Unresponsive system & black screen after returning from suspend

kapergagen · 2026-10-11 11:06:05

Every time I return from suspend/hibernation in Xorg I get unresponsive black screen. TTY (ctrl+alt+FN) switching doesn't work. SysRq works, so to recover from freezes I do ctrl+prtsc+e and then ctrl+alt+del. This started happening after I upgraded the kernel to version 7.2.6. Initially I downgraded to 7.2.4, fixing the issue. Recently I've decided to tackle this again and, after tweaking all kinds of settings I could find, ended up with no solution, except that downgrading now doesn't help either.

Hardware specs from neofetch:
Host: Pulse GL76 11UEK REV:1.0
CPU: 11th Gen Intel i7-11800H (16) @ 4.600GHz
GPU: NVIDIA GeForce RTX 3060 Mobile / Max-Q
GPU: Intel TigerLake-H GT1 [UHD Graphics]

I use bspwm + sxhkd. I start Xorg from the TTY using startx (rootless).

Tried nvidia-dkms, no luck (I didn't forget to install the right linux-headers when downgrading the kernel). Tried LTS kernel - same thing.

/etc/modprobe.d/nvidia.conf:

# options nvidia NVreg_EnableS0ixPowerManagement=1
options nvidia NVreg_PreserveVideoMemoryAllocations=1
# options nvidia NVreg_EnableMSAuthorization=0
# options nvidia-drm modeset=1
options nvidia NVreg_TemporaryFilePath=/var/tmp

Kernel arguments in /etc/default/grub:

GRUB_CMDLINE_LINUX_DEFAULT="resume=LABEL=ArchSwap loglevel=3 sysrq_always_enabled=1 acpi_osi=! \"acpi_osi=Windows 2015\""

I tried enabling and disabling these 3: nvidia-suspend, nvidia-hibernate, nvidia-resume.

nvidia-persistenced.service, whatever that is, is disabled.

Config produced by running nvidia-xconfig makes Xorg freeze on startup before even displaying the GUI (My TTY freezes with startx's initial stdout visible). However, my sxhkd xorg closure shortcut works.

I tried putting the following in /etc/systemd/system/systemd-suspend.service.d/disable-freeze.conf:

[Service]
Environment="SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=false"

Output of

cat .local/share/xorg/Xorg.0.log | grep -E "(\(EE\)|\(WW\))"

is:

	(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[    22.567] (WW) The directory "/usr/share/fonts/OTF" does not exist.
[    22.567] (WW) The directory "/usr/share/fonts/Type1" does not exist.
[    22.578] (WW) Warning, couldn't open module intel
[    22.578] (EE) Failed to load module "intel" (module does not exist, 0)
[    22.581] (WW) Warning, couldn't open module nouveau
[    22.581] (EE) Failed to load module "nouveau" (module does not exist, 0)
[    22.581] (WW) Warning, couldn't open module nv
[    22.581] (EE) Failed to load module "nv" (module does not exist, 0)
[    22.582] (WW) Warning, couldn't open module fbdev
[    22.582] (EE) Failed to load module "fbdev" (module does not exist, 0)
[    22.582] (WW) Warning, couldn't open module vesa
[    22.582] (EE) Failed to load module "vesa" (module does not exist, 0)
[    22.585] (WW) VGA arbiter: cannot open kernel arbiter, no multi-card support
[    22.837] (WW) NVIDIA(G0): Unable to get display device for DPI computation.
[    22.906] (WW) NVIDIA: Failed to bind sideband socket to
[    22.906] (WW) NVIDIA:     '/var/run/nvidia-xdriver-eb60b83a' Permission denied
[    30.994] (EE) libinput: Logitech G304: Parent device not available
[    45.948] (EE) systemd-logind: failed to release device: Connection was disconnected before a reply was received
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.948] (EE) systemd-logind: failed to release device: Connection is closed
[    45.949] (EE) systemd-logind: failed to release device: Connection is closed
[    45.977] (WW) NVIDIA(G0): Failed to set the display configuration
[    45.977] (WW) NVIDIA(G0):  - Setting a mode on head 0 failed: Insufficient permissions
[    45.977] (WW) NVIDIA(G0):  - Setting a mode on head 1 failed: Insufficient permissions
[    45.977] (WW) NVIDIA(G0):  - Setting a mode on head 2 failed: Insufficient permissions
[    45.977] (WW) NVIDIA(G0):  - Setting a mode on head 3 failed: Insufficient permissions
[    46.003] (WW) xf86CloseConsole: KDSETMODE failed: Input/output error
[    46.003] (WW) xf86CloseConsole: VT_GETMODE failed: Input/output error
[    46.003] (WW) xf86OpenConsole: VT_GETSTATE failed: Input/output error
[    46.004] (EE) systemd-logind: ReleaseControl failed: Connection is closed

Lines with 45.XXX timecode must be written on attempt to return from suspend. I really don't know what to do about these errors, though.

seth · 2026-10-11 13:29:58

Output of

https://porkmail.org/era/unix/award#cat

Also don't grep, post the entire Xorg log and if you can reboot using sysrq+REISUB, the journal from the previous (non-waking) boot might help

sudo journalctl -b -1 | nc termbin.com 9999

("-1" is th previous boot, you can access older boots by increasing the digit)

Resp. sleeping from and waking to the multi-user.target (no X11) works fine?

