[RFC][plugin]: fix concurrency problems in PAV.java

[Denis Lila]

> Deepak is right, ConcurrentHashMap doesn't lock on retrieval of value.
> So it is still possible that we could be trying to write and read at
> the same time.
> I propose to wrap the calls to status.get(identifier) in a
> synchronized method (preferably synchronizing on the object being
> affected [status]). Similarly, you would need to change updateStatus
> to synchronize on status as well. This would get rid of the problem
> with read while writing.

It doesn't lock, but that's ok. Concurrent reads and writes are
safe with ConcurrentHashMaps.

Regards,
Denis.

----- Original Message -----
> ----- Original Message -----
> > Hi.
> >
> > I already replied to this e-mail, but I didn't get it,
> > so I'm assuming there are others who didn't. The reply
> > is here:
> > http://mail.openjdk.java.net/pipermail/distro-pkg-dev/2011-April/013537.html
> >
> > In it I gave reasons for why I think the attached patch
> > is the correct way to do things.
> >
> > Is it good to go?
> 

> 
> -- snip --
> 
> Cheers,
> Andrew
> 
> >
> > Thank you,
> > Denis.
> >
> > ----- Original Message -----
> > > * Denis Lila <dlila at redhat.com> [2011-04-12 12:21]:
> > > > Hi.
> > > >
> > > > I found a few problems in PluginAppletViewer.java.
> > > > We were modifying/reading from two static hash maps
> > > > concurrently. The hash maps were "applets" and "status".
> > > > status was only modified in a synchronized method,
> > ...