Version-Release number of selected component: NetworkManager-0.9.9.0-29.git20140131.fc20 Additional info: reporter: libreport-2.1.12 backtrace_rating: 4 cmdline: nmcli d connect em1 crash_function: _g_log_abort executable: /usr/bin/nmcli kernel: 3.11.10-301.fc20.x86_64 runlevel: N 5 type: CCpp uid: 0 Truncated backtrace: Thread no. 1 (7 frames) #2 _g_log_abort at gmessages.c:255 #5 connect_device_cb at devices.c:1342 #6 activate_info_complete at nm-client.c:478 #7 recheck_pending_activations at nm-client.c:529 #13 g_object_notify_by_spec_internal at gobject.c:1141 #14 g_object_notify at gobject.c:1183 #15 deferred_notify_cb at nm-object.c:515
Created attachment 862269 [details] File: backtrace
Created attachment 862270 [details] File: cgroup
Created attachment 862271 [details] File: core_backtrace
Created attachment 862272 [details] File: dso_list
Created attachment 862273 [details] File: environ
Created attachment 862274 [details] File: limits
Created attachment 862275 [details] File: maps
Created attachment 862277 [details] File: open_fds
Created attachment 862278 [details] File: proc_pid_status
Created attachment 862279 [details] File: var_log_messages
It has been fixed upstream: 8472180 cli: fix crash if active connection doesn't have a device (rh #1064333) The code will propagate to Fedora in a next NM build.
The NetworkManager-0.9.9.0-29.git20140131.fc20 was a testing snapshot of upstream, but the code was later removed from Fedora 20. So the later Fedora 20 releases don't contain the problematic code. The rawhide releases are fixed.