jdk源码剖析二: 对象内存布局、synchronized终极原理

很多人一提到锁，自然第一个想到了synchronized，但一直不懂源码实现，现特地追踪到C++层来剥开synchronized的面纱。

网上的很多描述大都不全，让人看了不够爽，看完本章，你将彻底了解synchronized的核心原理。

一、启蒙知识预热

开启本文之前先介绍2个概念

1.1.cas操作

为了提高性能，JVM很多操作都依赖CAS实现，一种乐观锁的实现。本文锁优化中大量用到了CAS，故有必要先分析一下CAS的实现。

CAS：Compare and Swap。

JNI来完成CPU指令的操作：

unsafe.compareAndSwapInt(this, valueOffset, expect, update);

CAS有3个操作数，内存值V，旧的预期值A，要修改的新值B。如果A=V，那么把B赋值给V，返回V；如果A！=V，直接返回V。

打开源码：openjdk\hotspot\src\oscpu\windowsx86\vm\ atomicwindowsx86.inline.hpp，如下图：0

os::is_MP()  这个是runtime/os.hpp，实际就是返回是否多处理器，源码如下：

如上面源代码所示（看第一个int参数即可），LOCK_IF_MP:会根据当前处理器的类型来决定是否为cmpxchg指令添加lock前缀。如果程序是在多处理器上运行，就为cmpxchg指令加上lock前缀（lock cmpxchg）。反之，如果程序是在单处理器上运行，就省略lock前缀（单处理器自身会维护单处理器内的顺序一致性，不需要lock前缀提供的内存屏障效果）。

1.2.对象头

HotSpot虚拟机中，对象在内存中存储的布局可以分为三块区域：对象头（Header）、实例数据（Instance Data）和对齐填充（Padding）。

HotSpot虚拟机的对象头(Object
Header)包括两部分信息:

第一部分"Mark Word":用于存储对象自身的运行时数据，
如哈希码（HashCode）、GC分代年龄、锁状态标志、线程持有的锁、偏向线程ID、偏向时间戳等等.

第二部分"Klass Pointer"：对象指向它的类的元数据的指针，虚拟机通过这个指针来确定这个对象是哪个类的实例。(数组，对象头中还必须有一块用于记录数组长度的数据，因为虚拟机可以通过普通Java对象的元数据信息确定Java对象的大小，但是从数组的元数据中无法确定数组的大小。 )

32位的HotSpot虚拟机对象头存储结构：（下图摘自网络）

图1 32位的HotSpot虚拟机对象头Mark Word组成

为了证实上图的正确性，这里我们看openJDK--》hotspot源码markOop.hpp，虚拟机对象头存储结构：

图2 HotSpot源码markOop.hpp中注释

单词解释：

hash： 保存对象的哈希码
age： 保存对象的分代年龄
biased_lock： 偏向锁标识位
lock： 锁状态标识位
JavaThread*： 保存持有偏向锁的线程ID
epoch： 保存偏向时间戳

上图中有源码中对锁标志位这样枚举：

 enum {   locked_value             = ,//00 轻量级锁
          unlocked_value           = ,//01 无锁
          monitor_value            = ,//10 监视器锁，也叫膨胀锁，也叫重量级锁
          marked_value             = ,//11 GC标记
          biased_lock_pattern      = 5 //101 偏向锁
   };

下面是源码注释：

图3 HotSpot源码markOop.hpp中锁标志位注释

看图3，不管是32/64位JVM，都是1bit偏向锁+2bit锁标志位。上面红框是偏向锁（第一行是指向线程的显示偏向锁，第二行是匿名偏向锁）对应枚举biased_lock_pattern，下面红框是轻量级锁、无锁、监视器锁、GC标记，分别对应上面的前4种枚举。我们甚至能看见锁标志11时，是GC的markSweep(标记清除算法)使用的。（这里就不再拓展了）

