GNOME Bugzilla – Bug 494070
evolution crashes on startup (segmentation fault)
Last modified: 2009-03-16 23:27:11 UTC
Steps to reproduce: 1. start evolution from terminal 2. crashes immediately Stack trace:
+ Trace 175661
Other information: I reainstalled Ubuntu 6.10 and immediately updated to 7.04. (Before Evolution had worked fine under 7.04.) There is no difference, if I rename the directory ".evolution" and also if I try to start e. g. only the calendar or the mail with "evolution --component=mail". The output to the terminal is the following: (evolution-2.10:8000): GLib-GObject-WARNING **: instance of invalid non-instantiatable type `(null)' (evolution-2.10:8000): GLib-GObject-CRITICAL **: g_signal_connect_data: assertion `G_TYPE_CHECK_INSTANCE (instance)' failed (evolution-2.10:8000): e-data-server-CRITICAL **: e_source_group_peek_base_uri: assertion `E_IS_SOURCE_GROUP (group)' failed Program received signal SIGSEGV, Segmentation fault.
there's also a bug reported in Ubuntu here: https://bugs.launchpad.net/ubuntu/+source/evolution/+bug/161071
*** Bug 488772 has been marked as a duplicate of this bug. ***
*** Bug 495600 has been marked as a duplicate of this bug. ***
*** Bug 511505 has been marked as a duplicate of this bug. ***
*** Bug 515700 has been marked as a duplicate of this bug. ***
Program received signal SIGSEGV, Segmentation fault.
+ Trace 189820
Thread 47075017800800 (LWP 20753)
e_source_group* (gconf access) is what is causing the crash.
*** Bug 517801 has been marked as a duplicate of this bug. ***
(In reply to comment #6) > e_source_group* (gconf access) is what is causing the crash. GConf access? that function doesn't uses gconf at all. Based on the runtime warnings, it seems that the pointer has been freed before used, but the code didn't notice the change. But why the change has been invoked? Maybe some refcount error? I do not see any issue in the code, though. Maybe you tried with version before this revision, which fixes the issue, I believe. http://svn.gnome.org/viewvc/evolution-data-server?view=revision&revision=7730 The bad thing is that the above change is included in eds 1.10.3.1 in F7, where most of the duplicates come from. The patch is included in F7 since 1.10.1-4. I believe it's fixed. Any objections?
Apparently there are no dups. So I am closing as OBSOLETE.