ReentrantLock的实现网上有很多文章了,本篇文章会简单介绍下其java层实现,重点放在分析竞争锁失败后如何阻塞线程。
ReentrantLock的实现网上有很多文章了,本篇文章会简单介绍下其java层实现,重点放在分析竞争锁失败后如何阻塞线程。 因篇幅有限,synchronized的内容将会放到下篇文章。
ReentrantLock是jdk中常用的锁实现,其实现逻辑主语基于AQS(juc包中的大多数同步类实现都是基于AQS);接下来会简单介绍AQS的大致原理,关于其实现细节以及各种应用,之后会写一篇文章具体分析。
AQS是类AbstractQueuedSynchronizer.java的简称,JUC包下的ReentrantLock、CyclicBarrier、CountdownLatch都使用到了AQS。
其大致原理如下:
其中tryAcquire方法是抽象方法,具体实现取决于实现类,我们常说的公平锁和非公平锁的区别就在于该方法的实现。
ReentrantLock分为公平锁和非公平锁,我们只看公平锁。 ReentrantLock.lock会调用到ReentrantLock#FairSync.lock中:
FairSync.java
static final class FairSync extends Sync { final void lock() { acquire(1); } /** * Fair version of tryAcquire. Don't grant access unless * recursive call or no waiters or is first. */ protected final boolean tryAcquire(int acquires) { final Thread current = Thread.currentThread(); int c = getState(); if (c == 0) { if (!hasQueuedPredecessors() && compareAndSetState(0, acquires)) { setExclusiveOwnerThread(current); return true; } } else if (current == getExclusiveOwnerThread()) { int nextc = c + acquires; if (nextc < 0) throw new Error("Maximum lock count exceeded"); setState(nextc); return true; } return false; } }
AbstractQueuedSynchronizer.java
public final void acquire(int arg) { if (!tryAcquire(arg) && acquireQueued(addWaiter(Node.EXCLUSIVE), arg)) selfInterrupt(); }
可以看到FairSync.lock调用了AQS的acquire方法,而在acquire中首先调用tryAcquire尝试获得锁,以下两种情况返回true:
如果tryAcquire失败则调用acquireQueued阻塞当前线程。acquireQueued最终会调用到LockSupport.park()阻塞线程。
个人认为,要深入理解锁机制,一个很重要的点是理解系统是如何阻塞线程的。
LockSupport.java
public static void park(Object blocker) { Thread t = Thread.currentThread(); setBlocker(t, blocker); UNSAFE.park(false, 0L); setBlocker(t, null); }
park方法的参数blocker是用于负责这次阻塞的同步对象,在AQS的调用中,这个对象就是AQS本身。我们知道synchronized关键字是需要指定一个对象的(如果作用于方法上则是当前对象或当前类),与之类似blocker就是LockSupport指定的对象。
park方法调用了native方法UNSAFE.park,第一个参数代表第二个参数是否是绝对时间,第二个参数代表最长阻塞时间。
其实现如下,只保留核心代码,完整代码看查看unsafe.cpp
Unsafe_Park(JNIEnv *env, jobject unsafe, jboolean isAbsolute, jlong time){ ... thread->parker()->park(isAbsolute != 0, time); ... }
park方法在os_linux.cpp中(其他操作系统的实现在os_xxx中)
void Parker::park(bool isAbsolute, jlong time) { ... //获得当前线程 Thread* thread = Thread::current(); assert(thread->is_Java_thread(), "Must be JavaThread"); JavaThread *jt = (JavaThread *)thread; //如果当前线程被设置了interrupted标记,则直接返回 if (Thread::is_interrupted(thread, false)) { return; } if (time > 0) { //unpacktime中根据isAbsolute的值来填充absTime结构体,isAbsolute为true时,time代表绝对时间且单位是毫秒,否则time是相对时间且单位是纳秒 //absTime.tvsec代表了对于时间的秒 //absTime.tv_nsec代表对应时间的纳秒 unpackTime(&absTime, isAbsolute, time); } //调用mutex trylock方法 if (Thread::is_interrupted(thread, false) || pthread_mutex_trylock(_mutex) != 0) { return; } //_counter是一个许可的数量,跟ReentrantLock里定义的许可变量基本都是一个原理。 unpack方法调用时会将_counter赋值为1。 //_counter>0代表已经有人调用了unpark,所以不用阻塞 int status ; if (_counter > 0) { // no wait needed _counter = 0; //释放mutex锁 status = pthread_mutex_unlock(_mutex); return; } //设置线程状态为CONDVAR_WAIT OSThreadWaitState osts(thread->osthread(), false /* not Object.wait() */); ... //等待 _cur_index = isAbsolute ? ABS_INDEX : REL_INDEX; pthread_cond_timedwait(&_cond[_cur_index], _mutex, &absTime); ... //释放mutex锁 status = pthread_mutex_unlock(_mutex) ; }
