[mpich-discuss] MPI_Accumulate with MAXLOC does not error for incompatible types

[Jed Brown]

Thanks, Dave.


On Mon, Jan 14, 2013 at 6:39 PM, Dave Goodell <goodell at mcs.anl.gov> wrote:

> Fixed by http://git.mpich.org/mpich.git/commitdiff/ba2e444
>
> It looks like I accidentally put the wrong URL into the commit message for
> some reason.  The correct mailing list URL should be:
> http://lists.mpich.org/pipermail/discuss/2013-January/000227.html
>
> -Dave
>
> On Jan 1, 2013, at 1:09 PM CST, Jed Brown wrote:
>
> > With MPI_Reduce, we get a nice helpful error message:
> >
> > Fatal error in PMPI_Reduce: Invalid MPI_Op, error stack:
> > PMPI_Reduce(1217)...........: MPI_Reduce(sbuf=0x7fff3e39fa50,
> rbuf=0x7fff3e39fa40, count=1, dtype=USER<contig>, MPI_MAXLOC, root=0,
> MPI_COMM_WORLD) failed
> > MPIR_MAXLOC_check_dtype(151): MPI_Op MPI_MAXLOC operation not defined
> for this datatype
> >
> >
> > With MPI_Accumulate, we just get wrong results. See attached code using
> a user-defined contiguous results.
> >
> > I care about this because I'm stuck [1] dispatching into alternative
> mechanisms for type combinations that are not directly supported by MPI
> (e.g., struct {int u; long long k;} and __float128). I missed a spot and
> had a confusing bug that I think MPICH should be catching.
> >
> >
> > [1] As usual, the crud would evaporate if only MPI_Accumulate were
> extensible over datatypes.
> > <maxloc.c>_______________________________________________
> > discuss mailing list     discuss at mpich.org
> > To manage subscription options or unsubscribe:
> > https://lists.mpich.org/mailman/listinfo/discuss
>
> _______________________________________________
> discuss mailing list     discuss at mpich.org
> To manage subscription options or unsubscribe:
> https://lists.mpich.org/mailman/listinfo/discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mpich.org/pipermail/discuss/attachments/20130114/e68627cf/attachment.html>