example: -vf 'scale=256:256,mp=geq=gcd(X\,Y)-2:128:128'
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

useage example: -vf 'scale=256:256,mp=geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128'
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* qatar/master: (24 commits)
  mpegps: Use av_get_packet() instead of poorly emulating it.
  motionpixels: decode only the 111 complete frames for fate
  mpc8: Check out of bound bands limit
  xan: Prevent NULL dereference with missing palette
  xan: Check for out of bound reads in xan_huffman_decode()
  xan: Fixed out of bound accesses in xan_unpack()
  motionpixels: Prevent calling init_vlc() with invalid parameters
  shorten: Fix out of bound writes in fix_bitshift()
  dsicinav: Check for out of bounds writes
  tiertexseqv: Check for out of bound reads
  quickdraw: Check for out of bound reads
  dsicinav: Check for out of bounds reads
  motionpixels: Fix the size of workspace buffers
  motionpixels: Clear FF_INPUT_BUFFER_PADDING_SIZE bytes at the end of the temporary buffer
  wmavoice: Check for corrupted extra data
  wmavoice: Check for out of bound writes
  xan: Prevent NULL dereferences with missing reference frame
  bink: Prevent NULL dereferences with missing reference frame
  wavpack: Reset internal state on corrupted blocks
  wmapro: Validate the number of audio channels before using it
  ...

Conflicts:
	libavcodec/h264.c
	libavcodec/xan.c
	tests/ref/fate/motionpixels
Merged-by: Michael Niedermayer <michaelni@gmx.at>

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

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

Code based on ffdshow
Author unknown
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

This makes successive frames decode possible.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

The data pointers s->decoded[*] already take into account s->nwrap.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Some buffers must be mod 4 in width and/or height.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

wavpack_decode_block() supposes that it is called back with the exact
same buffer unless it has returned with an error. With multi-channels
files, wavpack_decode_frame() was breaking this assumption.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

* hexene/stagefright:
  libstagefright: start decode_thread() only after decode_frame() is called at least once.
  libstagefright: mark the dummy frame as keyframe.
  libstagefright: limit the output queue size
  libstagefright: return EOS if CustomSource::read() is called after decode_thread() returns
  libstagefright: set the correct frame size
Merged-by: Michael Niedermayer <michaelni@gmx.at>

This prevents the situation where EOS is passed as the first frame to the
h/w decoder and thus avoids a potential crash.

This makes sure decoder->read() returns if it is the first frame fed.

H264: Only wait before triggering ff_thread_setup_complete() until the next slice that contains a start-of-field/frame macroblock

This allows concurrent decoding of the last field/frame, rather than
only the last slice, of data packets with multiple NAL units packed
together.

This will fix the slowdown reported in e.g. bug 52.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

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

* qatar/master:
  vp6: Reset the internal state when aborting key frames header parsing
  vp56: Release old pictures after a resolution changes
  vp6: Check for huffman tree build errors
  vp56: Check for missing reference frame data
  cinepak: Fix invalid read access on extra data
  vmd: fix segfaults on corruped streams
  cook: Fix js_vlc_bits value validation for joint stereo
  segafilm: Check for memory allocation failures in segafilm demuxer.
  segafilm: Fix potential division by 0 on corrupted streams in the demuxer
  Fixed segfault on corrupted sega streams in the demuxer.
  Fixed deference of NULL pointer in motionpixels decoder.
  libx264: support 9- and 10-bit output.
  h264: correct implicit_weight for field-interlaced pictures.
  mpegvideo: set correct offset for edge emulation buffer.
  mpegvideo: fix position of bottom edge.

Conflicts:
	libavcodec/motionpixels.c
	libavcodec/mpegvideo.c
	libavcodec/version.h
	libavcodec/vmdav.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>