1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -aa-pipeline='basic-aa,scoped-noalias-aa' -passes=slp-vectorizer -mtriple=x86_64-apple-darwin -S %s | FileCheck %s
define void @version_multiple(ptr nocapture %out_block, ptr nocapture readonly %counter) {
; CHECK-LABEL: @version_multiple(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = load i32, ptr [[COUNTER:%.*]], align 4
; CHECK-NEXT: [[TMP1:%.*]] = load i32, ptr [[OUT_BLOCK:%.*]], align 4
; CHECK-NEXT: [[XOR:%.*]] = xor i32 [[TMP1]], [[TMP0]]
; CHECK-NEXT: store i32 [[XOR]], ptr [[OUT_BLOCK]], align 4
; CHECK-NEXT: [[ARRAYIDX_1:%.*]] = getelementptr inbounds i32, ptr [[COUNTER]], i64 1
; CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr [[ARRAYIDX_1]], align 4
; CHECK-NEXT: [[ARRAYIDX2_1:%.*]] = getelementptr inbounds i32, ptr [[OUT_BLOCK]], i64 1
; CHECK-NEXT: [[TMP3:%.*]] = load i32, ptr [[ARRAYIDX2_1]], align 4
; CHECK-NEXT: [[XOR_1:%.*]] = xor i32 [[TMP3]], [[TMP2]]
; CHECK-NEXT: store i32 [[XOR_1]], ptr [[ARRAYIDX2_1]], align 4
; CHECK-NEXT: [[ARRAYIDX_2:%.*]] = getelementptr inbounds i32, ptr [[COUNTER]], i64 2
; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr [[ARRAYIDX_2]], align 4
; CHECK-NEXT: [[ARRAYIDX2_2:%.*]] = getelementptr inbounds i32, ptr [[OUT_BLOCK]], i64 2
; CHECK-NEXT: [[TMP5:%.*]] = load i32, ptr [[ARRAYIDX2_2]], align 4
; CHECK-NEXT: [[XOR_2:%.*]] = xor i32 [[TMP5]], [[TMP4]]
; CHECK-NEXT: store i32 [[XOR_2]], ptr [[ARRAYIDX2_2]], align 4
; CHECK-NEXT: [[ARRAYIDX_3:%.*]] = getelementptr inbounds i32, ptr [[COUNTER]], i64 3
; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr [[ARRAYIDX_3]], align 4
; CHECK-NEXT: [[ARRAYIDX2_3:%.*]] = getelementptr inbounds i32, ptr [[OUT_BLOCK]], i64 3
; CHECK-NEXT: [[TMP7:%.*]] = load i32, ptr [[ARRAYIDX2_3]], align 4
; CHECK-NEXT: [[XOR_3:%.*]] = xor i32 [[TMP7]], [[TMP6]]
; CHECK-NEXT: store i32 [[XOR_3]], ptr [[ARRAYIDX2_3]], align 4
; CHECK-NEXT: ret void
;
entry:
%0 = load i32, ptr %counter, align 4
%1 = load i32, ptr %out_block, align 4
%xor = xor i32 %1, %0
store i32 %xor, ptr %out_block, align 4
%arrayidx.1 = getelementptr inbounds i32, ptr %counter, i64 1
%2 = load i32, ptr %arrayidx.1, align 4
%arrayidx2.1 = getelementptr inbounds i32, ptr %out_block, i64 1
%3 = load i32, ptr %arrayidx2.1, align 4
%xor.1 = xor i32 %3, %2
store i32 %xor.1, ptr %arrayidx2.1, align 4
%arrayidx.2 = getelementptr inbounds i32, ptr %counter, i64 2
%4 = load i32, ptr %arrayidx.2, align 4
%arrayidx2.2 = getelementptr inbounds i32, ptr %out_block, i64 2
%5 = load i32, ptr %arrayidx2.2, align 4
%xor.2 = xor i32 %5, %4
store i32 %xor.2, ptr %arrayidx2.2, align 4
