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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
; RUN: opt < %s -force-vector-width=2 -force-vector-interleave=2 -passes=loop-vectorize -S | FileCheck %s
target datalayout = "e-m:e-i64:64-i128:128-n32:64-S128"
;
; This test ensures that we don't generate trivially dead instructions prior to
; instruction simplification. We don't need to generate instructions
; corresponding to the original induction variable update or branch condition,
; since we rewrite the loop structure.
;
define i64 @dead_instructions_01(ptr %a, i64 %n) {
; CHECK-LABEL: define i64 @dead_instructions_01(
; CHECK-SAME: ptr [[A:%.*]], i64 [[N:%.*]]) {
; CHECK-NEXT: [[ENTRY:.*]]:
; CHECK-NEXT: [[SMAX:%.*]] = call i64 @llvm.smax.i64(i64 [[N]], i64 1)
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[SMAX]], 4
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; CHECK: [[VECTOR_PH]]:
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[SMAX]], 4
; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[SMAX]], [[N_MOD_VF]]
; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
; CHECK: [[VECTOR_BODY]]:
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP6:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT: [[VEC_PHI1:%.*]] = phi <2 x i64> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP7:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = add i64 [[INDEX]], 2
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[TMP0]]
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[TMP1]]
; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds i64, ptr [[TMP2]], i32 0
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds i64, ptr [[TMP2]], i32 2
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i64>, ptr [[TMP4]], align 8
; CHECK-NEXT: [[WIDE_LOAD2:%.*]] = load <2 x i64>, ptr [[TMP5]], align 8
; CHECK-NEXT: [[TMP6]] = add <2 x i64> [[WIDE_LOAD]], [[VEC_PHI]]
; CHECK-NEXT: [[TMP7]] = add <2 x i64> [[WIDE_LOAD2]], [[VEC_PHI1]]
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP8]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: [[MIDDLE_BLOCK]]:
; CHECK-NEXT: [[BIN_RDX:%.*]] = add <2 x i64> [[TMP7]], [[TMP6]]
; CHECK-NEXT: [[TMP9:%.*]] = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> [[BIN_RDX]])
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[SMAX]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label %[[FOR_END:.*]], label %[[SCALAR_PH]]
; CHECK: [[SCALAR_PH]]:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i64 [ [[TMP9]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; CHECK-NEXT: br label %[[FOR_BODY:.*]]
; CHECK: [[FOR_BODY]]:
; CHECK-NEXT: [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], %[[FOR_BODY]] ], [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ]
; CHECK-NEXT: [[R:%.*]] = phi i64 [ [[TMP2:%.*]], %[[FOR_BODY]] ], [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ]
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[I]]
; CHECK-NEXT: [[TMP1:%.*]] = load i64, ptr [[TMP0]], align 8
; CHECK-NEXT: [[TMP2]] = add i64 [[TMP1]], [[R]]
; CHECK-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1
; CHECK-NEXT: [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label %[[FOR_BODY]], label %[[FOR_END]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK: [[FOR_END]]:
; CHECK-NEXT: [[TMP3:%.*]] = phi i64 [ [[TMP2]], %[[FOR_BODY]] ], [ [[TMP9]], %[[MIDDLE_BLOCK]] ]
; CHECK-NEXT: ret i64 [[TMP3]]
;
entry:
br label %for.body
for.body:
%i = phi i64 [ %i.next, %for.body ], [ 0, %entry ]
%r = phi i64 [ %tmp2, %for.body ], [ 0, %entry ]
%tmp0 = getelementptr inbounds i64, ptr %a, i64 %i
%tmp1 = load i64, ptr %tmp0, align 8
%tmp2 = add i64 %tmp1, %r
%i.next = add nuw nsw i64 %i, 1
%cond = icmp slt i64 %i.next, %n
br i1 %cond, label %for.body, label %for.end
for.end:
%tmp3 = phi i64 [ %tmp2, %for.body ]
ret i64 %tmp3
}
;
; This test ensures that the primary induction is not considered dead when
; acting as the 'add' of another induction, and otherwise feeding only its own
; 'add' (recognized earlier as 'dead'), when the tail of the loop is folded by
; masking. Such masking uses the primary induction.
