This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Little C-benchmarks of gcc-2.95.4-ss,gcc-3.3.4-ss,gcc-3.4.1-ss & gcc-3.5.0-ss.
- From: jc-nospam at jr-pizarro dot jazztel dot es
- To: gcc at gcc dot gnu dot org
- Date: Tue, 01 Jun 2004 18:31:26 +0200 (CEST)
- Subject: Re: Little C-benchmarks of gcc-2.95.4-ss,gcc-3.3.4-ss,gcc-3.4.1-ss & gcc-3.5.0-ss.
> You're measuring on an AthlonXP and use -march=i486? That's not
> fair. Please schedule for Athlon with -march=athlon.
I did the bench for each -march= i386 i486 i586 i686 & athlon-xp,
but i wan't post this huge list because the results are very similar.
> Aren't those enabled by default?
Aren't because those options are for Novillo's tree-ssa (gcc-ssa-3.5).
> How often did you run these? CAn you be confident of any measurement
> errors?
There are measurement errors, it's normal at linux 2.4.x, but i did boot at
single mode (only 1 bash-session active).
> > gcc2 -O2 : 83160 bytes | 0m10.976s
> > gcc33 -O2 : 74944 bytes | 0m10.162s ; smaller and quicker
> > gcc34 -O2 : 83136 bytes | 0m10.377s
> > gcc35 -O2 : 83156 bytes | 0m10.863s
> >
> 83160 is larger than 83156, so gcc2 has the largest code size and all
> newer ones generate smaller code size, e.g. 4 bytes ;-)
83156 bytes of gcc35 to 74944 bytes of gcc33 is a big difference: 8212 bytes
(10.95%!!!)
>
> GCC 3.5 is not yet ready for production, we're still in stage 1 of the
> process. So your measurements will be encouragement to make 3.5
> better - but nobody promised you to any quality of the CVS snapshot of
> 3.5.
>
I know it xD