Steam doesn't start with GPU errors since 05/11 update

Diudid commented 1 week ago

Your system information

Steam client version (build number or date): updated 05/11/2024
Distribution (e.g. Ubuntu): Wolf
Opted into Steam client beta?: No
Have you checked for system updates?: Yes
Steam Logs: steam-logs.tar.gz
GPU: [AMD/ATI] Navi 31 [Radeon RX 7900 XT/7900 XTX/7900 GRE/7900M] (rev cc)

Please describe your issue in as much detail as possible:

I was playing yesterday and days before without issue. Since yesterday evening steam update, steam doesn't showup when starting the session. I looked in to the log and it seem that the helper have an issue with my amd GPU [2024-11-06 18:10:50] [1106/181050.645412:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139 [2024-11-06 18:10:50] [1106/181050.807937:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139 [2024-11-06 18:10:50] [1106/181050.961916:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139 [2024-11-06 18:10:51] [1106/181051.116925:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139

Steps for reproducing this issue:

Starting steam : crash

ABeltramo commented 1 week ago

Hi there! I'm the developer of Games on Whales we basically bundle Steam as a docker image so that we can then stream a virtual desktop via Moonlight.
A few users have reported this issue since yesterday, including @Diudid which was kind enough to open up this issue.

I can replicate the same problem when running the stack on my Intel ARC logs.tar.gz

The very same machine and the same stack can run the UI perfectly when using Nvidia with the proprietary drivers instead.

hnipps commented 1 week ago

I'm also experiencing the same issue on my Intel NUC. Here are my logs - wolf.log

inferrna commented 1 week ago

I have very similar error running Windows version under wine. [1107/103635.301:ERROR:viz_main_impl.cc(166)] Exiting GPU process due to errors during initialization

Native version runs fine.

higuita commented 1 week ago

Same, but i'm on amd radeon with mesa:


$ glxinfo | grep -i opengl
OpenGL vendor string: AMD
OpenGL renderer string: AMD Radeon RX 480 Graphics (radeonsi, polaris10, LLVM 19.1.3, DRM 3.57, 6.10.9-slack)
OpenGL core profile version string: 4.6 (Core Profile) Mesa 24.2.6
OpenGL core profile shading language version string: 4.60
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile
OpenGL core profile extensions:
OpenGL version string: 4.6 (Compatibility Profile) Mesa 24.2.6
OpenGL shading language version string: 4.60
OpenGL context flags: (none)
OpenGL profile mask: compatibility profile
OpenGL extensions:
OpenGL ES profile version string: OpenGL ES 3.2 Mesa 24.2.6
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20
OpenGL ES profile extensions:

I did several updates, mesa included, from mesa-24.2.4 to mesa-24.2.6... but i think steam also did some update, but not totally sure. opengl and vulkan are working correctly in games, so firefox and chrome, so if it is a mesa problem, it is just in some less used call

slackware-current, updated 2024/11/08 00h

ppascher commented 1 week ago

Running into the same issue as OP using wolf with amdgpu (6900xt):

cat /etc/wolf/11711553324472631402/Steam/.steam/debian-installation/logs/cef_log.txt
[1108/111117.620641:WARNING:alloy_main_delegate.cc(578)] Alloy bootstrap is deprecated and will be removed in ~M127. See https://github.com/chromiumembedded/cef/issues/3685
[1108/111117.635712:INFO:crash_reporting.cc(238)] Crash reporting enabled for process: browser
[1108/111117.636467:WARNING:task_impl.cc(31)] No task runner for threadId 0
[1108/111117.637069:WARNING:task_impl.cc(31)] No task runner for threadId 0
[1108/111117.682303:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111118.498028:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111118.498045:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 1 time(s)
[1108/111118.499514:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111118.880500:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: utility
[1108/111118.896832:INFO:crash_reporting.cc(238)] Crash reporting enabled for process: utility
[1108/111118.932294:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: renderer
[1108/111119.188331:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111119.188350:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 2 time(s)
[1108/111119.189665:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111119.881667:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111119.881685:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 3 time(s)
[1108/111119.882974:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111120.261946:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111120.261972:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 4 time(s)
[1108/111120.263182:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111120.659143:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111120.659160:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 5 time(s)
[1108/111120.660421:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111121.055653:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111121.055673:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 6 time(s)
[1108/111121.057064:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111121.231839:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111121.231858:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 7 time(s)
[1108/111121.233182:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111121.401456:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111121.401473:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 8 time(s)
[1108/111121.402701:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1108/111121.569469:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1108/111121.569488:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 9 time(s)
[1108/111121.569499:FATAL:gpu_data_manager_impl_private.cc(449)] GPU process isn't usable. Goodbye.

