Re: GNAT compiler switches and optimization

[Gautier <gautier@fakeaddress.nil>]

Jeffrey R. Carter:

>> But still, it won't spare the time lost to fill the C matrix with zeros.
> 
> I didn't include that in the timing.

But you have to! Otherwise you are cheating. The zeroing is a part of your 
changed algorithm; without zeroing the result is wrong.

>> If you want to do a benchmark with Fortran, it's really not a good 
>> idea to begin with "pessimizing" the Ada code.
> 
> I'm more interested in seeing what makes a difference in the Ada. In 
> this case, the high-level features that let you write less code.

It is less code... "on the paper". In fact, it is more work, or, at best, if 
the code optimizer is really smart enough to combine the ":= (others => 
(others => 0.0) );" with the loops, it is the same work. Maybe that's the 
problem with high-level features: one can zero a matrix in so few characters 
that it seems to be done instantly ;-).

Cheers
Gautier