Eduard Bloch <[EMAIL PROTECTED]> writes:
> #include <hallo.h>
> * Glenn McGrath [Sun, Nov 09 2003, 09:53:26AM]:
>
> > > We're all very interested in *real* evidence here, because there
> > > hasn't been any in the past. If you don't have any evidence, you can
> > > expect people to call bullshit on this.
> >
> > Gentoo
>
> What does that mean? Gentoo uses a heavily patched kernel which goes far
> beyound of what we dicuss
>
> > # time bzip2 -9k linux-2.6.0-test5.tar
> >
> > real 2m40.974s
> > user 2m33.679s
> ...
> > user 2m49.316s
>
> Even then, it's about only 10 percent. Let's compare them with vanilla
> kernel, optimised for P4:
>
> # time bzip2 -9k out.wav
>
> real 0m42.862s
> user 0m41.410s
> sys 0m0.310s
>
> The same running an unoptimised vanilla kernel:
>
> real 0m40.990s
> user 0m40.640s
> sys 0m0.350s
>
> Where is the performance gain you are talking about? Not even two
> percent, haha.
The unoptimized kernel is _clearly_ slower than the optimized one but
that actually makes bzip run _much_ faster.
This clearly prooves _beyond_any_reasonable_doubt_ we should not
optimize the kernel at all to get the best performance.
MfG
Goswin
