hello,

i see the fix, but why isn't line 339 changed?

https://svn.open-mpi.org/trac/ompi/browser/trunk/ompi/op/op_predefined.c?rev=10023#L339

bert

On Tue, 23 May 2006, Brian Barrett wrote:

> On May 23, 2006, at 11:49 AM, Bert Wesarg wrote:
>
> > Hello,
> >
> > the predfined mpi op MPI_PROD is apparently wrong:
> >
> > from openmpi-1.2a1r10020/ompi/op/op_predefined.c:
> >
> > OP_FUNC(prod, long_long_int, long long int, +=)
> > OP_FUNC(prod, unsigned_long_long, unsigned long long, +=)
> > :
> > OP_FUNC(prod, fortran_integer1, ompi_fortran_integer1_t, +=)
> > :
> > OP_FUNC(prod, fortran_integer2, ompi_fortran_integer2_t, +=)
> > :
> > OP_FUNC(prod, fortran_integer4, ompi_fortran_integer4_t, +=)
> > :
> > :
>
> Doh!  Thanks for finding that.  We'll have a fix in shortly,
> definitely before 1.1 releases.
>
> I've filed a bug report in our tracker, and you should get e-mail
> updates when the bug is fixed.
>
> Brian
> _______________________________________________
> users mailing list
> us...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/users
>

Reply via email to