG_PIPE_UPDATE_STATUS->OTG_VUPDATE_KEEPOUT_STATUS from VUPDATE keepout status (read-only) */
+	} optc[MAX_PIPES];
+
+	/* Metadata */
+	uint32_t active_pipe_count;
+	uint32_t active_stream_count;
+	bool state_valid;
+};
+
+/**
+ * dc_capture_register_software_state() - Capture software state for register programming
+ * @dc: DC context containing current display configuration
+ * @state: Pointer to dc_register_software_state structure to populate
+ *
+ * Extracts all software state variables that are used to program hardware register
+ * fields across the display driver pipeline. This provides a complete snapshot
+ * of the software configuration that drives hardware register programming.
+ *
+ * The function traverses the DC context and extracts values from:
+ * - Stream configurations (timing, format, DSC settings)
+ * - Plane states (surface format, rotation, scaling, cursor)
+ * - Pipe contexts (resource allocation, blending, viewport)
+ * - Clock manager (display clocks, DPP clocks, pixel clocks)
+ * - Resource context (DET buffer allocation, ODM configuration)
+ *
+ * This is essential for underflow debugging as it captures the exact software
+ * state that determines how registers are programmed, allowing analysis of
+ * whether underflow is caused by incorrect register programming or timing issues.
+ *
+ * Return: true if state was successfully captured, false on error
+ */
+bool dc_capture_register_software_state(struct dc *dc, struct dc_register_software_state *state);
+
 #endif /* DC_INTERFACE_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
index fffbf1983143..7b09af1cb306 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
@@ -2084,6 +2084,9 @@ bool dmub_lsdma_init(struct dc_dmub_srv *dc_dmub_srv)
 	struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
 	bool result;
 
+	if (!dc_dmub_srv->dmub->feature_caps.lsdma_support_in_dmu)
+		return false;
+
 	memset(&cmd, 0, sizeof(cmd));
 
 	cmd.cmd_common.header.type     = DMUB_CMD__LSDMA;
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 db669ccb1d58..79e1696def63 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
@@ -1157,6 +1157,16 @@ struct dprx_states {
 	bool cable_id_written;
 };
 
+union dpcd_panel_replay_capability_supported {
+	struct {
+		unsigned char PANEL_REPLAY_SUPPORT          :1;
+		unsigned char SELECTIVE_UPDATE_SUPPORT      :1;
+		unsigned char EARLY_TRANSPORT_SUPPORT       :1;
+		unsigned char RESERVED				        :5;
+	} bits;
+	unsigned char raw;
+};
+
 enum dpcd_downstream_port_max_bpc {
 	DOWN_STREAM_MAX_8BPC = 0,
 	DOWN_STREAM_MAX_10BPC,
@@ -1280,6 +1290,7 @@ struct dpcd_caps {
 	struct edp_psr_info psr_info;
 
 	struct replay_info pr_info;
+	union dpcd_panel_replay_capability_supported pr_caps_supported;
 	uint16_t edp_oled_emission_rate;
 	union dp_receive_port0_cap receive_port0_cap;
 	/* Indicates the number of SST links supported by MSO (Multi-Stream Output) */
@@ -1346,6 +1357,31 @@ union dpcd_replay_configuration {
 	unsigned char raw;
 };
 
+union panel_replay_enable_and_configuration_1 {
+	struct {
+		unsigned char PANEL_REPLAY_ENABLE                     :1;
+		unsigned char PANEL_REPLAY_CRC_ENABLE                 :1;
+		unsigned char IRQ_HPD_ASSDP_MISSING                   :1;
+		unsigned char IRQ_HPD_VSCSDP_UNCORRECTABLE_ERROR      :1;
+		unsigned char IRQ_HPD_RFB_ERROR                       :1;
+		unsigned char IRQ_HPD_ACTIVE_FRAME_CRC_ERROR          :1;
+		unsigned char PANEL_REPLAY_SELECTIVE_UPDATE_ENABLE    :1;
+		unsigned char PANEL_REPLAY_EARLY_TRANSPORT_ENABLE     :1;
+	} bits;
+	unsigned char raw;
+};
+
+union panel_replay_enable_and_configuration_2 {
+	struct {
+		unsigned char SINK_REFRESH_RATE_UNLOCK_GRANTED        :1;
+		unsigned char RESERVED                                :1;
+		unsigned char SU_Y_GRANULARITY_EXT_VALUE_ENABLED      :1;
+		unsigned char SU_Y_GRANULARITY_EXT_VALUE              :4;
+		unsigned char SU_REGION_SCAN_LINE_CAPTURE_INDICATION  :1;
+	} bits;
+	unsigned char raw;
+};
+
 union dpcd_alpm_configuration {
 	struct {
 		unsigned char ENABLE                    : 1;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index ea6b71c43d2c..f46039f64203 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -941,6 +941,12 @@ enum dc_psr_version {
 	DC_PSR_VERSION_UNSUPPORTED		= 0xFFFFFFFF,
 };
 
+enum dc_replay_version {
+	DC_FREESYNC_REPLAY = 0,
+	DC_VESA_PANEL_REPLAY = 1,
+	DC_REPLAY_VERSION_UNSUPPORTED = 0XFF,
+};
+
 /* Possible values of display_endpoint_id.endpoint */
 enum display_endpoint_type {
 	DISPLAY_ENDPOINT_PHY = 0, /* Physical connector. */
@@ -1093,6 +1099,7 @@ enum replay_FW_Message_type {
 	Replay_Set_Residency_Frameupdate_Timer,
 	Replay_Set_Pseudo_VTotal,
 	Replay_Disabled_Adaptive_Sync_SDP,
+	Replay_Set_Version,
 	Replay_Set_General_Cmd,
 };
 
@@ -1128,6 +1135,8 @@ union replay_low_refresh_rate_enable_options {
 };
 
 struct replay_config {
+	/* Replay version */
+	enum dc_replay_version replay_version;
 	/* Replay feature is supported */
 	bool replay_supported;
 	/* Replay caps support DPCD & EDID caps*/
@@ -1184,6 +1193,10 @@ struct replay_settings {
 	uint32_t coasting_vtotal_table[PR_COASTING_TYPE_NUM];
 	/* Defer Update Coasting vtotal table */
 	uint32_t defer_update_coasting_vtotal_table[PR_COASTING_TYPE_NUM];
+	/* Skip frame number table */
+	uint32_t frame_skip_number_table[PR_COASTING_TYPE_NUM];
+	/* Defer skip frame number table */
+	uint32_t defer_frame_skip_number_table[PR_COASTING_TYPE_NUM];
 	/* Maximum link off frame count */
 	uint32_t link_off_frame_count;
 	/* Replay pseudo vtotal for low refresh rate*/
@@ -1192,6 +1205,8 @@ struct replay_settings {
 	uint16_t last_pseudo_vtotal;
 	/* Replay desync error */
 	uint32_t replay_desync_error_fail_count;
+	/* The frame skip number dal send to DMUB */
+	uint16_t frame_skip_number;
 };
 
 /* To split out "global" and "per-panel" config settings.
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 eeed840073fe..fcad61c618a1 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
@@ -1143,7 +1143,8 @@ void dce_aud_wall_dto_setup(
 		REG_UPDATE(DCCG_AUDIO_DTO1_PHASE,
 				DCCG_AUDIO_DTO1_PHASE, clock_info.audio_dto_phase);
 
-		REG_UPDATE(DCCG_AUDIO_DTO_SOURCE,
+		if (aud->masks->DCCG_AUDIO_DTO2_USE_512FBR_DTO)
+			REG_UPDATE(DCCG_AUDIO_DTO_SOURCE,
 				DCCG_AUDIO_DTO2_USE_512FBR_DTO, 1);
 
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index f8996ee2856b..574618d5d4a4 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -1568,7 +1568,7 @@ void dce110_stream_encoder_construct(
 	enc110->se_mask = se_mask;
 }
 
-static const struct stream_encoder_funcs dce110_an_str_enc_funcs = {0};
+static const struct stream_encoder_funcs dce110_an_str_enc_funcs = {};
 
 void dce110_analog_stream_encoder_construct(
 	struct dce110_stream_encoder *enc110,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c
index 39f5fa73c43e..5bfa2b0d2afd 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c
@@ -61,31 +61,49 @@ void dmub_hw_lock_mgr_inbox0_cmd(struct dc_dmub_srv *dmub_srv,
 	dc_dmub_srv_wait_for_inbox0_ack(dmub_srv);
 }
 
-bool should_use_dmub_inbox1_lock(const struct dc *dc, const struct dc_link *link)
+bool dmub_hw_lock_mgr_does_link_require_lock(const struct dc *dc, const struct dc_link *link)
 {
-	/* ASIC doesn't support DMUB */
-	if (!dc->ctx->dmub_srv)
+	if (!link)
 		return false;
 
-	if (link) {
+	if (link->psr_settings.psr_version == DC_PSR_VERSION_SU_1)
+		return true;
 
-		if (link->psr_settings.psr_version == DC_PSR_VERSION_SU_1)
-			return true;
+	if (link->replay_settings.replay_feature_enabled)
+		return true;
 
-		if (link->replay_settings.replay_feature_enabled)
-			return true;
+	if (link->psr_settings.psr_version == DC_PSR_VERSION_1) {
+		struct dc_link *edp_links[MAX_NUM_EDP];
+		int edp_num;
 
-			/* only use HW lock for PSR1 on single eDP */
-		if (link->psr_settings.psr_version == DC_PSR_VERSION_1) {
-			struct dc_link *edp_links[MAX_NUM_EDP];
-			int edp_num;
+		dc_get_edp_links(dc, edp_links, &edp_num);
+		if (edp_num == 1)
+			return true;
+	}
+	return false;
+}
 
-			dc_get_edp_links(dc, edp_links, &edp_num);
+bool dmub_hw_lock_mgr_does_context_require_lock(const struct dc *dc, const struct dc_state *context)
+{
+	if (!context)
+		return false;
+	for (int i = 0; i < context->stream_count; i++) {
+		const struct dc_link *link = context->streams[i]->link;
 
-			if (edp_num == 1)
-				return true;
-		}
+		if (dmub_hw_lock_mgr_does_link_require_lock(dc, link))
+			return true;
 	}
-
 	return false;
 }
+
+bool should_use_dmub_inbox1_lock(const struct dc *dc, const struct dc_link *link)
+{
+	/* ASIC doesn't support DMUB */
+	if (!dc->ctx->dmub_srv)
+		return false;
+
+	if (dc->ctx->dce_version >= DCN_VERSION_4_01)
+		return false;
+
+	return dmub_hw_lock_mgr_does_link_require_lock(dc, link);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.h b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.h
index 9f53d2ea5fa5..4c80ca8484ad 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.h
@@ -46,5 +46,7 @@ void dmub_hw_lock_mgr_inbox0_cmd(struct dc_dmub_srv *dmub_srv,
  * Return: true if the inbox1 lock should be used, false otherwise
  */
 bool should_use_dmub_inbox1_lock(const struct dc *dc, const struct dc_link *link);
+bool dmub_hw_lock_mgr_does_link_require_lock(const struct dc *dc, const struct dc_link *link);
+bool dmub_hw_lock_mgr_does_context_require_lock(const struct dc *dc, const struct dc_state *context);
 
 #endif /*_DMUB_HW_LOCK_MGR_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c
index f9542edff14b..cf1372aaff6c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c
@@ -213,7 +213,8 @@ static bool dmub_replay_copy_settings(struct dmub_replay *dmub,
  */
 static void dmub_replay_set_coasting_vtotal(struct dmub_replay *dmub,
 		uint32_t coasting_vtotal,
-		uint8_t panel_inst)
+		uint8_t panel_inst,
+		uint16_t frame_skip_number)
 {
 	union dmub_rb_cmd cmd;
 	struct dc_context *dc = dmub->ctx;
@@ -227,6 +228,7 @@ static void dmub_replay_set_coasting_vtotal(struct dmub_replay *dmub,
 	pCmd->header.payload_bytes = sizeof(struct dmub_cmd_replay_set_coasting_vtotal_data);
 	pCmd->replay_set_coasting_vtotal_data.coasting_vtotal = (coasting_vtotal & 0xFFFF);
 	pCmd->replay_set_coasting_vtotal_data.coasting_vtotal_high = (coasting_vtotal & 0xFFFF0000) >> 16;
+	pCmd->replay_set_coasting_vtotal_data.frame_skip_number = frame_skip_number;
 
 	dc_wake_and_execute_dmub_cmd(dc, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
 }
@@ -283,7 +285,7 @@ static void dmub_replay_residency(struct dmub_replay *dmub, uint8_t panel_inst,
  * Set REPLAY power optimization flags and coasting vtotal.
  */
 static void dmub_replay_set_power_opt_and_coasting_vtotal(struct dmub_replay *dmub,
-		unsigned int power_opt, uint8_t panel_inst, uint32_t coasting_vtotal)
+		unsigned int power_opt, uint8_t panel_inst, uint32_t coasting_vtotal, uint16_t frame_skip_number)
 {
 	union dmub_rb_cmd cmd;
 	struct dc_context *dc = dmub->ctx;
@@ -301,6 +303,7 @@ static void dmub_replay_set_power_opt_and_coasting_vtotal(struct dmub_replay *dm
 	pCmd->replay_set_power_opt_data.panel_inst = panel_inst;
 	pCmd->replay_set_coasting_vtotal_data.coasting_vtotal = (coasting_vtotal & 0xFFFF);
 	pCmd->replay_set_coasting_vtotal_data.coasting_vtotal_high = (coasting_vtotal & 0xFFFF0000) >> 16;
+	pCmd->replay_set_coasting_vtotal_data.frame_skip_number = frame_skip_number;
 
 	dc_wake_and_execute_dmub_cmd(dc, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
 }
@@ -384,6 +387,19 @@ static void dmub_replay_send_cmd(struct dmub_replay *dmub,
 		cmd.replay_disabled_adaptive_sync_sdp.data.force_disabled =
 			cmd_element->disabled_adaptive_sync_sdp_data.force_disabled;
 		break;
+	case Replay_Set_Version:
+		//Header
+		cmd.replay_set_version.header.sub_type =
+			DMUB_CMD__REPLAY_SET_VERSION;
+		cmd.replay_set_version.header.payload_bytes =
+			sizeof(struct dmub_rb_cmd_replay_set_version) -
+			sizeof(struct dmub_cmd_header);
+		//Cmd Body
+		cmd.replay_set_version.replay_set_version_data.panel_inst =
+			cmd_element->version_data.panel_inst;
+		cmd.replay_set_version.replay_set_version_data.version =
+			cmd_element->version_data.version;
+		break;
 	case Replay_Set_General_Cmd:
 		//Header
 		cmd.replay_set_general_cmd.header.sub_type =
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h
index e6346c0ffc0e..07c79739a980 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h
@@ -27,11 +27,12 @@ struct dmub_replay_funcs {
 	void (*replay_send_cmd)(struct dmub_replay *dmub,
 		enum replay_FW_Message_type msg, union dmub_replay_cmd_set *cmd_element);
 	void (*replay_set_coasting_vtotal)(struct dmub_replay *dmub, uint32_t coasting_vtotal,
-		uint8_t panel_inst);
+		uint8_t panel_inst, uint16_t frame_skip_number);
 	void (*replay_residency)(struct dmub_replay *dmub,
 		uint8_t panel_inst, uint32_t *residency, const bool is_start, const enum pr_residency_mode mode);
 	void (*replay_set_power_opt_and_coasting_vtotal)(struct dmub_replay *dmub,
-		unsigned int power_opt, uint8_t panel_inst, uint32_t coasting_vtotal);
+		unsigned int power_opt, uint8_t panel_inst, uint32_t coasting_vtotal,
+		uint16_t frame_skip_number);
 };
 
 struct dmub_replay *dmub_replay_create(struct dc_context *ctx);
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core.c b/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core.c
index c468f492b876..09303c282495 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core.c
@@ -6711,6 +6711,76 @@ static noinline_for_stack void dml_prefetch_check(struct display_mode_lib_st *mo
 	} // for j
 }
 
+static noinline_for_stack void set_vm_row_and_swath_parameters(struct display_mode_lib_st *mode_lib)
+{
+	struct CalculateVMRowAndSwath_params_st *CalculateVMRowAndSwath_params = &mode_lib->scratch.CalculateVMRowAndSwath_params;
+	struct dml_core_mode_support_locals_st *s = &mode_lib->scratch.dml_core_mode_support_locals;
+
+	CalculateVMRowAndSwath_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
+	CalculateVMRowAndSwath_params->myPipe = s->SurfParameters;
+	CalculateVMRowAndSwath_params->SurfaceSizeInMALL = mode_lib->ms.SurfaceSizeInMALL;
+	CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsLuma = mode_lib->ms.ip.dpte_buffer_size_in_pte_reqs_luma;
+	CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsChroma = mode_lib->ms.ip.dpte_buffer_size_in_pte_reqs_chroma;
+	CalculateVMRowAndSwath_params->DCCMetaBufferSizeBytes = mode_lib->ms.ip.dcc_meta_buffer_size_bytes;
+	CalculateVMRowAndSwath_params->UseMALLForStaticScreen = mode_lib->ms.cache_display_cfg.plane.UseMALLForStaticScreen;
+	CalculateVMRowAndSwath_params->UseMALLForPStateChange = mode_lib->ms.cache_display_cfg.plane.UseMALLForPStateChange;
+	CalculateVMRowAndSwath_params->MALLAllocatedForDCN = mode_lib->ms.soc.mall_allocated_for_dcn_mbytes;
+	CalculateVMRowAndSwath_params->SwathWidthY = mode_lib->ms.SwathWidthYThisState;
+	CalculateVMRowAndSwath_params->SwathWidthC = mode_lib->ms.SwathWidthCThisState;
+	CalculateVMRowAndSwath_params->GPUVMEnable = mode_lib->ms.cache_display_cfg.plane.GPUVMEnable;
+	CalculateVMRowAndSwath_params->HostVMEnable = mode_lib->ms.cache_display_cfg.plane.HostVMEnable;
+	CalculateVMRowAndSwath_params->HostVMMaxNonCachedPageTableLevels = mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels;
+	CalculateVMRowAndSwath_params->GPUVMMaxPageTableLevels = mode_lib->ms.cache_display_cfg.plane.GPUVMMaxPageTableLevels;
+	CalculateVMRowAndSwath_params->GPUVMMinPageSizeKBytes = mode_lib->ms.cache_display_cfg.plane.GPUVMMinPageSizeKBytes;
+	CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024;
+	CalculateVMRowAndSwath_params->PTEBufferModeOverrideEn = mode_lib->ms.cache_display_cfg.plane.PTEBufferModeOverrideEn;
+	CalculateVMRowAndSwath_params->PTEBufferModeOverrideVal = mode_lib->ms.cache_display_cfg.plane.PTEBufferMode;
+	CalculateVMRowAndSwath_params->PTEBufferSizeNotExceeded = mode_lib->ms.PTEBufferSizeNotExceededPerState;
+	CalculateVMRowAndSwath_params->DCCMetaBufferSizeNotExceeded = mode_lib->ms.DCCMetaBufferSizeNotExceededPerState;
+	CalculateVMRowAndSwath_params->dpte_row_width_luma_ub = s->dummy_integer_array[0];
+	CalculateVMRowAndSwath_params->dpte_row_width_chroma_ub = s->dummy_integer_array[1];
+	CalculateVMRowAndSwath_params->dpte_row_height_luma = mode_lib->ms.dpte_row_height;
+	CalculateVMRowAndSwath_params->dpte_row_height_chroma = mode_lib->ms.dpte_row_height_chroma;
+	CalculateVMRowAndSwath_params->dpte_row_height_linear_luma = s->dummy_integer_array[2]; // VBA_DELTA
+	CalculateVMRowAndSwath_params->dpte_row_height_linear_chroma = s->dummy_integer_array[3]; // VBA_DELTA
+	CalculateVMRowAndSwath_params->meta_req_width = s->dummy_integer_array[4];
+	CalculateVMRowAndSwath_params->meta_req_width_chroma = s->dummy_integer_array[5];
+	CalculateVMRowAndSwath_params->meta_req_height = s->dummy_integer_array[6];
+	CalculateVMRowAndSwath_params->meta_req_height_chroma = s->dummy_integer_array[7];
+	CalculateVMRowAndSwath_params->meta_row_width = s->dummy_integer_array[8];
+	CalculateVMRowAndSwath_params->meta_row_width_chroma = s->dummy_integer_array[9];
+	CalculateVMRowAndSwath_params->meta_row_height = mode_lib->ms.meta_row_height;
+	CalculateVMRowAndSwath_params->meta_row_height_chroma = mode_lib->ms.meta_row_height_chroma;
+	CalculateVMRowAndSwath_params->vm_group_bytes = s->dummy_integer_array[10];
+	CalculateVMRowAndSwath_params->dpte_group_bytes = mode_lib->ms.dpte_group_bytes;
+	CalculateVMRowAndSwath_params->PixelPTEReqWidthY = s->dummy_integer_array[11];
+	CalculateVMRowAndSwath_params->PixelPTEReqHeightY = s->dummy_integer_array[12];
+	CalculateVMRowAndSwath_params->PTERequestSizeY = s->dummy_integer_array[13];
+	CalculateVMRowAndSwath_params->PixelPTEReqWidthC = s->dummy_integer_array[14];
+	CalculateVMRowAndSwath_params->PixelPTEReqHeightC = s->dummy_integer_array[15];
+	CalculateVMRowAndSwath_params->PTERequestSizeC = s->dummy_integer_array[16];
+	CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_l = s->dummy_integer_array[17];
+	CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_l = s->dummy_integer_array[18];
+	CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_c = s->dummy_integer_array[19];
+	CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_c = s->dummy_integer_array[20];
+	CalculateVMRowAndSwath_params->PrefetchSourceLinesY = mode_lib->ms.PrefetchLinesYThisState;
+	CalculateVMRowAndSwath_params->PrefetchSourceLinesC = mode_lib->ms.PrefetchLinesCThisState;
+	CalculateVMRowAndSwath_params->VInitPreFillY = mode_lib->ms.PrefillY;
+	CalculateVMRowAndSwath_params->VInitPreFillC = mode_lib->ms.PrefillC;
+	CalculateVMRowAndSwath_params->MaxNumSwathY = mode_lib->ms.MaxNumSwY;
+	CalculateVMRowAndSwath_params->MaxNumSwathC = mode_lib->ms.MaxNumSwC;
+	CalculateVMRowAndSwath_params->meta_row_bw = mode_lib->ms.meta_row_bandwidth_this_state;
+	CalculateVMRowAndSwath_params->dpte_row_bw = mode_lib->ms.dpte_row_bandwidth_this_state;
+	CalculateVMRowAndSwath_params->PixelPTEBytesPerRow = mode_lib->ms.DPTEBytesPerRowThisState;
+	CalculateVMRowAndSwath_params->PDEAndMetaPTEBytesFrame = mode_lib->ms.PDEAndMetaPTEBytesPerFrameThisState;
+	CalculateVMRowAndSwath_params->MetaRowByte = mode_lib->ms.MetaRowBytesThisState;
+	CalculateVMRowAndSwath_params->use_one_row_for_frame = mode_lib->ms.use_one_row_for_frame_this_state;
+	CalculateVMRowAndSwath_params->use_one_row_for_frame_flip = mode_lib->ms.use_one_row_for_frame_flip_this_state;
+	CalculateVMRowAndSwath_params->UsesMALLForStaticScreen = s->dummy_boolean_array[0];
+	CalculateVMRowAndSwath_params->PTE_BUFFER_MODE = s->dummy_boolean_array[1];
+	CalculateVMRowAndSwath_params->BIGK_FRAGMENT_SIZE = s->dummy_integer_array[21];
+}
+
 /// @brief The Mode Support function.
 dml_bool_t dml_core_mode_support(struct display_mode_lib_st *mode_lib)
 {
@@ -7683,69 +7753,7 @@ dml_bool_t dml_core_mode_support(struct display_mode_lib_st *mode_lib)
 			s->SurfParameters[k].SwathHeightC = mode_lib->ms.SwathHeightCThisState[k];
 		}
 
-		CalculateVMRowAndSwath_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
-		CalculateVMRowAndSwath_params->myPipe = s->SurfParameters;
-		CalculateVMRowAndSwath_params->SurfaceSizeInMALL = mode_lib->ms.SurfaceSizeInMALL;
-		CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsLuma = mode_lib->ms.ip.dpte_buffer_size_in_pte_reqs_luma;
-		CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsChroma = mode_lib->ms.ip.dpte_buffer_size_in_pte_reqs_chroma;
-		CalculateVMRowAndSwath_params->DCCMetaBufferSizeBytes = mode_lib->ms.ip.dcc_meta_buffer_size_bytes;
-		CalculateVMRowAndSwath_params->UseMALLForStaticScreen = mode_lib->ms.cache_display_cfg.plane.UseMALLForStaticScreen;
-		CalculateVMRowAndSwath_params->UseMALLForPStateChange = mode_lib->ms.cache_display_cfg.plane.UseMALLForPStateChange;
-		CalculateVMRowAndSwath_params->MALLAllocatedForDCN = mode_lib->ms.soc.mall_allocated_for_dcn_mbytes;
-		CalculateVMRowAndSwath_params->SwathWidthY = mode_lib->ms.SwathWidthYThisState;
-		CalculateVMRowAndSwath_params->SwathWidthC = mode_lib->ms.SwathWidthCThisState;
-		CalculateVMRowAndSwath_params->GPUVMEnable = mode_lib->ms.cache_display_cfg.plane.GPUVMEnable;
-		CalculateVMRowAndSwath_params->HostVMEnable = mode_lib->ms.cache_display_cfg.plane.HostVMEnable;
-		CalculateVMRowAndSwath_params->HostVMMaxNonCachedPageTableLevels = mode_lib->ms.cache_display_cfg.plane.HostVMMaxPageTableLevels;
-		CalculateVMRowAndSwath_params->GPUVMMaxPageTableLevels = mode_lib->ms.cache_display_cfg.plane.GPUVMMaxPageTableLevels;
-		CalculateVMRowAndSwath_params->GPUVMMinPageSizeKBytes = mode_lib->ms.cache_display_cfg.plane.GPUVMMinPageSizeKBytes;
-		CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->ms.soc.hostvm_min_page_size_kbytes * 1024;
-		CalculateVMRowAndSwath_params->PTEBufferModeOverrideEn = mode_lib->ms.cache_display_cfg.plane.PTEBufferModeOverrideEn;
-		CalculateVMRowAndSwath_params->PTEBufferModeOverrideVal = mode_lib->ms.cache_display_cfg.plane.PTEBufferMode;
-		CalculateVMRowAndSwath_params->PTEBufferSizeNotExceeded = mode_lib->ms.PTEBufferSizeNotExceededPerState;
-		CalculateVMRowAndSwath_params->DCCMetaBufferSizeNotExceeded = mode_lib->ms.DCCMetaBufferSizeNotExceededPerState;
-		CalculateVMRowAndSwath_params->dpte_row_width_luma_ub = s->dummy_integer_array[0];
-		CalculateVMRowAndSwath_params->dpte_row_width_chroma_ub = s->dummy_integer_array[1];
-		CalculateVMRowAndSwath_params->dpte_row_height_luma = mode_lib->ms.dpte_row_height;
-		CalculateVMRowAndSwath_params->dpte_row_height_chroma = mode_lib->ms.dpte_row_height_chroma;
-		CalculateVMRowAndSwath_params->dpte_row_height_linear_luma = s->dummy_integer_array[2]; // VBA_DELTA
-		CalculateVMRowAndSwath_params->dpte_row_height_linear_chroma = s->dummy_integer_array[3]; // VBA_DELTA
-		CalculateVMRowAndSwath_params->meta_req_width = s->dummy_integer_array[4];
-		CalculateVMRowAndSwath_params->meta_req_width_chroma = s->dummy_integer_array[5];
-		CalculateVMRowAndSwath_params->meta_req_height = s->dummy_integer_array[6];
-		CalculateVMRowAndSwath_params->meta_req_height_chroma = s->dummy_integer_array[7];
-		CalculateVMRowAndSwath_params->meta_row_width = s->dummy_integer_array[8];
-		CalculateVMRowAndSwath_params->meta_row_width_chroma = s->dummy_integer_array[9];
-		CalculateVMRowAndSwath_params->meta_row_height = mode_lib->ms.meta_row_height;
-		CalculateVMRowAndSwath_params->meta_row_height_chroma = mode_lib->ms.meta_row_height_chroma;
-		CalculateVMRowAndSwath_params->vm_group_bytes = s->dummy_integer_array[10];
-		CalculateVMRowAndSwath_params->dpte_group_bytes = mode_lib->ms.dpte_group_bytes;
-		CalculateVMRowAndSwath_params->PixelPTEReqWidthY = s->dummy_integer_array[11];
-		CalculateVMRowAndSwath_params->PixelPTEReqHeightY = s->dummy_integer_array[12];
-		CalculateVMRowAndSwath_params->PTERequestSizeY = s->dummy_integer_array[13];
-		CalculateVMRowAndSwath_params->PixelPTEReqWidthC = s->dummy_integer_array[14];
-		CalculateVMRowAndSwath_params->PixelPTEReqHeightC = s->dummy_integer_array[15];
-		CalculateVMRowAndSwath_params->PTERequestSizeC = s->dummy_integer_array[16];
-		CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_l = s->dummy_integer_array[17];
-		CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_l = s->dummy_integer_array[18];
-		CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_c = s->dummy_integer_array[19];
-		CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_c = s->dummy_integer_array[20];
-		CalculateVMRowAndSwath_params->PrefetchSourceLinesY = mode_lib->ms.PrefetchLinesYThisState;
-		CalculateVMRowAndSwath_params->PrefetchSourceLinesC = mode_lib->ms.PrefetchLinesCThisState;
-		CalculateVMRowAndSwath_params->VInitPreFillY = mode_lib->ms.PrefillY;
-		CalculateVMRowAndSwath_params->VInitPreFillC = mode_lib->ms.PrefillC;
-		CalculateVMRowAndSwath_params->MaxNumSwathY = mode_lib->ms.MaxNumSwY;
-		CalculateVMRowAndSwath_params->MaxNumSwathC = mode_lib->ms.MaxNumSwC;
-		CalculateVMRowAndSwath_params->meta_row_bw = mode_lib->ms.meta_row_bandwidth_this_state;
-		CalculateVMRowAndSwath_params->dpte_row_bw = mode_lib->ms.dpte_row_bandwidth_this_state;
-		CalculateVMRowAndSwath_params->PixelPTEBytesPerRow = mode_lib->ms.DPTEBytesPerRowThisState;
-		CalculateVMRowAndSwath_params->PDEAndMetaPTEBytesFrame = mode_lib->ms.PDEAndMetaPTEBytesPerFrameThisState;
-		CalculateVMRowAndSwath_params->MetaRowByte = mode_lib->ms.MetaRowBytesThisState;
-		CalculateVMRowAndSwath_params->use_one_row_for_frame = mode_lib->ms.use_one_row_for_frame_this_state;
-		CalculateVMRowAndSwath_params->use_one_row_for_frame_flip = mode_lib->ms.use_one_row_for_frame_flip_this_state;
-		CalculateVMRowAndSwath_params->UsesMALLForStaticScreen = s->dummy_boolean_array[0];
-		CalculateVMRowAndSwath_params->PTE_BUFFER_MODE = s->dummy_boolean_array[1];
-		CalculateVMRowAndSwath_params->BIGK_FRAGMENT_SIZE = s->dummy_integer_array[21];
+		set_vm_row_and_swath_parameters(mode_lib);
 
 		CalculateVMRowAndSwath(&mode_lib->scratch,
 			CalculateVMRowAndSwath_params);
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core_structs.h b/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core_structs.h
index 3b1d92e7697f..5b40dcdc4406 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/display_mode_core_structs.h
@@ -1893,7 +1893,7 @@ struct display_mode_lib_scratch_st {
 	struct CalculatePrefetchSchedule_params_st CalculatePrefetchSchedule_params;
 };
 
-/// @brief Represent the overall soc/ip enviroment. It contains data structure represent the soc/ip characteristic and also structures that hold calculation output
+/// @brief Represent the overall soc/ip environment. It contains data structure represent the soc/ip characteristic and also structures that hold calculation output
 struct display_mode_lib_st {
 	dml_uint_t project;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_dchub_registers.h b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_dchub_registers.h
index 8e5a30287220..bf57df42d1d9 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_dchub_registers.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_dchub_registers.h
@@ -121,6 +121,8 @@ struct dml2_display_rq_regs {
 	uint32_t crq_expansion_mode;
 	uint32_t plane1_base_address;
 	uint32_t unbounded_request_enabled;
+	bool pte_buffer_mode;
+	bool force_one_row_for_frame;
 
 	// MRQ
 	uint32_t mrq_expansion_mode;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_display_cfg_types.h b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_display_cfg_types.h
index 13749c9fcf18..35aa954248cd 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_display_cfg_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_display_cfg_types.h
@@ -87,6 +87,15 @@ enum dml2_output_link_dp_rate {
 	dml2_dp_rate_uhbr20 = 6
 };
 
+enum dml2_pstate_type {
+	dml2_pstate_type_uclk = 0,
+	dml2_pstate_type_fclk = 1,
+	dml2_pstate_type_ppt = 2,
+	dml2_pstate_type_temp_read = 3,
+	dml2_pstate_type_dummy_pstate = 4,
+	dml2_pstate_type_count = 5
+};
+
 enum dml2_uclk_pstate_change_strategy {
 	dml2_uclk_pstate_change_strategy_auto = 0,
 	dml2_uclk_pstate_change_strategy_force_vactive = 1,
@@ -393,8 +402,7 @@ struct dml2_plane_parameters {
 		// reserved_vblank_time_ns is the minimum time to reserve in vblank for Twait
 		// The actual reserved vblank time used for the corresponding stream in mode_programming would be at least as much as this per-plane override.
 		long reserved_vblank_time_ns;
-		unsigned int max_vactive_det_fill_delay_us; // 0 = no reserved time, +ve = explicit max delay
-		unsigned int vactive_latency_to_hide_for_pstate_admissibility_us;
+		unsigned int max_vactive_det_fill_delay_us[dml2_pstate_type_count]; // 0 = no reserved time, +ve = explicit max delay
 		unsigned int gpuvm_min_page_size_kbytes;
 		unsigned int hostvm_min_page_size_kbytes;
 
@@ -423,7 +431,6 @@ struct dml2_stream_parameters {
 		bool disable_dynamic_odm;
 		bool disable_subvp;
 		int minimum_vblank_idle_requirement_us;
-		bool minimize_active_latency_hiding;
 
 		struct {
 			struct {
@@ -489,7 +496,6 @@ struct dml2_display_cfg {
 		bool synchronize_ddr_displays_for_uclk_pstate_change;
 		bool max_outstanding_when_urgent_expected_disable;
 		bool enable_subvp_implicit_pmo; //enables PMO to switch pipe uclk strategy to subvp, and generate phantom programming
-		unsigned int best_effort_min_active_latency_hiding_us;
 		bool all_streams_blanked;
 	} overrides;
 };
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_soc_parameter_types.h b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_soc_parameter_types.h
index 4a9a0d5a09b7..1fbc520c2540 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_soc_parameter_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_soc_parameter_types.h
@@ -89,8 +89,8 @@ struct dml2_soc_qos_parameters {
 
 struct dml2_soc_power_management_parameters {
 	double dram_clk_change_blackout_us;
-	double dram_clk_change_read_only_us;
-	double dram_clk_change_write_only_us;
+	double dram_clk_change_read_only_us; // deprecated
+	double dram_clk_change_write_only_us; // deprecated
 	double fclk_change_blackout_us;
 	double g7_ppt_blackout_us;
 	double g7_temperature_read_blackout_us;
@@ -191,7 +191,7 @@ struct dml2_ip_capabilities {
 	unsigned int subvp_prefetch_end_to_mall_start_us;
 	unsigned int subvp_fw_processing_delay;
 	unsigned int max_vactive_det_fill_delay_us;
-	unsigned int ppt_max_allow_delay_ns;
+	unsigned int ppt_max_allow_delay_us;
 	unsigned int temp_read_max_allow_delay_us;
 	unsigned int dummy_pstate_max_allow_delay_us;
 	/* FAMS2 delays */
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_types.h b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_types.h
index 8646ce5f1c01..452e4a2e72c0 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/inc/dml_top_types.h
@@ -195,14 +195,6 @@ struct dml2_mcache_surface_allocation {
 	} informative;
 };
 
-enum dml2_pstate_type {
-	dml2_pstate_type_uclk,
-	dml2_pstate_type_ppt,
-	dml2_pstate_type_temp_read,
-	dml2_pstate_type_dummy_pstate,
-	dml2_pstate_type_count
-};
-
 enum dml2_pstate_method {
 	dml2_pstate_method_na = 0,
 	/* hw exclusive modes */
@@ -684,6 +676,8 @@ struct dml2_display_cfg_programming {
 			unsigned int PrefetchMode[DML2_MAX_PLANES]; // LEGACY_ONLY
 			bool ROBUrgencyAvoidance;
 			double LowestPrefetchMargin;
+
+			unsigned int pstate_recout_reduction_lines[DML2_MAX_PLANES];
 		} misc;
 
 		struct dml2_mode_support_info mode_support_info;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_dcn4_calcs.c b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_dcn4_calcs.c
index f809c4073b43..a02e9fd6b5ca 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_dcn4_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_dcn4_calcs.c
@@ -6972,7 +6972,7 @@ static void calculate_bytes_to_fetch_required_to_hide_latency(
 
 		stream_index = p->display_cfg->plane_descriptors[plane_index].stream_index;
 
