1 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --filter "LV: Found an estimated cost of [0-9]+ for VF [0-9]+ For instruction:\s*%v. = load i8, ptr %in."
2 ; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+sse2 --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefix=SSE2
3 ; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefix=AVX1
4 ; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2 --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefix=AVX2
5 ; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512vl --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefix=AVX512DQ
6 ; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512vl,+avx512bw --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefix=AVX512BW
9 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
10 target triple = "x86_64-unknown-linux-gnu"
12 @A = global [1024 x i8] zeroinitializer, align 128
13 @B = global [1024 x i8] zeroinitializer, align 128
17 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
18 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
19 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
20 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
21 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
22 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
23 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
24 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
25 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
26 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
27 ; SSE2: LV: Found an estimated cost of 38 for VF 2 For instruction: %v0 = load i8, ptr %in0, align 1
28 ; SSE2: LV: Found an estimated cost of 0 for VF 2 For instruction: %v1 = load i8, ptr %in1, align 1
29 ; SSE2: LV: Found an estimated cost of 0 for VF 2 For instruction: %v2 = load i8, ptr %in2, align 1
30 ; SSE2: LV: Found an estimated cost of 0 for VF 2 For instruction: %v3 = load i8, ptr %in3, align 1
31 ; SSE2: LV: Found an estimated cost of 0 for VF 2 For instruction: %v4 = load i8, ptr %in4, align 1
32 ; SSE2: LV: Found an estimated cost of 75 for VF 4 For instruction: %v0 = load i8, ptr %in0, align 1
33 ; SSE2: LV: Found an estimated cost of 0 for VF 4 For instruction: %v1 = load i8, ptr %in1, align 1
34 ; SSE2: LV: Found an estimated cost of 0 for VF 4 For instruction: %v2 = load i8, ptr %in2, align 1
35 ; SSE2: LV: Found an estimated cost of 0 for VF 4 For instruction: %v3 = load i8, ptr %in3, align 1
36 ; SSE2: LV: Found an estimated cost of 0 for VF 4 For instruction: %v4 = load i8, ptr %in4, align 1
37 ; SSE2: LV: Found an estimated cost of 155 for VF 8 For instruction: %v0 = load i8, ptr %in0, align 1
38 ; SSE2: LV: Found an estimated cost of 0 for VF 8 For instruction: %v1 = load i8, ptr %in1, align 1
39 ; SSE2: LV: Found an estimated cost of 0 for VF 8 For instruction: %v2 = load i8, ptr %in2, align 1
40 ; SSE2: LV: Found an estimated cost of 0 for VF 8 For instruction: %v3 = load i8, ptr %in3, align 1
41 ; SSE2: LV: Found an estimated cost of 0 for VF 8 For instruction: %v4 = load i8, ptr %in4, align 1
42 ; SSE2: LV: Found an estimated cost of 315 for VF 16 For instruction: %v0 = load i8, ptr %in0, align 1
43 ; SSE2: LV: Found an estimated cost of 0 for VF 16 For instruction: %v1 = load i8, ptr %in1, align 1
44 ; SSE2: LV: Found an estimated cost of 0 for VF 16 For instruction: %v2 = load i8, ptr %in2, align 1
45 ; SSE2: LV: Found an estimated cost of 0 for VF 16 For instruction: %v3 = load i8, ptr %in3, align 1
46 ; SSE2: LV: Found an estimated cost of 0 for VF 16 For instruction: %v4 = load i8, ptr %in4, align 1
49 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
50 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
51 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
52 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
