After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 101857 - GnomeClient signal die emitted spuriously.
GnomeClient signal die emitted spuriously.
Status: RESOLVED OBSOLETE
Product: gnome-session
Classification: Core
Component: gnome-session
2.2.x
Other other
: Normal normal
: ---
Assigned To: Session Maintainers
Session Maintainers
Depends on:
Blocks:
 
 
Reported: 2002-12-23 15:43 UTC by Matthew Tuck
Modified: 2014-12-15 21:10 UTC
See Also:
GNOME target: ---
GNOME version: 2.9/2.10

Attachments
Test case demonstrating the problem. (608 bytes, text/plain)
2002-12-23 15:45 UTC, Matthew Tuck 		Details

Description Matthew Tuck 2002-12-23 15:43:58 UTC 
So anyway, I was having trouble running my Bakery application under
Valgrind.  When under Valgrind, a window would pop up and then promptly
disappear, terminating the run.

10 hours later I had a test case and discovered a lot more about the problem.

As I reduced the test case, it didn't happen as often, it would sometimes
happen and sometimes not, on successive runs.

Do you know how hard it is to reduce a test case when the problem
disappears for no good reason, and you're not aware of it yet?

This behaviour and the fact it is only a problem under Valgrind loudly
suggests a race condition or timeout.  And yes, with appropriately placed
sleeps, you can make it occur outside Valgrind.

It seems to me that the GnomeClient signal die gets emitted if a
sufficiently large time occurs between doing some things.  From what I can
determine, this shouldn't happen, so I'm filing here.

I have reduced the test case to libgnomeui and will attach it.

I'm running the Debian package 2.0.6-1 of libgnomeui2.
Comment 1 Matthew Tuck 2002-12-23 15:45:03 UTC 
Created attachment 13188 [details]
Test case demonstrating the problem.
Comment 2 Per Kristian Gjermshus 2002-12-23 16:34:35 UTC 
Confirmed on Debian with libgnomeui 2.0.6. 
Seems to work on RH8 with libgnomeui 2.0.3.
Comment 3 Murray Cumming 2002-12-23 16:54:54 UTC 
Please tell us what the problem is and exactly how to reproduce it.
Comment 4 Matthew Tuck 2002-12-24 12:02:46 UTC 
MurrayC, I explained the problem in my first paragraph, and I don't
know how much more exact you can get than a test case.
Comment 5 Kjartan Maraas 2003-05-12 22:27:37 UTC 
Havoc, could you take a look at this?
Comment 6 Havoc Pennington 2003-05-12 22:35:09 UTC 
gnome-session times out apps and kills them after a while.
gnome-session rules! woohoo!

See
http://mail.gnome.org/archives/desktop-devel-list/2003-April/msg00090.html

Now is your chance to save us. ;-)

Anyway, not a libgnomeui bug, you could reassign to gnome-session, 
but the right behavior here is probably to add some nice dialog and 
so forth, and gnome-session is not easy to add that sort of thing 
to.
Comment 7 Anders Carlsson 2003-05-13 08:27:51 UTC 
IIRC, the Red Hat RPMS have a patch that increases the default timeout
value to 5 minutes instead of 30 seconds. Might be a good idea to
change that in the stock version too.
Comment 8 Kjartan Maraas 2003-11-30 20:39:01 UTC 
Moving to gnome-session then.
Comment 9 Lucas Rocha 2008-04-10 10:14:59 UTC 
This test case works well with gnome-session 2.22.x because the purge_delay is now 2 minutes. Need to check that on new code base.