amd-drm-fixes-6.9-2024-03-21:

amdgpu:
 - Freesync fixes
 - UAF IOCTL fixes
 - Fix mmhub client ID mapping
 - IH 7.0 fix
 - DML2 fixes
 - VCN 4.0.6 fix
 - GART bind fix
 - GPU reset fix
 - SR-IOV fix
 - OD table handling fixes
 - Fix TA handling on boards without display hardware
 - DML1 fix
 - ABM fix
 - eDP panel fix
 - DPPCLK fix
 - HDCP fix
 - Revert incorrect error case handling in ioremap
 - VPE fix
 - HDMI fixes
 - SDMA 4.4.2 fix
 - Other misc fixes
 
 amdkfd:
 - Fix duplicate BO handling in process restore
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQQgO5Idg2tXNTSZAr293/aFa7yZ2AUCZfyuKAAKCRC93/aFa7yZ
 2KkzAQD0AQbf7C2JB7fpSBPi/r8mzLBpWYs0VlbOfkrf8TnjQAD/XTv/KXnm1JDb
 dXVvyagc5TITqxlyjwcJMDuH7jqXIwI=
 =IZBa
 -----END PGP SIGNATURE-----

Merge tag 'amd-drm-fixes-6.9-2024-03-21' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-fixes-6.9-2024-03-21:

amdgpu:
- Freesync fixes
- UAF IOCTL fixes
- Fix mmhub client ID mapping
- IH 7.0 fix
- DML2 fixes
- VCN 4.0.6 fix
- GART bind fix
- GPU reset fix
- SR-IOV fix
- OD table handling fixes
- Fix TA handling on boards without display hardware
- DML1 fix
- ABM fix
- eDP panel fix
- DPPCLK fix
- HDCP fix
- Revert incorrect error case handling in ioremap
- VPE fix
- HDMI fixes
- SDMA 4.4.2 fix
- Other misc fixes

amdkfd:
- Fix duplicate BO handling in process restore

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240321220514.1418288-1-alexander.deucher@amd.com

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

@@ -146,7 +146,7 @@ int amdgpu_amdkfd_drm_client_create(struct amdgpu_device *adev)
 {
 	int ret;
 
-	if (!adev->kfd.init_complete)
+	if (!adev->kfd.init_complete || adev->kfd.client.dev)
 		return 0;
 
 	ret = drm_client_init(&adev->ddev, &adev->kfd.client, "kfd",

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

@@ -2869,14 +2869,16 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence __rcu *
 
 	mutex_lock(&process_info->lock);
 
-	drm_exec_init(&exec, 0, 0);
+	drm_exec_init(&exec, DRM_EXEC_IGNORE_DUPLICATES, 0);
 	drm_exec_until_all_locked(&exec) {
 		list_for_each_entry(peer_vm, &process_info->vm_list_head,
 				    vm_list_node) {
 			ret = amdgpu_vm_lock_pd(peer_vm, &exec, 2);
 			drm_exec_retry_on_contention(&exec);
-			if (unlikely(ret))
+			if (unlikely(ret)) {
+				pr_err("Locking VM PD failed, ret: %d\n", ret);
 				goto ttm_reserve_fail;
+			}
 		}
 
 		/* Reserve all BOs and page tables/directory. Add all BOs from
@@ -2889,8 +2891,10 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence __rcu *
 			gobj = &mem->bo->tbo.base;
 			ret = drm_exec_prepare_obj(&exec, gobj, 1);
 			drm_exec_retry_on_contention(&exec);
-			if (unlikely(ret))
+			if (unlikely(ret)) {
+				pr_err("drm_exec_prepare_obj failed, ret: %d\n", ret);
 				goto ttm_reserve_fail;
+			}
 		}
 	}
 
@@ -2950,8 +2954,10 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence __rcu *
 	 * validations above would invalidate DMABuf imports again.
 	 */
 	ret = process_validate_vms(process_info, &exec.ticket);
-	if (ret)
+	if (ret) {
+		pr_debug("Validating VMs failed, ret: %d\n", ret);
 		goto validate_map_fail;
+	}
 
 	/* Update mappings not managed by KFD */
 	list_for_each_entry(peer_vm, &process_info->vm_list_head,

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

@@ -4040,10 +4040,8 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	 * early on during init and before calling to RREG32.
 	 */
 	adev->reset_domain = amdgpu_reset_create_reset_domain(SINGLE_DEVICE, "amdgpu-reset-dev");
-	if (!adev->reset_domain) {
+	if (!adev->reset_domain)
-		r = -ENOMEM;
+		return -ENOMEM;
-		goto unmap_memory;
-	}
 
 	/* detect hw virtualization here */
 	amdgpu_detect_virtualization(adev);
@@ -4053,7 +4051,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_device_get_job_timeout_settings(adev);
 	if (r) {
 		dev_err(adev->dev, "invalid lockup_timeout parameter syntax\n");
-		goto unmap_memory;
+		return r;
 	}
 
 	amdgpu_device_set_mcbp(adev);
@@ -4061,12 +4059,12 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	/* early init functions */
 	r = amdgpu_device_ip_early_init(adev);
 	if (r)
-		goto unmap_memory;
+		return r;
 
 	/* Get rid of things like offb */
 	r = drm_aperture_remove_conflicting_pci_framebuffers(adev->pdev, &amdgpu_kms_driver);
 	if (r)
-		goto unmap_memory;
+		return r;
 
 	/* Enable TMZ based on IP_VERSION */
 	amdgpu_gmc_tmz_set(adev);
@@ -4076,7 +4074,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (adev->gmc.xgmi.supported) {
 		r = adev->gfxhub.funcs->get_xgmi_info(adev);
 		if (r)
-			goto unmap_memory;
+			return r;
 	}
 
 	/* enable PCIE atomic ops */
@@ -4345,8 +4343,6 @@ release_ras_con:
 failed:
 	amdgpu_vf_error_trans_all(adev);
 
-unmap_memory:
-	iounmap(adev->rmmio);
 	return r;
 }
 

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

@@ -2479,8 +2479,11 @@ static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work)
 	}
 	for (i = 0; i < mgpu_info.num_dgpu; i++) {
 		adev = mgpu_info.gpu_ins[i].adev;
-		if (!adev->kfd.init_complete)
+		if (!adev->kfd.init_complete) {
+			kgd2kfd_init_zone_device(adev);
 			amdgpu_amdkfd_device_init(adev);
+			amdgpu_amdkfd_drm_client_create(adev);
+		}
 		amdgpu_ttm_set_buffer_funcs_status(adev, true);
 	}
 }

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

@@ -687,7 +687,7 @@ int amdgpu_gfx_enable_kgq(struct amdgpu_device *adev, int xcc_id)
 	r = amdgpu_ring_test_helper(kiq_ring);
 	spin_unlock(&kiq->ring_lock);
 	if (r)
-		DRM_ERROR("KCQ enable failed\n");
+		DRM_ERROR("KGQ enable failed\n");
 
 	return r;
 }

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

@@ -129,13 +129,25 @@ static const struct mmu_interval_notifier_ops amdgpu_hmm_hsa_ops = {
  */
 int amdgpu_hmm_register(struct amdgpu_bo *bo, unsigned long addr)
 {
+	int r;
+
 	if (bo->kfd_bo)
-		return mmu_interval_notifier_insert(&bo->notifier, current->mm,
+		r = mmu_interval_notifier_insert(&bo->notifier, current->mm,
 						    addr, amdgpu_bo_size(bo),
 						    &amdgpu_hmm_hsa_ops);
-	return mmu_interval_notifier_insert(&bo->notifier, current->mm, addr,
+	else
-					    amdgpu_bo_size(bo),
+		r = mmu_interval_notifier_insert(&bo->notifier, current->mm, addr,
-					    &amdgpu_hmm_gfx_ops);
+							amdgpu_bo_size(bo),
+							&amdgpu_hmm_gfx_ops);
+	if (r)
+		/*
+		 * Make sure amdgpu_hmm_unregister() doesn't call
+		 * mmu_interval_notifier_remove() when the notifier isn't properly
+		 * initialized.
+		 */
+		bo->notifier.mm = NULL;
+
+	return r;
 }
 
 /**

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

@@ -1830,6 +1830,10 @@ static int psp_hdcp_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
+	/* bypass hdcp initialization if dmu is harvested */
+	if (!amdgpu_device_has_display_hardware(psp->adev))
+		return 0;
+
 	if (!psp->hdcp_context.context.bin_desc.size_bytes ||
 	    !psp->hdcp_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "HDCP: optional hdcp ta ucode is not available\n");
@@ -1862,6 +1866,9 @@ int psp_hdcp_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
+	if (!psp->hdcp_context.context.initialized)
+		return 0;
+
 	return psp_ta_invoke(psp, ta_cmd_id, &psp->hdcp_context.context);
 }
 
@@ -1897,6 +1904,10 @@ static int psp_dtm_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
+	/* bypass dtm initialization if dmu is harvested */
+	if (!amdgpu_device_has_display_hardware(psp->adev))
+		return 0;
+
 	if (!psp->dtm_context.context.bin_desc.size_bytes ||
 	    !psp->dtm_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "DTM: optional dtm ta ucode is not available\n");
@@ -1929,6 +1940,9 @@ int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
+	if (!psp->dtm_context.context.initialized)
+		return 0;
+
 	return psp_ta_invoke(psp, ta_cmd_id, &psp->dtm_context.context);
 }
 
