Function fast_math::log2_raw

pub fn log2_raw(x: f32) -> f32

Compute a fast approximation of the base-2 logarithm of positive, finite, non-denormal x.

This will return unspecified nonsense if x is doesn't not satisfy those constraints. Use log2 if correct handling is required (at the expense of some speed).

The maximum relative error across all valid input is less than 0.022. The maximum absolute error is less than 0.009.

	Time (ns)
x.log2()	14.3
log2(x)	4.0
log2_raw(x)	2.7