Newsgroups: comp.parallel.mpi
From: Marc Snir <snir>
Subject: Re: question about MPI_Probe
Organization: IBM T.J. Watson Research Center
Date: 23 Feb 1996 22:46:46 GMT
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <4glg4m$1205@watnews1.watson.ibm.com>

sagawa@hcrlgw.crl.hitachi.co.jp (Nobutoshi Sagawa) wrote:
>Forgive me if you received the same message more than once. I may have
>already sent it out inadvertently.
>---------------------------------------------------------------------
>
>Hello everybody
>
>I am a little bit confused with the following description of MPI_Probe
>which I found in the book, "MPI compelete reference". I would
>appreciate any help about this.
>
>"The semantics of MPI_PROBE guarantees progress, in the same way as a
>corresponding receive executed at the same point in the program."
>
>Does this mean that a call to MPI_Probe should complete a matching
>sync-send just as a matching MPI_Recv would do? 
The MPI_Probe call does not complete a matching MPI_Ssend -- only a matching
MPI_Recv will do so.  
 Does this also mean
>that a standard-send that has been blocked due to lack of the buffer
>space should progress with the presence of a matching MPI_Probe?

What the text says is that, in a situation where MPI guarantees that an
MPI_Recv would complete, then the corresponding MPI_Probe should also complete.
In a situation where  an MPI_Send has been posted at a process, and a matching
MPI_Recv has been posted at another process, then the communication must
complete eventually, irrespective of the amount of buffer space in the system.
Similarly, in a situation where an MPI_Send has been posted at a process, and a
matching MPI_Probe has been posted at another process, then the MPI_Probe call
must return (eventually).  This does not complete the send operation, which may
continue to be blocked until a matching receive is posted.
>
>Thank you.
>
>
>
>
>--
>--------------------------------------------------------------------
>                      Nobutoshi Sagawa (sagawa@crl.hitachi.co.jp)
>                      Processor Systems Dept.
>                      Central Research Laboratory, Hitachi Ltd.
>                      Higashi Koigakubo 1-280,
>                      Kokubunji, Tokyo 185, JAPAN.
>
>                      Tel: 0423-23-1111      Fax: 0423-27-7743
>                      Tel: +81-42323-1111    Fax: +81-42327-7743
>--------------------------------------------------------------------

-- 

Marc Snir  
IBM T.J. Watson Research Center
P.O. Box 218, Yorktown Heights, NY 10598
email: snir@watson.ibm.com
phone: 914-945-3204
fax:   914-945-4425
URL:   http://www.research.ibm.com/people/s/snir


