[mpich2-dev] issues Pack/Unpack_external with MPI_LONG and MPI_DOUBLE

Lisandro Dalcin dalcinl at gmail.com
Mon Dec 7 14:54:26 CST 2009


On Mon, Dec 7, 2009 at 5:34 PM, Dave Goodell <goodell at mcs.anl.gov> wrote:
> Thanks for the bug report.  I've filed this as a ticket in our bug tracking
> system and CCed you on it:
> https://trac.mcs.anl.gov/projects/mpich2/ticket/967
>

Many thanks. I've just created an account (now that you have OpenID
login enabled).

> AFAIK, external32 is not fully implemented and it is not surprising that it
> does not work.  We do not have any near-term plans to implement it.
>

However, It seems that it mostly works :-), at least for primitive datatypes.

My own testing shows that DOUBLE and LONG_DOUBLE are broken in 32/64
bits, and LONG and UNSIGNED_LONG on 32 bits.

Well, I'd have to test the new C99 datatypes introduced since MPI 2.2...

>
> We will update the status of that ticket as the status of external32 support
> in MPICH2 changes.
>

OK, many thanks.

-- 
Lisandro Dalcín
---------------
Centro Internacional de Métodos Computacionales en Ingeniería (CIMEC)
Instituto de Desarrollo Tecnológico para la Industria Química (INTEC)
Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
PTLC - Güemes 3450, (3000) Santa Fe, Argentina
Tel/Fax: +54-(0)342-451.1594


More information about the mpich2-dev mailing list