[AMD Official Use Only - Internal Distribution Only]
Ping......
-----Original Message-----
From: Jiawei <jiawei...@amd.com>
Sent: Wednesday, March 25, 2020 4:32 PM
To: amd-gfx@lists.freedesktop.org
Cc: Gu, JiaWei (Will) <jiawei...@amd.com>
Subject: [PATCH] SWDEV-227226 [AWS][Linux]ReallyQuick test failed, guest dmesg
and host dmesg have error
root cause: compute job timeout for sriov/passthrough is 10000 ms, which is too
short for some compute benchmark
solution: extend the default compute lockup timeout to 60000 ms
Signed-off-by: Jiawei <jiawei...@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 3607a63f48bb..88360b220a8f 100755
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2680,12 +2680,12 @@ static int
amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
* By default timeout for non compute jobs is 10000.
* And there is no timeout enforced on compute jobs.
* In SR-IOV or passthrough mode, timeout for compute
- * jobs are 10000 by default.
+ * jobs are 60000 by default.
*/
adev->gfx_timeout = msecs_to_jiffies(10000);
adev->sdma_timeout = adev->video_timeout = adev->gfx_timeout;
if (amdgpu_sriov_vf(adev) || amdgpu_passthrough(adev))
- adev->compute_timeout = adev->gfx_timeout;
+ adev->compute_timeout = msecs_to_jiffies(60000);
else
adev->compute_timeout = MAX_SCHEDULE_TIMEOUT;
--
2.20.1
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
