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
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
| ; RUN: llc < %s -march=sparcv9 -verify-machineinstrs | FileCheck %s
; CHECK-LABEL: test_atomic_i8
; CHECK: ldub [%o0]
; CHECK: membar
; CHECK: ldub [%o1]
; CHECK: membar
; CHECK: membar
; CHECK: stb {{.+}}, [%o2]
define i8 @test_atomic_i8(i8* %ptr1, i8* %ptr2, i8* %ptr3) {
entry:
%0 = load atomic i8, i8* %ptr1 acquire, align 1
%1 = load atomic i8, i8* %ptr2 acquire, align 1
%2 = add i8 %0, %1
store atomic i8 %2, i8* %ptr3 release, align 1
ret i8 %2
}
; CHECK-LABEL: test_atomic_i16
; CHECK: lduh [%o0]
; CHECK: membar
; CHECK: lduh [%o1]
; CHECK: membar
; CHECK: membar
; CHECK: sth {{.+}}, [%o2]
define i16 @test_atomic_i16(i16* %ptr1, i16* %ptr2, i16* %ptr3) {
entry:
%0 = load atomic i16, i16* %ptr1 acquire, align 2
%1 = load atomic i16, i16* %ptr2 acquire, align 2
%2 = add i16 %0, %1
store atomic i16 %2, i16* %ptr3 release, align 2
ret i16 %2
}
; CHECK-LABEL: test_atomic_i32
; CHECK: ld [%o0]
; CHECK: membar
; CHECK: ld [%o1]
; CHECK: membar
; CHECK: membar
; CHECK: st {{.+}}, [%o2]
define i32 @test_atomic_i32(i32* %ptr1, i32* %ptr2, i32* %ptr3) {
entry:
%0 = load atomic i32, i32* %ptr1 acquire, align 4
%1 = load atomic i32, i32* %ptr2 acquire, align 4
%2 = add i32 %0, %1
store atomic i32 %2, i32* %ptr3 release, align 4
ret i32 %2
}
; CHECK-LABEL: test_atomic_i64
; CHECK: ldx [%o0]
; CHECK: membar
; CHECK: ldx [%o1]
; CHECK: membar
; CHECK: membar
; CHECK: stx {{.+}}, [%o2]
define i64 @test_atomic_i64(i64* %ptr1, i64* %ptr2, i64* %ptr3) {
entry:
%0 = load atomic i64, i64* %ptr1 acquire, align 8
%1 = load atomic i64, i64* %ptr2 acquire, align 8
%2 = add i64 %0, %1
store atomic i64 %2, i64* %ptr3 release, align 8
ret i64 %2
}
;; TODO: the "move %icc" and related instructions are totally
;; redundant here. There's something weird happening in optimization
;; of the success value of cmpxchg.
; CHECK-LABEL: test_cmpxchg_i8
; CHECK: and %o1, -4, %o2
; CHECK: mov 3, %o3
; CHECK: andn %o3, %o1, %o1
; CHECK: sll %o1, 3, %o1
; CHECK: mov 255, %o3
; CHECK: sll %o3, %o1, %o5
; CHECK: xor %o5, -1, %o3
; CHECK: mov 123, %o4
; CHECK: ld [%o2], %g2
; CHECK: sll %o4, %o1, %o4
; CHECK: and %o0, 255, %o0
; CHECK: sll %o0, %o1, %o0
; CHECK: andn %g2, %o5, %g2
; CHECK: mov %g0, %o5
; CHECK: [[LABEL1:\.L.*]]:
; CHECK: or %g2, %o4, %g3
; CHECK: or %g2, %o0, %g4
; CHECK: cas [%o2], %g4, %g3
; CHECK: cmp %g3, %g4
; CHECK: mov %o5, %g4
; CHECK: move %icc, 1, %g4
