580.76.05 / 580.82.07 - Suspend & Hibernate Fails on Nvidia / Intel Hybrid Laptop

[sk0rabu]

NVIDIA Open GPU Kernel Modules Version

580.76.05 / 580.82.07

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

Arch Linux

Kernel Release

6.16.4-arch1-1

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

NVIDIA GeForce RTX 5000 Laptop GPU

Describe the bug

Recently decided to do a fresh install (Linux + Wayland) and stick with hybrid mode (dGPU mode has been okay for me), but couldn't get suspend / hibernate right. Suspend = wakes itself up in ~15 seconds. Hibernate = straight fail on reboot.

Log entries pertaining to suspend

Sep 02 23:55:03 ericarch systemd[1]: Finished NVIDIA system suspend actions.
Sep 02 23:55:03 ericarch systemd[1]: nvidia-suspend.service: Consumed 1.015s CPU time, 261.8M memory peak.
Sep 02 23:55:03 ericarch systemd[1]: Starting System Suspend...
Sep 02 23:55:03 ericarch systemd-sleep[1468]: User sess
ions remain unfrozen on explicit request ($SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=0).
Sep 02 23:55:03 ericarch systemd-sleep[1468]: This is not recommended, and might result in unexpected behavior, particularly
Sep 02 23:55:03 ericarch systemd-sleep[1468]: in suspend-then-hibernate operations or setups with encrypted home directories.
Sep 02 23:55:03 ericarch systemd-sleep[1468]: Performing sleep operation 'suspend'...
Sep 02 23:55:03 ericarch kernel: PM: suspend entry (s2idle)
Sep 02 23:55:03 ericarch kernel: Filesystems sync: 0.070 seconds

...

Sep 02 23:55:20 ericarch kernel: Freezing user space processes
Sep 02 23:55:20 ericarch kernel: Freezing user space processes completed (elapsed 0.001 seconds)
Sep 02 23:55:20 ericarch kernel: OOM killer disabled.
Sep 02 23:55:20 ericarch kernel: Freezing remaining freezable tasks
Sep 02 23:55:20 ericarch kernel: Freezing remaining freezable tasks completed (elapsed 0.001 seconds)
Sep 02 23:55:20 ericarch kernel: printk: Suspending console(s) (use no_console_
suspend to debug)
Sep 02 23:55:20 ericarch kernel: ACPI: EC: interrupt blocked
Sep 02 23:55:20 ericarch kernel: ACPI: EC: interrupt unblocked
Sep 02 23:55:20 ericarch kernel: nvidia 0000:02:00.0: Enabling HDA controller
Sep 02 23:55:20 ericarch kernel: pci 0000:00:08.0: Setting to D3hot
Sep 02 23:55:20 ericarch kernel: spd5118 5-0051: Failed to write b = 0: -6
Sep 02 23:55:20 ericarch kernel: spd5118 5-0051: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
Sep 02 23:55:20 ericarch kernel: spd5118 5-0051: PM: failed to resume async: error -6
Sep 02 23:55:20 ericarch kernel: spd5118 5-0053: Failed to write b = 0: -6
Sep 02 23:55:20 ericarch kernel: spd5118 5-0053: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
Sep 02 23:55:20 ericarch kernel: spd5118 5-0053: PM: failed to resume async: error -6

Log entries pertaining to hibernate

Sep 02 23:55:54 ericarch systemd[1]: Finished NVIDIA system hibernate actions.
Sep 02 23:55:54 ericarch systemd[1]: nvidia-hibernate.service: Consumed 998ms CPU time, 262.3M memory peak.
Sep 02 23:55:54 ericarch systemd[1]: Starting System Hibernate...
Sep 02 23:55:54 ericarch systemd-sleep[2964]: User sessions remain unfrozen on explicit request ($SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=0).
Sep 02 23:55:54 ericarch systemd-sleep[2964]: This is not recommended, and might result in unexpected behavior, particularly
Sep 02 23:55:54 ericarch systemd-sleep[2964]: in suspend-then-hibernate operations or setups with encrypted home directories.
Sep 02 23:55:54 ericarch systemd-sleep[2964]: Performing sleep operation 'hibernate'...
Sep 02 23:55:54 ericarch kernel: PM: hibernation: hibernation entry

...

