[SOLVED] GPU Driver keeps restarting when playing games

TheCrafters001 · 2026-06-24 01:00:22

Hello,

First of, I want to apologize if I put this in the wrong forum category. I am still relatively new to Arch specifically, so I am unsure if this problem should go here or not.

I have been using Linux for a few years now, and only fairly recently started using Arch, as the previous distros I have used haven't been able to provide the flexibility I want or need. I have loved my time on Arch. I come here with a specific problem I am having that I am having a hard time describing and time researching so I finally decided to bite the bullet and try posting here.

I play a few games that are a bit graphically intense, and throughout my time on Linux I have had a couple problems with crashing but none as bad as Arch has been giving me. And recently it has been getting a lot worse. Any time I have played Helldivers 2, DOOM Eternal, or anything that makes my graphics card sweat a bit, I can get at least 10 minutes of the game in before my screen freezes, goes black, then either 1) completely hard resets, 2) all my applications close and my window manager (KDE) restarts, or 3) My screen turns back on frozen on the last frame I saw, forcing me to hard reset. I wasn't having that many encounters with it up until starting this past week where it does it more and more frequently. I have even had it happen when I simply launch the game.

When I check the logs, I see something about the mesa driver and amdgpu crashing and resetting but I have been unable to determine the cause. At first, with the help of someone else, it seemed like it might have been Steam Game Recording, so I disabled that and the issue seemed to resolve itself but only temporarily, returning fairly quickly and somehow more frequently. Attached is the most recent journalctl log I have, and any assistance would be greatly helpful. Also below are my specs, should that be helpful.

JournalCTL Log: Due to file size limits, the log can be found here: https://bin.thecrafters001.com/?ab18458 … m2uBmu3UEN

Specs:

                  -`                     thecrafters001@archlinux
                 .o+`                    ------------------------
                `ooo/                    OS: Arch Linux x86_64
               `+oooo:                   Host: MS-7D75 (1.0)
              `+oooooo:                  Kernel: Linux 7.0.12-arch1-1
              -+oooooo+:                 Uptime: 1 hour, 14 mins
            `/:-:++oooo+:                Packages: 18 (flatpak), 1699 (pacman)
           `/++++/+++++++:               Shell: zsh 5.9.1
          `/++++++++++++++:              Display (MNN): 1920x1080 in 16", 60 Hz [External]
         `/+++ooooooooooooo/`            Display (LG ULTRAGEAR): 2560x1440 in 31", 165 Hz [External] *
        ./ooosssso++osssssso+`           DE: KDE Plasma 6.7.1
       .oossssso-````/ossssss+`          WM: KWin (Wayland)
      -osssssso.      :ssssssso.         WM Theme: Se7enAero
     :osssssss/        osssso+++.        Theme: Fusion (BreezeDark) [Qt], Breeze-Dark [GTK2], Breeze [GTK3/4]
    /ossssssss/        +ssssooo/-        Icons: breeze-dark [Qt], breeze-dark [GTK2/3/4]
  `/ossssso+/:-        -:/+osssso+-      Font: Noto Sans (10pt) [Qt], Noto Sans (10pt) [GTK2/3/4]
 `+sso+:-`                 `.-/+oso:     Cursor: Posy_Cursor_Black (32px)
`++:.                           `-/+/    Terminal: konsole 26.4.2
.`                                 `/    CPU: AMD Ryzen 9 7900X (24) @ 5.74 GHz
                                         GPU 1: AMD Radeon RX 7800 XT [Discrete]
                                         GPU 2: AMD Raphael [Integrated]

Last edited by TheCrafters001 (2026-06-27 00:33:12)

killertofus · 2026-06-24 02:56:48

does it crash on x11?

Nikolai5 · 2026-06-24 10:43:21

I can't access your logs, might just be me, but it won't load, if it's not just me can you consider resharing with another pastebin-like service? https://wiki.archlinux.org/title/List_o … n_services

TheCrafters001 · 2026-06-24 11:47:10

killertofus wrote:

does it crash on x11?

I haven't tried it on x11, honestly. I only use KDE on wayland. I don't even know if I have x11 KDE Plasma installed...

Nikolai5 wrote:

I can't access your logs, might just be me, but it won't load, if it's not just me can you consider resharing with another pastebin-like service? https://wiki.archlinux.org/title/List_o … n_services

Here is the log reuploaded elsewhere. I don't know why my self-hosted server has been giving me issues lately. https://paste.c-net.org/TenureFowler

Lone_Wolf · 2026-06-24 14:29:30

It looks like you used some extra parameters for the journalctl command please redo with just -b .

