Merge branch 'drm-next-5.2' of git://people.freedesktop.org/~agd5f/linux into drm-next

- SR-IOV fixes
- Raven flickering fix
- Misc spelling fixes
- Vega20 power fixes
- Freesync improvements
- DC fixes
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190502193020.3562-1-alexander.deucher@amd.com

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c

@@ -335,6 +335,43 @@ void amdgpu_amdkfd_free_gtt_mem(struct kgd_dev *kgd, void *mem_obj)
 	amdgpu_bo_unref(&(bo));
 }
 
+uint32_t amdgpu_amdkfd_get_fw_version(struct kgd_dev *kgd,
+				      enum kgd_engine_type type)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
+
+	switch (type) {
+	case KGD_ENGINE_PFP:
+		return adev->gfx.pfp_fw_version;
+
+	case KGD_ENGINE_ME:
+		return adev->gfx.me_fw_version;
+
+	case KGD_ENGINE_CE:
+		return adev->gfx.ce_fw_version;
+
+	case KGD_ENGINE_MEC1:
+		return adev->gfx.mec_fw_version;
+
+	case KGD_ENGINE_MEC2:
+		return adev->gfx.mec2_fw_version;
+
+	case KGD_ENGINE_RLC:
+		return adev->gfx.rlc_fw_version;
+
+	case KGD_ENGINE_SDMA1:
+		return adev->sdma.instance[0].fw_version;
+
+	case KGD_ENGINE_SDMA2:
+		return adev->sdma.instance[1].fw_version;
+
+	default:
+		return 0;
+	}
+
+	return 0;
+}
+
 void amdgpu_amdkfd_get_local_mem_info(struct kgd_dev *kgd,
 				      struct kfd_local_mem_info *mem_info)
 {

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h

@@ -81,6 +81,18 @@ struct amdgpu_kfd_dev {
 	uint64_t vram_used;
 };
 
+enum kgd_engine_type {
+	KGD_ENGINE_PFP = 1,
+	KGD_ENGINE_ME,
+	KGD_ENGINE_CE,
+	KGD_ENGINE_MEC1,
+	KGD_ENGINE_MEC2,
+	KGD_ENGINE_RLC,
+	KGD_ENGINE_SDMA1,
+	KGD_ENGINE_SDMA2,
+	KGD_ENGINE_MAX
+};
+
 struct amdgpu_amdkfd_fence *amdgpu_amdkfd_fence_create(u64 context,
 						       struct mm_struct *mm);
 bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm);
@@ -142,6 +154,8 @@ int amdgpu_amdkfd_alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 				void **mem_obj, uint64_t *gpu_addr,
 				void **cpu_ptr, bool mqd_gfx9);
 void amdgpu_amdkfd_free_gtt_mem(struct kgd_dev *kgd, void *mem_obj);
+uint32_t amdgpu_amdkfd_get_fw_version(struct kgd_dev *kgd,
+				      enum kgd_engine_type type);
 void amdgpu_amdkfd_get_local_mem_info(struct kgd_dev *kgd,
 				      struct kfd_local_mem_info *mem_info);
 uint64_t amdgpu_amdkfd_get_gpu_clock_counter(struct kgd_dev *kgd);

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c

@@ -22,14 +22,12 @@
 
 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
 #include "cikd.h"
 #include "cik_sdma.h"
-#include "amdgpu_ucode.h"
 #include "gfx_v7_0.h"
 #include "gca/gfx_7_2_d.h"
 #include "gca/gfx_7_2_enum.h"
@@ -139,7 +137,6 @@ static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd, uint8_t vmid);
 static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
 							uint8_t vmid);
 
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
 static void set_scratch_backing_va(struct kgd_dev *kgd,
 					uint64_t va, uint32_t vmid);
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
@@ -191,7 +188,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.address_watch_get_offset = kgd_address_watch_get_offset,
 	.get_atc_vmid_pasid_mapping_pasid = get_atc_vmid_pasid_mapping_pasid,
 	.get_atc_vmid_pasid_mapping_valid = get_atc_vmid_pasid_mapping_valid,
-	.get_fw_version = get_fw_version,
 	.set_scratch_backing_va = set_scratch_backing_va,
 	.get_tile_config = get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
@@ -792,63 +788,6 @@ static void set_scratch_backing_va(struct kgd_dev *kgd,
 	unlock_srbm(kgd);
 }
 
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-	const union amdgpu_firmware_header *hdr;
-
-	switch (type) {
-	case KGD_ENGINE_PFP:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.pfp_fw->data;
-		break;
-
-	case KGD_ENGINE_ME:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.me_fw->data;
-		break;
-
-	case KGD_ENGINE_CE:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.ce_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec2_fw->data;
-		break;
-
-	case KGD_ENGINE_RLC:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.rlc_fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[0].fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[1].fw->data;
-		break;
-
-	default:
-		return 0;
-	}
-
-	if (hdr == NULL)
-		return 0;
-
-	/* Only 12 bit in use*/
-	return hdr->common.ucode_version;
-}
-
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 			uint64_t page_table_base)
 {

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c

@@ -23,12 +23,10 @@
 #include <linux/module.h>
 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
-#include "amdgpu_ucode.h"
 #include "gfx_v8_0.h"
 #include "gca/gfx_8_0_sh_mask.h"
 #include "gca/gfx_8_0_d.h"
@@ -95,7 +93,6 @@ static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
 		uint8_t vmid);
 static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
 		uint8_t vmid);
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
 static void set_scratch_backing_va(struct kgd_dev *kgd,
 					uint64_t va, uint32_t vmid);
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
@@ -148,7 +145,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 			get_atc_vmid_pasid_mapping_pasid,
 	.get_atc_vmid_pasid_mapping_valid =
 			get_atc_vmid_pasid_mapping_valid,
-	.get_fw_version = get_fw_version,
 	.set_scratch_backing_va = set_scratch_backing_va,
 	.get_tile_config = get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
@@ -751,63 +747,6 @@ static void set_scratch_backing_va(struct kgd_dev *kgd,
 	unlock_srbm(kgd);
 }
 
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-	const union amdgpu_firmware_header *hdr;
-
-	switch (type) {
-	case KGD_ENGINE_PFP:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.pfp_fw->data;
-		break;
-
-	case KGD_ENGINE_ME:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.me_fw->data;
-		break;
-
-	case KGD_ENGINE_CE:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.ce_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec2_fw->data;
-		break;
-
-	case KGD_ENGINE_RLC:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.rlc_fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[0].fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[1].fw->data;
-		break;
-
-	default:
-		return 0;
-	}
-
-	if (hdr == NULL)
-		return 0;
-
-	/* Only 12 bit in use*/
-	return hdr->common.ucode_version;
-}
-
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base)
 {

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c

@@ -25,12 +25,10 @@
 #include <linux/module.h>
 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
-#include "amdgpu_ucode.h"
 #include "soc15_hw_ip.h"
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
@@ -111,7 +109,6 @@ static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
 		uint8_t vmid);
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base);
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
 static void set_scratch_backing_va(struct kgd_dev *kgd,
 					uint64_t va, uint32_t vmid);
 static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
@@ -158,7 +155,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 			get_atc_vmid_pasid_mapping_pasid,
 	.get_atc_vmid_pasid_mapping_valid =
 			get_atc_vmid_pasid_mapping_valid,
-	.get_fw_version = get_fw_version,
 	.set_scratch_backing_va = set_scratch_backing_va,
 	.get_tile_config = amdgpu_amdkfd_get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
@@ -874,56 +870,6 @@ static void set_scratch_backing_va(struct kgd_dev *kgd,
 	 */
 }
 
-/* FIXME: Does this need to be ASIC-specific code? */
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-	const union amdgpu_firmware_header *hdr;
-
-	switch (type) {
-	case KGD_ENGINE_PFP:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.pfp_fw->data;
-		break;
-
-	case KGD_ENGINE_ME:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.me_fw->data;
-		break;
-
-	case KGD_ENGINE_CE:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.ce_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC1:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC2:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec2_fw->data;
-		break;
-
-	case KGD_ENGINE_RLC:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.rlc_fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA1:
-		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[0].fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA2:
-		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[1].fw->data;
-		break;
-
-	default:
-		return 0;
-	}
-
-	if (hdr == NULL)
-		return 0;
-
-	/* Only 12 bit in use*/
-	return hdr->common.ucode_version;
-}
-
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base)
 {

