Red Hat Bugzilla – Bug 136518
Logout never exits
Last modified: 2012-06-20 12:05:30 EDT
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; rv:1.7.3) Gecko/20041001
Description of problem:
Both user and root user get stuck use CTL-ALT-DEL/Logout. The desktop
never exits forcing me to use CTL-ALT-Backspace.
This system is a new upgrade from enterprise 3es.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1.login runlevel 5
3.select logout and click the OK button.
Expected Results: Decktop closes and graphic login returns.
I've seen this intermittently on Fedora Core 2, but I haven't hit it
(yet??) with latest rawhide or with FC3 rc1.
If you want to try reproducing this with FC3 rc1, it's available here:
(I know this bug has been reported against RHEL 4 beta1, but FC3 and
RHEL4 are very related right now, so it might be helpful to see if
this bug still exists in FC3.)
I saw this, once, on one of my boxes with a fresh install of a
2004-11-05 rawhide snapshot. I can't reproduce it though. (Also, the
box that it happened on is only 300MHz, for what that's worth.)
Anyway, whatever the problem is, it's *far* less severe with recent
rawhide than with FC2 (with FC2 I was able to hit it somewhat
frequently on a pretty wide range of boxes). If I figure out a way of
reproducing this with recent rawhide, or RHEL 4b2, or FC3, then I'll
add another comment here.
There are several bugs similar to this one on bugzilla, but this seems to be the
one most similar to the problem I've encountered. This occurs on two separate
Fedora Core 3 (fresh install) running Gnome.
Problem: Delays on Log Out
On selecting Action>Log Out, the Log Out dialog box appears. After selecting
Log Out and clicking OK, nothing happens for two to three minutes. Eventually,
the log out occurs.
If you wait for the log out (that is, abstain from killing X with ctrl-alt-bs,
or opening a console with ctrl-alt-F1 and shutting down the machine), then log
in, then select Action>Log Out, nothing happens (no Log Out dialog box) for two
minutes. Then the box appears and log out can proceed immediately.
This symptom occurs on both machines.
The kernel is still running fine. An ssh connection seems to work just fine.
ps ax shows the last task running to be
/usr/bin/esd -terminate -nobeeps -as 2 -spawnfd 28
Fedora Core 3 - fresh install
1. Sony VGN-A270P (laptop)
2 GB RAM
ATI Technologies Inc RV350 [Mobility Radeon 9600 M10]
2. Dell Dimension V433C
384 GB RAM
ATI Technologies Inc 3D Rage Pro AGP 1X/2X (rev 5c)
I note that the commonality between the two machines is the ATI video card. I
wonder if there's a problem with the driver.
It looks like the delay between clicking OK on the Log Out dialog box
to the time the log out or shutdown occurs is exactly 120 seconds - at least
thatÂ´s as close as I can measure it on my wristwatch. This for the case where
the delay occurs between clicking OK and the logout occurs. IÂ´ll time the other
delay (where it occurs between selecting Action>Log Out and the Log Out dialog
box appears). If itÂ´s any different IÂ´ll post that as well.
I think it might be a race condition.
Select Actions>Log Out
BEFORE THE SCREEN FINISHES DIMMING - On the ensuing dialog box, click
Log Out or Shut down
In this case there will be a 120 second delay before the action takes
If you wait for the dimming to complete it will occur immediately.
Re: comments 3, 4, 5
That's not this bug. If this is the most similar bug in RH Bugzilla,
then please file a new bug rather than piggybacking onto this one.
This bug basically boils down to "click OK on the logout dialog, leave
the machine running overnight, come back the next day (or even a few
days later), and logout *still* hasn't finished yet". And IIRC one of
the boxes I've seen this on is a monster box where it's physically
impossible to click *anything* in the Log Out dialog before the screen
finishes dimming, because it dims that fast (and that's at 1920x1200
I have not seen this since the FC3 release, but the box that was most
afflicted by this is still running FC2. (That will be changing later
this month, but until then I won't know if this bug still exists.)
Thank you for submitting this issue for consideration in Red Hat Enterprise Linux. The release for which you requested us to review is now End of Life.
Please See https://access.redhat.com/support/policy/updates/errata/
If you would like Red Hat to re-consider your feature request for an active release, please re-open the request via appropriate support channels and provide additional supporting details about the importance of this issue.