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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; REQUIRES: asserts
; RUN: opt -passes=loop-vectorize -S -mtriple=aarch64 -mattr=+sve -debug-only=loop-vectorize \
; RUN: -prefer-predicate-over-epilogue=scalar-epilogue < %s 2>&1 | FileCheck %s
target triple = "aarch64-unknown-linux-gnu"
; CHECK-LABEL: LV: Checking a loop in 'pointer_induction_used_as_vector'
; CHECK-NOT: LV: Found {{.*}} scalar instruction: %ptr.iv.2.next = getelementptr inbounds i8, ptr %ptr.iv.2, i64 1
;
; CHECK: VPlan 'Initial VPlan for VF={vscale x 2},UF>=1' {
; CHECK-NEXT: Live-in vp<[[VFxUF:%.+]]> = VF * UF
; CHECK-NEXT: Live-in vp<[[VEC_TC:%.+]]> = vector-trip-count
; CHECK-NEXT: Live-in ir<%N> = original trip-count
; CHECK-EMPTY:
; CHECK-NEXT: ir-bb<entry>:
; CHECK-NEXT: Successor(s): vector.ph
; CHECK-EMPTY:
; CHECK-NEXT: vector.ph:
; CHECK-NEXT: vp<[[END1:%.+]]> = DERIVED-IV ir<%start.1> + vp<[[VEC_TC]]> * ir<8>
; CHECK-NEXT: vp<[[END2:%.+]]> = DERIVED-IV ir<%start.2> + vp<[[VEC_TC]]> * ir<1>
; CHECK-NEXT: Successor(s): vector loop
; CHECK-EMPTY:
; CHECK-NEXT: <x1> vector loop: {
; CHECK-NEXT: vector.body:
; CHECK-NEXT: EMIT vp<[[CAN_IV:%.+]]> = CANONICAL-INDUCTION
; CHECK-NEXT: EMIT ir<%ptr.iv.2> = WIDEN-POINTER-INDUCTION ir<%start.2>, ir<1>
; CHECK-NEXT: vp<[[PTR_IDX:%.+]]> = DERIVED-IV ir<0> + vp<[[CAN_IV]]> * ir<8>
; CHECK-NEXT: vp<[[PTR_IDX_STEPS:%.+]]> = SCALAR-STEPS vp<[[PTR_IDX]]>, ir<8>
; CHECK-NEXT: EMIT vp<[[PTR_IV_1:%.+]]> = ptradd ir<%start.1>, vp<[[PTR_IDX_STEPS]]>
; CHECK-NEXT: WIDEN-GEP Var[Inv] ir<%ptr.iv.2.next> = getelementptr inbounds ir<%ptr.iv.2>, ir<1>
; CHECK-NEXT: vp<[[VEC_PTR:%.+]]> = vector-pointer vp<[[PTR_IV_1]]>
; CHECK-NEXT: WIDEN store vp<[[VEC_PTR]]>, ir<%ptr.iv.2.next>
; CHECK-NEXT: vp<[[VEC_PTR2:%.+]]> = vector-pointer ir<%ptr.iv.2>
; CHECK-NEXT: WIDEN ir<%lv> = load vp<[[VEC_PTR2]]>
; CHECK-NEXT: WIDEN ir<%add> = add ir<%lv>, ir<1>
; CHECK-NEXT: vp<[[VEC_PTR3:%.+]]> = vector-pointer ir<%ptr.iv.2>
; CHECK-NEXT: WIDEN store vp<[[VEC_PTR3]]>, ir<%add>
; CHECK-NEXT: EMIT vp<[[CAN_IV_NEXT:%.+]]> = add nuw vp<[[CAN_IV]]>, vp<[[VFxUF]]>
; CHECK-NEXT: EMIT branch-on-count vp<[[CAN_IV_NEXT]]>, vp<[[VEC_TC]]>
; CHECK-NEXT: No successors
; CHECK-NEXT: }
; In the test below the pointer phi %ptr.iv.2 is used as
; 1. As a uniform address for the load, and
; 2. Non-uniform use by the getelementptr which is stored. This requires the
; vector value.
