You are not logged in.
Running on an R9 5950x and RX 7900XT on the KDE Plasma 6 desktop, and this only shows up on one specific program im trying to run through wine(and proton with uwu-run). This ran just fine prior to the update I performed earlier today and has a 100% fault rate with a 10 second lockup followed these exact errors when attempting to run after said update. The pre-update kernel was 6.18.8 and post update was 6.18.9. Rolling back the kernel alone was unsuccessful at reverting the issue. Neither firmware nor Mesa packages updated at this moment, though glibc and quite a number of other packages did as it has been almost a week since the previous time I went to update. I may even have to fully downgrade all 300 odd packages from that last update to check if it can be made to function again to narrow down suspects, if I can even manage
[ 1951.087019] gmc_v11_0_process_interrupt: 6 callbacks suppressed
[ 1951.087022] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087025] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087027] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c1000 from client 10
[ 1951.087028] amdgpu 0000:0f:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00541051
[ 1951.087029] amdgpu 0000:0f:00.0: amdgpu: Faulty UTCL2 client ID: TCP (0x8)
[ 1951.087030] amdgpu 0000:0f:00.0: amdgpu: MORE_FAULTS: 0x1
[ 1951.087031] amdgpu 0000:0f:00.0: amdgpu: WALKER_ERROR: 0x0
[ 1951.087031] amdgpu 0000:0f:00.0: amdgpu: PERMISSION_FAULTS: 0x5
[ 1951.087032] amdgpu 0000:0f:00.0: amdgpu: MAPPING_ERROR: 0x0
[ 1951.087032] amdgpu 0000:0f:00.0: amdgpu: RW: 0x1
[ 1951.087036] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087037] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087038] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c3000 from client 10
[ 1951.087042] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087043] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087043] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c0000 from client 10
[ 1951.087047] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087048] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087048] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c3000 from client 10
[ 1951.087052] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087053] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087054] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c1000 from client 10
[ 1951.087057] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087058] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087059] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c3000 from client 10
[ 1951.087062] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087063] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087064] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c0000 from client 10
[ 1951.087067] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087068] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087069] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c0000 from client 10
[ 1951.087072] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087073] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087074] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c1000 from client 10
[ 1951.087077] amdgpu 0000:0f:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:5 pasid:32818)
[ 1951.087078] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1951.087079] amdgpu 0000:0f:00.0: amdgpu: in page starting at address 0x00008001010c3000 from client 10
[ 1961.195212] amdgpu 0000:0f:00.0: amdgpu: Dumping IP State
[ 1961.196514] amdgpu 0000:0f:00.0: amdgpu: Dumping IP State Completed
[ 1961.196547] amdgpu 0000:0f:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
[ 1961.196547] amdgpu 0000:0f:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
[ 1961.196548] amdgpu 0000:0f:00.0: amdgpu: ring gfx_0.0.0 timeout, signaled seq=590042, emitted seq=590045
[ 1961.196549] amdgpu 0000:0f:00.0: amdgpu: Process CrGpuMain pid 20705 thread dxvk-submit pid 20712
[ 1961.196550] amdgpu 0000:0f:00.0: amdgpu: Starting gfx_0.0.0 ring reset
[ 1961.196682] amdgpu 0000:0f:00.0: amdgpu: Ring gfx_0.0.0 reset succeeded
[ 1961.196683] amdgpu 0000:0f:00.0: [drm] device wedged, but recovered through resetI was able to save the GPU coredump, but it only 'exists' for as long as the problematic application remains left open and hung. After closing it and letting the wine server close, that entire '/sys/class/drm/card1/device/devcoredump/data' location just disappears.
Beyond that, there are enough individual packages to check that I am uncertain how long it would even take to narrow it down.
It does seem to specifically involve DXVK. Running a completely fresh wineprefix with no DXVK installed through just base wine without built in DXVK skirts the issue
Last edited by Versa (2026-02-20 05:28:28)
Offline
There seems to be recurring isues with DXVK and amdgpu, you can see this for instance: https://github.com/doitsujin/dxvk/issues/5413
I don't know much more about it.
Offline
There seems to be recurring isues with DXVK and amdgpu, you can see this for instance: https://github.com/doitsujin/dxvk/issues/5413
I don't know much more about it.
That is a shame, but I am also not sure that alone is the issue. I downgraded the Kernel immediately after to test and the issue persisted anyways, and nothing else like Mesa updated along with all that. I almost have a feeling that it could have been glibc or some other dependency further down the line since the AMDgpu driver would have been the older one after the downgrade.
Offline