Unsafe类相当于是一个java语言中的后门类,提供了硬件级别的原子操作,所以在一些并发编程中被大量使用。jdk已经作出说明,该类对程序员而言不是一个安全操作,在后续的jdk升级过程中,可能会禁用该类。所以这个类的使用是一把双刃剑,实际项目中谨慎使用,以免造成jdk升级不兼容问题。
这里并不系统讲解Unsafe的所有功能,只介绍和接下来内容相关的操作
arrayBaseOffset
:获取数组的基础偏移量
arrayIndexScale
:获取数组中元素的偏移间隔,要获取对应所以的元素,将索引号和该值相乘,获得数组中指定角标元素的偏移量
getObjectVolatile
:获取对象上的属性值或者数组中的元素
getObject
:获取对象上的属性值或者数组中的元素,已过时
putOrderedObject
:设置对象的属性值或者数组中某个角标的元素,更高效
putObjectVolatile
:设置对象的属性值或者数组中某个角标的元素
putObject
:设置对象的属性值或者数组中某个角标的元素,已过时
public class Test02 {
public static void main(String[] args) throws Exception {
Integer[] arr = {
2,5,1,8,10};
//获取Unsafe对象
Unsafe unsafe = getUnsafe();
//获取Integer[]的基础偏移量
int baseOffset = unsafe.arrayBaseOffset(Integer[].class);
//获取Integer[]中元素的偏移间隔
int indexScale = unsafe.arrayIndexScale(Integer[].class);
//获取数组中索引为2的元素对象
Object o = unsafe.getObjectVolatile(arr, (2 * indexScale) + baseOffset);
System.out.println(o); //1
//设置数组中索引为2的元素值为100
unsafe.putOrderedObject(arr,(2 * indexScale) + baseOffset,100);
System.out.println(Arrays.toString(arr));//[2, 5, 100, 8, 10]
}
//反射获取Unsafe对象
public static Unsafe getUnsafe() throws Exception {
Field theUnsafe = Unsafe.class.getDeclaredField("theUnsafe");
theUnsafe.setAccessible(true);
return (Unsafe) theUnsafe.get(null);
}
}
无参构造
//空参构造
public ConcurrentHashMap() {
//调用本类的带参构造
//DEFAULT_INITIAL_CAPACITY = 16
//DEFAULT_LOAD_FACTOR = 0.75f
//int DEFAULT_CONCURRENCY_LEVEL = 16
this(DEFAULT_INITIAL_CAPACITY, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
}
三个参数的构造:一些非核心逻辑的代码已经省略
//initialCapacity 定义ConcurrentHashMap存放元素的容量
//concurrencyLevel 定义ConcurrentHashMap中Segment[]的大小
public ConcurrentHashMap(int initialCapacity,
float loadFactor, int concurrencyLevel) {
int sshift = 0;
int ssize = 1;
//计算Segment[]的大小,保证是2的幂次方数
while (ssize < concurrencyLevel) {
++sshift;
ssize <<= 1;
}
//这两个值用于后面计算Segment[]的角标
this.segmentShift = 32 - sshift;
this.segmentMask = ssize - 1;
//计算每个Segment中存储元素的个数
int c = initialCapacity / ssize;
if (c * ssize < initialCapacity)
++c;
//最小Segment中存储元素的个数为2
int cap = MIN_SEGMENT_TABLE_CAPACITY;
矫正每个Segment中存储元素的个数,保证是2的幂次方,最小为2
while (cap < c)
cap <<= 1;
//创建一个Segment对象,作为其他Segment对象的模板
Segment<K,V> s0 =
new Segment<K,V>(loadFactor, (int)(cap * loadFactor),
(HashEntry<K,V>[])new HashEntry[cap]);
Segment<K,V>[] ss = (Segment<K,V>[])new Segment[ssize];
//利用Unsafe类,将创建的Segment对象存入0角标位置
UNSAFE.putOrderedObject(ss, SBASE, s0); // ordered write of segments[0]
this.segments = ss;
}
综上:ConcurrentHashMap中保存了一个默认长度为16的Segment[],每个Segment元素中保存了一个默认长度为2的HashEntry[],我们添加的元素,是存入对应的Segment中的HashEntry[]中。所以ConcurrentHashMap中默认元素的长度是32个,而不是16个
static final class Segment<K,V> extends ReentrantLock implements Serializable {
...
