xref: /llvm-project/llvm/test/Transforms/SLPVectorizer/X86/arith-ucmp.ll (revision a156b5a47df58a1ac75cf67e26f557b1a4d26dc9) 
1; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
2; RUN: opt < %s -mtriple=x86_64-unknown -passes=slp-vectorizer -S | FileCheck %s --check-prefix=SSE
3; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=slm -passes=slp-vectorizer -S | FileCheck %s --check-prefix=SSE
4; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=corei7-avx -passes=slp-vectorizer -S | FileCheck %s --check-prefix=AVX
5; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=core-avx2 -passes=slp-vectorizer -S | FileCheck %s --check-prefix=AVX
6; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=knl -passes=slp-vectorizer -S | FileCheck %s --check-prefix=AVX512
7; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=skx -mattr=-prefer-256-bit -passes=slp-vectorizer -S | FileCheck %s --check-prefix=AVX512
8; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=skx -mattr=+prefer-256-bit -passes=slp-vectorizer -S | FileCheck %s --check-prefix=AVX
9
10@a64 = common global [8 x i64] zeroinitializer, align 64
11@b64 = common global [8 x i64] zeroinitializer, align 64
12@c64 = common global [8 x i64] zeroinitializer, align 64
13@a32 = common global [16 x i32] zeroinitializer, align 64
14@b32 = common global [16 x i32] zeroinitializer, align 64
15@c32 = common global [16 x i32] zeroinitializer, align 64
16@a16 = common global [32 x i16] zeroinitializer, align 64
17@b16 = common global [32 x i16] zeroinitializer, align 64
18@c16 = common global [32 x i16] zeroinitializer, align 64
19@a8  = common global [64 x i8] zeroinitializer, align 64
20@b8  = common global [64 x i8] zeroinitializer, align 64
21@c8  = common global [64 x i8] zeroinitializer, align 64
22
23define void @ucmp_v8i64() {
24; SSE-LABEL: @ucmp_v8i64(
25; SSE-NEXT:    [[A0:%.*]] = load i64, ptr @a64, align 8
26; SSE-NEXT:    [[A1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 1), align 8
27; SSE-NEXT:    [[A2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
28; SSE-NEXT:    [[A3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 3), align 8
29; SSE-NEXT:    [[A4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
30; SSE-NEXT:    [[A5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 5), align 8
31; SSE-NEXT:    [[A6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
32; SSE-NEXT:    [[A7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 7), align 8
33; SSE-NEXT:    [[B0:%.*]] = load i64, ptr @b64, align 8
34; SSE-NEXT:    [[B1:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 1), align 8
35; SSE-NEXT:    [[B2:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
36; SSE-NEXT:    [[B3:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 3), align 8
37; SSE-NEXT:    [[B4:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
38; SSE-NEXT:    [[B5:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 5), align 8
39; SSE-NEXT:    [[B6:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
40; SSE-NEXT:    [[B7:%.*]] = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 7), align 8
41; SSE-NEXT:    [[R0:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A0]], i64 [[B0]])
42; SSE-NEXT:    [[R1:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A1]], i64 [[B1]])
43; SSE-NEXT:    [[R2:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A2]], i64 [[B2]])
44; SSE-NEXT:    [[R3:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A3]], i64 [[B3]])
45; SSE-NEXT:    [[R4:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A4]], i64 [[B4]])
46; SSE-NEXT:    [[R5:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A5]], i64 [[B5]])
47; SSE-NEXT:    [[R6:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A6]], i64 [[B6]])
48; SSE-NEXT:    [[R7:%.*]] = call i64 @llvm.ucmp.i64.i64(i64 [[A7]], i64 [[B7]])
49; SSE-NEXT:    store i64 [[R0]], ptr @c64, align 8
50; SSE-NEXT:    store i64 [[R1]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 1), align 8
51; SSE-NEXT:    store i64 [[R2]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
52; SSE-NEXT:    store i64 [[R3]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 3), align 8
53; SSE-NEXT:    store i64 [[R4]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
54; SSE-NEXT:    store i64 [[R5]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 5), align 8
55; SSE-NEXT:    store i64 [[R6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
56; SSE-NEXT:    store i64 [[R7]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 7), align 8
57; SSE-NEXT:    ret void
58;
59; AVX-LABEL: @ucmp_v8i64(
60; AVX-NEXT:    [[TMP1:%.*]] = load <4 x i64>, ptr @a64, align 8
61; AVX-NEXT:    [[TMP2:%.*]] = load <4 x i64>, ptr @b64, align 8
62; AVX-NEXT:    [[TMP3:%.*]] = call <4 x i64> @llvm.ucmp.v4i64.v4i64(<4 x i64> [[TMP1]], <4 x i64> [[TMP2]])
63; AVX-NEXT:    store <4 x i64> [[TMP3]], ptr @c64, align 8
64; AVX-NEXT:    [[TMP4:%.*]] = load <4 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
65; AVX-NEXT:    [[TMP5:%.*]] = load <4 x i64>, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
66; AVX-NEXT:    [[TMP6:%.*]] = call <4 x i64> @llvm.ucmp.v4i64.v4i64(<4 x i64> [[TMP4]], <4 x i64> [[TMP5]])
67; AVX-NEXT:    store <4 x i64> [[TMP6]], ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
68; AVX-NEXT:    ret void
69;
70; AVX512-LABEL: @ucmp_v8i64(
71; AVX512-NEXT:    [[TMP1:%.*]] = load <8 x i64>, ptr @a64, align 8
72; AVX512-NEXT:    [[TMP2:%.*]] = load <8 x i64>, ptr @b64, align 8
73; AVX512-NEXT:    [[TMP3:%.*]] = call <8 x i64> @llvm.ucmp.v8i64.v8i64(<8 x i64> [[TMP1]], <8 x i64> [[TMP2]])
74; AVX512-NEXT:    store <8 x i64> [[TMP3]], ptr @c64, align 8
75; AVX512-NEXT:    ret void
76;
77  %a0 = load i64, ptr @a64, align 8
78  %a1 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 1), align 8
79  %a2 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 2), align 8
80  %a3 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 3), align 8
81  %a4 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 4), align 8
82  %a5 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 5), align 8
83  %a6 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 6), align 8
84  %a7 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @a64, i32 0, i64 7), align 8
