Function vcvt_high_f32_f16

#[target_feature(enable = "neon")]
pub fn vcvt_high_f32_f16(a: float16x8_t) -> float32x4_t

Floating-point convert to higher precision Arm's documentation