Hi Howard (et al),

 

I wrote down my thoughts on #710 (including the sessions viewpoint) here:

https://github.com/mpi-forum/mpi-issues/issues/710#issuecomment-1536306552

 

Addendum, just for sessions folks:

 

``` C

MPI_Init();

MPI_Recv_init(…, MPI_COMM_WORLD, &req);

MPI_Finalize();

MPI_Request_free(&req); // erroneous, MPI is no longer initialised, so calling MPI_Request_free is prohibited

MPI_Session_init();

MPI_Request_free(&req); // oh dear lord, what is this? Surely this is also erroneous, but where …flicks through the MPI Standard… is the prohibition in the standard?

MPI_Session_finalize();

```

 

Best wishes,

Dan.

 

 

From: mpiwg-sessions <mpiwg-sessions-bounces@lists.mpi-forum.org> On Behalf Of Pritchard Jr., Howard via mpiwg-sessions
Sent: 04 May 2023 23:46
To: MPI Sessions working group <mpiwg-sessions@lists.mpi-forum.org>
Cc: Pritchard Jr., Howard <howardp@lanl.gov>
Subject: [mpiwg-sessions] Agenda items for Sessions WG meeting - 5/8/23

 

Hi All,

 

Please send me agenda items you’d like to discuss at our WG meeting on 5/8/23.

 

I have one item already.  I’d like to discuss https://github.com/mpi-forum/mpi-issues/issues/710 and its possible impact on sessions.

 

Howard

 


 

signature_61897647

Howard Pritchard

Research Scientist

HPC-ENV

 

Los Alamos National Laboratory

howardp@lanl.gov

 

signature_2816248013signature_10222113signature_2944524454signature_3333939719