Mailinglist Archive: opensuse-factory (807 mails)

< Previous Next >
Re: [opensuse-factory] how to debug hibernation issue
Am Sonntag, den 12.02.2017, 19:42 +0100 schrieb Axel Braun:
Feb 12 19:20:33 eeePC kernel: PM: Image not found (code -22)
Feb 12 19:20:33 eeePC kernel: PM: Hibernation image not present or
could not be loaded.
Feb 12 19:20:33 eeePC systemd[1]: Started Resume from hibernation
using device /dev/sda1.
Feb 12 19:20:33 eeePC systemd[1]: Reached target Local File Systems
(Pre).
Feb 12 19:20:33 eeePC systemd[1]: Reached target Local File Systems.
Feb 12 19:20:33 eeePC systemd[1]: Reached target System
Initialization.
Feb 12 19:20:33 eeePC systemd[1]: Reached target Basic System.
Feb 12 19:20:33 eeePC systemd[1]: Starting File System Check on
/dev/disk/by-uuid/efb6a106-11a1-4537-baac-2d2

Obviously the image file was not written properly?
Any idea what I could try?

You are running into a safety feature. The "sacrificial" kernel clears
the hibernation signature on the swap partition. This is necessary
because if the resumption fails you'd be in an eternal boot loop.

Essentially a driver crashes when it needs to resume. Debugging this
is usually cumbersome, but straight forward.

1. You determine whether the the user space resumption code is faulty.
To do so you do a brute force suspension

echo "disk" > /sys/power/state

This will make the kernel do a brute force suspension with an
uncompressed image.

2. If that also fails you need to identify the failing driver.
As tedious as it sounds, blacklist more and more drivers until
your system survives a cycle.

HTH
Oliver

--
To unsubscribe, e-mail: opensuse-factory+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse-factory+owner@xxxxxxxxxxxx

< Previous Next >
Follow Ups
References