drm/amdgpu: Fix a bug in amdgpu_fill_buffer()
authorYong Zhao <yong.zhao@amd.com>
Fri, 15 Sep 2017 22:20:37 +0000 (18:20 -0400)
committerAlex Deucher <alexander.deucher@amd.com>
Tue, 26 Sep 2017 19:14:21 +0000 (15:14 -0400)
commit7bdc53f925af085ffa0580f10489f82b36cc2f1c
tree8b760f772a29969ecce509371d55e22111a19627
parentdfe5c2b76b2a32cd37283809737e55f9208f8346
drm/amdgpu: Fix a bug in amdgpu_fill_buffer()

When max_bytes is not 8 bytes aligned and bo size is larger than
max_bytes, the last 8 bytes in a ttm node may be left unchanged.
For example, on pre SDMA 4.0, max_bytes = 0x1fffff, and the bo size
is 0x200000, the problem will happen.

In order to fix the problem, we separately store the max nums of
PTEs/PDEs a single operation can set in amdgpu_vm_pte_funcs
structure, rather than inferring it from bytes limit of SDMA
constant fill, i.e. fill_max_bytes.

Together with the fix, we replace the hard code value "10" in
amdgpu_vm_bo_update_mapping() with the corresponding values from
structure amdgpu_vm_pte_funcs.

Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/amd/amdgpu/amdgpu.h
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
drivers/gpu/drm/amd/amdgpu/cik_sdma.c
drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
drivers/gpu/drm/amd/amdgpu/si_dma.c