Re: An interesting deadlock in ORBit2...

[Justin Schoeman <justin expertron co za>]

Michael Meeks wrote:
> Hi Justin,
> 
> On Mon, 2003-12-08 at 07:14, Justin Schoeman wrote:
> 
>>Hmmm... I did not realise that I needed to start up a glib mainloop in 
>>the CORBA client... Is CORBA_ORB_init() sufficient to start the main 
>>loop?  If so, then this is not the problem, as CORBA_ORB_init() is 
>>called before any threads are started.
> 
> 
> 	Nope; you need to do either a CORBA_ORB_run in 1 thread, or run a
> g_main loop elsewhere.

Hmmm... Interesting, I suppose that is why you should read the docs, 
instead of hacking the examples... None of the examples use 
CORBA_ORB_run in the CORBA client ;-)

>>BTW - I ran into a race in that last unref in IO thread patch...  It is 
>>rather difficult to trigger, but if the main thread adds another 
>>reference to the connection after the shutdown has been dispatched, then 
>>it will block indefinitely on g_cond_wait...
> 
> 
> 	Can you show me a stack trace for this; I just can't see how that could
> happen. OTOH, I've added some extra locking to the main thread unref
> case, since [ I suppose ] we're not certain it's the last unref anymore
> by the time we do the g_object_unref in the worker thread, so we have to
> check again :-)

Unfortunately I was in a bit of a hurry, and did not have time to save 
the dump. I am sure it will occur again though, and I will save it next 
time :-)

Thanks!

-justin