lavfi/colormatrix: switch to filter_frame.

vf_format: switch to filter_frame

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

vf_copy: switch to filter_frame

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

vf_null: switch to filter_frame

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

libavfilter: default to filter_frame when neither it nor start/slice/end is set.

this is needed for vf_null and a few others.

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

buffersrc: switch to filter_frame

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

buffersink: switch to filter_frame

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

lavfi/thumbnail: re-use ctx instead of inlink->dst.

lavfi/thumbnail: use avfilter_unref_bufferp() where appropriate.

lavfi/thumbnail: switch to filter_frame.

libavfilter/split: switch to filter_frame

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

vf_scale: switch to filter_frame

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

lavfi/sendcmd: expose the options for both filters.

lavfi/sendcmd: add FLAGS to AVOption array.

lavfi: add priv_class for some forgotten filters.

This allows to print the options available when using commands such as
ffmpeg -help full.

lavfi/mptestsrc: add FLAGS to AVOption array.

vf_blackframe: switch to filter_frame

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

vf_aspect: switch to filter_frame

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

vf_cropdetect: switch to filter_frame

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

vf_crop: switch to filter_frame

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

vf_drawbox: switch to filter frame

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

vflip: switch to filter_frame

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

Merge commit 'bb6c67bb36b136de10256f0999128df4a42f9ffc'

* commit 'bb6c67bb36b136de10256f0999128df4a42f9ffc':
  lavfi: remove vf_slicify

Conflicts:
doc/filters.texi
libavfilter/Makefile
libavfilter/allfilters.c
libavfilter/vf_slicify.c
tests/lavfi-regression.sh
tests/ref/lavfi/crop_scale_vflip

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

doc: add ffmpeg-bitstream-filters.texi file

vf_yadif: switch to filter_frame, this filter did not support slices

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

vf_fade: switch to filter_frame

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

lavfi: convert remaining input/output list compound literals to named objects.

This is following 568c70e79ee267426c15ef4603c69703f6a5884a.

vf_lut: switch to filter_frame

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

vf_hflip: switch to filter_frame

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

f_setpts: switch to filter_frame

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

f_settb: switch to filter_frame

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

src_movie: switch to filter_frame

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

vsrc_testsrc: switch to filter_frame

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

vsink_nullsink: switch to filter_frame

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

vf_unsharp: switch to filter_frame, this filter did not support slices

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

vf_transpose: switch to filter_frame

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vf_select: switch to filter_frame

The new code reads the input frame when its ready, the previous
code did read the input frame during start_frame at which point it
may not yet be available.

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

vf_showinfo: switch to filter_frame

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

vf_pad: switch to filter_frame

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Vivo demuxer

vf_pixdesctest: switch to filter_frame

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vf_libopencv: switch to filter_frame, this filter did not support slices

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

lavfi/geq: switch to filter_frame.

vf_gradfun: switch to filter_frame, this filter did not support slices

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

vf_frei0r: switch to filter_frame, the filter did not support slices

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

lavfi/edgedetect: switch to filter_frame.

vf_fps: switch to filter_frame

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

vf_fieldorder: switch to filter_frame, this filter did not support real slices

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

libavfilter/fifo: fix indent

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

vf_boxblur: cosmetics to reduce diff to qatar

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

vf_drawtext: switch to filter_frame, this filter did not support slices

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vf_delogo: switch to filter_frame, this filter did not support slices

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vf_hqdn3d: fix permissions

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

vf_delogo: fix order of operations in apply_delogo()

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

vf_boxblur: switch to filter_frame, boxblur did not support slices

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

libavfilter/fifo: switch to filter_frame, the filter did not support slices.

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

pgssubdec: remove unused variable

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

iff demuxer: 16SV support

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

vf_hqdn3d: switch to filter_frame

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

libavfilter: update doxy for filter_frame

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

libavfilter: Support using filter_frame for video

With this we can mix filters using filter_frame OR start/draw_slice/end

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

default_filter_frame: fix argument name

Based on code by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avfilter: check that filter_frame and slice variants are not both used

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

lavfi: replace filter_samples by filter_frame

Based on patch by Anton Khirnov
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vc1dec: add hacks to ensure static tables are initialized in open().

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Use err_out label for error-case cleanup.

Will simplify future changes (introducing proper
locking around opening/closing parsers).

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

atrac3: do not use init_static_data to init VLC data.