;
define void @pr47390(ptr %a) {
; CHECK-LABEL: define void @pr47390(
; CHECK-SAME: ptr [[A:%.*]]) {
; CHECK-NEXT: [[ENTRY:.*]]:
; CHECK-NEXT: br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; CHECK: [[VECTOR_PH]]:
; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
; CHECK: [[VECTOR_BODY]]:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4
; CHECK-NEXT: [[TMP0:%.*]] = icmp eq i32 [[INDEX_NEXT]], 8
; CHECK-NEXT: br i1 [[TMP0]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: [[MIDDLE_BLOCK]]:
; CHECK-NEXT: br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; CHECK: [[SCALAR_PH]]:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 8, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i32 [ 7, %[[MIDDLE_BLOCK]] ], [ -1, %[[ENTRY]] ]
; CHECK-NEXT: [[BC_RESUME_VAL2:%.*]] = phi i32 [ 9, %[[MIDDLE_BLOCK]] ], [ 1, %[[ENTRY]] ]
; CHECK-NEXT: br label %[[LOOP:.*]]
; CHECK: [[EXIT]]:
; CHECK-NEXT: ret void
; CHECK: [[LOOP]]:
; CHECK-NEXT: [[PRIMARY:%.*]] = phi i32 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[PRIMARY_ADD:%.*]], %[[LOOP]] ]
; CHECK-NEXT: [[USE_PRIMARY:%.*]] = phi i32 [ [[BC_RESUME_VAL1]], %[[SCALAR_PH]] ], [ [[PRIMARY]], %[[LOOP]] ]
; CHECK-NEXT: [[SECONDARY:%.*]] = phi i32 [ [[BC_RESUME_VAL2]], %[[SCALAR_PH]] ], [ [[SECONDARY_ADD:%.*]], %[[LOOP]] ]
; CHECK-NEXT: [[PRIMARY_ADD]] = add i32 [[PRIMARY]], 1
; CHECK-NEXT: [[SECONDARY_ADD]] = add i32 [[SECONDARY]], 1
; CHECK-NEXT: [[GEP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[SECONDARY]]
; CHECK-NEXT: [[LOAD:%.*]] = load i32, ptr [[GEP]], align 8
; CHECK-NEXT: [[CMP:%.*]] = icmp eq i32 [[SECONDARY]], 5
; CHECK-NEXT: br i1 [[CMP]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP5:![0-9]+]]
;
entry:
br label %loop
exit:
ret void
loop:
%primary = phi i32 [ 0, %entry ], [ %primary_add, %loop ]
%use_primary = phi i32 [ -1, %entry ], [ %primary, %loop ]
%secondary = phi i32 [ 1, %entry ], [ %secondary_add, %loop ]
%primary_add = add i32 %primary, 1
%secondary_add = add i32 %secondary, 1
%gep = getelementptr inbounds i32, ptr %a, i32 %secondary
%load = load i32, ptr %gep, align 8
%cmp = icmp eq i32 %secondary, 5
br i1 %cmp, label %exit, label %loop
}
; Test with a dead load and dead vector poiner.
define void @dead_load_and_vector_pointer(ptr %a, ptr %b) {
; CHECK-LABEL: define void @dead_load_and_vector_pointer(
; CHECK-SAME: ptr [[A:%.*]], ptr [[B:%.*]]) {
; CHECK-NEXT: [[ENTRY:.*]]:
; CHECK-NEXT: br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_MEMCHECK:.*]]
; CHECK: [[VECTOR_MEMCHECK]]:
; CHECK-NEXT: [[SCEVGEP:%.*]] = getelementptr i8, ptr [[A]], i64 516
; CHECK-NEXT: [[SCEVGEP1:%.*]] = getelementptr i8, ptr [[B]], i64 516
; CHECK-NEXT: [[BOUND0:%.*]] = icmp ult ptr [[A]], [[SCEVGEP1]]
; CHECK-NEXT: [[BOUND1:%.*]] = icmp ult ptr [[B]], [[SCEVGEP]]
; CHECK-NEXT: [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; CHECK-NEXT: br i1 [[FOUND_CONFLICT]], label %[[SCALAR_PH]], label %[[VECTOR_PH:.*]]
; CHECK: [[VECTOR_PH]]:
; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
; CHECK: [[VECTOR_BODY]]:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[INDEX]], 0
; CHECK-NEXT: [[TMP1:%.*]] = add i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP0]]
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP1]]
; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 0
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr [[TMP2]], i32 2
