Mail Archives: djgpp/1997/02/27/11:52:32
On Thu, 27 Feb 1997, John M. Aldrich wrote:
> Gurunandan R. Bhat wrote:
> >
> > greetings,
> > will a kind assembly guru pick this nit please: assuming that x+y
> > and x|y give me the same result, which one is faster?
>
> The logical OR is probably much faster, but those two are NOT equivalent
> unless x and y have no bits in common!
yes, thats precisely what i have. in my app (optimised fft actually), x
and y have no bits in common.
> If you can _guarantee_ that,
> then the OR is one of the fastest machine insructions there is. :)
thanks.
- Raw text -