ConcurrentHashMap
ConcurrentHashMap in Java8 uses a linked list + red-black tree approach. If the head node is of Node type, then it is an ordinary linked list; if the head node is of TreeNode type, then it is a red-black tree. (TreeNode is a subclass of Node)
Linked list and red-black tree can be converted to each other: initially it is a linked list, when the elements in the linked list exceed a certain threshold (8), the linked list is converted into a red-black tree; conversely, when the number of elements in the red-black tree When it is less than a certain threshold, it is converted into a linked list.
The maximum capacity is 2^30, and the default capacity is 16
The source code of ConcurrentHashMap is relatively complicated. Let's simply understand the general flow of the initialization and put methods:
public class ConcurrentHashMap<K,V> extends AbstractMap<K,V>
implements ConcurrentMap<K,V>, Serializable {
static class Node<K, V> implements Map.Entry<K, V> {
final int hash;
final K key;
volatile V val;
volatile Node<K, V> next;
}
private static final int MAXIMUM_CAPACITY = 1 << 30;
public ConcurrentHashMap(int initialCapacity) {
if (initialCapacity < 0)
throw new IllegalArgumentException();
//cap为Node数组长度,为2的整数次方
int cap = ((initialCapacity >= (MAXIMUM_CAPACITY >>> 1)) ?
MAXIMUM_CAPACITY :
// 1.5*初始容量+1,向上取最接近的2的整数次方
//保证数组长度为2的幂次方 好处是 xx%数组长度 = xx&(数组长度-1)
tableSizeFor(initialCapacity + (initialCapacity >>> 1) + 1));
//初始化时表示数组长度 或者 控制并发扩容的线程数
this.sizeCtl = cap;
}
transient volatile Node<K, V>[] table;
private transient volatile int sizeCtl;
private static final int DEFAULT_CAPACITY = 16;
static final int HASH_BITS = 0x7fffffff;
static final int TREEIFY_THRESHOLD = 8;
private final Node<K, V>[] initTable() {
Node<K, V>[] tab;
int sc;
while ((tab = table) == null || tab.length == 0) {
if ((sc = sizeCtl) < 0)
Thread.yield();//sizeCtl<0时自旋等待
//SIZECTL为sizeCtl的偏移量,CAS操作,将sizeCtl设置为-1,这样保证只有一个线程参与初始化
else if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
try {
if ((tab = table) == null || tab.length == 0) {
int n = (sc > 0) ? sc : DEFAULT_CAPACITY;
//初始化
@SuppressWarnings("unchecked")
Node<K, V>[] nt = (Node<K, V>[]) new Node<?, ?>[n];
table = tab = nt;
//sizeCtl初始化完后不再表示数组长度,而是变成扩容阈值0.75n,finally块中设置
sc = n - (n >>> 2);
}
} finally {
sizeCtl = sc;
}
break;
}
}
return tab;
}
private static final int tableSizeFor(int c) {
int n = c - 1;
n |= n >>> 1;
n |= n >>> 2;
n |= n >>> 4;
n |= n >>> 8;
n |= n >>> 16;
return (n < 0) ? 1 : (n >= MAXIMUM_CAPACITY) ? MAXIMUM_CAPACITY : n + 1;
}
static final int spread(int h) {
//将h无符号右移16位得到h',此时h'高16位为0,然后和h异或
return (h ^ (h >>> 16)) & HASH_BITS;
}
public V put(K key, V value) {
return putVal(key, value, false);
}
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();//如果table为空需要先初始化
else if ((f = tabAt(tab, i = (n - 1) & hash)) == null) {
//第i个元素为空需要初始化,对第i个元素进行CAS操作
if (casTabAt(tab, i, null, new Node<K,V>(hash, key, value, null)))
