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
| ; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=+bulk-memory | FileCheck %s --check-prefixes CHECK,BULK-MEM
; RUN: llc < %s -asm-verbose=false -verify-machineinstrs -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=-bulk-memory | FileCheck %s --check-prefixes CHECK,NO-BULK-MEM
; Test that basic bulk memory codegen works correctly
target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
target triple = "wasm32-unknown-unknown"
declare void @llvm.memcpy.p0i8.p0i8.i8(i8*, i8*, i8, i1)
declare void @llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i1)
declare void @llvm.memcpy.p0i32.p0i32.i32(i32*, i32*, i32, i1)
declare void @llvm.memmove.p0i8.p0i8.i8(i8*, i8*, i8, i1)
declare void @llvm.memmove.p0i8.p0i8.i32(i8*, i8*, i32, i1)
declare void @llvm.memmove.p0i32.p0i32.i32(i32*, i32*, i32, i1)
declare void @llvm.memset.p0i8.i8(i8*, i8, i8, i1)
declare void @llvm.memset.p0i8.i32(i8*, i8, i32, i1)
declare void @llvm.memset.p0i32.i32(i32*, i8, i32, i1)
; CHECK-LABEL: memcpy_i8:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memcpy_i8 (i32, i32, i32) -> ()
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $2
; BULK-MEM-NEXT: return
define void @memcpy_i8(i8* %dest, i8* %src, i8 zeroext %len) {
call void @llvm.memcpy.p0i8.p0i8.i8(i8* %dest, i8* %src, i8 %len, i1 0)
ret void
}
; CHECK-LABEL: memmove_i8:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memmove_i8 (i32, i32, i32) -> ()
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $2
; BULK-MEM-NEXT: return
define void @memmove_i8(i8* %dest, i8* %src, i8 zeroext %len) {
call void @llvm.memmove.p0i8.p0i8.i8(i8* %dest, i8* %src, i8 %len, i1 0)
ret void
}
; CHECK-LABEL: memset_i8:
; NO-BULK-MEM-NOT: memory.fill
; BULK-MEM-NEXT: .functype memset_i8 (i32, i32, i32) -> ()
; BULK-MEM-NEXT: memory.fill 0, $0, $1, $2
; BULK-MEM-NEXT: return
define void @memset_i8(i8* %dest, i8 %val, i8 zeroext %len) {
call void @llvm.memset.p0i8.i8(i8* %dest, i8 %val, i8 %len, i1 0)
ret void
}
; CHECK-LABEL: memcpy_i32:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memcpy_i32 (i32, i32, i32) -> ()
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $2
; BULK-MEM-NEXT: return
define void @memcpy_i32(i32* %dest, i32* %src, i32 %len) {
call void @llvm.memcpy.p0i32.p0i32.i32(i32* %dest, i32* %src, i32 %len, i1 0)
ret void
}
; CHECK-LABEL: memmove_i32:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memmove_i32 (i32, i32, i32) -> ()
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $2
; BULK-MEM-NEXT: return
define void @memmove_i32(i32* %dest, i32* %src, i32 %len) {
call void @llvm.memmove.p0i32.p0i32.i32(i32* %dest, i32* %src, i32 %len, i1 0)
ret void
}
; CHECK-LABEL: memset_i32:
; NO-BULK-MEM-NOT: memory.fill
; BULK-MEM-NEXT: .functype memset_i32 (i32, i32, i32) -> ()
; BULK-MEM-NEXT: memory.fill 0, $0, $1, $2
; BULK-MEM-NEXT: return
define void @memset_i32(i32* %dest, i8 %val, i32 %len) {
call void @llvm.memset.p0i32.i32(i32* %dest, i8 %val, i32 %len, i1 0)
ret void
}
; CHECK-LABEL: memcpy_1:
; CHECK-NEXT: .functype memcpy_1 (i32, i32) -> ()
; CHECK-NEXT: i32.load8_u $push[[L0:[0-9]+]]=, 0($1)
; CHECK-NEXT: i32.store8 0($0), $pop[[L0]]
; CHECK-NEXT: return
define void @memcpy_1(i8* %dest, i8* %src) {
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 1, i1 0)
ret void
}
; CHECK-LABEL: memmove_1:
; CHECK-NEXT: .functype memmove_1 (i32, i32) -> ()
; CHECK-NEXT: i32.load8_u $push[[L0:[0-9]+]]=, 0($1)
; CHECK-NEXT: i32.store8 0($0), $pop[[L0]]
; CHECK-NEXT: return
define void @memmove_1(i8* %dest, i8* %src) {
call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 1, i1 0)
ret void
}
; CHECK-LABEL: memset_1:
; NO-BULK-MEM-NOT: memory.fill
; BULK-MEM-NEXT: .functype memset_1 (i32, i32) -> ()
; BULK-MEM-NEXT: i32.store8 0($0), $1
; BULK-MEM-NEXT: return
define void @memset_1(i8* %dest, i8 %val) {
call void @llvm.memset.p0i8.i32(i8* %dest, i8 %val, i32 1, i1 0)
ret void
}
