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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -passes=instcombine -S %s | FileCheck %s
target datalayout = "p:32:32"
define i1 @alloca_argument_compare(ptr %arg) {
; CHECK-LABEL: @alloca_argument_compare(
; CHECK-NEXT: ret i1 false
;
%alloc = alloca i64
%cmp = icmp eq ptr %arg, %alloc
ret i1 %cmp
}
define i1 @alloca_argument_compare_swapped(ptr %arg) {
; CHECK-LABEL: @alloca_argument_compare_swapped(
; CHECK-NEXT: ret i1 false
;
%alloc = alloca i64
%cmp = icmp eq ptr %alloc, %arg
ret i1 %cmp
}
define i1 @alloca_argument_compare_ne(ptr %arg) {
; CHECK-LABEL: @alloca_argument_compare_ne(
; CHECK-NEXT: ret i1 true
;
%alloc = alloca i64
%cmp = icmp ne ptr %arg, %alloc
ret i1 %cmp
}
define i1 @alloca_argument_compare_derived_ptrs(ptr %arg, i64 %x) {
; CHECK-LABEL: @alloca_argument_compare_derived_ptrs(
; CHECK-NEXT: ret i1 false
;
%alloc = alloca i64, i64 8
%p = getelementptr i64, ptr %arg, i64 %x
%q = getelementptr i64, ptr %alloc, i64 3
%cmp = icmp eq ptr %p, %q
ret i1 %cmp
}
declare void @escape(ptr)
define i1 @alloca_argument_compare_escaped_alloca(ptr %arg) {
; CHECK-LABEL: @alloca_argument_compare_escaped_alloca(
; CHECK-NEXT: [[ALLOC:%.*]] = alloca i64, align 8
; CHECK-NEXT: call void @escape(ptr nonnull [[ALLOC]])
; CHECK-NEXT: [[CMP:%.*]] = icmp eq ptr [[ALLOC]], [[ARG:%.*]]
; CHECK-NEXT: ret i1 [[CMP]]
;
%alloc = alloca i64
call void @escape(ptr %alloc)
%cmp = icmp eq ptr %alloc, %arg
ret i1 %cmp
}
declare void @check_compares(i1, i1)
define void @alloca_argument_compare_two_compares(ptr %p) {
; CHECK-LABEL: @alloca_argument_compare_two_compares(
; CHECK-NEXT: call void @check_compares(i1 false, i1 false)
; CHECK-NEXT: ret void
;
%q = alloca i64, i64 8
%r = getelementptr i64, ptr %p, i64 1
%s = getelementptr i64, ptr %q, i64 2
%cmp1 = icmp eq ptr %p, %q
%cmp2 = icmp eq ptr %r, %s
call void @check_compares(i1 %cmp1, i1 %cmp2)
ret void
; We will only fold if there is a single cmp.
}
define i1 @alloca_argument_compare_escaped_through_store(ptr %arg, ptr %ptr) {
; CHECK-LABEL: @alloca_argument_compare_escaped_through_store(
; CHECK-NEXT: [[ALLOC:%.*]] = alloca i64, align 8
; CHECK-NEXT: [[CMP:%.*]] = icmp eq ptr [[ALLOC]], [[ARG:%.*]]
; CHECK-NEXT: [[P:%.*]] = getelementptr inbounds i64, ptr [[ALLOC]], i32 1
; CHECK-NEXT: store ptr [[P]], ptr [[PTR:%.*]], align 4
; CHECK-NEXT: ret i1 [[CMP]]
;
%alloc = alloca i64
%cmp = icmp eq ptr %alloc, %arg
%p = getelementptr i64, ptr %alloc, i64 1
store ptr %p, ptr %ptr
ret i1 %cmp
}
declare void @llvm.lifetime.start.p0(i64, ptr nocapture)
declare void @llvm.lifetime.end.p0(i64, ptr nocapture)
define i1 @alloca_argument_compare_benign_instrs(ptr %arg) {
; CHECK-LABEL: @alloca_argument_compare_benign_instrs(
; CHECK-NEXT: ret i1 false
;
%alloc = alloca i8
call void @llvm.lifetime.start.p0(i64 1, ptr %alloc)
%cmp = icmp eq ptr %arg, %alloc
%x = load i8, ptr %arg
store i8 %x, ptr %alloc
call void @llvm.lifetime.end.p0(i64 1, ptr %alloc)
ret i1 %cmp
}
declare ptr @allocator()
define i1 @alloca_call_compare() {
; CHECK-LABEL: @alloca_call_compare(
; CHECK-NEXT: [[Q:%.*]] = call ptr @allocator()
; CHECK-NEXT: ret i1 false
;
%p = alloca i64
%q = call ptr @allocator()
%cmp = icmp eq ptr %p, %q
ret i1 %cmp
}
; The next block of tests demonstrate a very subtle correctness requirement.
