Description of problem: Changed settings Version-Release number of selected component: evolution-3.11.90-2.fc21 Additional info: reporter: libreport-2.1.12 backtrace_rating: 4 cmdline: evolution crash_function: apply_and_free_fixups executable: /usr/bin/evolution kernel: 3.14.0-0.rc4.git1.1.fc21.x86_64 runlevel: N 5 type: CCpp uid: 1000 Truncated backtrace: Thread no. 1 (10 frames) #5 apply_and_free_fixups at dbus-marshal-recursive.c:113 #6 replacement_block_replace at dbus-marshal-recursive.c:1227 #7 reader_set_basic_variable_length at dbus-marshal-recursive.c:1300 #8 _dbus_type_reader_set_basic at dbus-marshal-recursive.c:1394 #9 set_basic_field at dbus-marshal-header.c:371 #10 _dbus_header_set_field_basic at dbus-marshal-header.c:1291 #11 _dbus_message_iter_close_signature at dbus-message.c:2465 #13 dbus_message_iter_append_basic at dbus-message.c:2671 #14 emit_event at event.c:505 #15 state_event_listener at event.c:688
Created attachment 869376 [details] File: backtrace
Created attachment 869377 [details] File: cgroup
Created attachment 869378 [details] File: core_backtrace
Created attachment 869379 [details] File: dso_list
Created attachment 869380 [details] File: environ
Created attachment 869381 [details] File: limits
Created attachment 869382 [details] File: maps
Created attachment 869383 [details] File: open_fds
Created attachment 869384 [details] File: proc_pid_status
Created attachment 869385 [details] File: var_log_messages
*** Bug 1086840 has been marked as a duplicate of this bug. ***
*** Bug 1087039 has been marked as a duplicate of this bug. ***
I've seen this several times in a row, each time when disconnecting my network cable. I had applied upgrades without restarting, so there may have been a discrepancy between libraries.
Actually I've seen it again when unplugging the cable, after having restarted the machine. So it should be worth investigating.
(In reply to Milan Bouchet-Valat from comment #14) > Actually I've seen it again when unplugging the cable, after having > restarted the machine. So it should be worth investigating. Are you running rawhide, or Fedora 20 with COPR repo with evolution 3.12.x?
Stock Fedora 20, with Evo 3.10.4, contrary to Igor (you have the details on the duplicate reports).
Aha, after reading the backtrace more carefully, evolution does gtk calls in other than the main thread, which can lead to memory corruption. Here a change for a mail account had been received from an evolution-source-registry process and an attempt to update account's editor UI was made immediately, which causes the gtk call being done in a dedicated thread, instead in the main thread. I moved this upstream as [1]. Please see it for any further updates. [1] https://bugzilla.gnome.org/show_bug.cgi?id=728414
*** Bug 1130877 has been marked as a duplicate of this bug. ***