| Lock.java |
1 /*
2 * %W% %E%
3 *
4 * Copyright (c) 2006, Oracle and/or its affiliates. All rights reserved.
5 * ORACLE PROPRIETARY/CONFIDENTIAL. Use is subject to license terms.
6 */
7
8 package java.util.concurrent.locks;
9 import java.util.concurrent.TimeUnit;
10
11 /**
12 * {@code Lock} implementations provide more extensive locking
13 * operations than can be obtained using {@code synchronized} methods
14 * and statements. They allow more flexible structuring, may have
15 * quite different properties, and may support multiple associated
16 * {@link Condition} objects.
17 *
18 * <p>A lock is a tool for controlling access to a shared resource by
19 * multiple threads. Commonly, a lock provides exclusive access to a
20 * shared resource: only one thread at a time can acquire the lock and
21 * all access to the shared resource requires that the lock be
22 * acquired first. However, some locks may allow concurrent access to
23 * a shared resource, such as the read lock of a {@link ReadWriteLock}.
24 *
25 * <p>The use of {@code synchronized} methods or statements provides
26 * access to the implicit monitor lock associated with every object, but
27 * forces all lock acquisition and release to occur in a block-structured way:
28 * when multiple locks are acquired they must be released in the opposite
29 * order, and all locks must be released in the same lexical scope in which
30 * they were acquired.
31 *
32 * <p>While the scoping mechanism for {@code synchronized} methods
33 * and statements makes it much easier to program with monitor locks,
34 * and helps avoid many common programming errors involving locks,
35 * there are occasions where you need to work with locks in a more
36 * flexible way. For example, some algorithms for traversing
37 * concurrently accessed data structures require the use of
38 * "hand-over-hand" or "chain locking": you
39 * acquire the lock of node A, then node B, then release A and acquire
40 * C, then release B and acquire D and so on. Implementations of the
41 * {@code Lock} interface enable the use of such techniques by
42 * allowing a lock to be acquired and released in different scopes,
43 * and allowing multiple locks to be acquired and released in any
44 * order.
45 *
46 * <p>With this increased flexibility comes additional
47 * responsibility. The absence of block-structured locking removes the
48 * automatic release of locks that occurs with {@code synchronized}
49 * methods and statements. In most cases, the following idiom
50 * should be used:
51 *
52 * <pre><tt> Lock l = ...;
53 * l.lock();
54 * try {
55 * // access the resource protected by this lock
56 * } finally {
57 * l.unlock();
58 * }
59 * </tt></pre>
60 *
61 * When locking and unlocking occur in different scopes, care must be
62 * taken to ensure that all code that is executed while the lock is
63 * held is protected by try-finally or try-catch to ensure that the
64 * lock is released when necessary.
65 *
66 * <p>{@code Lock} implementations provide additional functionality
67 * over the use of {@code synchronized} methods and statements by
68 * providing a non-blocking attempt to acquire a lock ({@link
69 * #tryLock()}), an attempt to acquire the lock that can be
70 * interrupted ({@link #lockInterruptibly}, and an attempt to acquire
71 * the lock that can timeout ({@link #tryLock(long, TimeUnit)}).
72 *
73 * <p>A {@code Lock} class can also provide behavior and semantics
74 * that is quite different from that of the implicit monitor lock,
75 * such as guaranteed ordering, non-reentrant usage, or deadlock
76 * detection. If an implementation provides such specialized semantics
77 * then the implementation must document those semantics.
78 *
79 * <p>Note that {@code Lock} instances are just normal objects and can
80 * themselves be used as the target in a {@code synchronized} statement.
81 * Acquiring the
82 * monitor lock of a {@code Lock} instance has no specified relationship
83 * with invoking any of the {@link #lock} methods of that instance.
84 * It is recommended that to avoid confusion you never use {@code Lock}
85 * instances in this way, except within their own implementation.
86 *
87 * <p>Except where noted, passing a {@code null} value for any
88 * parameter will result in a {@link NullPointerException} being
89 * thrown.
90 *
91 * <h3>Memory Synchronization</h3>
92 *
93 * <p>All {@code Lock} implementations <em>must</em> enforce the same
94 * memory synchronization semantics as provided by the built-in monitor
95 * lock, as described in <a href="http://java.sun.com/docs/books/jls/">
96 * The Java Language Specification, Third Edition (17.4 Memory Model)</a>:
97 * <ul>
98 * <li>A successful {@code lock} operation has the same memory
99 * synchronization effects as a successful <em>Lock</em> action.
