[FFmpeg-devel] PATCH: COOK audio decode infastructure to support fixpoint optimization
Mon Jul 16 00:26:48 CEST 2007
Marc Hoffman writes:
> On 7/15/07, Guillaume POIRIER <poirierg at gmail.com> wrote:
> > Hi,
> > On 7/15/07, Marc Hoffman <mmhoffm at gmail.com> wrote:
> > > On 7/15/07, Guillaume POIRIER <poirierg at gmail.com> wrote:
> > > > On 7/15/07, Marc Hoffman <mmh at pleasantst.com> wrote:
> > > > >
> > > > > This patch changes the way the data flows by introducing a vec_t
> > data
> > > > > type which is just unsigned char.
> > > >
> > > > Please use a different prefix. We already have vec_* types for Altivec
> > > > intrinsics.
> > >
> > >
> > >
> > > No problem how is SIG_T?
> > Ok with me, but how about sig_t instead? All-uppercase symbols are
> > more meant to be used for constants right?
> I guess I was planning to do this:
> #define SIG_T uint32_t
> And not use a typedef, the real intent here is so you can easily see what
> pieces of the code are moving the audio signal around.
break out the signal flow so that the data path is obvious and the
type is easily changeable. This is only important at the toplevel of
the system, lower levels will have implementation specific details
like floating or fixed point. this is a cosmetic.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 2663 bytes
Desc: not available
More information about the ffmpeg-devel