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -3437,7 +3437,7 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,
 
 				vram_lost = amdgpu_device_check_vram_lost(tmp_adev);
 				if (vram_lost) {
-					DRM_ERROR("VRAM is lost!\n");
+					DRM_INFO("VRAM is lost due to GPU reset!\n");
 					atomic_inc(&tmp_adev->vram_lost_counter);
 				}
 

drivers/gpu/drm/amd/amdgpu/amdgpu_object.c

@@ -88,12 +88,14 @@ static void amdgpu_bo_destroy(struct ttm_buffer_object *tbo)
 	if (bo->gem_base.import_attach)
 		drm_prime_gem_destroy(&bo->gem_base, bo->tbo.sg);
 	drm_gem_object_release(&bo->gem_base);
-	amdgpu_bo_unref(&bo->parent);
+	/* in case amdgpu_device_recover_vram got NULL of bo->parent */
 	if (!list_empty(&bo->shadow_list)) {
 		mutex_lock(&adev->shadow_list_lock);
 		list_del_init(&bo->shadow_list);
 		mutex_unlock(&adev->shadow_list_lock);
 	}
+	amdgpu_bo_unref(&bo->parent);
+
 	kfree(bo->metadata);
 	kfree(bo);
 }

drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c

@@ -144,7 +144,7 @@ static ssize_t amdgpu_get_dpm_state(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	enum amd_pm_state_type pm;
 
-	if (adev->smu.ppt_funcs->get_current_power_state)
+	if (is_support_sw_smu(adev) && adev->smu.ppt_funcs->get_current_power_state)
 		pm = amdgpu_smu_get_current_power_state(adev);
 	else if (adev->powerplay.pp_funcs->get_current_power_state)
 		pm = amdgpu_dpm_get_current_power_state(adev);

drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c

@@ -36,6 +36,7 @@ void amdgpu_virt_init_setting(struct amdgpu_device *adev)
 	/* enable virtual display */
 	adev->mode_info.num_crtc = 1;
 	adev->enable_virtual_display = true;
+	adev->ddev->driver->driver_features &= ~DRIVER_ATOMIC;
 	adev->cg_flags = 0;
 	adev->pg_flags = 0;
 }

drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c

@@ -515,7 +515,7 @@ static void xgpu_vi_mailbox_flr_work(struct work_struct *work)
 
 	/* wait until RCV_MSG become 3 */
 	if (xgpu_vi_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL)) {
-		pr_err("failed to recieve FLR_CMPL\n");
+		pr_err("failed to receive FLR_CMPL\n");
 		return;
 	}
 

drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c

@@ -156,7 +156,6 @@ static const struct soc15_reg_golden golden_settings_sdma0_4_2[] =
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC7_RB_RPTR_ADDR_LO, 0xfffffffd, 0x00000001),
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC7_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0),
-	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_WATERMK, 0xFE000000, 0x00000000),
 };
 
 static const struct soc15_reg_golden golden_settings_sdma1_4_2[] = {
@@ -186,7 +185,6 @@ static const struct soc15_reg_golden golden_settings_sdma1_4_2[] = {
 	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC7_RB_RPTR_ADDR_LO, 0xfffffffd, 0x00000001),
 	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC7_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
 	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_UTCL1_PAGE, 0x000003ff, 0x000003c0),
-	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_UTCL1_WATERMK, 0xFE000000, 0x00000000),
 };
 
 static const struct soc15_reg_golden golden_settings_sdma_rv1[] =
@@ -851,7 +849,7 @@ static void sdma_v4_0_gfx_resume(struct amdgpu_device *adev, unsigned int i)
 	wptr_poll_cntl = RREG32_SDMA(i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL);
 	wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl,
 				       SDMA0_GFX_RB_WPTR_POLL_CNTL,
-				       F32_POLL_ENABLE, amdgpu_sriov_vf(adev));
+				       F32_POLL_ENABLE, amdgpu_sriov_vf(adev)? 1 : 0);
 	WREG32_SDMA(i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL, wptr_poll_cntl);
 
 	/* enable DMA RB */
@@ -942,7 +940,7 @@ static void sdma_v4_0_page_resume(struct amdgpu_device *adev, unsigned int i)
 	wptr_poll_cntl = RREG32_SDMA(i, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL);
 	wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl,
 				       SDMA0_PAGE_RB_WPTR_POLL_CNTL,
-				       F32_POLL_ENABLE, amdgpu_sriov_vf(adev));
+				       F32_POLL_ENABLE, amdgpu_sriov_vf(adev)? 1 : 0);
 	WREG32_SDMA(i, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL, wptr_poll_cntl);
 
 	/* enable DMA RB */

drivers/gpu/drm/amd/amdgpu/soc15.c

@@ -470,6 +470,12 @@ static int soc15_asic_reset(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		soc15_asic_get_baco_capability(adev, &baco_reset);
 		break;
+	case CHIP_VEGA20:
+		if (adev->psp.sos_fw_version >= 0x80067)
+			soc15_asic_get_baco_capability(adev, &baco_reset);
+		else
+			baco_reset = false;
+		break;
 	default:
 		baco_reset = false;
 		break;
@@ -895,7 +901,8 @@ static int soc15_common_early_init(void *handle)
 
 			adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN;
 		} else if (adev->pdev->device == 0x15d8) {
-			adev->cg_flags = AMD_CG_SUPPORT_GFX_MGLS |
+			adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG |
+				AMD_CG_SUPPORT_GFX_MGLS |
 				AMD_CG_SUPPORT_GFX_CP_LS |
 				AMD_CG_SUPPORT_GFX_3D_CGCG |
 				AMD_CG_SUPPORT_GFX_3D_CGLS |

drivers/gpu/drm/amd/amdgpu/vce_v2_0.c

@@ -283,7 +283,7 @@ static int vce_v2_0_stop(struct amdgpu_device *adev)
 	}
 
 	if (vce_v2_0_wait_for_idle(adev)) {
-		DRM_INFO("VCE is busy, Can't set clock gateing");
+		DRM_INFO("VCE is busy, Can't set clock gating");
 		return 0;
 	}
 

drivers/gpu/drm/amd/amdgpu/vce_v4_0.c

@@ -382,6 +382,7 @@ static int vce_v4_0_start(struct amdgpu_device *adev)
 static int vce_v4_0_stop(struct amdgpu_device *adev)
 {
 
+	/* Disable VCPU */
 	WREG32_P(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CNTL), 0, ~0x200001);
 
 	/* hold on ECPU */
@@ -389,8 +390,8 @@ static int vce_v4_0_stop(struct amdgpu_device *adev)
 			VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
 			~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
 
-	/* clear BUSY flag */
-	WREG32_P(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), 0, ~VCE_STATUS__JOB_BUSY_MASK);
+	/* clear VCE_STATUS */
+	WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), 0);
 
 	/* Set Clock-Gating off */
 	/* if (adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)
@@ -922,6 +923,7 @@ static int vce_v4_0_set_clockgating_state(void *handle,
 
 	return 0;
 }
+#endif
 
 static int vce_v4_0_set_powergating_state(void *handle,
 					  enum amd_powergating_state state)
@@ -935,16 +937,11 @@ static int vce_v4_0_set_powergating_state(void *handle,
 	 */
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!(adev->pg_flags & AMD_PG_SUPPORT_VCE))
-		return 0;
-
 	if (state == AMD_PG_STATE_GATE)
-		/* XXX do we need a vce_v4_0_stop()? */
-		return 0;
+		return vce_v4_0_stop(adev);
 	else
 		return vce_v4_0_start(adev);
 }
-#endif
 
 static void vce_v4_0_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_job *job,
 					struct amdgpu_ib *ib, uint32_t flags)
@@ -1059,7 +1056,7 @@ const struct amd_ip_funcs vce_v4_0_ip_funcs = {
 	.soft_reset = NULL /* vce_v4_0_soft_reset */,
 	.post_soft_reset = NULL /* vce_v4_0_post_soft_reset */,
 	.set_clockgating_state = vce_v4_0_set_clockgating_state,
-	.set_powergating_state = NULL /* vce_v4_0_set_powergating_state */,
+	.set_powergating_state = vce_v4_0_set_powergating_state,
 };
 
 static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = {

drivers/gpu/drm/amd/amdkfd/kfd_device.c

@@ -494,9 +494,9 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 {
 	unsigned int size;
 
-	kfd->mec_fw_version = kfd->kfd2kgd->get_fw_version(kfd->kgd,
+	kfd->mec_fw_version = amdgpu_amdkfd_get_fw_version(kfd->kgd,
 			KGD_ENGINE_MEC1);
-	kfd->sdma_fw_version = kfd->kfd2kgd->get_fw_version(kfd->kgd,
+	kfd->sdma_fw_version = amdgpu_amdkfd_get_fw_version(kfd->kgd,
 			KGD_ENGINE_SDMA1);
 	kfd->shared_resources = *gpu_resources;
 

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h

@@ -271,6 +271,9 @@ struct dm_crtc_state {
 	struct drm_crtc_state base;
 	struct dc_stream_state *stream;
 
+	int active_planes;
+	bool interrupts_enabled;
+
 	int crc_skip_count;
 	bool crc_enabled;
 

drivers/gpu/drm/amd/display/dc/core/dc_link.c

@@ -514,6 +514,40 @@ static void link_disconnect_remap(struct dc_sink *prev_sink, struct dc_link *lin
 }
 
