Waking up from sleep often results in blank screen, forceful shutdown

[9a3eedi]

I recently built a new PC with the following specs:

• AMD 7800X3D
• Asus ROG STRIX B650E-I Gaming Wifi motherboard
• Nvidia GTX 1080Ti (recycled from my old PC)

I’m running Manjaro Linux with the 6.11.2 kernel, although I’ve recently downgraded to 6.10.13 in an effort to solve this issue, but no dice. I am using the proprietary Nvidia drivers.

In my previous PC, which was a 7th gen Intel CPU built in 2017, sleep would work perfectly well. I was using the 5.15 or the 5.18 kernel (not sure) in the old PC). In this new PC, I often wake up to a blank screen.

Looking at the system log with journalctl, I notice these messages

Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: The canary thread is apparently starving. Taking action.
Nov 13 19:00:40 9a3eedi-linux kernel: PM: hibernation: hibernation exit
Nov 13 19:00:40 9a3eedi-linux NetworkManager[714]: <info>  [1731510040.5156] dhcp4 (eno1): activation: beginning transaction (timeout in 45 seconds)
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Demoting known real-time threads.
Nov 13 19:00:40 9a3eedi-linux NetworkManager[714]: <info>  [1731510040.5157] dhcp4 (eno1): state changed no lease
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Successfully demoted thread 1150 of process 1131.
Nov 13 19:00:40 9a3eedi-linux dbus-broker[713]: A security policy denied :1.9 to send method call /midi/profile:org.bluez.GattProfile1.Release to :1.55.
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Successfully demoted thread 1131 of process 1131.
Nov 13 19:00:40 9a3eedi-linux systemd-sleep[16863]: System returned from sleep operation 'hybrid-sleep'.
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Successfully demoted thread 1141 of process 1130.
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Successfully demoted thread 1130 of process 1130.
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Successfully demoted thread 1142 of process 1132.
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Successfully demoted thread 1132 of process 1132.
Nov 13 19:00:40 9a3eedi-linux rtkit-daemon[1043]: Demoted 6 threads.
Nov 13 19:00:40 9a3eedi-linux bluetoothd[795]: Battery Provider Manager destroyed
Nov 13 19:00:40 9a3eedi-linux systemd[1]: Starting Load/Save RF Kill Switch Status...
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: ERROR: GPU:0: Failed detecting connected display devices
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: ERROR: GPU:0: Failure reading maximum pixel clock value for display device HDMI-0.
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: ERROR: GPU:0: Failed detecting connected display devices
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: ERROR: GPU:0: Failed detecting connected display devices
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: WARNING: GPU:0: Failure processing EDID for display device Ancor Communications Inc ROG PG278Q (DP-2).
Nov 13 19:00:40 9a3eedi-linux systemd[1]: Starting Rotate log files...
Nov 13 19:00:40 9a3eedi-linux systemd[1]: Started Verify integrity of password and group files.
Nov 13 19:00:40 9a3eedi-linux systemd[1]: systemd-hybrid-sleep.service: Deactivated successfully.
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: WARNING: GPU:0: Unable to read EDID for display device Ancor Communications Inc ROG PG278Q (DP-2)
Nov 13 19:00:40 9a3eedi-linux kernel: nvidia-modeset: ERROR: GPU:0: Failure reading maximum pixel clock value for display device DP-2.
Nov 13 19:00:40 9a3eedi-linux kernel: BUG: kernel NULL pointer dereference, address: 000000000000005c
Nov 13 19:00:40 9a3eedi-linux kernel: #PF: supervisor read access in kernel mode
Nov 13 19:00:40 9a3eedi-linux kernel: #PF: error_code(0x0000) - not-present page

Apart from the the NULL pointer dereference, notice the failure in detecting display devices. I think this might be the main reason behind the blank screen. However this wasn’t happening with the old PC. I also have 2 different displays plugged into the same monitor, but it’s not able to detect either of them when it wakes up from sleep.
Has anyone encountered this issue as well?

I realize the nvidia drivers are proprietary but I was wondering if there’s a solution or something else I can try.

[brahma]

maybe this will help, from the stable update annoucement:

@philm

System freezes for 60 seconds and then wakes back up or hangs after waking up

Since systemd v256, systemd freezes user.slice before sleeping. This process can fail due to kernel bugs, particularly when KVM is in use.[13] [14]

Messages in the logs will contain Failed to freeze unit 'user.slice' before sleep. When such an issue occurs, trying to login (start another session) would fail with pam_systemd(process:session): Failed to create session: Job 9876 for unit 'session-6.scope' failed with 'frozen'.

To temporarily revert back to the old behavior, edit systemd-suspend.service, systemd-hibernate.service, systemd-hybrid-sleep.service, and systemd-suspend-then-hibernate.service with the following drop-in:

[Service]
Environment="SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=false"

– Power management/Suspend and hibernate - ArchWiki

[9a3eedi]

I remember trying this solution, but it didn’t work. Also, I wasn’t getting that “Failed to freeze unit” message, rather I seem to be getting other errors

Regardless though, I decided to try it again, lets see how that goes.

EDIT: it worked! Haven’t had a single sleep issue for a week. not sure why it didn’t work the first time. I’m running 6.10.13-3 kernel right now.

[brahma]

yes, give it a try…
and kernel 6.10, so if you still using it, switch to another one…
if it doesnt work check:
https://wiki.archlinux.org/title/NVIDIA/Tips_and_tricks#Preserve_video_memory_after_suspend