-		dst_lines_to_hide = (unsigned int)math_ceil(p->latency_to_hide_us /
+		dst_lines_to_hide = (unsigned int)math_ceil(p->latency_to_hide_us[0] /
 			((double)p->display_cfg->stream_descriptors[stream_index].timing.h_total /
 				(double)p->display_cfg->stream_descriptors[stream_index].timing.pixel_clock_khz * 1000.0));
 
@@ -7069,9 +7069,9 @@ static void calculate_excess_vactive_bandwidth_required(
 		excess_vactive_fill_bw_l[plane_index] = 0.0;
 		excess_vactive_fill_bw_c[plane_index] = 0.0;
 
-		if (display_cfg->plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us > 0) {
-			excess_vactive_fill_bw_l[plane_index] = (double)bytes_required_l[plane_index] / (double)display_cfg->plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us;
-			excess_vactive_fill_bw_c[plane_index] = (double)bytes_required_c[plane_index] / (double)display_cfg->plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us;
+		if (display_cfg->plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us[dml2_pstate_type_uclk] > 0) {
+			excess_vactive_fill_bw_l[plane_index] = (double)bytes_required_l[plane_index] / (double)display_cfg->plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us[dml2_pstate_type_uclk];
+			excess_vactive_fill_bw_c[plane_index] = (double)bytes_required_c[plane_index] / (double)display_cfg->plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us[dml2_pstate_type_uclk];
 		}
 	}
 }
@@ -9051,11 +9051,11 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
 	calculate_bytes_to_fetch_required_to_hide_latency_params->swath_width_c = mode_lib->ms.SwathWidthC;
 	calculate_bytes_to_fetch_required_to_hide_latency_params->swath_height_l = mode_lib->ms.SwathHeightY;
 	calculate_bytes_to_fetch_required_to_hide_latency_params->swath_height_c = mode_lib->ms.SwathHeightC;
-	calculate_bytes_to_fetch_required_to_hide_latency_params->latency_to_hide_us = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
+	calculate_bytes_to_fetch_required_to_hide_latency_params->latency_to_hide_us[0] = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
 
 	/* outputs */
-	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_l = s->pstate_bytes_required_l;
-	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_c = s->pstate_bytes_required_c;
+	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_l = s->pstate_bytes_required_l[dml2_pstate_type_uclk];
+	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_c = s->pstate_bytes_required_c[dml2_pstate_type_uclk];
 
 	calculate_bytes_to_fetch_required_to_hide_latency(calculate_bytes_to_fetch_required_to_hide_latency_params);
 
@@ -9063,8 +9063,8 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
 	calculate_excess_vactive_bandwidth_required(
 			display_cfg,
 			mode_lib->ms.num_active_planes,
-			s->pstate_bytes_required_l,
-			s->pstate_bytes_required_c,
+			s->pstate_bytes_required_l[dml2_pstate_type_uclk],
+			s->pstate_bytes_required_c[dml2_pstate_type_uclk],
 			/* outputs */
 			mode_lib->ms.excess_vactive_fill_bw_l,
 			mode_lib->ms.excess_vactive_fill_bw_c);
@@ -9506,8 +9506,8 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
 	calculate_vactive_det_fill_latency(
 			display_cfg,
 			mode_lib->ms.num_active_planes,
-			s->pstate_bytes_required_l,
-			s->pstate_bytes_required_c,
+			s->pstate_bytes_required_l[dml2_pstate_type_uclk],
+			s->pstate_bytes_required_c[dml2_pstate_type_uclk],
 			mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0,
 			mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1,
 			mode_lib->ms.vactive_sw_bw_l,
@@ -9515,7 +9515,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
 			mode_lib->ms.surface_avg_vactive_required_bw,
 			mode_lib->ms.surface_peak_required_bw,
 			/* outputs */
-			mode_lib->ms.dram_change_vactive_det_fill_delay_us);
+			mode_lib->ms.pstate_vactive_det_fill_delay_us[dml2_pstate_type_uclk]);
 
 #ifdef __DML_VBA_DEBUG__
 	DML_LOG_VERBOSE("DML::%s: max_urgent_latency_us = %f\n", __func__, s->mSOCParameters.max_urgent_latency_us);
@@ -11009,11 +11009,11 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
 	calculate_bytes_to_fetch_required_to_hide_latency_params->swath_width_c = mode_lib->mp.SwathWidthC;
 	calculate_bytes_to_fetch_required_to_hide_latency_params->swath_height_l = mode_lib->mp.SwathHeightY;
 	calculate_bytes_to_fetch_required_to_hide_latency_params->swath_height_c = mode_lib->mp.SwathHeightC;
-	calculate_bytes_to_fetch_required_to_hide_latency_params->latency_to_hide_us = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
+	calculate_bytes_to_fetch_required_to_hide_latency_params->latency_to_hide_us[0] = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
 
 	/* outputs */
-	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_l = s->pstate_bytes_required_l;
-	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_c = s->pstate_bytes_required_c;
+	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_l = s->pstate_bytes_required_l[dml2_pstate_type_uclk];
+	calculate_bytes_to_fetch_required_to_hide_latency_params->bytes_required_c = s->pstate_bytes_required_c[dml2_pstate_type_uclk];
 
 	calculate_bytes_to_fetch_required_to_hide_latency(calculate_bytes_to_fetch_required_to_hide_latency_params);
 
@@ -11021,8 +11021,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
 	calculate_excess_vactive_bandwidth_required(
 			display_cfg,
 			s->num_active_planes,
-			s->pstate_bytes_required_l,
-			s->pstate_bytes_required_c,
+			s->pstate_bytes_required_l[dml2_pstate_type_uclk],
+			s->pstate_bytes_required_c[dml2_pstate_type_uclk],
 			/* outputs */
 			mode_lib->mp.excess_vactive_fill_bw_l,
 			mode_lib->mp.excess_vactive_fill_bw_c);
@@ -12943,7 +12943,8 @@ void dml2_core_calcs_get_plane_support_info(const struct dml2_display_cfg *displ
 
 	out->active_latency_hiding_us = (int)mode_lib->ms.VActiveLatencyHidingUs[plane_idx];
 
-	out->dram_change_vactive_det_fill_delay_us = (unsigned int)math_ceil(mode_lib->ms.dram_change_vactive_det_fill_delay_us[plane_idx]);
+	out->vactive_det_fill_delay_us[dml2_pstate_type_uclk] =
+			(unsigned int)math_ceil(mode_lib->ms.pstate_vactive_det_fill_delay_us[dml2_pstate_type_uclk][plane_idx]);
 }
 
 void dml2_core_calcs_get_stream_support_info(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct core_stream_support_info *out, int plane_index)
@@ -13024,7 +13025,7 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
 	out->informative.mode_support_info.InvalidCombinationOfMALLUseForPState = mode_lib->ms.support.InvalidCombinationOfMALLUseForPState;
 	out->informative.mode_support_info.ExceededMALLSize = mode_lib->ms.support.ExceededMALLSize;
 	out->informative.mode_support_info.EnoughWritebackUnits = mode_lib->ms.support.EnoughWritebackUnits;
-	out->informative.mode_support_info.temp_read_or_ppt_support = mode_lib->ms.support.temp_read_or_ppt_support;
+	out->informative.mode_support_info.temp_read_or_ppt_support = mode_lib->ms.support.global_temp_read_or_ppt_supported;
 	out->informative.mode_support_info.g6_temp_read_support = mode_lib->ms.support.g6_temp_read_support;
 
 	out->informative.mode_support_info.ExceededMultistreamSlots = mode_lib->ms.support.ExceededMultistreamSlots;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_shared_types.h b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_shared_types.h
index 051c31ec2f0e..1087a8c926ff 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_shared_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_shared_types.h
@@ -264,8 +264,11 @@ struct dml2_core_internal_mode_support_info {
 	bool DCCMetaBufferSizeNotExceeded;
 	enum dml2_pstate_change_support DRAMClockChangeSupport[DML2_MAX_PLANES];
 	enum dml2_pstate_change_support FCLKChangeSupport[DML2_MAX_PLANES];
+	enum dml2_pstate_change_support temp_read_or_ppt_support[DML2_MAX_PLANES];
+	bool global_dram_clock_change_support_required;
 	bool global_dram_clock_change_supported;
 	bool global_fclk_change_supported;
+	bool global_temp_read_or_ppt_supported;
 	bool USRRetrainingSupport;
 	bool AvgBandwidthSupport;
 	bool UrgVactiveBandwidthSupport;
@@ -336,7 +339,6 @@ struct dml2_core_internal_mode_support_info {
 	bool incorrect_imall_usage;
 
 	bool g6_temp_read_support;
-	bool temp_read_or_ppt_support;
 
 	struct dml2_core_internal_watermarks watermarks;
 	bool dcfclk_support;
@@ -591,7 +593,7 @@ struct dml2_core_internal_mode_support {
 	double VActiveLatencyHidingMargin[DML2_MAX_PLANES];
 	double VActiveLatencyHidingUs[DML2_MAX_PLANES];
 	unsigned int MaxVStartupLines[DML2_MAX_PLANES];
-	double dram_change_vactive_det_fill_delay_us[DML2_MAX_PLANES];
+	double pstate_vactive_det_fill_delay_us[dml2_pstate_type_count][DML2_MAX_PLANES];
 
 	unsigned int num_mcaches_l[DML2_MAX_PLANES];
 	unsigned int mcache_row_bytes_l[DML2_MAX_PLANES];
@@ -621,8 +623,8 @@ struct dml2_core_internal_mode_support {
 	unsigned int dpte_row_bytes_per_row_l[DML2_MAX_PLANES];
 	unsigned int dpte_row_bytes_per_row_c[DML2_MAX_PLANES];
 
-	unsigned int pstate_bytes_required_l[DML2_MAX_PLANES];
-	unsigned int pstate_bytes_required_c[DML2_MAX_PLANES];
+	unsigned int pstate_bytes_required_l[dml2_pstate_type_count][DML2_MAX_PLANES];
+	unsigned int pstate_bytes_required_c[dml2_pstate_type_count][DML2_MAX_PLANES];
 	unsigned int cursor_bytes_per_chunk[DML2_MAX_PLANES];
 	unsigned int cursor_bytes_per_line[DML2_MAX_PLANES];
 
@@ -646,7 +648,7 @@ struct dml2_core_internal_mode_support {
 	unsigned int DSTYAfterScaler[DML2_MAX_PLANES];
 	unsigned int DSTXAfterScaler[DML2_MAX_PLANES];
 
-	enum dml2_pstate_method pstate_switch_modes[DML2_MAX_PLANES];
+	enum dml2_pstate_method uclk_pstate_switch_modes[DML2_MAX_PLANES];
 };
 
 /// @brief A mega structure that houses various info for model programming step.
@@ -837,6 +839,7 @@ struct dml2_core_internal_mode_program {
 	double max_urgent_latency_us;
 	double df_response_time_us;
 
+	enum dml2_pstate_method uclk_pstate_switch_modes[DML2_MAX_PLANES];
 	// -------------------
 	// Output
 	// -------------------
@@ -963,11 +966,12 @@ struct dml2_core_internal_mode_program {
 	double MaxActiveFCLKChangeLatencySupported;
 	bool USRRetrainingSupport;
 	bool g6_temp_read_support;
-	bool temp_read_or_ppt_support;
 	enum dml2_pstate_change_support FCLKChangeSupport[DML2_MAX_PLANES];
 	enum dml2_pstate_change_support DRAMClockChangeSupport[DML2_MAX_PLANES];
+	enum dml2_pstate_change_support temp_read_or_ppt_support[DML2_MAX_PLANES];
 	bool global_dram_clock_change_supported;
 	bool global_fclk_change_supported;
+	bool global_temp_read_or_ppt_supported;
 	double MaxActiveDRAMClockChangeLatencySupported[DML2_MAX_PLANES];
 	double WritebackAllowFCLKChangeEndPosition[DML2_MAX_PLANES];
 	double WritebackAllowDRAMClockChangeEndPosition[DML2_MAX_PLANES];
@@ -1134,8 +1138,8 @@ struct dml2_core_calcs_mode_support_locals {
 	unsigned int cursor_bytes[DML2_MAX_PLANES];
 	bool stream_visited[DML2_MAX_PLANES];
 
-	unsigned int pstate_bytes_required_l[DML2_MAX_PLANES];
-	unsigned int pstate_bytes_required_c[DML2_MAX_PLANES];
+	unsigned int pstate_bytes_required_l[dml2_pstate_type_count][DML2_MAX_PLANES];
+	unsigned int pstate_bytes_required_c[dml2_pstate_type_count][DML2_MAX_PLANES];
 
 	double prefetch_sw_bytes[DML2_MAX_PLANES];
 	double Tpre_rounded[DML2_MAX_PLANES];
@@ -1226,8 +1230,8 @@ struct dml2_core_calcs_mode_programming_locals {
 	double Tr0_trips_flip_rounded[DML2_MAX_PLANES];
 	unsigned int per_pipe_flip_bytes[DML2_MAX_PLANES];
 
-	unsigned int pstate_bytes_required_l[DML2_MAX_PLANES];
-	unsigned int pstate_bytes_required_c[DML2_MAX_PLANES];
+	unsigned int pstate_bytes_required_l[dml2_pstate_type_count][DML2_MAX_PLANES];
+	unsigned int pstate_bytes_required_c[dml2_pstate_type_count][DML2_MAX_PLANES];
 
 	double prefetch_sw_bytes[DML2_MAX_PLANES];
 	double Tpre_rounded[DML2_MAX_PLANES];
@@ -1313,7 +1317,7 @@ struct dml2_core_calcs_CalculateVMRowAndSwath_params {
 	unsigned int HostVMMinPageSize;
 	unsigned int DCCMetaBufferSizeBytes;
 	bool mrq_present;
-	enum dml2_pstate_method *pstate_switch_modes;
+	enum dml2_pstate_method *uclk_pstate_switch_modes;
 
 	// Output
 	bool *PTEBufferSizeNotExceeded;
@@ -1740,10 +1744,12 @@ struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_param
 	unsigned int max_request_size_bytes;
 	unsigned int *meta_row_height_l;
 	unsigned int *meta_row_height_c;
+	enum dml2_pstate_method *uclk_pstate_switch_modes;
 
 	// Output
 	struct dml2_core_internal_watermarks *Watermark;
 	enum dml2_pstate_change_support *DRAMClockChangeSupport;
+	bool *global_dram_clock_change_support_required;
 	bool *global_dram_clock_change_supported;
 	double *MaxActiveDRAMClockChangeLatencySupported;
 	unsigned int *SubViewportLinesNeededInMALL;
@@ -1754,10 +1760,10 @@ struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_param
 	double *VActiveLatencyHidingMargin;
 	double *VActiveLatencyHidingUs;
 	bool *g6_temp_read_support;
-	bool *temp_read_or_ppt_support;
+	enum dml2_pstate_change_support *temp_read_or_ppt_support;
+	bool *global_temp_read_or_ppt_supported;
 };
 
-
 struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params {
 	const struct dml2_display_cfg *display_cfg;
 	unsigned int ConfigReturnBufferSizeInKByte;
@@ -2247,7 +2253,7 @@ struct dml2_core_calcs_calculate_bytes_to_fetch_required_to_hide_latency_params
 	unsigned int *swath_width_c;
 	unsigned int *swath_height_l;
 	unsigned int *swath_height_c;
-	double latency_to_hide_us;
+	double latency_to_hide_us[DML2_MAX_PLANES];
 
 	/* outputs */
 	unsigned int *bytes_required_l;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_utils.c b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_utils.c
index 5f301befed16..b57d0f6ea6a1 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_utils.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_utils.c
@@ -306,6 +306,8 @@ void dml2_core_utils_print_mode_support_info(const struct dml2_core_internal_mod
 		DML_LOG_VERBOSE("DML: support: ExceededMALLSize = %d\n", support->ExceededMALLSize);
 	if (!fail_only || support->g6_temp_read_support == 0)
 		DML_LOG_VERBOSE("DML: support: g6_temp_read_support = %d\n", support->g6_temp_read_support);
+	if (!fail_only || (support->global_dram_clock_change_supported == 0 && support->global_dram_clock_change_support_required))
+		DML_LOG_VERBOSE("DML: support: dram_clock_change_support = %d\n", support->global_dram_clock_change_supported);
 	if (!fail_only || support->ImmediateFlipSupport == 0)
 		DML_LOG_VERBOSE("DML: support: ImmediateFlipSupport = %d\n", support->ImmediateFlipSupport);
 	if (!fail_only || support->LinkCapacitySupport == 0)
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c
index 5769c2638f9a..c26e100fcaf2 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c
@@ -1087,7 +1087,7 @@ static bool all_timings_support_drr(const struct dml2_pmo_instance *pmo,
 
 			/* check required stretch is allowed */
 			if (stream_descriptor->timing.drr_config.max_instant_vtotal_delta > 0 &&
-					stream_pstate_meta->method_drr.stretched_vtotal - stream_pstate_meta->nom_vtotal > stream_descriptor->timing.drr_config.max_instant_vtotal_delta) {
+					stream_pstate_meta->method_drr.stretched_vtotal - stream_pstate_meta->nom_vtotal > (int)stream_descriptor->timing.drr_config.max_instant_vtotal_delta) {
 				return false;
 			}
 		}
@@ -1669,15 +1669,15 @@ static int get_vactive_pstate_margin(const struct display_configuation_with_meta
 	return min_vactive_margin_us;
 }
 
-static unsigned int get_vactive_det_fill_latency_delay_us(const struct display_configuation_with_meta *display_cfg, int plane_mask)
+static int get_vactive_det_fill_latency_delay_us(const struct display_configuation_with_meta *display_cfg, int plane_mask)
 {
 	unsigned char i;
-	unsigned int max_vactive_fill_us = 0;
+	int max_vactive_fill_us = 0;
 
 	for (i = 0; i < DML2_MAX_PLANES; i++) {
 		if (is_bit_set_in_bitfield(plane_mask, i)) {
-			if (display_cfg->mode_support_result.cfg_support_info.plane_support_info[i].dram_change_vactive_det_fill_delay_us > max_vactive_fill_us)
-				max_vactive_fill_us = display_cfg->mode_support_result.cfg_support_info.plane_support_info[i].dram_change_vactive_det_fill_delay_us;
+			if (display_cfg->mode_support_result.cfg_support_info.plane_support_info[i].vactive_det_fill_delay_us[dml2_pstate_type_uclk] > max_vactive_fill_us)
+				max_vactive_fill_us = display_cfg->mode_support_result.cfg_support_info.plane_support_info[i].vactive_det_fill_delay_us[dml2_pstate_type_uclk];
 		}
 	}
 
@@ -1962,9 +1962,6 @@ static void reset_display_configuration(struct display_configuation_with_meta *d
 
 	for (stream_index = 0; stream_index < display_config->display_config.num_streams; stream_index++) {
 		display_config->stage3.stream_svp_meta[stream_index].valid = false;
-
-		display_config->display_config.stream_descriptors[stream_index].overrides.minimize_active_latency_hiding = false;
-		display_config->display_config.overrides.best_effort_min_active_latency_hiding_us = 0;
 	}
 
 	for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
@@ -1997,7 +1994,6 @@ static void setup_planes_for_drr_by_mask(struct display_configuation_with_meta *
 			plane->overrides.uclk_pstate_change_strategy = dml2_uclk_pstate_change_strategy_force_drr;
 
 			display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_drr;
-
 		}
 	}
 }
@@ -2063,7 +2059,6 @@ static void setup_planes_for_vblank_by_mask(struct display_configuation_with_met
 					plane->overrides.reserved_vblank_time_ns);
 
 			display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_vblank;
-
 		}
 	}
 }
@@ -2078,6 +2073,7 @@ static void setup_planes_for_vblank_drr_by_mask(struct display_configuation_with
 	for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
 		if (is_bit_set_in_bitfield(plane_mask, plane_index)) {
 			plane = &display_config->display_config.plane_descriptors[plane_index];
+
 			plane->overrides.reserved_vblank_time_ns = (long)(pmo->soc_bb->power_management_parameters.dram_clk_change_blackout_us * 1000);
 
 			display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_vblank_drr;
@@ -2099,7 +2095,7 @@ static void setup_planes_for_vactive_by_mask(struct display_configuation_with_me
 			display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_vactive;
 
 			if (!pmo->options->disable_vactive_det_fill_bw_pad) {
-				display_config->display_config.plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us =
+				display_config->display_config.plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us[dml2_pstate_type_uclk] =
 					(unsigned int)math_floor(pmo->scratch.pmo_dcn4.stream_pstate_meta[stream_index].method_vactive.max_vactive_det_fill_delay_us);
 			}
 		}
@@ -2120,7 +2116,7 @@ static void setup_planes_for_vactive_drr_by_mask(struct display_configuation_wit
 			display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_vactive_drr;
 
 			if (!pmo->options->disable_vactive_det_fill_bw_pad) {
-				display_config->display_config.plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us =
+				display_config->display_config.plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us[dml2_pstate_type_uclk] =
 					(unsigned int)math_floor(pmo->scratch.pmo_dcn4.stream_pstate_meta[stream_index].method_vactive.max_vactive_det_fill_delay_us);
 			}
 		}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/inc/dml2_internal_shared_types.h b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/inc/dml2_internal_shared_types.h
index 9f562f0c4797..1a6c0727cd2a 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/inc/dml2_internal_shared_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/inc/dml2_internal_shared_types.h
@@ -152,7 +152,7 @@ struct core_plane_support_info {
 	int active_latency_hiding_us;
 	int mall_svp_size_requirement_ways;
 	int nominal_vblank_pstate_latency_hiding_us;
-	unsigned int dram_change_vactive_det_fill_delay_us;
+	int vactive_det_fill_delay_us[dml2_pstate_type_count];
 };
 
 struct core_stream_support_info {
@@ -209,6 +209,7 @@ struct dml2_core_mode_support_result {
 
 		unsigned int uclk_pstate_supported;
 		unsigned int fclk_pstate_supported;
+		struct dml2_core_internal_watermarks watermarks;
 	} global;
 
 	struct {
@@ -257,8 +258,8 @@ struct dml2_implicit_svp_meta {
 
 struct dml2_pstate_per_method_common_meta {
 	/* generic params */
-	unsigned int allow_start_otg_vline;
-	unsigned int allow_end_otg_vline;
+	int allow_start_otg_vline;
+	int allow_end_otg_vline;
 	/* scheduling params */
 	double allow_time_us;
 	double disallow_time_us;
@@ -268,39 +269,44 @@ struct dml2_pstate_per_method_common_meta {
 struct dml2_pstate_meta {
 	bool valid;
 	double otg_vline_time_us;
-	unsigned int scheduling_delay_otg_vlines;
-	unsigned int vertical_interrupt_ack_delay_otg_vlines;
-	unsigned int allow_to_target_delay_otg_vlines;
-	unsigned int contention_delay_otg_vlines;
-	unsigned int min_allow_width_otg_vlines;
-	unsigned int nom_vtotal;
-	unsigned int vblank_start;
+	int scheduling_delay_otg_vlines;
+	int vertical_interrupt_ack_delay_otg_vlines;
+	int allow_to_target_delay_otg_vlines;
+	int contention_delay_otg_vlines;
+	int min_allow_width_otg_vlines;
+	int nom_vtotal;
+	int vblank_start;
 	double nom_refresh_rate_hz;
 	double nom_frame_time_us;
-	unsigned int max_vtotal;
+	int max_vtotal;
 	double min_refresh_rate_hz;
 	double max_frame_time_us;
-	unsigned int blackout_otg_vlines;
+	int blackout_otg_vlines;
+	int max_allow_delay_otg_vlines;
+	double nom_vblank_time_us;
 	struct {
 		double max_vactive_det_fill_delay_us;
-		unsigned int max_vactive_det_fill_delay_otg_vlines;
+		double vactive_latency_hiding_us;
+		double reserved_vblank_required_us;
+		int max_vactive_det_fill_delay_otg_vlines;
+		int reserved_blank_required_vlines;
 		struct dml2_pstate_per_method_common_meta common;
 	} method_vactive;
 	struct {
 		struct dml2_pstate_per_method_common_meta common;
 	} method_vblank;
 	struct {
-		unsigned int programming_delay_otg_vlines;
-		unsigned int df_throttle_delay_otg_vlines;
-		unsigned int prefetch_to_mall_delay_otg_vlines;
+		int programming_delay_otg_vlines;
+		int df_throttle_delay_otg_vlines;
+		int prefetch_to_mall_delay_otg_vlines;
 		unsigned long phantom_vactive;
 		unsigned long phantom_vfp;
 		unsigned long phantom_vtotal;
 		struct dml2_pstate_per_method_common_meta common;
 	} method_subvp;
 	struct {
-		unsigned int programming_delay_otg_vlines;
-		unsigned int stretched_vtotal;
+		int programming_delay_otg_vlines;
+		int stretched_vtotal;
 		struct dml2_pstate_per_method_common_meta common;
 	} method_drr;
 };
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dcn35/dcn35_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dcn35/dcn35_dsc.c
index f9c6377ac66c..e712985f7abd 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dcn35/dcn35_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dcn35/dcn35_dsc.c
@@ -28,9 +28,9 @@
 #include "reg_helper.h"
 
 static void dsc35_enable(struct display_stream_compressor *dsc, int opp_pipe);
+static void dsc35_get_single_enc_caps(struct dsc_enc_caps *dsc_enc_caps, unsigned int max_dscclk_khz);
 
 static const struct dsc_funcs dcn35_dsc_funcs = {
-	.dsc_get_enc_caps = dsc2_get_enc_caps,
 	.dsc_read_state = dsc2_read_state,
 	.dsc_read_reg_state = dsc2_read_reg_state,
 	.dsc_validate_stream = dsc2_validate_stream,
@@ -40,6 +40,7 @@ static const struct dsc_funcs dcn35_dsc_funcs = {
 	.dsc_disable = dsc2_disable,
 	.dsc_disconnect = dsc2_disconnect,
 	.dsc_wait_disconnect_pending_clear = dsc2_wait_disconnect_pending_clear,
+	.dsc_get_single_enc_caps = dsc35_get_single_enc_caps,
 };
 
 /* Macro definitios for REG_SET macros*/
@@ -111,3 +112,31 @@ void dsc35_set_fgcg(struct dcn20_dsc *dsc20, bool enable)
 {
 	REG_UPDATE(DSC_TOP_CONTROL, DSC_FGCG_REP_DIS, !enable);
 }
+
+void dsc35_get_single_enc_caps(struct dsc_enc_caps *dsc_enc_caps, unsigned int max_dscclk_khz)
+{
+	dsc_enc_caps->dsc_version = 0x21; /* v1.2 - DP spec defined it in reverse order and we kept it */
+
+	dsc_enc_caps->slice_caps.bits.NUM_SLICES_1 = 1;
+	dsc_enc_caps->slice_caps.bits.NUM_SLICES_2 = 1;
+	dsc_enc_caps->slice_caps.bits.NUM_SLICES_3 = 1;
+	dsc_enc_caps->slice_caps.bits.NUM_SLICES_4 = 1;
+
+	dsc_enc_caps->lb_bit_depth = 13;
+	dsc_enc_caps->is_block_pred_supported = true;
+
+	dsc_enc_caps->color_formats.bits.RGB = 1;
+	dsc_enc_caps->color_formats.bits.YCBCR_444 = 1;
+	dsc_enc_caps->color_formats.bits.YCBCR_SIMPLE_422 = 1;
+	dsc_enc_caps->color_formats.bits.YCBCR_NATIVE_422 = 0;
+	dsc_enc_caps->color_formats.bits.YCBCR_NATIVE_420 = 1;
+
+	dsc_enc_caps->color_depth.bits.COLOR_DEPTH_8_BPC = 1;
+	dsc_enc_caps->color_depth.bits.COLOR_DEPTH_10_BPC = 1;
+	dsc_enc_caps->color_depth.bits.COLOR_DEPTH_12_BPC = 1;
+
+	dsc_enc_caps->max_total_throughput_mps = max_dscclk_khz * 3 / 1000;
+
+	dsc_enc_caps->max_slice_width = 5184; /* (including 64 overlap pixels for eDP MSO mode) */
+	dsc_enc_caps->bpp_increment_div = 16; /* 1/16th of a bit */
+}
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c
index d1aaa58b7db3..5a03758e3de6 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c
@@ -933,8 +933,8 @@ int hubbub31_init_dchub_sys_ctx(struct hubbub *hubbub,
 
 		dcn20_vmid_setup(&hubbub2->vmid[15], &phys_config);
 	}
-
-	dcn21_dchvm_init(hubbub);
+	if (hubbub->funcs->dchvm_init)
+		hubbub->funcs->dchvm_init(hubbub);
 
 	return NUM_VMID;
 }
@@ -1071,7 +1071,8 @@ static const struct hubbub_funcs hubbub31_funcs = {
 	.program_compbuf_size = dcn31_program_compbuf_size,
 	.init_crb = dcn31_init_crb,
 	.hubbub_read_state = hubbub2_read_state,
-	.hubbub_read_reg_state = hubbub3_read_reg_state
+	.hubbub_read_reg_state = hubbub3_read_reg_state,
+	.dchvm_init  = dcn21_dchvm_init
 };
 
 void hubbub31_construct(struct dcn20_hubbub *hubbub31,
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c
index 1b7746a6549a..43ba399f4822 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c
@@ -549,6 +549,55 @@ void hubbub35_init(struct hubbub *hubbub)
 	memset(&hubbub2->watermarks.a.cstate_pstate, 0, sizeof(hubbub2->watermarks.a.cstate_pstate));
 }
 
+void dcn35_dchvm_init(struct hubbub *hubbub)
+{
+	struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub);
+	uint32_t riommu_active;
+	int i;
+
+	//Init DCHVM block
+	REG_UPDATE(DCHVM_CTRL0, HOSTVM_INIT_REQ, 1);
+
+	//Poll until RIOMMU_ACTIVE = 1
+	for (i = 0; i < 100; i++) {
+		REG_GET(DCHVM_RIOMMU_STAT0, RIOMMU_ACTIVE, &riommu_active);
+
+		if (riommu_active)
+			break;
+		else
+			udelay(5);
+	}
+
+	if (riommu_active) {
+		// Disable gating and memory power requests
+		REG_UPDATE(DCHVM_MEM_CTRL, HVM_GPUVMRET_PWR_REQ_DIS, 1);
+		REG_UPDATE_4(DCHVM_CLK_CTRL,
+						HVM_DISPCLK_R_GATE_DIS, 1,
+						HVM_DISPCLK_G_GATE_DIS, 1,
+						HVM_DCFCLK_R_GATE_DIS, 1,
+						HVM_DCFCLK_G_GATE_DIS, 1);
+
+		//Reflect the power status of DCHUBBUB
+		REG_UPDATE(DCHVM_RIOMMU_CTRL0, HOSTVM_POWERSTATUS, 1);
+
+		//Start rIOMMU prefetching
+		REG_UPDATE(DCHVM_RIOMMU_CTRL0, HOSTVM_PREFETCH_REQ, 1);
+
+		//Poll until HOSTVM_PREFETCH_DONE = 1
+		REG_WAIT(DCHVM_RIOMMU_STAT0, HOSTVM_PREFETCH_DONE, 1, 5, 100);
+
+		//Enable memory power requests
+		REG_UPDATE(DCHVM_MEM_CTRL, HVM_GPUVMRET_PWR_REQ_DIS, 0);
+		// Enable dynamic clock gating
+		REG_UPDATE_4(DCHVM_CLK_CTRL,
+						HVM_DISPCLK_R_GATE_DIS, 0,
+						HVM_DISPCLK_G_GATE_DIS, 0,
+						HVM_DCFCLK_R_GATE_DIS, 0,
+						HVM_DCFCLK_G_GATE_DIS, 0);
+		hubbub->riommu_active = true;
+	}
+}
+
 /*static void hubbub35_set_request_limit(struct hubbub *hubbub,
 				       int memory_channel_count,
 				       int words_per_channel)
@@ -589,7 +638,8 @@ static const struct hubbub_funcs hubbub35_funcs = {
 	.hubbub_read_state = hubbub2_read_state,
 	.force_usr_retraining_allow = hubbub32_force_usr_retraining_allow,
 	.dchubbub_init = hubbub35_init,
-	.hubbub_read_reg_state = hubbub3_read_reg_state
+	.hubbub_read_reg_state = hubbub3_read_reg_state,
+	.dchvm_init = dcn35_dchvm_init
 };
 
 void hubbub35_construct(struct dcn20_hubbub *hubbub2,
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h
index 23fecf88556c..9f65fff1bd4d 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h
@@ -168,4 +168,5 @@ void dcn35_program_compbuf_size(struct hubbub *hubbub,
 	unsigned int compbuf_size_kb, bool safe_to_increase);
 void dcn35_init_crb(struct hubbub *hubbub);
 void hubbub35_init(struct hubbub *hubbub);
+void dcn35_dchvm_init(struct hubbub *hubbub);
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c
index 3005115c8505..4986f12dc9df 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c
@@ -685,7 +685,6 @@ void dce110_enable_stream(struct pipe_ctx *pipe_ctx)
 	uint32_t early_control = 0;
 	struct timing_generator *tg = pipe_ctx->stream_res.tg;
 
-	link_hwss->setup_stream_attribute(pipe_ctx);
 	link_hwss->setup_stream_encoder(pipe_ctx);
 
 	dc->hwss.update_info_frame(pipe_ctx);
@@ -1103,6 +1102,9 @@ void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx)
 	if (!pipe_ctx->stream)
 		return;
 
+	if (dc_is_rgb_signal(pipe_ctx->stream->signal))
+		return;
+
 	dc = pipe_ctx->stream->ctx->dc;
 	clk_mgr = dc->clk_mgr;
 	link_hwss = get_link_hwss(pipe_ctx->stream->link, &pipe_ctx->link_res);
@@ -1139,6 +1141,9 @@ void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx)
 	if (!pipe_ctx || !pipe_ctx->stream)
 		return;
 
+	if (dc_is_rgb_signal(pipe_ctx->stream->signal))
+		return;
+
 	dc = pipe_ctx->stream->ctx->dc;
 	clk_mgr = dc->clk_mgr;
 	link_hwss = get_link_hwss(pipe_ctx->stream->link, &pipe_ctx->link_res);
@@ -1193,8 +1198,7 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx)
 		pipe_ctx->stream_res.stream_enc->funcs->stop_dp_info_packets(
 			pipe_ctx->stream_res.stream_enc);
 
-	if (!dc_is_rgb_signal(pipe_ctx->stream->signal))
-		dc->hwss.disable_audio_stream(pipe_ctx);
+	dc->hwss.disable_audio_stream(pipe_ctx);
 
 	link_hwss->reset_stream_encoder(pipe_ctx);
 
@@ -1480,9 +1484,6 @@ void build_audio_output(
 						state->clk_mgr);
 	}
 
-	audio_output->pll_info.feed_back_divider =
-			pipe_ctx->pll_settings.feedback_divider;
-
 	audio_output->pll_info.dto_source =
 		translate_to_dto_source(
 			pipe_ctx->stream_res.tg->inst + 1);
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c
index 6bd905905984..c8ff8ae85a03 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c
@@ -614,6 +614,14 @@ void dcn20_dpp_pg_control(
 		 *		DOMAIN11_PGFSM_PWR_STATUS, pwr_status,
 		 * 		1, 1000);
 		 */
+
+		/* Force disable cursor on plane powerdown on DPP 5 using dpp_force_disable_cursor */
+		if (!power_on) {
+			struct dpp *dpp5 = hws->ctx->dc->res_pool->dpps[dpp_inst];
+			if (dpp5 && dpp5->funcs->dpp_force_disable_cursor)
+				dpp5->funcs->dpp_force_disable_cursor(dpp5);
+		}
+
 		break;
 	default:
 		BREAK_TO_DEBUGGER();
@@ -3055,8 +3063,6 @@ void dcn20_enable_stream(struct pipe_ctx *pipe_ctx)
 						      link_enc->transmitter - TRANSMITTER_UNIPHY_A);
 	}
 
-	link_hwss->setup_stream_attribute(pipe_ctx);
-
 	if (dc->res_pool->dccg->funcs->set_pixel_rate_div)
 		dc->res_pool->dccg->funcs->set_pixel_rate_div(
 			dc->res_pool->dccg,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c
index f02edc9371b0..2fbc22afb89c 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c
@@ -974,8 +974,6 @@ void dcn401_enable_stream(struct pipe_ctx *pipe_ctx)
 		}
 	}
 
-	link_hwss->setup_stream_attribute(pipe_ctx);
-
 	if (dc->res_pool->dccg->funcs->set_pixel_rate_div) {
 		dc->res_pool->dccg->funcs->set_pixel_rate_div(
 			dc->res_pool->dccg,
@@ -3675,6 +3673,8 @@ void dcn401_update_dchubp_dpp_sequence(struct dc *dc,
 	     pipe_ctx->update_flags.bits.scaler || viewport_changed == true) &&
 	    pipe_ctx->stream->cursor_attributes.address.quad_part != 0) {
 