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP4]], align 8, !alias.scope [[META6:![0-9]+]], !noalias [[META9:![0-9]+]]
; CHECK-NEXT: [[WIDE_LOAD2:%.*]] = load <2 x i32>, ptr [[TMP5]], align 8, !alias.scope [[META6]], !noalias [[META9]]
; CHECK-NEXT: [[TMP6:%.*]] = add <2 x i32> [[WIDE_LOAD]], <i32 1, i32 1>
; CHECK-NEXT: [[TMP7:%.*]] = add <2 x i32> [[WIDE_LOAD2]], <i32 1, i32 1>
; CHECK-NEXT: store <2 x i32> [[TMP6]], ptr [[TMP4]], align 4, !alias.scope [[META6]], !noalias [[META9]]
; CHECK-NEXT: store <2 x i32> [[TMP7]], ptr [[TMP5]], align 4, !alias.scope [[META6]], !noalias [[META9]]
; CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[TMP0]]
; CHECK-NEXT: [[TMP9:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[TMP1]]
; CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds i32, ptr [[TMP8]], i32 0
; CHECK-NEXT: [[TMP11:%.*]] = getelementptr inbounds i32, ptr [[TMP8]], i32 2
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i32 [[INDEX_NEXT]], 128
; CHECK-NEXT: br i1 [[TMP12]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP11:![0-9]+]]
; CHECK: [[MIDDLE_BLOCK]]:
; CHECK-NEXT: br i1 false, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; CHECK: [[SCALAR_PH]]:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ 128, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ], [ 0, %[[VECTOR_MEMCHECK]] ]
; CHECK-NEXT: br label %[[LOOP:.*]]
; CHECK: [[EXIT]]:
; CHECK-NEXT: ret void
; CHECK: [[LOOP]]:
; CHECK-NEXT: [[PRIMARY:%.*]] = phi i32 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[PRIMARY_ADD:%.*]], %[[LOOP]] ]
; CHECK-NEXT: [[PRIMARY_ADD]] = add i32 [[PRIMARY]], 1
; CHECK-NEXT: [[GEP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[PRIMARY]]
; CHECK-NEXT: [[LOAD:%.*]] = load i32, ptr [[GEP]], align 8
; CHECK-NEXT: [[ADD:%.*]] = add i32 [[LOAD]], 1
; CHECK-NEXT: store i32 [[ADD]], ptr [[GEP]], align 4
; CHECK-NEXT: [[GEP_B:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[PRIMARY]]
; CHECK-NEXT: [[LOAD2:%.*]] = load i32, ptr [[GEP_B]], align 4
; CHECK-NEXT: [[CMP:%.*]] = icmp eq i32 [[PRIMARY]], 128
; CHECK-NEXT: br i1 [[CMP]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP12:![0-9]+]]
;
entry:
br label %loop
exit:
ret void
loop:
%primary = phi i32 [ 0, %entry ], [ %primary_add, %loop ]
%primary_add = add i32 %primary, 1
%gep = getelementptr inbounds i32, ptr %a, i32 %primary
%load = load i32, ptr %gep, align 8
%add = add i32 %load, 1
store i32 %add, ptr %gep
%gep.b = getelementptr inbounds i32, ptr %b, i32 %primary
%load2 = load i32, ptr %gep.b
%cmp = icmp eq i32 %primary, 128
br i1 %cmp, label %exit, label %loop
}
;.
; CHECK: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]], [[META2:![0-9]+]]}
; CHECK: [[META1]] = !{!"llvm.loop.isvectorized", i32 1}
; CHECK: [[META2]] = !{!"llvm.loop.unroll.runtime.disable"}
; CHECK: [[LOOP3]] = distinct !{[[LOOP3]], [[META2]], [[META1]]}
; CHECK: [[LOOP4]] = distinct !{[[LOOP4]], [[META1]], [[META2]]}
; CHECK: [[LOOP5]] = distinct !{[[LOOP5]], [[META2]], [[META1]]}
; CHECK: [[META6]] = !{[[META7:![0-9]+]]}
; CHECK: [[META7]] = distinct !{[[META7]], [[META8:![0-9]+]]}
; CHECK: [[META8]] = distinct !{[[META8]], !"LVerDomain"}
; CHECK: [[META9]] = !{[[META10:![0-9]+]]}
; CHECK: [[META10]] = distinct !{[[META10]], [[META8]]}
; CHECK: [[LOOP11]] = distinct !{[[LOOP11]], [[META1]], [[META2]]}
; CHECK: [[LOOP12]] = distinct !{[[LOOP12]], [[META1]]}
;.
|