%arrayidx.3 = getelementptr inbounds i32, ptr %counter, i64 3
%6 = load i32, ptr %arrayidx.3, align 4
%arrayidx2.3 = getelementptr inbounds i32, ptr %out_block, i64 3
%7 = load i32, ptr %arrayidx2.3, align 4
%xor.3 = xor i32 %7, %6
store i32 %xor.3, ptr %arrayidx2.3, align 4
ret void
}
declare void @use(<8 x float>)
define void @delete_pointer_bound(ptr %a, ptr %b, i1 %c) #0 {
; CHECK-LABEL: @delete_pointer_bound(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[B_10:%.*]] = getelementptr inbounds float, ptr [[B:%.*]], i64 10
; CHECK-NEXT: [[B_14:%.*]] = getelementptr inbounds float, ptr [[B]], i64 14
; CHECK-NEXT: br i1 [[C:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]]
; CHECK: else:
; CHECK-NEXT: [[L0:%.*]] = load float, ptr [[B_10]], align 4
; CHECK-NEXT: [[L1:%.*]] = load float, ptr [[B_14]], align 4
; CHECK-NEXT: [[I2:%.*]] = insertelement <8 x float> undef, float [[L0]], i32 2
; CHECK-NEXT: [[I3:%.*]] = insertelement <8 x float> [[I2]], float [[L0]], i32 3
; CHECK-NEXT: [[I4:%.*]] = insertelement <8 x float> [[I3]], float [[L1]], i32 4
; CHECK-NEXT: [[I7:%.*]] = insertelement <8 x float> [[I4]], float [[L1]], i32 7
; CHECK-NEXT: call void @use(<8 x float> [[I7]])
; CHECK-NEXT: ret void
; CHECK: then:
; CHECK-NEXT: [[A_8:%.*]] = getelementptr inbounds float, ptr [[A:%.*]], i64 8
; CHECK-NEXT: store float 0.000000e+00, ptr [[A_8]], align 4
; CHECK-NEXT: [[L6:%.*]] = load float, ptr [[B_14]], align 4
; CHECK-NEXT: [[A_5:%.*]] = getelementptr inbounds float, ptr [[A]], i64 5
; CHECK-NEXT: store float [[L6]], ptr [[A_5]], align 4
; CHECK-NEXT: [[A_6:%.*]] = getelementptr inbounds float, ptr [[A]], i64 6
; CHECK-NEXT: store <2 x float> zeroinitializer, ptr [[A_6]], align 4
; CHECK-NEXT: ret void
;
entry:
%b.10 = getelementptr inbounds float, ptr %b, i64 10
%b.14 = getelementptr inbounds float, ptr %b, i64 14
br i1 %c, label %then, label %else
else:
%l0 = load float, ptr %b.10, align 4
%l1 = load float, ptr %b.14, align 4
%i2 = insertelement <8 x float> undef, float %l0, i32 2
%i3 = insertelement <8 x float> %i2, float %l0, i32 3
%i4 = insertelement <8 x float> %i3, float %l1, i32 4
%i7 = insertelement <8 x float> %i4, float %l1, i32 7
call void @use(<8 x float> %i7)
ret void
then:
%a.8 = getelementptr inbounds float, ptr %a, i64 8
store float 0.0, ptr %a.8, align 4
%l6 = load float, ptr %b.14, align 4
%a.5 = getelementptr inbounds float, ptr %a, i64 5
store float %l6, ptr %a.5, align 4
%a.6 = getelementptr inbounds float, ptr %a, i64 6
store float 0.0, ptr %a.6, align 4
%a.7 = getelementptr inbounds float, ptr %a, i64 7
store float 0.0, ptr %a.7, align 4
ret void
}
