ViewVC Help

View File | Revision Log | Show Annotations | Download File | Root Listing

root/jsr166/jsr166/src/jsr166e/StripedAdder.java

Comparing jsr166/src/jsr166e/StripedAdder.java (file contents):
Revision 1.6 by dl, Tue Jul 26 17:16:36 2011 UTC vs.
Revision 1.10 by dl, Fri Jul 29 13:50:54 2011 UTC

#	Line 5 | Line 5
5		*/
6
7		package jsr166e;
8	–	import java.util.Arrays;
8		import java.util.Random;
9		import java.util.concurrent.atomic.AtomicInteger;
10		import java.util.concurrent.atomic.AtomicLong;
#	Line 27 | Line 26 | import java.io.ObjectOutputStream;
26		* update a common sum that is used for purposes such as collecting
27		* statistics. In this case, performance may be significantly faster
28		* than using a shared {@link AtomicLong}, at the expense of using
29	<	* much more space.  On the other hand, if it is known that only one
30	<	* thread can ever update the sum, performance may be significantly
31	<	* slower than just updating a local variable.
29	>	* more space.  On the other hand, if it is known that only one thread
30	>	* can ever update the sum, performance may be significantly slower
31	>	* than just updating a local variable.
32		*
33		* <p>A StripedAdder may optionally be constructed with a given
34		* expected contention level; i.e., the number of threads that are
#	Line 46 | Line 45 | public class StripedAdder implements Ser
45		* A StripedAdder maintains a table of Atomic long variables. The
46		* table is indexed by per-thread hash codes.
47		*
48	<	* By default, the table is lazily initialized, to minimize
49	<	* footprint until adders are used. On first use, the table is set
50	<	* to size DEFAULT_INITIAL_SIZE (currently 8). Table size is
51	<	* bounded by the number of CPUS (if larger than the default
52	<	* size).
48	>	* Table entries are of class Adder; a variant of AtomicLong
49	>	* padded to reduce cache contention on most processors. Padding
50	>	* is overkill for most Atomics because they are usually
51	>	* irregularly scattered in memory and thus don't interfere much
52	>	* with each other. But Atomic objects residing in arrays will
53	>	* tend to be placed adjacent to each other, and so will most
54	>	* often share cache lines (with a huge negative performance
55	>	* impact) without this precaution.
56	>	*
57	>	* Because Adders are relatively large, we avoid creating them
58	>	* until they are needed. On the other hand, we try to create them
59	>	* on any sign of contention.
60		*
61		* Per-thread hash codes are initialized to random values.
62		* Collisions are indicated by failed CASes when performing an add
#	Line 62 | Line 68 | public class StripedAdder implements Ser
68		* hashing", using a secondary hash (Marsaglia XorShift) to try to
69		* find a free slot.
70		*
71	<	* The table size is capped because, when there are more threads
72	<	* than CPUs, supposing that each thread were bound to a CPU,
73	<	* there would exist a perfect hash function mapping threads to
74	<	* slots that eliminates collisions. When we reach capacity, we
75	<	* search for this mapping by randomly varying the hash codes of
76	<	* colliding threads.  Because search is random, and failures only
77	<	* become known via CAS failures, convergence will be slow, and
78	<	* because threads are typically not bound to CPUS forever, may
79	<	* not occur at all. However, despite these limitations, observed
80	<	* contention is typically low in these cases.
81	<	*
82	<	* Table entries are of class Adder; a form of AtomicLong padded
83	<	* to reduce cache contention on most processors. Padding is
84	<	* overkill for most Atomics because they are usually irregularly
79	<	* scattered in memory and thus don't interfere much with each
80	<	* other. But Atomic objects residing in arrays will tend to be
81	<	* placed adjacent to each other, and so will most often share
82	<	* cache lines without this precaution.  Adders are by default
83	<	* constructed upon first use, which further improves per-thread
84	<	* locality and helps reduce footprint.