break;
}
else if ((fh = f.hash) == MOVED)
//改槽正在扩容,协助扩容
tab = helpTransfer(tab, f);
else {
//放入元素
V oldVal = null;
//对table[i]元素加锁,也就是说table每个元素都作为一把锁
synchronized (f) {
//f=table[i]才继续,否则就重新循环
if (tabAt(tab, i) == f) {
if (fh >= 0) {//是链表
//链表元素个数
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;
}
}
}
else if (f instanceof TreeBin) {//是红黑树
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;
}
}
}
}
//是链表的话,binCount变量会从1一直累加
if (binCount != 0) {
//如果binCount超出阈值,则转为红黑树
if (binCount >= TREEIFY_THRESHOLD)
treeifyBin(tab, i);
if (oldVal != null)
return oldVal;
break;
}
}
}
//总元素个数+1
addCount(1L, binCount);
return null;
}
static final int MIN_TREEIFY_CAPACITY = 64;
private final void treeifyBin(Node<K,V>[] tab, int index) {
Node<K,V> b; int n, sc;
if (tab != null) {
if ((n = tab.length) < MIN_TREEIFY_CAPACITY)
//数组长度<64,不转成红黑树,直接扩容
tryPresize(n << 1);
else if ((b = tabAt(tab, index)) != null && b.hash >= 0) {
//链表转红黑树
synchronized (b) {
if (tabAt(tab, index) == b) {
TreeNode<K,V> hd = null, tl = null;
//遍历链表,构造红黑树
for (Node<K,V> e = b; e != null; e = e.next) {
TreeNode<K,V> p = new TreeNode<K,V>(e.hash, e.key, e.val, null, null);
if ((p.prev = tl) == null)
hd = p;
else
tl.next = p;
tl = p;
}
setTabAt(tab, index, new TreeBin<K,V>(hd));
}
}
}
}
}
private final void tryPresize(int size) {
//根据元素个数重新计算数组长度
int c = (size >= (MAXIMUM_CAPACITY >>> 1)) ? MAXIMUM_CAPACITY :
tableSizeFor(size + (size >>> 1) + 1);
int sc;
while ((sc = sizeCtl) >= 0) {
Node<K,V>[] tab = table; int n;
//如果table为空则先初始化
if (tab == null || (n = tab.length) == 0) {
n = (sc > c) ? sc : c;
if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
try {
if (table == tab) {
@SuppressWarnings("unchecked")
Node<K,V>[] nt = (Node<K,V>[])new Node<?,?>[n];
table = nt;
sc = n - (n >>> 2);//0.75n,下次扩容的阈值
}
} finally {
sizeCtl = sc;
}
}
}
//如果sizeCtl>=新算出的长度 或者 超过数组最大长度,则不扩容
else if (c <= sc || n >= MAXIMUM_CAPACITY)
break;
else if (tab == table) {
int rs = resizeStamp(n);
//多个线程进行并发扩容
if (sc < 0) {
Node<K,V>[] nt;
if ((sc >>> RESIZE_STAMP_SHIFT) != rs || sc == rs + 1 ||
sc == rs + MAX_RESIZERS || (nt = nextTable) == null ||
transferIndex <= 0)
//扩容结束
break;
if (U.compareAndSwapInt(this, SIZECTL, sc, sc + 1))
//协助扩容
transfer(tab, nt);
}
else if (U.compareAndSwapInt(this, SIZECTL, sc,
(rs << RESIZE_STAMP_SHIFT) + 2))
//第一次扩容
transfer(tab, null);
}
}
}
static final int NCPU = Runtime.getRuntime().availableProcessors();
private static final int MIN_TRANSFER_STRIDE = 16;
private transient volatile Node<K,V>[] nextTable;
//表示整个数组扩容的进度
private transient volatile int transferIndex;
/**
* 迁移元素
* @param tab 旧数组
* @param nextTab 新数组
*/
private final void transfer(Node<K,V>[] tab, Node<K,V>[] nextTab) {
//此方法可以被多线程调用,每个线程负责扩容一部分
/**
* 1.先计算每个线程扩容的步长
* 2.如果新数组nextTab为空则初始化,并设置transferIndex初始为旧数组长度,从大到小扩容,每次减去stride个位置,
* 直到n<=0,扩容完成。[0,transferIndex]位置表示还没分配到线程扩容;[transferIndex,n]表示已经分配线程扩容,
* 状态为正在扩容或者扩容成功。
*/
int n = tab.length, stride;
