Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf

Cross-merge bpf fixes after downstream PR.

Trivial conflict:
tools/testing/selftests/bpf/prog_tests/verifier.c

Adjacent changes in:
Auto-merging kernel/bpf/verifier.c
Auto-merging samples/bpf/Makefile
Auto-merging tools/testing/selftests/bpf/.gitignore
Auto-merging tools/testing/selftests/bpf/Makefile
Auto-merging tools/testing/selftests/bpf/prog_tests/verifier.c

Signed-off-by: Alexei Starovoitov <ast@kernel.org>

76 files changed, 1281 insertions, 746 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 9095c05e0269..96316111300a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -145,7 +145,7 @@ const char *amdgpu_asic_name[] = {
 	"LAST",
 };
 
-#define AMDGPU_IP_BLK_MASK_ALL GENMASK(AMDGPU_MAX_IP_NUM - 1, 0)
+#define AMDGPU_IP_BLK_MASK_ALL GENMASK(AMDGPU_MAX_IP_NUM, 0)
 /*
  * Default init level where all blocks are expected to be initialized. This is
  * the level of initialization expected by default and also after a full reset
@@ -3670,9 +3670,11 @@ static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev)
 				continue;
 
 			r = block->version->funcs->hw_init(&adev->ip_blocks[i]);
-			DRM_INFO("RE-INIT-early: %s %s\n", block->version->funcs->name, r?"failed":"succeeded");
-			if (r)
+			if (r) {
+				dev_err(adev->dev, "RE-INIT-early: %s failed\n",
+					 block->version->funcs->name);
 				return r;
+			}
 			block->status.hw = true;
 		}
 	}
@@ -3682,7 +3684,8 @@ static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev)
 
 static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev)
 {
-	int i, r;
+	struct amdgpu_ip_block *block;
+	int i, r = 0;
 
 	static enum amd_ip_block_type ip_order[] = {
 		AMD_IP_BLOCK_TYPE_SMC,
@@ -3697,34 +3700,28 @@ static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev)
 	};
 
 	for (i = 0; i < ARRAY_SIZE(ip_order); i++) {
-		int j;
-		struct amdgpu_ip_block *block;
-
-		for (j = 0; j < adev->num_ip_blocks; j++) {
-			block = &adev->ip_blocks[j];
+		block = amdgpu_device_ip_get_ip_block(adev, ip_order[i]);
 
-			if (block->version->type != ip_order[i] ||
-				!block->status.valid ||
-				block->status.hw)
-				continue;
+		if (!block)
+			continue;
 
+		if (block->status.valid && !block->status.hw) {
 			if (block->version->type == AMD_IP_BLOCK_TYPE_SMC) {
-				r = amdgpu_ip_block_resume(&adev->ip_blocks[i]);
-				if (r)
-					return r;
+				r = amdgpu_ip_block_resume(block);
 			} else {
-				r = block->version->funcs->hw_init(&adev->ip_blocks[i]);
-				if (r) {
-					DRM_ERROR("hw_init of IP block <%s> failed %d\n",
-						  adev->ip_blocks[i].version->funcs->name, r);
-					return r;
-				}
-				block->status.hw = true;
+				r = block->version->funcs->hw_init(block);
 			}
+
+			if (r) {
+				dev_err(adev->dev, "RE-INIT-late: %s failed\n",
+					 block->version->funcs->name);
+				break;
+			}
+			block->status.hw = true;
 		}
 	}
 
-	return 0;
+	return r;
 }
 
 /**
@@ -3765,7 +3762,7 @@ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev)
  *
  * @adev: amdgpu_device pointer
  *
- * First resume function for hardware IPs.  The list of all the hardware
+ * Second resume function for hardware IPs.  The list of all the hardware
  * IPs that make up the asic is walked and the resume callbacks are run for
  * all blocks except COMMON, GMC, and IH.  resume puts the hardware into a
  * functional state after a suspend and updates the software state as
@@ -3783,6 +3780,7 @@ static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev)
 		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON ||
 		    adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC ||
 		    adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH ||
+		    adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE ||
 		    adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP)
 			continue;
 		r = amdgpu_ip_block_resume(&adev->ip_blocks[i]);
@@ -3794,6 +3792,36 @@ static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev)
 }
 
 /**
+ * amdgpu_device_ip_resume_phase3 - run resume for hardware IPs
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Third resume function for hardware IPs.  The list of all the hardware
+ * IPs that make up the asic is walked and the resume callbacks are run for
+ * all DCE.  resume puts the hardware into a functional state after a suspend
+ * and updates the software state as necessary.  This function is also used
+ * for restoring the GPU after a GPU reset.
+ *
+ * Returns 0 on success, negative error code on failure.
+ */
+static int amdgpu_device_ip_resume_phase3(struct amdgpu_device *adev)
+{
+	int i, r;
+
+	for (i = 0; i < adev->num_ip_blocks; i++) {
+		if (!adev->ip_blocks[i].status.valid || adev->ip_blocks[i].status.hw)
+			continue;
+		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE) {
+			r = amdgpu_ip_block_resume(&adev->ip_blocks[i]);
+			if (r)
+				return r;
+		}
+	}
+
+	return 0;
+}
+
+/**
  * amdgpu_device_ip_resume - run resume for hardware IPs
  *
  * @adev: amdgpu_device pointer
@@ -3822,6 +3850,13 @@ static int amdgpu_device_ip_resume(struct amdgpu_device *adev)
 	if (adev->mman.buffer_funcs_ring->sched.ready)
 		amdgpu_ttm_set_buffer_funcs_status(adev, true);
 
+	if (r)
+		return r;
+
+	amdgpu_fence_driver_hw_init(adev);
+
+	r = amdgpu_device_ip_resume_phase3(adev);
+
 	return r;
 }
 
@@ -4902,7 +4937,6 @@ int amdgpu_device_resume(struct drm_device *dev, bool notify_clients)
 		dev_err(adev->dev, "amdgpu_device_ip_resume failed (%d).\n", r);
 		goto exit;
 	}
-	amdgpu_fence_driver_hw_init(adev);
 
 	if (!adev->in_s0ix) {
 		r = amdgpu_amdkfd_resume(adev, adev->in_runpm);
@@ -5487,6 +5521,10 @@ int amdgpu_device_reinit_after_reset(struct amdgpu_reset_context *reset_context)
 				if (tmp_adev->mman.buffer_funcs_ring->sched.ready)
 					amdgpu_ttm_set_buffer_funcs_status(tmp_adev, true);
 
+				r = amdgpu_device_ip_resume_phase3(tmp_adev);
+				if (r)
+					goto out;
+
 				if (vram_lost)
 					amdgpu_device_fill_reset_magic(tmp_adev);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 9f922ec50ea2..c8180cad0abd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -61,7 +61,7 @@
 #include "amdgpu_res_cursor.h"
 #include "bif/bif_4_1_d.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 #define AMDGPU_TTM_VRAM_MAX_DW_READ	((size_t)128)
 
diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c
index e019249883fb..194026e9be33 100644
--- a/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c
@@ -40,10 +40,12 @@
 static void hdp_v4_0_flush_hdp(struct amdgpu_device *adev,
 				struct amdgpu_ring *ring)
 {
-	if (!ring || !ring->funcs->emit_wreg)
+	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
-	else
+		RREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+	} else {
 		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
+	}
 }
 
 static void hdp_v4_0_invalidate_hdp(struct amdgpu_device *adev,
@@ -54,11 +56,13 @@ static void hdp_v4_0_invalidate_hdp(struct amdgpu_device *adev,
 	    amdgpu_ip_version(adev, HDP_HWIP, 0) == IP_VERSION(4, 4, 5))
 		return;
 
-	if (!ring || !ring->funcs->emit_wreg)
+	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32_SOC15_NO_KIQ(HDP, 0, mmHDP_READ_CACHE_INVALIDATE, 1);
-	else
+		RREG32_SOC15_NO_KIQ(HDP, 0, mmHDP_READ_CACHE_INVALIDATE);
+	} else {
 		amdgpu_ring_emit_wreg(ring, SOC15_REG_OFFSET(
 			HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1);
+	}
 }
 
 static void hdp_v4_0_query_ras_error_count(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c
index ed7facacf2fe..d3962d469088 100644
--- a/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c
@@ -31,10 +31,12 @@
 static void hdp_v5_0_flush_hdp(struct amdgpu_device *adev,
 				struct amdgpu_ring *ring)
 {
-	if (!ring || !ring->funcs->emit_wreg)
+	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
-	else
+		RREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+	} else {
 		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
+	}
 }
 
 static void hdp_v5_0_invalidate_hdp(struct amdgpu_device *adev,
@@ -42,6 +44,7 @@ static void hdp_v5_0_invalidate_hdp(struct amdgpu_device *adev,
 {
 	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32_SOC15_NO_KIQ(HDP, 0, mmHDP_READ_CACHE_INVALIDATE, 1);
+		RREG32_SOC15_NO_KIQ(HDP, 0, mmHDP_READ_CACHE_INVALIDATE);
 	} else {
 		amdgpu_ring_emit_wreg(ring, SOC15_REG_OFFSET(
 					HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1);
diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v5_2.c b/drivers/gpu/drm/amd/amdgpu/hdp_v5_2.c
index 29c3484ae1f1..f52552c5fa27 100644
--- a/drivers/gpu/drm/amd/amdgpu/hdp_v5_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/hdp_v5_2.c
@@ -31,13 +31,15 @@
 static void hdp_v5_2_flush_hdp(struct amdgpu_device *adev,
 				struct amdgpu_ring *ring)
 {
-	if (!ring || !ring->funcs->emit_wreg)
+	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32_NO_KIQ((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2,
 			0);
-	else
+		RREG32_NO_KIQ((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+	} else {
 		amdgpu_ring_emit_wreg(ring,
 			(adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2,
 			0);
+	}
 }
 
 static void hdp_v5_2_update_mem_power_gating(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v6_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v6_0.c
index 33736d361dd0..6948fe9956ce 100644
--- a/drivers/gpu/drm/amd/amdgpu/hdp_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/hdp_v6_0.c
@@ -34,10 +34,12 @@
 static void hdp_v6_0_flush_hdp(struct amdgpu_device *adev,
 				struct amdgpu_ring *ring)
 {
-	if (!ring || !ring->funcs->emit_wreg)
+	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
-	else
+		RREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+	} else {
 		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
+	}
 }
 
 static void hdp_v6_0_update_clock_gating(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v7_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v7_0.c
index 1c99bb09e2a1..63820329f67e 100644
--- a/drivers/gpu/drm/amd/amdgpu/hdp_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/hdp_v7_0.c
@@ -31,10 +31,12 @@
 static void hdp_v7_0_flush_hdp(struct amdgpu_device *adev,
 				struct amdgpu_ring *ring)
 {
-	if (!ring || !ring->funcs->emit_wreg)
+	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
-	else
+		RREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+	} else {
 		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
+	}
 }
 
 static void hdp_v7_0_update_clock_gating(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c
index 7319299f25ae..03b8b7cd5229 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v1_0.c
@@ -604,7 +604,7 @@ static void jpeg_v1_0_set_irq_funcs(struct amdgpu_device *adev)
 static void jpeg_v1_0_ring_begin_use(struct amdgpu_ring *ring)
 {
 	struct	amdgpu_device *adev = ring->adev;
-	bool	set_clocks = !cancel_delayed_work_sync(&adev->jpeg.idle_work);
+	bool	set_clocks = !cancel_delayed_work_sync(&adev->vcn.idle_work);
 	int		cnt = 0;
 
 	mutex_lock(&adev->vcn.vcn1_jpeg1_workaround);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index 723f1220e1cc..7b826a136ceb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -1510,6 +1510,8 @@ static int kfd_fill_gpu_cache_info_from_gfx_config_v2(struct kfd_dev *kdev,
 	if (adev->gfx.config.gc_tcp_size_per_cu) {
 		pcache_info[i].cache_size = adev->gfx.config.gc_tcp_size_per_cu;
 		pcache_info[i].cache_level = 1;
+		/* Cacheline size not available in IP discovery for gc943,gc944 */
+		pcache_info[i].cache_line_size = 128;
 		pcache_info[i].flags = (CRAT_CACHE_FLAGS_ENABLED |
 					CRAT_CACHE_FLAGS_DATA_CACHE |
 					CRAT_CACHE_FLAGS_SIMD_CACHE);
@@ -1521,6 +1523,7 @@ static int kfd_fill_gpu_cache_info_from_gfx_config_v2(struct kfd_dev *kdev,
 		pcache_info[i].cache_size =
 			adev->gfx.config.gc_l1_instruction_cache_size_per_sqc;
 		pcache_info[i].cache_level = 1;
+		pcache_info[i].cache_line_size = 64;
 		pcache_info[i].flags = (CRAT_CACHE_FLAGS_ENABLED |
 					CRAT_CACHE_FLAGS_INST_CACHE |
 					CRAT_CACHE_FLAGS_SIMD_CACHE);
@@ -1531,6 +1534,7 @@ static int kfd_fill_gpu_cache_info_from_gfx_config_v2(struct kfd_dev *kdev,
 	if (adev->gfx.config.gc_l1_data_cache_size_per_sqc) {
 		pcache_info[i].cache_size = adev->gfx.config.gc_l1_data_cache_size_per_sqc;
 		pcache_info[i].cache_level = 1;
+		pcache_info[i].cache_line_size = 64;
 		pcache_info[i].flags = (CRAT_CACHE_FLAGS_ENABLED |
 					CRAT_CACHE_FLAGS_DATA_CACHE |
 					CRAT_CACHE_FLAGS_SIMD_CACHE);
@@ -1541,6 +1545,7 @@ static int kfd_fill_gpu_cache_info_from_gfx_config_v2(struct kfd_dev *kdev,
 	if (adev->gfx.config.gc_tcc_size) {
 		pcache_info[i].cache_size = adev->gfx.config.gc_tcc_size;
 		pcache_info[i].cache_level = 2;
+		pcache_info[i].cache_line_size = 128;
 		pcache_info[i].flags = (CRAT_CACHE_FLAGS_ENABLED |
 					CRAT_CACHE_FLAGS_DATA_CACHE |
 					CRAT_CACHE_FLAGS_SIMD_CACHE);
@@ -1551,6 +1556,7 @@ static int kfd_fill_gpu_cache_info_from_gfx_config_v2(struct kfd_dev *kdev,
 	if (adev->gmc.mall_size) {
 		pcache_info[i].cache_size = adev->gmc.mall_size / 1024;
 		pcache_info[i].cache_level = 3;
+		pcache_info[i].cache_line_size = 64;
 		pcache_info[i].flags = (CRAT_CACHE_FLAGS_ENABLED |
 					CRAT_CACHE_FLAGS_DATA_CACHE |
 					CRAT_CACHE_FLAGS_SIMD_CACHE);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 956198da7859..9b51dd75fefc 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -235,6 +235,9 @@ static void kfd_device_info_init(struct kfd_dev *kfd,
 			 */
 			kfd->device_info.needs_pci_atomics = true;
 			kfd->device_info.no_atomic_fw_version = kfd->adev->gfx.rs64_enable ? 509 : 0;
+		} else if (gc_version < IP_VERSION(13, 0, 0)) {
+			kfd->device_info.needs_pci_atomics = true;
+			kfd->device_info.no_atomic_fw_version = 2090;
 		} else {
 			kfd->device_info.needs_pci_atomics = true;
 		}
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 f0a6816709ca..48be917e7bc5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3481,6 +3481,8 @@ static void update_connector_ext_caps(struct amdgpu_dm_connector *aconnector)
 		caps->aux_support = false;
 	else if (amdgpu_backlight == 1)
 		caps->aux_support = true;
+	if (caps->aux_support)
+		aconnector->dc_link->backlight_control_type = BACKLIGHT_CONTROL_AMD_AUX;
 
 	luminance_range = &conn_base->display_info.luminance_range;
 
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 b0fea0856866..6cbbb71d752b 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
@@ -907,14 +907,14 @@ dm_helpers_probe_acpi_edid(void *data, u8 *buf, unsigned int block, size_t len)
 	struct drm_connector *connector = data;
 	struct acpi_device *acpidev = ACPI_COMPANION(connector->dev->dev);
 	unsigned char start = block * EDID_LENGTH;
-	void *edid;
+	struct edid *edid;
 	int r;
 
 	if (!acpidev)
 		return -ENODEV;
 
 	/* fetch the entire edid from BIOS */
-	r = acpi_video_get_edid(acpidev, ACPI_VIDEO_DISPLAY_LCD, -1, &edid);
+	r = acpi_video_get_edid(acpidev, ACPI_VIDEO_DISPLAY_LCD, -1, (void *)&edid);
 	if (r < 0) {
 		drm_dbg(connector->dev, "Failed to get EDID from ACPI: %d\n", r);
 		return r;
@@ -924,7 +924,14 @@ dm_helpers_probe_acpi_edid(void *data, u8 *buf, unsigned int block, size_t len)
 		goto cleanup;
 	}
 
