[MPlayer-dev-eng] [PATCH] speedup and little bugfix for -vop eq2

[Arpi]

Hi,

> It fixes a small bug where a byte value is divided by 255.0 to convert
> to a float within [0.0, 1.0] and later multiplied by 256.0 to convert
> back. This makes the luminance lookup table more correct, although the
> visual difference is relatively small.
ok

> More importantly, it speeds up the inner loop a good bit, at least on
> x86 systems, the only one I have and can test on. Of course the code
> should work on all architectures, there's no asm stuff, just faster C.
> Comments on how it is done in detail are in the file.

i have problems with it. teh idea of using src[index[ instead of src++ is
ok, but the reverse counting isn't. If you do -dr to video ram, it will hurt
more than helps, as PCI/AGP doesn't like reverse/random indexing.
anyway that extra cmp doesn't matter, the loop will be slow due to memory
waits anyway.

i'll apply the patch wiht modifications


A'rpi / Astral & ESP-team

--
Developer of MPlayer, the Movie Player for Linux - http://www.MPlayerHQ.hu