My personal opinion is yes, it is worth investigating. There are vendors (HPE/Cray, not to name names, perhaps others) who leverage the MPICH ABI for their tools to readily support a bunch of MPI implementations. OMPI gets left off the list because we don't follow the ABI. I think it would be good for OMPI to be able to integrate more naturally if possible.
That's not to say that I'm signing up ORNL to do it. At least not at present. We have higher priority fish to fry, at least for the rest of the CY. From: devel <devel-boun...@lists.open-mpi.org> On Behalf Of Gilles Gouaillardet via devel Sent: Saturday, August 26, 2023 2:21 AM To: Open MPI Developers <devel@lists.open-mpi.org> Cc: Gilles Gouaillardet <gilles.gouaillar...@gmail.com> Subject: [EXTERNAL] [OMPI devel] MPI ABI effort Folks, Jeff Hammond and al. published "MPI Application Binary Interface Standardization" las week https://arxiv.org/abs/2308.11214<https://urldefense.us/v2/url?u=https-3A__arxiv.org_abs_2308.11214&d=DwMFaQ&c=v4IIwRuZAmwupIjowmMWUmLasxPEgYsgNI-O7C4ViYc&r=Oofpa2piyIoH9-ZgjiSR5uCfILs91wYVTpkZQO5bfCY&m=OR9GUNWuorFYu3zkYrgUhz-5Jv6LvXhILyzBBa3JnS6HpM-HXv2vqMMVw4T--1qd&s=Yw5btNSA7XvzReIAFuTsOAcsZTzW2kcjH1feCd8I0Fo&e=> The paper reads the (C) ABI has already been prototyped natively in MPICH. Is there any current interest into prototyping this ABI into Open MPI? Cheers, Gilles
