Re: MMX speedup for Floyd Steinberg error diffusion

"Maarten Kronenburg"  wrote in message
>
> The 4 32-bit words can be stored in a single 128-bit xmm register.
> Then you copy and ****ft left and add/subtract, because
> 7 = 8 - 1 = 2^3 - 1 and 5 = 4 + 1 = 2^2 + 1 etc.
> Then you ****ft right because 16 = 2^4.
> The ****ft is the PSLLD/PSRLD and the add/subtract is the PADDD/PSUBD.

Now I see the data is in bytes. In that case it seems better to put 16
bytes
in an 128-bit xmm register, then put 8 bytes each time into 8 16-bit words
by ****fting and anding, and do the above with PSLLW/PSRLW and PADDW/PSUBW
in
the 16-bit words. Then the scaling mentioned is not needed because the
upper
8 bits in the 16-bit words should be zero.
Maarten.