}
我们发现Segment是继承自ReentrantLock的,学过线程的兄弟都知道,它可以实现同步操作,从而保证多线程下的安全。因为每个Segment之间的锁互不影响,所以我们也将ConcurrentHashMap中的这种锁机制称之为分段锁,这比HashTable的线程安全操作高效的多。
//ConcurrentHashMap中真正存储数据的对象
static final class HashEntry<K,V> {
final int hash; //通过运算,得到的键的hash值
final K key; // 存入的键
volatile V value; //存入的值
volatile HashEntry<K,V> next; //记录下一个元素,形成单向链表
HashEntry(int hash, K key, V value, HashEntry<K,V> next) {
this.hash = hash;
this.key = key;
this.value = value;
this.next = next;
}
}
public V put(K key, V value) {
Segment<K,V> s;
if (value == null)
throw new NullPointerException();
//基于key,计算hash值
int hash = hash(key);
//因为一个键要计算两个数组的索引,为了避免冲突,这里取高位计算Segment[]的索引
int j = (hash >>> segmentShift) & segmentMask;
//判断该索引位的Segment对象是否创建,没有就创建
if ((s = (Segment<K,V>)UNSAFE.getObject // nonvolatile; recheck
(segments, (j << SSHIFT) + SBASE)) == null) // in ensureSegment
s = ensureSegment(j);
//调用Segmetn的put方法实现元素添加
return s.put(key, hash, value, false);
}
//创建对应索引位的Segment对象,并返回
private Segment<K,V> ensureSegment(int k) {
final Segment<K,V>[] ss = this.segments;
long u = (k << SSHIFT) + SBASE; // raw offset
Segment<K,V> seg;
//获取,如果为null,即创建
if ((seg = (Segment<K,V>)UNSAFE.getObjectVolatile(ss, u)) == null) {
//以0角标位的Segment为模板
Segment<K,V> proto = ss[0]; // use segment 0 as prototype
int cap = proto.table.length;
float lf = proto.loadFactor;
int threshold = (int)(cap * lf);
HashEntry<K,V>[] tab = (HashEntry<K,V>[])new HashEntry[cap];
//获取,如果为null,即创建
if ((seg = (Segment<K,V>)UNSAFE.getObjectVolatile(ss, u))
== null) {
// recheck
//创建
Segment<K,V> s = new Segment<K,V>(lf, threshold, tab);
//自旋方式,将创建的Segment对象放到Segment[]中,确保线程安全
while ((seg = (Segment<K,V>)UNSAFE.getObjectVolatile(ss, u))
== null) {
if (UNSAFE.compareAndSwapObject(ss, u, null, seg = s))
break;
}
}
}
//返回
return seg;
}
final V put(K key, int hash, V value, boolean onlyIfAbsent) {
//尝试获取锁,获取成功,node为null,代码向下执行
//如果有其他线程占据锁对象,那么去做别的事情,而不是一直等待,提升效率
//scanAndLockForPut 稍后分析
HashEntry<K,V> node = tryLock() ? null :
scanAndLockForPut(key, hash, value);
V oldValue;
try {
HashEntry<K,V>[] tab = table;
//取hash的低位,计算HashEntry[]的索引
int index = (tab.length - 1) & hash;
//获取索引位的元素对象
HashEntry<K,V> first = entryAt(tab, index);
for (HashEntry<K,V> e = first;;) {
//获取的元素对象不为空
if (e != null) {
K k;
//如果是重复元素,覆盖原值
if ((k = e.key) == key ||
(e.hash == hash && key.equals(k))) {
oldValue = e.value;
if (!onlyIfAbsent) {
e.value = value;
++modCount;
}
break;
}
//如果不是重复元素,获取链表的下一个元素,继续循环遍历链表
e = e.next;
}
else {
//如果获取到的元素为空
//当前添加的键值对的HashEntry对象已经创建
if (node != null)
node.setNext(first); //头插法关联即可
else
//创建当前添加的键值对的HashEntry对象
node = new HashEntry<K,V>(hash, key, value, first);
//添加的元素数量递增
int c = count + 1;
//判断是否需要扩容
if (c > threshold && tab.length < MAXIMUM_CAPACITY)
//需要扩容
rehash(node);
else
//不需要扩容
//将当前添加的元素对象,存入数组角标位,完成头插法添加元素
setEntryAt(tab, index, node);
++modCount;
count = c;
oldValue = null;
break;
}
}
} finally {
//释放锁
unlock();
}
return oldValue;
}
该方法在线程没有获取到锁的情况下,去完成HashEntry对象的创建,提升效率
但是这个操作个人感觉有点累赘了。
