1 files changed, 33 insertions, 29 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 57abf7abd7a9..4cea9ab237ac 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -90,6 +90,12 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
                 goto free_chunk;
         }
 
+        /* skip guilty context job */
+        if (atomic_read(&p->ctx->guilty) == 1) {
+                ret = -ECANCELED;
+                goto free_chunk;
+        }
+
         mutex_lock(&p->ctx->lock);
 
         /* get chunks */
@@ -337,7 +343,7 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
                                  struct amdgpu_bo *bo)
 {
         struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-        u64 initial_bytes_moved, bytes_moved;
+        struct ttm_operation_ctx ctx = { true, false };
         uint32_t domain;
         int r;
 
@@ -367,15 +373,13 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
 
 retry:
         amdgpu_ttm_placement_from_domain(bo, domain);
-        initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
-        r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
-        bytes_moved = atomic64_read(&adev->num_bytes_moved) -
-                      initial_bytes_moved;
-        p->bytes_moved += bytes_moved;
+        r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+
+        p->bytes_moved += ctx.bytes_moved;
         if (adev->mc.visible_vram_size < adev->mc.real_vram_size &&
             bo->tbo.mem.mem_type == TTM_PL_VRAM &&
             bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT)
-                p->bytes_moved_vis += bytes_moved;
+                p->bytes_moved_vis += ctx.bytes_moved;
 
         if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
                 domain = bo->allowed_domains;
@@ -390,6 +394,7 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
                                 struct amdgpu_bo *validated)
 {
         uint32_t domain = validated->allowed_domains;
+        struct ttm_operation_ctx ctx = { true, false };
         int r;
 
         if (!p->evictable)
@@ -431,7 +436,7 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
                         bo->tbo.mem.mem_type == TTM_PL_VRAM &&
                         bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT;
                 initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
-                r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+                r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
                 bytes_moved = atomic64_read(&adev->num_bytes_moved) -
                         initial_bytes_moved;
                 p->bytes_moved += bytes_moved;
@@ -470,6 +475,7 @@ static int amdgpu_cs_validate(void *param, struct amdgpu_bo *bo)
 static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
                             struct list_head *validated)
 {
+        struct ttm_operation_ctx ctx = { true, false };
         struct amdgpu_bo_list_entry *lobj;
         int r;
 
@@ -487,8 +493,7 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
                     lobj->user_pages) {
                         amdgpu_ttm_placement_from_domain(bo,
                                                          AMDGPU_GEM_DOMAIN_CPU);
-                        r = ttm_bo_validate(&bo->tbo, &bo->placement, true,
-                                            false);
+                        r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
                         if (r)
                                 return r;
                         amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm,
@@ -678,7 +683,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
         if (!r && p->uf_entry.robj) {
                 struct amdgpu_bo *uf = p->uf_entry.robj;
 
-                r = amdgpu_ttm_bind(&uf->tbo, &uf->tbo.mem);
+                r = amdgpu_ttm_alloc_gart(&uf->tbo);
                 p->job->uf_addr += amdgpu_bo_gpu_offset(uf);
         }
 
@@ -781,7 +786,7 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
                 return r;
 
         r = amdgpu_sync_fence(adev, &p->job->sync,
-                              fpriv->prt_va->last_pt_update);
+                              fpriv->prt_va->last_pt_update, false);
         if (r)
                 return r;
 
@@ -795,7 +800,7 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
                         return r;
 
                 f = bo_va->last_pt_update;
-                r = amdgpu_sync_fence(adev, &p->job->sync, f);
+                r = amdgpu_sync_fence(adev, &p->job->sync, f, false);
                 if (r)
                         return r;
         }
@@ -818,7 +823,7 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
                                 return r;
 
                         f = bo_va->last_pt_update;
-                        r = amdgpu_sync_fence(adev, &p->job->sync, f);
+                        r = amdgpu_sync_fence(adev, &p->job->sync, f, false);
                         if (r)
                                 return r;
                 }
@@ -829,7 +834,7 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
         if (r)
                 return r;
 
-        r = amdgpu_sync_fence(adev, &p->job->sync, vm->last_update);
+        r = amdgpu_sync_fence(adev, &p->job->sync, vm->last_update, false);
         if (r)
                 return r;
 
