Commit | Line | Data |
---|---|---|
7ac06cef | 1 | Userspace RCU Implementation |
c97ae6eb | 2 | by Mathieu Desnoyers and Paul E. McKenney |
6991f61a | 3 | |
c97ae6eb PMF |
4 | BUILDING |
5 | -------- | |
6991f61a | 6 | |
48d848c7 PMF |
7 | ./bootstrap (skip if using tarball) |
8 | ./configure | |
c97ae6eb PMF |
9 | make |
10 | make install | |
e197ac6f | 11 | ldconfig |
9ca52251 | 12 | |
7d413817 | 13 | Hints: Forcing 32-bit build: |
c4c18179 | 14 | * CFLAGS="-m32 -g -O2" ./configure |
9ca52251 MD |
15 | |
16 | Forcing 64-bit build: | |
c4c18179 | 17 | * CFLAGS="-m64 -g -O2" ./configure |
aa8c36e0 | 18 | |
f39cd442 | 19 | Forcing a 32-bit build with 386 backward compatibility: |
c5b9f8ff | 20 | * CFLAGS="-m32 -g -O2" ./configure --host=i386-pc-linux-gnu |
7d413817 | 21 | |
795d506a MD |
22 | Forcing a 32-bit build for Sparcv9 (typical for Sparc v9) |
23 | * CFLAGS="-m32 -Wa,-Av9a -g -O2" ./configure | |
24 | ||
c51e75e6 MD |
25 | ARCHITECTURES SUPPORTED |
26 | ----------------------- | |
27 | ||
ac6454bc JW |
28 | Currently, x86 (i386, i486, i586, i686), x86 64-bit, PowerPC 32/64, S390, S390x, |
29 | ARM, Alpha, ia64 and Sparcv9 32/64 are supported. Only tested on Linux so | |
7bcbcfb2 | 30 | far, but should theoretically work on other operating systems. |
c51e75e6 | 31 | |
ac6454bc | 32 | ARM depends on running a Linux kernel 2.6.15 or better. |
3b36a2e9 | 33 | |
3b38cfe1 MD |
34 | The gcc compiler versions 3.3, 3.4, 4.0, 4.1, 4.2, 4.3, 4.4 and 4.5 are |
35 | supported, with the following exceptions: | |
36 | ||
37 | - gcc 3.3 and 3.4 have a bug that prevents them from generating volatile | |
38 | accesses to offsets in a TLS structure on 32-bit x86. These versions are | |
39 | therefore not compatible with liburcu on x86 32-bit (i386, i486, i586, i686). | |
40 | The problem has been reported to the gcc community: | |
41 | http://www.mail-archive.com/gcc-bugs@gcc.gnu.org/msg281255.html | |
83e8be52 MD |
42 | - gcc 3.3 cannot match the "xchg" instruction on 32-bit x86 build. |
43 | See: http://kerneltrap.org/node/7507 | |
ac6454bc | 44 | - Alpha, ia64 and ARM architectures depend on 4.x gcc with atomic builtins |
7bcbcfb2 MD |
45 | support. |
46 | ||
3b38cfe1 | 47 | |
c97ae6eb PMF |
48 | QUICK START GUIDE |
49 | ----------------- | |
aa8c36e0 | 50 | |
0a1d290b MD |
51 | Usage of all urcu libraries |
52 | ||
53 | * Define _LGPL_SOURCE (only) if your code is LGPL or GPL compatible | |
54 | before including the urcu.h or urcu-qsbr.h header. If your application | |
55 | is distributed under another license, function calls will be generated | |
56 | instead of inlines, so your application can link with the library. | |
57 | * Linking with one of the libraries below is always necessary even for | |
58 | LGPL and GPL applications. | |
59 | ||
60 | Usage of liburcu | |
61 | ||
62 | * #include <urcu.h> | |
63 | * Link the application with "-lurcu". | |
fdf01eed MD |
64 | * This is the preferred version of the library, in terms of |
65 | grace-period detection speed, read-side speed and flexibility. | |
66 | Dynamically detects kernel support for sys_membarrier(). Falls back | |
67 | on urcu-mb scheme if support is not present, which has slower | |
68 | read-side. | |
0a1d290b MD |
69 | |
70 | Usage of liburcu-qsbr | |
71 | ||
72 | * #include <urcu-qsbr.h> | |
73 | * Link with "-lurcu-qsbr". | |
74 | * The QSBR flavor of RCU needs to have each reader thread executing | |
75 | rcu_quiescent_state() periodically to progress. rcu_thread_online() | |
76 | and rcu_thread_offline() can be used to mark long periods for which | |
77 | the threads are not active. It provides the fastest read-side at the | |
78 | expense of more intrusiveness in the application code. | |
79 | ||
fdf01eed MD |
80 | Usage of liburcu-mb |
81 | ||
82 | * #include <urcu.h> | |
83 | * Compile any _LGPL_SOURCE code using this library with "-DRCU_MB". | |
84 | * Link with "-lurcu-mb". | |
85 | * This version of the urcu library uses memory barriers on the writer | |
86 | and reader sides. This results in faster grace-period detection, but | |
87 | results in slower reads. | |
88 | ||
89 | Usage of liburcu-signal | |
