__lasx_xvpickod_d

LASX intrinsic · group: Misc

Synopsis

__m256i __lasx_xvpickod_d (__m256i a, __m256i b)
#include <lasxintrin.h>
Instruction: xvpickod.d xr, xr, xr
CPU Flags: LASX

Description

For each 128-bit lane, place odd-positioned 64-bit elements of b into its lower half and odd-positioned 64-bit elements of a into its upper half.

Operation

for (int i = 0; i < 2; i++) {
  dst.dword[i] = (i < 1) ? b.dword[i * 2 + 1] : a.dword[(i - 1) * 2 + 1];
}
for (int i = 2; i < 4; i++) {
  dst.dword[i] = (i < 3) ? b.dword[(i - 1) * 2 + 1] : a.dword[(i - 2) * 2 + 1];
}

// Expands to:

if (0) {
  dst.dword[0] = b.dword[1];
  dst.dword[1] = a.dword[1];
  dst.dword[2] = b.dword[3];
  dst.dword[3] = a.dword[3];
}

Tested on real machine.

Latency and Throughput

CPU	µarch	Latency	Throughput (IPC)
3C5000	LA464	1	2
3A6000	LA664	1	4
3C6000	LA664	1	4

---

All intrinsics in alphabetical order