ConcurrentHashMap源码分析(基于JDK1.8)

1. 首先来看一下ConcurrentHashMap类的定义:

public class ConcurrentHashMap<K,V> extends AbstractMap<K,V>
    implements ConcurrentMap<K,V>, Serializable {

由上述代码可见, ConcurrentHashMap扩展了AbstractMap类, 实现了ConcurrentMap接口和Serializable接口.

2. 键值对的存储

//ConcurrentHashMap类内部采用Node类存储键值对
    static class Node implements Map.Entry {
        final int hash;
        final K key;
        volatile V val;  //采用volatile关键字修饰
        volatile Node next;   //采用volatile关键字修饰

        Node(int hash, K key, V val, Node next) {
            this.hash = hash;
            this.key = key;
            this.val = val;
            this.next = next;
        }

        public final K getKey()       { return key; }
        public final V getValue()     { return val; }
        public final int hashCode()   { return key.hashCode() ^ val.hashCode(); }
        public final String toString(){ return key + "=" + val; }
        public final V setValue(V value) {   //不支持修改value, 否则将会抛出异常
            throw new UnsupportedOperationException();
        }

        public final boolean equals(Object o) {
            Object k, v, u; Map.Entry e;
            return ((o instanceof Map.Entry) &&
                    (k = (e = (Map.Entry)o).getKey()) != null &&
                    (v = e.getValue()) != null &&
                    (k == key || k.equals(key)) &&
                    (v == (u = val) || v.equals(u)));
        }
        //查找当前节点之后的链表,若是存在则返回相应的Node; 否则返回Null.
        Node find(int h, Object k) {
            Node e = this;
            if (k != null) {
                do {
                    K ek;
                    if (e.hash == h &&
                        ((ek = e.key) == k || (ek != null && k.equals(ek))))
                        return e;
                } while ((e = e.next) != null);
            }
            return null;
        }
    }

3. ConcurrentHashMap类中的成员变量

//当值为-1时, 代表数组正在被初始化;
//按照源码注释翻译,当值为-(1+扩容线程数), 代表数组正在被多个线程扩容。但是其实不是这样的,当线程进行扩容时,会根据resizeStamp函数生成一个基数戳rs,然后((rs<
//当table为null时, 代表要初始化的容量大小; 否则代表下次要扩容的容量
private transient volatile int sizeCtl;

//ConcurrentHashMap的最大容量 2^30
private static final int MAXIMUM_CAPACITY = 1 << 30;

//ConcurrentHashMap的默认容量 2^4
private static final int DEFAULT_CAPACITY = 16;

//hash值为-1处的节点代表forwarding node
static final int MOVED     = -1; 

//和key对应hash值进行与操作, 将hash值最高位置0
static final int HASH_BITS = 0x7fffffff; // usable bits of normal node hash

//用于生成当前数组对应的基数戳
private static int RESIZE_STAMP_BITS = 16;

//将基数戳左移的位数,保证左移后的基数戳为负值,然后再加上n+1,表示n个线程正在扩容
private static final int RESIZE_STAMP_SHIFT = 32 - RESIZE_STAMP_BITS;

//表示最多能有多少个线程能够帮助进行扩容,因为sizeCtl只有低16位用于标识,所以最多只有2^16-1个线程帮助扩容
private static final int MAX_RESIZERS = (1 << (32 - RESIZE_STAMP_BITS)) - 1;

//数组位置中红黑树根节点的hash值为-2,小于0
static final int TREEBIN   = -2; 

//将HASH_BITS和普通节点的hash相与,将hash值最高位置0,从而保证普通节点的hash值都是>=0的
static final int HASH_BITS = 0x7fffffff; 

//扩容线程所负责的区间大小最低为16,避免发生大量的内存冲突
private static final int MIN_TRANSFER_STRIDE = 16;

//用于扩容过程中,指示原数组下一个分割区间的上界位置
private transient volatile int transferIndex;

//只有当数组处于扩容过程时,nextTable才不为null;否则其他时刻,nextTable为null;
//nextTable主要用于扩容过程中指向扩容后的新数组
private transient volatile Node[] nextTable;

//节点数组,用于存储键值对,当第一次插入时进行初始化。
transient volatile Node[] table;

4. 构造方法

//默认构造方法
    public ConcurrentHashMap() {
    }

//用户自定义初始化容量作为参数
    public ConcurrentHashMap(int initialCapacity) {
        if (initialCapacity < 0)
            throw new IllegalArgumentException();
        int cap = ((initialCapacity >= (MAXIMUM_CAPACITY >>> 1)) ?
                   MAXIMUM_CAPACITY :
                   tableSizeFor(initialCapacity + (initialCapacity >>> 1) + 1));   //对用户输入的初始化容量修剪为2^n次方, 
        this.sizeCtl = cap;
    }