-	memcpy(buf, edid + start, len);
+	/* sanity check */
+	if (edid->revision < 4 || !(edid->input & DRM_EDID_INPUT_DIGITAL) ||
+	    (edid->input & DRM_EDID_DIGITAL_TYPE_MASK) == DRM_EDID_DIGITAL_TYPE_UNDEF) {
+		r = -EINVAL;
+		goto cleanup;
+	}
+
+	memcpy(buf, (void *)edid + start, len);
 	r = 0;
 
 cleanup:
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 1dd26d5df6b9..49fe7dcf9372 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -6109,3 +6109,21 @@ struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state
 		profile.power_level = dc->res_pool->funcs->get_power_profile(context);
 	return profile;
 }
+
+/*
+ **********************************************************************************
+ * dc_get_det_buffer_size_from_state() - extracts detile buffer size from dc state
+ *
+ * Called when DM wants to log detile buffer size from dc_state
+ *
+ **********************************************************************************
+ */
+unsigned int dc_get_det_buffer_size_from_state(const struct dc_state *context)
+{
+	struct dc *dc = context->clk_mgr->ctx->dc;
+
+	if (dc->res_pool->funcs->get_det_buffer_size)
+		return dc->res_pool->funcs->get_det_buffer_size(context);
+	else
+		return 0;
+}
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 619fad17de55..626f75b6ad00 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -2094,7 +2094,8 @@ int resource_get_odm_slice_dst_width(struct pipe_ctx *otg_master,
 	count = resource_get_odm_slice_count(otg_master);
 	h_active = timing->h_addressable +
 			timing->h_border_left +
-			timing->h_border_right;
+			timing->h_border_right +
+			otg_master->hblank_borrow;
 	width = h_active / count;
 
 	if (otg_master->stream_res.tg)
@@ -4027,6 +4028,41 @@ fail:
 }
 
 /**
+ * decide_hblank_borrow - Decides the horizontal blanking borrow value for a given pipe context.
+ * @pipe_ctx: Pointer to the pipe context structure.
+ *
+ * This function calculates the horizontal blanking borrow value for a given pipe context based on the
+ * display stream compression (DSC) configuration. If the horizontal active pixels (hactive) are less
+ * than the total width of the DSC slices, it sets the hblank_borrow value to the difference. If the
+ * total horizontal timing minus the hblank_borrow value is less than 32, it resets the hblank_borrow
+ * value to 0.
+ */
+static void decide_hblank_borrow(struct pipe_ctx *pipe_ctx)
+{
+	uint32_t hactive;
+	uint32_t ceil_slice_width;
+	struct dc_stream_state *stream = NULL;
+
+	if (!pipe_ctx)
+		return;
+
+	stream = pipe_ctx->stream;
+
+	if (stream->timing.flags.DSC) {
+		hactive = stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right;
+
+		/* Assume if determined slices does not divide Hactive evenly, Hborrow is needed for padding*/
+		if (hactive % stream->timing.dsc_cfg.num_slices_h != 0) {
+			ceil_slice_width = (hactive / stream->timing.dsc_cfg.num_slices_h) + 1;
+			pipe_ctx->hblank_borrow = ceil_slice_width * stream->timing.dsc_cfg.num_slices_h - hactive;
+
+			if (stream->timing.h_total - hactive - pipe_ctx->hblank_borrow < 32)
+				pipe_ctx->hblank_borrow = 0;
+		}
+	}
+}
+
+/**
  * dc_validate_global_state() - Determine if hardware can support a given state
  *
  * @dc: dc struct for this driver
@@ -4064,6 +4100,10 @@ enum dc_status dc_validate_global_state(
 			if (pipe_ctx->stream != stream)
 				continue;
 
+			/* Decide whether hblank borrow is needed and save it in pipe_ctx */
+			if (dc->debug.enable_hblank_borrow)
+				decide_hblank_borrow(pipe_ctx);
+
 			if (dc->res_pool->funcs->patch_unknown_plane_state &&
 					pipe_ctx->plane_state &&
 					pipe_ctx->plane_state->tiling_info.gfx9.swizzle == DC_SW_UNKNOWN) {
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 104051935884..e9b9126c0401 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -290,6 +290,7 @@ struct dc_caps {
 	uint16_t subvp_vertical_int_margin_us;
 	bool seamless_odm;
 	uint32_t max_v_total;
+	bool vtotal_limited_by_fp2;
 	uint32_t max_disp_clock_khz_at_vmin;
 	uint8_t subvp_drr_vblank_start_margin_us;
 	bool cursor_not_scaled;
@@ -1068,6 +1069,7 @@ struct dc_debug_options {
 	unsigned int scale_to_sharpness_policy;
 	bool skip_full_updated_if_possible;
 	unsigned int enable_oled_edp_power_up_opt;
+	bool enable_hblank_borrow;
 };
 
 
@@ -2550,6 +2552,8 @@ struct dc_power_profile {
 
 struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state *context);
 
+unsigned int dc_get_det_buffer_size_from_state(const struct dc_state *context);
+
 /* DSC Interfaces */
 #include "dc_dsc.h"
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c b/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c
index c8d8e335fa37..0e310fd48b5c 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c
@@ -120,7 +120,7 @@ void translate_SPL_in_params_from_pipe_ctx(struct pipe_ctx *pipe_ctx, struct spl
 	spl_in->odm_slice_index = resource_get_odm_slice_index(pipe_ctx);
 	// Make spl input basic out info output_size width point to stream h active
 	spl_in->basic_out.output_size.width =
-		stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right;
+		stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right + pipe_ctx->hblank_borrow;
 	// Make spl input basic out info output_size height point to v active
 	spl_in->basic_out.output_size.height =
 		stream->timing.v_addressable + stream->timing.v_border_bottom + stream->timing.v_border_top;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c
index d851c081e376..8dabb1ac0b68 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c
@@ -1222,6 +1222,7 @@ static dml_bool_t CalculatePrefetchSchedule(struct display_mode_lib_scratch_st *
 	s->dst_y_prefetch_oto = s->Tvm_oto_lines + 2 * s->Tr0_oto_lines + s->Lsw_oto;
 
 	s->dst_y_prefetch_equ = p->VStartup - (*p->TSetup + dml_max(p->TWait + p->TCalc, *p->Tdmdl)) / s->LineTime - (*p->DSTYAfterScaler + (dml_float_t) *p->DSTXAfterScaler / (dml_float_t)p->myPipe->HTotal);
+	s->dst_y_prefetch_equ = dml_min(s->dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
 
 #ifdef __DML_VBA_DEBUG__
 	dml_print("DML::%s: HTotal = %u\n", __func__, p->myPipe->HTotal);
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c
index 138b4b1e42ed..c6a5a8614679 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c
@@ -339,11 +339,22 @@ void dml21_apply_soc_bb_overrides(struct dml2_initialize_instance_in_out *dml_in
 	// }
 }
 
+static unsigned int calc_max_hardware_v_total(const struct dc_stream_state *stream)
+{
+	unsigned int max_hw_v_total = stream->ctx->dc->caps.max_v_total;
+
+	if (stream->ctx->dc->caps.vtotal_limited_by_fp2) {
+		max_hw_v_total -= stream->timing.v_front_porch + 1;
+	}
+
+	return max_hw_v_total;
+}
+
 static void populate_dml21_timing_config_from_stream_state(struct dml2_timing_cfg *timing,
 		struct dc_stream_state *stream,
 		struct dml2_context *dml_ctx)
 {
-	unsigned int hblank_start, vblank_start;
+	unsigned int hblank_start, vblank_start, min_hardware_refresh_in_uhz;
 
 	timing->h_active = stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right;
 	timing->v_active = stream->timing.v_addressable + stream->timing.v_border_bottom + stream->timing.v_border_top;
@@ -371,11 +382,23 @@ static void populate_dml21_timing_config_from_stream_state(struct dml2_timing_cf
 		- stream->timing.v_border_top - stream->timing.v_border_bottom;
 
 	timing->drr_config.enabled = stream->ignore_msa_timing_param;
-	timing->drr_config.min_refresh_uhz = stream->timing.min_refresh_in_uhz;
 	timing->drr_config.drr_active_variable = stream->vrr_active_variable;
 	timing->drr_config.drr_active_fixed = stream->vrr_active_fixed;
 	timing->drr_config.disallowed = !stream->allow_freesync;
 
+	/* limit min refresh rate to DC cap */
+	min_hardware_refresh_in_uhz = stream->timing.min_refresh_in_uhz;
+	if (stream->ctx->dc->caps.max_v_total != 0) {
+		min_hardware_refresh_in_uhz = div64_u64((stream->timing.pix_clk_100hz * 100000000ULL),
+				(stream->timing.h_total * (long long)calc_max_hardware_v_total(stream)));
+	}
+
+	if (stream->timing.min_refresh_in_uhz > min_hardware_refresh_in_uhz) {
+		timing->drr_config.min_refresh_uhz = stream->timing.min_refresh_in_uhz;
+	} else {
+		timing->drr_config.min_refresh_uhz = min_hardware_refresh_in_uhz;
+	}
+
 	if (dml_ctx->config.callbacks.get_max_flickerless_instant_vtotal_increase &&
 			stream->ctx->dc->config.enable_fpo_flicker_detection == 1)
 		timing->drr_config.max_instant_vtotal_delta = dml_ctx->config.callbacks.get_max_flickerless_instant_vtotal_increase(stream, false);
@@ -422,6 +445,21 @@ static void populate_dml21_timing_config_from_stream_state(struct dml2_timing_cf
 	timing->vblank_nom = timing->v_total - timing->v_active;
 }
 
+/**
+ * adjust_dml21_hblank_timing_config_from_pipe_ctx - Adjusts the horizontal blanking timing configuration
+ *                                                   based on the pipe context.
+ * @timing: Pointer to the dml2_timing_cfg structure to be adjusted.
+ * @pipe: Pointer to the pipe_ctx structure containing the horizontal blanking borrow value.
+ *
+ * This function modifies the horizontal active and blank end timings by adding and subtracting
+ * the horizontal blanking borrow value from the pipe context, respectively.
+ */
+static void adjust_dml21_hblank_timing_config_from_pipe_ctx(struct dml2_timing_cfg *timing, struct pipe_ctx *pipe)
+{
+	timing->h_active += pipe->hblank_borrow;
+	timing->h_blank_end -= pipe->hblank_borrow;
+}
+
 static void populate_dml21_output_config_from_stream_state(struct dml2_link_output_cfg *output,
 		struct dc_stream_state *stream, const struct pipe_ctx *pipe)
 {
@@ -709,6 +747,7 @@ static const struct scaler_data *get_scaler_data_for_plane(
 			temp_pipe->plane_state = pipe->plane_state;
 			temp_pipe->plane_res.scl_data.taps = pipe->plane_res.scl_data.taps;
 			temp_pipe->stream_res = pipe->stream_res;
+			temp_pipe->hblank_borrow = pipe->hblank_borrow;
 			dml_ctx->config.callbacks.build_scaling_params(temp_pipe);
 			break;
 		}
@@ -973,6 +1012,7 @@ bool dml21_map_dc_state_into_dml_display_cfg(const struct dc *in_dc, struct dc_s
 
 		ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
 		populate_dml21_timing_config_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location].timing, context->streams[stream_index], dml_ctx);
+		adjust_dml21_hblank_timing_config_from_pipe_ctx(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location].timing, &context->res_ctx.pipe_ctx[stream_index]);
 		populate_dml21_output_config_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location].output, context->streams[stream_index], &context->res_ctx.pipe_ctx[stream_index]);
 		populate_dml21_stream_overrides_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location], context->streams[stream_index]);
 
@@ -1111,12 +1151,12 @@ void dml21_populate_pipe_ctx_dlg_params(struct dml2_context *dml_ctx, struct dc_
 	struct dc_crtc_timing *timing = &pipe_ctx->stream->timing;
 	union dml2_global_sync_programming *global_sync = &stream_programming->global_sync;
 
-	hactive = timing->h_addressable + timing->h_border_left + timing->h_border_right;
+	hactive = timing->h_addressable + timing->h_border_left + timing->h_border_right + pipe_ctx->hblank_borrow;
 	vactive = timing->v_addressable + timing->v_border_bottom + timing->v_border_top;
 	hblank_start = pipe_ctx->stream->timing.h_total - pipe_ctx->stream->timing.h_front_porch;
 	vblank_start = pipe_ctx->stream->timing.v_total - pipe_ctx->stream->timing.v_front_porch;
 
-	hblank_end = hblank_start - timing->h_addressable - timing->h_border_left - timing->h_border_right;
+	hblank_end = hblank_start - timing->h_addressable - timing->h_border_left - timing->h_border_right - pipe_ctx->hblank_borrow;
 	vblank_end = vblank_start - timing->v_addressable - timing->v_border_top - timing->v_border_bottom;
 
 	if (dml_ctx->config.svp_pstate.callbacks.get_pipe_subvp_type(context, pipe_ctx) == SUBVP_PHANTOM) {
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c
index d7f8b2dcaa6b..fa11f075d1f9 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c
@@ -1049,7 +1049,8 @@ void dcn32_update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
 		}
 
 		/* Enable DSC hw block */
-		dsc_cfg.pic_width = (stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
+		dsc_cfg.pic_width = (stream->timing.h_addressable + pipe_ctx->hblank_borrow +
+				stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
 		dsc_cfg.pic_height = stream->timing.v_addressable + stream->timing.v_border_top + stream->timing.v_border_bottom;
 		dsc_cfg.pixel_encoding = stream->timing.pixel_encoding;
 		dsc_cfg.color_depth = stream->timing.display_color_depth;
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 5de11e2837c0..307782592789 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
@@ -820,6 +820,7 @@ enum dc_status dcn401_enable_stream_timing(
 	int opp_cnt = 1;
 	int opp_inst[MAX_PIPES] = {0};
 	struct pipe_ctx *opp_heads[MAX_PIPES] = {0};
+	struct dc_crtc_timing patched_crtc_timing = stream->timing;
 	bool manual_mode;
 	unsigned int tmds_div = PIXEL_RATE_DIV_NA;
 	unsigned int unused_div = PIXEL_RATE_DIV_NA;
@@ -874,9 +875,13 @@ enum dc_status dcn401_enable_stream_timing(
 	if (dc->hwseq->funcs.PLAT_58856_wa && (!dc_is_dp_signal(stream->signal)))
 		dc->hwseq->funcs.PLAT_58856_wa(context, pipe_ctx);
 
+	/* if we are borrowing from hblank, h_addressable needs to be adjusted */
+	if (dc->debug.enable_hblank_borrow)
+		patched_crtc_timing.h_addressable = patched_crtc_timing.h_addressable + pipe_ctx->hblank_borrow;
+
 	pipe_ctx->stream_res.tg->funcs->program_timing(
 			pipe_ctx->stream_res.tg,
-			&stream->timing,
+			&patched_crtc_timing,
 			pipe_ctx->pipe_dlg_param.vready_offset,
 			pipe_ctx->pipe_dlg_param.vstartup_start,
 			pipe_ctx->pipe_dlg_param.vupdate_offset,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index 8597e866bfe6..2edd5b38ce4f 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -219,6 +219,7 @@ struct resource_funcs {
 	 * Get indicator of power from a context that went through full validation
 	 */
 	int (*get_power_profile)(const struct dc_state *context);
+	unsigned int (*get_det_buffer_size)(const struct dc_state *context);
 };
 
 struct audio_support{
@@ -477,6 +478,8 @@ struct pipe_ctx {
 	/* subvp_index: only valid if the pipe is a SUBVP_MAIN*/
 	uint8_t subvp_index;
 	struct pixel_rate_divider pixel_rate_divider;
+	/* pixels borrowed from hblank to hactive */
+	uint8_t hblank_borrow;
 };
 
 /* Data used for dynamic link encoder assignment.
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 41cab9ad6885..5d66bfc7fe6e 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
@@ -808,7 +808,8 @@ void link_set_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
 		enum optc_dsc_mode optc_dsc_mode;
 
 		/* Enable DSC hw block */
-		dsc_cfg.pic_width = (stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
+		dsc_cfg.pic_width = (stream->timing.h_addressable + pipe_ctx->hblank_borrow +
+				stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
 		dsc_cfg.pic_height = stream->timing.v_addressable + stream->timing.v_border_top + stream->timing.v_border_bottom;
 		dsc_cfg.pixel_encoding = stream->timing.pixel_encoding;
 		dsc_cfg.color_depth = stream->timing.display_color_depth;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c
index 189d0c85872e..7a5b9aa5292c 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c
@@ -1510,6 +1510,7 @@ bool dcn20_split_stream_for_odm(
 
 	if (prev_odm_pipe->plane_state) {
 		struct scaler_data *sd = &prev_odm_pipe->plane_res.scl_data;
+		struct output_pixel_processor *opp = next_odm_pipe->stream_res.opp;
 		int new_width;
 
 		/* HACTIVE halved for odm combine */
@@ -1543,7 +1544,28 @@ bool dcn20_split_stream_for_odm(
 		sd->viewport_c.x += dc_fixpt_floor(dc_fixpt_mul_int(
 				sd->ratios.horz_c, sd->h_active - sd->recout.x));
 		sd->recout.x = 0;
+
+		/*
+		 * When odm is used in YcbCr422 or 420 colour space, a split screen
+		 * will be seen with the previous calculations since the extra left
+		 *  edge pixel is accounted for in fmt but not in viewport.
+		 *
+		 * Below are calculations which fix the split by fixing the calculations
+		 * if there is an extra left edge pixel.
+		 */
+		if (opp && opp->funcs->opp_get_left_edge_extra_pixel_count
+				&& opp->funcs->opp_get_left_edge_extra_pixel_count(
+					opp, next_odm_pipe->stream->timing.pixel_encoding,
+					resource_is_pipe_type(next_odm_pipe, OTG_MASTER)) == 1) {
+			sd->h_active += 1;
+			sd->recout.width += 1;
+			sd->viewport.x -= dc_fixpt_ceil(dc_fixpt_mul_int(sd->ratios.horz, 1));
+			sd->viewport_c.x -= dc_fixpt_ceil(dc_fixpt_mul_int(sd->ratios.horz, 1));
+			sd->viewport_c.width += dc_fixpt_ceil(dc_fixpt_mul_int(sd->ratios.horz, 1));
+			sd->viewport.width += dc_fixpt_ceil(dc_fixpt_mul_int(sd->ratios.horz, 1));
+		}
 	}
+
 	if (!next_odm_pipe->top_pipe)
 		next_odm_pipe->stream_res.opp = pool->opps[next_odm_pipe->pipe_idx];
 	else
@@ -2132,6 +2154,7 @@ bool dcn20_fast_validate_bw(
 			ASSERT(0);
 		}
 	}
+
 	/* Actual dsc count per stream dsc validation*/
 	if (!dcn20_validate_dsc(dc, context)) {
 		context->bw_ctx.dml.vba.ValidationStatus[context->bw_ctx.dml.vba.soc.num_states] =
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c
index cd31e4f16c14..bfd0eccbed28 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c
@@ -2353,6 +2353,7 @@ static bool dcn30_resource_construct(
 
 	dc->caps.dp_hdmi21_pcon_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* read VBIOS LTTPR caps */
 	{
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c
index 02af8b8f4d27..7baefc910a3d 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c
@@ -1233,6 +1233,7 @@ static bool dcn302_resource_construct(
 	dc->caps.extended_aux_timeout_support = true;
 	dc->caps.dmcub_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* Color pipeline capabilities */
 	dc->caps.color.dpp.dcn_arch = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c
index 7002a8dd358a..8a57d46ad15f 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c
@@ -1178,6 +1178,7 @@ static bool dcn303_resource_construct(
 	dc->caps.extended_aux_timeout_support = true;
 	dc->caps.dmcub_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* Color pipeline capabilities */
 	dc->caps.color.dpp.dcn_arch = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c
index c16cf1c8f7f9..54ec3d8e920c 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c
@@ -1720,6 +1720,12 @@ int dcn31_populate_dml_pipes_from_context(
 	return pipe_cnt;
 }
 