 
+static void read_edp_current_link_settings_on_detect(struct dc_link *link)
+{
+	union lane_count_set lane_count_set = { {0} };
+	uint8_t link_bw_set;
+	uint8_t link_rate_set;
+
+	// Read DPCD 00101h to find out the number of lanes currently set
+	core_link_read_dpcd(link, DP_LANE_COUNT_SET,
+			&lane_count_set.raw, sizeof(lane_count_set));
+	link->cur_link_settings.lane_count = lane_count_set.bits.LANE_COUNT_SET;
+
+	// Read DPCD 00100h to find if standard link rates are set
+	core_link_read_dpcd(link, DP_LINK_BW_SET,
+			&link_bw_set, sizeof(link_bw_set));
+
+	if (link_bw_set == 0) {
+		/* If standard link rates are not being used,
+		 * Read DPCD 00115h to find the link rate set used
+		 */
+		core_link_read_dpcd(link, DP_LINK_RATE_SET,
+				&link_rate_set, sizeof(link_rate_set));
+
+		if (link_rate_set < link->dpcd_caps.edp_supported_link_rates_count) {
+			link->cur_link_settings.link_rate =
+				link->dpcd_caps.edp_supported_link_rates[link_rate_set];
+			link->cur_link_settings.link_rate_set = link_rate_set;
+			link->cur_link_settings.use_link_rate_set = true;
+		}
+	} else {
+		link->cur_link_settings.link_rate = link_bw_set;
+		link->cur_link_settings.use_link_rate_set = false;
+	}
+}
+
 static bool detect_dp(
 	struct dc_link *link,
 	struct display_sink_capability *sink_caps,
@@ -648,9 +682,14 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		return false;
 	}
 
-	if (link->connector_signal == SIGNAL_TYPE_EDP &&
-			link->local_sink)
-		return true;
+	if (link->connector_signal == SIGNAL_TYPE_EDP) {
+		/* On detect, we want to make sure current link settings are
+		 * up to date, especially if link was powered on by GOP.
+		 */
+		read_edp_current_link_settings_on_detect(link);
+		if (link->local_sink)
+			return true;
+	}
 
 	if (link->connector_signal == SIGNAL_TYPE_LVDS &&
 			link->local_sink)
@@ -1396,13 +1435,19 @@ static enum dc_status enable_link_dp(
 	/* get link settings for video mode timing */
 	decide_link_settings(stream, &link_settings);
 
-	/* If link settings are different than current and link already enabled
-	 * then need to disable before programming to new rate.
-	 */
-	if (link->link_status.link_active &&
-		(link->cur_link_settings.lane_count != link_settings.lane_count ||
-		 link->cur_link_settings.link_rate != link_settings.link_rate)) {
-		dp_disable_link_phy(link, pipe_ctx->stream->signal);
+	if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP) {
+		/* If link settings are different than current and link already enabled
+		 * then need to disable before programming to new rate.
+		 */
+		if (link->link_status.link_active &&
+			(link->cur_link_settings.lane_count != link_settings.lane_count ||
+			 link->cur_link_settings.link_rate != link_settings.link_rate)) {
+			dp_disable_link_phy(link, pipe_ctx->stream->signal);
+		}
+
+		/*in case it is not on*/
+		link->dc->hwss.edp_power_control(link, true);
+		link->dc->hwss.edp_wait_for_hpd_ready(link, true);
 	}
 
 	pipe_ctx->stream_res.pix_clk_params.requested_sym_clk =
@@ -1448,15 +1493,9 @@ static enum dc_status enable_link_edp(
 		struct pipe_ctx *pipe_ctx)
 {
 	enum dc_status status;
-	struct dc_stream_state *stream = pipe_ctx->stream;
-	struct dc_link *link = stream->link;
-	/*in case it is not on*/
-	link->dc->hwss.edp_power_control(link, true);
-	link->dc->hwss.edp_wait_for_hpd_ready(link, true);
 
 	status = enable_link_dp(state, pipe_ctx);
 
-
 	return status;
 }
 

drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c

@@ -2185,6 +2185,30 @@ static int translate_dpcd_max_bpc(enum dpcd_downstream_port_max_bpc bpc)
 	return -1;
 }
 
+static void read_dp_device_vendor_id(struct dc_link *link)
+{
+	struct dp_device_vendor_id dp_id;
+
+	/* read IEEE branch device id */
+	core_link_read_dpcd(
+		link,
+		DP_BRANCH_OUI,
+		(uint8_t *)&dp_id,
+		sizeof(dp_id));
+
+	link->dpcd_caps.branch_dev_id =
+		(dp_id.ieee_oui[0] << 16) +
+		(dp_id.ieee_oui[1] << 8) +
+		dp_id.ieee_oui[2];
+
+	memmove(
+		link->dpcd_caps.branch_dev_name,
+		dp_id.ieee_device_id,
+		sizeof(dp_id.ieee_device_id));
+}
+
+
+
 static void get_active_converter_info(
 	uint8_t data, struct dc_link *link)
 {
@@ -2270,27 +2294,6 @@ static void get_active_converter_info(
 
 	ddc_service_set_dongle_type(link->ddc, link->dpcd_caps.dongle_type);
 
-	{
-		struct dp_device_vendor_id dp_id;
-
-		/* read IEEE branch device id */
-		core_link_read_dpcd(
-			link,
-			DP_BRANCH_OUI,
-			(uint8_t *)&dp_id,
-			sizeof(dp_id));
-
-		link->dpcd_caps.branch_dev_id =
-			(dp_id.ieee_oui[0] << 16) +
-			(dp_id.ieee_oui[1] << 8) +
-			dp_id.ieee_oui[2];
-
-		memmove(
-			link->dpcd_caps.branch_dev_name,
-			dp_id.ieee_device_id,
-			sizeof(dp_id.ieee_device_id));
-	}
-
 	{
 		struct dp_sink_hw_fw_revision dp_hw_fw_revision;
 
@@ -2455,6 +2458,8 @@ static bool retrieve_link_cap(struct dc_link *link)
 	ds_port.byte = dpcd_data[DP_DOWNSTREAMPORT_PRESENT -
 				 DP_DPCD_REV];
 
+	read_dp_device_vendor_id(link);
+
 	get_active_converter_info(ds_port.byte, link);
 
 	dp_wa_power_up_0010FA(link, dpcd_data, sizeof(dpcd_data));
@@ -2586,9 +2591,6 @@ void detect_edp_sink_caps(struct dc_link *link)
 	uint32_t entry;
 	uint32_t link_rate_in_khz;
 	enum dc_link_rate link_rate = LINK_RATE_UNKNOWN;
-	union lane_count_set lane_count_set = { {0} };
-	uint8_t link_bw_set;
-	uint8_t link_rate_set;
 
 	retrieve_link_cap(link);
 	link->dpcd_caps.edp_supported_link_rates_count = 0;
@@ -2614,33 +2616,6 @@ void detect_edp_sink_caps(struct dc_link *link)
 		}
 	}
 	link->verified_link_cap = link->reported_link_cap;
-
-	// Read DPCD 00101h to find out the number of lanes currently set
-	core_link_read_dpcd(link, DP_LANE_COUNT_SET,
-			&lane_count_set.raw, sizeof(lane_count_set));
-	link->cur_link_settings.lane_count = lane_count_set.bits.LANE_COUNT_SET;
-
-	// Read DPCD 00100h to find if standard link rates are set
-	core_link_read_dpcd(link, DP_LINK_BW_SET,
-			&link_bw_set, sizeof(link_bw_set));
-
-	if (link_bw_set == 0) {
-		/* If standard link rates are not being used,
-		 * Read DPCD 00115h to find the link rate set used
-		 */
-		core_link_read_dpcd(link, DP_LINK_RATE_SET,
-				&link_rate_set, sizeof(link_rate_set));
-
-		if (link_rate_set < link->dpcd_caps.edp_supported_link_rates_count) {
-			link->cur_link_settings.link_rate =
-				link->dpcd_caps.edp_supported_link_rates[link_rate_set];
-			link->cur_link_settings.link_rate_set = link_rate_set;
-			link->cur_link_settings.use_link_rate_set = true;
-		}
-	} else {
-		link->cur_link_settings.link_rate = link_bw_set;
-		link->cur_link_settings.use_link_rate_set = false;
-	}
 }
 
 void dc_link_dp_enable_hpd(const struct dc_link *link)

drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c

@@ -58,6 +58,8 @@ void dp_enable_link_phy(
 	const struct dc_link_settings *link_settings)
 {
 	struct link_encoder *link_enc = link->link_enc;
+	struct dc  *core_dc = link->ctx->dc;
+	struct dmcu *dmcu = core_dc->res_pool->dmcu;
 
 	struct pipe_ctx *pipes =
 			link->dc->current_state->res_ctx.pipe_ctx;
@@ -84,6 +86,9 @@ void dp_enable_link_phy(
 		}
 	}
 