85  %b0 = load i64, ptr @b64, align 8
86  %b1 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 1), align 8
87  %b2 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 2), align 8
88  %b3 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 3), align 8
89  %b4 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 4), align 8
90  %b5 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 5), align 8
91  %b6 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 6), align 8
92  %b7 = load i64, ptr getelementptr inbounds ([8 x i64], ptr @b64, i32 0, i64 7), align 8
93  %r0 = call i64 @llvm.ucmp.i64.i64(i64 %a0, i64 %b0)
94  %r1 = call i64 @llvm.ucmp.i64.i64(i64 %a1, i64 %b1)
95  %r2 = call i64 @llvm.ucmp.i64.i64(i64 %a2, i64 %b2)
96  %r3 = call i64 @llvm.ucmp.i64.i64(i64 %a3, i64 %b3)
97  %r4 = call i64 @llvm.ucmp.i64.i64(i64 %a4, i64 %b4)
98  %r5 = call i64 @llvm.ucmp.i64.i64(i64 %a5, i64 %b5)
99  %r6 = call i64 @llvm.ucmp.i64.i64(i64 %a6, i64 %b6)
100  %r7 = call i64 @llvm.ucmp.i64.i64(i64 %a7, i64 %b7)
101  store i64 %r0, ptr @c64, align 8
102  store i64 %r1, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 1), align 8
103  store i64 %r2, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 2), align 8
104  store i64 %r3, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 3), align 8
105  store i64 %r4, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 4), align 8
106  store i64 %r5, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 5), align 8
107  store i64 %r6, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 6), align 8
108  store i64 %r7, ptr getelementptr inbounds ([8 x i64], ptr @c64, i32 0, i64 7), align 8
109  ret void
110}
111
112define void @ucmp_v16i32() {
113; SSE-LABEL: @ucmp_v16i32(
114; SSE-NEXT:    [[TMP1:%.*]] = load <4 x i32>, ptr @a32, align 4
115; SSE-NEXT:    [[TMP2:%.*]] = load <4 x i32>, ptr @b32, align 4
116; SSE-NEXT:    [[TMP3:%.*]] = call <4 x i32> @llvm.ucmp.v4i32.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP2]])
117; SSE-NEXT:    store <4 x i32> [[TMP3]], ptr @c32, align 4
118; SSE-NEXT:    [[TMP4:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4), align 4
119; SSE-NEXT:    [[TMP5:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4), align 4
120; SSE-NEXT:    [[TMP6:%.*]] = call <4 x i32> @llvm.ucmp.v4i32.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP5]])
121; SSE-NEXT:    store <4 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4), align 4
122; SSE-NEXT:    [[TMP7:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
123; SSE-NEXT:    [[TMP8:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
124; SSE-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.ucmp.v4i32.v4i32(<4 x i32> [[TMP7]], <4 x i32> [[TMP8]])
125; SSE-NEXT:    store <4 x i32> [[TMP9]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
126; SSE-NEXT:    [[TMP10:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
127; SSE-NEXT:    [[TMP11:%.*]] = load <4 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
128; SSE-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.ucmp.v4i32.v4i32(<4 x i32> [[TMP10]], <4 x i32> [[TMP11]])
129; SSE-NEXT:    store <4 x i32> [[TMP12]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
130; SSE-NEXT:    ret void
131;
132; AVX-LABEL: @ucmp_v16i32(
133; AVX-NEXT:    [[TMP1:%.*]] = load <8 x i32>, ptr @a32, align 4
134; AVX-NEXT:    [[TMP2:%.*]] = load <8 x i32>, ptr @b32, align 4
135; AVX-NEXT:    [[TMP3:%.*]] = call <8 x i32> @llvm.ucmp.v8i32.v8i32(<8 x i32> [[TMP1]], <8 x i32> [[TMP2]])
136; AVX-NEXT:    store <8 x i32> [[TMP3]], ptr @c32, align 4
137; AVX-NEXT:    [[TMP4:%.*]] = load <8 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8), align 4
138; AVX-NEXT:    [[TMP5:%.*]] = load <8 x i32>, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8), align 4
139; AVX-NEXT:    [[TMP6:%.*]] = call <8 x i32> @llvm.ucmp.v8i32.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
140; AVX-NEXT:    store <8 x i32> [[TMP6]], ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8), align 4
141; AVX-NEXT:    ret void
142;
143; AVX512-LABEL: @ucmp_v16i32(
144; AVX512-NEXT:    [[TMP1:%.*]] = load <16 x i32>, ptr @a32, align 4
145; AVX512-NEXT:    [[TMP2:%.*]] = load <16 x i32>, ptr @b32, align 4
146; AVX512-NEXT:    [[TMP3:%.*]] = call <16 x i32> @llvm.ucmp.v16i32.v16i32(<16 x i32> [[TMP1]], <16 x i32> [[TMP2]])
147; AVX512-NEXT:    store <16 x i32> [[TMP3]], ptr @c32, align 4
148; AVX512-NEXT:    ret void
149;
150  %a0  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 0 ), align 4
151  %a1  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 1 ), align 4
152  %a2  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 2 ), align 4
153  %a3  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 3 ), align 4
154  %a4  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 4 ), align 4
155  %a5  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 5 ), align 4
156  %a6  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 6 ), align 4
157  %a7  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 7 ), align 4
158  %a8  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 8 ), align 4
159  %a9  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 9 ), align 4
160  %a10 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 10), align 4
161  %a11 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 11), align 4
162  %a12 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 12), align 4
163  %a13 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 13), align 4
164  %a14 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 14), align 4
165  %a15 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @a32, i32 0, i64 15), align 4
166  %b0  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 0 ), align 4
167  %b1  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 1 ), align 4
168  %b2  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 2 ), align 4
169  %b3  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 3 ), align 4
170  %b4  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 4 ), align 4
171  %b5  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 5 ), align 4
172  %b6  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 6 ), align 4