+		hwss_add_abort_cursor_offload_update(seq_state, dc, pipe_ctx);
+
 		hwss_add_set_cursor_attribute(seq_state, dc, pipe_ctx);
 
 		/* Step 15: Cursor position setup */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h
index 3772b4aa11cc..8ed9eea40c56 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h
@@ -696,6 +696,11 @@ struct hubp_program_mcache_id_and_split_coordinate_params {
 	struct mcache_regs_struct *mcache_regs;
 };
 
+struct abort_cursor_offload_update_params {
+	struct dc *dc;
+	struct pipe_ctx *pipe_ctx;
+};
+
 struct set_cursor_attribute_params {
 	struct dc *dc;
 	struct pipe_ctx *pipe_ctx;
@@ -842,6 +847,7 @@ union block_sequence_params {
 	struct mpc_insert_plane_params mpc_insert_plane_params;
 	struct dpp_set_scaler_params dpp_set_scaler_params;
 	struct hubp_mem_program_viewport_params hubp_mem_program_viewport_params;
+	struct abort_cursor_offload_update_params abort_cursor_offload_update_params;
 	struct set_cursor_attribute_params set_cursor_attribute_params;
 	struct set_cursor_position_params set_cursor_position_params;
 	struct set_cursor_sdr_white_level_params set_cursor_sdr_white_level_params;
@@ -960,6 +966,7 @@ enum block_sequence_func {
 	MPC_INSERT_PLANE,
 	DPP_SET_SCALER,
 	HUBP_MEM_PROGRAM_VIEWPORT,
+	ABORT_CURSOR_OFFLOAD_UPDATE,
 	SET_CURSOR_ATTRIBUTE,
 	SET_CURSOR_POSITION,
 	SET_CURSOR_SDR_WHITE_LEVEL,
@@ -1565,6 +1572,8 @@ void hwss_dpp_set_scaler(union block_sequence_params *params);
 
 void hwss_hubp_mem_program_viewport(union block_sequence_params *params);
 
+void hwss_abort_cursor_offload_update(union block_sequence_params *params);
+
 void hwss_set_cursor_attribute(union block_sequence_params *params);
 
 void hwss_set_cursor_position(union block_sequence_params *params);
@@ -1961,6 +1970,10 @@ void hwss_add_hubp_mem_program_viewport(struct block_sequence_state *seq_state,
 		const struct rect *viewport,
 		const struct rect *viewport_c);
 
+void hwss_add_abort_cursor_offload_update(struct block_sequence_state *seq_state,
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx);
+
 void hwss_add_set_cursor_attribute(struct block_sequence_state *seq_state,
 		struct dc *dc,
 		struct pipe_ctx *pipe_ctx);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
index dafc8490efb5..1ddfa30411c8 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
@@ -252,6 +252,8 @@ struct hubbub_funcs {
 	void (*program_compbuf_segments)(struct hubbub *hubbub, unsigned compbuf_size_seg, bool safe_to_increase);
 	void (*wait_for_det_update)(struct hubbub *hubbub, int hubp_inst);
 	bool (*program_arbiter)(struct hubbub *hubbub, struct dml2_display_arb_regs *arb_regs, bool safe_to_lower);
+	void (*dchvm_init)(struct hubbub *hubbub);
+
 	struct hubbub_perfmon_funcs {
 		void (*reset)(struct hubbub *hubbub);
 		void (*start_measuring_max_memory_latency_ns)(
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
index 5e2813e9ae2f..a61d12ec61bc 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -75,11 +75,36 @@
  */
 #define MAX_LINK_ENCODERS (MAX_DIG_LINK_ENCODERS + MAX_DAC_LINK_ENCODERS)
 
-#define MAX_DIG_LINK_ENCODERS 7
 #define MAX_DWB_PIPES	1
 #define MAX_HPO_DP2_ENCODERS	4
 #define MAX_HPO_DP2_LINK_ENCODERS	4
 
+/* Pipe topology snapshot structures */
+#define MAX_TOPOLOGY_SNAPSHOTS 4
+
+struct pipe_topology_line {
+	bool is_phantom_pipe;
+	int plane_idx;
+	int slice_idx;
+	int stream_idx;
+	int dpp_inst;
+	int opp_inst;
+	int tg_inst;
+};
+
+struct pipe_topology_snapshot {
+	struct pipe_topology_line pipe_log_lines[MAX_PIPES];
+	int line_count;
+	uint64_t timestamp_us;
+	int stream_count;
+	int phantom_stream_count;
+};
+
+struct pipe_topology_history {
+	struct pipe_topology_snapshot snapshots[MAX_TOPOLOGY_SNAPSHOTS];
+	int current_snapshot_index;
+};
+
 struct gamma_curve {
 	uint32_t offset;
 	uint32_t segments_num;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/link_service.h b/drivers/gpu/drm/amd/display/dc/inc/link_service.h
index 1e34e84160aa..6f94e48a24d1 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/link_service.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/link_service.h
@@ -292,12 +292,12 @@ struct link_service {
 			enum replay_FW_Message_type msg,
 			union dmub_replay_cmd_set *cmd_data);
 	bool (*edp_set_coasting_vtotal)(
-			struct dc_link *link, uint32_t coasting_vtotal);
+			struct dc_link *link, uint32_t coasting_vtotal, uint16_t frame_skip_number);
 	bool (*edp_replay_residency)(const struct dc_link *link,
 			unsigned int *residency, const bool is_start,
 			const enum pr_residency_mode mode);
 	bool (*edp_set_replay_power_opt_and_coasting_vtotal)(struct dc_link *link,
-			const unsigned int *power_opts, uint32_t coasting_vtotal);
+			const unsigned int *power_opts, uint32_t coasting_vtotal, uint16_t frame_skip_number);
 
 	bool (*edp_wait_for_t12)(struct dc_link *link);
 	bool (*edp_is_ilr_optimization_required)(struct dc_link *link,
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_detection.c b/drivers/gpu/drm/amd/display/dc/link/link_detection.c
index c417780f37bc..6d31f4967f1a 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_detection.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_detection.c
@@ -868,6 +868,11 @@ static void verify_link_capability(struct dc_link *link, struct dc_sink *sink,
  * Evaluates an 8-byte EDID header to check if it's good enough
  * for the purpose of determining whether a display is connected
  * without reading the full EDID.
+ *
+ * @edid_header: The first 8 bytes of the EDID read from DDC.
+ *
+ * Return: true if the header looks valid (>= 6 of 8 bytes match the
+ *         expected 00/FF pattern), false otherwise.
  */
 static bool link_detect_evaluate_edid_header(uint8_t edid_header[8])
 {
@@ -886,6 +891,11 @@ static bool link_detect_evaluate_edid_header(uint8_t edid_header[8])
  * Detect whether a display is connected to DDC without reading full EDID.
  * Reads only the EDID header (the first 8 bytes of EDID) from DDC and
  * evaluates whether that matches.
+ *
+ * @link: DC link whose DDC/I2C is probed for the EDID header.
+ *
+ * Return: true if the EDID header was read and passes validation,
+ *         false otherwise.
  */
 static bool link_detect_ddc_probe(struct dc_link *link)
 {
@@ -910,6 +920,11 @@ static bool link_detect_ddc_probe(struct dc_link *link)
  * Load detection can be used to detect the presence of an
  * analog display when we can't read DDC. This causes a visible
  * visual glitch so it should be used sparingly.
+ *
+ * @link: DC link to test using the DAC load-detect path.
+ *
+ * Return: true if the VBIOS load-detect call reports OK, false
+ *         otherwise.
  */
 static bool link_detect_dac_load_detect(struct dc_link *link)
 {
@@ -1209,8 +1224,6 @@ static bool detect_link_and_local_sink(struct dc_link *link,
 			break;
 		}
 
-		sink->edid_caps.analog &= dc_connector_supports_analog(link->link_id.id);
-
 		// Check if edid is the same
 		if ((prev_sink) &&
 		    (edid_status == EDID_THE_SAME || edid_status == EDID_OK))
@@ -1257,6 +1270,7 @@ static bool detect_link_and_local_sink(struct dc_link *link,
 		    !sink->edid_caps.edid_hdmi)
 			sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK;
 		else if (dc_is_dvi_signal(sink->sink_signal) &&
+			 dc_is_dvi_signal(link->connector_signal) &&
 			 aud_support->hdmi_audio_native &&
 			 sink->edid_caps.edid_hdmi)
 			sink->sink_signal = SIGNAL_TYPE_HDMI_TYPE_A;
@@ -1350,6 +1364,14 @@ static bool detect_link_and_local_sink(struct dc_link *link,
 
 /**
  * link_detect_analog() - Determines if an analog sink is connected.
+ *
+ * @link: DC link to evaluate (must support analog signalling).
+ * @type: Updated with the detected connection type:
+ *        dc_connection_single (analog via DDC),
+ *        dc_connection_dac_load (via load-detect),
+ *        or dc_connection_none.
+ *
+ * Return: true if detection completed.
  */
 static bool link_detect_analog(struct dc_link *link, enum dc_connection_type *type)
 {
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
index 4ddcdc222913..6ae134147617 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
@@ -2226,7 +2226,11 @@ static enum dc_status enable_link(
 {
 	enum dc_status status = DC_ERROR_UNEXPECTED;
 	struct dc_stream_state *stream = pipe_ctx->stream;
-	struct dc_link *link = stream->link;
+	struct dc_link *link = NULL;
+
+	if (stream == NULL)
+		return DC_ERROR_UNEXPECTED;
+	link = stream->link;
 
 	/* There's some scenarios where driver is unloaded with display
 	 * still enabled. When driver is reloaded, it may cause a display
@@ -2374,8 +2378,7 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx)
 			set_avmute(pipe_ctx, true);
 	}
 
-	if (!dc_is_rgb_signal(pipe_ctx->stream->signal))
-		dc->hwss.disable_audio_stream(pipe_ctx);
+	dc->hwss.disable_audio_stream(pipe_ctx);
 
 	update_psp_stream_config(pipe_ctx, true);
 	dc->hwss.blank_stream(pipe_ctx);
@@ -2464,6 +2467,7 @@ void link_set_dpms_on(
 	struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
 	enum otg_out_mux_dest otg_out_dest = OUT_MUX_DIO;
 	struct vpg *vpg = pipe_ctx->stream_res.stream_enc->vpg;
+	const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res);
 	bool apply_edp_fast_boot_optimization =
 		pipe_ctx->stream->apply_edp_fast_boot_optimization;
 
@@ -2508,6 +2512,8 @@ void link_set_dpms_on(
 		pipe_ctx->stream_res.tg->funcs->set_out_mux(pipe_ctx->stream_res.tg, otg_out_dest);
 	}
 
+	link_hwss->setup_stream_attribute(pipe_ctx);
+
 	pipe_ctx->stream->apply_edp_fast_boot_optimization = false;
 
 	// Enable VPG before building infoframe
@@ -2660,8 +2666,7 @@ void link_set_dpms_on(
 		enable_stream_features(pipe_ctx);
 	update_psp_stream_config(pipe_ctx, false);
 
-	if (!dc_is_rgb_signal(pipe_ctx->stream->signal))
-		dc->hwss.enable_audio_stream(pipe_ctx);
+	dc->hwss.enable_audio_stream(pipe_ctx);
 
 	if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) {
 		set_avmute(pipe_ctx, false);
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_factory.c b/drivers/gpu/drm/amd/display/dc/link/link_factory.c
index 7989baf3843c..a6e2b0821969 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_factory.c
@@ -529,16 +529,16 @@ static bool construct_phy(struct dc_link *link,
 	enc_init_data.transmitter = translate_encoder_to_transmitter(enc_init_data.encoder);
 	enc_init_data.analog_engine = find_analog_engine(link);
 
-	if (!transmitter_supported(enc_init_data.transmitter) &&
-		!analog_engine_supported(enc_init_data.analog_engine)) {
-		DC_LOG_WARNING("link_id %d has unsupported encoder\n", link->link_id.id);
-		return false;
-	}
-
 	link->ep_type = DISPLAY_ENDPOINT_PHY;
 
 	DC_LOG_DC("BIOS object table - link_id: %d", link->link_id.id);
 
+	if (!transmitter_supported(enc_init_data.transmitter) &&
+	    !analog_engine_supported(enc_init_data.analog_engine)) {
+		DC_LOG_WARNING("link_id %d has unsupported encoder\n", link->link_id.id);
+		goto unsupported_fail;
+	}
+
 	if (bios->funcs->get_disp_connector_caps_info) {
 		bios->funcs->get_disp_connector_caps_info(bios, link->link_id, &disp_connect_caps_info);
 		link->is_internal_display = disp_connect_caps_info.INTERNAL_DISPLAY;
@@ -787,6 +787,7 @@ static bool construct_phy(struct dc_link *link,
 
 	link->psr_settings.psr_vtotal_control_support = false;
 	link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED;
+	link->replay_settings.config.replay_version = DC_REPLAY_VERSION_UNSUPPORTED;
 
 	DC_LOG_DC("BIOS object table - %s finished successfully.\n", __func__);
 	return true;
@@ -805,6 +806,7 @@ create_fail:
 		link->hpd_gpio = NULL;
 	}
 
+unsupported_fail:
 	DC_LOG_DC("BIOS object table - %s failed.\n", __func__);
 	return false;
 }
@@ -868,6 +870,7 @@ static bool construct_dpia(struct dc_link *link,
 	/* TODO: Create link encoder */
 
 	link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED;
+	link->replay_settings.config.replay_version = DC_REPLAY_VERSION_UNSUPPORTED;
 
 	return true;
 
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
index 750147c52c8a..ad90a0106938 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
@@ -1713,7 +1713,7 @@ static bool retrieve_link_cap(struct dc_link *link)
 	union edp_configuration_cap edp_config_cap;
 	union dp_downstream_port_present ds_port = { 0 };
 	enum dc_status status = DC_ERROR_UNEXPECTED;
-	uint32_t read_dpcd_retry_cnt = 3;
+	uint32_t read_dpcd_retry_cnt = 20;
 	int i;
 	struct dp_sink_hw_fw_revision dp_hw_fw_revision;
 	const uint32_t post_oui_delay = 30; // 30ms
@@ -1756,12 +1756,13 @@ static bool retrieve_link_cap(struct dc_link *link)
 	}
 
 	dpcd_set_source_specific_data(link);
-	/* Sink may need to configure internals based on vendor, so allow some
-	 * time before proceeding with possibly vendor specific transactions
-	 */
-	msleep(post_oui_delay);
 
 	for (i = 0; i < read_dpcd_retry_cnt; i++) {
+		/*
+		 * Sink may need to configure internals based on vendor, so allow some
+		 * time before proceeding with possibly vendor specific transactions
+		 */
+		msleep(post_oui_delay);
 		status = core_link_read_dpcd(
 				link,
 				DP_DPCD_REV,
@@ -2091,6 +2092,11 @@ static bool retrieve_link_cap(struct dc_link *link)
 			link->dpcd_caps.max_uncompressed_pixel_rate_cap.raw,
 			sizeof(link->dpcd_caps.max_uncompressed_pixel_rate_cap.raw));
 
+	core_link_read_dpcd(link,
+			DP_PANEL_REPLAY_CAPABILITY_SUPPORT,
+			&link->dpcd_caps.pr_caps_supported.raw,
+			sizeof(link->dpcd_caps.pr_caps_supported.raw));
+
 	/* Read DP tunneling information. */
 	status = dpcd_get_tunneling_device_data(link);
 	if (status != DC_OK)
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
index 5e806edbb9f6..c56e69eb27ef 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
@@ -949,7 +949,7 @@ bool edp_set_replay_allow_active(struct dc_link *link, const bool *allow_active,
 	/* Set power optimization flag */
 	if (power_opts && link->replay_settings.replay_power_opt_active != *power_opts) {
 		if (replay != NULL && link->replay_settings.replay_feature_enabled &&
-		    replay->funcs->replay_set_power_opt) {
+			replay->funcs->replay_set_power_opt) {
 			replay->funcs->replay_set_power_opt(replay, *power_opts, panel_inst);
 			link->replay_settings.replay_power_opt_active = *power_opts;
 		}
@@ -984,7 +984,117 @@ bool edp_get_replay_state(const struct dc_link *link, uint64_t *state)
 	return true;
 }
 
-bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *stream)
+static bool edp_setup_panel_replay(struct dc_link *link, const struct dc_stream_state *stream)
+{
+	/* To-do: Setup Replay */
+	struct dc *dc;
+	struct dmub_replay *replay;
+	int i;
+	unsigned int panel_inst;
+	struct replay_context replay_context = { 0 };
+	unsigned int lineTimeInNs = 0;
+
+	union panel_replay_enable_and_configuration_1 pr_config_1 = { 0 };
+	union panel_replay_enable_and_configuration_2 pr_config_2 = { 0 };
+
+	union dpcd_alpm_configuration alpm_config;
+
+	replay_context.controllerId = CONTROLLER_ID_UNDEFINED;
+
+	if (!link)
+		return false;
+
+	//Clear Panel Replay enable & config
+	dm_helpers_dp_write_dpcd(link->ctx, link,
+		DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_1,
+		(uint8_t *)&(pr_config_1.raw), sizeof(uint8_t));
+
+	dm_helpers_dp_write_dpcd(link->ctx, link,
+		DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_2,
+		(uint8_t *)&(pr_config_2.raw), sizeof(uint8_t));
+
+	if (!(link->replay_settings.config.replay_supported))
+		return false;
+
+	dc = link->ctx->dc;
+
+	//not sure should keep or not
+	replay = dc->res_pool->replay;
+
+	if (!replay)
+		return false;
+
+	if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst))
+		return false;
+
+	replay_context.aux_inst = link->ddc->ddc_pin->hw_info.ddc_channel;
+	replay_context.digbe_inst = link->link_enc->transmitter;
+	replay_context.digfe_inst = link->link_enc->preferred_engine;
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		if (dc->current_state->res_ctx.pipe_ctx[i].stream
+				== stream) {
+			/* dmcu -1 for all controller id values,
+			 * therefore +1 here
+			 */
+			replay_context.controllerId =
+				dc->current_state->res_ctx.pipe_ctx[i].stream_res.tg->inst + 1;
+			break;
+		}
+	}
+
+	lineTimeInNs =
+		((stream->timing.h_total * 1000000) /
+			(stream->timing.pix_clk_100hz / 10)) + 1;
+
+	replay_context.line_time_in_ns = lineTimeInNs;
+
+	link->replay_settings.replay_feature_enabled =
+			replay->funcs->replay_copy_settings(replay, link, &replay_context, panel_inst);
+
+	if (link->replay_settings.replay_feature_enabled) {
+		pr_config_1.bits.PANEL_REPLAY_ENABLE = 1;
+		pr_config_1.bits.PANEL_REPLAY_CRC_ENABLE = 1;
+		pr_config_1.bits.IRQ_HPD_ASSDP_MISSING = 1;
+		pr_config_1.bits.IRQ_HPD_VSCSDP_UNCORRECTABLE_ERROR = 1;
+		pr_config_1.bits.IRQ_HPD_RFB_ERROR = 1;
+		pr_config_1.bits.IRQ_HPD_ACTIVE_FRAME_CRC_ERROR = 1;
+		pr_config_1.bits.PANEL_REPLAY_SELECTIVE_UPDATE_ENABLE = 1;
+		pr_config_1.bits.PANEL_REPLAY_EARLY_TRANSPORT_ENABLE = 1;
+
+		pr_config_2.bits.SINK_REFRESH_RATE_UNLOCK_GRANTED = 0;
+		pr_config_2.bits.SU_Y_GRANULARITY_EXT_VALUE_ENABLED = 0;
+		pr_config_2.bits.SU_REGION_SCAN_LINE_CAPTURE_INDICATION = 0;
+
+		dm_helpers_dp_write_dpcd(link->ctx, link,
+			DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_1,
+			(uint8_t *)&(pr_config_1.raw), sizeof(uint8_t));
+
+		dm_helpers_dp_write_dpcd(link->ctx, link,
+			DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_2,
+			(uint8_t *)&(pr_config_2.raw), sizeof(uint8_t));
+
+		//ALPM Setup
+		memset(&alpm_config, 0, sizeof(alpm_config));
+		alpm_config.bits.ENABLE = link->replay_settings.config.alpm_mode != DC_ALPM_UNSUPPORTED ? 1 : 0;
+
+		if (link->replay_settings.config.alpm_mode == DC_ALPM_AUXLESS) {
+			alpm_config.bits.ALPM_MODE_SEL = 1;
+			alpm_config.bits.ACDS_PERIOD_DURATION = 1;
+		}
+
+		dm_helpers_dp_write_dpcd(
+			link->ctx,
+			link,
+			DP_RECEIVER_ALPM_CONFIG,
+			&alpm_config.raw,
+			sizeof(alpm_config.raw));
+	}
+
+	return true;
+}
+
+static bool edp_setup_freesync_replay(struct dc_link *link, const struct dc_stream_state *stream)
 {
 	/* To-do: Setup Replay */
 	struct dc *dc;
@@ -1080,6 +1190,18 @@ bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *stream
 	return true;
 }
 
+bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *stream)
+{
+	if (!link)
+		return false;
+	if (link->replay_settings.config.replay_version == DC_VESA_PANEL_REPLAY)
+		return edp_setup_panel_replay(link, stream);
+	else if (link->replay_settings.config.replay_version == DC_FREESYNC_REPLAY)
+		return edp_setup_freesync_replay(link, stream);
+	else
+		return false;
+}
+
 /*
  * This is general Interface for Replay to set an 32 bit variable to dmub
  * replay_FW_Message_type: Indicates which instruction or variable pass to DMUB
@@ -1110,7 +1232,7 @@ bool edp_send_replay_cmd(struct dc_link *link,
 	return true;
 }
 
-bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal)
+bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal, uint16_t frame_skip_number)
 {
 	struct dc *dc = link->ctx->dc;
 	struct dmub_replay *replay = dc->res_pool->replay;
@@ -1122,9 +1244,11 @@ bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal)
 	if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst))
 		return false;
 
-	if (coasting_vtotal && link->replay_settings.coasting_vtotal != coasting_vtotal) {
-		replay->funcs->replay_set_coasting_vtotal(replay, coasting_vtotal, panel_inst);
+	if (coasting_vtotal && (link->replay_settings.coasting_vtotal != coasting_vtotal ||
+		link->replay_settings.frame_skip_number != frame_skip_number)) {
+		replay->funcs->replay_set_coasting_vtotal(replay, coasting_vtotal, panel_inst, frame_skip_number);
 		link->replay_settings.coasting_vtotal = coasting_vtotal;
+		link->replay_settings.frame_skip_number = frame_skip_number;
 	}
 
 	return true;
@@ -1152,7 +1276,7 @@ bool edp_replay_residency(const struct dc_link *link,
 }
 
 bool edp_set_replay_power_opt_and_coasting_vtotal(struct dc_link *link,
-	const unsigned int *power_opts, uint32_t coasting_vtotal)
+	const unsigned int *power_opts, uint32_t coasting_vtotal, uint16_t frame_skip_number)
 {
 	struct dc  *dc = link->ctx->dc;
 	struct dmub_replay *replay = dc->res_pool->replay;
@@ -1163,13 +1287,16 @@ bool edp_set_replay_power_opt_and_coasting_vtotal(struct dc_link *link,
 
 	/* Only both power and coasting vtotal changed, this func could return true */
 	if (power_opts && link->replay_settings.replay_power_opt_active != *power_opts &&
-		coasting_vtotal && link->replay_settings.coasting_vtotal != coasting_vtotal) {
+		(coasting_vtotal &&
+		(link->replay_settings.coasting_vtotal != coasting_vtotal ||
+		link->replay_settings.frame_skip_number != frame_skip_number))) {
 		if (link->replay_settings.replay_feature_enabled &&
 			replay->funcs->replay_set_power_opt_and_coasting_vtotal) {
 			replay->funcs->replay_set_power_opt_and_coasting_vtotal(replay,
-				*power_opts, panel_inst, coasting_vtotal);
+				*power_opts, panel_inst, coasting_vtotal, frame_skip_number);
 			link->replay_settings.replay_power_opt_active = *power_opts;
 			link->replay_settings.coasting_vtotal = coasting_vtotal;
+			link->replay_settings.frame_skip_number = frame_skip_number;
 		} else
 			return false;
 	} else
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h
index 62a6344e613e..dd79c7cd2828 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h
@@ -59,12 +59,12 @@ bool edp_setup_replay(struct dc_link *link,
 bool edp_send_replay_cmd(struct dc_link *link,
 			enum replay_FW_Message_type msg,
 			union dmub_replay_cmd_set *cmd_data);
-bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal);
+bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal, uint16_t frame_skip_number);
 bool edp_replay_residency(const struct dc_link *link,
 	unsigned int *residency, const bool is_start, const enum pr_residency_mode mode);
 bool edp_get_replay_state(const struct dc_link *link, uint64_t *state);
 bool edp_set_replay_power_opt_and_coasting_vtotal(struct dc_link *link,
-	const unsigned int *power_opts, uint32_t coasting_vtotal);
+	const unsigned int *power_opts, uint32_t coasting_vtotal, uint16_t frame_skip_number);
 bool edp_wait_for_t12(struct dc_link *link);
 bool edp_is_ilr_optimization_required(struct dc_link *link,
        struct dc_crtc_timing *crtc_timing);
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c
index 81e64e17d0cb..b276fec3e479 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c
@@ -1847,7 +1847,7 @@ enum dc_status dcn32_validate_bandwidth(struct dc *dc,
 				dc_state_set_stream_cursor_subvp_limit(stream, context, true);
 				status = DC_FAIL_HW_CURSOR_SUPPORT;
 			}
-		};
+		}
 	}
 
 	if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING && status == DC_FAIL_HW_CURSOR_SUPPORT) {
@@ -2200,7 +2200,8 @@ static bool dcn32_resource_construct(
 	dc->caps.i2c_speed_in_khz_hdcp = 100; /*1.4 w/a applied by default*/
 	/* TODO: Bring max_cursor_size back to 256 after subvp cursor corruption is fixed*/
 	dc->caps.max_cursor_size = 64;
-	dc->caps.max_buffered_cursor_size = 64; // sqrt(16 * 1024 / 4)
+	/* floor(sqrt(buf_size_bytes / bpp ) * bpp, fixed_req_size) / bpp = max_width */
+	dc->caps.max_buffered_cursor_size = 64; // floor(sqrt(16 * 1024 / 4) * 4, 256) / 4 = 64
 	dc->caps.min_horizontal_blanking_period = 80;
 	dc->caps.dmdata_alloc_size = 2048;
 	dc->caps.mall_size_per_mem_channel = 4;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c
index 130058d7a70c..875ae97489d3 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c
@@ -1671,7 +1671,7 @@ enum dc_status dcn401_validate_bandwidth(struct dc *dc,
 				dc_state_set_stream_cursor_subvp_limit(stream, context, true);
 				status = DC_FAIL_HW_CURSOR_SUPPORT;
 			}
-		};
+		}
 	}
 
 	if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING && status == DC_FAIL_HW_CURSOR_SUPPORT) {
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h
index 0fc66487d800..e1fa2e80a15a 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h
@@ -227,7 +227,8 @@ void dcn401_prepare_mcache_programming(struct dc *dc, struct dc_state *context);
 #define LE_DCN401_REG_LIST_RI(id)                                            \
 	LE_DCN3_REG_LIST_RI(id), \
 	SRI_ARR(DP_DPHY_INTERNAL_CTRL, DP, id), \
-	SRI_ARR(DIG_BE_CLK_CNTL, DIG, id)
+	SRI_ARR(DIG_BE_CLK_CNTL, DIG, id),\
+	SR_ARR(DIO_CLK_CNTL, id)
 
 /* DPP */
 #define DPP_REG_LIST_DCN401_COMMON_RI(id)                                    \
diff --git a/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c
index 6ffc74fc9dcd..ad088d70e189 100644
--- a/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/virtual/virtual_stream_encoder.c
@@ -44,11 +44,6 @@ static void virtual_stream_encoder_dvi_set_stream_attribute(
 	struct dc_crtc_timing *crtc_timing,
 	bool is_dual_link) {}
 
-static void virtual_stream_encoder_lvds_set_stream_attribute(
-	struct stream_encoder *enc,
-	struct dc_crtc_timing *crtc_timing)
-{}
-
 static void virtual_stream_encoder_set_throttled_vcp_size(
 	struct stream_encoder *enc,
 	struct fixed31_32 avg_time_slots_per_mtp)
@@ -120,8 +115,6 @@ static const struct stream_encoder_funcs virtual_str_enc_funcs = {
 		virtual_stream_encoder_hdmi_set_stream_attribute,
 	.dvi_set_stream_attribute =
 		virtual_stream_encoder_dvi_set_stream_attribute,
-	.lvds_set_stream_attribute =
-		virtual_stream_encoder_lvds_set_stream_attribute,
 	.set_throttled_vcp_size =
 		virtual_stream_encoder_set_throttled_vcp_size,
 	.update_hdmi_info_packets =
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 772e07a1a959..3f2a0ed02c59 100644
--- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
+++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
@@ -491,7 +491,13 @@ union replay_debug_flags {
 		 */
 		uint32_t debug_log_enabled : 1;
 
-		uint32_t reserved : 17;
+		/**
+		 * 0x8000 (bit 15)
+		 * @enable_sub_feature_visual_confirm: Enable Sub Feature Visual Confirm
+		 */
+		uint32_t enable_sub_feature_visual_confirm : 1;
+
+		uint32_t reserved : 16;
 	} bitfields;
 
 	uint32_t u32All;
@@ -599,6 +605,104 @@ union replay_hw_flags {
 	uint32_t u32All;
 };
 
+/**
+ * Flags that can be set by driver to change some Panel Replay behaviour.
+ */
+union pr_debug_flags {
+	struct {
+		/**
+		 * 0x1 (bit 0)
+		 * Enable visual confirm in FW.
+		 */
+		uint32_t visual_confirm : 1;
+
+		/**
+		 * 0x2 (bit 1)
+		 * @skip_crc: Set if need to skip CRC.
+		 */
+		uint32_t skip_crc : 1;
+
+		/**
+		 * 0x4 (bit 2)
+		 * @force_link_power_on: Force disable ALPM control
+		 */
+		uint32_t force_link_power_on : 1;
+
+		/**
+		 * 0x8 (bit 3)
+		 * @force_phy_power_on: Force phy power on
+		 */
+		uint32_t force_phy_power_on : 1;
+
+		/**
+		 * 0x10 (bit 4)
+		 * @skip_crtc_disabled: CRTC disable skipped
+		 */
+		uint32_t skip_crtc_disabled : 1;
+
+		/*
+		 * 0x20 (bit 5)
+		 * @visual_confirm_rate_control: Enable Visual Confirm rate control detection
+		 */
+		uint32_t visual_confirm_rate_control : 1;
+
+		uint32_t reserved : 26;
+	} bitfields;
+
+	uint32_t u32All;
+};
+
+union pr_hw_flags {
+	struct {
+		/**
+		 * @allow_alpm_fw_standby_mode: To indicate whether the
+		 * ALPM FW standby mode is allowed
+		 */
+		uint32_t allow_alpm_fw_standby_mode : 1;
+
+		/*
+		 * @dsc_enable_status: DSC enable status in driver
+		 */
+		uint32_t dsc_enable_status : 1;
+
+		/**
+		 * @fec_enable_status: receive fec enable/disable status from driver
+		 */
+		uint32_t fec_enable_status : 1;
+
+		/*
+		 * @smu_optimizations_en: SMU power optimization.
+		 * Only when active display is Replay capable and display enters Replay.
+		 * Trigger interrupt to SMU to powerup/down.
+		 */
+		uint32_t smu_optimizations_en : 1;
+
+		/**
+		 * @phy_power_state: Indicates current phy power state
+		 */
+		uint32_t phy_power_state : 1;
+
+		/**
+		 * @link_power_state: Indicates current link power state
+		 */
+		uint32_t link_power_state : 1;
+		/**
+		 * Use TPS3 signal when restore main link.
+		 */
+		uint32_t force_wakeup_by_tps3 : 1;
+		/**
+		 * @is_alpm_initialized: Indicates whether ALPM is initialized
+		 */
+		uint32_t is_alpm_initialized : 1;
+		/**
+		 * @alpm_mode: Indicates ALPM mode selected
+		 */
+		uint32_t alpm_mode : 2;
+	} bitfields;
+
+	uint32_t u32All;
+};
+
 union fw_assisted_mclk_switch_version {
 	struct {
 		uint8_t minor : 5;
@@ -623,6 +727,7 @@ struct dmub_feature_caps {
 	uint8_t replay_supported;
 	uint8_t replay_reserved[3];
 	uint8_t abm_aux_backlight_support;
+	uint8_t lsdma_support_in_dmu;
 };
 
 struct dmub_visual_confirm_color {
@@ -1732,9 +1837,15 @@ enum dmub_cmd_type {
 	DMUB_CMD__CURSOR_OFFLOAD = 92,
 
 	/**
-	 * Command type used for all SMART_POWER_HDR commands.
+	 * Command type used for all SMART_POWER_OLED commands.
+	 */
+	DMUB_CMD__SMART_POWER_OLED = 93,
+
+	/**
+	 * Command type use for all Panel Replay commands.
 	 */
-	DMUB_CMD__SMART_POWER_HDR = 93,
+	DMUB_CMD__PR = 94,
+
 
 	/**
 	 * Command type use for VBIOS shared commands.
@@ -2536,6 +2647,7 @@ struct dmub_cmd_fams2_global_config {
 
 union dmub_cmd_fams2_config {
 	struct dmub_cmd_fams2_global_config global;
+// coverity[cert_dcl37_c_violation:FALSE]  errno.h, stddef.h, stdint.h not included in atombios.h
 	struct dmub_fams2_stream_static_state stream; //v0
 	union {
 		struct dmub_fams2_cmd_stream_static_base_state base;
@@ -4148,6 +4260,33 @@ enum replay_state {
 };
 
 /**
+ * Definition of a panel replay state
+ */
+enum pr_state {
+	PR_STATE_0									= 0x00, // State 0 steady state
+	// Pending SDP and Unlock before back to State 0
+	PR_STATE_0_PENDING_SDP_AND_UNLOCK			= 0x01,
+	PR_STATE_1									= 0x10, // State 1
+	PR_STATE_2									= 0x20, // State 2 steady state
+	// Pending frame transmission before transition to State 2
+	PR_STATE_2_PENDING_FRAME_TRANSMISSION		= 0x30,
+	// Active and Powered Up
+	PR_STATE_2_POWERED							= 0x31,
+	// Active and Powered Down, but need to blank HUBP after DPG_EN latch
+	PR_STATE_2_PENDING_HUBP_BLANK				= 0x32,
+	// Active and Pending Power Up
+	PR_STATE_2_PENDING_POWER_UP					= 0x33,
+	// Active and Powered Up, Pending DPG latch
+	PR_STATE_2_PENDING_LOCK_FOR_DPG_POWER_ON	= 0x34,
+	// Active and Powered Up, Pending SDP and Unlock
+	PR_STATE_2_PENDING_SDP_AND_UNLOCK			= 0x35,
+	// Pending transmission of AS SDP for timing sync, but no rfb update
+	PR_STATE_2_PENDING_AS_SDP					= 0x36,
+	// Invalid
+	PR_STATE_INVALID							= 0xFF,
+};
+
+/**
  * Replay command sub-types.
  */
 enum dmub_cmd_replay_type {
@@ -4197,6 +4336,25 @@ enum dmub_cmd_replay_type {
 	DMUB_CMD__REPLAY_SET_GENERAL_CMD = 16,
 };
 
+/*
+ * Panel Replay sub-types
+ */
+enum dmub_cmd_panel_replay_type {
+	DMUB_CMD__PR_ENABLE = 0,
+	DMUB_CMD__PR_COPY_SETTINGS = 1,
+	DMUB_CMD__PR_UPDATE_STATE = 2,
+	DMUB_CMD__PR_GENERAL_CMD = 3,
+};
+
+enum dmub_cmd_panel_replay_state_update_subtype {
+	PR_STATE_UPDATE_COASTING_VTOTAL = 0x1,
+	PR_STATE_UPDATE_SYNC_MODE = 0x2,
+};
+
+enum dmub_cmd_panel_replay_general_subtype {
+	PR_GENERAL_CMD_DEBUG_OPTION = 0x1,
+};
+
 /**
  * Replay general command sub-types.
  */
@@ -4212,6 +4370,7 @@ enum dmub_cmd_replay_general_subtype {
 	REPLAY_GENERAL_CMD_DISABLED_DESYNC_ERROR_DETECTION,
 	REPLAY_GENERAL_CMD_UPDATE_ERROR_STATUS,
 	REPLAY_GENERAL_CMD_SET_LOW_RR_ACTIVATE,
+	REPLAY_GENERAL_CMD_VIDEO_CONFERENCING,
 };
 
 struct dmub_alpm_auxless_data {
@@ -4349,17 +4508,13 @@ struct dmub_cmd_replay_set_version_data {
 	 */
 	uint8_t panel_inst;
 	/**
-	 * PSR version that FW should implement.
+	 * Replay version that FW should implement.
 	 */
 	enum replay_version version;
 	/**
-	 * PSR control version.
-	 */
-	uint8_t cmd_version;
-	/**
 	 * Explicit padding to 4 byte boundary.
 	 */
-	uint8_t pad[2];
+	uint8_t pad[3];
 };
 