53 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
54 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
55 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
56 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
57 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
58 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
59 ; AVX1: LV: Found an estimated cost of 24 for VF 2 For instruction: %v0 = load i8, ptr %in0, align 1
60 ; AVX1: LV: Found an estimated cost of 0 for VF 2 For instruction: %v1 = load i8, ptr %in1, align 1
61 ; AVX1: LV: Found an estimated cost of 0 for VF 2 For instruction: %v2 = load i8, ptr %in2, align 1
62 ; AVX1: LV: Found an estimated cost of 0 for VF 2 For instruction: %v3 = load i8, ptr %in3, align 1
63 ; AVX1: LV: Found an estimated cost of 0 for VF 2 For instruction: %v4 = load i8, ptr %in4, align 1
64 ; AVX1: LV: Found an estimated cost of 45 for VF 4 For instruction: %v0 = load i8, ptr %in0, align 1
65 ; AVX1: LV: Found an estimated cost of 0 for VF 4 For instruction: %v1 = load i8, ptr %in1, align 1
66 ; AVX1: LV: Found an estimated cost of 0 for VF 4 For instruction: %v2 = load i8, ptr %in2, align 1
67 ; AVX1: LV: Found an estimated cost of 0 for VF 4 For instruction: %v3 = load i8, ptr %in3, align 1
68 ; AVX1: LV: Found an estimated cost of 0 for VF 4 For instruction: %v4 = load i8, ptr %in4, align 1
69 ; AVX1: LV: Found an estimated cost of 83 for VF 8 For instruction: %v0 = load i8, ptr %in0, align 1
70 ; AVX1: LV: Found an estimated cost of 0 for VF 8 For instruction: %v1 = load i8, ptr %in1, align 1
71 ; AVX1: LV: Found an estimated cost of 0 for VF 8 For instruction: %v2 = load i8, ptr %in2, align 1
72 ; AVX1: LV: Found an estimated cost of 0 for VF 8 For instruction: %v3 = load i8, ptr %in3, align 1
73 ; AVX1: LV: Found an estimated cost of 0 for VF 8 For instruction: %v4 = load i8, ptr %in4, align 1
74 ; AVX1: LV: Found an estimated cost of 165 for VF 16 For instruction: %v0 = load i8, ptr %in0, align 1
75 ; AVX1: LV: Found an estimated cost of 0 for VF 16 For instruction: %v1 = load i8, ptr %in1, align 1
76 ; AVX1: LV: Found an estimated cost of 0 for VF 16 For instruction: %v2 = load i8, ptr %in2, align 1
77 ; AVX1: LV: Found an estimated cost of 0 for VF 16 For instruction: %v3 = load i8, ptr %in3, align 1
78 ; AVX1: LV: Found an estimated cost of 0 for VF 16 For instruction: %v4 = load i8, ptr %in4, align 1
79 ; AVX1: LV: Found an estimated cost of 335 for VF 32 For instruction: %v0 = load i8, ptr %in0, align 1
80 ; AVX1: LV: Found an estimated cost of 0 for VF 32 For instruction: %v1 = load i8, ptr %in1, align 1
81 ; AVX1: LV: Found an estimated cost of 0 for VF 32 For instruction: %v2 = load i8, ptr %in2, align 1
82 ; AVX1: LV: Found an estimated cost of 0 for VF 32 For instruction: %v3 = load i8, ptr %in3, align 1
83 ; AVX1: LV: Found an estimated cost of 0 for VF 32 For instruction: %v4 = load i8, ptr %in4, align 1
86 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
87 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
88 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
89 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
90 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
91 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
92 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
93 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
94 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
95 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
96 ; AVX2: LV: Found an estimated cost of 4 for VF 2 For instruction: %v0 = load i8, ptr %in0, align 1
97 ; AVX2: LV: Found an estimated cost of 4 for VF 2 For instruction: %v1 = load i8, ptr %in1, align 1
98 ; AVX2: LV: Found an estimated cost of 4 for VF 2 For instruction: %v2 = load i8, ptr %in2, align 1