+unsigned int dcn31_get_det_buffer_size(
+	const struct dc_state *context)
+{
+	return context->bw_ctx.dml.ip.det_buffer_size_kbytes;
+}
+
 void dcn31_calculate_wm_and_dlg(
 		struct dc *dc, struct dc_state *context,
 		display_e2e_pipe_params_st *pipes,
@@ -1842,6 +1848,7 @@ static struct resource_funcs dcn31_res_pool_funcs = {
 	.update_bw_bounding_box = dcn31_update_bw_bounding_box,
 	.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
 	.get_panel_config_defaults = dcn31_get_panel_config_defaults,
+	.get_det_buffer_size = dcn31_get_det_buffer_size,
 };
 
 static struct clock_source *dcn30_clock_source_create(
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h
index 901436591ed4..551ad912f7be 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h
@@ -63,6 +63,9 @@ struct resource_pool *dcn31_create_resource_pool(
 		const struct dc_init_data *init_data,
 		struct dc *dc);
 
+unsigned int dcn31_get_det_buffer_size(
+	const struct dc_state *context);
+
 /*temp: B0 specific before switch to dcn313 headers*/
 #ifndef regPHYPLLF_PIXCLK_RESYNC_CNTL
 #define regPHYPLLF_PIXCLK_RESYNC_CNTL 0x007e
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c
index c0f48c78e968..2794473f2aff 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c
@@ -1777,6 +1777,7 @@ static struct resource_funcs dcn314_res_pool_funcs = {
 	.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
 	.get_panel_config_defaults = dcn314_get_panel_config_defaults,
 	.get_preferred_eng_id_dpia = dcn314_get_preferred_eng_id_dpia,
+	.get_det_buffer_size = dcn31_get_det_buffer_size,
 };
 
 static struct clock_source *dcn30_clock_source_create(
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c
index 6c3295259a81..4ee33eb3381d 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c
@@ -1845,6 +1845,7 @@ static struct resource_funcs dcn315_res_pool_funcs = {
 	.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
 	.get_panel_config_defaults = dcn315_get_panel_config_defaults,
 	.get_power_profile = dcn315_get_power_profile,
+	.get_det_buffer_size = dcn31_get_det_buffer_size,
 };
 
 static bool dcn315_resource_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c
index 6edaaadcb173..79eddbafe3c2 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c
@@ -1719,6 +1719,7 @@ static struct resource_funcs dcn316_res_pool_funcs = {
 	.update_bw_bounding_box = dcn316_update_bw_bounding_box,
 	.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
 	.get_panel_config_defaults = dcn316_get_panel_config_defaults,
+	.get_det_buffer_size = dcn31_get_det_buffer_size,
 };
 
 static bool dcn316_resource_construct(
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 01d1a11d5545..12d247a7ec45 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
@@ -2189,6 +2189,7 @@ static bool dcn32_resource_construct(
 	dc->caps.dmcub_support = true;
 	dc->caps.seamless_odm = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* Color pipeline capabilities */
 	dc->caps.color.dpp.dcn_arch = 1;
@@ -2803,6 +2804,7 @@ struct pipe_ctx *dcn32_acquire_free_pipe_as_secondary_opp_head(
 		free_pipe->plane_res.xfm = pool->transforms[free_pipe_idx];
 		free_pipe->plane_res.dpp = pool->dpps[free_pipe_idx];
 		free_pipe->plane_res.mpcc_inst = pool->dpps[free_pipe_idx]->inst;
+		free_pipe->hblank_borrow = otg_master->hblank_borrow;
 		if (free_pipe->stream->timing.flags.DSC == 1) {
 			dcn20_acquire_dsc(free_pipe->stream->ctx->dc,
 					&new_ctx->res_ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
index 5cb74fd9cb7d..06b9479c8bd3 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
@@ -1742,6 +1742,7 @@ static bool dcn321_resource_construct(
 	dc->caps.extended_aux_timeout_support = true;
 	dc->caps.dmcub_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* Color pipeline capabilities */
 	dc->caps.color.dpp.dcn_arch = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c
index 6cc2960b6104..89e2adcf2a28 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c
@@ -1778,6 +1778,7 @@ static struct resource_funcs dcn35_res_pool_funcs = {
 	.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
 	.get_panel_config_defaults = dcn35_get_panel_config_defaults,
 	.get_preferred_eng_id_dpia = dcn35_get_preferred_eng_id_dpia,
+	.get_det_buffer_size = dcn31_get_det_buffer_size,
 };
 
 static bool dcn35_resource_construct(
@@ -1849,6 +1850,7 @@ static bool dcn35_resource_construct(
 	dc->caps.zstate_support = true;
 	dc->caps.ips_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* Color pipeline capabilities */
 	dc->caps.color.dpp.dcn_arch = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c
index d87e2641cda1..263a37c1cd3a 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c
@@ -1757,6 +1757,7 @@ static struct resource_funcs dcn351_res_pool_funcs = {
 	.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
 	.get_panel_config_defaults = dcn35_get_panel_config_defaults,
 	.get_preferred_eng_id_dpia = dcn351_get_preferred_eng_id_dpia,
+	.get_det_buffer_size = dcn31_get_det_buffer_size,
 };
 
 static bool dcn351_resource_construct(
@@ -1828,6 +1829,7 @@ static bool dcn351_resource_construct(
 	dc->caps.zstate_support = true;
 	dc->caps.ips_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	/* Color pipeline capabilities */
 	dc->caps.color.dpp.dcn_arch = 1;
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 db93bac247c0..2a3dabfe3cea 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
@@ -1864,6 +1864,7 @@ static bool dcn401_resource_construct(
 	dc->caps.extended_aux_timeout_support = true;
 	dc->caps.dmcub_support = true;
 	dc->caps.max_v_total = (1 << 15) - 1;
+	dc->caps.vtotal_limited_by_fp2 = true;
 
 	if (ASICREV_IS_GC_12_0_1_A0(dc->ctx->asic_id.hw_internal_rev))
 		dc->caps.dcc_plane_width_limit = 7680;
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index f980a84dceef..2b3964529539 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -122,6 +122,17 @@ static unsigned int calc_duration_in_us_from_v_total(
 	return duration_in_us;
 }
 
+static unsigned int calc_max_hardware_v_total(const struct dc_stream_state *stream)
+{
+	unsigned int max_hw_v_total = stream->ctx->dc->caps.max_v_total;
+
+	if (stream->ctx->dc->caps.vtotal_limited_by_fp2) {
+		max_hw_v_total -= stream->timing.v_front_porch + 1;
+	}
+
+	return max_hw_v_total;
+}
+
 unsigned int mod_freesync_calc_v_total_from_refresh(
 		const struct dc_stream_state *stream,
 		unsigned int refresh_in_uhz)
@@ -1016,7 +1027,7 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync,
 
 	if (stream->ctx->dc->caps.max_v_total != 0 && stream->timing.h_total != 0) {
 		min_hardware_refresh_in_uhz = div64_u64((stream->timing.pix_clk_100hz * 100000000ULL),
-			(stream->timing.h_total * (long long)stream->ctx->dc->caps.max_v_total));
+			(stream->timing.h_total * (long long)calc_max_hardware_v_total(stream)));
 	}
 	/* Limit minimum refresh rate to what can be supported by hardware */
 	min_refresh_in_uhz = min_hardware_refresh_in_uhz > in_config->min_refresh_in_uhz ?
diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
index 136e8193867c..e8ae7681bf0a 100644
--- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
@@ -1361,7 +1361,11 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev,
  * create a custom set of heuristics, write a string of numbers to the file
  * starting with the number of the custom profile along with a setting
  * for each heuristic parameter.  Due to differences across asic families
- * the heuristic parameters vary from family to family.
+ * the heuristic parameters vary from family to family. Additionally,
+ * you can apply the custom heuristics to different clock domains.  Each
+ * clock domain is considered a distinct operation so if you modify the
+ * gfxclk heuristics and then the memclk heuristics, the all of the
+ * custom heuristics will be retained until you switch to another profile.
  *
  */
 
diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
index b8355293518f..5eae14fe79f1 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
@@ -72,6 +72,10 @@ static int smu_set_power_limit(void *handle, uint32_t limit);
 static int smu_set_fan_speed_rpm(void *handle, uint32_t speed);
 static int smu_set_gfx_cgpg(struct smu_context *smu, bool enabled);
 static int smu_set_mp1_state(void *handle, enum pp_mp1_state mp1_state);
+static void smu_power_profile_mode_get(struct smu_context *smu,
+				       enum PP_SMC_POWER_PROFILE profile_mode);
+static void smu_power_profile_mode_put(struct smu_context *smu,
+				       enum PP_SMC_POWER_PROFILE profile_mode);
 
 static int smu_sys_get_pp_feature_mask(void *handle,
 				       char *buf)
@@ -1259,42 +1263,19 @@ static int smu_sw_init(struct amdgpu_ip_block *ip_block)
 	INIT_WORK(&smu->interrupt_work, smu_interrupt_work_fn);
 	atomic64_set(&smu->throttle_int_counter, 0);
 	smu->watermarks_bitmap = 0;
-	smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
-	smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
-	smu->user_dpm_profile.user_workload_mask = 0;
 
 	atomic_set(&smu->smu_power.power_gate.vcn_gated, 1);
 	atomic_set(&smu->smu_power.power_gate.jpeg_gated, 1);
 	atomic_set(&smu->smu_power.power_gate.vpe_gated, 1);
 	atomic_set(&smu->smu_power.power_gate.umsch_mm_gated, 1);
 
-	smu->workload_priority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0;
-	smu->workload_priority[PP_SMC_POWER_PROFILE_FULLSCREEN3D] = 1;
-	smu->workload_priority[PP_SMC_POWER_PROFILE_POWERSAVING] = 2;
-	smu->workload_priority[PP_SMC_POWER_PROFILE_VIDEO] = 3;
-	smu->workload_priority[PP_SMC_POWER_PROFILE_VR] = 4;
-	smu->workload_priority[PP_SMC_POWER_PROFILE_COMPUTE] = 5;
-	smu->workload_priority[PP_SMC_POWER_PROFILE_CUSTOM] = 6;
-
 	if (smu->is_apu ||
-	    !smu_is_workload_profile_available(smu, PP_SMC_POWER_PROFILE_FULLSCREEN3D)) {
-		smu->driver_workload_mask =
-			1 << smu->workload_priority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT];
-	} else {
-		smu->driver_workload_mask =
-			1 << smu->workload_priority[PP_SMC_POWER_PROFILE_FULLSCREEN3D];
-		smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_FULLSCREEN3D;
-	}
-
-	smu->workload_mask = smu->driver_workload_mask |
-							smu->user_dpm_profile.user_workload_mask;
-	smu->workload_setting[0] = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
-	smu->workload_setting[1] = PP_SMC_POWER_PROFILE_FULLSCREEN3D;
-	smu->workload_setting[2] = PP_SMC_POWER_PROFILE_POWERSAVING;
-	smu->workload_setting[3] = PP_SMC_POWER_PROFILE_VIDEO;
-	smu->workload_setting[4] = PP_SMC_POWER_PROFILE_VR;
-	smu->workload_setting[5] = PP_SMC_POWER_PROFILE_COMPUTE;
-	smu->workload_setting[6] = PP_SMC_POWER_PROFILE_CUSTOM;
+	    !smu_is_workload_profile_available(smu, PP_SMC_POWER_PROFILE_FULLSCREEN3D))
+		smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
+	else
+		smu->power_profile_mode = PP_SMC_POWER_PROFILE_FULLSCREEN3D;
+	smu_power_profile_mode_get(smu, smu->power_profile_mode);
+
 	smu->display_config = &adev->pm.pm_display_cfg;
 
 	smu->smu_dpm.dpm_level = AMD_DPM_FORCED_LEVEL_AUTO;
@@ -1347,6 +1328,11 @@ static int smu_sw_fini(struct amdgpu_ip_block *ip_block)
 		return ret;
 	}
 
+	if (smu->custom_profile_params) {
+		kfree(smu->custom_profile_params);
+		smu->custom_profile_params = NULL;
+	}
+
 	smu_fini_microcode(smu);
 
 	return 0;
@@ -2131,6 +2117,9 @@ static int smu_suspend(struct amdgpu_ip_block *ip_block)
 	if (!ret)
 		adev->gfx.gfx_off_entrycount = count;
 
+	/* clear this on suspend so it will get reprogrammed on resume */
+	smu->workload_mask = 0;
+
 	return 0;
 }
 
@@ -2243,25 +2232,49 @@ static int smu_enable_umd_pstate(void *handle,
 }
 
 static int smu_bump_power_profile_mode(struct smu_context *smu,
-					   long *param,
-					   uint32_t param_size)
+				       long *custom_params,
+				       u32 custom_params_max_idx)
 {
-	int ret = 0;
+	u32 workload_mask = 0;
+	int i, ret = 0;
+
+	for (i = 0; i < PP_SMC_POWER_PROFILE_COUNT; i++) {
+		if (smu->workload_refcount[i])
+			workload_mask |= 1 << i;
+	}
+
+	if (smu->workload_mask == workload_mask)
+		return 0;
 
 	if (smu->ppt_funcs->set_power_profile_mode)
-		ret = smu->ppt_funcs->set_power_profile_mode(smu, param, param_size);
+		ret = smu->ppt_funcs->set_power_profile_mode(smu, workload_mask,
+							     custom_params,
+							     custom_params_max_idx);
+
+	if (!ret)
+		smu->workload_mask = workload_mask;
 
 	return ret;
 }
 
+static void smu_power_profile_mode_get(struct smu_context *smu,
+				       enum PP_SMC_POWER_PROFILE profile_mode)
+{
+	smu->workload_refcount[profile_mode]++;
+}
+
+static void smu_power_profile_mode_put(struct smu_context *smu,
+				       enum PP_SMC_POWER_PROFILE profile_mode)
+{
+	if (smu->workload_refcount[profile_mode])
+		smu->workload_refcount[profile_mode]--;
+}
+
 static int smu_adjust_power_state_dynamic(struct smu_context *smu,
 					  enum amd_dpm_forced_level level,
-					  bool skip_display_settings,
-					  bool init)
+					  bool skip_display_settings)
 {
 	int ret = 0;
-	int index = 0;
-	long workload[1];
 	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
 
 	if (!skip_display_settings) {
@@ -2298,14 +2311,8 @@ static int smu_adjust_power_state_dynamic(struct smu_context *smu,
 	}
 
 	if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL &&
-		smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PERF_DETERMINISM) {
-		index = fls(smu->workload_mask);
-		index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0;
-		workload[0] = smu->workload_setting[index];
-
-		if (init || smu->power_profile_mode != workload[0])
-			smu_bump_power_profile_mode(smu, workload, 0);
-	}
+	    smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PERF_DETERMINISM)
+		smu_bump_power_profile_mode(smu, NULL, 0);
 
 	return ret;
 }
@@ -2324,13 +2331,13 @@ static int smu_handle_task(struct smu_context *smu,
 		ret = smu_pre_display_config_changed(smu);
 		if (ret)
 			return ret;
-		ret = smu_adjust_power_state_dynamic(smu, level, false, false);
+		ret = smu_adjust_power_state_dynamic(smu, level, false);
 		break;
 	case AMD_PP_TASK_COMPLETE_INIT:
-		ret = smu_adjust_power_state_dynamic(smu, level, true, true);
+		ret = smu_adjust_power_state_dynamic(smu, level, true);
 		break;
 	case AMD_PP_TASK_READJUST_POWER_STATE:
-		ret = smu_adjust_power_state_dynamic(smu, level, true, false);
+		ret = smu_adjust_power_state_dynamic(smu, level, true);
 		break;
 	default:
 		break;
@@ -2352,12 +2359,11 @@ static int smu_handle_dpm_task(void *handle,
 
 static int smu_switch_power_profile(void *handle,
 				    enum PP_SMC_POWER_PROFILE type,
-				    bool en)
+				    bool enable)
 {
 	struct smu_context *smu = handle;
 	struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
-	long workload[1];
-	uint32_t index;
+	int ret;
 
 	if (!smu->pm_enabled || !smu->adev->pm.dpm_enabled)
 		return -EOPNOTSUPP;
@@ -2365,24 +2371,21 @@ static int smu_switch_power_profile(void *handle,
 	if (!(type < PP_SMC_POWER_PROFILE_CUSTOM))
 		return -EINVAL;
 