173  %b7  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 7 ), align 4
174  %b8  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 8 ), align 4
175  %b9  = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 9 ), align 4
176  %b10 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 10), align 4
177  %b11 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 11), align 4
178  %b12 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 12), align 4
179  %b13 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 13), align 4
180  %b14 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 14), align 4
181  %b15 = load i32, ptr getelementptr inbounds ([16 x i32], ptr @b32, i32 0, i64 15), align 4
182  %r0  = call i32 @llvm.ucmp.i32.i32(i32 %a0 , i32 %b0 )
183  %r1  = call i32 @llvm.ucmp.i32.i32(i32 %a1 , i32 %b1 )
184  %r2  = call i32 @llvm.ucmp.i32.i32(i32 %a2 , i32 %b2 )
185  %r3  = call i32 @llvm.ucmp.i32.i32(i32 %a3 , i32 %b3 )
186  %r4  = call i32 @llvm.ucmp.i32.i32(i32 %a4 , i32 %b4 )
187  %r5  = call i32 @llvm.ucmp.i32.i32(i32 %a5 , i32 %b5 )
188  %r6  = call i32 @llvm.ucmp.i32.i32(i32 %a6 , i32 %b6 )
189  %r7  = call i32 @llvm.ucmp.i32.i32(i32 %a7 , i32 %b7 )
190  %r8  = call i32 @llvm.ucmp.i32.i32(i32 %a8 , i32 %b8 )
191  %r9  = call i32 @llvm.ucmp.i32.i32(i32 %a9 , i32 %b9 )
192  %r10 = call i32 @llvm.ucmp.i32.i32(i32 %a10, i32 %b10)
193  %r11 = call i32 @llvm.ucmp.i32.i32(i32 %a11, i32 %b11)
194  %r12 = call i32 @llvm.ucmp.i32.i32(i32 %a12, i32 %b12)
195  %r13 = call i32 @llvm.ucmp.i32.i32(i32 %a13, i32 %b13)
196  %r14 = call i32 @llvm.ucmp.i32.i32(i32 %a14, i32 %b14)
197  %r15 = call i32 @llvm.ucmp.i32.i32(i32 %a15, i32 %b15)
198  store i32 %r0 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 0 ), align 4
199  store i32 %r1 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 1 ), align 4
200  store i32 %r2 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 2 ), align 4
201  store i32 %r3 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 3 ), align 4
202  store i32 %r4 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 4 ), align 4
203  store i32 %r5 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 5 ), align 4
204  store i32 %r6 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 6 ), align 4
205  store i32 %r7 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 7 ), align 4
206  store i32 %r8 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 8 ), align 4
207  store i32 %r9 , ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 9 ), align 4
208  store i32 %r10, ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 10), align 4
209  store i32 %r11, ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 11), align 4
210  store i32 %r12, ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 12), align 4
211  store i32 %r13, ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 13), align 4
212  store i32 %r14, ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 14), align 4
213  store i32 %r15, ptr getelementptr inbounds ([16 x i32], ptr @c32, i32 0, i64 15), align 4
214  ret void
215}
216
217define void @ucmp_v32i16() {
218; SSE-LABEL: @ucmp_v32i16(
219; SSE-NEXT:    [[TMP1:%.*]] = load <8 x i16>, ptr @a16, align 2
220; SSE-NEXT:    [[TMP2:%.*]] = load <8 x i16>, ptr @b16, align 2
221; SSE-NEXT:    [[TMP3:%.*]] = call <8 x i16> @llvm.ucmp.v8i16.v8i16(<8 x i16> [[TMP1]], <8 x i16> [[TMP2]])
222; SSE-NEXT:    store <8 x i16> [[TMP3]], ptr @c16, align 2
223; SSE-NEXT:    [[TMP4:%.*]] = load <8 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 8), align 2
224; SSE-NEXT:    [[TMP5:%.*]] = load <8 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 8), align 2
225; SSE-NEXT:    [[TMP6:%.*]] = call <8 x i16> @llvm.ucmp.v8i16.v8i16(<8 x i16> [[TMP4]], <8 x i16> [[TMP5]])
226; SSE-NEXT:    store <8 x i16> [[TMP6]], ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 8), align 2
227; SSE-NEXT:    [[TMP7:%.*]] = load <8 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 16), align 2
228; SSE-NEXT:    [[TMP8:%.*]] = load <8 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 16), align 2
229; SSE-NEXT:    [[TMP9:%.*]] = call <8 x i16> @llvm.ucmp.v8i16.v8i16(<8 x i16> [[TMP7]], <8 x i16> [[TMP8]])
230; SSE-NEXT:    store <8 x i16> [[TMP9]], ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 16), align 2
231; SSE-NEXT:    [[TMP10:%.*]] = load <8 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 24), align 2
232; SSE-NEXT:    [[TMP11:%.*]] = load <8 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 24), align 2
233; SSE-NEXT:    [[TMP12:%.*]] = call <8 x i16> @llvm.ucmp.v8i16.v8i16(<8 x i16> [[TMP10]], <8 x i16> [[TMP11]])
234; SSE-NEXT:    store <8 x i16> [[TMP12]], ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 24), align 2
235; SSE-NEXT:    ret void
236;
237; AVX-LABEL: @ucmp_v32i16(
238; AVX-NEXT:    [[TMP1:%.*]] = load <16 x i16>, ptr @a16, align 2
239; AVX-NEXT:    [[TMP2:%.*]] = load <16 x i16>, ptr @b16, align 2
240; AVX-NEXT:    [[TMP3:%.*]] = call <16 x i16> @llvm.ucmp.v16i16.v16i16(<16 x i16> [[TMP1]], <16 x i16> [[TMP2]])
241; AVX-NEXT:    store <16 x i16> [[TMP3]], ptr @c16, align 2
242; AVX-NEXT:    [[TMP4:%.*]] = load <16 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 16), align 2
243; AVX-NEXT:    [[TMP5:%.*]] = load <16 x i16>, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 16), align 2
244; AVX-NEXT:    [[TMP6:%.*]] = call <16 x i16> @llvm.ucmp.v16i16.v16i16(<16 x i16> [[TMP4]], <16 x i16> [[TMP5]])
245; AVX-NEXT:    store <16 x i16> [[TMP6]], ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 16), align 2
246; AVX-NEXT:    ret void
247;
248; AVX512-LABEL: @ucmp_v32i16(
249; AVX512-NEXT:    [[TMP1:%.*]] = load <32 x i16>, ptr @a16, align 2
250; AVX512-NEXT:    [[TMP2:%.*]] = load <32 x i16>, ptr @b16, align 2
251; AVX512-NEXT:    [[TMP3:%.*]] = call <32 x i16> @llvm.ucmp.v32i16.v32i16(<32 x i16> [[TMP1]], <32 x i16> [[TMP2]])
252; AVX512-NEXT:    store <32 x i16> [[TMP3]], ptr @c16, align 2
253; AVX512-NEXT:    ret void
254;
255  %a0  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 0 ), align 2
256  %a1  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 1 ), align 2
257  %a2  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 2 ), align 2
