Wake from suspend not working

binaryanomaly · 2025-08-03 19:53:52

Hey there,

It's been a while since I last used arch but it feels already like back in the old days ;-)

Unfortunately my machine can not properly restore from suspend again. The machine wakes up and then the fans go up to full speed but otherwise no reaction it requires a cold boot to recover. I can't find any clues as all I see with journalctl -k -b -1 is:

[Aug 03 21:47:01 arch-desk systemd-journald[498]: Successfully sent stream file descriptor to service manager.
Aug 03 21:47:01 arch-desk systemd-journald[498]: Successfully sent stream file descriptor to service manager.
Aug 03 21:47:01 arch-desk systemd-journald[498]: Successfully sent stream file descriptor to service manager.
Aug 03 21:47:01 arch-desk kernel: PM: suspend entry (deep)
Aug 03 21:47:01 arch-desk kernel: Filesystems sync: 0.047 seconds

as the last entries. No entries from wakeup.

My hardware:
- Intel(R) Core(TM) Ultra 9 285K
- ASUS TUF GAMING Z890-PLUS WIFI
- ASUS GeForce RTX™ 4080 16GB GDDR6X Noctua OC Edition

I use nvidia-dkms but had the exact same behavior with the nouveau drivers. I mostly followed the guide here https://wiki.archlinux.org/title/NVIDIA#Installation and also ahve added the modules to /etc/mkinitcpio.conf

MODULES=(btrfs nvidia nvidia_modeset nvidia_uvm nvidia_drm)

I am a bit lost where to look further as journalctl provides no further hints on the potential issue.
Any ideas? Thx

seth · 2025-08-04 16:00:26

If there's a parallel windows, see the 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

Can you s2idle?
https://wiki.archlinux.org/title/Power_ … end_method

Can you suspend from the multi-user.target (2nd link below)?
Can you reboot w/ the https://wiki.archlinux.org/title/Keyboa … el_(SysRq) instead of holding the power button?

binaryanomaly · 2025-08-07 11:51:18

Hey seth, thanks a lot for your help, much appreciated.

I can s2idle but it has the same effect/problem.
When I boot without GUI and suspend, same effect/problem.

I can reboot with Alt+SysRq+b - so it seems to be quite low level but kernel still somewhat functional...

seth · 2025-08-07 13:05:33

I can reboot with Alt+SysRq+b - so it seems to be quite low level but kernel still somewhat functional...

\o/
Please post your complete system journal for that boot, eg.

sudo journalctl -b -1 | curl -F 'file=@-' 0x0.st

for the previous ("-1") one.

binaryanomaly · 2025-08-07 15:53:23

Here it is: (removed)
(removed mac and ipv6 addresses)

Last edited by binaryanomaly (2025-08-08 06:38:57)

seth · 2025-08-07 19:09:44

Ends at

Aug 07 17:33:02 arch-desk kernel: PM: suspend entry (deep)

I can reboot with Alt+SysRq+b

To be clear, the entire Alt+SysRq+reiSUb sequence is required, specifically the S and the U parts to sync the journal to disk and preserve it across the reboot.
Iff you've actually done that (and gave every step like 3 seconds), you should look into https://wiki.archlinux.org/title/Solid_ … leshooting

binaryanomaly · 2025-08-07 19:46:39

Sorry, I got that wrong. I was able to capture more. Seems to work only sometimes...

(removed)

Last edited by binaryanomaly (2025-08-08 06:39:17)

