Public bug reported:

## Summary
NVIDIA GSP firmware enters a power management reset loop, causing a hard system
freeze. The GPU repeatedly resets itself (~every 30s) while the GSP firmware
keeps firing PFM_REQ_HNDLR_STATE_SYNC_CALLBACK RPCs during its own "bootup"
state, failing the API lock assertion, and resetting again. After ~14 minutes
the display subsystem gives up (drmModeAtomicCommit: Invalid argument) and
the system hard-freezes, requiring a power cycle.

## Environment
- Ubuntu 26.04 LTS (Resolute)
- Kernel: 7.0.0-22-generic
- NVIDIA driver: 595.71.05-0ubuntu0.26.04.1 (open kernel module, 
nvidia-dkms-595-open)
- GPU setup: Hybrid Intel i915 (primary/builtin panel) + NVIDIA (secondary)
- Machine: Lenovo laptop

## Sequence of events (from journalctl -b -1)

Starting at ~10:04, repeating every 30–60 seconds until freeze:

  kernel: nvidia 0000:01:00.0: Enabling HDA controller
  kernel: NVRM: GPU0 _kgspProcessRpcEvent: Attempted to process RPC event
    from GPU0: 0x101a (PFM_REQ_HNDLR_STATE_SYNC_CALLBACK) during bootup
    without API lock
  kernel: NVRM: GPU0 nvAssertFailedNoLog: Assertion failed: 0 @ 
kernel_gsp.c:1447

  [cycle repeats ~every 30-60s]

At 10:04:59:
  kernel: workqueue: output_poll_execute hogged CPU for >10000us 67 times

At 10:17:35 (final display failure):
  gnome-shell: Cursor update failed: drmModeAtomicCommit: Invalid argument

Last log entry at 10:18:08 — hard freeze, required power cycle.

## Root cause
The NVIDIA GSP firmware (GPU System Processor) gets stuck in a power management
state sync loop. Each iteration: fires PFM_REQ_HNDLR_STATE_SYNC_CALLBACK RPC →
fails API lock assertion at kernel_gsp.c:1447 → GPU resets → firmware restarts
→ repeat. The GPU becomes unavailable to the display driver after ~14 minutes
of this loop.

This bug has caused multiple freezes on this machine (also observed on 
2026-05-27
and earlier).

## Workaround
  echo 'options nvidia NVreg_EnableGpuFirmware=0' | sudo tee 
/etc/modprobe.d/nvidia-gsp.conf
  sudo update-initramfs -u
  sudo reboot

Disabling GSP firmware prevents the loop entirely.

## Related
- nvidia-powerd also crashes with SIGSEGV on every boot (separate bug, same 
driver)
- /var/crash/_usr_bin_nvidia-powerd.0.crash present on system

ProblemType: Bug
DistroRelease: Ubuntu 26.04
Package: nvidia-kernel-common-595 595.71.05-0ubuntu0.26.04.1
ProcVersionSignature: Ubuntu 7.0.0-22.22-generic 7.0.0
Uname: Linux 7.0.0-22-generic x86_64
ApportVersion: 2.34.0-0ubuntu2
Architecture: amd64
CasperMD5CheckResult: pass
CurrentDesktop: ubuntu:GNOME
Date: Wed Jun  3 10:41:20 2026
InstallationDate: Installed on 2026-05-05 (29 days ago)
InstallationMedia: Ubuntu 26.04 "Resolute Raccoon" - Release amd64 (20260423.1)
ProcEnviron:
 LANG=en_US.UTF-8
 PATH=(custom, no user)
 SHELL=/bin/bash
 TERM=xterm-256color
 XDG_RUNTIME_DIR=<set>
SourcePackage: nvidia-graphics-drivers-595
UpgradeStatus: No upgrade log present (probably fresh install)

** Affects: nvidia-graphics-drivers-595 (Ubuntu)
     Importance: Undecided
         Status: New


** Tags: amd64 apport-bug resolute wayland-session

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2155159

Title:
  NVIDIA GSP firmware enters a power management reset loop, causing a
  hard system freeze

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-595/+bug/2155159/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to