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
|
; RUN: opt -S -mtriple=amdgcn-- -amdgpu-lower-module-lds --amdgpu-super-align-lds-globals=true --amdgpu-lower-module-lds-strategy=module < %s | FileCheck --check-prefixes=CHECK,SUPER-ALIGN_ON %s
; RUN: opt -S -mtriple=amdgcn-- -passes=amdgpu-lower-module-lds --amdgpu-super-align-lds-globals=true --amdgpu-lower-module-lds-strategy=module < %s | FileCheck --check-prefixes=CHECK,SUPER-ALIGN_ON %s
; RUN: opt -S -mtriple=amdgcn-- -amdgpu-lower-module-lds --amdgpu-super-align-lds-globals=false --amdgpu-lower-module-lds-strategy=module < %s | FileCheck --check-prefixes=CHECK,SUPER-ALIGN_OFF %s
; RUN: opt -S -mtriple=amdgcn-- -passes=amdgpu-lower-module-lds --amdgpu-super-align-lds-globals=false --amdgpu-lower-module-lds-strategy=module < %s | FileCheck --check-prefixes=CHECK,SUPER-ALIGN_OFF %s
; CHECK: %llvm.amdgcn.kernel.k1.lds.t = type { [32 x i8] }
; CHECK: %llvm.amdgcn.kernel.k2.lds.t = type { i16, [2 x i8], i16 }
; CHECK: %llvm.amdgcn.kernel.k3.lds.t = type { [32 x i64], [32 x i32] }
; CHECK: %llvm.amdgcn.kernel.k4.lds.t = type { [2 x ptr addrspace(3)] }
; SUPER-ALIGN_ON: @lds.unused = addrspace(3) global i32 undef, align 4
; SUPER-ALIGN_OFF: @lds.unused = addrspace(3) global i32 undef, align 2
@lds.unused = addrspace(3) global i32 undef, align 2
@llvm.used = appending global [1 x ptr] [ptr addrspacecast (ptr addrspace(3) @lds.unused to ptr)], section "llvm.metadata"
; CHECK-NOT: @lds.1
@lds.1 = internal unnamed_addr addrspace(3) global [32 x i8] undef, align 1
; SUPER-ALIGN_ON: @llvm.amdgcn.kernel.k1.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k1.lds.t undef, align 16
; SUPER-ALIGN_OFF: @llvm.amdgcn.kernel.k1.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k1.lds.t undef, align 1
; CHECK: @llvm.amdgcn.kernel.k2.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k2.lds.t undef, align 4
; SUPER-ALIGN_ON: @llvm.amdgcn.kernel.k3.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k3.lds.t undef, align 16
; SUPER-ALIGN_OFF: @llvm.amdgcn.kernel.k3.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k3.lds.t undef, align 8
; SUPER-ALIGN_ON: @llvm.amdgcn.kernel.k4.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k4.lds.t undef, align 16
; SUPER-ALIGN_OFF: @llvm.amdgcn.kernel.k4.lds = internal addrspace(3) global %llvm.amdgcn.kernel.k4.lds.t undef, align 4
; CHECK-LABEL: @k1
; CHECK: %1 = addrspacecast ptr addrspace(3) @llvm.amdgcn.kernel.k1.lds to ptr
; CHECK: %ptr = getelementptr inbounds i8, ptr %1, i64 %x
; CHECK: store i8 1, ptr %ptr, align 1
define amdgpu_kernel void @k1(i64 %x) {
%ptr = getelementptr inbounds i8, ptr addrspacecast (ptr addrspace(3) @lds.1 to ptr), i64 %x
store i8 1, ptr addrspace(0) %ptr, align 1
ret void
}
@lds.2 = internal unnamed_addr addrspace(3) global i16 undef, align 4
@lds.3 = internal unnamed_addr addrspace(3) global i16 undef, align 4
; Check that alignment is propagated to uses for scalar variables.
; CHECK-LABEL: @k2
; CHECK: store i16 1, ptr addrspace(3) @llvm.amdgcn.kernel.k2.lds, align 4
; CHECK: store i16 2, ptr addrspace(3) getelementptr inbounds (%llvm.amdgcn.kernel.k2.lds.t, ptr addrspace(3) @llvm.amdgcn.kernel.k2.lds, i32 0, i32 2), align 4
define amdgpu_kernel void @k2() {
store i16 1, ptr addrspace(3) @lds.2, align 2
store i16 2, ptr addrspace(3) @lds.3, align 2
ret void
}
@lds.4 = internal unnamed_addr addrspace(3) global [32 x i64] undef, align 8
@lds.5 = internal unnamed_addr addrspace(3) global [32 x i32] undef, align 4
; Check that alignment is propagated to uses for arrays.
