[mpiwg-coll] ticket 353

Rajeev Thakur thakur at mcs.anl.gov
Tue Oct 22 11:37:52 CDT 2013


MPI_REDUCE_LOCAL is not the same as MPI_REDUCE on MPI_COMM_SELF. The latter just copies inbuf to outbuf without doing any op. Reduce_local performs an op between inbuf and inoutbuf. If inbuf is missing, it doesn't make sense to call Reduce_local.

On Oct 22, 2013, at 9:51 AM, Jeff Hammond wrote:

> https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/353 is trivial but
> perhaps I am the only person that sees the point.  Can the WG give a
> thumbs up or down on proceeding?
> 
> Thanks,
> 
> Jeff
> 
> -- 
> Jeff Hammond
> jeff.science at gmail.com
> _______________________________________________
> mpiwg-coll mailing list
> mpiwg-coll at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-coll




More information about the mpiwg-coll mailing list