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
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -S -passes=early-cse -earlycse-debug-hash | FileCheck %s
; RUN: opt < %s -S -passes='early-cse<memssa>' | FileCheck %s
; RUN: opt < %s -S -passes=early-cse | FileCheck %s
declare void @llvm.assume(i1) nounwind
define void @test1(i8 %V, ptr%P) {
; CHECK-LABEL: @test1(
; CHECK-NEXT: store i32 23, ptr [[P:%.*]], align 4
; CHECK-NEXT: [[C:%.*]] = zext i8 [[V:%.*]] to i32
; CHECK-NEXT: store volatile i32 [[C]], ptr [[P]], align 4
; CHECK-NEXT: store volatile i32 [[C]], ptr [[P]], align 4
; CHECK-NEXT: [[E:%.*]] = add i32 [[C]], [[C]]
; CHECK-NEXT: store volatile i32 [[E]], ptr [[P]], align 4
; CHECK-NEXT: store volatile i32 [[E]], ptr [[P]], align 4
; CHECK-NEXT: store volatile i32 [[E]], ptr [[P]], align 4
; CHECK-NEXT: ret void
;
%A = bitcast i64 42 to double ;; dead
%B = add i32 4, 19 ;; constant folds
store i32 %B, ptr %P
%C = zext i8 %V to i32
%D = zext i8 %V to i32 ;; CSE
store volatile i32 %C, ptr %P
store volatile i32 %D, ptr %P
%E = add i32 %C, %C
%F = add i32 %C, %C
store volatile i32 %E, ptr %P
store volatile i32 %F, ptr %P
%G = add nuw i32 %C, %C
store volatile i32 %G, ptr %P
ret void
}
;; Simple load value numbering.
define i32 @test2(ptr%P) {
; CHECK-LABEL: @test2(
; CHECK-NEXT: [[V1:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: ret i32 0
;
%V1 = load i32, ptr %P
%V2 = load i32, ptr %P
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
define i32 @test2a(ptr%P, i1 %b) {
; CHECK-LABEL: @test2a(
; CHECK-NEXT: [[V1:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: tail call void @llvm.assume(i1 [[B:%.*]])
; CHECK-NEXT: ret i32 0
;
%V1 = load i32, ptr %P
tail call void @llvm.assume(i1 %b)
%V2 = load i32, ptr %P
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
;; Cross block load value numbering.
define i32 @test3(ptr%P, i1 %Cond) {
; CHECK-LABEL: @test3(
; CHECK-NEXT: [[V1:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: br i1 [[COND:%.*]], label [[T:%.*]], label [[F:%.*]]
; CHECK: T:
; CHECK-NEXT: store i32 4, ptr [[P]], align 4
; CHECK-NEXT: ret i32 42
; CHECK: F:
; CHECK-NEXT: ret i32 0
;
%V1 = load i32, ptr %P
br i1 %Cond, label %T, label %F
T:
store i32 4, ptr %P
ret i32 42
F:
%V2 = load i32, ptr %P
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
define i32 @test3a(ptr%P, i1 %Cond, i1 %b) {
; CHECK-LABEL: @test3a(
; CHECK-NEXT: [[V1:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: br i1 [[COND:%.*]], label [[T:%.*]], label [[F:%.*]]
; CHECK: T:
; CHECK-NEXT: store i32 4, ptr [[P]], align 4
; CHECK-NEXT: ret i32 42
; CHECK: F:
; CHECK-NEXT: tail call void @llvm.assume(i1 [[B:%.*]])
; CHECK-NEXT: ret i32 0
;
%V1 = load i32, ptr %P
br i1 %Cond, label %T, label %F
T:
store i32 4, ptr %P
ret i32 42
F:
tail call void @llvm.assume(i1 %b)
%V2 = load i32, ptr %P
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
;; Cross block load value numbering stops when stores happen.
define i32 @test4(ptr%P, i1 %Cond) {
; CHECK-LABEL: @test4(
; CHECK-NEXT: [[V1:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: br i1 [[COND:%.*]], label [[T:%.*]], label [[F:%.*]]
; CHECK: T:
; CHECK-NEXT: ret i32 42
; CHECK: F:
; CHECK-NEXT: store i32 42, ptr [[P]], align 4
; CHECK-NEXT: [[DIFF:%.*]] = sub i32 [[V1]], 42
; CHECK-NEXT: ret i32 [[DIFF]]
;
%V1 = load i32, ptr %P
br i1 %Cond, label %T, label %F
T:
ret i32 42
F:
; Clobbers V1
store i32 42, ptr %P
%V2 = load i32, ptr %P
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
declare i32 @func(ptr%P) readonly
;; Simple call CSE'ing.