-	if (!en) {
-		smu->driver_workload_mask &= ~(1 << smu->workload_priority[type]);
-		index = fls(smu->workload_mask);
-		index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0;
-		workload[0] = smu->workload_setting[index];
-	} else {
-		smu->driver_workload_mask |= (1 << smu->workload_priority[type]);
-		index = fls(smu->workload_mask);
-		index = index <= WORKLOAD_POLICY_MAX ? index - 1 : 0;
-		workload[0] = smu->workload_setting[index];
-	}
-
-	smu->workload_mask = smu->driver_workload_mask |
-						 smu->user_dpm_profile.user_workload_mask;
-
 	if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL &&
-		smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PERF_DETERMINISM)
-		smu_bump_power_profile_mode(smu, workload, 0);
+	    smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PERF_DETERMINISM) {
+		if (enable)
+			smu_power_profile_mode_get(smu, type);
+		else
+			smu_power_profile_mode_put(smu, type);
+		ret = smu_bump_power_profile_mode(smu, NULL, 0);
+		if (ret) {
+			if (enable)
+				smu_power_profile_mode_put(smu, type);
+			else
+				smu_power_profile_mode_get(smu, type);
+			return ret;
+		}
+	}
 
 	return 0;
 }
@@ -3074,21 +3077,33 @@ static int smu_set_power_profile_mode(void *handle,
 				      uint32_t param_size)
 {
 	struct smu_context *smu = handle;
-	int ret;
+	bool custom = false;
+	int ret = 0;
 
 	if (!smu->pm_enabled || !smu->adev->pm.dpm_enabled ||
 	    !smu->ppt_funcs->set_power_profile_mode)
 		return -EOPNOTSUPP;
 
-	if (smu->user_dpm_profile.user_workload_mask &
-	   (1 << smu->workload_priority[param[param_size]]))
-	   return 0;
+	if (param[param_size] == PP_SMC_POWER_PROFILE_CUSTOM) {
+		custom = true;
+		/* clear frontend mask so custom changes propogate */
+		smu->workload_mask = 0;
+	}
 
-	smu->user_dpm_profile.user_workload_mask =
-		(1 << smu->workload_priority[param[param_size]]);
-	smu->workload_mask = smu->user_dpm_profile.user_workload_mask |
-		smu->driver_workload_mask;
-	ret = smu_bump_power_profile_mode(smu, param, param_size);
+	if ((param[param_size] != smu->power_profile_mode) || custom) {
+		/* clear the old user preference */
+		smu_power_profile_mode_put(smu, smu->power_profile_mode);
+		/* set the new user preference */
+		smu_power_profile_mode_get(smu, param[param_size]);
+		ret = smu_bump_power_profile_mode(smu,
+						  custom ? param : NULL,
+						  custom ? param_size : 0);
+		if (ret)
+			smu_power_profile_mode_put(smu, param[param_size]);
+		else
+			/* store the user's preference */
+			smu->power_profile_mode = param[param_size];
+	}
 
 	return ret;
 }
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 d665c47f19b7..3925815358ce 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
@@ -240,7 +240,6 @@ struct smu_user_dpm_profile {
 	/* user clock state information */
 	uint32_t clk_mask[SMU_CLK_COUNT];
 	uint32_t clk_dependency;
-	uint32_t user_workload_mask;
 };
 
 #define SMU_TABLE_INIT(tables, table_id, s, a, d)	\
@@ -557,12 +556,13 @@ struct smu_context {
 	uint32_t hard_min_uclk_req_from_dal;
 	bool disable_uclk_switch;
 
+	/* asic agnostic workload mask */
 	uint32_t workload_mask;
-	uint32_t driver_workload_mask;
-	uint32_t workload_priority[WORKLOAD_POLICY_MAX];
-	uint32_t workload_setting[WORKLOAD_POLICY_MAX];
+	/* default/user workload preference */
 	uint32_t power_profile_mode;
-	uint32_t default_power_profile_mode;
+	uint32_t workload_refcount[PP_SMC_POWER_PROFILE_COUNT];
+	/* backend specific custom workload settings */
+	long *custom_profile_params;
 	bool pm_enabled;
 	bool is_apu;
 
@@ -733,9 +733,12 @@ struct pptable_funcs {
 	 * @set_power_profile_mode: Set a power profile mode. Also used to
 	 *                          create/set custom power profile modes.
 	 * &input: Power profile mode parameters.
-	 * &size: Size of &input.
+	 * &workload_mask: mask of workloads to enable
+	 * &custom_params: custom profile parameters
+	 * &custom_params_max_idx: max valid idx into custom_params
 	 */
-	int (*set_power_profile_mode)(struct smu_context *smu, long *input, uint32_t size);
+	int (*set_power_profile_mode)(struct smu_context *smu, u32 workload_mask,
+				      long *custom_params, u32 custom_params_max_idx);
 
 	/**
 	 * @dpm_set_vcn_enable: Enable/disable VCN engine dynamic power
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c
index 12125303bb79..8aa61a9f7778 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c
@@ -1445,97 +1445,120 @@ static int arcturus_get_power_profile_mode(struct smu_context *smu,
 	return size;
 }
 
-static int arcturus_set_power_profile_mode(struct smu_context *smu,
-					   long *input,
-					   uint32_t size)
+#define ARCTURUS_CUSTOM_PARAMS_COUNT 10
+#define ARCTURUS_CUSTOM_PARAMS_CLOCK_COUNT 2
+#define ARCTURUS_CUSTOM_PARAMS_SIZE (ARCTURUS_CUSTOM_PARAMS_CLOCK_COUNT * ARCTURUS_CUSTOM_PARAMS_COUNT * sizeof(long))
+
+static int arcturus_set_power_profile_mode_coeff(struct smu_context *smu,
+						 long *input)
 {
 	DpmActivityMonitorCoeffInt_t activity_monitor;
-	int workload_type = 0;
-	uint32_t profile_mode = input[size];
-	int ret = 0;
+	int ret, idx;
 
-	if (profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", profile_mode);
-		return -EINVAL;
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
+				   WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor),
+				   false);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
+		return ret;
 	}
 
-	if ((profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) &&
-	     (smu->smc_fw_version >= 0x360d00)) {
-		if (size != 10)
-			return -EINVAL;
+	idx = 0 * ARCTURUS_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Gfxclk */
+		activity_monitor.Gfx_FPS = input[idx + 1];
+		activity_monitor.Gfx_UseRlcBusy = input[idx + 2];
+		activity_monitor.Gfx_MinActiveFreqType = input[idx + 3];
+		activity_monitor.Gfx_MinActiveFreq = input[idx + 4];
+		activity_monitor.Gfx_BoosterFreqType = input[idx + 5];
+		activity_monitor.Gfx_BoosterFreq = input[idx + 6];
+		activity_monitor.Gfx_PD_Data_limit_c = input[idx + 7];
+		activity_monitor.Gfx_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor.Gfx_PD_Data_error_rate_coeff = input[idx + 9];
+	}
+	idx = 1 * ARCTURUS_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Uclk */
+		activity_monitor.Mem_FPS = input[idx + 1];
+		activity_monitor.Mem_UseRlcBusy = input[idx + 2];
+		activity_monitor.Mem_MinActiveFreqType = input[idx + 3];
+		activity_monitor.Mem_MinActiveFreq = input[idx + 4];
+		activity_monitor.Mem_BoosterFreqType = input[idx + 5];
+		activity_monitor.Mem_BoosterFreq = input[idx + 6];
+		activity_monitor.Mem_PD_Data_limit_c = input[idx + 7];
+		activity_monitor.Mem_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor.Mem_PD_Data_error_rate_coeff = input[idx + 9];
+	}
 
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF,
-				       WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor),
-				       false);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
-			return ret;
-		}
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
+				   WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor),
+				   true);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+		return ret;
+	}
 
-		switch (input[0]) {
-		case 0: /* Gfxclk */
-			activity_monitor.Gfx_FPS = input[1];
-			activity_monitor.Gfx_UseRlcBusy = input[2];
-			activity_monitor.Gfx_MinActiveFreqType = input[3];
-			activity_monitor.Gfx_MinActiveFreq = input[4];
-			activity_monitor.Gfx_BoosterFreqType = input[5];
-			activity_monitor.Gfx_BoosterFreq = input[6];
-			activity_monitor.Gfx_PD_Data_limit_c = input[7];
-			activity_monitor.Gfx_PD_Data_error_coeff = input[8];
-			activity_monitor.Gfx_PD_Data_error_rate_coeff = input[9];
-			break;
-		case 1: /* Uclk */
-			activity_monitor.Mem_FPS = input[1];
-			activity_monitor.Mem_UseRlcBusy = input[2];
-			activity_monitor.Mem_MinActiveFreqType = input[3];
-			activity_monitor.Mem_MinActiveFreq = input[4];
-			activity_monitor.Mem_BoosterFreqType = input[5];
-			activity_monitor.Mem_BoosterFreq = input[6];
-			activity_monitor.Mem_PD_Data_limit_c = input[7];
-			activity_monitor.Mem_PD_Data_error_coeff = input[8];
-			activity_monitor.Mem_PD_Data_error_rate_coeff = input[9];
-			break;
-		default:
+	return ret;
+}
+
+static int arcturus_set_power_profile_mode(struct smu_context *smu,
+					   u32 workload_mask,
+					   long *custom_params,
+					   u32 custom_params_max_idx)
+{
+	u32 backend_workload_mask = 0;
+	int ret, idx = -1, i;
+
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
+
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_CUSTOM)) {
+		if (smu->smc_fw_version < 0x360d00)
 			return -EINVAL;
+		if (!smu->custom_profile_params) {
+			smu->custom_profile_params =
+				kzalloc(ARCTURUS_CUSTOM_PARAMS_SIZE, GFP_KERNEL);
+			if (!smu->custom_profile_params)
+				return -ENOMEM;
 		}
-
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF,
-				       WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor),
-				       true);
+		if (custom_params && custom_params_max_idx) {
+			if (custom_params_max_idx != ARCTURUS_CUSTOM_PARAMS_COUNT)
+				return -EINVAL;
+			if (custom_params[0] >= ARCTURUS_CUSTOM_PARAMS_CLOCK_COUNT)
+				return -EINVAL;
+			idx = custom_params[0] * ARCTURUS_CUSTOM_PARAMS_COUNT;
+			smu->custom_profile_params[idx] = 1;
+			for (i = 1; i < custom_params_max_idx; i++)
+				smu->custom_profile_params[idx + i] = custom_params[i];
+		}
+		ret = arcturus_set_power_profile_mode_coeff(smu,
+							    smu->custom_profile_params);
 		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+			if (idx != -1)
+				smu->custom_profile_params[idx] = 0;
 			return ret;
 		}
-	}
-
-	/*
-	 * Conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT
-	 * Not all profile modes are supported on arcturus.
-	 */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       profile_mode);
-	if (workload_type < 0) {
-		dev_dbg(smu->adev->dev, "Unsupported power profile mode %d on arcturus\n", profile_mode);
-		return -EINVAL;
+	} else if (smu->custom_profile_params) {
+		memset(smu->custom_profile_params, 0, ARCTURUS_CUSTOM_PARAMS_SIZE);
 	}
 
 	ret = smu_cmn_send_smc_msg_with_param(smu,
-					  SMU_MSG_SetWorkloadMask,
-					  smu->workload_mask,
-					  NULL);
+					      SMU_MSG_SetWorkloadMask,
+					      backend_workload_mask,
+					      NULL);
 	if (ret) {
-		dev_err(smu->adev->dev, "Fail to set workload type %d\n", workload_type);
+		dev_err(smu->adev->dev, "Failed to set workload mask 0x%08x\n",
+			workload_mask);
+		if (idx != -1)
+			smu->custom_profile_params[idx] = 0;
 		return ret;
 	}
 
-	smu_cmn_assign_power_profile(smu);
-
-	return 0;
+	return ret;
 }
 
 static int arcturus_set_performance_level(struct smu_context *smu,
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 211635dabed8..7fad5dfb39c4 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c
@@ -2006,90 +2006,122 @@ static int navi10_get_power_profile_mode(struct smu_context *smu, char *buf)
 	return size;
 }
 
-static int navi10_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
+#define NAVI10_CUSTOM_PARAMS_COUNT 10
+#define NAVI10_CUSTOM_PARAMS_CLOCKS_COUNT 3
+#define NAVI10_CUSTOM_PARAMS_SIZE (NAVI10_CUSTOM_PARAMS_CLOCKS_COUNT * NAVI10_CUSTOM_PARAMS_COUNT * sizeof(long))
+
+static int navi10_set_power_profile_mode_coeff(struct smu_context *smu,
+					       long *input)
 {
 	DpmActivityMonitorCoeffInt_t activity_monitor;
-	int workload_type, ret = 0;
+	int ret, idx;
 
-	smu->power_profile_mode = input[size];
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor), false);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
+		return ret;
+	}
 
-	if (smu->power_profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", smu->power_profile_mode);
-		return -EINVAL;
+	idx = 0 * NAVI10_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Gfxclk */
+		activity_monitor.Gfx_FPS = input[idx + 1];
+		activity_monitor.Gfx_MinFreqStep = input[idx + 2];
+		activity_monitor.Gfx_MinActiveFreqType = input[idx + 3];
+		activity_monitor.Gfx_MinActiveFreq = input[idx + 4];
+		activity_monitor.Gfx_BoosterFreqType = input[idx + 5];
+		activity_monitor.Gfx_BoosterFreq = input[idx + 6];
+		activity_monitor.Gfx_PD_Data_limit_c = input[idx + 7];
+		activity_monitor.Gfx_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor.Gfx_PD_Data_error_rate_coeff = input[idx + 9];
+	}
+	idx = 1 * NAVI10_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Socclk */
+		activity_monitor.Soc_FPS = input[idx + 1];
+		activity_monitor.Soc_MinFreqStep = input[idx + 2];
+		activity_monitor.Soc_MinActiveFreqType = input[idx + 3];
+		activity_monitor.Soc_MinActiveFreq = input[idx + 4];
+		activity_monitor.Soc_BoosterFreqType = input[idx + 5];
+		activity_monitor.Soc_BoosterFreq = input[idx + 6];
+		activity_monitor.Soc_PD_Data_limit_c = input[idx + 7];
+		activity_monitor.Soc_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor.Soc_PD_Data_error_rate_coeff = input[idx + 9];
+	}
+	idx = 2 * NAVI10_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Memclk */
+		activity_monitor.Mem_FPS = input[idx + 1];
+		activity_monitor.Mem_MinFreqStep = input[idx + 2];
+		activity_monitor.Mem_MinActiveFreqType = input[idx + 3];
+		activity_monitor.Mem_MinActiveFreq = input[idx + 4];
+		activity_monitor.Mem_BoosterFreqType = input[idx + 5];
+		activity_monitor.Mem_BoosterFreq = input[idx + 6];
+		activity_monitor.Mem_PD_Data_limit_c = input[idx + 7];
+		activity_monitor.Mem_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor.Mem_PD_Data_error_rate_coeff = input[idx + 9];
+	}
+
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor), true);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+		return ret;
 	}
 
-	if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) {
-		if (size != 10)
-			return -EINVAL;
+	return ret;
+}
 
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor), false);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
-			return ret;
-		}
+static int navi10_set_power_profile_mode(struct smu_context *smu,
+					 u32 workload_mask,
+					 long *custom_params,
+					 u32 custom_params_max_idx)
+{
+	u32 backend_workload_mask = 0;
+	int ret, idx = -1, i;
 
-		switch (input[0]) {
-		case 0: /* Gfxclk */
-			activity_monitor.Gfx_FPS = input[1];
-			activity_monitor.Gfx_MinFreqStep = input[2];
-			activity_monitor.Gfx_MinActiveFreqType = input[3];
-			activity_monitor.Gfx_MinActiveFreq = input[4];
-			activity_monitor.Gfx_BoosterFreqType = input[5];
-			activity_monitor.Gfx_BoosterFreq = input[6];
-			activity_monitor.Gfx_PD_Data_limit_c = input[7];
-			activity_monitor.Gfx_PD_Data_error_coeff = input[8];
-			activity_monitor.Gfx_PD_Data_error_rate_coeff = input[9];
-			break;
-		case 1: /* Socclk */
-			activity_monitor.Soc_FPS = input[1];
-			activity_monitor.Soc_MinFreqStep = input[2];
-			activity_monitor.Soc_MinActiveFreqType = input[3];
-			activity_monitor.Soc_MinActiveFreq = input[4];
-			activity_monitor.Soc_BoosterFreqType = input[5];
-			activity_monitor.Soc_BoosterFreq = input[6];
-			activity_monitor.Soc_PD_Data_limit_c = input[7];
-			activity_monitor.Soc_PD_Data_error_coeff = input[8];
-			activity_monitor.Soc_PD_Data_error_rate_coeff = input[9];
-			break;
-		case 2: /* Memclk */
-			activity_monitor.Mem_FPS = input[1];
-			activity_monitor.Mem_MinFreqStep = input[2];
-			activity_monitor.Mem_MinActiveFreqType = input[3];
-			activity_monitor.Mem_MinActiveFreq = input[4];
-			activity_monitor.Mem_BoosterFreqType = input[5];
-			activity_monitor.Mem_BoosterFreq = input[6];
-			activity_monitor.Mem_PD_Data_limit_c = input[7];
-			activity_monitor.Mem_PD_Data_error_coeff = input[8];
-			activity_monitor.Mem_PD_Data_error_rate_coeff = input[9];
-			break;
-		default:
-			return -EINVAL;
-		}
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
 
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor), true);
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_CUSTOM)) {
+		if (!smu->custom_profile_params) {
+			smu->custom_profile_params = kzalloc(NAVI10_CUSTOM_PARAMS_SIZE, GFP_KERNEL);
+			if (!smu->custom_profile_params)
+				return -ENOMEM;
+		}
+		if (custom_params && custom_params_max_idx) {
+			if (custom_params_max_idx != NAVI10_CUSTOM_PARAMS_COUNT)
+				return -EINVAL;
+			if (custom_params[0] >= NAVI10_CUSTOM_PARAMS_CLOCKS_COUNT)
+				return -EINVAL;
+			idx = custom_params[0] * NAVI10_CUSTOM_PARAMS_COUNT;
+			smu->custom_profile_params[idx] = 1;
+			for (i = 1; i < custom_params_max_idx; i++)
+				smu->custom_profile_params[idx + i] = custom_params[i];
+		}
+		ret = navi10_set_power_profile_mode_coeff(smu,
+							  smu->custom_profile_params);
 		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+			if (idx != -1)
+				smu->custom_profile_params[idx] = 0;
 			return ret;
 		}
+	} else if (smu->custom_profile_params) {
+		memset(smu->custom_profile_params, 0, NAVI10_CUSTOM_PARAMS_SIZE);
 	}
 
