questions on gfortran

[lucatrv]

> You could easily check whether your vectorization option has changed the
> generated code in the sin() test loop, using the command line options,
> or by the -S option, for example.

You're right, even if I don't understand assembly, it's convinient to look
at the .s files to understand if flags had any effect on the compilation.
So, -O2 and -O3 produce the same assembly from this code, while all other
flags change something in the assebly (with little effect on the speed
though) with the exclusion of -ftree-vectorize.

> Among Fortran compilers for linux x86[32/64] , I think only ifort would
> vectorize sin().  It does it by automatically substituting calls to
> "short vector math library."  All functions to be vectorized must have
> pre-built library support.
> No compiler rebuilds the math library, or vectorizes automatically those
> functions which aren't available as built-in parallel instructions. The
> only built-in for sin() on x86 architectures is the original x86
> instruction, dating back to i386 and beyond.

In this case, unfortunally, I think this is an important shortcoming, which
can make a big speed difference, especially with numerical code in
fortran... I've actually to confirm that ifort reaches much better results
with this little test code...

Luca