258  %a3  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 3 ), align 2
259  %a4  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 4 ), align 2
260  %a5  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 5 ), align 2
261  %a6  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 6 ), align 2
262  %a7  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 7 ), align 2
263  %a8  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 8 ), align 2
264  %a9  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 9 ), align 2
265  %a10 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 10), align 2
266  %a11 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 11), align 2
267  %a12 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 12), align 2
268  %a13 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 13), align 2
269  %a14 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 14), align 2
270  %a15 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 15), align 2
271  %a16 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 16), align 2
272  %a17 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 17), align 2
273  %a18 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 18), align 2
274  %a19 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 19), align 2
275  %a20 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 20), align 2
276  %a21 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 21), align 2
277  %a22 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 22), align 2
278  %a23 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 23), align 2
279  %a24 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 24), align 2
280  %a25 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 25), align 2
281  %a26 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 26), align 2
282  %a27 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 27), align 2
283  %a28 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 28), align 2
284  %a29 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 29), align 2
285  %a30 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 30), align 2
286  %a31 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @a16, i32 0, i64 31), align 2
287  %b0  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 0 ), align 2
288  %b1  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 1 ), align 2
289  %b2  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 2 ), align 2
290  %b3  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 3 ), align 2
291  %b4  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 4 ), align 2
292  %b5  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 5 ), align 2
293  %b6  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 6 ), align 2
294  %b7  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 7 ), align 2
295  %b8  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 8 ), align 2
296  %b9  = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 9 ), align 2
297  %b10 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 10), align 2
298  %b11 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 11), align 2
299  %b12 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 12), align 2
300  %b13 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 13), align 2
301  %b14 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 14), align 2
302  %b15 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 15), align 2
303  %b16 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 16), align 2
304  %b17 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 17), align 2
305  %b18 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 18), align 2
306  %b19 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 19), align 2
307  %b20 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 20), align 2
308  %b21 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 21), align 2
309  %b22 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 22), align 2
310  %b23 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 23), align 2
311  %b24 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 24), align 2
312  %b25 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 25), align 2
313  %b26 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 26), align 2
314  %b27 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 27), align 2
315  %b28 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 28), align 2
316  %b29 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 29), align 2
317  %b30 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 30), align 2
318  %b31 = load i16, ptr getelementptr inbounds ([32 x i16], ptr @b16, i32 0, i64 31), align 2
319  %r0  = call i16 @llvm.ucmp.i16.i16(i16 %a0 , i16 %b0 )
320  %r1  = call i16 @llvm.ucmp.i16.i16(i16 %a1 , i16 %b1 )
321  %r2  = call i16 @llvm.ucmp.i16.i16(i16 %a2 , i16 %b2 )
322  %r3  = call i16 @llvm.ucmp.i16.i16(i16 %a3 , i16 %b3 )
323  %r4  = call i16 @llvm.ucmp.i16.i16(i16 %a4 , i16 %b4 )
324  %r5  = call i16 @llvm.ucmp.i16.i16(i16 %a5 , i16 %b5 )
325  %r6  = call i16 @llvm.ucmp.i16.i16(i16 %a6 , i16 %b6 )
326  %r7  = call i16 @llvm.ucmp.i16.i16(i16 %a7 , i16 %b7 )
327  %r8  = call i16 @llvm.ucmp.i16.i16(i16 %a8 , i16 %b8 )
328  %r9  = call i16 @llvm.ucmp.i16.i16(i16 %a9 , i16 %b9 )
329  %r10 = call i16 @llvm.ucmp.i16.i16(i16 %a10, i16 %b10)
330  %r11 = call i16 @llvm.ucmp.i16.i16(i16 %a11, i16 %b11)
331  %r12 = call i16 @llvm.ucmp.i16.i16(i16 %a12, i16 %b12)
332  %r13 = call i16 @llvm.ucmp.i16.i16(i16 %a13, i16 %b13)
333  %r14 = call i16 @llvm.ucmp.i16.i16(i16 %a14, i16 %b14)
334  %r15 = call i16 @llvm.ucmp.i16.i16(i16 %a15, i16 %b15)
335  %r16 = call i16 @llvm.ucmp.i16.i16(i16 %a16, i16 %b16)
336  %r17 = call i16 @llvm.ucmp.i16.i16(i16 %a17, i16 %b17)
337  %r18 = call i16 @llvm.ucmp.i16.i16(i16 %a18, i16 %b18)
338  %r19 = call i16 @llvm.ucmp.i16.i16(i16 %a19, i16 %b19)
339  %r20 = call i16 @llvm.ucmp.i16.i16(i16 %a20, i16 %b20)
340  %r21 = call i16 @llvm.ucmp.i16.i16(i16 %a21, i16 %b21)
341  %r22 = call i16 @llvm.ucmp.i16.i16(i16 %a22, i16 %b22)
342  %r23 = call i16 @llvm.ucmp.i16.i16(i16 %a23, i16 %b23)
343  %r24 = call i16 @llvm.ucmp.i16.i16(i16 %a24, i16 %b24)