Is x265 installed ?
When was your last full pacman -Syu ?

seth · 2026-06-24 15:35:46

There're two resets

Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: Dumping IP State Completed
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: [drm] AMDGPU device coredump file has been created
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: ring gfx_0.0.0 timeout, signaled seq=2911905, emitted seq=2911907
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 5286 thread vkd3d_queue pid 6179
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: Starting gfx_0.0.0 ring reset
Jun 23 17:54:32 archlinux kernel: [drm:gfx_v11_0_bad_op_irq [amdgpu]] *ERROR* Illegal opcode in command stream
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: MES failed to respond to msg=RESET
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: failed to reset legacy queue
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: reset via MES failed and try pipe reset -110
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: The CPFW hasn't support pipe reset yet.
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: Dumping IP State Completed
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: [drm] AMDGPU device coredump file has been created
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: ring gfx_0.0.0 timeout, signaled seq=24440020, emitted seq=24440022
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 11077 thread vkd3d_queue pid 11241
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: Starting gfx_0.0.0 ring reset
Jun 23 19:21:51 archlinux kernel: [drm:gfx_v11_0_bad_op_irq [amdgpu]] *ERROR* Illegal opcode in command stream
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: MES failed to respond to msg=RESET
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: failed to reset legacy queue
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: reset via MES failed and try pipe reset -110
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: Ring gfx_0.0.0 reset failed

The second one requires

Jun 23 19:21:56 archlinux kernel: amdgpu 0000:03:00.0: [drm] device wedged, but recovered through reset

This is apparently caused by

Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 5286 thread vkd3d_queue pid 6179
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 11077 thread vkd3d_queue pid 11241

https://wiki.archlinux.org/title/Steam/ … _emulation ?

AMD will require /sys/class/drm/card1/device/devcoredump/data (nb this won't survive a reboot, you'll have to copy it into a permanent location after the crash)

TheCrafters001 · 2026-06-24 22:49:28

Lone_Wolf wrote:

It looks like you used some extra parameters for the journalctl command please redo with just -b .
Is x265 installed ?
When was your last full pacman -Syu ?

I redid the command: https://paste.c-net.org/TravestyWisest

x256 is installed, last pacman -Syu was the night before last, but the problem still was occurring before then.

seth wrote:

There're two resets

Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: Dumping IP State Completed
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: [drm] AMDGPU device coredump file has been created
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: ring gfx_0.0.0 timeout, signaled seq=2911905, emitted seq=2911907
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 5286 thread vkd3d_queue pid 6179
Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0: Starting gfx_0.0.0 ring reset
Jun 23 17:54:32 archlinux kernel: [drm:gfx_v11_0_bad_op_irq [amdgpu]] *ERROR* Illegal opcode in command stream
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: MES failed to respond to msg=RESET
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: failed to reset legacy queue
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: reset via MES failed and try pipe reset -110
Jun 23 17:54:34 archlinux kernel: amdgpu 0000:03:00.0: The CPFW hasn't support pipe reset yet.
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: Dumping IP State Completed
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: [drm] AMDGPU device coredump file has been created
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: ring gfx_0.0.0 timeout, signaled seq=24440020, emitted seq=24440022
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 11077 thread vkd3d_queue pid 11241
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0: Starting gfx_0.0.0 ring reset
Jun 23 19:21:51 archlinux kernel: [drm:gfx_v11_0_bad_op_irq [amdgpu]] *ERROR* Illegal opcode in command stream
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: MES failed to respond to msg=RESET
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: failed to reset legacy queue
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: reset via MES failed and try pipe reset -110
Jun 23 19:21:53 archlinux kernel: amdgpu 0000:03:00.0: Ring gfx_0.0.0 reset failed

The second one requires

Jun 23 19:21:56 archlinux kernel: amdgpu 0000:03:00.0: [drm] device wedged, but recovered through reset

This is apparently caused by

Jun 23 17:54:32 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 5286 thread vkd3d_queue pid 6179
Jun 23 19:21:51 archlinux kernel: amdgpu 0000:03:00.0:  Process main pid 11077 thread vkd3d_queue pid 11241

https://wiki.archlinux.org/title/Steam/ … _emulation ?

AMD will require /sys/class/drm/card1/device/devcoredump/data (nb this won't survive a reboot, you'll have to copy it into a permanent location after the crash)

I tried what was recommended in the wiki and the games I am trying to play refuse to launch. I'll try to get a copy of that file the next crash I experience. Is there a recommended way to try and grab it before it gets deleted from a reboot?

seth · 2026-06-25 05:37:31

Do they refuse to launch with an error message?

TheCrafters001 · 2026-06-25 22:15:50

seth wrote:

Do they refuse to launch with an error message?

Something about DirectX 12 not being supported. I would assume this is because adding

PROTON_USE_WINED3D=1 %command%

to the launch arguments on steam causes the game to run in DirectX 11? As far as I know, Helldivers 2 doesn't officially support DirectX 11, in fact I have had issues running the game in DirectX 11 in the first place so I just don't really use it. And I just got Ready or Not which seems to give me that same error about DirectX 12 not being supported.

seth · 2026-06-26 05:58:06

https://wiki.archlinux.org/title/Wine#VKD3D-Proton - DX12 requires vkd3d-proton
https://gitlab.freedesktop.org/search?g … e%20pid%22
Eg https://gitlab.freedesktop.org/drm/amd/ … items/4819 ?

