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 112958 - Always crash at the every exit.
Always crash at the every exit.
Status: RESOLVED FIXED
Product: seahorse
Classification: Applications
Component: general
0.7.x
Other All
: Normal critical
: ---
Assigned To: Seahorse Maintainer
Seahorse Maintainer
: 113536 122455 128103 128653 141041 (view as bug list)
Depends on:
Blocks:
 
 
Reported: 2003-05-14 08:43 UTC by Jeremy Messenger
Modified: 2004-12-22 21:47 UTC
See Also:
GNOME target: ---
GNOME version: ---


Description Jeremy Messenger 2003-05-14 08:43:43 UTC 
Info: seahorse 0.7.3 on Gnome 2.3.1 with FreeBSD 5.0-CURRENT

It always crash at the every exit, which I just run and exit it. Here's 
info that I ran it under the gdb:

====================================
(gdb) r
Starting program: /usr/X11R6/bin/seahorse
 Program received signal SIGBUS, Bus error.
0x08059bbc in seahorse_context_destroy (sctx=0xd0d0d0d0)
    at seahorse-context.c:300
300     seahorse-context.c: No such file or directory.
        in seahorse-context.c
 
(gdb) bt

+ Trace 36858

  • #0 seahorse_context_destroy
    at seahorse-context.c line 300
  • #1 quit
    at seahorse-key-manager.c line 52
  • #2 delete_event
    at seahorse-key-manager.c line 60
  • #3 _gtk_marshal_BOOLEAN__BOXED
    at gtkmarshalers.c line 82
  • #4 g_closure_invoke
    at gclosure.c line 437
  • #5 signal_emit_unlocked_R
    at gsignal.c line 2822
  • #6 g_signal_emit_valist
    at gsignal.c line 2564
  • #7 g_signal_emit
    at gsignal.c line 2612
  • #8 gtk_widget_event_internal
    at gtkwidget.c line 3143
  • #9 gtk_widget_event
  • #10 gtk_main_do_event
    at gtkmain.c line 1437
  • #11 gdk_event_dispatch
    at gdkevents-x11.c line 2018
  • #12 g_main_dispatch
    at gmain.c line 1653
  • #13 g_main_context_dispatch
  • #14 g_main_context_iterate
    at gmain.c line 2278
  • #15 g_main_loop_run
    at gmain.c line 2498
  • #16 gtk_main
    at gtkmain.c line 1092
  • #17 main
    at main.c line 207
  • #18 _start 

Joe Marcus Clarke <marcus@FreeBSD.org> help with me and here's his hints. 
Let me copy and paste in here.

====================================
I thought it might be trying to double-free the sctx, but, in
fact, it looks like they have a bit of a race condition when quitting
the app.  They unref the main widget, then try to destroy the context it
contained.  If you comment out line 288 in
libseahorse/seahorse-widget.c, the problem goes away.  File this with
the developers if a bug doesn't exist already.  They either need to not
destroy the context, or fix line 545 of src/seahorse-key-manager.c not
to use seahorse_widget_destroy for the data_destroy...or something else
that tidies this up.
====================================
Comment 1 Jacob Perkins 2003-05-14 17:05:41 UTC 
Thanks for the heads up, I'm currently trying to refactor the context 
partly to fix issues like this.
Comment 2 Jacob Perkins 2003-09-20 20:15:03 UTC 
*** Bug 122455 has been marked as a duplicate of this bug. ***
Comment 3 Elijah Newren 2003-12-08 04:32:00 UTC 
*** Bug 128653 has been marked as a duplicate of this bug. ***
Comment 4 Elijah Newren 2003-12-08 04:33:11 UTC 
Crash in latest duplicate was on SuSE; changing OS to all.  Since it's
a crasher, I'm also going to change severity->critical, and I'll also
attach the bugsquad keyword.
Comment 5 Samuel Stringham 2004-10-04 00:05:22 UTC 
*** Bug 128103 has been marked as a duplicate of this bug. ***
Comment 6 Samuel Stringham 2004-10-04 00:05:35 UTC 
*** Bug 113536 has been marked as a duplicate of this bug. ***
Comment 7 Samuel Stringham 2004-10-04 00:12:53 UTC 
*** Bug 141041 has been marked as a duplicate of this bug. ***
Comment 8 Stef Walter 2004-10-08 23:00:49 UTC 
Can anyone duplicate this in a recent CVS build? I reviewed the relevant
portions of code and I believe that the behavior has been fixed due to code
cleanup.
Comment 9 Stef Walter 2004-10-14 18:27:41 UTC 
This seems to be fixed. There's a check in seahorse_context_destroy that makes
sure that we don't destroy twice. If it continues to occur please reopen this bug.