+	if (dmcu != NULL && dmcu->funcs->lock_phy)
+		dmcu->funcs->lock_phy(dmcu);
+
 	if (dc_is_dp_sst_signal(signal)) {
 		link_enc->funcs->enable_dp_output(
 						link_enc,
@@ -95,6 +100,10 @@ void dp_enable_link_phy(
 						link_settings,
 						clock_source);
 	}
+
+	if (dmcu != NULL && dmcu->funcs->unlock_phy)
+		dmcu->funcs->unlock_phy(dmcu);
+
 	link->cur_link_settings = *link_settings;
 
 	dp_receiver_power_ctrl(link, true);
@@ -150,15 +159,25 @@ bool edp_receiver_ready_T7(struct dc_link *link)
 
 void dp_disable_link_phy(struct dc_link *link, enum signal_type signal)
 {
+	struct dc  *core_dc = link->ctx->dc;
+	struct dmcu *dmcu = core_dc->res_pool->dmcu;
+
 	if (!link->wa_flags.dp_keep_receiver_powered)
 		dp_receiver_power_ctrl(link, false);
 
 	if (signal == SIGNAL_TYPE_EDP) {
 		link->link_enc->funcs->disable_output(link->link_enc, signal);
 		link->dc->hwss.edp_power_control(link, false);
-	} else
+	} else {
+		if (dmcu != NULL && dmcu->funcs->lock_phy)
+			dmcu->funcs->lock_phy(dmcu);
+
 		link->link_enc->funcs->disable_output(link->link_enc, signal);
 
+		if (dmcu != NULL && dmcu->funcs->unlock_phy)
+			dmcu->funcs->unlock_phy(dmcu);
+	}
+
 	/* Clear current link setting.*/
 	memset(&link->cur_link_settings, 0,
 			sizeof(link->cur_link_settings));

drivers/gpu/drm/amd/display/dc/core/dc_stream.c

@@ -163,6 +163,27 @@ struct dc_stream_state *dc_create_stream_for_sink(
 	return stream;
 }
 
+struct dc_stream_state *dc_copy_stream(const struct dc_stream_state *stream)
+{
+	struct dc_stream_state *new_stream;
+
+	new_stream = kzalloc(sizeof(struct dc_stream_state), GFP_KERNEL);
+	if (!new_stream)
+		return NULL;
+
+	memcpy(new_stream, stream, sizeof(struct dc_stream_state));
+
+	if (new_stream->sink)
+		dc_sink_retain(new_stream->sink);
+
+	if (new_stream->out_transfer_func)
+		dc_transfer_func_retain(new_stream->out_transfer_func);
+
+	kref_init(&new_stream->refcount);
+
+	return new_stream;
+}
+
 /**
  * dc_stream_get_status_from_state - Get stream status from given dc state
  * @state: DC state to find the stream status in
@@ -312,7 +333,7 @@ bool dc_stream_set_cursor_position(
 				(!pipe_ctx->plane_res.mi  && !pipe_ctx->plane_res.hubp) ||
 				!pipe_ctx->plane_state ||
 				(!pipe_ctx->plane_res.xfm && !pipe_ctx->plane_res.dpp) ||
-				!pipe_ctx->plane_res.ipp)
+				(!pipe_ctx->plane_res.ipp && !pipe_ctx->plane_res.dpp))
 			continue;
 
 		if (!pipe_to_program) {

drivers/gpu/drm/amd/display/dc/dc.h

@@ -39,7 +39,7 @@
 #include "inc/hw/dmcu.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.2.26"
+#define DC_VER "3.2.27"
 
 #define MAX_SURFACES 3
 #define MAX_PLANES 6
@@ -204,6 +204,7 @@ struct dc_config {
 	bool optimize_edp_link_rate;
 	bool disable_fractional_pwm;
 	bool allow_seamless_boot_optimization;
+	bool power_down_display_on_boot;
 };
 
 enum visual_confirm {

drivers/gpu/drm/amd/display/dc/dc_link.h

@@ -120,6 +120,7 @@ struct dc_link {
 	/* MST record stream using this link */
 	struct link_flags {
 		bool dp_keep_receiver_powered;
+		bool dp_skip_DID2;
 	} wa_flags;
 	struct link_mst_stream_allocation_table mst_stream_alloc_table;
 

drivers/gpu/drm/amd/display/dc/dc_stream.h

@@ -307,6 +307,8 @@ enum surface_update_type dc_check_update_surfaces_for_stream(
  */
 struct dc_stream_state *dc_create_stream_for_sink(struct dc_sink *dc_sink);
 
+struct dc_stream_state *dc_copy_stream(const struct dc_stream_state *stream);
+
 void update_stream_signal(struct dc_stream_state *stream, struct dc_sink *sink);
 
 void dc_stream_retain(struct dc_stream_state *dc_stream);

drivers/gpu/drm/amd/display/dc/dce/dce_abm.c

@@ -50,7 +50,6 @@
 #define MCP_ABM_LEVEL_SET 0x65
 #define MCP_ABM_PIPE_SET 0x66
 #define MCP_BL_SET 0x67
-#define MCP_BL_SET_PWM_FRAC     0x6A  /* Enable or disable Fractional PWM */
 
 #define MCP_DISABLE_ABM_IMMEDIATELY 255
 
@@ -391,23 +390,6 @@ static bool dce_abm_init_backlight(struct abm *abm)
 	REG_UPDATE(BL_PWM_GRP1_REG_LOCK,
 			BL_PWM_GRP1_REG_LOCK, 0);
 
-	/* Wait until microcontroller is ready to process interrupt */
-	REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 100, 800);
-
-	/* Set PWM fractional enable/disable */
-	value = (abm->ctx->dc->config.disable_fractional_pwm == false) ? 1 : 0;
-	REG_WRITE(MASTER_COMM_DATA_REG1, value);
-
-	/* Set command to enable or disable fractional PWM microcontroller */
-	REG_UPDATE(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE0,
-			MCP_BL_SET_PWM_FRAC);
-
-	/* Notify microcontroller of new command */
-	REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1);
-
-	/* Ensure command has been executed before continuing */
-	REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 100, 800);
-
 	return true;
 }
 

drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c

@@ -51,6 +51,9 @@
 #define PSR_SET_WAITLOOP 0x31
 #define MCP_INIT_DMCU 0x88
 #define MCP_INIT_IRAM 0x89
+#define MCP_SYNC_PHY_LOCK 0x90
+#define MCP_SYNC_PHY_UNLOCK 0x91
+#define MCP_BL_SET_PWM_FRAC 0x6A  /* Enable or disable Fractional PWM */
 #define MASTER_COMM_CNTL_REG__MASTER_COMM_INTERRUPT_MASK   0x00000001L
 
 static bool dce_dmcu_init(struct dmcu *dmcu)
@@ -339,9 +342,32 @@ static void dcn10_get_dmcu_version(struct dmcu *dmcu)
 			IRAM_RD_ADDR_AUTO_INC, 0);
 }
 
+static void dcn10_dmcu_enable_fractional_pwm(struct dmcu *dmcu,
+		uint32_t fractional_pwm)
+{
+	struct dce_dmcu *dmcu_dce = TO_DCE_DMCU(dmcu);
+
+	/* Wait until microcontroller is ready to process interrupt */
+	REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 100, 800);
+
+	/* Set PWM fractional enable/disable */
+	REG_WRITE(MASTER_COMM_DATA_REG1, fractional_pwm);
+
+	/* Set command to enable or disable fractional PWM microcontroller */
+	REG_UPDATE(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE0,
+			MCP_BL_SET_PWM_FRAC);
+
+	/* Notify microcontroller of new command */
+	REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1);
+
+	/* Ensure command has been executed before continuing */
+	REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 100, 800);
+}
+
 static bool dcn10_dmcu_init(struct dmcu *dmcu)
 {
 	struct dce_dmcu *dmcu_dce = TO_DCE_DMCU(dmcu);
+	const struct dc_config *config = &dmcu->ctx->dc->config;
 	bool status = false;
 
 	/*  Definition of DC_DMCU_SCRATCH
@@ -379,9 +405,14 @@ static bool dcn10_dmcu_init(struct dmcu *dmcu)
 		if (dmcu->dmcu_state == DMCU_RUNNING) {
 			/* Retrieve and cache the DMCU firmware version. */
 			dcn10_get_dmcu_version(dmcu);
+
+			/* Initialize DMCU to use fractional PWM or not */
+			dcn10_dmcu_enable_fractional_pwm(dmcu,
+				(config->disable_fractional_pwm == false) ? 1 : 0);
 			status = true;
-		} else
+		} else {
 			status = false;
+		}
 
 		break;
 	case DMCU_RUNNING:
@@ -690,7 +721,7 @@ static bool dcn10_is_dmcu_initialized(struct dmcu *dmcu)
 	return true;
 }
 
-#endif
+#endif //(CONFIG_DRM_AMD_DC_DCN1_0)
 
 static const struct dmcu_funcs dce_funcs = {
 	.dmcu_init = dce_dmcu_init,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_clk_mgr.c