对象头中的Mark Word，synchronized源码实现就用了Mark Word来标识对象加锁状态。

二、JVM中synchronized锁实现原理（优化）

大家都知道java中锁synchronized性能较差，线程会阻塞。本节将以图文形式来描述JVM的synchronized锁优化。

在jdk1.6中对锁的实现引入了大量的优化来减少锁操作的开销：

锁粗化（Lock Coarsening）：将多个连续的锁扩展成一个范围更大的锁，用以减少频繁互斥同步导致的性能损耗。

锁消除（Lock Elimination）：JVM及时编译器在运行时，通过逃逸分析，如果判断一段代码中，堆上的所有数据不会逃逸出去从来被其他线程访问到，就可以去除这些锁。

轻量级锁（Lightweight Locking）：JDK1.6引入。在没有多线程竞争的情况下避免重量级互斥锁，只需要依靠一条CAS原子指令就可以完成锁的获取及释放。

偏向锁（Biased Locking）：JDK1.6引入。目的是消除数据再无竞争情况下的同步原语。使用CAS记录获取它的线程。下一次同一个线程进入则偏向该线程，无需任何同步操作。

适应性自旋（Adaptive Spinning）：为了避免线程频繁挂起、恢复的状态切换消耗。产生了忙循环（循环时间固定），即自旋。JDK1.6引入了自适应自旋。自旋时间根据之前锁自旋时间和线程状态，动态变化，用以期望能减少阻塞的时间。

锁升级：偏向锁--》轻量级锁--》重量级锁

2.1.偏向锁

　　按照之前的HotSpot设计，每次加锁/解锁都会涉及到一些CAS操作（比如对等待队列的CAS操作），CAS操作会延迟本地调用，因此偏向锁的想法是一旦线程第一次获得了监视对象，之后让监视对象“偏向”这个线程，之后的多次调用则可以避免CAS操作。

　　简单的讲，就是在锁对象的对象头（开篇讲的对象头数据存储结构）中有个ThreaddId字段，这个字段如果是空的，第一次获取锁的时候，就将自身的ThreadId写入到锁的ThreadId字段内，将锁头内的是否偏向锁的状态位置1.这样下次获取锁的时候，直接检查ThreadId是否和自身线程Id一致，如果一致，则认为当前线程已经获取了锁，因此不需再次获取锁，略过了轻量级锁和重量级锁的加锁阶段。提高了效率。
注意：当锁有竞争关系的时候，需要解除偏向锁，进入轻量级锁。

每一个线程在准备获取共享资源时：

第一步，检查MarkWord里面是不是放的自己的ThreadId ,如果是，表示当前线程是处于 “偏向锁”.跳过轻量级锁直接执行同步体。

获得偏向锁如下图：

2.2.轻量级锁和重量级锁

如上图所示：

第二步，如果MarkWord不是自己的ThreadId,锁升级，这时候，用CAS来执行切换，新的线程根据MarkWord里面现有的ThreadId，通知之前线程暂停，之前线程将Markword的内容置为空。

第三步，两个线程都把对象的HashCode复制到自己新建的用于存储锁的记录空间，接着开始通过CAS操作，把共享对象的MarKword的内容修改为自己新建的记录空间的地址的方式竞争MarkWord.

第四步，第三步中成功执行CAS的获得资源，失败的则进入自旋.

第五步，自旋的线程在自旋过程中，成功获得资源(即之前获的资源的线程执行完成并释放了共享资源)，则整个状态依然处于轻量级锁的状态，如果自旋失败 第六步，进入重量级锁的状态，这个时候，自旋的线程进行阻塞，等待之前线程执行完成并唤醒自己.