Sep 02 23:56:44 archlinux kernel: NVRM: GPU 0000:02:00.0: PreserveVideoMemoryAllocations module parameter is set. System Power Management attempted without driver procfs suspend interface. Please refer to the 'Configuring Power Management Support' section in the driver README.
Sep 02 23:56:44 archlinux kernel: nvidia 0000:02:00.0: PM: pci_pm_freeze(): nv_pmops_freeze [nvidia] returns -5
Sep 02 23:56:44 archlinux kernel: nvidia 0000:02:00.0: PM: dpm_run_callback(): pci_pm_freeze returns -5
Sep 02 23:56:44 archlinux kernel: nvidia 0000:02:00.0: PM: failed to quiesce async: error -5
Sep 02 23:56:44 archlinux kernel: PM: hibernation: Failed to load image, recovering.
Sep 02 23:56:44 archlinux kernel: PM: hibernation: Basic memory bitmaps freed
Sep 02 23:56:44 archlinux kernel: OOM killer enabled.
Sep 02 23:56:44 archlinux kernel: Restarting tasks: Starting
Sep 02 23:56:44 archlinux kernel: Restarting tasks: Done
Sep 02 23:56:44 archlinux kernel: PM: hibernation: resume failed (-5)

mkinitcpio.conf

MODULES=(btrfs nvidia nvidia_modeset nvidia_uvm nvidia_drm)

bootloader entries

options root=xxx zswap.enabled=0 rootflags=subvol=@ rw rootfstype=btrfs nvidia-drm.modeset=1 nvidia-drm.fbdev=1 resume=xxx

modprobe conf (dunno if it makes any difference)

options nvidia NVreg_PreserveVideoMemoryAllocations=1
options nvidia NVreg_TemporaryFilePath=/var/tmp
options nvidia NVreg_EnableGpuFirmware=0

Thank you very much in advance.

To Reproduce

• Switch laptop to hybrid graphics mode
• Suspend / hibernate

Bug Incidence

Always

nvidia-bug-report.log.gz

nvidia-bug-report.log.gz

More Info

Processor / iGPU: Intel Arrow Lake 285HX

[sk0rabu]

Tried setting NVreg_PreserveVideoMemoryAllocations=0 according to a few posts online.

Suspend still exhibits the same behavior but this time with no video output to monitors. Extra log entries a few seconds after errors from spd5118:

Sep 04 04:49:14 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: Shader Program Header 11 Error
Sep 04 04:49:14 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: Shader Program Header 18 Error
Sep 04 04:49:14 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: ESR 0x405840=0xa2040800
Sep 04 04:49:14 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: ESR 0x405848=0x80000000
Sep 04 04:49:14 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, pid=904, name=Hyprland, Graphics Exception: channel 0x00000003, Class 0000ce97, Offset 00000000, Data 00000000
Sep 04 04:49:14 ericarch kernel: [drm:nv_drm_atomic_apply_modeset_config [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to initialize semaphore for plane fence
Sep 04 04:49:14 ericarch kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to apply atomic modeset.  Error code: -11
Sep 04 04:49:14 ericarch kernel: [drm:nv_drm_atomic_apply_modeset_config [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to initialize semaphore for plane fence
Sep 04 04:49:14 ericarch kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to apply atomic modeset.  Error code: -11

...

Sep 04 04:51:13 ericarch kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Flip event timeout on head 0

Sep 04 04:51:16 ericarch kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Flip event timeout on head 1

Had to switch to a different TTY.

OTOH hibernate now works (Edit: some of the time):

Below is what the journal looks like after a successful wakeup -

# error messages similar to ones from suspend, on resume; but no more error -5s

Sep 04 05:01:39 ericarch kernel: spd5118 5-0051: Failed to write b = 0: -6
Sep 04 05:01:39 ericarch kernel: spd5118 5-0051: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
Sep 04 05:01:39 ericarch kernel: spd5118 5-0051: PM: failed to restore async: error -6
Sep 04 05:01:39 ericarch kernel: spd5118 5-0053: Failed to write b = 0: -6
Sep 04 05:01:39 ericarch kernel: spd5118 5-0053: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
Sep 04 05:01:39 ericarch kernel: spd5118 5-0053: PM: failed to restore async: error -6

# ... with following warnings
Sep 04 05:01:41 ericarch kernel: NVRM: GPU Board Serial Number: 0
Sep 04 05:01:41 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: Shader Program Header 11 Error
Sep 04 05:01:41 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: Shader Program Header 18 Error
Sep 04 05:01:41 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: ESR 0x405840=0xa2040800
Sep 04 05:01:41 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, Graphics Exception: ESR 0x405848=0x80000000
Sep 04 05:01:41 ericarch kernel: NVRM: Xid (PCI:0000:02:00): 13, pid=897, name=Hyprland, Graphics Exception: channel 0x00000003, Class 0000ce97, Offset 00000000, Data 00000000

