[MPlayer-dev-eng] Re: [SURVEY] change default CPU target when compiled with runtime CPU detection

[Zuxy]

2005/11/15, Rich Felker <dalias at aerifal.cx>:
> On Tue, Nov 15, 2005 at 05:57:32AM +0200, Jan Knutar wrote:
> > On Monday 14 November 2005 18:21, Matthias Wieser wrote:
> >
> >
> > Although, the Athlon is made for running unoptimized code and wont do too
> > badly on  P4 optimized code :-)
>
> P4-optimized code is not "unoptimized", it's "anti-optimized". i.e.
> stupid things which should be slow are done because they're the only
> way to make it perform well on the P4. :)
>

Well I'd say that's a rumor. It's true that P4s are more sensitive to
instruction selection, and imply stricter rules for compilers to do
that. However, it's still far from so called 'anti-optimized', since
these rules have minimal negative impact on earlier CPUs, i.e. they
usually don't care. Besides, instruction selection is only a small
part of the whole optimization, and most important optimizations like
loop unrolling and constant propagation are not CPU specific.:-)

> IMO the -mcpu setting should be either i686 or athlon. Definitely not
> P4.

Then be more conservative and leave it i686. I was just inspired by
the Fedora team, who optimize the i686 Linux core using -march=i686
-mtune=pentium4.

--
Zuxy
Beauty is truth,
While truth is beauty.
PGP KeyID: E8555ED6