Bug 1034500

Summary: [abrt] network-manager-applet-0.9.9.0-7.git20131028.fc21: g_object_finalize: Process /usr/bin/nm-applet was killed by signal 5 (SIGTRAP)
Product: [Fedora] Fedora Reporter: Christopher Meng <i>
Component: network-manager-appletAssignee: Dan Winship <danw>
Status: CLOSED UPSTREAM QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: rawhideCC: danw, dcbw, jklimes, thaller
Target Milestone: ---   
Target Release: ---   
Hardware: i686   
OS: Unspecified   
URL: https://retrace.fedoraproject.org/faf/reports/bthash/f125ed96a4785dc6c401b0a6c799d1f2d67644b0
Whiteboard: abrt_hash:7cffefac38b76b2ec03a697c33335053d965c4ca
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-02-11 04:52:05 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
File: backtrace
none
File: cgroup
none
File: core_backtrace
none
File: dso_list
none
File: environ
none
File: limits
none
File: maps
none
File: open_fds
none
File: proc_pid_status
none
File: var_log_messages
none
patch none

Description Christopher Meng 2013-11-26 01:27:34 UTC 
Version-Release number of selected component:
network-manager-applet-0.9.9.0-7.git20131028.fc21

Additional info:
reporter:       libreport-2.1.9
backtrace_rating: 4
cmdline:        /bin/nm-applet
crash_function: g_object_finalize
executable:     /usr/bin/nm-applet
kernel:         3.12.0-1.fc21.i686
runlevel:       N 5
type:           CCpp
uid:            0

Truncated backtrace:
Thread no. 1 (6 frames)
 #2 g_object_finalize at gobject.c:1030
 #4 constructor at applet.c:3727
 #5 g_object_new_with_custom_constructor at gobject.c:1643
 #6 g_object_new_internal at gobject.c:1720
 #7 g_object_new_valist at gobject.c:1978
 #9 nm_applet_new at applet.c:3851
Comment 1 Christopher Meng 2013-11-26 01:27:45 UTC 
Created attachment 828990 [details]
File: backtrace
Comment 2 Christopher Meng 2013-11-26 01:27:49 UTC 
Created attachment 828991 [details]
File: cgroup
Comment 3 Christopher Meng 2013-11-26 01:27:54 UTC 
Created attachment 828993 [details]
File: core_backtrace
Comment 4 Christopher Meng 2013-11-26 01:27:59 UTC 
Created attachment 828994 [details]
File: dso_list
Comment 5 Christopher Meng 2013-11-26 01:28:06 UTC 
Created attachment 828995 [details]
File: environ
Comment 6 Christopher Meng 2013-11-26 01:28:10 UTC 
Created attachment 828996 [details]
File: limits
Comment 7 Christopher Meng 2013-11-26 01:28:16 UTC 
Created attachment 828997 [details]
File: maps
Comment 8 Christopher Meng 2013-11-26 01:28:21 UTC 
Created attachment 828998 [details]
File: open_fds
Comment 9 Christopher Meng 2013-11-26 01:28:25 UTC 
Created attachment 828999 [details]
File: proc_pid_status
Comment 10 Christopher Meng 2013-11-26 01:28:29 UTC 
Created attachment 829000 [details]
File: var_log_messages
Comment 11 Dan Winship 2013-11-26 14:20:22 UTC 
Created attachment 829301 [details]
patch
Comment 12 Dan Williams 2013-11-26 23:08:20 UTC 
(In reply to Dan Winship from comment #11)
> Created attachment 829301 [details]
> patch

Patch looks good; also tested normal (works) and without info.ui (correctly fails with warning  message and no glib spew).
Comment 13 Jirka Klimes 2013-11-27 08:41:11 UTC 
The patch looks good, just fix indentation in #ifdef BUILD_MIGRATION_TOOL block.
Comment 14 Thomas Haller 2013-11-27 15:57:35 UTC 
(Looks good to me too)



Just wondering, maybe the code in "#ifdef BUILD_MIGRATION_TOOL" should check

  if (g_file_test (argv[0], G_FILE_TEST_IS_EXECUTABLE)) {
     ...

and only to spawn the process in that case? Also, isn't it a bit unthrifty to spawn the migration process at every start?
Comment 15 Dan Williams 2013-11-27 16:13:38 UTC 
(In reply to Thomas Haller from comment #14)
> (Looks good to me too)
> 
> 
> 
> Just wondering, maybe the code in "#ifdef BUILD_MIGRATION_TOOL" should check
> 
>   if (g_file_test (argv[0], G_FILE_TEST_IS_EXECUTABLE)) {
>      ...
> 
> and only to spawn the process in that case? Also, isn't it a bit unthrifty
> to spawn the migration process at every start?

Yeah, it is, but we don't want to link the main applet to GConf at all, and that would be required to do even a small check to see if migration happened.

But also, we don't build the migration stuff in Fedora 20 or later, or in RHEL7.  The last time Fedora shipped with NM 0.8.x, which would require the migration on upgrade, was Fedora 14 (May 2011), and anyone upgrading from F14 to F20 can just re-create their network settings...
Comment 16 Dan Winship 2013-12-05 19:08:12 UTC 
committed upstream