park方法用POSIX的pthread_cond_timedwait方法阻塞线程,调用pthread_cond_timedwait前需要先获得锁,因此park主要流程为:
另外,在阻塞当前线程前,会调用OSThreadWaitState的构造方法将线程状态设置为CONDVAR_WAIT,在Jvm中Thread状态枚举如下
enum ThreadState { ALLOCATED, // Memory has been allocated but not initialized INITIALIZED, // The thread has been initialized but yet started RUNNABLE, // Has been started and is runnable, but not necessarily running MONITOR_WAIT, // Waiting on a contended monitor lock CONDVAR_WAIT, // Waiting on a condition variable OBJECT_WAIT, // Waiting on an Object.wait() call BREAKPOINTED, // Suspended at breakpoint SLEEPING, // Thread.sleep() ZOMBIE // All done, but not reclaimed yet };
由上文我们可以知道LockSupport.park方法最终是由POSIX的 pthread_cond_timedwait的方法实现的。 我们现在就进一步看看pthread_mutex_trylock,pthread_cond_timedwait,pthread_mutex_unlock这几个方法是如何实现的。
Linux系统中相关代码在glibc库中。
先看trylock的实现, 代码在glibc的pthread_mutex_trylock.c文件中,该方法代码很多,我们只看主要代码
//pthread_mutex_t是posix中的互斥锁结构体 int __pthread_mutex_trylock (mutex) pthread_mutex_t *mutex; { int oldval; pid_t id = THREAD_GETMEM (THREAD_SELF, tid); switch (__builtin_expect (PTHREAD_MUTEX_TYPE (mutex), PTHREAD_MUTEX_TIMED_NP)) { case PTHREAD_MUTEX_ERRORCHECK_NP: case PTHREAD_MUTEX_TIMED_NP: case PTHREAD_MUTEX_ADAPTIVE_NP: /* Normal mutex. */ if (lll_trylock (mutex->__data.__lock) != 0) break; /* Record the ownership. */ mutex->__data.__owner = id; ++mutex->__data.__nusers; return 0; } } //以下代码在lowlevellock.h中 #define __lll_trylock(futex) / (atomic_compare_and_exchange_val_acq (futex, 1, 0) != 0) #define lll_trylock(futex) __lll_trylock (&(futex))
mutex默认用的是PTHREAD_MUTEX_NORMAL类型(与PTHREAD_MUTEX_TIMED_NP相同); 因此会先调用lll_trylock方法,lll_trylock实际上是一个cas操作,如果mutex->data.lock==0则将其修改为1并返回0,否则返回1。
如果成功,则更改mutex中的owner为当前线程。
pthread_mutex_unlock.c
int internal_function attribute_hidden __pthread_mutex_unlock_usercnt (mutex, decr) pthread_mutex_t *mutex; int decr; { if (__builtin_expect (type, PTHREAD_MUTEX_TIMED_NP) == PTHREAD_MUTEX_TIMED_NP) { /* Always reset the owner field. */ normal: mutex->__data.__owner = 0; if (decr) /* One less user. */ --mutex->__data.__nusers; /* Unlock. */ lll_unlock (mutex->__data.__lock, PTHREAD_MUTEX_PSHARED (mutex)); return 0; } }
pthread_mutex_unlock将mutex中的owner清空,并调用了lll_unlock方法
lowlevellock.h
#define __lll_unlock(futex, private) / ((void) ({ / int *__futex = (futex); / int __val = atomic_exchange_rel (__futex, 0); / / if (__builtin_expect (__val > 1, 0)) / lll_futex_wake (__futex, 1, private); / })) #define lll_unlock(futex, private) __lll_unlock(&(futex), private) #define lll_futex_wake(ftx, nr, private) / ({ / DO_INLINE_SYSCALL(futex, 3, (long) (ftx), / __lll_private_flag (FUTEX_WAKE, private), / (int) (nr)); / _r10 == -1 ? -_retval : _retval; / })
lll_unlock分为两个步骤:
FUTEX_WAKE在上一篇文章有分析,futex机制的核心是当获得锁时,尝试cas更改一个int型变量(用户态操作),如果integer原始值是0,则修改成功,该线程获得锁,否则就将当期线程放入到 wait queue中,wait queue中的线程不会被系统调度(内核态操作)。
