Request for comments: mpi and using multiple compilers

[Sean Farley]

eborisch at macports.org writes:

> On Thursday, July 25, 2013, Sean Farley wrote:
>
>>
>> eborisch at ieee.org <javascript:;> writes:
>>
>> > On Thursday, July 25, 2013, Sean Farley wrote:
>> >>
>> >> But really, we're at the whim of what the macports community whats to do
>> >> in this situation. Since my Ph.D is riding on getting a working mpi +
>> >> fortran, I'd very much like to iron out these issues and get the ports
>> >> chugging along!
>> >>
>> >
>> > Does mpich +gccXX not get you to working fortran and MPI?
>> >
>> > I'll try to read through some of this thread later, but just looking for
>> > clarification on that point.
>>
>> Again, the issue is when using libraries dependent on mpi and
>> exacerbated on dependents of those dependents. This usually results in
>> breakage with the inability to specify which compiler to use in the n-th
>> dependent.
>>
>
> Again, I haven't scoured the whole thread, but would making sub-ports
> rather than variants for the different compilers help? The
> dependent's +gcc44+mpich could require the mpich-gcc44 package, for example.

Not really. It just changes the name to the same problem.

> I've also been playing with having the fortran compiler specified
> separately with a gfortNN variant, so one could build +clangXX+gfortNN.
> This opens up an entire new realm of possible breakage, however.

Well, I'd suggest just taking the highest stable gfortran and forcing it
as a dependency when using clang (which is the only compiler suite to
not have a fortran compiler).