[MPI3 Fortran] What to do with mpif.h in MPI-3?
jsquyres at cisco.com
Mon Jan 25 11:26:25 CST 2010
Just a short note -- I agree with just about everything you and Bill said.
On Jan 22, 2010, at 6:32 PM, N.M. Maclaren wrote:
> On Jan 22 2010, Bill Long wrote:
> >INTEGER*8 is not part of any Fortran standard - though the * notation is
> >popular in old codes and supported by many vendors.
> With a variety of different meanings! I have seen all of word count,
> byte count, and a simple sequence from 1 upwards.
> >I'm sure there is some difference of opinion on what is meant by
> >"Fortran 77 style interfaces". My opinion is that such interfaces are
> >free to use more modern syntax and parameterized intrinsic types. The
> >restriction is that only argument passing mechanisms, data types, and
> >kinds of arrays that were in f77 should be allowed. Basically, derived
> >types and dummy argument declarations that would force the caller to
> >have a visible interface are not allowed.
> Yes, I agree.
> >> 2. Don't deprecate mpif.h, but do not include any new MPI-3
> >> functionality in it. This is really only a minor distinction from
> >> deprecating; it's just a slightly less-strong statement meant to assure
> >> Fortran programmers that we are NOT abandoning mpif.h. However, the road
> >> to all new MPI-3 functionality is via the proposed MPI-3 explicit
> >> bindings (i.e., "use mpi3").
> >I like this better than (1) - it is more honest. It makes it clear that
> >mpif.h will continue to work in legacy codes, which I think is an
> >important message.
> Again, I agree. The reason to preserve the old interfaces is for old code.
> There's no need to be dogmatic about this, and trivial extensions could be
> added where necessary, but the general rule would be that they wouldn't be.
> Also, I think that this will be by far the least hassle.
> >> 3. Include all new MPI-3 functionality in mpif.h *except* for the
> >> "large count" functionality. This allows mpif.h-using applications to
> >> get most of the MPI-3 goodness, with the exception that only INTEGER
> >> counts will be supported. If your app needs larger counts than that, you
> >> must "use mpi3".
> >I suspect there could be more "exceptions" uncovered as work progressed.
> > I think it is simpler to be all one way or the other.
> Non-blocking buffer attributes, to name just one.
jsquyres at cisco.com
More information about the mpiwg-fortran