@@ -151,9 +151,6 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 	struct dc *dc = clk_mgr->ctx->dc;
 	struct dc_debug_options *debug = &dc->debug;
 	struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk;
-	struct pp_smu_display_requirement_rv *smu_req_cur =
-			&dc->res_pool->pp_smu_req;
-	struct pp_smu_display_requirement_rv smu_req = *smu_req_cur;
 	struct pp_smu_funcs_rv *pp_smu = NULL;
 	bool send_request_to_increase = false;
 	bool send_request_to_lower = false;
@@ -175,8 +172,6 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 		 */
 		if (pp_smu && pp_smu->set_display_count)
 			pp_smu->set_display_count(&pp_smu->pp_smu, display_count);
-
-		smu_req.display_count = display_count;
 	}
 
 	if (new_clocks->dispclk_khz > clk_mgr->clks.dispclk_khz
@@ -187,7 +182,6 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 
 	if (should_set_clock(safe_to_lower, new_clocks->phyclk_khz, clk_mgr->clks.phyclk_khz)) {
 		clk_mgr->clks.phyclk_khz = new_clocks->phyclk_khz;
-
 		send_request_to_lower = true;
 	}
 
@@ -197,24 +191,18 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 
 	if (should_set_clock(safe_to_lower, new_clocks->fclk_khz, clk_mgr->clks.fclk_khz)) {
 		clk_mgr->clks.fclk_khz = new_clocks->fclk_khz;
-		smu_req.hard_min_fclk_mhz = new_clocks->fclk_khz / 1000;
-
 		send_request_to_lower = true;
 	}
 
 	//DCF Clock
 	if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, clk_mgr->clks.dcfclk_khz)) {
 		clk_mgr->clks.dcfclk_khz = new_clocks->dcfclk_khz;
-		smu_req.hard_min_dcefclk_mhz = new_clocks->dcfclk_khz / 1000;
-
 		send_request_to_lower = true;
 	}
 
 	if (should_set_clock(safe_to_lower,
 			new_clocks->dcfclk_deep_sleep_khz, clk_mgr->clks.dcfclk_deep_sleep_khz)) {
 		clk_mgr->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz;
-		smu_req.min_deep_sleep_dcefclk_mhz = (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000;
-
 		send_request_to_lower = true;
 	}
 
@@ -227,9 +215,9 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 				pp_smu->set_hard_min_dcfclk_by_freq &&
 				pp_smu->set_min_deep_sleep_dcfclk) {
 
-			pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_fclk_mhz);
-			pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_dcefclk_mhz);
-			pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, smu_req.min_deep_sleep_dcefclk_mhz);
+			pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, new_clocks->fclk_khz / 1000);
+			pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, new_clocks->dcfclk_khz / 1000);
+			pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000);
 		}
 	}
 
@@ -239,7 +227,6 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 			|| new_clocks->dispclk_khz == clk_mgr->clks.dispclk_khz) {
 		dcn1_ramp_up_dispclk_with_dpp(clk_mgr, new_clocks);
 		clk_mgr->clks.dispclk_khz = new_clocks->dispclk_khz;
-
 		send_request_to_lower = true;
 	}
 
@@ -249,13 +236,11 @@ static void dcn1_update_clocks(struct clk_mgr *clk_mgr,
 				pp_smu->set_hard_min_dcfclk_by_freq &&
 				pp_smu->set_min_deep_sleep_dcfclk) {
 
-			pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_fclk_mhz);
-			pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, smu_req.hard_min_dcefclk_mhz);
-			pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, smu_req.min_deep_sleep_dcefclk_mhz);
+			pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, new_clocks->fclk_khz / 1000);
+			pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, new_clocks->dcfclk_khz / 1000);
+			pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000);
 		}
 	}
-
-	*smu_req_cur = smu_req;
 }
 static const struct clk_mgr_funcs dcn1_funcs = {
 	.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c

@@ -283,7 +283,8 @@ void hubbub1_program_watermarks(
 		hubbub1->watermarks.a.urgent_ns = watermarks->a.urgent_ns;
 		prog_wm_value = convert_and_clamp(watermarks->a.urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, 0,
+				DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value);
 
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n",
@@ -310,7 +311,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, 0,
+					DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
@@ -323,7 +325,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->a.cstate_pstate.cstate_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, 0,
+					DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value);
@@ -337,7 +340,8 @@ void hubbub1_program_watermarks(
 		prog_wm_value = convert_and_clamp(
 				watermarks->a.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, 0,
+				DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value);
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value);
@@ -348,7 +352,8 @@ void hubbub1_program_watermarks(
 		hubbub1->watermarks.b.urgent_ns = watermarks->b.urgent_ns;
 		prog_wm_value = convert_and_clamp(watermarks->b.urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, 0,
+				DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value);
 
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n",
@@ -375,7 +380,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, 0,
+					DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_B calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
@@ -388,7 +394,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->b.cstate_pstate.cstate_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, 0,
+					DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value);
@@ -402,7 +409,8 @@ void hubbub1_program_watermarks(
 		prog_wm_value = convert_and_clamp(
 				watermarks->b.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, 0,
+				DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value);
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value);
@@ -413,7 +421,8 @@ void hubbub1_program_watermarks(
 		hubbub1->watermarks.c.urgent_ns = watermarks->c.urgent_ns;
 		prog_wm_value = convert_and_clamp(watermarks->c.urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, 0,
+				DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value);
 
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n",
@@ -440,7 +449,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, 0,
+					DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_C calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
@@ -453,7 +463,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->c.cstate_pstate.cstate_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, 0,
+					DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value);
@@ -467,7 +478,8 @@ void hubbub1_program_watermarks(
 		prog_wm_value = convert_and_clamp(
 				watermarks->c.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, 0,
+				DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value);
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value);
@@ -478,7 +490,8 @@ void hubbub1_program_watermarks(
 		hubbub1->watermarks.d.urgent_ns = watermarks->d.urgent_ns;
 		prog_wm_value = convert_and_clamp(watermarks->d.urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, 0,
+				DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value);
 
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n",
@@ -505,7 +518,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, 0,
+					DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_D calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
@@ -518,7 +532,8 @@ void hubbub1_program_watermarks(
 			prog_wm_value = convert_and_clamp(
 					watermarks->d.cstate_pstate.cstate_exit_ns,
 					refclk_mhz, 0x1fffff);
-			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value);
+			REG_SET(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, 0,
+					DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value);
 			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n"
 				"HW register value = 0x%x\n",
 				watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value);
@@ -532,7 +547,8 @@ void hubbub1_program_watermarks(
 		prog_wm_value = convert_and_clamp(
 				watermarks->d.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value);
+		REG_SET(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, 0,
+				DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value);
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value);
@@ -867,6 +883,7 @@ static const struct hubbub_funcs hubbub1_funcs = {
 	.dcc_support_pixel_format = hubbub1_dcc_support_pixel_format,
 	.get_dcc_compression_cap = hubbub1_get_dcc_compression_cap,
 	.wm_read_state = hubbub1_wm_read_state,
+	.program_watermarks = hubbub1_program_watermarks,
 };
 
 void hubbub1_construct(struct hubbub *hubbub,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h

@@ -32,18 +32,14 @@
 #define TO_DCN10_HUBBUB(hubbub)\
 	container_of(hubbub, struct dcn10_hubbub, base)
 
-#define HUBHUB_REG_LIST_DCN()\
+#define HUBBUB_REG_LIST_DCN_COMMON()\
 	SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A),\
-	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A),\
 	SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A),\
 	SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B),\
-	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B),\
 	SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B),\
 	SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C),\
-	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C),\
 	SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C),\
 	SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D),\
-	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D),\
 	SR(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D),\
 	SR(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL),\
 	SR(DCHUBBUB_ARB_DRAM_STATE_CNTL),\
@@ -54,6 +50,12 @@
 	SR(DCHUBBUB_TEST_DEBUG_DATA),\
 	SR(DCHUBBUB_SOFT_RESET)
 
+#define HUBBUB_VM_REG_LIST() \
+	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A),\
+	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B),\
+	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C),\
+	SR(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D)
+
 #define HUBBUB_SR_WATERMARK_REG_LIST()\
 	SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A),\
 	SR(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A),\
@@ -65,7 +67,8 @@
 	SR(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D)
 
 #define HUBBUB_REG_LIST_DCN10(id)\
-	HUBHUB_REG_LIST_DCN(), \
+	HUBBUB_REG_LIST_DCN_COMMON(), \
+	HUBBUB_VM_REG_LIST(), \
 	HUBBUB_SR_WATERMARK_REG_LIST(), \
 	SR(DCHUBBUB_SDPIF_FB_TOP),\
 	SR(DCHUBBUB_SDPIF_FB_BASE),\
@@ -122,8 +125,7 @@ struct dcn_hubbub_registers {
 #define HUBBUB_SF(reg_name, field_name, post_fix)\
 	.field_name = reg_name ## __ ## field_name ## post_fix
 
