Re: [JBoss-dev] Unable to passivate due to ctx with option c

[Adrian B. <ad...@jb...>]

On Tue, 2003-09-30 at 15:14, Scott M Stark wrote:
> Right, but does this not mean we are leaking cache entries in this
> scenario? This did not appear to be a recursive call so why are there
> more than one lock ref here?
> 

The synchronization takes one when it registers - it releases it
at the end of afterCompletion, a few lines after the cache remove().
The cache takes one during the passivation.

Two lock refs.

Regards,
Adrian

> Adrian Brock wrote:
> 
> > That is because the instance synchronization still holds the lock
> > reference.
> > The cache also takes a lock reference at passivation.
> > 
> > 1+1 == 2 > 1 which is what tryToPassivate checks.
> > 
> > In plain English, the Cache is saying I cannot passivate this
> > instance because somebody else besides me has a lock reference.
> > 
> > The fact that "the somebody else" has told the Cache to remove it
> > is not taken into account.
> > 
> > Regards,
> > Adrian
> > 
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> JBoss-Development mailing list
> JBo...@li...
> https://lists.sourceforge.net/lists/listinfo/jboss-development
-- 
xxxxxxxxxxxxxxxxxxxxxxxx 
Adrian Brock
Director of Support
Back Office
JBoss Group, LLC 
xxxxxxxxxxxxxxxxxxxxxxxx