注意点：JVM加锁流程

偏向锁--》轻量级锁--》重量级锁

从左往右可以升级，从右往左不能降级

三、从C++源码看synchronized

前两节讲了synchronized锁实现原理，这一节我们从C++源码来剖析synchronized。

3.1 同步和互斥

同步：多个线程并发访问共享资源时，保证同一时刻只有一个（信号量可以多个）线程使用。

实现同步的方法有很多，常见四种如下：

1）临界区（CriticalSection，又叫关键段）：通过对多线程的串行化来访问公共资源或一段代码，速度快，适合控制数据访问。进程内可用。

2）互斥量：互斥量用于线程的互斥。只能为0/1。一个互斥量只能用于一个资源的互斥访问，可跨进程使用。

3）信号量：信号线用于线程的同步。可以为非负整数，可实现多个同类资源的多线程互斥和同步。当信号量为单值信号量是，也可以完成一个资源的互斥访问。可跨进程使用。

4）事件：用来通知线程有一些事件已发生，从而启动后继任务的开始,可跨进程使用。

synchronized的底层实现就用到了临界区和互斥锁（重量级锁的情况下）这两个概念。

3.2 synchronized  C++源码

重点来了，之前在第一节中的图1，看过了对象头Mark Word。现在我们从C++源码来剖析具体的数据结构和获取释放锁的过程。

2.2.1 C++中的监视器锁数据结构

oopDesc--继承-->markOopDesc--方法monitor()-->ObjectMonitor-->enter、exit 获取、释放锁

1.oopDesc类

openjdk\hotspot\src\share\vm\oops\oop.hpp下oopDesc类是JVM对象的顶级基类,故每个object都包含markOop。如下图所示：

 class oopDesc {
   friend class VMStructs;
  private:
   volatile markOop  _mark;//markOop:Mark Word标记字段
   union _metadata {
     Klass*      _klass;//对象类型元数据的指针
     narrowKlass _compressed_klass;
   } _metadata;

   // Fast access to barrier set.  Must be initialized.
   static BarrierSet* _bs;

  public:
   markOop  mark() const         { return _mark; }
   markOop* mark_addr() const    { return (markOop*) &_mark; }

   void set_mark(volatile markOop m)      { _mark = m;   }

   void    release_set_mark(markOop m);
   markOop cas_set_mark(markOop new_mark, markOop old_mark);

   // Used only to re-initialize the mark word (e.g., of promoted
   // objects during a GC) -- requires a valid klass pointer
   void init_mark();

   Klass* klass() const;
   Klass* klass_or_null() const volatile;
   Klass** klass_addr();
   narrowKlass* compressed_klass_addr();
....省略...
}

2.markOopDesc类

openjdk\hotspot\src\share\vm\oops\markOop.hpp下markOopDesc继承自oopDesc，并拓展了自己的方法monitor(),如下图

 ObjectMonitor* monitor() const {
     assert(has_monitor(), "check");
     // Use xor instead of &~ to provide one extra tag-bit check.
     return (ObjectMonitor*) (value() ^ monitor_value);
   }

该方法返回一个ObjectMonitor*对象指针。

其中value()这样定义：

uintptr_t value() const { return (uintptr_t) this; }

可知：将this转换成一个指针宽度的整数（uintptr_t），然后进行"异或"位操作。

monitor_value是常量

 enum {   locked_value             = ,//00轻量级锁
          unlocked_value           = ,//01无锁
          monitor_value            = 2,//10监视器锁，又叫重量级锁
          marked_value             = ,//11GC标记
          biased_lock_pattern      = 5 //101偏向锁
   };