 /**
@@ -4405,11 +4560,11 @@ enum replay_enable {
 };
 
 /**
- * Data passed from driver to FW in a DMUB_CMD__SMART_POWER_HDR_ENABLE command.
+ * Data passed from driver to FW in a DMUB_CMD__SMART_POWER_OLED_ENABLE command.
  */
-struct dmub_rb_cmd_smart_power_hdr_enable_data {
+struct dmub_rb_cmd_smart_power_oled_enable_data {
 	/**
-	 * SMART_POWER_HDR enable or disable.
+	 * SMART_POWER_OLED enable or disable.
 	 */
 	uint8_t enable;
 	/**
@@ -4777,53 +4932,53 @@ union dmub_replay_cmd_set {
 };
 
 /**
- * SMART POWER HDR command sub-types.
+ * SMART POWER OLED command sub-types.
  */
-enum dmub_cmd_smart_power_hdr_type {
+enum dmub_cmd_smart_power_oled_type {
 
 	/**
-	 * Enable/Disable SMART_POWER_HDR.
+	 * Enable/Disable SMART_POWER_OLED.
 	 */
-	DMUB_CMD__SMART_POWER_HDR_ENABLE = 1,
+	DMUB_CMD__SMART_POWER_OLED_ENABLE = 1,
 	/**
-	 * Get current MaxCLL value if SMART POWER HDR is enabled.
+	 * Get current MaxCLL value if SMART POWER OLED is enabled.
 	 */
-	DMUB_CMD__SMART_POWER_HDR_GETMAXCLL = 2,
+	DMUB_CMD__SMART_POWER_OLED_GETMAXCLL = 2,
 };
 
 /**
- * Definition of a DMUB_CMD__SMART_POWER_HDR command.
+ * Definition of a DMUB_CMD__SMART_POWER_OLED command.
  */
-struct dmub_rb_cmd_smart_power_hdr_enable {
+struct dmub_rb_cmd_smart_power_oled_enable {
 	/**
 	 * Command header.
 	 */
 	struct dmub_cmd_header header;
 
-	struct dmub_rb_cmd_smart_power_hdr_enable_data data;
+	struct dmub_rb_cmd_smart_power_oled_enable_data data;
 };
 
-struct dmub_cmd_smart_power_hdr_getmaxcll_input {
+struct dmub_cmd_smart_power_oled_getmaxcll_input {
 	uint8_t panel_inst;
 	uint8_t pad[3];
 };
 
-struct dmub_cmd_smart_power_hdr_getmaxcll_output {
+struct dmub_cmd_smart_power_oled_getmaxcll_output {
 	uint16_t current_max_cll;
 	uint8_t pad[2];
 };
 
 /**
- * Definition of a DMUB_CMD__SMART_POWER_HDR command.
+ * Definition of a DMUB_CMD__SMART_POWER_OLED command.
  */
-struct dmub_rb_cmd_smart_power_hdr_getmaxcll {
+struct dmub_rb_cmd_smart_power_oled_getmaxcll {
 	struct dmub_cmd_header header; /**< Command header */
 	/**
-	 * Data passed from driver to FW in a DMUB_CMD__SMART_POWER_HDR_GETMAXCLL command.
+	 * Data passed from driver to FW in a DMUB_CMD__SMART_POWER_OLED_GETMAXCLL command.
 	 */
-	union dmub_cmd_smart_power_hdr_getmaxcll_data {
-		struct dmub_cmd_smart_power_hdr_getmaxcll_input input; /**< Input */
-		struct dmub_cmd_smart_power_hdr_getmaxcll_output output; /**< Output */
+	union dmub_cmd_smart_power_oled_getmaxcll_data {
+		struct dmub_cmd_smart_power_oled_getmaxcll_input input; /**< Input */
+		struct dmub_cmd_smart_power_oled_getmaxcll_output output; /**< Output */
 		uint32_t output_raw; /**< Raw data output */
 	} data;
 };
@@ -6357,6 +6512,223 @@ struct dmub_rb_cmd_cursor_offload_stream_cntl {
 };
 
 /**
+ * Data passed from driver to FW in a DMUB_CMD__PR_ENABLE command.
+ */
+struct dmub_cmd_pr_enable_data {
+	/**
+	 * Panel Replay enable or disable.
+	 */
+	uint8_t enable;
+	/**
+	 * Panel Instance.
+	 * Panel isntance to identify which replay_state to use
+	 * Currently the support is only for 0 or 1
+	 */
+	uint8_t panel_inst;
+	/**
+	 * Phy state to enter.
+	 * Values to use are defined in dmub_phy_fsm_state
+	 */
+	uint8_t phy_fsm_state;
+	/**
+	 * Phy rate for DP - RBR/HBR/HBR2/HBR3.
+	 * Set this using enum phy_link_rate.
+	 * This does not support HDMI/DP2 for now.
+	 */
+	uint8_t phy_rate;
+	/**
+	 * @hpo_stream_enc_inst: HPO stream encoder instance
+	 */
+	uint8_t hpo_stream_enc_inst;
+	/**
+	 * @hpo_link_enc_inst: HPO link encoder instance
+	 */
+	uint8_t hpo_link_enc_inst;
+	/**
+	 * @pad: Align structure to 4 byte boundary.
+	 */
+	uint8_t pad[2];
+};
+
+/**
+ * Definition of a DMUB_CMD__PR_ENABLE command.
+ * Panel Replay enable/disable is controlled using action in data.
+ */
+struct dmub_rb_cmd_pr_enable {
+	/**
+	 * Command header.
+	 */
+	struct dmub_cmd_header header;
+
+	struct dmub_cmd_pr_enable_data data;
+};
+
+/**
+ * Data passed from driver to FW in a DMUB_CMD__PR_COPY_SETTINGS command.
+ */
+struct dmub_cmd_pr_copy_settings_data {
+	/**
+	 * Flags that can be set by driver to change some replay behaviour.
+	 */
+	union pr_debug_flags debug;
+
+	/**
+	 * @flags: Flags used to determine feature functionality.
+	 */
+	union pr_hw_flags flags;
+
+	/**
+	 * DPP HW instance.
+	 */
+	uint8_t dpp_inst;
+	/**
+	 * OTG HW instance.
+	 */
+	uint8_t otg_inst;
+	/**
+	 * DIG FE HW instance.
+	 */
+	uint8_t digfe_inst;
+	/**
+	 * DIG BE HW instance.
+	 */
+	uint8_t digbe_inst;
+	/**
+	 * AUX HW instance.
+	 */
+	uint8_t aux_inst;
+	/**
+	 * Panel Instance.
+	 * Panel isntance to identify which psr_state to use
+	 * Currently the support is only for 0 or 1
+	 */
+	uint8_t panel_inst;
+	/**
+	 * Length of each horizontal line in ns.
+	 */
+	uint32_t line_time_in_ns;
+	/**
+	 * PHY instance.
+	 */
+	uint8_t dpphy_inst;
+	/**
+	 * Determines if SMU optimzations are enabled/disabled.
+	 */
+	uint8_t smu_optimizations_en;
+	/*
+	 * Use FSM state for Replay power up/down
+	 */
+	uint8_t use_phy_fsm;
+	/*
+	 * Use FSFT afftet pixel clk
+	 */
+	uint32_t pix_clk_100hz;
+	/*
+	 * Use Original pixel clock
+	 */
+	uint32_t sink_pix_clk_100hz;
+	/**
+	 * Use for AUX-less ALPM LFPS wake operation
+	 */
+	struct dmub_alpm_auxless_data auxless_alpm_data;
+	/**
+	 * @hpo_stream_enc_inst: HPO stream encoder instance
+	 */
+	uint8_t hpo_stream_enc_inst;
+	/**
+	 * @hpo_link_enc_inst: HPO link encoder instance
+	 */
+	uint8_t hpo_link_enc_inst;
+	/**
+	 * @pad: Align structure to 4 byte boundary.
+	 */
+	uint8_t pad[2];
+};
+
+/**
+ * Definition of a DMUB_CMD__PR_COPY_SETTINGS command.
+ */
+struct dmub_rb_cmd_pr_copy_settings {
+	/**
+	 * Command header.
+	 */
+	struct dmub_cmd_header header;
+	/**
+	 * Data passed from driver to FW in a DMUB_CMD__PR_COPY_SETTINGS command.
+	 */
+	struct dmub_cmd_pr_copy_settings_data data;
+};
+
+struct dmub_cmd_pr_update_state_data {
+	/**
+	 * Panel Instance.
+	 * Panel isntance to identify which psr_state to use
+	 * Currently the support is only for 0 or 1
+	 */
+	uint8_t panel_inst;
+
+	uint8_t pad[3]; // align to 4-byte boundary
+	/*
+	 * Update flags to control the update behavior.
+	 */
+	uint32_t update_flag;
+	/**
+	 * state/data to set.
+	 */
+	uint32_t coasting_vtotal;
+	uint32_t sync_mode;
+};
+
+struct dmub_cmd_pr_general_cmd_data {
+	/**
+	 * Panel Instance.
+	 * Panel isntance to identify which psr_state to use
+	 * Currently the support is only for 0 or 1
+	 */
+	uint8_t panel_inst;
+	/**
+	 * subtype: PR general cmd sub type
+	 */
+	uint8_t subtype;
+
+	uint8_t pad[2];
+	/**
+	 * config data by different subtypes
+	 */
+	union {
+		uint32_t u32All;
+	} data;
+};
+
+/**
+ * Definition of a DMUB_CMD__PR_UPDATE_STATE command.
+ */
+struct dmub_rb_cmd_pr_update_state {
+	/**
+	 * Command header.
+	 */
+	struct dmub_cmd_header header;
+	/**
+	 * Data passed from driver to FW in a DMUB_CMD__PR_UPDATE_STATE command.
+	 */
+	struct dmub_cmd_pr_update_state_data data;
+};
+
+/**
+ * Definition of a DMUB_CMD__PR_GENERAL_CMD command.
+ */
+struct dmub_rb_cmd_pr_general_cmd {
+	/**
+	 * Command header.
+	 */
+	struct dmub_cmd_header header;
+	/**
+	 * Data passed from driver to FW in a DMUB_CMD__PR_GENERAL_CMD command.
+	 */
+	struct dmub_cmd_pr_general_cmd_data data;
+};
+
+/**
  * union dmub_rb_cmd - DMUB inbox command.
  */
 union dmub_rb_cmd {
@@ -6698,13 +7070,25 @@ union dmub_rb_cmd {
 	 */
 	struct dmub_rb_cmd_cursor_offload_stream_cntl cursor_offload_stream_ctnl;
 	/**
-	 * Definition of a DMUB_CMD__SMART_POWER_HDR_ENABLE command.
+	 * Definition of a DMUB_CMD__SMART_POWER_OLED_ENABLE command.
 	 */
-	struct dmub_rb_cmd_smart_power_hdr_enable smart_power_hdr_enable;
+	struct dmub_rb_cmd_smart_power_oled_enable smart_power_oled_enable;
 	/**
-	 * Definition of a DMUB_CMD__DMUB_CMD__SMART_POWER_HDR_GETMAXCLL command.
+	 * Definition of a DMUB_CMD__DMUB_CMD__SMART_POWER_OLED_GETMAXCLL command.
 	 */
-	struct dmub_rb_cmd_smart_power_hdr_getmaxcll smart_power_hdr_getmaxcll;
+	struct dmub_rb_cmd_smart_power_oled_getmaxcll smart_power_oled_getmaxcll;
+	/*
+	 * Definition of a DMUB_CMD__REPLAY_COPY_SETTINGS command.
+	 */
+	struct dmub_rb_cmd_pr_copy_settings pr_copy_settings;
+	/**
+	 * Definition of a DMUB_CMD__REPLAY_ENABLE command.
+	 */
+	struct dmub_rb_cmd_pr_enable pr_enable;
+
+	struct dmub_rb_cmd_pr_update_state pr_update_state;
+
+	struct dmub_rb_cmd_pr_general_cmd pr_general_cmd;
 };
 
 /**
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 a657efda89ce..a6ae1d2e9685 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
@@ -66,7 +66,7 @@
 #define DMUB_SCRATCH_MEM_SIZE (1024)
 
 /* Default indirect buffer size. */
-#define DMUB_IB_MEM_SIZE (1280)
+#define DMUB_IB_MEM_SIZE (2560)
 
 /* Default LSDMA ring buffer size. */
 #define DMUB_LSDMA_RB_SIZE (64 * 1024)
diff --git a/drivers/gpu/drm/amd/display/include/audio_types.h b/drivers/gpu/drm/amd/display/include/audio_types.h
index e4a26143f14c..6699ad4fa825 100644
--- a/drivers/gpu/drm/amd/display/include/audio_types.h
+++ b/drivers/gpu/drm/amd/display/include/audio_types.h
@@ -47,15 +47,15 @@ struct audio_crtc_info {
 	uint32_t h_total;
 	uint32_t h_active;
 	uint32_t v_active;
-	uint32_t pixel_repetition;
 	uint32_t requested_pixel_clock_100Hz; /* in 100Hz */
 	uint32_t calculated_pixel_clock_100Hz; /* in 100Hz */
-	uint32_t refresh_rate;
+	uint32_t dsc_bits_per_pixel;
+	uint32_t dsc_num_slices;
 	enum dc_color_depth color_depth;
 	enum dc_pixel_encoding pixel_encoding;
+	uint16_t refresh_rate;
+	uint8_t pixel_repetition;
 	bool interlaced;
-	uint32_t dsc_bits_per_pixel;
-	uint32_t dsc_num_slices;
 };
 struct azalia_clock_info {
 	uint32_t pixel_clock_in_10khz;
@@ -78,11 +78,9 @@ enum audio_dto_source {
 
 struct audio_pll_info {
 	uint32_t audio_dto_source_clock_in_khz;
-	uint32_t feed_back_divider;
+	uint32_t ss_percentage;
 	enum audio_dto_source dto_source;
 	bool ss_enabled;
-	uint32_t ss_percentage;
-	uint32_t ss_percentage_divider;
 };
 
 struct audio_channel_associate_info {
diff --git a/drivers/gpu/drm/amd/display/include/dpcd_defs.h b/drivers/gpu/drm/amd/display/include/dpcd_defs.h
index de8f3cfed6c8..07b937b92efc 100644
--- a/drivers/gpu/drm/amd/display/include/dpcd_defs.h
+++ b/drivers/gpu/drm/amd/display/include/dpcd_defs.h
@@ -30,6 +30,22 @@
 #ifndef DP_SINK_HW_REVISION_START // can remove this once the define gets into linux drm_dp_helper.h
 #define DP_SINK_HW_REVISION_START 0x409
 #endif
+/* Panel Replay*/
+#ifndef DP_PANEL_REPLAY_CAPABILITY_SUPPORT // can remove this once the define gets into linux drm_dp_helper.h
+#define DP_PANEL_REPLAY_CAPABILITY_SUPPORT 0x0b0
+#endif /* DP_PANEL_REPLAY_CAPABILITY_SUPPORT */
+#ifndef DP_PANEL_REPLAY_CAPABILITY // can remove this once the define gets into linux drm_dp_helper.h
+#define DP_PANEL_REPLAY_CAPABILITY 0x0b1
+#endif /* DP_PANEL_REPLAY_CAPABILITY */
+#ifndef DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_1  // can remove this once the define gets into linux drm_dp_helper.h
+#define DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_1  0x1b0
+#endif /* DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_1 */
+#ifndef DP_PANEL_REPLAY_ENABLE // can remove this once the define gets into linux drm_dp_helper.h
+#define DP_PANEL_REPLAY_ENABLE (1 << 0)
+#endif /* DP_PANEL_REPLAY_ENABLE */
+#ifndef DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_2 // can remove this once the define gets into linux drm_dp_helper.h
+#define DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_2 0x1b1
+#endif /* DP_PANEL_REPLAY_ENABLE_AND_CONFIGURATION_2 */
 
 enum dpcd_revision {
 	DPCD_REV_10 = 0x10,
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index ce421bcddcb0..1aae46d703ba 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -1260,6 +1260,17 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync,
 		update_v_total_for_static_ramp(
 				core_freesync, stream, in_out_vrr);
 	}
+
+	/*
+	 * If VRR is inactive, set vtotal min and max to nominal vtotal
+	 */
+	 if (in_out_vrr->state == VRR_STATE_INACTIVE) {
+		in_out_vrr->adjust.v_total_min =
+			mod_freesync_calc_v_total_from_refresh(stream,
+				in_out_vrr->max_refresh_in_uhz);
+		in_out_vrr->adjust.v_total_max = in_out_vrr->adjust.v_total_min;
+		return;
+	}
 }
 
 unsigned long long mod_freesync_calc_nominal_field_rate(
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
index b883d626f1c3..26a351a184f3 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
@@ -88,6 +88,7 @@ struct mod_hdcp_transition_input_hdcp2 {
 	uint8_t lc_init_write;
 	uint8_t l_prime_available_poll;
 	uint8_t l_prime_read;
+	uint8_t l_prime_combo_read;
 	uint8_t l_prime_validation;
 	uint8_t eks_prepare;
 	uint8_t eks_write;
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c
index 5628f0ef73fd..27500abf9fee 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c
@@ -465,54 +465,11 @@ out:
 	return status;
 }
 
-static enum mod_hdcp_status locality_check_sw(struct mod_hdcp *hdcp,
-		struct mod_hdcp_event_context *event_ctx,
-		struct mod_hdcp_transition_input_hdcp2 *input)
-{
-	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
-
-	if (!mod_hdcp_execute_and_set(mod_hdcp_write_lc_init,
-			&input->lc_init_write, &status,
-			 hdcp, "lc_init_write"))
-		goto out;
-	if (is_dp_hdcp(hdcp))
-		msleep(16);
-	else
-		if (!mod_hdcp_execute_and_set(poll_l_prime_available,
-				&input->l_prime_available_poll, &status,
-				hdcp, "l_prime_available_poll"))
-			goto out;
-	if (!mod_hdcp_execute_and_set(mod_hdcp_read_l_prime,
-			&input->l_prime_read, &status,
-			hdcp, "l_prime_read"))
-		goto out;
-out:
-	return status;
-}
-
-static enum mod_hdcp_status locality_check_fw(struct mod_hdcp *hdcp,
-		struct mod_hdcp_event_context *event_ctx,
-		struct mod_hdcp_transition_input_hdcp2 *input)
-{
-	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
-
-	if (!mod_hdcp_execute_and_set(mod_hdcp_write_poll_read_lc_fw,
-			&input->l_prime_read, &status,
-			hdcp, "l_prime_read"))
-		goto out;
-
-out:
-	return status;
-}
-
 static enum mod_hdcp_status locality_check(struct mod_hdcp *hdcp,
 		struct mod_hdcp_event_context *event_ctx,
 		struct mod_hdcp_transition_input_hdcp2 *input)
 {
 	enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
-	const bool use_fw = hdcp->config.ddc.funcs.atomic_write_poll_read_i2c
-			&& hdcp->config.ddc.funcs.atomic_write_poll_read_aux
-			&& !hdcp->connection.link.adjust.hdcp2.force_sw_locality_check;
 
 	if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
 		event_ctx->unexpected_event = 1;
@@ -524,9 +481,28 @@ static enum mod_hdcp_status locality_check(struct mod_hdcp *hdcp,
 			hdcp, "lc_init_prepare"))
 		goto out;
 
-	status = (use_fw ? locality_check_fw : locality_check_sw)(hdcp, event_ctx, input);
-	if (status != MOD_HDCP_STATUS_SUCCESS)
-		goto out;
+	if (hdcp->connection.link.adjust.hdcp2.use_fw_locality_check) {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_write_poll_read_lc_fw,
+				&input->l_prime_combo_read, &status,
+				hdcp, "l_prime_combo_read"))
+			goto out;
+	} else {
+		if (!mod_hdcp_execute_and_set(mod_hdcp_write_lc_init,
+				&input->lc_init_write, &status,
+				hdcp, "lc_init_write"))
+			goto out;
+		if (is_dp_hdcp(hdcp))
+			msleep(16);
+		else
+			if (!mod_hdcp_execute_and_set(poll_l_prime_available,
+					&input->l_prime_available_poll, &status,
+					hdcp, "l_prime_available_poll"))
+				goto out;
+		if (!mod_hdcp_execute_and_set(mod_hdcp_read_l_prime,
+				&input->l_prime_read, &status,
+				hdcp, "l_prime_read"))
+			goto out;
+	}
 
 	if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp2_validate_l_prime,
 			&input->l_prime_validation, &status,
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c
index 89ffb89e1932..9316312a4df5 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c
@@ -184,31 +184,33 @@ enum mod_hdcp_status mod_hdcp_hdcp2_transition(struct mod_hdcp *hdcp,
 		callback_in_ms(0, output);
 		set_state_id(hdcp, output, H2_A2_LOCALITY_CHECK);
 		break;
-	case H2_A2_LOCALITY_CHECK: {
-		const bool use_fw = hdcp->config.ddc.funcs.atomic_write_poll_read_i2c
-				&& !adjust->hdcp2.force_sw_locality_check;
-
-		/*
-		 * 1A-05: consider disconnection after LC init a failure
-		 * 1A-13-1: consider invalid l' a failure
-		 * 1A-13-2: consider l' timeout a failure
-		 */
+	case H2_A2_LOCALITY_CHECK:
+		/* 1A-05: consider disconnection after LC init a failure */
 		if (hdcp->state.stay_count > 10 ||
-				input->lc_init_prepare != PASS ||
-				(!use_fw && input->lc_init_write != PASS) ||
-				(!use_fw && input->l_prime_available_poll != PASS)) {
+				input->lc_init_prepare != PASS) {
 			fail_and_restart_in_ms(0, &status, output);
 			break;
-		} else if (input->l_prime_read != PASS) {
-			if (use_fw && hdcp->config.debug.lc_enable_sw_fallback) {
-				adjust->hdcp2.force_sw_locality_check = true;
+		} else if (adjust->hdcp2.use_fw_locality_check &&
+				input->l_prime_combo_read != PASS) {
+			/* 1A-13-2: consider l' timeout a failure */
+			if (adjust->hdcp2.use_sw_locality_fallback) {
+				/* switch to software locality check */
+				adjust->hdcp2.use_fw_locality_check = 0;
 				callback_in_ms(0, output);
+				increment_stay_counter(hdcp);
 				break;
 			}
-
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (!adjust->hdcp2.use_fw_locality_check &&
+					(input->lc_init_write != PASS ||
+					input->l_prime_available_poll != PASS ||
+					input->l_prime_read != PASS)) {
+			/* 1A-13-2: consider l' timeout a failure */
 			fail_and_restart_in_ms(0, &status, output);
 			break;
 		} else if (input->l_prime_validation != PASS) {
+			/* 1A-13-1: consider invalid l' a failure */
 			callback_in_ms(0, output);
 			increment_stay_counter(hdcp);
 			break;
@@ -216,7 +218,6 @@ enum mod_hdcp_status mod_hdcp_hdcp2_transition(struct mod_hdcp *hdcp,
 		callback_in_ms(0, output);
 		set_state_id(hdcp, output, H2_A3_EXCHANGE_KS_AND_TEST_FOR_REPEATER);
 		break;
-	}
 	case H2_A3_EXCHANGE_KS_AND_TEST_FOR_REPEATER:
 		if (input->eks_prepare != PASS ||
 				input->eks_write != PASS) {
@@ -510,26 +511,29 @@ enum mod_hdcp_status mod_hdcp_hdcp2_dp_transition(struct mod_hdcp *hdcp,
 		callback_in_ms(0, output);
 		set_state_id(hdcp, output, D2_A2_LOCALITY_CHECK);
 		break;
-	case D2_A2_LOCALITY_CHECK: {
-		const bool use_fw = hdcp->config.ddc.funcs.atomic_write_poll_read_aux
-				&& !adjust->hdcp2.force_sw_locality_check;
-
+	case D2_A2_LOCALITY_CHECK:
 		if (hdcp->state.stay_count > 10 ||
-				input->lc_init_prepare != PASS ||
-				(!use_fw && input->lc_init_write != PASS)) {
-			/* 1A-12: consider invalid l' a failure */
+				input->lc_init_prepare != PASS) {
 			fail_and_restart_in_ms(0, &status, output);
 			break;
-		} else if (input->l_prime_read != PASS) {
-			if (use_fw && hdcp->config.debug.lc_enable_sw_fallback) {
-				adjust->hdcp2.force_sw_locality_check = true;
+		} else if (adjust->hdcp2.use_fw_locality_check &&
+				input->l_prime_combo_read != PASS) {
+			if (adjust->hdcp2.use_sw_locality_fallback) {
+				/* switch to software locality check */
+				adjust->hdcp2.use_fw_locality_check = 0;
 				callback_in_ms(0, output);
+				increment_stay_counter(hdcp);
 				break;
 			}
-
+			fail_and_restart_in_ms(0, &status, output);
+			break;
+		} else if (!adjust->hdcp2.use_fw_locality_check &&
+					(input->lc_init_write != PASS ||
+					input->l_prime_read != PASS)) {
 			fail_and_restart_in_ms(0, &status, output);
 			break;
 		} else if (input->l_prime_validation != PASS) {
+			/* 1A-12: consider invalid l' a failure */
 			callback_in_ms(0, output);
 			increment_stay_counter(hdcp);
 			break;
@@ -537,7 +541,6 @@ enum mod_hdcp_status mod_hdcp_hdcp2_dp_transition(struct mod_hdcp *hdcp,
 		callback_in_ms(0, output);
 		set_state_id(hdcp, output, D2_A34_EXCHANGE_KS_AND_TEST_FOR_REPEATER);
 		break;
-	}
 	case D2_A34_EXCHANGE_KS_AND_TEST_FOR_REPEATER:
 		if (input->eks_prepare != PASS ||
 				input->eks_write != PASS) {
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
index 2e6408579194..0ca39873f807 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
@@ -758,6 +758,6 @@ enum mod_hdcp_status mod_hdcp_write_poll_read_lc_fw(struct mod_hdcp *hdcp)
 {
 	const bool success = (is_dp_hdcp(hdcp) ? write_stall_read_lc_fw_aux : write_poll_read_lc_fw_i2c)(hdcp);
 
-	return success ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_DDC_FAILURE;
+	return success ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_HDCP2_LOCALITY_COMBO_READ_FAILURE;
 }
 
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c
index 6b3b5f610907..5cb979c2cf8c 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_log.c
@@ -125,129 +125,11 @@ void mod_hdcp_log_ddc_trace(struct mod_hdcp *hdcp)
 	}
 }
 
+#define CASE_FORMAT(entry) case entry: return #entry;
 char *mod_hdcp_status_to_str(int32_t status)
 {
 	switch (status) {
-	case MOD_HDCP_STATUS_SUCCESS:
-		return "MOD_HDCP_STATUS_SUCCESS";
-	case MOD_HDCP_STATUS_FAILURE:
-		return "MOD_HDCP_STATUS_FAILURE";
-	case MOD_HDCP_STATUS_RESET_NEEDED:
-		return "MOD_HDCP_STATUS_RESET_NEEDED";
-	case MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND:
-		return "MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND";
-	case MOD_HDCP_STATUS_DISPLAY_NOT_FOUND:
-		return "MOD_HDCP_STATUS_DISPLAY_NOT_FOUND";
-	case MOD_HDCP_STATUS_INVALID_STATE:
-		return "MOD_HDCP_STATUS_INVALID_STATE";
-	case MOD_HDCP_STATUS_NOT_IMPLEMENTED:
-		return "MOD_HDCP_STATUS_NOT_IMPLEMENTED";
-	case MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE:
-		return "MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE";
-	case MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE:
-		return "MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE";
-	case MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE:
-		return "MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE";
-	case MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE:
-		return "MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER:
-		return "MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER";
-	case MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE:
-		return "MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE";
-	case MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING:
-		return "MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING";
-	case MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_BKSV_REVOKED:
-		return "MOD_HDCP_STATUS_HDCP1_BKSV_REVOKED";
-	case MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY:
-		return "MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY";
-	case MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_KSV_LIST_REVOKED:
-		return "MOD_HDCP_STATUS_HDCP1_KSV_LIST_REVOKED";
-	case MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED:
-		return "MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED";
-	case MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE";
-	case MOD_HDCP_STATUS_HDCP1_INVALID_BKSV:
-		return "MOD_HDCP_STATUS_HDCP1_INVALID_BKSV";
-	case MOD_HDCP_STATUS_DDC_FAILURE:
-		return "MOD_HDCP_STATUS_DDC_FAILURE";
-	case MOD_HDCP_STATUS_INVALID_OPERATION:
-		return "MOD_HDCP_STATUS_INVALID_OPERATION";
-	case MOD_HDCP_STATUS_HDCP2_NOT_CAPABLE:
-		return "MOD_HDCP_STATUS_HDCP2_NOT_CAPABLE";
-	case MOD_HDCP_STATUS_HDCP2_CREATE_SESSION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_CREATE_SESSION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_DESTROY_SESSION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_DESTROY_SESSION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_PREP_AKE_INIT_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_PREP_AKE_INIT_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_AKE_CERT_PENDING:
-		return "MOD_HDCP_STATUS_HDCP2_AKE_CERT_PENDING";
-	case MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING:
-		return "MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING";
-	case MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING:
-		return "MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING";
-	case MOD_HDCP_STATUS_HDCP2_VALIDATE_AKE_CERT_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_VALIDATE_AKE_CERT_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_AKE_CERT_REVOKED:
-		return "MOD_HDCP_STATUS_HDCP2_AKE_CERT_REVOKED";
-	case MOD_HDCP_STATUS_HDCP2_VALIDATE_H_PRIME_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_VALIDATE_H_PRIME_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_VALIDATE_PAIRING_INFO_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_VALIDATE_PAIRING_INFO_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_PREP_LC_INIT_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_PREP_LC_INIT_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_L_PRIME_PENDING:
-		return "MOD_HDCP_STATUS_HDCP2_L_PRIME_PENDING";
-	case MOD_HDCP_STATUS_HDCP2_VALIDATE_L_PRIME_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_VALIDATE_L_PRIME_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_PREP_EKS_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_PREP_EKS_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_ENABLE_ENCRYPTION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_ENABLE_ENCRYPTION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_VALIDATE_RX_ID_LIST_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_VALIDATE_RX_ID_LIST_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_REVOKED:
-		return "MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_REVOKED";
-	case MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_NOT_READY:
-		return "MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_NOT_READY";
-	case MOD_HDCP_STATUS_HDCP2_ENABLE_STREAM_ENCRYPTION_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_ENABLE_STREAM_ENCRYPTION_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_STREAM_READY_PENDING:
-		return "MOD_HDCP_STATUS_HDCP2_STREAM_READY_PENDING";
-	case MOD_HDCP_STATUS_HDCP2_VALIDATE_STREAM_READY_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_VALIDATE_STREAM_READY_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_PREPARE_STREAM_MANAGEMENT_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_PREPARE_STREAM_MANAGEMENT_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_REAUTH_REQUEST:
-		return "MOD_HDCP_STATUS_HDCP2_REAUTH_REQUEST";
-	case MOD_HDCP_STATUS_HDCP2_REAUTH_LINK_INTEGRITY_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_REAUTH_LINK_INTEGRITY_FAILURE";
-	case MOD_HDCP_STATUS_HDCP2_DEVICE_COUNT_MISMATCH_FAILURE:
-		return "MOD_HDCP_STATUS_HDCP2_DEVICE_COUNT_MISMATCH_FAILURE";
-	case MOD_HDCP_STATUS_UNSUPPORTED_PSP_VER_FAILURE:
-		return "MOD_HDCP_STATUS_UNSUPPORTED_PSP_VER_FAILURE";
+	MOD_HDCP_STATUS_LIST(CASE_FORMAT)
 	default:
 		return "MOD_HDCP_STATUS_UNKNOWN";
 	}
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 46e52fb3a118..835467225458 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
@@ -35,69 +35,74 @@ struct mod_hdcp;
 #define MAX_NUM_OF_DISPLAYS 6
 #define MAX_NUM_OF_ATTEMPTS 4
 #define MAX_NUM_OF_ERROR_TRACE 10
+#define MOD_HDCP_STATUS_LIST(FORMAT) \
+	FORMAT(MOD_HDCP_STATUS_SUCCESS) \
+	FORMAT(MOD_HDCP_STATUS_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_RESET_NEEDED) \
+	FORMAT(MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND) \
+	FORMAT(MOD_HDCP_STATUS_DISPLAY_NOT_FOUND) \
+	FORMAT(MOD_HDCP_STATUS_INVALID_STATE) \
+	FORMAT(MOD_HDCP_STATUS_NOT_IMPLEMENTED) \
+	FORMAT(MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_BKSV_REVOKED) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_KSV_LIST_REVOKED) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP1_INVALID_BKSV) \
+	FORMAT(MOD_HDCP_STATUS_DDC_FAILURE) /* TODO: specific errors */ \
+	FORMAT(MOD_HDCP_STATUS_INVALID_OPERATION) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_NOT_CAPABLE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_CREATE_SESSION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_DESTROY_SESSION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_PREP_AKE_INIT_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_AKE_CERT_PENDING) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_VALIDATE_AKE_CERT_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_AKE_CERT_REVOKED) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_VALIDATE_H_PRIME_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_VALIDATE_PAIRING_INFO_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_PREP_LC_INIT_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_L_PRIME_PENDING) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_VALIDATE_L_PRIME_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_PREP_EKS_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_ENABLE_ENCRYPTION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_NOT_READY) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_VALIDATE_RX_ID_LIST_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_REVOKED) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_ENABLE_STREAM_ENCRYPTION_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_STREAM_READY_PENDING) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_VALIDATE_STREAM_READY_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_PREPARE_STREAM_MANAGEMENT_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_REAUTH_REQUEST) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_REAUTH_LINK_INTEGRITY_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_DEVICE_COUNT_MISMATCH_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_UNSUPPORTED_PSP_VER_FAILURE) \
+	FORMAT(MOD_HDCP_STATUS_HDCP2_LOCALITY_COMBO_READ_FAILURE)
+
+#define ENUM_FORMAT(entry) entry,
 
 /* detailed return status */
 enum mod_hdcp_status {
-	MOD_HDCP_STATUS_SUCCESS = 0,
-	MOD_HDCP_STATUS_FAILURE,
-	MOD_HDCP_STATUS_RESET_NEEDED,
-	MOD_HDCP_STATUS_DISPLAY_OUT_OF_BOUND,
-	MOD_HDCP_STATUS_DISPLAY_NOT_FOUND,
-	MOD_HDCP_STATUS_INVALID_STATE,
-	MOD_HDCP_STATUS_NOT_IMPLEMENTED,
-	MOD_HDCP_STATUS_INTERNAL_POLICY_FAILURE,
-	MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE,
-	MOD_HDCP_STATUS_CREATE_PSP_SERVICE_FAILURE,
-	MOD_HDCP_STATUS_DESTROY_PSP_SERVICE_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_CREATE_SESSION_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_DESTROY_SESSION_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_VALIDATE_ENCRYPTION_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_NOT_HDCP_REPEATER,
-	MOD_HDCP_STATUS_HDCP1_NOT_CAPABLE,
-	MOD_HDCP_STATUS_HDCP1_R0_PRIME_PENDING,
-	MOD_HDCP_STATUS_HDCP1_VALIDATE_RX_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_BKSV_REVOKED,
-	MOD_HDCP_STATUS_HDCP1_KSV_LIST_NOT_READY,
-	MOD_HDCP_STATUS_HDCP1_VALIDATE_KSV_LIST_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_KSV_LIST_REVOKED,
-	MOD_HDCP_STATUS_HDCP1_ENABLE_ENCRYPTION_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_ENABLE_STREAM_ENCRYPTION_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_MAX_CASCADE_EXCEEDED_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_MAX_DEVS_EXCEEDED_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_DEVICE_COUNT_MISMATCH_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_LINK_INTEGRITY_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_REAUTH_REQUEST_ISSUED,
-	MOD_HDCP_STATUS_HDCP1_LINK_MAINTENANCE_FAILURE,
-	MOD_HDCP_STATUS_HDCP1_INVALID_BKSV,
-	MOD_HDCP_STATUS_DDC_FAILURE, /* TODO: specific errors */
-	MOD_HDCP_STATUS_INVALID_OPERATION,
-	MOD_HDCP_STATUS_HDCP2_NOT_CAPABLE,
-	MOD_HDCP_STATUS_HDCP2_CREATE_SESSION_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_DESTROY_SESSION_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_PREP_AKE_INIT_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_AKE_CERT_PENDING,
-	MOD_HDCP_STATUS_HDCP2_H_PRIME_PENDING,
-	MOD_HDCP_STATUS_HDCP2_PAIRING_INFO_PENDING,
-	MOD_HDCP_STATUS_HDCP2_VALIDATE_AKE_CERT_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_AKE_CERT_REVOKED,
-	MOD_HDCP_STATUS_HDCP2_VALIDATE_H_PRIME_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_VALIDATE_PAIRING_INFO_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_PREP_LC_INIT_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_L_PRIME_PENDING,
-	MOD_HDCP_STATUS_HDCP2_VALIDATE_L_PRIME_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_PREP_EKS_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_ENABLE_ENCRYPTION_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_NOT_READY,
-	MOD_HDCP_STATUS_HDCP2_VALIDATE_RX_ID_LIST_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_RX_ID_LIST_REVOKED,
-	MOD_HDCP_STATUS_HDCP2_ENABLE_STREAM_ENCRYPTION_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_STREAM_READY_PENDING,
-	MOD_HDCP_STATUS_HDCP2_VALIDATE_STREAM_READY_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_PREPARE_STREAM_MANAGEMENT_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_REAUTH_REQUEST,
-	MOD_HDCP_STATUS_HDCP2_REAUTH_LINK_INTEGRITY_FAILURE,
-	MOD_HDCP_STATUS_HDCP2_DEVICE_COUNT_MISMATCH_FAILURE,
-	MOD_HDCP_STATUS_UNSUPPORTED_PSP_VER_FAILURE,
+	MOD_HDCP_STATUS_LIST(ENUM_FORMAT)
 };
 
 struct mod_hdcp_displayport {
@@ -214,8 +219,9 @@ struct mod_hdcp_link_adjustment_hdcp2 {
 	uint8_t force_type		: 2;
 	uint8_t force_no_stored_km	: 1;
 	uint8_t increase_h_prime_timeout: 1;
-	uint8_t force_sw_locality_check : 1;
-	uint8_t reserved		: 2;
+	uint8_t use_fw_locality_check 	: 1;
+	uint8_t use_sw_locality_fallback: 1;
+	uint8_t reserved		: 1;
 };
 
 struct mod_hdcp_link_adjustment {
@@ -317,10 +323,6 @@ struct mod_hdcp_display_query {
 struct mod_hdcp_config {
 	struct mod_hdcp_psp psp;
 	struct mod_hdcp_ddc ddc;
-	struct {
-		uint8_t lc_enable_sw_fallback : 1;
-		uint8_t reserved : 7;
-	} debug;
 	uint8_t index;
 };
 
diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
index 29ccd3532d13..fd139b219bf9 100644
--- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
+++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
@@ -975,6 +975,34 @@ bool psr_su_set_dsc_slice_height(struct dc *dc, struct dc_link *link,
 	return true;
 }
 
