Switch to mpi_f08

[fstein93]

This PR replaces the old Fortran 90-style MPI module with its Fortran 2008 counterpart. This solves a bunch of issues with the old Fortran wrapper.

Because the mpi_f08 module is not always available or is buggy with certain combinations of compilers and libraries, the use of the new module has to be actively enabled by adding the preprocessor flag -D__MPI_F08.

It will be used automatically via the toolchain if OpenMPI or IntelMPI is used. Probably, a few adjustments on the Dashboard will be necessary later.

[fstein93]

I am not really sure why Intel or MPICH have these issues. There are not too many changes from CP2K side.

[fstein93]

I suspect a bug on either MPICH or compiler side. I will consult the developers.

Furthermore, there is a bug in the Intel implementation which does not allow to pass non-contiguous arrays which is allowed to the MPI standard. I will also open a bug report there.

[oschuett]

does not allow to pass non-contiguous arrays which is allowed to the MPI standard

Do we actually pass non-contiguous arrays to MPI? Generally, I think it would be good if we started using CONTIGUOUS wherever possible.

[fstein93]

There is one spot. I forgot which one. This triggered the Intel compiler. What we could do is to add the contiguous keyword if we do not use the mpi_f08.
I have just done some research. There is a compile-time variable called MPI_SUBARRAYS_SUPPORTED which enables this behavior. If we do not set it, we have to enforce contiguity of all arrays in our wrapper. The devil is in the details. But that should be easy to implement.

[oschuett]

There is one spot. ...we have to enforce contiguity of all arrays in our wrapper.

Yes, I think the best solution would be to refactor that one spot an then simply declare all arrays in our wrapper as CONTIGUOUS. This should spare us a lot of hassle and produce more performance portable code.

[alazzaro]

For the record, I've already spent some time to add CONTIGUOUS keyword in DBCSR...

[fstein93]

I added the contiguous keyword wherever possible except from the non-blocking routines where I used the is_contiguous function to prevent accidental programming mistakes (see commit message).

[oschuett]

I added the contiguous keyword wherever possible except from the non-blocking routines ...

Oh, good thinking!

As an alternative we could use CONTIGUOUS POINTER or ALLOCATABLE for those async routines because their are treated as a distinct types and get checked at compile time. However, it's probably not worth the effort.

[fstein93]

As an alternative we could use CONTIGUOUS POINTER or ALLOCATABLE for those async routines because their are treated as a distinct types and get checked at compile time. However, it's probably not worth the effort.

That does not work if the array has a constant size or the caller has to artificially use an allocatable instead.

[fstein93]

The issue with MPICH looks to me like a bug on the library or the compiler-side. If there are no complaints, I will just add a commit to enforce the use of the old MPI interface with MPICH in our install script and merge.

[oschuett]

Looks like this GCC bug which I found via Stackoverflow.

[oschuett]

I will just add a commit to enforce the use of the old MPI interface with MPICH in our install script and merge.

I don't know. It feels like this could still trip up a lot of people. Maybe we make it opt-in for now via a -D__MPI_F08 flag?

[fstein93]

Just for reference: #1030