It allows to search for options only with AVClass, without allocating
the corresponding context.

1<<31 overflows because 1 is signed, so force it to unsigned.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

These caused scrambled error messages to be printed and
floating point exceptions.
For example when there was no decoder available for a
stream.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

FF_API_LAME_GLOBAL_OPTIONS => FF_API_LAME_GLOBAL_OPTS

Thanks to Clément Bœsch for spotting.

Deprecate CODEC_FLAG_SVCD_SCAN_OFFSET

Deprecate CODEC_FLAG_H263P_SLICE_STRUCT

Deprecate CODEC_FLAG_OBMC

Deprecate CODEC_FLAG_H263P_AIV

Deprecate CODEC_FLAG_H263P_UMV

Deprecate CODEC_FLAG_ALT_SCAN

Deprecate CODEC_FLAG_EXTERN_HUFF

Deprecate CODEC_FLAG_PART

Deprecate CODEC_FLAG2_MEMC_ONLY

Deprecate CODEC_FLAG2_MBTREE

Deprecate CODEC_FLAG2_PSY

Deprecate CODEC_FLAG2_BIT_RESERVOIR

Deprecate CODEC_FLAG2_NON_LINEAR_QUANT

Deprecate CODEC_FLAG2_DROP_FRAME_TIMECODE

Deprecate CODEC_FLAG2_INTRA_VLC.

Fixes SA00072, SA00073, SA10150, SA10151, Issue2076

Improves SA10153
Signed-off-by: Anton Khirnov <anton@khirnov.net>

mjpeg: treat external huffman table setup failure as codec init failure if external huffman table use requested
Signed-off-by: Anton Khirnov <anton@khirnov.net>

It's been unused for the last 3 years.

Deprecate CODEC_FLAG2_AUD.

Deprecate CODEC_FLAG2_FASTPSKIP.

Deprecate CODEC_FLAG2_8X8DCT.

Deprecate CODEC_FLAG2_MIXED_REFS.

Deprecate CODEC_FLAG2_BPYRAMID.

Deprecate CODEC_FLAG2_BPYRAMID.

Deprecate CODEC_FLAG2_INTRA_REFRESH.

Deprecate CODEC_FLAG2_SSIM.

Replace '_' with '-', merge psy-trellis into psy-rd and rename cqp->qp
Also fix typo pdy-rd -> psy-rd

coded_width is the width before lowres scaling,
coded_height is the height before lowres scaling.

It's only shadowing the AVOption with the same name.

Since RV40 is based on H.264, it uses the same interpolation filters and
functions for halfpel interpolation in one direction can be reused.
Unfortunately other cases are implemented differently and produce slightly
different output because of rounding.
Signed-off-by: Mans Rullgard <mans@mansr.com>