+void set_replay_frame_skip_number(struct dc_link *link,
+	enum replay_coasting_vtotal_type type,
+	uint32_t coasting_vtotal_refresh_rate_mhz,
+	uint32_t flicker_free_refresh_rate_mhz,
+	bool is_defer)
+{
+	uint32_t *frame_skip_number_array = NULL;
+	uint32_t frame_skip_number = 0;
+
+	if (link == NULL || flicker_free_refresh_rate_mhz == 0 || coasting_vtotal_refresh_rate_mhz == 0)
+		return;
+
+	if (is_defer)
+		frame_skip_number_array = link->replay_settings.defer_frame_skip_number_table;
+	else
+		frame_skip_number_array = link->replay_settings.frame_skip_number_table;
+
+	if (frame_skip_number_array == NULL)
+		return;
+
+	frame_skip_number = coasting_vtotal_refresh_rate_mhz / flicker_free_refresh_rate_mhz;
+
+	if (frame_skip_number >= 1)
+		frame_skip_number_array[type] = frame_skip_number - 1;
+	else
+		frame_skip_number_array[type] = 0;
+}
+
 void set_replay_defer_update_coasting_vtotal(struct dc_link *link,
 	enum replay_coasting_vtotal_type type,
 	uint32_t vtotal)
@@ -987,6 +1015,8 @@ void update_replay_coasting_vtotal_from_defer(struct dc_link *link,
 {
 	link->replay_settings.coasting_vtotal_table[type] =
 		link->replay_settings.defer_update_coasting_vtotal_table[type];
+	link->replay_settings.frame_skip_number_table[type] =
+		link->replay_settings.defer_frame_skip_number_table[type];
 }
 
 void set_replay_coasting_vtotal(struct dc_link *link,
@@ -1007,6 +1037,9 @@ void calculate_replay_link_off_frame_count(struct dc_link *link,
 	uint8_t max_link_off_frame_count = 0;
 	uint16_t max_deviation_line = 0,  pixel_deviation_per_line = 0;
 
+	if (!link || link->replay_settings.config.replay_version != DC_FREESYNC_REPLAY)
+		return;
+
 	max_deviation_line = link->dpcd_caps.pr_info.max_deviation_line;
 	pixel_deviation_per_line = link->dpcd_caps.pr_info.pixel_deviation_per_line;
 
diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
index 391209a3bf29..87d31d9dce5a 100644
--- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
+++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
@@ -60,6 +60,11 @@ void set_replay_coasting_vtotal(struct dc_link *link,
 void set_replay_defer_update_coasting_vtotal(struct dc_link *link,
 	enum replay_coasting_vtotal_type type,
 	uint32_t vtotal);
+void set_replay_frame_skip_number(struct dc_link *link,
+	enum replay_coasting_vtotal_type type,
+	uint32_t coasting_vtotal_refresh_rate_Mhz,
+	uint32_t flicker_free_refresh_rate_Mhz,
+	bool is_defer);
 void update_replay_coasting_vtotal_from_defer(struct dc_link *link,
 	enum replay_coasting_vtotal_type type);
 void set_replay_low_rr_full_screen_video_src_vtotal(struct dc_link *link, uint16_t vtotal);
diff --git a/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_d.h b/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_d.h
index 2176548e9203..9778822dd2a0 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_d.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_d.h
@@ -60,5 +60,10 @@
 #define mmVCE_VCPU_CACHE_SIZE1 0x800C
 #define mmVCE_VCPU_CACHE_SIZE2 0x800E
 #define mmVCE_VCPU_CNTL 0x8005
+#define mmVCE_VCPU_SCRATCH7 0x8037
+#define mmVCE_FW_REG_STATUS 0x8384
+#define mmVCE_LMI_FW_PERIODIC_CTRL 0x8388
+#define mmVCE_LMI_FW_START_KEYSEL 0x8386
+
 
 #endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_sh_mask.h
index ea5b26b11cb1..1f82d6f5abde 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/vce/vce_1_0_sh_mask.h
@@ -61,6 +61,8 @@
 #define VCE_RB_WPTR__RB_WPTR__SHIFT 0x00000004
 #define VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK 0x00000001L
 #define VCE_SOFT_RESET__ECPU_SOFT_RESET__SHIFT 0x00000000
+#define VCE_SOFT_RESET__FME_SOFT_RESET_MASK 0x00000004L
+#define VCE_SOFT_RESET__FME_SOFT_RESET__SHIFT 0x00000002
 #define VCE_STATUS__JOB_BUSY_MASK 0x00000001L
 #define VCE_STATUS__JOB_BUSY__SHIFT 0x00000000
 #define VCE_STATUS__UENC_BUSY_MASK 0x00000100L
@@ -95,5 +97,13 @@
 #define VCE_VCPU_CNTL__CLK_EN__SHIFT 0x00000000
 #define VCE_VCPU_CNTL__RBBM_SOFT_RESET_MASK 0x00040000L
 #define VCE_VCPU_CNTL__RBBM_SOFT_RESET__SHIFT 0x00000012
+#define VCE_CLOCK_GATING_A__CGC_DYN_CLOCK_MODE_MASK 0x00010000
+#define VCE_CLOCK_GATING_A__CGC_DYN_CLOCK_MODE_SHIFT 0x00000010
+#define VCE_FW_REG_STATUS__BUSY_MASK 0x0000001
+#define VCE_FW_REG_STATUS__BUSY__SHIFT 0x0000001
+#define VCE_FW_REG_STATUS__PASS_MASK 0x0000008
+#define VCE_FW_REG_STATUS__PASS__SHIFT 0x0000003
+#define VCE_FW_REG_STATUS__DONE_MASK 0x0000800
+#define VCE_FW_REG_STATUS__DONE__SHIFT 0x000000b
 
 #endif
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index f92f78d5d330..2366e68262e6 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -1820,4 +1820,10 @@ struct amdgpu_partition_metrics_v1_0 {
 	uint64_t gfx_below_host_limit_total_acc[MAX_XCC];
 };
 
+struct amdgpu_partition_metrics_v1_1 {
+	struct metrics_table_header common_header;
+	int attr_count;
+	struct gpu_metrics_attr metrics_attrs[];
+};
+
 #endif
diff --git a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c
index 5d08dc3b7110..79b174e5326d 100644
--- a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c
@@ -195,24 +195,6 @@ int amdgpu_dpm_set_mp1_state(struct amdgpu_device *adev,
 	return ret;
 }
 
-int amdgpu_dpm_notify_rlc_state(struct amdgpu_device *adev, bool en)
-{
-	int ret = 0;
-	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
-
-	if (pp_funcs && pp_funcs->notify_rlc_state) {
-		mutex_lock(&adev->pm.mutex);
-
-		ret = pp_funcs->notify_rlc_state(
-				adev->powerplay.pp_handle,
-				en);
-
-		mutex_unlock(&adev->pm.mutex);
-	}
-
-	return ret;
-}
-
 int amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev)
 {
 	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
@@ -1205,8 +1187,11 @@ int amdgpu_dpm_get_pp_table(struct amdgpu_device *adev, char **table)
 	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
 	int ret = 0;
 
-	if (!pp_funcs->get_pp_table)
-		return 0;
+	if (!table)
+		return -EINVAL;
+
+	if (amdgpu_sriov_vf(adev) || !pp_funcs->get_pp_table || adev->scpm_enabled)
+		return -EOPNOTSUPP;
 
 	mutex_lock(&adev->pm.mutex);
 	ret = pp_funcs->get_pp_table(adev->powerplay.pp_handle,
@@ -1733,7 +1718,10 @@ int amdgpu_dpm_set_pp_table(struct amdgpu_device *adev,
 	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
 	int ret = 0;
 
-	if (!pp_funcs->set_pp_table)
+	if (!buf || !size)
+		return -EINVAL;
+
+	if (amdgpu_sriov_vf(adev) || !pp_funcs->set_pp_table || adev->scpm_enabled)
 		return -EOPNOTSUPP;
 
 	mutex_lock(&adev->pm.mutex);
@@ -2140,3 +2128,10 @@ ssize_t amdgpu_dpm_get_xcp_metrics(struct amdgpu_device *adev, int xcp_id,
 
 	return ret;
 }
+
+const struct ras_smu_drv *amdgpu_dpm_get_ras_smu_driver(struct amdgpu_device *adev)
+{
+	void *pp_handle = adev->powerplay.pp_handle;
+
+	return smu_get_ras_smu_driver(pp_handle);
+}
diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
index c88a76cce401..65296a819e6a 100644
--- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
@@ -2506,7 +2506,7 @@ static struct amdgpu_device_attr amdgpu_device_attrs[] = {
 	AMDGPU_DEVICE_ATTR_RO(pp_num_states,				ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF),
 	AMDGPU_DEVICE_ATTR_RO(pp_cur_state,				ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF),
 	AMDGPU_DEVICE_ATTR_RW(pp_force_state,				ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF),
-	AMDGPU_DEVICE_ATTR_RW(pp_table,					ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF),
+	AMDGPU_DEVICE_ATTR_RW(pp_table,					ATTR_FLAG_BASIC),
 	AMDGPU_DEVICE_ATTR_RW(pp_dpm_sclk,				ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF,
 			      .attr_update = pp_dpm_clk_default_attr_update),
 	AMDGPU_DEVICE_ATTR_RW(pp_dpm_mclk,				ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF,
@@ -2638,6 +2638,15 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_
 		if (amdgpu_dpm_get_apu_thermal_limit(adev, &limit) ==
 		    -EOPNOTSUPP)
 			*states = ATTR_STATE_UNSUPPORTED;
+	} else if (DEVICE_ATTR_IS(pp_table)) {
+		int ret;
+		char *tmp = NULL;
+
+		ret = amdgpu_dpm_get_pp_table(adev, &tmp);
+		if (ret == -EOPNOTSUPP || !tmp)
+			*states = ATTR_STATE_UNSUPPORTED;
+		else
+			*states = ATTR_STATE_SUPPORTED;
 	}
 
 	switch (gc_ver) {
@@ -3372,7 +3381,9 @@ static ssize_t amdgpu_hwmon_show_power_label(struct device *dev,
 				  to_sensor_dev_attr(attr)->index == PP_PWR_TYPE_FAST ?
 				  "fastPPT" : "slowPPT");
 	else
-		return sysfs_emit(buf, "PPT\n");
+		return sysfs_emit(buf, "%s\n",
+				  to_sensor_dev_attr(attr)->index == PP_PWR_TYPE_FAST ?
+				  "PPT1" : "PPT");
 }
 
 static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
@@ -3577,7 +3588,6 @@ static SENSOR_DEVICE_ATTR(power1_cap_min, S_IRUGO, amdgpu_hwmon_show_power_cap_m
 static SENSOR_DEVICE_ATTR(power1_cap, S_IRUGO | S_IWUSR, amdgpu_hwmon_show_power_cap, amdgpu_hwmon_set_power_cap, 0);
 static SENSOR_DEVICE_ATTR(power1_cap_default, S_IRUGO, amdgpu_hwmon_show_power_cap_default, NULL, 0);
 static SENSOR_DEVICE_ATTR(power1_label, S_IRUGO, amdgpu_hwmon_show_power_label, NULL, 0);
-static SENSOR_DEVICE_ATTR(power2_average, S_IRUGO, amdgpu_hwmon_show_power_avg, NULL, 1);
 static SENSOR_DEVICE_ATTR(power2_cap_max, S_IRUGO, amdgpu_hwmon_show_power_cap_max, NULL, 1);
 static SENSOR_DEVICE_ATTR(power2_cap_min, S_IRUGO, amdgpu_hwmon_show_power_cap_min, NULL, 1);
 static SENSOR_DEVICE_ATTR(power2_cap, S_IRUGO | S_IWUSR, amdgpu_hwmon_show_power_cap, amdgpu_hwmon_set_power_cap, 1);
@@ -3626,7 +3636,6 @@ static struct attribute *hwmon_attributes[] = {
 	&sensor_dev_attr_power1_cap.dev_attr.attr,
 	&sensor_dev_attr_power1_cap_default.dev_attr.attr,
 	&sensor_dev_attr_power1_label.dev_attr.attr,
-	&sensor_dev_attr_power2_average.dev_attr.attr,
 	&sensor_dev_attr_power2_cap_max.dev_attr.attr,
 	&sensor_dev_attr_power2_cap_min.dev_attr.attr,
 	&sensor_dev_attr_power2_cap.dev_attr.attr,
@@ -3825,13 +3834,14 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 		return 0;
 
 	/* only Vangogh has fast PPT limit and power labels */
-	if (!(gc_ver == IP_VERSION(10, 3, 1)) &&
-	    (attr == &sensor_dev_attr_power2_average.dev_attr.attr ||
-	     attr == &sensor_dev_attr_power2_cap_max.dev_attr.attr ||
+	if ((attr == &sensor_dev_attr_power2_cap_max.dev_attr.attr ||
 	     attr == &sensor_dev_attr_power2_cap_min.dev_attr.attr ||
 	     attr == &sensor_dev_attr_power2_cap.dev_attr.attr ||
 	     attr == &sensor_dev_attr_power2_cap_default.dev_attr.attr ||
-	     attr == &sensor_dev_attr_power2_label.dev_attr.attr))
+	     attr == &sensor_dev_attr_power2_label.dev_attr.attr) &&
+	     (amdgpu_dpm_get_power_limit(adev, &tmp,
+					 PP_PWR_LIMIT_MAX,
+					 PP_PWR_TYPE_FAST) == -EOPNOTSUPP))
 		return 0;
 
 	return effective_mode;
@@ -4723,14 +4733,14 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
 		ret = devm_device_add_group(adev->dev,
 					    &amdgpu_pm_policy_attr_group);
 		if (ret)
-			goto err_out0;
+			goto err_out1;
 	}
 
 	if (amdgpu_dpm_is_temp_metrics_supported(adev, SMU_TEMP_METRIC_GPUBOARD)) {
 		ret = devm_device_add_group(adev->dev,
 					    &amdgpu_board_attr_group);
 		if (ret)
-			goto err_out0;
+			goto err_out1;
 		if (amdgpu_pm_get_sensor_generic(adev, AMDGPU_PP_SENSOR_MAXNODEPOWERLIMIT,
 						 (void *)&tmp) != -EOPNOTSUPP) {
 			sysfs_add_file_to_group(&adev->dev->kobj,
diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h
index 3bce74f8bb0a..aa3f427819a0 100644
--- a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h
@@ -424,8 +424,6 @@ int amdgpu_dpm_mode1_reset(struct amdgpu_device *adev);
 int amdgpu_dpm_set_mp1_state(struct amdgpu_device *adev,
 			     enum pp_mp1_state mp1_state);
 
-int amdgpu_dpm_notify_rlc_state(struct amdgpu_device *adev, bool en);
-
 int amdgpu_dpm_set_gfx_power_up_by_imu(struct amdgpu_device *adev);
 
 int amdgpu_dpm_baco_exit(struct amdgpu_device *adev);
@@ -614,5 +612,6 @@ int amdgpu_dpm_reset_vcn(struct amdgpu_device *adev, uint32_t inst_mask);
 bool amdgpu_dpm_reset_vcn_is_supported(struct amdgpu_device *adev);
 bool amdgpu_dpm_is_temp_metrics_supported(struct amdgpu_device *adev,
 					  enum smu_temp_metric_type type);
+const struct ras_smu_drv *amdgpu_dpm_get_ras_smu_driver(struct amdgpu_device *adev);
 
 #endif
diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c
index 3a9522c17fee..1f539cc65f41 100644
--- a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c
+++ b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c
@@ -2558,18 +2558,13 @@ static int si_enable_power_containment(struct amdgpu_device *adev,
 		if (enable) {
 			if (!si_should_disable_uvd_powertune(adev, amdgpu_new_state)) {
 				smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_TDPClampingActive);
-				if (smc_result != PPSMC_Result_OK) {
+				if (smc_result != PPSMC_Result_OK)
 					ret = -EINVAL;
-					ni_pi->pc_enabled = false;
-				} else {
-					ni_pi->pc_enabled = true;
-				}
 			}
 		} else {
 			smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_TDPClampingInactive);
 			if (smc_result != PPSMC_Result_OK)
 				ret = -EINVAL;
-			ni_pi->pc_enabled = false;
 		}
 	}
 
@@ -7051,13 +7046,20 @@ static void si_set_vce_clock(struct amdgpu_device *adev,
 	if ((old_rps->evclk != new_rps->evclk) ||
 	    (old_rps->ecclk != new_rps->ecclk)) {
 		/* Turn the clocks on when encoding, off otherwise */
+		dev_dbg(adev->dev, "set VCE clocks: %u, %u\n", new_rps->evclk, new_rps->ecclk);
+
 		if (new_rps->evclk || new_rps->ecclk) {
-			/* Place holder for future VCE1.0 porting to amdgpu
-			vce_v1_0_enable_mgcg(adev, false, false);*/
+			amdgpu_asic_set_vce_clocks(adev, new_rps->evclk, new_rps->ecclk);
+			amdgpu_device_ip_set_clockgating_state(
+				adev, AMD_IP_BLOCK_TYPE_VCE, AMD_CG_STATE_UNGATE);
+			amdgpu_device_ip_set_powergating_state(
+				adev, AMD_IP_BLOCK_TYPE_VCE, AMD_PG_STATE_UNGATE);
 		} else {
-			/* Place holder for future VCE1.0 porting to amdgpu
-			vce_v1_0_enable_mgcg(adev, true, false);
-			amdgpu_asic_set_vce_clocks(adev, new_rps->evclk, new_rps->ecclk);*/
+			amdgpu_device_ip_set_powergating_state(
+				adev, AMD_IP_BLOCK_TYPE_VCE, AMD_PG_STATE_GATE);
+			amdgpu_device_ip_set_clockgating_state(
+				adev, AMD_IP_BLOCK_TYPE_VCE, AMD_CG_STATE_GATE);
+			amdgpu_asic_set_vce_clocks(adev, 0, 0);
 		}
 	}
 }
@@ -7509,8 +7511,6 @@ static int si_dpm_init(struct amdgpu_device *adev)
 	pi->pasi = CYPRESS_HASI_DFLT;
 	pi->vrc = SISLANDS_VRC_DFLT;
 
-	pi->gfx_clock_gating = true;
-
 	eg_pi->sclk_deep_sleep = true;
 	si_pi->sclk_deep_sleep_above_low = false;
 
@@ -7521,7 +7521,6 @@ static int si_dpm_init(struct amdgpu_device *adev)
 
 	eg_pi->dynamic_ac_timing = true;
 
-	eg_pi->light_sleep = true;
 #if defined(CONFIG_ACPI)
 	eg_pi->pcie_performance_request =
 		amdgpu_acpi_is_pcie_performance_request_supported(adev);
@@ -7582,6 +7581,7 @@ static void si_dpm_debugfs_print_current_performance_level(void *handle,
 	} else {
 		pl = &ps->performance_levels[current_index];
 		seq_printf(m, "uvd    vclk: %d dclk: %d\n", rps->vclk, rps->dclk);
+		seq_printf(m, "vce    evclk: %d ecclk: %d\n", rps->evclk, rps->ecclk);
 		seq_printf(m, "power level %d    sclk: %u mclk: %u vddc: %u vddci: %u pcie gen: %u\n",
 			   current_index, pl->sclk, pl->mclk, pl->vddc, pl->vddci, pl->pcie_gen + 1);
 	}
diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.h b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.h
index 11cb7874a6bb..3aed75fbf913 100644
--- a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.h
+++ b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.h
@@ -38,11 +38,7 @@
 #define	MC_ARB_DRAM_TIMING2_2				0xa00
 #define	MC_ARB_DRAM_TIMING2_3				0xa01
 
-#define MAX_NO_OF_MVDD_VALUES 2
-#define MAX_NO_VREG_STEPS 32
 #define NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16
-#define SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE 32
-#define SMC_NISLANDS_MC_REGISTER_ARRAY_SET_COUNT 20
 #define RV770_ASI_DFLT                                1000
 #define CYPRESS_HASI_DFLT                               400000
 #define PCIE_PERF_REQ_PECI_GEN1         2
@@ -51,11 +47,6 @@
 #define RV770_DEFAULT_VCLK_FREQ  53300 /* 10 khz */
 #define RV770_DEFAULT_DCLK_FREQ  40000 /* 10 khz */
 
-#define SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE 16
-
-#define RV770_SMC_TABLE_ADDRESS 0xB000
-#define RV770_SMC_PERFORMANCE_LEVELS_PER_SWSTATE    3
-
 #define SMC_STROBE_RATIO    0x0F
 #define SMC_STROBE_ENABLE   0x10
 
@@ -64,27 +55,6 @@
 #define SMC_MC_RTT_ENABLE   0x04
 #define SMC_MC_STUTTER_EN   0x08
 
-#define RV770_SMC_VOLTAGEMASK_VDDC 0
-#define RV770_SMC_VOLTAGEMASK_MVDD 1
-#define RV770_SMC_VOLTAGEMASK_VDDCI 2
-#define RV770_SMC_VOLTAGEMASK_MAX  4
-
-#define NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16
-#define NISLANDS_SMC_STROBE_RATIO    0x0F
-#define NISLANDS_SMC_STROBE_ENABLE   0x10
-
-#define NISLANDS_SMC_MC_EDC_RD_FLAG  0x01
-#define NISLANDS_SMC_MC_EDC_WR_FLAG  0x02
-#define NISLANDS_SMC_MC_RTT_ENABLE   0x04
-#define NISLANDS_SMC_MC_STUTTER_EN   0x08
-
-#define MAX_NO_VREG_STEPS 32
-
-#define NISLANDS_SMC_VOLTAGEMASK_VDDC  0
-#define NISLANDS_SMC_VOLTAGEMASK_MVDD  1
-#define NISLANDS_SMC_VOLTAGEMASK_VDDCI 2
-#define NISLANDS_SMC_VOLTAGEMASK_MAX   4
-
 #define SISLANDS_MCREGISTERTABLE_INITIAL_SLOT               0
 #define SISLANDS_MCREGISTERTABLE_ACPI_SLOT                  1
 #define SISLANDS_MCREGISTERTABLE_ULV_SLOT                   2
@@ -219,32 +189,6 @@ enum si_cac_config_reg_type
 	SISLANDS_CACCONFIG_MAX
 };
 
-enum si_power_level {
-	SI_POWER_LEVEL_LOW = 0,
-	SI_POWER_LEVEL_MEDIUM = 1,
-	SI_POWER_LEVEL_HIGH = 2,
-	SI_POWER_LEVEL_CTXSW = 3,
-};
-
-enum si_td {
-	SI_TD_AUTO,
-	SI_TD_UP,
-	SI_TD_DOWN,
-};
-
-enum si_display_watermark {
-	SI_DISPLAY_WATERMARK_LOW = 0,
-	SI_DISPLAY_WATERMARK_HIGH = 1,
-};
-
-enum si_display_gap
-{
-    SI_PM_DISPLAY_GAP_VBLANK_OR_WM = 0,
-    SI_PM_DISPLAY_GAP_VBLANK       = 1,
-    SI_PM_DISPLAY_GAP_WATERMARK    = 2,
-    SI_PM_DISPLAY_GAP_IGNORE       = 3,
-};
-
 extern const struct amdgpu_ip_block_version si_smu_ip_block;
 
 struct ni_leakage_coeffients
@@ -258,56 +202,6 @@ struct ni_leakage_coeffients
 	u32 t_ref;
 };
 
-struct SMC_Evergreen_MCRegisterAddress
-{
-    uint16_t s0;
-    uint16_t s1;
-};
-
-typedef struct SMC_Evergreen_MCRegisterAddress SMC_Evergreen_MCRegisterAddress;
-
-struct evergreen_mc_reg_entry {
-	u32 mclk_max;
-	u32 mc_data[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
-};
-
-struct evergreen_mc_reg_table {
-	u8 last;
-	u8 num_entries;
-	u16 valid_flag;
-	struct evergreen_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
-	SMC_Evergreen_MCRegisterAddress mc_reg_address[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
-};
-
-struct SMC_Evergreen_MCRegisterSet
-{
-    uint32_t value[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
-};
-
-typedef struct SMC_Evergreen_MCRegisterSet SMC_Evergreen_MCRegisterSet;
-
-struct SMC_Evergreen_MCRegisters
-{
-    uint8_t                             last;
-    uint8_t                             reserved[3];
-    SMC_Evergreen_MCRegisterAddress     address[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
-    SMC_Evergreen_MCRegisterSet         data[5];
-};
-
-typedef struct SMC_Evergreen_MCRegisters SMC_Evergreen_MCRegisters;
-
-struct SMC_NIslands_MCRegisterSet
-{
-    uint32_t value[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
-};
-
-typedef struct SMC_NIslands_MCRegisterSet SMC_NIslands_MCRegisterSet;
-
-struct ni_mc_reg_entry {
-	u32 mclk_max;
-	u32 mc_data[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
-};
-
 struct SMC_NIslands_MCRegisterAddress
 {
     uint16_t s0;
@@ -316,257 +210,20 @@ struct SMC_NIslands_MCRegisterAddress
 
 typedef struct SMC_NIslands_MCRegisterAddress SMC_NIslands_MCRegisterAddress;
 
-struct SMC_NIslands_MCRegisters
-{
-    uint8_t                             last;
-    uint8_t                             reserved[3];
-    SMC_NIslands_MCRegisterAddress      address[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
-    SMC_NIslands_MCRegisterSet          data[SMC_NISLANDS_MC_REGISTER_ARRAY_SET_COUNT];
-};
-
-typedef struct SMC_NIslands_MCRegisters SMC_NIslands_MCRegisters;
-
-struct evergreen_ulv_param {
-	bool supported;
-	struct rv7xx_pl *pl;
-};
-
-struct evergreen_arb_registers {
-	u32 mc_arb_dram_timing;
-	u32 mc_arb_dram_timing2;
-	u32 mc_arb_rfsh_rate;
-	u32 mc_arb_burst_time;
-};
-
-struct at {
-	u32 rlp;
-	u32 rmp;
-	u32 lhp;
-	u32 lmp;
-};
-
-struct ni_clock_registers {
-	u32 cg_spll_func_cntl;
-	u32 cg_spll_func_cntl_2;
-	u32 cg_spll_func_cntl_3;
-	u32 cg_spll_func_cntl_4;
-	u32 cg_spll_spread_spectrum;
-	u32 cg_spll_spread_spectrum_2;
-	u32 mclk_pwrmgt_cntl;
-	u32 dll_cntl;
-	u32 mpll_ad_func_cntl;
-	u32 mpll_ad_func_cntl_2;
-	u32 mpll_dq_func_cntl;
-	u32 mpll_dq_func_cntl_2;
-	u32 mpll_ss1;
-	u32 mpll_ss2;
-};
-
-struct RV770_SMC_SCLK_VALUE
-{
-    uint32_t        vCG_SPLL_FUNC_CNTL;
-    uint32_t        vCG_SPLL_FUNC_CNTL_2;
-    uint32_t        vCG_SPLL_FUNC_CNTL_3;
-    uint32_t        vCG_SPLL_SPREAD_SPECTRUM;
-    uint32_t        vCG_SPLL_SPREAD_SPECTRUM_2;
-    uint32_t        sclk_value;
-};
-
-typedef struct RV770_SMC_SCLK_VALUE RV770_SMC_SCLK_VALUE;
-
-struct RV770_SMC_MCLK_VALUE
-{
-    uint32_t        vMPLL_AD_FUNC_CNTL;
-    uint32_t        vMPLL_AD_FUNC_CNTL_2;
-    uint32_t        vMPLL_DQ_FUNC_CNTL;
-    uint32_t        vMPLL_DQ_FUNC_CNTL_2;
-    uint32_t        vMCLK_PWRMGT_CNTL;
-    uint32_t        vDLL_CNTL;
-    uint32_t        vMPLL_SS;
-    uint32_t        vMPLL_SS2;
-    uint32_t        mclk_value;
-};
-
-typedef struct RV770_SMC_MCLK_VALUE RV770_SMC_MCLK_VALUE;
-
-
-struct RV730_SMC_MCLK_VALUE
-{
-    uint32_t        vMCLK_PWRMGT_CNTL;
-    uint32_t        vDLL_CNTL;
-    uint32_t        vMPLL_FUNC_CNTL;
-    uint32_t        vMPLL_FUNC_CNTL2;
-    uint32_t        vMPLL_FUNC_CNTL3;
-    uint32_t        vMPLL_SS;
-    uint32_t        vMPLL_SS2;
-    uint32_t        mclk_value;
-};
-
-typedef struct RV730_SMC_MCLK_VALUE RV730_SMC_MCLK_VALUE;
-
-struct RV770_SMC_VOLTAGE_VALUE
-{
-    uint16_t             value;
-    uint8_t              index;
-    uint8_t              padding;
-};
-
-typedef struct RV770_SMC_VOLTAGE_VALUE RV770_SMC_VOLTAGE_VALUE;
-
-union RV7XX_SMC_MCLK_VALUE
-{
-    RV770_SMC_MCLK_VALUE    mclk770;
-    RV730_SMC_MCLK_VALUE    mclk730;
-};
-
-typedef union RV7XX_SMC_MCLK_VALUE RV7XX_SMC_MCLK_VALUE, *LPRV7XX_SMC_MCLK_VALUE;
-
-struct RV770_SMC_HW_PERFORMANCE_LEVEL
-{
-    uint8_t                 arbValue;
-    union{
-        uint8_t             seqValue;
-        uint8_t             ACIndex;
-    };
-    uint8_t                 displayWatermark;
-    uint8_t                 gen2PCIE;
-    uint8_t                 gen2XSP;
-    uint8_t                 backbias;
-    uint8_t                 strobeMode;
-    uint8_t                 mcFlags;
-    uint32_t                aT;
-    uint32_t                bSP;
-    RV770_SMC_SCLK_VALUE    sclk;
-    RV7XX_SMC_MCLK_VALUE    mclk;
-    RV770_SMC_VOLTAGE_VALUE vddc;
-    RV770_SMC_VOLTAGE_VALUE mvdd;
-    RV770_SMC_VOLTAGE_VALUE vddci;
-    uint8_t                 reserved1;
-    uint8_t                 reserved2;
-    uint8_t                 stateFlags;
-    uint8_t                 padding;
-};
-
-typedef struct RV770_SMC_HW_PERFORMANCE_LEVEL RV770_SMC_HW_PERFORMANCE_LEVEL;
-
-struct RV770_SMC_SWSTATE
-{
-    uint8_t           flags;
-    uint8_t           padding1;
-    uint8_t           padding2;
-    uint8_t           padding3;
-    RV770_SMC_HW_PERFORMANCE_LEVEL levels[RV770_SMC_PERFORMANCE_LEVELS_PER_SWSTATE];
-};
-
-typedef struct RV770_SMC_SWSTATE RV770_SMC_SWSTATE;
-
-struct RV770_SMC_VOLTAGEMASKTABLE
-{
-    uint8_t  highMask[RV770_SMC_VOLTAGEMASK_MAX];
-    uint32_t lowMask[RV770_SMC_VOLTAGEMASK_MAX];
-};
-
-typedef struct RV770_SMC_VOLTAGEMASKTABLE RV770_SMC_VOLTAGEMASKTABLE;
-
-struct RV770_SMC_STATETABLE
-{
-    uint8_t             thermalProtectType;
-    uint8_t             systemFlags;
-    uint8_t             maxVDDCIndexInPPTable;
-    uint8_t             extraFlags;
-    uint8_t             highSMIO[MAX_NO_VREG_STEPS];
-    uint32_t            lowSMIO[MAX_NO_VREG_STEPS];
-    RV770_SMC_VOLTAGEMASKTABLE voltageMaskTable;
-    RV770_SMC_SWSTATE   initialState;
-    RV770_SMC_SWSTATE   ACPIState;
-    RV770_SMC_SWSTATE   driverState;
-    RV770_SMC_SWSTATE   ULVState;
-};
-
-typedef struct RV770_SMC_STATETABLE RV770_SMC_STATETABLE;
-
-struct vddc_table_entry {
-	u16 vddc;
-	u8 vddc_index;
-	u8 high_smio;
-	u32 low_smio;
-};
-
-struct rv770_clock_registers {
-	u32 cg_spll_func_cntl;
-	u32 cg_spll_func_cntl_2;
-	u32 cg_spll_func_cntl_3;
-	u32 cg_spll_spread_spectrum;
-	u32 cg_spll_spread_spectrum_2;
-	u32 mpll_ad_func_cntl;
-	u32 mpll_ad_func_cntl_2;
-	u32 mpll_dq_func_cntl;
-	u32 mpll_dq_func_cntl_2;
-	u32 mclk_pwrmgt_cntl;
-	u32 dll_cntl;
-	u32 mpll_ss1;
-	u32 mpll_ss2;
-};
-
-struct rv730_clock_registers {
-	u32 cg_spll_func_cntl;
-	u32 cg_spll_func_cntl_2;
-	u32 cg_spll_func_cntl_3;
-	u32 cg_spll_spread_spectrum;
-	u32 cg_spll_spread_spectrum_2;
-	u32 mclk_pwrmgt_cntl;
-	u32 dll_cntl;
-	u32 mpll_func_cntl;
-	u32 mpll_func_cntl2;
-	u32 mpll_func_cntl3;
-	u32 mpll_ss;
-	u32 mpll_ss2;
-};
-
-union r7xx_clock_registers {
-	struct rv770_clock_registers rv770;
-	struct rv730_clock_registers rv730;
-};
-
 struct rv7xx_power_info {
 	/* flags */
-	bool mem_gddr5;
-	bool pcie_gen2;
-	bool dynamic_pcie_gen2;
-	bool acpi_pcie_gen2;
-	bool boot_in_gen2;
 	bool voltage_control; /* vddc */
 	bool mvdd_control;
 	bool sclk_ss;
 	bool mclk_ss;
 	bool dynamic_ss;
-	bool gfx_clock_gating;
-	bool mg_clock_gating;
-	bool mgcgtssm;
-	bool power_gating;
 	bool thermal_protection;
-	bool display_gap;
-	bool dcodt;
-	bool ulps;
-	/* registers */
-	union r7xx_clock_registers clk_regs;
-	u32 s0_vid_lower_smio_cntl;
 	/* voltage */
-	u32 vddc_mask_low;
-	u32 mvdd_mask_low;
 	u32 mvdd_split_frequency;
-	u32 mvdd_low_smio[MAX_NO_OF_MVDD_VALUES];
 	u16 max_vddc;
 	u16 max_vddc_in_table;
 	u16 min_vddc_in_table;
-	struct vddc_table_entry vddc_table[MAX_NO_VREG_STEPS];
-	u8 valid_vddc_entries;
-	/* dc odt */
-	u32 mclk_odt_threshold;
-	u8 odt_value_0[2];
-	u8 odt_value_1[2];
 	/* stored values */
-	u32 boot_sclk;
 	u16 acpi_vddc;
 	u32 ref_div;
 	u32 active_auto_throttle_sources;
@@ -582,17 +239,6 @@ struct rv7xx_power_info {
 	u32 asi;
 	u32 pasi;
 	u32 vrc;
-	u32 restricted_levels;
-	u32 rlp;
-	u32 rmp;
-	u32 lhp;
-	u32 lmp;
-	/* smc offsets */
-	u16 state_table_start;
-	u16 soft_regs_start;
-	u16 sram_end;
-	/* scratch structs */
-	RV770_SMC_STATETABLE smc_statetable;
 };
 
 enum si_pcie_gen {
@@ -611,44 +257,12 @@ struct rv7xx_pl {
 	enum si_pcie_gen pcie_gen; /* si+ only */
 };
 
