[TRACKING] AMD, Fedora 39 Beta and Firefox

Thanks. I still get these errors regularly. Another site that triggers it is gog.com, when browsing the game thumbnails.

[ 1833.240545] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, but soft recovered
[ 1833.510718] amdgpu 0000:c1:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32773)
[ 1833.510736] amdgpu 0000:c1:00.0: amdgpu:  in process firefox pid 3806 thread firefox:cs0 pid 3906)
[ 1833.510744] amdgpu 0000:c1:00.0: amdgpu:   in page starting at address 0x0000000000000000 from client 10
[ 1833.510751] amdgpu 0000:c1:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00601430
[ 1833.510757] amdgpu 0000:c1:00.0: amdgpu: 	 Faulty UTCL2 client ID: SQC (data) (0xa)
[ 1833.510763] amdgpu 0000:c1:00.0: amdgpu: 	 MORE_FAULTS: 0x0
[ 1833.510769] amdgpu 0000:c1:00.0: amdgpu: 	 WALKER_ERROR: 0x0
[ 1833.510772] amdgpu 0000:c1:00.0: amdgpu: 	 PERMISSION_FAULTS: 0x3
[ 1833.510775] amdgpu 0000:c1:00.0: amdgpu: 	 MAPPING_ERROR: 0x0
[ 1833.510778] amdgpu 0000:c1:00.0: amdgpu: 	 RW: 0x0
[ 1843.992485] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, but soft recovered

Can’t reproduce on either gog.com or openstreetmap.org - any particular page/content on GOG that reproduces this on your machine?

One difference I notice is that I don’t use RPM Fusion, so my about:support includes this mesa version:

I switched back to the Fedora mesa for now. Let’s see if it is still reproducible.

Interestingly enough, I can still reproduce it with the Fedora mesa:

[11415.063404] amdgpu 0000:c1:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:2 pasid:32777)
[11415.063413] amdgpu 0000:c1:00.0: amdgpu:  in process firefox pid 17576 thread firefox:cs0 pid 17668)
[11415.063416] amdgpu 0000:c1:00.0: amdgpu:   in page starting at address 0x0000000000000000 from client 10
[11415.063421] amdgpu 0000:c1:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00201430
[11415.063425] amdgpu 0000:c1:00.0: amdgpu: 	 Faulty UTCL2 client ID: SQC (data) (0xa)
[11415.063428] amdgpu 0000:c1:00.0: amdgpu: 	 MORE_FAULTS: 0x0
[11415.063430] amdgpu 0000:c1:00.0: amdgpu: 	 WALKER_ERROR: 0x0
[11415.063432] amdgpu 0000:c1:00.0: amdgpu: 	 PERMISSION_FAULTS: 0x3
[11415.063440] amdgpu 0000:c1:00.0: amdgpu: 	 MAPPING_ERROR: 0x0
[11415.063442] amdgpu 0000:c1:00.0: amdgpu: 	 RW: 0x0
[11425.472365] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, but soft recovered

What could effect this? I’m running fairly standard apart from:

  1. sudo dnf swap ffmpeg-free ffmpeg --allowerasing
  2. sudo dnf update @multimedia --setopt="install_weak_deps=False" --exclude=PackageKit-gstreamer-plugin
  3. sudo dnf update @sound-and-video
  4. media.ffmpeg.vaapi.enabled enabled in Firefox