-
-#define HUBBUB_MASK_SH_LIST_DCN(mask_sh)\
+#define HUBBUB_MASK_SH_LIST_DCN_COMMON(mask_sh)\
 		HUBBUB_SF(DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_SOFT_RESET, DCHUBBUB_GLOBAL_SOFT_RESET, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, mask_sh), \
@@ -133,10 +135,29 @@ struct dcn_hubbub_registers {
 		HUBBUB_SF(DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_VALUE, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_PSTATE_CHANGE_FORCE_ENABLE, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_SAT_LEVEL, DCHUBBUB_ARB_SAT_LEVEL, mask_sh), \
-		HUBBUB_SF(DCHUBBUB_ARB_DF_REQ_OUTSTAND, DCHUBBUB_ARB_MIN_REQ_OUTSTAND, mask_sh)
+		HUBBUB_SF(DCHUBBUB_ARB_DF_REQ_OUTSTAND, DCHUBBUB_ARB_MIN_REQ_OUTSTAND, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, mask_sh)
+
+#define HUBBUB_MASK_SH_LIST_STUTTER(mask_sh) \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, mask_sh)
 
 #define HUBBUB_MASK_SH_LIST_DCN10(mask_sh)\
-		HUBBUB_MASK_SH_LIST_DCN(mask_sh), \
+		HUBBUB_MASK_SH_LIST_DCN_COMMON(mask_sh), \
+		HUBBUB_MASK_SH_LIST_STUTTER(mask_sh), \
 		HUBBUB_SF(DCHUBBUB_SDPIF_FB_TOP, SDPIF_FB_TOP, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_SDPIF_FB_BASE, SDPIF_FB_BASE, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_SDPIF_FB_OFFSET, SDPIF_FB_OFFSET, mask_sh), \
@@ -167,15 +188,35 @@ struct dcn_hubbub_registers {
 		type FB_OFFSET;\
 		type AGP_BOT;\
 		type AGP_TOP;\
-		type AGP_BASE
+		type AGP_BASE;\
+		type DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A;\
+		type DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B;\
+		type DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C;\
+		type DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D;\
+		type DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A;\
+		type DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B;\
+		type DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C;\
+		type DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D
+
+#define HUBBUB_STUTTER_REG_FIELD_LIST(type) \
+		type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A;\
+		type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B;\
+		type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C;\
+		type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D;\
+		type DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A;\
+		type DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B;\
+		type DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C;\
+		type DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D
 
 
 struct dcn_hubbub_shift {
 	DCN_HUBBUB_REG_FIELD_LIST(uint8_t);
+	HUBBUB_STUTTER_REG_FIELD_LIST(uint8_t);
 };
 
 struct dcn_hubbub_mask {
 	DCN_HUBBUB_REG_FIELD_LIST(uint32_t);
+	HUBBUB_STUTTER_REG_FIELD_LIST(uint32_t);
 };
 
 struct dc;

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c

@@ -1178,6 +1178,10 @@ void hubp1_vtg_sel(struct hubp *hubp, uint32_t otg_inst)
 	REG_UPDATE(DCHUBP_CNTL, HUBP_VTG_SEL, otg_inst);
 }
 
+void hubp1_init(struct hubp *hubp)
+{
+	//do nothing
+}
 static const struct hubp_funcs dcn10_hubp_funcs = {
 	.hubp_program_surface_flip_and_addr =
 			hubp1_program_surface_flip_and_addr,
@@ -1201,7 +1205,7 @@ static const struct hubp_funcs dcn10_hubp_funcs = {
 	.hubp_clear_underflow = hubp1_clear_underflow,
 	.hubp_disable_control =  hubp1_disable_control,
 	.hubp_get_underflow_status = hubp1_get_underflow_status,
-
+	.hubp_init = hubp1_init,
 };
 
 /*****************************************/

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h

@@ -34,6 +34,7 @@
 #define HUBP_REG_LIST_DCN(id)\
 	SRI(DCHUBP_CNTL, HUBP, id),\
 	SRI(HUBPREQ_DEBUG_DB, HUBP, id),\
+	SRI(HUBPREQ_DEBUG, HUBP, id),\
 	SRI(DCSURF_ADDR_CONFIG, HUBP, id),\
 	SRI(DCSURF_TILING_CONFIG, HUBP, id),\
 	SRI(DCSURF_SURFACE_PITCH, HUBPREQ, id),\
@@ -138,6 +139,7 @@
 #define HUBP_COMMON_REG_VARIABLE_LIST \
 	uint32_t DCHUBP_CNTL; \
 	uint32_t HUBPREQ_DEBUG_DB; \
+	uint32_t HUBPREQ_DEBUG; \
 	uint32_t DCSURF_ADDR_CONFIG; \
 	uint32_t DCSURF_TILING_CONFIG; \
 	uint32_t DCSURF_SURFACE_PITCH; \
@@ -749,4 +751,6 @@ enum cursor_pitch hubp1_get_cursor_pitch(unsigned int pitch);
 void hubp1_vready_workaround(struct hubp *hubp,
 		struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest);
 
+void hubp1_init(struct hubp *hubp);
+
 #endif

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c

@@ -1118,14 +1118,17 @@ static void dcn10_init_hw(struct dc *dc)
 	 * Otherwise, if taking control is not possible, we need to power
 	 * everything down.
 	 */
-	if (dcb->funcs->is_accelerated_mode(dcb)) {
+	if (dcb->funcs->is_accelerated_mode(dcb) || dc->config.power_down_display_on_boot) {
 		for (i = 0; i < dc->res_pool->pipe_count; i++) {
 			struct hubp *hubp = dc->res_pool->hubps[i];
 			struct dpp *dpp = dc->res_pool->dpps[i];
 
+			hubp->funcs->hubp_init(hubp);
 			dc->res_pool->opps[i]->mpc_tree_params.opp_id = dc->res_pool->opps[i]->inst;
 			plane_atomic_power_down(dc, dpp, hubp);
 		}
+
+		apply_DEGVIDCN10_253_wa(dc);
 	}
 
 	for (i = 0; i < dc->res_pool->audio_count; i++) {
@@ -2436,6 +2439,8 @@ static void dcn10_prepare_bandwidth(
 		struct dc *dc,
 		struct dc_state *context)
 {
+	struct hubbub *hubbub = dc->res_pool->hubbub;
+
 	if (dc->debug.sanity_checks)
 		dcn10_verify_allow_pstate_change_high(dc);
 
@@ -2449,7 +2454,7 @@ static void dcn10_prepare_bandwidth(
 				false);
 	}
 
-	hubbub1_program_watermarks(dc->res_pool->hubbub,
+	hubbub->funcs->program_watermarks(hubbub,
 			&context->bw_ctx.bw.dcn.watermarks,
 			dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000,
 			true);
@@ -2466,6 +2471,8 @@ static void dcn10_optimize_bandwidth(
 		struct dc *dc,
 		struct dc_state *context)
 {
+	struct hubbub *hubbub = dc->res_pool->hubbub;
+
 	if (dc->debug.sanity_checks)
 		dcn10_verify_allow_pstate_change_high(dc);
 
@@ -2479,7 +2486,7 @@ static void dcn10_optimize_bandwidth(
 				true);
 	}
 
-	hubbub1_program_watermarks(dc->res_pool->hubbub,
+	hubbub->funcs->program_watermarks(hubbub,
 			&context->bw_ctx.bw.dcn.watermarks,
 			dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000,
 			true);

drivers/gpu/drm/amd/display/dc/dm_pp_smu.h

@@ -74,29 +74,6 @@ struct pp_smu_wm_range_sets {
 	struct pp_smu_wm_set_range writer_wm_sets[MAX_WATERMARK_SETS];
 };
 
-struct pp_smu_display_requirement_rv {
-	/* PPSMC_MSG_SetDisplayCount: count
-	 *  0 triggers S0i2 optimization
-	 */
-	unsigned int display_count;
-
-	/* PPSMC_MSG_SetHardMinFclkByFreq: mhz
-	 *  FCLK will vary with DPM, but never below requested hard min
-	 */
-	unsigned int hard_min_fclk_mhz;
-
-	/* PPSMC_MSG_SetHardMinDcefclkByFreq: mhz
-	 *  fixed clock at requested freq, either from FCH bypass or DFS
-	 */
-	unsigned int hard_min_dcefclk_mhz;
-
-	/* PPSMC_MSG_SetMinDeepSleepDcefclk: mhz
-	 *  when DF is in cstate, dcf clock is further divided down
-	 *  to just above given frequency
-	 */
-	unsigned int min_deep_sleep_dcefclk_mhz;
-};
-
 struct pp_smu_funcs_rv {
 	struct pp_smu pp_smu;
 

drivers/gpu/drm/amd/display/dc/inc/core_types.h

@@ -145,7 +145,6 @@ struct resource_pool {
 	struct hubbub *hubbub;
 	struct mpc *mpc;
 	struct pp_smu_funcs *pp_smu;
-	struct pp_smu_display_requirement_rv pp_smu_req;
 	struct dce_aux *engines[MAX_PIPES];
 	struct dce_i2c_hw *hw_i2cs[MAX_PIPES];
 	struct dce_i2c_sw *sw_i2cs[MAX_PIPES];

drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h

@@ -77,6 +77,12 @@ struct hubbub_funcs {
 	void (*get_dchub_ref_freq)(struct hubbub *hubbub,
 			unsigned int dccg_ref_freq_inKhz,
 			unsigned int *dchub_ref_freq_inKhz);
+
+	void (*program_watermarks)(
+			struct hubbub *hubbub,
+			struct dcn_watermark_set *watermarks,
+			unsigned int refclk_mhz,
+			bool safe_to_lower);
 };
 
 struct hubbub {

drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h

@@ -70,6 +70,8 @@ struct dmcu_funcs {
 	void (*get_psr_wait_loop)(struct dmcu *dmcu,
 			unsigned int *psr_wait_loop_number);
 	bool (*is_dmcu_initialized)(struct dmcu *dmcu);
+	bool (*lock_phy)(struct dmcu *dmcu);
+	bool (*unlock_phy)(struct dmcu *dmcu);
 };
 