-struct rv7xx_ps {
-	struct rv7xx_pl high;
-	struct rv7xx_pl medium;
-	struct rv7xx_pl low;
-	bool dc_compatible;
-};
-
 struct si_ps {
 	u16 performance_level_count;
 	bool dc_compatible;
 	struct rv7xx_pl performance_levels[NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE];
 };
 
-struct ni_mc_reg_table {
-	u8 last;
-	u8 num_entries;
-	u16 valid_flag;
-	struct ni_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
-	SMC_NIslands_MCRegisterAddress mc_reg_address[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
-};
-
-struct ni_cac_data
-{
-	struct ni_leakage_coeffients leakage_coefficients;
-	u32 i_leakage;
-	s32 leakage_minimum_temperature;
-	u32 pwr_const;
-	u32 dc_cac_value;
-	u32 bif_cac_value;
-	u32 lkge_pwr;
-	u8 mc_wr_weight;
-	u8 mc_rd_weight;
-	u8 allow_ovrflw;
-	u8 num_win_tdp;
-	u8 l2num_win_tdp;
-	u8 lts_truncate_n;
-};
-
 struct evergreen_power_info {
 	/* must be first! */
 	struct rv7xx_power_info rv7xx;
@@ -657,203 +271,33 @@ struct evergreen_power_info {
 	bool dynamic_ac_timing;
 	bool abm;
 	bool mcls;
-	bool light_sleep;
-	bool memory_transition;
 	bool pcie_performance_request;
-	bool pcie_performance_request_registered;
 	bool sclk_deep_sleep;
-	bool dll_default_on;
-	bool ls_clock_gating;
 	bool smu_uvd_hs;
 	bool uvd_enabled;
 	/* stored values */
 	u16 acpi_vddci;
-	u8 mvdd_high_index;
-	u8 mvdd_low_index;
 	u32 mclk_edc_wr_enable_threshold;
-	struct evergreen_mc_reg_table mc_reg_table;
 	struct atom_voltage_table vddc_voltage_table;
 	struct atom_voltage_table vddci_voltage_table;
-	struct evergreen_arb_registers bootup_arb_registers;
-	struct evergreen_ulv_param ulv;
-	struct at ats[2];
-	/* smc offsets */
-	u16 mc_reg_table_start;
 	struct amdgpu_ps current_rps;
-	struct rv7xx_ps current_ps;
 	struct amdgpu_ps requested_rps;
-	struct rv7xx_ps requested_ps;
-};
-
-struct PP_NIslands_Dpm2PerfLevel
-{
-    uint8_t     MaxPS;
-    uint8_t     TgtAct;
-    uint8_t     MaxPS_StepInc;
-    uint8_t     MaxPS_StepDec;
-    uint8_t     PSST;
-    uint8_t     NearTDPDec;
-    uint8_t     AboveSafeInc;
-    uint8_t     BelowSafeInc;
-    uint8_t     PSDeltaLimit;
-    uint8_t     PSDeltaWin;
-    uint8_t     Reserved[6];
-};
-
-typedef struct PP_NIslands_Dpm2PerfLevel PP_NIslands_Dpm2PerfLevel;
-
-struct PP_NIslands_DPM2Parameters
-{
-    uint32_t    TDPLimit;
-    uint32_t    NearTDPLimit;
-    uint32_t    SafePowerLimit;
-    uint32_t    PowerBoostLimit;
-};
-typedef struct PP_NIslands_DPM2Parameters PP_NIslands_DPM2Parameters;
-
-struct NISLANDS_SMC_SCLK_VALUE
-{
-    uint32_t        vCG_SPLL_FUNC_CNTL;
-    uint32_t        vCG_SPLL_FUNC_CNTL_2;
-    uint32_t        vCG_SPLL_FUNC_CNTL_3;
-    uint32_t        vCG_SPLL_FUNC_CNTL_4;
-    uint32_t        vCG_SPLL_SPREAD_SPECTRUM;
-    uint32_t        vCG_SPLL_SPREAD_SPECTRUM_2;
-    uint32_t        sclk_value;
-};
-
-typedef struct NISLANDS_SMC_SCLK_VALUE NISLANDS_SMC_SCLK_VALUE;
-
-struct NISLANDS_SMC_MCLK_VALUE
-{
-    uint32_t        vMPLL_FUNC_CNTL;
-    uint32_t        vMPLL_FUNC_CNTL_1;
-    uint32_t        vMPLL_FUNC_CNTL_2;
-    uint32_t        vMPLL_AD_FUNC_CNTL;
-    uint32_t        vMPLL_AD_FUNC_CNTL_2;
-    uint32_t        vMPLL_DQ_FUNC_CNTL;
-    uint32_t        vMPLL_DQ_FUNC_CNTL_2;
-    uint32_t        vMCLK_PWRMGT_CNTL;
-    uint32_t        vDLL_CNTL;
-    uint32_t        vMPLL_SS;
-    uint32_t        vMPLL_SS2;
-    uint32_t        mclk_value;
-};
-
-typedef struct NISLANDS_SMC_MCLK_VALUE NISLANDS_SMC_MCLK_VALUE;
-
-struct NISLANDS_SMC_VOLTAGE_VALUE
-{
-    uint16_t             value;
-    uint8_t              index;
-    uint8_t              padding;
-};
-
-typedef struct NISLANDS_SMC_VOLTAGE_VALUE NISLANDS_SMC_VOLTAGE_VALUE;
-
-struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL
-{
-    uint8_t                     arbValue;
-    uint8_t                     ACIndex;
-    uint8_t                     displayWatermark;
-    uint8_t                     gen2PCIE;
-    uint8_t                     reserved1;
-    uint8_t                     reserved2;
-    uint8_t                     strobeMode;
-    uint8_t                     mcFlags;
-    uint32_t                    aT;
-    uint32_t                    bSP;
-    NISLANDS_SMC_SCLK_VALUE     sclk;
-    NISLANDS_SMC_MCLK_VALUE     mclk;
-    NISLANDS_SMC_VOLTAGE_VALUE  vddc;
-    NISLANDS_SMC_VOLTAGE_VALUE  mvdd;
-    NISLANDS_SMC_VOLTAGE_VALUE  vddci;
-    NISLANDS_SMC_VOLTAGE_VALUE  std_vddc;
-    uint32_t                    powergate_en;
-    uint8_t                     hUp;
-    uint8_t                     hDown;
-    uint8_t                     stateFlags;
-    uint8_t                     arbRefreshState;
-    uint32_t                    SQPowerThrottle;
-    uint32_t                    SQPowerThrottle_2;
-    uint32_t                    reserved[2];
-    PP_NIslands_Dpm2PerfLevel   dpm2;
-};
-
-typedef struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL NISLANDS_SMC_HW_PERFORMANCE_LEVEL;
-
-struct NISLANDS_SMC_SWSTATE
-{
-    uint8_t                             flags;
-    uint8_t                             levelCount;
-    uint8_t                             padding2;
-    uint8_t                             padding3;
-    NISLANDS_SMC_HW_PERFORMANCE_LEVEL   levels[];
-};
-
-typedef struct NISLANDS_SMC_SWSTATE NISLANDS_SMC_SWSTATE;
-
-struct NISLANDS_SMC_VOLTAGEMASKTABLE
-{
-    uint8_t  highMask[NISLANDS_SMC_VOLTAGEMASK_MAX];
-    uint32_t lowMask[NISLANDS_SMC_VOLTAGEMASK_MAX];
-};
-
-typedef struct NISLANDS_SMC_VOLTAGEMASKTABLE NISLANDS_SMC_VOLTAGEMASKTABLE;
-
-#define NISLANDS_MAX_NO_VREG_STEPS 32
-
-struct NISLANDS_SMC_STATETABLE
-{
-    uint8_t                             thermalProtectType;
-    uint8_t                             systemFlags;
-    uint8_t                             maxVDDCIndexInPPTable;
-    uint8_t                             extraFlags;
-    uint8_t                             highSMIO[NISLANDS_MAX_NO_VREG_STEPS];
-    uint32_t                            lowSMIO[NISLANDS_MAX_NO_VREG_STEPS];
-    NISLANDS_SMC_VOLTAGEMASKTABLE       voltageMaskTable;
-    PP_NIslands_DPM2Parameters          dpm2Params;
-    NISLANDS_SMC_SWSTATE                initialState;
-    NISLANDS_SMC_SWSTATE                ACPIState;
-    NISLANDS_SMC_SWSTATE                ULVState;
-    NISLANDS_SMC_SWSTATE                driverState;
-    NISLANDS_SMC_HW_PERFORMANCE_LEVEL   dpmLevels[NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE - 1];
 };
 
-typedef struct NISLANDS_SMC_STATETABLE NISLANDS_SMC_STATETABLE;
-
 struct ni_power_info {
 	/* must be first! */
 	struct evergreen_power_info eg;
-	struct ni_clock_registers clock_registers;
-	struct ni_mc_reg_table mc_reg_table;
 	u32 mclk_rtt_mode_threshold;
 	/* flags */
-	bool use_power_boost_limit;
 	bool support_cac_long_term_average;
 	bool cac_enabled;
 	bool cac_configuration_required;
 	bool driver_calculate_cac_leakage;
-	bool pc_enabled;
 	bool enable_power_containment;
 	bool enable_cac;
 	bool enable_sq_ramping;
-	/* smc offsets */
-	u16 arb_table_start;
-	u16 fan_table_start;
-	u16 cac_table_start;
-	u16 spll_table_start;
-	/* CAC stuff */
-	struct ni_cac_data cac_data;
-	u32 dc_cac_table[NISLANDS_DCCAC_MAX_LEVELS];
-	const struct ni_cac_weights *cac_weights;
-	u8 lta_window_size;
-	u8 lts_truncate;
 	struct si_ps current_ps;
 	struct si_ps requested_ps;
-	/* scratch structs */
-	SMC_NIslands_MCRegisters smc_mc_reg_table;
-	NISLANDS_SMC_STATETABLE smc_statetable;
 };
 
 struct si_cac_config_reg
@@ -952,7 +396,6 @@ struct si_leakage_voltage
 	struct si_leakage_voltage_entry entries[SISLANDS_MAX_LEAKAGE_COUNT];
 };
 
-
 struct si_ulv_param {
 	bool supported;
 	u32 cg_ulv_control;
diff --git a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c
index 76a5353d7f4a..3aaf3dd71868 100644
--- a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c
@@ -631,9 +631,12 @@ static int pp_dpm_get_pp_table(void *handle, char **table)
 {
 	struct pp_hwmgr *hwmgr = handle;
 
-	if (!hwmgr || !hwmgr->pm_en || !hwmgr->soft_pp_table)
+	if (!hwmgr || !hwmgr->pm_en || !table)
 		return -EINVAL;
 
+	if (!hwmgr->soft_pp_table)
+		return -EOPNOTSUPP;
+
 	*table = (char *)hwmgr->soft_pp_table;
 	return hwmgr->soft_pp_table_size;
 }
diff --git a/drivers/gpu/drm/amd/pm/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/pm/powerplay/smumgr/smu10_smumgr.c
index ac9ec8257f82..38e19e5cad4d 100644
--- a/drivers/gpu/drm/amd/pm/powerplay/smumgr/smu10_smumgr.c
+++ b/drivers/gpu/drm/amd/pm/powerplay/smumgr/smu10_smumgr.c
@@ -139,7 +139,7 @@ static int smu10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			priv->smu_tables.entry[table_id].table_id,
 			NULL);
 
-	amdgpu_asic_invalidate_hdp(adev, NULL);
+	amdgpu_hdp_invalidate(adev, NULL);
 
 	memcpy(table, (uint8_t *)priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
@@ -164,7 +164,7 @@ static int smu10_copy_table_to_smc(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[table_id].table, table,
 			priv->smu_tables.entry[table_id].size);
 
-	amdgpu_asic_flush_hdp(adev, NULL);
+	amdgpu_hdp_flush(adev, NULL);
 
 	smum_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
diff --git a/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega10_smumgr.c
index f9c0f117725d..0bf1bf5528c2 100644
--- a/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega10_smumgr.c
+++ b/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega10_smumgr.c
@@ -60,7 +60,7 @@ static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			priv->smu_tables.entry[table_id].table_id,
 			NULL);
 
-	amdgpu_asic_invalidate_hdp(adev, NULL);
+	amdgpu_hdp_invalidate(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
@@ -90,7 +90,7 @@ static int vega10_copy_table_to_smc(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[table_id].table, table,
 			priv->smu_tables.entry[table_id].size);
 
-	amdgpu_asic_flush_hdp(adev, NULL);
+	amdgpu_hdp_flush(adev, NULL);
 
 	smum_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
diff --git a/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega12_smumgr.c b/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega12_smumgr.c
index d3ff6a831ed5..e2ba593faa5d 100644
--- a/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega12_smumgr.c
+++ b/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega12_smumgr.c
@@ -68,7 +68,7 @@ static int vega12_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			"[CopyTableFromSMC] Attempt to Transfer Table From SMU Failed!",
 			return -EINVAL);
 
-	amdgpu_asic_invalidate_hdp(adev, NULL);
+	amdgpu_hdp_invalidate(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
@@ -98,7 +98,7 @@ static int vega12_copy_table_to_smc(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[table_id].table, table,
 			priv->smu_tables.entry[table_id].size);
 
-	amdgpu_asic_flush_hdp(adev, NULL);
+	amdgpu_hdp_flush(adev, NULL);
 
 	PP_ASSERT_WITH_CODE(smum_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
diff --git a/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega20_smumgr.c b/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega20_smumgr.c
index a5c95b180672..e3515156d26f 100644
--- a/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega20_smumgr.c
+++ b/drivers/gpu/drm/amd/pm/powerplay/smumgr/vega20_smumgr.c
@@ -192,7 +192,7 @@ static int vega20_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			"[CopyTableFromSMC] Attempt to Transfer Table From SMU Failed!",
 			return ret);
 
-	amdgpu_asic_invalidate_hdp(adev, NULL);
+	amdgpu_hdp_invalidate(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[table_id].table,
 			priv->smu_tables.entry[table_id].size);
@@ -223,7 +223,7 @@ static int vega20_copy_table_to_smc(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[table_id].table, table,
 			priv->smu_tables.entry[table_id].size);
 
-	amdgpu_asic_flush_hdp(adev, NULL);
+	amdgpu_hdp_flush(adev, NULL);
 
 	PP_ASSERT_WITH_CODE((ret = smum_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
@@ -256,7 +256,7 @@ int vega20_set_activity_monitor_coeff(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[TABLE_ACTIVITY_MONITOR_COEFF].table, table,
 			priv->smu_tables.entry[TABLE_ACTIVITY_MONITOR_COEFF].size);
 
-	amdgpu_asic_flush_hdp(adev, NULL);
+	amdgpu_hdp_flush(adev, NULL);
 
 	PP_ASSERT_WITH_CODE((ret = smum_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
@@ -306,7 +306,7 @@ int vega20_get_activity_monitor_coeff(struct pp_hwmgr *hwmgr,
 			"[GetActivityMonitor] Attempt to Transfer Table From SMU Failed!",
 			return ret);
 
-	amdgpu_asic_invalidate_hdp(adev, NULL);
+	amdgpu_hdp_invalidate(adev, NULL);
 
 	memcpy(table, priv->smu_tables.entry[TABLE_ACTIVITY_MONITOR_COEFF].table,
 			priv->smu_tables.entry[TABLE_ACTIVITY_MONITOR_COEFF].size);
diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
index 4317da6f7c38..f51fa265230b 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
@@ -634,7 +634,7 @@ static int smu_sys_get_pp_table(void *handle,
 		return -EOPNOTSUPP;
 
 	if (!smu_table->power_play_table && !smu_table->hardcode_pptable)
-		return -EINVAL;
+		return -EOPNOTSUPP;
 
 	if (smu_table->hardcode_pptable)
 		*table = smu_table->hardcode_pptable;
@@ -1669,9 +1669,12 @@ static int smu_smc_hw_setup(struct smu_context *smu)
 		if (adev->in_suspend && smu_is_dpm_running(smu)) {
 			dev_info(adev->dev, "dpm has been enabled\n");
 			ret = smu_system_features_control(smu, true);
-			if (ret)
+			if (ret) {
 				dev_err(adev->dev, "Failed system features control!\n");
-			return ret;
+				return ret;
+			}
+
+			return smu_enable_thermal_alert(smu);
 		}
 		break;
 	default:
@@ -2054,6 +2057,12 @@ static int smu_disable_dpms(struct smu_context *smu)
 	    smu->is_apu && (amdgpu_in_reset(adev) || adev->in_s0ix))
 		return 0;
 
+	/* vangogh s0ix */
+	if ((amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(11, 5, 0) ||
+	     amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(11, 5, 2)) &&
+	    adev->in_s0ix)
+		return 0;
+
 	/*
 	 * For gpu reset, runpm and hibernation through BACO,
 	 * BACO feature has to be kept enabled.
@@ -2797,6 +2806,17 @@ const struct amdgpu_ip_block_version smu_v14_0_ip_block = {
 	.funcs = &smu_ip_funcs,
 };
 
+const struct ras_smu_drv *smu_get_ras_smu_driver(void *handle)
+{
+	struct smu_context *smu = (struct smu_context *)handle;
+	const struct ras_smu_drv *tmp = NULL;
+	int ret;
+
+	ret = smu_get_ras_smu_drv(smu, &tmp);
+
+	return ret ? NULL : tmp;
+}
+
 static int smu_load_microcode(void *handle)
 {
 	struct smu_context *smu = handle;
@@ -2890,6 +2910,9 @@ int smu_get_power_limit(void *handle,
 	if (!smu->pm_enabled || !smu->adev->pm.dpm_enabled)
 		return -EOPNOTSUPP;
 
+	if  (!limit)
+		return -EINVAL;
+
 	switch (pp_power_type) {
 	case PP_PWR_TYPE_SUSTAINED:
 		limit_type = SMU_DEFAULT_PPT_LIMIT;
@@ -2921,6 +2944,8 @@ int smu_get_power_limit(void *handle,
 	if (limit_type != SMU_DEFAULT_PPT_LIMIT) {
 		if (smu->ppt_funcs->get_ppt_limit)
 			ret = smu->ppt_funcs->get_ppt_limit(smu, limit, limit_type, limit_level);
+		else
+			return -EOPNOTSUPP;
 	} else {
 		switch (limit_level) {
 		case SMU_PPT_LIMIT_CURRENT:
diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
index c48028abc8c4..8815fc70b63b 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
@@ -1531,6 +1531,12 @@ struct pptable_funcs {
 	int (*ras_send_msg)(struct smu_context *smu,
 			    enum smu_message_type msg, uint32_t param, uint32_t *read_arg);
 
+
+	/**
+	 * @get_ras_smu_drv: Get RAS smu driver interface
+	 * Return: ras_smu_drv *
+	 */
+	int (*get_ras_smu_drv)(struct smu_context *smu, const struct ras_smu_drv **ras_smu_drv);
 };
 
 typedef enum {
@@ -1795,6 +1801,7 @@ int smu_set_pm_policy(struct smu_context *smu, enum pp_pm_policy p_type,
 		      int level);
 ssize_t smu_get_pm_policy_info(struct smu_context *smu,
 			       enum pp_pm_policy p_type, char *sysbuf);
+const struct ras_smu_drv *smu_get_ras_smu_driver(void *handle);
 
 int amdgpu_smu_ras_send_msg(struct amdgpu_device *adev, enum smu_message_type msg,
 			    uint32_t param, uint32_t *readarg);
diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_pmfw.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_pmfw.h
index bf6aa9620911..dd30d96e1ca2 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_pmfw.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_pmfw.h
@@ -87,7 +87,7 @@ typedef enum {
 /*37*/  FEATURE_DVO                         = 37,
 /*38*/  FEATURE_XVMINORPSM_CLKSTOP_DS       = 38,
 /*39*/  FEATURE_GLOBAL_DPM                  = 39,
-/*40*/  FEATURE_NODE_POWER_MANAGER          = 40,
+/*40*/  FEATURE_HROM_EN                     = 40,
 
 /*41*/  NUM_FEATURES                        = 41
 } FEATURE_LIST_e;
@@ -189,7 +189,7 @@ typedef enum {
   SVI_MAX_TEMP_ENTRIES,   // 13
 } SVI_TEMP_e;
 
-#define SMU_METRICS_TABLE_VERSION 0x14
+#define SMU_METRICS_TABLE_VERSION 0x15
 
 #define SMU_SYSTEM_METRICS_TABLE_VERSION 0x1
 
@@ -367,6 +367,11 @@ typedef struct {
 
   //Node Power Limit
   uint32_t MaxNodePowerLimit;
+
+  // PPT1 Configuration
+  uint32_t PPT1Max;
+  uint32_t PPT1Min;
+  uint32_t PPT1Default;
 } StaticMetricsTable_t;
 #pragma pack(pop)
 
diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_ppsmc.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_ppsmc.h
index 4b066c42e0ec..d09b6ae9827e 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_ppsmc.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_12_ppsmc.h
@@ -105,23 +105,21 @@
 #define PPSMC_MSG_UpdatePccWaitDecMaxStr            0x4C
 #define PPSMC_MSG_ResetSDMA                         0x4D
 #define PPSMC_MSG_GetRasTableVersion                0x4E
-#define PPSMC_MSG_GetRmaStatus                      0x4F
-#define PPSMC_MSG_GetErrorCount                     0x50
-#define PPSMC_MSG_GetBadPageCount                   0x51
-#define PPSMC_MSG_GetBadPageInfo                    0x52
-#define PPSMC_MSG_GetBadPagePaAddrLoHi              0x53
-#define PPSMC_MSG_SetTimestampLoHi                  0x54
-#define PPSMC_MSG_GetTimestampLoHi                  0x55
-#define PPSMC_MSG_GetRasPolicy                      0x56
-#define PPSMC_MSG_DumpErrorRecord                   0x57
+#define PPSMC_MSG_GetBadPageCount                   0x50
+#define PPSMC_MSG_GetBadPageMcaAddress              0x51
+#define PPSMC_MSG_SetTimestamp                      0x53
+#define PPSMC_MSG_SetTimestampHi                    0x54
+#define PPSMC_MSG_GetTimestamp                      0x55
+#define PPSMC_MSG_GetBadPageIpIdLoHi                0x57
 #define PPSMC_MSG_EraseRasTable                     0x58
 #define PPSMC_MSG_GetStaticMetricsTable             0x59
 #define PPSMC_MSG_ResetVfArbitersByIndex            0x5A
-#define PPSMC_MSG_GetBadPageSeverity                0x5B
 #define PPSMC_MSG_GetSystemMetricsTable             0x5C
 #define PPSMC_MSG_GetSystemMetricsVersion           0x5D
 #define PPSMC_MSG_ResetVCN                          0x5E
-#define PPSMC_Message_Count                         0x5F
+#define PPSMC_MSG_SetFastPptLimit                   0x5F
+#define PPSMC_MSG_GetFastPptLimit                   0x60
+#define PPSMC_Message_Count                         0x61
 
 //PPSMC Reset Types for driver msg argument
 #define PPSMC_RESET_TYPE_DRIVER_MODE_1_RESET        0x1
diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
index 2256c77da636..9b71a8afdd35 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
@@ -279,7 +279,16 @@
 	__SMU_DUMMY_MAP(ResetSDMA), \
 	__SMU_DUMMY_MAP(ResetVCN), \
 	__SMU_DUMMY_MAP(GetStaticMetricsTable), \
-	__SMU_DUMMY_MAP(GetSystemMetricsTable),
+	__SMU_DUMMY_MAP(GetSystemMetricsTable), \
+	__SMU_DUMMY_MAP(GetRASTableVersion), \
+	__SMU_DUMMY_MAP(GetBadPageCount), \
+	__SMU_DUMMY_MAP(GetBadPageMcaAddr), \
+	__SMU_DUMMY_MAP(SetTimestamp), \
+	__SMU_DUMMY_MAP(GetTimestamp), \
+	__SMU_DUMMY_MAP(GetBadPageIpid), \
+	__SMU_DUMMY_MAP(EraseRasTable),  \
+	__SMU_DUMMY_MAP(SetFastPptLimit), \
+	__SMU_DUMMY_MAP(GetFastPptLimit),
 
 #undef __SMU_DUMMY_MAP
 #define __SMU_DUMMY_MAP(type)	SMU_MSG_##type
@@ -458,7 +467,8 @@ enum smu_clk_type {
        __SMU_DUMMY_MAP(GFX_EDC_XVMIN),				\
        __SMU_DUMMY_MAP(GFX_DIDT_XVMIN),				\
        __SMU_DUMMY_MAP(FAN_ABNORMAL),				\
-       __SMU_DUMMY_MAP(PIT),
+       __SMU_DUMMY_MAP(PIT),				\
+       __SMU_DUMMY_MAP(HROM_EN),
 
 #undef __SMU_DUMMY_MAP
 #define __SMU_DUMMY_MAP(feature)	SMU_FEATURE_##feature##_BIT
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c
index bbf09aec9152..7c9f77124ab2 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c
@@ -2889,7 +2889,7 @@ static int navi10_set_dummy_pstates_table_location(struct smu_context *smu)
 		dummy_table += 0x1000;
 	}
 
-	amdgpu_asic_flush_hdp(smu->adev, NULL);
+	amdgpu_hdp_flush(smu->adev, NULL);
 
 	ret = smu_cmn_send_smc_msg_with_param(smu,
 					      SMU_MSG_SET_DRIVER_DUMMY_TABLE_DRAM_ADDR_HIGH,
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c
index 78e4186d06cc..b0d6487171d7 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c
@@ -1022,7 +1022,12 @@ int smu_v11_0_enable_thermal_alert(struct smu_context *smu)
 
 int smu_v11_0_disable_thermal_alert(struct smu_context *smu)
 {
-	return amdgpu_irq_put(smu->adev, &smu->irq_source, 0);
+	int ret = 0;
+
+	if (smu->smu_table.thermal_controller_type)
+		ret = amdgpu_irq_put(smu->adev, &smu->irq_source, 0);
+
+	return ret;
 }
 
 static uint16_t convert_to_vddc(uint8_t vid)
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
index 53579208cffb..9626da2dba58 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
@@ -2219,6 +2219,9 @@ static int vangogh_post_smu_init(struct smu_context *smu)
 	uint32_t total_cu = adev->gfx.config.max_cu_per_sh *
 		adev->gfx.config.max_sh_per_se * adev->gfx.config.max_shader_engines;
 
+	if (adev->in_s0ix)
+		return 0;
+
 	/* allow message will be sent after enable message on Vangogh*/
 	if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_GFXCLK_BIT) &&
 			(adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) {
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_12_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_12_ppt.c
index cb3fea9e8cf3..9e635f733fbf 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_12_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_12_ppt.c
@@ -34,6 +34,7 @@
 #include "amdgpu_fru_eeprom.h"
 #include <linux/pci.h>
 #include "smu_cmn.h"
+#include "amdgpu_ras.h"
 
 #undef MP1_Public
 #undef smnMP1_FIRMWARE_FLAGS
@@ -58,7 +59,7 @@
 
 #define NUM_JPEG_RINGS_FW	10
 #define NUM_JPEG_RINGS_GPU_METRICS(gpu_metrics) \
-	(ARRAY_SIZE(gpu_metrics->xcp_stats[0].jpeg_busy) / 4)
+	(ARRAY_SIZE(gpu_metrics->jpeg_busy) / 4)
 
 const struct cmn2asic_mapping smu_v13_0_12_feature_mask_map[SMU_FEATURE_COUNT] = {
 	SMU_13_0_12_FEA_MAP(SMU_FEATURE_DATA_CALCULATIONS_BIT, 		FEATURE_DATA_CALCULATION),
@@ -81,6 +82,7 @@ const struct cmn2asic_mapping smu_v13_0_12_feature_mask_map[SMU_FEATURE_COUNT] =
 	SMU_13_0_12_FEA_MAP(SMU_FEATURE_DS_MPIOCLK_BIT,			FEATURE_DS_MPIOCLK),
 	SMU_13_0_12_FEA_MAP(SMU_FEATURE_DS_MP0CLK_BIT,			FEATURE_DS_MP0CLK),
 	SMU_13_0_12_FEA_MAP(SMU_FEATURE_PIT_BIT,			FEATURE_PIT),
+	SMU_13_0_12_FEA_MAP(SMU_FEATURE_HROM_EN_BIT,			FEATURE_HROM_EN),
 };
 
 const struct cmn2asic_msg_mapping smu_v13_0_12_message_map[SMU_MSG_MAX_COUNT] = {
@@ -139,6 +141,15 @@ const struct cmn2asic_msg_mapping smu_v13_0_12_message_map[SMU_MSG_MAX_COUNT] =
 	MSG_MAP(ResetVCN,                            PPSMC_MSG_ResetVCN,                        0),
 	MSG_MAP(GetStaticMetricsTable,               PPSMC_MSG_GetStaticMetricsTable,           1),
 	MSG_MAP(GetSystemMetricsTable,               PPSMC_MSG_GetSystemMetricsTable,           1),
+	MSG_MAP(GetRASTableVersion,                  PPSMC_MSG_GetRasTableVersion,              0),
+	MSG_MAP(GetBadPageCount,                     PPSMC_MSG_GetBadPageCount,                 0),
+	MSG_MAP(GetBadPageMcaAddr,                   PPSMC_MSG_GetBadPageMcaAddress,            0),
+	MSG_MAP(SetTimestamp,                        PPSMC_MSG_SetTimestamp,                    0),
+	MSG_MAP(GetTimestamp,                        PPSMC_MSG_GetTimestamp,                    0),
+	MSG_MAP(GetBadPageIpid,                      PPSMC_MSG_GetBadPageIpIdLoHi,              0),
+	MSG_MAP(EraseRasTable,                       PPSMC_MSG_EraseRasTable,                   0),
+	MSG_MAP(SetFastPptLimit,		     PPSMC_MSG_SetFastPptLimit,			1),
+	MSG_MAP(GetFastPptLimit,		     PPSMC_MSG_GetFastPptLimit,			1),
 };
 
 int smu_v13_0_12_tables_init(struct smu_context *smu)
@@ -345,6 +356,12 @@ int smu_v13_0_12_setup_driver_pptable(struct smu_context *smu)
 		if (smu_v13_0_6_cap_supported(smu, SMU_CAP(NPM_METRICS)))
 			pptable->MaxNodePowerLimit =
 				SMUQ10_ROUND(static_metrics->MaxNodePowerLimit);
+		if (smu_v13_0_6_cap_supported(smu, SMU_CAP(FAST_PPT)) &&
+		    static_metrics->PPT1Max) {
+			pptable->PPT1Max = static_metrics->PPT1Max;
+			pptable->PPT1Min = static_metrics->PPT1Min;
+			pptable->PPT1Default = static_metrics->PPT1Default;
+		}
 		smu_v13_0_12_init_xgmi_data(smu, static_metrics);
 		pptable->Init = true;
 	}
@@ -449,7 +466,7 @@ static int smu_v13_0_12_get_system_metrics_table(struct smu_context *smu)
 		return ret;
 	}
 
-	amdgpu_asic_invalidate_hdp(smu->adev, NULL);
+	amdgpu_hdp_invalidate(smu->adev, NULL);
 	smu_table_cache_update_time(sys_table, jiffies);
 	memcpy(sys_table->cache.buffer, table->cpu_addr,
 	       smu_v13_0_12_get_system_metrics_size());
@@ -719,15 +736,14 @@ static ssize_t smu_v13_0_12_get_temp_metrics(struct smu_context *smu,
 ssize_t smu_v13_0_12_get_xcp_metrics(struct smu_context *smu, struct amdgpu_xcp *xcp, void *table, void *smu_metrics)
 {
 	const u8 num_jpeg_rings = NUM_JPEG_RINGS_FW;
-	struct amdgpu_partition_metrics_v1_0 *xcp_metrics;
+	struct smu_v13_0_6_partition_metrics *xcp_metrics;
 	struct amdgpu_device *adev = smu->adev;
 	MetricsTable_t *metrics;
 	int inst, j, k, idx;
 	u32 inst_mask;
 
 	metrics = (MetricsTable_t *)smu_metrics;
-	xcp_metrics = (struct amdgpu_partition_metrics_v1_0 *) table;
-	smu_cmn_init_partition_metrics(xcp_metrics, 1, 0);
+	xcp_metrics = (struct smu_v13_0_6_partition_metrics *)table;
 	amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_VCN, &inst_mask);
 	idx = 0;
 	for_each_inst(k, inst_mask) {
@@ -772,22 +788,17 @@ ssize_t smu_v13_0_12_get_xcp_metrics(struct smu_context *smu, struct amdgpu_xcp
 	return sizeof(*xcp_metrics);
 }
 
-ssize_t smu_v13_0_12_get_gpu_metrics(struct smu_context *smu, void **table, void *smu_metrics)
+void smu_v13_0_12_get_gpu_metrics(struct smu_context *smu, void **table,
+				  void *smu_metrics,
+				  struct smu_v13_0_6_gpu_metrics *gpu_metrics)
 {
-	struct smu_table_context *smu_table = &smu->smu_table;
-	struct gpu_metrics_v1_8 *gpu_metrics =
-		(struct gpu_metrics_v1_8 *)smu_table->gpu_metrics_table;
-	int ret = 0, xcc_id, inst, i, j, k, idx;
 	struct amdgpu_device *adev = smu->adev;
+	int ret = 0, xcc_id, inst, i, j;
 	u8 num_jpeg_rings_gpu_metrics;
 	MetricsTable_t *metrics;
-	struct amdgpu_xcp *xcp;
-	u32 inst_mask;
 
 	metrics = (MetricsTable_t *)smu_metrics;
 
-	smu_cmn_init_soft_gpu_metrics(gpu_metrics, 1, 8);
-
 	gpu_metrics->temperature_hotspot =
 		SMUQ10_ROUND(metrics->MaxSocketTemperature);
 	/* Individual HBM stack temperature is not reported */
@@ -877,60 +888,186 @@ ssize_t smu_v13_0_12_get_gpu_metrics(struct smu_context *smu, void **table, void
 			gpu_metrics->xgmi_link_status[j] = ret;
 	}
 
-	gpu_metrics->num_partition = adev->xcp_mgr->num_xcps;
-
 	num_jpeg_rings_gpu_metrics = NUM_JPEG_RINGS_GPU_METRICS(gpu_metrics);
-	for_each_xcp(adev->xcp_mgr, xcp, i) {
-		amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_VCN, &inst_mask);
-		idx = 0;
-		for_each_inst(k, inst_mask) {
-			/* Both JPEG and VCN has same instances */
-			inst = GET_INST(VCN, k);
-
-			for (j = 0; j < num_jpeg_rings_gpu_metrics; ++j) {
-				gpu_metrics->xcp_stats[i].jpeg_busy
-					[(idx * num_jpeg_rings_gpu_metrics) + j] =
-					SMUQ10_ROUND(metrics->JpegBusy
-							[(inst * NUM_JPEG_RINGS_FW) + j]);
-			}
-			gpu_metrics->xcp_stats[i].vcn_busy[idx] =
-			       SMUQ10_ROUND(metrics->VcnBusy[inst]);
-			idx++;
+	for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
+		inst = GET_INST(VCN, i);
+
+		for (j = 0; j < num_jpeg_rings_gpu_metrics; ++j) {
+			gpu_metrics->jpeg_busy[(i * num_jpeg_rings_gpu_metrics) +
+					       j] =
+				SMUQ10_ROUND(
+					metrics->JpegBusy[(inst *
+							   NUM_JPEG_RINGS_FW) +
+							  j]);
 		}
+		gpu_metrics->vcn_busy[i] = SMUQ10_ROUND(metrics->VcnBusy[inst]);
+	}
 
