__lsx_vssrlrni_d_q
LSX intrinsic · group: Shift
Synopsis
__m128i __lsx_vssrlrni_d_q (__m128i a, __m128i b, imm0_127 imm)
#include <lsxintrin.h>
Instruction: vssrlrni.d.q vr, vr, imm
CPU Flags: LSX
Description
Logical right shift with rounding the unsigned 128-bit elements in a and b by imm, saturate the result to signed 64-bit, write the 64-bit results into dst.
Operation
for (int i = 0; i < 2; i++) {
if (i < 1) {
u128 temp;
if (imm == 0) {
temp = (u128)b.qword[i];
} else {
temp = ((u128)b.qword[i] >> imm) + (((u128)b.qword[i] >> (imm - 1)) & 1);
}
dst.dword[i] = clamp<u128>(temp, 0, 9223372036854775807);
} else {
u128 temp;
if (imm == 0) {
temp = (u128)a.qword[i - 1];
} else {
temp = ((u128)a.qword[i - 1] >> imm) +
(((u128)a.qword[i - 1] >> (imm - 1)) & 1);
}
dst.dword[i] = clamp<u128>(temp, 0, 9223372036854775807);
}
}
Tested on real machine.
Latency and Throughput
| CPU | µarch | Latency | Throughput (IPC) |
|---|---|---|---|
| 3C5000 | LA464 | 3 | 2 |
| 3A6000 | LA664 | 3 | 2 |
| 3C6000 | LA664 | 3 | 2 |
| 2K1000LA | LA264 | 2 | 1 |
| 2K3000 | LA364E | 2 | 1 |