You are not logged in.
After upgrading my ASUS UX31A automatically reboots after some time (about one minute after boot).
I've downgraded the package linux-4.12.3-1 to linux-4.11.9-1 and the problem vanished.
Still, I'd like to investigate what the problem caused as I don't want to be stuck on that kernel version.
Is this reboot issue already a known problem? How can I investigate what exactly caused the reboot?
Offline
Read your journal.
Offline
I assume that you're referring to the output of journalctl.The output looks normal, which is not surprising as booting up works fine. Just *something* triggers a reboot some time after a successful boot.
How do I find out what triggered a reboot?
Offline
journalctl by default stores logs across boots persistently on disk. When you look at the output from journalctl, you should see logs from different boots, including all the way from boot to crash. Hopefully there is some message from just before the reboot indicating a problem. Could you post an example output starting from a previous boot, and leading up until right before you get rebooted again? Also, since you don't experience it with the other kernel, sounds like it's indeed a kernel problem (and not hardware, for example). So if there is no hints from journalctl, and if you want to investigate for yourself what is causing the issue, probably your best bet is to use git bisect to find which kernel commit causes the regression. See https://wiki.archlinux.org/index.php/Bisecting_bugs and search online for "linux git bisect kernel".
Last edited by GenkiSky (2017-08-05 00:45:34)
Offline
Same here. There is nothing notable in my system's journal. It appears to be a normal boot, and then the log just ends.
Offline
the timeout for the watchdog is too low [2]. you can fix it in your systemd configuration, see [1], until it gets fixed upstream.
[1] https://www.reddit.com/r/archlinux/comm … e/dkuy4pm/
[2] https://bugzilla.kernel.org/show_bug.cgi?id=196509
Offline
Wow, that is some impressive debugging! How did you know?
Offline
Yes! Changing the watchdog timeout fixes it for me.
Thanks, nem!
Offline