MPI needs to know about the Fortran type. That's what this routine does. It does not create the Fortran type. Just view this as a special case of MPI_Type_create_struct. We could just use BIND(C) but then I have to add the pair type constructor for C. I was planning to do that but haven't typed it. Jeff On Fri, Nov 10, 2017 at 11:53 AM, Bill Long <[email protected]> wrote:

Fortran 90 had derived types, and Fortran 2003 had BIND(C) derived types - essentially C structs. I don’t see the logic of inventing an archaic “create type” routine when the user could just declared the required type directly.

Cheers, Bill

...

On Nov 10, 2017, at 1:34 PM, Jeff Hammond <[email protected]> wrote:

Given that Fortran 90+ has user-defined types, I don't see any reason why we can't allow these to be used to make MPI_M**LOC reductions more useful in Fortran. See the following ticket for details. Please add comments to the GitHub issue.

https://github.com/mpi-forum/mpi-issues/issues/70

Thanks,

Jeff

-- Jeff Hammond [email protected] http://jeffhammond.github.io/ _______________________________________________ mpiwg-fortran mailing list [email protected] https://lists.mpi-forum.org/mailman/listinfo/mpiwg-fortran

Bill Long [email protected] Principal Engineer, Fortran Technical Support & voice: 651-605-9024 Bioinformatics Software Development fax: 651-605-9143 Cray Inc./ 2131 Lindau Lane/ Suite 1000/ Bloomington, MN 55425

_______________________________________________ mpiwg-fortran mailing list [email protected] https://lists.mpi-forum.org/mailman/listinfo/mpiwg-fortran

-- Jeff Hammond [email protected] http://jeffhammond.github.io/