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; |
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 |
43 |
|
|
44 |
|
/* |
45 |
|
* A StripedAdder maintains a table of Atomic long variables. The |
46 |
< |
* table is indexed by per-thread hash codes that are initialized |
48 |
< |
* to random values. |
46 |
> |
* table is indexed by per-thread hash codes. |
47 |
|
* |
48 |
< |
* The table doubles in size upon contention (as indicated by |
49 |
< |
* failed CASes when performing add()), but is capped at the |
50 |
< |
* nearest power of two >= #CPUS. This reflects the idea that, |
51 |
< |
* when there are more threads than CPUs, then if each thread were |
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 |
63 |
> |
* operation (see method retryAdd). Upon a collision, if the table |
64 |
> |
* size is less than the capacity, it is doubled in size unless |
65 |
> |
* some other thread holds lock. If a hashed slot is empty, and |
66 |
> |
* lock is available, a new Adder is created. Otherwise, if the |
67 |
> |
* slot exists, a CAS is tried. Retries proceed by "double |
68 |
> |
* hashing", using a secondary hash (Marsaglia XorShift) to try to |
69 |
> |
* find a free slot. |
70 |
> |
* |
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 |
83 |
|
* limitations, observed contention is typically low in these |
84 |
|
* cases. |
85 |
|
* |
64 |
– |
* Table entries are of class Adder; a form of AtomicLong padded |
65 |
– |
* to reduce cache contention on most processors. Padding is |
66 |
– |
* overkill for most Atomics because they are most often |
67 |
– |
* irregularly scattered in memory and thus don't interfere much |
68 |
– |
* with each other. But Atomic objects residing in arrays will |
69 |
– |
* tend to be placed adjacent to each other, and so will most |
70 |
– |
* often share cache lines without this precaution. Adders are |
71 |
– |
* constructed upon first use, which further improves per-thread |
72 |
– |
* locality and helps reduce (an already large) footprint. |
73 |
– |
* |
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 |
91 |
> |
* candidate Adder. During these retries, there is increased |
92 |
|
* contention and reduced locality, which is still better than |
93 |
|
* alternatives. |
94 |
|
*/ |
95 |
|
|
96 |
< |
/** |
84 |
< |
* Number of processors, to place a cap on table growth. |
85 |
< |
*/ |
86 |
< |
static final int NCPU = Runtime.getRuntime().availableProcessors(); |
87 |
< |
|
88 |
< |
/** |
89 |
< |
* The table size set upon first use when default-constructed |
90 |
< |
*/ |
91 |
< |
private static final int DEFAULT_ARRAY_SIZE = 8; |
96 |
> |
private static final int NCPU = Runtime.getRuntime().availableProcessors(); |
97 |
|
|
98 |
|
/** |
99 |
< |
* Padded version of AtomicLong |
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 extends AtomicLong { |
104 |
< |
long p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, pa, pb, pc, pd; |
105 |
< |
Adder(long x) { super(x); } |
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 starts off with |
112 |
< |
* a given random value, but may be set to a different value upon |
104 |
< |
* collisions in retryAdd. |
111 |
> |
* Holder for the thread-local hash code. The code is initially |
112 |
> |
* random, but may be set to a different value upon collisions. |
113 |
|
*/ |
114 |
|
static final class HashCode { |
115 |
+ |
static final Random rng = new Random(); |
116 |
|
int code; |
117 |
< |
HashCode(int h) { code = h; } |
117 |
> |
HashCode() { |
118 |
> |
int h = rng.nextInt(); // Avoid zero, because of xorShift rehash |
