Merge tag 'amd-drm-fixes-6.1-2022-11-23' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.1-2022-11-23:

amdgpu:
- DCN 3.1.4 fixes
- DP MST DSC deadlock fixes
- HMM userptr fixes
- Fix Aldebaran CU occupancy reporting
- GFX11 fixes
- PSP suspend/resume fix
- DCE12 KASAN fix
- DCN 3.2.x fixes
- Rotated cursor fix
- SMU 13.x fix
- DELL platform suspend/resume fixes
- VCN4 SR-IOV fix
- Display regression fix for polled connectors

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221123143453.8977-1-alexander.deucher@amd.com

Author: airlied

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

@@ -41,5 +41,6 @@ const struct kfd2kgd_calls aldebaran_kfd2kgd = {
 	.get_atc_vmid_pasid_mapping_info =
 				kgd_gfx_v9_get_atc_vmid_pasid_mapping_info,
 	.set_vm_context_page_table_base = kgd_gfx_v9_set_vm_context_page_table_base,
+	.get_cu_occupancy = kgd_gfx_v9_get_cu_occupancy,
 	.program_trap_handler_settings = kgd_gfx_v9_program_trap_handler_settings
 };

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

@@ -986,6 +986,7 @@ static int init_user_pages(struct kgd_mem *mem, uint64_t user_addr,
 	struct amdkfd_process_info *process_info = mem->process_info;
 	struct amdgpu_bo *bo = mem->bo;
 	struct ttm_operation_ctx ctx = { true, false };
+	struct hmm_range *range;
 	int ret = 0;
 
 	mutex_lock(&process_info->lock);
@@ -1015,7 +1016,7 @@ static int init_user_pages(struct kgd_mem *mem, uint64_t user_addr,
 		return 0;
 	}
 
-	ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages);
+	ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages, &range);
 	if (ret) {
 		pr_err("%s: Failed to get user pages: %d\n", __func__, ret);
 		goto unregister_out;
@@ -1033,7 +1034,7 @@ static int init_user_pages(struct kgd_mem *mem, uint64_t user_addr,
 	amdgpu_bo_unreserve(bo);
 
 release_out:
-	amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
+	amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm, range);
 unregister_out:
 	if (ret)
 		amdgpu_mn_unregister(bo);
@@ -2370,6 +2371,8 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
 	/* Go through userptr_inval_list and update any invalid user_pages */
 	list_for_each_entry(mem, &process_info->userptr_inval_list,
 			    validate_list.head) {
+		struct hmm_range *range;
+
 		invalid = atomic_read(&mem->invalid);
 		if (!invalid)
 			/* BO hasn't been invalidated since the last
@@ -2380,7 +2383,8 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
 		bo = mem->bo;
 
 		/* Get updated user pages */
-		ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages);
+		ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages,
+						   &range);
 		if (ret) {
 			pr_debug("Failed %d to get user pages\n", ret);
 
@@ -2399,7 +2403,7 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
 			 * FIXME: Cannot ignore the return code, must hold
 			 * notifier_lock
 			 */
-			amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
+			amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm, range);
 		}
 
 		/* Mark the BO as valid unless it was invalidated

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

@@ -209,6 +209,7 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
 			list_add_tail(&e->tv.head, &bucket[priority]);
 
 		e->user_pages = NULL;
+		e->range = NULL;
 	}
 
 	/* Connect the sorted buckets in the output list. */

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

@@ -26,6 +26,8 @@
 #include <drm/ttm/ttm_execbuf_util.h>
 #include <drm/amdgpu_drm.h>
 
+struct hmm_range;
+
 struct amdgpu_device;
 struct amdgpu_bo;
 struct amdgpu_bo_va;
@@ -36,6 +38,7 @@ struct amdgpu_bo_list_entry {
 	struct amdgpu_bo_va		*bo_va;
 	uint32_t			priority;
 	struct page			**user_pages;
+	struct hmm_range		*range;
 	bool				user_invalidated;
 };
 

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

@@ -328,7 +328,6 @@ static void amdgpu_connector_free_edid(struct drm_connector *connector)
 
 	kfree(amdgpu_connector->edid);
 	amdgpu_connector->edid = NULL;
-	drm_connector_update_edid_property(connector, NULL);
 }
 
 static int amdgpu_connector_ddc_get_modes(struct drm_connector *connector)

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

@@ -913,7 +913,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 			goto out_free_user_pages;
 		}
 
-		r = amdgpu_ttm_tt_get_user_pages(bo, e->user_pages);
+		r = amdgpu_ttm_tt_get_user_pages(bo, e->user_pages, &e->range);
 		if (r) {
 			kvfree(e->user_pages);
 			e->user_pages = NULL;
@@ -991,9 +991,10 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 
 		if (!e->user_pages)
 			continue;
-		amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
+		amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm, e->range);
 		kvfree(e->user_pages);
 		e->user_pages = NULL;
+		e->range = NULL;
 	}
 	mutex_unlock(&p->bo_list->bo_list_mutex);
 	return r;
@@ -1273,7 +1274,8 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) {
 		struct amdgpu_bo *bo = ttm_to_amdgpu_bo(e->tv.bo);
 