private HashEntry<K,V> scanAndLockForPut(K key, int hash, V value) {
//获取头部元素
HashEntry<K,V> first = entryForHash(this, hash);
HashEntry<K,V> e = first;
HashEntry<K,V> node = null;
int retries = -1; // negative while locating node
while (!tryLock()) {
//获取锁失败
HashEntry<K,V> f; // to recheck first below
if (retries < 0) {
//没有下一个节点,并且也不是重复元素,创建HashEntry对象,不再遍历
if (e == null) {
if (node == null) // speculatively create node
node = new HashEntry<K,V>(hash, key, value, null);
retries = 0;
}
else if (key.equals(e.key))
//重复元素,不创建HashEntry对象,不再遍历
retries = 0;
else
//继续遍历下一个节点
e = e.next;
}
else if (++retries > MAX_SCAN_RETRIES) {
//如果尝试获取锁的次数过多,直接阻塞
//MAX_SCAN_RETRIES会根据可用cpu核数来确定
lock();
break;
}
else if ((retries & 1) == 0 &&
(f = entryForHash(this, hash)) != first) {
//如果期间有别的线程获取锁,重新遍历
e = first = f; // re-traverse if entry changed
retries = -1;
}
}
return node;
}
这里“通话”和“重地”的哈希值是一样的,那么他们添加时,会存入同一个Segment对象,必然会存在锁竞争
public static void main(String[] args) throws Exception {
final ConcurrentHashMap chm = new ConcurrentHashMap();
new Thread(){
@Override
public void run() {
chm.put("通话","11");
System.out.println("-----------");
}
}.start();
//让第一个线程先启动,进入put方法
Thread.sleep(1000);
new Thread(){
@Override
public void run() {
chm.put("重地","22");
System.out.println("===========");
}
}.start();
}
会发现两个线程,分别停在不同的断点位置,这就是多线程锁互斥产生的结果
然后就可以分别让不同的线程向下执行,查看代码走向了。
private void rehash(HashEntry<K,V> node) {
HashEntry<K,V>[] oldTable = table;
int oldCapacity = oldTable.length;
//两倍容量
int newCapacity = oldCapacity << 1;
threshold = (int)(newCapacity * loadFactor);
//基于新容量,创建HashEntry数组
HashEntry<K,V>[] newTable =
(HashEntry<K,V>[]) new HashEntry[newCapacity];
int sizeMask = newCapacity - 1;
//实现数据迁移
for (int i = 0; i < oldCapacity ; i++) {
HashEntry<K,V> e = oldTable[i];
if (e != null) {
HashEntry<K,V> next = e.next;
int idx = e.hash & sizeMask;
if (next == null) // Single node on list
//原位置只有一个元素,直接放到新数组即可
newTable[idx] = e;
else {
// Reuse consecutive sequence at same slot
//=========图一=====================
HashEntry<K,V> lastRun = e;
int lastIdx = idx;
for (HashEntry<K,V> last = next;
last != null;
last = last.next) {
int k = last.hash & sizeMask;
if (k != lastIdx) {
lastIdx = k;
lastRun = last;
}
}
//=========图一=====================
//=========图二=====================
newTable[lastIdx] = lastRun;
//=========图二=====================
// Clone remaining nodes
//=========图三=====================
for (HashEntry<K,V> p = e; p != lastRun; p = p.next) {
V v = p.value;
int h = p.hash;
int k = h & sizeMask;
HashEntry<K,V> n = newTable[k];
//这里旧的HashEntry不会放到新数组
//而是基于原来的数据创建了一个新的HashEntry对象,放入新数组
newTable[k] = new HashEntry<K,V>(h, p.key, v, n);
}
//=========图三=====================
}
}
}
//采用头插法,将新元素加入到数组中
int nodeIndex = node.hash & sizeMask; // add the new node
node.setNext(newTable[nodeIndex]);
newTable[nodeIndex] = node;
table = newTable;
}
图一
图二
图三
public int size() {
// Try a few times to get accurate count. On failure due to
// continuous async changes in table, resort to locking.
