-> Date: Mon, 29 Apr 1996 18:44:25 -0400
-> From: Gary Oberbrunner <garyo@avs.com>
-> To: "William C. Saphir" <wcs@nas.nasa.gov>
-> Cc: mpi-dynamic@mcs.anl.gov
-> Subject: Re: Dynamic Process comment: using MPI_Init() without mpirun
-> > It may be that current implementations require it to be started by
-> > mpirun, but this is just an implementation detail, [n]ot a
-> > fundamental problem. Then you can spawn the rest of the processes
-> > with MPI_Spawn. So I understand correctly?
->
-> Yup, exactly. All I really want to do is get a sentence in the
-> standard that says what happens if you call MPI_Init() with no mpirun
-> or spawn above it (it becomes a singleton, assuming the implementation
-> supports dynamic processes).
Since the MPI standard makes no mention of mpirun, I can't see how it can
state what happens if you don't use it.
I can assure you on my system (Paragon) no mpirun is used and when MPI_Init()
is called the process will block until one process from the same processs
group from each "node" in its "partition" also calls MPI_Init(). When all
process(es) in the application (whether that is 1 or 4000) call MPI_Init(), each
pro cess becomes an MPI process without mpirun or MPI_Spawn() or support for
dynamic processes.
joel clark
->
-> --
-> Gary Oberbrunner garyo@avs.com
-> Advanced Visual Systems, Inc. http://www.avs.com/~garyo
-> 300 Fifth Avenue (617)890-8192 x2133 TEL
-> Waltham, MA 02154 (617)890-8287 FAX
--- End Of Your Message