提交 b7456c03 编写于 作者: P psandoz

8190974: Parallel stream execution within a custom ForkJoinPool should obey the parallelism

Reviewed-by: martin, tvaleev
上级 3bbef61d
/*
* Copyright (c) 2012, 2013, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2017, Oracle and/or its affiliates. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
* This code is free software; you can redistribute it and/or modify it
......@@ -27,6 +27,7 @@ package java.util.stream;
import java.util.Spliterator;
import java.util.concurrent.CountedCompleter;
import java.util.concurrent.ForkJoinPool;
import java.util.concurrent.ForkJoinWorkerThread;
/**
* Abstract base class for most fork-join tasks used to implement stream ops.
......@@ -88,13 +89,7 @@ abstract class AbstractTask<P_IN, P_OUT, R,
K extends AbstractTask<P_IN, P_OUT, R, K>>
extends CountedCompleter<R> {
/**
* Default target factor of leaf tasks for parallel decomposition.
* To allow load balancing, we over-partition, currently to approximately
* four tasks per processor, which enables others to help out
* if leaf tasks are uneven or some processors are otherwise busy.
*/
static final int LEAF_TARGET = ForkJoinPool.getCommonPoolParallelism() << 2;
private static final int LEAF_TARGET = ForkJoinPool.getCommonPoolParallelism() << 2;
/** The pipeline helper, common to all tasks in a computation */
protected final PipelineHelper<P_OUT> helper;
......@@ -156,6 +151,22 @@ abstract class AbstractTask<P_IN, P_OUT, R,
this.targetSize = parent.targetSize;
}
/**
* Default target of leaf tasks for parallel decomposition.
* To allow load balancing, we over-partition, currently to approximately
* four tasks per processor, which enables others to help out
* if leaf tasks are uneven or some processors are otherwise busy.
*/
public static int getLeafTarget() {
Thread t = Thread.currentThread();
if (t instanceof ForkJoinWorkerThread) {
return ((ForkJoinWorkerThread) t).getPool().getParallelism() << 2;
}
else {
return LEAF_TARGET;
}
}
/**
* Constructs a new node of type T whose parent is the receiver; must call
* the AbstractTask(T, Spliterator) constructor with the receiver and the
......@@ -181,7 +192,7 @@ abstract class AbstractTask<P_IN, P_OUT, R,
* @return suggested target leaf size
*/
public static long suggestTargetSize(long sizeEstimate) {
long est = sizeEstimate / LEAF_TARGET;
long est = sizeEstimate / getLeafTarget();
return est > 0L ? est : 1L;
}
......
/*
* Copyright (c) 2012, 2013, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2017, Oracle and/or its affiliates. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
* This code is free software; you can redistribute it and/or modify it
......@@ -28,7 +28,6 @@ import java.util.Objects;
import java.util.Spliterator;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.CountedCompleter;
import java.util.concurrent.ForkJoinTask;
import java.util.function.Consumer;
import java.util.function.DoubleConsumer;
import java.util.function.IntConsumer;
......@@ -378,7 +377,7 @@ final class ForEachOps {
this.spliterator = spliterator;
this.targetSize = AbstractTask.suggestTargetSize(spliterator.estimateSize());
// Size map to avoid concurrent re-sizes
this.completionMap = new ConcurrentHashMap<>(Math.max(16, AbstractTask.LEAF_TARGET << 1));
this.completionMap = new ConcurrentHashMap<>(Math.max(16, AbstractTask.getLeafTarget() << 1));
this.action = action;
this.leftPredecessor = null;
}
......
/*
* Copyright (c) 2012, 2013, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2017, Oracle and/or its affiliates. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
* This code is free software; you can redistribute it and/or modify it
......@@ -898,7 +898,7 @@ class StreamSpliterators {
* Note: The source spliterator may report {@code ORDERED} since that
* spliterator be the result of a previous pipeline stage that was
* collected to a {@code Node}. It is the order of the pipeline stage
* that governs whether the this slice spliterator is to be used or not.
* that governs whether this slice spliterator is to be used or not.
*/
static abstract class UnorderedSliceSpliterator<T, T_SPLITR extends Spliterator<T>> {
static final int CHUNK_SIZE = 1 << 7;
......@@ -915,7 +915,7 @@ class StreamSpliterators {
this.unlimited = limit < 0;
this.skipThreshold = limit >= 0 ? limit : 0;
this.chunkSize = limit >= 0 ? (int)Math.min(CHUNK_SIZE,
((skip + limit) / AbstractTask.LEAF_TARGET) + 1) : CHUNK_SIZE;
((skip + limit) / AbstractTask.getLeafTarget()) + 1) : CHUNK_SIZE;
this.permits = new AtomicLong(limit >= 0 ? skip + limit : skip);
}
......
