Re: GNAT compiler switches and optimization

[Jeffrey Creem <jeff@thecreems.com>]

Alinabi wrote:
> I ran your test programs compiled with gcc 4.0.3 and the following
> optimizations:
> COMMON_FLAGS=-g -march=opteron -mtune=opteron -mfpmath=sse
> -fomit-frame-pointer -O2 -fdump-tree-optimized
> and I cannot reproduce the large differences in performance everyone
> else talks about. Here are the times I get:
> 
> N        Ada            Fortran
> ====================
> 64      0.002029   0.000000
> 128    0.016321   0.016000
> 256    0.214143   0.204013
> 512    3.125888   3.124195
> 800    6.374982   5.864366
> 1024  34.10479   35.22620
> 2048  277.3071   283.2417
> 

That is interesting. The question then becomes has FORTRAN improved on 
the way to 4.2.0 or has Ada regressed.

Try doing a make dis_all which should produce annotated assembly output. 
The Ada version can be a little daunting in the way we have setup the 
files since the generic instantiations at the top full the .S files 
(woops, looked like I named them .dis) with the generic instatance.

Even if you are not an assembly guru, if you start from the bottom of 
the files you can usually pretty quickly find that inner loop and 
compare the number of statements.