Merge tag 'amd-drm-next-5.14-2021-06-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-5.14-2021-06-16:

amdgpu:
- Aldebaran fixes
- Expose asic independent throttler status
- BACO fixes for navi1x
- Smartshift fixes
- Misc code cleanups
- RAS fixes for Sienna Cichlid
- Gamma verificaton fixes
- DC LTTPR fixes
- DP AUX timeout handling fixes
- GFX9, 10 powergating fixes

amdkfd:
- TLB flush fixes when using SDMA
- Locking fixes
- SVM fixes

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210617031719.4013-1-alexander.deucher@amd.com

Author: airlied

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

@@ -47,12 +47,6 @@ static struct {
 	spinlock_t mem_limit_lock;
 } kfd_mem_limit;
 
-/* Struct used for amdgpu_amdkfd_bo_validate */
-struct amdgpu_vm_parser {
-	uint32_t        domain;
-	bool            wait;
-};
-
 static const char * const domain_bit_to_string[] = {
 		"CPU",
 		"GTT",
@@ -348,11 +342,9 @@ static int amdgpu_amdkfd_bo_validate(struct amdgpu_bo *bo, uint32_t domain,
 	return ret;
 }
 
-static int amdgpu_amdkfd_validate(void *param, struct amdgpu_bo *bo)
+static int amdgpu_amdkfd_validate_vm_bo(void *_unused, struct amdgpu_bo *bo)
 {
-	struct amdgpu_vm_parser *p = param;
-
-	return amdgpu_amdkfd_bo_validate(bo, p->domain, p->wait);
+	return amdgpu_amdkfd_bo_validate(bo, bo->allowed_domains, false);
 }
 
 /* vm_validate_pt_pd_bos - Validate page table and directory BOs
@@ -364,28 +356,23 @@ static int amdgpu_amdkfd_validate(void *param, struct amdgpu_bo *bo)
  */
 static int vm_validate_pt_pd_bos(struct amdgpu_vm *vm)
 {
-	struct amdgpu_bo *pd = vm->root.base.bo;
+	struct amdgpu_bo *pd = vm->root.bo;
 	struct amdgpu_device *adev = amdgpu_ttm_adev(pd->tbo.bdev);
-	struct amdgpu_vm_parser param;
 	int ret;
 
-	param.domain = AMDGPU_GEM_DOMAIN_VRAM;
-	param.wait = false;
-
-	ret = amdgpu_vm_validate_pt_bos(adev, vm, amdgpu_amdkfd_validate,
-					&param);
+	ret = amdgpu_vm_validate_pt_bos(adev, vm, amdgpu_amdkfd_validate_vm_bo, NULL);
 	if (ret) {
 		pr_err("failed to validate PT BOs\n");
 		return ret;
 	}
 
-	ret = amdgpu_amdkfd_validate(&param, pd);
+	ret = amdgpu_amdkfd_validate_vm_bo(NULL, pd);
 	if (ret) {
 		pr_err("failed to validate PD\n");
 		return ret;
 	}
 
-	vm->pd_phys_addr = amdgpu_gmc_pd_addr(vm->root.base.bo);
+	vm->pd_phys_addr = amdgpu_gmc_pd_addr(vm->root.bo);
 
 	if (vm->use_cpu_for_update) {
 		ret = amdgpu_bo_kmap(pd, NULL);
@@ -400,7 +387,7 @@ static int vm_validate_pt_pd_bos(struct amdgpu_vm *vm)
 
 static int vm_update_pds(struct amdgpu_vm *vm, struct amdgpu_sync *sync)
 {
-	struct amdgpu_bo *pd = vm->root.base.bo;
+	struct amdgpu_bo *pd = vm->root.bo;
 	struct amdgpu_device *adev = amdgpu_ttm_adev(pd->tbo.bdev);
 	int ret;
 
@@ -652,7 +639,7 @@ kfd_mem_attach_dmabuf(struct amdgpu_device *adev, struct kgd_mem *mem,
 		}
 	}
 
-	gobj = amdgpu_gem_prime_import(&adev->ddev, mem->dmabuf);
+	gobj = amdgpu_gem_prime_import(adev_to_drm(adev), mem->dmabuf);
 	if (IS_ERR(gobj))
 		return PTR_ERR(gobj);
 
