It is proposed that
(a) text be added to the description of MPI_RMA_ALLOC() which states
both 2-sided and 1-sided communication could potentially benefit
from using the memory returned by this function.
(b) the 'bytes' argument to MPI_RMA_ALLOC() be replaced with 'count'
(c) MPI_RMA_ALLOC() be made collective.
(d) MPI_RMA_ALLOC() and MPI_RMA_FREE() be moved out of the 1-sided
chapter and renamed as MPI_MEM_ALLOC() and MPI_MEM_FREE(). (For
lack of a better alternative, we could just move them into the
misc chapter for now.)
The only "operative" parts of the proposal are (b) and (c). I have no problems
with either. I shall add the text for discussion at next meeting . The text
should make clear that this "collective" call may not involve any collective
communication at all.
If malloc is collective shouldn't free be collective as well?