I start Xorg from the TTY using startx

Sanity check, last link below - 2nd blue note about mandatory includes.

kapergagen · 2026-10-11 21:13:16

Full Xorg log: https://paste.c-net.org/PresentSpiral
journalctl log: https://paste.c-net.org/HarshLaura
(Tried termbin.com, but the command would exit with code 0 and no link provided after a while)

Suspend and hibernate work fine with Xorg off.

Resp. sleeping from and waking to the multi-user.target (no X11) works fine?

I don't quite understand what you meant here. Anyway, initially systemctl get-default printed "graphical.target", then I ran "sudo systemctl set-default multi-user.target. That didn't help.

I made sure Windows fast boot is disabled (I dual boot with Windows).

I also made sure that the last if block from /etc/X11/xinit/xinitrc is present in my .xinitrc.

seth · 2026-10-11 21:42:44

https://www.msi.com/Laptop/Pulse-GL76-1 … cification
Optimus, only output is eDP, wired to i915
There're an exteranal SONiX keyboard and logitech G304 mouse.

No indication of wakeup issues in the journal whatsoever.

I ran "sudo systemctl set-default multi-user.target. That didn't help

The command in and by itself won't do much (immediately and w/ your setup "at all") - the idea was to test sleep behavior when no X11 session is/has ever been running during the boot.
Can you
* disable the nvidia GPU (in the UEFI settings)
* remove external keyboard and mouse ( https://bbs.archlinux.org/viewtopic.php … 2#p2311232 - though your HW doesn't fit the general pattern)
to see whether the failure is coming from either of this?

kapergagen · Yesterday 05:45:11

* remove external keyboard and mouse ( https://bbs.archlinux.org/viewtopic.php … 2#p2311232 - though your HW doesn't fit the general pattern)

I found out that going to sleep with my mouse dongle plugged in causes the issue. My main mouse is wireless. Then I tried to plug in an older, wired mouse and it's still the same. Also it doesn't matter if the mouse is on or off. Apparently my keyboard is not involved.

No indication of wakeup issues in the journal whatsoever.

I noticed that commands following suspend are ran indeed. For instance, I ran "systemctl suspend && sleep 10 && echo 123 > abc", waited a little with black screen and after reboot the file was present. It says about finishing a suspend and then "sysrq: Terminate All Tasks" so it must be one of my attempted wakeup boots. I guess it's clean because the issue is mostly Xorg related.

seth · Yesterday 07:21:48

I guess it's clean because the issue is mostly Xorg related.

Did you

seth wrote:

test sleep behavior when no X11 session is/has ever been running during the boot

?

older, wired mouse and it's still the same

There's no issue w/o any mouse but using the G304 dongle or an older (what model specifically) rodent results in wakeup failures?
What does the xorg log look like w/ the old mouse?

Fwiw

[    16.170] (II) XINPUT: Adding extended input device "Logitech G304" (type: KEYBOARD, id 23)
[    36.583] (EE) libinput: Logitech G304: Parent device not available

kapergagen · Yesterday 09:33:15

Having a dongle plugged in, it is enough to remove it right before waking to avoid wakeup failure (and plug it back after successful wakeup).

Xorg log with G304 dongle plugged in for the entire time (failure): https://paste.c-net.org/DemonsKinky
Xorg log with my older Dell mouse (it's connected via USB)(failure, too): https://paste.c-net.org/GottaGirdle
Xorg log with G304 dongle plugged out right before waking (success): https://paste.c-net.org/HechtCaspar
Xorg log with Dell mouse plugged out before waking (success): https://paste.c-net.org/BlushEntirely

Suspension and wakeup with Xorg off (doesn't matter if it was never started or was exited from) work correctly regardless of whether a mouse is plugged in.

Fwiw

Logitech G304 is my mouse, why does it say "type: KEYBOARD"? My external keyboard is Red Square IO, definitely not Logitech...

Last edited by kapergagen (Yesterday 09:33:52)

seth · Yesterday 17:55:13

why does it say "type: KEYBOARD"?

That's not even that uncommon - the mouse provides more buttons as keyboard device or in this case it might also just be an artifact of the dongle, but

[   183.180] (II) XINPUT: Adding extended input device "Logitech USB Optical Mouse" (type: MOUSE, id 12)

your other mouse (might say dell, is logitech) only shows up as mouse.

It very much sounds like the the udev bug but the unclear device patterns don't allow an obvious mitigation
See https://bbs.archlinux.org/viewtopic.php … 8#p2297968 - since the system properly you might be able to catch the troublesome udev action this way …

Arch Linux

#1 2026-10-11 11:06:05

Unresponsive system & black screen after returning from suspend

#2 2026-10-11 13:29:58

Re: Unresponsive system & black screen after returning from suspend

#3 2026-10-11 21:13:16

Re: Unresponsive system & black screen after returning from suspend

#4 2026-10-11 21:42:44

Re: Unresponsive system & black screen after returning from suspend

#5 Yesterday 05:45:11

Re: Unresponsive system & black screen after returning from suspend

#6 Yesterday 07:21:48

Re: Unresponsive system & black screen after returning from suspend

#7 Yesterday 09:33:15

Re: Unresponsive system & black screen after returning from suspend

#8 Yesterday 17:55:13

Re: Unresponsive system & black screen after returning from suspend

Board footer