指针低2位00，异或10，结果还是10.（拿一个模板为00的数，异或一个二位数=数本身。因为异或：“相同为0，不同为1”.操作）

3.ObjectMonitor类

在HotSpot虚拟机中，最终采用ObjectMonitor类实现monitor。

openjdk\hotspot\src\share\vm\runtime\objectMonitor.hpp源码如下：

 1 ObjectMonitor() {
 2     _header       = NULL;//markOop对象头
 3     _count        = 0;
 4     _waiters      = 0,//等待线程数
 5     _recursions   = 0;//重入次数
 6     _object       = NULL;//监视器锁寄生的对象。锁不是平白出现的，而是寄托存储于对象中。
 7     _owner        = NULL;//指向获得ObjectMonitor对象的线程或基础锁
 8     _WaitSet      = NULL;//处于wait状态的线程，会被加入到wait set；
 9     _WaitSetLock  = 0 ;
10     _Responsible  = NULL ;
11     _succ         = NULL ;
12     _cxq          = NULL ;
13     FreeNext      = NULL ;
14     _EntryList    = NULL ;//处于等待锁block状态的线程，会被加入到entry set；
15     _SpinFreq     = 0 ;
16     _SpinClock    = 0 ;
17     OwnerIsThread = 0 ;// _owner is (Thread *) vs SP/BasicLock
18     _previous_owner_tid = 0;// 监视器前一个拥有者线程的ID
19   }

每个线程都有两个ObjectMonitor对象列表，分别为free和used列表，如果当前free列表为空，线程将向全局global list请求分配ObjectMonitor。

ObjectMonitor对象中有两个队列：_WaitSet 和 _EntryList，用来保存ObjectWaiter对象列表；

2.获取锁流程

synchronized关键字修饰的代码段，在JVM被编译为monitorenter、monitorexit指令来获取和释放互斥锁.。

解释器执行monitorenter时会进入到InterpreterRuntime.cpp的InterpreterRuntime::monitorenter函数，具体实现如下：

 IRT_ENTRY_NO_ASYNC(void, InterpreterRuntime::monitorenter(JavaThread* thread, BasicObjectLock* elem))
 #ifdef ASSERT
   thread->last_frame().interpreter_frame_verify_monitor(elem);
 #endif
   if (PrintBiasedLockingStatistics) {
     Atomic::inc(BiasedLocking::slow_path_entry_count_addr());
   }
   Handle h_obj(thread, elem->obj());
   assert(Universe::heap()->is_in_reserved_or_null(h_obj()),
          "must be NULL or an object");
   if (UseBiasedLocking) {//标识虚拟机是否开启偏向锁功能,默认开启
     // Retry fast entry if bias is revoked to avoid unnecessary inflation
     ObjectSynchronizer::fast_enter(h_obj, elem->lock(), true, CHECK);
   } else {
     ObjectSynchronizer::slow_enter(h_obj, elem->lock(), CHECK);
   }
   assert(Universe::heap()->is_in_reserved_or_null(elem->obj()),
          "must be NULL or an object");
 #ifdef ASSERT
   thread->last_frame().interpreter_frame_verify_monitor(elem);
 #endif
 IRT_END

先看一下入参：

1、JavaThread thread指向java中的当前线程；
2、BasicObjectLock基础对象锁：包含一个BasicLock和一个指向Object对象的指针oop。

openjdk\hotspot\src\share\vm\runtime\basicLock.hpp中BasicObjectLock类源码如下：

 class BasicObjectLock VALUE_OBJ_CLASS_SPEC {
   friend class VMStructs;
  private:
   BasicLock _lock;                                    // the lock, must be double word aligned
   oop       _obj;                                     // object holds the lock;

  public:
   // Manipulation
   oop      obj() const                                { return _obj;  }
   void set_obj(oop obj)                               { _obj = obj; }
   BasicLock* lock()                                   { return &_lock; }

   // Note: Use frame::interpreter_frame_monitor_size() for the size of BasicObjectLocks
   //       in interpreter activation frames since it includes machine-specific padding.
   static int size()                                   { return sizeof(BasicObjectLock)/wordSize; }

   // GC support
   void oops_do(OopClosure* f) { f->do_oop(&_obj); }

   static int obj_offset_in_bytes()                    { return offset_of(BasicObjectLock, _obj);  }
   static int lock_offset_in_bytes()                   { return offset_of(BasicObjectLock, _lock); }
 };

