This Message Is From an External Sender

This message came from outside your organization.

Hi folks, During the meeting in Stuttgart, we discovered some weirdness where we thought MPI_Psend_init_c and MPI_Precv_init_c should have been present in MPI 4. 1, but weren’t. Failing to find any reason why that was true, we assumed it was

ZjQcmQRYFpfptBannerStart

ZjQcmQRYFpfptBannerEnd

During the meeting in Stuttgart, we discovered some weirdness where we thought MPI_Psend_init_c and MPI_Precv_init_c should have been present in MPI 4.1, but weren’t. Failing to find any reason why that was true, we assumed it was a problem with a Git merge and I replayed those changes in PRs 1108 and 1110.

Yesterday, Ken found the place where we had discussed these specific topics in October, 2023 and decided not to put them in MPI 4.1 here. At that time, we decided to push them out of 4.1 and address them later (which we never circled back to do) and removed the changes here.

Now that we know what happened there, I’m going to revert the PRs to put these changes back in and put us back to the situation we were in before the RCM and put this back in the Forum’s queue to discuss at some point post-MPI 5.0. You can find the PR for that here.

I’ll leave that up for a few days before proposing it gets merged just so everyone has a chance to take a look.

Sorry about all the mess. Hopefully we’re all back to the point that we previously agreed on.