%struct.zot = type { i16, i16, i16, i32, float, float, float, ptr, ptr, ptr, %struct.wombat.0 }
%struct.quux = type { i16, ptr, ptr }
%struct.wombat = type { i32, i16, i8, i8, ptr }
%struct.eggs = type { float, i8, %struct.ham }
%struct.ham = type { [2 x double], [8 x i8] }
%struct.wombat.0 = type { %struct.bar }
%struct.bar = type { [3 x double], [3 x double], double, double, i16, ptr, i32, [3 x double] }
define double @preserve_loop_info(ptr %arg) {
; CHECK-LABEL: @preserve_loop_info(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP:%.*]] = alloca [3 x double], align 16
; CHECK-NEXT: br label [[OUTER_HEADER:%.*]]
; CHECK: outer.header:
; CHECK-NEXT: br label [[INNER:%.*]]
; CHECK: inner:
; CHECK-NEXT: br i1 undef, label [[OUTER_LATCH:%.*]], label [[INNER]]
; CHECK: outer.latch:
; CHECK-NEXT: br i1 undef, label [[BB:%.*]], label [[OUTER_HEADER]]
; CHECK: bb:
; CHECK-NEXT: [[TMP5:%.*]] = load ptr, ptr undef, align 8
; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds [3 x double], ptr [[TMP]], i64 0, i64 1
; CHECK-NEXT: br label [[LOOP_3HEADER:%.*]]
; CHECK: loop.3header:
; CHECK-NEXT: br i1 undef, label [[LOOP_3LATCH:%.*]], label [[BB9:%.*]]
; CHECK: bb9:
; CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds [3 x double], ptr [[TMP5]], i64 undef, i64 1
; CHECK-NEXT: store double undef, ptr [[TMP]], align 16
; CHECK-NEXT: [[TMP12:%.*]] = load double, ptr [[TMP10]], align 8
; CHECK-NEXT: store double [[TMP12]], ptr [[TMP7]], align 8
; CHECK-NEXT: br label [[LOOP_3LATCH]]
; CHECK: loop.3latch:
; CHECK-NEXT: br i1 undef, label [[BB14:%.*]], label [[LOOP_3HEADER]]
; CHECK: bb14:
; CHECK-NEXT: [[TMP15:%.*]] = call double undef(ptr [[TMP]], ptr [[ARG:%.*]])
; CHECK-NEXT: ret double undef
;
entry:
%tmp = alloca [3 x double], align 16
br label %outer.header
outer.header: ; preds = %bb3, %bb
br label %inner
inner:
br i1 undef, label %outer.latch, label %inner
outer.latch: ; preds = %bb16
br i1 undef, label %bb, label %outer.header
bb: ; preds = %bb3
%tmp5 = load ptr, ptr undef, align 8
%tmp7 = getelementptr inbounds [3 x double], ptr %tmp, i64 0, i64 1
br label %loop.3header
loop.3header: ; preds = %bb13, %bb4
br i1 undef, label %loop.3latch, label %bb9
bb9: ; preds = %bb8
%tmp10 = getelementptr inbounds [3 x double], ptr %tmp5, i64 undef, i64 1
store double undef, ptr %tmp, align 16
%tmp12 = load double, ptr %tmp10, align 8
store double %tmp12, ptr %tmp7, align 8
br label %loop.3latch
loop.3latch: ; preds = %bb11, %bb8
br i1 undef, label %bb14, label %loop.3header
bb14: ; preds = %bb13
%tmp15 = call double undef(ptr %tmp, ptr %arg)
ret double undef
}
define void @gather_sequence_crash(<2 x float> %arg, ptr %arg1, float %arg2, ptr %arg3, ptr %arg4, ptr %arg5, i1 %c.1, i1 %c.2) {
