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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -passes=instcombine -S | FileCheck %s
target datalayout = "E-p:64:64:64-p1:32:32:32-a0:0:8-f32:32:32-f64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-v64:64:64-v128:128:128"
; Instcombine should be able to prove vector alignment in the
; presence of a few mild address computation tricks.
define void @test0(ptr %b, i64 %n, i64 %u, i64 %y) nounwind {
; CHECK-LABEL: @test0(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[C:%.*]] = ptrtoint ptr [[B:%.*]] to i64
; CHECK-NEXT: [[D:%.*]] = and i64 [[C]], -16
; CHECK-NEXT: [[E:%.*]] = inttoptr i64 [[D]] to ptr
; CHECK-NEXT: [[V:%.*]] = shl i64 [[U:%.*]], 1
; CHECK-NEXT: [[Z:%.*]] = and i64 [[Y:%.*]], -2
; CHECK-NEXT: [[T1421:%.*]] = icmp eq i64 [[N:%.*]], 0
; CHECK-NEXT: br i1 [[T1421]], label [[RETURN:%.*]], label [[BB:%.*]]
; CHECK: bb:
; CHECK-NEXT: [[I:%.*]] = phi i64 [ [[INDVAR_NEXT:%.*]], [[BB]] ], [ 20, [[ENTRY:%.*]] ]
; CHECK-NEXT: [[J:%.*]] = mul i64 [[I]], [[V]]
; CHECK-NEXT: [[H:%.*]] = add i64 [[J]], [[Z]]
; CHECK-NEXT: [[T8:%.*]] = getelementptr double, ptr [[E]], i64 [[H]]
; CHECK-NEXT: store <2 x double> zeroinitializer, ptr [[T8]], align 16
; CHECK-NEXT: [[INDVAR_NEXT]] = add i64 [[I]], 1
; CHECK-NEXT: [[EXITCOND:%.*]] = icmp eq i64 [[INDVAR_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[EXITCOND]], label [[RETURN]], label [[BB]]
; CHECK: return:
; CHECK-NEXT: ret void
;
entry:
%c = ptrtoint ptr %b to i64
%d = and i64 %c, -16
%e = inttoptr i64 %d to ptr
%v = mul i64 %u, 2
%z = and i64 %y, -2
%t1421 = icmp eq i64 %n, 0
br i1 %t1421, label %return, label %bb
bb:
%i = phi i64 [ %indvar.next, %bb ], [ 20, %entry ]
%j = mul i64 %i, %v
%h = add i64 %j, %z
%t8 = getelementptr double, ptr %e, i64 %h
store <2 x double><double 0.0, double 0.0>, ptr %t8, align 8
%indvar.next = add i64 %i, 1
%exitcond = icmp eq i64 %indvar.next, %n
br i1 %exitcond, label %return, label %bb
return:
ret void
}
; When we see a unaligned load from an insufficiently aligned global or
; alloca, increase the alignment of the load, turning it into an aligned load.
@GLOBAL = internal global [4 x i32] zeroinitializer
define <16 x i8> @test1(<2 x i64> %x) {
; CHECK-LABEL: @test1(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP:%.*]] = load <16 x i8>, ptr @GLOBAL, align 16
; CHECK-NEXT: ret <16 x i8> [[TMP]]
;
entry:
%tmp = load <16 x i8>, ptr @GLOBAL, align 1
ret <16 x i8> %tmp
}
@GLOBAL_as1 = internal addrspace(1) global [4 x i32] zeroinitializer
define <16 x i8> @test1_as1(<2 x i64> %x) {
; CHECK-LABEL: @test1_as1(
; CHECK-NEXT: [[TMP:%.*]] = load <16 x i8>, ptr addrspace(1) @GLOBAL_as1, align 16
; CHECK-NEXT: ret <16 x i8> [[TMP]]
;
%tmp = load <16 x i8>, ptr addrspace(1) @GLOBAL_as1, align 1
ret <16 x i8> %tmp
}
@GLOBAL_as1_gep = internal addrspace(1) global [8 x i32] zeroinitializer
define <16 x i8> @test1_as1_gep(<2 x i64> %x) {
; CHECK-LABEL: @test1_as1_gep(
; CHECK-NEXT: [[TMP:%.*]] = load <16 x i8>, ptr addrspace(1) getelementptr inbounds ([8 x i32], ptr addrspace(1) @GLOBAL_as1_gep, i32 0, i32 4), align 16
; CHECK-NEXT: ret <16 x i8> [[TMP]]
;
%tmp = load <16 x i8>, ptr addrspace(1) getelementptr ([8 x i32], ptr addrspace(1) @GLOBAL_as1_gep, i16 0, i16 4), align 1
ret <16 x i8> %tmp
}
; When a load or store lacks an explicit alignment, add one.
