use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
authorMichael Niedermayer <michaelni@gmx.at>
Sat, 25 Aug 2007 03:00:51 +0000 (03:00 +0000)
committerMichael Niedermayer <michaelni@gmx.at>
Sat, 25 Aug 2007 03:00:51 +0000 (03:00 +0000)
commitd593e329832a432777218eb92469bad10594a3c5
tree729557ca1d10a888887e887be4c86c87d84e8a10
parent830bf1f20d8087e76e3070840c02be7056cf781f
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with
the old 32bit code)
disable mmx/sse2 optimizations as they need a rewrite now

Originally committed as revision 10218 to svn://svn.ffmpeg.org/ffmpeg/trunk
libavcodec/dsputil.h
libavcodec/i386/dsputil_mmx.c
libavcodec/snow.c
libavcodec/snow.h
tests/ffmpeg.regression.ref
tests/rotozoom.regression.ref
tests/seek.regression.ref