71	>	* By default, the table is lazily initialized.  Upon first use,
72	>	* the table is set to size 1, and contains a single Adder. The
73	>	* maximum table size is bounded by nearest power of two >= the
74	>	* number of CPUS.  The table size is capped because, when there
75	>	* are more threads than CPUs, supposing that each thread were
76	>	* bound to a CPU, there would exist a perfect hash function
77	>	* mapping threads to slots that eliminates collisions. When we
78	>	* reach capacity, we search for this mapping by randomly varying
79	>	* the hash codes of colliding threads.  Because search is random,
80	>	* and failures only become known via CAS failures, convergence
81	>	* will be slow, and because threads are typically not bound to
82	>	* CPUS forever, may not occur at all. However, despite these
83	>	* limitations, observed contention is typically low in these
84	>	* cases.
85		*
86		* A single spinlock is used for resizing the table as well as
87	<	* populating slots with new Adders. Upon lock contention, threads
87	>	* populating slots with new Adders. After initialization, there
88	>	* is no need for a blocking lock: Upon lock contention, threads
89		* try other slots rather than blocking. After initialization, at
90		* least one slot exists, so retries will eventually find a
91		* candidate Adder.  During these retries, there is increased
#	Line 92 | Line 93 | public class StripedAdder implements Ser
93		* alternatives.
94		*/
95
95	–	/**
96	–	* Padded version of AtomicLong
97	–	*/
98	–	static final class Adder extends AtomicLong {
99	–	long p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pa, pb, pc, pd, pe;
100	–	Adder(long x) { super(x); }
101	–	}
102	–
96		private static final int NCPU = Runtime.getRuntime().availableProcessors();
97
98		/**
99	<	* Table bounds. DEFAULT_INITIAL_SIZE is the table size set upon
100	<	* first use under default constructor, and must be a power of
101	<	* two. There is not much point in making size a lot smaller than
102	<	* that of Adders though.  CAP is the maximum allowed table size.
103	<	*/
104	<	private static final int DEFAULT_INITIAL_SIZE = 8;
105	<	private static final int CAP = Math.max(NCPU, DEFAULT_INITIAL_SIZE);
99	>	* Padded variant of AtomicLong.  The value field is placed
100	>	* between pads, hoping that the JVM doesn't reorder them.
101	>	* Updates are via inlined CAS in methods add and retryAdd.
102	>	*/
103	>	static final class Adder {
104	>	volatile long p0, p1, p2, p3, p4, p5, p6;
105	>	volatile long value;
106	>	volatile long q0, q1, q2, q3, q4, q5, q6;
107	>	Adder(long x) { value = x; }
108	>	}
109
110		/**
111		* Holder for the thread-local hash code. The code is initially
#	Line 119 | Line 115 | public class StripedAdder implements Ser
115		static final Random rng = new Random();
116		int code;
117		HashCode() {
118	<	int h = rng.nextInt();
119	<	code = (h == 0) ? 1 : h; // ensure nonzero
118	>	int h = rng.nextInt(); // Avoid zero, because of xorShift rehash
119	>	code = (h == 0) ? 1 : h;
120		}
121		}
122
#	Line 140 | Line 136 | public class StripedAdder implements Ser
136		static final ThreadHashCode threadHashCode = new ThreadHashCode();
137
138		/**
139	<	* Table of adders. Size is power of two, grows to be at most CAP.
139	>	* Table of adders. When non-null, size is a power of 2.
140		*/
141		private transient volatile Adder[] adders;
142
143		/**
144	<	* Serves as a lock when resizing and/or creating Adders.  There
149	<	* is no need for a blocking lock: Except during initialization
150	<	* races, when busy, other threads try other slots. However,
151	<	* during (double-checked) initializations, we use the
152	<	* "synchronized" lock on this object.
144	>	* Spinlock (locked via CAS) used when resizing and/or creating Adders.
