1 .\" Copyright (c) 2000-2001 John H. Baldwin <jhb@FreeBSD.org>
2 .\" All rights reserved.
4 .\" Redistribution and use in source and binary forms, with or without
5 .\" modification, are permitted provided that the following conditions
7 .\" 1. Redistributions of source code must retain the above copyright
8 .\" notice, this list of conditions and the following disclaimer.
9 .\" 2. Redistributions in binary form must reproduce the above copyright
10 .\" notice, this list of conditions and the following disclaimer in the
11 .\" documentation and/or other materials provided with the distribution.
13 .\" THIS SOFTWARE IS PROVIDED BY THE DEVELOPERS ``AS IS'' AND ANY EXPRESS OR
14 .\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
15 .\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
16 .\" IN NO EVENT SHALL THE DEVELOPERS BE LIABLE FOR ANY DIRECT, INDIRECT,
17 .\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
18 .\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
19 .\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
20 .\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
21 .\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
22 .\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
24 .\" $FreeBSD: src/share/man/man9/atomic.9,v 1.17 2010/05/27 13:56:27 uqs Exp $
35 .Nm atomic_readandclear ,
44 .Fn atomic_add_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
46 .Fn atomic_clear_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
48 .Fo atomic_cmpset_[acq_|rel_]<type>
49 .Fa "volatile <type> *dst"
54 .Fn atomic_fetchadd_<type> "volatile <type> *p" "<type> v"
56 .Fn atomic_load_acq_<type> "volatile <type> *p"
58 .Fn atomic_readandclear_<type> "volatile <type> *p"
60 .Fn atomic_set_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
62 .Fn atomic_subtract_[acq_|rel_]<type> "volatile <type> *p" "<type> v"
64 .Fn atomic_store_rel_<type> "volatile <type> *p" "<type> v"
66 Each of the atomic operations is guaranteed to be atomic in the presence of
68 They can be used to implement reference counts or as building blocks for more
69 advanced synchronization primitives such as mutexes.
71 Each atomic operation operates on a specific
73 The type to use is indicated in the function name.
74 The available types that can be used are:
76 .Bl -tag -offset indent -width short -compact
84 unsigned integer the size of a pointer
86 unsigned 32-bit integer
88 .\"unsigned 64-bit integer
91 For example, the function to atomically add two integers is called
94 Certain architectures also provide operations for types smaller than
97 .Bl -tag -offset indent -width short -compact
101 unsigned short integer
103 unsigned 8-bit integer
105 unsigned 16-bit integer
108 These must not be used in MI code because the instructions to implement them
109 efficiently may not be available.
111 Memory barriers are used to guarantee the order of data accesses in
113 First, they specify hints to the compiler to not re-order or optimize the
115 Second, on architectures that do not guarantee ordered data accesses,
116 special instructions or special variants of instructions are used to indicate
117 to the processor that data accesses need to occur in a certain order.
118 As a result, most of the atomic operations have three variants in order to
119 include optional memory barriers.
120 The first form just performs the operation without any explicit barriers.
121 The second form uses a read memory barrier, and the third variant uses a write
124 The second variant of each operation includes a read memory barrier.
125 This barrier ensures that the effects of this operation are completed before the
126 effects of any later data accesses.
127 As a result, the operation is said to have acquire semantics as it acquires a
128 pseudo-lock requiring further operations to wait until it has completed.
129 To denote this, the suffix
131 is inserted into the function name immediately prior to the
132 .Dq Li _ Ns Aq Fa type
134 For example, to subtract two integers ensuring that any later writes will
135 happen after the subtraction is performed, use
136 .Fn atomic_subtract_acq_int .
138 The third variant of each operation includes a write memory barrier.
139 This ensures that all effects of all previous data accesses are completed
140 before this operation takes place.
141 As a result, the operation is said to have release semantics as it releases
142 any pending data accesses to be completed before its operation is performed.
143 To denote this, the suffix
145 is inserted into the function name immediately prior to the
146 .Dq Li _ Ns Aq Fa type
148 For example, to add two long integers ensuring that all previous
149 writes will happen first, use
150 .Fn atomic_add_rel_long .
152 A practical example of using memory barriers is to ensure that data accesses
153 that are protected by a lock are all performed while the lock is held.
154 To achieve this, one would use a read barrier when acquiring the lock to
155 guarantee that the lock is held before any protected operations are performed.
156 Finally, one would use a write barrier when releasing the lock to ensure that
157 all of the protected operations are completed before the lock is released.
158 .Ss Multiple Processors
159 The current set of atomic operations do not necessarily guarantee atomicity
160 across multiple processors.
161 To guarantee atomicity across processors, not only does the individual
162 operation need to be atomic on the processor performing the operation, but
163 the result of the operation needs to be pushed out to stable storage and the
164 caches of all other processors on the system need to invalidate any cache
165 lines that include the affected memory region.
168 architecture, the cache coherency model requires that the hardware perform
169 this task, thus the atomic operations are atomic across multiple processors.
172 .\"architecture, coherency is only guaranteed for pages that are configured to
173 .\"using a caching policy of either uncached or write back.
175 This section describes the semantics of each operation using a C like notation.
177 .It Fn atomic_add p v
178 .Bd -literal -compact
185 functions are not implemented for the type
188 .It Fn atomic_clear p v
189 .Bd -literal -compact
192 .It Fn atomic_cmpset dst old new
193 .Bd -literal -compact
204 functions are not implemented for the types
211 .It Fn atomic_fetchadd p v
212 .Bd -literal -compact
221 functions are only implemented for the types
226 and do not have any variants with memory barriers at this time.
228 .It Fn atomic_load addr
229 .Bd -literal -compact
236 functions are only provided with acquire memory barriers.
238 .It Fn atomic_readandclear addr
239 .Bd -literal -compact
247 .Fn atomic_readandclear
248 functions are not implemented for the types
257 not have any variants with memory barriers at this time.
259 .It Fn atomic_set p v
260 .Bd -literal -compact
263 .It Fn atomic_subtract p v
264 .Bd -literal -compact
271 functions are not implemented for the type
274 .It Fn atomic_store p v
275 .Bd -literal -compact
282 functions are only provided with release memory barriers.
286 .\"is currently not implemented for any of the atomic operations on the
296 returns the result of the compare operation.
298 .Fn atomic_fetchadd ,
301 .Fn atomic_readandclear
303 return the value at the specified address.
305 .\"This example uses the
306 .\".Fn atomic_cmpset_acq_ptr
308 .\".Fn atomic_set_ptr
309 .\"functions to obtain a sleep mutex and handle recursion.
318 .\"/* Try to obtain mtx_lock once. */
319 .\"#define _obtain_lock(mp, tid) \\
320 .\" atomic_cmpset_acq_ptr(&(mp)->mtx_lock, MTX_UNOWNED, (tid))
322 .\"/* Get a sleep lock, deal with recursion inline. */
323 .\"#define _get_sleep_lock(mp, tid, opts, file, line) do { \\
324 .\" uintptr_t _tid = (uintptr_t)(tid); \\
326 .\" if (!_obtain_lock(mp, tid)) { \\
327 .\" if (((mp)->mtx_lock & MTX_FLAGMASK) != _tid) \\
328 .\" _mtx_lock_sleep((mp), _tid, (opts), (file), (line));\\
330 .\" atomic_set_ptr(&(mp)->mtx_lock, MTX_RECURSE); \\
331 .\" (mp)->mtx_recurse++; \\
343 operations were first introduced in
345 This first set only supported the types
354 .Fn atomic_readandclear ,
357 operations were added in
366 and all of the acquire and release variants
372 operations were added in