On Wed, Jan 04, 2012 at 01:13:22AM -0700, John Brooks wrote:
> On Wed, Jan 4, 2012 at 1:04 AM, Kostya Shishkov
> <[email protected]> wrote:
> > On Wed, Jan 04, 2012 at 12:56:09AM -0700, John Brooks wrote:
> >> On Wed, Jan 4, 2012 at 12:44 AM, Kostya Shishkov
> >> <[email protected]> wrote:
> >> > On Wed, Jan 04, 2012 at 12:23:01AM -0700, John Brooks wrote:
> >> >> @@ -799,8 +800,9 @@ static void vc1_mc_4mv_luma(VC1Context *v, int n, 
> >> >> int dir)
> >> >>      if (fieldmv && (src_y & 1) && src_y < 4)
> >> >>          src_y--;
> >> >>      if (v->rangeredfrm || (v->mv_mode == MV_PMODE_INTENSITY_COMP)
> >> >> +        || s->h_edge_pos - s->mspel * 2 < 11
> >> >>          || (unsigned)(src_x - s->mspel) > s->h_edge_pos - (mx & 3) - 8 
> >> >> - s->mspel * 2
> >> >> -        || (unsigned)(src_y - (s->mspel << fieldmv)) > v_edge_pos - 
> >> >> (my & 3) - ((8 + s->mspel * 2) << fieldmv)) {
> >> >> +        || (unsigned)(src_y - (s->mspel << fieldmv)) > 
> >> >> FFMAX(v_edge_pos - (my & 3) - ((8 + s->mspel * 2) << fieldmv), 0)) {
> >> >
> >> > this particular change should be not needed anymore or what?
> >> >
> >>
> >> This case is particularly complex, and I couldn't find any succinct
> >> way to test it without making assumptions about s->mspel or fieldmv
> >> (which I'm not personally comfortable doing). If either of those is
> >> just a flag (I haven't looked), it should be easy enough, but I don't
> >> mind leaving one FFMAX, either.
> >
> > Just assume it, it shouldn't do much harm (it's only 23 in the most pessimal
> > case).
> 
> Attached.

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

Reply via email to