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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -S -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -passes=amdgpu-promote-alloca < %s | FileCheck %s
; RUN: opt -S -mtriple=amdgcn-unknown-amdhsa -mcpu=kaveri -passes=amdgpu-promote-alloca -disable-promote-alloca-to-lds< %s | FileCheck -check-prefix=NOLDS %s
; This normally would be fixed by instcombine to be compare to the GEP
; indices
define amdgpu_kernel void @lds_promoted_alloca_icmp_same_derived_pointer(ptr addrspace(1) %out, i32 %a, i32 %b) #0 {
; CHECK-LABEL: @lds_promoted_alloca_icmp_same_derived_pointer(
; CHECK-NEXT: [[TMP1:%.*]] = call noalias nonnull dereferenceable(64) ptr addrspace(4) @llvm.amdgcn.dispatch.ptr()
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[TMP1]], i64 1
; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr addrspace(4) [[TMP3]], align 4, !invariant.load !0
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[TMP1]], i64 2
; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(4) [[TMP5]], align 4, !range [[RNG1:![0-9]+]], !invariant.load !0
; CHECK-NEXT: [[TMP7:%.*]] = lshr i32 [[TMP4]], 16
; CHECK-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.workitem.id.x(), !range [[RNG2:![0-9]+]]
; CHECK-NEXT: [[TMP9:%.*]] = call i32 @llvm.amdgcn.workitem.id.y(), !range [[RNG2]]
; CHECK-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.workitem.id.z(), !range [[RNG2]]
; CHECK-NEXT: [[TMP11:%.*]] = mul nuw nsw i32 [[TMP7]], [[TMP6]]
; CHECK-NEXT: [[TMP12:%.*]] = mul i32 [[TMP11]], [[TMP8]]
; CHECK-NEXT: [[TMP13:%.*]] = mul nuw nsw i32 [[TMP9]], [[TMP6]]
; CHECK-NEXT: [[TMP14:%.*]] = add i32 [[TMP12]], [[TMP13]]
; CHECK-NEXT: [[TMP15:%.*]] = add i32 [[TMP14]], [[TMP10]]
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds [256 x [16 x i32]], ptr addrspace(3) @lds_promoted_alloca_icmp_same_derived_pointer.alloca, i32 0, i32 [[TMP15]]
; CHECK-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(3) [[TMP16]], i32 0, i32 [[A:%.*]]
; CHECK-NEXT: [[PTR1:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(3) [[TMP16]], i32 0, i32 [[B:%.*]]
; CHECK-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(3) [[PTR0]], [[PTR1]]
; CHECK-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; CHECK-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; CHECK-NEXT: ret void
;
; NOLDS-LABEL: @lds_promoted_alloca_icmp_same_derived_pointer(
; NOLDS-NEXT: [[ALLOCA:%.*]] = alloca [16 x i32], align 4, addrspace(5)
; NOLDS-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[A:%.*]]
; NOLDS-NEXT: [[PTR1:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[B:%.*]]
; NOLDS-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(5) [[PTR0]], [[PTR1]]
; NOLDS-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; NOLDS-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; NOLDS-NEXT: ret void
;
%alloca = alloca [16 x i32], align 4, addrspace(5)
%ptr0 = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 0, i32 %a
%ptr1 = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 0, i32 %b
%cmp = icmp eq ptr addrspace(5) %ptr0, %ptr1
%zext = zext i1 %cmp to i32
store volatile i32 %zext, ptr addrspace(1) %out
ret void
}
define amdgpu_kernel void @lds_promoted_alloca_icmp_null_rhs(ptr addrspace(1) %out, i32 %a, i32 %b) #0 {
; CHECK-LABEL: @lds_promoted_alloca_icmp_null_rhs(
; CHECK-NEXT: [[TMP1:%.*]] = call noalias nonnull dereferenceable(64) ptr addrspace(4) @llvm.amdgcn.dispatch.ptr()
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[TMP1]], i64 1
; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr addrspace(4) [[TMP3]], align 4, !invariant.load !0
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[TMP1]], i64 2
; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(4) [[TMP5]], align 4, !range [[RNG1]], !invariant.load !0
; CHECK-NEXT: [[TMP7:%.*]] = lshr i32 [[TMP4]], 16
; CHECK-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.workitem.id.x(), !range [[RNG2]]
; CHECK-NEXT: [[TMP9:%.*]] = call i32 @llvm.amdgcn.workitem.id.y(), !range [[RNG2]]
; CHECK-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.workitem.id.z(), !range [[RNG2]]
; CHECK-NEXT: [[TMP11:%.*]] = mul nuw nsw i32 [[TMP7]], [[TMP6]]
; CHECK-NEXT: [[TMP12:%.*]] = mul i32 [[TMP11]], [[TMP8]]
; CHECK-NEXT: [[TMP13:%.*]] = mul nuw nsw i32 [[TMP9]], [[TMP6]]
; CHECK-NEXT: [[TMP14:%.*]] = add i32 [[TMP12]], [[TMP13]]
; CHECK-NEXT: [[TMP15:%.*]] = add i32 [[TMP14]], [[TMP10]]
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds [256 x [16 x i32]], ptr addrspace(3) @lds_promoted_alloca_icmp_null_rhs.alloca, i32 0, i32 [[TMP15]]
; CHECK-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(3) [[TMP16]], i32 0, i32 [[A:%.*]]
; CHECK-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(3) [[PTR0]], null
