System Shutdown and resume hangs since Kernel linux-image-5.13.0-23-generic

Status changed to 'Confirmed' because the bug affects multiple users.

I believe I'm having the same/similar issue. I cannot shut down or restart my system unless I hold the power button to force shut down.

I've tried using two different terminal commands in order to shut down the system, those being `sudo shutdown now' and '$ poweroff'.

Both commands end with the same result: the system status message shows: 'systemd-shutdown[1]: Waiting for process: systemd-udevd,systemd-udevd' and then nothing happens. The screen stays like this until I force shut down using the power button.

Ubuntu 20.04.3 LTS
Uname: 5.13.0-27-generic

I believe I'm having the same issue as well. Have tried both GUI and terminal commands.

Ubuntu 20.04.3
Uname: 5.13.0-27-generic

I've confirmed that by rolling back to a prior kernel (5.11.0-46-generic), the system shuts down normally.

Temp work around I used was to

>sudo vi /etc/default/grub

GRUB_DEFAULT=0
#GRUB_TIMEOUT_STYLE=hidden <-- comment this line
GRUB_TIMEOUT=10 <-- change this line to a value between 5 and 10
GRUB_DISTRIBUTOR=`lsb_release -i -s 2> /dev/null || echo Debian`
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
GRUB_CMDLINE_LINUX=""

>sudo update-grub

This allows me to see the GRUB boot menu and then select "Advanced options for Ubuntu" and then the kernel I want to boot

I can confirm that it works until 5.13.22 and from 5.13.23 onward shutting down does not work.

Here is the danger of this bug on a laptop (as I've experienced):

The system, from all appearances looks powered down (with the exception of the power indicator which remains lit on my laptop). Unfortunately, the power indicator is not in a conspicuous place which means that in order to *verify* the system is off, I have to lean over and look at the side of the laptop.

When a user gets into the habit of a system just working, checking the power indicator slips the mind. Hence, a powered on laptop may be placed inside a tightly insulated laptop bag. When this happened to me, I didn't discover it for about 2 hours. When I went to take the laptop out of the bag, the fan was now running trying to keep the internals cool, but the bag was insulated (padded). This resulted in the system being so hot that I could not even touch the keyboard in places without getting mildly burned.

I have no idea what will happen to a discharging lithium battery that is being heated rather than cooled in that situation, but I imagine there may be a component of definite risk to a system.

I believe I’m having the same issue. I see the same failure regarding systemd-udevd, systemd-udevd. This did not happen on the last version I had - 5.11.0-46-generic.

Also it tells me that systemd-udevd worker (541)processing SEQNUM=3327 killed.
And that worker 541 terminated by signal 9 (kill) and worker (541) failed.

All this happens as I’m trying to reboot. It just shows this error while the system hangs indefinitely.

System runs AMD Ryzen 7 4800U
Ubuntu 20.04.3
Uname: 5.13.0-27-generic

I think this bug specifically affects AMD processor based systems. Whoever picks it up should look into it from that angle IMO.

This is the same as https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1956821 . Not marking either as duplicates of one another to maintain proper visibility (though I think they definitely are).

Thanks Dan

I've rolled back to kernal 5.11.0-46-generic and shut-down/restart is re-established normally.
I've updated the grub to save this as default for the time being.

I'm new to the way Linux distros release their updates/bug fixes. My system is fixed (stable) for now, but how will I know if a new kernal release will address the bug in future? Any advice?

John

Hi John,

Ideally you’ll be able to read specific references to fixes within the kernel release notes but in practice they are not that organized per se. That said, in the other ticket I referenced above, one or two users mentioned that kernel 5.14.X fixes this issue (which seems to be part of ubuntu 20.04.4 that’s suppose to be released on February 24). So I’m guessing unless the kernel is updated earlier, at the latest this should be fixed by end of February.

Ok Dan - thanks very much.

I will patiently await 5.14.X as 5.11.X works perfectly.

I guess this type of rollback is the obvious first check for any suspected kernal bug. Noted for future.
John

Thank you very much. Until Kernel 5.14.X / Ubuntu 20.04.4 there is a workaround, I am curios about the fix in Ubuntu 20.04.4.
Martin

Hey everyone! kernel version 5.13.0-28-generic is out and it seems to fix this issue! Myself and another user were already able to confirm that. Good luck everyone!

Great, it works now. Thank you so much.
Martin

Good morning!
Yesterday my Kernel was updated to 5.13.0-28-generic via the regular
update cycle: I can connfirm that my problem is fixed too,
Thanks to you all!

Am 31.01.22 um 21:42 schrieb Dan:
> Hey everyone! kernel version 5.13.0-28-generic is out and it seems to
> fix this issue! Myself and another user were already able to confirm
> that. Good luck everyone!
>

--
Kontakt:
Ing. Anton Lettenbichler
Prof.-Böhm-Straße 6
4870 Vöcklamarkt

Fon: +43(0)699 1105 76 69
E-mail: <email address hidden>
Web: www.geosono.com

Firma:
geosono - Anton Lettenbichler e.U.
Sitz des Unternehmens: Vöcklamarkt
FirmenbuchNr: FN 311210 m
Firmenbuchgericht: Landesgericht Wels

UID/VAT-ID: ATU64165423

 Same here, cleared with 5.13.0-28 update.
Thanks