futex变量的值有3种:0代表当前锁空闲,1代表有线程持有当前锁,2代表存在锁冲突。futex的值初始化时是0;当调用try_lock的时候会利用cas操作改为1(见上面的trylock函数);当调用lll_lock时,如果不存在锁冲突,则将其改为1,否则改为2。
#define __lll_lock(futex, private) / ((void) ({ / int *__futex = (futex); / if (__builtin_expect (atomic_compare_and_exchange_bool_acq (__futex, / 1, 0), 0)) / { / if (__builtin_constant_p (private) && (private) == LLL_PRIVATE) / __lll_lock_wait_private (__futex); / else / __lll_lock_wait (__futex, private); / } / })) #define lll_lock(futex, private) __lll_lock (&(futex), private) void __lll_lock_wait_private (int *futex) { //第一次进来的时候futex==1,所以不会走这个if if (*futex == 2) lll_futex_wait (futex, 2, LLL_PRIVATE); //在这里会把futex设置成2,并调用futex_wait让当前线程等待 while (atomic_exchange_acq (futex, 2) != 0) lll_futex_wait (futex, 2, LLL_PRIVATE); }
pthread_cond_timedwait用于阻塞线程,实现线程等待, 代码在glibc的pthread_cond_timedwait.c文件中,代码较长,你可以先简单过一遍,看完下面的分析再重新读一遍代码
int int __pthread_cond_timedwait (cond, mutex, abstime) pthread_cond_t *cond; pthread_mutex_t *mutex; const struct timespec *abstime; { struct _pthread_cleanup_buffer buffer; struct _condvar_cleanup_buffer cbuffer; int result = 0; /* Catch invalid parameters. */ if (abstime->tv_nsec < 0 || abstime->tv_nsec >= 1000000000) return EINVAL; int pshared = (cond->__data.__mutex == (void *) ~0l) ? LLL_SHARED : LLL_PRIVATE; //1.获得cond锁 lll_lock (cond->__data.__lock, pshared); //2.释放mutex锁 int err = __pthread_mutex_unlock_usercnt (mutex, 0); if (err) { lll_unlock (cond->__data.__lock, pshared); return err; } /* We have one new user of the condvar. */ //每执行一次wait(pthread_cond_timedwait/pthread_cond_wait),__total_seq就会+1 ++cond->__data.__total_seq; //用来执行futex_wait的变量 ++cond->__data.__futex; //标识该cond还有多少线程在使用,pthread_cond_destroy需要等待所有的操作完成 cond->__data.__nwaiters += 1 << COND_NWAITERS_SHIFT; /* Remember the mutex we are using here. If there is already a different address store this is a bad user bug. Do not store anything for pshared condvars. */ //保存mutex锁 if (cond->__data.__mutex != (void *) ~0l) cond->__data.__mutex = mutex; /* Prepare structure passed to cancellation handler. */ cbuffer.cond = cond; cbuffer.mutex = mutex; /* Before we block we enable cancellation. Therefore we have to install a cancellation handler. */ __pthread_cleanup_push (&buffer, __condvar_cleanup, &cbuffer); /* The current values of the wakeup counter. The "woken" counter must exceed this value. */ //记录futex_wait前的__wakeup_seq(为该cond上执行了多少次sign操作+timeout次数)和__broadcast_seq(代表在该cond上执行了多少次broadcast) unsigned long long int val; unsigned long long int seq; val = seq = cond->__data.__wakeup_seq; /* Remember the broadcast counter. */ cbuffer.bc_seq = cond->__data.__broadcast_seq; while (1) { //3.计算要wait的相对时间 struct timespec rt; { #ifdef __NR_clock_gettime INTERNAL_SYSCALL_DECL (err); int ret; ret = INTERNAL_VSYSCALL (clock_gettime, err, 2, (cond->__data.