Function vcvtxd_f32_f64

#[target_feature(enable = "neon")]
pub fn vcvtxd_f32_f64(a: f64) -> f32

Floating-point convert to lower precision narrow, rounding to odd Arm's documentation