Bug 600366 - KDM or X crashes on startup
Summary: KDM or X crashes on startup
Keywords:
Status: CLOSED WORKSFORME
Alias: None
Product: Fedora
Classification: Fedora
Component: kdebase-workspace
Version: 13
Hardware: x86_64
OS: Linux
low
medium
Target Milestone: ---
Assignee: Than Ngo
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2010-06-04 15:01 UTC by Patrick O'Callaghan
Modified: 2010-06-15 23:28 UTC (History)
8 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2010-06-15 23:22:15 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
/var/log/messages (65.28 KB, text/plain)
2010-06-04 15:01 UTC, Patrick O'Callaghan
no flags Details
dmesg output (40.23 KB, text/plain)
2010-06-04 15:04 UTC, Patrick O'Callaghan
no flags Details
lsmod output (2.19 KB, text/plain)
2010-06-04 15:06 UTC, Patrick O'Callaghan
no flags Details

Description Patrick O'Callaghan 2010-06-04 15:01:24 UTC
Created attachment 421251 [details]
/var/log/messages

Description of problem:
On rebooting, either kdm or X apparently crashes, leaving only a console login.

Version-Release number of selected component (if applicable):
kdm-4.4.3-1.fc13.1.x86_64
xorg-x11-drv-intel-2.11.0-4.fc13.x86_64
xorg-x11-server-Xorg-1.8.0-12.fc13.x86_64

How reproducible:
Solid

Steps to Reproduce:
1.Reboot system
2.X spinner appears
3.X terminates, leaving a console
  
Actual results:
No X session

Expected results:
KDE session

Additional info:
/var/log/messages contains:
Jun  3 00:35:09 bree kdm: :0[27754]: Fatal X server IO error: Interrupted system call
Video is Intel 82Q963/Q965.

Comment 1 Patrick O'Callaghan 2010-06-04 15:04:59 UTC
Created attachment 421253 [details]
dmesg output

Comment 2 Patrick O'Callaghan 2010-06-04 15:06:12 UTC
Created attachment 421255 [details]
lsmod output

Comment 3 Steven M. Parrish 2010-06-05 02:46:40 UTC
This is a issue which needs to be reported upstream.  Please file a report at
http://bugs.kde.org  

Please add upstream report info to this report.  We will monitor the upstream
report for a resolution.

Thanks for the report



-- 
Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

Comment 4 Rex Dieter 2010-06-05 04:55:10 UTC
By the accounts so far, this is looks to be an X server crash, so let's keep this here for now.

Comment 5 Patrick O'Callaghan 2010-06-05 14:12:07 UTC
(In reply to comment #4)
> By the accounts so far, this is looks to be an X server crash, so let's keep
> this here for now.    

I concur, in fact I probably should have reported it against X, except that the bug doesn't happen with startx and does happen with kdm, so it looks like it's kdm that's tickling it.

Let me know if there's any additional info I can collect, such as stack traces.

Comment 6 Patrick O'Callaghan 2010-06-08 20:41:15 UTC
Oddly enough, the problem seems to have fixed itself. I ran "yum update" and now kdm starts without problems, even though no KDE or X rpms were affected. The only things that might be relevant were glibc and the deja-vu fonts.

Comment 7 Patrick O'Callaghan 2010-06-15 00:27:54 UTC
And it's back. This time there's no error in /var/log/messages, but the KDM session simply doesn't start, leaving me looking at a console. Again, startx works fine.

Comment 8 Patrick O'Callaghan 2010-06-15 23:16:36 UTC
I realized that the default init state was set to 3, so that explains the latest manifestation, i.e. not a bug. Note that the initial report (including the fatal X error) was correct, but hasn't reappeared in this form.

Also, I've no idea how the state got set to 3 as I'm the only one with root access and /etc/inittab has a mod time of over 6 weeks ago. Go figure.

Comment 9 Kevin Kofler 2010-06-15 23:22:15 UTC
Since the original problem appears to be fixed and the new one isn't a bug, let's close this.

Comment 10 Patrick O'Callaghan 2010-06-15 23:28:53 UTC
(In reply to comment #9)
> Since the original problem appears to be fixed and the new one isn't a bug,
> let's close this.    

Sure.


Note You need to log in before you can comment on or make changes to this bug.