__nwaiters & ((1 << COND_NWAITERS_SHIFT) - 1)), &rt); # ifndef __ASSUME_POSIX_TIMERS if (__builtin_expect (INTERNAL_SYSCALL_ERROR_P (ret, err), 0)) { struct timeval tv; (void) gettimeofday (&tv, NULL); /* Convert the absolute timeout value to a relative timeout. */ rt.tv_sec = abstime->tv_sec - tv.tv_sec; rt.tv_nsec = abstime->tv_nsec - tv.tv_usec * 1000; } else # endif { /* Convert the absolute timeout value to a relative timeout. */ rt.tv_sec = abstime->tv_sec - rt.tv_sec; rt.tv_nsec = abstime->tv_nsec - rt.tv_nsec; } #else /* Get the current time. So far we support only one clock. */ struct timeval tv; (void) gettimeofday (&tv, NULL); /* Convert the absolute timeout value to a relative timeout. */ rt.tv_sec = abstime->tv_sec - tv.tv_sec; rt.tv_nsec = abstime->tv_nsec - tv.tv_usec * 1000; #endif } if (rt.tv_nsec < 0) { rt.tv_nsec += 1000000000; --rt.tv_sec; } /*---计算要wait的相对时间 end---- */ //是否超时 /* Did we already time out? */ if (__builtin_expect (rt.tv_sec < 0, 0)) { //被broadcast唤醒,这里疑问的是,为什么不需要判断__wakeup_seq? if (cbuffer.bc_seq != cond->__data.__broadcast_seq) goto bc_out; goto timeout; } unsigned int futex_val = cond->__data.__futex; //4.释放cond锁,准备wait lll_unlock (cond->__data.__lock, pshared); /* Enable asynchronous cancellation. Required by the standard. */ cbuffer.oldtype = __pthread_enable_asynccancel (); //5.调用futex_wait /* Wait until woken by signal or broadcast. */ err = lll_futex_timed_wait (&cond->__data.__futex, futex_val, &rt, pshared); /* Disable asynchronous cancellation. */ __pthread_disable_asynccancel (cbuffer.oldtype); //6.重新获得cond锁,因为又要访问&修改cond的数据了 lll_lock (cond->__data.__lock, pshared); //__broadcast_seq值发生改变,代表发生了有线程调用了广播 if (cbuffer.bc_seq != cond->__data.__broadcast_seq) goto bc_out; //判断是否是被sign唤醒的,sign会增加__wakeup_seq //第二个条件cond->__data.__woken_seq != val的意义在于 //可能两个线程A、B在wait,一个线程调用了sign导致A被唤醒,这时B因为超时被唤醒 //对于B线程来说,执行到这里时第一个条件也是满足的,从而导致上层拿到的result不是超时 //所以这里需要判断下__woken_seq(即该cond已经被唤醒的线程数)是否等于__wakeup_seq(sign执行次数+timeout次数) val = cond->__data.__wakeup_seq; if (val != seq && cond->__data.__woken_seq != val) break; /* Not woken yet. Maybe the time expired? */ if (__builtin_expect (err == -ETIMEDOUT, 0)) { timeout: /* Yep. Adjust the counters. */ ++cond->__data.__wakeup_seq; ++cond->__data.__futex; /* The error value. */ result = ETIMEDOUT; break; } } //一个线程已经醒了所以这里__woken_seq +1 ++cond->__data.__woken_seq; bc_out: // cond->__data.__nwaiters -= 1 << COND_NWAITERS_SHIFT; /* If pthread_cond_destroy was called on this variable already, notify the pthread_cond_destroy caller all waiters have left and it can be successfully destroyed. */ if (cond->__data.__total_seq == -1ULL && cond->__data.__nwaiters < (1 << COND_NWAITERS_SHIFT)) lll_futex_wake (&cond->__data.__nwaiters, 1, pshared); //9.cond数据修改完毕,释放锁 lll_unlock (cond->__data.__lock, pshared); /* The cancellation handling is back to normal, remove the handler. */ __pthread_cleanup_pop (&buffer, 0); //10.重新获得mutex锁 err = __pthread_mutex_cond_lock (mutex); return err ?: result; }
上面的代码虽然加了注释,但相信大多数人第一次看都看不懂。 我们来简单梳理下,上面代码有两把锁,一把是mutex锁,一把cond锁。另外,在调用pthread_cond_timedwait前后必须调用pthread_mutex_lock(&mutex);和pthread_mutex_unlock(&mutex);加/解mutex锁。
因此pthread_cond_timedwait的使用大致分为几个流程:
看到这里,你可能有几点疑问:为什么需要两把锁?mutex锁和cond锁的作用是什么?