344  %r25 = call i16 @llvm.ucmp.i16.i16(i16 %a25, i16 %b25)
345  %r26 = call i16 @llvm.ucmp.i16.i16(i16 %a26, i16 %b26)
346  %r27 = call i16 @llvm.ucmp.i16.i16(i16 %a27, i16 %b27)
347  %r28 = call i16 @llvm.ucmp.i16.i16(i16 %a28, i16 %b28)
348  %r29 = call i16 @llvm.ucmp.i16.i16(i16 %a29, i16 %b29)
349  %r30 = call i16 @llvm.ucmp.i16.i16(i16 %a30, i16 %b30)
350  %r31 = call i16 @llvm.ucmp.i16.i16(i16 %a31, i16 %b31)
351  store i16 %r0 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 0 ), align 2
352  store i16 %r1 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 1 ), align 2
353  store i16 %r2 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 2 ), align 2
354  store i16 %r3 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 3 ), align 2
355  store i16 %r4 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 4 ), align 2
356  store i16 %r5 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 5 ), align 2
357  store i16 %r6 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 6 ), align 2
358  store i16 %r7 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 7 ), align 2
359  store i16 %r8 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 8 ), align 2
360  store i16 %r9 , ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 9 ), align 2
361  store i16 %r10, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 10), align 2
362  store i16 %r11, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 11), align 2
363  store i16 %r12, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 12), align 2
364  store i16 %r13, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 13), align 2
365  store i16 %r14, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 14), align 2
366  store i16 %r15, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 15), align 2
367  store i16 %r16, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 16), align 2
368  store i16 %r17, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 17), align 2
369  store i16 %r18, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 18), align 2
370  store i16 %r19, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 19), align 2
371  store i16 %r20, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 20), align 2
372  store i16 %r21, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 21), align 2
373  store i16 %r22, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 22), align 2
374  store i16 %r23, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 23), align 2
375  store i16 %r24, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 24), align 2
376  store i16 %r25, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 25), align 2
377  store i16 %r26, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 26), align 2
378  store i16 %r27, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 27), align 2
379  store i16 %r28, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 28), align 2
380  store i16 %r29, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 29), align 2
381  store i16 %r30, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 30), align 2
382  store i16 %r31, ptr getelementptr inbounds ([32 x i16], ptr @c16, i32 0, i64 31), align 2
383  ret void
384}
385
386define void @ucmp_v64i8() {
387; SSE-LABEL: @ucmp_v64i8(
388; SSE-NEXT:    [[TMP1:%.*]] = load <16 x i8>, ptr @a8, align 1
389; SSE-NEXT:    [[TMP2:%.*]] = load <16 x i8>, ptr @b8, align 1
390; SSE-NEXT:    [[TMP3:%.*]] = call <16 x i8> @llvm.ucmp.v16i8.v16i8(<16 x i8> [[TMP1]], <16 x i8> [[TMP2]])
391; SSE-NEXT:    store <16 x i8> [[TMP3]], ptr @c8, align 1
392; SSE-NEXT:    [[TMP4:%.*]] = load <16 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 16), align 1
393; SSE-NEXT:    [[TMP5:%.*]] = load <16 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 16), align 1
394; SSE-NEXT:    [[TMP6:%.*]] = call <16 x i8> @llvm.ucmp.v16i8.v16i8(<16 x i8> [[TMP4]], <16 x i8> [[TMP5]])
395; SSE-NEXT:    store <16 x i8> [[TMP6]], ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 16), align 1
396; SSE-NEXT:    [[TMP7:%.*]] = load <16 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 32), align 1
397; SSE-NEXT:    [[TMP8:%.*]] = load <16 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 32), align 1
398; SSE-NEXT:    [[TMP9:%.*]] = call <16 x i8> @llvm.ucmp.v16i8.v16i8(<16 x i8> [[TMP7]], <16 x i8> [[TMP8]])
399; SSE-NEXT:    store <16 x i8> [[TMP9]], ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 32), align 1
400; SSE-NEXT:    [[TMP10:%.*]] = load <16 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 48), align 1
401; SSE-NEXT:    [[TMP11:%.*]] = load <16 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 48), align 1
402; SSE-NEXT:    [[TMP12:%.*]] = call <16 x i8> @llvm.ucmp.v16i8.v16i8(<16 x i8> [[TMP10]], <16 x i8> [[TMP11]])
403; SSE-NEXT:    store <16 x i8> [[TMP12]], ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 48), align 1
404; SSE-NEXT:    ret void
405;
406; AVX-LABEL: @ucmp_v64i8(
407; AVX-NEXT:    [[TMP1:%.*]] = load <32 x i8>, ptr @a8, align 1
408; AVX-NEXT:    [[TMP2:%.*]] = load <32 x i8>, ptr @b8, align 1
409; AVX-NEXT:    [[TMP3:%.*]] = call <32 x i8> @llvm.ucmp.v32i8.v32i8(<32 x i8> [[TMP1]], <32 x i8> [[TMP2]])
410; AVX-NEXT:    store <32 x i8> [[TMP3]], ptr @c8, align 1
411; AVX-NEXT:    [[TMP4:%.*]] = load <32 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 32), align 1
412; AVX-NEXT:    [[TMP5:%.*]] = load <32 x i8>, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 32), align 1
413; AVX-NEXT:    [[TMP6:%.*]] = call <32 x i8> @llvm.ucmp.v32i8.v32i8(<32 x i8> [[TMP4]], <32 x i8> [[TMP5]])
414; AVX-NEXT:    store <32 x i8> [[TMP6]], ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 32), align 1
415; AVX-NEXT:    ret void
416;
417; AVX512-LABEL: @ucmp_v64i8(
418; AVX512-NEXT:    [[TMP1:%.*]] = load <64 x i8>, ptr @a8, align 1
419; AVX512-NEXT:    [[TMP2:%.*]] = load <64 x i8>, ptr @b8, align 1
420; AVX512-NEXT:    [[TMP3:%.*]] = call <64 x i8> @llvm.ucmp.v64i8.v64i8(<64 x i8> [[TMP1]], <64 x i8> [[TMP2]])
421; AVX512-NEXT:    store <64 x i8> [[TMP3]], ptr @c8, align 1
422; AVX512-NEXT:    ret void
423;
424  %a0  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 0 ), align 1
425  %a1  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 1 ), align 1
