GNOME Bugzilla – Bug 405646
crash in Evolution: Exited Evolution. Hadn't...
Last modified: 2008-03-20 18:11:30 UTC
What were you doing when the application crashed? Exited Evolution. Hadn't done anything more than start it, check a POP account, and quit. Distribution: Unknown Gnome Release: 2.16.3 2007-01-31 (GNOME.Org) BugBuddy Version: 2.16.1 System: Linux 2.6.20 #2 PREEMPT Wed Feb 7 20:05:35 NZDT 2007 i686 X Vendor: The X.Org Foundation X Vendor Release: 70199903 Selinux: No Accessibility: Disabled Memory status: size: 90718208 vsize: 0 resident: 90718208 share: 0 rss: 24891392 rss_rlim: 0 CPU usage: start_time: 1170917730 rtime: 0 utime: 160 stime: 0 cutime:148 cstime: 0 timeout: 12 it_real_value: 0 frequency: 0 Backtrace was generated from '/usr/bin/evolution-2.8' Using host libthread_db library "/lib/libthread_db.so.1". [Thread debugging using libthread_db enabled] [New Thread -1232496960 (LWP 3321)] [New Thread -1274025072 (LWP 3335)] [Thread debugging using libthread_db enabled] [New Thread -1232496960 (LWP 3321)] [New Thread -1274025072 (LWP 3335)] [Thread debugging using libthread_db enabled] [New Thread -1232496960 (LWP 3321)] [New Thread -1274025072 (LWP 3335)] [New Thread -1254880368 (LWP 3334)] [New Thread -1263744112 (LWP 3332)] [New Thread -1238058096 (LWP 3328)] 0xffffe410 in ?? ()
+ Trace 108947
Thread 1 (Thread -1232496960 (LWP 3321))
----------- .xsession-errors --------------------- ** (evolution-2.8:3321): DEBUG: mailto URL command: evolution %s ** (evolution-2.8:3321): DEBUG: mailto URL program: evolution (evolution-2.8:3321): e-utils-WARNING **: can't load plugin '/usr/lib/evolution/2.8/plugins/liborg-gnome-groupwise-features.so: undefined symbol: _' process 3321: The last reference on a connection was dropped without closing the connection. This is a bug in an application. See dbus_connection_unref() documentation for details. Most likely, the application was supposed to call dbus_connection_close(), since this is a private connection. D-Bus not built with -rdynamic so unable to print a backtrace ** (bug-buddy:3340): WARNING **: Couldn't load icon for Bonobo Component Browser ** (bug-buddy:3340): WARNING **: Couldn't load icon for F-Spot Photo Manager ** (bug-buddy:3340): WARNING **: Couldn't load icon for F-Spot Photo Viewer ** (bug-buddy:3340): WARNING **: Couldn't load icon for Open Folder --------------------------------------------------
Thanks for taking the time to report this bug. Unfortunately, that stack trace is missing some elements that will help a lot to solve the problem, so it will be hard for the developers to fix that crash. Can you get us a stack trace with debugging symbols? Please see http://live.gnome.org/GettingTraces for more information on how to do so. Thanks in advance!
Everything here is built from sources, so rebuilding packages with debug symbols will take some time. I'm rebuilding Evolution and E-D-S now - is there anything else which might be required as well?
Ok, disregard the last message - rebuilding relevant packages didn't take as long as I thought it would. Updated stack trace is below... Looks like the problem is something to do with Evolution's use of NetworkManager to monitor network availability. If it matters, this is using NM 0.6.4 and dbus-1.0.2. Backtrace was generated from '/usr/bin/evolution-2.8' Using host libthread_db library "/lib/libthread_db.so.1". [Thread debugging using libthread_db enabled] [New Thread -1232841024 (LWP 6105)] [New Thread -1283040368 (LWP 6122)] [Thread debugging using libthread_db enabled] [New Thread -1232841024 (LWP 6105)] [New Thread -1283040368 (LWP 6122)] [Thread debugging using libthread_db enabled] [New Thread -1232841024 (LWP 6105)] [New Thread -1283040368 (LWP 6122)] [New Thread -1274377328 (LWP 6120)] [New Thread -1265984624 (LWP 6119)] [New Thread -1255228528 (LWP 6115)] [New Thread -1238402160 (LWP 6111)] 0xffffe410 in ?? ()
+ Trace 109243
Thread 6 (Thread -1238402160 (LWP 6111))
I'm hitting the same crasher with latest trunk. An updated stacktrace follows: (gdb) thread apply all bt
+ Trace 113365
Seems to be a problem in libnm: "The last reference on a connection was dropped without closing the connection. This is a bug in an application. See dbus_connection_unref() documentation for details." The problem is that libnm is getting a private connection, which it seems should be closed before unrefered. Moving to NetworkManager.
