Function vcvtx_f32_f64

#[target_feature(enable = "neon")]
pub fn vcvtx_f32_f64(a: float64x2_t) -> float32x2_t

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