define void @pointer_induction_used_as_vector(ptr noalias %start.1, ptr noalias %start.2, i64 %N) {
; CHECK-LABEL: @pointer_induction_used_as_vector(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP1:%.*]] = mul i64 [[TMP0]], 2
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[N:%.*]], [[TMP1]]
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP3:%.*]] = mul i64 [[TMP2]], 2
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N]], [[TMP3]]
; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[N]], [[N_MOD_VF]]
; CHECK-NEXT: [[TMP5:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP6:%.*]] = mul i64 [[TMP5]], 2
; CHECK-NEXT: [[TMP4:%.*]] = mul i64 [[N_VEC]], 8
; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[START_1:%.*]], i64 [[TMP4]]
; CHECK-NEXT: [[IND_END2:%.*]] = getelementptr i8, ptr [[START_2:%.*]], i64 [[N_VEC]]
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[POINTER_PHI:%.*]] = phi ptr [ [[START_2]], [[VECTOR_PH]] ], [ [[PTR_IND:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP7:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP8:%.*]] = mul i64 [[TMP7]], 2
; CHECK-NEXT: [[TMP9:%.*]] = mul i64 [[TMP8]], 1
; CHECK-NEXT: [[TMP10:%.*]] = mul i64 1, [[TMP9]]
; CHECK-NEXT: [[TMP11:%.*]] = mul i64 [[TMP8]], 0
; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TMP11]], i64 0
; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
; CHECK-NEXT: [[TMP12:%.*]] = call <vscale x 2 x i64> @llvm.stepvector.nxv2i64()
; CHECK-NEXT: [[TMP13:%.*]] = add <vscale x 2 x i64> [[DOTSPLAT]], [[TMP12]]
; CHECK-NEXT: [[TMP14:%.*]] = mul <vscale x 2 x i64> [[TMP13]], splat (i64 1)
; CHECK-NEXT: [[VECTOR_GEP:%.*]] = getelementptr i8, ptr [[POINTER_PHI]], <vscale x 2 x i64> [[TMP14]]
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = mul i64 [[INDEX]], 8
; CHECK-NEXT: [[TMP15:%.*]] = add i64 [[OFFSET_IDX]], 0
; CHECK-NEXT: [[NEXT_GEP:%.*]] = getelementptr i8, ptr [[START_1]], i64 [[TMP15]]
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds i8, <vscale x 2 x ptr> [[VECTOR_GEP]], i64 1
; CHECK-NEXT: [[TMP17:%.*]] = getelementptr ptr, ptr [[NEXT_GEP]], i32 0
; CHECK-NEXT: store <vscale x 2 x ptr> [[TMP16]], ptr [[TMP17]], align 8
; CHECK-NEXT: [[TMP18:%.*]] = extractelement <vscale x 2 x ptr> [[VECTOR_GEP]], i32 0
; CHECK-NEXT: [[TMP19:%.*]] = getelementptr i8, ptr [[TMP18]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 2 x i8>, ptr [[TMP19]], align 1
; CHECK-NEXT: [[TMP20:%.*]] = add <vscale x 2 x i8> [[WIDE_LOAD]], splat (i8 1)
; CHECK-NEXT: store <vscale x 2 x i8> [[TMP20]], ptr [[TMP19]], align 1
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP6]]
; CHECK-NEXT: [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 [[TMP10]]
; CHECK-NEXT: [[TMP21:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP21]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[N]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[START_1]], [[ENTRY]] ]
; CHECK-NEXT: [[BC_RESUME_VAL3:%.*]] = phi ptr [ [[IND_END2]], [[MIDDLE_BLOCK]] ], [ [[START_2]], [[ENTRY]] ]
