Merge tag 'amd-drm-fixes-6.9-2024-03-21' of https://gitlab.freedeskto…

…p.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 <[email protected]>

From: Alex Deucher <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

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) {
-		r = -ENOMEM;
-		goto unmap_memory;
-	}
+	if (!adev->reset_domain)
+		return -ENOMEM;
 
 	/* detect hw virtualization here */
 	amdgpu_detect_virtualization(adev);
@@ -4053,20 +4051,20 @@ 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);
 
 	/* 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 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 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,
-					    amdgpu_bo_size(bo),
-					    &amdgpu_hmm_gfx_ops);
+	else
+		r = mmu_interval_notifier_insert(&bo->notifier, current->mm, addr,
+							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));
-	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", ucode_prefix);
-	r = amdgpu_ucode_request(adev, &adev->vcn.fw, fw_name);
-	if (r)
-		amdgpu_ucode_release(&adev->vcn.fw);
+	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
+		amdgpu_ucode_ip_version_decode(adev, UVD_HWIP, ucode_prefix, sizeof(ucode_prefix));
+		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", ucode_prefix);
+		if (amdgpu_ip_version(adev, UVD_HWIP, 0) ==  IP_VERSION(4, 0, 6) &&
+			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);
 				}
@@ -1043,19 +1055,19 @@ 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");
 				break;
 			}
 			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);
 }