]> asedeno.scripts.mit.edu Git - linux.git/commit
drm/amdgpu: Set VM_L2_CNTL.PDE_FAULT_CLASSIFICATION to 0 for GFX10
authorYong Zhao <Yong.Zhao@amd.com>
Tue, 13 Aug 2019 18:38:03 +0000 (14:38 -0400)
committerAlex Deucher <alexander.deucher@amd.com>
Thu, 15 Aug 2019 15:58:14 +0000 (10:58 -0500)
commit8b7d6157f2188ce94040ccb367490dbca929104a
treebe1ba7fd0391ea82caccb67511fa287a017bc278
parent5d36d4c97635f16ade47c86f4b3dd5843c79a7b2
drm/amdgpu: Set VM_L2_CNTL.PDE_FAULT_CLASSIFICATION to 0 for GFX10

We have done this for pre-GFX10 asics, but GFX10 did not pick up the
new change. The below is the commit message for that change.

This is recommended by HW designers. Previously when it was set to 1,
the PDE walk error in VM fault will be treated as
PERMISSION_OR_INVALID_PAGE_FAULT rather than usually expected OTHER_FAULT.
As a result, the retry control in VM_CONTEXT*_CNTL will change accordingly.

The above behavior is kind of abnormal. Furthermore, the
PDE_FAULT_CLASSIFICATION == 1 feature was targeted for very old ASICs
and it never made it way to production. Therefore, we should set it to 0.

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/gfxhub_v2_0.c
drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c