system is broken after waking up from suspend

Stella · 2026-05-14 15:02:41

After waking up from a suspend, the system is broken in a strange state where
- all systemd services seem to be blocked
- shells are extremely slow or completely unresponsive, all other applications are frozen

After a reboot everything is fixed.

Here is the journalctl output from a boot where I
- booted like normal
- suspended and woke up (using the keyboard)
- tried and failed to log into tty3
- eventually managed to reboot through kitty on tty1, which took a few minutes to respond

https://paste.c-net.org/HerselfGrows

I've tried various display managers and ways of activating the suspend but it's the same issue across the board, though sometimes there is no graphical output at all after the wake up.

seth · 2026-05-14 15:07:32

May 14 16:41:04 mother-brain kernel: nvme nvme0: I/O tag 585 (c249) QID 5 timeout, completion polled
May 14 16:41:04 mother-brain kernel: nvme nvme0: I/O tag 603 (c25b) QID 5 timeout, completion polled
May 14 16:41:04 mother-brain kernel: nvme nvme0: I/O tag 618 (d26a) QID 5 timeout, completion polled
May 14 16:41:04 mother-brain systemd[1]: systemd-journald.service: start operation timed out. Terminating.
May 14 16:41:04 mother-brain kernel: INFO: task NetworkManager:873 blocked for more than 122 seconds.
May 14 16:41:04 mother-brain kernel:       Tainted: G           OE       7.0.3-zen1-2-zen #1
May 14 16:41:04 mother-brain kernel: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
May 14 16:41:04 mother-brain kernel: task:NetworkManager  state:D stack:0     pid:873   tgid:873   ppid:1      task_flags:0x400100 flags:0x00080800
May 14 16:41:04 mother-brain kernel: Call Trace:
May 14 16:41:04 mother-brain kernel:  <TASK>

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

Edit: seems there's also a problem w/ the RAM temp sensor

May 14 16:29:29 mother-brain kernel: spd5118 25-0051: DDR5 temperature sensor: vendor 0x00:0xb3 revision 2.2
May 14 16:29:29 mother-brain kernel: spd5118 25-0053: DDR5 temperature sensor: vendor 0x00:0xb3 revision 2.2
May 14 16:30:28 mother-brain kernel: spd5118 25-0051: Failed to write b = 0: -6
May 14 16:30:28 mother-brain kernel: spd5118 25-0051: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
May 14 16:30:28 mother-brain kernel: spd5118 25-0051: PM: failed to resume async: error -6
May 14 16:30:28 mother-brain kernel: spd5118 25-0053: Failed to write b = 0: -6
May 14 16:30:28 mother-brain kernel: spd5118 25-0053: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
May 14 16:30:28 mother-brain kernel: spd5118 25-0053: PM: failed to resume async: error -6

Last edited by seth (2026-05-14 15:08:22)

Stella · 2026-05-14 16:01:30

Thank you for the reply, I tried adding the kernel parameters (in full rw root=UUID=c1c4fc4f-b6ab-4077-bb58-d4bc1d8e95e8 nvme_core.default_ps_max_latency_us=0 pcie_aspm=off pcie_port_pm=off iommu=soft initrd=boot\initramfs-linux-zen.img) but it didn't appear to change anything.
I could investigate the RAM temp sensor but I'm not entirely sure what to do there, worst case is a faulty motherboard I assume?

seth · 2026-05-14 18:37:28

ASPM btw. isn't available anyway, you'd bet on "nvme_core.default_ps_max_latency_us=0 iommu=soft"

Does

cat /proc/cmdline

reflect that those parameters are applied?
(The nvme errors are a bit too much on the nose and the drive is certainly the reason for the symptoms)

You can "module_blacklist=spd5118" to get that out of the equation and maybe reloading it restores its function after the S3, but we'll need to get the drive issues sorted out.

Do you have the same problem w/ the lts or main kernel?

Stella · 2026-05-14 22:59:40

- /proc/cmdline did have the correct kernel parameters
- the same issue persists on main, but not on lts 6.18.29, which seems to have completely fixed it (even without the kernel parameters)

I can bisect the version where this breaks, if needed

seth · 2026-05-15 07:36:56

The https://wiki.archlinux.org/title/Arch_Linux_Archive may speed up the process (basically test whether this is already on 6.19)
Do you have the spd5118 errors with the LTS kernel?

Random sanity checks:
there seem to be residuals of nvidia drivers and you're running supergfxd but this doesn't seem to be a hybrid graphics system?
vfio loads but doesn't seem to take control of/forward anything … and are you building the vbox modules for the LTS kernel?

Arch Linux

#1 2026-05-14 15:02:41

system is broken after waking up from suspend

#2 2026-05-14 15:07:32

Re: system is broken after waking up from suspend

#3 2026-05-14 16:01:30

Re: system is broken after waking up from suspend

#4 2026-05-14 18:37:28

Re: system is broken after waking up from suspend

#5 2026-05-14 22:59:40

Re: system is broken after waking up from suspend

#6 2026-05-15 07:36:56

Re: system is broken after waking up from suspend

Board footer