[FFmpeg-devel] [RFC] cpu specific implementation of libavutil functions
Mon Sep 10 04:25:23 CEST 2007
On Mon, Sep 10, 2007 at 12:27:02AM +0200, Michael Niedermayer wrote:
> On Sun, Sep 09, 2007 at 11:23:09PM +0200, Luca Barbato wrote:
> > Michael Niedermayer wrote:
> > > Hi
> > >
> > > On Wed, Sep 05, 2007 at 05:10:50PM +0200, Luca Barbato wrote:
> > >> Recently I got an altivec adler32 implementation from Sigbj?rn Skj?ret
> > >> and the main issue is how to plant it into libavutils and find a sane
> > >> general way.
> > >
> > > how much faster is that adler32 code?
> > > adler32 really should be limited by memory speed not by cpu
> > 4x
> > > 2. do we even have a working altivec detection routine? that what we do
> > > in dsputil is not what i would call working and i will not accept such
> > > hacks in libavutil
> > Currently altivec is checked by system dependent reporting or basic
> > sigill catching on altivec instruction. I could have an even more crazy
> > version like using fork() to execute the altivec code. Still this is to
> > my knowledge the only portable way to check.
> id say let configure detect it and forget about runtime detection
> also solves the problem of how to interface to the code, its just per
> #ifdef then
the alternative to that would be to let the user app do the detection
fork(), sigill catching or anything similar is not ok in libavutil
and actually not in libavcodec (or any other lib) either ...
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
The educated differ from the uneducated as much as the living from the
dead. -- Aristotle
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: Digital signature
More information about the ffmpeg-devel