Red Hat Bugzilla – Bug 991539
[abrt] evolution-data-server-3.8.4-1.fc19: block_source: Process /usr/libexec/evolution-addressbook-factory was killed by signal 11 (SIGSEGV)
Last modified: 2014-06-18 12:40:52 EDT
Description of problem:
I read and closed a message on my IMAP server. I _might_ have moved from the sent folder to Inbox, I am not certain, but it's nothing that I don't do a 100 times a day. Then the address books crashed.
Version-Release number of selected component:
runlevel: N 5
Thread no. 1 (7 frames)
#0 block_source at gmain.c:2947
#6 dbus_server_run_server at e-dbus-server.c:222
#7 ffi_call_unix64 at ../src/x86/unix64.S:76
#8 ffi_call at ../src/x86/ffi64.c:522
#9 g_cclosure_marshal_generic_va at gclosure.c:1550
#10 _g_closure_invoke_va at gclosure.c:840
#13 e_dbus_server_run at e-dbus-server.c:414
Created attachment 782073 [details]
Created attachment 782074 [details]
Created attachment 782075 [details]
Created attachment 782076 [details]
Created attachment 782077 [details]
Created attachment 782078 [details]
Created attachment 782079 [details]
Created attachment 782080 [details]
Created attachment 782081 [details]
Created attachment 782082 [details]
Thanks for a bug report. The backtrace doesn't show much, most threads are idle, only the main thread crashed in GLib code, which doesn't necessarily mean the issue is there, because with memory corruption (if this is it) it can strike anywhere. I see in the backtrace that you use evolution-ews, there is one soup thread for it and two delta threads, but as I said, they are currently idle, not doing anything.
It seems to me that you've updated evolution-data-server package, but the background processes were not restarted yet (this is not forced in RPM), which might or might not be related. I think so because of the first line in the backtrace:
> warning: core file may not match specified executable file.
I would try to restart the machine, and keep using it for some time. Then we'll see whether you get the crash again.
With respect of a message selection, if you've configured evolution to search for contact photos, then the most of the message selections fires a request to the address books to find the sender.
I have updated to F20 so this is no longer a testable issue to me.