On Aug 10, 2010, at 3:59 PM, Gus Correa wrote:
> Thank you for opening a ticket and taking care of this.
Sorry -- I missed your inline questions when I first read this mail...
> > That being said, we didn't previously find any correctness
> > issues with using an alignment of 1.
>
> Does it affect only the information
> provided by ompi_info, as Martin Siegert suggested?
> Or does it really affect the actual alignment of
> MPI types when OpenMPI is compiled with Intel,
> as Martin, Ake Sandgren, Hugo Gagnon, and myself
> thought it might?
It affects both: I do see the size and alignment of double precisions (as determined by configure) used in the OMPI datatype engine (and real8, etc.). We had a copy-n-paste error in ompi_info such that some Fortran types were displaying the wrong sizes / alignments.
--
Jeff Squyres
jsquyres_at_[hidden]
For corporate legal information go to:
http://www.cisco.com/web/about/doing_business/legal/cri/
|
![[Flag]](../../../../../community/mirrors/flag-us.gif){kind=small}