Bug 1083939

Summary:	Libreoffice Writer stalled, X server restarted
Product:	[openSUSE] openSUSE Distribution	Reporter:	Thomas Schmid <thomi.schmid>
Component:	X.Org	Assignee:	E-mail List <xorg-maintainer-bugs>
Status:	RESOLVED INVALID	QA Contact:	E-mail List <xorg-maintainer-bugs>
Severity:	Normal
Priority:	P5 - None	CC:	mstaudt, thomi.schmid
Version:	Leap 15.0
Target Milestone:	---
Hardware:	x86-64
OS:	openSUSE Factory
Whiteboard:
Found By:	---	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---
Attachments:	Lenovo X260 HW Description provided by YaST Syslog during problem 2018-05-03T09:12 and T09:13

Description Thomas Schmid 2018-03-05 11:49:51 UTC

Created attachment 762677 [details]
Lenovo X260 HW Description provided by YaST

I was writing in LO Writer, when Plasma 5 Workspace froze. After maybe 30s Plasma workspace restarted itself, presenting me with the login screen. Afer logging in again, I tried to start LO again, but to nothing happend. I then went to konsole, checked the running process (2 soffice splash, 1 soffice.bin) and killed the running soffice.bin. After that LO started up automatically.

Software:
openSuSE 15.0 Beta, Mid Februar 2018 Update
Plasma 5 Workspace
LO 6.0.1.1

Hardware:
Lenovo X260

Attached HW description and syslog: the interesting part is between 2018-03-05T09:12 (mark "# PROBLEM STARTS HERE") and 2018-03-05T09:13 (mark "# PROBLEM ENDS HERE").

I have now updated to the latest openSuSE 15.0 beta update.

There is a possible connection to https://bugzilla.opensuse.org/show_bug.cgi?id=1048858, which sounds very similar.

Comment 1 Thomas Schmid 2018-03-05 11:51:17 UTC

Created attachment 762680 [details]
Syslog during problem 2018-05-03T09:12 and T09:13

Comment 2 Stefan Dirsch 2018-03-05 13:42:33 UTC

Please provide your /var/log/Xorg.0.log* files right after the issue has happenend.

Comment 3 Thomas Schmid 2018-03-05 13:59:15 UTC

Unfortunately this is not possible: I have a Xorg.0.log.old which is from the right period. But for unknown reasons, it starts with log entries having kernel time stamps pointing at a system time 3h *after* the event.
As I wrote, 3+h after the incident, I did a openSuSE update and restarted the X260 labtop.

Comment 4 Thomas Schmid 2018-03-05 14:03:19 UTC

I guess at the restart of the X.org, it copied the Xorg.0.log to Xorg.0.log.old. Then when I restarted the labtop, Xorg log got rotated again, overwriting Xorg.0.log.old.

Comment 5 Stefan Dirsch 2018-03-05 14:36:10 UTC

(In reply to Thomas Schmid from comment #4)
> I guess at the restart of the X.org, it copied the Xorg.0.log to
> Xorg.0.log.old. Then when I restarted the labtop, Xorg log got rotated
> again, overwriting Xorg.0.log.old.

Yes. That's correct. If you can't access the system any longer after the issue has happened and need to reboot, you can  reboot in runlevel 3, so X isn't started. Just add

  3

to the kernel command line in grub. Then save Xorg.0.log.old somehow.

Comment 6 Thomas Schmid 2018-03-05 18:04:43 UTC

Ok, I understand. As I wrote, the reboot happened after the event after I updated my Leap 15.0 installation, and thought that after such a massive update (1100+ packages) a reboot would be prudent...
So in summary, without the Xorg logs nothing can be done ?

Comment 7 Stefan Dirsch 2018-03-06 10:52:29 UTC

I'm not sure I understand correctly. As long as you can reproduce the issue, you can also provide the appropriate logfiles.

Comment 8 Thomas Schmid 2018-03-06 18:06:01 UTC

Well I don't know if it is reproducable: I have not seen it again, but have not uswd LO Writer yet either. We'll see..

Comment 9 Max Staudt 2018-03-22 14:54:01 UTC

Okay, it's been two weeks, and no further complaints. It may just as well have been a glitch in Plasma.

Let's close this for now, and if it becomes reproducible, please feel free to REOPEN the ticket, so we can have a look.

For lack of a better label, I'll close this as RESOLVED INVALID for now.