On Tue, Mar 19, 2002 at 11:00:55AM +0100, Thilo Manske wrote:
> On Tue, Mar 19 2002 at 01:10:15 -0500, Charles Shannon Hendrix wrote:
> > On Thu, 14 Mar 2002 11:14:12 +0100
> > "Thilo Manske" <Thilo.Manske@HEH.Uni-Oldenburg.DE> wrote:
> > > BTW: You get sightly better results using -mv8 than -msupersparc on s SS5 (I
> > > think -msupersparc doesn't only use the v8 instcructions, it tunes the code
> > > for superscalar execution for supersparc cpus as well).
> > 
> > This isn't true on either of my SS5 machines (a 110MHz and a 170 turboSPARC).
> Well, I haven't mentioned it, but I was talking about -current and gcc
> 2.95.3. Maybe the picture is different with older versions of gcc.

Yeah, the ups and downs of compiler performance will make your head spin.

I'm running egcs-1.1.2 on NetBSD 1.5.2.

> With gcc 2.9.5 you get around ~>3% more dhrystones per second (with
> optimization of libc as well probably more) on a SS5 with 110MHz ?Sparc II.



-- 
UNIX/Perl/C/Pizza__________________________________shannon@widomaker.com