90 | ||
ee39cfb6 MD |
91 | * #include <urcu.h> |
92 | * Compile any _LGPL_SOURCE code using this library with "-DRCU_SIGNAL". | |
fdf01eed MD |
93 | * Link the application with "-lurcu-signal". |
94 | * Version of the library that requires a signal, typically SIGUSR1. Can | |
95 | be overridden with -DSIGRCU by modifying Makefile.build.inc. | |
96 | ||
fdee2e6d MD |
97 | Usage of liburcu-bp |
98 | ||
99 | * #include <urcu-bp.h> | |
100 | * Link with "-lurcu-bp". | |
101 | * The BP library flavor stands for "bulletproof". It is specifically | |
102 | designed to help tracing library to hook on applications without | |
02be5561 | 103 | requiring to modify these applications. rcu_init(), |
fdee2e6d MD |
104 | rcu_register_thread() and rcu_unregister_thread() all become nops. |
105 | The state is dealt with by the library internally at the expense of | |
106 | read-side and write-side performance. | |
107 | ||
c97ae6eb PMF |
108 | Initialization |
109 | ||
110 | Each thread that has reader critical sections (that uses | |
111 | rcu_read_lock()/rcu_read_unlock() must first register to the URCU | |
4c1471de MD |
112 | library. This is done by calling rcu_register_thread(). Unregistration |
113 | must be performed before exiting the thread by using | |
114 | rcu_unregister_thread(). | |
c97ae6eb PMF |
115 | |
116 | Reading | |
117 | ||
118 | Reader critical sections must be protected by locating them between | |
119 | calls to rcu_read_lock() and rcu_read_unlock(). Inside that lock, | |
120 | rcu_dereference() may be called to read an RCU protected pointer. | |
121 | ||
122 | Writing | |
123 | ||
124 | rcu_assign_pointer() and rcu_xchg_pointer() may be called anywhere. | |
9fb223da MD |
125 | After, synchronize_rcu() must be called. When it returns, the old |
126 | values are not in usage anymore. | |
c97ae6eb | 127 | |
ec4e58a3 MD |
128 | Usage of liburcu-defer |
129 | ||
0376e7b2 PM |
130 | * Follow instructions for either liburcu, liburcu-qsbr, |
131 | liburcu-mb, liburcu-signal, or liburcu-bp above. | |
132 | The liburcu-defer functionality is pulled into each of | |
133 | those library modules. | |
632dd6ba | 134 | * Provides defer_rcu() primitive to enqueue delayed callbacks. Queued |
ec4e58a3 | 135 | callbacks are executed in batch periodically after a grace period. |
632dd6ba | 136 | Do _not_ use defer_rcu() within a read-side critical section, because |
ec4e58a3 | 137 | it may call synchronize_rcu() if the thread queue is full. |
0376e7b2 | 138 | This can lead to deadlock or worse. |
83dd659a MD |
139 | * Requires that rcu_defer_barrier() must be called in library destructor |
140 | if a library queues callbacks and is expected to be unloaded with | |
141 | dlclose(). | |
9c55af9f MD |
142 | * Its API is currently experimental. It may change in future library |
143 | releases. | |
ec4e58a3 | 144 | |
26ba798a PM |
145 | Usage of urcu-call-rcu |
146 | ||
147 | * Follow instructions for either liburcu, liburcu-qsbr, | |
148 | liburcu-mb, liburcu-signal, or liburcu-bp above. | |
149 | The urcu-call-rcu functionality is provided for each of | |
150 | these library modules. | |
151 | * Provides the call_rcu() primitive to enqueue delayed callbacks | |
152 | in a manner similar to defer_rcu(), but without ever delaying | |
153 | for a grace period. On the other hand, call_rcu()'s best-case | |
154 | overhead is not quite as good as that of defer_rcu(). | |
155 | * Provides call_rcu() to allow asynchronous handling of RCU | |
156 | grace periods. A number of additional functions are provided | |
157 | to manage the helper threads used by call_rcu(), but reasonable | |
158 | defaults are used if these additional functions are not invoked. | |
159 | See API.txt for more details. | |
160 | ||
dd052bd3 PMF |
161 | Being careful with signals |
162 | ||
0a1d290b | 163 | The liburcu library uses signals internally. The signal handler is |
dd052bd3 PMF |
164 | registered with the SA_RESTART flag. However, these signals may cause |
165 | some non-restartable system calls to fail with errno = EINTR. Care | |
166 | should be taken to restart system calls manually if they fail with this | |
