Function mul_div_u64 

pub fn mul_div_u64(value: u64, numerator: u64, denom: u64) -> u64

Computes (value*numerator)/denom without overflow, as long as both numerator*denom and the overall result fit into u64 (which is the case for our time conversions).