drm/amdgpu: remove the GART copy hack
authorChristian König <christian.koenig@amd.com>
Tue, 22 Aug 2017 10:50:46 +0000 (12:50 +0200)
committerAlex Deucher <alexander.deucher@amd.com>
Thu, 24 Aug 2017 15:48:47 +0000 (11:48 -0400)
This isn't used since we don't map evicted BOs to GART any more.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Roger He <Hongbo.He@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

index 96ec4e2..3bd430e 100644 (file)
@@ -1616,7 +1616,6 @@ error_free:
  *
  * @adev: amdgpu_device pointer
  * @exclusive: fence we need to sync to
- * @gtt_flags: flags as they are used for GTT
  * @pages_addr: DMA addresses to use for mapping
  * @vm: requested vm
  * @mapping: mapped range and flags to use for the update
@@ -1630,7 +1629,6 @@ error_free:
  */
 static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
                                      struct dma_fence *exclusive,
-                                     uint64_t gtt_flags,
                                      dma_addr_t *pages_addr,
                                      struct amdgpu_vm *vm,
                                      struct amdgpu_bo_va_mapping *mapping,
@@ -1685,11 +1683,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
                }
 
                if (pages_addr) {
-                       if (flags == gtt_flags)
-                               src = adev->gart.table_addr +
-                                       (addr >> AMDGPU_GPU_PAGE_SHIFT) * 8;
-                       else
-                               max_entries = min(max_entries, 16ull * 1024ull);
+                       max_entries = min(max_entries, 16ull * 1024ull);
                        addr = 0;
                } else if (flags & AMDGPU_PTE_VALID) {
                        addr += adev->vm_manager.vram_base_offset;
@@ -1734,10 +1728,10 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
        struct amdgpu_vm *vm = bo_va->base.vm;
        struct amdgpu_bo_va_mapping *mapping;
        dma_addr_t *pages_addr = NULL;
-       uint64_t gtt_flags, flags;
        struct ttm_mem_reg *mem;
        struct drm_mm_node *nodes;
        struct dma_fence *exclusive;
+       uint64_t flags;
        int r;
 
        if (clear || !bo_va->base.bo) {
@@ -1757,15 +1751,10 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
                exclusive = reservation_object_get_excl(bo->tbo.resv);
        }
 
-       if (bo) {
+       if (bo)
                flags = amdgpu_ttm_tt_pte_flags(adev, bo->tbo.ttm, mem);
-               gtt_flags = (amdgpu_ttm_is_bound(bo->tbo.ttm) &&
-                       adev == amdgpu_ttm_adev(bo->tbo.bdev)) ?
-                       flags : 0;
-       } else {
+       else
                flags = 0x0;
-               gtt_flags = ~0x0;
-       }
 
        spin_lock(&vm->status_lock);
        if (!list_empty(&bo_va->base.vm_status))
@@ -1773,8 +1762,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
        spin_unlock(&vm->status_lock);
 
        list_for_each_entry(mapping, &bo_va->invalids, list) {
-               r = amdgpu_vm_bo_split_mapping(adev, exclusive,
-                                              gtt_flags, pages_addr, vm,
+               r = amdgpu_vm_bo_split_mapping(adev, exclusive, pages_addr, vm,
                                               mapping, flags, nodes,
                                               &bo_va->last_pt_update);
                if (r)