On Oct 21, 2009, at 5:43 PM, Lisandro Dalcin wrote:
> Currently (trunk, just svn update'd), the following call fails
> (because of the ranks=NULL pointer)
>
> MPI_Group_{incl|excl}(group, 0, NULL, &newgroup)
>
Agreed; bug. I think there's a little wiggle room in the spec wording
that it may or may not be an error, but I think that it's certainly
more social to accept NULL if n==0.
> BTW, MPI_Group_translate_ranks() has similar issues...
>
> Provided that Open MPI accept the combination (int_array_size=0,
> int_array_ptr=NULL) in other calls, I think it should also accept the
> NULL's in the calls above... What do you think?
>
The spec isn't quite a clear for MPI_Group_translate_ranks as it is
for MPI_Group_incl|excl (e.g., those both have defined behavior for
when n==0), but I think it falls into "it's more social" category to
just do nothing and return MPI_SUCCESS when n==0.
I'll fix both -- thanks!
--
Jeff Squyres
jsquyres_at_[hidden]
|
![[Flag]](../../../../../community/mirrors/flag-us.gif){kind=small}