"Failed to apply atomic modeset" + display hang on fullscreen game launch with 530.41.03, 1070 GTX, Wayland, Gnome 43.4, Manjaro

Graphics / Linux Linux
22

I am now running the 535.86.05 driver and got this hang yesterday running https://theforceengine.github.io fullscreen. Disabling direct scanout still works as a workaround.

23

the most consistent results i get is when i launch a unity game via lutris/wine. it crashes everytime without fail.

24

Unfortunately I still get this crash with the new 535.98 drivers.

This time, the errors I see in the systemd log are:

Aug 12 09:57:39 kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to apply atomic modeset. Error code: -22
Aug 12 09:57:47 kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Flip event timeout on head 0

Actually looking at them now, I’m pretty sure they’re the exact some ones I saw at the top of this thread.

Satisfactory and GZDoom both launch without triggering this hang, though, so that’s weird. I guess one difference is TheForceEngine is an OpenGL application whereas the other two are Vulkan?

25

Update: a locally compiled build of The Force Engine still crashes for me with the 535.104.05 drivers.

26

Issue has been root caused and fix is integrated in future release drivers.

27

Is it in the 545 drivers that just released, or a release yet to come?

28

Fix is available in 545.29.02 driver which is released few hours back.
Please help to test and share feedback.

29

Unfortunately I still see this issue on the latest 545.29.02 driver, on an Ampere GPU (RTX 3070 Ti).

Relevant dmesg logs:

[Tue Nov  7 13:01:23 2023] [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to apply atomic modeset.  Error code: -22
[Tue Nov  7 13:01:51 2023] [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Flip event timeout on head 0
[Tue Nov  7 13:01:54 2023] [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Flip event timeout on head 0
30

I can confirm this on x11 as well
Lenovo P1 gen6 with RTX Ada 2000

31

same having this error as well. with “[ 12.960303] [drm:nv_drm_atomic_commit [nvidia_drm]] ERROR [nvidia-drm] [GPU ID 0x00000a00] Flip event timeout on head 0” (somtimes happening in flip 1 or 2 instead of 0) in driver 545, although I am unsure if it’s for the same reason. Can’t figure why it happens but it’s make the system almost unbootable.

Please , this is getting really annoying.

32

I was able to resolve the [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000800] Failed to apply atomic modeset. Error code: -22 issue by adding the kernel parameter initcall_blacklist=simpledrm_platform_driver_init. I only see [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000800] Flip event timeout on head 0 when doing a reboot. I don’t see this error on a cold boot. My screen is blank until the desktop shows up during a reboot, which is the difference in behavior for me compared to a cold boot and maybe explains the error, but not sure.

1 Like