3、BasicLock类型_lock对象主要用来保存：指向Object对象的对象头数据；

basicLock.hpp中BasicLock源码如下：

 class BasicLock VALUE_OBJ_CLASS_SPEC {
   friend class VMStructs;
  private:
   volatile markOop _displaced_header;//markOop是不是很熟悉？1.2节中讲解对象头时就是分析的markOop源码
  public:
   markOop      displaced_header() const               { return _displaced_header; }
   void         set_displaced_header(markOop header)   { _displaced_header = header; }

   void print_on(outputStream* st) const;

   // move a basic lock (used during deoptimization
   void move_to(oop obj, BasicLock* dest);

   static int displaced_header_offset_in_bytes()       { return offset_of(BasicLock, _displaced_header); }
 };

偏向锁的获取ObjectSynchronizer::fast_enter

在HotSpot中，偏向锁的入口位于openjdk\hotspot\src\share\vm\runtime\synchronizer.cpp文件的ObjectSynchronizer::fast_enter函数：

 void ObjectSynchronizer::fast_enter(Handle obj, BasicLock* lock, bool attempt_rebias, TRAPS) {
  if (UseBiasedLocking) {
     if (!SafepointSynchronize::is_at_safepoint()) {
       BiasedLocking::Condition cond = BiasedLocking::revoke_and_rebias(obj, attempt_rebias, THREAD);
       if (cond == BiasedLocking::BIAS_REVOKED_AND_REBIASED) {
         return;
       }
     } else {
       assert(!attempt_rebias, "can not rebias toward VM thread");
       BiasedLocking::revoke_at_safepoint(obj);
     }
     assert(!obj->mark()->has_bias_pattern(), "biases should be revoked by now");
  }
  //轻量级锁
  slow_enter (obj, lock, THREAD) ;
 }

偏向锁的获取由BiasedLocking::revoke_and_rebias方法实现，由于实现比较长，就不贴代码了，实现逻辑如下：
1、通过markOop mark = obj->mark()获取对象的markOop数据mark，即对象头的Mark Word；
2、判断mark是否为可偏向状态，即mark的偏向锁标志位为 1，锁标志位为 01；
3、判断mark中JavaThread的状态：如果为空，则进入步骤（4）；如果指向当前线程，则执行同步代码块；如果指向其它线程，进入步骤（5）；
4、通过CAS原子指令设置mark中JavaThread为当前线程ID，如果执行CAS成功，则执行同步代码块，否则进入步骤（5）；
5、如果执行CAS失败，表示当前存在多个线程竞争锁，当达到全局安全点（safepoint），获得偏向锁的线程被挂起，撤销偏向锁，并升级为轻量级，升级完成后被阻塞在安全点的线程继续执行同步代码块；

偏向锁的撤销

只有当其它线程尝试竞争偏向锁时，持有偏向锁的线程才会释放锁，偏向锁的撤销由BiasedLocking::revoke_at_safepoint方法实现：

 void BiasedLocking::revoke_at_safepoint(Handle h_obj) {
   assert(SafepointSynchronize::is_at_safepoint(), "must only be called while at safepoint");//校验全局安全点
   oop obj = h_obj();
   HeuristicsResult heuristics = update_heuristics(obj, false);
   if (heuristics == HR_SINGLE_REVOKE) {
     revoke_bias(obj, false, false, NULL);
   } else if ((heuristics == HR_BULK_REBIAS) ||
              (heuristics == HR_BULK_REVOKE)) {
     bulk_revoke_or_rebias_at_safepoint(obj, (heuristics == HR_BULK_REBIAS), false, NULL);
   }
   clean_up_cached_monitor_info();
 }

1、偏向锁的撤销动作必须等待全局安全点；
2、暂停拥有偏向锁的线程，判断锁对象是否处于被锁定状态；
3、撤销偏向锁，恢复到无锁（标志位为 01）或轻量级锁（标志位为 00）的状态；

偏向锁在Java 1.6之后是默认启用的，但在应用程序启动几秒钟之后才激活，可以使用-XX:BiasedLockingStartupDelay=0参数关闭延迟，如果确定应用程序中所有锁通常情况下处于竞争状态，可以通过XX:-UseBiasedLocking=false参数关闭偏向锁。