@@ -2063,6 +2077,10 @@ static int psp_securedisplay_initialize(struct psp_context *psp)
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
+	/* bypass securedisplay initialization if dmu is harvested */
+	if (!amdgpu_device_has_display_hardware(psp->adev))
+		return 0;
+
 	if (!psp->securedisplay_context.context.bin_desc.size_bytes ||
 	    !psp->securedisplay_context.context.bin_desc.start_addr) {
 		dev_info(psp->adev->dev, "SECUREDISPLAY: securedisplay ta ucode is not available\n");

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

@@ -864,6 +864,7 @@ static void amdgpu_ttm_gart_bind(struct amdgpu_device *adev,
 		amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages,
 				 gtt->ttm.dma_address, flags);
 	}
+	gtt->bound = true;
 }
 
 /*

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

@@ -60,6 +60,7 @@
 #define FIRMWARE_VCN4_0_4		"amdgpu/vcn_4_0_4.bin"
 #define FIRMWARE_VCN4_0_5		"amdgpu/vcn_4_0_5.bin"
 #define FIRMWARE_VCN4_0_6		"amdgpu/vcn_4_0_6.bin"
+#define FIRMWARE_VCN4_0_6_1		"amdgpu/vcn_4_0_6_1.bin"
 #define FIRMWARE_VCN5_0_0		"amdgpu/vcn_5_0_0.bin"
 
 MODULE_FIRMWARE(FIRMWARE_RAVEN);
@@ -85,6 +86,7 @@ MODULE_FIRMWARE(FIRMWARE_VCN4_0_3);
 MODULE_FIRMWARE(FIRMWARE_VCN4_0_4);
 MODULE_FIRMWARE(FIRMWARE_VCN4_0_5);
 MODULE_FIRMWARE(FIRMWARE_VCN4_0_6);
+MODULE_FIRMWARE(FIRMWARE_VCN4_0_6_1);
 MODULE_FIRMWARE(FIRMWARE_VCN5_0_0);
 
 static void amdgpu_vcn_idle_work_handler(struct work_struct *work);
@@ -93,14 +95,22 @@ int amdgpu_vcn_early_init(struct amdgpu_device *adev)
 {
 	char ucode_prefix[30];
 	char fw_name[40];
-	int r;
+	int r, i;
 
-	amdgpu_ucode_ip_version_decode(adev, UVD_HWIP, ucode_prefix, sizeof(ucode_prefix));
+	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
-	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", ucode_prefix);
+		amdgpu_ucode_ip_version_decode(adev, UVD_HWIP, ucode_prefix, sizeof(ucode_prefix));
-	r = amdgpu_ucode_request(adev, &adev->vcn.fw, fw_name);
+		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", ucode_prefix);
-	if (r)
+		if (amdgpu_ip_version(adev, UVD_HWIP, 0) ==  IP_VERSION(4, 0, 6) &&
-		amdgpu_ucode_release(&adev->vcn.fw);
+			i == 1) {
+			snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_%d.bin", ucode_prefix, i);
+		}
 
+		r = amdgpu_ucode_request(adev, &adev->vcn.fw[i], fw_name);
+		if (r) {
+			amdgpu_ucode_release(&adev->vcn.fw[i]);
+			return r;
+		}
+	}
 	return r;
 }
 
@@ -141,7 +151,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 		}
 	}
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[0]->data;
 	adev->vcn.fw_version = le32_to_cpu(hdr->ucode_version);
 
 	/* Bit 20-23, it is encode major and non-zero for new naming convention.
@@ -256,9 +266,10 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
 
 		for (i = 0; i < adev->vcn.num_enc_rings; ++i)
 			amdgpu_ring_fini(&adev->vcn.inst[j].ring_enc[i]);
+
+		amdgpu_ucode_release(&adev->vcn.fw[j]);
 	}
 
-	amdgpu_ucode_release(&adev->vcn.fw);
 	mutex_destroy(&adev->vcn.vcn1_jpeg1_workaround);
 	mutex_destroy(&adev->vcn.vcn_pg_lock);
 
@@ -354,11 +365,12 @@ int amdgpu_vcn_resume(struct amdgpu_device *adev)
 			const struct common_firmware_header *hdr;
 			unsigned int offset;
 
-			hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+			hdr = (const struct common_firmware_header *)adev->vcn.fw[i]->data;
 			if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
 				offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
 				if (drm_dev_enter(adev_to_drm(adev), &idx)) {
-					memcpy_toio(adev->vcn.inst[i].cpu_addr, adev->vcn.fw->data + offset,
+					memcpy_toio(adev->vcn.inst[i].cpu_addr,
+						    adev->vcn.fw[i]->data + offset,
 						    le32_to_cpu(hdr->ucode_size_bytes));
 					drm_dev_exit(idx);
 				}
@@ -1043,11 +1055,11 @@ void amdgpu_vcn_setup_ucode(struct amdgpu_device *adev)
 	if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 		const struct common_firmware_header *hdr;
 
-		hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
-
 		for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
 			if (adev->vcn.harvest_config & (1 << i))
 				continue;
+
+			hdr = (const struct common_firmware_header *)adev->vcn.fw[i]->data;
 			/* currently only support 2 FW instances */
 			if (i >= 2) {
 				dev_info(adev->dev, "More then 2 VCN FW instances!\n");
@@ -1055,7 +1067,7 @@ void amdgpu_vcn_setup_ucode(struct amdgpu_device *adev)
 			}
 			idx = AMDGPU_UCODE_ID_VCN + i;
 			adev->firmware.ucode[idx].ucode_id = idx;
-			adev->firmware.ucode[idx].fw = adev->vcn.fw;
+			adev->firmware.ucode[idx].fw = adev->vcn.fw[i];
 			adev->firmware.fw_size +=
 				ALIGN(le32_to_cpu(hdr->ucode_size_bytes), PAGE_SIZE);
 

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

@@ -306,7 +306,7 @@ struct amdgpu_vcn_ras {
 struct amdgpu_vcn {
 	unsigned		fw_version;
 	struct delayed_work	idle_work;
-	const struct firmware	*fw;	/* VCN firmware */
+	const struct firmware	*fw[AMDGPU_MAX_VCN_INSTANCES];	/* VCN firmware */
 	unsigned		num_enc_rings;
 	enum amd_powergating_state cur_state;
 	bool			indirect_sram;

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

@@ -575,9 +575,6 @@ static unsigned int vpe_ring_init_cond_exec(struct amdgpu_ring *ring,
 {
 	unsigned int ret;
 
-	if (ring->adev->vpe.collaborate_mode)
-		return ~0;
-
 	amdgpu_ring_write(ring, VPE_CMD_HEADER(VPE_CMD_OPCODE_COND_EXE, 0));
 	amdgpu_ring_write(ring, lower_32_bits(addr));
 	amdgpu_ring_write(ring, upper_32_bits(addr));

drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c

@@ -3657,6 +3657,9 @@ static void gfx_v10_0_init_spm_golden_registers(struct amdgpu_device *adev)
 
 static void gfx_v10_0_init_golden_registers(struct amdgpu_device *adev)
 {
+	if (amdgpu_sriov_vf(adev))
+		return;
+
 	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(10, 1, 10):
 		soc15_program_register_sequence(adev,
@@ -4982,7 +4985,8 @@ static void gfx_v10_0_constants_init(struct amdgpu_device *adev)
 	u32 tmp;
 	int i;
 
-	WREG32_FIELD15(GC, 0, GRBM_CNTL, READ_TIMEOUT, 0xff);
+	if (!amdgpu_sriov_vf(adev))
+		WREG32_FIELD15(GC, 0, GRBM_CNTL, READ_TIMEOUT, 0xff);
 
 	gfx_v10_0_setup_rb(adev);
 	gfx_v10_0_get_cu_info(adev, &adev->gfx.cu_info);
@@ -7163,7 +7167,7 @@ static int gfx_v10_0_hw_init(void *handle)
 	if (amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(10, 3, 0))
 		gfx_v10_3_program_pbb_mode(adev);
 
-	if (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(10, 3, 0))
+	if (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(10, 3, 0) && !amdgpu_sriov_vf(adev))
 		gfx_v10_3_set_power_brake_sequence(adev);
 
 	return r;

drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c

@@ -155,6 +155,9 @@ static void gfxhub_v2_1_init_system_aperture_regs(struct amdgpu_device *adev)
 {
 	uint64_t value;
 
+	if (amdgpu_sriov_vf(adev))
+		return;
+
 	/* Program the AGP BAR */
 	WREG32_SOC15(GC, 0, mmGCMC_VM_AGP_BASE, 0);
 	WREG32_SOC15(GC, 0, mmGCMC_VM_AGP_BOT, adev->gmc.agp_start >> 24);

drivers/gpu/drm/amd/amdgpu/ih_v7_0.c

@@ -418,6 +418,12 @@ static u32 ih_v7_0_get_wptr(struct amdgpu_device *adev,
 	tmp = RREG32_NO_KIQ(ih_regs->ih_rb_cntl);
 	tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 1);
 	WREG32_NO_KIQ(ih_regs->ih_rb_cntl, tmp);
+
+	/* Unset the CLEAR_OVERFLOW bit immediately so new overflows
+	 * can be detected.
+	 */
+	tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 0);
+	WREG32_NO_KIQ(ih_regs->ih_rb_cntl, tmp);
 out:
 	return (wptr & ih->ptr_mask);
 }

drivers/gpu/drm/amd/amdgpu/mmhub_v3_3.c

@@ -99,16 +99,15 @@ mmhub_v3_3_print_l2_protection_fault_status(struct amdgpu_device *adev,
 	switch (amdgpu_ip_version(adev, MMHUB_HWIP, 0)) {
 	case IP_VERSION(3, 3, 0):
 	case IP_VERSION(3, 3, 1):
-		mmhub_cid = mmhub_client_ids_v3_3[cid][rw];
+		mmhub_cid = cid < ARRAY_SIZE(mmhub_client_ids_v3_3) ?
+			    mmhub_client_ids_v3_3[cid][rw] :
+			    cid == 0x140 ? "UMSCH" : NULL;
 		break;
 	default:
 		mmhub_cid = NULL;
 		break;
 	}
 
