Deprecate avcodec_alloc_context/2.

It should be replaced with proper per-codec defaults.

Deprecate corresponding AVCodecContext option.

This is the first test of decoder private options.

It supports passing options to codecs.

Adds support for decoder-private options and makes setting other options
simpler.

In testing, the file output plugin gave a max buffer size of about 20 million
frames, which is way more than what is really needed and causes a memory
allocation error on my system.

buffer_size/4 is the value used by aplay. This fixes output to null
devices, e.g. writing ALSA output to a file.

Instead, just decrease it to a valid value and use that.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Set DV packet durations using fields_per_frame.
This requires turning gxf_stream_info into the demuxer's context for access to the value in gxf_packet().
Since MPEG-2 seems to work fine this done only for DV.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

libavcodec/libxvidff.c:752: warning: initialization discards qualifiers from pointer target type

Eliminate redundant check in filter_mb_fast, consider bit depth in calculating qp_thresh.

Inline the relevant parts of fill_decode_caches into P-SKIP mv prediction to
avoid calling the whole thing.

These weren't getting inlined all the time in all gcc versions.

Mainly ported from 8-bit H.264 predict.

Some code ported from x264. LGPL ok by author.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Based on a somewhat similar idea in FFmpeg's swscale.

Based on a somewhat similar idea in FFmpeg's swscale copy.

We operated on 31-bits, but with e.g. lanczos scaling, values can
add up to beyond 0x80000000, thus leading to output of zeroes. Drop
one bit of precision fixes this.

We accessed the alpha array even it wasn't used and didn't
exist, hence leading to a NULL pointer segfault.

When using e.g. lanczos scaling, values can drop below 0, so they
should never be unsigned.

We would use the second half of the U plane buffer, rather than the
V plane buffer, to output the V plane pixels.

It's required to be non-zero since
cc58300e

There's no point in keeping them separate.

Those are reserved for the libs.

They were replaced by (de)muxer private options.

Deprecate AVFormatContext.loop_output.

Deprecate AVFormatContext.loop_input.

That is, unless the caller explicitly asks for them.

Prevents conflict between e.g. the 'loop' option in img2 demuxer and
'loop' flag in AVCodecContext.