99 ; AVX2: LV: Found an estimated cost of 4 for VF 2 For instruction: %v3 = load i8, ptr %in3, align 1
100 ; AVX2: LV: Found an estimated cost of 4 for VF 2 For instruction: %v4 = load i8, ptr %in4, align 1
101 ; AVX2: LV: Found an estimated cost of 8 for VF 4 For instruction: %v0 = load i8, ptr %in0, align 1
102 ; AVX2: LV: Found an estimated cost of 8 for VF 4 For instruction: %v1 = load i8, ptr %in1, align 1
103 ; AVX2: LV: Found an estimated cost of 8 for VF 4 For instruction: %v2 = load i8, ptr %in2, align 1
104 ; AVX2: LV: Found an estimated cost of 8 for VF 4 For instruction: %v3 = load i8, ptr %in3, align 1
105 ; AVX2: LV: Found an estimated cost of 8 for VF 4 For instruction: %v4 = load i8, ptr %in4, align 1
106 ; AVX2: LV: Found an estimated cost of 16 for VF 8 For instruction: %v0 = load i8, ptr %in0, align 1
107 ; AVX2: LV: Found an estimated cost of 16 for VF 8 For instruction: %v1 = load i8, ptr %in1, align 1
108 ; AVX2: LV: Found an estimated cost of 16 for VF 8 For instruction: %v2 = load i8, ptr %in2, align 1
109 ; AVX2: LV: Found an estimated cost of 16 for VF 8 For instruction: %v3 = load i8, ptr %in3, align 1
110 ; AVX2: LV: Found an estimated cost of 16 for VF 8 For instruction: %v4 = load i8, ptr %in4, align 1
111 ; AVX2: LV: Found an estimated cost of 32 for VF 16 For instruction: %v0 = load i8, ptr %in0, align 1
112 ; AVX2: LV: Found an estimated cost of 32 for VF 16 For instruction: %v1 = load i8, ptr %in1, align 1
113 ; AVX2: LV: Found an estimated cost of 32 for VF 16 For instruction: %v2 = load i8, ptr %in2, align 1
114 ; AVX2: LV: Found an estimated cost of 32 for VF 16 For instruction: %v3 = load i8, ptr %in3, align 1
115 ; AVX2: LV: Found an estimated cost of 32 for VF 16 For instruction: %v4 = load i8, ptr %in4, align 1
116 ; AVX2: LV: Found an estimated cost of 65 for VF 32 For instruction: %v0 = load i8, ptr %in0, align 1
117 ; AVX2: LV: Found an estimated cost of 65 for VF 32 For instruction: %v1 = load i8, ptr %in1, align 1
118 ; AVX2: LV: Found an estimated cost of 65 for VF 32 For instruction: %v2 = load i8, ptr %in2, align 1
119 ; AVX2: LV: Found an estimated cost of 65 for VF 32 For instruction: %v3 = load i8, ptr %in3, align 1
120 ; AVX2: LV: Found an estimated cost of 65 for VF 32 For instruction: %v4 = load i8, ptr %in4, align 1
122 ; AVX512DQ-LABEL: 'test'
123 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
124 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
125 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
126 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
127 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
128 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
129 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
130 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
131 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
132 ; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
133 ; AVX512DQ: LV: Found an estimated cost of 24 for VF 2 For instruction: %v0 = load i8, ptr %in0, align 1
134 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 2 For instruction: %v1 = load i8, ptr %in1, align 1
135 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 2 For instruction: %v2 = load i8, ptr %in2, align 1
136 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 2 For instruction: %v3 = load i8, ptr %in3, align 1
137 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 2 For instruction: %v4 = load i8, ptr %in4, align 1
138 ; AVX512DQ: LV: Found an estimated cost of 45 for VF 4 For instruction: %v0 = load i8, ptr %in0, align 1
139 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 4 For instruction: %v1 = load i8, ptr %in1, align 1
140 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 4 For instruction: %v2 = load i8, ptr %in2, align 1