Created attachment 84286 [details] [review] close the connection before unrefering it This fixes the issue for me.
Just hit this in Fedora Devel evolution-2.10.0-2.fc7.x86_64
Shouldn't it check that the refcount is 1 before calling close? Or can it never be > 1?
Created attachment 85239 [details] [review] Revised patch Comment #5 is correct, but it looks like there's a second place where dbus_connection_close() needs to be called.
hmm, need to get johnp's comments on this... j5?
If you have a private connection you need to close it at some point before the last unref becuase your app needs to be able to respond to the disconnect messages before the object is destroyed. Or something to that effect. In any case libnm should know when it no longer needs the private connection and close it at that point (regaurdless of if you are unreffing at that point though in practice you will always unref after a close).
Should this bug still be in NEEDINFO?
committed to stable, thanks!
*** Bug 424026 has been marked as a duplicate of this bug. ***
This has been fixed already. Thanks for the patch, Matt.
*** Bug 432796 has been marked as a duplicate of this bug. ***
*** Bug 449174 has been marked as a duplicate of this bug. ***
*** Bug 449659 has been marked as a duplicate of this bug. ***
*** Bug 450544 has been marked as a duplicate of this bug. ***
*** Bug 452304 has been marked as a duplicate of this bug. ***
*** Bug 447125 has been marked as a duplicate of this bug. ***
*** Bug 454498 has been marked as a duplicate of this bug. ***
*** Bug 461644 has been marked as a duplicate of this bug. ***
*** Bug 455158 has been marked as a duplicate of this bug. ***
*** Bug 457830 has been marked as a duplicate of this bug. ***
Reopening. The patch in that fixed this bug (comment #9) got removed somehow.
Looks like it's still there on both the stable branch and trunk. Even though this is true I saw that exact crash myself recently with a current rawhide setup. Reported in bug 478992
*** Bug 478992 has been marked as a duplicate of this bug. ***
*** Bug 484452 has been marked as a duplicate of this bug. ***
*** Bug 475393 has been marked as a duplicate of this bug. ***
*** Bug 473508 has been marked as a duplicate of this bug. ***
*** Bug 470472 has been marked as a duplicate of this bug. ***
*** Bug 465528 has been marked as a duplicate of this bug. ***
*** Bug 467771 has been marked as a duplicate of this bug. ***
*** Bug 464531 has been marked as a duplicate of this bug. ***
*** Bug 459958 has been marked as a duplicate of this bug. ***
*** Bug 485200 has been marked as a duplicate of this bug. ***
*** Bug 485883 has been marked as a duplicate of this bug. ***
*** Bug 485924 has been marked as a duplicate of this bug. ***
*** Bug 486400 has been marked as a duplicate of this bug. ***
*** Bug 488224 has been marked as a duplicate of this bug. ***
*** Bug 488530 has been marked as a duplicate of this bug. ***
*** Bug 489697 has been marked as a duplicate of this bug. ***
*** Bug 490285 has been marked as a duplicate of this bug. ***
*** Bug 491634 has been marked as a duplicate of this bug. ***
*** Bug 493400 has been marked as a duplicate of this bug. ***
*** Bug 493393 has been marked as a duplicate of this bug. ***
*** Bug 493852 has been marked as a duplicate of this bug. ***
*** Bug 495760 has been marked as a duplicate of this bug. ***
*** Bug 495353 has been marked as a duplicate of this bug. ***
*** Bug 496338 has been marked as a duplicate of this bug. ***
should be fixed for a while in stable and head.
*** Bug 497514 has been marked as a duplicate of this bug. ***
*** Bug 498731 has been marked as a duplicate of this bug. ***
*** Bug 499236 has been marked as a duplicate of this bug. ***
*** Bug 499143 has been marked as a duplicate of this bug. ***
*** Bug 499079 has been marked as a duplicate of this bug. ***
*** Bug 499310 has been marked as a duplicate of this bug. ***
*** Bug 502134 has been marked as a duplicate of this bug. ***
*** Bug 502838 has been marked as a duplicate of this bug. ***
*** Bug 503830 has been marked as a duplicate of this bug. ***
*** Bug 503994 has been marked as a duplicate of this bug. ***
*** Bug 515333 has been marked as a duplicate of this bug. ***
*** Bug 517187 has been marked as a duplicate of this bug. ***
*** Bug 518335 has been marked as a duplicate of this bug. ***
*** Bug 521474 has been marked as a duplicate of this bug. ***
*** Bug 521852 has been marked as a duplicate of this bug. ***
*** Bug 419885 has been marked as a duplicate of this bug. ***