Sporadic but common Xid 119 on most driver versions

NVIDIA Open GPU Kernel Modules Version

580.126.18

Please confirm this issue does not happen with the proprietary driver (of the same version). This issue tracker is only for bugs specific to the open kernel driver.

• I confirm that this does not happen with the proprietary driver package.

Operating System and Version

Fedora Linux 42 (KDE Plasma Desktop Edition)

Kernel Release

Linux RZS-OLYSK 6.18.13-100.fc42.x86_64 #1 SMP PREEMPT_DYNAMIC Thu Feb 19 20:58:08 UTC 2026 x86_64 GNU/Linux

Please confirm you are running a stable release kernel (e.g. not a -rc). We do not accept bug reports for unreleased kernels.

• I am running on a stable kernel release.

Hardware: GPU

GPU 0: NVIDIA GeForce RTX 3060 Laptop GPU (UUID: GPU-0d85ac9d-a4c9-70b6-ad1e-c86ba7bda742)

Describe the bug

After some amount of time, the dedicated GPU (RTX 3060 Laptop) on my laptop experiences an Xid 119 and ceases working. When this happens, the GPU becomes unuseable and nvidia-smi either reports ERR! for most fields or just hangs.

The system overall remains useable as the DE runs on the iGPU, but apps that try to use the dGPU either fail or hang. Attempting to suspend the system in this state causes it to lock up entirely, requiring it to be powered off via holding the power button (as even SysRq + O fails; previous versions/kernels reported soft CPU lockups).

A full system reset is required to restore functionality.

To Reproduce

This bug occurs often, usually one or more times per day during casual use. Unfortunately, I have not been able to determine the exact conditions required to reproduce it; in some cases it doesn't occur across multiple days and several suspend-resume cycles, and other times it's occurred within two minutes after a (re)boot.

I have tried reproducing the problem by repeatedly sleeping and waking the GPU (by calling nvidia-smi every time it enters a D3cold state), but doing that didn't appear to make crash occur any sooner.

More importantly, I have been observing this (or a similar) crash for most driver versions. So far, the only driver version that I haven't had any incidents on was 580.82.x. The problem was also much less frequent on 580.95.05. However, it's occurred consistently both before 580.82.07 (down to at least 565.77) and after 580.95.05 (up to at least 580.126.18).

I also recall that upgrading to 580.105.08 from 580.95.05 decreased performance in some games by ~10 FPS, but I don't have any data besides vague memory to back that up and it may not be related.

Bug Incidence

Sometimes

nvidia-bug-report.log.gz

Shortly after the GPU died: nvidia-bug-report-immediate.log.gz. This log actually also captures an earlier instance of the problem that occurred a few days prior.

After a reboot: nvidia-bug-report-after-reboot.log.gz

More Info

Not sure if it's related as my DE runs on the iGPU, but I'm using Wayland.

I've looked at the other reported issues mentioning Xid 119 and did not find any that contained similar log lines (e.g. nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY]) so I'm reporting this on the chance that this is a separate issue.

Please let me know if I can provide any more information or test anything.

Start with the attached immediate and post-reboot nvidia-bug-report logs, focusing on the Xid 119 and nvCheckOkFailedNoLog entries. Compare the reported behavior across the listed driver versions and suspend-resume attempts. Done would require identifying a reproducible trigger or confirmed cause and verifying that the GPU no longer becomes unusable.