[MPI3 Fortran] What to do with mpif.h in MPI-3?

Torsten Hoefler htor at cs.indiana.edu
Mon Jan 25 21:46:41 CST 2010


On Mon, Jan 25, 2010 at 12:29:03PM -0500, Jeff Squyres wrote:
> On Jan 24, 2010, at 6:38 PM, Torsten Hoefler wrote:
> 
> > > What is everyone's opinion on which (1-4) the Forum should go with?
> > It seems like the route to go with the C-bindings (and also the general
> > bindings) is not clear yet and this would decide the outcome. Because if
> > we decide to take the "64-bit" (tm) approach and double the number of
> > function symbols, then the issue would be much simpler. However, if we
> > decide to take the int -> MPI_Count approach, then I don't see how we
> > can "extend" F77 in a standard-conforming and portable manner (but I
> > might miss something).
> 
> Hmm -- maybe I missed something in ATL, but I thought we agreed that
> we would do *both* things:
> 
> - Add new versions of various MPI functions with a <suffix> appended (e.g., MPI_File_write<suffix>) and take the large count parameter
> - The large count parameter in these new functions would have (in C) a parameter type of MPI_Count
This is exactly what I meant with "new symbols" (such symbols would of
course have a different count datatype, otherwise it wouldn't make
sense). I must have missed the consensus there; all I remember is that
there were strong reservations about the number of new functions. I do
not remember a straw vote?

> There has been some discussion on the mailing list (by 1 person) since
> ATL about *not* using an "MPI_Count" datatype (rather, use uint64_t or
> somesuch), but my $0.02 is that we should use MPI_Count.
Sure, the naming is an orthogonal discussion. I tend to agree to use the
standard-conforming approach for F77 (i.e., *no* MPI_COUNT but rather
INTERGER*8).

> > So depending on the outcome, I would vote:
> > 
> > - if we choose the extra function symbol for long types: option 4
> > - if we choose the MPI_Count approach: option 3
> 
> Based on my understanding, I'm not sure how to parse your answer
> because the current direction of the Forum is to do *both* things: new
> functions *and* use MPI_Count.
So *if* we decided for the extra symbols: option 4 :-).

Best,
  Torsten

-- 
 bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ -----
Torsten Hoefler       | Postdoctoral Fellow
Open Systems Lab      | Indiana University    
150 S. Woodlawn Ave.  | Bloomington, IN, 474045, USA
Lindley Hall Room 135 | +01 (812) 856-0501



More information about the mpiwg-fortran mailing list