Function core::arch::aarch64::vcvtx_high_f32_f64

pub unsafe fn vcvtx_high_f32_f64(a: float32x2_t, b: float64x2_t) -> float32x4_t

Available on AArch64 and target feature neon only.

浮点数转换为低精度 narrow，四舍五入为奇数

Arm’s documentation