Add add_pixels4/8() to h264dsp, and remove add_pixels4 from dsputil.

These functions are mostly H264-specific (the only other user I can
spot is bink), and this allows us to special-case some functionality
for H264. Also remove the 16-bit-coeff with >8bpp versions (unused)
and merge the duplicate 32-bit-coeff for >8bpp (identical).

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

sanm: Use the correct height variable in the decoded_size checks

Fixes integer overflow and out of array accesses

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

sanm: init sizes with the full frame size

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

sanm: reset frame sizes on buffer destruction

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

sanm: add forgotten check for decoded_size in old_codec37()

Fixes out of array accesses

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffmpeg: fix streamcopy of lsf mp3.

Fixes streamcopy of bear_mpeg4_mp3.avi
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

doc/filters: rework idet filter description

doc/filters: apply minor fixes to il filter docs

doc/filters: fix position of il filter docs

Use simple nop codes for <= sse (rather than <= mmx).

The "CPU: CentaurHauls family 6 model 9 stepping 8" family of CPUs
(flags: fpu vme de pse tsc msr cx8 sep mtrr pge mov pat mmx fxsr sse
up rng rng_en ace ace_en) SIGILLs on long nop codes.

Change-Id: I7e7c52a2191006df30a9aadbc40d481a1db89106

h264chroma: remove duplicate 9/10 bit functions.

Also use the resulting 16bpp functions for anything >8 and <=16, not just
9 and 10. This fixes 12 and 14bpp H264 support.

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

ffv1enc: better heuristic to calculate initial states

Slightly improves compression of 2pass files

Tested-by: "Peter B." <pb@das-werkstatt.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

libfdk-aacenc: Actually check for upper bounds of cutoff

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

lavfi/amerge: make description consistent with documentation

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

Print mxf Universal Label if the user requested verbose logging.

Fixes ticket #2198

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  sparc: dsputil_vis: Fix silly variable name search and replace typo

Conflicts:
libavcodec/sparc/dsputil_vis.c

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

evrcdec: fix wrong condition

Fixes CID980005 and CID980004.

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

get_bits: return pointer to buffer that is the result of the alignment.

This allows more transparent mixing of get_bits and whole-byte access
without having to touch get_bits internals.

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

bit_depth_template: don't depend on dsputil.h.

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

Move ff_emulated_edge_mc prototypes to videodsp.

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

h264 intra pred: add missing includes.

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

mlpdec: check ch_assign more completely

Fixes CID970924

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

sparc: dsputil_vis: Fix silly variable name search and replace typo

lavfi/il: filter out hwaccels formats from the list of supported formats

lavf/matroskaenc: provide more feedback in case of unsupported codec

Also return meaningful error code.

lavd/v4l2: fix error logic when enumerating standards

Set ret to 0 before enumerating standards in v4l2_set_parameters(), avoid
use of uninitialized variable.

Regression introduced in 60950adc18fe145a235211e75da68ab07123fcaa.

tiff: dont leave geotag_count in an invalid state on errors.

Fixes out of array reads

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

tiff: check for failure in search_keyval()

Fixes null pointer dereference

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

pictordec: fix cga palette index limit

Fixes out of array read

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Fix division by 0 due to audio frame size

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

dpx: include offset in the  total_size calculation

Fixes out of array reads

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

dpx: Fix rounding in 10bit total_size calculation

This could have caused out of array reads

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

pcx: Add missing padding to scanline buffer

Fixes out of array read

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

pcx: use init_get_bits8()

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

lavd/v4l2: implement consistent error handling

In particular, avoid use of non-thread-safe strerror(), and store errno
before calling av_log().

tests: add fate-lavfi-il test

pcx: fix rounding in bytes_per_scanline check

Fixes out of array reads

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Support more atrac3-in-mkv samples.

The mkv demuxer sometimes finds 12 bytes of realmedia
extradata after the matroska real audio properties.

Revert "Merge commit 'db0a943266be29ff0596872ebb418dfed75d00de'"

This reverts commit d80820655362645ae1c65d457287d7d12c04a44f, reversing
changes made to 30085bd43e35acf931429dc445db5bc1869be022.

This is not needed

Requested-by: Marton Balint

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  sws: disable yuv2rgb warning for planar rgb.

Conflicts:
libswscale/yuv2rgb.c

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

Merge commit '7b41c24c5fa221b55a9302efaf1a9eeb38b12551'

* commit '7b41c24c5fa221b55a9302efaf1a9eeb38b12551':
  sws: dont enable chrSrcHSubSample for planar RGB

Conflicts:
libswscale/utils.c

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

lavfi/il: move query_formats() definition up

Respect order of call during the filter lifecycle, improve readability.

lavfi/il: simplify/generalize linesize computation

Rely on generic utilities for computing each plane linesize.

In particular, add support to NV12/21 formats and avoid use of
PIX_FMT_PLANAR pixdesc flag, whose semantics is questionable.

It also fixes various crashes.

