At 05:47 AM 11/28/2003, Doug Lea wrote:
> > So I'm arguing that if notify/interrupt reordering is allowed within a
> > thread, then the proposed specification change does not fit the bill. What
> > should it say instead?
>
>It should suffice to add to the specs for both notify and interrupt:
>
> The effects of a series of invocations of Object.notify and/or
> Thread.interrupt by a given thread occur in the order in which they
> are invoked.
I think the correct thing to say is that Object.notify and Thread.interrupt
respect the "happens before" relationship. Because "happens before" totally
orders events in a given thread this implies the above form.
>(Or maybe some less awkward wording.)
>
>This way, the spec for wait can still only add the phrase "for a
>notification".
>
>Do you see any problems with this?
>
>-Doug
>
>-------------------------------
>JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel
-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel
This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:55 EDT