; CHECK: cmp %g4, 0
; CHECK: bne [[LABEL2:\.L.*]]
; CHECK: nop
; CHECK: and %g3, %o3, %g4
; CHECK: cmp %g2, %g4
; CHECK: bne [[LABEL1]]
; CHECK: mov %g4, %g2
; CHECK: [[LABEL2]]:
; CHECK: retl
; CHECK: srl %g3, %o1, %o0
define i8 @test_cmpxchg_i8(i8 %a, i8* %ptr) {
entry:
%pair = cmpxchg i8* %ptr, i8 %a, i8 123 monotonic monotonic
%b = extractvalue { i8, i1 } %pair, 0
ret i8 %b
}
; CHECK-LABEL: test_cmpxchg_i16
; CHECK: and %o1, -4, %o2
; CHECK: and %o1, 3, %o1
; CHECK: xor %o1, 2, %o1
; CHECK: sll %o1, 3, %o1
; CHECK: sethi 63, %o3
; CHECK: or %o3, 1023, %o4
; CHECK: sll %o4, %o1, %o5
; CHECK: xor %o5, -1, %o3
; CHECK: and %o0, %o4, %o4
; CHECK: ld [%o2], %g2
; CHECK: mov 123, %o0
; CHECK: sll %o0, %o1, %o0
; CHECK: sll %o4, %o1, %o4
; CHECK: andn %g2, %o5, %g2
; CHECK: mov %g0, %o5
; CHECK: [[LABEL1:\.L.*]]:
; CHECK: or %g2, %o0, %g3
; CHECK: or %g2, %o4, %g4
; CHECK: cas [%o2], %g4, %g3
; CHECK: cmp %g3, %g4
; CHECK: mov %o5, %g4
; CHECK: move %icc, 1, %g4
; CHECK: cmp %g4, 0
; CHECK: bne [[LABEL2:\.L.*]]
; CHECK: nop
; CHECK: and %g3, %o3, %g4
; CHECK: cmp %g2, %g4
; CHECK: bne [[LABEL1]]
; CHECK: mov %g4, %g2
; CHECK: [[LABEL2]]:
; CHECK: retl
; CHECK: srl %g3, %o1, %o0
define i16 @test_cmpxchg_i16(i16 %a, i16* %ptr) {
entry:
%pair = cmpxchg i16* %ptr, i16 %a, i16 123 monotonic monotonic
%b = extractvalue { i16, i1 } %pair, 0
ret i16 %b
}
; CHECK-LABEL: test_cmpxchg_i32
; CHECK: mov 123, [[R:%[gilo][0-7]]]
; CHECK: cas [%o1], %o0, [[R]]
define i32 @test_cmpxchg_i32(i32 %a, i32* %ptr) {
entry:
%pair = cmpxchg i32* %ptr, i32 %a, i32 123 monotonic monotonic
%b = extractvalue { i32, i1 } %pair, 0
ret i32 %b
}
; CHECK-LABEL: test_cmpxchg_i64
; CHECK: mov 123, [[R:%[gilo][0-7]]]
; CHECK: casx [%o1], %o0, [[R]]
define i64 @test_cmpxchg_i64(i64 %a, i64* %ptr) {
entry:
%pair = cmpxchg i64* %ptr, i64 %a, i64 123 monotonic monotonic
%b = extractvalue { i64, i1 } %pair, 0
ret i64 %b
}
; CHECK-LABEL: test_swap_i8
; CHECK: mov 42, [[R:%[gilo][0-7]]]
; CHECK: cas
define i8 @test_swap_i8(i8 %a, i8* %ptr) {
entry:
%b = atomicrmw xchg i8* %ptr, i8 42 monotonic
ret i8 %b
}
; CHECK-LABEL: test_swap_i16
; CHECK: mov 42, [[R:%[gilo][0-7]]]
; CHECK: cas
define i16 @test_swap_i16(i16 %a, i16* %ptr) {
entry:
%b = atomicrmw xchg i16* %ptr, i16 42 monotonic
ret i16 %b
}
; CHECK-LABEL: test_swap_i32
; CHECK: mov 42, [[R:%[gilo][0-7]]]
; CHECK: swap [%o1], [[R]]
define i32 @test_swap_i32(i32 %a, i32* %ptr) {
entry:
%b = atomicrmw xchg i32* %ptr, i32 42 monotonic
