Description of problem: No idea; was just typing an email. System was very slow and swappy; perhaps exacerbated a race condition. There are global variables in this code... Version-Release number of selected component: evolution-3.8.5-2.fc19 Additional info: reporter: libreport-2.1.9 backtrace_rating: 4 cmdline: evolution crash_function: call_old_file_Sync executable: /usr/bin/evolution kernel: 3.11.7-200.fc19.x86_64 runlevel: N 5 type: CCpp uid: 1000 Truncated backtrace: Thread no. 1 (3 frames) #0 call_old_file_Sync at camel-db.c:67 #1 sync_request_thread_cb at camel-db.c:94 #3 g_thread_proxy at gthread.c:798 Potential duplicate: bug 859691
Created attachment 826524 [details] File: backtrace
Created attachment 826525 [details] File: cgroup
Created attachment 826526 [details] File: core_backtrace
Created attachment 826527 [details] File: dso_list
Created attachment 826528 [details] File: environ
Created attachment 826529 [details] File: limits
Created attachment 826530 [details] File: maps
Created attachment 826531 [details] File: open_fds
Created attachment 826532 [details] File: proc_pid_status
Created attachment 826533 [details] File: var_log_messages
Also filed upstream, but it's called "Obsolete" there, while in Fedora it is still the latest bleeding-edge, shipped in the very newest stable version of the distribution. So I suppose it needs to be handled here rather than CLOSED/UPSTREAM.
Well, I do not like to duplicate the effort by filling basically the same comments in two bugzillas - it's time consuming and tends to break when I forget of the connection between them. *** This bug has been marked as a duplicate of bug 859691 ***
(In reply to Milan Crha from comment #12) > Well, I do not like to duplicate the effort by filling basically the same > comments in two bugzillas - it's time consuming and tends to break when I > forget of the connection between them. That's fair enough. I'm aware that we fulfil more than one rôle — sometimes acting as 'upstream' and sometimes 'downstream'. But in this case upstream has clearly marked 3.8 as obsolete, while Fedora cannot tolerate that since it's still the latest release we've shipped. So perhaps we should put on our 'downstream' hat (fedora) and argue with our 'upstream' hat that it would be better not to mark 3.8 as 'obsolete' in GNOME bugzilla just yet... :)
:) My point would be more about "even the bug was found/noticed in 3.8, maybe it's still relevant on 3.10", because I didn't find any directly related change in 3.10 which would, even marginally, address the crash.