深刻理解ConcurrentHashMap

一, 什麼是ConcurrentHashMap

ConcurrentHashMap和HashMap同樣是一個用來存儲鍵值對<key,value>的集合類,但和HashMap不一樣的是ConcurrentHashMap是線程安全的,也就是多個線程同時對ConcurrentHashMap進行修改或者刪除增長操做不會出現數據錯誤的問題.

二, 實現原理

和HashMap同樣採用數組+鏈表+紅黑樹實現

但和HashMap不一樣的是,數組中存儲的節點類型有所增長,包括Node<key,value>,TreeNode<key,value>,ForwardingNode<key,value>,新增這個節點的目的就是爲了線程併發協助擴容時使用

三, 基本屬性介紹

//01111111111111111111111111111111 該值能夠保證計算出來的哈希值爲正數
static final int HASH_BITS = 0x7fffffff; // usable bits of normal node hash

//該屬性用在擴容時生成一個負值,表示正在擴容
//The number of bits used for generation stamp in sizeCtl.
//sizeCtl中用於生成戳記的位數。
//Must be at least 6 for 32bit arrays.
//對於32位數組，必須至少爲6。
private static int RESIZE_STAMP_BITS = 16;

//和上面同樣,也是爲了在擴容時生成一個負值,具體在代碼中解釋
//The bit shift for recording size stamp in sizeCtl.
//在sizeCtl中記錄大小戳的位移位。
private static final int RESIZE_STAMP_SHIFT = 32 - RESIZE_STAMP_BITS;

//表示當前桶位正在被遷移
//Encodings for Node hash fields. See above for explanation.
static final int MOVED = -1;

//表示當前桶是以樹來存儲節點的
static final int TREEBIN = -2;

//Number of CPUS, to place bounds on some sizings
//cpu的數量,用來計算元素數量時限制CounterCell數組大小
static final int NCPU = Runtime.getRuntime().availableProcessors();

/**
 * The next table to use; non-null only while resizing.
 * 用來擴容的哈希表
 */
private transient volatile Node<K, V>[] nextTable;

/**
 * Base counter value, used mainly when there is no contention, but also as a fallback during table    initialization 
 * races. Updated via CAS.
 * 哈希表元素數量,經過longAdder來維護
 */
private transient volatile long baseCount;

/**
 * Table initialization and resizing control.
 * 哈希表初始化和擴容大小控制.
 * When negative, the table is being initialized or resized:
 * 當這個值爲負數時,表示哈希表正在初始化或從新計算大小
 * -1 for initialization,
 * -1 表示正在初始化了
 * else -(1 + the number of active resizing threads).
 * 表示哈希表正在擴容,-(1+n),表示此時有n個線程正在共同完成哈希表的擴容
 * Otherwise, when table is null, holds the initial table size to use upon creation,or 0 for default.
 * 不然,當哈希表爲空時, 保留要建立哈希表的大小0或默認(16)
 * After initialization, holds the next element count value upon which to resize the table.
 * 初始化完成以後,保留下一次須要擴容的閾值
 */
private transient volatile int sizeCtl;

/**
 * The next table index (plus one) to split while resizing.
 * 擴容時的當前轉移下標
 */
private transient volatile int transferIndex;

/**
 * Spinlock (locked via CAS) used when resizing and/or creating CounterCells.
 * 獲取計算集合元素容量的CounterCell對象的鎖
 */
private transient volatile int cellsBusy;

/**
 * Table of counter cells. When non-null, size is a power of 2.
 * 計算元素數量的數組
 */
private transient volatile CounterCell[] counterCells;

四, 構造函數

/**
 * 和HashMap構造函數不一樣的是,數組容量的計算老是大於傳入容量的2的冪
 * 即若是傳入32則數組初始容量爲64,而不是32,而HashMap計算出來爲32
 */
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;
}

五, 經常使用方法介紹

/**
 * 這個方法就是HashMap中的hash方法,用來計算哈希值
 */
static final int spread(int h) {
    return (h ^ (h >>> 16)) & HASH_BITS;
}

獲取節點

public V get(Object key) {
    Node<K, V>[] tab;
    Node<K, V> e, p;
    int n, eh;
    K ek;
    //計算散列值
    int h = spread(key.hashCode());
    //計算下標(這一塊同HashMap再也不贅述)
    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)
            //哈希值小於0,表示爲樹節點,從樹中尋找,這一步和HashMap一致
            return (p = e.find(h, key)) != null ? p.val : null;
        //在鏈表中尋找
        while ((e = e.next) != null) {
            if (e.hash == h && ((ek = e.key) == key || (ek != null && key.equals(ek))))
                return e.val;
        }
    }
    return null;
}

