# Some quick-n-dirty ways to create floats.
fn fill-in-rational _out: (addr float), nr: int, dr: int {
var out/edi: (addr float) <- copy _out
var result/xmm0: float <- convert nr
var divisor/xmm1: float <- convert dr
result <- divide divisor
copy-to *out, result
}
fn fill-in-sqrt _out: (addr float), n: int {
var out/edi: (addr float) <- copy _out
var result/xmm0: float <- convert n
result <- square-root result
copy-to *out, result
}
fn rational nr: int, dr: int -> _/xmm0: float {
var result/xmm0: float <- convert nr
var divisor/xmm1: float <- convert dr
result <- divide divisor
return result
}