__lasx_xvsubwod_w_h

LASX intrinsic · group: Integer Computation

Synopsis

__m256i __lasx_xvsubwod_w_h (__m256i a, __m256i b)
#include <lasxintrin.h>
Instruction: xvsubwod.w.h xr, xr, xr
CPU Flags: LASX

Description

Subtract the odd-positioned signed 16-bit elements in b from the corresponding elements in a, extending both to 32-bit, and save the 32-bit result in dst.

Operation

for (int i = 0; i < 8; i++) {
  dst.word[i] = (s32)(s16)a.half[2 * i + 1] - (s32)(s16)b.half[2 * i + 1];
}

// Expands to:

if (0) {
  dst.word[0] = ((s32)((s16)a.half[1])) - ((s32)((s16)b.half[1]));
  dst.word[1] = ((s32)((s16)a.half[3])) - ((s32)((s16)b.half[3]));
  dst.word[2] = ((s32)((s16)a.half[5])) - ((s32)((s16)b.half[5]));
  dst.word[3] = ((s32)((s16)a.half[7])) - ((s32)((s16)b.half[7]));
  dst.word[4] = ((s32)((s16)a.half[9])) - ((s32)((s16)b.half[9]));
  dst.word[5] = ((s32)((s16)a.half[11])) - ((s32)((s16)b.half[11]));
  dst.word[6] = ((s32)((s16)a.half[13])) - ((s32)((s16)b.half[13]));
  dst.word[7] = ((s32)((s16)a.half[15])) - ((s32)((s16)b.half[15]));
}

Tested on real machine.

Latency and Throughput

CPU	µarch	Latency	Throughput (IPC)
3C5000	LA464	2	2
3A6000	LA664	2	2
3C6000	LA664	2	2

---

All intrinsics in alphabetical order