define i32 @test5(ptr%P) {
; CHECK-LABEL: @test5(
; CHECK-NEXT: [[V1:%.*]] = call i32 @func(ptr [[P:%.*]]), !prof !0
; CHECK-NEXT: ret i32 0
;
%V1 = call i32 @func(ptr %P), !prof !0
%V2 = call i32 @func(ptr %P), !prof !1
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
!0 = !{!"branch_weights", i32 95}
!1 = !{!"branch_weights", i32 95}
;; Trivial Store->load forwarding
define i32 @test6(ptr%P) {
; CHECK-LABEL: @test6(
; CHECK-NEXT: store i32 42, ptr [[P:%.*]], align 4
; CHECK-NEXT: ret i32 42
;
store i32 42, ptr %P
%V1 = load i32, ptr %P
ret i32 %V1
}
define i32 @test6a(ptr%P, i1 %b) {
; CHECK-LABEL: @test6a(
; CHECK-NEXT: store i32 42, ptr [[P:%.*]], align 4
; CHECK-NEXT: tail call void @llvm.assume(i1 [[B:%.*]])
; CHECK-NEXT: ret i32 42
;
store i32 42, ptr %P
tail call void @llvm.assume(i1 %b)
%V1 = load i32, ptr %P
ret i32 %V1
}
;; Trivial dead store elimination.
define void @test7(ptr%P) {
; CHECK-LABEL: @test7(
; CHECK-NEXT: store i32 45, ptr [[P:%.*]], align 4
; CHECK-NEXT: ret void
;
store i32 42, ptr %P
store i32 45, ptr %P
ret void
}
;; Readnone functions aren't invalidated by stores.
define i32 @test8(ptr%P) {
; CHECK-LABEL: @test8(
; CHECK-NEXT: [[V1:%.*]] = call i32 @func(ptr [[P:%.*]]) #[[ATTR2:[0-9]+]]
; CHECK-NEXT: store i32 4, ptr [[P]], align 4
; CHECK-NEXT: ret i32 0
;
%V1 = call i32 @func(ptr %P) readnone
store i32 4, ptr %P
%V2 = call i32 @func(ptr %P) readnone
%Diff = sub i32 %V1, %V2
ret i32 %Diff
}
;; Trivial DSE can't be performed across a readonly call. The call
;; can observe the earlier write.
define i32 @test9(ptr%P) {
; CHECK-LABEL: @test9(
; CHECK-NEXT: store i32 4, ptr [[P:%.*]], align 4
; CHECK-NEXT: [[V1:%.*]] = call i32 @func(ptr [[P]]) #[[ATTR1:[0-9]+]]
; CHECK-NEXT: store i32 5, ptr [[P]], align 4
; CHECK-NEXT: ret i32 [[V1]]
;
store i32 4, ptr %P
%V1 = call i32 @func(ptr %P) readonly
store i32 5, ptr %P
ret i32 %V1
}
;; Trivial DSE can be performed across a readnone call.
define i32 @test10(ptr%P) {
; CHECK-LABEL: @test10(
; CHECK-NEXT: [[V1:%.*]] = call i32 @func(ptr [[P:%.*]]) #[[ATTR2]]
; CHECK-NEXT: store i32 5, ptr [[P]], align 4
; CHECK-NEXT: ret i32 [[V1]]
;
store i32 4, ptr %P
%V1 = call i32 @func(ptr %P) readnone
store i32 5, ptr %P
ret i32 %V1
}
;; Trivial dead store elimination - should work for an entire series of dead stores too.
