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
|
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
#include "textflag.h"
#include "funcdata.h"
// bool Cas(int32 *val, int32 old, int32 new)
// Atomically:
// if(*val == old){
// *val = new;
// return 1;
// }else
// return 0;
TEXT ·Cas(SB), NOSPLIT, $0-13
MOVL ptr+0(FP), BX
MOVL old+4(FP), AX
MOVL new+8(FP), CX
LOCK
CMPXCHGL CX, 0(BX)
SETEQ ret+12(FP)
RET
TEXT ·Casint32(SB), NOSPLIT, $0-13
JMP ·Cas(SB)
TEXT ·Casint64(SB), NOSPLIT, $0-21
JMP ·Cas64(SB)
TEXT ·Casuintptr(SB), NOSPLIT, $0-13
JMP ·Cas(SB)
TEXT ·CasRel(SB), NOSPLIT, $0-13
JMP ·Cas(SB)
TEXT ·Loaduintptr(SB), NOSPLIT, $0-8
JMP ·Load(SB)
TEXT ·Loaduint(SB), NOSPLIT, $0-8
JMP ·Load(SB)
TEXT ·Storeint32(SB), NOSPLIT, $0-8
JMP ·Store(SB)
TEXT ·Storeint64(SB), NOSPLIT, $0-12
JMP ·Store64(SB)
TEXT ·Storeuintptr(SB), NOSPLIT, $0-8
JMP ·Store(SB)
TEXT ·Xadduintptr(SB), NOSPLIT, $0-12
JMP ·Xadd(SB)
TEXT ·Loadint32(SB), NOSPLIT, $0-8
JMP ·Load(SB)
TEXT ·Loadint64(SB), NOSPLIT, $0-12
JMP ·Load64(SB)
TEXT ·Xaddint32(SB), NOSPLIT, $0-12
JMP ·Xadd(SB)
TEXT ·Xaddint64(SB), NOSPLIT, $0-20
JMP ·Xadd64(SB)
// bool ·Cas64(uint64 *val, uint64 old, uint64 new)
// Atomically:
// if(*val == old){
// *val = new;
// return 1;
// } else {
// return 0;
// }
TEXT ·Cas64(SB), NOSPLIT, $0-21
NO_LOCAL_POINTERS
MOVL ptr+0(FP), BP
TESTL $7, BP
JZ 2(PC)
CALL ·panicUnaligned(SB)
MOVL old_lo+4(FP), AX
MOVL old_hi+8(FP), DX
MOVL new_lo+12(FP), BX
MOVL new_hi+16(FP), CX
LOCK
CMPXCHG8B 0(BP)
SETEQ ret+20(FP)
RET
// bool Casp1(void **p, void *old, void *new)
// Atomically:
// if(*p == old){
// *p = new;
// return 1;
// }else
// return 0;
TEXT ·Casp1(SB), NOSPLIT, $0-13
MOVL ptr+0(FP), BX
MOVL old+4(FP), AX
MOVL new+8(FP), CX
LOCK
CMPXCHGL CX, 0(BX)
SETEQ ret+12(FP)
RET
// uint32 Xadd(uint32 volatile *val, int32 delta)
// Atomically:
// *val += delta;
// return *val;
TEXT ·Xadd(SB), NOSPLIT, $0-12
MOVL ptr+0(FP), BX
MOVL delta+4(FP), AX
MOVL AX, CX
LOCK
XADDL AX, 0(BX)
ADDL CX, AX
MOVL AX, ret+8(FP)
RET
TEXT ·Xadd64(SB), NOSPLIT, $0-20
NO_LOCAL_POINTERS
// no XADDQ so use CMPXCHG8B loop
MOVL ptr+0(FP), BP
TESTL $7, BP
JZ 2(PC)
CALL ·panicUnaligned(SB)
// DI:SI = delta
MOVL delta_lo+4(FP), SI
MOVL delta_hi+8(FP), DI
// DX:AX = *addr
MOVL 0(BP), AX
MOVL 4(BP), DX
addloop:
// CX:BX = DX:AX (*addr) + DI:SI (delta)
MOVL AX, BX
MOVL DX, CX
ADDL SI, BX
ADCL DI, CX
// if *addr == DX:AX {
// *addr = CX:BX
// } else {
