Source for a small fire routine. Doesn't actually work on my computer. To answer a question proposed by the author about size optimization, inc al is bigger than inc ax because intel doesn't have an inc /r8 instruction. inc ax is of the form inc /r16 which compiles to db 40+rw, where rw = 0 for ax, 1 for cx, 2 for dx, 3 for bx, ... inc al is of the form inc r/m8 and compiles to db FE /1 where /1 = c0 for al, c4 for ah, c1 for bl..., 05 for a memory reference.... See an intel document for more details.
|