119 |
> |
code = (h == 0) ? 1 : h; |
120 |
> |
} |
121 |
|
} |
122 |
|
|
123 |
|
/** |
124 |
|
* The corresponding ThreadLocal class |
125 |
|
*/ |
126 |
|
static final class ThreadHashCode extends ThreadLocal<HashCode> { |
127 |
< |
static final Random rng = new Random(); |
116 |
< |
public HashCode initialValue() { |
117 |
< |
int h = rng.nextInt(); |
118 |
< |
return new HashCode((h == 0) ? 1 : h); // ensure nonzero |
119 |
< |
} |
127 |
> |
public HashCode initialValue() { return new HashCode(); } |
128 |
|
} |
129 |
|
|
130 |
|
/** |
131 |
|
* Static per-thread hash codes. Shared across all StripedAdders |
132 |
< |
* because adjustments due to collisions in one table are likely |
133 |
< |
* to be appropriate for others. |
132 |
> |
* to reduce ThreadLocal pollution and because adjustments due to |
133 |
> |
* collisions in one table are likely to be appropriate for |
134 |
> |
* others. |
135 |
|
*/ |
136 |
|
static final ThreadHashCode threadHashCode = new ThreadHashCode(); |
137 |
|
|
138 |
|
/** |
139 |
< |
* Table of adders. Size is power of two, grows to be at most NCPU. |
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 |
136 |
< |
* is no need for a blocking lock: Except during initialization |
137 |
< |
* races, when busy, other threads try other slots. |
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() { |
145 |
– |
this.mutex = new AtomicInteger(); |
146 |
– |
// remaining initialization on first call to add. |
152 |
|
} |
153 |
|
|
154 |
|
/** |
159 |
|
* will concurrently update the sum. |
160 |
|
*/ |
161 |
|
public StripedAdder(int expectedContention) { |
162 |
< |
int size; |
163 |
< |
if (expectedContention > 0) { |
164 |
< |
int cap = (expectedContention < NCPU) ? expectedContention : NCPU; |
165 |
< |
size = 1; |
161 |
< |
while (size < cap) |
162 |
< |
size <<= 1; |
163 |
< |
} |
164 |
< |
else |
165 |
< |
size = 0; |
162 |
> |
int cap = (expectedContention < NCPU) ? expectedContention : NCPU; |
163 |
> |
int size = 1; |
164 |
> |
while (size < cap) |
165 |
> |
size <<= 1; |
166 |
|
Adder[] as = new Adder[size]; |
167 |
|
for (int i = 0; i < size; ++i) |
168 |
|
as[i] = new Adder(0); |
169 |
|
this.adders = as; |
170 |
– |
this.mutex = new AtomicInteger(); |
170 |
|
} |
171 |
|
|
172 |
|
/** |
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 |
< |
if ((as = adders) == null || (n = as.length) < 1 || |
181 |
< |
(a = as[hc.code & (n - 1)]) == null || |
182 |
< |
!a.compareAndSet(v = a.get(), v + x)) |
183 |
< |
retryAdd(x, hc); |
180 |
> |
int h = hc.code; |
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. |
196 |
> |
* creating new Adders, and/or contention. See above for |
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; |
208 |
< |
for (boolean retried = false; ; retried = true) { |
209 |
< |
Adder[] as; Adder a; long v; int n, k; // Locals for volatiles |
210 |
< |
if ((as = adders) == null || (n = as.length) < 1) { |
211 |
< |
if (mutex.get() == 0 && mutex.compareAndSet(0, 1)) { |
212 |
< |
try { |
213 |
< |
if (adders == null) // Default-initialize |
214 |
< |
adders = new Adder[DEFAULT_ARRAY_SIZE]; |
215 |
< |
} finally { |
216 |
< |
mutex.set(0); |
207 |
> |
boolean collide = false; // true if last slot nonempty |
208 |
> |
for (;;) { |
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 |
> |
Adder r = new Adder(x); // Optimistically create |
214 |
> |
if (busy == 0 && |
215 |
> |
UNSAFE.compareAndSwapInt(this, busyOffset, 0, 1)) { |
216 |
> |
boolean created = false; |
217 |
> |
try { // Recheck under lock |
218 |
> |
Adder[] rs; int m, j; |
219 |