426  %a2  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 2 ), align 1
427  %a3  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 3 ), align 1
428  %a4  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 4 ), align 1
429  %a5  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 5 ), align 1
430  %a6  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 6 ), align 1
431  %a7  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 7 ), align 1
432  %a8  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 8 ), align 1
433  %a9  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 9 ), align 1
434  %a10 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 10), align 1
435  %a11 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 11), align 1
436  %a12 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 12), align 1
437  %a13 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 13), align 1
438  %a14 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 14), align 1
439  %a15 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 15), align 1
440  %a16 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 16), align 1
441  %a17 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 17), align 1
442  %a18 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 18), align 1
443  %a19 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 19), align 1
444  %a20 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 20), align 1
445  %a21 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 21), align 1
446  %a22 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 22), align 1
447  %a23 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 23), align 1
448  %a24 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 24), align 1
449  %a25 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 25), align 1
450  %a26 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 26), align 1
451  %a27 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 27), align 1
452  %a28 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 28), align 1
453  %a29 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 29), align 1
454  %a30 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 30), align 1
455  %a31 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 31), align 1
456  %a32 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 32), align 1
457  %a33 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 33), align 1
458  %a34 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 34), align 1
459  %a35 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 35), align 1
460  %a36 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 36), align 1
461  %a37 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 37), align 1
462  %a38 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 38), align 1
463  %a39 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 39), align 1
464  %a40 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 40), align 1
465  %a41 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 41), align 1
466  %a42 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 42), align 1
467  %a43 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 43), align 1
468  %a44 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 44), align 1
469  %a45 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 45), align 1
470  %a46 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 46), align 1
471  %a47 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 47), align 1
472  %a48 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 48), align 1
473  %a49 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 49), align 1
474  %a50 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 50), align 1
475  %a51 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 51), align 1
476  %a52 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 52), align 1
477  %a53 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 53), align 1
478  %a54 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 54), align 1
479  %a55 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 55), align 1
480  %a56 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 56), align 1
481  %a57 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 57), align 1
482  %a58 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 58), align 1
483  %a59 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 59), align 1
484  %a60 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 60), align 1
485  %a61 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 61), align 1
486  %a62 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 62), align 1
487  %a63 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @a8, i32 0, i64 63), align 1
488  %b0  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 0 ), align 1
489  %b1  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 1 ), align 1
490  %b2  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 2 ), align 1
491  %b3  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 3 ), align 1
492  %b4  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 4 ), align 1
493  %b5  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 5 ), align 1
494  %b6  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 6 ), align 1
495  %b7  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 7 ), align 1
496  %b8  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 8 ), align 1
497  %b9  = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 9 ), align 1
498  %b10 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 10), align 1
499  %b11 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 11), align 1
500  %b12 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 12), align 1
501  %b13 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 13), align 1
502  %b14 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 14), align 1
503  %b15 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 15), align 1
504  %b16 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 16), align 1
505  %b17 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 17), align 1
506  %b18 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 18), align 1
