No speed difference was meassureable
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* qatar/master:
  lavfi: unref AVFilterLink.out_buf in ff_end_frame().
  lavfi: unref AVFilterLink.cur_buf in ff_end_frame().
  vsrc_testsrc: avoid an unnecessary avfilter_ref_buffer().
  vf_slicify: clear AVFilterLink.cur_buf in start_frame().
  vf_settb: simplify start_frame().
  vf_fieldorder: don't give up its own reference to the output buffer.
  vf_pad: don't give up its own reference to the output buffer.
  vf_overlay: don't access a buffer reference that's been given away.
  vf_drawtext: don't give up its own reference to the input buffer.
  vf_gradfun: don't store two pointers to one AVFilterBufferRef.
  vf_delogo: don't store two pointers to one AVFilterBufferRef.
  vf_aspect: clear AVFilterLink.cur_buf in start_frame().
  lavfi: add avfilter_unref_bufferp()

Conflicts:
	doc/APIchanges
	libavfilter/avfilter.h
	libavfilter/buffer.c
	libavfilter/vf_aspect.c
	libavfilter/vf_drawtext.c
	libavfilter/vf_overlay.c
	libavfilter/vf_pad.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Set picture type before calling get_buffer.
This allows the DR application to make better decisions.
It also fixes a resource leak in case of missing reference frames
since it would call get_buffer but never release_buffer.
Also use FFSWAP to ensure that the AVFrame is properly initialized
in the next get_buffer (in particular that data[0] is NULL).
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Reveiwed-by: compn <tempn@twmi.rr.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Marton Balint <cus@passwd.hu>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* qatar/master:
  libopenjpeg: introduce lowres and lowqual private options
  FATE: add a test for flac cover art.
  cafdec: allow larger ALAC magic cookie
  alac: fix channel pointer assignment for 24 and 32-bit

Conflicts:
	libavcodec/alac.c
	libavcodec/libopenjpegdec.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

This reduces code duplication and prevents stale pointers from remaining
on the link.

v4l2_fd_open is already called at the end of v4l2_open;
calling a second time leaks structures inside libv4l2.

Fix trac ticket #1560.

This reduces code duplication and prevents stale pointers from remaining
on the link.

The buffer is passed on to the next filter, so we shouldn't keep any
pointers to it.

Avoid pointless filter ref/unref. Just use the reference passed to
start_frame().

Directly forwarding the input buffer to the next filter means that
drawtext no longer owns any references to it and thus shouldn't refer to
it in any way.

The buffer is passed on to the next filter, so we shouldn't keep any
pointers to it.

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Do not continue decoding if one is missing.
Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Convert them to zigzag order, as the rest of them are.

When I was adding support for 10-bit DNxHD, I just copy-pasted the
missing quant matrices from the spec. Now it turns out the existing
matrices in dnxhddata.c were in zigzag order. This resulted in wrong
quantization for 10-bit DNxHD. The attached patch fixes the problem by
converting 10-bit quant matrices to zigzag order.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

This happens with for example mplayer.
Fixing it in ffmpeg allows new ffmpeg to be compiled with older mplayer
which would not be possible if the fix was just in mplayer.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>