[Mpi-forum] C++ types inaccessible after #281
jedbrown at mcs.anl.gov
Tue Jun 26 11:39:57 CDT 2012
On Tue, Jun 26, 2012 at 8:19 AM, Jeff Squyres <jsquyres at cisco.com> wrote:
> On Jun 26, 2012, at 12:14 PM, Jed Brown wrote:
> > On Tue, Jun 26, 2012 at 6:13 AM, Jeff Squyres <jsquyres at cisco.com>
> > There are a few C++-specific datatypes for which it would be worth
> introducing new C datatypes (e.g., MPI_CXX_COMPLEX). It's too bad that no
> one realized this until now, because this won't make the MPI-3 deadline.
> But it could be part of MPI-3.1 or errata, or some such.
> > Two people in offline discussions have suggested that C bindings to the
> std::complex types (I'm not aware of anything else that #281 will lose)
> might qualify as ticket 0.
> I'd have to see the exact changes to get a better feel for it, but I would
> be a little surprised if these changes could be considered ticket 0, for
> two reasons:
> 1. "ticket 0" is reserved for typos or other insignificant changes that do
> not change the meaning of the standard. Indeed, the standard would still
> be fine without those ticket 0 changes -- the idea is that ticket 0 changes
> make the standard read better, etc.
> 2. Adding MPI datatypes for uin64_t (and friends) was a standalone ticket.
> I would think that this would be very similar.
Okay. FWIW, I think this request is only to preserve the table "C++-only
Named Predefined Datatypes" as in the MPI-2.2 standard, but with namespaced
MPI:: names replaced by the MPI_CXX_ prefix (and s/Complex/std::complex/
because that was apparently a typo). From the discussion, it sounds like
removing support for these types was an accidental/unrecognized side-effect
of bulk-removing the C++ bindings in #281.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the mpi-forum