; CHECK-LABEL: @k3
; CHECK: store i32 1, ptr addrspace(3) %ptr1, align 8
; CHECK: store i32 2, ptr addrspace(3) %ptr2, align 4
; SUPER-ALIGN_ON: store i32 3, ptr addrspace(3) %ptr3, align 16
; SUPER-ALIGN_OFF: store i32 3, ptr addrspace(3) %ptr3, align 8
; CHECK: store i32 4, ptr addrspace(3) %ptr4, align 4
; CHECK: store i32 5, ptr addrspace(3) %ptr5, align 4
; CHECK: %load1 = load i32, ptr addrspace(3) %ptr1, align 8
; CHECK: %load2 = load i32, ptr addrspace(3) %ptr2, align 4
; SUPER-ALIGN_ON: %load3 = load i32, ptr addrspace(3) %ptr3, align 16
; SUPER-ALIGN_OFF: %load3 = load i32, ptr addrspace(3) %ptr3, align 8
; CHECK: %load4 = load i32, ptr addrspace(3) %ptr4, align 4
; CHECK: %load5 = load i32, ptr addrspace(3) %ptr5, align 4
; CHECK: %val1 = atomicrmw volatile add ptr addrspace(3) %ptr1, i32 1 monotonic, align 8
; CHECK: %val2 = cmpxchg volatile ptr addrspace(3) %ptr1, i32 1, i32 2 monotonic monotonic, align 8
; CHECK: store i16 11, ptr addrspace(3) %ptr1, align 8
; CHECK: store i16 12, ptr addrspace(3) %ptr2, align 4
; SUPER-ALIGN_ON: store i16 13, ptr addrspace(3) %ptr3, align 16
; SUPER-ALIGN_OFF: store i16 13, ptr addrspace(3) %ptr3, align 8
; CHECK: store i16 14, ptr addrspace(3) %ptr4, align 4
; CHECK: %ptr1.ac = addrspacecast ptr addrspace(3) %ptr1 to ptr
; CHECK: %ptr2.ac = addrspacecast ptr addrspace(3) %ptr2 to ptr
; CHECK: %ptr3.ac = addrspacecast ptr addrspace(3) %ptr3 to ptr
; CHECK: %ptr4.ac = addrspacecast ptr addrspace(3) %ptr4 to ptr
; CHECK: store i32 21, ptr %ptr1.ac, align 8
; CHECK: store i32 22, ptr %ptr2.ac, align 4
; SUPER-ALIGN_ON: store i32 23, ptr %ptr3.ac, align 16
; SUPER-ALIGN_OFF: store i32 23, ptr %ptr3.ac, align 8
; CHECK: store i32 24, ptr %ptr4.ac, align 4
define amdgpu_kernel void @k3(i64 %x) {
store i64 0, ptr addrspace(3) @lds.4, align 8
%ptr1 = getelementptr inbounds i32, ptr addrspace(3) @lds.5, i64 2
%ptr2 = getelementptr inbounds i32, ptr addrspace(3) @lds.5, i64 3
%ptr3 = getelementptr inbounds i32, ptr addrspace(3) @lds.5, i64 4
%ptr4 = getelementptr inbounds i32, ptr addrspace(3) @lds.5, i64 5
%ptr5 = getelementptr inbounds i32, ptr addrspace(3) @lds.5, i64 %x
store i32 1, ptr addrspace(3) %ptr1, align 4
store i32 2, ptr addrspace(3) %ptr2, align 4
store i32 3, ptr addrspace(3) %ptr3, align 4
store i32 4, ptr addrspace(3) %ptr4, align 4
store i32 5, ptr addrspace(3) %ptr5, align 4
%load1 = load i32, ptr addrspace(3) %ptr1, align 4
%load2 = load i32, ptr addrspace(3) %ptr2, align 4
%load3 = load i32, ptr addrspace(3) %ptr3, align 4
%load4 = load i32, ptr addrspace(3) %ptr4, align 4
%load5 = load i32, ptr addrspace(3) %ptr5, align 4
%val1 = atomicrmw volatile add ptr addrspace(3) %ptr1, i32 1 monotonic, align 4
%val2 = cmpxchg volatile ptr addrspace(3) %ptr1, i32 1, i32 2 monotonic monotonic, align 4
store i16 11, ptr addrspace(3) %ptr1, align 2
store i16 12, ptr addrspace(3) %ptr2, align 2
store i16 13, ptr addrspace(3) %ptr3, align 2
store i16 14, ptr addrspace(3) %ptr4, align 2
%ptr1.ac = addrspacecast ptr addrspace(3) %ptr1 to ptr
%ptr2.ac = addrspacecast ptr addrspace(3) %ptr2 to ptr
%ptr3.ac = addrspacecast ptr addrspace(3) %ptr3 to ptr
%ptr4.ac = addrspacecast ptr addrspace(3) %ptr4 to ptr
store i32 21, ptr %ptr1.ac, align 4
store i32 22, ptr %ptr2.ac, align 4
store i32 23, ptr %ptr3.ac, align 4
store i32 24, ptr %ptr4.ac, align 4
ret void
}
@lds.6 = internal unnamed_addr addrspace(3) global [2 x ptr addrspace(3)] undef, align 4
; Check that aligment is not propagated if use is not a pointer operand.
; CHECK-LABEL: @k4
; SUPER-ALIGN_ON: store i32 undef, ptr addrspace(3) %gep, align 8
; SUPER-ALIGN_OFF: store i32 undef, ptr addrspace(3) %gep, align 4
; CHECK: store ptr addrspace(3) %gep, ptr undef, align 4
; SUPER-ALIGN_ON: %val1 = cmpxchg volatile ptr addrspace(3) %gep, i32 1, i32 2 monotonic monotonic, align 8
; SUPER-ALIGN_OFF: %val1 = cmpxchg volatile ptr addrspace(3) %gep, i32 1, i32 2 monotonic monotonic, align 4
; CHECK: %val2 = cmpxchg volatile ptr undef, ptr addrspace(3) %gep, ptr addrspace(3) undef monotonic monotonic, align 4
define amdgpu_kernel void @k4() {
%gep = getelementptr inbounds ptr addrspace(3), ptr addrspace(3) @lds.6, i64 1
store i32 undef, ptr addrspace(3) %gep, align 4
store ptr addrspace(3) %gep, ptr undef, align 4
%val1 = cmpxchg volatile ptr addrspace(3) %gep, i32 1, i32 2 monotonic monotonic, align 4
%val2 = cmpxchg volatile ptr undef, ptr addrspace(3) %gep, ptr addrspace(3) undef monotonic monotonic, align 4
ret void
}
|
