Van Snyder wrote:
>
> I haven't been following this discussion carefully, so this comment may
> miss the mark.
>
> Ian Bush was quoted as having written:
>
> > Certainly, and it does nothing for you when calling libraries, e.g.
> > for BLAS it's DGEMM I generally want on my workstation here, but for
> > the Cray it's SGEMM.
>
> This problem is already solved.
>
> Define your problem in terms of the number of digits you're using.
>
> Define a generic interface, say GEMM, that includes DGEMM and SGEMM.
>
> Use the generic name.
>
> If your real number kind is, say, SELECTED_REAL_KIND(P=13) you'll get DGEMM
> on your Sun workstation, and SGEMM on your Cray.
>
> --
Hum, nothing proves that selected_real_kind(p=13) be either REAL or DOUBLE PRECISION. I already know what you're going to reply: "On all machines I use, it is the case". Well, I have to agree, but I thought that F90 was designed to have a higher degree of abstraction than assuming that selected_real_kind(13) is necessarily the REAL ou the DOUBLE PRECISION kind ?
Pierre
--
+-----------------------------------+----------------------------+
| Pierre Hugonnet | mail....CGG |
| | 1, rue Leon Migaux |
| R&D Data Processing | 91341 MASSY cedex |
| | FRANCE |
| COMPAGNIE GENERALE DE GEOPHYSIQUE | phone...(33) 164 47 45 59 |
| Paris Processing Centre | fax.....(33) 164 47 32 49 |
| http://www.cgg.com | [log in to unmask] |
+-----------------------------------+----------------------------+
My opinions are not necessarily those of CGG
--------------------------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|