-	if (!mmhub_cid && cid == 0x140)
-		mmhub_cid = "UMSCH";
-
 	dev_err(adev->dev, "\t Faulty UTCL2 client ID: %s (0x%x)\n",
 		mmhub_cid ? mmhub_cid : "unknown", cid);
 	dev_err(adev->dev, "\t MORE_FAULTS: 0x%lx\n",

drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c

@@ -431,16 +431,11 @@ static void sdma_v4_4_2_inst_gfx_stop(struct amdgpu_device *adev,
 	struct amdgpu_ring *sdma[AMDGPU_MAX_SDMA_INSTANCES];
 	u32 doorbell_offset, doorbell;
 	u32 rb_cntl, ib_cntl;
-	int i, unset = 0;
+	int i;
 
 	for_each_inst(i, inst_mask) {
 		sdma[i] = &adev->sdma.instance[i].ring;
 
-		if ((adev->mman.buffer_funcs_ring == sdma[i]) && unset != 1) {
-			amdgpu_ttm_set_buffer_funcs_status(adev, false);
-			unset = 1;
-		}
-
 		rb_cntl = RREG32_SDMA(i, regSDMA_GFX_RB_CNTL);
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA_GFX_RB_CNTL, RB_ENABLE, 0);
 		WREG32_SDMA(i, regSDMA_GFX_RB_CNTL, rb_cntl);
@@ -487,20 +482,10 @@ static void sdma_v4_4_2_inst_rlc_stop(struct amdgpu_device *adev,
 static void sdma_v4_4_2_inst_page_stop(struct amdgpu_device *adev,
 				       uint32_t inst_mask)
 {
-	struct amdgpu_ring *sdma[AMDGPU_MAX_SDMA_INSTANCES];
 	u32 rb_cntl, ib_cntl;
 	int i;
-	bool unset = false;
 
 	for_each_inst(i, inst_mask) {
-		sdma[i] = &adev->sdma.instance[i].page;
-
-		if ((adev->mman.buffer_funcs_ring == sdma[i]) &&
-			(!unset)) {
-			amdgpu_ttm_set_buffer_funcs_status(adev, false);
-			unset = true;
-		}
-
 		rb_cntl = RREG32_SDMA(i, regSDMA_PAGE_RB_CNTL);
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA_PAGE_RB_CNTL,
 					RB_ENABLE, 0);
@@ -950,13 +935,7 @@ static int sdma_v4_4_2_inst_start(struct amdgpu_device *adev,
 			r = amdgpu_ring_test_helper(page);
 			if (r)
 				return r;
-
-			if (adev->mman.buffer_funcs_ring == page)
-				amdgpu_ttm_set_buffer_funcs_status(adev, true);
 		}
-
-		if (adev->mman.buffer_funcs_ring == ring)
-			amdgpu_ttm_set_buffer_funcs_status(adev, true);
 	}
 
 	return r;

drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c

@@ -304,7 +304,7 @@ static int vcn_v1_0_resume(void *handle)
  */
 static void vcn_v1_0_mc_resume_spg_mode(struct amdgpu_device *adev)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[0]->size + 4);
 	uint32_t offset;
 
 	/* cache window 0: fw */
@@ -371,7 +371,7 @@ static void vcn_v1_0_mc_resume_spg_mode(struct amdgpu_device *adev)
 
 static void vcn_v1_0_mc_resume_dpg_mode(struct amdgpu_device *adev)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[0]->size + 4);
 	uint32_t offset;
 
 	/* cache window 0: fw */

drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c

@@ -330,7 +330,7 @@ static int vcn_v2_0_resume(void *handle)
  */
 static void vcn_v2_0_mc_resume(struct amdgpu_device *adev)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[0]->size + 4);
 	uint32_t offset;
 
 	if (amdgpu_sriov_vf(adev))
@@ -386,7 +386,7 @@ static void vcn_v2_0_mc_resume(struct amdgpu_device *adev)
 
 static void vcn_v2_0_mc_resume_dpg_mode(struct amdgpu_device *adev, bool indirect)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[0]->size + 4);
 	uint32_t offset;
 
 	/* cache window 0: fw */
@@ -1878,7 +1878,7 @@ static int vcn_v2_0_start_sriov(struct amdgpu_device *adev)
 
 		init_table += header->vcn_table_offset;
 
-		size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+		size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[0]->size + 4);
 
 		MMSCH_V2_0_INSERT_DIRECT_RD_MOD_WT(
 			SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS),

drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c

@@ -414,13 +414,15 @@ static int vcn_v2_5_resume(void *handle)
  */
 static void vcn_v2_5_mc_resume(struct amdgpu_device *adev)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size;
 	uint32_t offset;
 	int i;
 
 	for (i = 0; i < adev->vcn.num_vcn_inst; ++i) {
 		if (adev->vcn.harvest_config & (1 << i))
 			continue;
+
+		size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[i]->size + 4);
 		/* cache window 0: fw */
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 			WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
@@ -469,7 +471,7 @@ static void vcn_v2_5_mc_resume(struct amdgpu_device *adev)
 
 static void vcn_v2_5_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_idx, bool indirect)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[inst_idx]->size + 4);
 	uint32_t offset;
 
 	/* cache window 0: fw */
@@ -1240,7 +1242,7 @@ static int vcn_v2_5_sriov_start(struct amdgpu_device *adev)
 			SOC15_REG_OFFSET(VCN, i, mmUVD_STATUS),
 			~UVD_STATUS__UVD_BUSY, UVD_STATUS__UVD_BUSY);
 
-		size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+		size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[i]->size + 4);
 		/* mc resume*/
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 			MMSCH_V1_0_INSERT_DIRECT_WT(

drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c

@@ -449,7 +449,7 @@ static int vcn_v3_0_resume(void *handle)
  */
 static void vcn_v3_0_mc_resume(struct amdgpu_device *adev, int inst)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[inst]->size + 4);
 	uint32_t offset;
 
 	/* cache window 0: fw */
@@ -499,7 +499,7 @@ static void vcn_v3_0_mc_resume(struct amdgpu_device *adev, int inst)
 
 static void vcn_v3_0_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_idx, bool indirect)
 {
-	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+	uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[inst_idx]->size + 4);
 	uint32_t offset;
 
 	/* cache window 0: fw */
@@ -1332,7 +1332,7 @@ static int vcn_v3_0_start_sriov(struct amdgpu_device *adev)
 			mmUVD_STATUS),
 			~UVD_STATUS__UVD_BUSY, UVD_STATUS__UVD_BUSY);
 
-		cache_size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+		cache_size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[i]->size + 4);
 
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 			MMSCH_V3_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCN, i,

drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c

@@ -382,7 +382,7 @@ static void vcn_v4_0_mc_resume(struct amdgpu_device *adev, int inst)
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */
@@ -442,7 +442,7 @@ static void vcn_v4_0_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_idx
 {
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst_idx]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */
@@ -1289,7 +1289,7 @@ static int vcn_v4_0_start_sriov(struct amdgpu_device *adev)
 			regUVD_STATUS),
 			~UVD_STATUS__UVD_BUSY, UVD_STATUS__UVD_BUSY);
 
-		cache_size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+		cache_size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[i]->size + 4);
 
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 			MMSCH_V4_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCN, i,

drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c

@@ -332,7 +332,7 @@ static void vcn_v4_0_3_mc_resume(struct amdgpu_device *adev, int inst_idx)
 	uint32_t offset, size, vcn_inst;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst_idx]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	vcn_inst = GET_INST(VCN, inst_idx);
@@ -407,7 +407,7 @@ static void vcn_v4_0_3_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_i
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst_idx]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */
@@ -894,7 +894,7 @@ static int vcn_v4_0_3_start_sriov(struct amdgpu_device *adev)
 		MMSCH_V4_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCN, 0, regUVD_STATUS),
 			~UVD_STATUS__UVD_BUSY, UVD_STATUS__UVD_BUSY);
 
-		cache_size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
+		cache_size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw[i]->size + 4);
 
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 			MMSCH_V4_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCN, 0,

drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c

@@ -45,7 +45,7 @@
 #define mmUVD_DPG_LMA_DATA_BASE_IDX					regUVD_DPG_LMA_DATA_BASE_IDX
 
 #define VCN_VID_SOC_ADDRESS_2_0						0x1fb00
-#define VCN1_VID_SOC_ADDRESS_3_0					0x48300
+#define VCN1_VID_SOC_ADDRESS_3_0					(0x48300 + 0x38000)
 
 #define VCN_HARVEST_MMSCH							0
 
@@ -329,7 +329,7 @@ static void vcn_v4_0_5_mc_resume(struct amdgpu_device *adev, int inst)
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */
@@ -390,7 +390,7 @@ static void vcn_v4_0_5_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_i
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst_idx]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */
@@ -486,7 +486,8 @@ static void vcn_v4_0_5_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_i
 
 	/* VCN global tiling registers */
 	WREG32_SOC15_DPG_MODE(inst_idx, SOC15_DPG_MODE_OFFSET(
-		VCN, 0, regUVD_GFX10_ADDR_CONFIG), adev->gfx.config.gb_addr_config, 0, indirect);
+		VCN, inst_idx, regUVD_GFX10_ADDR_CONFIG),
+		adev->gfx.config.gb_addr_config, 0, indirect);
 }
 
