[SOLVED] X is randomly dying with a useless backtrace

incertia · 2014-10-27 19:23:18

This happens occasionally when I make edits in vim, always when I run emacs without -nw, and most of the time when I change a split size in tmux. It's gotten to the point where it becomes extremely annoying to do any work because I have to keep reloading and deleting swap files. This is a fairly fresh install of arch. Anybody know what could be the issue? The only not so standard packages I believe I've installed are steam and a few fonts. This is in xfce4.

tmux does not crash anything under the standard xserver, but emacs still does.

Last edited by incertia (2014-10-29 19:39:41)

Trilby · 2014-10-27 19:28:04

The backtrace might be useless for you, but for us it'd be the only thing to go on. You haven't given us anything that could help.

The backtrace and the xorg.logs would definitely be usefull. You could also check/share your journal and dmesg.

Without any inforation to go on, the best random guess I could make would be that it is a video driver issue. So also please share what video driver you are using, what you video card is (lspci output), and whether you've tried alternatives (proprietary vs FOSS drivers for ati or nvidia).

incertia · 2014-10-27 19:36:54

This is anything that is tagged with EE in /var/log/Xorg.0.log, it appears as if something is trying to access NULL.

[    69.560] (EE) 
[    69.561] (EE) Backtrace:
[    69.602] (EE) 0: /usr/bin/Xorg.bin (xorg_backtrace+0x56) [0x593a46]
[    69.602] (EE) 1: /usr/bin/Xorg.bin (0x400000+0x197b99) [0x597b99]
[    69.603] (EE) 2: /usr/lib/libc.so.6 (0x7fe6b5f61000+0x339f0) [0x7fe6b5f949f0]
[    69.603] (EE) 
[    69.603] (EE) Segmentation fault at address 0x0
[    69.603] (EE) 
[    69.603] (EE) Caught signal 11 (Segmentation fault). Server aborting
[    69.603] (EE) 
[    69.603] (EE) 
[    69.603] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    69.603] (EE) 
[    71.036] (EE) Server terminated with error (1). Closing log file.

What should I be looking for in the output of dmesg and journalctl?

My laptop is has switchable graphics, so I've installed xf86-video-{ati,intel} (not sure how to check which one is currently in use). lspci gives the card names as

00:02.0 VGA compatible controller: Intel Corporation Haswell-ULT Integrated Graphics Controller (rev 0b)
05:00.0 Display controller: Advanced Micro Devices, Inc. [AMD/ATI] Venus XTX [Radeon HD 8890M / R9 M275X] (rev ff)

Last edited by incertia (2014-10-27 19:40:16)

Stebalien · 2014-10-28 05:40:02

This is the signal handler:

[ 69.602] (EE) 1: /usr/bin/Xorg.bin (0x400000+0x197b99) [0x597b99]

This is (as far as I can tell, __restore_rt (from glibc):

[    69.603] (EE) 2: /usr/lib/libc.so.6 (0x7fe6b5f61000+0x339f0) [0x7fe6b5f949f0]

The good news is you may not be alone: https://bbs.archlinux.org/viewtopic.php?id=187235

Unfortunately, I'm pretty sure that this part of the crash is only the symptom and the problem is elsewhere and I have no idea where to look (any stack traces in dmesg?). At this point, I'd report the bug upstream.

Last edited by Stebalien (2014-10-28 05:40:25)

incertia · 2014-10-29 19:38:29

So I managed to fix my problem. I just needed to configure the intel driver with /etc/X11/xorg.conf.d/20-intel.conf. Here are the contents for anyone else who is interested

Section "Device"
    Identifier "Intel Graphic"
    Driver "intel"
    Option "AccelMethod" "uxa"
EndSection

Arch Linux

#1 2014-10-27 19:23:18

[SOLVED] X is randomly dying with a useless backtrace

#2 2014-10-27 19:28:04

Re: [SOLVED] X is randomly dying with a useless backtrace

#3 2014-10-27 19:36:54

Re: [SOLVED] X is randomly dying with a useless backtrace

#4 2014-10-28 05:40:02

Re: [SOLVED] X is randomly dying with a useless backtrace

#5 2014-10-29 19:38:29

Re: [SOLVED] X is randomly dying with a useless backtrace

Board footer