To huh
bitRake is right:


@@:
xor edx,edx ;!!! why do you need to do it in each iteration?!
mov dl,byte ptr [esi]
inc esi
xor dl,al
shr eax,8
xor eax,dword ptr [edi+(edx*4)]
dec ecx
jnz @B

You have to do this way:


xor edx,edx
@@:
mov dl,byte ptr [esi]
inc esi
xor dl,al
shr eax,8
xor eax,dword ptr [edi+(edx*4)]
dec ecx
jnz @B

This way you will save as many clocks as (number in ecx before the loop) - 1
Posted on 2002-02-15 19:46:19 by The Svin
bitRAKE = Optimizing bastard

(j/k RAKE ;))
Posted on 2002-02-15 20:03:50 by Torch