Big MPI Support for the RMA Chapter

[tonyskjellum]

Currently, we are working to offer the "Big MPI" interface across the MPI Standard in the Large Count WG as most of you know. It is a cross-cutting change to the MPI standard, as discussed in June and September meetings. It started with Forum Issue #80 but now we are working to offer the comparable (_X) interface standard wide for further consideration by the Forum in December.

Forum Issue #99 was created for the RMA chapter in June (by me) as part of expanding the effort. It was tagged with this WG, but we didn't really discuss it in anyway till Barcelona, and then offline. According to the RMA working group rules (and my conversation/email with @wgropp), an issue for the RMA should start here as an RMA WG issue. We had not been following such an approach in other WGs, so this has happened out of order. I hope all will be understanding of this lack of understanding on my part.

If the WG would be willing to promote The Forum's issue #99, there is already full text and a PR for this interface. Please let me know, and I will also be happy to join an RMA con call in near future to explain this request and the status of the proposal to the WG.

For complete reference, the current text I've written up (with Bill's recent encouragement) for this WG, is as follows:

mpi32-report-ticket99-29sep18-1207.pdf

which has an associated PR. This is how we have been going chapter-by-chapter since June.

Please see also:

mpi-forum/mpi-issues#99
https://github.com/mpi-forum/mpi-standard/pull/62

Thank you!

[jeffhammond]

Other than https://github.com/mpi-forum/mpi-standard/pull/62/files#r221774804, this looks good to me.

[devreal]

@tonyskjellum @jeffhammond Can this be closed?

[devreal]

I believe the question of embiggening has been resolved, closing.