; We can generally assume any *single* stack layout we chose for the result of
; an alloca, but we can't simultanious assume two different ones. As a
; result, we must make sure that we only fold conditions if we can ensure that
; we fold *all* potentially address capturing compares the same.
; These two functions represents either a) forging a pointer via inttoptr or
; b) indexing off an adjacent allocation. In either case, the operation is
; obscured by an uninlined helper and not visible to instcombine.
declare ptr @hidden_inttoptr()
declare ptr @hidden_offset(ptr %other)
define i1 @ptrtoint_single_cmp() {
; CHECK-LABEL: @ptrtoint_single_cmp(
; CHECK-NEXT: ret i1 false
;
%m = alloca i8, i32 4
%rhs = inttoptr i64 2048 to ptr
%cmp = icmp eq ptr %m, %rhs
ret i1 %cmp
}
define i1 @offset_single_cmp() {
; CHECK-LABEL: @offset_single_cmp(
; CHECK-NEXT: ret i1 false
;
%m = alloca i8, i32 4
%n = alloca i8, i32 4
%rhs = getelementptr i8, ptr %n, i32 4
%cmp = icmp eq ptr %m, %rhs
ret i1 %cmp
}
declare void @witness(i1, i1)
define void @consistent_fold1() {
; CHECK-LABEL: @consistent_fold1(
; CHECK-NEXT: [[RHS2:%.*]] = call ptr @hidden_inttoptr()
; CHECK-NEXT: call void @witness(i1 false, i1 false)
; CHECK-NEXT: ret void
;
%m = alloca i8, i32 4
%rhs = inttoptr i64 2048 to ptr
%rhs2 = call ptr @hidden_inttoptr()
%cmp1 = icmp eq ptr %m, %rhs
%cmp2 = icmp eq ptr %m, %rhs2
call void @witness(i1 %cmp1, i1 %cmp2)
ret void
}
define void @consistent_fold2() {
; CHECK-LABEL: @consistent_fold2(
; CHECK-NEXT: [[N2:%.*]] = alloca [4 x i8], align 1
; CHECK-NEXT: [[RHS2:%.*]] = call ptr @hidden_offset(ptr nonnull [[N2]])
; CHECK-NEXT: call void @witness(i1 false, i1 false)
; CHECK-NEXT: ret void
;
%m = alloca i8, i32 4
%n = alloca i8, i32 4
%rhs = getelementptr i8, ptr %n, i32 4
%rhs2 = call ptr @hidden_offset(ptr %n)
%cmp1 = icmp eq ptr %m, %rhs
%cmp2 = icmp eq ptr %m, %rhs2
call void @witness(i1 %cmp1, i1 %cmp2)
ret void
}
define void @consistent_fold3() {
; CHECK-LABEL: @consistent_fold3(
; CHECK-NEXT: [[RHS2:%.*]] = call ptr @hidden_inttoptr()
; CHECK-NEXT: call void @witness(i1 false, i1 false)
; CHECK-NEXT: ret void
;
%m = alloca i8, i32 4
%lgp = load ptr, ptr @gp, align 8
%rhs2 = call ptr @hidden_inttoptr()
%cmp1 = icmp eq ptr %m, %lgp
%cmp2 = icmp eq ptr %m, %rhs2
call void @witness(i1 %cmp1, i1 %cmp2)
ret void
}
define void @neg_consistent_fold4() {
; CHECK-LABEL: @neg_consistent_fold4(
; CHECK-NEXT: call void @witness(i1 false, i1 false)
; CHECK-NEXT: ret void
;
%m = alloca i8, i32 4
%lgp = load ptr, ptr @gp, align 8
%cmp1 = icmp eq ptr %m, %lgp
%cmp2 = icmp eq ptr %m, %lgp
call void @witness(i1 %cmp1, i1 %cmp2)
ret void
}
; A nocapture call can't cause a consistent result issue as it is (by
; assumption) not able to contain a comparison which might capture the
; address.