@@ -865,8 +870,8 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
                         struct amdgpu_bo_va_mapping *m;
                         struct amdgpu_bo *aobj = NULL;
                         struct amdgpu_cs_chunk *chunk;
+                        uint64_t offset, va_start;
                         struct amdgpu_ib *ib;
-                        uint64_t offset;
                         uint8_t *kptr;
 
                         chunk = &p->chunks[i];
@@ -876,14 +881,14 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
                         if (chunk->chunk_id != AMDGPU_CHUNK_ID_IB)
                                 continue;
 
-                        r = amdgpu_cs_find_mapping(p, chunk_ib->va_start,
-                                                   &aobj, &m);
+                        va_start = chunk_ib->va_start & AMDGPU_VA_HOLE_MASK;
+                        r = amdgpu_cs_find_mapping(p, va_start, &aobj, &m);
                         if (r) {
                                 DRM_ERROR("IB va_start is invalid\n");
                                 return r;
                         }
 
-                        if ((chunk_ib->va_start + chunk_ib->ib_bytes) >
+                        if ((va_start + chunk_ib->ib_bytes) >
                             (m->last + 1) * AMDGPU_GPU_PAGE_SIZE) {
                                 DRM_ERROR("IB va_start+ib_bytes is invalid\n");
                                 return -EINVAL;
@@ -896,7 +901,7 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
                         }
 
                         offset = m->start * AMDGPU_GPU_PAGE_SIZE;
-                        kptr += chunk_ib->va_start - offset;
+                        kptr += va_start - offset;
 
                         memcpy(ib->ptr, kptr, chunk_ib->ib_bytes);
                         amdgpu_bo_kunmap(aobj);
@@ -1033,8 +1038,8 @@ static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
                         amdgpu_ctx_put(ctx);
                         return r;
                 } else if (fence) {
-                        r = amdgpu_sync_fence(p->adev, &p->job->sync,
-                                              fence);
+                        r = amdgpu_sync_fence(p->adev, &p->job->sync, fence,
+                                        true);
                         dma_fence_put(fence);
                         amdgpu_ctx_put(ctx);
                         if (r)
@@ -1053,7 +1058,7 @@ static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p,
         if (r)
                 return r;
 
-        r = amdgpu_sync_fence(p->adev, &p->job->sync, fence);
+        r = amdgpu_sync_fence(p->adev, &p->job->sync, fence, true);
         dma_fence_put(fence);
 
         return r;
@@ -1194,11 +1199,10 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
         job->uf_sequence = seq;
 
         amdgpu_job_free_resources(job);
-        amdgpu_ring_priority_get(job->ring,
-                                 amd_sched_get_job_priority(&job->base));
+        amdgpu_ring_priority_get(job->ring, job->base.s_priority);
 
         trace_amdgpu_cs_ioctl(job);
-        amd_sched_entity_push_job(&job->base);
+        amd_sched_entity_push_job(&job->base, entity);
 
         ttm_eu_fence_buffer_objects(&p->ticket, &p->validated, p->fence);
         amdgpu_mn_unlock(p->mn);
@@ -1570,6 +1574,7 @@ int amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser,
                            struct amdgpu_bo_va_mapping **map)
 {
         struct amdgpu_fpriv *fpriv = parser->filp->driver_priv;
+        struct ttm_operation_ctx ctx = { false, false };
         struct amdgpu_vm *vm = &fpriv->vm;
         struct amdgpu_bo_va_mapping *mapping;
         int r;
@@ -1590,11 +1595,10 @@ int amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser,
         if (!((*bo)->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)) {
                 (*bo)->flags |= AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
                 amdgpu_ttm_placement_from_domain(*bo, (*bo)->allowed_domains);
-                r = ttm_bo_validate(&(*bo)->tbo, &(*bo)->placement, false,
-                                    false);
+                r = ttm_bo_validate(&(*bo)->tbo, &(*bo)->placement, &ctx);
                 if (r)
                         return r;
         }
 
-        return amdgpu_ttm_bind(&(*bo)->tbo, &(*bo)->tbo.mem);
+        return amdgpu_ttm_alloc_gart(&(*bo)->tbo);
 }