-		amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_GFX, &inst_mask);
-		idx = 0;
-		for_each_inst(k, inst_mask) {
-			inst = GET_INST(GC, k);
-			gpu_metrics->xcp_stats[i].gfx_busy_inst[idx] =
-				SMUQ10_ROUND(metrics->GfxBusy[inst]);
-			gpu_metrics->xcp_stats[i].gfx_busy_acc[idx] =
-				SMUQ10_ROUND(metrics->GfxBusyAcc[inst]);
-			if (smu_v13_0_6_cap_supported(smu, SMU_CAP(HST_LIMIT_METRICS))) {
-				gpu_metrics->xcp_stats[i].gfx_below_host_limit_ppt_acc[idx] =
-					SMUQ10_ROUND(metrics->GfxclkBelowHostLimitPptAcc[inst]);
-				gpu_metrics->xcp_stats[i].gfx_below_host_limit_thm_acc[idx] =
-					SMUQ10_ROUND(metrics->GfxclkBelowHostLimitThmAcc[inst]);
-				gpu_metrics->xcp_stats[i].gfx_low_utilization_acc[idx] =
-					SMUQ10_ROUND(metrics->GfxclkLowUtilizationAcc[inst]);
-				gpu_metrics->xcp_stats[i].gfx_below_host_limit_total_acc[idx] =
-					SMUQ10_ROUND(metrics->GfxclkBelowHostLimitTotalAcc[inst]);
-			}
-			idx++;
-		}
+	for (i = 0; i < NUM_XCC(adev->gfx.xcc_mask); ++i) {
+		inst = GET_INST(GC, i);
+		gpu_metrics->gfx_busy_inst[i] =
+			SMUQ10_ROUND(metrics->GfxBusy[inst]);
+		gpu_metrics->gfx_busy_acc[i] =
+			SMUQ10_ROUND(metrics->GfxBusyAcc[inst]);
+		if (smu_v13_0_6_cap_supported(smu,
+					      SMU_CAP(HST_LIMIT_METRICS))) {
+			gpu_metrics
+				->gfx_below_host_limit_ppt_acc[i] = SMUQ10_ROUND(
+				metrics->GfxclkBelowHostLimitPptAcc[inst]);
+			gpu_metrics
+				->gfx_below_host_limit_thm_acc[i] = SMUQ10_ROUND(
+				metrics->GfxclkBelowHostLimitThmAcc[inst]);
+			gpu_metrics->gfx_low_utilization_acc[i] = SMUQ10_ROUND(
+				metrics->GfxclkLowUtilizationAcc[inst]);
+			gpu_metrics->gfx_below_host_limit_total_acc
+				[i] = SMUQ10_ROUND(
+				metrics->GfxclkBelowHostLimitTotalAcc[inst]);
+		};
 	}
 
 	gpu_metrics->xgmi_link_width = metrics->XgmiWidth;
 	gpu_metrics->xgmi_link_speed = metrics->XgmiBitrate;
 
 	gpu_metrics->firmware_timestamp = metrics->Timestamp;
-
-	*table = (void *)gpu_metrics;
-
-	return sizeof(*gpu_metrics);
 }
 
 const struct smu_temp_funcs smu_v13_0_12_temp_funcs = {
 	.temp_metrics_is_supported = smu_v13_0_12_is_temp_metrics_supported,
 	.get_temp_metrics = smu_v13_0_12_get_temp_metrics,
 };
+
+static int smu_v13_0_12_get_ras_table_version(struct amdgpu_device *adev,
+					      uint32_t *table_version)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+
+	return smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetRASTableVersion, 0, table_version);
+}
+
+static int smu_v13_0_12_get_badpage_count(struct amdgpu_device *adev, uint32_t *count,
+					  uint32_t timeout)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+	uint64_t end, now;
+	int ret = 0;
+
+	now = (uint64_t)ktime_to_ms(ktime_get());
+	end = now + timeout;
+	do {
+		ret = smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetBadPageCount, 0, count);
+		/* eeprom is not ready */
+		if (ret != -EBUSY)
+			return ret;
+		mdelay(10);
+		now = (uint64_t)ktime_to_ms(ktime_get());
+	} while (now < end);
+
+	dev_err(adev->dev,
+			 "smu get bad page count timeout!\n");
+	return ret;
+}
+
+static int smu_v13_0_12_set_timestamp(struct amdgpu_device *adev, uint64_t timestamp)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+
+	return smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_SetTimestamp, (uint32_t)timestamp, 0);
+}
+
+static int smu_v13_0_12_get_timestamp(struct amdgpu_device *adev,
+				      uint16_t index, uint64_t *timestamp)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+	uint32_t temp;
+	int ret;
+
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetTimestamp, index, &temp);
+	if (!ret)
+		*timestamp = temp;
+
+	return ret;
+}
+
+static int smu_v13_0_12_get_badpage_ipid(struct amdgpu_device *adev,
+					 uint16_t index, uint64_t *ipid)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+	uint32_t temp_arg, temp_ipid_lo, temp_ipid_high;
+	int ret;
+
+	temp_arg = index | (1 << 16);
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetBadPageIpid, temp_arg, &temp_ipid_lo);
+	if (ret)
+		return ret;
+
+	temp_arg = index | (2 << 16);
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetBadPageIpid, temp_arg, &temp_ipid_high);
+	if (!ret)
+		*ipid = (uint64_t)temp_ipid_high << 32 | temp_ipid_lo;
+	return ret;
+}
+
+static int smu_v13_0_12_erase_ras_table(struct amdgpu_device *adev,
+					uint32_t *result)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+
+	return smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_EraseRasTable, 0, result);
+}
+
+static int smu_v13_0_12_get_badpage_mca_addr(struct amdgpu_device *adev,
+					     uint16_t index, uint64_t *mca_addr)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+	uint32_t temp_arg, temp_addr_lo, temp_addr_high;
+	int ret;
+
+	temp_arg = index | (1 << 16);
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetBadPageMcaAddr, temp_arg, &temp_addr_lo);
+	if (ret)
+		return ret;
+
+	temp_arg = index | (2 << 16);
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+			SMU_MSG_GetBadPageMcaAddr, temp_arg, &temp_addr_high);
+	if (!ret)
+		*mca_addr = (uint64_t)temp_addr_high << 32 | temp_addr_lo;
+	return ret;
+}
+
+static const struct ras_eeprom_smu_funcs smu_v13_0_12_eeprom_smu_funcs = {
+	.get_ras_table_version = smu_v13_0_12_get_ras_table_version,
+	.get_badpage_count = smu_v13_0_12_get_badpage_count,
+	.get_badpage_mca_addr = smu_v13_0_12_get_badpage_mca_addr,
+	.set_timestamp = smu_v13_0_12_set_timestamp,
+	.get_timestamp = smu_v13_0_12_get_timestamp,
+	.get_badpage_ipid = smu_v13_0_12_get_badpage_ipid,
+	.erase_ras_table = smu_v13_0_12_erase_ras_table,
+};
+
+static void smu_v13_0_12_ras_smu_feature_flags(struct amdgpu_device *adev, uint64_t *flags)
+{
+	struct smu_context *smu = adev->powerplay.pp_handle;
+
+	if (!flags)
+		return;
+
+	*flags = 0ULL;
+
+	if (smu_v13_0_6_cap_supported(smu, SMU_CAP(RAS_EEPROM)))
+		*flags |= RAS_SMU_FEATURE_BIT__RAS_EEPROM;
+
+}
+
+const struct ras_smu_drv smu_v13_0_12_ras_smu_drv = {
+	.smu_eeprom_funcs = &smu_v13_0_12_eeprom_smu_funcs,
+	.ras_smu_feature_flags = smu_v13_0_12_ras_smu_feature_flags,
+};
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
index 0a7d2cea7dc6..44e1cd821eec 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
@@ -356,6 +356,9 @@ static void smu_v13_0_12_init_caps(struct smu_context *smu)
 	if (fw_ver > 0x04560900)
 		smu_v13_0_6_cap_set(smu, SMU_CAP(VCN_RESET));
 
+	if (fw_ver >= 0x04560D00)
+		smu_v13_0_6_cap_set(smu, SMU_CAP(FAST_PPT));
+
 	if (fw_ver >= 0x04560700) {
 		if (fw_ver >= 0x04560900) {
 			smu_v13_0_6_cap_set(smu, SMU_CAP(TEMP_METRICS));
@@ -549,7 +552,7 @@ static int smu_v13_0_6_tables_init(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
 	struct smu_table *tables = smu_table->tables;
-	void *gpu_metrics_table __free(kfree) = NULL;
+	struct smu_v13_0_6_gpu_metrics *gpu_metrics;
 	void *driver_pptable __free(kfree) = NULL;
 	void *metrics_table __free(kfree) = NULL;
 	struct amdgpu_device *adev = smu->adev;
@@ -579,24 +582,28 @@ static int smu_v13_0_6_tables_init(struct smu_context *smu)
 		return -ENOMEM;
 	smu_table->metrics_time = 0;
 
-	smu_table->gpu_metrics_table_size = sizeof(struct gpu_metrics_v1_8);
-	gpu_metrics_table =
-		kzalloc(smu_table->gpu_metrics_table_size, GFP_KERNEL);
-	if (!gpu_metrics_table)
-		return -ENOMEM;
-
 	driver_pptable = kzalloc(sizeof(struct PPTable_t), GFP_KERNEL);
 	if (!driver_pptable)
 		return -ENOMEM;
 
+	ret = smu_table_cache_init(smu, SMU_TABLE_SMU_METRICS,
+				   sizeof(struct smu_v13_0_6_gpu_metrics), 1);
+	if (ret)
+		return ret;
+
+	gpu_metrics = (struct smu_v13_0_6_gpu_metrics
+			       *)(tables[SMU_TABLE_SMU_METRICS].cache.buffer);
+
+	smu_v13_0_6_gpu_metrics_init(gpu_metrics, 1, 9);
 	if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) ==
 	    IP_VERSION(13, 0, 12)) {
 		ret = smu_v13_0_12_tables_init(smu);
-		if (ret)
+		if (ret) {
+			smu_table_cache_fini(smu, SMU_TABLE_SMU_METRICS);
 			return ret;
+		}
 	}
 
-	smu_table->gpu_metrics_table = no_free_ptr(gpu_metrics_table);
 	smu_table->metrics_table = no_free_ptr(metrics_table);
 	smu_table->driver_pptable = no_free_ptr(driver_pptable);
 
@@ -732,6 +739,7 @@ static int smu_v13_0_6_fini_smc_tables(struct smu_context *smu)
 {
 	if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 12))
 		smu_v13_0_12_tables_fini(smu);
+	smu_table_cache_fini(smu, SMU_TABLE_SMU_METRICS);
 	return smu_v13_0_fini_smc_tables(smu);
 }
 
@@ -766,7 +774,7 @@ int smu_v13_0_6_get_metrics_table(struct smu_context *smu, void *metrics_table,
 			return ret;
 		}
 
-		amdgpu_asic_invalidate_hdp(smu->adev, NULL);
+		amdgpu_hdp_invalidate(smu->adev, NULL);
 		memcpy(smu_table->metrics_table, table->cpu_addr, table_size);
 
 		smu_table->metrics_time = jiffies;
@@ -845,12 +853,23 @@ int smu_v13_0_6_get_static_metrics_table(struct smu_context *smu)
 		return ret;
 	}
 
-	amdgpu_asic_invalidate_hdp(smu->adev, NULL);
+	amdgpu_hdp_invalidate(smu->adev, NULL);
 	memcpy(smu_table->metrics_table, table->cpu_addr, table_size);
 
 	return 0;
 }
 
+static void smu_v13_0_6_update_caps(struct smu_context *smu)
+{
+	struct smu_table_context *smu_table = &smu->smu_table;
+	struct PPTable_t *pptable =
+		(struct PPTable_t *)smu_table->driver_pptable;
+
+	if (smu_v13_0_6_cap_supported(smu, SMU_CAP(FAST_PPT)) &&
+	    !pptable->PPT1Max)
+		smu_v13_0_6_cap_clear(smu, SMU_CAP(FAST_PPT));
+}
+
 static int smu_v13_0_6_setup_driver_pptable(struct smu_context *smu)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
@@ -867,8 +886,12 @@ static int smu_v13_0_6_setup_driver_pptable(struct smu_context *smu)
 	uint8_t max_width;
 
 	if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 12) &&
-	    smu_v13_0_6_cap_supported(smu, SMU_CAP(STATIC_METRICS)))
-		return smu_v13_0_12_setup_driver_pptable(smu);
+	    smu_v13_0_6_cap_supported(smu, SMU_CAP(STATIC_METRICS))) {
+		ret = smu_v13_0_12_setup_driver_pptable(smu);
+		if (ret)
+			return ret;
+		goto out;
+	}
 
 	/* Store one-time values in driver PPTable */
 	if (!pptable->Init) {
@@ -948,7 +971,8 @@ static int smu_v13_0_6_setup_driver_pptable(struct smu_context *smu)
 			smu_v13_0_6_fill_static_metrics_table(smu, static_metrics);
 		}
 	}
-
+out:
+	smu_v13_0_6_update_caps(smu);
 	return 0;
 }
 
@@ -1394,7 +1418,7 @@ static int smu_v13_0_6_print_clks(struct smu_context *smu, char *buf, int size,
 		return -EINVAL;
 
 	if (curr_clk < SMU_13_0_6_DSCLK_THRESHOLD) {
-		size = sysfs_emit_at(buf, size, "S: %uMhz *\n", curr_clk);
+		size += sysfs_emit_at(buf, size, "S: %uMhz *\n", curr_clk);
 		for (i = 0; i < clocks.num_levels; i++)
 			size += sysfs_emit_at(buf, size, "%d: %uMhz\n", i,
 					      clocks.data[i].clocks_in_khz /
@@ -1514,9 +1538,13 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu,
 
 		single_dpm_table = &(dpm_context->dpm_tables.uclk_table);
 
-		return smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
-					      now, "mclk");
+		ret = smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
+					     now, "mclk");
+		if (ret < 0)
+			return ret;
 
+		size += ret;
+		break;
 	case SMU_SOCCLK:
 		ret = smu_v13_0_6_get_current_clk_freq_by_table(smu, SMU_SOCCLK,
 								&now);
@@ -1528,9 +1556,13 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu,
 
 		single_dpm_table = &(dpm_context->dpm_tables.soc_table);
 
-		return smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
-					      now, "socclk");
+		ret = smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
+					     now, "socclk");
+		if (ret < 0)
+			return ret;
 
+		size += ret;
+		break;
 	case SMU_FCLK:
 		ret = smu_v13_0_6_get_current_clk_freq_by_table(smu, SMU_FCLK,
 								&now);
@@ -1542,9 +1574,13 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu,
 
 		single_dpm_table = &(dpm_context->dpm_tables.fclk_table);
 
-		return smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
-					      now, "fclk");
+		ret = smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
+					     now, "fclk");
+		if (ret < 0)
+			return ret;
 
+		size += ret;
+		break;
 	case SMU_VCLK:
 		ret = smu_v13_0_6_get_current_clk_freq_by_table(smu, SMU_VCLK,
 								&now);
@@ -1556,9 +1592,13 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu,
 
 		single_dpm_table = &(dpm_context->dpm_tables.vclk_table);
 
-		return smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
-					      now, "vclk");
+		ret = smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
+					     now, "vclk");
+		if (ret < 0)
+			return ret;
 
+		size += ret;
+		break;
 	case SMU_DCLK:
 		ret = smu_v13_0_6_get_current_clk_freq_by_table(smu, SMU_DCLK,
 							       &now);
@@ -1570,9 +1610,13 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu,
 
 		single_dpm_table = &(dpm_context->dpm_tables.dclk_table);
 
-		return smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
-					      now, "dclk");
+		ret = smu_v13_0_6_print_clks(smu, buf, size, single_dpm_table,
+					     now, "dclk");
+		if (ret < 0)
+			return ret;
 
+		size += ret;
+		break;
 	default:
 		break;
 	}
@@ -1847,7 +1891,7 @@ static int smu_v13_0_6_get_power_limit(struct smu_context *smu,
 	if (current_power_limit)
 		*current_power_limit = power_limit;
 	if (default_power_limit)
-		*default_power_limit = power_limit;
+		*default_power_limit = pptable->MaxSocketPowerLimit;
 
 	if (max_power_limit) {
 		*max_power_limit = pptable->MaxSocketPowerLimit;
@@ -1862,9 +1906,66 @@ static int smu_v13_0_6_set_power_limit(struct smu_context *smu,
 				       enum smu_ppt_limit_type limit_type,
 				       uint32_t limit)
 {
+	struct smu_table_context *smu_table = &smu->smu_table;
+	struct PPTable_t *pptable =
+		(struct PPTable_t *)smu_table->driver_pptable;
+	int ret;
+
+	if (limit_type == SMU_FAST_PPT_LIMIT) {
+		if (!smu_v13_0_6_cap_supported(smu, SMU_CAP(FAST_PPT)))
+			return -EOPNOTSUPP;
+		if (limit > pptable->PPT1Max || limit < pptable->PPT1Min) {
+			dev_err(smu->adev->dev,
+				"New power limit (%d) should be between min %d max %d\n",
+				limit, pptable->PPT1Min, pptable->PPT1Max);
+			return -EINVAL;
+		}
+		ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetFastPptLimit,
+						      limit, NULL);
+		if (ret)
+			dev_err(smu->adev->dev, "Set fast PPT limit failed!\n");
+		return ret;
+	}
+
 	return smu_v13_0_set_power_limit(smu, limit_type, limit);
 }
 
+static int smu_v13_0_6_get_ppt_limit(struct smu_context *smu,
+				     uint32_t *ppt_limit,
+				     enum smu_ppt_limit_type type,
+				     enum smu_ppt_limit_level level)
+{
+	struct smu_table_context *smu_table = &smu->smu_table;
+	struct PPTable_t *pptable =
+		(struct PPTable_t *)smu_table->driver_pptable;
+	int ret = 0;
+
+	if (type == SMU_FAST_PPT_LIMIT) {
+		if (!smu_v13_0_6_cap_supported(smu, SMU_CAP(FAST_PPT)))
+			return -EOPNOTSUPP;
+		switch (level) {
+		case SMU_PPT_LIMIT_MAX:
+			*ppt_limit = pptable->PPT1Max;
+			break;
+		case SMU_PPT_LIMIT_CURRENT:
+			ret = smu_cmn_send_smc_msg(smu, SMU_MSG_GetFastPptLimit, ppt_limit);
+			if (ret)
+				dev_err(smu->adev->dev, "Get fast PPT limit failed!\n");
+			break;
+		case SMU_PPT_LIMIT_DEFAULT:
+			*ppt_limit = pptable->PPT1Default;
+			break;
+		case SMU_PPT_LIMIT_MIN:
+			*ppt_limit = pptable->PPT1Min;
+			break;
+		default:
+			return -EOPNOTSUPP;
+		}
+		return ret;
+	}
+	return -EOPNOTSUPP;
+}
+
 static int smu_v13_0_6_irq_process(struct amdgpu_device *adev,
 				   struct amdgpu_irq_src *source,
 				   struct amdgpu_iv_entry *entry)
@@ -2385,7 +2486,7 @@ static int smu_v13_0_6_request_i2c_xfer(struct smu_context *smu,
 
 	memcpy(table->cpu_addr, table_data, table_size);
 	/* Flush hdp cache */
-	amdgpu_asic_flush_hdp(adev, NULL);
+	amdgpu_hdp_flush(adev, NULL);
 	ret = smu_cmn_send_smc_msg(smu, SMU_MSG_RequestI2cTransaction,
 					  NULL);
 
@@ -2629,7 +2730,7 @@ static ssize_t smu_v13_0_6_get_xcp_metrics(struct smu_context *smu, int xcp_id,
 {
 	const u8 num_jpeg_rings = AMDGPU_MAX_JPEG_RINGS_4_0_3;
 	int version = smu_v13_0_6_get_metrics_version(smu);
-	struct amdgpu_partition_metrics_v1_0 *xcp_metrics;
+	struct smu_v13_0_6_partition_metrics *xcp_metrics;
 	MetricsTableV0_t *metrics_v0 __free(kfree) = NULL;
 	struct amdgpu_device *adev = smu->adev;
 	int ret, inst, i, j, k, idx;
@@ -2649,8 +2750,8 @@ static ssize_t smu_v13_0_6_get_xcp_metrics(struct smu_context *smu, int xcp_id,
 	if (i == adev->xcp_mgr->num_xcps)
 		return -EINVAL;
 
-	xcp_metrics = (struct amdgpu_partition_metrics_v1_0 *)table;
-	smu_cmn_init_partition_metrics(xcp_metrics, 1, 0);
+	xcp_metrics = (struct smu_v13_0_6_partition_metrics *)table;
+	smu_v13_0_6_partition_metrics_init(xcp_metrics, 1, 1);
 
 	metrics_v0 = kzalloc(METRICS_TABLE_SIZE, GFP_KERNEL);
 	if (!metrics_v0)
@@ -2742,18 +2843,16 @@ static ssize_t smu_v13_0_6_get_xcp_metrics(struct smu_context *smu, int xcp_id,
 static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table)
 {
 	struct smu_table_context *smu_table = &smu->smu_table;
-	struct gpu_metrics_v1_8 *gpu_metrics =
-		(struct gpu_metrics_v1_8 *)smu_table->gpu_metrics_table;
+	struct smu_table *tables = smu_table->tables;
+	struct smu_v13_0_6_gpu_metrics *gpu_metrics;
 	int version = smu_v13_0_6_get_metrics_version(smu);
 	MetricsTableV0_t *metrics_v0 __free(kfree) = NULL;
-	int ret = 0, xcc_id, inst, i, j, k, idx;
 	struct amdgpu_device *adev = smu->adev;
+	int ret = 0, xcc_id, inst, i, j;
 	MetricsTableV1_t *metrics_v1;
 	MetricsTableV2_t *metrics_v2;
-	struct amdgpu_xcp *xcp;
 	u16 link_width_level;
 	u8 num_jpeg_rings;
-	u32 inst_mask;
 	bool per_inst;
 
 	metrics_v0 = kzalloc(METRICS_TABLE_SIZE, GFP_KERNEL);
@@ -2761,16 +2860,20 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table
 	if (ret)
 		return ret;
 
-	if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) ==
-		    IP_VERSION(13, 0, 12) &&
-	    smu_v13_0_6_cap_supported(smu, SMU_CAP(STATIC_METRICS)))
-		return smu_v13_0_12_get_gpu_metrics(smu, table, metrics_v0);
+	metrics_v2 = (MetricsTableV2_t *)metrics_v0;
+	gpu_metrics = (struct smu_v13_0_6_gpu_metrics
+			       *)(tables[SMU_TABLE_SMU_METRICS].cache.buffer);
+
+	if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 12) &&
+	    smu_v13_0_6_cap_supported(smu, SMU_CAP(STATIC_METRICS))) {
+		smu_v13_0_12_get_gpu_metrics(smu, table, metrics_v0,
+					     gpu_metrics);
+		goto fill;
+	}
 
 	metrics_v1 = (MetricsTableV1_t *)metrics_v0;
 	metrics_v2 = (MetricsTableV2_t *)metrics_v0;
 
-	smu_cmn_init_soft_gpu_metrics(gpu_metrics, 1, 8);
-
 	gpu_metrics->temperature_hotspot =
 		SMUQ10_ROUND(GET_METRIC_FIELD(MaxSocketTemperature, version));
 	/* Individual HBM stack temperature is not reported */
@@ -2891,55 +2994,49 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table
 			gpu_metrics->xgmi_link_status[j] = ret;
 	}
 
-	gpu_metrics->num_partition = adev->xcp_mgr->num_xcps;
-
 	per_inst = smu_v13_0_6_cap_supported(smu, SMU_CAP(PER_INST_METRICS));
 
 	num_jpeg_rings = AMDGPU_MAX_JPEG_RINGS_4_0_3;
-	for_each_xcp(adev->xcp_mgr, xcp, i) {
-		amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_VCN, &inst_mask);
-		idx = 0;
-		for_each_inst(k, inst_mask) {
-			/* Both JPEG and VCN has same instances */
-			inst = GET_INST(VCN, k);
-
-			for (j = 0; j < num_jpeg_rings; ++j) {
-				gpu_metrics->xcp_stats[i].jpeg_busy
-					[(idx * num_jpeg_rings) + j] =
-					SMUQ10_ROUND(GET_METRIC_FIELD(JpegBusy, version)
-							[(inst * num_jpeg_rings) + j]);
-			}
-			gpu_metrics->xcp_stats[i].vcn_busy[idx] =
-			       SMUQ10_ROUND(GET_METRIC_FIELD(VcnBusy, version)[inst]);
-			idx++;
-
-		}
+	for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
+		inst = GET_INST(JPEG, i);
+		for (j = 0; j < num_jpeg_rings; ++j)
+			gpu_metrics->jpeg_busy[(i * num_jpeg_rings) + j] =
+				SMUQ10_ROUND(GET_METRIC_FIELD(
+					JpegBusy,
+					version)[(inst * num_jpeg_rings) + j]);
+	}
+	for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
+		inst = GET_INST(VCN, i);
+		gpu_metrics->vcn_busy[i] =
+			SMUQ10_ROUND(GET_METRIC_FIELD(VcnBusy, version)[inst]);
+	}
 
-		if (per_inst) {
-			amdgpu_xcp_get_inst_details(xcp, AMDGPU_XCP_GFX, &inst_mask);
-			idx = 0;
-			for_each_inst(k, inst_mask) {
-				inst = GET_INST(GC, k);
-				gpu_metrics->xcp_stats[i].gfx_busy_inst[idx] =
-					SMUQ10_ROUND(GET_GPU_METRIC_FIELD(GfxBusy, version)[inst]);
-				gpu_metrics->xcp_stats[i].gfx_busy_acc[idx] =
-					SMUQ10_ROUND(GET_GPU_METRIC_FIELD(GfxBusyAcc,
-									  version)[inst]);
-				if (smu_v13_0_6_cap_supported(smu, SMU_CAP(HST_LIMIT_METRICS))) {
-					gpu_metrics->xcp_stats[i].gfx_below_host_limit_ppt_acc[idx] =
-						SMUQ10_ROUND
-						(metrics_v0->GfxclkBelowHostLimitPptAcc[inst]);
-					gpu_metrics->xcp_stats[i].gfx_below_host_limit_thm_acc[idx] =
-						SMUQ10_ROUND
-						(metrics_v0->GfxclkBelowHostLimitThmAcc[inst]);
-					gpu_metrics->xcp_stats[i].gfx_low_utilization_acc[idx] =
-						SMUQ10_ROUND
-						(metrics_v0->GfxclkLowUtilizationAcc[inst]);
-					gpu_metrics->xcp_stats[i].gfx_below_host_limit_total_acc[idx] =
-						SMUQ10_ROUND
-						(metrics_v0->GfxclkBelowHostLimitTotalAcc[inst]);
-				}
-				idx++;
+	if (per_inst) {
+		for (i = 0; i < NUM_XCC(adev->gfx.xcc_mask); ++i) {
+			inst = GET_INST(GC, i);
+			gpu_metrics->gfx_busy_inst[i] = SMUQ10_ROUND(
+				GET_GPU_METRIC_FIELD(GfxBusy, version)[inst]);
+			gpu_metrics->gfx_busy_acc[i] = SMUQ10_ROUND(
+				GET_GPU_METRIC_FIELD(GfxBusyAcc,
+						     version)[inst]);
+			if (smu_v13_0_6_cap_supported(
+				    smu, SMU_CAP(HST_LIMIT_METRICS))) {
+				gpu_metrics->gfx_below_host_limit_ppt_acc
+					[i] = SMUQ10_ROUND(
+					metrics_v0->GfxclkBelowHostLimitPptAcc
+						[inst]);
+				gpu_metrics->gfx_below_host_limit_thm_acc
+					[i] = SMUQ10_ROUND(
+					metrics_v0->GfxclkBelowHostLimitThmAcc
+						[inst]);
+				gpu_metrics->gfx_low_utilization_acc
+					[i] = SMUQ10_ROUND(
+					metrics_v0
+						->GfxclkLowUtilizationAcc[inst]);
+				gpu_metrics->gfx_below_host_limit_total_acc
+					[i] = SMUQ10_ROUND(
+					metrics_v0->GfxclkBelowHostLimitTotalAcc
+						[inst]);
 			}
 		}
 	}
@@ -2949,7 +3046,8 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table
 
 	gpu_metrics->firmware_timestamp = GET_METRIC_FIELD(Timestamp, version);
 
-	*table = (void *)gpu_metrics;
+fill:
+	*table = tables[SMU_TABLE_SMU_METRICS].cache.buffer;
 
 	return sizeof(*gpu_metrics);
 }
@@ -3883,6 +3981,29 @@ static void smu_v13_0_6_set_temp_funcs(struct smu_context *smu)
 			== IP_VERSION(13, 0, 12)) ? &smu_v13_0_12_temp_funcs : NULL;
 }
 
+static int smu_v13_0_6_get_ras_smu_drv(struct smu_context *smu, const struct ras_smu_drv **ras_smu_drv)
+{
+	if (!ras_smu_drv)
+		return -EINVAL;
+
+	if (amdgpu_sriov_vf(smu->adev))
+		return -EOPNOTSUPP;
+
+	if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_HROM_EN_BIT))
+		smu_v13_0_6_cap_set(smu, SMU_CAP(RAS_EEPROM));
+
+	switch (amdgpu_ip_version(smu->adev, MP1_HWIP, 0)) {
+	case IP_VERSION(13, 0, 12):
+		*ras_smu_drv = &smu_v13_0_12_ras_smu_drv;
+		break;
+	default:
+		*ras_smu_drv = NULL;
+		break;
+	}
+
+	return 0;
+}
+
 static const struct pptable_funcs smu_v13_0_6_ppt_funcs = {
 	/* init dpm */
 	.get_allowed_feature_mask = smu_v13_0_6_get_allowed_feature_mask,
@@ -3914,6 +4035,7 @@ static const struct pptable_funcs smu_v13_0_6_ppt_funcs = {
 	.get_enabled_mask = smu_v13_0_6_get_enabled_mask,
 	.feature_is_enabled = smu_cmn_feature_is_enabled,
 	.set_power_limit = smu_v13_0_6_set_power_limit,
+	.get_ppt_limit = smu_v13_0_6_get_ppt_limit,
 	.set_xgmi_pstate = smu_v13_0_set_xgmi_pstate,
 	.register_irq_handler = smu_v13_0_6_register_irq_handler,
 	.enable_thermal_alert = smu_v13_0_enable_thermal_alert,
@@ -3942,6 +4064,7 @@ static const struct pptable_funcs smu_v13_0_6_ppt_funcs = {
 	.dpm_reset_vcn = smu_v13_0_6_reset_vcn,
 	.post_init = smu_v13_0_6_post_init,
 	.ras_send_msg = smu_v13_0_6_ras_send_msg,
+	.get_ras_smu_drv = smu_v13_0_6_get_ras_smu_drv,
 };
 
 void smu_v13_0_6_set_ppt_funcs(struct smu_context *smu)
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.h b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.h
index 7ef5f3e66c27..6cbdd7c5ded9 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.h
@@ -50,6 +50,9 @@ struct PPTable_t {
 	uint32_t MinLclkDpmRange;
 	uint64_t PublicSerialNumber_AID;
 	uint32_t MaxNodePowerLimit;
+	uint32_t PPT1Max;
+	uint32_t PPT1Min;
+	uint32_t PPT1Default;
 	bool Init;
 };
 
@@ -72,9 +75,18 @@ enum smu_v13_0_6_caps {
 	SMU_CAP(PLDM_VERSION),
 	SMU_CAP(TEMP_METRICS),
 	SMU_CAP(NPM_METRICS),
+	SMU_CAP(RAS_EEPROM),
+	SMU_CAP(FAST_PPT),
 	SMU_CAP(ALL),
 };
 
