Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>

This will make multi-channel implementation simpler.
Based partially on a patch by Andrew D'Addesio <modchipv12@gmail.com>.

Use get_bits_long() in decode_scalar().
Use unsigned int for decoded value.

Avoids unneeded interleaving and allows for reusing the AVFrame output buffer
as the internal buffer for 24-bit and 32-bit sample size.

Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>

This will simplify the multi-channel implementation.

It does not need to be unsigned.

It is already unconditionally set to 0 prior to this, so we can modify it
only when needed.

x_modified is just unnecessary, and final_val can be removed by simplifying
the unsigned-to-signed conversion.

avoids some unnecessary arithmetic in certain situations

reduces the number of parameters to decode_scalar() and slightly simplifies
the code

Use the ALACContext fields directly instead.

It is used to store a value read by avio_rb64().

according to the CAF specification:
"... the value for mChunkSize can be greater than the actual valid content
of the packet table chunk"

This provides a more precise identification of the compiler used.
Signed-off-by: Mans Rullgard <mans@mansr.com>

Reversing the lpc coefficient order simplifies indexing in
the filter.
Signed-off-by: Mans Rullgard <mans@mansr.com>

Signed-off-by: Mans Rullgard <mans@mansr.com>

This moves all VP3-specific function pointers from dsputil to a
new vp3dsp context.  There is no reason to ever use the VP3 IDCT
where an MPEG2 IDCT is expected or vice versa.
Signed-off-by: Mans Rullgard <mans@mansr.com>

Signed-off-by: Mans Rullgard <mans@mansr.com>