make_uint8x16_t

createsimd.c

namespace {
/**
 * make_uint8x16_t initializes a SIMD register (uint8x16_t).
 * This is needed because, incredibly, the syntax uint8x16_t x = {1,2,3...}
 * is not recognized under Visual Studio! This is a workaround.
 * Using a std::initializer_list<uint8_t>  as a parameter resulted in
 * inefficient code. With the current approach, if the parameters are
 * compile-time constants,
 * GNU GCC compiles it to ldr, the same as uint8x16_t x = {1,2,3...}.
 * You should not use this function except for compile-time constant:
 * it is not efficient.
 */
uint8x16_t make_uint8x16_t(uint8_t x1, uint8_t x2, uint8_t x3, uint8_t x4,
                           uint8_t x5, uint8_t x6, uint8_t x7, uint8_t x8,
                           uint8_t x9, uint8_t x10, uint8_t x11, uint8_t x12,
                           uint8_t x13, uint8_t x14, uint8_t x15, uint8_t x16) {
  // Doing a load like so end ups generating worse code.
  // uint8_t array[16] = {x1, x2, x3, x4, x5, x6, x7, x8,
  //                     x9, x10,x11,x12,x13,x14,x15,x16};
  // return vld1q_u8(array);
  uint8x16_t x{};
  // incredibly, Visual Studio does not allow x[0] = x1
  x = vsetq_lane_u8(x1, x, 0);
  x = vsetq_lane_u8(x2, x, 1);
  x = vsetq_lane_u8(x3, x, 2);
  x = vsetq_lane_u8(x4, x, 3);
  x = vsetq_lane_u8(x5, x, 4);
  x = vsetq_lane_u8(x6, x, 5);
  x = vsetq_lane_u8(x7, x, 6);
  x = vsetq_lane_u8(x8, x, 7);
  x = vsetq_lane_u8(x1, x, 8);
  x = vsetq_lane_u8(x1, x, 9);
  x = vsetq_lane_u8(x1, x, 10);
  x = vsetq_lane_u8(x1, x, 11);
  x = vsetq_lane_u8(x1, x, 12);
  x = vsetq_lane_u8(x1, x, 13);
  x = vsetq_lane_u8(x1, x, 14);
  x = vsetq_lane_u8(x1, x, 15);
  return x;
}
} // namespace