__lasx_xvfrstpi_h

LASX intrinsic · group: Misc

Synopsis

__m256i __lasx_xvfrstpi_h (__m256i a, __m256i b, imm0_31 imm)
#include <lasxintrin.h>
Instruction: xvfrstpi.h xr, xr, imm
CPU Flags: LASX

Description

For each 128-bit lane, find the first negative 16-bit element in b. Write its index into the lane of a selected by imm modulo 8, or write the maximum valid index plus one if no such element exists. Other elements of a remain unchanged.

Operation

for (int i = 0; i < 16; i++) {
  dst.half[i] = a.half[i];
}
int i;
for (i = 0; i < 8; i++) {
  if ((s16)b.half[i] < 0) {
    break;
  }
}
dst.half[imm % 8] = i;
for (i = 8; i < 16; i++) {
  if ((s16)b.half[i] < 0) {
    break;
  }
}
dst.half[(imm % 8) + 8] = i - 8;

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