 #endif

drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h

@@ -130,6 +130,7 @@ struct hubp_funcs {
 	void (*hubp_clear_underflow)(struct hubp *hubp);
 	void (*hubp_disable_control)(struct hubp *hubp, bool disable_hubp);
 	unsigned int (*hubp_get_underflow_status)(struct hubp *hubp);
+	void (*hubp_init)(struct hubp *hubp);
 
 };
 

drivers/gpu/drm/amd/display/modules/freesync/freesync.c

@@ -437,10 +437,8 @@ static void apply_below_the_range(struct core_freesync *core_freesync,
 			inserted_frame_duration_in_us = last_render_time_in_us /
 							frames_to_insert;
 
-		if (inserted_frame_duration_in_us <
-			(1000000 / in_out_vrr->max_refresh_in_uhz))
-			inserted_frame_duration_in_us =
-				(1000000 / in_out_vrr->max_refresh_in_uhz);
+		if (inserted_frame_duration_in_us < in_out_vrr->min_duration_in_us)
+			inserted_frame_duration_in_us = in_out_vrr->min_duration_in_us;
 
 		/* Cache the calculated variables */
 		in_out_vrr->btr.inserted_duration_in_us =

drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h

@@ -2347,6 +2347,8 @@
 #define mmHUBP0_DCHUBP_VMPG_CONFIG_BASE_IDX                                                            2
 #define mmHUBP0_HUBPREQ_DEBUG_DB                                                                       0x0569
 #define mmHUBP0_HUBPREQ_DEBUG_DB_BASE_IDX                                                              2
+#define mmHUBP0_HUBPREQ_DEBUG                                                                          0x056a
+#define mmHUBP0_HUBPREQ_DEBUG_BASE_IDX                                                                 2
 #define mmHUBP0_HUBP_MEASURE_WIN_CTRL_DCFCLK                                                           0x056e
 #define mmHUBP0_HUBP_MEASURE_WIN_CTRL_DCFCLK_BASE_IDX                                                  2
 #define mmHUBP0_HUBP_MEASURE_WIN_CTRL_DPPCLK                                                           0x056f
@@ -2631,6 +2633,8 @@
 #define mmHUBP1_DCHUBP_VMPG_CONFIG_BASE_IDX                                                            2
 #define mmHUBP1_HUBPREQ_DEBUG_DB                                                                       0x062d
 #define mmHUBP1_HUBPREQ_DEBUG_DB_BASE_IDX                                                              2
+#define mmHUBP1_HUBPREQ_DEBUG                                                                          0x062e
+#define mmHUBP1_HUBPREQ_DEBUG_BASE_IDX                                                                 2
 #define mmHUBP1_HUBP_MEASURE_WIN_CTRL_DCFCLK                                                           0x0632
 #define mmHUBP1_HUBP_MEASURE_WIN_CTRL_DCFCLK_BASE_IDX                                                  2
 #define mmHUBP1_HUBP_MEASURE_WIN_CTRL_DPPCLK                                                           0x0633
@@ -2915,6 +2919,8 @@
 #define mmHUBP2_DCHUBP_VMPG_CONFIG_BASE_IDX                                                            2
 #define mmHUBP2_HUBPREQ_DEBUG_DB                                                                       0x06f1
 #define mmHUBP2_HUBPREQ_DEBUG_DB_BASE_IDX                                                              2
+#define mmHUBP2_HUBPREQ_DEBUG                                                                          0x06f2
+#define mmHUBP2_HUBPREQ_DEBUG_BASE_IDX                                                                 2
 #define mmHUBP2_HUBP_MEASURE_WIN_CTRL_DCFCLK                                                           0x06f6
 #define mmHUBP2_HUBP_MEASURE_WIN_CTRL_DCFCLK_BASE_IDX                                                  2
 #define mmHUBP2_HUBP_MEASURE_WIN_CTRL_DPPCLK                                                           0x06f7
@@ -3199,6 +3205,8 @@
 #define mmHUBP3_DCHUBP_VMPG_CONFIG_BASE_IDX                                                            2
 #define mmHUBP3_HUBPREQ_DEBUG_DB                                                                       0x07b5
 #define mmHUBP3_HUBPREQ_DEBUG_DB_BASE_IDX                                                              2
+#define mmHUBP3_HUBPREQ_DEBUG                                                                          0x07b6
+#define mmHUBP3_HUBPREQ_DEBUG_BASE_IDX                                                                 2
 #define mmHUBP3_HUBP_MEASURE_WIN_CTRL_DCFCLK                                                           0x07ba
 #define mmHUBP3_HUBP_MEASURE_WIN_CTRL_DCFCLK_BASE_IDX                                                  2
 #define mmHUBP3_HUBP_MEASURE_WIN_CTRL_DPPCLK                                                           0x07bb

drivers/gpu/drm/amd/include/atomfirmware.h

@@ -718,6 +718,7 @@ enum atom_encoder_caps_def
   ATOM_ENCODER_CAP_RECORD_HBR2_EN               =0x02,         // DP1.2 HBR2 setting is qualified and HBR2 can be enabled 
   ATOM_ENCODER_CAP_RECORD_HDMI6Gbps_EN          =0x04,         // HDMI2.0 6Gbps enable or not. 
   ATOM_ENCODER_CAP_RECORD_HBR3_EN               =0x08,         // DP1.3 HBR3 is supported by board. 
+  ATOM_ENCODER_CAP_RECORD_USB_C_TYPE            =0x100,        // the DP connector is a USB-C type.
 };
 
 struct  atom_encoder_caps_record

drivers/gpu/drm/amd/include/kgd_kfd_interface.h

@@ -85,18 +85,6 @@ enum kgd_memory_pool {
 	KGD_POOL_FRAMEBUFFER = 3,
 };
 
-enum kgd_engine_type {
-	KGD_ENGINE_PFP = 1,
-	KGD_ENGINE_ME,
-	KGD_ENGINE_CE,
-	KGD_ENGINE_MEC1,
-	KGD_ENGINE_MEC2,
-	KGD_ENGINE_RLC,
-	KGD_ENGINE_SDMA1,
-	KGD_ENGINE_SDMA2,
-	KGD_ENGINE_MAX
-};
-
 /**
  * enum kfd_sched_policy
  *
@@ -230,8 +218,6 @@ struct tile_config {
  * @hqd_sdma_destroy: Destructs and preempts the SDMA queue assigned to that
  * SDMA hqd slot.
  *
- * @get_fw_version: Returns FW versions from the header
- *
  * @set_scratch_backing_va: Sets VA for scratch backing memory of a VMID.
  * Only used for no cp scheduling mode
  *
@@ -311,8 +297,6 @@ struct kfd2kgd_calls {
 					struct kgd_dev *kgd,
 					uint8_t vmid);
 
-	uint16_t (*get_fw_version)(struct kgd_dev *kgd,
-				enum kgd_engine_type type);
 	void (*set_scratch_backing_va)(struct kgd_dev *kgd,
 				uint64_t va, uint32_t vmid);
 	int (*get_tile_config)(struct kgd_dev *kgd, struct tile_config *config);

drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c

@@ -35,6 +35,7 @@
 #include "smu10_hwmgr.h"
 #include "power_state.h"
 #include "soc15_common.h"
+#include "smu10.h"
 
 #define SMU10_MAX_DEEPSLEEP_DIVIDER_ID     5
 #define SMU10_MINIMUM_ENGINE_CLOCK         800   /* 8Mhz, the low boundary of engine clock allowed on this chip */
@@ -204,18 +205,13 @@ static int smu10_set_clock_limit(struct pp_hwmgr *hwmgr, const void *input)
 	return 0;
 }
 
