1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gn -mcpu=knl | FileCheck %s --check-prefix=ALL --check-prefix=KNL
3 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gn -mcpu=skx | FileCheck %s --check-prefix=ALL --check-prefix=SKX
6 define void @any_extend_load_v8i64(<8 x i8> * %ptr) {
7 ; ALL-LABEL: any_extend_load_v8i64:
9 ; ALL-NEXT: vpmovzxbq {{.*#+}} zmm0 = mem[0],zero,zero,zero,zero,zero,zero,zero,mem[1],zero,zero,zero,zero,zero,zero,zero,mem[2],zero,zero,zero,zero,zero,zero,zero,mem[3],zero,zero,zero,zero,zero,zero,zero,mem[4],zero,zero,zero,zero,zero,zero,zero,mem[5],zero,zero,zero,zero,zero,zero,zero,mem[6],zero,zero,zero,zero,zero,zero,zero,mem[7],zero,zero,zero,zero,zero,zero,zero
10 ; ALL-NEXT: vpaddq {{.*}}(%rip){1to8}, %zmm0, %zmm0
11 ; ALL-NEXT: vpmovqb %zmm0, (%rdi)
13 %wide.load = load <8 x i8>, <8 x i8>* %ptr, align 1
14 %1 = zext <8 x i8> %wide.load to <8 x i64>
15 %2 = add nuw nsw <8 x i64> %1, <i64 4, i64 4, i64 4, i64 4, i64 4, i64 4, i64 4, i64 4>
16 %3 = xor <8 x i64> %2, zeroinitializer
17 %4 = trunc <8 x i64> %3 to <8 x i8>
18 store <8 x i8> %4, <8 x i8>* %ptr, align 1
22 define void @any_extend_load_v8i32(<8 x i8> * %ptr) {
23 ; KNL-LABEL: any_extend_load_v8i32:
25 ; KNL-NEXT: vpmovzxbw {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
26 ; KNL-NEXT: vpaddw {{.*}}(%rip), %xmm0, %xmm0
27 ; KNL-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
28 ; KNL-NEXT: vmovq %xmm0, (%rdi)
31 ; SKX-LABEL: any_extend_load_v8i32:
33 ; SKX-NEXT: vpmovzxbd {{.*#+}} ymm0 = mem[0],zero,zero,zero,mem[1],zero,zero,zero,mem[2],zero,zero,zero,mem[3],zero,zero,zero,mem[4],zero,zero,zero,mem[5],zero,zero,zero,mem[6],zero,zero,zero,mem[7],zero,zero,zero
34 ; SKX-NEXT: vpaddd {{.*}}(%rip){1to8}, %ymm0, %ymm0
35 ; SKX-NEXT: vpmovdb %ymm0, (%rdi)
37 %wide.load = load <8 x i8>, <8 x i8>* %ptr, align 1
38 %1 = zext <8 x i8> %wide.load to <8 x i32>
39 %2 = add nuw nsw <8 x i32> %1, <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
40 %3 = xor <8 x i32> %2, zeroinitializer
41 %4 = trunc <8 x i32> %3 to <8 x i8>
42 store <8 x i8> %4, <8 x i8>* %ptr, align 1
46 define void @any_extend_load_v8i16(<8 x i8> * %ptr) {
47 ; KNL-LABEL: any_extend_load_v8i16:
49 ; KNL-NEXT: vpmovzxbw {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
50 ; KNL-NEXT: vpaddb {{.*}}(%rip), %xmm0, %xmm0
51 ; KNL-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
52 ; KNL-NEXT: vmovq %xmm0, (%rdi)
55 ; SKX-LABEL: any_extend_load_v8i16:
57 ; SKX-NEXT: vpmovzxbw {{.*#+}} xmm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
58 ; SKX-NEXT: vpaddw {{.*}}(%rip), %xmm0, %xmm0
59 ; SKX-NEXT: vpmovwb %xmm0, (%rdi)
61 %wide.load = load <8 x i8>, <8 x i8>* %ptr, align 1
62 %1 = zext <8 x i8> %wide.load to <8 x i16>
63 %2 = add nuw nsw <8 x i16> %1, <i16 4, i16 4, i16 4, i16 4, i16 4, i16 4, i16 4, i16 4>
64 %3 = xor <8 x i16> %2, zeroinitializer
65 %4 = trunc <8 x i16> %3 to <8 x i8>
66 store <8 x i8> %4, <8 x i8>* %ptr, align 1