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       smu->power_profile_mode);
-	if (workload_type < 0)
-		return -EINVAL;
-
 	ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask,
-				    smu->workload_mask, NULL);
-	if (ret)
-		dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__);
-	else
-		smu_cmn_assign_power_profile(smu);
+					      backend_workload_mask, NULL);
+	if (ret) {
+		dev_err(smu->adev->dev, "Failed to set workload mask 0x%08x\n",
+			workload_mask);
+		if (idx != -1)
+			smu->custom_profile_params[idx] = 0;
+		return ret;
+	}
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c
index d0ed0d060a8a..286777ada1df 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c
@@ -1708,93 +1708,126 @@ static int sienna_cichlid_get_power_profile_mode(struct smu_context *smu, char *
 	return size;
 }
 
-static int sienna_cichlid_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
+#define SIENNA_CICHLID_CUSTOM_PARAMS_COUNT 10
+#define SIENNA_CICHLID_CUSTOM_PARAMS_CLOCK_COUNT 3
+#define SIENNA_CICHLID_CUSTOM_PARAMS_SIZE (SIENNA_CICHLID_CUSTOM_PARAMS_CLOCK_COUNT * SIENNA_CICHLID_CUSTOM_PARAMS_COUNT * sizeof(long))
+
+static int sienna_cichlid_set_power_profile_mode_coeff(struct smu_context *smu,
+						       long *input)
 {
 
 	DpmActivityMonitorCoeffIntExternal_t activity_monitor_external;
 	DpmActivityMonitorCoeffInt_t *activity_monitor =
 		&(activity_monitor_external.DpmActivityMonitorCoeffInt);
-	int workload_type, ret = 0;
+	int ret, idx;
 
-	smu->power_profile_mode = input[size];
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external), false);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
+		return ret;
+	}
 
-	if (smu->power_profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", smu->power_profile_mode);
-		return -EINVAL;
+	idx = 0 * SIENNA_CICHLID_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Gfxclk */
+		activity_monitor->Gfx_FPS = input[idx + 1];
+		activity_monitor->Gfx_MinFreqStep = input[idx + 2];
+		activity_monitor->Gfx_MinActiveFreqType = input[idx + 3];
+		activity_monitor->Gfx_MinActiveFreq = input[idx + 4];
+		activity_monitor->Gfx_BoosterFreqType = input[idx + 5];
+		activity_monitor->Gfx_BoosterFreq = input[idx + 6];
+		activity_monitor->Gfx_PD_Data_limit_c = input[idx + 7];
+		activity_monitor->Gfx_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor->Gfx_PD_Data_error_rate_coeff = input[idx + 9];
+	}
+	idx = 1 * SIENNA_CICHLID_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Socclk */
+		activity_monitor->Fclk_FPS = input[idx + 1];
+		activity_monitor->Fclk_MinFreqStep = input[idx + 2];
+		activity_monitor->Fclk_MinActiveFreqType = input[idx + 3];
+		activity_monitor->Fclk_MinActiveFreq = input[idx + 4];
+		activity_monitor->Fclk_BoosterFreqType = input[idx + 5];
+		activity_monitor->Fclk_BoosterFreq = input[idx + 6];
+		activity_monitor->Fclk_PD_Data_limit_c = input[idx + 7];
+		activity_monitor->Fclk_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor->Fclk_PD_Data_error_rate_coeff = input[idx + 9];
+	}
+	idx = 2 * SIENNA_CICHLID_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Memclk */
+		activity_monitor->Mem_FPS = input[idx + 1];
+		activity_monitor->Mem_MinFreqStep = input[idx + 2];
+		activity_monitor->Mem_MinActiveFreqType = input[idx + 3];
+		activity_monitor->Mem_MinActiveFreq = input[idx + 4];
+		activity_monitor->Mem_BoosterFreqType = input[idx + 5];
+		activity_monitor->Mem_BoosterFreq = input[idx + 6];
+		activity_monitor->Mem_PD_Data_limit_c = input[idx + 7];
+		activity_monitor->Mem_PD_Data_error_coeff = input[idx + 8];
+		activity_monitor->Mem_PD_Data_error_rate_coeff = input[idx + 9];
 	}
 
-	if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) {
-		if (size != 10)
-			return -EINVAL;
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external), true);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+		return ret;
+	}
 
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor_external), false);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
-			return ret;
-		}
+	return ret;
+}
 
-		switch (input[0]) {
-		case 0: /* Gfxclk */
-			activity_monitor->Gfx_FPS = input[1];
-			activity_monitor->Gfx_MinFreqStep = input[2];
-			activity_monitor->Gfx_MinActiveFreqType = input[3];
-			activity_monitor->Gfx_MinActiveFreq = input[4];
-			activity_monitor->Gfx_BoosterFreqType = input[5];
-			activity_monitor->Gfx_BoosterFreq = input[6];
-			activity_monitor->Gfx_PD_Data_limit_c = input[7];
-			activity_monitor->Gfx_PD_Data_error_coeff = input[8];
-			activity_monitor->Gfx_PD_Data_error_rate_coeff = input[9];
-			break;
-		case 1: /* Socclk */
-			activity_monitor->Fclk_FPS = input[1];
-			activity_monitor->Fclk_MinFreqStep = input[2];
-			activity_monitor->Fclk_MinActiveFreqType = input[3];
-			activity_monitor->Fclk_MinActiveFreq = input[4];
-			activity_monitor->Fclk_BoosterFreqType = input[5];
-			activity_monitor->Fclk_BoosterFreq = input[6];
-			activity_monitor->Fclk_PD_Data_limit_c = input[7];
-			activity_monitor->Fclk_PD_Data_error_coeff = input[8];
-			activity_monitor->Fclk_PD_Data_error_rate_coeff = input[9];
-			break;
-		case 2: /* Memclk */
-			activity_monitor->Mem_FPS = input[1];
-			activity_monitor->Mem_MinFreqStep = input[2];
-			activity_monitor->Mem_MinActiveFreqType = input[3];
-			activity_monitor->Mem_MinActiveFreq = input[4];
-			activity_monitor->Mem_BoosterFreqType = input[5];
-			activity_monitor->Mem_BoosterFreq = input[6];
-			activity_monitor->Mem_PD_Data_limit_c = input[7];
-			activity_monitor->Mem_PD_Data_error_coeff = input[8];
-			activity_monitor->Mem_PD_Data_error_rate_coeff = input[9];
-			break;
-		default:
-			return -EINVAL;
-		}
+static int sienna_cichlid_set_power_profile_mode(struct smu_context *smu,
+						 u32 workload_mask,
+						 long *custom_params,
+						 u32 custom_params_max_idx)
+{
+	u32 backend_workload_mask = 0;
+	int ret, idx = -1, i;
+
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
 
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor_external), true);
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_CUSTOM)) {
+		if (!smu->custom_profile_params) {
+			smu->custom_profile_params =
+				kzalloc(SIENNA_CICHLID_CUSTOM_PARAMS_SIZE, GFP_KERNEL);
+			if (!smu->custom_profile_params)
+				return -ENOMEM;
+		}
+		if (custom_params && custom_params_max_idx) {
+			if (custom_params_max_idx != SIENNA_CICHLID_CUSTOM_PARAMS_COUNT)
+				return -EINVAL;
+			if (custom_params[0] >= SIENNA_CICHLID_CUSTOM_PARAMS_CLOCK_COUNT)
+				return -EINVAL;
+			idx = custom_params[0] * SIENNA_CICHLID_CUSTOM_PARAMS_COUNT;
+			smu->custom_profile_params[idx] = 1;
+			for (i = 1; i < custom_params_max_idx; i++)
+				smu->custom_profile_params[idx + i] = custom_params[i];
+		}
+		ret = sienna_cichlid_set_power_profile_mode_coeff(smu,
+								  smu->custom_profile_params);
 		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+			if (idx != -1)
+				smu->custom_profile_params[idx] = 0;
 			return ret;
 		}
+	} else if (smu->custom_profile_params) {
+		memset(smu->custom_profile_params, 0, SIENNA_CICHLID_CUSTOM_PARAMS_SIZE);
 	}
 
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       smu->power_profile_mode);
-	if (workload_type < 0)
-		return -EINVAL;
-
 	ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask,
-				    smu->workload_mask, NULL);
-	if (ret)
-		dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__);
-	else
-		smu_cmn_assign_power_profile(smu);
+					      backend_workload_mask, NULL);
+	if (ret) {
+		dev_err(smu->adev->dev, "Failed to set workload mask 0x%08x\n",
+			workload_mask);
+		if (idx != -1)
+			smu->custom_profile_params[idx] = 0;
+		return ret;
+	}
 
 	return ret;
 }
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 f89c487dce72..a55ea76d7399 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
@@ -1056,42 +1056,27 @@ static int vangogh_get_power_profile_mode(struct smu_context *smu,
 	return size;
 }
 
-static int vangogh_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
+static int vangogh_set_power_profile_mode(struct smu_context *smu,
+					  u32 workload_mask,
+					  long *custom_params,
+					  u32 custom_params_max_idx)
 {
-	int workload_type, ret;
-	uint32_t profile_mode = input[size];
+	u32 backend_workload_mask = 0;
+	int ret;
 
-	if (profile_mode >= PP_SMC_POWER_PROFILE_COUNT) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", profile_mode);
-		return -EINVAL;
-	}
-
-	if (profile_mode == PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT ||
-			profile_mode == PP_SMC_POWER_PROFILE_POWERSAVING)
-		return 0;
-
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       profile_mode);
-	if (workload_type < 0) {
-		dev_dbg(smu->adev->dev, "Unsupported power profile mode %d on VANGOGH\n",
-					profile_mode);
-		return -EINVAL;
-	}
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
 
 	ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ActiveProcessNotify,
-				    smu->workload_mask,
-				    NULL);
+					      backend_workload_mask,
+					      NULL);
 	if (ret) {
-		dev_err_once(smu->adev->dev, "Fail to set workload type %d\n",
-					workload_type);
+		dev_err_once(smu->adev->dev, "Fail to set workload mask 0x%08x\n",
+			     workload_mask);
 		return ret;
 	}
 
-	smu_cmn_assign_power_profile(smu);
-
-	return 0;
+	return ret;
 }
 
 static int vangogh_set_soft_freq_limited_range(struct smu_context *smu,
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c
index 75a9ea87f419..37d82a71a2d7 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c
@@ -864,44 +864,27 @@ static int renoir_force_clk_levels(struct smu_context *smu,
 	return ret;
 }
 
-static int renoir_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
+static int renoir_set_power_profile_mode(struct smu_context *smu,
+					 u32 workload_mask,
+					 long *custom_params,
+					 u32 custom_params_max_idx)
 {
-	int workload_type, ret;
-	uint32_t profile_mode = input[size];
+	int ret;
+	u32 backend_workload_mask = 0;
 
-	if (profile_mode > PP_SMC_POWER_PROFILE_CUSTOM) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", profile_mode);
-		return -EINVAL;
-	}
-
-	if (profile_mode == PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT ||
-			profile_mode == PP_SMC_POWER_PROFILE_POWERSAVING)
-		return 0;
-
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       profile_mode);
-	if (workload_type < 0) {
-		/*
-		 * TODO: If some case need switch to powersave/default power mode
-		 * then can consider enter WORKLOAD_COMPUTE/WORKLOAD_CUSTOM for power saving.
-		 */
-		dev_dbg(smu->adev->dev, "Unsupported power profile mode %d on RENOIR\n", profile_mode);
-		return -EINVAL;
-	}
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
 
 	ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ActiveProcessNotify,
-				    smu->workload_mask,
-				    NULL);
+					      backend_workload_mask,
+					      NULL);
 	if (ret) {
-		dev_err_once(smu->adev->dev, "Fail to set workload type %d\n", workload_type);
+		dev_err_once(smu->adev->dev, "Failed to set workload mask 0x08%x\n",
+			     workload_mask);
 		return ret;
 	}
 
-	smu_cmn_assign_power_profile(smu);
-
-	return 0;
+	return ret;
 }
 
 static int renoir_set_peak_clock_by_device(struct smu_context *smu)
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
index 80c6b1e523aa..3aa705aae4c0 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
@@ -2571,82 +2571,76 @@ static int smu_v13_0_0_get_power_profile_mode(struct smu_context *smu,
 	return size;
 }
 
-static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu,
-					      long *input,
-					      uint32_t size)
+#define SMU_13_0_0_CUSTOM_PARAMS_COUNT 9
+#define SMU_13_0_0_CUSTOM_PARAMS_CLOCK_COUNT 2
+#define SMU_13_0_0_CUSTOM_PARAMS_SIZE (SMU_13_0_0_CUSTOM_PARAMS_CLOCK_COUNT * SMU_13_0_0_CUSTOM_PARAMS_COUNT * sizeof(long))
+
+static int smu_v13_0_0_set_power_profile_mode_coeff(struct smu_context *smu,
+						    long *input)
 {
 	DpmActivityMonitorCoeffIntExternal_t activity_monitor_external;
 	DpmActivityMonitorCoeffInt_t *activity_monitor =
 		&(activity_monitor_external.DpmActivityMonitorCoeffInt);
-	int workload_type, ret = 0;
-	u32 workload_mask;
-
-	smu->power_profile_mode = input[size];
+	int ret, idx;
 
-	if (smu->power_profile_mode >= PP_SMC_POWER_PROFILE_COUNT) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", smu->power_profile_mode);
-		return -EINVAL;
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
+				   WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external),
+				   false);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
+		return ret;
 	}
 
-	if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) {
-		if (size != 9)
-			return -EINVAL;
-
-		ret = smu_cmn_update_table(smu,
-					   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
-					   WORKLOAD_PPLIB_CUSTOM_BIT,
-					   (void *)(&activity_monitor_external),
-					   false);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
-			return ret;
-		}
-
-		switch (input[0]) {
-		case 0: /* Gfxclk */
-			activity_monitor->Gfx_FPS = input[1];
-			activity_monitor->Gfx_MinActiveFreqType = input[2];
-			activity_monitor->Gfx_MinActiveFreq = input[3];
-			activity_monitor->Gfx_BoosterFreqType = input[4];
-			activity_monitor->Gfx_BoosterFreq = input[5];
-			activity_monitor->Gfx_PD_Data_limit_c = input[6];
-			activity_monitor->Gfx_PD_Data_error_coeff = input[7];
-			activity_monitor->Gfx_PD_Data_error_rate_coeff = input[8];
-			break;
-		case 1: /* Fclk */
-			activity_monitor->Fclk_FPS = input[1];
-			activity_monitor->Fclk_MinActiveFreqType = input[2];
-			activity_monitor->Fclk_MinActiveFreq = input[3];
-			activity_monitor->Fclk_BoosterFreqType = input[4];
-			activity_monitor->Fclk_BoosterFreq = input[5];
-			activity_monitor->Fclk_PD_Data_limit_c = input[6];
-			activity_monitor->Fclk_PD_Data_error_coeff = input[7];
-			activity_monitor->Fclk_PD_Data_error_rate_coeff = input[8];
-			break;
-		default:
-			return -EINVAL;
-		}
+	idx = 0 * SMU_13_0_0_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Gfxclk */
+		activity_monitor->Gfx_FPS = input[idx + 1];
+		activity_monitor->Gfx_MinActiveFreqType = input[idx + 2];
+		activity_monitor->Gfx_MinActiveFreq = input[idx + 3];
+		activity_monitor->Gfx_BoosterFreqType = input[idx + 4];
+		activity_monitor->Gfx_BoosterFreq = input[idx + 5];
+		activity_monitor->Gfx_PD_Data_limit_c = input[idx + 6];
+		activity_monitor->Gfx_PD_Data_error_coeff = input[idx + 7];
+		activity_monitor->Gfx_PD_Data_error_rate_coeff = input[idx + 8];
+	}
+	idx = 1 * SMU_13_0_0_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Fclk */
+		activity_monitor->Fclk_FPS = input[idx + 1];
+		activity_monitor->Fclk_MinActiveFreqType = input[idx + 2];
+		activity_monitor->Fclk_MinActiveFreq = input[idx + 3];
+		activity_monitor->Fclk_BoosterFreqType = input[idx + 4];
+		activity_monitor->Fclk_BoosterFreq = input[idx + 5];
+		activity_monitor->Fclk_PD_Data_limit_c = input[idx + 6];
+		activity_monitor->Fclk_PD_Data_error_coeff = input[idx + 7];
+		activity_monitor->Fclk_PD_Data_error_rate_coeff = input[idx + 8];
+	}
 
-		ret = smu_cmn_update_table(smu,
-					   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
-					   WORKLOAD_PPLIB_CUSTOM_BIT,
-					   (void *)(&activity_monitor_external),
-					   true);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
-			return ret;
-		}
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
+				   WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external),
+				   true);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+		return ret;
 	}
 
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       smu->power_profile_mode);
+	return ret;
+}
 
-	if (workload_type < 0)
-		return -EINVAL;
+static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu,
+					      u32 workload_mask,
+					      long *custom_params,
+					      u32 custom_params_max_idx)
+{
+	u32 backend_workload_mask = 0;
+	int workload_type, ret, idx = -1, i;
 
-	workload_mask = 1 << workload_type;
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
 
 	/* Add optimizations for SMU13.0.0/10.  Reuse the power saving profile */
 	if ((amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 0) &&
@@ -2658,24 +2652,47 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu,
 							       CMN2ASIC_MAPPING_WORKLOAD,
 							       PP_SMC_POWER_PROFILE_POWERSAVING);
 		if (workload_type >= 0)
-			workload_mask |= 1 << workload_type;
+			backend_workload_mask |= 1 << workload_type;
 	}
 