; CHECK-LABEL: @gather_sequence_crash(
; CHECK-NEXT: bb:
; CHECK-NEXT: br i1 [[C_1:%.*]], label [[BB16:%.*]], label [[BB6:%.*]]
; CHECK: bb6:
; CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds float, ptr [[ARG1:%.*]], i32 3
; CHECK-NEXT: [[TMP0:%.*]] = shufflevector <2 x float> [[ARG:%.*]], <2 x float> poison, <4 x i32> <i32 poison, i32 0, i32 1, i32 poison>
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x float> [[TMP0]], <4 x float> <float poison, float poison, float poison, float 0.000000e+00>, <4 x i32> <i32 poison, i32 1, i32 2, i32 7>
; CHECK-NEXT: [[TMP2:%.*]] = insertelement <4 x float> [[TMP1]], float [[ARG2:%.*]], i32 0
; CHECK-NEXT: [[TMP3:%.*]] = fmul <4 x float> [[TMP2]], zeroinitializer
; CHECK-NEXT: store <4 x float> [[TMP3]], ptr [[TMP8]], align 4
; CHECK-NEXT: ret void
; CHECK: bb16:
; CHECK-NEXT: br label [[BB17:%.*]]
; CHECK: bb17:
; CHECK-NEXT: br label [[BB18:%.*]]
; CHECK: bb18:
; CHECK-NEXT: br label [[BB19:%.*]]
; CHECK: bb19:
; CHECK-NEXT: br label [[BB20:%.*]]
; CHECK: bb20:
; CHECK-NEXT: br label [[BB21:%.*]]
; CHECK: bb21:
; CHECK-NEXT: br label [[BB22:%.*]]
; CHECK: bb22:
; CHECK-NEXT: [[TMP24:%.*]] = getelementptr float, ptr [[ARG4:%.*]], i64 7
; CHECK-NEXT: br i1 [[C_2:%.*]], label [[BB25:%.*]], label [[BB22]]
; CHECK: bb25:
; CHECK-NEXT: [[TMP26:%.*]] = getelementptr float, ptr [[ARG4]], i64 6
; CHECK-NEXT: store float 0.000000e+00, ptr [[TMP24]], align 4
; CHECK-NEXT: [[TMP27:%.*]] = load float, ptr [[ARG5:%.*]], align 4
; CHECK-NEXT: [[TMP29:%.*]] = fadd float 0.000000e+00, 0.000000e+00
; CHECK-NEXT: store float 0.000000e+00, ptr [[TMP26]], align 4
; CHECK-NEXT: [[TMP30:%.*]] = getelementptr float, ptr [[ARG4]], i64 4
; CHECK-NEXT: [[TMP31:%.*]] = fadd float 0.000000e+00, 0.000000e+00
; CHECK-NEXT: store <2 x float> zeroinitializer, ptr [[TMP30]], align 4
; CHECK-NEXT: br label [[BB33:%.*]]
; CHECK: bb33:
; CHECK-NEXT: br label [[BB34:%.*]]
; CHECK: bb34:
; CHECK-NEXT: [[TMP35:%.*]] = getelementptr float, ptr [[ARG4]], i64 3
; CHECK-NEXT: [[TMP37:%.*]] = load float, ptr [[TMP35]], align 4
; CHECK-NEXT: [[TMP38:%.*]] = fadd float 0.000000e+00, [[TMP37]]
; CHECK-NEXT: store float [[TMP38]], ptr [[TMP35]], align 4
; CHECK-NEXT: [[TMP39:%.*]] = getelementptr float, ptr [[ARG4]], i64 1
; CHECK-NEXT: [[TMP4:%.*]] = load <2 x float>, ptr [[TMP39]], align 4
; CHECK-NEXT: [[TMP5:%.*]] = fadd <2 x float> zeroinitializer, [[TMP4]]
; CHECK-NEXT: store <2 x float> [[TMP5]], ptr [[TMP39]], align 4
; CHECK-NEXT: [[TMP44:%.*]] = load float, ptr [[ARG3:%.*]], align 4
; CHECK-NEXT: [[TMP45:%.*]] = load float, ptr [[ARG4]], align 4
