===================================
From: Eric Salo[SMTP:salo@mrjones.engr.sgi.com]
Sent: Monday, March 17, 1997 2:50 PM
To: 'mpi-external@mcs.anl.gov'
Subject: Re: MPI and Threads
> We think it is quite clear by now that all this business with multiple thread
> support options stems from the fact that some vendor thread libraries ARE NOT
> GOOD ENOUGH to be used with a portable communication library (such as MPI).
And pthreads is? We know this?
I gotta disagree. But to some extent this is a moot issue. I see no benefit at
all to officially blessing any one threads package in the MPI standard. If
====================
The main point of our proposal is that POSIX is a standard but not yet another
thread package. If one wants to do portable programming with threads and
MPI, one uses POSIX threads (and time will tell how viable this idea is).
If one wants to use any other package, just specify its name in the info argument
while writing your library with MPI, and user applications that will use the library
are able to check if the right thread package is used.
If yes, they proceed, if no - they abort.
====================
users want pthreads in sufficient numbers then the market will adapt to fill
this need. And if they don't, then we would be better off not attempting to
dictate otherwise in the standard.
====================
It probably has been missed, but in our original proposal we said that MPI should
consider POSIX threads for PORTABLE programming;
we did not say that other thread packages cannot be used. Quite to the contrary,
we provided a means to check if the right thread package is used in order
to give users an opportunity to abort gracefully in case of the mismatch.
====================
-- Eric Salo Silicon Graphics Inc. salo@sgi.com
==================== Thanks, Boris.