MJPEG Tools Svn

lavplay works correctly now with the latest libquicktime.

add reference fdct implementation
fix rounding bug in basic C/integer fdct implementation
add fdct_test
rename fdctref.c -> fdct.c for consistency with idct*.c

Separate out testing code into new func

fix -M 0 (don't initialize Despatcher if parallelism==0)

Needed a debug point for a problem - but the new code's a bit cleaner too.

improved mmx idct (faster and more accurate), add sse idct (slower but very accurate)

better quant_weight_coeff_sum_mmx -- explicit parallelism, better abs func

use movd_g2r/movd_r2g whenever possible

Convert predict_mmx from NASM to inline C assembly (from Gert Vervoort)

get rid of SSE quant_non_intra (which was broken, and abusing
program-wide SSE rounding flags which affected the ability to write
other SSE routines). Totally rewrite MMX quant_non_intra so that it
is 1) correct, and 2) faster than the SSE code was!