5. 扩容过程

在扩容过程中, 正在扩容的线程会将正在转移的table节点标记为ForwardingNode, 其他线程若是查找到某个节点为ForwardingNode类型节点, 则查找下一个table节点辅助进行扩容操作, ForwardingNode源代码如下:

//ForwardingNode是Node的子类型
static final class ForwardingNode<K,V> extends Node<K,V> {
        final Node[] nextTable;    //设置辅助扩容线程的下一段table
        ForwardingNode(Node[] tab) {
            super(MOVED, null, null, null);
            this.nextTable = tab;
        }

        Node find(int h, Object k) {
            outer: for (Node[] tab = nextTable;;) {
                Node e; int n;
                if (k == null || tab == null || (n = tab.length) == 0 ||
                    (e = tabAt(tab, (n - 1) & h)) == null)  //查找hash数组位置h处的Node
                    return null;
                for (;;) {
                    int eh; K ek;
                    if ((eh = e.hash) == h &&
                        ((ek = e.key) == k || (ek != null && k.equals(ek))))    //查找到key相同的Node
                        return e;
                    if (eh < 0) {
                        if (e instanceof ForwardingNode) {
                            tab = ((ForwardingNode)e).nextTable;   //递归查询下一个ForwardingNode
                            continue outer;
                        }
                        else
                            return e.find(h, k);   //查找链表
                    }
                    if ((e = e.next) == null)
                        return null;
                }
            }
        }
    }
//扩容详细过程
private final void transfer(Node[] tab, Node[] nextTab) {
        int n = tab.length, stride;
        if ((stride = (NCPU > 1) ? (n >>> 3) / NCPU : n) < MIN_TRANSFER_STRIDE)
            stride = MIN_TRANSFER_STRIDE; // 每个线程所负责转移的数组的区间最少为MIN_TRANSFER_STRIDE=16,也就是说数组的连续16个位置都是由这个线程来进行转移,其他线程不允许接触这连续的16个位置,必须发生线程之间大量的内存冲突。换另一个角度来说,每个线程负责连续16个大小区间的数组转移。
        if (nextTab == null) {            // 初始化生成新的扩容数组
            try {
                @SuppressWarnings("unchecked")
                Node[] nt = (Node[])new Node[n << 1];  //新创建两倍原数组大小的新数组
                nextTab = nt;
            } catch (Throwable ex) {      // try to cope with OOME
                sizeCtl = Integer.MAX_VALUE;
                return;
            }
            nextTable = nextTab;  //nextTable为类成员变量,只有在扩容的过程中有作用,在其他时刻都是null值。nextTable指向新数组
            transferIndex = n;   //转移后的节点偏移量
        }
        int nextn = nextTab.length;
        ForwardingNode fwd = new ForwardingNode(nextTab);
        boolean advance = true;   //遍历
        boolean finishing = false; //保证在提交扩容后的新数组时,原数组中的所有元素都已经被遍历
        for (int i = 0, bound = 0;;) {
            Node f; int fh;
            while (advance) {
                int nextIndex, nextBound;
                if (--i >= bound || finishing)   //bound为数组区间下限值,i为当前转移数组的位置,--i处理转移下一个节点位置,从后往前处理
                    advance = false;  //退出while循环
                else if ((nextIndex = transferIndex) <= 0) {   //表示原数组已经分割完了
                    i = -1;
                    advance = false;  //退出while循环
                }
                else if (U.compareAndSwapInt
                         (this, TRANSFERINDEX, nextIndex,    
                          nextBound = (nextIndex > stride ?
                                       nextIndex - stride : 0))) {  //CAS操作修改transferIndex值,代表下一个线程转移原数组的节点的位置
                    bound = nextBound;  //设置当前线程转移原数组区间的下限值
                    i = nextIndex - 1;  //从后往前处理
                    advance = false;  //退出while循环
                }
            }
            if (i < 0 || i >= n || i + n >= nextn) {
                int sc;
                if (finishing) {   //扩容完成
                    nextTable = null;   //将nextTable置为null,表示当前扩容过程完成
                    table = nextTab;    //table指向扩容后的新数组
                    sizeCtl = (n << 1) - (n >>> 1);  //将szieCtl设置为正数,设置为原数组的3/2,即新数组的3/4
                    return;
                }
                if (U.compareAndSwapInt(this, SIZECTL, sc = sizeCtl, sc - 1)) {
                    if ((sc - 2) != resizeStamp(n) << RESIZE_STAMP_SHIFT)   //因为只有一个线程扩容时sc=resizeStamp(n)+2,所以该if语句是在最后一个线程完成扩容操作时,将finishing置为true,表示正确完成。
                        return;
                    finishing = advance = true;
                    i = n; // recheck before commit
                }
            }
            else if ((f = tabAt(tab, i)) == null)
