On 2012-01-05 08:57:53 -0800, Ronald S. Bultje wrote: > Hi, > > On Thu, Jan 5, 2012 at 8:50 AM, Janne Grunau <janne-li...@jannau.net> wrote: > > On 2012-01-03 14:32:34 +0100, Janne Grunau wrote: > >> Fixes a regression introduced in 8b94df0f2047e972. > >> --- > >> libavcodec/vp3.c | 2 ++ > >> 1 files changed, 2 insertions(+), 0 deletions(-) > >> > >> diff --git a/libavcodec/vp3.c b/libavcodec/vp3.c > >> index f44d084..602b5fa 100644 > >> --- a/libavcodec/vp3.c > >> +++ b/libavcodec/vp3.c > >> @@ -1378,6 +1378,8 @@ static inline int vp3_dequant(Vp3DecodeContext *s, > >> Vp3Fragment *frag, > >> return i; > >> } > >> } while (i < 64); > >> + // return value is expected to be a valid level > >> + i--; > >> end: > >> // the actual DC+prediction is in the fragment structure > >> block[0] = frag->dc * s->qmat[0][inter][plane][0]; > > > > ping > > Why not just return i - 1?
that would return -1 instead of 0 if the first token & 3 is 0 Not sure if that is a valid thing but it changes behavior of vp3_dequant and we test if the return value is < 0 Janne _______________________________________________ libav-devel mailing list libav-devel@libav.org https://lists.libav.org/mailman/listinfo/libav-devel
