Michael Hall wrote:
>...To
>completely eliminate the possibility of a deadlock between the
>application thread and either of the AWT threads wouldn't it need a
>priority of 4?
It's impossible to eliminate deadlock possibilities by juggling priorities.
I mean that in both an analytical/theoretical sense, as well as a
real-world practical sense. At best, juggling priorities will only shift
when or where the conditions arise that lead to deadlock. The problem is
only shifted, though, not eliminated. I have seen such attempts actually
cause the deadlock to occur more often ("The best-laid schemes..." and all).
To eliminate the problem, the same sequence of lock acquisition must be
used everywhere. That means code has to be changed.
Ben Galbraith wrote:
>Have we already discussed what exactly the Runnable submitted by
>Thread-39 is doing? Can you submit the source of that Runnable, please?
See the list archives:
<http://lists.apple.com/archives/java-dev/2006/Jan/msg00142.html>
4. "It doesn't matter what the code in the Runnable() is."
-- GG
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Java-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/java-dev/email@hidden
This email sent to email@hidden