167 | error. A list of non-restartable system calls may be found in | |
0a1d290b MD |
168 | signal(7). The liburcu-mb and liburcu-qsbr versions of the Userspace RCU |
169 | library do not require any signal. | |
c97ae6eb | 170 | |
0a1d290b | 171 | Read-side critical sections are allowed in a signal handler with |
7ac06cef MD |
172 | liburcu and liburcu-mb. Be careful, however, to disable these signals |
173 | between thread creation and calls to rcu_register_thread(), because a | |
174 | signal handler nesting on an unregistered thread would not be allowed to | |
175 | call rcu_read_lock(). | |
cee02f0a | 176 | |
0a1d290b MD |
177 | Read-side critical sections are _not_ allowed in a signal handler with |
178 | liburcu-qsbr, unless signals are disabled explicitly around each | |
179 | rcu_quiescent_state() calls, when threads are put offline and around | |
180 | calls to synchronize_rcu(). Even then, we do not recommend it. | |
c97ae6eb | 181 | |
955f5e52 MD |
182 | Interaction with mutexes |
183 | ||
184 | One must be careful to do not cause deadlocks due to interaction of | |
185 | synchronize_rcu() and RCU read-side with mutexes. If synchronize_rcu() | |
186 | is called with a mutex held, this mutex (or any mutex which has this | |
187 | mutex in its dependency chain) should not be acquired from within a RCU | |
188 | read-side critical section. | |
189 | ||
30890270 MD |
190 | This is especially important to understand in the context of the |
191 | QSBR flavor: a registered reader thread being "online" by | |
192 | default should be considered as within a RCU read-side critical | |
193 | section unless explicitly put "offline". Therefore, if | |
194 | synchronize_rcu() is called with a mutex held, this mutex, as | |
195 | well as any mutex which has this mutex in its dependency chain | |
196 | should only be taken when the RCU reader thread is "offline" | |
197 | (this can be performed by calling rcu_thread_offline()). | |
198 | ||
cee02f0a MD |
199 | Usage of DEBUG_RCU |
200 | ||
201 | DEBUG_RCU is used to add internal debugging self-checks to the | |
0a1d290b | 202 | RCU library. This define adds a performance penalty when enabled. |
fb6e510b MD |
203 | Can be enabled by uncommenting the corresponding line in |
204 | Makefile.build.inc. | |
c97ae6eb PMF |
205 | |
206 | Usage of DEBUG_YIELD | |
207 | ||
208 | DEBUG_YIELD is used to add random delays in the code for testing | |
209 | purposes. | |
7d413817 MD |
210 | |
211 | SMP support | |
212 | ||
213 | By default the library is configured to use synchronization primitives | |
214 | adequate for SMP systems. On uniprocessor systems, support for SMP | |
215 | systems can be disabled with: | |
216 | ||
217 | ./configure --disable-smp-support | |
218 | ||
219 | theoretically yielding slightly better performance. | |
47c5a84f MD |
220 | |
221 | Interaction with fork() | |
222 | ||
223 | Special care must be taken for applications performing fork() without | |
224 | any following exec(). This is caused by the fact that Linux only clones | |
225 | the thread calling fork(), and thus never replicates any of the other | |
226 | parent thread into the child process. Most liburcu implementations | |
227 | require that all registrations (as reader, defer_rcu and call_rcu | |
228 | threads) should be released before a fork() is performed, except for the | |
229 | rather common scenario where fork() is immediately followed by exec() in | |
230 | the child process. The only implementation not subject to that rule is | |
4cf1675f MD |
231 | liburcu-bp, which is designed to handle fork() by calling |
232 | rcu_bp_before_fork, rcu_bp_after_fork_parent and | |
233 | rcu_bp_after_fork_child. | |
81ad2e19 | 234 | |
ef84facf PM |
235 | Applications that use call_rcu() and that fork() without |
236 | doing an immediate exec() must take special action. The parent | |
237 | must invoke call_rcu_before_fork() before the fork() and | |
238 | call_rcu_after_fork_parent() after the fork(). The child | |
239 | process must invoke call_rcu_after_fork_child(). | |
240 | These three APIs are suitable for passing to pthread_atfork(). |