seth · 2025-08-07 21:14:38

Aug 07 21:22:31 arch-desk kernel: Freezing user space processes
Aug 07 21:22:31 arch-desk kernel: Freezing user space processes failed after 20.006 seconds (1 tasks refusing to freeze, wq_busy=0):
Aug 07 21:22:31 arch-desk kernel: task:gnome-shell     state:R  running task     stack:0     pid:1855  tgid:1855  ppid:1729   task_flags:0x400000 flags:0x00004006
Aug 07 21:22:31 arch-desk kernel: Call Trace:
Aug 07 21:22:31 arch-desk kernel:  <TASK>
Aug 07 21:22:31 arch-desk kernel:  ? __schedule+0x411/0x1330
Aug 07 21:22:31 arch-desk kernel:  ? sysvec_reschedule_ipi+0x28/0xf0
Aug 07 21:22:31 arch-desk kernel:  ? asm_sysvec_reschedule_ipi+0x1a/0x20
Aug 07 21:22:31 arch-desk kernel:  ? _nv030487rm+0x10/0x10 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv030593rm+0x2e/0x130 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? os_get_current_tick+0x3a/0xa0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv039794rm+0x8a/0xd0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv039997rm+0x107/0x360 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv033915rm+0xed/0x1f0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv033915rm+0xbd/0x1f0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv033883rm+0x6d5/0x11e0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv025529rm+0x147c/0x1f70 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv013701rm+0x100/0x1f0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv038695rm+0x1e5/0x4a0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv038695rm+0x18f/0x4a0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv042608rm+0xb64/0xf00 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv055051rm+0x289/0x3a0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv053015rm+0xfd/0x160 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv053013rm+0x5c/0x90 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv053013rm+0x32/0x90 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv013972rm+0x64/0xa0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? _nv013972rm+0x28/0xa0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? rm_kernel_rmapi_op+0x92/0x273 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? nvkms_call_rm+0x49/0x80 [nvidia_modeset c49b3c97f3b364172b65cdba3c6a4d5d81b4d15c]
Aug 07 21:22:31 arch-desk kernel:  ? _nv003140kms+0x4c/0x60 [nvidia_modeset c49b3c97f3b364172b65cdba3c6a4d5d81b4d15c]
Aug 07 21:22:31 arch-desk kernel:  ? _nv000593kms+0xb4/0x110 [nvidia_modeset c49b3c97f3b364172b65cdba3c6a4d5d81b4d15c]
Aug 07 21:22:31 arch-desk kernel:  ? _nv000593kms+0x8e/0x110 [nvidia_modeset c49b3c97f3b364172b65cdba3c6a4d5d81b4d15c]
Aug 07 21:22:31 arch-desk kernel:  ? __nv_drm_gem_nvkms_map+0x71/0xe0 [nvidia_drm d731060723a59be20e0e0cd2adebbbcd245001e8]
Aug 07 21:22:31 arch-desk kernel:  ? __nv_drm_gem_nvkms_mmap+0x1a/0x50 [nvidia_drm d731060723a59be20e0e0cd2adebbbcd245001e8]
Aug 07 21:22:31 arch-desk kernel:  ? nv_drm_mmap+0xdb/0x170 [nvidia_drm d731060723a59be20e0e0cd2adebbbcd245001e8]
Aug 07 21:22:31 arch-desk kernel:  ? __mmap_region+0x6cd/0xd40
Aug 07 21:22:31 arch-desk kernel:  ? _nv013969rm+0x6e/0xb0 [nvidia e39dfbdac0a63d83349b17043b3dfbf7e2485d30]
Aug 07 21:22:31 arch-desk kernel:  ? mmap_region+0x82/0xb0
Aug 07 21:22:31 arch-desk kernel:  ? do_mmap+0x470/0x640
Aug 07 21:22:31 arch-desk kernel:  ? vm_mmap_pgoff+0x110/0x200
Aug 07 21:22:31 arch-desk kernel:  ? ksys_mmap_pgoff+0x15c/0x200

Aug 07 21:21:34 arch-desk kernel: i915 0000:00:02.0: [drm] Cannot find any crtc or sizes
Aug 07 21:21:34 arch-desk kernel: i915 0000:00:02.0: [drm] Cannot find any crtc or sizes

Aug 07 21:21:36 arch-desk gnome-shell[1308]: Device '/dev/dri/card0' prefers shadow buffer
Aug 07 21:21:36 arch-desk gnome-shell[1308]: Added device '/dev/dri/card0' (nvidia-drm) using atomic mode setting.
Aug 07 21:21:36 arch-desk gnome-shell[1308]: Created gbm renderer for '/dev/dri/card0'
Aug 07 21:21:36 arch-desk gnome-shell[1308]: Boot VGA GPU /dev/dri/card0 selected as primary

gnome-shell is running on the nvidia GPU (and apparently all outputs are attached to that) and hangs there when you're *trying* to sleep (never actually entered)
What if you use "nvidia_drm.fbdev=0"? (You can btw. remove " loglevel=7 debug systemd.show_status=true")
What if you completely *dis*able https://wiki.archlinux.org/title/NVIDIA … er_suspend (kernel paramter "nvidia.NVreg_PreserveVideoMemoryAllocations=0" and disable the system services)?

