On Sat, Mar 15, 2014 at 12:08:56PM +0100, Anton Khirnov wrote:
> 
> On Fri, 14 Mar 2014 05:42:23 -0700, Diego Biurrun <[email protected]> wrote:
> > ---
> >  libavcodec/bfin/vp3dsp_init.c   | 2 +-
> >  libavcodec/bit_depth_template.c | 1 +
> >  libavcodec/cavsdsp.c            | 2 ++
> >  libavcodec/dsputil.h            | 7 +------
> >  libavcodec/imgconvert.c         | 2 +-
> >  libavcodec/indeo4data.h         | 3 ++-
> >  libavcodec/mathops.h            | 4 ++++
> >  libavcodec/mathtables.c         | 4 +++-
> >  libavcodec/mss3.c               | 2 +-
> >  libavcodec/pgssubdec.c          | 2 +-
> >  libavcodec/rv30dsp.c            | 1 +
> >  libavcodec/rv40dsp.c            | 1 +
> >  libavcodec/vp3.c                | 2 +-
> >  libavcodec/vp8dsp.c             | 2 +-
> >  14 files changed, 21 insertions(+), 14 deletions(-)
> 
> Why is that a better place?

Anything is better than dsputil.h ;)

The tables are defined in mathtables.c and all other declarations of the
tables defined in mathtables.c are in mathops.h.  I don't see why these
shouldn't live there as well.

> > --- a/libavcodec/dsputil.h
> > +++ b/libavcodec/dsputil.h
> > @@ -36,16 +36,11 @@
> >  /* encoding scans */
> >  extern const uint8_t ff_alternate_horizontal_scan[64];
> >  extern const uint8_t ff_alternate_vertical_scan[64];
> > -extern const uint8_t ff_zigzag_direct[64];
> >  extern const uint8_t ff_zigzag248_direct[64];
> 
> Why one and not the other?

The 248 table is only used in DV, I'm moving it to dvdata in a later commit.

Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to