Public bug reported:

Ubuntu 26.04 LTS (Resolute Raccoon), kernel 7.0.0-30-generic (Ubuntu
7.0.0-30.30-generic / upstream 7.0.12) — running as Linux Mint 22.3.

System information:
- Distro: Linux Mint 22.3, based on Ubuntu 26.04 (Resolute Raccoon)
- Host: ASUS TUF Dash F15 FX517ZE (Optimus hybrid graphics)
- CPU: Intel 12th Gen i7-12650H
- GPU: NVIDIA GeForce RTX 3050 Ti Mobile + Intel Alder Lake-P UHD Graphics
- NVIDIA driver: 595.84 (Open Kernel Module)
- PRIME mode: on-demand
- Affected kernel: 7.0.0-30-generic (working: 7.0.0-28-generic, same driver)

Description:

Hibernation worked reliably until approximately three weeks ago. After a
kernel upgrade from 7.0.0-28-generic to 7.0.0-30-generic, hibernation
resume consistently fails.

The hibernation image is written successfully. On the next boot, the
kernel correctly detects and loads the saved image (100% progress,
"Image successfully loaded"), but immediately afterward the NVIDIA
driver fails to freeze the PCI device during the atomic restore step:

nvidia 0000:01:00.0: PM: pci_pm_freeze(): nv_pmops_freeze [nvidia] returns -5
nvidia 0000:01:00.0: PM: dpm_run_callback(): pci_pm_freeze returns -5
nvidia 0000:01:00.0: PM: failed to quiesce async: error -5
PM: hibernation: Failed to load image, recovering.
PM: hibernation: resume failed (-5)

The kernel then aborts the restore and boots fresh, discarding the saved
session entirely (all open applications are lost).

Steps to reproduce:
1. Open several applications (tested with ParaView, browser, terminal)
2. sudo systemctl hibernate
3. Power the machine back on
4. Observe: fresh boot instead of restored session; dmesg shows the error above

Regression testing performed:
- Kernel 7.0.0-28-generic + driver 595.84 -> hibernation/resume works 
perfectly, all windows restored
- Kernel 7.0.0-30-generic + driver 595.84 -> fails every time with the error 
above

Additional tests performed (none resolved the issue on kernel 30):
- NVreg_PreserveVideoMemoryAllocations=1 already set by default (via 
/lib/modprobe.d/nvidia-runtimepm.conf) -- no effect
- Disabling GPU firmware (NVreg_EnableGpuFirmware=0) -- no effect, identical 
error
- nvidia-hibernate.service, nvidia-resume.service, nvidia-suspend.service all 
confirmed enabled

Related bug (different symptom, same kernel/driver pair):
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2164816 -- describes a 
hibernation write/poweroff failure on the same kernel 7.0.0-30-generic with a 
newer NVIDIA GPU (RTX 5060 Ti). My issue occurs during resume/restore instead, 
which may indicate a broader PM regression affecting the NVIDIA driver's PCI 
freeze/thaw callbacks introduced in this kernel version.

Full dmesg output attached showing the failure at the "nv_pmops_freeze"
line.

** Affects: linux (Ubuntu)
     Importance: Undecided
         Status: New


** Tags: hibernate nvidia optimus regression resume

** Attachment added: "dmesg hibernate bug"
   
https://bugs.launchpad.net/bugs/2165927/+attachment/5996600/+files/dmesg-hibernate-bug.txt

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2165927

Title:
  Hibernation resume fails with nv_pmops_freeze returns -5 on kernel
  7.0.0-30-generic (regression from 7.0.0-28-generic)

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2165927/+subscriptions


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

Reply via email to