binaryanomaly · 2025-08-08 05:17:23

seth wrote:

...and hangs there when you're *trying* to sleep (never actually entered)

Ok seems I caught an oultlier then. It is almost certainly in 9 out of 10 cases going to suspend - I can hear the click of the PSU relay.
I'll try to capture another one but was so far only able to catch this one with entries after

arch-desk kernel: PM: suspend entry (deep)

seth · 2025-08-08 06:50:36

gnome-shell hanging in some ioctl involving the nvidia module is however a bad sign enough, so if you end up not getting "better" journals, I'd certainly look there.

binaryanomaly · 2025-08-08 07:18:51

gnome-shell seems like an additional issue happening from time to time. But if that was the main issue it shouldn´t occur when booting withou gui.

It is however as I suspected not able to get much more information from journalctl. (Below is with "nvidia_drm.fbdev=0")

Aug 08 09:08:06 arch-desk sudo[2610]: pam_unix(sudo:session): session opened for user root(uid=0) by rha(uid=1000)
Aug 08 09:08:06 arch-desk sudo[2610]: pam_unix(sudo:session): session closed for user root
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC callback: terminal.osc.Command{ .end_of_
command = terminal.osc.Command.Command__struct_97953{ .exit_code = 0 } }
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC command: end_of_command
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC callback: terminal.osc.Command{ .end_of_
command = terminal.osc.Command.Command__struct_97953{ .exit_code = 0 } }
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC command: end_of_command
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC callback: terminal.osc.Command{ .end_of_
command = terminal.osc.Command.Command__struct_97953{ .exit_code = null } }
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC command: end_of_command
Aug 08 09:08:06 arch-desk com.mitchellh.ghostty.desktop[2422]: info(osc): unknown semantic prompts option: special_key
Aug 08 09:08:09 arch-desk com.mitchellh.ghostty.desktop[2422]: info(grid): reallocating GPU buffer old=255 new=390
Aug 08 09:08:14 arch-desk systemd-resolved[570]: Clock change detected. Flushing caches.
Aug 08 09:08:14 arch-desk systemd-timesyncd[571]: Contacted time server [2a02:168:420b:d::12]:123 (2.arch.pool.ntp.org).
Aug 08 09:08:14 arch-desk systemd-timesyncd[571]: Initial clock synchronization to Fri 2025-08-08 09:08:14.123606 CEST.
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: info(osc): unknown semantic prompts option: cmdline_url
Aug 08 09:08:17 arch-desk polkitd[810]: Registered Authentication Agent for unix-process:2666:5002 (system bus nam
e :1.109 [/usr/bin/pkttyagent --notify-fd 6 --fallback], object path /org/freedesktop/PolicyKit1/AuthenticationAgent, locale en_US.UTF-8)