gonnalose commented 1 week ago

Same issue on Ubuntu 24.04.1, up-to-date, AMD Ryzen 7 5800X CPU, AMD Radeon RX 6600 XT GPU

$ glxinfo | grep -i opengl
OpenGL vendor string: AMD
OpenGL renderer string: AMD Radeon RX 6600 XT (radeonsi, navi23, LLVM 17.0.6, DRM 3.58, 6.8.0-48-generic)
OpenGL core profile version string: 4.6 (Core Profile) Mesa 24.0.9-0ubuntu0.2
OpenGL core profile shading language version string: 4.60
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile
OpenGL core profile extensions:
OpenGL version string: 4.6 (Compatibility Profile) Mesa 24.0.9-0ubuntu0.2
OpenGL shading language version string: 4.60
OpenGL context flags: (none)
OpenGL profile mask: compatibility profile
OpenGL extensions:
OpenGL ES profile version string: OpenGL ES 3.2 Mesa 24.0.9-0ubuntu0.2
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20
OpenGL ES profile extensions:

$ tail .steam/logs/cef_log.txt
[1110/150429.063765:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1110/150429.214261:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1110/150429.214280:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 4 time(s)
[1110/150429.216071:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1110/150429.367407:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1110/150429.367435:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 5 time(s)
[1110/150429.369326:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1110/150429.524783:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1110/150429.524808:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 6 time(s)
[1110/150429.524820:FATAL:gpu_data_manager_impl_private.cc(449)] GPU process isn't usable. Goodbye.

SebbeM commented 6 days ago

I'm also affected by this and get the same logs as the rest of you. Is there any way to downgrade and lock the Steam version until this is resolved?

ABeltramo commented 6 days ago

There's an easier way to reproduce the issue without using Wolf+Moonlight, on a Wayland host you can directly run the Steam image with:

docker run --rm -it --device=/dev/dri/renderD128 --device=/dev/dri/card0 --ipc=host --cap-add=ALL --security-opt seccomp=unconfined -e XDG_RUNTIME_DIR=/tmp -v ${XDG_RUNTIME_DIR}/${WAYLAND_DISPLAY}:/tmp/${WAYLAND_DISPLAY}:rw -e XDG_SESSION_TYPE=wayland -e WAYLAND_DISPLAY=${WAYLAND_DISPLAY} -e RUN_SWAY=true -v /tmp/SteamGOWData:/home/retro/ ghcr.io/games-on-whales/steam:edge

(On X11 should work as well by just changing the WAYLAND_ variables)

Resulting in: Screenshot_20241112_201134

You can access the exported home folder under /tmp/SteamGOWData/, which will contain the incriminating:

cat /tmp/SteamGOWData/.steam/debian-installation/logs/cef_log.txt
[1112/114848.161731:WARNING:alloy_main_delegate.cc(578)] Alloy bootstrap is deprecated and will be removed in ~M127. See https://github.com/chromiumembedded/cef/issues/3685
[1112/114848.178415:INFO:crash_reporting.cc(238)] Crash reporting enabled for process: browser
[1112/114848.180572:WARNING:task_impl.cc(31)] No task runner for threadId 0
[1112/114848.181150:WARNING:task_impl.cc(31)] No task runner for threadId 0
[1112/114848.202414:WARNING:discardable_shared_memory_manager.cc(197)] Less than 64MB of free space in temporary directory for shared memory files: 60
[1112/114848.226382:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process
[1112/114848.864053:ERROR:gpu_process_host.cc(1002)] GPU process exited unexpectedly: exit_code=139
[1112/114848.864066:WARNING:gpu_process_host.cc(1436)] The GPU process has crashed 1 time(s)
[1112/114848.865586:INFO:crash_reporting.cc(255)] Crash reporting enabled for process: gpu-process

This has been tested on latest Arch (6.11.6-arch1-1, Plasma 6.2.3) with an Intel Arc A380; if instead I use a Nvidia 3070 (with the Nvidia container toolkit) on the same system everything works correctly.

Hope this helps!

higuita commented 6 days ago

Notice that i'm not using wolf or any docker solution, my setup is a plain desktop starting steam right now i'm unable to start steam and can only use games that do not require steam running

InfectedBytes commented 4 days ago

I'm also using Steam directly on my machine (without docker) and get the same issue (using Radeon RX 6750xt). I also tried to start the game directly with proton, but to no avail. Anyone got any progress on this? a workaround to still run games with proton?

ppascher commented 4 days ago

For me it does not work with the wolf docker image (as mentioned above) but steam works fine starting from my desktop (sway) directly using a 6900xt. This is with the beta client and mesa-git driver.

ValveSoftware / steam-for-linux