[libvirt] libvirtd segfault

[Jim Fehlig]

AL13N wrote:
>> AL13N wrote:
>>     
>>> Op woensdag 20 maart 2013 08:42:52 schreef Jim Fehlig:
>>>
>>>       
>>>> AL13N wrote:
>>>>
>>>>         
>>>>> Thread 1 (Thread 0x7fdef683b800 (LWP 20522)):
>>>>> #0  0x0000000000000000 in ?? ()
>>>>> #1  0x00007fdee9a72dc7 in libxl_osevent_occurred_timeout
>>>>> (ctx=<optimized
>>>>> out>, for_libxl=0x7fdedc001608) at libxl_event.c:1039
>>>>> #2  0x00007fdee9c9ff87 in libxlTimerCallback (timer=<optimized out>,
>>>>> timer_info=0x7fdedc001730) at libxl/libxl_driver.c:259
>>>>> #3  0x00007fdef5dd0f1a in virEventPollDispatchTimeouts () at
>>>>> util/vireventpoll.c:450
>>>>> #4  virEventPollRunOnce () at util/vireventpoll.c:643
>>>>> #5  0x00007fdef5dcf88d in virEventRunDefaultImpl () at
>>>>> util/virevent.c:273
>>>>> #6  0x00007fdef5ec96c5 in virNetServerRun (srv=0x1146220) at
>>>>> rpc/virnetserver.c:1108
>>>>> #7  0x000000000040c8e0 in main (argc=<optimized out>, argv=<optimized
>>>>> out>) at libvirtd.c:1481
>>>>>
>>>>>           
>>>> I received a similar report on #xendevel just yesterday.  I'll take a
>>>> look at this in the next few days after finishing my current project.
>>>>
>>>> Regards,
>>>> Jim
>>>>
>>>>         
>>> restesting with libvirt-1.0.4 confirmed that it's still unfixed
>>>
>>>       
>> I would have been really surprised if it was fixed :).
>>     
>
> /me too
>   

FYI, Bamvor fixed a remaining (known) race in the libxl driver, which
I've pushed for libvirt 1.0.5

https://www.redhat.com/archives/libvir-list/2013-April/msg01926.html

I still want to cleanup handling of domain destruction events as
mentioned elsewhere in this thread, but that can wait until support for
jobs is added to the driver.

Regards,
Jim