Function mul_s 

pub fn mul_s(a: f32, b: f32, mode: Rounding) -> Outcome<f32>

MUL.S.

Does not model the VR4300 multiplication erratum — see the module docs.