 /**
@@ -911,7 +912,6 @@ static int vcn_v4_0_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, b
 		VCN, inst_idx, regUVD_MASTINT_EN),
 		UVD_MASTINT_EN__VCPU_EN_MASK, 0, indirect);
 
-
 	if (indirect)
 		amdgpu_vcn_psp_update_sram(adev, inst_idx, 0);
 

drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c

@@ -290,7 +290,7 @@ static void vcn_v5_0_0_mc_resume(struct amdgpu_device *adev, int inst)
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */
@@ -351,7 +351,7 @@ static void vcn_v5_0_0_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_i
 	uint32_t offset, size;
 	const struct common_firmware_header *hdr;
 
-	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	hdr = (const struct common_firmware_header *)adev->vcn.fw[inst_idx]->data;
 	size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	/* cache window 0: fw */

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

@@ -1767,6 +1767,9 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 	if (amdgpu_dc_debug_mask & DC_FORCE_SUBVP_MCLK_SWITCH)
 		adev->dm.dc->debug.force_subvp_mclk_switch = true;
 
+	if (amdgpu_dc_debug_mask & DC_ENABLE_DML2)
+		adev->dm.dc->debug.using_dml2 = true;
+
 	adev->dm.dc->debug.visual_confirm = amdgpu_dc_visual_confirm;
 
 	/* TODO: Remove after DP2 receiver gets proper support of Cable ID feature */
@@ -11271,18 +11274,24 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector,
 	if (!adev->dm.freesync_module)
 		goto update;
 
-	if (sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT
+	if (edid && (sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT ||
-		|| sink->sink_signal == SIGNAL_TYPE_EDP) {
+		     sink->sink_signal == SIGNAL_TYPE_EDP)) {
 		bool edid_check_required = false;
 
-		if (edid) {
+		if (is_dp_capable_without_timing_msa(adev->dm.dc,
-			edid_check_required = is_dp_capable_without_timing_msa(
+						     amdgpu_dm_connector)) {
-						adev->dm.dc,
+			if (edid->features & DRM_EDID_FEATURE_CONTINUOUS_FREQ) {
-						amdgpu_dm_connector);
+				freesync_capable = true;
+				amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq;
+				amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq;
+			} else {
+				edid_check_required = edid->version > 1 ||
+						      (edid->version == 1 &&
+						       edid->revision > 1);
+			}
 		}
 
-		if (edid_check_required == true && (edid->version > 1 ||
+		if (edid_check_required) {
-		   (edid->version == 1 && edid->revision > 1))) {
 			for (i = 0; i < 4; i++) {
 
 				timing	= &edid->detailed_timings[i];

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c

@@ -67,6 +67,8 @@ static void apply_edid_quirks(struct edid *edid, struct dc_edid_caps *edid_caps)
 	/* Workaround for some monitors that do not clear DPCD 0x317 if FreeSync is unsupported */
 	case drm_edid_encode_panel_id('A', 'U', 'O', 0xA7AB):
 	case drm_edid_encode_panel_id('A', 'U', 'O', 0xE69B):
+	case drm_edid_encode_panel_id('B', 'O', 'E', 0x092A):
+	case drm_edid_encode_panel_id('L', 'G', 'D', 0x06D1):
 		DRM_DEBUG_DRIVER("Clearing DPCD 0x317 on monitor with panel id %X\n", panel_id);
 		edid_caps->panel_patch.remove_sink_ext_caps = true;
 		break;
@@ -120,6 +122,8 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
 
 	edid_caps->edid_hdmi = connector->display_info.is_hdmi;
 
+	apply_edid_quirks(edid_buf, edid_caps);
+
 	sad_count = drm_edid_to_sad((struct edid *) edid->raw_edid, &sads);
 	if (sad_count <= 0)
 		return result;
@@ -146,8 +150,6 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
 	else
 		edid_caps->speaker_flags = DEFAULT_SPEAKER_LOCATION;
 
-	apply_edid_quirks(edid_buf, edid_caps);
-
 	kfree(sads);
 	kfree(sadb);
 

drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c

@@ -216,6 +216,16 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base)
 	if (clk_mgr_base->bw_params->dc_mode_limit.dispclk_mhz > 1950)
 		clk_mgr_base->bw_params->dc_mode_limit.dispclk_mhz = 1950;
 
+	/* DPPCLK */
+	dcn32_init_single_clock(clk_mgr, PPCLK_DPPCLK,
+			&clk_mgr_base->bw_params->clk_table.entries[0].dppclk_mhz,
+			&num_entries_per_clk->num_dppclk_levels);
+	num_levels = num_entries_per_clk->num_dppclk_levels;
+	clk_mgr_base->bw_params->dc_mode_limit.dppclk_mhz = dcn30_smu_get_dc_mode_max_dpm_freq(clk_mgr, PPCLK_DPPCLK);
+	//HW recommends limit of 1950 MHz in display clock for all DCN3.2.x
+	if (clk_mgr_base->bw_params->dc_mode_limit.dppclk_mhz > 1950)
+		clk_mgr_base->bw_params->dc_mode_limit.dppclk_mhz = 1950;
+
 	if (num_entries_per_clk->num_dcfclk_levels &&
 			num_entries_per_clk->num_dtbclk_levels &&
 			num_entries_per_clk->num_dispclk_levels)
@@ -240,6 +250,10 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base)
 					= khz_to_mhz_ceil(clk_mgr_base->ctx->dc->debug.min_dpp_clk_khz);
 	}
 
+	for (i = 0; i < num_levels; i++)
+		if (clk_mgr_base->bw_params->clk_table.entries[i].dppclk_mhz > 1950)
+			clk_mgr_base->bw_params->clk_table.entries[i].dppclk_mhz = 1950;
+
 	/* Get UCLK, update bounding box */
 	clk_mgr_base->funcs->get_memclk_states_from_smu(clk_mgr_base);
 

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

@@ -1302,6 +1302,54 @@ static void disable_vbios_mode_if_required(
 	}
 }
 
+/**
+ * wait_for_blank_complete - wait for all active OPPs to finish pending blank
+ * pattern updates
+ *
+ * @dc: [in] dc reference
+ * @context: [in] hardware context in use
+ */
+static void wait_for_blank_complete(struct dc *dc,
+		struct dc_state *context)
+{
+	struct pipe_ctx *opp_head;
+	struct dce_hwseq *hws = dc->hwseq;
+	int i;
+
+	if (!hws->funcs.wait_for_blank_complete)
+		return;
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		opp_head = &context->res_ctx.pipe_ctx[i];
+
+		if (!resource_is_pipe_type(opp_head, OPP_HEAD) ||
+				dc_state_get_pipe_subvp_type(context, opp_head) == SUBVP_PHANTOM)
+			continue;
+
+		hws->funcs.wait_for_blank_complete(opp_head->stream_res.opp);
+	}
+}
+
+static void wait_for_odm_update_pending_complete(struct dc *dc, struct dc_state *context)
+{
+	struct pipe_ctx *otg_master;
+	struct timing_generator *tg;
+	int i;
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		otg_master = &context->res_ctx.pipe_ctx[i];
+		if (!resource_is_pipe_type(otg_master, OTG_MASTER) ||
+				dc_state_get_pipe_subvp_type(context, otg_master) == SUBVP_PHANTOM)
+			continue;
+		tg = otg_master->stream_res.tg;
+		if (tg->funcs->wait_odm_doublebuffer_pending_clear)
+			tg->funcs->wait_odm_doublebuffer_pending_clear(tg);
+	}
+
+	/* ODM update may require to reprogram blank pattern for each OPP */
+	wait_for_blank_complete(dc, context);
+}
+
 static void wait_for_no_pipes_pending(struct dc *dc, struct dc_state *context)
 {
 	int i;
@@ -1993,6 +2041,11 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 		context->stream_count == 0) {
 		/* Must wait for no flips to be pending before doing optimize bw */
 		wait_for_no_pipes_pending(dc, context);
+		/*
+		 * optimized dispclk depends on ODM setup. Need to wait for ODM
+		 * update pending complete before optimizing bandwidth.
+		 */
+		wait_for_odm_update_pending_complete(dc, context);
 		/* pplib is notified if disp_num changed */
 		dc->hwss.optimize_bandwidth(dc, context);
 		/* Need to do otg sync again as otg could be out of sync due to otg
@@ -3270,6 +3323,9 @@ static bool dc_dmub_should_send_dirty_rect_cmd(struct dc *dc, struct dc_stream_s
 	if (stream->link->replay_settings.config.replay_supported)
 		return true;
 
+	if (stream->ctx->dce_version >= DCN_VERSION_3_5 && stream->abm_level)
+		return true;
+
 	return false;
 }
 
@@ -3493,7 +3549,7 @@ static void commit_planes_for_stream_fast(struct dc *dc,
 		top_pipe_to_program->stream->update_flags.raw = 0;
 }
 
-static void wait_for_outstanding_hw_updates(struct dc *dc, const struct dc_state *dc_context)
+static void wait_for_outstanding_hw_updates(struct dc *dc, struct dc_state *dc_context)
 {
 /*
  * This function calls HWSS to wait for any potentially double buffered
@@ -3531,6 +3587,7 @@ static void wait_for_outstanding_hw_updates(struct dc *dc, const struct dc_state
 			}
 		}
 	}
+	wait_for_odm_update_pending_complete(dc, dc_context);
 }
 
 static void commit_planes_for_stream(struct dc *dc,
@@ -4844,22 +4901,16 @@ void dc_exit_ips_for_hw_access(struct dc *dc)
 
 bool dc_dmub_is_ips_idle_state(struct dc *dc)
 {
-	uint32_t idle_state = 0;
-
 	if (dc->debug.disable_idle_power_optimizations)
 		return false;
 
 	if (!dc->caps.ips_support || (dc->config.disable_ips == DMUB_IPS_DISABLE_ALL))
 		return false;
 
-	if (dc->hwss.get_idle_state)
+	if (!dc->ctx->dmub_srv)
-		idle_state = dc->hwss.get_idle_state(dc);
+		return false;
 
-	if (!(idle_state & DMUB_IPS1_ALLOW_MASK) ||
+	return dc->ctx->dmub_srv->idle_allowed;
-		!(idle_state & DMUB_IPS2_ALLOW_MASK))
-		return true;
-
-	return false;
 }
 
 /* set min and max memory clock to lowest and highest DPM level, respectively */

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

@@ -334,7 +334,8 @@ static void dc_state_free(struct kref *kref)
 
 void dc_state_release(struct dc_state *state)
 {
-	kref_put(&state->refcount, dc_state_free);
+	if (state != NULL)
+		kref_put(&state->refcount, dc_state_free);
 }
 /*
  * dc_state_add_stream() - Add a new dc_stream_state to a dc_state.

