| Summary: | ~/.xinitrc crashes server | ||||||
|---|---|---|---|---|---|---|---|
| Product: | xorg | Reporter: | Felix Miata <mrmazda> | ||||
| Component: | Server/General | Assignee: | Xorg Project Team <xorg-team> | ||||
| Status: | RESOLVED NOTABUG | QA Contact: | Xorg Project Team <xorg-team> | ||||
| Severity: | normal | ||||||
| Priority: | medium | ||||||
| Version: | unspecified | ||||||
| Hardware: | x86 (IA32) | ||||||
| OS: | Linux (All) | ||||||
| Whiteboard: | |||||||
| i915 platform: | i915 features: | ||||||
| Attachments: |
|
||||||
Is this reproducible with server 1.9? Can you get a backtrace from gdb? Note that having an empty .xinitrc, or just xrandr there, is probably not what you want, since the session exits when .xinitrc is done... (In reply to comment #1) > Is this reproducible with server 1.9? It is reproducible with today's Rawhide 1.8.99.904/xrandr 1.3 on i845G. > Can you get a backtrace from gdb? Only with help. I've not used gdb in many moons, and then only given specific instructions. I'm a-865 on freenode's #xorg if someone would care to explain how to do it outside this bug or email. > Note that having an empty .xinitrc, or just xrandr there, is probably not what > you want, since the session exits when .xinitrc is done... That seems true, but I had to start somewhere, and I've yet to find uncryptic docs comprehensively explaining how to configure what I need. > --- Comment #2 from Felix Miata <mrmazda@earthlink.net> 2010-07-17 12:02:58 PDT --- > > Can you get a backtrace from gdb? > > Only with help. I've not used gdb in many moons, and then only given specific > instructions. I'm a-865 on freenode's #xorg if someone would care to explain > how to do it outside this bug or email. > Maybe the easiest is to use 'startx -- -core', then when X crashes it should leave a 'core' file in its working directory. Run 'gdb /usr/bin/Xorg core', then 'bt full' at the gdb prompt. There's some more explanations at http://www.x.org/wiki/Development/Documentation/ServerDebugging (In reply to comment #2) > (In reply to comment #1) > > Is this reproducible with server 1.9? > It is reproducible with today's Rawhide 1.8.99.904/xrandr 1.3 on i845G. I think I mistook what I was seeing on Rawhide. Now I think it looks like it's just a normal exit, but immediate, instead of following KDE startup and normal exit. > --- Comment #4 from Felix Miata <mrmazda@earthlink.net> 2010-07-17 12:15:55 PDT ---
> (In reply to comment #2)
> > (In reply to comment #1)
> > > Is this reproducible with server 1.9?
> > It is reproducible with today's Rawhide 1.8.99.904/xrandr 1.3 on i845G.
>
> I think I mistook what I was seeing on Rawhide. Now I think it looks like it's
> just a normal exit, but immediate, instead of following KDE startup and normal
> exit.
that's the expected behaviour when your .xinitrc doesn't start kde...
On Sun, Jul 18, 2010 at 04:11:22AM -0700, bugzilla-daemon@freedesktop.org wrote: > --- Comment #5 from Julien Cristau <jcristau@debian.org> 2010-07-18 04:11:22 PDT --- > > --- Comment #4 from Felix Miata <mrmazda@earthlink.net> 2010-07-17 12:15:55 PDT --- > > I think I mistook what I was seeing on Rawhide. Now I think it looks like it's > > just a normal exit, but immediate, instead of following KDE startup and normal > > exit. > > that's the expected behaviour when your .xinitrc doesn't start kde... Right. .xinitrc contains the _complete list_ of clients to be run by startx, and nothing else will be run if it's present. So if you want to autostart something in addition to KDE, use the KDE autostart preferences. (In reply to comment #6) > Right. .xinitrc contains the _complete list_ of clients to be run by > startx, and nothing else will be run if it's present. Your reponse was premature. I've not finished testing all previous configurations reported to see which segfault and which WAD. BTW, I didn't change the resolution. This bug page was still open in a tab waiting on me to finish testing, then respond. On reload after seeing comment 6 bugmail, Bugzilla changed resolution, but I never looked at that part of the page before submitting. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Created attachment 37146 [details] Xorg.X.log Attachment is Xorg.1.log from rv380 radeon on openSUSE 11.3 (1.8.0/1.3.2), but this happens also with rv200 radeon on openSUSE 11.2 (1.6.5/1.3), i845G on openSUSE 11.3M4 (1.8/1.3), and i845G on Mandriva 2010.1 (1.7.7/1.3). To reproduce: 1-create ~/.xinitrc (either empty, or with such valid content as 'xrandr --output VGA-0 --panning 1920x1200') 2a-startx, or 2b-log in from KDM Actual results: 1-X immediately segfaults @ 0x4 Expected results: 1-X runs normally Comments: 1-I created .xinitrc in order to begin customization per user so that like in pre-xrandr days, where one could use a different X config file for each of multiple X sessions, I could today use a different effective screen resolution under 1.8.x/1.3.x for each X session. I started by merely putting 'xrandr --output VGA-0 --panning 1920x1200' to emulate WUXGA on a UXGA CRT, but found even an empty .xinitrc causes the server to crash. 2-My expectation is to find a way to not only pan or not per X instance, but also run a different actual resolution for each X instance, just like in pre-xrandr versions, or in post-xrandr versions in which 'Option "NoRandr"' in xorg.conf worked.