Function vcvtnh_s32_f16

#[target_feature(enable = "neon", enable = "fp16")]
pub fn vcvtnh_s32_f16(a: f16) -> i32

Floating-point convert to integer, rounding to nearest with ties to even Arm's documentation