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

@@ -1085,9 +1085,9 @@ struct replay_settings {
 	/* SMU optimization is enabled */
 	bool replay_smu_opt_enable;
 	/* Current Coasting vtotal */
-	uint16_t coasting_vtotal;
+	uint32_t coasting_vtotal;
 	/* Coasting vtotal table */
-	uint16_t coasting_vtotal_table[PR_COASTING_TYPE_NUM];
+	uint32_t coasting_vtotal_table[PR_COASTING_TYPE_NUM];
 	/* Maximum link off frame count */
 	enum replay_link_off_frame_count_level link_off_frame_count_level;
 	/* Replay pseudo vtotal for abm + ips on full screen video which can improve ips residency */

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c

@@ -384,6 +384,7 @@ static const struct opp_funcs dcn10_opp_funcs = {
 		.opp_set_disp_pattern_generator = NULL,
 		.opp_program_dpg_dimensions = NULL,
 		.dpg_is_blanked = NULL,
+		.dpg_is_pending = NULL,
 		.opp_destroy = opp1_destroy
 };
 

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_opp.c

@@ -337,6 +337,19 @@ bool opp2_dpg_is_blanked(struct output_pixel_processor *opp)
 		(double_buffer_pending == 0);
 }
 
+bool opp2_dpg_is_pending(struct output_pixel_processor *opp)
+{
+	struct dcn20_opp *oppn20 = TO_DCN20_OPP(opp);
+	uint32_t double_buffer_pending;
+	uint32_t dpg_en;
+
+	REG_GET(DPG_CONTROL, DPG_EN, &dpg_en);
+
+	REG_GET(DPG_STATUS, DPG_DOUBLE_BUFFER_PENDING, &double_buffer_pending);
+
+	return (dpg_en == 1 && double_buffer_pending == 1);
+}
+
 void opp2_program_left_edge_extra_pixel (
 		struct output_pixel_processor *opp,
 		bool count)
@@ -363,6 +376,7 @@ static struct opp_funcs dcn20_opp_funcs = {
 		.opp_set_disp_pattern_generator = opp2_set_disp_pattern_generator,
 		.opp_program_dpg_dimensions = opp2_program_dpg_dimensions,
 		.dpg_is_blanked = opp2_dpg_is_blanked,
+		.dpg_is_pending = opp2_dpg_is_pending,
 		.opp_dpg_set_blank_color = opp2_dpg_set_blank_color,
 		.opp_destroy = opp1_destroy,
 		.opp_program_left_edge_extra_pixel = opp2_program_left_edge_extra_pixel,

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_opp.h

@@ -159,6 +159,8 @@ void opp2_program_dpg_dimensions(
 
 bool opp2_dpg_is_blanked(struct output_pixel_processor *opp);
 
+bool opp2_dpg_is_pending(struct output_pixel_processor *opp);
+
 void opp2_dpg_set_blank_color(
 		struct output_pixel_processor *opp,
 		const struct tg_color *color);

drivers/gpu/drm/amd/display/dc/dcn201/dcn201_opp.c

@@ -50,6 +50,7 @@ static struct opp_funcs dcn201_opp_funcs = {
 		.opp_set_disp_pattern_generator = opp2_set_disp_pattern_generator,
 		.opp_program_dpg_dimensions = opp2_program_dpg_dimensions,
 		.dpg_is_blanked = opp2_dpg_is_blanked,
+		.dpg_is_pending = opp2_dpg_is_pending,
 		.opp_dpg_set_blank_color = opp2_dpg_set_blank_color,
 		.opp_destroy = opp1_destroy,
 		.opp_program_left_edge_extra_pixel = opp2_program_left_edge_extra_pixel,

drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c

@@ -782,3 +782,9 @@ void dcn32_update_dml_pipes_odm_policy_based_on_context(struct dc *dc, struct dc
 		pipe_cnt++;
 	}
 }
+
+void dcn32_override_min_req_dcfclk(struct dc *dc, struct dc_state *context)
+{
+	if (dcn32_subvp_in_use(dc, context) && context->bw_ctx.bw.dcn.clk.dcfclk_khz <= MIN_SUBVP_DCFCLK_KHZ)
+		context->bw_ctx.bw.dcn.clk.dcfclk_khz = MIN_SUBVP_DCFCLK_KHZ;
+}

drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c

@@ -623,6 +623,7 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc,
 		 * - Not TMZ surface
 		 */
 		if (pipe->plane_state && !pipe->top_pipe && !dcn32_is_center_timing(pipe) &&
+				!(pipe->stream->timing.pix_clk_100hz / 10000 > DCN3_2_MAX_SUBVP_PIXEL_RATE_MHZ) &&
 				(!dcn32_is_psr_capable(pipe) || (context->stream_count == 1 && dc->caps.dmub_caps.subvp_psr)) &&
 				dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_NONE &&
 				(refresh_rate < 120 || dcn32_allow_subvp_high_refresh_rate(dc, context, pipe)) &&

drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c

@@ -824,13 +824,25 @@ static struct scaler_data get_scaler_data_for_plane(const struct dc_plane_state
 
 static void populate_dummy_dml_plane_cfg(struct dml_plane_cfg_st *out, unsigned int location, const struct dc_stream_state *in)
 {
+	dml_uint_t width, height;
+
+	if (in->timing.h_addressable > 3840)
+		width = 3840;
+	else
+		width = in->timing.h_addressable;	// 4K max
+
+	if (in->timing.v_addressable > 2160)
+		height = 2160;
+	else
+		height = in->timing.v_addressable;	// 4K max
+
 	out->CursorBPP[location] = dml_cur_32bit;
 	out->CursorWidth[location] = 256;
 
 	out->GPUVMMinPageSizeKBytes[location] = 256;
 
-	out->ViewportWidth[location] = in->timing.h_addressable;
+	out->ViewportWidth[location] = width;
-	out->ViewportHeight[location] = in->timing.v_addressable;
+	out->ViewportHeight[location] = height;
 	out->ViewportStationary[location] = false;
 	out->ViewportWidthChroma[location] = 0;
 	out->ViewportHeightChroma[location] = 0;
@@ -849,7 +861,7 @@ static void populate_dummy_dml_plane_cfg(struct dml_plane_cfg_st *out, unsigned
 	out->HTapsChroma[location] = 0;
 	out->VTapsChroma[location] = 0;
 	out->SourceScan[location] = dml_rotation_0;
-	out->ScalerRecoutWidth[location] = in->timing.h_addressable;
+	out->ScalerRecoutWidth[location] = width;
 
 	out->LBBitPerPixel[location] = 57;
 

drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c

@@ -703,13 +703,8 @@ static inline struct dml2_context *dml2_allocate_memory(void)
 	return (struct dml2_context *) kzalloc(sizeof(struct dml2_context), GFP_KERNEL);
 }
 
-bool dml2_create(const struct dc *in_dc, const struct dml2_configuration_options *config, struct dml2_context **dml2)
+static void dml2_init(const struct dc *in_dc, const struct dml2_configuration_options *config, struct dml2_context **dml2)
 {
-	// Allocate Mode Lib Ctx
-	*dml2 = dml2_allocate_memory();
-
-	if (!(*dml2))
-		return false;
 
 	// Store config options
 	(*dml2)->config = *config;
@@ -737,9 +732,18 @@ bool dml2_create(const struct dc *in_dc, const struct dml2_configuration_options
 	initialize_dml2_soc_bbox(*dml2, in_dc, &(*dml2)->v20.dml_core_ctx.soc);
 
 	initialize_dml2_soc_states(*dml2, in_dc, &(*dml2)->v20.dml_core_ctx.soc, &(*dml2)->v20.dml_core_ctx.states);
+}
+
+bool dml2_create(const struct dc *in_dc, const struct dml2_configuration_options *config, struct dml2_context **dml2)
+{
+	// Allocate Mode Lib Ctx
+	*dml2 = dml2_allocate_memory();
+
+	if (!(*dml2))
+		return false;
+
+	dml2_init(in_dc, config, dml2);
 
-	/*Initialize DML20 instance which calls dml2_core_create, and core_dcn3_populate_informative*/
-	//dml2_initialize_instance(&(*dml_ctx)->v20.dml_init);
 	return true;
 }
 
@@ -779,3 +783,11 @@ bool dml2_create_copy(struct dml2_context **dst_dml2,
 
 	return true;
 }
+
+void dml2_reinit(const struct dc *in_dc,
+				 const struct dml2_configuration_options *config,
+				 struct dml2_context **dml2)
+{
+
+	dml2_init(in_dc, config, dml2);
+}

drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h

@@ -214,6 +214,9 @@ void dml2_copy(struct dml2_context *dst_dml2,
 	struct dml2_context *src_dml2);
 bool dml2_create_copy(struct dml2_context **dst_dml2,
 	struct dml2_context *src_dml2);
+void dml2_reinit(const struct dc *in_dc,
+				 const struct dml2_configuration_options *config,
+				 struct dml2_context **dml2);
 
