drm/amdgpu: skipping SDMA hw_init and hw_fini for S0ix.
authorRajib Mahapatra <rajib.mahapatra@amd.com>
Thu, 10 Feb 2022 13:16:40 +0000 (18:46 +0530)
committerAlex Deucher <alexander.deucher@amd.com>
Mon, 14 Feb 2022 19:59:46 +0000 (14:59 -0500)
[Why]
SDMA ring buffer test failed if suspend is aborted during
S0i3 resume.

[How]
If suspend is aborted for some reason during S0i3 resume
cycle, it follows SDMA ring test failing and errors in amdgpu
resume. For RN/CZN/Picasso, SMU saves and restores SDMA
registers during S0ix cycle. So, skipping SDMA suspend and
resume from driver solves the issue. This time, the system
is able to resume gracefully even the suspend is aborted.

Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rajib Mahapatra <rajib.mahapatra@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c

index e8e4749..f0638db 100644 (file)
@@ -2057,6 +2057,10 @@ static int sdma_v4_0_suspend(void *handle)
 {
        struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+       /* SMU saves SDMA state for us */
+       if (adev->in_s0ix)
+               return 0;
+
        return sdma_v4_0_hw_fini(adev);
 }
 
@@ -2064,6 +2068,10 @@ static int sdma_v4_0_resume(void *handle)
 {
        struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+       /* SMU restores SDMA state for us */
+       if (adev->in_s0ix)
+               return 0;
+
        return sdma_v4_0_hw_init(adev);
 }