final Segment<K,V>[] segments = this.segments;
int size;
boolean overflow; // true if size overflows 32 bits
long sum; // sum of modCounts
long last = 0L; // previous sum
int retries = -1; // first iteration isn't retry
try {
for (;;) {
//当第5次走到这个地方时,会将整个Segment[]的所有Segment对象锁住
if (retries++ == RETRIES_BEFORE_LOCK) {
for (int j = 0; j < segments.length; ++j)
ensureSegment(j).lock(); // force creation
}
sum = 0L;
size = 0;
overflow = false;
for (int j = 0; j < segments.length; ++j) {
Segment<K,V> seg = segmentAt(segments, j);
if (seg != null) {
//累加所有Segment的操作次数
sum += seg.modCount;
int c = seg.count;
//累加所有segment中的元素个数 size+=c
if (c < 0 || (size += c) < 0)
overflow = true;
}
}
//当这次累加值和上一次累加值一样,证明没有进行新的增删改操作,返回sum
//第一次last为0,如果有元素的话,这个for循环最少循环两次的
if (sum == last)
break;
//记录累加的值
last = sum;
}
} finally {
//如果之前有锁住,解锁
if (retries > RETRIES_BEFORE_LOCK) {
for (int j = 0; j < segments.length; ++j)
segmentAt(segments, j).unlock();
}
}
//溢出,返回int的最大值,否则返回累加的size
return overflow ? Integer.MAX_VALUE : size;
}
在jdk8的ConcurrentHashMap中一共有5个构造方法,这四个构造方法中都没有对内部的数组做初始化, 只是对一些变量的初始值做了处理
jdk8的ConcurrentHashMap的数组初始化是在第一次添加元素时完成
//没有维护任何变量的操作,如果调用该方法,数组长度默认是16
public ConcurrentHashMap() {
}
//传递进来一个初始容量,ConcurrentHashMap会基于这个值计算一个比这个值大的2的幂次方数作为初始容量
public ConcurrentHashMap(int initialCapacity) {
if (initialCapacity < 0)
throw new IllegalArgumentException();
int cap = ((initialCapacity >= (MAXIMUM_CAPACITY >>> 1)) ?
MAXIMUM_CAPACITY :
tableSizeFor(initialCapacity + (initialCapacity >>> 1) + 1));
this.sizeCtl = cap;
}
注意,调用这个方法,得到的初始容量和我们之前讲的HashMap以及jdk7的ConcurrentHashMap不同,即使你传递的是一个2的幂次方数,该方法计算出来的初始容量依然是比这个值大的2的幂次方数
//调用四个参数的构造
public ConcurrentHashMap(int initialCapacity, float loadFactor) {
this(initialCapacity, loadFactor, 1);
}
//计算一个大于或者等于给定的容量值,该值是2的幂次方数作为初始容量
public ConcurrentHashMap(int initialCapacity,
float loadFactor, int concurrencyLevel) {
if (!(loadFactor > 0.0f) || initialCapacity < 0 || concurrencyLevel <= 0)
throw new IllegalArgumentException();
if (initialCapacity < concurrencyLevel) // Use at least as many bins
initialCapacity = concurrencyLevel; // as estimated threads
long size = (long)(1.0 + (long)initialCapacity / loadFactor);
int cap = (size >= (long)MAXIMUM_CAPACITY) ?
MAXIMUM_CAPACITY : tableSizeFor((int)size);
this.sizeCtl = cap;
}
//基于一个Map集合,构建一个ConcurrentHashMap
//初始容量为16
public ConcurrentHashMap(Map<? extends K, ? extends V> m) {
this.sizeCtl = DEFAULT_CAPACITY;
putAll(m);
}
sizeCtl
含义解释注意:以上这些构造方法中,都涉及到一个变量
sizeCtl
,这个变量是一个非常重要的变量,而且具有非常丰富的含义,它的值不同,对应的含义也不一样,这里我们先对这个变量不同的值的含义做一下说明,后续源码分析过程中,进一步解释
sizeCtl
为0,代表数组未初始化, 且数组的初始容量为16
sizeCtl
为正数,如果数组未初始化,那么其记录的是数组的初始容量,如果数组已经初始化,那么其记录的是数组的扩容阈值
sizeCtl
为-1,表示数组正在进行初始化
sizeCtl
小于0,并且不是-1,表示数组正在扩容, -(1+n),表示此时有n个线程正在共同完成数组的扩容操作
public V put(K key, V value) {
return putVal(key, value, false);
}
final V putVal(K key, V value, boolean onlyIfAbsent) {
//如果有空值或者空键,直接抛异常
if (key == null || value == null) throw new NullPointerException();
//基于key计算hash值,并进行一定的扰动
int hash = spread(key.hashCode());
//记录某个桶上元素的个数,如果超过8个,会转成红黑树
int binCount = 0;
for (Node<K,V>[] tab = table;;) {
Node<K,V> f; int n, i, fh;
//如果数组还未初始化,先对数组进行初始化
if (tab == null || (n = tab.length) == 0)
tab = initTable();
//如果hash计算得到的桶位置没有元素,利用cas将元素添加
else if ((f = tabAt(tab, i = (n - 1) & hash)) == null) {
