[FFmpeg-cvslog] r12512 - trunk/libavcodec/i386/dsputil_mmx.c

[Aurelien Jacobs]

?smail D?nmez wrote:

> On Fri, Mar 21, 2008 at 12:27 AM, Mike Melanson <mike at multimedia.cx>
> wrote:
> > ?smail D?nmez wrote:
> >  > On Fri, Mar 21, 2008 at 12:23 AM, Mike Melanson
> >  > <mike at multimedia.cx> wrote:
> >  >> Uoti Urpala wrote:
> >  >>  > On Thu, 2008-03-20 at 14:52 -0700, Mike Melanson wrote:
> >  >>  >> michael wrote:
> >  >>  >>> Log:
> >  >>  >>> Hardcode register to prevent aparent miscompilation.
> >  >>  >>> Fixes regression tests with gcc 2.95.
> >  >>  >> Awesome. FATE thanks you. :)
> >  >>  >
> >  >>  > Not so awesome for people with working compilers. The change
> >  >>  > makes the code worse for the sake of working around bugs in
> >  >>  > an obsolete compiler.
> >  >>
> >  >>  Tradeoffs, tradeoffs. Do we drop official support for 2.95.3
> >  >> on x86_32?
> >  >
> >  > I would be surprised if it works on x86_64 either.
> >
> >  If what works? If this change works correctly on x86_64? Or gcc
> > 2.95.3 on x86_64? We don't support the latter config.
> 
> Hmm didn't know we didn't support x86_64 + gcc 2.95.3 (which is good
> news) .

You got it somewhat backwards. In fact, that's gcc-2.95 which don't
support x86_64 at all. So obviously ffmpeg can't support this
non-existing combination.

Aurel