GNOME Bugzilla – Bug 349816
Marking mail not junk caused hang/crash
Last modified: 2007-10-07 21:53:53 UTC
Subject: Marking mail not junk caused hang/crash Distribution: Debian testing/unstable Package: Evolution Severity: Normal Version: GNOME2.14.2 2.6.3 Gnome-Distributor: Debian Synopsis: Marking mail not junk caused hang/crash Bugzilla-Product: Evolution Bugzilla-Component: Miscellaneous Bugzilla-Version: 2.6.3 BugBuddy-GnomeVersion: 2.0 (2.14.1) Description: Description of the crash: While reading my junk mail folder on a Groupwise backend, I marked a message "not junk". The message disappeared from the Junk Mail folder. I switched back to the Mailbox folder, and the folder was highlighted in the left frame, but the right frame and the preview pane were empy. I clicked on a few more folders, which also did not display. Then evolution crashed and bug buddy launched. Steps to reproduce the crash: 1. Wait for junk mail on a Groupwise system 2. Mark the mail not junk. 3. Try to find the junk mail in the Mailbox folder. 4. The Mailbox folder will not display. 5. Change to other folders that also will not display 5. Evolution crashes in about 30 seconds. Expected Results: I should be able to find the mail marked not junk in the Mailbox folder. Evolution should not hang or crash. How often does this happen? Twice so far. Additional Information: I'll attach the backtrace later. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ------- Bug created by bug-buddy at 2006-08-03 17:49 ------- Unknown version 2.6.3 in product Evolution. Setting version to "2.6.x".
Created attachment 70150 [details] Backtrace of crash
Backtrace was generated from '/usr/bin/evolution-2.6' Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1". [Thread debugging using libthread_db enabled] [New Thread -1500899648 (LWP 1787)] [New Thread -1562543184 (LWP 1863)] [New Thread -1562211408 (LWP 1862)] [New Thread -1546269776 (LWP 1838)] [New Thread -1537483856 (LWP 1831)] [New Thread -1529091152 (LWP 1828)] [New Thread -1520698448 (LWP 1827)] [New Thread -1512109136 (LWP 1826)] 0xffffe410 in __kernel_vsyscall ()
+ Trace 69948
Comment 2 shows the top-most, crashing thread from the attached stacktrace. Probably GTK+ or gtk-engines. Moving over.
I don't think I understand this stack trace. gtk_gc_get takes a depth and colormap, yet the error being raised is during checking a drawable for its depth. The drawable can't be coming from smooth, it must be something in the way gtk handles getting a gc. So I have no idea what could be causing this.
Further Ben pointed out to me - in _gdk_x11_gc_new (drawable=0x80b9550, values=0x8799ca0, in IA__gdk_gc_new_with_values (drawable=0x60001, The drawable is different and(whats with 0x60001).... perhaps some sort of memory corruption caused by evolution? I really don't see how anything Smooth (or GTK+ for that matter), could do in this stack could cause the problem. it really sounds like memory corruption and the rest is just caught up in the middle of it.
Can you get us a stack trace with debugging symbols of the gtk-engines package? Please see http://live.gnome.org/GettingTraces for more information on how to do so.
Closing this bug report as no further information has been provided. Please feel free to reopen this bug if you can provide the information asked for. Thanks!