[mpich-discuss] Why do predefined MPI_Ops function elementwise in MPI_Accumulate, but not in MPI-1 routines?

[Jed Brown]

On Tue, Apr 24, 2012 at 14:52, Jim Dinan <dinan at mcs.anl.gov> wrote:

> Another thing to point out is that the Accumulate behavior on complex data
> will do what you need for MPI_SUM, but it will give you the wrong result
> for MPI_PROD (product) since complex multiplication is not just multiplying
> real and imaginary parts separately.


I know, MPI_Accumulate has the same problem. I don't care because I compute
additive instead of geometric norms (and even if I did geometric norms,
they would be computed using addition).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20120424/a20d2553/attachment.htm>