-	smu->workload_mask |= workload_mask;
-	ret = smu_cmn_send_smc_msg_with_param(smu,
-					       SMU_MSG_SetWorkloadMask,
-					       smu->workload_mask,
-					       NULL);
-	if (!ret) {
-		smu_cmn_assign_power_profile(smu);
-		if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_POWERSAVING) {
-			workload_type = smu_cmn_to_asic_specific_index(smu,
-							       CMN2ASIC_MAPPING_WORKLOAD,
-							       PP_SMC_POWER_PROFILE_FULLSCREEN3D);
-			smu->power_profile_mode = smu->workload_mask & (1 << workload_type)
-										? PP_SMC_POWER_PROFILE_FULLSCREEN3D
-										: PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_CUSTOM)) {
+		if (!smu->custom_profile_params) {
+			smu->custom_profile_params =
+				kzalloc(SMU_13_0_0_CUSTOM_PARAMS_SIZE, GFP_KERNEL);
+			if (!smu->custom_profile_params)
+				return -ENOMEM;
 		}
+		if (custom_params && custom_params_max_idx) {
+			if (custom_params_max_idx != SMU_13_0_0_CUSTOM_PARAMS_COUNT)
+				return -EINVAL;
+			if (custom_params[0] >= SMU_13_0_0_CUSTOM_PARAMS_CLOCK_COUNT)
+				return -EINVAL;
+			idx = custom_params[0] * SMU_13_0_0_CUSTOM_PARAMS_COUNT;
+			smu->custom_profile_params[idx] = 1;
+			for (i = 1; i < custom_params_max_idx; i++)
+				smu->custom_profile_params[idx + i] = custom_params[i];
+		}
+		ret = smu_v13_0_0_set_power_profile_mode_coeff(smu,
+							       smu->custom_profile_params);
+		if (ret) {
+			if (idx != -1)
+				smu->custom_profile_params[idx] = 0;
+			return ret;
+		}
+	} else if (smu->custom_profile_params) {
+		memset(smu->custom_profile_params, 0, SMU_13_0_0_CUSTOM_PARAMS_SIZE);
+	}
+
+	ret = smu_cmn_send_smc_msg_with_param(smu,
+					      SMU_MSG_SetWorkloadMask,
+					      backend_workload_mask,
+					      NULL);
+	if (ret) {
+		dev_err(smu->adev->dev, "Failed to set workload mask 0x%08x\n",
+			workload_mask);
+		if (idx != -1)
+			smu->custom_profile_params[idx] = 0;
+		return ret;
 	}
 
 	return ret;
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
index 4fd0354bd312..f4ac403b8b36 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
@@ -2530,79 +2530,110 @@ out:
 	return result;
 }
 
-static int smu_v13_0_7_set_power_profile_mode(struct smu_context *smu, long *input, uint32_t size)
+#define SMU_13_0_7_CUSTOM_PARAMS_COUNT 8
+#define SMU_13_0_7_CUSTOM_PARAMS_CLOCK_COUNT 2
+#define SMU_13_0_7_CUSTOM_PARAMS_SIZE (SMU_13_0_7_CUSTOM_PARAMS_CLOCK_COUNT * SMU_13_0_7_CUSTOM_PARAMS_COUNT * sizeof(long))
+
+static int smu_v13_0_7_set_power_profile_mode_coeff(struct smu_context *smu,
+						    long *input)
 {
 
 	DpmActivityMonitorCoeffIntExternal_t activity_monitor_external;
 	DpmActivityMonitorCoeffInt_t *activity_monitor =
 		&(activity_monitor_external.DpmActivityMonitorCoeffInt);
-	int workload_type, ret = 0;
+	int ret, idx;
 
-	smu->power_profile_mode = input[size];
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external), false);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
+		return ret;
+	}
 
-	if (smu->power_profile_mode > PP_SMC_POWER_PROFILE_WINDOW3D) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", smu->power_profile_mode);
-		return -EINVAL;
+	idx = 0 * SMU_13_0_7_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Gfxclk */
+		activity_monitor->Gfx_ActiveHystLimit = input[idx + 1];
+		activity_monitor->Gfx_IdleHystLimit = input[idx + 2];
+		activity_monitor->Gfx_FPS = input[idx + 3];
+		activity_monitor->Gfx_MinActiveFreqType = input[idx + 4];
+		activity_monitor->Gfx_BoosterFreqType = input[idx + 5];
+		activity_monitor->Gfx_MinActiveFreq = input[idx + 6];
+		activity_monitor->Gfx_BoosterFreq = input[idx + 7];
+	}
+	idx = 1 * SMU_13_0_7_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Fclk */
+		activity_monitor->Fclk_ActiveHystLimit = input[idx + 1];
+		activity_monitor->Fclk_IdleHystLimit = input[idx + 2];
+		activity_monitor->Fclk_FPS = input[idx + 3];
+		activity_monitor->Fclk_MinActiveFreqType = input[idx + 4];
+		activity_monitor->Fclk_BoosterFreqType = input[idx + 5];
+		activity_monitor->Fclk_MinActiveFreq = input[idx + 6];
+		activity_monitor->Fclk_BoosterFreq = input[idx + 7];
 	}
 
-	if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) {
-		if (size != 8)
-			return -EINVAL;
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external), true);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+		return ret;
+	}
 
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor_external), false);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
-			return ret;
-		}
+	return ret;
+}
 
-		switch (input[0]) {
-		case 0: /* Gfxclk */
-			activity_monitor->Gfx_ActiveHystLimit = input[1];
-			activity_monitor->Gfx_IdleHystLimit = input[2];
-			activity_monitor->Gfx_FPS = input[3];
-			activity_monitor->Gfx_MinActiveFreqType = input[4];
-			activity_monitor->Gfx_BoosterFreqType = input[5];
-			activity_monitor->Gfx_MinActiveFreq = input[6];
-			activity_monitor->Gfx_BoosterFreq = input[7];
-			break;
-		case 1: /* Fclk */
-			activity_monitor->Fclk_ActiveHystLimit = input[1];
-			activity_monitor->Fclk_IdleHystLimit = input[2];
-			activity_monitor->Fclk_FPS = input[3];
-			activity_monitor->Fclk_MinActiveFreqType = input[4];
-			activity_monitor->Fclk_BoosterFreqType = input[5];
-			activity_monitor->Fclk_MinActiveFreq = input[6];
-			activity_monitor->Fclk_BoosterFreq = input[7];
-			break;
-		default:
-			return -EINVAL;
+static int smu_v13_0_7_set_power_profile_mode(struct smu_context *smu,
+					      u32 workload_mask,
+					      long *custom_params,
+					      u32 custom_params_max_idx)
+{
+	u32 backend_workload_mask = 0;
+	int ret, idx = -1, i;
+
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
+
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_CUSTOM)) {
+		if (!smu->custom_profile_params) {
+			smu->custom_profile_params =
+				kzalloc(SMU_13_0_7_CUSTOM_PARAMS_SIZE, GFP_KERNEL);
+			if (!smu->custom_profile_params)
+				return -ENOMEM;
 		}
-
-		ret = smu_cmn_update_table(smu,
-				       SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT,
-				       (void *)(&activity_monitor_external), true);
+		if (custom_params && custom_params_max_idx) {
+			if (custom_params_max_idx != SMU_13_0_7_CUSTOM_PARAMS_COUNT)
+				return -EINVAL;
+			if (custom_params[0] >= SMU_13_0_7_CUSTOM_PARAMS_CLOCK_COUNT)
+				return -EINVAL;
+			idx = custom_params[0] * SMU_13_0_7_CUSTOM_PARAMS_COUNT;
+			smu->custom_profile_params[idx] = 1;
+			for (i = 1; i < custom_params_max_idx; i++)
+				smu->custom_profile_params[idx + i] = custom_params[i];
+		}
+		ret = smu_v13_0_7_set_power_profile_mode_coeff(smu,
+							       smu->custom_profile_params);
 		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+			if (idx != -1)
+				smu->custom_profile_params[idx] = 0;
 			return ret;
 		}
+	} else if (smu->custom_profile_params) {
+		memset(smu->custom_profile_params, 0, SMU_13_0_7_CUSTOM_PARAMS_SIZE);
 	}
 
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       smu->power_profile_mode);
-	if (workload_type < 0)
-		return -EINVAL;
-
 	ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask,
-				    smu->workload_mask, NULL);
+					      backend_workload_mask, NULL);
 
-	if (ret)
-		dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__);
-	else
-		smu_cmn_assign_power_profile(smu);
+	if (ret) {
+		dev_err(smu->adev->dev, "Failed to set workload mask 0x%08x\n",
+			workload_mask);
+		if (idx != -1)
+			smu->custom_profile_params[idx] = 0;
+		return ret;
+	}
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c
index 687a0f5ac94f..6a565ce74d5b 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c
@@ -1739,89 +1739,120 @@ static int smu_v14_0_2_get_power_profile_mode(struct smu_context *smu,
 	return size;
 }
 
-static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu,
-					      long *input,
-					      uint32_t size)
+#define SMU_14_0_2_CUSTOM_PARAMS_COUNT 9
+#define SMU_14_0_2_CUSTOM_PARAMS_CLOCK_COUNT 2
+#define SMU_14_0_2_CUSTOM_PARAMS_SIZE (SMU_14_0_2_CUSTOM_PARAMS_CLOCK_COUNT * SMU_14_0_2_CUSTOM_PARAMS_COUNT * sizeof(long))
+
+static int smu_v14_0_2_set_power_profile_mode_coeff(struct smu_context *smu,
+						    long *input)
 {
 	DpmActivityMonitorCoeffIntExternal_t activity_monitor_external;
 	DpmActivityMonitorCoeffInt_t *activity_monitor =
 		&(activity_monitor_external.DpmActivityMonitorCoeffInt);
-	int workload_type, ret = 0;
-	uint32_t current_profile_mode = smu->power_profile_mode;
-	smu->power_profile_mode = input[size];
+	int ret, idx;
 
-	if (smu->power_profile_mode >= PP_SMC_POWER_PROFILE_COUNT) {
-		dev_err(smu->adev->dev, "Invalid power profile mode %d\n", smu->power_profile_mode);
-		return -EINVAL;
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
+				   WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external),
+				   false);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
+		return ret;
 	}
 
-	if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) {
-		if (size != 9)
-			return -EINVAL;
+	idx = 0 * SMU_14_0_2_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Gfxclk */
+		activity_monitor->Gfx_FPS = input[idx + 1];
+		activity_monitor->Gfx_MinActiveFreqType = input[idx + 2];
+		activity_monitor->Gfx_MinActiveFreq = input[idx + 3];
+		activity_monitor->Gfx_BoosterFreqType = input[idx + 4];
+		activity_monitor->Gfx_BoosterFreq = input[idx + 5];
+		activity_monitor->Gfx_PD_Data_limit_c = input[idx + 6];
+		activity_monitor->Gfx_PD_Data_error_coeff = input[idx + 7];
+		activity_monitor->Gfx_PD_Data_error_rate_coeff = input[idx + 8];
+	}
+	idx = 1 * SMU_14_0_2_CUSTOM_PARAMS_COUNT;
+	if (input[idx]) {
+		/* Fclk */
+		activity_monitor->Fclk_FPS = input[idx + 1];
+		activity_monitor->Fclk_MinActiveFreqType = input[idx + 2];
+		activity_monitor->Fclk_MinActiveFreq = input[idx + 3];
+		activity_monitor->Fclk_BoosterFreqType = input[idx + 4];
+		activity_monitor->Fclk_BoosterFreq = input[idx + 5];
+		activity_monitor->Fclk_PD_Data_limit_c = input[idx + 6];
+		activity_monitor->Fclk_PD_Data_error_coeff = input[idx + 7];
+		activity_monitor->Fclk_PD_Data_error_rate_coeff = input[idx + 8];
+	}
 
-		ret = smu_cmn_update_table(smu,
-					   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
-					   WORKLOAD_PPLIB_CUSTOM_BIT,
-					   (void *)(&activity_monitor_external),
-					   false);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__);
-			return ret;
-		}
+	ret = smu_cmn_update_table(smu,
+				   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
+				   WORKLOAD_PPLIB_CUSTOM_BIT,
+				   (void *)(&activity_monitor_external),
+				   true);
+	if (ret) {
+		dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
+		return ret;
+	}
 
-		switch (input[0]) {
-		case 0: /* Gfxclk */
-			activity_monitor->Gfx_FPS = input[1];
-			activity_monitor->Gfx_MinActiveFreqType = input[2];
-			activity_monitor->Gfx_MinActiveFreq = input[3];
-			activity_monitor->Gfx_BoosterFreqType = input[4];
-			activity_monitor->Gfx_BoosterFreq = input[5];
-			activity_monitor->Gfx_PD_Data_limit_c = input[6];
-			activity_monitor->Gfx_PD_Data_error_coeff = input[7];
-			activity_monitor->Gfx_PD_Data_error_rate_coeff = input[8];
-			break;
-		case 1: /* Fclk */
-			activity_monitor->Fclk_FPS = input[1];
-			activity_monitor->Fclk_MinActiveFreqType = input[2];
-			activity_monitor->Fclk_MinActiveFreq = input[3];
-			activity_monitor->Fclk_BoosterFreqType = input[4];
-			activity_monitor->Fclk_BoosterFreq = input[5];
-			activity_monitor->Fclk_PD_Data_limit_c = input[6];
-			activity_monitor->Fclk_PD_Data_error_coeff = input[7];
-			activity_monitor->Fclk_PD_Data_error_rate_coeff = input[8];
-			break;
-		default:
-			return -EINVAL;
-		}
+	return ret;
+}
 
-		ret = smu_cmn_update_table(smu,
-					   SMU_TABLE_ACTIVITY_MONITOR_COEFF,
-					   WORKLOAD_PPLIB_CUSTOM_BIT,
-					   (void *)(&activity_monitor_external),
-					   true);
-		if (ret) {
-			dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__);
-			return ret;
-		}
-	}
+static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu,
+					      u32 workload_mask,
+					      long *custom_params,
+					      u32 custom_params_max_idx)
+{
+	u32 backend_workload_mask = 0;
+	int ret, idx = -1, i;
+
+	smu_cmn_get_backend_workload_mask(smu, workload_mask,
+					  &backend_workload_mask);
 
-	if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_COMPUTE)
+	/* disable deep sleep if compute is enabled */
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_COMPUTE))
 		smu_v14_0_deep_sleep_control(smu, false);
-	else if (current_profile_mode == PP_SMC_POWER_PROFILE_COMPUTE)
+	else
 		smu_v14_0_deep_sleep_control(smu, true);
 
-	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
-	workload_type = smu_cmn_to_asic_specific_index(smu,
-						       CMN2ASIC_MAPPING_WORKLOAD,
-						       smu->power_profile_mode);
-	if (workload_type < 0)
-		return -EINVAL;
+	if (workload_mask & (1 << PP_SMC_POWER_PROFILE_CUSTOM)) {
+		if (!smu->custom_profile_params) {
+			smu->custom_profile_params =
+				kzalloc(SMU_14_0_2_CUSTOM_PARAMS_SIZE, GFP_KERNEL);
+			if (!smu->custom_profile_params)
+				return -ENOMEM;
+		}
+		if (custom_params && custom_params_max_idx) {
+			if (custom_params_max_idx != SMU_14_0_2_CUSTOM_PARAMS_COUNT)
+				return -EINVAL;
+			if (custom_params[0] >= SMU_14_0_2_CUSTOM_PARAMS_CLOCK_COUNT)
+				return -EINVAL;
+			idx = custom_params[0] * SMU_14_0_2_CUSTOM_PARAMS_COUNT;
+			smu->custom_profile_params[idx] = 1;
+			for (i = 1; i < custom_params_max_idx; i++)
+				smu->custom_profile_params[idx + i] = custom_params[i];
+		}
+		ret = smu_v14_0_2_set_power_profile_mode_coeff(smu,
+							       smu->custom_profile_params);
+		if (ret) {
+			if (idx != -1)
+				smu->custom_profile_params[idx] = 0;
+			return ret;
+		}
+	} else if (smu->custom_profile_params) {
+		memset(smu->custom_profile_params, 0, SMU_14_0_2_CUSTOM_PARAMS_SIZE);
+	}
 
 	ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask,
-										  smu->workload_mask, NULL);
-
-	if (!ret)
-		smu_cmn_assign_power_profile(smu);
+					      backend_workload_mask, NULL);
+	if (ret) {
+		dev_err(smu->adev->dev, "Failed to set workload mask 0x%08x\n",
+			workload_mask);
+		if (idx != -1)
+			smu->custom_profile_params[idx] = 0;
+		return ret;
+	}
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c
index dbbd3759bff3..9f55207ea9bc 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c
@@ -1144,14 +1144,6 @@ int smu_cmn_set_mp1_state(struct smu_context *smu,
 	return ret;
 }
 
-void smu_cmn_assign_power_profile(struct smu_context *smu)
-{
-	uint32_t index;
-	index = fls(smu->workload_mask);
-	index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0;
-	smu->power_profile_mode = smu->workload_setting[index];
-}
-
 bool smu_cmn_is_audio_func_enabled(struct amdgpu_device *adev)
 {
 	struct pci_dev *p = NULL;
@@ -1229,3 +1221,28 @@ void smu_cmn_generic_plpd_policy_desc(struct smu_dpm_policy *policy)
 {
 	policy->desc = &xgmi_plpd_policy_desc;
 }
+
+void smu_cmn_get_backend_workload_mask(struct smu_context *smu,
+				       u32 workload_mask,
+				       u32 *backend_workload_mask)
+{
+	int workload_type;
+	u32 profile_mode;
+
+	*backend_workload_mask = 0;
+
+	for (profile_mode = 0; profile_mode < PP_SMC_POWER_PROFILE_COUNT; profile_mode++) {
+		if (!(workload_mask & (1 << profile_mode)))
+			continue;
+
+		/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
+		workload_type = smu_cmn_to_asic_specific_index(smu,
+							       CMN2ASIC_MAPPING_WORKLOAD,
+							       profile_mode);
+
+		if (workload_type < 0)
+			continue;
+
+		*backend_workload_mask |= 1 << workload_type;
+	}
+}
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h
index 8a801e389659..a020277dec3e 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h
@@ -130,8 +130,6 @@ void smu_cmn_init_soft_gpu_metrics(void *table, uint8_t frev, uint8_t crev);
 int smu_cmn_set_mp1_state(struct smu_context *smu,
 			  enum pp_mp1_state mp1_state);
 