 /*
  * dml2_validate - Determines if a display configuration is supported or not.

drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c

@@ -1498,6 +1498,11 @@ static void dcn20_detect_pipe_changes(struct dc_state *old_state,
 		return;
 	}
 
+	if (resource_is_pipe_type(new_pipe, OTG_MASTER) &&
+			resource_is_odm_topology_changed(new_pipe, old_pipe))
+		/* Detect odm changes */
+		new_pipe->update_flags.bits.odm = 1;
+
 	/* Exit on unchanged, unused pipe */
 	if (!old_pipe->plane_state && !new_pipe->plane_state)
 		return;
@@ -1551,10 +1556,6 @@ static void dcn20_detect_pipe_changes(struct dc_state *old_state,
 
 	/* Detect top pipe only changes */
 	if (resource_is_pipe_type(new_pipe, OTG_MASTER)) {
-		/* Detect odm changes */
-		if (resource_is_odm_topology_changed(new_pipe, old_pipe))
-			new_pipe->update_flags.bits.odm = 1;
-
 		/* Detect global sync changes */
 		if (old_pipe->pipe_dlg_param.vready_offset != new_pipe->pipe_dlg_param.vready_offset
 				|| old_pipe->pipe_dlg_param.vstartup_start != new_pipe->pipe_dlg_param.vstartup_start
@@ -1999,19 +2000,20 @@ void dcn20_program_front_end_for_ctx(
 	DC_LOGGER_INIT(dc->ctx->logger);
 	unsigned int prev_hubp_count = 0;
 	unsigned int hubp_count = 0;
+	struct pipe_ctx *pipe;
 
 	if (resource_is_pipe_topology_changed(dc->current_state, context))
 		resource_log_pipe_topology_update(dc, context);
 
 	if (dc->hwss.program_triplebuffer != NULL && dc->debug.enable_tri_buf) {
 		for (i = 0; i < dc->res_pool->pipe_count; i++) {
-			struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+			pipe = &context->res_ctx.pipe_ctx[i];
 
-			if (!pipe_ctx->top_pipe && !pipe_ctx->prev_odm_pipe && pipe_ctx->plane_state) {
+			if (!pipe->top_pipe && !pipe->prev_odm_pipe && pipe->plane_state) {
-				ASSERT(!pipe_ctx->plane_state->triplebuffer_flips);
+				ASSERT(!pipe->plane_state->triplebuffer_flips);
 				/*turn off triple buffer for full update*/
 				dc->hwss.program_triplebuffer(
-						dc, pipe_ctx, pipe_ctx->plane_state->triplebuffer_flips);
+						dc, pipe, pipe->plane_state->triplebuffer_flips);
 			}
 		}
 	}
@@ -2085,12 +2087,22 @@ void dcn20_program_front_end_for_ctx(
 			DC_LOG_DC("Reset mpcc for pipe %d\n", dc->current_state->res_ctx.pipe_ctx[i].pipe_idx);
 		}
 
+	/* update ODM for blanked OTG master pipes */
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		pipe = &context->res_ctx.pipe_ctx[i];
+		if (resource_is_pipe_type(pipe, OTG_MASTER) &&
+				!resource_is_pipe_type(pipe, DPP_PIPE) &&
+				pipe->update_flags.bits.odm &&
+				hws->funcs.update_odm)
+			hws->funcs.update_odm(dc, context, pipe);
+	}
+
 	/*
 	 * Program all updated pipes, order matters for mpcc setup. Start with
 	 * top pipe and program all pipes that follow in order
 	 */
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+		pipe = &context->res_ctx.pipe_ctx[i];
 
 		if (pipe->plane_state && !pipe->top_pipe) {
 			while (pipe) {
@@ -2129,17 +2141,6 @@ void dcn20_program_front_end_for_ctx(
 			context->stream_status[0].plane_count > 1) {
 			pipe->plane_res.hubp->funcs->hubp_wait_pipe_read_start(pipe->plane_res.hubp);
 		}
-
-		/* when dynamic ODM is active, pipes must be reconfigured when all planes are
-		 * disabled, as some transitions will leave software and hardware state
-		 * mismatched.
-		 */
-		if (dc->debug.enable_single_display_2to1_odm_policy &&
-			pipe->stream &&
-			pipe->update_flags.bits.disable &&
-			!pipe->prev_odm_pipe &&
-			hws->funcs.update_odm)
-			hws->funcs.update_odm(dc, context, pipe);
 	}
 }
 
@@ -2451,7 +2452,7 @@ bool dcn20_wait_for_blank_complete(
 	int counter;
 
 	for (counter = 0; counter < 1000; counter++) {
-		if (opp->funcs->dpg_is_blanked(opp))
+		if (!opp->funcs->dpg_is_pending(opp))
 			break;
 
 		udelay(100);
@@ -2462,7 +2463,7 @@ bool dcn20_wait_for_blank_complete(
 		return false;
 	}
 
-	return true;
+	return opp->funcs->dpg_is_blanked(opp);
 }
 
 bool dcn20_dmdata_status_done(struct pipe_ctx *pipe_ctx)

drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c

@@ -812,10 +812,20 @@ void dcn30_set_avmute(struct pipe_ctx *pipe_ctx, bool enable)
 	if (pipe_ctx == NULL)
 		return;
 
-	if (dc_is_hdmi_signal(pipe_ctx->stream->signal) && pipe_ctx->stream_res.stream_enc != NULL)
+	if (dc_is_hdmi_signal(pipe_ctx->stream->signal) && pipe_ctx->stream_res.stream_enc != NULL) {
 		pipe_ctx->stream_res.stream_enc->funcs->set_avmute(
 				pipe_ctx->stream_res.stream_enc,
 				enable);
+
+		/* Wait for two frame to make sure AV mute is sent out */
+		if (enable) {
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VBLANK);
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VBLANK);
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
+		}
+	}
 }
 
 void dcn30_update_info_frame(struct pipe_ctx *pipe_ctx)

drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c

@@ -1156,6 +1156,13 @@ void dcn32_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx *
 			dsc->funcs->dsc_disconnect(dsc);
 		}
 	}
+
+	if (!resource_is_pipe_type(pipe_ctx, DPP_PIPE))
+		/*
+		 * blank pattern is generated by OPP, reprogram blank pattern
+		 * due to OPP count change
+		 */
+		dc->hwseq->funcs.blank_pixel_data(dc, pipe_ctx, true);
 }
 
 unsigned int dcn32_calculate_dccg_k1_k2_values(struct pipe_ctx *pipe_ctx, unsigned int *k1_div, unsigned int *k2_div)
@@ -1778,3 +1785,26 @@ void dcn32_prepare_bandwidth(struct dc *dc,
 		context->bw_ctx.bw.dcn.clk.p_state_change_support = p_state_change_support;
 	}
 }
+
+void dcn32_interdependent_update_lock(struct dc *dc,
+		struct dc_state *context, bool lock)
+{
+	unsigned int i;
+	struct pipe_ctx *pipe;
+	struct timing_generator *tg;
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		pipe = &context->res_ctx.pipe_ctx[i];
+		tg = pipe->stream_res.tg;
+
+		if (!resource_is_pipe_type(pipe, OTG_MASTER) ||
+				!tg->funcs->is_tg_enabled(tg) ||
+				dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_PHANTOM)
+			continue;
+
+		if (lock)
+			dc->hwss.pipe_control_lock(dc, pipe, true);
+		else
+			dc->hwss.pipe_control_lock(dc, pipe, false);
+	}
+}

drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.h

@@ -129,4 +129,6 @@ bool dcn32_is_pipe_topology_transition_seamless(struct dc *dc,
 void dcn32_prepare_bandwidth(struct dc *dc,
 	struct dc_state *context);
 
+void dcn32_interdependent_update_lock(struct dc *dc,
+		struct dc_state *context, bool lock);
 #endif /* __DC_HWSS_DCN32_H__ */

drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c

@@ -58,7 +58,7 @@ static const struct hw_sequencer_funcs dcn32_funcs = {
 	.disable_plane = dcn20_disable_plane,
 	.disable_pixel_data = dcn20_disable_pixel_data,
 	.pipe_control_lock = dcn20_pipe_control_lock,
-	.interdependent_update_lock = dcn10_lock_all_pipes,
+	.interdependent_update_lock = dcn32_interdependent_update_lock,
 	.cursor_lock = dcn10_cursor_lock,
 	.prepare_bandwidth = dcn32_prepare_bandwidth,
 	.optimize_bandwidth = dcn20_optimize_bandwidth,

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

@@ -337,6 +337,9 @@ struct opp_funcs {
 	bool (*dpg_is_blanked)(
 			struct output_pixel_processor *opp);
 
+	bool (*dpg_is_pending)(struct output_pixel_processor *opp);
+
+
 	void (*opp_dpg_set_blank_color)(
 			struct output_pixel_processor *opp,
 			const struct tg_color *color);

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

@@ -331,6 +331,7 @@ struct timing_generator_funcs {
 
 	void (*init_odm)(struct timing_generator *tg);
 	void (*wait_drr_doublebuffer_pending_clear)(struct timing_generator *tg);
+	void (*wait_odm_doublebuffer_pending_clear)(struct timing_generator *tg);
 };
 