/*
* Copyright (c) 2017, Oracle and/or its affiliates. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
* This code is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License version 2 only, as
* published by the Free Software Foundation. Oracle designates this
* particular file as subject to the "Classpath" exception as provided
* by Oracle in the LICENSE file that accompanied this code.
*
* This code is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
* version 2 for more details (a copy is included in the LICENSE file that
* accompanied this code).
*
* You should have received a copy of the GNU General Public License version
* 2 along with this work; if not, write to the Free Software Foundation,
* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
*
* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
* or visit www.oracle.com if you need additional information or have any
* questions.
*/
package org.openjdk.tests.java.util.stream;
import org.testng.annotations.Test;
import java.util.Comparator;
import java.util.Spliterator;
import java.util.concurrent.ForkJoinPool;
import java.util.concurrent.ForkJoinTask;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.function.Consumer;
import java.util.stream.IntStream;
import java.util.stream.StreamSupport;
import static org.testng.Assert.assertEquals;
import static org.testng.Assert.assertTrue;
/**
* Tests stream execution in a custom ForkJoinPool. See JDK-8190974.
*/
@Test
public class CustomFJPoolTest {
// A Spliterator that counts the number of spliterators created
// including itself, thus the count starts at 1
static class SplitCountingSpliterator<T> implements Spliterator<T> {
final Spliterator<T> s;
final AtomicInteger nsplits;
// Top-level constructor
public SplitCountingSpliterator(Spliterator<T> s) {
this.s = s;
nsplits = new AtomicInteger(1);
}
// Splitting constructor
SplitCountingSpliterator(Spliterator<T> s, AtomicInteger nsplits) {
this.s = s;
this.nsplits = nsplits;
}
int splits() {
return nsplits.get();
}
@Override
public boolean tryAdvance(Consumer<? super T> action) {
return s.tryAdvance(action);
}
@Override
public void forEachRemaining(Consumer<? super T> action) {
s.forEachRemaining(action);
}
@Override
public Spliterator<T> trySplit() {
Spliterator<T> split = s.trySplit();
if (split != null) {
nsplits.incrementAndGet();
return new SplitCountingSpliterator<>(split, nsplits);
}
else {
return null;
}
}
@Override
public long estimateSize() {
return s.estimateSize();
}
@Override
public long getExactSizeIfKnown() {
return s.getExactSizeIfKnown();
}
@Override
public int characteristics() {
return s.characteristics();
}
@Override
public boolean hasCharacteristics(int characteristics) {
return s.hasCharacteristics(characteristics);
}
@Override
public Comparator<? super T> getComparator() {
return s.getComparator();
}
}
public void testCustomPools() throws Exception {
int splitsForP1 = countSplits(new ForkJoinPool(1));
int splitsForP2 = countSplits(new ForkJoinPool(2));
assertEquals(splitsForP2, splitsForP1 * 2);
int commonParallelism = ForkJoinPool.getCommonPoolParallelism();
if (commonParallelism > 1 && commonParallelism < 128) {
int splitsForPHalfC = countSplits(new ForkJoinPool(commonParallelism / 2));
int splitsForPC = countSplits(ForkJoinPool.commonPool());
assertTrue(splitsForPHalfC < splitsForPC);
assertEquals(splitsForPC / splitsForPHalfC,
nearestPowerOfTwo(commonParallelism) / nearestPowerOfTwo(commonParallelism / 2));
}
}
static int countSplits(ForkJoinPool fjp) throws Exception {
// The number of splits will be equivalent to the number of leaf nodes
// and will be a power of 2
ForkJoinTask<Integer> fInteger = fjp.submit(() -> {
Spliterator<Integer> s = IntStream.range(0, 1024).boxed().parallel().spliterator();
SplitCountingSpliterator<Integer> cs = new SplitCountingSpliterator<>(s);
StreamSupport.stream(cs, true).forEach(e -> {});
return cs.splits();
});
return fInteger.get();
}
static int nearestPowerOfTwo(int i) {
return (i & (i - 1)) == 0
? i
: 1 << (32 - Integer.numberOfLeadingZeros(i));
}
}
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册