Nikolai5 · 2026-06-26 09:43:21

TheCrafters001 wrote:

seth wrote:
Do they refuse to launch with an error message?
Something about DirectX 12 not being supported. I would assume this is because adding
PROTON_USE_WINED3D=1 %command%
to the launch arguments on steam causes the game to run in DirectX 11? As far as I know, Helldivers 2 doesn't officially support DirectX 11, in fact I have had issues running the game in DirectX 11 in the first place so I just don't really use it. And I just got Ready or Not which seems to give me that same error about DirectX 12 not being supported.

I have the same GPU as you, and have no issues with any games, including Ready Or Not which runs perfectly fine.
I run it in DX12 mode on Proton Experimental without any launch options. (experimental is my default).

All I have done IIRC is install the packages as recommended here: https://wiki.archlinux.org/title/AMDGPU#Installation
and adjusted the max map count as per here: https://wiki.archlinux.org/title/Gaming … _map_count

I've also enabled Resizable BAR or rather ensured that it was enabled in my UEFI (I think it was "auto" for me, older boards might be an off/on toggle).

I run a fairly simple setup with just the recommended packages mentioned in the installation sections of the wiki, and have no issues playing DX9 to DX12, OpenGL or Vulkan games. No custom configuration or kernel parameters, and I don't supply steam with any parameters to run any games, I default to Proton Experimental, that works for 99%, anything else, I'll try other versions.

Given that, I would conclude that the issue is not the GPU and instead your configuration or the packages being used, we need to figure out what is causing these issues.

Last edited by Nikolai5 (2026-06-26 09:44:04)

seth · 2026-06-26 10:58:13

There's an additional

amdgpu 0000:13:00.0: [drm] ATOM BIOS: 102-RAPHAEL-008

which has no outputs attached.
Can you disable that in the BIOS/UEFI?
Since vulkan somewhat notoriously has vblank problems, what outputs are attached?

for OUT in /sys/class/drm/card*; do echo $OUT; edid-decode $OUT/edid; echo "================="; done

You'll need https://archlinux.org/packages/extra/x86_64/v4l-utils/

TheCrafters001 · 2026-06-27 00:33:00

seth wrote:

There's an additional
amdgpu 0000:13:00.0: [drm] ATOM BIOS: 102-RAPHAEL-008
which has no outputs attached.
Can you disable that in the BIOS/UEFI?
Since vulkan somewhat notoriously has vblank problems, what outputs are attached?
for OUT in /sys/class/drm/card*; do echo $OUT; edid-decode $OUT/edid; echo "================="; done
You'll need https://archlinux.org/packages/extra/x86_64/v4l-utils/

You know, I feel so dumb now that I have done some things.

First, the output of the command you have provided me https://paste.c-net.org/MikeyImagined

Next, I have played Ready or Not without issues. I recently tried playing Doom Eternal again and had some freezing that would happen only when I went into combat and did some research. Turns out it was caused by having Async Compute (or whatever it is called in Doom Eternal) turned on. So I turned it off and had no problems.
I tried turning off Async Compute in Helldivers 2 and that seems to have fixed the issue. I apologize if I feel like I wasted time by posting here.

Lone_Wolf · 2026-06-27 13:54:39

TheCrafters001 wrote:

I apologize if I feel like I wasted time by posting here

Not needed, such restarts are often blamed on the driver and its refreshing to see a case where it could b e solved by an application setting.

Arch Linux

#1 2026-06-24 01:00:22

[SOLVED] GPU Driver keeps restarting when playing games

#2 2026-06-24 02:56:48

Re: [SOLVED] GPU Driver keeps restarting when playing games

#3 2026-06-24 10:43:21

Re: [SOLVED] GPU Driver keeps restarting when playing games

#4 2026-06-24 11:47:10

Re: [SOLVED] GPU Driver keeps restarting when playing games

#5 2026-06-24 14:29:30

Re: [SOLVED] GPU Driver keeps restarting when playing games

#6 2026-06-24 15:35:46

Re: [SOLVED] GPU Driver keeps restarting when playing games

#7 2026-06-24 22:49:28

Re: [SOLVED] GPU Driver keeps restarting when playing games

#8 2026-06-25 05:37:31

Re: [SOLVED] GPU Driver keeps restarting when playing games

#9 2026-06-25 22:15:50

Re: [SOLVED] GPU Driver keeps restarting when playing games

#10 2026-06-26 05:58:06

Re: [SOLVED] GPU Driver keeps restarting when playing games

#11 2026-06-26 09:43:21

Re: [SOLVED] GPU Driver keeps restarting when playing games

#12 2026-06-26 10:58:13

Re: [SOLVED] GPU Driver keeps restarting when playing games

#13 2026-06-27 00:33:00

Re: [SOLVED] GPU Driver keeps restarting when playing games

#14 2026-06-27 13:54:39

Re: [SOLVED] GPU Driver keeps restarting when playing games

Board footer