It would be called while registering the codec, which means
it needlessly wastes memory when it is never used.
Instead do the work when first opening the codec.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

avrdec: implement seeking

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

Merge commit 'af7d13ee4a4bf8d708f9b0598abb8f6e22b76de1'

* commit 'af7d13ee4a4bf8d708f9b0598abb8f6e22b76de1':
  asink_nullsink: plug a memory leak.
  x86: h264_idct: port to cpuflags
  x86: cpu: Drop unused HAVE_RWEFLAGS condition

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

Merge commit 'f5fa03660db16f9d78abc5a626438b4d0b54f563'

* commit 'f5fa03660db16f9d78abc5a626438b4d0b54f563':
  vble: Do not abort decoding when version is not 1
  lavr: do not pass consumed samples as a parameter to ff_audio_resample()
  lavr: correct the documentation for the ff_audio_resample() return value
  lavr: do not pass sample count as a parameter to ff_audio_convert()
  x86: h264_weight: port to cpuflags
  configure: Enable avconv filter dependencies automatically

Conflicts:
configure
libavcodec/x86/h264_weight.asm

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

avrdec: sample rate takes 3 bytes not 4

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

avrdec: fix typo in .long_name

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

lavfi: remove vf_slicify

The following commit will make it useless.

The crop_scale_vflip FATE test changes because of off-by-one differences
in output when vflipped slices are passed to sws.

asink_nullsink: plug a memory leak.

qpeg: reset palette on seeks

Fixes Ticket1921

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

nut: put nut PCM audio tags in front of avi tags.

This fixes correctly storing and identifying PCM in nut.

Based on patch by Luca Barbato
Found-by: durandal_1707
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

x86: h264_idct: port to cpuflags

x86: cpu: Drop unused HAVE_RWEFLAGS condition

The test for rweflags was dropped in a previous commit.

doc: add ffmpeg-protocols.texi manual

doc: move syntax and eval chapters from libavutil to ffmpeg-utils

Move utilities description to a dedicated manual, consistent with what I
did with the rest of the documentation.

doc: add libavdevice.texi and ffmpeg-devices.texi files

doc: add libavformat.texi and ffmpeg-formats.texi files

lavfi/drawtext: add support to expansion of generic expressions

vble: Do not abort decoding when version is not 1

Some combinations of OS, VirtualDub, and VBLE can accidentally
set the version to a value other than 1. Since no other version
of VBLE was ever released, simply warn about it.

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

lavr: do not pass consumed samples as a parameter to ff_audio_resample()

Since the resampler handles buffering of unconsumed samples internally, the
caller does not need this information.

lavr: correct the documentation for the ff_audio_resample() return value

lavr: do not pass sample count as a parameter to ff_audio_convert()

It will always be the number of samples in the input buffer, so just use that
directly instead of passing it as a separate parameter.

x86: h264_weight: port to cpuflags

mjpegdec: dont set got_picture to 0 per frame.

Fixes Ticket1915

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

mjpegdec: reset got_picture on errors, successful return, init and flush.

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

mjpegdec: replace returns by goto fail in frame decode.

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

mjpegdec: make sure ret is set before its used.

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

fate: add ADPCM IMA Dialogic OKI decoder test

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

h264: support invalid annex B in mp4

Fixes Ticket1914

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

configure: Enable avconv filter dependencies automatically

This makes sure minimal configurations such as
"--disable-everything --enable-avconv" will enable the filters
necessary for running avconv, instead of just keeping avconv
disabled (even if the user specified "--enable-avconv").

Signed-off-by: Martin Storsjö <martin@martin.st>

ffserver remove CONFIG_NOCUTILS check

Nothing sets CONFIG_NOCUTILS, i also dont remember anything setting it in
the past.

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

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  h264: set Picture.owner2 to the current thread

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

Merge commit '6e5cdf26281945ddea3aaf5eca4d127791f23ca8'

* commit '6e5cdf26281945ddea3aaf5eca4d127791f23ca8':
  h264: check ref_count validity for num_ref_idx_active_override_flag
  h264: add missing new line to log message
  dcadec: skip QMF on unused channels
  wavenc: write fact chunk sample count at the correct file position
  riff: do not add empty metadata tags in INFO chunk

Conflicts:
libavcodec/dcadec.c
libavcodec/h264.c
libavformat/riff.c

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

Merge commit '3c370f5abc55739a261534b9f9bdc739cedbbbb9'

* commit '3c370f5abc55739a261534b9f9bdc739cedbbbb9':
  riff: only warn on a bad INFO chunk code size instead of failing
  configure: Add separate list for libraries and use where appropriate
  x86: float_dsp: add SSE version of vector_fmul_scalar()

Conflicts:
configure
libavformat/riff.c
libavutil/x86/float_dsp.asm

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

Merge commit '284ea790d89441fa1e6b2d72d3c1ed6d61972f0b'

* commit '284ea790d89441fa1e6b2d72d3c1ed6d61972f0b':
  dsputil: move vector_fmul_scalar() to AVFloatDSPContext in libavutil
  aacenc: use the correct output buffer
  aacdec: fix signed overflows in lcg_random()
  base64: fix signed overflow in shift

Conflicts:
libavcodec/dsputil.c
libavutil/base64.c

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