GNOME Bugzilla – Bug 541872
crash in Evolution Mail and Calendar:
Last modified: 2008-11-11 22:44:40 UTC
What were you doing when the application crashed? Distribution: Debian lenny/sid Gnome Release: 2.22.3 2008-06-30 (Debian) BugBuddy Version: 2.22.0 System: Linux 2.6.25-2-686 #1 SMP Fri Jun 27 03:23:20 UTC 2008 i686 X Vendor: The X.Org Foundation X Vendor Release: 10402000 Selinux: No Accessibility: Disabled GTK+ Theme: Clearlooks Icon Theme: gnome Memory status: size: 112824320 vsize: 112824320 resident: 42639360 share: 25518080 rss: 42639360 rss_rlim: 4294967295 CPU usage: start_time: 1215431733 rtime: 677 utime: 612 stime: 65 cutime:0 cstime: 0 timeout: 0 it_real_value: 0 frequency: 100 Backtrace was generated from '/usr/bin/evolution' [Thread debugging using libthread_db enabled] [New Thread 0xb6676740 (LWP 5538)] [New Thread 0xb26b6b90 (LWP 5584)] [New Thread 0xb36b8b90 (LWP 5583)] 0xb7ee9424 in __kernel_vsyscall ()
+ Trace 202219
Thread 1 (Thread 0xb6676740 (LWP 5538))
----------- .xsession-errors (28 sec old) --------------------- Parameter name: path (evolution:5538): GLib-GObject-WARNING **: instance of invalid non-instantiatable type `(null)' (evolution:5538): GLib-GObject-CRITICAL **: g_signal_handlers_disconnect_matched: assertion `G_TYPE_CHECK_INSTANCE (instance)' failed (evolution:5538): GLib-GObject-WARNING **: instance of invalid non-instantiatable type `(null)' (evolution:5538): GLib-GObject-CRITICAL **: g_signal_handlers_disconnect_matched: assertion `G_TYPE_CHECK_INSTANCE (instance)' failed (evolution:5538): GLib-GObject-WARNING **: instance of invalid non-instantiatable type `(null)' (evolution:5538): GLib-GObject-CRITICAL **: g_signal_handlers_disconnect_matched: assertion `G_TYPE_CHECK_INSTANCE (instance)' failed RSS Plugin enabled (evolution 2.22, evolution-rss 0.1.0) RSS: cleaning all remaining sessions ...done --------------------------------------------------
doesn't look like crash in evolution ...some hints of rss plugin so moving it there
*** Bug 542509 has been marked as a duplicate of this bug. ***
*** Bug 542579 has been marked as a duplicate of this bug. ***
*** Bug 542521 has been marked as a duplicate of this bug. ***
*** Bug 542524 has been marked as a duplicate of this bug. ***
*** Bug 542703 has been marked as a duplicate of this bug. ***
*** Bug 542776 has been marked as a duplicate of this bug. ***
*** Bug 543133 has been marked as a duplicate of this bug. ***
*** Bug 543353 has been marked as a duplicate of this bug. ***
*** Bug 543512 has been marked as a duplicate of this bug. ***
*** Bug 543921 has been marked as a duplicate of this bug. ***
*** Bug 543884 has been marked as a duplicate of this bug. ***
*** Bug 543740 has been marked as a duplicate of this bug. ***
Created attachment 114909 [details] [review] temporary fix Apparently crash is in nsSocketTransportService::Shutdown from xulrunner and has something to do with this invoked from main thread or not. Here is a temporary fix for the problem.
see bug 544064
*** Bug 544074 has been marked as a duplicate of this bug. ***
*** Bug 544191 has been marked as a duplicate of this bug. ***
*** Bug 544417 has been marked as a duplicate of this bug. ***
*** Bug 544635 has been marked as a duplicate of this bug. ***
*** Bug 544619 has been marked as a duplicate of this bug. ***
*** Bug 544532 has been marked as a duplicate of this bug. ***
*** Bug 544801 has been marked as a duplicate of this bug. ***
*** Bug 544755 has been marked as a duplicate of this bug. ***
*** Bug 544789 has been marked as a duplicate of this bug. ***
*** Bug 544821 has been marked as a duplicate of this bug. ***
*** Bug 544834 has been marked as a duplicate of this bug. ***
*** Bug 544329 has been marked as a duplicate of this bug. ***
*** Bug 543815 has been marked as a duplicate of this bug. ***
*** Bug 543662 has been marked as a duplicate of this bug. ***
*** Bug 542765 has been marked as a duplicate of this bug. ***
*** Bug 543430 has been marked as a duplicate of this bug. ***
*** Bug 542536 has been marked as a duplicate of this bug. ***
*** Bug 542450 has been marked as a duplicate of this bug. ***
*** Bug 542519 has been marked as a duplicate of this bug. ***
*** Bug 542441 has been marked as a duplicate of this bug. ***
*** Bug 542310 has been marked as a duplicate of this bug. ***
*** Bug 544563 has been marked as a duplicate of this bug. ***
*** Bug 544852 has been marked as a duplicate of this bug. ***
Created attachment 115324 [details] [review] better shutdown method This is a better method to handle gecko_shutdown