                advance = casTabAt(tab, i, null, fwd);   //将原数组相应位置直接设置为fwd,表示该位置已经遍历过
            else if ((fh = f.hash) == MOVED)
                advance = true; // 表示该数组位置已经被其他线程处理过了
            else {  //否则需要将原数组位置相应元素复制到新数组上
                synchronized (f) {   //上锁
                    if (tabAt(tab, i) == f) {   //再次核对,防止其他线程对该hash值进行修改
                        Node ln, hn;
                        if (fh >= 0) {   //说明该位置存放的是普通节点
                            int runBit = fh & n;  //判断原数组中的节点的hash的 log(n)位为0或者1
                            Node lastRun = f;
                            for (Node 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;  //指向链表的最后出现连续log(n)位为0的第一个节点
                                hn = null;
                            }
                            else {     
                                hn = lastRun;   //指向链表的最后出现连续log(n)位为1的第一个节点
                                ln = null;
                            }
                            for (Node 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(ph, pk, pv, ln);
                                else
                                    hn = new Node(ph, pk, pv, hn);
                            }
                            setTabAt(nextTab, i, ln);   //将hash值的 log(n) 位为0的节点链表复制到新数组对应原来数组的位置
                            setTabAt(nextTab, i + n, hn);  //将Hash值的 log(n) 位为1的节点链表复制到新数组对应原来数组位置+n
                            setTabAt(tab, i, fwd);  //将该数组位置设置为已处理
                            advance = true;
                        }
                        else if (f instanceof TreeBin) {   //说明该数组位置是红黑树根节点
                            TreeBin t = (TreeBin)f;
                            TreeNode lo = null, loTail = null;
                            TreeNode hi = null, hiTail = null;
                            int lc = 0, hc = 0;
                            for (Node e = t.first; e != null; e = e.next) {
                                int h = e.hash;
                                TreeNode p = new TreeNode
                                    (h, e.key, e.val, null, null);
                                if ((h & n) == 0) {   //判断红黑树中节点的hash值的 log(n) 位为0,说明该节点应该存放到新数组中对应原数组的位置
                                    if ((p.prev = loTail) == null)
                                        lo = p;
                                    else
                                        loTail.next = p;
                                    loTail = p;
                                    ++lc;
                                }
                                else {    //判断红黑树中节点的hash值的 log(n) 位为1,说明该节点应该存放到新数组中对应原数组位置+n
                                    if ((p.prev = hiTail) == null)
                                        hi = p;
                                    else
                                        hiTail.next = p;
                                    hiTail = p;
                                    ++hc;
                                }
                            }
                            //根据链表中节点的个数和UNTREEIFY_THRESHOLD进行比较,如果小于等于,则不需要将链表转换为红黑树;如果大于,则需要将链表转换为红黑树
                            ln = (lc <= UNTREEIFY_THRESHOLD) ? untreeify(lo) :   
                                (hc != 0) ? new TreeBin(lo) : t;
                            hn = (hc <= UNTREEIFY_THRESHOLD) ? untreeify(hi) :
                                (lc != 0) ? new TreeBin(hi) : t;
                            setTabAt(nextTab, i, ln);   //复制到新数组中
                            setTabAt(nextTab, i + n, hn);  //复制到新数组中
                            setTabAt(tab, i, fwd);  //将原数组中相应位置为fwd,表示该位置已经被处理过
                            advance = true;  //继续进行遍历
                        }
                    }
                }
            }
        }
    }
//helpTransfer函数的主要作用是如果有线程正在进行扩容操作,则帮助其他线程进行扩容操作
    final Node[] helpTransfer(Node[] tab, Node f) {
        Node[] nextTab; int sc;
        if (tab != null && (f instanceof ForwardingNode) &&
            (nextTab = ((ForwardingNode)f).nextTable) != null) {   //帮助进行扩容
            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)) {  //CAS修改sizeCtl=sizeCtl+1,表示新增加一个线程辅助扩容
                    transfer(tab, nextTab);
                    break;
                }
            }
            return nextTab;
        }
        return table;
    }