Aug 08 09:08:17 arch-desk systemd-logind[796]: The system will suspend now!
Aug 08 09:08:17 arch-desk polkitd[810]: Unregistered Authentication Agent for unix-process:2666:5002 (system bus n
ame :1.109, object path /org/freedesktop/PolicyKit1/AuthenticationAgent, locale en_US.UTF-8) (disconnected from bus)
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC callback: terminal.osc.Command{ .end_of_
command = terminal.osc.Command.Command__struct_97953{ .exit_code = 0 } }
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC command: end_of_command
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC callback: terminal.osc.Command{ .end_of_
command = terminal.osc.Command.Command__struct_97953{ .exit_code = 0 } }
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC command: end_of_command
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC callback: terminal.osc.Command{ .end_of_
command = terminal.osc.Command.Command__struct_97953{ .exit_code = null } }
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: warning(stream): unimplemented OSC command: end_of_command
Aug 08 09:08:17 arch-desk com.mitchellh.ghostty.desktop[2422]: info(osc): unknown semantic prompts option: special_key
Aug 08 09:08:17 arch-desk gnome-shell[1847]: Type gint32 of property AccountsService.User::password-mo
de does not match return type interface of getter get_password_mode. Falling back to slow path
Aug 08 09:08:20 arch-desk systemd[1]: systemd-timedated.service: Deactivated successfully.
Aug 08 09:08:22 arch-desk systemd[1]: systemd-hostnamed.service: Deactivated successfully.
Aug 08 09:08:22 arch-desk systemd-logind[796]: Delay lock is active (UID 1000/rha, PID 1847/gnome-shell) but inhib
itor timeout is reached.
Aug 08 09:08:22 arch-desk systemd[1]: Reached target Sleep.
Aug 08 09:08:22 arch-desk systemd[1]: Starting NVIDIA system suspend actions...
Aug 08 09:08:22 arch-desk suspend[2692]: nvidia-suspend.service
Aug 08 09:08:22 arch-desk logger[2692]: <13>Aug  8 09:08:22 suspend: nvidia-suspend.service
Aug 08 09:08:22 arch-desk kernel: rfkill: input handler enabled
Aug 08 09:08:22 arch-desk gsd-media-keys[1960]: Unable to get default sink
Aug 08 09:08:22 arch-desk gsd-media-keys[1960]: Unable to get default source
Aug 08 09:08:23 arch-desk systemd[1]: nvidia-suspend.service: Deactivated successfully.
Aug 08 09:08:23 arch-desk systemd[1]: Finished NVIDIA system suspend actions.
Aug 08 09:08:23 arch-desk systemd[1]: nvidia-suspend.service: Consumed 825ms CPU time, 376.2M memory peak.
Aug 08 09:08:23 arch-desk systemd[1]: Starting System Suspend...
Aug 08 09:08:23 arch-desk systemd-sleep[2722]: User sessions remain unfrozen on explicit request ($SYSTEMD_SLEEP_F
REEZE_USER_SESSIONS=0).
Aug 08 09:08:23 arch-desk systemd-sleep[2722]: This is not recommended, and might result in unexpected behavior, p
articularly
Aug 08 09:08:23 arch-desk systemd-sleep[2722]: in suspend-then-hibernate operations or setups with encrypted home
directories.
Aug 08 09:08:23 arch-desk systemd-sleep[2722]: Performing sleep operation 'suspend'...
Aug 08 09:08:23 arch-desk kernel: PM: suspend entry (deep)
Aug 08 09:08:23 arch-desk kernel: Filesystems sync: 0.335 seconds

Last edited by binaryanomaly (2025-08-08 07:19:29)

binaryanomaly · 2025-08-08 08:40:00

Got the same behavior with hyprland so we can exclude gnome as the general issue. Also masking nvidia suspend scripts and disabling vidia.NVreg_PreserveVideoMemoryAllocations had no impact.

I suspect an nvidia / mainboard (z890) issue - most likely mainboard being the villain.

Aug 08 10:27:54 arch-desk kernel: ACPI BIOS Error (bug): Failure creating named object [\_SB.PC00.RP01.PXSX._DSM.USRG], AE_ALREADY_EXISTS (20240827/dsfield-184)
Aug 08 10:27:54 arch-desk kernel: ACPI Error: AE_ALREADY_EXISTS, CreateBufferField failure (20240827/dswload2-477)
Aug 08 10:27:54 arch-desk kernel: ACPI Error: Aborting method \_SB.PC00.RP01.PXSX._DSM due to previous error (AE_ALREADY_EXISTS) (20240827/psparse-529)

Might have to wait for newer firmware / kernel / drivers...

Thanks for your help though!

seth · 2025-08-08 13:06:50

It'd always been more on the nvidia than the gnome side.
nb. that ACPI errors during the boot are normal-ish, you cannot read too much into only them.

The boot in #11 either did not wake up or the wakeup wasn't logged or the log wasn't written to disk.