507  %b19 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 19), align 1
508  %b20 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 20), align 1
509  %b21 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 21), align 1
510  %b22 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 22), align 1
511  %b23 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 23), align 1
512  %b24 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 24), align 1
513  %b25 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 25), align 1
514  %b26 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 26), align 1
515  %b27 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 27), align 1
516  %b28 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 28), align 1
517  %b29 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 29), align 1
518  %b30 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 30), align 1
519  %b31 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 31), align 1
520  %b32 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 32), align 1
521  %b33 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 33), align 1
522  %b34 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 34), align 1
523  %b35 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 35), align 1
524  %b36 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 36), align 1
525  %b37 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 37), align 1
526  %b38 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 38), align 1
527  %b39 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 39), align 1
528  %b40 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 40), align 1
529  %b41 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 41), align 1
530  %b42 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 42), align 1
531  %b43 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 43), align 1
532  %b44 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 44), align 1
533  %b45 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 45), align 1
534  %b46 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 46), align 1
535  %b47 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 47), align 1
536  %b48 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 48), align 1
537  %b49 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 49), align 1
538  %b50 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 50), align 1
539  %b51 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 51), align 1
540  %b52 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 52), align 1
541  %b53 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 53), align 1
542  %b54 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 54), align 1
543  %b55 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 55), align 1
544  %b56 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 56), align 1
545  %b57 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 57), align 1
546  %b58 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 58), align 1
547  %b59 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 59), align 1
548  %b60 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 60), align 1
549  %b61 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 61), align 1
550  %b62 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 62), align 1
551  %b63 = load i8, ptr getelementptr inbounds ([64 x i8], ptr @b8, i32 0, i64 63), align 1
552  %r0  = call i8 @llvm.ucmp.i8.i8(i8 %a0 , i8 %b0 )
553  %r1  = call i8 @llvm.ucmp.i8.i8(i8 %a1 , i8 %b1 )
554  %r2  = call i8 @llvm.ucmp.i8.i8(i8 %a2 , i8 %b2 )
555  %r3  = call i8 @llvm.ucmp.i8.i8(i8 %a3 , i8 %b3 )
556  %r4  = call i8 @llvm.ucmp.i8.i8(i8 %a4 , i8 %b4 )
557  %r5  = call i8 @llvm.ucmp.i8.i8(i8 %a5 , i8 %b5 )
558  %r6  = call i8 @llvm.ucmp.i8.i8(i8 %a6 , i8 %b6 )
559  %r7  = call i8 @llvm.ucmp.i8.i8(i8 %a7 , i8 %b7 )
560  %r8  = call i8 @llvm.ucmp.i8.i8(i8 %a8 , i8 %b8 )
561  %r9  = call i8 @llvm.ucmp.i8.i8(i8 %a9 , i8 %b9 )
562  %r10 = call i8 @llvm.ucmp.i8.i8(i8 %a10, i8 %b10)
563  %r11 = call i8 @llvm.ucmp.i8.i8(i8 %a11, i8 %b11)
564  %r12 = call i8 @llvm.ucmp.i8.i8(i8 %a12, i8 %b12)
565  %r13 = call i8 @llvm.ucmp.i8.i8(i8 %a13, i8 %b13)
566  %r14 = call i8 @llvm.ucmp.i8.i8(i8 %a14, i8 %b14)
567  %r15 = call i8 @llvm.ucmp.i8.i8(i8 %a15, i8 %b15)
568  %r16 = call i8 @llvm.ucmp.i8.i8(i8 %a16, i8 %b16)
569  %r17 = call i8 @llvm.ucmp.i8.i8(i8 %a17, i8 %b17)
570  %r18 = call i8 @llvm.ucmp.i8.i8(i8 %a18, i8 %b18)
571  %r19 = call i8 @llvm.ucmp.i8.i8(i8 %a19, i8 %b19)
572  %r20 = call i8 @llvm.ucmp.i8.i8(i8 %a20, i8 %b20)
573  %r21 = call i8 @llvm.ucmp.i8.i8(i8 %a21, i8 %b21)
574  %r22 = call i8 @llvm.ucmp.i8.i8(i8 %a22, i8 %b22)
575  %r23 = call i8 @llvm.ucmp.i8.i8(i8 %a23, i8 %b23)
576  %r24 = call i8 @llvm.ucmp.i8.i8(i8 %a24, i8 %b24)
577  %r25 = call i8 @llvm.ucmp.i8.i8(i8 %a25, i8 %b25)
578  %r26 = call i8 @llvm.ucmp.i8.i8(i8 %a26, i8 %b26)
579  %r27 = call i8 @llvm.ucmp.i8.i8(i8 %a27, i8 %b27)
580  %r28 = call i8 @llvm.ucmp.i8.i8(i8 %a28, i8 %b28)
581  %r29 = call i8 @llvm.ucmp.i8.i8(i8 %a29, i8 %b29)
582  %r30 = call i8 @llvm.ucmp.i8.i8(i8 %a30, i8 %b30)
583  %r31 = call i8 @llvm.ucmp.i8.i8(i8 %a31, i8 %b31)
584  %r32 = call i8 @llvm.ucmp.i8.i8(i8 %a32, i8 %b32)
585  %r33 = call i8 @llvm.ucmp.i8.i8(i8 %a33, i8 %b33)
586  %r34 = call i8 @llvm.ucmp.i8.i8(i8 %a34, i8 %b34)
587  %r35 = call i8 @llvm.ucmp.i8.i8(i8 %a35, i8 %b35)
588  %r36 = call i8 @llvm.ucmp.i8.i8(i8 %a36, i8 %b36)
589  %r37 = call i8 @llvm.ucmp.i8.i8(i8 %a37, i8 %b37)
590  %r38 = call i8 @llvm.ucmp.i8.i8(i8 %a38, i8 %b38)
591  %r39 = call i8 @llvm.ucmp.i8.i8(i8 %a39, i8 %b39)
592  %r40 = call i8 @llvm.ucmp.i8.i8(i8 %a40, i8 %b40)
593  %r41 = call i8 @llvm.ucmp.i8.i8(i8 %a41, i8 %b41)
594  %r42 = call i8 @llvm.ucmp.i8.i8(i8 %a42, i8 %b42)
595  %r43 = call i8 @llvm.ucmp.i8.i8(i8 %a43, i8 %b43)
596  %r44 = call i8 @llvm.ucmp.i8.i8(i8 %a44, i8 %b44)
597  %r45 = call i8 @llvm.ucmp.i8.i8(i8 %a45, i8 %b45)
598  %r46 = call i8 @llvm.ucmp.i8.i8(i8 %a46, i8 %b46)
599  %r47 = call i8 @llvm.ucmp.i8.i8(i8 %a47, i8 %b47)
600  %r48 = call i8 @llvm.ucmp.i8.i8(i8 %a48, i8 %b48)
601  %r49 = call i8 @llvm.ucmp.i8.i8(i8 %a49, i8 %b49)