*** Bug 544879 has been marked as a duplicate of this bug. ***
*** Bug 544899 has been marked as a duplicate of this bug. ***
*** Bug 544912 has been marked as a duplicate of this bug. ***
*** Bug 544941 has been marked as a duplicate of this bug. ***
*** Bug 545028 has been marked as a duplicate of this bug. ***
*** Bug 544995 has been marked as a duplicate of this bug. ***
*** Bug 545080 has been marked as a duplicate of this bug. ***
*** Bug 545105 has been marked as a duplicate of this bug. ***
*** Bug 545257 has been marked as a duplicate of this bug. ***
*** Bug 545396 has been marked as a duplicate of this bug. ***
*** Bug 545416 has been marked as a duplicate of this bug. ***
*** Bug 545425 has been marked as a duplicate of this bug. ***
*** Bug 545588 has been marked as a duplicate of this bug. ***
*** Bug 545628 has been marked as a duplicate of this bug. ***
*** Bug 545760 has been marked as a duplicate of this bug. ***
*** Bug 545755 has been marked as a duplicate of this bug. ***
*** Bug 545736 has been marked as a duplicate of this bug. ***
*** Bug 544064 has been marked as a duplicate of this bug. ***
*** Bug 545719 has been marked as a duplicate of this bug. ***
*** Bug 545680 has been marked as a duplicate of this bug. ***
*** Bug 545788 has been marked as a duplicate of this bug. ***
*** Bug 545789 has been marked as a duplicate of this bug. ***
*** Bug 545928 has been marked as a duplicate of this bug. ***
*** Bug 545907 has been marked as a duplicate of this bug. ***
*** Bug 545906 has been marked as a duplicate of this bug. ***
*** Bug 545905 has been marked as a duplicate of this bug. ***
*** Bug 545939 has been marked as a duplicate of this bug. ***
*** Bug 546002 has been marked as a duplicate of this bug. ***
*** Bug 546024 has been marked as a duplicate of this bug. ***
*** Bug 546041 has been marked as a duplicate of this bug. ***
*** Bug 546151 has been marked as a duplicate of this bug. ***
*** Bug 546103 has been marked as a duplicate of this bug. ***
*** Bug 546164 has been marked as a duplicate of this bug. ***
>--- src/gecko-utils.cpp (revision 342) >+++ src/gecko-utils.cpp (working copy) >@@ -97,7 +97,13 @@ > NS_IF_RELEASE (gPrefBranch); > gPrefBranch = nsnull; > >+#ifdef XPCOM_GLUE >+ XPCOMGlueShutdown(); >+ NS_ShutdownXPCOM (nsnull); >+ PR_ProcessExit (0); I think pop_startup + XPCOMGlueShutdown (in that order) should be enough here. Did you try that? Of course, if you dont call push_startup for the XPCOM_GLUE you wont need it; however XPCOMGlueShutdown probably has to be called _after_ the NS_ShutdownXPCOM call.
*** Bug 546287 has been marked as a duplicate of this bug. ***
*** Bug 546246 has been marked as a duplicate of this bug. ***
*** Bug 546356 has been marked as a duplicate of this bug. ***
*** Bug 546472 has been marked as a duplicate of this bug. ***
*** Bug 546580 has been marked as a duplicate of this bug. ***
*** Bug 546581 has been marked as a duplicate of this bug. ***
*** Bug 546667 has been marked as a duplicate of this bug. ***
*** Bug 546509 has been marked as a duplicate of this bug. ***
*** Bug 546536 has been marked as a duplicate of this bug. ***
*** Bug 546709 has been marked as a duplicate of this bug. ***
Would someone please be so kind to address Alexander's concerns? In my experience, pop_shutdown() causes the crash on exit, but we want some kind of clean shutdown to serialize the profile on disk.
*** Bug 546773 has been marked as a duplicate of this bug. ***
*** Bug 546827 has been marked as a duplicate of this bug. ***
*** Bug 547002 has been marked as a duplicate of this bug. ***
*** Bug 546928 has been marked as a duplicate of this bug. ***
*** Bug 547001 has been marked as a duplicate of this bug. ***
*** Bug 547115 has been marked as a duplicate of this bug. ***
*** Bug 546858 has been marked as a duplicate of this bug. ***
*** Bug 547238 has been marked as a duplicate of this bug. ***
*** Bug 547436 has been marked as a duplicate of this bug. ***
*** Bug 547394 has been marked as a duplicate of this bug. ***
*** Bug 547529 has been marked as a duplicate of this bug. ***
*** Bug 547591 has been marked as a duplicate of this bug. ***
*** Bug 547897 has been marked as a duplicate of this bug. ***
> I think pop_startup + XPCOMGlueShutdown (in that order) should be enough here. > Did you try that? This is exactly what this bug is about. calling pop_startup crashes evolution. > In my > experience, pop_shutdown() causes the crash on exit, but we want some kind of > clean shutdown to serialize the profile on disk. Latest proposed patch cleanly shuts down gecko. +#ifdef XPCOM_GLUE + XPCOMGlueShutdown(); + NS_ShutdownXPCOM (nsnull); + PR_ProcessExit (0); You can check $HOME/.evolution/mail/rss/mozembed-rss.
