basecamp
latest version of chromium crashes the machine when hardware acceleration is enabled
System details
Thinkpad Z13, CPU AMD Ryzen 7 PRO 6850U, GPU AMD Radeon 680M, Omarchy 3.1.4, kernel 6.17.5-arch1-1 Chromium 141.0.7390.122
What's wrong?
This is probably an upstream bug, but I still think it might be useful to track it here, feel free to close if you think otherwise.
With one of the latest updates executed via the Update/Omarchy menu, everytime I open a Google Meet Chromium will crash and bring the whole machine to a halt, the only option to recover from this is to perform a hard reset by holding the power button for a few seconds and rebooting again.
If i run chromium with --disable-gpu this does not happen.
If I install and run Google Chrome for the AUR, this does not happen.
I have attached the report created by chrome://gpu when hardware accelaration is enabled and disabled.
about-gpu-enabled-2025-10-29T11-57-32-857Z.txt about-gpu-disabled-2025-10-29T11-55-47-173Z.txt
I am able to reproduce this bug with a fresh install on two different computers, both with AMD graphics (both are a bit dated):
- 2011 21" iMac with AMD Radeon HD 6750M (causes a hard crash of the computer after a few seconds)
- HP ProDesk 600 G3 SFF with i5-7500 and AMD R5 340X (low profile card installed). Everything runs fine when the AMD card is removed and running on the integrated Intel i5-7500 graphics.
This seems to not only affect Chromium, but all preinstalled web apps such as Figma.
It looks like this is probably a duplicate of:
https://github.com/basecamp/omarchy/issues/1813
As mentioned in that report, "--use-gl=desktop" option to ~/.config/chromium-flags.conf and any Electron apps and other browsers also seems to be a work-around for the issue.
Using this flag on ~/.config/chromium-flags.conf solved this problem for me, while being able to keep hardware acceleration on after multiple system crashes:
--use-angle=vulkan
GPU: AMD RX Radeon 6750 XT
This happens for me too, randomly, when using Google Meet (Chromium 142). Sometimes after 5min, sometimes after 30min or so: video freezes, screen goes black and the only way to recover is a hard-reset.
CPU: AMD Ryzen 7 8745HS (16) @ 4.97 GHz GPU: AMD Radeon 780M Graphics [Integrated]
I ran various CPU/GPU stress tests using stress-ng and ffmpeg and couldn't reproduce the issue. Looks Chromium-specific.
So far, I've added --use-gl=desktop to chromium-flags.conf, but can't yet tell it helped.
Here's what Claude says:
November 19, 10:20 AM - Critical GPU Crash
This was a serious AMD GPU hardware failure that cascaded into system crashes:
10:20:49 AM - GPU Ring Timeout:
- AMD GPU graphics ring (gfx_0.0.0) timed out
- Chromium (pid 252864) was using the GPU when it hung
- GPU experienced multiple page faults with PERMISSION_FAULTS
- GPU generated a coredump (saved to /sys/class/drm/card1/device/devcoredump/data)
- The GPU driver initiated a ring reset to recover
10:20:54 AM - Hyprland crashed (SIGABRT) as a result of the GPU failure
- Same stack trace as Nov 27: crash in CHyprOpenGLImpl::begin
- All GPU threads blocked in libgallium (AMD Mesa driver)
10:20:55 AM - xdg-desktop-portal-hyprland segfaulted
This was a GPU hardware hang caused by Chromium, which then crashed your entire Wayland compositor. The GPU ring timeout is a serious issue indicating the GPU stopped responding.
Here are the complete logs from November 19, 10:20-10:21 AM showing the GPU crash and Hyprland failure:
GPU Crash Sequence (10:20:49)
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Dumping IP State
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Dumping IP State Completed
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: ring gfx_0.0.0 timeout, signaled seq=2059389, emitted seq=2059391
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Process chromium pid 252864 thread chromium:cs0 pid 252867
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Starting gfx_0.0.0 ring reset
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:88 vmid:5 pasid:32782)
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Process chromium pid 252864 thread chromium:cs0 pid 252867
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: in page starting at address 0x000080da1ffcf000 from client 10
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x005012B1
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Faulty UTCL2 client ID: SQC (inst) (0x9)
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: MORE_FAULTS: 0x1
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: WALKER_ERROR: 0x0
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: PERMISSION_FAULTS: 0xb
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: MAPPING_ERROR: 0x0
Nov 19 10:20:49 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: RW: 0x0
[Multiple repeated page faults - Chromium causing GPU memory access violations]
GPU Reset Failure (10:20:51)
Nov 19 10:20:51 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: MES failed to respond to msg=RESET
Nov 19 10:20:51 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: failed to reset legacy queue
Nov 19 10:20:51 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: reset via MES failed and try pipe reset -110
Nov 19 10:20:51 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: Ring gfx_0.0.0 reset failed
Nov 19 10:20:51 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: GPU reset begin!
Full GPU Reset (10:20:54)
Nov 19 10:20:54 archlinux kernel: [drm:gfx_v11_0_hw_fini [amdgpu]] *ERROR* failed to halt cp gfx
Nov 19 10:20:54 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: MODE2 reset
Nov 19 10:20:54 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: GPU reset succeeded, trying to resume
Nov 19 10:20:54 archlinux kernel: [drm] PCIE GART of 512M enabled (table at 0x00000080FFD00000).
Nov 19 10:20:54 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: SMU is resuming...
Nov 19 10:20:54 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: SMU is resumed successfully!
Nov 19 10:20:54 archlinux kernel: amdgpu 0000:65:00.0: amdgpu: [drm] DMUB hardware initialized: version=0x08005400
Hyprland Crash
Nov 19 10:20:54 archlinux uwsm_hyprland.desktop[3589]: Hyprland has crashed :(
Consult the crash report at /home/pch/.cache/hyprland/hyprlandCrashReport3589.txt for more information.
Nov 19 10:20:55 archlinux kernel: xdg-desktop-por[3948]: segfault at 6c ip 00007fd7da596c01
sp 00007ffe047292a0 error 6 in libwayland-client.so.0.24.0
Key Details:
- Chromium triggered GPU page faults with permission violations
- GPU ring timed out and could not be soft-reset
- Required full MODE2 GPU reset (hardware reset)
- Hyprland crashed during the GPU reset
- 26+ threads in Hyprland all blocked waiting on AMD GPU driver (libgallium)
This is a severe GPU hang requiring hardware-level reset.
What Happened
10:20:49 - GPU Hardware Hang
- Chromium caused GPU page faults and ring timeout
- GPU driver attempted soft reset → FAILED
- GPU driver performed full MODE2 hardware reset
10:20:54 - Hyprland Detected GPU Reset
- Hyprland's OpenGL context became invalid during GPU reset
- OpenGL returned GL_INNOCENT_CONTEXT_RESET status
- Hyprland intentionally aborted because it doesn't have proper GPU reset recovery
- 26+ rendering threads all blocked waiting on AMD driver (libgallium)
