[PATCH] drm/amdgpu: fix error handling in amdgpu_bo_do_create

[michel@xxxxxxxxxxx (Michel Dänzer)]

On 31/10/17 04:03 PM, Christian König wrote:
> Am 31.10.2017 um 09:43 schrieb Michel Dänzer:
>> On 31/10/17 09:37 AM, Christian König wrote:
>>> From: Christian König <christian.koenig at amd.com>
>>>
>>> The bo structure is freed up in case of an error, so we can't do any
>>> accounting if that happens.
>>>
>>> Signed-off-by: Christian König <christian.koenig at amd.com>
>>> ---
>>> Â  drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 6 +++---
>>> Â  1 file changed, 3 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
>>> index e44b880eefdd..ff6f842655d1 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
>>> @@ -369,6 +369,9 @@ static int amdgpu_bo_do_create(struct
>>> amdgpu_device *adev,
>>> Â Â Â Â Â  r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
>>> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â  &bo->placement, page_align, !kernel, NULL,
>>> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â  acc_size, sg, resv, &amdgpu_ttm_bo_destroy);
>>> +Â Â Â  if (unlikely(r != 0))
>>> +Â Â Â Â Â Â Â  return r;
>>> +
>>> Â Â Â Â Â  bytes_moved = atomic64_read(&adev->num_bytes_moved) -
>>> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â  initial_bytes_moved;
>>> Â Â Â Â Â  if (adev->mc.visible_vram_size < adev->mc.real_vram_size &&
>>> @@ -378,9 +381,6 @@ static int amdgpu_bo_do_create(struct
>>> amdgpu_device *adev,
>>> Â Â Â Â Â  else
>>> Â Â Â Â Â Â Â Â Â  amdgpu_cs_report_moved_bytes(adev, bytes_moved, 0);
>>> Â  -Â Â Â  if (unlikely(r != 0))
>>> -Â Â Â Â Â Â Â  return r;
>>> -
>>> Â Â Â Â Â  if (kernel)
>>> Â Â Â Â Â Â Â Â Â  bo->tbo.priority = 1;
>>> Â 
>> Hmm. If it's at least theoretically possible that ttm_bo_init_reserved
>> moves some BOs before returning an error, we should instead make the
>> accounting safe WRT bo having been freed. But assuming it's not possible,
> 
> I have made patches for this back in April and send them to the list,
> but the value is rather low and you need to change a lot of places in
> other drivers as well.

At least in this case, simply moving the

	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)

test before the ttm_bo_init_reserved call, and then passing bytes_moved
or 0 for the third parameter of amdgpu_cs_report_moved_bytes depending
on its result, would work as well, wouldn't it?


-- 
Earthling Michel Dänzer               |               http://www.amd.com
Libre software enthusiast             |             Mesa and X developer