ret i32 %b
}
; CHECK-LABEL: test_swap_i64
; CHECK: casx [%o1],
define i64 @test_swap_i64(i64 %a, i64* %ptr) {
entry:
%b = atomicrmw xchg i64* %ptr, i64 42 monotonic
ret i64 %b
}
; CHECK-LABEL: test_load_sub_i8
; CHECK: membar
; CHECK: .L{{.*}}:
; CHECK: sub
; CHECK: cas [{{%[gilo][0-7]}}]
; CHECK: membar
define zeroext i8 @test_load_sub_i8(i8* %p, i8 zeroext %v) {
entry:
%0 = atomicrmw sub i8* %p, i8 %v seq_cst
ret i8 %0
}
; CHECK-LABEL: test_load_sub_i16
; CHECK: membar
; CHECK: .L{{.*}}:
; CHECK: sub
; CHECK: cas [{{%[gilo][0-7]}}]
; CHECK: membar
define zeroext i16 @test_load_sub_i16(i16* %p, i16 zeroext %v) {
entry:
%0 = atomicrmw sub i16* %p, i16 %v seq_cst
ret i16 %0
}
; CHECK-LABEL: test_load_add_i32
; CHECK: membar
; CHECK: mov %g0
; CHECK: mov [[U:%[gilo][0-7]]], [[V:%[gilo][0-7]]]
; CHECK: add [[U:%[gilo][0-7]]], %o1, [[V2:%[gilo][0-7]]]
; CHECK: cas [%o0], [[V]], [[V2]]
; CHECK: membar
define zeroext i32 @test_load_add_i32(i32* %p, i32 zeroext %v) {
entry:
%0 = atomicrmw add i32* %p, i32 %v seq_cst
ret i32 %0
}
; CHECK-LABEL: test_load_sub_64
; CHECK: membar
; CHECK: sub
; CHECK: casx [%o0]
; CHECK: membar
define zeroext i64 @test_load_sub_64(i64* %p, i64 zeroext %v) {
entry:
%0 = atomicrmw sub i64* %p, i64 %v seq_cst
ret i64 %0
}
; CHECK-LABEL: test_load_xor_32
; CHECK: membar
; CHECK: xor
; CHECK: cas [%o0]
; CHECK: membar
define zeroext i32 @test_load_xor_32(i32* %p, i32 zeroext %v) {
entry:
%0 = atomicrmw xor i32* %p, i32 %v seq_cst
ret i32 %0
}
; CHECK-LABEL: test_load_and_32
; CHECK: membar
; CHECK: and
; CHECK-NOT: xor
; CHECK: cas [%o0]
; CHECK: membar
define zeroext i32 @test_load_and_32(i32* %p, i32 zeroext %v) {
entry:
%0 = atomicrmw and i32* %p, i32 %v seq_cst
ret i32 %0
}
; CHECK-LABEL: test_load_nand_32
; CHECK: membar
; CHECK: and
; CHECK: xor
; CHECK: cas [%o0]
; CHECK: membar
define zeroext i32 @test_load_nand_32(i32* %p, i32 zeroext %v) {
entry:
%0 = atomicrmw nand i32* %p, i32 %v seq_cst
ret i32 %0
}
; CHECK-LABEL: test_load_max_64
; CHECK: membar
; CHECK: cmp
; CHECK: movg %xcc
; CHECK: casx [%o0]
; CHECK: membar
define zeroext i64 @test_load_max_64(i64* %p, i64 zeroext %v) {
entry:
%0 = atomicrmw max i64* %p, i64 %v seq_cst
ret i64 %0
}
; CHECK-LABEL: test_load_umin_32
; CHECK: membar
; CHECK: cmp
; CHECK: movleu %icc
; CHECK: cas [%o0]
; CHECK: membar
define zeroext i32 @test_load_umin_32(i32* %p, i32 zeroext %v) {
entry:
%0 = atomicrmw umin i32* %p, i32 %v seq_cst
ret i32 %0
}
|