@@ -1166,7 +1153,7 @@ static int process_sync_pds_resv(struct amdkfd_process_info *process_info,
 
 	list_for_each_entry(peer_vm, &process_info->vm_list_head,
 			    vm_list_node) {
-		struct amdgpu_bo *pd = peer_vm->root.base.bo;
+		struct amdgpu_bo *pd = peer_vm->root.bo;
 
 		ret = amdgpu_sync_resv(NULL, sync, pd->tbo.base.resv,
 				       AMDGPU_SYNC_NE_OWNER,
@@ -1233,24 +1220,24 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 	vm->process_info = *process_info;
 
 	/* Validate page directory and attach eviction fence */
-	ret = amdgpu_bo_reserve(vm->root.base.bo, true);
+	ret = amdgpu_bo_reserve(vm->root.bo, true);
 	if (ret)
 		goto reserve_pd_fail;
 	ret = vm_validate_pt_pd_bos(vm);
 	if (ret) {
 		pr_err("validate_pt_pd_bos() failed\n");
 		goto validate_pd_fail;
 	}
-	ret = amdgpu_bo_sync_wait(vm->root.base.bo,
+	ret = amdgpu_bo_sync_wait(vm->root.bo,
 				  AMDGPU_FENCE_OWNER_KFD, false);
 	if (ret)
 		goto wait_pd_fail;
-	ret = dma_resv_reserve_shared(vm->root.base.bo->tbo.base.resv, 1);
+	ret = dma_resv_reserve_shared(vm->root.bo->tbo.base.resv, 1);
 	if (ret)
 		goto reserve_shared_fail;
-	amdgpu_bo_fence(vm->root.base.bo,
+	amdgpu_bo_fence(vm->root.bo,
 			&vm->process_info->eviction_fence->base, true);
-	amdgpu_bo_unreserve(vm->root.base.bo);
+	amdgpu_bo_unreserve(vm->root.bo);
 
 	/* Update process info */
 	mutex_lock(&vm->process_info->lock);
@@ -1264,7 +1251,7 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 reserve_shared_fail:
 wait_pd_fail:
 validate_pd_fail:
-	amdgpu_bo_unreserve(vm->root.base.bo);
+	amdgpu_bo_unreserve(vm->root.bo);
 reserve_pd_fail:
 	vm->process_info = NULL;
 	if (info) {
@@ -1319,7 +1306,7 @@ void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
 				    struct amdgpu_vm *vm)
 {
 	struct amdkfd_process_info *process_info = vm->process_info;
-	struct amdgpu_bo *pd = vm->root.base.bo;
+	struct amdgpu_bo *pd = vm->root.bo;
 
 	if (!process_info)
 		return;
@@ -1375,7 +1362,7 @@ void amdgpu_amdkfd_gpuvm_release_process_vm(struct kgd_dev *kgd, void *drm_priv)
 uint64_t amdgpu_amdkfd_gpuvm_get_process_page_dir(void *drm_priv)
 {
 	struct amdgpu_vm *avm = drm_priv_to_vm(drm_priv);
-	struct amdgpu_bo *pd = avm->root.base.bo;
+	struct amdgpu_bo *pd = avm->root.bo;
 	struct amdgpu_device *adev = amdgpu_ttm_adev(pd->tbo.bdev);
 
 	if (adev->asic_type < CHIP_VEGA10)
@@ -2402,7 +2389,7 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence **ef)
 	/* Attach eviction fence to PD / PT BOs */
 	list_for_each_entry(peer_vm, &process_info->vm_list_head,
 			    vm_list_node) {
-		struct amdgpu_bo *bo = peer_vm->root.base.bo;
+		struct amdgpu_bo *bo = peer_vm->root.bo;
 
 		amdgpu_bo_fence(bo, &process_info->eviction_fence->base, true);
 	}

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

@@ -832,7 +832,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p)
 	if (r)
 		return r;
 
-	p->job->vm_pd_addr = amdgpu_gmc_pd_addr(vm->root.base.bo);
+	p->job->vm_pd_addr = amdgpu_gmc_pd_addr(vm->root.bo);
 
 	if (amdgpu_vm_debug) {
 		/* Invalidate all BOs to test for userspace bugs */

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

@@ -1304,11 +1304,11 @@ static int amdgpu_debugfs_vm_info_show(struct seq_file *m, void *unused)
 
 		seq_printf(m, "pid:%d\tProcess:%s ----------\n",
 				vm->task_info.pid, vm->task_info.process_name);
-		r = amdgpu_bo_reserve(vm->root.base.bo, true);
+		r = amdgpu_bo_reserve(vm->root.bo, true);
 		if (r)
 			break;
 		amdgpu_debugfs_vm_bo_info(vm, m);
-		amdgpu_bo_unreserve(vm->root.base.bo);
+		amdgpu_bo_unreserve(vm->root.bo);
 	}
 
 	mutex_unlock(&dev->filelist_mutex);

