1; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2; RUN: llc < %s -mtriple=x86_64-unknown-linux-gn -mattr=+avx512f | FileCheck %s --check-prefix=ALL --check-prefix=KNL
3; RUN: llc < %s -mtriple=x86_64-unknown-linux-gn -mattr=+avx512f,+avx512vl,+avx512dq,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=SKX
4
5
6define void @any_extend_load_v8i64(<8 x i8> * %ptr) {
7; ALL-LABEL: any_extend_load_v8i64:
8; ALL:       # %bb.0:
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 {{\.?LCPI[0-9]+_[0-9]+}}(%rip){1to8}, %zmm0, %zmm0
11; ALL-NEXT:    vpmovqb %zmm0, (%rdi)
12; ALL-NEXT:    vzeroupper
13; ALL-NEXT:    retq
14  %wide.load = load <8 x i8>, <8 x i8>* %ptr, align 1
15  %1 = zext <8 x i8> %wide.load to <8 x i64>
16  %2 = add nuw nsw <8 x i64> %1, <i64 4, i64 4, i64 4, i64 4, i64 4, i64 4, i64 4, i64 4>
17  %3 = xor <8 x i64> %2, zeroinitializer
18  %4 = trunc <8 x i64> %3 to <8 x i8>
19  store <8 x i8> %4, <8 x i8>* %ptr, align 1
20  ret void
21}
22
23define void @any_extend_load_v8i32(<8 x i8> * %ptr) {
24; KNL-LABEL: any_extend_load_v8i32:
25; KNL:       # %bb.0:
26; KNL-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
27; KNL-NEXT:    vpbroadcastd {{.*#+}} ymm1 = [4,4,4,4,4,4,4,4]
28; KNL-NEXT:    vpaddd %ymm1, %ymm0, %ymm0
29; KNL-NEXT:    vpmovdb %zmm0, %xmm0
30; KNL-NEXT:    vmovq %xmm0, (%rdi)
31; KNL-NEXT:    vzeroupper
32; KNL-NEXT:    retq
33;
34; SKX-LABEL: any_extend_load_v8i32:
35; SKX:       # %bb.0:
36; 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
37; SKX-NEXT:    vpaddd {{\.?LCPI[0-9]+_[0-9]+}}(%rip){1to8}, %ymm0, %ymm0
38; SKX-NEXT:    vpmovdb %ymm0, (%rdi)
39; SKX-NEXT:    vzeroupper
40; SKX-NEXT:    retq
41  %wide.load = load <8 x i8>, <8 x i8>* %ptr, align 1
42  %1 = zext <8 x i8> %wide.load to <8 x i32>
43  %2 = add nuw nsw <8 x i32> %1, <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
44  %3 = xor <8 x i32> %2, zeroinitializer
45  %4 = trunc <8 x i32> %3 to <8 x i8>
46  store <8 x i8> %4, <8 x i8>* %ptr, align 1
47  ret void
48}
49
50define void @any_extend_load_v8i16(<8 x i8> * %ptr) {
51; KNL-LABEL: any_extend_load_v8i16:
52; KNL:       # %bb.0:
53; 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
54; KNL-NEXT:    vpaddw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
55; KNL-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
56; KNL-NEXT:    vmovq %xmm0, (%rdi)
57; KNL-NEXT:    retq
58;
59; SKX-LABEL: any_extend_load_v8i16:
60; SKX:       # %bb.0:
61; 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
62; SKX-NEXT:    vpaddw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
63; SKX-NEXT:    vpmovwb %xmm0, (%rdi)
64; SKX-NEXT:    retq
65  %wide.load = load <8 x i8>, <8 x i8>* %ptr, align 1
66  %1 = zext <8 x i8> %wide.load to <8 x i16>
67  %2 = add nuw nsw <8 x i16> %1, <i16 4, i16 4, i16 4, i16 4, i16 4, i16 4, i16 4, i16 4>
68  %3 = xor <8 x i16> %2, zeroinitializer
69  %4 = trunc <8 x i16> %3 to <8 x i8>
70  store <8 x i8> %4, <8 x i8>* %ptr, align 1
71  ret void
72}
73