[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [mpi-21] MPI Receive buffers



Any thoughts on the standard choice of words?

-----Original Message-----
From: Erez Haba
Sent: Monday, December 17, 2007 2:45 PM
To: mpi-21@xxxxxxxxxxxxx
Subject: RE: [mpi-21] MPI Receive buffers

Can you elaborate?

-----Original Message-----
From: owner-mpi-21@xxxxxxxxxxxxx [mailto:owner-mpi-21@xxxxxxxxxxxxx] On Behalf Of Steven Ericsson-Zenith
Sent: Monday, December 17, 2007 12:14 PM
To: mpi-21@xxxxxxxxxxxxx
Subject: Re: [mpi-21] MPI Receive buffers

Yes, it is on purpose.

With respect,
Steven


On Dec 17, 2007, at 10:22 AM, Erez Haba wrote:

> Reading the MPI 1.1 standard I noticed that the wording prohibits
> read access to the receive buffer rather than warning that the
> result is undefined. (or simply prohibit modifying the buffer).
>
> Is this on purpose?
>
> There are patterns where programs would like to,
>       * poll on memory (longshot)
>       * read some region of the buffer that is outside the non-contig
> buffer.
>
> Thoughts?
>
> Thanks,
> .Erez
>
>