说mutex锁的作用之前,我们回顾一下java的Object.wait的使用。Object.wait必须是在synchronized同步块中使用。试想下如果不加synchronized也能运行Object.wait的话会存在什么问题?
Object condObj=new Object(); voilate int flag = 0; public void waitTest(){ if(flag == 0){ condObj.wait(); } } public void notifyTest(){ flag=1; condObj.notify(); }
如上代码,A线程调用waitTest,这时flag==0,所以准备调用wait方法进行休眠,这时B线程开始执行,调用notifyTest将flag置为1,并调用notify方法,注意:此时A线程还没调用wait,所以notfiy没有唤醒任何线程。然后A线程继续执行,调用wait方法进行休眠,而之后不会有人来唤醒A线程,A线程将永久wait下去!
Object condObj=new Object(); voilate int flag = 0; public void waitTest(){ synchronized(condObj){ if(flag == 0){ condObj.wait(); } } } public void notifyTest(){ synchronized(condObj){ flag=1; condObj.notify(); } }
在有锁保护下的情况下, 当调用condObj.wait时,flag一定是等于0的,不会存在一直wait的问题。
回到pthread_cond_timedwait,其需要加mutex锁的原因就呼之欲出了: 保证wait和其wait条件的原子性
不管是glibc的pthread_cond_timedwait/pthread_cond_signal还是java层的Object.wait/Object.notify,Jdk AQS的Condition.await/Condition.signal,所有的Condition机制都需要在加锁环境下才能使用,其根本原因就是要保证进行线程休眠时,条件变量是没有被篡改的。
注意下mutex锁释放的时机,回顾上文中pthread_cond_timedwait的流程,在第2步时就释放了mutex锁,之后调用futex_wait进行休眠,为什么要在休眠前就释放mutex锁呢?原因也很简单:如果不释放mutex锁就开始休眠,那其他线程就永远无法调用signal方法将休眠线程唤醒(因为调用signal方法前需要获得mutex锁)。
在线程被唤醒之后还要在第10步中重新获得mutex锁是为了保证锁的语义(思考下如果不重新获得mutex锁会发生什么)。
cond锁的作用其实很简单: 保证对象cond->data的线程安全。 在pthread_cond_timedwait时需要修改cond->data的数据,如增加total_seq(在这个cond上一共执行过多少次wait)增加nwaiters(现在还有多少个线程在wait这个cond),所有在修改及访问cond->data时需要加cond锁。
这里我没想明白的一点是,用mutex锁也能保证cond->data修改的线程安全,只要晚一点释放mutex锁就行了。为什么要先释放mutex,重新获得cond来保证线程安全? 是为了避免mutex锁住的范围太大吗?
该问题的答案可以见评论区@11800222 的回答:
mutex锁不能保护cond->data修改的线程安全,调用signal的线程没有用mutex锁保护修改cond的那段临界区。
pthread_cond_wait/signal这一对本身用cond锁同步就能睡眠唤醒。 wait的时候需要传入mutex是因为睡眠前需要释放mutex锁,但睡眠之前又不能有无锁的空隙,解决办法是让mutex锁在cond锁上之后再释放。 而signal前不需要释放mutex锁,在持有mutex的情况下signal,之后再释放mutex锁。
唤醒休眠线程的代码比较简单,主要就是调用lll_futex_wake。
int __pthread_cond_signal (cond) pthread_cond_t *cond; { int pshared = (cond->__data.__mutex == (void *) ~0l) ? LLL_SHARED : LLL_PRIVATE; //因为要操作cond的数据,所以要加锁 lll_lock (cond->__data.__lock, pshared); /* Are there any waiters to be woken? */ if (cond->__data.__total_seq > cond->__data.__wakeup_seq) { //__wakeup_seq为执行sign与timeout次数的和 ++cond->__data.__wakeup_seq; ++cond->__data.__futex; ... //唤醒wait的线程 lll_futex_wake (&cond->__data.__futex, 1, pshared); } /* We are done. */ lll_unlock (cond->__data.__lock, pshared); return 0; }
本文对Java简单介绍了ReentrantLock实现原理,对LockSupport.park底层实现pthread_cond_timedwait机制做了详细分析。
【编辑推荐】
【责任编辑:华轩 TEL:(010)68476606】