//cas+自旋(和外侧的for构成自旋循环),保证元素添加安全
if (casTabAt(tab, i, null,
new Node<K,V>(hash, key, value, null)))
break; // no lock when adding to empty bin
}
//如果hash计算得到的桶位置元素的hash值为MOVED,证明正在扩容,那么协助扩容
else if ((fh = f.hash) == MOVED)
tab = helpTransfer(tab, f);
else {
//hash计算的桶位置元素不为空,且当前没有处于扩容操作,进行元素添加
V oldVal = null;
//对当前桶进行加锁,保证线程安全,执行元素添加操作
synchronized (f) {
if (tabAt(tab, i) == f) {
//普通链表节点
if (fh >= 0) {
binCount = 1;
for (Node<K,V> e = f;; ++binCount) {
K ek;
if (e.hash == hash &&
((ek = e.key) == key ||
(ek != null && key.equals(ek)))) {
oldVal = e.val;
if (!onlyIfAbsent)
e.val = value;
break;
}
Node<K,V> pred = e;
if ((e = e.next) == null) {
pred.next = new Node<K,V>(hash, key,
value, null);
break;
}
}
}
//树节点,将元素添加到红黑树中
else if (f instanceof TreeBin) {
Node<K,V> p;
binCount = 2;
if ((p = ((TreeBin<K,V>)f).putTreeVal(hash, key,
value)) != null) {
oldVal = p.val;
if (!onlyIfAbsent)
p.val = value;
}
}
}
}
if (binCount != 0) {
//链表长度大于/等于8,将链表转成红黑树
if (binCount >= TREEIFY_THRESHOLD)
treeifyBin(tab, i);
//如果是重复键,直接将旧值返回
if (oldVal != null)
return oldVal;
break;
}
}
}
//添加的是新元素,维护集合长度,并判断是否要进行扩容操作
addCount(1L, binCount);
return null;
}
通过以上源码,我们可以看到,当需要添加元素时,会针对当前元素所对应的桶位进行加锁操作,这样一方面保证元素添加时,多线程的安全,同时对某个桶位加锁不会影响其他桶位的操作,进一步提升多线程的并发效率
在一个桶上,当没有节点时,是使用CAS进行添加,当有了节点后,就使用synchronized对链表的头节点或者树的根节点进行解锁,即锁了一个桶。
private final Node<K,V>[] initTable() {
Node<K,V>[] tab; int sc;
//cas+自旋,保证线程安全,对数组进行初始化操作
while ((tab = table) == null || tab.length == 0) {
//如果sizeCtl的值(-1)小于0,说明此时正在初始化, 让出cpu
if ((sc = sizeCtl) < 0)
Thread.yield(); // lost initialization race; just spin
//cas修改sizeCtl的值为-1,修改成功,进行数组初始化,失败,继续自旋
else if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
try {
if ((tab = table) == null || tab.length == 0) {
//sizeCtl为0,取默认长度16,否则去sizeCtl的值
int n = (sc > 0) ? sc : DEFAULT_CAPACITY;
@SuppressWarnings("unchecked")
//基于初始长度,构建数组对象
Node<K,V>[] nt = (Node<K,V>[])new Node<?,?>[n];
table = tab = nt;
//计算扩容阈值,并赋值给sc
sc = n - (n >>> 2);
}
} finally {
//将扩容阈值,赋值给sizeCtl
sizeCtl = sc;
}
break;
}
}
return tab;
}
private final void transfer(Node<K,V>[] tab, Node<K,V>[] nextTab) {
int n = tab.length, stride;
//如果是多cpu,那么每个线程划分任务,最小任务量是16个桶位的迁移
if ((stride = (NCPU > 1) ? (n >>> 3) / NCPU : n) < MIN_TRANSFER_STRIDE)
stride = MIN_TRANSFER_STRIDE; // subdivide range
//如果是扩容线程,此时新数组为null
if (nextTab == null) {
// initiating
try {
@SuppressWarnings("unchecked")
//两倍扩容创建新数组
Node<K,V>[] nt = (Node<K,V>[])new Node<?,?>[n << 1];
nextTab = nt;
} catch (Throwable ex) {
// try to cope with OOME
sizeCtl = Integer.MAX_VALUE;
return;
}
nextTable = nextTab;
//记录线程开始迁移的桶位,从后往前迁移
transferIndex = n;
}
//记录新数组的末尾
int nextn = nextTab.length;
//已经迁移的桶位,会用这个节点占位(这个节点的hash值为-1--MOVED)
ForwardingNode<K,V> fwd = new ForwardingNode<K,V>(nextTab);
boolean advance = true;
boolean finishing = false; // to ensure sweep before committing nextTab
for (int i = 0, bound = 0;;) {
Node<K,V> f; int fh;
while (advance) {
int nextIndex, nextBound;
//i记录当前正在迁移桶位的索引值
//bound记录下一次任务迁移的开始桶位
//--i >= bound 成立表示当前线程分配的迁移任务还没有完成
if (--i >= bound || finishing)
advance = false;
//没有元素需要迁移 -- 后续会去将扩容线程数减1,并判断扩容是否完成
else if ((nextIndex = transferIndex) <= 0) {
i = -1;
advance = false;
}
//计算下一次任务迁移的开始桶位,并将这个值赋值给transferIndex
else if (U.compareAndSwapInt
(this, TRANSFERINDEX, nextIndex,
nextBound = (nextIndex > stride ?