Below's another log dump for failed wakeup from hibernate

Sep 06 11:16:21 archlinux kernel: ACPI Warning: Time parameter 250 us > 100 us violating ACPI spec, please fix the firmware. (20250404/exsystem-142)
Sep 06 11:16:21 archlinux kernel: nvidia 0000:02:00.0: Enabling HDA controller
Sep 06 11:16:21 archlinux kernel: nvidia 0000:02:00.0: PM: pci_pm_freeze(): nv_pmops_freeze [nvidia] returns -5
Sep 06 11:16:21 archlinux kernel: nvidia 0000:02:00.0: PM: dpm_run_callback(): pci_pm_freeze returns -5
Sep 06 11:16:21 archlinux kernel: nvidia 0000:02:00.0: PM: failed to quiesce async: error -5
# new
Sep 06 11:16:21 archlinux kernel: PM: hibernation: Failed to load image, recovering.
Sep 06 11:16:21 archlinux systemd-modules-load[257]: Failed to insert module 'nvidia_modeset': Interrupted system call
Sep 06 11:16:21 archlinux kernel: NVRM: testIfDsmSubFunctionEnabled: GPS ACPI DSM called before _acpiDsmSupportedFuncCacheInit subfunction = 1

Sep 06 11:16:24 archlinux kernel: NVRM: testIfDsmSubFunctionEnabled: GPS ACPI DSM called before _acpiDsmSupportedFuncCacheInit subfunction = 11.
Sep 06 11:16:24 archlinux kernel: [drm] Initialized nvidia-drm 0.0.0 for 0000:02:00.0 on minor 1
Sep 06 11:16:24 archlinux kernel: Console: switching to colour dummy device 80x25
Sep 06 11:16:24 archlinux kernel: Console: switching to colour frame buffer device 240x67
Sep 06 11:16:24 archlinux kernel: nvidia 0000:02:00.0: [drm] fb0: nvidia-drmdrmfb frame buffer device
Sep 06 11:16:24 archlinux systemd-modules-load[257]: Inserted module 'nvidia_drm'
Sep 06 11:16:24 archlinux systemd[1]: systemd-modules-load.service: Main process exited, code=exited, status=1/FAILURE
Sep 06 11:16:24 archlinux systemd[1]: systemd-modules-load.service: Failed with result 'exit-code'.
Sep 06 11:16:24 archlinux systemd[1]: Failed to start Load Kernel Modules.

...

Sep 06 11:16:25 ericarch kernel: pcieport 0000:00:06.0: AER: Multiple Uncorrectable (Non-Fatal) error message received from 0000:02:00.0
Sep 06 11:16:25 ericarch kernel: nvidia 0000:02:00.0: PCIe Bus Error: severity=Uncorrectable (Non-Fatal), type=Transaction Layer, (Requester ID)
Sep 06 11:16:25 ericarch kernel: nvidia 0000:02:00.0:   device [10de:2c18] error status/mask=00100000/00400000
Sep 06 11:16:25 ericarch kernel: nvidia 0000:02:00.0:    [20] UnsupReq               (First)
Sep 06 11:16:25 ericarch kernel: nvidia 0000:02:00.0: AER:   TLP Header: 0x40000001 0x0000000c 0x8408000c 0x00000000
Sep 06 11:16:26 ericarch kernel: nvidia 0000:02:00.0: AER: can't recover (no error_detected callback)
Sep 06 11:16:26 ericarch kernel: snd_hda_intel 0000:02:00.1: AER: can't recover (no error_detected callback)
Sep 06 11:16:26 ericarch kernel: pcieport 0000:00:06.0: AER: device recovery failed

(Repeat)

I see the NVreg_PreserveVideoMemoryAllocations=1 has been made default as per /usr/lib/modprobe.d/nvidia-sleep.conf, although it seems glitchy for now?

Some other posts suggest just getting rid of everything nvidia from MODULES in the mkinitcpio.conf; will give it a try when I got time.

[sk0rabu]

Issue persists on 580.82.07 .

Someone mentioned setting fbdev to 0 in response to the User sessions remain unfrozen on explicit request ($SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=0) log entry, tried that; suspend still didn't work and it broke things (after sign-in, laptop screen via eDP was no longer detected but remained on, thus unable to disable / power off)