Same signature here on Ubuntu 26.04 LTS: `[gfxhub] page fault`, `Faulty
UTCL2 client ID: SQC (data) (0xa)`, `PERMISSION_FAULTS: 0x3`, ring 24,
followed by a gfx ring timeout. Hardware: Ryzen 7 7700X iGPU
(Raphael/gfx1036, RDNA2) on a Gigabyte B650M DS3H (same B650 platform as
the original report).

Data points that may help triage:

1. **Reproducible in seconds with a public URL**: opening 
https://pantograph.com/journal/pan-1 in Google Chrome (32 `<video>` elements, 
several autoplaying) triggers the fault within seconds, reliably. Faulting 
process is always `chrome`.
2. **Reproduced on Mesa 26.0.3 (archive) AND 26.1.4 (kisak PPA)**, kernels 
7.0.0-27 and 7.0.0-28 — consistent with this not being fixed in the 26.1.x 
series either.
3. 3 of 4 faults captured on 2026-07-16 were at address `0x000000003f800000` 
(the bit pattern of float 1.0), suggesting a shader dereferencing a float 
constant as a pointer.
4. Disabling Chrome's hardware video decode does not prevent it. BIOS update 
(FB3j → FC3a latest) did not help either.
5. Whether the session survives depends on which ring the fault lands on: 
`gfx_0.0.0` soft-resets fine; `gfx_0.1.0` soft reset fails → MODE2 reset → 
"VRAM is lost" → compositor context lost → forced logout.

Filed upstream with full logs:
https://gitlab.freedesktop.org/mesa/mesa/-/work_items/15890

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2125697

Title:
  amdgpu crashes with page fault

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/2125697/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to