6. put方法

    public V put(K key, V value) {
        return putVal(key, value, false);
    }

由上述代码可见, put方法调用了putVal方法, putVal方法如下:

    final V putVal(K key, V value, boolean onlyIfAbsent) {
        if (key == null || value == null) throw new NullPointerException();
        //获取key的hash值, 并将hash值传递给spread函数.
        //spread函数的主要作用是将hash值高16位和低16位进行异或操作, 对hash值进行优化, 避免在生成hash值位置时只考虑低16位.
        int hash = spread(key.hashCode());   
        int binCount = 0;
        for (Node[] tab = table;;) {   //类似死循环,直到插入成功
            Node f; int n, i, fh;
            if (tab == null || (n = tab.length) == 0)
                tab = initTable();   //如果tab为null, 则需要对tab进行初始化.
            else if ((f = tabAt(tab, i = (n - 1) & hash)) == null) {   //如果hash值对应位置处为null, 直接添加即可
                if (casTabAt(tab, i, null,
                             new Node(hash, key, value, null)))    //无需加锁, 进行CAS操作, 在i位置处添加新hash对应的键值对
                    break;                   
            }
            else if ((fh = f.hash) == MOVED)   //f.hash==-1说明其他线程正在进行扩容操作
                tab = helpTransfer(tab, f);   //调用helpTransfer函数进行扩容操作
            else {  //否则进行插入操作
                V oldVal = null;
                synchronized (f) {   //对f节点加锁
                    if (tabAt(tab, i) == f) {   //重复检查,避免多线程导致的修改
                        if (fh >= 0) {  //说明该节点为普通节点
                            binCount = 1;
                            for (Node 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 pred = e;
                                if ((e = e.next) == null) {
                                    pred.next = new Node(hash, key,
                                                              value, null);   //插入到链表的末尾
                                    break;
                                }
                            }
                        }
                        else if (f instanceof TreeBin) {  //说明该节点为红黑树的根节点
                            Node p;
                            binCount = 2;
                            if ((p = ((TreeBin)f).putTreeVal(hash, key,
                                                           value)) != null) {
                                oldVal = p.val;
                                if (!onlyIfAbsent)
                                    p.val = value;
                            }
                        }
                    }
                }
                if (binCount != 0) {
                    if (binCount >= TREEIFY_THRESHOLD)  //根据链表的长度判断是否需要将链表转换为红黑树结构
                        treeifyBin(tab, i);  //调用treeifyBin方法将链表改为红黑树结构
                    if (oldVal != null)
                        return oldVal;
                    break;
                }
            }
        }
        addCount(1L, binCount);  //调用addCount函数,将容器大小加1,并判断是否需要进行扩容
        return null;
    }

addCount函数的源代码如下:

private final void addCount(long x, int check) {
        CounterCell[] as; long b, s;
        if ((as = counterCells) != null ||
            !U.compareAndSwapLong(this, BASECOUNT, b = baseCount, s = b + x)) {   //利用CAS操作更新baseCount
            CounterCell a; long v; int m;
            boolean uncontended = true;
            if (as == null || (m = as.length - 1) < 0 ||
                (a = as[ThreadLocalRandom.getProbe() & m]) == null ||
                !(uncontended =
                  U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))) {
                fullAddCount(x, uncontended);
                return;
            }
            if (check <= 1)  
                return;
            s = sumCount();
        }
        if (check >= 0) {   //判断是否需要扩容
            Node[] tab, nt; int n, sc;
            while (s >= (long)(sc = sizeCtl) && (tab = table) != null &&
                   (n = tab.length) < MAXIMUM_CAPACITY) {
                int rs = resizeStamp(n);   //生成一个基数戳
                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))   //将sizeCtl加1,表示新增加一个线程进行辅助操作
                        transfer(tab, nt);
                }
                else if (U.compareAndSwapInt(this, SIZECTL, sc,
                                             (rs << RESIZE_STAMP_SHIFT) + 2))   //基数戳rs<
                    transfer(tab, null);
                s = sumCount();
            }
        }
    }

treeifyBin方法的源代码如下:

    private final void treeifyBin(Node[] tab, int index) {
        Node b; int n, sc;
        if (tab != null) {
            if ((n = tab.length) < MIN_TREEIFY_CAPACITY)
                tryPresize(n << 1);   //如果数组的长度小于 MIN_TREEIFY_CAPACITY=64,则调用tryPresize方法进行扩容,而不是直接改为红黑树结构
            else if ((b = tabAt(tab, index)) != null && b.hash >= 0) {   //需要改为红黑树结构
                synchronized (b) {  //将node b加锁
                    if (tabAt(tab, index) == b) {  //重复检查,避免多线程导致的修改
                        TreeNode hd = null, tl = null;
                        for (Node e = b; e != null; e = e.next) {
                            TreeNode p =
                                new TreeNode(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(hd));   //将TreeNode链表封装到TreeBin对象中,由TreeBin负责红黑树的生成,将数组相应位置设置为TreeBin对象
                    }
                }
            }
        }
    }

tryPresize方法源代码如下:

//将原数组进行两倍扩容
    private final void tryPresize(int size) {
        int c = (size >= (MAXIMUM_CAPACITY >>> 1)) ? MAXIMUM_CAPACITY :
            tableSizeFor(size + (size >>> 1) + 1);
        int sc;
        while ((sc = sizeCtl) >= 0) {   //说明数组不是处于扩容状态
            Node[] tab = table; int n;
            if (tab == null || (n = tab.length) == 0) {   //如果数组为null
                n = (sc > c) ? sc : c;
                if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {   //将sc设置为-1,表示当前数组正在进行扩容操作
                    try {
                        if (table == tab) {
                            @SuppressWarnings("unchecked")
                            Node[] nt = (Node[])new Node[n];   //生成新的数组
                            table = nt;  //table指向新数组
                            sc = n - (n >>> 2);  //sc保存新数组的上限值
                        }
                    } finally {
                        sizeCtl = sc;
                    }
                }
            }
            else if (c <= sc || n >= MAXIMUM_CAPACITY)
                break;
            else if (tab == table) {
                int rs = resizeStamp(n);
                if (sc < 0) {
                    Node[] nt;
                    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))  //辅助扩容操作,将sizeCtl加1,表示新增加一个线程辅助扩容
                        transfer(tab, nt);
                }
                else if (U.compareAndSwapInt(this, SIZECTL, sc,
                                             (rs << RESIZE_STAMP_SHIFT) + 2))   //开始进行扩容,通过CAS操作将sizeCtl置为负值,代表只要一个线程在进行扩容操作。
                    transfer(tab, null);
            }
        }
    }

table的初始化函数initTable过程如下:

private final Node[] initTable() {
        Node[] tab; int sc;
        while ((tab = table) == null || tab.length == 0) {
            if ((sc = sizeCtl) < 0)   //如果sizeCtl<0, 根据规定, 这代表有其他线程正在初始化或者扩容
                Thread.yield(); // 暂停初始化步骤, 让出处理器, 进行旋转
            else if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {      // 否则进行CAS操作, 将sizeCtl置为-1, 代表当前线程正在进行初始化操作
                try {
                    if ((tab = table) == null || tab.length == 0) {
                        int n = (sc > 0) ? sc : DEFAULT_CAPACITY;
                        @SuppressWarnings("unchecked")
                        Node[] nt = (Node[])new Node[n];
                        table = tab = nt;
                        sc = n - (n >>> 2);   //减去1/4, 剩下3/4
                    }
                } finally {
                    sizeCtl = sc;   // 作为下一次扩容的临界值
                }
                break;
            }
        }
        return tab;
    }

7. get方法

public V get(Object key) {
        Node[] tab; Node e, p; int n, eh; K ek;
        int h = spread(key.hashCode());   //获取相应的hash值
        if ((tab = table) != null && (n = tab.length) > 0 &&
            (e = tabAt(tab, (n - 1) & h)) != null) {
            if ((eh = e.hash) == h) {
                if ((ek = e.key) == key || (ek != null && key.equals(ek)))
                    return e.val;
            }
            else if (eh < 0)   //说明该节点位置为红黑树节点
                return (p = e.find(h, key)) != null ? p.val : null;  //调用find方法在红黑树中进行查找
            while ((e = e.next) != null) {   //遍历链表
                if (e.hash == h &&
                    ((ek = e.key) == key || (ek != null && key.equals(ek))))
                    return e.val;
            }
        }
        return null;
    }

8. size方法

public int size() {
        long n = sumCount();  //调用内部sumCount方法
        return ((n < 0L) ? 0 :
                (n > (long)Integer.MAX_VALUE) ? Integer.MAX_VALUE :
                (int)n);
}

final long sumCount() {
        CounterCell[] as = counterCells; CounterCell a;
        long sum = baseCount;
        if (as != null) {
            for (int i = 0; i < as.length; ++i) {
                if ((a = as[i]) != null)
                    sum += a.value;
            }
        }
        return sum;  
}

实际上在ConcurrentHashMap内部使用了如下变量来保存map中键值对个数

private transient volatile long baseCount;

因为在调用size()获取当前ConcurrentHashMap对象中的键值对个数时,返回的值是估算值,不是精确值,因为在查询个数的同时可能存在多个线程在进行插入、删除操作,不能将所有线程停下进行统计。

你可能感兴趣的:(JDK1.8集合类源代码)