Fixes Ticket #3642.

Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

also make a const array static
Fixes 2 warnings
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The used value should not matter as long as its within the array
as it is multiplied by 0
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

Found-by: Katerina Barone-Adesi
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

From 133 (unrolled av_intfloat32 C) to 59 cycles on Arrandale/Win64.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

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

This is not correct

Requested-by: mraulet
This reverts commit 973de9eb.

* commit 'd3f5b947':
  aarch64: opus NEON iMDCT and FFT
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '7c5ca546':
  configure: fix enable-libopus help string

See: 434ba17bMerged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '5e2ba41d':
  build: add avresample after avcodec to FFLIBS

Conflicts:
	Makefile
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Describe the option added in commit f1c16749Signed-off-by: Olivier Langlois <olivier@trillion01.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The value is already copied in the generic code so it should always match
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* commit '3b2fbe67':
  lavc: properly handle subtitle_header in avcodec_copy_context()

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

* commit 'efc7df6c':
  lavc: preserve the original private data in avcodec_copy_context()
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit 'c9281a01':
  lavf: drop the zero-sized packets hack

Conflicts:
	libavformat/mux.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '0c1959b0':
  lavf: add AVFMT_FLAG_BITEXACT.

Conflicts:
	doc/APIchanges
	libavformat/avformat.h
	libavformat/flacenc.c
	libavformat/movenc.c
	libavformat/oggenc.c
	libavformat/options_table.h
	libavformat/version.h
	tests/fate/video.mak
Merged-by: Michael Niedermayer <michaelni@gmx.at>

This also fixes linking failures in doc/examples which where apparently
caused by the linking order between avcodec and avresample
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

* commit 'b70d7a4a':
  lavc: add a native Opus decoder.

Conflicts:
	Changelog
	configure
	libavcodec/version.h

Fate tests pass with both avresample as well as swresample based opus decoder, but
are disabled (reference files are very large so i want to think a day or 2 about
if theres an alternative or if they could be avoided, they also dont match the
official samples)
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Opus celt decoding 11% faster and the iMDCT over 2.5 times faster on
Apple's A7.

avcodec might depend on avresample and with --as-needed required symbols
might be get removed if avresample is linked before avcodec.
Fixes link failures of avprobe and avplay on aarch64 with
--enable-neon-clobber-test.

If a non-NULL codec was passed to avcodec_alloc_context3(), private data
will be already allocated in dest.

There should not be any valid reason anymore for passing completely
empty packets to lavf.
OTOH side data-only packets can be useful.

Use it instead of checking CODEC_FLAG_BITEXACT in the first stream's
codec context.

Using codec options inside lavf is fragile and can easily break when the
muxing codec context is not the encoding context.

Initial implementation by Andrew D'Addesio <modchipv12@gmail.com> during
GSoC 2012.

Completion by Anton Khirnov <anton@khirnov.net>, sponsored by the
Mozilla Corporation.

Further contributions by:
Christophe Gisquet <christophe.gisquet@gmail.com>
Janne Grunau <janne-libav@jannau.net>
Luca Barbato <lu_zero@gentoo.org>

Problem: ffmpeg tries to read COVR atom data twice if MOV_EXPORT_ALL_METADATA is enabled.
If COVR atom is the last in the stream, a parsing of such file fails.

Solution: just return immediatelly after mov_read_covr
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>

With this the fate clients runtime cpu feature set should be visible
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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