插入節點

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();
    //計算哈希值
    int hash = spread(key.hashCode());
    //插入桶的節點數量
    int binCount = 0;
    //使用死循環,目的是可能有的線程正在協助擴容,以後還須要插入或者更新,或者須要操做的節點所在的桶已經被其餘線程鎖定,須要等待其餘線程執行完以後再執行
    for (Node<K, V>[] tab = table; ; ) {
        Node<K, V> f;
        int n, i, fh;
        if (tab == null || (n = tab.length) == 0)
            //初始化哈希表
            tab = initTable();
        else if
            //計算下標,而且計算該下標是否有元素
            ((f = tabAt(tab, i = (n - 1) & hash)) == null) {
            //cas插入,這一步不須要鎖,由於當前桶爲空
            if (casTabAt(tab, i, null, new Node<K, V>(hash, key, value, null)))
                break;
        } else if
            //表明當前節點已經被移動,正在擴容,須要當前線程協助擴容
            ((fh = f.hash) == MOVED)
            tab = helpTransfer(tab, f);
        else {
            V oldVal = null;
            //鎖住頭節點,保證全部線程的插入都是線程安全的
            synchronized (f) {
                //這一步判斷的緣由是,可能插入元素以後會形成鏈表樹化,須要插入的位置已經發生了變化
                if (tabAt(tab, i) == f) {
                    if (fh >= 0) {
                        //當前桶上有多少個節點
                        binCount = 1;
                        for (Node<K, V> e = f; ; ++binCount) {
                            K ek;
                            //查找到了key相同的節點,直接修改值並返回舊的值
                            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;
                            //沒有找到相同的key,直接向鏈表尾部插入節點
                            if ((e = e.next) == null) {
                                pred.next = new Node<K, V>(hash, key, value, null);
                                break;
                            }
                        }
                    } else if (f instanceof TreeBin) {
                        //給樹裏面插入節點
                        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) {
                //須要樹化
                if (binCount >= TREEIFY_THRESHOLD)
                    treeifyBin(tab, i);
                //修改舊值,直接將舊值返回
                if (oldVal != null)
                    return oldVal;
                break;
            }
        }
    }
    //計算節點數量
    addCount(1L, binCount);
    return null;
}

初始化哈希表

private final Node<K, V>[] initTable() {
    Node<K, V>[] tab;
    int sc;
    while ((tab = table) == null || tab.length == 0) {
        //小於0表示正在初始化或者正在擴容,讓出cpu
        if ((sc = sizeCtl) < 0)
            Thread.yield(); // lost initialization race; just spin
        else if
            //判斷sc是否與SIZECTL是否相等,若是相等,則將SIZECTL設置爲-1,表示當前正在初始化(只有一個線程能進行此操做,其餘線程會被擋在前面的判斷上)
            (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
            try {
                //防止有線程已經初始化了1
                if ((tab = table) == null || tab.length == 0) {
                    //該sc若是在構造器上傳入了,則會被計算爲大於其的2次冪,不然會按照默認值初始化
                    int n = (sc > 0) ? sc : DEFAULT_CAPACITY;
                    @SuppressWarnings("unchecked")
                    Node<K, V>[] nt = (Node<K, V>[]) new Node<?, ?>[n];
                    table = tab = nt;
                    //設置下一次擴容的閾值 n - (n >>> 2) = n - n / 4 = (3 / 4) * n = 0.75n,即下一次的擴容閾值爲當前哈希表數量的0.75*n
                    sc = n - (n >>> 2);
                }
            } finally {
                //設置sizeCtl爲-1,表示初始化動做已經有線程在執行了
                sizeCtl = sc;
            }
            break;
        }
    }
    return tab;
}