; CHECK-NEXT: br label [[LOOP_BODY:%.*]]
; CHECK: loop.body:
; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP_BODY]] ]
; CHECK-NEXT: [[PTR_IV_1:%.*]] = phi ptr [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ], [ [[PTR_IV_1_NEXT:%.*]], [[LOOP_BODY]] ]
; CHECK-NEXT: [[PTR_IV_2:%.*]] = phi ptr [ [[BC_RESUME_VAL3]], [[SCALAR_PH]] ], [ [[PTR_IV_2_NEXT:%.*]], [[LOOP_BODY]] ]
; CHECK-NEXT: [[PTR_IV_1_NEXT]] = getelementptr inbounds ptr, ptr [[PTR_IV_1]], i64 1
; CHECK-NEXT: [[PTR_IV_2_NEXT]] = getelementptr inbounds i8, ptr [[PTR_IV_2]], i64 1
; CHECK-NEXT: store ptr [[PTR_IV_2_NEXT]], ptr [[PTR_IV_1]], align 8
; CHECK-NEXT: [[LV:%.*]] = load i8, ptr [[PTR_IV_2]], align 1
; CHECK-NEXT: [[ADD:%.*]] = add i8 [[LV]], 1
; CHECK-NEXT: store i8 [[ADD]], ptr [[PTR_IV_2]], align 1
; CHECK-NEXT: [[IV_NEXT]] = add nuw i64 [[IV]], 1
; CHECK-NEXT: [[C:%.*]] = icmp ne i64 [[IV_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[C]], label [[LOOP_BODY]], label [[EXIT]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK: exit:
; CHECK-NEXT: ret void
;
entry:
br label %loop.body
loop.body: ; preds = %loop.body, %entry
%iv = phi i64 [ 0, %entry ], [ %iv.next, %loop.body ]
%ptr.iv.1 = phi ptr [ %start.1, %entry ], [ %ptr.iv.1.next, %loop.body ]
%ptr.iv.2 = phi ptr [ %start.2, %entry ], [ %ptr.iv.2.next, %loop.body ]
%ptr.iv.1.next = getelementptr inbounds ptr, ptr %ptr.iv.1, i64 1
%ptr.iv.2.next = getelementptr inbounds i8, ptr %ptr.iv.2, i64 1
store ptr %ptr.iv.2.next, ptr %ptr.iv.1, align 8
%lv = load i8, ptr %ptr.iv.2, align 1
%add = add i8 %lv, 1
store i8 %add, ptr %ptr.iv.2, align 1
%iv.next = add nuw i64 %iv, 1
%c = icmp ne i64 %iv.next, %N
br i1 %c, label %loop.body, label %exit, !llvm.loop !0
exit: ; preds = %loop.body
ret void
}
define void @pointer_induction(ptr noalias %start, i64 %N) {
; CHECK-LABEL: @pointer_induction(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[N:%.*]], 1
; CHECK-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP2:%.*]] = mul i64 [[TMP1]], 2
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], [[TMP2]]
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP4:%.*]] = mul i64 [[TMP3]], 2
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[TMP0]], [[TMP4]]
; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[TMP0]], [[N_MOD_VF]]
; CHECK-NEXT: [[TMP5:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP6:%.*]] = mul i64 [[TMP5]], 2
; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[START:%.*]], i64 [[N_VEC]]
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[POINTER_PHI:%.*]] = phi ptr [ [[START]], [[VECTOR_PH]] ], [ [[PTR_IND:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[INDEX2:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP7:%.*]] = call i64 @llvm.vscale.i64()
; CHECK-NEXT: [[TMP8:%.*]] = mul i64 [[TMP7]], 2
; CHECK-NEXT: [[TMP9:%.*]] = mul i64 [[TMP8]], 1
; CHECK-NEXT: [[TMP10:%.*]] = mul i64 1, [[TMP9]]
; CHECK-NEXT: [[TMP11:%.*]] = mul i64 [[TMP8]], 0
; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[TMP11]], i64 0
; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
; CHECK-NEXT: [[TMP12:%.*]] = call <vscale x 2 x i64> @llvm.stepvector.nxv2i64()
; CHECK-NEXT: [[TMP13:%.*]] = add <vscale x 2 x i64> [[DOTSPLAT]], [[TMP12]]
; CHECK-NEXT: [[TMP14:%.*]] = mul <vscale x 2 x i64> [[TMP13]], splat (i64 1)
; CHECK-NEXT: [[VECTOR_GEP:%.*]] = getelementptr i8, ptr [[POINTER_PHI]], <vscale x 2 x i64> [[TMP14]]
; CHECK-NEXT: [[TMP15:%.*]] = extractelement <vscale x 2 x ptr> [[VECTOR_GEP]], i32 0
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr i8, ptr [[TMP15]], i32 0
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 2 x i8>, ptr [[TMP16]], align 1
; CHECK-NEXT: [[TMP17:%.*]] = add <vscale x 2 x i8> [[WIDE_LOAD]], splat (i8 1)
; CHECK-NEXT: store <vscale x 2 x i8> [[TMP17]], ptr [[TMP16]], align 1
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX2]], [[TMP6]]
; CHECK-NEXT: [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 [[TMP10]]
; CHECK-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP0]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[END:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi ptr [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ [[START]], [[ENTRY:%.*]] ]
; CHECK-NEXT: [[BC_RESUME_VAL1:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ]
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
; CHECK: for.body:
; CHECK-NEXT: [[PTR_PHI:%.*]] = phi ptr [ [[PTR_PHI_NEXT:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ [[INDEX_NXT:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL1]], [[SCALAR_PH]] ]
; CHECK-NEXT: [[INDEX_NXT]] = add i64 [[INDEX]], 1
; CHECK-NEXT: [[TMP19:%.*]] = load i8, ptr [[PTR_PHI]], align 1
; CHECK-NEXT: [[ADD:%.*]] = add i8 [[TMP19]], 1
; CHECK-NEXT: store i8 [[ADD]], ptr [[PTR_PHI]], align 1
; CHECK-NEXT: [[PTR_PHI_NEXT]] = getelementptr inbounds i8, ptr [[PTR_PHI]], i64 1
; CHECK-NEXT: [[CMP_I_NOT:%.*]] = icmp eq ptr [[PTR_PHI_NEXT]], [[START]]
; CHECK-NEXT: [[CMP:%.*]] = icmp ult i64 [[INDEX]], [[N]]
; CHECK-NEXT: br i1 [[CMP]], label [[FOR_BODY]], label [[END]], !llvm.loop [[LOOP5:![0-9]+]]
; CHECK: end:
; CHECK-NEXT: ret void
;
entry:
br label %for.body
for.body:
%ptr.phi = phi ptr [ %ptr.phi.next, %for.body ], [ %start, %entry ]
%index = phi i64 [ %index_nxt, %for.body ], [ 0, %entry ]
%index_nxt = add i64 %index, 1
%0 = load i8, ptr %ptr.phi, align 1
%add = add i8 %0, 1
store i8 %add, ptr %ptr.phi
%ptr.phi.next = getelementptr inbounds i8, ptr %ptr.phi, i64 1
%cmp.i.not = icmp eq ptr %ptr.phi.next, %start
%cmp = icmp ult i64 %index, %N
br i1 %cmp, label %for.body, label %end, !llvm.loop !0
end:
ret void
}
attributes #0 = {"target-features"="+sve"}
!0 = distinct !{!0, !1, !2, !3}
!1 = !{!"llvm.loop.interleave.count", i32 1}
!2 = !{!"llvm.loop.vectorize.width", i32 2}
!3 = !{!"llvm.loop.vectorize.scalable.enable", i1 true}
!4 = !{ !5 }
!5 = distinct !{ !5, !6 }
!6 = distinct !{ !7 }
!7 = distinct !{ !7, !6 }
|