> |
if ((rs = adders) != null && |
220 |
> |
(m = rs.length) > 0 && |
221 |
> |
rs[j = (m - 1) & h] == null) { |
222 |
> |
rs[j] = r; |
223 |
> |
created = true; |
224 |
> |
} |
225 |
> |
} finally { |
226 |
> |
busy = 0; |
227 |
> |
} |
228 |
> |
if (created) |
229 |
> |
break; |
230 |
> |
continue; // Slot is now non-empty |
231 |
> |
} |
232 |
|
} |
233 |
+ |
collide = false; |
234 |
|
} |
235 |
< |
else |
236 |
< |
Thread.yield(); // initialization race |
237 |
< |
} |
238 |
< |
else if ((a = as[k = h & (n - 1)]) != null && |
239 |
< |
retried && a.compareAndSet(v = a.get(), v + x)) |
240 |
< |
break; |
241 |
< |
else if ((a == null || n < NCPU) && |
242 |
< |
mutex.get() == 0 && mutex.compareAndSet(0, 1)) { |
243 |
< |
boolean created = false; |
244 |
< |
try { |
245 |
< |
if (adders == as) { |
246 |
< |
if (as[k] == null) { |
247 |
< |
as[k] = new Adder(x); |
248 |
< |
created = true; |
235 |
> |
else if (precontended) // CAS already known to fail |
236 |
> |
precontended = false; // Continue after rehash |
237 |
> |
else { |
238 |
> |
long v = a.value; |
239 |
> |
if (UNSAFE.compareAndSwapLong(a, valueOffset, v, v + x)) |
240 |
> |
break; |
241 |
> |
if (!collide) |
242 |
> |
collide = true; |
243 |
> |
else if (n >= NCPU || adders != as) |
244 |
> |
collide = false; // Can't expand |
245 |
> |
else if (busy == 0 && |
246 |
> |
UNSAFE.compareAndSwapInt(this, busyOffset, 0, 1)) { |
247 |
> |
collide = false; |
248 |
> |
try { |
249 |
> |
if (adders == as) { // Expand table |
250 |
> |
Adder[] rs = new Adder[n << 1]; |
251 |
> |
for (int i = 0; i < n; ++i) |
252 |
> |
rs[i] = as[i]; |
253 |
> |
adders = rs; |
254 |
> |
} |
255 |
> |
} finally { |
256 |
> |
busy = 0; |
257 |
|
} |
258 |
< |
else { // Expand table |
259 |
< |
Adder[] rs = new Adder[n << 1]; |
260 |
< |
for (int i = 0; i < n; ++i) |
261 |
< |
rs[i] = as[i]; |
262 |
< |
adders = rs; |
258 |
> |
continue; |
259 |
> |
} |
260 |
> |
} |
261 |
> |
h ^= h << 13; // Rehash |
262 |
> |
h ^= h >>> 17; |
263 |
> |
h ^= h << 5; |
264 |
> |
} |
265 |
> |
else if (adders == as) { // Try to default-initialize |
266 |
> |
Adder[] rs = new Adder[1]; |
267 |
> |
rs[0] = new Adder(x); |
268 |
> |
boolean init = false; |
269 |
> |
while (adders == as) { |
270 |
> |
if (UNSAFE.compareAndSwapInt(this, busyOffset, 0, 1)) { |
271 |
> |
try { |
272 |
> |
if (adders == as) { |
273 |
> |
adders = rs; |
274 |
> |
init = true; |
275 |
> |
} |
276 |
> |
} finally { |
277 |
> |
busy = 0; |
278 |
|
} |
279 |
+ |
break; |
280 |
|
} |
281 |
< |
} finally { |
282 |
< |
mutex.set(0); |
281 |
> |
if (adders != as) |
282 |
> |
break; |
283 |
> |
Thread.yield(); // Back off |
284 |
|
} |
285 |
< |
if (created) |
285 |
> |
if (init) |
286 |
|
break; |
287 |
|
} |
233 |
– |
else { // Try elsewhere |
234 |
– |
h ^= h << 13; |
235 |
– |
h ^= h >>> 17; // Marsaglia XorShift |
236 |
– |
h ^= h << 5; |
237 |
– |
} |
288 |
|
} |
289 |
< |
hc.code = h; |
289 |
> |
hc.code = h; // Record index for next time |
290 |
> |
} |
291 |
> |
|
292 |
> |
/** |
293 |
> |
* Equivalent to {@code add(1)}. |
294 |
> |
*/ |
295 |
> |
public void increment() { |
296 |
> |
add(1L); |
297 |
> |
} |
298 |
> |
|
299 |
> |
/** |
300 |
> |
* Equivalent to {@code add(-1)}. |
301 |
> |
*/ |
302 |
> |
public void decrement() { |
303 |
> |
add(-1L); |
304 |
|
} |
305 |
|
|
306 |
|
/** |
318 |
|
for (int i = 0; i < n; ++i) { |
319 |
|
Adder a = as[i]; |
320 |
|
if (a != null) |
321 |
< |
sum += a.get(); |
321 |
> |
sum += a.value; |
322 |
|
} |
323 |
|
} |
324 |
|
return sum; |
325 |
|
} |
326 |
|
|
327 |
|
/** |
328 |
< |
