Red Hat Bugzilla – Bug 4846
gdm.pid...is this a bug?
Last modified: 2008-05-01 11:37:51 EDT
Some background: I'm running RedHat 6 on an old DEC Pentium
90 computer with
onboard S3 video.
While working to set up Samba on my computer, I tried to
Enlightenment. The system crashed, and I had no choice but
to power down in
order to reboot. Now it's doing something utterly bizarre.
I'm set to runlevel 5, so it's supposed to go into x
computer tries to do so...or at least I think that's what
it's doing. The
screen just keeps flashing on and off, over and over, for
about a minute.
When it finally stops, this message appears, over and over:
"According to /var/run/gdm.pid, gdm was already running
(697), but seems to
have been murdered mysteriously."
This appears about a dozen times, with different job
numbers each time.
Then at the end:
"INIT: Id "x" respawning too fast: disabled for 5 minutes."
Sure enough, the cycle repeats itself every five minutes.
When I try to go into X manually, I get an error message
Fatal server error:
Cannot open mouse (No such device)
I do have a mouse, and it worked fine up till the system
Now...what does all this mean? Help! Please....?
Have you tried removing /var/run/gdm.pid?
(We're probably moving to a newer version of gdm for the
upcoming release of Red Hat Linux 6.1, which may be the
best way to fix the bug.)
------- Additional Comments From 10/29/99 23:11 -------
This also happens on my AlphaStation 4/200 machine running RedHat 6.0
Linux. Something is killing the X font server (xfs), and gdm can't
start a new session until xfs is restarted. If you log in through
another terminal when this is happening as root and
type "/etc/rc.d/init.d/xfs start" then the problem disappears. (once
anyways). I do not know why xfs is dieing. There is no good
information in the logs.
The ``murdered mysteriously'' message is informal and not critial. gdm
gracefully handles when a pidfile exists but no daemon is running.
I have implemented /sbin/init-like repawn control to my working tree
to avoid restarting X servers.
Regarding the xfs bug: No clue, but I've seen it on several machines.
A compiler bug causes xfs to crash under certain conditions. This one
is fixed in 6.1.