merge from libav: 585dc1ae

If the metadata packet is corrupted, flv_read_metabody can accidentally
read past the start of the next packet. If the start of the  next packet
had been flushed out of the IO buffer, we would be unable to seek to
the right position (on a nonseekable stream).

Prefer to clearly error out instead of silently  trying to read from a
desynced stream which will only be interpreted as garbage.

This reverts commit ef7fe81b.

Signed-off-by: Marton Balint <cus@passwd.hu>

Signed-off-by: Marton Balint <cus@passwd.hu>

Deprecated since October 2015.

* commit '193b0918':
  thread: Define ff_mutex_* macros as stub functions when threads are disabled
Merged-by: James Almer <jamrial@gmail.com>

* commit 'b200a2c8':
  examples: Fixed and extended Doxygen documentation
Merged-by: James Almer <jamrial@gmail.com>

* commit 'efddf2c0':
  decode: Initialize ret before using it
  filtfmts-test: Mark filter as const

This commit is a noop, see
3c14547e
960b4d47Merged-by: James Almer <jamrial@gmail.com>

* commit 'f5950b8f':
  lavfi: Drop unused and empty header file
Merged-by: James Almer <jamrial@gmail.com>

* commit 'd32d59bc':
  matroska: Read only the data written in the scratch buffer

This commit is a noop, see fd59207cMerged-by: James Almer <jamrial@gmail.com>

* commit '6ac0e781':
  mpeg4videodec: raise an error if sprite_trajectory.table is NULL
Merged-by: James Almer <jamrial@gmail.com>

* commit 'b62ed687':
  configure: Better names for functions that sanitize input
Merged-by: James Almer <jamrial@gmail.com>

id3v2 code is compiled unconditionally and uses zlib when available.
Signed-off-by: James Almer <jamrial@gmail.com>

* commit '604fbb31':
  mov: Move code shared with CAF to a separate file
Merged-by: James Almer <jamrial@gmail.com>

* commit '3d197514':
  qsvenc: Allow use of hw_device_ctx to make the internal session
  qsvdec: Allow use of hw_device_ctx to make the internal session
  qsv: Add ability to create a session from a device
  doc: Add VAAPI encoders
  vaapi_encode: Add VP9 support
  vp9: Add bsf to fix reordering in raw streams

This commit is a noop, see
49419925
bde04604
dc81f1a2
91c3b50d
8aa3c2df
28aedeedMerged-by: James Almer <jamrial@gmail.com>

* commit '0cf949a0':
  vp9: Add bsf to merge superframes

[14:25:37] <@BBB> the functional change is wrong
[14:25:48] <@BBB> it allows recusrively packing superframes in superframes
[14:26:37] <@BBB> so I would ignore the remainder

See 2e6636aaMerged-by: James Almer <jamrial@gmail.com>

* commit 'f64d1100':
  avconv: Flush output BSFs when encode reaches EOF
  vf_deinterlace_vaapi: Add support for field rate output
  pthread_frame: Propagate sw_pix_fmt across threads

This commit is a noop, see
ebce1332
bff7bec1
bc4e33ceMerged-by: James Almer <jamrial@gmail.com>

* commit '10f4511f':
  libavutil: Make LOCAL_ALIGNED(xx be equal to LOCAL_ALIGNED_xx(

Also added LOCAL_ALIGNED_4 as it's used in vp8 decoder, and
simplified the configure defines.
Merged-by: James Almer <jamrial@gmail.com>

Signed-off-by: Mateusz Brzostek <mateuszb@poczta.onet.pl>

Fixes the following gcc warning:
libavcodec/bitstream_filter.c:39:12: warning: return discards 'const' qualifier from pointer target type

According to EBU tech 3285 supplement 3 the dwPosPeakOfPeaks field
should contain the absolute position to the maximum audio sample value,
but the current implementation writes the relative peak frame index
instead.

Fix the issue by writing the "unknown" value (-1) for now until the
feature is implemented correctly.

