Re: Proposal to modify two-phase...

Raja Daoud (raja@tbag.rsn.hp.com)
Sun, 02 Feb 1997 15:06:19 CST

> That is why, in part, persistent collectives are of value for
> optimization. _Tony

:-) Or better yet, not add tagged-collectives and have to "solve" the
complexity they bring by adding persistent operations.

If it's the limit of one two-phased collective that is of concern, this
can be relaxed without using tags (at most N concurrent two-phased
collectives are optimized in some implementations, the others use
point-to-point with a system-generated generation tag per communicator).

--Raja

-=-
Raja Daoud Hewlett-Packard Co.
raja@rsn.hp.com Convex Division