[MPlayer-dev-eng] [BUG] Mplayer 1.0 pre5 ppc/altivec issues

[Michael Niedermayer]

Hi

On Thursday 15 July 2004 22:31, Luca Barbato wrote:
> Magnus Damm wrote:
> > On Thu, 2004-07-15 at 15:12, Michael Niedermayer wrote:
> >>On Thursday 15 July 2004 09:58, Magnus Damm wrote:
> >>>I'm unable to compile pre5 too. My PPC box runs Linux.
> >>>Adding a "#include <altivec.h> in "swscale_internal.h" helped a bit,
> >>>but more fixes are needed for "yuv2rgb_altivec.c".
> >>>I'm using a gcc from FSF.
> >>
> >>does reversal of the attached patches fix the issue? they should have
> >> never been applied anyway, but alex and diego repeatedly asked me to
> >> apply them so i did
> >
> > Well, yes, the attached patches solves the problem. Thank you. But it
> > all depends on what you want to do. I mean, yes, it successfully
> > disables the altivec code. But maybe it is a good thing to have altivec
> > acceleration? I rather see FSF-gcc support in the code. We can at least
> > find a solution where it is enabled for Darwin and disabled otherwise.
> >
> > I added the file "libavcodec/ppc/gcc_fixes.h" to solve things like this.
> > But it is probably not a good idea to include that one from code under
> > "postproc/"...
> >
> > lu, are you working on FSF-gcc support in "yuv2rgb_altivec.c"?
> >
> > / magnus
>
> I almost finished a "kludge patch" it should at least let you build
> everything, I could bet that that code will break something somewhere,
> please audit the patch.
including altivec.h at random places like vo_x11.c is IMHO unacceptable, not 
to mention the #undef vector stuff which is needed then too
u replace several constants with vec_splat*(constant), is this generating 
slower code?

btw, another question as u seem to be fairly interrested in this code, are u 
interrested in fixing it so beside compiling it actually outputs correct 
data?
and yes, theres plenty of "stray" code in there, dont fear to remove it

-- 
Michael
level[i]= get_vlc(); i+=get_vlc();		(violates patent EP0266049)
median(mv[y-1][x], mv[y][x-1], mv[y+1][x+1]);	(violates patent #5,905,535)
buf[i]= qp - buf[i-1];				(violates patent #?)
for more examples, see http://mplayerhq.hu/~michael/patent.html
stop it, see http://petition.eurolinux.org & http://petition.ffii.org/eubsa/en