drm/amdgpu: fix another potential cause of VM faults

"Linux Kernel Mailing List" <[email protected]>
Newsgroups gmane.linux.kernel.commits.head
Message-ID <[email protected]>
Web:        https://git.kernel.org/torvalds/c/d3aab672de4902ce25e9858fbb5bf67e8254b23c
Commit:     d3aab672de4902ce25e9858fbb5bf67e8254b23c
Parent:     59fd27cd2fd762b08a1f54de8bbe3b82aba06a13
Refname:    refs/heads/master
Author:     Christian König <[email protected]>
AuthorDate: Wed Jan 24 14:57:02 2018 +0100
Committer:  Alex Deucher <[email protected]>
CommitDate: Mon Jan 29 23:15:26 2018 -0500

    drm/amdgpu: fix another potential cause of VM faults
    
    The root PD can be evicted directly after allocating it, just validate
    it on first use.
    
    Signed-off-by: Christian König <[email protected]>
    Acked-by: Felix Kuehling <[email protected]>
    Signed-off-by: Alex Deucher <[email protected]>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 32 ++++++++++++--------------------
 1 file changed, 12 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 6fc16eecf2dc..5afbc5e714d0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -2262,12 +2262,12 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 {
 	const unsigned align = min(AMDGPU_VM_PTB_ALIGN_SIZE,
 		AMDGPU_VM_PTE_COUNT(adev) * 8);
+	uint64_t init_pde_value = 0, flags;
 	unsigned ring_instance;
 	struct amdgpu_ring *ring;
 	struct drm_sched_rq *rq;
+	unsigned long size;
 	int r, i;
-	u64 flags;
-	uint64_t init_pde_value = 0;
 
 	vm->va = RB_ROOT_CACHED;
 	for (i = 0; i < AMDGPU_MAX_VMHUBS; i++)
@@ -2318,29 +2318,21 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		flags |= (AMDGPU_GEM_CREATE_NO_CPU_ACCESS |
 				AMDGPU_GEM_CREATE_SHADOW);
 
-	r = amdgpu_bo_create(adev,
-			     amdgpu_vm_bo_size(adev, adev->vm_manager.root_level),
-			     align, true,
-			     AMDGPU_GEM_DOMAIN_VRAM,
-			     flags,
-			     NULL, NULL, init_pde_value, &vm->root.base.bo);
+	size = amdgpu_vm_bo_size(adev, adev->vm_manager.root_level);
+	r = amdgpu_bo_create(adev, size, align, true, AMDGPU_GEM_DOMAIN_VRAM,
+			     flags, NULL, NULL, init_pde_value,
+			     &vm->root.base.bo);
 	if (r)
 		goto error_free_sched_entity;
 
+	r = amdgpu_bo_reserve(vm->root.base.bo, true);
+	if (r)
+		goto error_free_root;
+
 	vm->root.base.vm = vm;
 	list_add_tail(&vm->root.base.bo_list, &vm->root.base.bo->va);
-	INIT_LIST_HEAD(&vm->root.base.vm_status);
-
-	if (vm->use_cpu_for_update) {
-		r = amdgpu_bo_reserve(vm->root.base.bo, false);
-		if (r)
-			goto error_free_root;
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.