-void smu_cmn_assign_power_profile(struct smu_context *smu);
-
 /*
  * Helper function to make sysfs_emit_at() happy. Align buf to
  * the current page boundary and record the offset.
@@ -149,5 +147,9 @@ bool smu_cmn_is_audio_func_enabled(struct amdgpu_device *adev);
 void smu_cmn_generic_soc_policy_desc(struct smu_dpm_policy *policy);
 void smu_cmn_generic_plpd_policy_desc(struct smu_dpm_policy *policy);
 
+void smu_cmn_get_backend_workload_mask(struct smu_context *smu,
+				       u32 workload_mask,
+				       u32 *backend_workload_mask);
+
 #endif
 #endif
diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c
index 26d10065d534..1a1680d71486 100644
--- a/drivers/gpu/drm/armada/armada_gem.c
+++ b/drivers/gpu/drm/armada/armada_gem.c
@@ -15,7 +15,7 @@
 #include "armada_gem.h"
 #include "armada_ioctlP.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 static vm_fault_t armada_gem_vm_fault(struct vm_fault *vmf)
 {
diff --git a/drivers/gpu/drm/display/drm_dp_mst_topology.c b/drivers/gpu/drm/display/drm_dp_mst_topology.c
index ac90118b9e7a..dc4446d589e7 100644
--- a/drivers/gpu/drm/display/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/display/drm_dp_mst_topology.c
@@ -320,6 +320,9 @@ static bool drm_dp_decode_sideband_msg_hdr(const struct drm_dp_mst_topology_mgr
 	hdr->broadcast = (buf[idx] >> 7) & 0x1;
 	hdr->path_msg = (buf[idx] >> 6) & 0x1;
 	hdr->msg_len = buf[idx] & 0x3f;
+	if (hdr->msg_len < 1)		/* min space for body CRC */
+		return false;
+
 	idx++;
 	hdr->somt = (buf[idx] >> 7) & 0x1;
 	hdr->eomt = (buf[idx] >> 6) & 0x1;
@@ -3697,8 +3700,7 @@ int drm_dp_mst_topology_mgr_set_mst(struct drm_dp_mst_topology_mgr *mgr, bool ms
 		ret = 0;
 		mgr->payload_id_table_cleared = false;
 
-		memset(&mgr->down_rep_recv, 0, sizeof(mgr->down_rep_recv));
-		memset(&mgr->up_req_recv, 0, sizeof(mgr->up_req_recv));
+		mgr->reset_rx_state = true;
 	}
 
 out_unlock:
@@ -3856,6 +3858,11 @@ out_fail:
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_resume);
 
+static void reset_msg_rx_state(struct drm_dp_sideband_msg_rx *msg)
+{
+	memset(msg, 0, sizeof(*msg));
+}
+
 static bool
 drm_dp_get_one_sb_msg(struct drm_dp_mst_topology_mgr *mgr, bool up,
 		      struct drm_dp_mst_branch **mstb)
@@ -3934,6 +3941,34 @@ drm_dp_get_one_sb_msg(struct drm_dp_mst_topology_mgr *mgr, bool up,
 	return true;
 }
 
+static int get_msg_request_type(u8 data)
+{
+	return data & 0x7f;
+}
+
+static bool verify_rx_request_type(struct drm_dp_mst_topology_mgr *mgr,
+				   const struct drm_dp_sideband_msg_tx *txmsg,
+				   const struct drm_dp_sideband_msg_rx *rxmsg)
+{
+	const struct drm_dp_sideband_msg_hdr *hdr = &rxmsg->initial_hdr;
+	const struct drm_dp_mst_branch *mstb = txmsg->dst;
+	int tx_req_type = get_msg_request_type(txmsg->msg[0]);
+	int rx_req_type = get_msg_request_type(rxmsg->msg[0]);
+	char rad_str[64];
+
+	if (tx_req_type == rx_req_type)
+		return true;
+
+	drm_dp_mst_rad_to_str(mstb->rad, mstb->lct, rad_str, sizeof(rad_str));
+	drm_dbg_kms(mgr->dev,
+		    "Got unexpected MST reply, mstb: %p seqno: %d lct: %d rad: %s rx_req_type: %s (%02x) != tx_req_type: %s (%02x)\n",
+		    mstb, hdr->seqno, mstb->lct, rad_str,
+		    drm_dp_mst_req_type_str(rx_req_type), rx_req_type,
+		    drm_dp_mst_req_type_str(tx_req_type), tx_req_type);
+
+	return false;
+}
+
 static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr)
 {
 	struct drm_dp_sideband_msg_tx *txmsg;
@@ -3949,9 +3984,9 @@ static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr)
 
 	/* find the message */
 	mutex_lock(&mgr->qlock);
+
 	txmsg = list_first_entry_or_null(&mgr->tx_msg_downq,
 					 struct drm_dp_sideband_msg_tx, next);
-	mutex_unlock(&mgr->qlock);
 
 	/* Were we actually expecting a response, and from this mstb? */
 	if (!txmsg || txmsg->dst != mstb) {
@@ -3960,6 +3995,15 @@ static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr)
 		hdr = &msg->initial_hdr;
 		drm_dbg_kms(mgr->dev, "Got MST reply with no msg %p %d %d %02x %02x\n",
 			    mstb, hdr->seqno, hdr->lct, hdr->rad[0], msg->msg[0]);
+
+		mutex_unlock(&mgr->qlock);
+
+		goto out_clear_reply;
+	}
+
+	if (!verify_rx_request_type(mgr, txmsg, msg)) {
+		mutex_unlock(&mgr->qlock);
+
 		goto out_clear_reply;
 	}
 
@@ -3975,20 +4019,15 @@ static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr)
 			    txmsg->reply.u.nak.nak_data);
 	}
 
-	memset(msg, 0, sizeof(struct drm_dp_sideband_msg_rx));
-	drm_dp_mst_topology_put_mstb(mstb);
-
-	mutex_lock(&mgr->qlock);
 	txmsg->state = DRM_DP_SIDEBAND_TX_RX;
 	list_del(&txmsg->next);
+
 	mutex_unlock(&mgr->qlock);
 
 	wake_up_all(&mgr->tx_waitq);
 
-	return 0;
-
 out_clear_reply:
-	memset(msg, 0, sizeof(struct drm_dp_sideband_msg_rx));
+	reset_msg_rx_state(msg);
 out:
 	if (mstb)
 		drm_dp_mst_topology_put_mstb(mstb);
@@ -4070,16 +4109,20 @@ static void drm_dp_mst_up_req_work(struct work_struct *work)
 static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
 {
 	struct drm_dp_pending_up_req *up_req;
+	struct drm_dp_mst_branch *mst_primary;
+	int ret = 0;
 
 	if (!drm_dp_get_one_sb_msg(mgr, true, NULL))
-		goto out;
+		goto out_clear_reply;
 
 	if (!mgr->up_req_recv.have_eomt)
 		return 0;
 
 	up_req = kzalloc(sizeof(*up_req), GFP_KERNEL);
-	if (!up_req)
-		return -ENOMEM;
+	if (!up_req) {
+		ret = -ENOMEM;
+		goto out_clear_reply;
+	}
 
 	INIT_LIST_HEAD(&up_req->next);
 
@@ -4090,10 +4133,19 @@ static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
 		drm_dbg_kms(mgr->dev, "Received unknown up req type, ignoring: %x\n",
 			    up_req->msg.req_type);
 		kfree(up_req);
-		goto out;
+		goto out_clear_reply;
 	}
 
-	drm_dp_send_up_ack_reply(mgr, mgr->mst_primary, up_req->msg.req_type,
+	mutex_lock(&mgr->lock);
+	mst_primary = mgr->mst_primary;
+	if (!mst_primary || !drm_dp_mst_topology_try_get_mstb(mst_primary)) {
+		mutex_unlock(&mgr->lock);
+		kfree(up_req);
+		goto out_clear_reply;
+	}
+	mutex_unlock(&mgr->lock);
+
+	drm_dp_send_up_ack_reply(mgr, mst_primary, up_req->msg.req_type,
 				 false);
 
 	if (up_req->msg.req_type == DP_CONNECTION_STATUS_NOTIFY) {
@@ -4110,13 +4162,13 @@ static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
 			    conn_stat->peer_device_type);
 
 		mutex_lock(&mgr->probe_lock);
-		handle_csn = mgr->mst_primary->link_address_sent;
+		handle_csn = mst_primary->link_address_sent;
 		mutex_unlock(&mgr->probe_lock);
 
 		if (!handle_csn) {
 			drm_dbg_kms(mgr->dev, "Got CSN before finish topology probing. Skip it.");
 			kfree(up_req);
-			goto out;
+			goto out_put_primary;
 		}
 	} else if (up_req->msg.req_type == DP_RESOURCE_STATUS_NOTIFY) {
 		const struct drm_dp_resource_status_notify *res_stat =
@@ -4133,9 +4185,22 @@ static int drm_dp_mst_handle_up_req(struct drm_dp_mst_topology_mgr *mgr)
 	mutex_unlock(&mgr->up_req_lock);
 	queue_work(system_long_wq, &mgr->up_req_work);
 
-out:
-	memset(&mgr->up_req_recv, 0, sizeof(struct drm_dp_sideband_msg_rx));
-	return 0;
+out_put_primary:
+	drm_dp_mst_topology_put_mstb(mst_primary);
+out_clear_reply:
+	reset_msg_rx_state(&mgr->up_req_recv);
+	return ret;
+}
+
+static void update_msg_rx_state(struct drm_dp_mst_topology_mgr *mgr)
+{
+	mutex_lock(&mgr->lock);
+	if (mgr->reset_rx_state) {
+		mgr->reset_rx_state = false;
+		reset_msg_rx_state(&mgr->down_rep_recv);
+		reset_msg_rx_state(&mgr->up_req_recv);
+	}
+	mutex_unlock(&mgr->lock);
 }
 
 /**
@@ -4172,6 +4237,8 @@ int drm_dp_mst_hpd_irq_handle_event(struct drm_dp_mst_topology_mgr *mgr, const u
 		*handled = true;
 	}
 
+	update_msg_rx_state(mgr);
+
 	if (esi[1] & DP_DOWN_REP_MSG_RDY) {
 		ret = drm_dp_mst_handle_down_rep(mgr);
 		*handled = true;
diff --git a/drivers/gpu/drm/drm_gem_dma_helper.c b/drivers/gpu/drm/drm_gem_dma_helper.c
index 870b90b78bc4..16988d316a6d 100644
--- a/drivers/gpu/drm/drm_gem_dma_helper.c
+++ b/drivers/gpu/drm/drm_gem_dma_helper.c
@@ -600,5 +600,5 @@ drm_gem_dma_prime_import_sg_table_vmap(struct drm_device *dev,
 EXPORT_SYMBOL(drm_gem_dma_prime_import_sg_table_vmap);
 
 MODULE_DESCRIPTION("DRM DMA memory-management helpers");
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/drm_gem_framebuffer_helper.c b/drivers/gpu/drm/drm_gem_framebuffer_helper.c
index 3bdb6ba37ff4..185534f56bab 100644
--- a/drivers/gpu/drm/drm_gem_framebuffer_helper.c
+++ b/drivers/gpu/drm/drm_gem_framebuffer_helper.c
@@ -18,7 +18,7 @@
 
 #include "drm_internal.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 #define AFBC_HEADER_SIZE		16
 #define AFBC_TH_LAYOUT_ALIGNMENT	8
diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_gem_shmem_helper.c
index 8508060a1a95..5ab351409312 100644
--- a/drivers/gpu/drm/drm_gem_shmem_helper.c
+++ b/drivers/gpu/drm/drm_gem_shmem_helper.c
@@ -22,7 +22,7 @@
 #include <drm/drm_prime.h>
 #include <drm/drm_print.h>
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 /**
  * DOC: overview
@@ -800,5 +800,5 @@ drm_gem_shmem_prime_import_sg_table(struct drm_device *dev,
 EXPORT_SYMBOL_GPL(drm_gem_shmem_prime_import_sg_table);
 
 MODULE_DESCRIPTION("DRM SHMEM memory-management helpers");
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 0e3f8adf162f..32a8781cfd67 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -40,7 +40,7 @@
 
 #include "drm_internal.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 /**
  * DOC: overview and lifetime rules
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
index 6b98200068e4..42e57d142554 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
@@ -10,7 +10,7 @@
 #include "etnaviv_drv.h"
 #include "etnaviv_gem.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 static struct lock_class_key etnaviv_prime_lock_class;
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 638ca96830e9..4787fee4696f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -18,7 +18,7 @@
 #include "exynos_drm_drv.h"
 #include "exynos_drm_gem.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 static int exynos_drm_gem_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
index 1df74f7aa3dc..9473050ac842 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
@@ -16,7 +16,7 @@
 #include "i915_gem_object.h"
 #include "i915_scatterlist.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 I915_SELFTEST_DECLARE(static bool force_different_devices;)
 
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index ca0fb126b02d..b27ff77bfb50 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -53,8 +53,8 @@
 #include "intel_gvt.h"
 #include "gvt.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
-MODULE_IMPORT_NS(I915_GVT);
+MODULE_IMPORT_NS("DMA_BUF");
+MODULE_IMPORT_NS("I915_GVT");
 
 /* helper macros copied from vfio-pci */
 #define VFIO_PCI_OFFSET_SHIFT   40
diff --git a/drivers/gpu/drm/i915/intel_gvt.c b/drivers/gpu/drm/i915/intel_gvt.c
index 5a01d60e5186..a5383a2bc64b 100644
--- a/drivers/gpu/drm/i915/intel_gvt.c
+++ b/drivers/gpu/drm/i915/intel_gvt.c
@@ -204,7 +204,7 @@ int intel_gvt_set_ops(const struct intel_vgpu_ops *ops)
 
 	return 0;
 }
-EXPORT_SYMBOL_NS_GPL(intel_gvt_set_ops, I915_GVT);
+EXPORT_SYMBOL_NS_GPL(intel_gvt_set_ops, "I915_GVT");
 
 void intel_gvt_clear_ops(const struct intel_vgpu_ops *ops)
 {
@@ -222,7 +222,7 @@ void intel_gvt_clear_ops(const struct intel_vgpu_ops *ops)
 	intel_gvt_ops = NULL;
 	mutex_unlock(&intel_gvt_mutex);
 }
-EXPORT_SYMBOL_NS_GPL(intel_gvt_clear_ops, I915_GVT);
+EXPORT_SYMBOL_NS_GPL(intel_gvt_clear_ops, "I915_GVT");
 
 /**
  * intel_gvt_init - initialize GVT components
@@ -284,40 +284,40 @@ void intel_gvt_resume(struct drm_i915_private *dev_priv)
  * Exported here so that the exports only get created when GVT support is
  * actually enabled.
  */
-EXPORT_SYMBOL_NS_GPL(i915_gem_object_alloc, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_object_create_shmem, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_object_init, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_object_ggtt_pin_ww, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_object_pin_map, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_object_set_to_cpu_domain, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(__i915_gem_object_flush_map, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(__i915_gem_object_set_pages, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_gtt_insert, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_prime_export, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_ww_ctx_init, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_ww_ctx_backoff, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_gem_ww_ctx_fini, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_ppgtt_create, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_request_add, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_request_create, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_request_wait, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_reserve_fence, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_unreserve_fence, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_vm_release, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(_i915_vma_move_to_active, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(intel_context_create, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(__intel_context_do_pin, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(__intel_context_do_unpin, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(intel_ring_begin, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(intel_runtime_pm_get, I915_GVT);
+EXPORT_SYMBOL_NS_GPL(i915_gem_object_alloc, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_object_create_shmem, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_object_init, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_object_ggtt_pin_ww, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_object_pin_map, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_object_set_to_cpu_domain, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(__i915_gem_object_flush_map, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(__i915_gem_object_set_pages, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_gtt_insert, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_prime_export, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_ww_ctx_init, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_ww_ctx_backoff, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_gem_ww_ctx_fini, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_ppgtt_create, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_request_add, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_request_create, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_request_wait, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_reserve_fence, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_unreserve_fence, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_vm_release, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(_i915_vma_move_to_active, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(intel_context_create, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(__intel_context_do_pin, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(__intel_context_do_unpin, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(intel_ring_begin, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(intel_runtime_pm_get, "I915_GVT");
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG_RUNTIME_PM)
-EXPORT_SYMBOL_NS_GPL(intel_runtime_pm_put, I915_GVT);
+EXPORT_SYMBOL_NS_GPL(intel_runtime_pm_put, "I915_GVT");
 #endif
-EXPORT_SYMBOL_NS_GPL(intel_runtime_pm_put_unchecked, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(intel_uncore_forcewake_for_reg, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(intel_uncore_forcewake_get, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(intel_uncore_forcewake_put, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(shmem_pin_map, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(shmem_unpin_map, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(__px_dma, I915_GVT);
-EXPORT_SYMBOL_NS_GPL(i915_fence_ops, I915_GVT);
+EXPORT_SYMBOL_NS_GPL(intel_runtime_pm_put_unchecked, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(intel_uncore_forcewake_for_reg, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(intel_uncore_forcewake_get, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(intel_uncore_forcewake_put, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(shmem_pin_map, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(shmem_unpin_map, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(__px_dma, "I915_GVT");
+EXPORT_SYMBOL_NS_GPL(i915_fence_ops, "I915_GVT");
diff --git a/drivers/gpu/drm/i915/intel_gvt_mmio_table.c b/drivers/gpu/drm/i915/intel_gvt_mmio_table.c
index 955c9a33212a..aa51f366626c 100644
--- a/drivers/gpu/drm/i915/intel_gvt_mmio_table.c
+++ b/drivers/gpu/drm/i915/intel_gvt_mmio_table.c
@@ -1308,4 +1308,4 @@ int intel_gvt_iterate_mmio_table(struct intel_gvt_mmio_table_iter *iter)
 err:
 	return ret;
 }
-EXPORT_SYMBOL_NS_GPL(intel_gvt_iterate_mmio_table, I915_GVT);
+EXPORT_SYMBOL_NS_GPL(intel_gvt_iterate_mmio_table, "I915_GVT");
diff --git a/drivers/gpu/drm/imagination/pvr_drv.c b/drivers/gpu/drm/imagination/pvr_drv.c
index a3e737f91b19..85ee9abd1811 100644
--- a/drivers/gpu/drm/imagination/pvr_drv.c
+++ b/drivers/gpu/drm/imagination/pvr_drv.c
@@ -1497,5 +1497,5 @@ module_platform_driver(pvr_driver);
 MODULE_AUTHOR("Imagination Technologies Ltd.");
 MODULE_DESCRIPTION(PVR_DRIVER_DESC);
 MODULE_LICENSE("Dual MIT/GPL");
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 MODULE_FIRMWARE("powervr/rogue_33.15.11.3_v1.fw");
diff --git a/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c b/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
index 36f9ee4baad3..30cf1cdc1aa3 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
@@ -11,7 +11,7 @@
 