// DX:AX = *addr
// }
// all in one instruction
LOCK
CMPXCHG8B 0(BP)
JNZ addloop
// success
// return CX:BX
MOVL BX, ret_lo+12(FP)
MOVL CX, ret_hi+16(FP)
RET
TEXT ·Xchg(SB), NOSPLIT, $0-12
MOVL ptr+0(FP), BX
MOVL new+4(FP), AX
XCHGL AX, 0(BX)
MOVL AX, ret+8(FP)
RET
TEXT ·Xchgint32(SB), NOSPLIT, $0-12
JMP ·Xchg(SB)
TEXT ·Xchgint64(SB), NOSPLIT, $0-20
JMP ·Xchg64(SB)
TEXT ·Xchguintptr(SB), NOSPLIT, $0-12
JMP ·Xchg(SB)
TEXT ·Xchg64(SB),NOSPLIT,$0-20
NO_LOCAL_POINTERS
// no XCHGQ so use CMPXCHG8B loop
MOVL ptr+0(FP), BP
TESTL $7, BP
JZ 2(PC)
CALL ·panicUnaligned(SB)
// CX:BX = new
MOVL new_lo+4(FP), BX
MOVL new_hi+8(FP), CX
// DX:AX = *addr
MOVL 0(BP), AX
MOVL 4(BP), DX
swaploop:
// if *addr == DX:AX
// *addr = CX:BX
// else
// DX:AX = *addr
// all in one instruction
LOCK
CMPXCHG8B 0(BP)
JNZ swaploop
// success
// return DX:AX
MOVL AX, ret_lo+12(FP)
MOVL DX, ret_hi+16(FP)
RET
TEXT ·StorepNoWB(SB), NOSPLIT, $0-8
MOVL ptr+0(FP), BX
MOVL val+4(FP), AX
XCHGL AX, 0(BX)
RET
TEXT ·Store(SB), NOSPLIT, $0-8
MOVL ptr+0(FP), BX
MOVL val+4(FP), AX
XCHGL AX, 0(BX)
RET
TEXT ·StoreRel(SB), NOSPLIT, $0-8
JMP ·Store(SB)
TEXT ·StoreReluintptr(SB), NOSPLIT, $0-8
JMP ·Store(SB)
// uint64 atomicload64(uint64 volatile* addr);
TEXT ·Load64(SB), NOSPLIT, $0-12
NO_LOCAL_POINTERS
MOVL ptr+0(FP), AX
TESTL $7, AX
JZ 2(PC)
CALL ·panicUnaligned(SB)
MOVQ (AX), M0
MOVQ M0, ret+4(FP)
EMMS
RET
// void ·Store64(uint64 volatile* addr, uint64 v);
TEXT ·Store64(SB), NOSPLIT, $0-12
NO_LOCAL_POINTERS
MOVL ptr+0(FP), AX
TESTL $7, AX
JZ 2(PC)
CALL ·panicUnaligned(SB)
// MOVQ and EMMS were introduced on the Pentium MMX.
MOVQ val+4(FP), M0
MOVQ M0, (AX)
EMMS
// This is essentially a no-op, but it provides required memory fencing.
// It can be replaced with MFENCE, but MFENCE was introduced only on the Pentium4 (SSE2).
XORL AX, AX
LOCK
XADDL AX, (SP)
RET
// void ·Or8(byte volatile*, byte);
TEXT ·Or8(SB), NOSPLIT, $0-5
MOVL ptr+0(FP), AX
MOVB val+4(FP), BX
LOCK
ORB BX, (AX)
RET
// void ·And8(byte volatile*, byte);
TEXT ·And8(SB), NOSPLIT, $0-5
MOVL ptr+0(FP), AX
MOVB val+4(FP), BX
LOCK
ANDB BX, (AX)
RET
TEXT ·Store8(SB), NOSPLIT, $0-5
MOVL ptr+0(FP), BX
MOVB val+4(FP), AX
XCHGB AX, 0(BX)
RET
// func Or(addr *uint32, v uint32)
TEXT ·Or(SB), NOSPLIT, $0-8
MOVL ptr+0(FP), AX
MOVL val+4(FP), BX
LOCK
ORL BX, (AX)
RET
// func And(addr *uint32, v uint32)
TEXT ·And(SB), NOSPLIT, $0-8
MOVL ptr+0(FP), AX
MOVL val+4(FP), BX
LOCK
ANDL BX, (AX)
RET
|