[ehl] systemd-udevd

[EricLiu]

Dear all,

Platform : Elkhart Lake (x6425RE)
OS : Ubuntu 20.04 (5.13.0-1008-intel),[iso file name : ubuntu-20.04-desktop-amd64+intel-iot.iso]

After click power off option, the board is still running, it can't power down.
Few minutes later, the screen will show the following mesg:
=====================================================================
systemd-shutdown[1]: Waiting for process: systemd-udevd, systemd-udevd, systemd-udevd
systemd-udevd[303]: INTC1020:00: Worker [335] processing SEQNUM=3118 killed
systemd-udevd[303]: Worker [335] terminated by signal 9 (KILL)
systemd-udevd[303]: INTC1020:00: Worker [335] failed
systemd-udevd[303]: INTC1020:00: Worker [334] processing SEQNUM=3593 killed
systemd-udevd[303]: Worker [334] terminated by signal 9 (KILL)
systemd-udevd[303]: INTC1020:00: Worker [334] failed
=====================================================================

How can i fix this issue?

Thanks.

BR,
Eric.

[EricLiu (ericliu16663)]

Add the journalctl log in attachment

[mlsvrts (mlsvrts)]

It's possible to workaround this issue using kernel commandline arguments by either:

Blacklisting the pinctrl_elkhartlake driver (modprobe.blacklist=pinctrl_elkhartlake) or
Lowering the udev worker timeout (udev.event-timeout=10)

I can't comment on the impact to platform features if you do disable the driver. The root-cause may be a BIOS/ACPI issue, so you could first try updating your system firmware.

Hope this helps,

mlsvrts

[EricLiu (ericliu16663)]

Hi mlsvrts,

Our bios is based on latest intel rc code(MR2, v.1.4.0), even run with pure code with intel crb board, it still has the same problem (can't power down). DO you know which intel rc code version you verified?

More about the previous comment:
Q1. About these driver(pinctrl_elkhartlake) and setting(udev worker timeout), could you provide the steps to do for me to try?
Q2. Do you have any suggestion about bios' acpi setting?

Thanks.

BR,
Eric.

[EricLiu (ericliu16663)]

Hi mlsvrts,

I modified the blacklist file which is in "/etc/modprobe.d/" as following :

blacklist pinctrl_elkhartlake

After reboot, then try to power off. It seems not work.
btw, i don;t know how to set udev.event-timeout=10.

Thanks.

BR,
Eric.

[EricLiu (ericliu16663)]

Hi mlsvrts,

I added "modprobe.blacklist=pinctrl_elkhartlake udev.event-timeout=10" in /etc/default/grub file:
GRUB_CMDLINE_LINUX_DEFAULT = "quiet splash modprobe.blacklist=pinctrl_elkhartlake udev.event-timeout=10"
and run "grub-mkconfig -o /boot/grub/grub.cfg" in terminal.

After force reboot, it seems the power on issue Bug #1955469 won't happen(no more delay 120s with fbcon).
But it still can't power down.

Thanks.

BR,
Eric.

[EricLiu (ericliu16663)]

Hi mlsvrts,

Update the modified result:
1. I added "modprobe.blacklist=pinctrl_elkhartlake" in /etc/default/grub file:
GRUB_CMDLINE_LINUX_DEFAULT = "quiet splash modprobe.blacklist=pinctrl_elkhartlake"
and run "grub-mkconfig -o /boot/grub/grub.cfg" in terminal.

2. I modified /etc/udev/udev.conf file:
event-timeout=10

After force reboot, it seems the power on issue Bug #1955469 won't happen(no more delay 120s with fbcon).
But it still can't power down.

Thanks.

BR,
Eric.

[Andy Shevchenko (andy.shevchenko)]

https://bugzilla.redhat.com/show_bug.cgi?id=1948468
https://bugzilla.kernel.org/show_bug.cgi?id=213365

I believe second one has a link to a workaround patch.

[Rex Tsai (chihchun)]

Changed to won't fix, based on https://bugzilla.redhat.com/show_bug.cgi?id=1948468#c11 Intel is deploying a new BIOS for OEMs/ODMs/customers. Nothing to fix in upstream.