Replacing the _mm_cvtsi128_si64 Function

This function is used to copy the lower 64-bit integer in vector a.

For details about _mm_cvtsi128_si64, see Intrinsics Guide.

• Code on x86:

  all_zeros = _mm_cvtsi128_si64(zero_bytes) == 0;

• Alternative for Kunpeng processors:

  #include <arm_neon.h>
  all_zeros = vgetq_lane_u64(vreinterpretq_u64_s32(zero_bytes),0) == 0;