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

* jamrial/adpcm:
  ADPCM IMA Radical decoder
  RedSpark demuxer
  RSD demuxer
  adpcm_thp: Allow the use of extradata for the adpcm table
  ADP demuxer
  ADPCM DTK decoder
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Thanks to ubitux for spotting that
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Somehow i managed to loose this fix before pushing
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

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

This is about 20-30% faster than slice threading
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

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

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

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

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

* qatar/master:
  indeo3: avoid writes without necessary alignment in copy_cell()
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '2c299d41':
  x86: sbrdsp: implement SSE2 qmf_pre_shuffle

Conflicts:
	libavcodec/x86/sbrdsp.asm
	libavcodec/x86/sbrdsp_init.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '16a4a18d':
  af_asyncts: fix offset calculation
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '72fbc968':
  avconv: use current syntax for the scale filter
Merged-by: Michael Niedermayer <michaelni@gmx.at>

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

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

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

There are several containers that support adpcm_thp (Also known as Gamecube DSP)
streams, but only thp files contain the coeff table and previous sample inside
each frame.
Some don't even contain previous sample information at all.

This change will make it easier to implement demuxers for said containers
without having to create a new decoder.
Signed-off-by: James Almer <jamrial@gmail.com>

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

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

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

clang did not show this one, so i missed it in the last batch
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

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

This should fix several issues with odd dimensions inputs.

lut, vflip, pad and crop video filters also need to be checked for such
issues. It's possible sws is also affected.

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

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

Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Cells starting at a position aligned to 8 pixels but wider than
4 blocks are copied with 3 blocks per loop. This creates problems on the
next loop iterations since the routine copying 2 blocks requires the
same alignment on some architectures like ARM NEON.