why no floats in kernel?

I was wondering why floats cannot be used in the kernel, I did read it some where, but I also notice befs and some other driver does use floats.

anyways, the problem is, I need to know how to turn this math into ints, yes I know I can bit shift but it doesn't help initially till I get around the floating point result.

( sound rate / sample rate ) * (note freq / sample freq)

Thanks in advance
Chris Ison