; CHECK-LABEL: memcpy_1024:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memcpy_1024 (i32, i32) -> ()
; BULK-MEM-NEXT: i32.const $push[[L0:[0-9]+]]=, 1024
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L0]]
; BULK-MEM-NEXT: return
define void @memcpy_1024(i8* %dest, i8* %src) {
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 1024, i1 0)
ret void
}
; CHECK-LABEL: memmove_1024:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memmove_1024 (i32, i32) -> ()
; BULK-MEM-NEXT: i32.const $push[[L0:[0-9]+]]=, 1024
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L0]]
; BULK-MEM-NEXT: return
define void @memmove_1024(i8* %dest, i8* %src) {
call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 1024, i1 0)
ret void
}
; CHECK-LABEL: memset_1024:
; NO-BULK-MEM-NOT: memory.fill
; BULK-MEM-NEXT: .functype memset_1024 (i32, i32) -> ()
; BULK-MEM-NEXT: i32.const $push[[L0:[0-9]+]]=, 1024
; BULK-MEM-NEXT: memory.fill 0, $0, $1, $pop[[L0]]
; BULK-MEM-NEXT: return
define void @memset_1024(i8* %dest, i8 %val) {
call void @llvm.memset.p0i8.i32(i8* %dest, i8 %val, i32 1024, i1 0)
ret void
}
; The following tests check that frame index elimination works for
; bulk memory instructions. The stack pointer is bumped by 112 instead
; of 100 because the stack pointer in WebAssembly is currently always
; 16-byte aligned, even in leaf functions, although it is not written
; back to the global in this case.
; TODO: Change TransientStackAlignment to 1 to avoid this extra
; arithmetic. This will require forcing the use of StackAlignment in
; PrologEpilogEmitter.cpp when
; WebAssemblyFrameLowering::needsSPWriteback would be true.
; CHECK-LABEL: memcpy_alloca_src:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memcpy_alloca_src (i32) -> ()
; BULK-MEM-NEXT: global.get $push[[L0:[0-9]+]]=, __stack_pointer
; BULK-MEM-NEXT: i32.const $push[[L1:[0-9]+]]=, 112
; BULK-MEM-NEXT: i32.sub $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]
; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 12
; BULK-MEM-NEXT: i32.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]]
; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 100
; BULK-MEM-NEXT: memory.copy 0, 0, $0, $pop[[L4]], $pop[[L5]]
; BULK-MEM-NEXT: return
define void @memcpy_alloca_src(i8* %dst) {
%a = alloca [100 x i8]
%p = bitcast [100 x i8]* %a to i8*
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dst, i8* %p, i32 100, i1 false)
ret void
}
; CHECK-LABEL: memcpy_alloca_dst:
; NO-BULK-MEM-NOT: memory.copy
; BULK-MEM-NEXT: .functype memcpy_alloca_dst (i32) -> ()
; BULK-MEM-NEXT: global.get $push[[L0:[0-9]+]]=, __stack_pointer
; BULK-MEM-NEXT: i32.const $push[[L1:[0-9]+]]=, 112
; BULK-MEM-NEXT: i32.sub $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]
; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 12
; BULK-MEM-NEXT: i32.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]]
; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 100
; BULK-MEM-NEXT: memory.copy 0, 0, $pop[[L4]], $0, $pop[[L5]]
; BULK-MEM-NEXT: return
define void @memcpy_alloca_dst(i8* %src) {
%a = alloca [100 x i8]
%p = bitcast [100 x i8]* %a to i8*
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %src, i32 100, i1 false)
ret void
}
; CHECK-LABEL: memset_alloca:
; NO-BULK-MEM-NOT: memory.fill
; BULK-MEM-NEXT: .functype memset_alloca (i32) -> ()
; BULK-MEM-NEXT: global.get $push[[L0:[0-9]+]]=, __stack_pointer
; BULK-MEM-NEXT: i32.const $push[[L1:[0-9]+]]=, 112
; BULK-MEM-NEXT: i32.sub $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]]
; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 12
; BULK-MEM-NEXT: i32.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]]
; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 100
; BULK-MEM-NEXT: memory.fill 0, $pop[[L4]], $0, $pop[[L5]]
; BULK-MEM-NEXT: return
define void @memset_alloca(i8 %val) {
%a = alloca [100 x i8]
%p = bitcast [100 x i8]* %a to i8*
call void @llvm.memset.p0i8.i32(i8* %p, i8 %val, i32 100, i1 false)
ret void
}
|