__lsx_vextrins_w
LSX intrinsic · group: Misc
Synopsis
__m128i __lsx_vextrins_w (__m128i a, __m128i b, imm0_255 imm)
#include <lsxintrin.h>
Instruction: vextrins.w vr, vr, imm
CPU Flags: LSX
Description
Copy the 32-bit element of b selected by imm[1:0] into the element of a selected by imm[5:4]; other elements remain unchanged.
Operation
for (int i = 0; i < 4; i++) {
dst.word[i] = (i == ((imm >> 4) & 3)) ? b.word[imm & 3] : a.word[i];
}
// Expands to:
if (0) {
dst.word[0] = (0 == ((imm >> 4) & 3)) ? (b.word[imm & 3]) : (a.word[0]);
dst.word[1] = (1 == ((imm >> 4) & 3)) ? (b.word[imm & 3]) : (a.word[1]);
dst.word[2] = (2 == ((imm >> 4) & 3)) ? (b.word[imm & 3]) : (a.word[2]);
dst.word[3] = (3 == ((imm >> 4) & 3)) ? (b.word[imm & 3]) : (a.word[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 |
| 2K1000LA | LA264 | 1 | 1 |
| 2K3000 | LA364E | 1 | 1 |