debcrafters-packages team mailing list archive

Thread
Date
[Bug 2112311] Re: Radeon 780M crashes on wake from suspend

To: debcrafters-packages@xxxxxxxxxxxxxxxxxxx
From: Timo Aaltonen <2112311@xxxxxxxxxxxxxxxxxx>
Date: Tue, 10 Jun 2025 17:50:31 -0000
Reply-to: Bug 2112311 <2112311@xxxxxxxxxxxxxxxxxx>
Sender: noreply@xxxxxxxxxxxxx
this is actually fixed in questing, which has 25.0.7

** Changed in: mesa (Ubuntu)
       Status: New => Fix Released

-- 
You received this bug notification because you are a member of
Debcrafters packages, which is subscribed to mesa in Ubuntu.
https://bugs.launchpad.net/bugs/2112311

Title:
  Radeon 780M crashes on wake from suspend

Status in Mesa:
  Fix Released
Status in mesa package in Ubuntu:
  Fix Released

Bug description:
  When I wake up from sleep (s2idle or hibernate), occasionally all
  graphics freeze, either leading to the Wayland compositor crashing or
  just being unresponsive until rebooted.

  Excerpt from crash log:

  ```
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:8 vmid:5 pasid:32772)
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:  in process kwin_wayland pid 520216 thread kwin_wayla:cs0 pid 520248)
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:   in page starting at address 0x0000800000000000 from client 10
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00541010
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          Faulty UTCL2 client ID: TCP (0x8)
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          MORE_FAULTS: 0x0
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          WALKER_ERROR: 0x0
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          PERMISSION_FAULTS: 0x1
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          MAPPING_ERROR: 0x0
  Jun 02 10:43:37 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          RW: 0x1
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: Dumping IP State
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: Dumping IP State Completed
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:8 vmid:5 pasid:32772)
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:  in process kwin_wayland pid 520216 thread kwin_wayla:cs0 pid 520248)
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:   in page starting at address 0x0000800000000000 from client 10
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00501E11
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          Faulty UTCL2 client ID: GCR (0xf)
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          MORE_FAULTS: 0x1
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          WALKER_ERROR: 0x0
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          PERMISSION_FAULTS: 0x1
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          MAPPING_ERROR: 0x0
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:          RW: 0x0
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: ring gfx_0.0.0 timeout, but soft recovered
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:8 vmid:5 pasid:32772)
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:  in process kwin_wayland pid 520216 thread kwin_wayla:cs0 pid 520248)
  Jun 02 10:43:48 nessaesq kernel: amdgpu 0000:65:00.0: amdgpu:   in page starting at address 0x0000800000000000 from client 10
  ```

  system information:
  Kubuntu 25.04
  mesa 25.0.3-1ubuntu2
  kernel 6.14.0-15.15-generic 6.14.0
  CPU/GPU: AMD Ryzen 7 8845HS w/ Radeon 780M Graphics
  Laptop: TongFang GX4HRXL

  I am aware of https://gitlab.freedesktop.org/drm/amd/-/issues/3388 and
  bug #2103854 and setting dcdebugmask=0x10 has not helped me this time
  - it was already set when I encountered this issue.

  My suspicion is that it's
  https://gitlab.freedesktop.org/mesa/mesa/-/issues/12809, fixable with
  disabling VM_ALWAYS_VALID - see patches attached to upstream issue.
  Manually building and installing Mesa 25.0.5 from Debian Unstable has
  successfully fixed this issue for me.

To manage notifications about this bug go to:
https://bugs.launchpad.net/mesa/+bug/2112311/+subscriptions