602  %r50 = call i8 @llvm.ucmp.i8.i8(i8 %a50, i8 %b50)
603  %r51 = call i8 @llvm.ucmp.i8.i8(i8 %a51, i8 %b51)
604  %r52 = call i8 @llvm.ucmp.i8.i8(i8 %a52, i8 %b52)
605  %r53 = call i8 @llvm.ucmp.i8.i8(i8 %a53, i8 %b53)
606  %r54 = call i8 @llvm.ucmp.i8.i8(i8 %a54, i8 %b54)
607  %r55 = call i8 @llvm.ucmp.i8.i8(i8 %a55, i8 %b55)
608  %r56 = call i8 @llvm.ucmp.i8.i8(i8 %a56, i8 %b56)
609  %r57 = call i8 @llvm.ucmp.i8.i8(i8 %a57, i8 %b57)
610  %r58 = call i8 @llvm.ucmp.i8.i8(i8 %a58, i8 %b58)
611  %r59 = call i8 @llvm.ucmp.i8.i8(i8 %a59, i8 %b59)
612  %r60 = call i8 @llvm.ucmp.i8.i8(i8 %a60, i8 %b60)
613  %r61 = call i8 @llvm.ucmp.i8.i8(i8 %a61, i8 %b61)
614  %r62 = call i8 @llvm.ucmp.i8.i8(i8 %a62, i8 %b62)
615  %r63 = call i8 @llvm.ucmp.i8.i8(i8 %a63, i8 %b63)
616  store i8 %r0 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 0 ), align 1
617  store i8 %r1 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 1 ), align 1
618  store i8 %r2 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 2 ), align 1
619  store i8 %r3 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 3 ), align 1
620  store i8 %r4 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 4 ), align 1
621  store i8 %r5 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 5 ), align 1
622  store i8 %r6 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 6 ), align 1
623  store i8 %r7 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 7 ), align 1
624  store i8 %r8 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 8 ), align 1
625  store i8 %r9 , ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 9 ), align 1
626  store i8 %r10, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 10), align 1
627  store i8 %r11, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 11), align 1
628  store i8 %r12, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 12), align 1
629  store i8 %r13, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 13), align 1
630  store i8 %r14, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 14), align 1
631  store i8 %r15, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 15), align 1
632  store i8 %r16, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 16), align 1
633  store i8 %r17, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 17), align 1
634  store i8 %r18, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 18), align 1
635  store i8 %r19, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 19), align 1
636  store i8 %r20, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 20), align 1
637  store i8 %r21, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 21), align 1
638  store i8 %r22, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 22), align 1
639  store i8 %r23, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 23), align 1
640  store i8 %r24, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 24), align 1
641  store i8 %r25, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 25), align 1
642  store i8 %r26, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 26), align 1
643  store i8 %r27, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 27), align 1
644  store i8 %r28, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 28), align 1
645  store i8 %r29, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 29), align 1
646  store i8 %r30, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 30), align 1
647  store i8 %r31, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 31), align 1
648  store i8 %r32, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 32), align 1
649  store i8 %r33, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 33), align 1
650  store i8 %r34, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 34), align 1
651  store i8 %r35, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 35), align 1
652  store i8 %r36, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 36), align 1
653  store i8 %r37, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 37), align 1
654  store i8 %r38, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 38), align 1
655  store i8 %r39, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 39), align 1
656  store i8 %r40, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 40), align 1
657  store i8 %r41, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 41), align 1
658  store i8 %r42, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 42), align 1
659  store i8 %r43, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 43), align 1
660  store i8 %r44, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 44), align 1
661  store i8 %r45, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 45), align 1
662  store i8 %r46, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 46), align 1
663  store i8 %r47, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 47), align 1
664  store i8 %r48, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 48), align 1
665  store i8 %r49, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 49), align 1
666  store i8 %r50, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 50), align 1
667  store i8 %r51, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 51), align 1
668  store i8 %r52, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 52), align 1
669  store i8 %r53, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 53), align 1
670  store i8 %r54, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 54), align 1
671  store i8 %r55, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 55), align 1
672  store i8 %r56, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 56), align 1
673  store i8 %r57, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 57), align 1
674  store i8 %r58, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 58), align 1
675  store i8 %r59, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 59), align 1
676  store i8 %r60, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 60), align 1
677  store i8 %r61, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 61), align 1
678  store i8 %r62, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 62), align 1
679  store i8 %r63, ptr getelementptr inbounds ([64 x i8], ptr @c8, i32 0, i64 63), align 1
680  ret void
681}
682