100 * <li>A successful {@code unlock} operation has the same
101 * memory synchronization effects as a successful <em>Unlock</em> action.
102 * </ul>
103 *
104 * Unsuccessful locking and unlocking operations, and reentrant
105 * locking/unlocking operations, do not require any memory
106 * synchronization effects.
107 *
108 * <h3>Implementation Considerations</h3>
109 *
110 * <p> The three forms of lock acquisition (interruptible,
111 * non-interruptible, and timed) may differ in their performance
112 * characteristics, ordering guarantees, or other implementation
113 * qualities. Further, the ability to interrupt the <em>ongoing</em>
114 * acquisition of a lock may not be available in a given {@code Lock}
115 * class. Consequently, an implementation is not required to define
116 * exactly the same guarantees or semantics for all three forms of
117 * lock acquisition, nor is it required to support interruption of an
118 * ongoing lock acquisition. An implementation is required to clearly
119 * document the semantics and guarantees provided by each of the
120 * locking methods. It must also obey the interruption semantics as
121 * defined in this interface, to the extent that interruption of lock
122 * acquisition is supported: which is either totally, or only on
123 * method entry.
124 *
125 * <p>As interruption generally implies cancellation, and checks for
126 * interruption are often infrequent, an implementation can favor responding
127 * to an interrupt over normal method return. This is true even if it can be
128 * shown that the interrupt occurred after another action may have unblocked
129 * the thread. An implementation should document this behavior.
130 *
131 * @see ReentrantLock
132 * @see Condition
133 * @see ReadWriteLock
134 *
135 * @since 1.5
136 * @author Doug Lea
137 */
138 public interface Lock {
139
140 /**
141 * Acquires the lock.
142 *
143 * <p>If the lock is not available then the current thread becomes
144 * disabled for thread scheduling purposes and lies dormant until the
145 * lock has been acquired.
146 *
147 * <p><b>Implementation Considerations</b>
148 *
149 * <p>A {@code Lock} implementation may be able to detect erroneous use
150 * of the lock, such as an invocation that would cause deadlock, and
151 * may throw an (unchecked) exception in such circumstances. The
152 * circumstances and the exception type must be documented by that
153 * {@code Lock} implementation.
154 */
155 void lock();
156
157 /**
158 * Acquires the lock unless the current thread is
159 * {@linkplain Thread#interrupt interrupted}.
160 *
161 * <p>Acquires the lock if it is available and returns immediately.
162 *
163 * <p>If the lock is not available then the current thread becomes
164 * disabled for thread scheduling purposes and lies dormant until
165 * one of two things happens:
166 *
167 * <ul>
168 * <li>The lock is acquired by the current thread; or
169 * <li>Some other thread {@linkplain Thread#interrupt interrupts} the
170 * current thread, and interruption of lock acquisition is supported.
171 * </ul>
172 *
173 * <p>If the current thread:
174 * <ul>
175 * <li>has its interrupted status set on entry to this method; or
176 * <li>is {@linkplain Thread#interrupt interrupted} while acquiring the
177 * lock, and interruption of lock acquisition is supported,
178 * </ul>
179 * then {@link InterruptedException} is thrown and the current thread's
180 * interrupted status is cleared.
181 *
182 * <p><b>Implementation Considerations</b>
183 *
184 * <p>The ability to interrupt a lock acquisition in some
185 * implementations may not be possible, and if possible may be an
186 * expensive operation. The programmer should be aware that this
187 * may be the case. An implementation should document when this is
188 * the case.
189 *
190 * <p>An implementation can favor responding to an interrupt over
191 * normal method return.
192 *
193 * <p>A {@code Lock} implementation may be able to detect
194 * erroneous use of the lock, such as an invocation that would
195 * cause deadlock, and may throw an (unchecked) exception in such
196 * circumstances. The circumstances and the exception type must
197 * be documented by that {@code Lock} implementation.
198 *
199 * @throws InterruptedException if the current thread is
200 * interrupted while acquiring the lock (and interruption
201 * of lock acquisition is supported).
202 */
203 void lockInterruptibly() throws InterruptedException;
204
205 /**
206 * Acquires the lock only if it is free at the time of invocation.
207 *
208 * <p>Acquires the lock if it is available and returns immediately
209 * with the value {@code true}.
