[mpich-discuss] Help with mpich2 1.3a1 on i7-920

[Gaetano Bellanca]

Hi Pavan,

thanks for your specification.  I'm going to change this accordingly to 
your suggestion.
And begin to use mpich2-1.3a1

Regards.

Gaetano

Il 19/02/2010 15:16, Pavan Balaji ha scritto:
> Gaetano,
>
> MPI_IN_PLACE should only be specified at the root:
>
> if (rank == root)
> 	MPI_Gather(MPI_IN_PLACE, scount, stype, rbuf, rcount, rtype, root, comm);
> else
> 	MPI_Gather(sbuf, scount, stype, NULL, 0, MPI_DATATYPE_NULL, root, comm);
>
>   -- Pavan
>
> On 02/19/2010 01:10 AM, Gaetano Bellanca wrote:
>    
>> Dear Rajeev and Pavan,
>>
>> thanks for your answers. I tried using MPI_IN_PLACE in MPI_GATHER in the
>> position of the send buffer, but probably I didn't understand correctly
>> its use. In fact, I'm still having errors.
>>
>> Fatal error in MPI_Gather: Invalid buffer pointer, error stack:
>> MPI_Gather(761): MPI_Gather(sbuf=MPI_IN_PLACE, scount=1, MPI_LOGICAL,
>> rbuf=0x187f0b0, rcount=1, MPI_LOGICAL, root=0, comm=0x84000002) failed
>> MPI_Gather(682): sendbuf cannot be MPI_IN_PLACE
>>
>> However, it works using separate buffers as input and output.
>>
>> Regards.
>>
>> Gaetano
>> _______________________________________________
>> mpich-discuss mailing list
>> mpich-discuss at mcs.anl.gov
>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>>      
>    

-- 
Gaetano Bellanca - Department of Engineering - University of Ferrara
Via Saragat, 1 - 44122 - Ferrara - ITALY
Voice (VoIP): +39 0532 974809 Fax: +39 0532 974870
mailto:gaetano.bellanca at unife.it

L'istruzione costa? Stanno provando con l'ignoranza!