* Resets each of the variables to zero. This is effective in |
329 |
< |
* fully resetting the sum only if there are no concurrent |
330 |
< |
* updates. |
267 |
< |
*/ |
268 |
< |
public void reset() { |
269 |
< |
Adder[] as = adders; |
270 |
< |
if (as != null) { |
271 |
< |
int n = as.length; |
272 |
< |
for (int i = 0; i < n; ++i) { |
273 |
< |
Adder a = as[i]; |
274 |
< |
if (a != null) |
275 |
< |
a.set(0L); |
276 |
< |
} |
277 |
< |
} |
278 |
< |
} |
279 |
< |
|
280 |
< |
/** |
281 |
< |
* Equivalent to {@code add(1)}. |
282 |
< |
*/ |
283 |
< |
public void increment() { |
284 |
< |
add(1L); |
285 |
< |
} |
286 |
< |
|
287 |
< |
/** |
288 |
< |
* Equivalent to {@code add(-1)}. |
289 |
< |
*/ |
290 |
< |
public void decrement() { |
291 |
< |
add(-1L); |
292 |
< |
} |
293 |
< |
|
294 |
< |
/** |
295 |
< |
* Equivalent to {@link #sum} followed by {@link #reset}. |
328 |
> |
* Resets each of the variables to zero, returning the estimated |
329 |
> |
* previous sum. This is effective in fully resetting the sum only |
330 |
> |
* if there are no concurrent updates. |
331 |
|
* |
332 |
< |
* @return the estimated sum |
332 |
> |
* @return the estimated previous sum |
333 |
|
*/ |
334 |
< |
public long sumAndReset() { |
334 |
> |
public long reset() { |
335 |
|
long sum = 0L; |
336 |
|
Adder[] as = adders; |
337 |
|
if (as != null) { |
339 |
|
for (int i = 0; i < n; ++i) { |
340 |
|
Adder a = as[i]; |
341 |
|
if (a != null) { |
342 |
< |
sum += a.get(); |
343 |
< |
a.set(0L); |
342 |
> |
sum += a.value; |
343 |
> |
a.value = 0L; |
344 |
|
} |
345 |
|
} |
346 |
|
} |
356 |
|
private void readObject(ObjectInputStream s) |
357 |
|
throws IOException, ClassNotFoundException { |
358 |
|
s.defaultReadObject(); |
359 |
< |
mutex.set(0); |
359 |
> |
busy = 0; |
360 |
|
add(s.readLong()); |
361 |
|
} |
362 |
|
|
363 |
+ |
// Unsafe mechanics |
364 |
+ |
private static final sun.misc.Unsafe UNSAFE; |
365 |
+ |
private static final long busyOffset; |
366 |
+ |
private static final long valueOffset; |
367 |
+ |
static { |
368 |
+ |
try { |
369 |
+ |
UNSAFE = getUnsafe(); |
370 |
+ |
Class<?> sk = StripedAdder.class; |
371 |
+ |
busyOffset = UNSAFE.objectFieldOffset |
372 |
+ |
(sk.getDeclaredField("busy")); |
373 |
+ |
Class<?> ak = Adder.class; |
374 |
+ |
valueOffset = UNSAFE.objectFieldOffset |
375 |
+ |
(ak.getDeclaredField("value")); |
376 |
+ |
} catch (Exception e) { |
377 |
+ |
throw new Error(e); |
378 |
+ |
} |
379 |
+ |
} |
380 |
+ |
|
381 |
+ |
/** |
382 |
+ |
* Returns a sun.misc.Unsafe. Suitable for use in a 3rd party package. |
383 |
+ |
* Replace with a simple call to Unsafe.getUnsafe when integrating |
384 |
+ |
* into a jdk. |
385 |
+ |
* |
386 |
+ |
* @return a sun.misc.Unsafe |
387 |
+ |
*/ |
388 |
+ |
private static sun.misc.Unsafe getUnsafe() { |
389 |
+ |
try { |
390 |
+ |
return sun.misc.Unsafe.getUnsafe(); |
391 |
+ |
} catch (SecurityException se) { |
392 |
+ |
try { |
393 |
+ |
return java.security.AccessController.doPrivileged |
394 |
+ |
(new java.security |
395 |
+ |
.PrivilegedExceptionAction<sun.misc.Unsafe>() { |
396 |
+ |
public sun.misc.Unsafe run() throws Exception { |
397 |
+ |
java.lang.reflect.Field f = sun.misc |
398 |
+ |
.Unsafe.class.getDeclaredField("theUnsafe"); |
399 |
+ |
f.setAccessible(true); |
400 |
+ |
return (sun.misc.Unsafe) f.get(null); |
401 |
+ |
}}); |
402 |
+ |
} catch (java.security.PrivilegedActionException e) { |
403 |
+ |
throw new RuntimeException("Could not initialize intrinsics", |
404 |
+ |
e.getCause()); |
405 |
+ |
} |
406 |
+ |
} |
407 |
+ |
} |
408 |
+ |
|
409 |
|
} |