210 * If the lock is not available then this method will return
211 * immediately with the value {@code false}.
212 *
213 * <p>A typical usage idiom for this method would be:
214 * <pre>
215 * Lock lock = ...;
216 * if (lock.tryLock()) {
217 * try {
218 * // manipulate protected state
219 * } finally {
220 * lock.unlock();
221 * }
222 * } else {
223 * // perform alternative actions
224 * }
225 * </pre>
226 * This usage ensures that the lock is unlocked if it was acquired, and
227 * doesn't try to unlock if the lock was not acquired.
228 *
229 * @return {@code true} if the lock was acquired and
230 * {@code false} otherwise
231 */
232 boolean tryLock();
233
234 /**
235 * Acquires the lock if it is free within the given waiting time and the
236 * current thread has not been {@linkplain Thread#interrupt interrupted}.
237 *
238 * <p>If the lock is available this method returns immediately
239 * with the value {@code true}.
240 * If the lock is not available then
241 * the current thread becomes disabled for thread scheduling
242 * purposes and lies dormant until one of three things happens:
243 * <ul>
244 * <li>The lock is acquired by the current thread; or
245 * <li>Some other thread {@linkplain Thread#interrupt interrupts} the
246 * current thread, and interruption of lock acquisition is supported; or
247 * <li>The specified waiting time elapses
248 * </ul>
249 *
250 * <p>If the lock is acquired then the value {@code true} is returned.
251 *
252 * <p>If the current thread:
253 * <ul>
254 * <li>has its interrupted status set on entry to this method; or
255 * <li>is {@linkplain Thread#interrupt interrupted} while acquiring
256 * the lock, and interruption of lock acquisition is supported,
257 * </ul>
258 * then {@link InterruptedException} is thrown and the current thread's
259 * interrupted status is cleared.
260 *
261 * <p>If the specified waiting time elapses then the value {@code false}
262 * is returned.
263 * If the time is
264 * less than or equal to zero, the method will not wait at all.
265 *
266 * <p><b>Implementation Considerations</b>
267 *
268 * <p>The ability to interrupt a lock acquisition in some implementations
269 * may not be possible, and if possible may
270 * be an expensive operation.
271 * The programmer should be aware that this may be the case. An
272 * implementation should document when this is the case.
273 *
274 * <p>An implementation can favor responding to an interrupt over normal
275 * method return, or reporting a timeout.
276 *
277 * <p>A {@code Lock} implementation may be able to detect
278 * erroneous use of the lock, such as an invocation that would cause
279 * deadlock, and may throw an (unchecked) exception in such circumstances.
280 * The circumstances and the exception type must be documented by that
281 * {@code Lock} implementation.
282 *
283 * @param time the maximum time to wait for the lock
284 * @param unit the time unit of the {@code time} argument
285 * @return {@code true} if the lock was acquired and {@code false}
286 * if the waiting time elapsed before the lock was acquired
287 *
288 * @throws InterruptedException if the current thread is interrupted
289 * while acquiring the lock (and interruption of lock
290 * acquisition is supported)
291 */
292 boolean tryLock(long time, TimeUnit unit) throws InterruptedException;
293
294 /**
295 * Releases the lock.
296 *
297 * <p><b>Implementation Considerations</b>
298 *
299 * <p>A {@code Lock} implementation will usually impose
300 * restrictions on which thread can release a lock (typically only the
301 * holder of the lock can release it) and may throw
302 * an (unchecked) exception if the restriction is violated.
303 * Any restrictions and the exception
304 * type must be documented by that {@code Lock} implementation.
305 */
306 void unlock();
307
308 /**
309 * Returns a new {@link Condition} instance that is bound to this
310 * {@code Lock} instance.
311 *
312 * <p>Before waiting on the condition the lock must be held by the
313 * current thread.
314 * A call to {@link Condition#await()} will atomically release the lock
315 * before waiting and re-acquire the lock before the wait returns.
316 *
317 * <p><b>Implementation Considerations</b>
318 *
319 * <p>The exact operation of the {@link Condition} instance depends on
320 * the {@code Lock} implementation and must be documented by that
321 * implementation.
322 *
323 * @return A new {@link Condition} instance for this {@code Lock} instance
324 * @throws UnsupportedOperationException if this {@code Lock}
325 * implementation does not support conditions
326 */
327 Condition newCondition();
328 }
329