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

@@ -4124,6 +4124,7 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev)
 {
 	struct dma_fence *fence = NULL, *next = NULL;
 	struct amdgpu_bo *shadow;
+	struct amdgpu_bo_vm *vmbo;
 	long r = 1, tmo;
 
 	if (amdgpu_sriov_runtime(adev))
@@ -4133,8 +4134,8 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev)
 
 	dev_info(adev->dev, "recover vram bo from shadow start\n");
 	mutex_lock(&adev->shadow_list_lock);
-	list_for_each_entry(shadow, &adev->shadow_list, shadow_list) {
-
+	list_for_each_entry(vmbo, &adev->shadow_list, shadow_list) {
+		shadow = &vmbo->bo;
 		/* No need to recover an evicted BO */
 		if (shadow->tbo.resource->mem_type != TTM_PL_TT ||
 		    shadow->tbo.resource->start == AMDGPU_BO_INVALID_OFFSET ||

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

@@ -325,7 +325,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
 	return 0;
 }
 
-int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
+int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id, int number_instance,
 				    int *major, int *minor, int *revision)
 {
 	struct binary_header *bhdr;
@@ -357,7 +357,7 @@ int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
 		for (j = 0; j < num_ips; j++) {
 			ip = (struct ip *)(adev->mman.discovery_bin + ip_offset);
 
-			if (le16_to_cpu(ip->hw_id) == hw_id) {
+			if ((le16_to_cpu(ip->hw_id) == hw_id) && (ip->number_instance == number_instance)) {
 				if (major)
 					*major = ip->major;
 				if (minor)

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

@@ -30,7 +30,7 @@
 void amdgpu_discovery_fini(struct amdgpu_device *adev);
 int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev);
 void amdgpu_discovery_harvest_ip(struct amdgpu_device *adev);
-int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
+int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id, int number_instance,
                                     int *major, int *minor, int *revision);
 int amdgpu_discovery_get_gfx_info(struct amdgpu_device *adev);
 

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

@@ -448,7 +448,7 @@ amdgpu_dma_buf_move_notify(struct dma_buf_attachment *attach)
 
 	for (bo_base = bo->vm_bo; bo_base; bo_base = bo_base->next) {
 		struct amdgpu_vm *vm = bo_base->vm;
-		struct dma_resv *resv = vm->root.base.bo->tbo.base.resv;
+		struct dma_resv *resv = vm->root.bo->tbo.base.resv;
 
 		if (ticket) {
 			/* When we get an error here it means that somebody

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

@@ -69,13 +69,13 @@ void amdgpu_show_fdinfo(struct seq_file *m, struct file *f)
 	dev = PCI_SLOT(adev->pdev->devfn);
 	fn = PCI_FUNC(adev->pdev->devfn);
 
-	ret = amdgpu_bo_reserve(fpriv->vm.root.base.bo, false);
+	ret = amdgpu_bo_reserve(fpriv->vm.root.bo, false);
 	if (ret) {
 		DRM_ERROR("Fail to reserve bo\n");
 		return;
 	}
 	amdgpu_vm_get_memory(&fpriv->vm, &vram_mem, &gtt_mem, &cpu_mem);
-	amdgpu_bo_unreserve(fpriv->vm.root.base.bo);
+	amdgpu_bo_unreserve(fpriv->vm.root.bo);
 	seq_printf(m, "pdev:\t%04x:%02x:%02x.%d\npasid:\t%u\n", domain, bus,
 			dev, fn, fpriv->vm.pasid);
 	seq_printf(m, "vram mem:\t%llu kB\n", vram_mem/1024UL);

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

@@ -170,7 +170,7 @@ static int amdgpu_gem_object_open(struct drm_gem_object *obj,
 		return -EPERM;
 
 	if (abo->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID &&
-	    abo->tbo.base.resv != vm->root.base.bo->tbo.base.resv)
+	    abo->tbo.base.resv != vm->root.bo->tbo.base.resv)
 		return -EPERM;
 
 	r = amdgpu_bo_reserve(abo, false);
@@ -320,11 +320,11 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
-		r = amdgpu_bo_reserve(vm->root.base.bo, false);
+		r = amdgpu_bo_reserve(vm->root.bo, false);
 		if (r)
 			return r;
 