nextIndex - stride : 0))) {
bound = nextBound;
i = nextIndex - 1;
advance = false;
}
}
//如果没有更多的需要迁移的桶位,就进入该if
if (i < 0 || i >= n || i + n >= nextn) {
int sc;
//扩容结束后,保存新数组,并重新计算扩容阈值,赋值给sizeCtl
if (finishing) {
nextTable = null;
table = nextTab;
sizeCtl = (n << 1) - (n >>> 1);
return;
}
//扩容任务线程数减1
if (U.compareAndSwapInt(this, SIZECTL, sc = sizeCtl, sc - 1)) {
//判断当前所有扩容任务线程是否都执行完成
if ((sc - 2) != resizeStamp(n) << RESIZE_STAMP_SHIFT)
return;
//所有扩容线程都执行完,标识结束
finishing = advance = true;
i = n; // recheck before commit
}
}
//当前迁移的桶位没有元素,直接在该位置添加一个fwd节点
else if ((f = tabAt(tab, i)) == null)
advance = casTabAt(tab, i, null, fwd);
//当前节点已经被迁移
else if ((fh = f.hash) == MOVED)
advance = true; // already processed
else {
//当前节点需要迁移,加锁迁移,保证多线程安全
//此处迁移逻辑和jdk7的ConcurrentHashMap相同,不再赘述
synchronized (f) {
if (tabAt(tab, i) == f) {
Node<K,V> ln, hn;
if (fh >= 0) {
int runBit = fh & n;
Node<K,V> lastRun = f;
for (Node<K,V> p = f.next; p != null; p = p.next) {
int b = p.hash & n;
if (b != runBit) {
runBit = b;
lastRun = p;
}
}
if (runBit == 0) {
ln = lastRun;
hn = null;
}
else {
hn = lastRun;
ln = null;
}
for (Node<K,V> p = f; p != lastRun; p = p.next) {
int ph = p.hash; K pk = p.key; V pv = p.val;
if ((ph & n) == 0)
ln = new Node<K,V>(ph, pk, pv, ln);
else
hn = new Node<K,V>(ph, pk, pv, hn);
}
setTabAt(nextTab, i, ln);
setTabAt(nextTab, i + n, hn);
setTabAt(tab, i, fwd);
advance = true;
}
else if (f instanceof TreeBin) {
TreeBin<K,V> t = (TreeBin<K,V>)f;
TreeNode<K,V> lo = null, loTail = null;
TreeNode<K,V> hi = null, hiTail = null;
int lc = 0, hc = 0;
for (Node<K,V> e = t.first; e != null; e = e.next) {
int h = e.hash;
TreeNode<K,V> p = new TreeNode<K,V>
(h, e.key, e.val, null, null);
if ((h & n) == 0) {
if ((p.prev = loTail) == null)
lo = p;
else
loTail.next = p;
loTail = p;
++lc;
}
else {
if ((p.prev = hiTail) == null)
hi = p;
else
hiTail.next = p;
hiTail = p;
++hc;
}
}
ln = (lc <= UNTREEIFY_THRESHOLD) ? untreeify(lo) :
(hc != 0) ? new TreeBin<K,V>(lo) : t;
hn = (hc <= UNTREEIFY_THRESHOLD) ? untreeify(hi) :
(lc != 0) ? new TreeBin<K,V>(hi) : t;
setTabAt(nextTab, i, ln);
setTabAt(nextTab, i + n, hn);
setTabAt(tab, i, fwd);
advance = true;
}
}
}
}
}
}
多线程协助扩容的操作会在两个地方被触发:
① 当添加元素时,发现添加的元素对用的桶位为fwd节点,就会先去协助扩容,然后再添加元素
② 当添加完元素后,判断当前元素个数达到了扩容阈值,此时发现sizeCtl的值小于0,并且新数组不为空,这个时候,会去协助扩容
final V putVal(K key, V value, boolean onlyIfAbsent) {
if (key == null || value == null) throw new NullPointerException();
int hash = spread(key.hashCode());
int binCount = 0;
for (Node<K,V>[] tab = table;;) {
Node<K,V> f; int n, i, fh;
if (tab == null || (n = tab.length) == 0)
tab = initTable();
else if ((f = tabAt(tab, i = (n - 1) & hash)) == null) {
if (casTabAt(tab, i, null,
new Node<K,V>(hash, key, value, null)))
break; // no lock when adding to empty bin
}
//发现此处为fwd节点,协助扩容,扩容结束后,再循环回来添加元素
else if ((fh = f.hash) == MOVED)
tab = helpTransfer(tab, f);
//省略代码
final Node<K,V>[] helpTransfer(Node<K,V>[] tab, Node<K,V> f) {
Node<K,V>[] nextTab; int sc;