 #include "omap_drv.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 /* -----------------------------------------------------------------------------
  * DMABUF Export
diff --git a/drivers/gpu/drm/solomon/ssd130x-i2c.c b/drivers/gpu/drm/solomon/ssd130x-i2c.c
index f2ccab9c06d9..941a2eb44c57 100644
--- a/drivers/gpu/drm/solomon/ssd130x-i2c.c
+++ b/drivers/gpu/drm/solomon/ssd130x-i2c.c
@@ -123,4 +123,4 @@ module_i2c_driver(ssd130x_i2c_driver);
 MODULE_DESCRIPTION(DRIVER_DESC);
 MODULE_AUTHOR("Javier Martinez Canillas <javierm@redhat.com>");
 MODULE_LICENSE("GPL v2");
-MODULE_IMPORT_NS(DRM_SSD130X);
+MODULE_IMPORT_NS("DRM_SSD130X");
diff --git a/drivers/gpu/drm/solomon/ssd130x-spi.c b/drivers/gpu/drm/solomon/ssd130x-spi.c
index 84bfde31d172..08334be38694 100644
--- a/drivers/gpu/drm/solomon/ssd130x-spi.c
+++ b/drivers/gpu/drm/solomon/ssd130x-spi.c
@@ -192,4 +192,4 @@ module_spi_driver(ssd130x_spi_driver);
 MODULE_DESCRIPTION(DRIVER_DESC);
 MODULE_AUTHOR("Javier Martinez Canillas <javierm@redhat.com>");
 MODULE_LICENSE("GPL");
-MODULE_IMPORT_NS(DRM_SSD130X);
+MODULE_IMPORT_NS("DRM_SSD130X");
diff --git a/drivers/gpu/drm/solomon/ssd130x.c b/drivers/gpu/drm/solomon/ssd130x.c
index 29b2f82d81f8..486d8f5282f9 100644
--- a/drivers/gpu/drm/solomon/ssd130x.c
+++ b/drivers/gpu/drm/solomon/ssd130x.c
@@ -209,7 +209,7 @@ const struct ssd130x_deviceinfo ssd130x_variants[] = {
 		.family_id = SSD133X_FAMILY,
 	}
 };
-EXPORT_SYMBOL_NS_GPL(ssd130x_variants, DRM_SSD130X);
+EXPORT_SYMBOL_NS_GPL(ssd130x_variants, "DRM_SSD130X");
 
 struct ssd130x_crtc_state {
 	struct drm_crtc_state base;
diff --git a/drivers/gpu/drm/sti/sti_mixer.c b/drivers/gpu/drm/sti/sti_mixer.c
index 7e5f14646625..06c1b81912f7 100644
--- a/drivers/gpu/drm/sti/sti_mixer.c
+++ b/drivers/gpu/drm/sti/sti_mixer.c
@@ -137,7 +137,7 @@ static void mixer_dbg_crb(struct seq_file *s, int val)
 	}
 }
 
-static void mixer_dbg_mxn(struct seq_file *s, void *addr)
+static void mixer_dbg_mxn(struct seq_file *s, void __iomem *addr)
 {
 	int i;
 
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index d275404ad0e9..ace3e5a805cf 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -22,7 +22,7 @@
 #include "drm.h"
 #include "gem.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 static unsigned int sg_dma_count_chunks(struct scatterlist *sgl, unsigned int nents)
 {
diff --git a/drivers/gpu/drm/v3d/v3d_perfmon.c b/drivers/gpu/drm/v3d/v3d_perfmon.c
index 156be13ab2ef..924814cab46a 100644
--- a/drivers/gpu/drm/v3d/v3d_perfmon.c
+++ b/drivers/gpu/drm/v3d/v3d_perfmon.c
@@ -254,9 +254,9 @@ void v3d_perfmon_start(struct v3d_dev *v3d, struct v3d_perfmon *perfmon)
 		V3D_CORE_WRITE(0, V3D_V4_PCTR_0_SRC_X(source), channel);
 	}
 
+	V3D_CORE_WRITE(0, V3D_V4_PCTR_0_EN, mask);
 	V3D_CORE_WRITE(0, V3D_V4_PCTR_0_CLR, mask);
 	V3D_CORE_WRITE(0, V3D_PCTR_0_OVERFLOW, mask);
-	V3D_CORE_WRITE(0, V3D_V4_PCTR_0_EN, mask);
 
 	v3d->active_perfmon = perfmon;
 }
diff --git a/drivers/gpu/drm/vmwgfx/ttm_object.c b/drivers/gpu/drm/vmwgfx/ttm_object.c
index a17e62867f3b..36d46b79562a 100644
--- a/drivers/gpu/drm/vmwgfx/ttm_object.c
+++ b/drivers/gpu/drm/vmwgfx/ttm_object.c
@@ -54,7 +54,7 @@
 #include <linux/module.h>
 #include <linux/hashtable.h>
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 #define VMW_TTM_OBJECT_REF_HT_ORDER 10
 
diff --git a/drivers/gpu/drm/xe/tests/xe_live_test_mod.c b/drivers/gpu/drm/xe/tests/xe_live_test_mod.c
index 5f14737c8210..0d36ab864ec0 100644
--- a/drivers/gpu/drm/xe/tests/xe_live_test_mod.c
+++ b/drivers/gpu/drm/xe/tests/xe_live_test_mod.c
@@ -18,4 +18,4 @@ kunit_test_suite(xe_mocs_test_suite);
 MODULE_AUTHOR("Intel Corporation");
 MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION("xe live kunit tests");
-MODULE_IMPORT_NS(EXPORTED_FOR_KUNIT_TESTING);
+MODULE_IMPORT_NS("EXPORTED_FOR_KUNIT_TESTING");
diff --git a/drivers/gpu/drm/xe/tests/xe_test_mod.c b/drivers/gpu/drm/xe/tests/xe_test_mod.c
index 875f3e6f965e..93081bcf2ab0 100644
--- a/drivers/gpu/drm/xe/tests/xe_test_mod.c
+++ b/drivers/gpu/drm/xe/tests/xe_test_mod.c
@@ -7,4 +7,4 @@
 MODULE_AUTHOR("Intel Corporation");
 MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION("xe kunit tests");
-MODULE_IMPORT_NS(EXPORTED_FOR_KUNIT_TESTING);
+MODULE_IMPORT_NS("EXPORTED_FOR_KUNIT_TESTING");
diff --git a/drivers/gpu/drm/xe/xe_devcoredump.c b/drivers/gpu/drm/xe/xe_devcoredump.c
index 0b0cd6aa1d9f..f8947e7e917e 100644
--- a/drivers/gpu/drm/xe/xe_devcoredump.c
+++ b/drivers/gpu/drm/xe/xe_devcoredump.c
@@ -155,36 +155,6 @@ static void xe_devcoredump_snapshot_free(struct xe_devcoredump_snapshot *ss)
 	ss->vm = NULL;
 }
 
-static void xe_devcoredump_deferred_snap_work(struct work_struct *work)
-{
-	struct xe_devcoredump_snapshot *ss = container_of(work, typeof(*ss), work);
-	struct xe_devcoredump *coredump = container_of(ss, typeof(*coredump), snapshot);
-	struct xe_device *xe = coredump_to_xe(coredump);
-	unsigned int fw_ref;
-
-	xe_pm_runtime_get(xe);
-
-	/* keep going if fw fails as we still want to save the memory and SW data */
-	fw_ref = xe_force_wake_get(gt_to_fw(ss->gt), XE_FORCEWAKE_ALL);
-	if (!xe_force_wake_ref_has_domain(fw_ref, XE_FORCEWAKE_ALL))
-		xe_gt_info(ss->gt, "failed to get forcewake for coredump capture\n");
-	xe_vm_snapshot_capture_delayed(ss->vm);
-	xe_guc_exec_queue_snapshot_capture_delayed(ss->ge);
-	xe_force_wake_put(gt_to_fw(ss->gt), fw_ref);
-
-	xe_pm_runtime_put(xe);
-
-	/* Calculate devcoredump size */
-	ss->read.size = __xe_devcoredump_read(NULL, INT_MAX, coredump);
-
-	ss->read.buffer = kvmalloc(ss->read.size, GFP_USER);
-	if (!ss->read.buffer)
-		return;
-
-	__xe_devcoredump_read(ss->read.buffer, ss->read.size, coredump);
-	xe_devcoredump_snapshot_free(ss);
-}
-
 static ssize_t xe_devcoredump_read(char *buffer, loff_t offset,
 				   size_t count, void *data, size_t datalen)
 {
@@ -234,6 +204,45 @@ static void xe_devcoredump_free(void *data)
 		 "Xe device coredump has been deleted.\n");
 }
 
+static void xe_devcoredump_deferred_snap_work(struct work_struct *work)
+{
+	struct xe_devcoredump_snapshot *ss = container_of(work, typeof(*ss), work);
+	struct xe_devcoredump *coredump = container_of(ss, typeof(*coredump), snapshot);
+	struct xe_device *xe = coredump_to_xe(coredump);
+	unsigned int fw_ref;
+
+	/*
+	 * NB: Despite passing a GFP_ flags parameter here, more allocations are done
+	 * internally using GFP_KERNEL expliictly. Hence this call must be in the worker
+	 * thread and not in the initial capture call.
+	 */
+	dev_coredumpm_timeout(gt_to_xe(ss->gt)->drm.dev, THIS_MODULE, coredump, 0, GFP_KERNEL,
+			      xe_devcoredump_read, xe_devcoredump_free,
+			      XE_COREDUMP_TIMEOUT_JIFFIES);
+
+	xe_pm_runtime_get(xe);
+
+	/* keep going if fw fails as we still want to save the memory and SW data */
+	fw_ref = xe_force_wake_get(gt_to_fw(ss->gt), XE_FORCEWAKE_ALL);
+	if (!xe_force_wake_ref_has_domain(fw_ref, XE_FORCEWAKE_ALL))
+		xe_gt_info(ss->gt, "failed to get forcewake for coredump capture\n");
+	xe_vm_snapshot_capture_delayed(ss->vm);
+	xe_guc_exec_queue_snapshot_capture_delayed(ss->ge);
+	xe_force_wake_put(gt_to_fw(ss->gt), fw_ref);
+
+	xe_pm_runtime_put(xe);
+
+	/* Calculate devcoredump size */
+	ss->read.size = __xe_devcoredump_read(NULL, INT_MAX, coredump);
+
+	ss->read.buffer = kvmalloc(ss->read.size, GFP_USER);
+	if (!ss->read.buffer)
+		return;
+
+	__xe_devcoredump_read(ss->read.buffer, ss->read.size, coredump);
+	xe_devcoredump_snapshot_free(ss);
+}
+
 static void devcoredump_snapshot(struct xe_devcoredump *coredump,
 				 struct xe_sched_job *job)
 {
@@ -310,10 +319,6 @@ void xe_devcoredump(struct xe_sched_job *job)
 	drm_info(&xe->drm, "Xe device coredump has been created\n");
 	drm_info(&xe->drm, "Check your /sys/class/drm/card%d/device/devcoredump/data\n",
 		 xe->drm.primary->index);
-
-	dev_coredumpm_timeout(xe->drm.dev, THIS_MODULE, coredump, 0, GFP_KERNEL,
-			      xe_devcoredump_read, xe_devcoredump_free,
-			      XE_COREDUMP_TIMEOUT_JIFFIES);
 }
 
 static void xe_driver_devcoredump_fini(void *arg)
diff --git a/drivers/gpu/drm/xe/xe_dma_buf.c b/drivers/gpu/drm/xe/xe_dma_buf.c
index 68f309f5e981..c5b95470fa32 100644
--- a/drivers/gpu/drm/xe/xe_dma_buf.c
+++ b/drivers/gpu/drm/xe/xe_dma_buf.c
@@ -20,7 +20,7 @@
 #include "xe_ttm_vram_mgr.h"
 #include "xe_vm.h"
 
-MODULE_IMPORT_NS(DMA_BUF);
+MODULE_IMPORT_NS("DMA_BUF");
 
 static int xe_dma_buf_attach(struct dma_buf *dmabuf,
 			     struct dma_buf_attachment *attach)
diff --git a/drivers/gpu/drm/xe/xe_guc_capture.c b/drivers/gpu/drm/xe/xe_guc_capture.c
index cc72446a5de1..d63912d28246 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture.c
+++ b/drivers/gpu/drm/xe/xe_guc_capture.c
@@ -102,6 +102,7 @@ struct __guc_capture_parsed_output {
  *                   A 64 bit register define requires 2 consecutive entries,
  *                   with low dword first and hi dword the second.
  *     2. Register name: null for incompleted define
+ *     3. Incorrect order will trigger XE_WARN.
  */
 #define COMMON_XELP_BASE_GLOBAL \
 	{ FORCEWAKE_GT,			REG_32BIT,	0,	0,	"FORCEWAKE_GT"}
@@ -1675,10 +1676,10 @@ snapshot_print_by_list_order(struct xe_hw_engine_snapshot *snapshot, struct drm_
 	struct xe_devcoredump *devcoredump = &xe->devcoredump;
 	struct xe_devcoredump_snapshot *devcore_snapshot = &devcoredump->snapshot;
 	struct gcap_reg_list_info *reginfo = NULL;
-	u32 last_value, i;
-	bool is_ext;
+	u32 i, last_value = 0;
+	bool is_ext, low32_ready = false;
 
-	if (!list || list->num_regs == 0)
+	if (!list || !list->list || list->num_regs == 0)
 		return;
 	XE_WARN_ON(!devcore_snapshot->matched_node);
 
@@ -1701,29 +1702,75 @@ snapshot_print_by_list_order(struct xe_hw_engine_snapshot *snapshot, struct drm_
 			continue;
 
 		value = reg->value;
-		if (reg_desc->data_type == REG_64BIT_LOW_DW) {
+		switch (reg_desc->data_type) {
+		case REG_64BIT_LOW_DW:
 			last_value = value;
+
+			/*
+			 * A 64 bit register define requires 2 consecutive
+			 * entries in register list, with low dword first
+			 * and hi dword the second, like:
+			 *  { XXX_REG_LO(0), REG_64BIT_LOW_DW, 0, 0, NULL},
+			 *  { XXX_REG_HI(0), REG_64BIT_HI_DW,  0, 0, "XXX_REG"},
+			 *
+			 * Incorrect order will trigger XE_WARN.
+			 *
+			 * Possible double low here, for example:
+			 *  { XXX_REG_LO(0), REG_64BIT_LOW_DW, 0, 0, NULL},
+			 *  { XXX_REG_LO(0), REG_64BIT_LOW_DW, 0, 0, NULL},
+			 */
+			XE_WARN_ON(low32_ready);
+			low32_ready = true;
 			/* Low 32 bit dword saved, continue for high 32 bit */
-			continue;
-		} else if (reg_desc->data_type == REG_64BIT_HI_DW) {
+			break;
+
+		case REG_64BIT_HI_DW: {
 			u64 value_qw = ((u64)value << 32) | last_value;
 
+			/*
+			 * Incorrect 64bit register order. Possible missing low.
+			 * for example:
+			 *  { XXX_REG(0), REG_32BIT, 0, 0, NULL},
+			 *  { XXX_REG_HI(0), REG_64BIT_HI_DW, 0, 0, NULL},
+			 */
+			XE_WARN_ON(!low32_ready);
+			low32_ready = false;
+
 			drm_printf(p, "\t%s: 0x%016llx\n", reg_desc->regname, value_qw);
-			continue;
+			break;
 		}
 
-		if (is_ext) {
-			int dss, group, instance;
+		case REG_32BIT:
+			/*
+			 * Incorrect 64bit register order. Possible missing high.
+			 * for example:
+			 *  { XXX_REG_LO(0), REG_64BIT_LOW_DW, 0, 0, NULL},
+			 *  { XXX_REG(0), REG_32BIT, 0, 0, "XXX_REG"},
+			 */
+			XE_WARN_ON(low32_ready);
+
+			if (is_ext) {
+				int dss, group, instance;
 
-			group = FIELD_GET(GUC_REGSET_STEERING_GROUP, reg_desc->flags);
-			instance = FIELD_GET(GUC_REGSET_STEERING_INSTANCE, reg_desc->flags);
-			dss = xe_gt_mcr_steering_info_to_dss_id(gt, group, instance);
+				group = FIELD_GET(GUC_REGSET_STEERING_GROUP, reg_desc->flags);
+				instance = FIELD_GET(GUC_REGSET_STEERING_INSTANCE, reg_desc->flags);
+				dss = xe_gt_mcr_steering_info_to_dss_id(gt, group, instance);
 
-			drm_printf(p, "\t%s[%u]: 0x%08x\n", reg_desc->regname, dss, value);
-		} else {
-			drm_printf(p, "\t%s: 0x%08x\n", reg_desc->regname, value);
+				drm_printf(p, "\t%s[%u]: 0x%08x\n", reg_desc->regname, dss, value);
+			} else {
+				drm_printf(p, "\t%s: 0x%08x\n", reg_desc->regname, value);
+			}
+			break;
 		}
 	}
+
+	/*
+	 * Incorrect 64bit register order. Possible missing high.
+	 * for example:
+	 *  { XXX_REG_LO(0), REG_64BIT_LOW_DW, 0, 0, NULL},
+	 *  } // <- Register list end
+	 */
+	XE_WARN_ON(low32_ready);
 }
 
 /**