141 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 4 For instruction: %v3 = load i8, ptr %in3, align 1
142 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 4 For instruction: %v4 = load i8, ptr %in4, align 1
143 ; AVX512DQ: LV: Found an estimated cost of 85 for VF 8 For instruction: %v0 = load i8, ptr %in0, align 1
144 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 8 For instruction: %v1 = load i8, ptr %in1, align 1
145 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 8 For instruction: %v2 = load i8, ptr %in2, align 1
146 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 8 For instruction: %v3 = load i8, ptr %in3, align 1
147 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 8 For instruction: %v4 = load i8, ptr %in4, align 1
148 ; AVX512DQ: LV: Found an estimated cost of 165 for VF 16 For instruction: %v0 = load i8, ptr %in0, align 1
149 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 16 For instruction: %v1 = load i8, ptr %in1, align 1
150 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 16 For instruction: %v2 = load i8, ptr %in2, align 1
151 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 16 For instruction: %v3 = load i8, ptr %in3, align 1
152 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 16 For instruction: %v4 = load i8, ptr %in4, align 1
153 ; AVX512DQ: LV: Found an estimated cost of 335 for VF 32 For instruction: %v0 = load i8, ptr %in0, align 1
154 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 32 For instruction: %v1 = load i8, ptr %in1, align 1
155 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 32 For instruction: %v2 = load i8, ptr %in2, align 1
156 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 32 For instruction: %v3 = load i8, ptr %in3, align 1
157 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 32 For instruction: %v4 = load i8, ptr %in4, align 1
158 ; AVX512DQ: LV: Found an estimated cost of 675 for VF 64 For instruction: %v0 = load i8, ptr %in0, align 1
159 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 64 For instruction: %v1 = load i8, ptr %in1, align 1
160 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 64 For instruction: %v2 = load i8, ptr %in2, align 1
161 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 64 For instruction: %v3 = load i8, ptr %in3, align 1
162 ; AVX512DQ: LV: Found an estimated cost of 0 for VF 64 For instruction: %v4 = load i8, ptr %in4, align 1
164 ; AVX512BW-LABEL: 'test'
165 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
166 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
167 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
168 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
169 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
170 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i8, ptr %in0, align 1
171 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v1 = load i8, ptr %in1, align 1
172 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v2 = load i8, ptr %in2, align 1
173 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v3 = load i8, ptr %in3, align 1
174 ; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: %v4 = load i8, ptr %in4, align 1
175 ; AVX512BW: LV: Found an estimated cost of 6 for VF 2 For instruction: %v0 = load i8, ptr %in0, align 1
176 ; AVX512BW: LV: Found an estimated cost of 0 for VF 2 For instruction: %v1 = load i8, ptr %in1, align 1
177 ; AVX512BW: LV: Found an estimated cost of 0 for VF 2 For instruction: %v2 = load i8, ptr %in2, align 1
178 ; AVX512BW: LV: Found an estimated cost of 0 for VF 2 For instruction: %v3 = load i8, ptr %in3, align 1
179 ; AVX512BW: LV: Found an estimated cost of 0 for VF 2 For instruction: %v4 = load i8, ptr %in4, align 1
