Function compute_clip 

pub fn compute_clip(
    input: &[f32],
    output: &mut [f32],
    minimum: f32,
    maximum: f32,
)

Compute contiguous Float32 clipping with MLAS SIMD.