[MPlayer-dev-eng] [PATCH] Saturation & PSWAPD bugfix in mp3lib/dct64_3dnow.c & mp3lib/dct64_k7.c

Zuxy Meng zuxy.meng at gmail.com
Fri Jul 7 16:00:30 CEST 2006


Hi,

2006/7/7, Guillaume POIRIER <poirierg at gmail.com>:
> Hi,
>
> On 7/7/06, Diego Biurrun <diego at biurrun.de> wrote:
> > On Tue, Jun 27, 2006 at 04:37:52PM +0800, Zuxy Meng wrote:
> > > 2006/6/27, Guillaume Poirier <gpoirier at mplayerhq.hu>:
>
> [..]
>
> > > dct_MMX_3dnowex will raise SIGILL on a K6. For a K6-III+, changes
> > > should be made as
> > >
> > >    if (gCpuCaps.has3DNowExt)
> > >    {
> > >        dct36_func=dct36_3dnowex;
> > > -        dct64_MMX_func= (gCpuCaps.hasMMX2) ? dct64_MMX_3dnowex :
> > > dct64_MMX_3dnow;
> > > +       dct64_MMX_func=dct64_MMX_3dnowex;
> > >        mp_msg(MSGT_DECAUDIO,MSGL_V,"mp3lib: using 3DNow!Ex optimized
> > > decore!\n");
> > >    }
> >
> > It seems to work on my K6-III.  I played an MP3 file without problems.
> >
> > Any clever ideas how to integrate this?
>
> Well, you've got the patch right there. Now the question is: is it
> faster on K6III to use the 3dnowext version rather than the 3Dnow
> version.
>

K6-3 doesn't support 3dnowext, K6-3+ does. The 3dnowext version should
always be used when supported.
-- 
Zuxy
Beauty is truth,
While truth is beauty.
PGP KeyID: E8555ED6



More information about the MPlayer-dev-eng mailing list