Talk:Assembler code example of parity calculation
|
|
Does this really belong in an encyclopedia? Egil 14:04 Mar 10, 2003 (UTC)
- Yes. however, the calculation is wrong. Even parity means an even number of 1 bits; odd parity means an odd number of them. You can't test this by looking at the least significant bit, only by looking at the least-significant bit of the sum of all the bits. Waveguy
- Well, by the definition he is using, the first one at the top of the article Parity, the calculation is correct. However, as you said, this is not the definition applicable to computing and the article needs to be rewritten. -- RTC 01:28, 12 Sep 2003 (UTC)
- Ok, here's the updated code fragment. Caveat: I haven't tried running this code yet, but here it is...
- Ok, here's the updated code fragment. Caveat: I haven't tried running this code yet, but here it is...
- calculate parity of an 8-bit value in the bx register
mov ax,bx
shr ax,1
xor ax,bx
shr ax,2
xor ax,bx
shr ax,4
xor ax,bx
test ax,1
- the zero flag is now set if parity is even, and clear if parity is odd.
- Of course on the x86 architecture there is an even simpler algorithm: use the Parity Flag of the status register and let the hardware do it automatically for you :-) I know this works.
; test the parity of the 8-bit value in the AL register:
or al,al
; the parity flag (PF) is now set if parity is even, and clear if parity is odd.
- -- RTC 17:50, 15 Sep 2003 (UTC)