define double @test2(ptr %p, double %n) nounwind {
; CHECK-LABEL: @test2(
; CHECK-NEXT: [[T:%.*]] = load double, ptr [[P:%.*]], align 8
; CHECK-NEXT: store double [[N:%.*]], ptr [[P]], align 8
; CHECK-NEXT: ret double [[T]]
;
%t = load double, ptr %p
store double %n, ptr %p
ret double %t
}
declare void @llvm.memset.p0.i64(ptr nocapture, i8, i64, i1) nounwind
declare void @use(ptr)
%struct.s = type { i32, i32, i32, i32 }
define void @test3(ptr sret(%struct.s) %a4) {
; Check that the alignment is bumped up the alignment of the sret type.
; CHECK-LABEL: @test3(
; CHECK-NEXT: call void @llvm.memset.p0.i64(ptr noundef nonnull align 4 dereferenceable(16) [[A4:%.*]], i8 0, i64 16, i1 false)
; CHECK-NEXT: call void @use(ptr [[A4]])
; CHECK-NEXT: ret void
;
call void @llvm.memset.p0.i64(ptr %a4, i8 0, i64 16, i1 false)
call void @use(ptr %a4)
ret void
}
declare ptr @llvm.ptrmask.p0.i64(ptr, i64)
declare ptr @llvm.ptrmask.p0.i32(ptr, i32)
declare ptr @llvm.ptrmask.p0.i128(ptr, i128)
define <16 x i8> @ptrmask_align_unknown_ptr_align1(ptr align 1 %ptr, i64 %mask) {
; CHECK-LABEL: @ptrmask_align_unknown_ptr_align1(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 [[MASK:%.*]])
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 1
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 %mask)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
define <16 x i8> @ptrmask_align_unknown_ptr_align8(ptr align 8 %ptr, i64 %mask) {
; CHECK-LABEL: @ptrmask_align_unknown_ptr_align8(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 [[MASK:%.*]])
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 8
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 %mask)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Increase load align from 1 to 2
define <16 x i8> @ptrmask_align2_ptr_align1(ptr align 1 %ptr) {
; CHECK-LABEL: @ptrmask_align2_ptr_align1(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 -2)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 2
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 -2)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Increase load align from 1 to 4
define <16 x i8> @ptrmask_align4_ptr_align1(ptr align 1 %ptr) {
; CHECK-LABEL: @ptrmask_align4_ptr_align1(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 -4)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 4
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 -4)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Increase load align from 1 to 8
define <16 x i8> @ptrmask_align8_ptr_align1(ptr align 1 %ptr) {
; CHECK-LABEL: @ptrmask_align8_ptr_align1(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 -8)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 8
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 -8)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Underlying alignment already the same as forced alignment by ptrmask
; TODO: Should be able to drop the ptrmask
define <16 x i8> @ptrmask_align8_ptr_align8(ptr align 8 %ptr) {
; CHECK-LABEL: @ptrmask_align8_ptr_align8(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 -8)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 8
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 -8)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Underlying alignment greater than alignment forced by ptrmask
; TODO: Should be able to drop the ptrmask
define <16 x i8> @ptrmask_align8_ptr_align16(ptr align 16 %ptr) {
; CHECK-LABEL: @ptrmask_align8_ptr_align16(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i64(ptr [[PTR:%.*]], i64 -8)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 16
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i64(ptr %ptr, i64 -8)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Increase load align from 1 to 8, and the mask type is smaller
; than the pointer size.
define <16 x i8> @ptrmask_align8_ptr_align1_smallmask(ptr align 1 %ptr) {
; CHECK-LABEL: @ptrmask_align8_ptr_align1_smallmask(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[PTR:%.*]], i32 -8)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 8
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i32(ptr %ptr, i32 -8)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
; Increase load align from 1 to 8, and the mask type is larger
; than the pointer size.
define <16 x i8> @ptrmask_align8_ptr_align1_bigmask(ptr align 1 %ptr) {
; CHECK-LABEL: @ptrmask_align8_ptr_align1_bigmask(
; CHECK-NEXT: [[ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i128(ptr [[PTR:%.*]], i128 -8)
; CHECK-NEXT: [[LOAD:%.*]] = load <16 x i8>, ptr [[ALIGNED]], align 8
; CHECK-NEXT: ret <16 x i8> [[LOAD]]
;
%aligned = call ptr @llvm.ptrmask.p0.i128(ptr %ptr, i128 -8)
%load = load <16 x i8>, ptr %aligned, align 1
ret <16 x i8> %load
}
|
