[Mpi-21] [MPI3 Fortran] MPI-2.1: Fortran 90 bindings
crasmussen at [hidden]
Thu Mar 6 17:20:37 CST 2008
I should warn everyone that the problem becomes worse for Fortran
2008 because the number of possible dimensions has been increased to 15.
By the way, I've been officially appointed by the Fortran J3
standards committee to liaison with the MPI committee (if the MPI
Forum will accept me :-) Now would be a very good time to decide
what should be in new Fortran bindings and if they still don't seem
to work well, you can instruct me to ask the Fortran committee to
make changes to the proposed F2008 standard. I have one suggestion
already (but I want to keep you all in suspense until next week when
I can explain it).
On Mar 5, 2008, at 7:15 PM, Jeff Squyres wrote:
> The current MPI-2 F90 bindings are un-implementable because they
> require almost 7 *million* interface functions: with 15 intrinsic
> Fortran types, each with 7 possible dimensions, the current F90
> interface *requires*:
> - 50 MPI functions with one choice buffer: 15 * 7 * 50 = 5,250
> - 25 MPI functions with two choice buffers: (15 * 7 * 25)^2 = 6.8M
> - ...and a few hundred more MPI functions with no choice buffers
> This is clearly broken.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Mpi-21