180 ; AVX512BW: LV: Found an estimated cost of 21 for VF 4 For instruction: %v0 = load i8, ptr %in0, align 1
181 ; AVX512BW: LV: Found an estimated cost of 0 for VF 4 For instruction: %v1 = load i8, ptr %in1, align 1
182 ; AVX512BW: LV: Found an estimated cost of 0 for VF 4 For instruction: %v2 = load i8, ptr %in2, align 1
183 ; AVX512BW: LV: Found an estimated cost of 0 for VF 4 For instruction: %v3 = load i8, ptr %in3, align 1
184 ; AVX512BW: LV: Found an estimated cost of 0 for VF 4 For instruction: %v4 = load i8, ptr %in4, align 1
185 ; AVX512BW: LV: Found an estimated cost of 41 for VF 8 For instruction: %v0 = load i8, ptr %in0, align 1
186 ; AVX512BW: LV: Found an estimated cost of 0 for VF 8 For instruction: %v1 = load i8, ptr %in1, align 1
187 ; AVX512BW: LV: Found an estimated cost of 0 for VF 8 For instruction: %v2 = load i8, ptr %in2, align 1
188 ; AVX512BW: LV: Found an estimated cost of 0 for VF 8 For instruction: %v3 = load i8, ptr %in3, align 1
189 ; AVX512BW: LV: Found an estimated cost of 0 for VF 8 For instruction: %v4 = load i8, ptr %in4, align 1
190 ; AVX512BW: LV: Found an estimated cost of 99 for VF 16 For instruction: %v0 = load i8, ptr %in0, align 1
191 ; AVX512BW: LV: Found an estimated cost of 0 for VF 16 For instruction: %v1 = load i8, ptr %in1, align 1
192 ; AVX512BW: LV: Found an estimated cost of 0 for VF 16 For instruction: %v2 = load i8, ptr %in2, align 1
193 ; AVX512BW: LV: Found an estimated cost of 0 for VF 16 For instruction: %v3 = load i8, ptr %in3, align 1
194 ; AVX512BW: LV: Found an estimated cost of 0 for VF 16 For instruction: %v4 = load i8, ptr %in4, align 1
195 ; AVX512BW: LV: Found an estimated cost of 198 for VF 32 For instruction: %v0 = load i8, ptr %in0, align 1
196 ; AVX512BW: LV: Found an estimated cost of 0 for VF 32 For instruction: %v1 = load i8, ptr %in1, align 1
197 ; AVX512BW: LV: Found an estimated cost of 0 for VF 32 For instruction: %v2 = load i8, ptr %in2, align 1
198 ; AVX512BW: LV: Found an estimated cost of 0 for VF 32 For instruction: %v3 = load i8, ptr %in3, align 1
199 ; AVX512BW: LV: Found an estimated cost of 0 for VF 32 For instruction: %v4 = load i8, ptr %in4, align 1
200 ; AVX512BW: LV: Found an estimated cost of 395 for VF 64 For instruction: %v0 = load i8, ptr %in0, align 1
201 ; AVX512BW: LV: Found an estimated cost of 0 for VF 64 For instruction: %v1 = load i8, ptr %in1, align 1
202 ; AVX512BW: LV: Found an estimated cost of 0 for VF 64 For instruction: %v2 = load i8, ptr %in2, align 1
203 ; AVX512BW: LV: Found an estimated cost of 0 for VF 64 For instruction: %v3 = load i8, ptr %in3, align 1
204 ; AVX512BW: LV: Found an estimated cost of 0 for VF 64 For instruction: %v4 = load i8, ptr %in4, align 1
210 %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
212 %iv.0 = add nuw nsw i64 %iv, 0
213 %iv.1 = add nuw nsw i64 %iv, 1
214 %iv.2 = add nuw nsw i64 %iv, 2
215 %iv.3 = add nuw nsw i64 %iv, 3
216 %iv.4 = add nuw nsw i64 %iv, 4
218 %in0 = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.0
219 %in1 = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.1
220 %in2 = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.2
221 %in3 = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.3
222 %in4 = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.4
224 %v0 = load i8, ptr %in0
225 %v1 = load i8, ptr %in1
226 %v2 = load i8, ptr %in2
227 %v3 = load i8, ptr %in3
228 %v4 = load i8, ptr %in4
230 %reduce.add.0 = add i8 %v0, %v1
231 %reduce.add.1 = add i8 %reduce.add.0, %v2
232 %reduce.add.2 = add i8 %reduce.add.1, %v3
233 %reduce.add.3 = add i8 %reduce.add.2, %v4
235 %out = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.0
236 store i8 %reduce.add.3, ptr %out
238 %iv.next = add nuw nsw i64 %iv.0, 5
239 %cmp = icmp ult i64 %iv.next, 1024
240 br i1 %cmp, label %for.body, label %for.cond.cleanup