> > Right. Maybe we need "CPULocals" in Java.
> Does that really buy you performance at user level? Unlike thread locals,
> the semantics of "CPULocals" seem nontrivial, since you can ge preempted
> halfway through an update ...
A CPULocal would have to be maintained entirely by the system, updated
via tricky mechanics upon every remapping of thread to CPU. Dave Dice
and Alex Garthwaite at Sun Labs have been working on some of these
techniques on sparcs. Maybe they could post something on JMM list
outlining how and why to do it? :-)
I'm not sure how realistic this would be in Java, but maybe a more
useful and general construct along these lines will emerge.
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel
This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:30 EDT