-		resv = vm->root.base.bo->tbo.base.resv;
+		resv = vm->root.bo->tbo.base.resv;
 	}
 
 	initial_domain = (u32)(0xffffffff & args->in.domains);
@@ -353,9 +353,9 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 		if (!r) {
 			struct amdgpu_bo *abo = gem_to_amdgpu_bo(gobj);
 
-			abo->parent = amdgpu_bo_ref(vm->root.base.bo);
+			abo->parent = amdgpu_bo_ref(vm->root.bo);
 		}
-		amdgpu_bo_unreserve(vm->root.base.bo);
+		amdgpu_bo_unreserve(vm->root.bo);
 	}
 	if (r)
 		return r;
@@ -841,7 +841,7 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data,
 		}
 		for (base = robj->vm_bo; base; base = base->next)
 			if (amdgpu_xgmi_same_hive(amdgpu_ttm_adev(robj->tbo.bdev),
-				amdgpu_ttm_adev(base->vm->root.base.bo->tbo.bdev))) {
+				amdgpu_ttm_adev(base->vm->root.bo->tbo.bdev))) {
 				r = -EINVAL;
 				amdgpu_bo_unreserve(robj);
 				goto out;

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

@@ -124,6 +124,22 @@ void amdgpu_register_gpu_instance(struct amdgpu_device *adev)
 	mutex_unlock(&mgpu_info.mutex);
 }
 
+static void amdgpu_get_audio_func(struct amdgpu_device *adev)
+{
+	struct pci_dev *p = NULL;
+
+	p = pci_get_domain_bus_and_slot(pci_domain_nr(adev->pdev->bus),
+			adev->pdev->bus->number, 1);
+	if (p) {
+		pm_runtime_get_sync(&p->dev);
+
+		pm_runtime_mark_last_busy(&p->dev);
+		pm_runtime_put_autosuspend(&p->dev);
+
+		pci_dev_put(p);
+	}
+}
+
 /**
  * amdgpu_driver_load_kms - Main load function for KMS.
  *
@@ -213,9 +229,35 @@ int amdgpu_driver_load_kms(struct amdgpu_device *adev, unsigned long flags)
 						DPM_FLAG_MAY_SKIP_RESUME);
 		pm_runtime_use_autosuspend(dev->dev);
 		pm_runtime_set_autosuspend_delay(dev->dev, 5000);
+
 		pm_runtime_allow(dev->dev);
+
 		pm_runtime_mark_last_busy(dev->dev);
 		pm_runtime_put_autosuspend(dev->dev);
+
+		/*
+		 * For runpm implemented via BACO, PMFW will handle the
+		 * timing for BACO in and out:
+		 *   - put ASIC into BACO state only when both video and
+		 *     audio functions are in D3 state.
+		 *   - pull ASIC out of BACO state when either video or
+		 *     audio function is in D0 state.
+		 * Also, at startup, PMFW assumes both functions are in
+		 * D0 state.
+		 *
+		 * So if snd driver was loaded prior to amdgpu driver
+		 * and audio function was put into D3 state, there will
+		 * be no PMFW-aware D-state transition(D0->D3) on runpm
+		 * suspend. Thus the BACO will be not correctly kicked in.
+		 *
+		 * Via amdgpu_get_audio_func(), the audio dev is put
+		 * into D0 state. Then there will be a PMFW-aware D-state
+		 * transition(D0->D3) on runpm suspend.
+		 */
+		if (amdgpu_device_supports_baco(dev) &&
+		    !(adev->flags & AMD_IS_APU) &&
+		    (adev->asic_type >= CHIP_NAVI10))
+			amdgpu_get_audio_func(adev);
 	}
 
 	if (amdgpu_acpi_smart_shift_update(dev, AMDGPU_SS_DRV_LOAD))
@@ -1220,7 +1262,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 	}
 
 	pasid = fpriv->vm.pasid;
-	pd = amdgpu_bo_ref(fpriv->vm.root.base.bo);
+	pd = amdgpu_bo_ref(fpriv->vm.root.bo);
 
 	amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr);
 	amdgpu_vm_fini(adev, &fpriv->vm);