; CHECK-NEXT: [[TMP46:%.*]] = fadd float 0.000000e+00, [[TMP45]]
; CHECK-NEXT: store float [[TMP46]], ptr [[ARG4]], align 4
; CHECK-NEXT: call void @quux()
; CHECK-NEXT: br label [[BB47:%.*]]
; CHECK: bb47:
; CHECK-NEXT: br label [[BB17]]
;
bb:
br i1 %c.1, label %bb16, label %bb6
bb6: ; preds = %bb
%tmp = getelementptr inbounds float, ptr %arg1, i32 4
%tmp7 = getelementptr inbounds float, ptr %arg1, i32 5
%tmp8 = getelementptr inbounds float, ptr %arg1, i32 3
%tmp9 = getelementptr inbounds float, ptr %arg1, i32 6
%tmp10 = extractelement <2 x float> %arg, i32 0
%tmp11 = fmul float %tmp10, 0.000000e+00
store float %tmp11, ptr %tmp, align 4
%tmp12 = extractelement <2 x float> %arg, i32 1
%tmp13 = fmul float %tmp12, 0.000000e+00
store float %tmp13, ptr %tmp7, align 4
%tmp14 = fmul float %arg2, 0.000000e+00
store float %tmp14, ptr %tmp8, align 4
%tmp15 = fmul float 0.000000e+00, 0.000000e+00
store float %tmp15, ptr %tmp9, align 4
ret void
bb16: ; preds = %bb
br label %bb17
bb17: ; preds = %bb47, %bb16
br label %bb18
bb18: ; preds = %bb17
br label %bb19
bb19: ; preds = %bb18
br label %bb20
bb20: ; preds = %bb19
br label %bb21
bb21: ; preds = %bb20
br label %bb22
bb22: ; preds = %bb22, %bb21
%tmp24 = getelementptr float, ptr %arg4, i64 7
br i1 %c.2, label %bb25, label %bb22
bb25: ; preds = %bb22
%tmp26 = getelementptr float, ptr %arg4, i64 6
store float 0.000000e+00, ptr %tmp24, align 4
%tmp27 = load float, ptr %arg5, align 4
%tmp28 = getelementptr float, ptr %arg4, i64 5
%tmp29 = fadd float 0.000000e+00, 0.000000e+00
store float 0.000000e+00, ptr %tmp26, align 4
%tmp30 = getelementptr float, ptr %arg4, i64 4
store float 0.000000e+00, ptr %tmp28, align 4
%tmp31 = fadd float 0.000000e+00, 0.000000e+00
store float 0.000000e+00, ptr %tmp30, align 4
br label %bb33
bb33: ; preds = %bb25
br label %bb34
bb34: ; preds = %bb33
%tmp35 = getelementptr float, ptr %arg4, i64 3
%tmp36 = getelementptr float, ptr %arg4, i64 2
%tmp37 = load float, ptr %tmp35, align 4
%tmp38 = fadd float 0.000000e+00, %tmp37
store float %tmp38, ptr %tmp35, align 4
%tmp39 = getelementptr float, ptr %arg4, i64 1
%tmp40 = load float, ptr %tmp36, align 4
%tmp41 = fadd float 0.000000e+00, %tmp40
store float %tmp41, ptr %tmp36, align 4
%tmp42 = load float, ptr %tmp39, align 4
%tmp43 = fadd float 0.000000e+00, %tmp42
store float %tmp43, ptr %tmp39, align 4
%tmp44 = load float, ptr %arg3, align 4
%tmp45 = load float, ptr %arg4, align 4
%tmp46 = fadd float 0.000000e+00, %tmp45
store float %tmp46, ptr %arg4, align 4
call void @quux()
br label %bb47
bb47: ; preds = %bb34
br label %bb17
}
declare void @quux()
attributes #0 = { "target-features"="+avx2" }
|