 #endif

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

@@ -285,12 +285,12 @@ struct link_service {
 			enum replay_FW_Message_type msg,
 			union dmub_replay_cmd_set *cmd_data);
 	bool (*edp_set_coasting_vtotal)(
-			struct dc_link *link, uint16_t coasting_vtotal);
+			struct dc_link *link, uint32_t coasting_vtotal);
 	bool (*edp_replay_residency)(const struct dc_link *link,
 			unsigned int *residency, const bool is_start,
 			const bool is_alpm);
 	bool (*edp_set_replay_power_opt_and_coasting_vtotal)(struct dc_link *link,
-			const unsigned int *power_opts, uint16_t coasting_vtotal);
+			const unsigned int *power_opts, uint32_t coasting_vtotal);
 
 	bool (*edp_wait_for_t12)(struct dc_link *link);
 	bool (*edp_is_ilr_optimization_required)(struct dc_link *link,

drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c

@@ -1034,7 +1034,7 @@ bool edp_send_replay_cmd(struct dc_link *link,
 	return true;
 }
 
-bool edp_set_coasting_vtotal(struct dc_link *link, uint16_t coasting_vtotal)
+bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal)
 {
 	struct dc *dc = link->ctx->dc;
 	struct dmub_replay *replay = dc->res_pool->replay;
@@ -1073,7 +1073,7 @@ bool edp_replay_residency(const struct dc_link *link,
 }
 
 bool edp_set_replay_power_opt_and_coasting_vtotal(struct dc_link *link,
-	const unsigned int *power_opts, uint16_t coasting_vtotal)
+	const unsigned int *power_opts, uint32_t coasting_vtotal)
 {
 	struct dc  *dc = link->ctx->dc;
 	struct dmub_replay *replay = dc->res_pool->replay;

drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h

@@ -59,12 +59,12 @@ bool edp_setup_replay(struct dc_link *link,
 bool edp_send_replay_cmd(struct dc_link *link,
 			enum replay_FW_Message_type msg,
 			union dmub_replay_cmd_set *cmd_data);
-bool edp_set_coasting_vtotal(struct dc_link *link, uint16_t coasting_vtotal);
+bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal);
 bool edp_replay_residency(const struct dc_link *link,
 	unsigned int *residency, const bool is_start, const bool is_alpm);
 bool edp_get_replay_state(const struct dc_link *link, uint64_t *state);
 bool edp_set_replay_power_opt_and_coasting_vtotal(struct dc_link *link,
-	const unsigned int *power_opts, uint16_t coasting_vtotal);
+	const unsigned int *power_opts, uint32_t coasting_vtotal);
 bool edp_wait_for_t12(struct dc_link *link);
 bool edp_is_ilr_optimization_required(struct dc_link *link,
        struct dc_crtc_timing *crtc_timing);

drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h

@@ -557,7 +557,8 @@ struct dcn_optc_registers {
 	type OTG_CRC_DATA_STREAM_SPLIT_MODE;\
 	type OTG_CRC_DATA_FORMAT;\
 	type OTG_V_TOTAL_LAST_USED_BY_DRR;\
-	type OTG_DRR_TIMING_DBUF_UPDATE_PENDING;
+	type OTG_DRR_TIMING_DBUF_UPDATE_PENDING;\
+	type OTG_H_TIMING_DIV_MODE_DB_UPDATE_PENDING;
 
 #define TG_REG_FIELD_LIST_DCN3_2(type) \
 	type OTG_H_TIMING_DIV_MODE_MANUAL;

drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c

@@ -122,6 +122,13 @@ void optc32_get_odm_combine_segments(struct timing_generator *tg, int *odm_combi
 	}
 }
 
+void optc32_wait_odm_doublebuffer_pending_clear(struct timing_generator *tg)
+{
+	struct optc *optc1 = DCN10TG_FROM_TG(tg);
+
+	REG_WAIT(OTG_DOUBLE_BUFFER_CONTROL, OTG_H_TIMING_DIV_MODE_DB_UPDATE_PENDING, 0, 2, 50000);
+}
+
 void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -345,6 +352,7 @@ static struct timing_generator_funcs dcn32_tg_funcs = {
 		.set_odm_bypass = optc32_set_odm_bypass,
 		.set_odm_combine = optc32_set_odm_combine,
 		.get_odm_combine_segments = optc32_get_odm_combine_segments,
+		.wait_odm_doublebuffer_pending_clear = optc32_wait_odm_doublebuffer_pending_clear,
 		.set_h_timing_div_manual_mode = optc32_set_h_timing_div_manual_mode,
 		.get_optc_source = optc2_get_optc_source,
 		.set_out_mux = optc3_set_out_mux,

drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h

@@ -183,5 +183,6 @@ void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool man
 void optc32_get_odm_combine_segments(struct timing_generator *tg, int *odm_combine_segments);
 void optc32_set_odm_bypass(struct timing_generator *optc,
 		const struct dc_crtc_timing *dc_crtc_timing);
+void optc32_wait_odm_doublebuffer_pending_clear(struct timing_generator *tg);
 
 #endif /* __DC_OPTC_DCN32_H__ */

drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c

@@ -1771,6 +1771,7 @@ static bool dml1_validate(struct dc *dc, struct dc_state *context, bool fast_val
 	dc->res_pool->funcs->calculate_wm_and_dlg(dc, context, pipes, pipe_cnt, vlevel);
 
 	dcn32_override_min_req_memclk(dc, context);
+	dcn32_override_min_req_dcfclk(dc, context);
 
 	BW_VAL_TRACE_END_WATERMARKS();
 
@@ -1930,6 +1931,8 @@ static void dcn32_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw
 {
 	DC_FP_START();
 	dcn32_update_bw_bounding_box_fpu(dc, bw_params);
+	if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2)
+		dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2);
 	DC_FP_END();
 }
 

drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h

@@ -42,6 +42,7 @@
 #define SUBVP_ACTIVE_MARGIN_LIST_LEN 2
 #define DCN3_2_MAX_SUBVP_PIXEL_RATE_MHZ 1800
 #define DCN3_2_VMIN_DISPCLK_HZ 717000000
+#define MIN_SUBVP_DCFCLK_KHZ 400000
 
 #define TO_DCN32_RES_POOL(pool)\
 	container_of(pool, struct dcn32_resource_pool, base)
@@ -181,6 +182,8 @@ bool dcn32_subvp_vblank_admissable(struct dc *dc, struct dc_state *context, int
 
 void dcn32_update_dml_pipes_odm_policy_based_on_context(struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes);
 
+void dcn32_override_min_req_dcfclk(struct dc *dc, struct dc_state *context);
+
 /* definitions for run time init of reg offsets */
 
 /* CLK SRC */

drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c

@@ -1581,6 +1581,8 @@ static void dcn321_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *b
 {
 	DC_FP_START();
 	dcn321_update_bw_bounding_box_fpu(dc, bw_params);
+	if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2)
+		dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2);
 	DC_FP_END();
 }
 

drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h

@@ -3238,6 +3238,14 @@ struct dmub_cmd_replay_set_coasting_vtotal_data {
 	 * Currently the support is only for 0 or 1
 	 */
 	uint8_t panel_inst;
+	/**
+	 * 16-bit value dicated by driver that indicates the coasting vtotal high byte part.
+	 */
+	uint16_t coasting_vtotal_high;
+	/**
+	 * Explicit padding to 4 byte boundary.
+	 */
+	uint8_t pad[2];
 };
 
 /**

drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c

@@ -513,6 +513,9 @@ enum mod_hdcp_status mod_hdcp_hdcp2_create_session(struct mod_hdcp *hdcp)
 	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.context.mem_context.shared_buf;
 	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
 
+	if (!display)
+		return MOD_HDCP_STATUS_DISPLAY_NOT_FOUND;
+
 	hdcp_cmd->in_msg.hdcp2_create_session_v2.display_handle = display->index;
 
 	if (hdcp->connection.link.adjust.hdcp2.force_type == MOD_HDCP_FORCE_TYPE_0)

drivers/gpu/drm/amd/display/modules/power/power_helpers.c

@@ -975,7 +975,7 @@ bool psr_su_set_dsc_slice_height(struct dc *dc, struct dc_link *link,
 
 void set_replay_coasting_vtotal(struct dc_link *link,
 	enum replay_coasting_vtotal_type type,
-	uint16_t vtotal)
+	uint32_t vtotal)
 {
 	link->replay_settings.coasting_vtotal_table[type] = vtotal;
 }

drivers/gpu/drm/amd/display/modules/power/power_helpers.h

@@ -56,7 +56,7 @@ bool dmub_init_abm_config(struct resource_pool *res_pool,
 void init_replay_config(struct dc_link *link, struct replay_config *pr_config);
 void set_replay_coasting_vtotal(struct dc_link *link,
 	enum replay_coasting_vtotal_type type,
-	uint16_t vtotal);
+	uint32_t vtotal);
 void set_replay_ips_full_screen_video_src_vtotal(struct dc_link *link, uint16_t vtotal);
 void calculate_replay_link_off_frame_count(struct dc_link *link,
 	uint16_t vtotal, uint16_t htotal);

drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c

@@ -1283,10 +1283,8 @@ static int arcturus_get_power_limit(struct smu_context *smu,
 					uint32_t *max_power_limit,
 					uint32_t *min_power_limit)
 {
-	struct smu_11_0_powerplay_table *powerplay_table =
-		(struct smu_11_0_powerplay_table *)smu->smu_table.power_play_table;
 	PPTable_t *pptable = smu->smu_table.driver_pptable;
-	uint32_t power_limit, od_percent_upper, od_percent_lower;
+	uint32_t power_limit;
 