+#define SMU_13_0_6_NUM_XGMI_LINKS 8
+#define SMU_13_0_6_MAX_GFX_CLKS 8
+#define SMU_13_0_6_MAX_CLKS 4
+#define SMU_13_0_6_MAX_XCC 8
+#define SMU_13_0_6_MAX_VCN 4
+#define SMU_13_0_6_MAX_JPEG 40
+
 extern void smu_v13_0_6_set_ppt_funcs(struct smu_context *smu);
 bool smu_v13_0_6_cap_supported(struct smu_context *smu, enum smu_v13_0_6_caps cap);
 int smu_v13_0_6_get_static_metrics_table(struct smu_context *smu);
@@ -87,7 +99,6 @@ size_t smu_v13_0_12_get_system_metrics_size(void);
 int smu_v13_0_12_setup_driver_pptable(struct smu_context *smu);
 int smu_v13_0_12_get_smu_metrics_data(struct smu_context *smu,
 				      MetricsMember_t member, uint32_t *value);
-ssize_t smu_v13_0_12_get_gpu_metrics(struct smu_context *smu, void **table, void *smu_metrics);
 ssize_t smu_v13_0_12_get_xcp_metrics(struct smu_context *smu,
 				     struct amdgpu_xcp *xcp, void *table,
 				     void *smu_metrics);
@@ -99,4 +110,156 @@ int smu_v13_0_12_get_npm_data(struct smu_context *smu,
 extern const struct cmn2asic_mapping smu_v13_0_12_feature_mask_map[];
 extern const struct cmn2asic_msg_mapping smu_v13_0_12_message_map[];
 extern const struct smu_temp_funcs smu_v13_0_12_temp_funcs;
+extern const struct ras_smu_drv smu_v13_0_12_ras_smu_drv;
+
+#if defined(SWSMU_CODE_LAYER_L2)
+#include "smu_cmn.h"
+
+/* SMUv 13.0.6 GPU metrics*/
+#define SMU_13_0_6_METRICS_FIELDS(SMU_SCALAR, SMU_ARRAY)                       \
+	SMU_SCALAR(SMU_MATTR(TEMPERATURE_HOTSPOT), SMU_MUNIT(TEMP_1),          \
+		   SMU_MTYPE(U16), temperature_hotspot);                       \
+	SMU_SCALAR(SMU_MATTR(TEMPERATURE_MEM), SMU_MUNIT(TEMP_1),              \
+		   SMU_MTYPE(U16), temperature_mem);                           \
+	SMU_SCALAR(SMU_MATTR(TEMPERATURE_VRSOC), SMU_MUNIT(TEMP_1),            \
+		   SMU_MTYPE(U16), temperature_vrsoc);                         \
+	SMU_SCALAR(SMU_MATTR(CURR_SOCKET_POWER), SMU_MUNIT(POWER_1),           \
+		   SMU_MTYPE(U16), curr_socket_power);                         \
+	SMU_SCALAR(SMU_MATTR(AVERAGE_GFX_ACTIVITY), SMU_MUNIT(PERCENT),        \
+		   SMU_MTYPE(U16), average_gfx_activity);                      \
+	SMU_SCALAR(SMU_MATTR(AVERAGE_UMC_ACTIVITY), SMU_MUNIT(PERCENT),        \
+		   SMU_MTYPE(U16), average_umc_activity);                      \
+	SMU_SCALAR(SMU_MATTR(MEM_MAX_BANDWIDTH), SMU_MUNIT(BW_1),              \
+		   SMU_MTYPE(U64), mem_max_bandwidth);                         \
+	SMU_SCALAR(SMU_MATTR(ENERGY_ACCUMULATOR), SMU_MUNIT(NONE),             \
+		   SMU_MTYPE(U64), energy_accumulator);                        \
+	SMU_SCALAR(SMU_MATTR(SYSTEM_CLOCK_COUNTER), SMU_MUNIT(TIME_1),         \
+		   SMU_MTYPE(U64), system_clock_counter);                      \
+	SMU_SCALAR(SMU_MATTR(ACCUMULATION_COUNTER), SMU_MUNIT(NONE),           \
+		   SMU_MTYPE(U32), accumulation_counter);                      \
+	SMU_SCALAR(SMU_MATTR(PROCHOT_RESIDENCY_ACC), SMU_MUNIT(NONE),          \
+		   SMU_MTYPE(U32), prochot_residency_acc);                     \
+	SMU_SCALAR(SMU_MATTR(PPT_RESIDENCY_ACC), SMU_MUNIT(NONE),              \
+		   SMU_MTYPE(U32), ppt_residency_acc);                         \
+	SMU_SCALAR(SMU_MATTR(SOCKET_THM_RESIDENCY_ACC), SMU_MUNIT(NONE),       \
+		   SMU_MTYPE(U32), socket_thm_residency_acc);                  \
+	SMU_SCALAR(SMU_MATTR(VR_THM_RESIDENCY_ACC), SMU_MUNIT(NONE),           \
+		   SMU_MTYPE(U32), vr_thm_residency_acc);                      \
+	SMU_SCALAR(SMU_MATTR(HBM_THM_RESIDENCY_ACC), SMU_MUNIT(NONE),          \
+		   SMU_MTYPE(U32), hbm_thm_residency_acc);                     \
+	SMU_SCALAR(SMU_MATTR(GFXCLK_LOCK_STATUS), SMU_MUNIT(NONE),             \
+		   SMU_MTYPE(U32), gfxclk_lock_status);                        \
+	SMU_SCALAR(SMU_MATTR(PCIE_LINK_WIDTH), SMU_MUNIT(NONE),                \
+		   SMU_MTYPE(U16), pcie_link_width);                           \
+	SMU_SCALAR(SMU_MATTR(PCIE_LINK_SPEED), SMU_MUNIT(SPEED_2),             \
+		   SMU_MTYPE(U16), pcie_link_speed);                           \
+	SMU_SCALAR(SMU_MATTR(XGMI_LINK_WIDTH), SMU_MUNIT(NONE),                \
+		   SMU_MTYPE(U16), xgmi_link_width);                           \
+	SMU_SCALAR(SMU_MATTR(XGMI_LINK_SPEED), SMU_MUNIT(SPEED_1),             \
+		   SMU_MTYPE(U16), xgmi_link_speed);                           \
+	SMU_SCALAR(SMU_MATTR(GFX_ACTIVITY_ACC), SMU_MUNIT(PERCENT),            \
+		   SMU_MTYPE(U32), gfx_activity_acc);                          \
+	SMU_SCALAR(SMU_MATTR(MEM_ACTIVITY_ACC), SMU_MUNIT(PERCENT),            \
+		   SMU_MTYPE(U32), mem_activity_acc);                          \
+	SMU_SCALAR(SMU_MATTR(PCIE_BANDWIDTH_ACC), SMU_MUNIT(PERCENT),          \
+		   SMU_MTYPE(U64), pcie_bandwidth_acc);                        \
+	SMU_SCALAR(SMU_MATTR(PCIE_BANDWIDTH_INST), SMU_MUNIT(BW_1),            \
+		   SMU_MTYPE(U64), pcie_bandwidth_inst);                       \
+	SMU_SCALAR(SMU_MATTR(PCIE_L0_TO_RECOV_COUNT_ACC), SMU_MUNIT(NONE),     \
+		   SMU_MTYPE(U64), pcie_l0_to_recov_count_acc);                \
+	SMU_SCALAR(SMU_MATTR(PCIE_REPLAY_COUNT_ACC), SMU_MUNIT(NONE),          \
+		   SMU_MTYPE(U64), pcie_replay_count_acc);                     \
+	SMU_SCALAR(SMU_MATTR(PCIE_REPLAY_ROVER_COUNT_ACC), SMU_MUNIT(NONE),    \
+		   SMU_MTYPE(U64), pcie_replay_rover_count_acc);               \
+	SMU_SCALAR(SMU_MATTR(PCIE_NAK_SENT_COUNT_ACC), SMU_MUNIT(NONE),        \
+		   SMU_MTYPE(U32), pcie_nak_sent_count_acc);                   \
+	SMU_SCALAR(SMU_MATTR(PCIE_NAK_RCVD_COUNT_ACC), SMU_MUNIT(NONE),        \
+		   SMU_MTYPE(U32), pcie_nak_rcvd_count_acc);                   \
+	SMU_ARRAY(SMU_MATTR(XGMI_READ_DATA_ACC), SMU_MUNIT(DATA_1),            \
+		  SMU_MTYPE(U64), xgmi_read_data_acc,                          \
+		  SMU_13_0_6_NUM_XGMI_LINKS);                                  \
+	SMU_ARRAY(SMU_MATTR(XGMI_WRITE_DATA_ACC), SMU_MUNIT(DATA_1),           \
+		  SMU_MTYPE(U64), xgmi_write_data_acc,                         \
+		  SMU_13_0_6_NUM_XGMI_LINKS);                                  \
+	SMU_ARRAY(SMU_MATTR(XGMI_LINK_STATUS), SMU_MUNIT(NONE),                \
+		  SMU_MTYPE(U16), xgmi_link_status,                            \
+		  SMU_13_0_6_NUM_XGMI_LINKS);                                  \
+	SMU_SCALAR(SMU_MATTR(FIRMWARE_TIMESTAMP), SMU_MUNIT(TIME_2),           \
+		   SMU_MTYPE(U64), firmware_timestamp);                        \
+	SMU_ARRAY(SMU_MATTR(CURRENT_GFXCLK), SMU_MUNIT(CLOCK_1),               \
+		  SMU_MTYPE(U16), current_gfxclk, SMU_13_0_6_MAX_GFX_CLKS);    \
+	SMU_ARRAY(SMU_MATTR(CURRENT_SOCCLK), SMU_MUNIT(CLOCK_1),               \
+		  SMU_MTYPE(U16), current_socclk, SMU_13_0_6_MAX_CLKS);        \
+	SMU_ARRAY(SMU_MATTR(CURRENT_VCLK0), SMU_MUNIT(CLOCK_1),                \
+		  SMU_MTYPE(U16), current_vclk0, SMU_13_0_6_MAX_CLKS);         \
+	SMU_ARRAY(SMU_MATTR(CURRENT_DCLK0), SMU_MUNIT(CLOCK_1),                \
+		  SMU_MTYPE(U16), current_dclk0, SMU_13_0_6_MAX_CLKS);         \
+	SMU_SCALAR(SMU_MATTR(CURRENT_UCLK), SMU_MUNIT(CLOCK_1),                \
+		   SMU_MTYPE(U16), current_uclk);                              \
+	SMU_SCALAR(SMU_MATTR(PCIE_LC_PERF_OTHER_END_RECOVERY),                 \
+		   SMU_MUNIT(NONE), SMU_MTYPE(U32),                            \
+		   pcie_lc_perf_other_end_recovery);                           \
+	SMU_ARRAY(SMU_MATTR(GFX_BUSY_INST), SMU_MUNIT(PERCENT),                \
+		  SMU_MTYPE(U32), gfx_busy_inst, SMU_13_0_6_MAX_XCC);          \
+	SMU_ARRAY(SMU_MATTR(JPEG_BUSY), SMU_MUNIT(PERCENT), SMU_MTYPE(U16),    \
+		  jpeg_busy, SMU_13_0_6_MAX_JPEG);                             \
+	SMU_ARRAY(SMU_MATTR(VCN_BUSY), SMU_MUNIT(PERCENT), SMU_MTYPE(U16),     \
+		  vcn_busy, SMU_13_0_6_MAX_VCN);                               \
+	SMU_ARRAY(SMU_MATTR(GFX_BUSY_ACC), SMU_MUNIT(PERCENT), SMU_MTYPE(U64), \
+		  gfx_busy_acc, SMU_13_0_6_MAX_XCC);                           \
+	SMU_ARRAY(SMU_MATTR(GFX_BELOW_HOST_LIMIT_PPT_ACC), SMU_MUNIT(NONE),    \
+		  SMU_MTYPE(U64), gfx_below_host_limit_ppt_acc,                \
+		  SMU_13_0_6_MAX_XCC);                                         \
+	SMU_ARRAY(SMU_MATTR(GFX_BELOW_HOST_LIMIT_THM_ACC), SMU_MUNIT(NONE),    \
+		  SMU_MTYPE(U64), gfx_below_host_limit_thm_acc,                \
+		  SMU_13_0_6_MAX_XCC);                                         \
+	SMU_ARRAY(SMU_MATTR(GFX_LOW_UTILIZATION_ACC), SMU_MUNIT(NONE),         \
+		  SMU_MTYPE(U64), gfx_low_utilization_acc,                     \
+		  SMU_13_0_6_MAX_XCC);                                         \
+	SMU_ARRAY(SMU_MATTR(GFX_BELOW_HOST_LIMIT_TOTAL_ACC), SMU_MUNIT(NONE),  \
+		  SMU_MTYPE(U64), gfx_below_host_limit_total_acc,              \
+		  SMU_13_0_6_MAX_XCC);
+
+DECLARE_SMU_METRICS_CLASS(smu_v13_0_6_gpu_metrics, SMU_13_0_6_METRICS_FIELDS);
+void smu_v13_0_12_get_gpu_metrics(struct smu_context *smu, void **table,
+				  void *smu_metrics,
+				  struct smu_v13_0_6_gpu_metrics *gpu_metrics);
+
+#define SMU_13_0_6_PARTITION_METRICS_FIELDS(SMU_SCALAR, SMU_ARRAY)             \
+	SMU_ARRAY(SMU_MATTR(CURRENT_GFXCLK), SMU_MUNIT(CLOCK_1),               \
+		  SMU_MTYPE(U16), current_gfxclk, SMU_13_0_6_MAX_XCC);         \
+	SMU_ARRAY(SMU_MATTR(CURRENT_SOCCLK), SMU_MUNIT(CLOCK_1),               \
+		  SMU_MTYPE(U16), current_socclk, SMU_13_0_6_MAX_CLKS);        \
+	SMU_ARRAY(SMU_MATTR(CURRENT_VCLK0), SMU_MUNIT(CLOCK_1),                \
+		  SMU_MTYPE(U16), current_vclk0, SMU_13_0_6_MAX_CLKS);         \
+	SMU_ARRAY(SMU_MATTR(CURRENT_DCLK0), SMU_MUNIT(CLOCK_1),                \
+		  SMU_MTYPE(U16), current_dclk0, SMU_13_0_6_MAX_CLKS);         \
+	SMU_SCALAR(SMU_MATTR(CURRENT_UCLK), SMU_MUNIT(CLOCK_1),                \
+		   SMU_MTYPE(U16), current_uclk);                              \
+	SMU_ARRAY(SMU_MATTR(GFX_BUSY_INST), SMU_MUNIT(PERCENT),                \
+		  SMU_MTYPE(U32), gfx_busy_inst, SMU_13_0_6_MAX_XCC);          \
+	SMU_ARRAY(SMU_MATTR(JPEG_BUSY), SMU_MUNIT(PERCENT), SMU_MTYPE(U16),    \
+		  jpeg_busy, SMU_13_0_6_MAX_JPEG);                             \
+	SMU_ARRAY(SMU_MATTR(VCN_BUSY), SMU_MUNIT(PERCENT), SMU_MTYPE(U16),     \
+		  vcn_busy, SMU_13_0_6_MAX_VCN);                               \
+	SMU_ARRAY(SMU_MATTR(GFX_BUSY_ACC), SMU_MUNIT(PERCENT), SMU_MTYPE(U64), \
+		  gfx_busy_acc, SMU_13_0_6_MAX_XCC);                           \
+	SMU_ARRAY(SMU_MATTR(GFX_BELOW_HOST_LIMIT_PPT_ACC), SMU_MUNIT(NONE),    \
+		  SMU_MTYPE(U64), gfx_below_host_limit_ppt_acc,                \
+		  SMU_13_0_6_MAX_XCC);                                         \
+	SMU_ARRAY(SMU_MATTR(GFX_BELOW_HOST_LIMIT_THM_ACC), SMU_MUNIT(NONE),    \
+		  SMU_MTYPE(U64), gfx_below_host_limit_thm_acc,                \
+		  SMU_13_0_6_MAX_XCC);                                         \
+	SMU_ARRAY(SMU_MATTR(GFX_LOW_UTILIZATION_ACC), SMU_MUNIT(NONE),         \
+		  SMU_MTYPE(U64), gfx_low_utilization_acc,                     \
+		  SMU_13_0_6_MAX_XCC);                                         \
+	SMU_ARRAY(SMU_MATTR(GFX_BELOW_HOST_LIMIT_TOTAL_ACC), SMU_MUNIT(NONE),  \
+		  SMU_MTYPE(U64), gfx_below_host_limit_total_acc,              \
+		  SMU_13_0_6_MAX_XCC);
+
+DECLARE_SMU_METRICS_CLASS(smu_v13_0_6_partition_metrics,
+			  SMU_13_0_6_PARTITION_METRICS_FIELDS);
+
+#endif /* SWSMU_CODE_LAYER_L2 */
+
 #endif
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c
index a8961a8f5c42..4040ff926544 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c
@@ -164,9 +164,13 @@ static void __smu_cmn_reg_print_error(struct smu_context *smu,
 				    msg_index, param, message);
 		break;
 	case SMU_RESP_BUSY_OTHER:
-		dev_err_ratelimited(adev->dev,
-				    "SMU: I'm very busy for your command: index:%d param:0x%08X message:%s",
-				    msg_index, param, message);
+		/* It is normal for SMU_MSG_GetBadPageCount to return busy
+		 * so don't print error at this case.
+		 */
+		if (msg != SMU_MSG_GetBadPageCount)
+			dev_err_ratelimited(adev->dev,
+						"SMU: I'm very busy for your command: index:%d param:0x%08X message:%s",
+						msg_index, param, message);
 		break;
 	case SMU_RESP_DEBUG_END:
 		dev_err_ratelimited(adev->dev,
@@ -980,7 +984,7 @@ int smu_cmn_update_table(struct smu_context *smu,
 		 * Flush hdp cache: to guard the content seen by
 		 * GPU is consitent with CPU.
 		 */
-		amdgpu_asic_flush_hdp(adev, NULL);
+		amdgpu_hdp_flush(adev, NULL);
 	}
 
 	ret = smu_cmn_send_smc_msg_with_param(smu, drv2smu ?
@@ -992,7 +996,7 @@ int smu_cmn_update_table(struct smu_context *smu,
 		return ret;
 
 	if (!drv2smu) {
-		amdgpu_asic_invalidate_hdp(adev, NULL);
+		amdgpu_hdp_invalidate(adev, NULL);
 		memcpy(table_data, table->cpu_addr, table_size);
 	}
 
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h
index 0ae91c8b6d72..8d7c4814c68f 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h
@@ -202,5 +202,72 @@ void smu_cmn_get_backend_workload_mask(struct smu_context *smu,
 				       u32 workload_mask,
 				       u32 *backend_workload_mask);
 
+/*SMU gpu metrics */
+
+/* Attribute ID mapping */
+#define SMU_MATTR(X) AMDGPU_METRICS_ATTR_ID_##X
+/* Type ID mapping */
+#define SMU_MTYPE(X) AMDGPU_METRICS_TYPE_##X
+/* Unit ID mapping */
+#define SMU_MUNIT(X) AMDGPU_METRICS_UNIT_##X
+
+/* Map TYPEID to C type */
+#define SMU_CTYPE(TYPEID) SMU_CTYPE_##TYPEID
+
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_U8 u8
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_S8 s8
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_U16 u16
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_S16 s16
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_U32 u32
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_S32 s32
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_U64 u64
+#define SMU_CTYPE_AMDGPU_METRICS_TYPE_S64 s64
+
+/* struct members */
+#define SMU_METRICS_SCALAR(ID, UNIT, TYPEID, NAME) \
+	u64 NAME##_ftype;                          \
+	SMU_CTYPE(TYPEID) NAME
+
+#define SMU_METRICS_ARRAY(ID, UNIT, TYPEID, NAME, SIZE) \
+	u64 NAME##_ftype;                               \
+	SMU_CTYPE(TYPEID) NAME[SIZE]
+
+/* Init functions for scalar/array fields - init to 0xFFs */
+#define SMU_METRICS_INIT_SCALAR(ID, UNIT, TYPEID, NAME)               \
+	do {                                                          \
+		obj->NAME##_ftype =                                   \
+			AMDGPU_METRICS_ENC_ATTR(UNIT, TYPEID, ID, 1); \
+		obj->NAME = (SMU_CTYPE(TYPEID)) ~0;                   \
+		count++;                                              \
+	} while (0)
+
+#define SMU_METRICS_INIT_ARRAY(ID, UNIT, TYPEID, NAME, SIZE)             \
+	do {                                                             \
+		obj->NAME##_ftype =                                      \
+			AMDGPU_METRICS_ENC_ATTR(UNIT, TYPEID, ID, SIZE); \
+		memset(obj->NAME, 0xFF, sizeof(obj->NAME));              \
+		count++;                                                 \
+	} while (0)
+
+/* Declare Metrics Class and Template object */
+#define DECLARE_SMU_METRICS_CLASS(CLASSNAME, SMU_METRICS_FIELD_LIST)           \
+	struct __packed CLASSNAME {                                            \
+		struct metrics_table_header header;                            \
+		int attr_count;                                                \
+		SMU_METRICS_FIELD_LIST(SMU_METRICS_SCALAR, SMU_METRICS_ARRAY); \
+	};                                                                     \
+	static inline void CLASSNAME##_init(struct CLASSNAME *obj,             \
+					    uint8_t frev, uint8_t crev)        \
+	{                                                                      \
+		int count = 0;                                                 \
+		memset(obj, 0xFF, sizeof(*obj));                               \
+		obj->header.format_revision = frev;                            \
+		obj->header.content_revision = crev;                           \
+		obj->header.structure_size = sizeof(*obj);                     \
+		SMU_METRICS_FIELD_LIST(SMU_METRICS_INIT_SCALAR,                \
+				       SMU_METRICS_INIT_ARRAY)                 \
+		obj->attr_count = count;                                       \
+	}
+
 #endif
 #endif
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h b/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h
index c09ecf1a68a0..34f6b4b1c3ba 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu_internal.h
@@ -100,6 +100,7 @@
 #define smu_is_asic_wbrf_supported(smu)			smu_ppt_funcs(is_asic_wbrf_supported, false, smu)
 #define smu_enable_uclk_shadow(smu, enable)		smu_ppt_funcs(enable_uclk_shadow, 0, smu, enable)
 #define smu_set_wbrf_exclusion_ranges(smu, freq_band_range)		smu_ppt_funcs(set_wbrf_exclusion_ranges, -EOPNOTSUPP, smu, freq_band_range)
+#define smu_get_ras_smu_drv(smu, ras_smu_drv)			smu_ppt_funcs(get_ras_smu_drv, -EOPNOTSUPP, smu, ras_smu_drv)
 
 #endif
 #endif
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_eeprom_i2c.c b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_eeprom_i2c.c
index 1bb7b7001ec7..3ed3ff42b7e1 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_eeprom_i2c.c
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_eeprom_i2c.c
@@ -85,6 +85,7 @@ static int ras_eeprom_i2c_config(struct ras_core_context *ras_core)
 	case IP_VERSION(13, 0, 5):
 	case IP_VERSION(13, 0, 6):
 	case IP_VERSION(13, 0, 10):
+	case IP_VERSION(13, 0, 12):
 	case IP_VERSION(13, 0, 14):
 		control->i2c_address = EEPROM_I2C_MADDR_4;
 		return 0;
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c
index 8007e49951d8..afe8135b6258 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.c
@@ -37,7 +37,7 @@
 #define MAX_XCD_NUM_PER_AID			2
 
 /* typical ECC bad page rate is 1 bad page per 100MB VRAM */
-#define ESTIMATE_BAD_PAGE_THRESHOLD(size)         ((size)/(100 * 1024 * 1024ULL))
+#define TYPICAL_ECC_BAD_PAGE_RATE (100ULL * SZ_1M)
 
 #define COUNT_BAD_PAGE_THRESHOLD(size) (((size) >> 21) << 4)
 
@@ -129,7 +129,7 @@ static int amdgpu_ras_mgr_init_eeprom_config(struct amdgpu_device *adev,
 	 */
 	if (amdgpu_bad_page_threshold == NONSTOP_OVER_THRESHOLD)
 		eeprom_cfg->eeprom_record_threshold_count =
-				ESTIMATE_BAD_PAGE_THRESHOLD(adev->gmc.mc_vram_size);
+			div64_u64(adev->gmc.mc_vram_size, TYPICAL_ECC_BAD_PAGE_RATE);
 	else if (amdgpu_bad_page_threshold == WARN_NONSTOP_OVER_THRESHOLD)
 		eeprom_cfg->eeprom_record_threshold_count =
 				COUNT_BAD_PAGE_THRESHOLD(RAS_RESERVED_VRAM_SIZE_DEFAULT);
@@ -172,12 +172,13 @@ static int amdgpu_ras_mgr_init_nbio_config(struct amdgpu_device *adev,
 
 	switch (config->nbio_ip_version) {
 	case IP_VERSION(7, 9, 0):
+	case IP_VERSION(7, 9, 1):
 		nbio_cfg->nbio_sys_fn = &amdgpu_ras_nbio_sys_func_v7_9;
 		break;
 	default:
 		RAS_DEV_ERR(adev,
 			"The nbio(0x%x) ras config is not right!\n",
-			config->mp1_ip_version);
+			config->nbio_ip_version);
 		ret = -EINVAL;
 		break;
 	}
@@ -258,7 +259,8 @@ static struct ras_core_context *amdgpu_ras_mgr_create_ras_core(struct amdgpu_dev
 	init_config.nbio_ip_version = amdgpu_ip_version(adev, NBIO_HWIP, 0);
 	init_config.psp_ip_version = amdgpu_ip_version(adev, MP1_HWIP, 0);
 
-	if (init_config.umc_ip_version == IP_VERSION(12, 0, 0))
+	if (init_config.umc_ip_version == IP_VERSION(12, 0, 0) ||
+	    init_config.umc_ip_version == IP_VERSION(12, 5, 0))
 		init_config.aca_ip_version = IP_VERSION(1, 0, 0);
 
 	init_config.sys_fn = &amdgpu_ras_sys_fn;
@@ -283,6 +285,18 @@ static int amdgpu_ras_mgr_sw_init(struct amdgpu_ip_block *ip_block)
 	struct amdgpu_ras_mgr *ras_mgr;
 	int ret = 0;
 
+	/* Disabled by default */
+	con->uniras_enabled = false;
+
+	/* Enabled only in debug mode */
+	if (adev->debug_enable_ras_aca) {
+		con->uniras_enabled = true;
+		RAS_DEV_INFO(adev, "Debug amdgpu uniras!");
+	}
+
+	if (!con->uniras_enabled)
+		return 0;
+
 	ras_mgr = kzalloc(sizeof(*ras_mgr), GFP_KERNEL);
 	if (!ras_mgr)
 		return -EINVAL;
@@ -315,6 +329,9 @@ static int amdgpu_ras_mgr_sw_fini(struct amdgpu_ip_block *ip_block)
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct amdgpu_ras_mgr *ras_mgr = (struct amdgpu_ras_mgr *)con->ras_mgr;
 
+	if (!con->uniras_enabled)
+		return 0;
+
 	if (!ras_mgr)
 		return 0;
 
@@ -332,12 +349,11 @@ static int amdgpu_ras_mgr_sw_fini(struct amdgpu_ip_block *ip_block)
 static int amdgpu_ras_mgr_hw_init(struct amdgpu_ip_block *ip_block)
 {
 	struct amdgpu_device *adev = ip_block->adev;
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
 	int ret;
 
-	/* Currently only debug mode can enable the ras module
-	 */
-	if (!adev->debug_enable_ras_aca)
+	if (!con->uniras_enabled)
 		return 0;
 
 	if (!ras_mgr || !ras_mgr->ras_core)
@@ -360,11 +376,10 @@ static int amdgpu_ras_mgr_hw_init(struct amdgpu_ip_block *ip_block)
 static int amdgpu_ras_mgr_hw_fini(struct amdgpu_ip_block *ip_block)
 {
 	struct amdgpu_device *adev = ip_block->adev;
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
 
-	/* Currently only debug mode can enable the ras module
-	 */
-	if (!adev->debug_enable_ras_aca)
+	if (!con->uniras_enabled)
 		return 0;
 
 	if (!ras_mgr || !ras_mgr->ras_core)
@@ -609,3 +624,25 @@ int amdgpu_ras_mgr_handle_ras_cmd(struct amdgpu_device *adev,
 
 	return ret;
 }
+
+int amdgpu_ras_mgr_pre_reset(struct amdgpu_device *adev)
+{
+	if (!amdgpu_ras_mgr_is_ready(adev)) {
+		RAS_DEV_ERR(adev, "Invalid ras suspend!\n");
+		return -EPERM;
+	}
+
+	amdgpu_ras_process_pre_reset(adev);
+	return 0;
+}
+
+int amdgpu_ras_mgr_post_reset(struct amdgpu_device *adev)
+{
+	if (!amdgpu_ras_mgr_is_ready(adev)) {
+		RAS_DEV_ERR(adev, "Invalid ras resume!\n");
+		return -EPERM;
+	}
+
+	amdgpu_ras_process_post_reset(adev);
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h
index 42f190a8feb9..8fb7eb4b8f13 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_mgr.h
@@ -52,6 +52,9 @@ struct amdgpu_ras_mgr {
 	struct ras_event_manager ras_event_mgr;
 	uint64_t last_poison_consumption_seqno;
 	bool ras_is_ready;
+
+	bool is_paused;
+	struct completion ras_event_done;
 };
 
 extern const struct amdgpu_ip_block_version ras_v1_0_ip_block;
@@ -75,4 +78,6 @@ bool amdgpu_ras_mgr_is_rma(struct amdgpu_device *adev);
 int amdgpu_ras_mgr_handle_ras_cmd(struct amdgpu_device *adev,
 		uint32_t cmd_id, void *input, uint32_t input_size,
 		void *output, uint32_t out_size);
+int amdgpu_ras_mgr_pre_reset(struct amdgpu_device *adev);
+int amdgpu_ras_mgr_post_reset(struct amdgpu_device *adev);
 #endif
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.c b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.c
index 6727fc9a2b9b..5782c007de71 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.c
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.c
@@ -29,6 +29,7 @@
 #include "amdgpu_ras_process.h"
 
 #define RAS_MGR_RETIRE_PAGE_INTERVAL  100
+#define RAS_EVENT_PROCESS_TIMEOUT  1200
 
 static void ras_process_retire_page_dwork(struct work_struct *work)
 {
@@ -57,6 +58,9 @@ int amdgpu_ras_process_init(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
 
+	ras_mgr->is_paused = false;
+	init_completion(&ras_mgr->ras_event_done);
+
 	INIT_DELAYED_WORK(&ras_mgr->retire_page_dwork, ras_process_retire_page_dwork);
 
 	return 0;
@@ -66,6 +70,7 @@ int amdgpu_ras_process_fini(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
 
+	ras_mgr->is_paused = false;
 	/* Save all cached bad pages to eeprom */
 	flush_delayed_work(&ras_mgr->retire_page_dwork);
 	cancel_delayed_work_sync(&ras_mgr->retire_page_dwork);
@@ -124,3 +129,62 @@ int amdgpu_ras_process_handle_consumption_interrupt(struct amdgpu_device *adev,
 
 	return ras_process_add_interrupt_req(ras_mgr->ras_core, &req, false);
 }
+
+int amdgpu_ras_process_begin(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
+
+	if (ras_mgr->is_paused)
+		return -EAGAIN;
+
+	reinit_completion(&ras_mgr->ras_event_done);
+	return 0;
+}
+
+int amdgpu_ras_process_end(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
+
+	complete(&ras_mgr->ras_event_done);
+	return 0;
+}
+
+int amdgpu_ras_process_pre_reset(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
+	long rc;
+
+	if (!ras_mgr || !ras_mgr->ras_core)
+		return -EINVAL;
+
+	if (!ras_mgr->ras_core->is_initialized)
+		return -EPERM;
+
+	ras_mgr->is_paused = true;
+
+	/* Wait for RAS event processing to complete */
+	rc = wait_for_completion_interruptible_timeout(&ras_mgr->ras_event_done,
+			msecs_to_jiffies(RAS_EVENT_PROCESS_TIMEOUT));
+	if (rc <= 0)
+		RAS_DEV_WARN(adev, "Waiting for ras process to complete %s\n",
+			 rc ? "interrupted" : "timeout");
+
+	flush_delayed_work(&ras_mgr->retire_page_dwork);
+	return 0;
+}
+
+int amdgpu_ras_process_post_reset(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras_mgr *ras_mgr = amdgpu_ras_mgr_get_context(adev);
+
+	if (!ras_mgr || !ras_mgr->ras_core)
+		return -EINVAL;
+
+	if (!ras_mgr->ras_core->is_initialized)
+		return -EPERM;
+
+	ras_mgr->is_paused = false;
+
+	schedule_delayed_work(&ras_mgr->retire_page_dwork, 0);
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.h b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.h
index b9502bd21beb..d55cdaeac441 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.h
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_process.h
@@ -34,4 +34,8 @@ int amdgpu_ras_process_handle_unexpected_interrupt(struct amdgpu_device *adev,
 		void *data);
 int amdgpu_ras_process_handle_consumption_interrupt(struct amdgpu_device *adev,
 		void *data);
+int amdgpu_ras_process_begin(struct amdgpu_device *adev);
+int amdgpu_ras_process_end(struct amdgpu_device *adev);
+int amdgpu_ras_process_pre_reset(struct amdgpu_device *adev);
+int amdgpu_ras_process_post_reset(struct amdgpu_device *adev);
 #endif
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c
index f21cd55a25be..45ed8c3b5563 100644
--- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c
+++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c
@@ -142,6 +142,12 @@ static int amdgpu_ras_sys_event_notifier(struct ras_core_context *ras_core,
 	case RAS_EVENT_ID__RESET_GPU:
 		ret = amdgpu_ras_mgr_reset_gpu(ras_core->dev, *(uint32_t *)data);
 		break;
+	case RAS_EVENT_ID__RAS_EVENT_PROC_BEGIN:
+		ret = amdgpu_ras_process_begin(ras_core->dev);
+		break;
+	case RAS_EVENT_ID__RAS_EVENT_PROC_END:
+		ret = amdgpu_ras_process_end(ras_core->dev);
+		break;
 	default:
 		RAS_DEV_WARN(ras_core->dev, "Invalid ras notify event:%d\n", event_id);
 		break;
diff --git a/drivers/gpu/drm/amd/ras/rascore/ras.h b/drivers/gpu/drm/amd/ras/rascore/ras.h
index fa224b36e3f2..3396b2e0949d 100644
--- a/drivers/gpu/drm/amd/ras/rascore/ras.h
+++ b/drivers/gpu/drm/amd/ras/rascore/ras.h
@@ -115,6 +115,8 @@ enum ras_notify_event {
 	RAS_EVENT_ID__FATAL_ERROR_DETECTED,
 	RAS_EVENT_ID__RESET_GPU,
 	RAS_EVENT_ID__RESET_VF,
+	RAS_EVENT_ID__RAS_EVENT_PROC_BEGIN,
+	RAS_EVENT_ID__RAS_EVENT_PROC_END,
 };
 
 enum ras_gpu_status {
diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_cper.c b/drivers/gpu/drm/amd/ras/rascore/ras_cper.c
index 2343991adccf..0fc7522b7ab6 100644
--- a/drivers/gpu/drm/amd/ras/rascore/ras_cper.c
+++ b/drivers/gpu/drm/amd/ras/rascore/ras_cper.c
@@ -54,7 +54,7 @@ static void fill_section_hdr(struct ras_core_context *ras_core,
 				enum ras_cper_severity sev, struct ras_log_info *trace)
 {
 	struct device_system_info dev_info = {0};
-	char record_id[16];
+	char record_id[32];
 
 	hdr->signature[0]		= 'C';
 	hdr->signature[1]		= 'P';
@@ -62,17 +62,16 @@ static void fill_section_hdr(struct ras_core_context *ras_core,
 	hdr->signature[3]		= 'R';
 	hdr->revision			= CPER_HDR__REV_1;
 	hdr->signature_end		= 0xFFFFFFFF;
-	hdr->error_severity		= sev;
+	hdr->error_severity		= (sev == RAS_CPER_SEV_RMA ? RAS_CPER_SEV_FATAL_UE : sev);
 
 	hdr->valid_bits.platform_id	= 1;
-	hdr->valid_bits.partition_id	= 1;
 	hdr->valid_bits.timestamp	= 1;
 
 	ras_core_get_device_system_info(ras_core, &dev_info);
 
 	cper_get_timestamp(ras_core, &hdr->timestamp, trace->timestamp);
 
-	snprintf(record_id, 9, "%d:%llX", dev_info.socket_id,
+	snprintf(record_id, sizeof(record_id), "%d:%llX", dev_info.socket_id,
 		    RAS_LOG_SEQNO_TO_BATCH_IDX(trace->seqno));
 	memcpy(hdr->record_id, record_id, 8);
 
@@ -116,7 +115,7 @@ static int fill_section_descriptor(struct ras_core_context *ras_core,
 	descriptor->sec_length		= section_length;
 	descriptor->valid_bits.fru_text	= 1;
 	descriptor->flag_bits.primary	= 1;
-	descriptor->severity			= sev;
+	descriptor->severity = (sev == RAS_CPER_SEV_RMA ? RAS_CPER_SEV_FATAL_UE : sev);
 	descriptor->sec_type			= sec_type;
 
 	ras_core_get_device_system_info(ras_core, &dev_info);
@@ -147,13 +146,19 @@ static int fill_section_fatal(struct ras_core_context *ras_core,
 }
 
 static int fill_section_runtime(struct ras_core_context *ras_core,
-		struct cper_section_runtime *runtime, struct ras_log_info *trace)
+		struct cper_section_runtime *runtime, struct ras_log_info *trace,
+		enum ras_cper_severity sev)
 {
 	runtime->hdr.valid_bits.err_info_cnt = 1;
 	runtime->hdr.valid_bits.err_context_cnt = 1;
 
 	runtime->descriptor.error_type = RUNTIME;
 	runtime->descriptor.ms_chk_bits.err_type_valid = 1;
+	if (sev == RAS_CPER_SEV_RMA) {
+		runtime->descriptor.valid_bits.ms_chk = 1;
+		runtime->descriptor.ms_chk_bits.err_type = 1;
+		runtime->descriptor.ms_chk_bits.pcc = 1;
+	}
 
 	runtime->reg.reg_ctx_type = CPER_CTX_TYPE__CRASH;
 	runtime->reg.reg_arr_size = sizeof(runtime->reg.reg_dump);
@@ -189,7 +194,7 @@ static int cper_generate_runtime_record(struct ras_core_context *ras_core,
 		fill_section_descriptor(ras_core, descriptor, sev, RUNTIME,
 			RAS_NONSTD_SEC_OFFSET(hdr->sec_cnt, i),
 			sizeof(struct cper_section_runtime));
-		fill_section_runtime(ras_core, runtime, trace_arr[i]);
+		fill_section_runtime(ras_core, runtime, trace_arr[i], sev);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_log_ring.c b/drivers/gpu/drm/amd/ras/rascore/ras_log_ring.c
index d0621464f1a7..0a838fdcb2f6 100644
--- a/drivers/gpu/drm/amd/ras/rascore/ras_log_ring.c
+++ b/drivers/gpu/drm/amd/ras/rascore/ras_log_ring.c
@@ -226,7 +226,9 @@ void ras_log_ring_add_log_event(struct ras_core_context *ras_core,
 		enum ras_log_event event, void *data, struct ras_log_batch_tag *batch_tag)
 {
 	struct ras_log_ring *log_ring = &ras_core->ras_log_ring;
+	struct device_system_info dev_info = {0};
 	struct ras_log_info *log;
+	uint64_t socket_id;
 	void *obj;
 
 	obj = mempool_alloc_preallocated(log_ring->ras_log_mempool);
@@ -252,8 +254,13 @@ void ras_log_ring_add_log_event(struct ras_core_context *ras_core,
 	if (data)
 		memcpy(&log->aca_reg, data, sizeof(log->aca_reg));
 
-	if (event == RAS_LOG_EVENT_RMA)
+	if (event == RAS_LOG_EVENT_RMA) {
 		memcpy(&log->aca_reg, ras_rma_aca_reg, sizeof(log->aca_reg));
+		ras_core_get_device_system_info(ras_core, &dev_info);
+		socket_id = dev_info.socket_id;
+		log->aca_reg.regs[ACA_REG_IDX__IPID] |= ((socket_id / 4) & 0x01);
+		log->aca_reg.regs[ACA_REG_IDX__IPID] |= (((socket_id % 4) & 0x3) << 44);
+	}
 
 	ras_log_ring_add_data(ras_core, log, batch_tag);
 }
diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_nbio.c b/drivers/gpu/drm/amd/ras/rascore/ras_nbio.c
index 8bf1f35d595e..bfddd104d548 100644
--- a/drivers/gpu/drm/amd/ras/rascore/ras_nbio.c
+++ b/drivers/gpu/drm/amd/ras/rascore/ras_nbio.c
@@ -31,6 +31,7 @@ static const struct ras_nbio_ip_func *ras_nbio_get_ip_funcs(
 {
 	switch (ip_version) {
 	case IP_VERSION(7, 9, 0):
+	case IP_VERSION(7, 9, 1):
 		return &ras_nbio_v7_9;
 	default:
 		RAS_DEV_ERR(ras_core->dev,
diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_process.c b/drivers/gpu/drm/amd/ras/rascore/ras_process.c
index 02f0657f78a3..3267dcdb169c 100644
--- a/drivers/gpu/drm/amd/ras/rascore/ras_process.c
+++ b/drivers/gpu/drm/amd/ras/rascore/ras_process.c
@@ -162,6 +162,11 @@ int ras_process_handle_ras_event(struct ras_core_context *ras_core)
 	uint32_t umc_event_count;
 	int ret;
 
+	ret = ras_core_event_notify(ras_core,
+			RAS_EVENT_ID__RAS_EVENT_PROC_BEGIN, NULL);
+	if (ret)
+		return ret;
+
 	ras_aca_clear_fatal_flag(ras_core);
 	ras_umc_log_pending_bad_bank(ras_core);
 
@@ -185,6 +190,8 @@ int ras_process_handle_ras_event(struct ras_core_context *ras_core)
 		atomic_set(&ras_proc->umc_interrupt_count, 0);
 	}
 
+	ras_core_event_notify(ras_core,
+			RAS_EVENT_ID__RAS_EVENT_PROC_END, NULL);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_umc.c b/drivers/gpu/drm/amd/ras/rascore/ras_umc.c
index 4067359bb299..4dae64c424a2 100644
--- a/drivers/gpu/drm/amd/ras/rascore/ras_umc.c
+++ b/drivers/gpu/drm/amd/ras/rascore/ras_umc.c
@@ -71,6 +71,7 @@ static const struct ras_umc_ip_func *ras_umc_get_ip_func(
 {
 	switch (ip_version) {
 	case IP_VERSION(12, 0, 0):
+	case IP_VERSION(12, 5, 0):
 		return &ras_umc_func_v12_0;
 	default:
 		RAS_DEV_ERR(ras_core->dev,