On Mon, Jan 25, 2010 at 10:51:38PM -0500, Jeff Squyres wrote:

On Jan 25, 2010, at 10:46 PM, Torsten Hoefler wrote:

...

...

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).

Heh -- INTEGER*8 isn't standard. :-) D'oh, I didn't take my Fortran book on my travel but http://www.fortran.com/F77_std/f77_std.html doesn't list any mechanism to influence the size of an INTEGER. What am I missing?

I would think that a standards-based approach would be something like INTEGER(KIND=MPI_COUNT_KIND), right? Ok, I thought both was allowed, but I'm voting for whatever the F77 standard supports.

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