-		r |= !amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
+		r |= !amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm, e->range);
+		e->range = NULL;
 	}
 	if (r) {
 		r = -EAGAIN;

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

@@ -378,6 +378,7 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
 	struct amdgpu_device *adev = drm_to_adev(dev);
 	struct drm_amdgpu_gem_userptr *args = data;
 	struct drm_gem_object *gobj;
+	struct hmm_range *range;
 	struct amdgpu_bo *bo;
 	uint32_t handle;
 	int r;
@@ -413,14 +414,13 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
 	if (r)
 		goto release_object;
 
-	if (args->flags & AMDGPU_GEM_USERPTR_REGISTER) {
-		r = amdgpu_mn_register(bo, args->addr);
-		if (r)
-			goto release_object;
-	}
+	r = amdgpu_mn_register(bo, args->addr);
+	if (r)
+		goto release_object;
 
 	if (args->flags & AMDGPU_GEM_USERPTR_VALIDATE) {
-		r = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages);
+		r = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages,
+						 &range);
 		if (r)
 			goto release_object;
 
@@ -443,7 +443,7 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
 
 user_pages_done:
 	if (args->flags & AMDGPU_GEM_USERPTR_VALIDATE)
-		amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm);
+		amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm, range);
 
 release_object:
 	drm_gem_object_put(gobj);

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

@@ -479,6 +479,12 @@ int amdgpu_gmc_allocate_vm_inv_eng(struct amdgpu_device *adev)
 	unsigned i;
 	unsigned vmhub, inv_eng;
 
+	if (adev->enable_mes) {
+		/* reserve engine 5 for firmware */
+		for (vmhub = 0; vmhub < AMDGPU_MAX_VMHUBS; vmhub++)
+			vm_inv_engs[vmhub] &= ~(1 << 5);
+	}
+
 	for (i = 0; i < adev->num_rings; ++i) {
 		ring = adev->rings[i];
 		vmhub = ring->funcs->vmhub;

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

@@ -169,7 +169,11 @@ static void amdgpu_job_free_cb(struct drm_sched_job *s_job)
 	amdgpu_sync_free(&job->sync);
 	amdgpu_sync_free(&job->sched_sync);
 
-	dma_fence_put(&job->hw_fence);
+	/* only put the hw fence if has embedded fence */
+	if (!job->hw_fence.ops)
+		kfree(job);
+	else
+		dma_fence_put(&job->hw_fence);
 }
 
 void amdgpu_job_set_gang_leader(struct amdgpu_job *job,

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

@@ -172,6 +172,7 @@ void psp_ta_free_shared_buf(struct ta_mem_context *mem_ctx)
 {
 	amdgpu_bo_free_kernel(&mem_ctx->shared_bo, &mem_ctx->shared_mc_addr,
 			      &mem_ctx->shared_buf);
+	mem_ctx->shared_bo = NULL;
 }
 
 static void psp_free_shared_bufs(struct psp_context *psp)
@@ -182,6 +183,7 @@ static void psp_free_shared_bufs(struct psp_context *psp)
 	/* free TMR memory buffer */
 	pptr = amdgpu_sriov_vf(psp->adev) ? &tmr_buf : NULL;
 	amdgpu_bo_free_kernel(&psp->tmr_bo, &psp->tmr_mc_addr, pptr);
+	psp->tmr_bo = NULL;
 
 	/* free xgmi shared memory */
 	psp_ta_free_shared_buf(&psp->xgmi_context.context.mem_context);
@@ -743,7 +745,7 @@ static int psp_load_toc(struct psp_context *psp,
 /* Set up Trusted Memory Region */
 static int psp_tmr_init(struct psp_context *psp)
 {
-	int ret;
+	int ret = 0;
 	int tmr_size;
 	void *tmr_buf;
 	void **pptr;
@@ -770,10 +772,12 @@ static int psp_tmr_init(struct psp_context *psp)
 		}
 	}
 
-	pptr = amdgpu_sriov_vf(psp->adev) ? &tmr_buf : NULL;
-	ret = amdgpu_bo_create_kernel(psp->adev, tmr_size, PSP_TMR_ALIGNMENT,
-				      AMDGPU_GEM_DOMAIN_VRAM,
-				      &psp->tmr_bo, &psp->tmr_mc_addr, pptr);
+	if (!psp->tmr_bo) {
+		pptr = amdgpu_sriov_vf(psp->adev) ? &tmr_buf : NULL;
+		ret = amdgpu_bo_create_kernel(psp->adev, tmr_size, PSP_TMR_ALIGNMENT,
+					      AMDGPU_GEM_DOMAIN_VRAM,
+					      &psp->tmr_bo, &psp->tmr_mc_addr, pptr);
+	}
 
 	return ret;
 }
@@ -2732,8 +2736,6 @@ static int psp_suspend(void *handle)
 	}
 
 out:
-	psp_free_shared_bufs(psp);
-
 	return ret;
 }