145		*/
146	<	private final AtomicInteger mutex;
146	>	private volatile int busy;
147
148		/**
149		* Creates a new adder with zero sum.
150		*/
151		public StripedAdder() {
160	–	this.mutex = new AtomicInteger();
161	–	// remaining initialization on first call to add.
152		}
153
154		/**
#	Line 169 | Line 159 | public class StripedAdder implements Ser
159		* will concurrently update the sum.
160		*/
161		public StripedAdder(int expectedContention) {
162	<	int cap = (expectedContention < CAP) ? expectedContention : CAP;
162	>	int cap = (expectedContention < NCPU) ? expectedContention : NCPU;
163		int size = 1;
164		while (size < cap)
165		size <<= 1;
#	Line 177 | Line 167 | public class StripedAdder implements Ser
167		for (int i = 0; i < size; ++i)
168		as[i] = new Adder(0);
169		this.adders = as;
180	–	this.mutex = new AtomicInteger();
170		}
171
172		/**
#	Line 186 | Line 175 | public class StripedAdder implements Ser
175		* @param x the value to add
176		*/
177		public void add(long x) {
178	<	Adder[] as; Adder a; int n; long v; // locals to hold volatile reads
178	>	Adder[] as; Adder a; int n;  // locals to hold volatile reads
179		HashCode hc = threadHashCode.get();
180		int h = hc.code;
181	<	if ((as = adders) == null || (n = as.length) < 1 ||
182	<	(a = as[(n - 1) & h]) == null ||
183	<	!a.compareAndSet(v = a.get(), v + x))
184	<	retryAdd(x, hc);
181	>	boolean contended;
182	>	if ((as = adders) != null && (n = as.length) > 0 &&
183	>	(a = as[(n - 1) & h]) != null) {
184	>	long v = a.value;
185	>	if (UNSAFE.compareAndSwapLong(a, valueOffset, v, v + x))
186	>	return;
187	>	contended = true;
188	>	}
189	>	else
190	>	contended = false;
191	>	retryAdd(x, hc, contended);
192		}
193
194		/**
195		* Handle cases of add involving initialization, resizing,
196		* creating new Adders, and/or contention. See above for
197	<	* explanation.
197	>	* explanation. This method suffers the usual non-modularity
198	>	* problems of optimistic retry code, relying on rechecked sets of
199	>	* reads.
200	>	*
201	>	* @param x the value to add
202	>	* @param hc the hash code holder
203	>	* @param precontended true if CAS failed before call
204		*/
205	<	private void retryAdd(long x, HashCode hc) {
205	>	private void retryAdd(long x, HashCode hc, boolean precontended) {
206		int h = hc.code;
207	<	final AtomicInteger mutex = this.mutex;
206	<	int collisions = 1 - mutex.get(); // first guess: collides if not locked
207	>	boolean collide = false; // true if last slot nonempty
208		for (;;) {
209	<	Adder[] as; Adder a; long v; int k, n;
210	<	while ((as = adders) == null || (n = as.length) < 1) {
211	<	synchronized(mutex) {                // Try to initialize
212	<	if (adders == null) {
213	<	Adder[] rs = new Adder[DEFAULT_INITIAL_SIZE];
214	<	rs[h & (DEFAULT_INITIAL_SIZE - 1)] = new Adder(0);
215	<	adders = rs;
209	>	Adder[] as; Adder a; int n;
210	>	if ((as = adders) != null && (n = as.length) > 0) {
211	>	if ((a = as[(n - 1) & h]) == null) {
212	>	if (busy == 0 &&            // Try to attach new Adder
213	>	UNSAFE.compareAndSwapInt(this, busyOffset, 0, 1)) {
214	>	boolean created = false;
215	>	try {                   // Recheck under lock
216	>	Adder[] rs; int m, j;
217	>	if ((rs = adders) != null && (m = rs.length) > 0 &&