Previous version reviewed-by: Peter Bubestinger <p.bubestinger@av-rd.com>
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Tobias Rapp <t.rapp@noa-archive.com>

This prevents potential ABI issues with GetByteContext.
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>

This prevents potential ABI issues with GetBitContext.
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>

Signed-off-by: Zhong Li <zhong.li@intel.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

This patch uses dithering in DITHER_COPY macro only if
it was not used option '-sws_dither none'.
With option '-sws_dither none' it uses downshift.

For human eye dithering is OK, for video codecs not necessarily.
If user don't want to use dithering, we should respect that.
Signed-off-by: Mateusz Brzostek <mateuszb@poczta.onet.pl>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Use immediate unsigned saturation for clip to max saving one vector register.
Signed-off-by: Kaustubh Raste <kaustubh.raste@imgtec.com>
Reviewed-by: Manojkumar Bhosale <Manojkumar.Bhosale@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Replace generic with block size specific function.
Load the specific destination bytes instead of MSA load and pack.
Signed-off-by: Kaustubh Raste <kaustubh.raste@imgtec.com>
Reviewed-by: Manojkumar Bhosale <Manojkumar.Bhosale@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Remove loops and unroll as block sizes are known.
Signed-off-by: Kaustubh Raste <kaustubh.raste@imgtec.com>
Reviewed-by: Manojkumar Bhosale <Manojkumar.Bhosale@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Technically _tzcnt* intrinsics are only available when the BMI
instruction set is present. However the instruction encoding
degrades to "rep bsf" on older processors.

Clang for Windows debatably restricts the _tzcnt* instrinics behind
the __BMI__ architecture define, so check for its presence or
exclude the usage of these intrinics when clang is present.

See also:
https://ffmpeg.org/pipermail/ffmpeg-devel/2015-November/183404.html
https://bugs.llvm.org/show_bug.cgi?id=30506
http://lists.llvm.org/pipermail/cfe-dev/2016-October/051034.htmlSigned-off-by: Dale Curtis <dalecurtis@chromium.org>
Reviewed-by: Matt Oliver <protogonoi@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '26d9b603':
  hevc: Avoid using LOCAL_ALIGNED for 4 byte alignment
Merged-by: James Almer <jamrial@gmail.com>

* commit '163cc67b':
  takdec: Use ISO C printf conversion specifiers where appropriate
  dcadec: remove extra indirection
  hevcdec: Use LOCAL_ALIGNED_* for declaring local variables with alignment
  arm: Always build the hevcdsp_init_arm.c file

This commit is a noop.
Merged-by: James Almer <jamrial@gmail.com>

Signed-off-by: Zhong Li <zhong.li@intel.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Code suggested by ubitux
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Enables demuxing of Ambisonics content coded with channel mapping 2
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '5a969f64':
  jack: Drop support for old (2012) JACK versions
Merged-by: James Almer <jamrial@gmail.com>

* commit '1bd986ed':
  hwcontext: Move NONE to the be the first member of AVHWDeviceType
Merged-by: James Almer <jamrial@gmail.com>

* commit '0b9a237b':
  hevc: Add NEON 4x4 and 8x8 IDCT

[15:12:59] <@ubitux> hevc_idct_4x4_8_c: 389.1
[15:13:00] <@ubitux> hevc_idct_4x4_8_neon: 126.6
[15:13:02] <@ubitux> our ^
[15:13:06] <@ubitux> hevc_idct_4x4_8_c: 389.3
[15:13:08] <@ubitux> hevc_idct_4x4_8_neon: 107.8
[15:13:10] <@ubitux> hevc_idct_4x4_10_c: 418.6
[15:13:12] <@ubitux> hevc_idct_4x4_10_neon: 108.1
[15:13:14] <@ubitux> libav ^
[15:13:30] <@ubitux> so yeah, we can probably trash our versions here
Merged-by: James Almer <jamrial@gmail.com>