Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

What is a "host process"? What is "Machine-dependent absolute address" when referring to process rank? #569

Open
wgropp opened this issue Aug 1, 2022 · 4 comments
Assignees
Labels
chap-contexts Groups, Contexts, Communicators, Caching Chapter Committee mpi-5 For inclusion in the MPI 5.0 standard
Projects

Comments

@wgropp
Copy link

wgropp commented Aug 1, 2022

Section 9.1.2, p453 (MPI40) in describing the "MPI_HOST" predefined attribute key, calls this the "host process rank". This dates from the very, very early days of message passing systems, when some systems had a separate "host" computer and a process running on it, which might be part of the MPI job. I do not believe any system does this any more (so we might also want to deprecate MPI_HOST), but in any event, "host process" is undefined in the MPI standard.

Related to this, in 7.2.4 on page 316 (MPI40), is

MPI does not specify the correspondence between the process rank in MPI_COMM_WORLD and its (machine-dependent) absolute address.

What is a "(machine-dependent) absolute address"? Elsewhere in the standard, absolute addresses refer to the memory address space. Yes, I know what is intended here - but the standard should be precise and clear.

(Originally found in translating MPI 2.2 into Japanese)

@wesbland
Copy link
Member

wesbland commented Aug 3, 2022

In fact, we do have an issue in progress to deprecate MPI_HOST: #57

@wesbland
Copy link
Member

wesbland commented Aug 3, 2022

I think the part in Chapter 9 can be a duplicate of #57. For the rest of it, I'll hand this to the Chapter 7 CC: @GuillaumeMercier

@wesbland wesbland added chap-contexts Groups, Contexts, Communicators, Caching Chapter Committee mpi-4.1 For inclusion in the MPI 4.1 standard labels Aug 3, 2022
@wesbland wesbland added this to To Do in MPI 4.1 via automation Aug 3, 2022
@wesbland
Copy link
Member

wesbland commented Dec 8, 2022

This wasn't read at the December 2022 meeting. The last opportunity for MPI 4.1 is to have it ready at the March 2022 meeting (and it needs to "pass" the reading).

@GuillaumeMercier Are you (or is someone else) planning to push this forward or should we move it out of the plan for MPI 4.1?

@GuillaumeMercier
Copy link

I have to think about this one.

@wesbland wesbland added mpi-5 For inclusion in the MPI 5.0 standard and removed mpi-4.1 For inclusion in the MPI 4.1 standard labels May 3, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
chap-contexts Groups, Contexts, Communicators, Caching Chapter Committee mpi-5 For inclusion in the MPI 5.0 standard
Projects
Status: To Do
MPI 4.1
To Do
Development

No branches or pull requests

5 participants