if (tab != null && (f instanceof ForwardingNode) &&
(nextTab = ((ForwardingNode<K,V>)f).nextTable) != null) {
int rs = resizeStamp(tab.length);
while (nextTab == nextTable && table == tab &&
(sc = sizeCtl) < 0) {
if ((sc >>> RESIZE_STAMP_SHIFT) != rs || sc == rs + 1 ||
sc == rs + MAX_RESIZERS || transferIndex <= 0)
break;
if (U.compareAndSwapInt(this, SIZECTL, sc, sc + 1)) {
//扩容,传递一个不是null的nextTab
transfer(tab, nextTab);
break;
}
}
return nextTab;
}
return table;
}
private final void addCount(long x, int check) {
//省略代码
if (check >= 0) {
Node<K,V>[] tab, nt; int n, sc;
//元素个数达到扩容阈值
while (s >= (long)(sc = sizeCtl) && (tab = table) != null &&
(n = tab.length) < MAXIMUM_CAPACITY) {
int rs = resizeStamp(n);
//sizeCtl小于0,说明正在执行扩容,那么协助扩容
if (sc < 0) {
if ((sc >>> RESIZE_STAMP_SHIFT) != rs || sc == rs + 1 ||
sc == rs + MAX_RESIZERS || (nt = nextTable) == null ||
transferIndex <= 0)
break;
if (U.compareAndSwapInt(this, SIZECTL, sc, sc + 1))
transfer(tab, nt);
}
else if (U.compareAndSwapInt(this, SIZECTL, sc,
(rs << RESIZE_STAMP_SHIFT) + 2))
transfer(tab, null);
s = sumCount();
}
}
}
注意:扩容的代码都在transfer方法中,这里不再赘述
① CounterCell数组不为空,优先利用数组中的CounterCell记录数量
② 如果数组为空,尝试对baseCount进行累加,失败后,会执行fullAddCount逻辑
③ 如果是添加元素操作,会继续判断是否需要扩容
private final void addCount(long x, int check) {
CounterCell[] as; long b, s;
//当CounterCell数组不为空,则优先利用数组中的CounterCell记录数量
//或者当baseCount的累加操作失败,会利用数组中的CounterCell记录数量
if ((as = counterCells) != null ||
!U.compareAndSwapLong(this, BASECOUNT, b = baseCount, s = b + x)) {
CounterCell a; long v; int m;
//标识是否有多线程竞争
boolean uncontended = true;
//当as数组为空
//或者当as长度为0
//或者当前线程对应的as数组桶位的元素为空
//或者当前线程对应的as数组桶位不为空,但是累加失败
if (as == null || (m = as.length - 1) < 0 ||
(a = as[ThreadLocalRandom.getProbe() & m]) == null ||
!(uncontended =
U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))) {
//以上任何一种情况成立,都会进入该方法,传入的uncontended是false
fullAddCount(x, uncontended);
return;
}
if (check <= 1)
return;
//计算元素个数
s = sumCount();
}
if (check >= 0) {
Node<K,V>[] tab, nt; int n, sc;
//当元素个数达到扩容阈值
//并且数组不为空
//并且数组长度小于限定的最大值
//满足以上所有条件,执行扩容
while (s >= (long)(sc = sizeCtl) && (tab = table) != null &&
(n = tab.length) < MAXIMUM_CAPACITY) {
//这个是一个很大的正数
int rs = resizeStamp(n);
//sc小于0,说明有线程正在扩容,那么会协助扩容
if (sc < 0) {
//扩容结束或者扩容线程数达到最大值或者扩容后的数组为null或者没有更多的桶位需要转移,结束操作
if ((sc >>> RESIZE_STAMP_SHIFT) != rs || sc == rs + 1 ||
sc == rs + MAX_RESIZERS || (nt = nextTable) == null ||
transferIndex <= 0)
break;
//扩容线程加1,成功后,进行协助扩容操作
if (U.compareAndSwapInt(this, SIZECTL, sc, sc + 1))
//协助扩容,newTable不为null
transfer(tab, nt);
}
//没有其他线程在进行扩容,达到扩容阈值后,给sizeCtl赋了一个很大的负数
//1+1=2 --》 代表此时有一个线程在扩容
//rs << RESIZE_STAMP_SHIFT)是一个很大的负数
else if (U.compareAndSwapInt(this, SIZECTL, sc,
(rs << RESIZE_STAMP_SHIFT) + 2))
//扩容,newTable为null
transfer(tab, null);
s = sumCount();
}
}
}
① 当CounterCell数组不为空,优先对CounterCell数组中的CounterCell的value累加
② 当CounterCell数组为空,会去创建CounterCell数组,默认长度为2,并对数组中的CounterCell的value累加
③ 当数组为空,并且此时有别的线程正在创建数组,那么尝试对baseCount做累加,成功即返回,否则自旋
private final void fullAddCount(long x, boolean wasUncontended) {