//计算步长,单核CPU下为n,只要一个线程,多核下保证步长最小值为16,扩容需要的线程数为 n/stride
if ((stride = (NCPU > 1) ? (n >>> 3) / NCPU : n) < MIN_TRANSFER_STRIDE)
stride = MIN_TRANSFER_STRIDE;
//初始化新的数组
if (nextTab == null) {
try {
//扩容2倍,保证新数组长度仍是2的幂次方
@SuppressWarnings("unchecked")
Node<K,V>[] nt = (Node<K,V>[])new Node<?,?>[n << 1];
nextTab = nt;
} catch (Throwable ex) { // 处理内存溢出异常
sizeCtl = Integer.MAX_VALUE;
return;
}
nextTable = nextTab;
//起始的transferIndex为旧的数组长度
transferIndex = n;
}
int nextn = nextTab.length;
//转发节点是为了解决 部分槽扩容完成,但是在get调用的时候还是调用的原来的hashmap的情况,此节点保存新的数组的引用
//当调用get时,会通过此节点转发到新数组进行访问
ForwardingNode<K,V> fwd = new ForwardingNode<K,V>(nextTab);
boolean advance = true;
boolean finishing = false; // to ensure sweep before committing nextTab
//i为遍历索引,bound为遍历的边界。如果成功拿到一个任务,i=nextIndex-1,bound=nextIndex=stride;
//如果拿不到任务,i=0,bound=0;
for (int i = 0, bound = 0;;) {
Node<K,V> f; int fh;
//只要有一个if分支执行就不再死循环,用while的目的是确保CAS操作能成功(自旋),拿到一个stride迁移任务
while (advance) {
int nextIndex, nextBound;
//对数组遍历
if (--i >= bound || finishing)
advance = false;
//整个hashmap完成
else if ((nextIndex = transferIndex) <= 0) {
i = -1;
advance = false;
}
//为当前线程分配一个stride,CAS成功,拿到一个stride迁移任务,不成功,继续while循环
else if (U.compareAndSwapInt
(this, TRANSFERINDEX, nextIndex,
nextBound = (nextIndex > stride ?
nextIndex - stride : 0))) {
bound = nextBound;
i = nextIndex - 1;
advance = false;
}
}
//越界,说明整个数组已遍历完成
if (i < 0 || i >= n || i + n >= nextn) {
int sc;
//整个数组扩容完成
if (finishing) {
nextTable = null;
//当前table换成nextTab
table = nextTab;
sizeCtl = (n << 1) - (n >>> 1);
return;
}
if (U.compareAndSwapInt(this, SIZECTL, sc = sizeCtl, sc - 1)) {
if ((sc - 2) != resizeStamp(n) << RESIZE_STAMP_SHIFT)
return;
finishing = advance = true;
i = n; // recheck before commit
}
}
//tab[i]迁移完成,赋值一个ForwardingNode
else if ((f = tabAt(tab, i)) == null)
advance = casTabAt(tab, i, null, fwd);
//tab[i]的位置已经在迁移中
else if ((fh = f.hash) == MOVED)
advance = true; // already processed
//对tab[i]进行迁移,tab[i]可能是一个链表或红黑树
else {
synchronized (f) {
if (tabAt(tab, i) == f) {
Node<K,V> ln, hn;
if (fh >= 0) {//链表
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);
}
/**
* 因为数组长度为2的幂次方,所以 hashCode%tab.length = hashCode & (tab.length-1)。
* 说明原来处于第i个位置的元素在新的数组中也一定处于第i个位置或者i+n个位置
* 例如:数据长度为8,扩容后为16.
* 设hashCode=7; 7%8=7;7%16=7;位置不变
* 设hashCode=16; 16%8=0;16%16=0;位置不变
* 设hashCode=25; 25%8=1;25%16=9;后移8个位置
* 设hashCode=36; 36%8=4;36%16=4;位置不变
* 所以把tab[i]的链表或红黑树分为两部分,一部分到nextTab[i]位置,另一部分到nextTab[i+n]位置
* 然后把tab[i]的位置指向一个ForwardingNode节点
*
*/
setTabAt(nextTab, i, ln);
setTabAt(nextTab, i + n, hn);
setTabAt(tab, i, fwd);
advance = true;
}
else if (f instanceof TreeBin) {//红黑树
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;
//和链表分支中的代码效果相同
setTabAt(nextTab, i, ln);
setTabAt(nextTab, i + n, hn);
setTabAt(tab, i, fwd);
advance = true;
}
}
}
}
}
}
static final <K,V> void setTabAt(Node<K,V>[] tab, int i, Node<K,V> v) {
U.putObjectVolatile(tab, ((long)i << ASHIFT) + ABASE, v);
}
}