declare void @unknown(ptr)
define i1 @consistent_nocapture_inttoptr() {
; CHECK-LABEL: @consistent_nocapture_inttoptr(
; CHECK-NEXT: [[M1:%.*]] = alloca [4 x i8], align 1
; CHECK-NEXT: call void @unknown(ptr nocapture nonnull [[M1]])
; CHECK-NEXT: ret i1 false
;
%m = alloca i8, i32 4
call void @unknown(ptr nocapture %m)
%rhs = inttoptr i64 2048 to ptr
%cmp = icmp eq ptr %m, %rhs
ret i1 %cmp
}
define i1 @consistent_nocapture_offset() {
; CHECK-LABEL: @consistent_nocapture_offset(
; CHECK-NEXT: [[M1:%.*]] = alloca [4 x i8], align 1
; CHECK-NEXT: call void @unknown(ptr nocapture nonnull [[M1]])
; CHECK-NEXT: ret i1 false
;
%m = alloca i8, i32 4
call void @unknown(ptr nocapture %m)
%n = alloca i8, i32 4
%rhs = getelementptr i8, ptr %n, i32 4
%cmp = icmp eq ptr %m, %rhs
ret i1 %cmp
}
@gp = global ptr null, align 8
define i1 @consistent_nocapture_through_global() {
; CHECK-LABEL: @consistent_nocapture_through_global(
; CHECK-NEXT: [[M1:%.*]] = alloca [4 x i8], align 1
; CHECK-NEXT: call void @unknown(ptr nocapture nonnull [[M1]])
; CHECK-NEXT: ret i1 false
;
%m = alloca i8, i32 4
call void @unknown(ptr nocapture %m)
%lgp = load ptr, ptr @gp, align 8, !nonnull !{}
%cmp = icmp eq ptr %m, %lgp
ret i1 %cmp
}
define void @select_alloca_unrelated_ptr(i1 %c, ptr %p, ptr %p2) {
; CHECK-LABEL: @select_alloca_unrelated_ptr(
; CHECK-NEXT: [[M:%.*]] = alloca i8, align 1
; CHECK-NEXT: [[CMP1:%.*]] = icmp eq ptr [[M]], [[P:%.*]]
; CHECK-NEXT: [[S:%.*]] = select i1 [[C:%.*]], ptr [[M]], ptr [[P2:%.*]]
; CHECK-NEXT: [[CMP2:%.*]] = icmp eq ptr [[S]], [[P]]
; CHECK-NEXT: call void @witness(i1 [[CMP1]], i1 [[CMP2]])
; CHECK-NEXT: ret void
;
%m = alloca i8
%cmp1 = icmp eq ptr %m, %p
%s = select i1 %c, ptr %m, ptr %p2
%cmp2 = icmp eq ptr %s, %p
call void @witness(i1 %cmp1, i1 %cmp2)
ret void
}
define void @alloca_offset_icmp(ptr %p, i32 %offset) {
; CHECK-LABEL: @alloca_offset_icmp(
; CHECK-NEXT: [[CMP2:%.*]] = icmp eq i32 [[OFFSET:%.*]], 0
; CHECK-NEXT: call void @witness(i1 false, i1 [[CMP2]])
; CHECK-NEXT: ret void
;
%m = alloca [4 x i8]
%g = getelementptr i8, ptr %m, i32 %offset
%cmp1 = icmp eq ptr %m, %p
%cmp2 = icmp eq ptr %m, %g
call void @witness(i1 %cmp1, i1 %cmp2)
ret void
}
|
