[Pidgin] #7414: pidgin refuses to be killed after X server dies

Pidgin trac at pidgin.im
Thu Oct 30 02:19:54 EDT 2008


#7414: pidgin refuses to be killed after X server dies
--------------------------+-------------------------------------------------
 Reporter:  acli          |     Owner:  lschiere
     Type:  defect        |    Status:  new     
Component:  unclassified  |   Version:  2.5.2   
 Keywords:                |  
--------------------------+-------------------------------------------------
 I was running pidgin remotely and the box running the X server froze, so I
 turned off the box and restarted. When I got X running again and tried to
 run pidgin, I found that pidgin was still running (unaware that the X
 server died).

 I tried to kill it several times with some "normal" signals (SIGTERM,
 SIGHUP, SIGINT), but it still refuses to die. Then I took a look at the
 log and found this:
 {{{
 (02:07:57) plugins: Unloading plugin MySpaceIM
 (02:07:57) plugins: Unloading plugin GroupWise
 (02:07:57) plugins: Unloading plugin AIM
 (02:07:57) plugins: Unloading plugin ICQ
 (02:07:57) plugins: Unloading plugin QQ
 (02:07:57) plugins: Unloading plugin SIMPLE
 (02:07:57) plugins: Unloading plugin Yahoo
 (02:07:57) plugins: Unloading plugin Zephyr
 (02:07:57) sighandler: Caught signal 15
 (02:07:57) g_log: purple_core_quit: assertion `core != NULL' failed
 (02:07:58) sighandler: Caught signal 15
 (02:07:58) g_log: purple_core_quit: assertion `core != NULL' failed
 (02:07:59) sighandler: Caught signal 15
 (02:07:59) g_log: purple_core_quit: assertion `core != NULL' failed
 (02:08:00) sighandler: Caught signal 15
 (02:08:00) g_log: purple_core_quit: assertion `core != NULL' failed
 (02:08:01) sighandler: Caught signal 15
 (02:08:01) g_log: purple_core_quit: assertion `core != NULL' failed
 (02:08:13) sighandler: Caught signal 1
 (02:08:14) sighandler: Caught signal 1
 (02:08:15) sighandler: Caught signal 1
 (02:08:19) sighandler: Caught signal 2
 (02:08:19) g_log: purple_core_quit: assertion `core != NULL' failed
 }}}
 So it would appear that asserting for "core != NULL" before agreeing to
 exit is wrong, or at least wrong in some cases.

 (I finally killed it successfully with SIGBUS.)

-- 
Ticket URL: <http://developer.pidgin.im/ticket/7414>
Pidgin <http://pidgin.im>
Pidgin


More information about the Tracker mailing list