 	if (smu_v11_0_get_current_power_limit(smu, &power_limit)) {
 		/* the last hope to figure out the ppt limit */
@@ -1302,26 +1300,10 @@ static int arcturus_get_power_limit(struct smu_context *smu,
 		*current_power_limit = power_limit;
 	if (default_power_limit)
 		*default_power_limit = power_limit;
-
+	if (max_power_limit)
-	if (smu->od_enabled)
+		*max_power_limit = power_limit;
-		od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
+	if (min_power_limit)
-	else
+		*min_power_limit = power_limit;
-		od_percent_upper = 0;
-
-	od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
-
-	dev_dbg(smu->adev->dev, "od percent upper:%d, od percent lower:%d (default power: %d)\n",
-							od_percent_upper, od_percent_lower, power_limit);
-
-	if (max_power_limit) {
-		*max_power_limit = power_limit * (100 + od_percent_upper);
-		*max_power_limit /= 100;
-	}
-
-	if (min_power_limit) {
-		*min_power_limit = power_limit * (100 - od_percent_lower);
-		*min_power_limit /= 100;
-	}
 
 	return 0;
 }

drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c

@@ -2339,7 +2339,7 @@ static int navi10_get_power_limit(struct smu_context *smu,
 		(struct smu_11_0_powerplay_table *)smu->smu_table.power_play_table;
 	struct smu_11_0_overdrive_table *od_settings = smu->od_settings;
 	PPTable_t *pptable = smu->smu_table.driver_pptable;
-	uint32_t power_limit, od_percent_upper, od_percent_lower;
+	uint32_t power_limit, od_percent_upper = 0, od_percent_lower = 0;
 
 	if (smu_v11_0_get_current_power_limit(smu, &power_limit)) {
 		/* the last hope to figure out the ppt limit */
@@ -2356,13 +2356,16 @@ static int navi10_get_power_limit(struct smu_context *smu,
 	if (default_power_limit)
 		*default_power_limit = power_limit;
 
-	if (smu->od_enabled &&
+	if (powerplay_table) {
-		    navi10_od_feature_is_supported(od_settings, SMU_11_0_ODCAP_POWER_LIMIT))
+		if (smu->od_enabled &&
-		od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
+			    navi10_od_feature_is_supported(od_settings, SMU_11_0_ODCAP_POWER_LIMIT)) {
-	else
+			od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
-		od_percent_upper = 0;
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
-
+		} else if (navi10_od_feature_is_supported(od_settings, SMU_11_0_ODCAP_POWER_LIMIT)) {
-	od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
+			od_percent_upper = 0;
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
+		}
+	}
 
 	dev_dbg(smu->adev->dev, "od percent upper:%d, od percent lower:%d (default power: %d)\n",
 					od_percent_upper, od_percent_lower, power_limit);

drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c

@@ -617,6 +617,12 @@ static uint32_t sienna_cichlid_get_throttler_status_locked(struct smu_context *s
 	return throttler_status;
 }
 
+static bool sienna_cichlid_is_od_feature_supported(struct smu_11_0_7_overdrive_table *od_table,
+						   enum SMU_11_0_7_ODFEATURE_CAP cap)
+{
+	return od_table->cap[cap];
+}
+
 static int sienna_cichlid_get_power_limit(struct smu_context *smu,
 					  uint32_t *current_power_limit,
 					  uint32_t *default_power_limit,
@@ -625,7 +631,8 @@ static int sienna_cichlid_get_power_limit(struct smu_context *smu,
 {
 	struct smu_11_0_7_powerplay_table *powerplay_table =
 		(struct smu_11_0_7_powerplay_table *)smu->smu_table.power_play_table;
-	uint32_t power_limit, od_percent_upper, od_percent_lower;
+	struct smu_11_0_7_overdrive_table *od_settings = smu->od_settings;
+	uint32_t power_limit, od_percent_upper = 0, od_percent_lower = 0;
 	uint16_t *table_member;
 
 	GET_PPTABLE_MEMBER(SocketPowerLimitAc, &table_member);
@@ -640,12 +647,16 @@ static int sienna_cichlid_get_power_limit(struct smu_context *smu,
 	if (default_power_limit)
 		*default_power_limit = power_limit;
 
-	if (smu->od_enabled)
+	if (powerplay_table) {
-		od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_7_ODSETTING_POWERPERCENTAGE]);
+		if (smu->od_enabled &&
-	else
+				sienna_cichlid_is_od_feature_supported(od_settings, SMU_11_0_7_ODCAP_POWER_LIMIT)) {
-		od_percent_upper = 0;
+			od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_7_ODSETTING_POWERPERCENTAGE]);
-
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_7_ODSETTING_POWERPERCENTAGE]);
-	od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_7_ODSETTING_POWERPERCENTAGE]);
+		} else if ((sienna_cichlid_is_od_feature_supported(od_settings, SMU_11_0_7_ODCAP_POWER_LIMIT))) {
+			od_percent_upper = 0;
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_11_0_7_ODSETTING_POWERPERCENTAGE]);
+		}
+	}
 
 	dev_dbg(smu->adev->dev, "od percent upper:%d, od percent lower:%d (default power: %d)\n",
 					od_percent_upper, od_percent_lower, power_limit);
@@ -1250,12 +1261,6 @@ static bool sienna_cichlid_is_support_fine_grained_dpm(struct smu_context *smu,
 	return dpm_desc->SnapToDiscrete == 0;
 }
 
-static bool sienna_cichlid_is_od_feature_supported(struct smu_11_0_7_overdrive_table *od_table,
-						   enum SMU_11_0_7_ODFEATURE_CAP cap)
-{
-	return od_table->cap[cap];
-}
-
 static void sienna_cichlid_get_od_setting_range(struct smu_11_0_7_overdrive_table *od_table,
 						enum SMU_11_0_7_ODSETTING_ID setting,
 						uint32_t *min, uint32_t *max)

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c

@@ -2356,7 +2356,7 @@ static int smu_v13_0_0_get_power_limit(struct smu_context *smu,
 		(struct smu_13_0_0_powerplay_table *)table_context->power_play_table;
 	PPTable_t *pptable = table_context->driver_pptable;
 	SkuTable_t *skutable = &pptable->SkuTable;
-	uint32_t power_limit, od_percent_upper, od_percent_lower;
+	uint32_t power_limit, od_percent_upper = 0, od_percent_lower = 0;
 	uint32_t msg_limit = skutable->MsgLimits.Power[PPT_THROTTLER_PPT0][POWER_SOURCE_AC];
 
 	if (smu_v13_0_get_current_power_limit(smu, &power_limit))
@@ -2369,12 +2369,16 @@ static int smu_v13_0_0_get_power_limit(struct smu_context *smu,
 	if (default_power_limit)
 		*default_power_limit = power_limit;
 
-	if (smu->od_enabled)
+	if (powerplay_table) {
-		od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_13_0_0_ODSETTING_POWERPERCENTAGE]);
+		if (smu->od_enabled &&
-	else
+				smu_v13_0_0_is_od_feature_supported(smu, PP_OD_FEATURE_PPT_BIT)) {
-		od_percent_upper = 0;
+			od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_13_0_0_ODSETTING_POWERPERCENTAGE]);
-
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_13_0_0_ODSETTING_POWERPERCENTAGE]);
-	od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_13_0_0_ODSETTING_POWERPERCENTAGE]);
+		} else if (smu_v13_0_0_is_od_feature_supported(smu, PP_OD_FEATURE_PPT_BIT)) {
+			od_percent_upper = 0;
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_13_0_0_ODSETTING_POWERPERCENTAGE]);
+		}
+	}
 
 	dev_dbg(smu->adev->dev, "od percent upper:%d, od percent lower:%d (default power: %d)\n",
 					od_percent_upper, od_percent_lower, power_limit);

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c

@@ -2320,7 +2320,7 @@ static int smu_v13_0_7_get_power_limit(struct smu_context *smu,
 		(struct smu_13_0_7_powerplay_table *)table_context->power_play_table;
 	PPTable_t *pptable = table_context->driver_pptable;
 	SkuTable_t *skutable = &pptable->SkuTable;
-	uint32_t power_limit, od_percent_upper, od_percent_lower;
+	uint32_t power_limit, od_percent_upper = 0, od_percent_lower = 0;
 	uint32_t msg_limit = skutable->MsgLimits.Power[PPT_THROTTLER_PPT0][POWER_SOURCE_AC];
 
 	if (smu_v13_0_get_current_power_limit(smu, &power_limit))
@@ -2333,12 +2333,16 @@ static int smu_v13_0_7_get_power_limit(struct smu_context *smu,
 	if (default_power_limit)
 		*default_power_limit = power_limit;
 
-	if (smu->od_enabled)
+	if (powerplay_table) {
-		od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_13_0_7_ODSETTING_POWERPERCENTAGE]);
+		if (smu->od_enabled &&
-	else
+				(smu_v13_0_7_is_od_feature_supported(smu, PP_OD_FEATURE_PPT_BIT))) {
-		od_percent_upper = 0;
+			od_percent_upper = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_13_0_7_ODSETTING_POWERPERCENTAGE]);
-
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_13_0_7_ODSETTING_POWERPERCENTAGE]);
-	od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_13_0_7_ODSETTING_POWERPERCENTAGE]);
+		} else if (smu_v13_0_7_is_od_feature_supported(smu, PP_OD_FEATURE_PPT_BIT)) {
+			od_percent_upper = 0;
+			od_percent_lower = le32_to_cpu(powerplay_table->overdrive_table.min[SMU_13_0_7_ODSETTING_POWERPERCENTAGE]);
+		}
+	}
 
 	dev_dbg(smu->adev->dev, "od percent upper:%d, od percent lower:%d (default power: %d)\n",
 					od_percent_upper, od_percent_lower, power_limit);