Problem: There is no point to reserve bad pages before GPU reset as it was done untill now since we need to do it after ASIC was reset as we lose all reservation during ASIC reset. Fix: Call amdgpu_ras_reserve_bad_pages after ASIC is reset and not before. Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@xxxxxxx> --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index f2feff4..eaac2b7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -3804,6 +3804,10 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive, break; } } + + /* Mark vram pages with errors as bad after ASIC was reset */ + list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) + amdgpu_ras_reserve_bad_pages(tmp_adev); } } -- 2.7.4 _______________________________________________ amd-gfx mailing list amd-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/amd-gfx
