<html><body>
<p>I did not notice this until I was going over the Ballot 4 slides. I would prefer not to vote in the currently proposed wording<br>
<br>
The correction is for page 22, not 222<br>
<br>
I propose:<br>
<tt><br>
MPI 1.1, page 22, line 48 reads<br>
used after a call to MPI_PROBE. (End of rationale.)<br>
but should read<br>
used after a call to MPI_PROBE or MPI_IPROBE. </tt><br>
<tt> In this case the call to MPI_GET_COUNT should use the same</tt><br>
<tt> datatype as will be used in the MPI_RECV.<br>
(End of rationale.) <br>
<br>
(Advice to users.) </tt><br>
<tt> The buffer size required for the receive can be affected by </tt><br>
<tt> data conversions and by the stride of the receive </tt><br>
<tt> datatype. Calling MP_GET_COUNT with MPI_BYTE to calculate the</tt><br>
<tt> receive buffer size may not be portable unless the MPI_RECV </tt><br>
<tt> uses MPI_BYTE.<br>
(End of advice to users.)<br>
</tt><br>
<br>
Dick Treumann - MPI Team/TCEM <br>
IBM Systems & Technology Group<br>
Dept 0lva / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601<br>
Tele (845) 433-7846 Fax (845) 433-8363<br>
<br>
<br>
<tt>mpi-21-bounces@cs.uiuc.edu wrote on 01/18/2008 12:55:09 PM:<br>
<br>
> This is a proposal for MPI 2.1, Ballot 4.<br>
> <br>
> This is a follow up to:<br>
> Datatypes and MPI_PROBE <br>
> in </tt><tt><a href="http://www.cs.uiuc.edu/homes/wgropp/projects/parallel/MPI/mpi-">http://www.cs.uiuc.edu/homes/wgropp/projects/parallel/MPI/mpi-</a></tt><tt><br>
> errata/index.html<br>
> with mail discussion in<br>
> </tt><tt><a href="http://www.cs.uiuc.edu/homes/wgropp/projects/parallel/MPI/mpi-">http://www.cs.uiuc.edu/homes/wgropp/projects/parallel/MPI/mpi-</a></tt><tt><br>
> errata/discuss/probedatatype/<br>
> <br>
> <br>
> Proposal for MPI 2.1, Ballot 4:<br>
> -------------------------------<br>
> MPI 1.1, page 222, line 48 reads<br>
> used after a call to MPI_PROBE. (End of rationale.)<br>
> but should read<br>
> used after a call to MPI_PROBE or MPI_IPROBE. <br>
> With a status returned from MPI_PROBE or MPI_IPROBE, the same<br>
> dataypes are allowed as in a call to MPI_RECV to receive this message.<br>
> (End of rationale.) <br>
> <br>
> Advice to users. To allocate the appropriate amount of memory as receive<br>
> buffer, the same datatype as in the following receive call should be used<br>
> to determine the needed space. In portable programs due to possible data <br>
> conversions, it is not guaranteed that the count returned by MPI_GET_COUNT<br>
> with datatype MPI_BYTE is the correct amount of needed memory space in <br>
> the receive buffer (although MPI_BYTE is matching every datatype). <br>
> (End of advice to users.)<br>
> -------------------------------<br>
> <br>
> Reason for the first part: The current MPI-1.1 text says "The <br>
> datatype argument<br>
> should match the argument provided by the receive call that set the status <br>
> variable." With MPI_PROBE, there isn't such a receive call.<br>
> <br>
> Reason for the advice to users: It helps to write portable code. <br>
> Because malloc needs a byte count, users may write wrong programs <br>
> by using MPI_BYTE.<br>
> <br>
> -------------------------------<br>
> <br>
> Discussion should be done through the new mailing list<br>
> mpi-21@cs.uiuc.edu.<br>
> <br>
> I have sent out this mail with CC through the old general list<br>
> mpi-21@mpi-forum.org<br>
> <br>
> Best regards<br>
> Rolf<br>
> <br>
> <br>
> <br>
> Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner@hlrs.de<br>
> High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530<br>
> University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832<br>
> Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner<br>
> Nobelstr. 19, D-70550 Stuttgart, Germany . (Office: Allmandring 30)<br>
> _______________________________________________<br>
> mpi-21 mailing list<br>
> mpi-21@cs.uiuc.edu<br>
> </tt><tt><a href="http://lists.cs.uiuc.edu/mailman/listinfo/mpi-21">http://lists.cs.uiuc.edu/mailman/listinfo/mpi-21</a></tt><tt><br>
</tt></body></html>