Also as the lock is on the hashtable, no
new locks are added as the iteration
will use the same lock.
On 1/12/07, Matt Benson <[EMAIL PROTECTED]> wrote:
--- Steve Loughran <[EMAIL PROTECTED]> wrote:
> Matt Benson wrote:
> > I agree, Peter. However, the patch I am currently
> > testing adds synchro blocks to all Hashtable
> copies in
> > PropertyHelper. :)
> >
> > -Matt
>
> Just as long as we don't get deadlocks.
>
>
> When I was updating the diagnostics for Ant1.7.1
> when a task doesnt get
> found, the IDE (intellij) was flagging what it felt
> were concurrency
> issues, places inside ComponentHelper where stuff
> was being accessed in
> both syncrhonized and unsynchronized locations. I
> left it all alone,
> because IDEA is over-paranoid, and because I like to
> understand the code
> better before wrapping everything.
Yes, I looked at this stuff, but these were all cases
of
public Hashtable getFooProperties() {
return new Hashtable(fooProperties);
}
All I did was place synchronized (fooProperties) {}
around that statement (the constructor call being the
important piece, obviously), to ensure no wayward
put()s would hit fooProperties while Hashtable's
constructor was presumably iterating over its
contents. I can't see any danger there; that doesn't
mean somebody cooler than I won't, however. ;)
-Matt
>
> -steve
>
>
---------------------------------------------------------------------
> To unsubscribe, e-mail:
> [EMAIL PROTECTED]
> For additional commands, e-mail:
> [EMAIL PROTECTED]
>
>
____________________________________________________________________________________
Get your own web address.
Have a HUGE year through Yahoo! Small Business.
http://smallbusiness.yahoo.com/domains/?p=BESTDEAL
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]