[Mpi-forum] Fortran change for discussion
nmm1 at cam.ac.uk
Tue Sep 11 02:37:39 CDT 2012
On Sep 11 2012, Jeff Squyres wrote:
> You explain the situation in Fortran-eese, but the current text is an
> attempt to explain it in MPI-eese (i.e., in terms of copying). I think
> the MPI-reader-friendly text should be used.
Sort-of, I agree. The problem with the current description is precisely
that it uses copying rather than aliasing as a basis, but I assume that
is for C programmers. There is definitely a problem in trying to explain
something in C-friendly terms that is conceptually outside C's rudimentary
>> If the current text is kept, I think the clarification is important.
The other possibility is to stress that the copying can occur at any
time between the initial call and the Wait. That is, after, all precisely
what asynchronous means in this context!
More information about the mpi-forum