-static inline uint32_t convert_10k_to_mhz(uint32_t clock)
-{
-	return (clock + 99) / 100;
-}
-
 static int smu10_set_min_deep_sleep_dcefclk(struct pp_hwmgr *hwmgr, uint32_t clock)
 {
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 
 	if (smu10_data->need_min_deep_sleep_dcefclk &&
-		smu10_data->deep_sleep_dcefclk != convert_10k_to_mhz(clock)) {
-		smu10_data->deep_sleep_dcefclk = convert_10k_to_mhz(clock);
+		smu10_data->deep_sleep_dcefclk != clock) {
+		smu10_data->deep_sleep_dcefclk = clock;
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 					PPSMC_MSG_SetMinDeepSleepDcefclk,
 					smu10_data->deep_sleep_dcefclk);
@@ -228,8 +224,8 @@ static int smu10_set_hard_min_dcefclk_by_freq(struct pp_hwmgr *hwmgr, uint32_t c
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 
 	if (smu10_data->dcf_actual_hard_min_freq &&
-		smu10_data->dcf_actual_hard_min_freq != convert_10k_to_mhz(clock)) {
-		smu10_data->dcf_actual_hard_min_freq = convert_10k_to_mhz(clock);
+		smu10_data->dcf_actual_hard_min_freq != clock) {
+		smu10_data->dcf_actual_hard_min_freq = clock;
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 					PPSMC_MSG_SetHardMinDcefclkByFreq,
 					smu10_data->dcf_actual_hard_min_freq);
@@ -242,8 +238,8 @@ static int smu10_set_hard_min_fclk_by_freq(struct pp_hwmgr *hwmgr, uint32_t cloc
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 
 	if (smu10_data->f_actual_hard_min_freq &&
-		smu10_data->f_actual_hard_min_freq != convert_10k_to_mhz(clock)) {
-		smu10_data->f_actual_hard_min_freq = convert_10k_to_mhz(clock);
+		smu10_data->f_actual_hard_min_freq != clock) {
+		smu10_data->f_actual_hard_min_freq = clock;
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 					PPSMC_MSG_SetHardMinFclkByFreq,
 					smu10_data->f_actual_hard_min_freq);
@@ -572,7 +568,6 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 				enum amd_dpm_forced_level level)
 {
 	struct smu10_hwmgr *data = hwmgr->backend;
-	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t min_sclk = hwmgr->display_config->min_core_set_clock;
 	uint32_t min_mclk = hwmgr->display_config->min_mem_set_clock/100;
 
@@ -581,11 +576,6 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 		return 0;
 	}
 
-	/* Disable UMDPSTATE support on rv2 temporarily */
-	if ((adev->asic_type == CHIP_RAVEN) &&
-	    (adev->rev_id >= 8))
-		return 0;
-
 	if (min_sclk < data->gfx_min_freq_limit)
 		min_sclk = data->gfx_min_freq_limit;
 
@@ -1200,6 +1190,94 @@ static void smu10_powergate_vcn(struct pp_hwmgr *hwmgr, bool bgate)
 	}
 }
 
+static int conv_power_profile_to_pplib_workload(int power_profile)
+{
+	int pplib_workload = 0;
+
+	switch (power_profile) {
+	case PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT:
+		pplib_workload = WORKLOAD_DEFAULT_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_FULLSCREEN3D:
+		pplib_workload = WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_POWERSAVING:
+		pplib_workload = WORKLOAD_PPLIB_POWER_SAVING_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_VIDEO:
+		pplib_workload = WORKLOAD_PPLIB_VIDEO_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_VR:
+		pplib_workload = WORKLOAD_PPLIB_VR_BIT;
+		break;
+	case PP_SMC_POWER_PROFILE_COMPUTE:
+		pplib_workload = WORKLOAD_PPLIB_COMPUTE_BIT;
+		break;
+	}
+
+	return pplib_workload;
+}
+
+static int smu10_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
+{
+	uint32_t i, size = 0;
+	static const uint8_t
+		profile_mode_setting[6][4] = {{70, 60, 0, 0,},
+						{70, 60, 1, 3,},
+						{90, 60, 0, 0,},
+						{70, 60, 0, 0,},
+						{70, 90, 0, 0,},
+						{30, 60, 0, 6,},
+						};
+	static const char *profile_name[6] = {
+					"BOOTUP_DEFAULT",
+					"3D_FULL_SCREEN",
+					"POWER_SAVING",
+					"VIDEO",
+					"VR",
+					"COMPUTE"};
+	static const char *title[6] = {"NUM",
+			"MODE_NAME",
+			"BUSY_SET_POINT",
+			"FPS",
+			"USE_RLC_BUSY",
+			"MIN_ACTIVE_LEVEL"};
+
+	if (!buf)
+		return -EINVAL;
+
+	size += sprintf(buf + size, "%s %16s %s %s %s %s\n",title[0],
+			title[1], title[2], title[3], title[4], title[5]);
+
+	for (i = 0; i <= PP_SMC_POWER_PROFILE_COMPUTE; i++)
+		size += sprintf(buf + size, "%3d %14s%s: %14d %3d %10d %14d\n",
+			i, profile_name[i], (i == hwmgr->power_profile_mode) ? "*" : " ",
+			profile_mode_setting[i][0], profile_mode_setting[i][1],
+			profile_mode_setting[i][2], profile_mode_setting[i][3]);
+
+	return size;
+}
+
+static int smu10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, uint32_t size)
+{
+	int workload_type = 0;
+
+	if (input[size] > PP_SMC_POWER_PROFILE_COMPUTE) {
+		pr_err("Invalid power profile mode %ld\n", input[size]);
+		return -EINVAL;
+	}
+	hwmgr->power_profile_mode = input[size];
+
+	/* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */
+	workload_type =
+		conv_power_profile_to_pplib_workload(hwmgr->power_profile_mode);
+	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_ActiveProcessNotify,
+						1 << workload_type);
+
+	return 0;
+}
+
+
 static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.backend_init = smu10_hwmgr_backend_init,
 	.backend_fini = smu10_hwmgr_backend_fini,
@@ -1241,6 +1319,8 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.powergate_sdma = smu10_powergate_sdma,
 	.set_hard_min_dcefclk_by_freq = smu10_set_hard_min_dcefclk_by_freq,
 	.set_hard_min_fclk_by_freq = smu10_set_hard_min_fclk_by_freq,
+	.get_power_profile_mode = smu10_get_power_profile_mode,
+	.set_power_profile_mode = smu10_set_power_profile_mode,
 };
 
 int smu10_init_function_pointers(struct pp_hwmgr *hwmgr)

drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c

@@ -3460,7 +3460,18 @@ static void vega20_power_gate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 		return ;
 
 	data->vce_power_gated = bgate;
-	vega20_enable_disable_vce_dpm(hwmgr, !bgate);
+	if (bgate) {
+		vega20_enable_disable_vce_dpm(hwmgr, !bgate);
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
+						AMD_IP_BLOCK_TYPE_VCE,
+						AMD_PG_STATE_GATE);
+	} else {
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
+						AMD_IP_BLOCK_TYPE_VCE,
+						AMD_PG_STATE_UNGATE);
+		vega20_enable_disable_vce_dpm(hwmgr, !bgate);
+	}
+
 }
 
 static void vega20_power_gate_uvd(struct pp_hwmgr *hwmgr, bool bgate)

drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h

@@ -85,7 +85,6 @@
 #define PPSMC_MSG_SetRccPfcPmeRestoreRegister   0x36
 #define PPSMC_Message_Count                     0x37
 
-
 typedef uint16_t PPSMC_Result;
 typedef int      PPSMC_Msg;
 

drivers/gpu/drm/amd/powerplay/inc/smu10.h

@@ -136,12 +136,14 @@
 #define FEATURE_CORE_CSTATES_MASK     (1 << FEATURE_CORE_CSTATES_BIT)
 
 /* Workload bits */
-#define WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT 0
-#define WORKLOAD_PPLIB_VIDEO_BIT          2
-#define WORKLOAD_PPLIB_VR_BIT             3
-#define WORKLOAD_PPLIB_COMPUTE_BIT        4
-#define WORKLOAD_PPLIB_CUSTOM_BIT         5
-#define WORKLOAD_PPLIB_COUNT              6
+#define WORKLOAD_DEFAULT_BIT              0
+#define WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT 1
+#define WORKLOAD_PPLIB_POWER_SAVING_BIT   2
+#define WORKLOAD_PPLIB_VIDEO_BIT          3
+#define WORKLOAD_PPLIB_VR_BIT             4
+#define WORKLOAD_PPLIB_COMPUTE_BIT        5
+#define WORKLOAD_PPLIB_CUSTOM_BIT         6
+#define WORKLOAD_PPLIB_COUNT              7
 
 typedef struct {
 	/* MP1_EXT_SCRATCH0 */

drivers/gpu/drm/amd/powerplay/smu_v11_0.c

@@ -1896,8 +1896,13 @@ static int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
 static int smu_v11_0_set_xgmi_pstate(struct smu_context *smu,
 				     uint32_t pstate)
 {
-	/* send msg to SMU to set pstate */
-	return 0;
+	int ret = 0;
+	mutex_lock(&(smu->mutex));
+	ret = smu_send_smc_msg_with_param(smu,
+					  SMU_MSG_SetXgmiMode,
+					  pstate ? XGMI_STATE_D0 : XGMI_STATE_D3);
+	mutex_unlock(&(smu->mutex));
+	return ret;
 }
 
 static const struct smu_funcs smu_v11_0_funcs = {