Aug 07 21:21:33 arch-desk kernel: nvme nvme1: D3 entry latency set to 10 seconds
Aug 07 21:21:33 arch-desk kernel: nvme nvme0: D3 entry latency set to 10 seconds
Aug 07 21:21:33 arch-desk kernel: nvme nvme1: 8/0/0 default/read/poll queues
Aug 07 21:21:33 arch-desk kernel:  nvme1n1: p1 p2
Aug 07 21:21:33 arch-desk kernel: nvme nvme0: 15/0/0 default/read/poll queues
Aug 07 21:21:33 arch-desk kernel:  nvme0n1: p1 p2 p3 p4

seth wrote:

If there's a parallel windows, see the 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

Otherwise see https://wiki.archlinux.org/title/Solid_ … leshooting

binaryanomaly · 2025-08-08 15:17:24

seth wrote:

The boot in #11 either did not wake up or the wakeup wasn't logged or the log wasn't written to disk.

It does wake up physically and the fans spin up but the screen never comes back also everything is unresponsive Ctrl+Alt+Del reboot is not possible only Alt+SysReq+b

seth wrote:

If there's a parallel windows, see the 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

I do have dual boot but Windows is on a completely different drive, shouldn't interfere. Also grub boots into the last OS selection by default.

Otherwise see https://wiki.archlinux.org/title/Solid_ … leshooting

Arch is on a Sata disk.

Maybe the motherboard is messing with it. At the very beginning suspend was even unstable on W11. It's now stable since a few BIOS updates.

seth · 2025-08-08 15:23:29

The storage location does absolutely not matter, as long as they're using the same board, windows ACPI control will spill over.
Make sure to disable windows fast start.

binaryanomaly · 2025-08-08 16:02:13

Ah ok. I misread that because the wiki page was so much focussed on the hdds. Disabled fast start up now and turned the machine completely off/on. Behavior is the same though.

seth · 2025-08-08 21:13:10

seth wrote:

The boot in #11 either did not wake up or the wakeup wasn't logged or the log wasn't written to disk.

When you suspend from the multi-user.target, is there any output when you wakeup?
Can you post a journal w/o "loglevel=7 debug systemd.show_status=true" - so iff something sketchy shows up it's nut buried in 60k+ lines of noise?

binaryanomaly · 2025-08-17 09:20:29

No output. No additional info in journarctl.

Also since the most recent kernel/driver updates it seems to really hang: alt + sysreq + b no longer working.

loglevel=3, no debug
https://0x0.st/s/ZMGUO48993blC6lQUL6SnQ/Krlw.log

seth · 2025-08-17 13:19:44

Probably the 6.16 kernel - try to downgrade that or use the LTS kernel, but crucially

alt + sysreq + b

you need to do the the entire alt + sysreq + reisub dance, notably the "s" and "u" parts would be what would preserve the journal.
The posted one ends w/ entering the sleep, nothing to learn from that.

Arch Linux

#1 2025-08-03 19:53:52

Wake from suspend not working

#2 2025-08-04 16:00:26

Re: Wake from suspend not working

#3 2025-08-07 11:51:18

Re: Wake from suspend not working

#4 2025-08-07 13:05:33

Re: Wake from suspend not working

#5 2025-08-07 15:53:23

Re: Wake from suspend not working

#6 2025-08-07 19:09:44

Re: Wake from suspend not working

#7 2025-08-07 19:46:39

Re: Wake from suspend not working

#8 2025-08-07 21:14:38

Re: Wake from suspend not working

#9 2025-08-08 05:17:23

Re: Wake from suspend not working

#10 2025-08-08 06:50:36

Re: Wake from suspend not working

#11 2025-08-08 07:18:51

Re: Wake from suspend not working

#12 2025-08-08 08:40:00

Re: Wake from suspend not working

#13 2025-08-08 13:06:50

Re: Wake from suspend not working

#14 2025-08-08 15:17:24

Re: Wake from suspend not working

#15 2025-08-08 15:23:29

Re: Wake from suspend not working

#16 2025-08-08 16:02:13

Re: Wake from suspend not working

#17 2025-08-08 21:13:10

Re: Wake from suspend not working

#18 2025-08-17 09:20:29

Re: Wake from suspend not working

#19 2025-08-17 13:19:44

Re: Wake from suspend not working

Board footer