218	>	rs[j = (m - 1) & h] == null) {
219	>	rs[j] = new Adder(x);
220	>	created = true;
221	>	}
222	>	} finally {
223	>	busy = 0;
224	>	}
225	>	if (created)
226	>	break;
227	>	continue;               // Slot is now non-empty
228		}
229	+	collide = false;
230		}
231	<	collisions = 0;
232	<	}
233	<
234	<	if ((a = as[k = (n - 1) & h]) == null) { // Try to add slot
235	<	if (mutex.get() == 0 && mutex.compareAndSet(0, 1)) {
222	<	try {
223	<	if (adders == as && as[k] == null)
224	<	a = as[k] = new Adder(x);
225	<	} finally {
226	<	mutex.set(0);
227	<	}
228	<	if (a != null)
231	>	else if (precontended)          // CAS already known to fail
232	>	precontended = false;       // Continue after rehash
233	>	else {
234	>	long v = a.value;
235	>	if (UNSAFE.compareAndSwapLong(a, valueOffset, v, v + x))
236		break;
237	+	if (!collide)
238	+	collide = true;
239	+	else if (n >= NCPU || adders != as)
240	+	collide = false;        // Don't expand
241	+	else if (busy == 0 &&
242	+	UNSAFE.compareAndSwapInt(this, busyOffset, 0, 1)) {
243	+	collide = false;
244	+	try {
245	+	if (adders == as) { // Expand table
246	+	Adder[] rs = new Adder[n << 1];
247	+	for (int i = 0; i < n; ++i)
248	+	rs[i] = as[i];
249	+	adders = rs;
250	+	}
251	+	} finally {
252	+	busy = 0;
253	+	}
254	+	continue;
255	+	}
256		}
257	<	collisions = 0;
257	>	h ^= h << 13;                   // Rehash
258	>	h ^= h >>> 17;
259	>	h ^= h << 5;
260		}
261	<	else if (collisions != 0 && n < CAP &&   // Try to expand table
262	<	mutex.get() == 0 && mutex.compareAndSet(0, 1)) {
263	<	try {
264	<	if (adders == as) {
265	<	Adder[] rs = new Adder[n << 1];
266	<	for (int i = 0; i < n; ++i)
267	<	rs[i] = as[i];
268	<	adders = rs;
261	>	else if (adders == as) {            // Try to default-initialize
262	>	Adder[] rs = new Adder[1];
263	>	rs[0] = new Adder(x);
264	>	boolean init = false;
265	>	while (adders == as) {
266	>	if (UNSAFE.compareAndSwapInt(this, busyOffset, 0, 1)) {
267	>	try {
268	>	if (adders == as) {
269	>	adders = rs;
270	>	init = true;
271	>	}
272	>	} finally {
273	>	busy = 0;
274	>	}
275	>	break;
276		}
277	<	} finally {
278	<	mutex.set(0);
277	>	if (adders != as)
278	>	break;
279	>	Thread.yield();              // Back off
280		}
281	<	collisions = 0;
281	>	if (init)
282	>	break;
283		}
247	–	else if (a.compareAndSet(v = a.get(), v + x))
248	–	break;
249	–	else
250	–	collisions = 1;
251	–	h ^= h << 13;                            // Rehash
252	–	h ^= h >>> 17;
253	–	h ^= h << 5;
284		}
285	<	hc.code = h;
285	>	hc.code = h;                            // Record index for next time
286	>	}
287	>
288	>	/**
289	>	* Equivalent to {@code add(1)}.
290	>	*/
291	>	public void increment() {
292	>	add(1L);
293	>	}
294	>
295	>	/**
296	>	* Equivalent to {@code add(-1)}.
297	>	*/
298	>	public void decrement() {
299	>	add(-1L);
300		}
301
302		/**
#	Line 270 | Line 314 | public class StripedAdder implements Ser
314		for (int i = 0; i < n; ++i) {
315		Adder a = as[i];
316		if (a != null)
317	<	sum += a.get();
317	>	sum += a.value;
318		}
319		}
320		return sum;
321		}
322
323		/**
324	<	* Resets each of the variables to zero. This is effective in
325	<	* fully resetting the sum only if there are no concurrent
326	<	* updates.
