[Pidgin] #9024: delete prefs -> start pidgin -> freeze (infinite loop?)

Pidgin trac at pidgin.im
Thu Apr 23 01:07:18 EDT 2009


#9024: delete prefs -> start pidgin -> freeze (infinite loop?)
------------------------+---------------------------------------------------
 Reporter:  orbisvicis  |        Owner:  lschiere    
     Type:  defect      |       Status:  new         
Milestone:              |    Component:  unclassified
  Version:  2.5.5       |   Resolution:              
 Keywords:              |  
------------------------+---------------------------------------------------

Old description:

> 1) I deleted/renamed the ~/.purple directory
> 2) start pidgin. In approx ~1-2sec
> 3) permanent spike in CPU (pidgin) followed by
> 4) complete freeze, GUI fails to refresh
>
> Attached:
> 1) output of "pidgin -d"
> 2) 4 gdb logs attached to an already running pidgin instance
> 3) 1 gdb log starting pidgin
> 4) update log for that past 5 days
>
> Summary:
> Possibly an infinite loop??
> __kernel_vsyscall -> get_pc_thunk.bx -> strcmp -> strcmp at plt
>
> Background:
> Have been using pidgin 2.5.5 for a while without any problems. This
> abruptly started happening around two days ago. I have no idea why. Now
> it happens all the time.

New description:

 Procedure:

 (1) I deleted/renamed the ~/.purple directory

 (2) start pidgin. In approx ~1-2sec

 (3) permanent spike in CPU (pidgin) followed by

 (4) complete freeze, GUI fails to refresh


 Attached:

 (1) output of "pidgin -d"

 (2) multiple gdb logs attached to an already running pidgin instance

 (3) 1 gdb log starting pidgin

 (4) update log for that past 5 days


 Summary:

 Possibly an infinite loop?? __kernel_vsyscall -> get_pc_thunk.bx -> strcmp
 -> strcmp at plt

 Background:

 Have been using pidgin 2.5.5 for a while without any problems. This new
 behavior abruptly started around two days ago. I have no idea why. Now it
 happens all the time.

--

Comment(by orbisvicis):

 A correction:

 step (3) - permanent spike in CPU - in some rare cases does not happen,
 although step (4) always occurs. I am not sure why sometimes the CPU usage
 remains normal.

 Also, the archive is screwed up, so I'm uploading an augmented revision.

-- 
Ticket URL: <http://developer.pidgin.im/ticket/9024#comment:1>
Pidgin <http://pidgin.im>
Pidgin


More information about the Tracker mailing list