深入理解ThreadLocal的"内存溢出"
程序员文章站
2022-05-14 18:55:21
...
背景
对ThreadLocal的实际使用场景一直有点模糊。在code review中大家对ThreadLocal是否会出现内存泄漏问题提出不同看法。故上网一探究竟,但是发现网上的说法不一,有的说会导致内存泄漏有的说不会,很难发现实战的结晶。
分析
结构
一个简洁的ThreadLocal类的内部结构如下
public class ThreadLocal<T> { static class ThreadLocalMap { static class Entry extends WeakReference<ThreadLocal> { Object value; Entry(ThreadLocal k, Object v) { super(k); value = v; } private ThreadLocal.ThreadLocalMap.Entry[] table; } } }
ThreadLocal类中定义了一个静态内部类ThreadLocalMap,ThreadLocalMap并没有实现Map接口,而是自己"实现"了一个Map,在ThreadLocalMap内部定义了一个静态内部类Entry继承自WeakReference,寻找一下对WeakReference的记忆—当所引用的对象在JVM内不再有强引用指向时,GC后weak reference将会被自动回收。
流程
然后,我们从创建的流程来看一下
public void set(T value) { Thread t = Thread.currentThread(); ThreadLocalMap map = getMap(t); if (map != null) map.set(this, value); else createMap(t, value); }当线程首次调用set方法,并不能获取到ThreadLocalMap,于是ThreadLocalMap被创建
void createMap(Thread t, T firstValue) { t.threadLocals = new ThreadLocalMap(this, firstValue); } ThreadLocalMap(ThreadLocal firstKey, Object firstValue){ table = new Entry[INITIAL_CAPACITY]; int i = firstKey.threadLocalHashCode & (INITIAL_CAPACITY - 1); table[i] = new Entry(firstKey, firstValue); size = 1; setThreshold(INITIAL_CAPACITY); }
可以看到ThreadLocalMap以当前ThreadLocal对象为key被创建,其内部存储结构如上,将key进行hash计算后,再将key和value放入Entry中,
注意一下上面t.threadLocals = new ThreadLocalMap(this, firstValue),实际上是一个Thread的成员变量在引用着这个ThreadLocalMap如下
public class Thread{ ThreadLocal.ThreadLocalMap threadLocals = null; }所以我们可以分析,当Thread运行结束后(没有线程池):
- 这个ThreadLocalMap对象会被GC回收
- ThreadLocalMap的成员变量table所指向的对象会被gc回收,这时注意Entry是继承了WeakReference的,所以Entry对象也会被gc回收
- value作为Entry的成员变量自然也会被gc回收
结论
这样看来,较为严谨的说法是,在不使用线程池的前提下,即使不调用remove方法,线程的"变量副本"也会被gc回收,即不会造成内存泄漏的情况。
问题
1、那在使用线程池的情况下呢?会不会出现内存泄漏的问题呢?我做了这样一个简单的小测试
public static void testThreadLocalExist(){ ExecutorService service = Executors.newSingleThreadExecutor(); for (int i = 0; i < 10; i++) { if(i == 0){ service.execute(new Runnable() { public void run() { System.out.println("Thread id is " + Thread.currentThread().getId()); threadLocal.set("variable"); } }); } else if(i > 0){ service.execute(new Runnable() { public void run() { if("variable".equals(threadLocal.get())){ System.out.println("Thread id " + Thread.currentThread().getId() + " got it !"); } } }); } } } 输出: Thread id is 9 Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it ! Thread id 9 got it !如上测试,我初始化了一个线程数量为1的线程池,为了保证每次线程池中获取到的都是同一个线程
那么根据这个测试可以看出,当线程从线程池中再次被调用的时候,这个"变量副本"是可以获取到的,即内存可能会发生泄漏,但没有实战的情况下,无法预估其影响。
2、那么当使用线程池的情况下,出于安全起见如何避免发生内存泄漏呢?在上面的测试中,做一点小小的变化
public static void testThreadLocalExist() {
ExecutorService service = Executors.newSingleThreadExecutor();
for (int i = 0; i < 10; i++) {
if (i == 0) {
service.execute(new Runnable() {
public void run() {
System.out.println("Thread id is " + Thread.currentThread().getId());
threadLocal.set("variable");
threadLocal.remove();
}
});
} else {
service.execute(new Runnable() {
public void run() {
if ("variable".equals(threadLocal.get())) {
System.out.println("Thread id " + Thread.currentThread().getId() + " get it !");
} else {
System.out.println("Thread id " + Thread.currentThread().getId() + " can't get it !");
}
}
});
}
}
}
输出:
Thread id is 9
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
Thread id 9 can't get it !
如上测试,在原来的基础上,在线程第一次运行完之前调用ThreadLocal的remove方法,然后再将线程放回线程池,这样当这个线程再次被调用时,"变量副本"已经不存在了。
当ThreadLocal在调用remove方法的时候,其实就是调用ThreadLocalMap的remove方法
public void remove() { ThreadLocalMap m = getMap(Thread.currentThread()); if (m != null) m.remove(this); }那就深入看看ThreadLocalMap的remove方法吧
private void remove(ThreadLocal key) { Entry[] tab = table; int len = tab.length; int i = key.threadLocalHashCode & (len-1); for (Entry e = tab[i]; e != null; e = tab[i = nextIndex(i, len)]) { if (e.get() == key) { e.clear(); expungeStaleEntry(i); return; } } }可以看到在清空Entry之后,又调用了expungeStaleEntry方法
private int expungeStaleEntry(int staleSlot) {
Entry[] tab = table;
int len = tab.length;
// expunge entry at staleSlot
tab[staleSlot].value = null;
tab[staleSlot] = null;
size--;
// Rehash until we encounter null
Entry e;
int i;
for (i = nextIndex(staleSlot, len);
(e = tab[i]) != null;
i = nextIndex(i, len)) {
ThreadLocal k = e.get();
if (k == null) {
e.value = null;
tab[i] = null;
size--;
} else {
int h = k.threadLocalHashCode & (len - 1);
if (h != i) {
tab[i] = null;
// Unlike Knuth 6.4 Algorithm R, we must scan until
// null because multiple entries could have been stale.
while (tab[h] != null)
h = nextIndex(h, len);
tab[h] = e;
}
}
}
return i;
}
这里对防止内存泄漏做了一些处理,请注意红色的部分,手动将value的值赋为null,让下轮gc可以回收这个value对象。
以上内容为个人分析和测试,真实情况请以实战为准。如果对以上的分析,感觉有不合理的地方请大家指出,共同学习。