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

This is the remaining error, the output on the SPX samples,
respectively csi_miami_stereo_128_spx.eac3 and
csi_miami_5.1_256_spx.eac3, goes from:
stddev:    8.71 PSNR: 77.52 MAXDIFF:  235
stddev:24270.51 PSNR: 22.17 MAXDIFF:47166
to:
stddev:    0.12 PSNR:114.12 MAXDIFF:    1
stddev:    0.12 PSNR:114.73 MAXDIFF:    1
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

It was set to 1 instead of sqrt(3)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Should also improve decoding, but actually doesn't...
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>

When the decoder is deactivated, the x86-optimized versions are
not compiled, resulting in a link error.

The C version is unaffected, as it is part of the idctdsp
subsystem.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Also reduce the table duplication with SSE2 code, remove duplicated
macro parameters.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The AC3DecodeContext has a float (USE_FIXED=0) and an integer
(USE_FIXED=1) variant, both of which can be present in the same binary.
This is not only very confusing, but it also breaks horribly, when one
variant is used by code expecting the other.

This currently happens, because eac3dec.c is only compiled for the float
variant, but also used from ac3dec_fixed.c, which uses the integer
variant.

The result is memory corruption, leading to crashes.

So compile eac3dec.c once for each variant and adapt it, so that it
works with the integer variant.

A loss of precission and scaling bug has been fixed by the committer
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Found-by: Clang -fsanitize=shift
Reported-by: Thierry Foucu <tfoucu@google.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

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

Fixes assertion failure and race conditions
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* commit '247e370e':
  TDSC decoder

Conflicts:
	Changelog
	doc/general.texi
	libavcodec/Makefile
	libavcodec/allcodecs.c
	libavcodec/version.h
	tests/fate/video.mak
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '5a0e953c':
  mjpeg: Mark decoder family as thread safe

Conflicts:
	libavcodec/mjpegbdec.c
	libavcodec/mjpegdec.c
	libavcodec/mxpegdec.c
	libavcodec/sp5xdec.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '117b4327':
  lavc: Introduce AVCodec internal capabilities

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

* commit '9993a067':
  lavc: Improve thread locking error message

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

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>

No global variables are used and the VLC tables are allocated without
static elements. This will allow using a JPEG decoding context within
other decoders.

This field is designed for marking codec properties useful to lavc internals.
Two internal capabilities are added:
 - FF_CODEC_CAP_INIT_THREADSAFE: codec can be opened without locks;
 - FF_CODEC_CAP_INIT_CLEANUP: codec frees memory if initialization fails.

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

Found-by: kurosu
Reviewed-by: Andreas Cadhalpun <andreas.cadhalpun@googlemail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Previous-version-reviewed-by: Andreas Cadhalpun
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Found-by: Clang -fsanitize=shift
Reported-by: Thierry Foucu <tfoucu@google.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Instead check for all mov code-points when demuxing avi
and print a warning if a video codec is found like this.
Fixes a regression similar to the one described in ticket #4307.

The main difference consists in renaming properly labels, and
letting yasm select the gprs for skipping 1D transforms.
Previous-version-reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* commit '913aa9a4':
  libx264: Return more meaningful error codes
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '6bdae41d':
  matroskadec: Check memory allocations

Conflicts:
	libavformat/matroskadec.c

See: 3e2a5b33Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit 'c3bd1d60':
  formats: Check memory allocations

Conflicts:
	libavfilter/formats.c

See: 527ca398Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '5a2645ca':
  nutdec: Prevent leaks on memory error

Conflicts:
	libavformat/nutdec.c

See: 269845dbMerged-by: Michael Niedermayer <michaelni@gmx.at>

* commit 'edca1dd5':
  xcbgrab: Check av_strdup() allocation

Conflicts:
	libavdevice/xcbgrab.c

See: db442c87Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '17c45d4d':
  libtheora: Check frame allocation
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit 'a72d93da':
  mpegvideo_enc: Check AVCodecContext allocation

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

* commit '78c89228':
  dca: Document the permuted order in the lfe_fir_64 and lfe_fir_128 tables
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Also add av_cold to ff_hevcdsp_init_arm.
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>