alexander.supalov at intel.com
Thu Dec 5 11:18:24 CST 2013
I think that is about all that we had. Jeff Brown, who chaired the WG, as far as I can recall, might have more material, as probably do I in my private archive. Anyway, the point here is to re-merge that WG's work with the reignited discussion rather than to replace it.
The situation then was that after the initial assessment, we could not detect enough enthusiasm on the part of the forum and implementations to get any closer than they were before. The ABI WG then went dormant. I hope that the situation changed somewhat now that we have a joint MPICH ABI initiative.
I am afraid I cannot comment on particular snippets of other people's email 5-6 year after the events. If you wish, we can get this trail offline and go through the material with you.
From: Jed Brown [mailto:jedbrown at mcs.anl.gov]
Sent: Thursday, December 05, 2013 5:43 PM
To: Supalov, Alexander; Main MPI Forum mailing list
Subject: RE: [Mpi-forum] ABI
"Supalov, Alexander" <alexander.supalov at intel.com> writes:
> I hope so. Just click on the "Attachments" tab and look into the DOC files there.
That just defines some terms and scope, but Analysis, Solution, and Issues are TBD. Then there are various emails, but they are mainly high-level motivation. The pointer vs. integer sizes thing came up in one email, where Jeff Squyres said
> For an ABI, you need to standardize both the size and the value. It
> may not be difficult to agree on the sizes: sizeof(void*)/64 bits
> because you can treat it as an int or a pointer. But the value is
> where the religious debates come in. For example: MPI_COMM_WORLD --
> it is "0" or is it a pointer resolved at compile/link time?
which I think is a red herring since the value can be a link-time constant and an implementation is free to "optimize" internally.
Is there more technical content somewhere?
Dornacher Strasse 1
85622 Feldkirchen/Muenchen, Deutschland
Sitz der Gesellschaft: Feldkirchen bei Muenchen
Geschaeftsfuehrer: Christian Lamprechter, Hannes Schwaderer, Douglas Lusk
Registergericht: Muenchen HRB 47456
Ust.-IdNr./VAT Registration No.: DE129385895
Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052
More information about the mpi-forum