283	<	*/
284	<	public void reset() {
285	<	Adder[] as = adders;
286	<	if (as != null) {
287	<	int n = as.length;
288	<	for (int i = 0; i < n; ++i) {
289	<	Adder a = as[i];
290	<	if (a != null)
291	<	a.set(0L);
292	<	}
293	<	}
294	<	}
295	<
296	<	/**
297	<	* Equivalent to {@code add(1)}.
298	<	*/
299	<	public void increment() {
300	<	add(1L);
301	<	}
302	<
303	<	/**
304	<	* Equivalent to {@code add(-1)}.
305	<	*/
306	<	public void decrement() {
307	<	add(-1L);
308	<	}
309	<
310	<	/**
311	<	* Equivalent to {@link #sum} followed by {@link #reset}.
324	>	* Resets each of the variables to zero, returning the estimated
325	>	* previous sum. This is effective in fully resetting the sum only
326	>	* if there are no concurrent updates.
327		*
328	<	* @return the estimated sum
328	>	* @return the estimated previous sum
329		*/
330	<	public long sumAndReset() {
330	>	public long reset() {
331		long sum = 0L;
332		Adder[] as = adders;
333		if (as != null) {
#	Line 320 | Line 335 | public class StripedAdder implements Ser
335		for (int i = 0; i < n; ++i) {
336		Adder a = as[i];
337		if (a != null) {
338	<	sum += a.get();
339	<	a.set(0L);
338	>	sum += a.value;
339	>	a.value = 0L;
340		}
341		}
342		}
#	Line 337 | Line 352 | public class StripedAdder implements Ser
352		private void readObject(ObjectInputStream s)
353		throws IOException, ClassNotFoundException {
354		s.defaultReadObject();
355	<	mutex.set(0);
355	>	busy = 0;
356		add(s.readLong());
357		}
358
359	+	// Unsafe mechanics
360	+	private static final sun.misc.Unsafe UNSAFE;
361	+	private static final long busyOffset;
362	+	private static final long valueOffset;
363	+	static {
364	+	try {
365	+	UNSAFE = getUnsafe();
366	+	Class<?> sk = StripedAdder.class;
367	+	busyOffset = UNSAFE.objectFieldOffset
368	+	(sk.getDeclaredField("busy"));
369	+	Class<?> ak = Adder.class;
370	+	valueOffset = UNSAFE.objectFieldOffset
371	+	(ak.getDeclaredField("value"));
372	+	} catch (Exception e) {
373	+	throw new Error(e);
374	+	}
375	+	}
376	+
377	+	/**
378	+	* Returns a sun.misc.Unsafe.  Suitable for use in a 3rd party package.
379	+	* Replace with a simple call to Unsafe.getUnsafe when integrating
380	+	* into a jdk.
381	+	*
382	+	* @return a sun.misc.Unsafe
383	+	*/
384	+	private static sun.misc.Unsafe getUnsafe() {
385	+	try {
386	+	return sun.misc.Unsafe.getUnsafe();
387	+	} catch (SecurityException se) {
388	+	try {
389	+	return java.security.AccessController.doPrivileged
390	+	(new java.security
391	+	.PrivilegedExceptionAction<sun.misc.Unsafe>() {
392	+	public sun.misc.Unsafe run() throws Exception {
393	+	java.lang.reflect.Field f = sun.misc
394	+	.Unsafe.class.getDeclaredField("theUnsafe");
395	+	f.setAccessible(true);
396	+	return (sun.misc.Unsafe) f.get(null);
397	+	}});
398	+	} catch (java.security.PrivilegedActionException e) {
399	+	throw new RuntimeException("Could not initialize intrinsics",
400	+	e.getCause());
401	+	}
402	+	}
403	+	}
404	+
405		}

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines