* commit '577326d4':
  lavc: deprecate refcounted_frames field
Merged-by: James Almer <jamrial@gmail.com>

* commit '3ad82579':
  hwcontext_cuda: implement frames_get_constraints

This commit is a noop, c16fe143Merged-by: James Almer <jamrial@gmail.com>

* commit 'fd9212f2':
  Mark some arrays that never change as const.
Merged-by: James Almer <jamrial@gmail.com>

* commit 'b420a27e':
  avconv: allow -b to be used with streamcopy

This commit is a noop. We already have this functionality.
Merged-by: James Almer <jamrial@gmail.com>

* commit 'ab2539bd':
  ffv1: Convert to the new bitstream reader
  h261dec: Convert to the new bitstream reader
  shorten: Convert to the new bitstream reader
  ralf: Convert to the new bitstream reader
  loco: Convert to the new bitstream reader
  fic: Convert to the new bitstream reader
  dirac: Convert to the new bitstream reader
  cavs: Convert to the new bitstream reader
  aic: Convert to the new bitstream reader
  golomb: Convert to the new bitstream reader

This commit is a noop, see
http://ffmpeg.org/pipermail/ffmpeg-devel/2017-April/209609.htmlMerged-by: James Almer <jamrial@gmail.com>

* commit 'ab87af41':
  configure: Add proper weak dependency of avformat on network
Merged-by: James Almer <jamrial@gmail.com>

* commit '612cc071':
  pgssubdec: reset rle_data_len/rle_remaining_len on allocation error

This commit is a noop, see 842e98b4Merged-by: James Almer <jamrial@gmail.com>

* commit '708e84cd':
  mov: Avoid memcmp of uninitialised data
Merged-by: James Almer <jamrial@gmail.com>

* commit 'ca62236a':
  vaapi_encode: Add VP8 support
  vaapi_encode: Pass framerate parameters to driver
  vaapi_h264: Enable VBR mode
  vaapi_encode: Support VBR mode

This commit is a noop, see
ceb28c3c
2201c02e
be6546a4
d1acab82Merged-by: James Almer <jamrial@gmail.com>

* commit 'c5c66354':
  doc: add dash muxer
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit '7295b737':
  dashenc: add webm support
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit '01f1f017':
  dashenc: use avio_dynbuf instead of packet_write callback
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit 'dce2929e':
  dashenc: copy language and role metadata from streams assigned to sets
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit 'ca9bc9de':
  dashenc: default to one AdaptationSet per stream
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit 'efd2fc41':
  dashenc: allow assigning all streams of a media type to an AdaptationSet
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit '3d23a5f9':
  dashenc: add support for assigning streams to AdaptationSets
Merged-by: Rodger Combs <rodger.combs@gmail.com>

* commit '9df9309d':
  dashenc: calculate stream bitrate from first segment if not available
Merged-by: Rodger Combs <rodger.combs@gmail.com>

This reverts commit 89c0fda5.

A different solution will be committed instead.
Signed-off-by: James Almer <jamrial@gmail.com>

The GnuTLS version is checked through the macro GNUTLS_VERSION_NUMBER,
but this wasn't introduced before 2.7.2. Building with older versions
of GnuTLS (using icc) warns:

src/libavformat/tls_gnutls.c(38): warning #193: zero used for undefined preprocessing identifier "GNUTLS_VERSION_NUMBER"
  #if HAVE_THREADS && GNUTLS_VERSION_NUMBER <= 0x020b00

This adds a fallback to the older, deprecated LIBGNUTLS_VERSION_NUMBER
macro.
Signed-off-by: Moritz Barsnick <barsnick@gmx.net>

Commit 598e4168 added use of
GNUTLS_E_PREMATURE_TERMINATION, which wasn't introduced to GnuTLS
before 2.99.x / 3.x. This fixes compilation with older versions.
Signed-off-by: Moritz Barsnick <barsnick@gmx.net>

Fixes #6649.

It's unnecessary after a call to av_packet_ref().
Signed-off-by: James Almer <jamrial@gmail.com>

Signed-off-by: James Almer <jamrial@gmail.com>

Signed-off-by: James Almer <jamrial@gmail.com>

It does the same thing as av_packet_ref().
Signed-off-by: James Almer <jamrial@gmail.com>

Signed-off-by: Thomas Mundt <tmundt75@gmail.com>

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

Signed-off-by: Paras Chadha <paraschadha18@gmail.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>

Fixes: runtime error: signed integer overflow: 16748560 + 2143729712 cannot be represented in type 'int'
Fixes: 3202/clusterfuzz-testcase-minimized-4988291642294272

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: runtime error: shift exponent 42 is too large for 32-bit type 'unsigned int'
Fixes: 3410/clusterfuzz-testcase-minimized-5313377960198144

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: runtime error: signed integer overflow: -1562477869 + -691460395 cannot be represented in type 'int'
Fixes: 3196/clusterfuzz-testcase-minimized-4528307146063872

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Duration depends on the selected subsong and thus must be queried after
selecting the subsong. There is no compelling reason to query other
metadata earlier either.

Tested with libopenmpt version: 0.2.8760-beta27
Libopenmpt configure options: --without-ogg --without-vorbis
--without-vorbisfile --without-portaudio --without-portaudiocpp
--without-mpg123 --without-pulseaudio --without-sndfile --without-flac
Signed-off-by: Jörn Heusipp <osmanx@problemloesungsmaschine.de>
Signed-off-by: Josh de Kock <josh@itanimul.li>

Replace generic with block size specific function.
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: Kaustubh Raste <kaustubh.raste@imgtec.com>
Reviewed-by: Manojkumar Bhosale <Manojkumar.Bhosale@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Load the specific destination bytes instead of MSA load and pack.
Pack the data to half word before clipping.
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>

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

Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

In case we are short of queued buffers, at first v4l2_buffer was enqueued to kernel so it's not owned by
user-space anymore. After that it's timestamp field was read, but it might be overwritten by driver at
that moment. It resulted in invalid timestamp sometimes.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>