[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bayonne-devel] DEAD LOCK
From: |
David Sugar |
Subject: |
Re: [Bayonne-devel] DEAD LOCK |
Date: |
Wed, 02 Aug 2006 17:04:22 -0400 |
User-agent: |
Thunderbird 1.5.0.4 (X11/20060614) |
I had found a couple of strange issues in detach/hangup handing in
Bayonne, including one where it can receive and process a STOP_STATE
event even after detaching scripting engine, which then locks the port.
This was the remaining issue I had with the testing of the new dialogic
driver with outbound calling (if dialing fails in gc driver, it may kick
the runtime into a state that requires a line reset which then kicks
back a STOP_STATE in detach....), and is currently fixed in cvs for what
will be 1.2.17. I would like to see this additional issue identified as
well since it sounds similar; a scripting event signal received after
detach().
Julien Chavanton wrote:
> I think I have found a dead-lock in Bayonne.
>
>
>
> It does take place when script1 use the send command to notify script2
> while script2 is in detach somewhere.
>
>
>
>
>
> Script 1:
>
> Aug 2 14:49:18 192.168.0.165/192.168.0.165 bayonne: dx(18): step
> 20000003 send(%childid)
>
>
>
> Script 2:
>
> Is in hangup state more precisely in:
>
> void Trunk::detach(void)
>
> Even more precisely I think it is in: ScriptInterp::detach(); (not sure
> yet but I did set some breakpoint next.)
>
>
>
>
>
>>From here I have to found the mutex causing this lock, I know that when
> this take place the server does not receive more event.
>
>
>
> Lets see if I am in the right direction.
>
> Julien
>
>
>
> _______________________________________________
> Bayonne-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/bayonne-devel