(j3.2006) (SC22WG5.4744) Comments on comments on Ballot 3 on Fortran 2008 interpretations

[Dick Hendrickson]

On Fri, Aug 31, 2012 at 4:21 PM, Keith Bierman <khbkhb at gmail.com> wrote:
>
>
> On Aug 31, 2012, at 2:17 PM, Van Snyder <Van.Snyder at jpl.nasa.gov> wrote:
>
> On Fri, 2012-08-31 at 15:40 +0900, Malcolm Cohen wrote:
>
> SPEC benchmarks and code that gets the right answer are two different
> beasts.
>
>
> Well SPEC attempts to use real code with at least consistent results.
> Sometimes to the point of parody. The original SpeC spice was so constrained
> that instead of being an fp benchmark it turned out to be a memcopy ... Any
> real circuit generates different results  on different platforms
>
>
> REAL(x,kind(y)) is so rare that it would have no effect on performance
> in real programs to do it right.  If vendors absolutely need to get it
> wrong to make some arcane SPEC benchmark run faster, they can provide a
> command-line option to violate the intent of the standard.  Maybe even a
> secret option that ony they ever invoke -- for the purpose of compiling
>
> Can't be secret and comply with SPEC rules
>
> Sadly more people complain about performance than correctness

That's because performance is easy to measure.

Dick Hendrickson
>
> _______________________________________________
> J3 mailing list
> J3 at mailman.j3-fortran.org
> http://mailman.j3-fortran.org/mailman/listinfo/j3
>