Merge commit 'e58013dd8f00e17dd98ba64e41c72f1d152f6608'

* commit 'e58013dd8f00e17dd98ba64e41c72f1d152f6608':
  sws: use planarRgbToRgbWrapper only for 8bit per component

Conflicts:
libswscale/swscale_unscaled.c

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

Merge commit 'd6d5ef5534d582f9dbaf18ac2605cf5bb72cd821'

* commit 'd6d5ef5534d582f9dbaf18ac2605cf5bb72cd821':
  sws: GBRP9, GBRP10, and GBRP16 output support

Conflicts:
libswscale/output.c
libswscale/utils.c
tests/ref/lavfi/pixdesc
tests/ref/lavfi/pixfmts_copy
tests/ref/lavfi/pixfmts_null
tests/ref/lavfi/pixfmts_scale
tests/ref/lavfi/pixfmts_vflip

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

Merge commit '81726a4f0b8a43e19898e2a36fdde80583bafff0'

* commit '81726a4f0b8a43e19898e2a36fdde80583bafff0':
  FATE: add tests for additional flavors of asf cover art
  asfdec: do not assume every AVStream has a corresponding ASFStream

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

Merge commit 'f5fac6f77752931347ab302563802dcaa49c2419'

* commit 'f5fac6f77752931347ab302563802dcaa49c2419':
  asfdec: support reading ID3v2 tags in ASF files

Conflicts:
Changelog
libavformat/asfdec.c

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

Merge commit 'db0a943266be29ff0596872ebb418dfed75d00de'

* commit 'db0a943266be29ff0596872ebb418dfed75d00de':
  avplay: apply the stream sample_aspect_ratio to decoded video frames

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

Merge commit '84b721db366c0734fdfd23c8daaa7da7da21f761'

* commit '84b721db366c0734fdfd23c8daaa7da7da21f761':
  asfdec: also read Metadata Library Object

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

Merge commit '61f9ad2dfcb3f98b7ac5777d19d0e7b61d0be01e'

* commit '61f9ad2dfcb3f98b7ac5777d19d0e7b61d0be01e':
  asfdec: read the full Metadata Object, not just aspect ratio information

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

Merge commit '36fab50e90d15352e403e4cc210890810f2fb4e2'

* commit '36fab50e90d15352e403e4cc210890810f2fb4e2':
  asfdec: silence a warning
  mss4, ra288: Remove unused DSPContext local codec context members

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

Do not write invalid matroska files.

Some real codecs are supported by the matroska specification
(and FFmpeg's matroska demuxer) but not the FFmpeg muxer.

imgconvert: use get_pix_fmt_score in avcodec_find_best_pix_fmt_of_2()

This fixes various poorly choosen pixel format cases.

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

imgconvert: add get_pix_fmt_score()

get_pix_fmt_score() returns a score representing the amount
of loss when converting a pixel format

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

lavfi/il: fix typo in Copyright

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

lavfi: remove mp=il filter

Native il video filter is now available.

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

lavfi: port MP il filter

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

targa: Fix y check in advance_line

Fixes out of array accesses

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

lavf: add tee pseudo-muxer.

sws: disable yuv2rgb warning for planar rgb.

planar rgb formats do not use the table

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

rv10: remove check that is always true

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

sws: dont enable chrSrcHSubSample for planar RGB

This code path is not implemented and makes not much sense to implement
either.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

sws: use planarRgbToRgbWrapper only for 8bit per component

The function doesnt support >8bit currently

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

sws: GBRP9, GBRP10, and GBRP16 output support

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

FATE: add tests for additional flavors of asf cover art

asfdec: support reading ID3v2 tags in ASF files

Yes, these files do exist

Signed-off-by: Vladimir Pantelic <vladoman@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

avplay: apply the stream sample_aspect_ratio to decoded video frames

If there is a sample_aspect_ratio in the stream, then apply it to every
decoded frame in the same way as avconv does. This also makes sure that
the avfilter chain has access to the aspect ratio.

Signed-off-by: Vladimir Pantelic <vladoman@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

asfdec: also read Metadata Library Object

In some ASF files this objects holds cover art and other tags. Compared to
Metadata Object it can also hold GUIDs, but we ignore these for now.

Signed-off-by: Vladimir Pantelic <vladoman@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

asfdec: read the full Metadata Object, not just aspect ratio information

Use the same get_tag()/get_value() as for the Extended Content Description
but handle the 16 bit vs 32 bit difference for type 2 (BOOL)

Signed-off-by: Vladimir Pantelic <vladoman@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

asfdec: silence a warning

Signed-off-by: Vladimir Pantelic <vladoman@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

asfdec: do not assume every AVStream has a corresponding ASFStream

This won't be true for ID3 attached picture.

Also stop allocating now useless dummy ASFStreams for ASF native
attached pictures.

motionpixels: Check that the vlc table has been fully inited

Fixes out of array reads

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

motionpixels: Propagate errors in vlc table init

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind

Call the common 7.1 wide channel layout "7.1(wide)".

Rename the uncommon variant with side speakers to "7.1(wide-side)".

lavu/internal: rename macro E(x) to EXPAND(x)

Fixes bunch of warnings of E redefinition under clang.

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

lavfi/histogram: make waveform mode more useful

Now it displays all color components (not just luma)
either in parade or overlay and also works with RGB colorspace.

