Translated and optimized for the Scenix SX by Nikolai Golovchenko
From: Dwayne Reid
...here is an example that takes the value in R1H,M,L and adds it to the value contained in AD2H,M,L (result in AD2H,M,L). You should be able to see how to extend it to more than 24 bits.
There is a similar version that does subtraction. I often extend to 40 bits or more - its easier to do this when manipulating integer values than it is to try and figure out which bits I can discard.
;now add new sample mov W, R1L ; add AD2L, W ;LS byte mov W, R1M ;middle byte snb C ;take R1M+1 if carry movsz W, ++R1M ;if R1M+1==0 skip addition to preserve carry add AD2M, W ; mov W, R1H ;MS byte snb C movsz W, ++R1H add AD2H, W ;
And 24-bit substraction: AD2 = AD2 - R1
mov W, R1L ; sub AD2L, W ;LS byte mov W, R1M ;middle byte sb C ;take middle byte + 1 if borrow movsz W, ++R1M ;if R1M+1==0 skip substraction to preserve borrow sub AD2M, W ; mov W, R1H ;MS byte sb C movsz W, ++R1H sub AD2H, W ;