define void @test11(ptr%P) {
; CHECK-LABEL: @test11(
; CHECK-NEXT: store i32 45, ptr [[P:%.*]], align 4
; CHECK-NEXT: ret void
;
store i32 42, ptr %P
store i32 43, ptr %P
store i32 44, ptr %P
store i32 45, ptr %P
ret void
}
define i32 @test12(i1 %B, ptr %P1, ptr %P2) {
; CHECK-LABEL: @test12(
; CHECK-NEXT: [[LOAD0:%.*]] = load i32, ptr [[P1:%.*]], align 4
; CHECK-NEXT: [[TMP1:%.*]] = load atomic i32, ptr [[P2:%.*]] seq_cst, align 4
; CHECK-NEXT: [[LOAD1:%.*]] = load i32, ptr [[P1]], align 4
; CHECK-NEXT: [[SEL:%.*]] = select i1 [[B:%.*]], i32 [[LOAD0]], i32 [[LOAD1]]
; CHECK-NEXT: ret i32 [[SEL]]
;
%load0 = load i32, ptr %P1
%1 = load atomic i32, ptr %P2 seq_cst, align 4
%load1 = load i32, ptr %P1
%sel = select i1 %B, i32 %load0, i32 %load1
ret i32 %sel
}
define void @dse1(ptr%P) {
; CHECK-LABEL: @dse1(
; CHECK-NEXT: [[V:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: ret void
;
%v = load i32, ptr %P
store i32 %v, ptr %P
ret void
}
define void @dse2(ptr%P) {
; CHECK-LABEL: @dse2(
; CHECK-NEXT: [[V:%.*]] = load atomic i32, ptr [[P:%.*]] seq_cst, align 4
; CHECK-NEXT: ret void
;
%v = load atomic i32, ptr %P seq_cst, align 4
store i32 %v, ptr %P
ret void
}
define void @dse3(ptr%P) {
; CHECK-LABEL: @dse3(
; CHECK-NEXT: [[V:%.*]] = load atomic i32, ptr [[P:%.*]] seq_cst, align 4
; CHECK-NEXT: ret void
;
%v = load atomic i32, ptr %P seq_cst, align 4
store atomic i32 %v, ptr %P unordered, align 4
ret void
}
define i32 @dse4(ptr%P, ptr%Q) {
; CHECK-LABEL: @dse4(
; CHECK-NEXT: [[A:%.*]] = load i32, ptr [[Q:%.*]], align 4
; CHECK-NEXT: [[V:%.*]] = load atomic i32, ptr [[P:%.*]] unordered, align 4
; CHECK-NEXT: ret i32 0
;
%a = load i32, ptr %Q
%v = load atomic i32, ptr %P unordered, align 4
store atomic i32 %v, ptr %P unordered, align 4
%b = load i32, ptr %Q
%res = sub i32 %a, %b
ret i32 %res
}
; Note that in this example, %P and %Q could in fact be the same
; pointer. %v could be different than the value observed for %a
; and that's okay because we're using relaxed memory ordering.
; The only guarantee we have to provide is that each of the loads
; has to observe some value written to that location. We do
; not have to respect the order in which those writes were done.
define i32 @dse5(ptr%P, ptr%Q) {
; CHECK-LABEL: @dse5(
; CHECK-NEXT: [[V:%.*]] = load atomic i32, ptr [[P:%.*]] unordered, align 4
; CHECK-NEXT: [[A:%.*]] = load atomic i32, ptr [[Q:%.*]] unordered, align 4
; CHECK-NEXT: ret i32 0
;
%v = load atomic i32, ptr %P unordered, align 4
%a = load atomic i32, ptr %Q unordered, align 4
store atomic i32 %v, ptr %P unordered, align 4
%b = load atomic i32, ptr %Q unordered, align 4
%res = sub i32 %a, %b
ret i32 %res
}
define void @dse_neg1(ptr%P) {
; CHECK-LABEL: @dse_neg1(
; CHECK-NEXT: store i32 5, ptr [[P:%.*]], align 4
; CHECK-NEXT: ret void
;
%v = load i32, ptr %P
store i32 5, ptr %P
ret void
}
; Could remove the store, but only if ordering was somehow
; encoded.
define void @dse_neg2(ptr%P) {
; CHECK-LABEL: @dse_neg2(
; CHECK-NEXT: [[V:%.*]] = load i32, ptr [[P:%.*]], align 4
; CHECK-NEXT: store atomic i32 [[V]], ptr [[P]] seq_cst, align 4
; CHECK-NEXT: ret void
;
%v = load i32, ptr %P
store atomic i32 %v, ptr %P seq_cst, align 4
ret void
}
@c = external global i32, align 4
declare i32 @reads_c(i32 returned)
define void @pr28763() {
; CHECK-LABEL: @pr28763(
; CHECK-NEXT: entry:
; CHECK-NEXT: store i32 0, ptr @c, align 4
; CHECK-NEXT: [[CALL:%.*]] = call i32 @reads_c(i32 0)
; CHECK-NEXT: store i32 2, ptr @c, align 4
; CHECK-NEXT: ret void
;
entry:
%load = load i32, ptr @c, align 4
store i32 0, ptr @c, align 4
%call = call i32 @reads_c(i32 0)
store i32 2, ptr @c, align 4
ret void
}
define i1 @cse_freeze(i1 %a) {
; CHECK-LABEL: @cse_freeze(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[B:%.*]] = freeze i1 [[A:%.*]]
; CHECK-NEXT: ret i1 [[B]]
;
entry:
%b = freeze i1 %a
%c = freeze i1 %a
%and = and i1 %b, %c
ret i1 %and
}
|