; CHECK-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; CHECK-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; CHECK-NEXT: ret void
;
; NOLDS-LABEL: @lds_promoted_alloca_icmp_null_rhs(
; NOLDS-NEXT: [[ALLOCA:%.*]] = alloca [16 x i32], align 4, addrspace(5)
; NOLDS-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[A:%.*]]
; NOLDS-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(5) [[PTR0]], null
; NOLDS-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; NOLDS-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; NOLDS-NEXT: ret void
;
%alloca = alloca [16 x i32], align 4, addrspace(5)
%ptr0 = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 0, i32 %a
%cmp = icmp eq ptr addrspace(5) %ptr0, null
%zext = zext i1 %cmp to i32
store volatile i32 %zext, ptr addrspace(1) %out
ret void
}
define amdgpu_kernel void @lds_promoted_alloca_icmp_null_lhs(ptr addrspace(1) %out, i32 %a, i32 %b) #0 {
; CHECK-LABEL: @lds_promoted_alloca_icmp_null_lhs(
; CHECK-NEXT: [[TMP1:%.*]] = call noalias nonnull dereferenceable(64) ptr addrspace(4) @llvm.amdgcn.dispatch.ptr()
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[TMP1]], i64 1
; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr addrspace(4) [[TMP3]], align 4, !invariant.load !0
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, ptr addrspace(4) [[TMP1]], i64 2
; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(4) [[TMP5]], align 4, !range [[RNG1]], !invariant.load !0
; CHECK-NEXT: [[TMP7:%.*]] = lshr i32 [[TMP4]], 16
; CHECK-NEXT: [[TMP8:%.*]] = call i32 @llvm.amdgcn.workitem.id.x(), !range [[RNG2]]
; CHECK-NEXT: [[TMP9:%.*]] = call i32 @llvm.amdgcn.workitem.id.y(), !range [[RNG2]]
; CHECK-NEXT: [[TMP10:%.*]] = call i32 @llvm.amdgcn.workitem.id.z(), !range [[RNG2]]
; CHECK-NEXT: [[TMP11:%.*]] = mul nuw nsw i32 [[TMP7]], [[TMP6]]
; CHECK-NEXT: [[TMP12:%.*]] = mul i32 [[TMP11]], [[TMP8]]
; CHECK-NEXT: [[TMP13:%.*]] = mul nuw nsw i32 [[TMP9]], [[TMP6]]
; CHECK-NEXT: [[TMP14:%.*]] = add i32 [[TMP12]], [[TMP13]]
; CHECK-NEXT: [[TMP15:%.*]] = add i32 [[TMP14]], [[TMP10]]
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds [256 x [16 x i32]], ptr addrspace(3) @lds_promoted_alloca_icmp_null_lhs.alloca, i32 0, i32 [[TMP15]]
; CHECK-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(3) [[TMP16]], i32 0, i32 [[A:%.*]]
; CHECK-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(3) null, [[PTR0]]
; CHECK-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; CHECK-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; CHECK-NEXT: ret void
;
; NOLDS-LABEL: @lds_promoted_alloca_icmp_null_lhs(
; NOLDS-NEXT: [[ALLOCA:%.*]] = alloca [16 x i32], align 4, addrspace(5)
; NOLDS-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[A:%.*]]
; NOLDS-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(5) null, [[PTR0]]
; NOLDS-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; NOLDS-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; NOLDS-NEXT: ret void
;
%alloca = alloca [16 x i32], align 4, addrspace(5)
%ptr0 = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 0, i32 %a
%cmp = icmp eq ptr addrspace(5) null, %ptr0
%zext = zext i1 %cmp to i32
store volatile i32 %zext, ptr addrspace(1) %out
ret void
}
define amdgpu_kernel void @lds_promoted_alloca_icmp_unknown_ptr(ptr addrspace(1) %out, i32 %a, i32 %b) #0 {
; CHECK-LABEL: @lds_promoted_alloca_icmp_unknown_ptr(
; CHECK-NEXT: [[ALLOCA:%.*]] = alloca [16 x i32], align 4, addrspace(5)
; CHECK-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[A:%.*]]
; CHECK-NEXT: [[PTR1:%.*]] = call ptr addrspace(5) @get_unknown_pointer()
; CHECK-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(5) [[PTR0]], [[PTR1]]
; CHECK-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; CHECK-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; CHECK-NEXT: ret void
;
; NOLDS-LABEL: @lds_promoted_alloca_icmp_unknown_ptr(
; NOLDS-NEXT: [[ALLOCA:%.*]] = alloca [16 x i32], align 4, addrspace(5)
; NOLDS-NEXT: [[PTR0:%.*]] = getelementptr inbounds [16 x i32], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[A:%.*]]
; NOLDS-NEXT: [[PTR1:%.*]] = call ptr addrspace(5) @get_unknown_pointer()
; NOLDS-NEXT: [[CMP:%.*]] = icmp eq ptr addrspace(5) [[PTR0]], [[PTR1]]
; NOLDS-NEXT: [[ZEXT:%.*]] = zext i1 [[CMP]] to i32
; NOLDS-NEXT: store volatile i32 [[ZEXT]], ptr addrspace(1) [[OUT:%.*]], align 4
; NOLDS-NEXT: ret void
;
%alloca = alloca [16 x i32], align 4, addrspace(5)
%ptr0 = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 0, i32 %a
%ptr1 = call ptr addrspace(5) @get_unknown_pointer()
%cmp = icmp eq ptr addrspace(5) %ptr0, %ptr1
%zext = zext i1 %cmp to i32
store volatile i32 %zext, ptr addrspace(1) %out
ret void
}
declare ptr addrspace(5) @get_unknown_pointer() #0
attributes #0 = { nounwind "amdgpu-waves-per-eu"="1,1" "amdgpu-flat-work-group-size"="1,256" }
|
