RE: JavaMemoryModel: Thread model algorithm.

From: David Holmes (dholmes@dltech.com.au)
Date: Tue Nov 18 2003 - 22:05:06 EST


I forgot to add ...

Sylvia Else wrote:
> So, on the face of it, all that's required to make this
> structure work is to guarantee that the monitor will next
> be acquired by the notified thread, and not by any other.
> ...
> This rule would not conflict with the current specification, which
> says nothing at all about the order in which threads
> acquire monitors.

An implementation that does this conforms to the existing spec. But an
application that relies upon it is no longer portable. WORA.

David Holmes

-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel



This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:54 EDT