Re: dynamic counter-proposal

Eric Salo (salo@mrjones.engr.sgi.com)
Sat, 11 May 1996 16:24:45 -0700

> OK, here is the meat. Note that there are two communicator building
> requirements, the children's world and the inter-comm. You _do_ know, by
> way of a flag, whether or not to build the child world communicator
> in the MPI_Spawn function.

Well, almost. Recall that one of the benefits of splitting the spawn from the
creation of the intercommunicator is that it gives us the SPAWN_MULTIPLE
semantics basically for free. In the new proposal, we don't know at spawn time
whether we'll be spawning additional groups and merging with them, so we can't
really build the new MPI_COMM_WORLD for the children until the attach/detach.

But in the common case we will probably not(?) be spawning multiple groups, so
an implementation might simply create a communicator for every spawned group
and then just merge them later if needed.

-- 
Eric Salo         Silicon Graphics Inc.             "Do you know what the
(415)933-2998     2011 N. Shoreline Blvd, 7L-802     last Xon said, just
salo@sgi.com      Mountain View, CA   94043-1389     before he died?"