-
Notifications
You must be signed in to change notification settings - Fork 0
2020 03 18
- Jeff Squyres
- Martin Rüefenacht
- Dan Holmes
- Tony Skjellum
Wow. It's real. Sucks.
Martin to work on fixing the conflicts in https://github.com/mpi-forum/mpi-standard/pull/132.
We looked for the errata that we want to file.
See https://github.com/mpiwg-large-count/large-count-issues/wiki/problems-possible-errata.
There's also a bunch of "to do" items -- logistics and mechanical stuff.
See https://github.com/mpiwg-large-count/large-count-issues/wiki/TODOs.
Tony suggests that we need to have someone look at the IO chapter to verify the embiggening. --> Tony sends email to Rajeev.
Re-iterating the rationale about all the confusion with datatype constructors:
In general, datatypes have three uses:
- MPI_Aint used as number of bytes. You're using MPI_Aint simply as a large number.
- MPI_Aint used as an absolute address. Then MPI_Aint is a bit string that represents a location in memory.
- With MPI_Files, so you're potentially addressing something bigger than memory.
Bottom line: MPI doesn't necessarily know at construction/commit time whether you're in case 1 or case 2 because it depends on whether you use MPI_BOTTOM with the constructed datatype or not.
When embiggening:
- Keep the old signature for backwards compat reasons (may be int, may be MPI_Aint, ... etc.)
- If relevant, embiggen to MPI_Aint for use case 2 (i.e., absolute addresses)
- If relevant, embiggen to MPI_Count for use case 3 (i.e., files)
Today, embiggening has only resulted in 2 bindings (cases 1 and 2 or cases 1 and 3). We have not (yet?) found a case where embiggening resulted in 3 bindings (cases 1, 2, and 3).
Didn't talk about these today:
- Making official machine-readable file(s) from Python to publish on mpi-forum.org
- Python?
- JSON?
- mpi.h / mpif.h / mpi module / mpi_f08 module
- ...?
- Continuous integration with Pythonization
- Pytest
- Generate a list of changed / added / deleted bindings per PR
- ...
- All the "to do" items
- MPI-5 things -- all-new language stuff (C++, divorce 1:1 relationship to C bindings, ...etc.)
- Remember: there are several people who want to be involved in these discussions (e.g., Jeff H., Lisandro D., ...etc.).
Back to weekly meetings? Yes.
Can we do 8:30am US Eastern on Wednesdays? Seems like we can. Jeff will update the calendar invite.