計算節點數量

private final void addCount(long x, int check) {
    CounterCell[] as;
    long b, s;
    /*
     * 維護數組長度
     */
    //嘗試cas直接修改值,若是修改失敗
    if ((as = counterCells) != null || !U.compareAndSwapLong(this, BASECOUNT, b = baseCount, s = b + x)) {
        CounterCell a;
        long v;
        int m;
        boolean uncontended = true;
        //數組爲空或者長度小於0或者對應的位置爲空或者直接修改數組對應位置上的值失敗,則進行修改操做
        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;
        }
        //桶上的節點數量小於等於1,不須要判斷擴容,直接退出
        if (check <= 1)
            return;
        //獲取當前數組的元素數量
        s = sumCount();
    }
    /*
     * 判斷是否須要擴容
     */
    if (check >= 0) {
        Node<K, V>[] tab, nt;
        int n, sc;
        //當前節點數量大於擴容閾值,而且數組不爲空而且數組長度小於最大值則須要擴容
        while (s >= (long) (sc = sizeCtl) && (tab = table) != null && (n = tab.length) < MAXIMUM_CAPACITY) {
            //獲取一個負值
            int rs = resizeStamp(n);
            //若是sc小於0,說明正在擴容,須要協助擴容
            if (sc < 0) {
                //判斷擴容是否完成
                if ((sc >>> RESIZE_STAMP_SHIFT) != rs || sc == rs + 1 || sc == rs + MAX_RESIZERS || (nt = nextTable) == null || transferIndex <= 0)
                    break;
                //協助擴容,這裏sc+1表明新加入一個線程協助擴容
                if (U.compareAndSwapInt(this, SIZECTL, sc, sc + 1))
                    transfer(tab, nt);
            } else if
                /*
                 * 假設 rs = 00000000 00000000 10000000 00000000
                 * 將其向左移16位結果爲 10000000 00000000 00000000 00000000 能夠看出該值爲負
                 * 這一步嘗試將sc設置爲負數
                 */
                (U.compareAndSwapInt(this, SIZECTL, sc, (rs << RESIZE_STAMP_SHIFT) + 2))
                //將舊數組置空,裏面會建立一個新的數組
                transfer(tab, null);
            //計算集合元素數量
            s = sumCount();
        }
    }
}

private final void fullAddCount(long x, boolean wasUncontended) {
    int h;
    //獲取當前線程的hash值
    if ((h = ThreadLocalRandom.getProbe()) == 0) {
        ThreadLocalRandom.localInit();      // force initialization
        h = ThreadLocalRandom.getProbe();
        wasUncontended = true;
    }
    //檢測是否有衝突,若是最後一個桶不爲null,則爲true
    boolean collide = false;
    for (; ; ) {
        CounterCell[] as;
        CounterCell a;
        int n;
        long v;
        //數組若是不爲空,則優先對CounterCell裏面的counterCell的value進行累加
        if ((as = counterCells) != null && (n = as.length) > 0) {
            //當前位置爲空
            if ((a = as[(n - 1) & h]) == null) {
                //當前沒有線程嘗試修改該值
                if (cellsBusy == 0) {
                    CounterCell r = new CounterCell(x);
                    //搶佔修改的鎖
                    if (cellsBusy == 0 && U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
                        boolean created = false;
                        try {
                            CounterCell[] rs;
                            int m, j;
                            if ((rs = counterCells) != null && (m = rs.length) > 0 && rs[j = (m - 1) & h] == null) {
                                rs[j] = r;
                                created = true;
                            }
                        } finally {
                            //釋放鎖
                            cellsBusy = 0;
                        }
                        if (created)
                            break;
                        continue;           // Slot is now non-empty
                    }
                }
                //搶佔失敗
                collide = false;
            } else if
                //桶位不爲空,從新計算線程hash值,繼續循環
                (!wasUncontended)       // CAS already known to fail
                wasUncontended = true;      // Continue after rehash
            /*
                     * 從新計算hash值以後,對應的桶位仍是不爲空,對value進行累加
                     * 嘗試cas對value加值
                     */
            else if (U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))
                break;
            //數組長度已經大於等於CPU的核數了,不須要再擴容了
            else if (counterCells != as || n >= NCPU)
                collide = false;
            //當沒有衝突,修改成有衝突,從新計算hash值,繼續循環
            else if (!collide)
                collide = true;
            else if
                //屢次循環沒有設置成功值,則對原數組進行擴容
                (cellsBusy == 0 && U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
                try {
                    if (counterCells == as) {// Expand table unless stale
                        //數組長度沒有超過cpu核數,將數組擴容兩倍
                        CounterCell[] rs = new CounterCell[n << 1];
                        for (int i = 0; i < n; ++i)
                            //擴容使用
                            rs[i] = as[i];
                        counterCells = rs;
                    }
                } finally {
                    cellsBusy = 0;
                }
                collide = false;
                continue;                   // Retry with expanded table
            }
            //從新計算隨機值
            h = ThreadLocalRandom.advanceProbe(h);
        } else if
            //初始進來數組爲空,須要初始化數組
            (cellsBusy == 0 && counterCells == as && U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
            boolean init = false;
            try {
                if (counterCells == as) {
                    CounterCell[] rs = new CounterCell[2];
                    rs[h & 1] = new CounterCell(x);
                    counterCells = rs;
                    init = true;
                }
            } finally {
                cellsBusy = 0;
            }
            if (init)
                break;
        } else if
            //數組爲空而且有其餘線程正在建立數組,嘗試直接對baseCount進行累加
            (U.compareAndSwapLong(this, BASECOUNT, v = baseCount, v + x))
            break;                          // Fall back on using base
    }
}