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

paf, sanm: remove unused header after b05830169d

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

Fix libpostproc compilation on x86_32 with disable-optimizations.

doc/filters: end table in histogram docs

doc/demuxers: add docs for rawvideo demuxer

lavf/rawvideodec: fix/extend option descriptions

doc/demuxers: fix copy&paste error in "@c man end" line

Use proper name for ending chapter.

doc/demuxers: sort demuxers by name

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  dsputil: Move fdct function declarations to dct.h

Conflicts:
libavcodec/dsputil.h

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

Merge commit '218aefce4472dc02ee3f12830a9a894bf7916da9'

* commit '218aefce4472dc02ee3f12830a9a894bf7916da9':
  dsputil: Move LOCAL_ALIGNED macros to libavutil

Conflicts:
libavcodec/dvdec.c
libavcodec/imc.c
libavcodec/mpegvideo_motion.c

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

mss4, ra288: Remove unused DSPContext local codec context members

Merge commit '059866eb17cd3c66422a61d3e9c3305f55830a50'

* commit '059866eb17cd3c66422a61d3e9c3305f55830a50':
  dsputil: Move WRAPPER8_16_SQ macro to the only place it is used

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

rnd_avg: fix author attribution

Reference:
commit 41fda91d097b85542981a4779e36f8eefda7ee69
Author: BERO <bero@geocities.co.jp>
Date:   Wed May 14 17:46:55 2003 +0000

    aligned dsputil (for sh4) patch by (BERO <bero at geocities dot co dot jp>)

    Originally committed as revision 1880 to svn://svn.ffmpeg.org/ffmpeg/trunk

commit 8dbe5856410ee10ac4a9368105149efdacfac994
Author: Oskar Arvidsson <oskar@irock.se>
Date:   Tue Mar 29 17:48:59 2011 +0200

    Adds 8-, 9- and 10-bit versions of some of the functions used by the h264 decoder.

    This patch lets e.g. dsputil_init chose dsp functions with respect to
    the bit depth to decode. The naming scheme of bit depth dependent
    functions is <base name>_<bit depth>[_<prefix>] (i.e. the old
    clear_blocks_c is now named clear_blocks_8_c).

    Note: Some of the functions for high bit depth is not dependent on the
    bit depth, but only on the pixel size. This leaves some room for
    optimizing binary size.

    Preparatory patch for high bit depth h264 decoding support.

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

Merge commit 'bf6b3ec924b4fb64d14df33077f4d4541d525dbf'

* commit 'bf6b3ec924b4fb64d14df33077f4d4541d525dbf':
  dsputil: Move rnd_avg inline functions to a separate header

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

Merge commit 'b9c2408b19dc9408e9dc0ce4a099af529bbfba3b'

* commit 'b9c2408b19dc9408e9dc0ce4a099af529bbfba3b':
  dsputil: Remove commented-out, unused function declarations

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

Merge commit '35685a3c2a1ec09f3c62dcfc4368fe9e92bcddf6'

* commit '35685a3c2a1ec09f3c62dcfc4368fe9e92bcddf6':
  dsputil: Move ff_shrink* function declarations to separate header
  dsputil: Move ff_svq3 function declarations to a separate header

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

Merge commit '6a4430407404bfe55581e66a86a496c9adf6fb44'

* commit '6a4430407404bfe55581e66a86a496c9adf6fb44':
  dsputil: Move ff_h264_idct function declarations to a separate header

Conflicts:
libavcodec/dsputil.h

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

lavfi/histogram: make it C90 compliant

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

Merge commit 'b9ba5253dd1232be4b48cfe61c31ff4b3de3d10a'

* commit 'b9ba5253dd1232be4b48cfe61c31ff4b3de3d10a':
  dsputil: Move copy_block functions to a separate header
  dsputil: Drop unused functions copy_block{2|4|16}

No functions are droped because they are used in several places.

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

histogram filter

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

Merge commit 'b9ee5f2cab3ffe1c962e542346b1ed61394864ec'

* commit 'b9ee5f2cab3ffe1c962e542346b1ed61394864ec':
  indeo3: replace use of copy_block4 with put_pixels
  mjpegdec: use put_pixels instead of copy_block8

Conflicts:
libavcodec/mjpegdec.c

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

Merge commit 'b5b7b75e91e291c810ebd01868a42b53f267a87c'

* commit 'b5b7b75e91e291c810ebd01868a42b53f267a87c':
  dsputil: Move get_penalty_factor() to the only place it is used.

Conflicts:
libavcodec/dsputil.h

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

Merge commit '0b016eb99d38738e2c53e36549a4732a0f863b2e'

* commit '0b016eb99d38738e2c53e36549a4732a0f863b2e':
  dsputil: Move ff_block_permute to mpegvideo_enc

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

Merge remote-tracking branch 'rdp/dshow_shared'

* rdp/dshow_shared:
  dshow: attempt to build shared as well as static

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

sparc: Revert some parameter name changes in dsputil_vis.c

They were renamed by mistake in 25841df.
This should fix compilation failures on sparc.

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