Post by thread:STATUS flags

Hey all! This is just a little somthing that I wanted to verify... I am using 'subwf' on a 16C54 to compare two bytes, and I saw in the programmers manual how the Carry flag is set... the Zero flag is still set according to the result, correct? Really, I'm doing a greater than/less than/equal to thing, so, first I should check the Z flag to see if the result was zero, if it was, I'll branch, if not, then test the C flag and go from there, right? So, when using 'subwf': If (f) - (W) is positive, (f) > (W) / Z = 0, C = 1. If (f) - (W) is zero, (f) = (W) / Z = 1, C = 1. If (f) - (W) is negative, (f) < (W) / Z = 0, C = 0. Is everything I said about the flags correct? Thanks again, -- ~Keith spam_OUTtsk3000TakeThisOuTProdigy.Net http://pages.prodigy.net/tsk3000/

Keith Burzinski <.....tsk3000KILLspam@spam@Prodigy.Net> wrote: > I am using 'subwf' on a 16C54 to compare two bytes, and I saw in the > programmers manual how the Carry flag is set... the Zero flag is > still set according to the result, correct? Correct, Keith. > Really, I'm doing a greater than/less than/equal to thing, so, > first I should check the Z flag to see if the result was zero, if > it was, I'll branch, if not, then test the C flag and go from > there, right? Actually, it'll be faster to check the C flag first, then only check the Z flag if the C flag was set... That way, you only check one flag when the result is negative (a much more-likely occurence than a "0" result). > So, when using 'subwf': > If (f) - (W) is positive, > (f) > (W) / Z = 0, C = 1. > If (f) - (W) is zero, > (f) = (W) / Z = 1, C = 1. > If (f) - (W) is negative, > (f) < (W) / Z = 0, C = 0. > > Is everything I said about the flags correct? Yes, if you treat the contents of "f" and "w" as unsigned (i.e., non-negative) numbers. -Andy === Andrew Warren - fastfwdKILLspamix.netcom.com === Fast Forward Engineering - Vista, California === http://www.geocities.com/SiliconValley/2499

Andrew Warren wrote: {Quote hidden}> > > Really, I'm doing a greater than/less than/equal to thing, so, > > first I should check the Z flag to see if the result was zero, if > > it was, I'll branch, if not, then test the C flag and go from > > there, right? > > Actually, it'll be faster to check the C flag first, then only > check the Z flag if the C flag was set... That way, you only > check one flag when the result is negative (a much more-likely > occurence than a "0" result). > > > So, when using 'subwf': > > If (f) - (W) is positive, > > (f) > (W) / Z = 0, C = 1. > > If (f) - (W) is zero, > > (f) = (W) / Z = 1, C = 1. > > If (f) - (W) is negative, > > (f) < (W) / Z = 0, C = 0. > > > > Is everything I said about the flags correct? > > Yes, if you treat the contents of "f" and "w" as unsigned (i.e., > non-negative) numbers. Okay, this works! Thanks! Now I've run into another problem, though. What's the quickest way to check if a number is negative? Can I just test bit 7? I just tried subtracting 0 from the register using 'subwf', but that seems to _always_ leave the C flag set. And if I add 0 using 'addwf', it always leaves the C flag clear. Will just tesing bit 7 of the reg always work? Thanks again, -- ~Keith .....tsk3000KILLspam.....Prodigy.Net http://pages.prodigy.net/tsk3000/

Keith Burzinski <EraseMEtsk3000spam_OUTTakeThisOuTProdigy.Net> wrote: > What's the quickest way to check if a number is negative? Can I > just test bit 7? Yes. > I just tried subtracting 0 from the register using 'subwf', but > that seems to _always_ leave the C flag set. Of course it does; SUBWF treats the two numbers as UNSIGNED (i.e., non-negative), so subtracting 0 always gives (as far as SUBWF is concerned) a non-negative result. > And if I add 0 using 'addwf', it always leaves the C flag clear. This happens because the ADDWF instruction only sets the carry flag if the result is greater than 255, and adding 0 to an 8-bit number can NEVER result in an answer that's greater than 255. > Will just tesing bit 7 of the reg always work? Yes. -Andy === Andrew Warren - fastfwdspam_OUTix.netcom.com === Fast Forward Engineering - Vista, California === http://www.geocities.com/SiliconValley/2499