int h;
//获取当前线程的hash值
if ((h = ThreadLocalRandom.getProbe()) == 0) {
ThreadLocalRandom.localInit(); // force initialization
h = ThreadLocalRandom.getProbe();
wasUncontended = true;
}
//标识是否有冲突,如果最后一个桶不是null,那么为true
boolean collide = false; // True if last slot nonempty
for (;;) {
CounterCell[] as; CounterCell a; int n; long v;
//数组不为空,优先对数组中CouterCell的value累加
if ((as = counterCells) != null && (n = as.length) > 0) {
//线程对应的桶位为null
if ((a = as[(n - 1) & h]) == null) {
if (cellsBusy == 0) {
// Try to attach new Cell
//创建CounterCell对象
CounterCell r = new CounterCell(x); // Optimistic create
//利用CAS修改cellBusy状态为1,成功则将刚才创建的CounterCell对象放入数组中
if (cellsBusy == 0 &&
U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
boolean created = false;
try {
// Recheck under lock
CounterCell[] rs; int m, j;
//桶位为空, 将CounterCell对象放入数组
if ((rs = counterCells) != null &&
(m = rs.length) > 0 &&
rs[j = (m - 1) & h] == null) {
rs[j] = r;
//表示放入成功
created = true;
}
} finally {
cellsBusy = 0;
}
if (created) //成功退出循环
break;
//桶位已经被别的线程放置了已给CounterCell对象,继续循环
continue; // Slot is now non-empty
}
}
collide = false;
}
//桶位不为空,重新计算线程hash值,然后继续循环
else if (!wasUncontended) // CAS already known to fail
wasUncontended = true; // Continue after rehash
//重新计算了hash值后,对应的桶位依然不为空,对value累加
//成功则结束循环
//失败则继续下面判断
else if (U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))
break;
//数组被别的线程改变了,或者数组长度超过了可用cpu大小,重新计算线程hash值,否则继续下一个判断
else if (counterCells != as || n >= NCPU)
collide = false; // At max size or stale
//当没有冲突,修改为有冲突,并重新计算线程hash,继续循环
else if (!collide)
collide = true;
//如果CounterCell的数组长度没有超过cpu核数,对数组进行两倍扩容
//并继续循环
else if (cellsBusy == 0 &&
U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
try {
if (counterCells == as) {
// Expand table unless stale
CounterCell[] rs = new CounterCell[n << 1];
for (int i = 0; i < n; ++i)
rs[i] = as[i];
counterCells = rs;
}
} finally {
cellsBusy = 0;
}
collide = false;
continue; // Retry with expanded table
}
h = ThreadLocalRandom.advanceProbe(h);
}
//CounterCell数组为空,并且没有线程在创建数组,修改标记,并创建数组
else if (cellsBusy == 0 && counterCells == as &&
U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
boolean init = false;
try {
// Initialize table
if (counterCells == as) {
CounterCell[] rs = new CounterCell[2];
rs[h & 1] = new CounterCell(x);
counterCells = rs;
init = true;
}
} finally {
cellsBusy = 0;
}
if (init)
break;
}
//数组为空,并且有别的线程在创建数组,那么尝试对baseCount做累加,成功就退出循环,失败就继续循环
else if (U.compareAndSwapLong(this, BASECOUNT, v = baseCount, v + x))
break; // Fall back on using base
}
}
fullAddCount方法中,当as数组不为空的逻辑图解
public int size() {
long n = sumCount();
return ((n < 0L) ? 0 :
(n > (long)Integer.MAX_VALUE) ? Integer.MAX_VALUE :
(int)n);
}
final long sumCount() {
CounterCell[] as = counterCells; CounterCell a;
//获取baseCount的值
long sum = baseCount;
if (as != null) {
//遍历CounterCell数组,累加每一个CounterCell的value值
for (int i = 0; i < as.length; ++i) {
if ((a = as[i]) != null)
sum += a.value;
}
}
return sum;
}
注意:这个方法并不是线程安全的