On Fri, Feb 17, 2012 at 05:05:05PM +0000, Paul B Mahol wrote:
> On 2/17/12, Kostya Shishkov <[email protected]> wrote:
> > On Thu, Feb 16, 2012 at 09:48:58PM +0000, Paul B Mahol wrote:
> >> Width is padded for ham encodings too.
> >>
> >> Signed-off-by: Paul B Mahol <[email protected]>
> >> ---
> >>  libavcodec/cdxl.c |   13 ++++++-------
> >>  1 files changed, 6 insertions(+), 7 deletions(-)
> >>
> >> diff --git a/libavcodec/cdxl.c b/libavcodec/cdxl.c
> >> index a53a001..2e21836 100644
> >> --- a/libavcodec/cdxl.c
> >> +++ b/libavcodec/cdxl.c
> >> @@ -60,9 +60,9 @@ static void import_palette(CDXLVideoContext *c, uint32_t
> >> *new_palette)
> >>      }
> >>  }
> >>
> >> -static void bitplanar2chunky(CDXLVideoContext *c, int width,
> >> -                             int linesize, uint8_t *out)
> >> +static void bitplanar2chunky(CDXLVideoContext *c, int linesize, uint8_t
> >> *out)
> >>  {
> >> +    int padded_width = FFALIGN(c->avctx->width, 16);
> >>      GetBitContext gb;
> >>      int x, y, plane;
> >>
> >> @@ -70,17 +70,16 @@ static void bitplanar2chunky(CDXLVideoContext *c, int
> >> width,
> >>      memset(out, 0, linesize * c->avctx->height);
> >>      for (plane = 0; plane < c->bpp; plane++)
> >>          for (y = 0; y < c->avctx->height; y++)
> >> -            for (x = 0; x < width; x++)
> >> +            for (x = 0; x < padded_width; x++)
> >>                  out[linesize * y + x] |= get_bits1(&gb) << plane;
> >>  }
> >>
> >>  static void cdxl_decode_rgb(CDXLVideoContext *c)
> >>  {
> >>      uint32_t *new_palette = (uint32_t *)c->frame.data[1];
> >> -    int padded_width = FFALIGN(c->avctx->width, 16);
> >>
> >>      import_palette(c, new_palette);
> >> -    bitplanar2chunky(c, padded_width, c->frame.linesize[0],
> >> c->frame.data[0]);
> >> +    bitplanar2chunky(c, c->frame.linesize[0], c->frame.data[0]);
> >
> > second argument should be c->avctx->width instead
> >
> 
> No it should not, that breaks at least one sample.

indeed, got confused, sorry
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to