The PR_ProcessExit doesn't sound too nice, especially given that evolution-rss registers its shutdown in an atexit handler. I understand it's like exit(0); is it really needed?
(In reply to comment #99) > > I understand it's like exit(0); is it really needed? > Unfortunately that is required to avoid another problem with XPCOM and camel_shutdown. http://bugzilla.gnome.org/show_bug.cgi?id=548325
*** Bug 548531 has been marked as a duplicate of this bug. ***
*** Bug 548921 has been marked as a duplicate of this bug. ***
*** Bug 548968 has been marked as a duplicate of this bug. ***
*** Bug 549123 has been marked as a duplicate of this bug. ***
*** Bug 549317 has been marked as a duplicate of this bug. ***
*** Bug 549800 has been marked as a duplicate of this bug. ***
*** Bug 549868 has been marked as a duplicate of this bug. ***
*** Bug 549817 has been marked as a duplicate of this bug. ***
*** Bug 550233 has been marked as a duplicate of this bug. ***
*** Bug 550382 has been marked as a duplicate of this bug. ***
*** Bug 550674 has been marked as a duplicate of this bug. ***
*** Bug 547812 has been marked as a duplicate of this bug. ***
*** Bug 550893 has been marked as a duplicate of this bug. ***
*** Bug 551129 has been marked as a duplicate of this bug. ***
*** Bug 551174 has been marked as a duplicate of this bug. ***
*** Bug 551493 has been marked as a duplicate of this bug. ***
*** Bug 552530 has been marked as a duplicate of this bug. ***
*** Bug 552686 has been marked as a duplicate of this bug. ***
*** Bug 552692 has been marked as a duplicate of this bug. ***
*** Bug 553293 has been marked as a duplicate of this bug. ***
*** Bug 553236 has been marked as a duplicate of this bug. ***
*** Bug 553398 has been marked as a duplicate of this bug. ***
*** Bug 553858 has been marked as a duplicate of this bug. ***
*** Bug 553881 has been marked as a duplicate of this bug. ***
*** Bug 554190 has been marked as a duplicate of this bug. ***
*** Bug 554475 has been marked as a duplicate of this bug. ***
*** Bug 554747 has been marked as a duplicate of this bug. ***
*** Bug 555019 has been marked as a duplicate of this bug. ***
*** Bug 555026 has been marked as a duplicate of this bug. ***
*** Bug 555081 has been marked as a duplicate of this bug. ***
*** Bug 555245 has been marked as a duplicate of this bug. ***
*** Bug 555434 has been marked as a duplicate of this bug. ***
*** Bug 555638 has been marked as a duplicate of this bug. ***
can we get a fix soon please, so that at least we can commit in stable 2.24.0, we get so many dups per day :(
*** Bug 555812 has been marked as a duplicate of this bug. ***
*** Bug 556168 has been marked as a duplicate of this bug. ***
*** Bug 556595 has been marked as a duplicate of this bug. ***
*** Bug 553439 has been marked as a duplicate of this bug. ***
*** Bug 556882 has been marked as a duplicate of this bug. ***
*** Bug 556871 has been marked as a duplicate of this bug. ***
*** Bug 554909 has been marked as a duplicate of this bug. ***
*** Bug 557088 has been marked as a duplicate of this bug. ***
Debian packages that fixes this issue: i386 - http://gnome.eu.org/evo/debian/evolution-rss_0.1.0-1xul_i386.deb x86_64 - http://gnome.eu.org/evo/debian/evolution-rss_0.1.0-1xul_amd64.deb
*** Bug 557291 has been marked as a duplicate of this bug. ***
*** Bug 557503 has been marked as a duplicate of this bug. ***
*** Bug 557754 has been marked as a duplicate of this bug. ***
Lucian: Are they officially available in Debian as updates?
Unfortunately there was no answer from debian (patch submitted 3 months ago) so I decided to make those packages available on evolution-rss homepage: http://gnome.eu.org/evo/index.php/Debian_Packages Luckily ubuntu is not affected by those packages and soon there is going to be a new release which includes the patch.
Lucian: where did you contact us? someone directed us to this patch now ( http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=491703#31 ), and that is the first I've heard of the fix. I have to admit I've been to busy lately myself, to search out solutions for this and other stuff away from our mailing-list/bts but if it is available for all of the debian evo-team some of us should be able to grab it. I will try to get this ready for Lenny during the weekend. cheers
Lucian: Sorry, my bad. I hadn't updated that directory from subversion and didn't know that the patch was commited :( Will build and upload to debian tomorrow.
there is an official update for this issue evolution-rss (0.1.0-2) thanks Heikki .. closing this.
*** Bug 557750 has been marked as a duplicate of this bug. ***
*** Bug 559629 has been marked as a duplicate of this bug. ***
*** Bug 559574 has been marked as a duplicate of this bug. ***
*** Bug 559983 has been marked as a duplicate of this bug. ***
*** Bug 560384 has been marked as a duplicate of this bug. ***