擴容並遷移

private final void transfer(Node<K, V>[] tab, Node<K, V>[] nextTab) {
    //stride表示遷移數據的區間
    int n = tab.length, stride;
    /*
     * 這裏計算每一個CPU負責遷移元素的個數
     * 若是這裏的跨度區間小於16,則按照最小區間16來計算
     */
    if ((stride = (NCPU > 1) ? (n >>> 3) / NCPU : n) < MIN_TRANSFER_STRIDE)
        stride = MIN_TRANSFER_STRIDE;
    //這裏表示爲第一個線程來擴容
    if (nextTab == null) {
        try {
            //擴容爲兩倍
            @SuppressWarnings("unchecked")
            Node<K, V>[] nt = (Node<K, V>[]) new Node<?, ?>[n << 1];
            nextTab = nt;
        } catch (Throwable ex) {
            sizeCtl = Integer.MAX_VALUE;
            return;
        }
        nextTable = nextTab;
        //遷移數據的index
        transferIndex = n;
    }
    //新擴容數組的長度
    int nextn = nextTab.length;
    //建立頭節點,該節點會被標識爲MOVED表示數據正在遷移中
    ForwardingNode<K, V> fwd = new ForwardingNode<K, V>(nextTab);

    boolean advance = true;
    boolean finishing = false; // to ensure sweep before committing nextTab
    //從後向前遷移
    for (int i = 0, bound = 0; ; ) {
        Node<K, V> f;
        int fh;
        while (advance) {
            int nextIndex, nextBound;
            //不屬於本身的遷移位置或者已經遷移完成直接退出
            if (--i >= bound || finishing)
                advance = false;
            else if
                //下一個遷移位置小於等於0直接退出
                ((nextIndex = transferIndex) <= 0) {
                i = -1;
                advance = false;
            } else if
                //計算遷移位置(多線程會劃分多個區間)
                (U.compareAndSwapInt(this, TRANSFERINDEX, nextIndex, nextBound = (nextIndex > stride ? nextIndex - stride : 0))) {
                bound = nextBound;
                i = nextIndex - 1;
                advance = false;
            }
        }
        if (i < 0 || i >= n || i + n >= nextn) {
            int sc;
            //判斷是否全部的線程都作完了任務
            if (finishing) {
                nextTable = null;
                table = nextTab;
                //等於0.75 * 2n,也就是新數組擴容2倍*擴容因子
                sizeCtl = (n << 1) - (n >>> 1);
                return;
            }
            if (U.compareAndSwapInt(this, SIZECTL, sc = sizeCtl, sc - 1)) {
                //判斷擴容是否成功
                if ((sc - 2) != resizeStamp(n) << RESIZE_STAMP_SHIFT)
                    return;
                finishing = advance = true;
                i = n; // recheck before commit
            }
        } else if
            //若是當前位置爲空,直接插入fwd節點,表示當前節點正在被遷移
            ((f = tabAt(tab, i)) == null)
            advance = casTabAt(tab, i, null, fwd);
        else if ((fh = f.hash) == MOVED)
            //正在被遷移,須要從新計算位置
            advance = true; // already processed
        else {
            //遷移代碼
            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);
                        //遷移完成,設置頭節點爲fwd
                        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);
                        //遷移完成,設置頭節點爲fwd
                        setTabAt(tab, i, fwd);
                        //從新計算位置繼續遷移
                        advance = true;
                    }
                }
            }
        }
    }
}

獲取節點數量

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

/**
 * 獲取哈希表中節點的數量(非線程安全)
 * 這個方法返回的數據不必定準確,由於可能在調用該方法的時候,有其餘線程正在嘗試給數組中的節點加值
 */
final long sumCount() {
    CounterCell[] as = counterCells;
    CounterCell a;
    long sum = baseCount;
    if (as != null) {
        for (CounterCell counterCell : as) {
            if ((a = counterCell) != null)
                sum += a.value;
        }
    }
    return sum;
}