轻量级锁的获取

当关闭偏向锁功能，或多个线程竞争偏向锁导致偏向锁升级为轻量级锁，会尝试获取轻量级锁，其入口位于ObjectSynchronizer::slow_enter

 void ObjectSynchronizer::slow_enter(Handle obj, BasicLock* lock, TRAPS) {
   markOop mark = obj->mark();
   assert(!mark->has_bias_pattern(), "should not see bias pattern here");

   if (mark->is_neutral()) {//是否为无锁状态001
     // Anticipate successful CAS -- the ST of the displaced mark must
     // be visible <= the ST performed by the CAS.
     lock->set_displaced_header(mark);
     if (mark == (markOop) Atomic::cmpxchg_ptr(lock, obj()->mark_addr(), mark)) {//CAS成功，释放栈锁
       TEVENT (slow_enter: release stacklock) ;
       return ;
     }
     // Fall through to inflate() ...
   } else
   if (mark->has_locker() && THREAD->is_lock_owned((address)mark->locker())) {
     assert(lock != mark->locker(), "must not re-lock the same lock");
     assert(lock != (BasicLock*)obj->mark(), "don't relock with same BasicLock");
     lock->set_displaced_header(NULL);
     return;
   }

 #if 0
   // The following optimization isn't particularly useful.
   if (mark->has_monitor() && mark->monitor()->is_entered(THREAD)) {
     lock->set_displaced_header (NULL) ;
     return ;
   }
 #endif

   // The object header will never be displaced to this lock,
   // so it does not matter what the value is, except that it
   // must be non-zero to avoid looking like a re-entrant lock,
   // and must not look locked either.
   lock->set_displaced_header(markOopDesc::unused_mark());
   ObjectSynchronizer::inflate(THREAD, obj())->enter(THREAD);
 }

1、markOop mark = obj->mark()方法获取对象的markOop数据mark；
2、mark->is_neutral()方法判断mark是否为无锁状态：mark的偏向锁标志位为 0，锁标志位为 01；
3、如果mark处于无锁状态，则进入步骤（4），否则执行步骤（6）；
4、把mark保存到BasicLock对象的_displaced_header字段；
5、通过CAS尝试将Mark Word更新为指向BasicLock对象的指针，如果更新成功，表示竞争到锁，则执行同步代码，否则执行步骤（6）；
6、如果当前mark处于加锁状态，且mark中的ptr指针指向当前线程的栈帧，则执行同步代码，否则说明有多个线程竞争轻量级锁，轻量级锁需要膨胀升级为重量级锁；

假设线程A和B同时执行到临界区if (mark->is_neutral())：
1、线程AB都把Mark Word复制到各自的_displaced_header字段，该数据保存在线程的栈帧上，是线程私有的；
2、Atomic::cmpxchg_ptr原子操作保证只有一个线程可以把指向栈帧的指针复制到Mark Word，假设此时线程A执行成功，并返回继续执行同步代码块；
3、线程B执行失败，退出临界区，通过ObjectSynchronizer::inflate方法开始膨胀锁；

轻量级锁的释放

轻量级锁的释放通过ObjectSynchronizer::slow_exit--->调用ObjectSynchronizer::fast_exit完成。

 void ObjectSynchronizer::fast_exit(oop object, BasicLock* lock, TRAPS) {
   assert(!object->mark()->has_bias_pattern(), "should not see bias pattern here");
   // if displaced header is null, the previous enter is recursive enter, no-op
   markOop dhw = lock->displaced_header();
   markOop mark ;
   if (dhw == NULL) {
      // Recursive stack-lock.
      // Diagnostics -- Could be: stack-locked, inflating, inflated.
      mark = object->mark() ;
      assert (!mark->is_neutral(), "invariant") ;
      if (mark->has_locker() && mark != markOopDesc::INFLATING()) {
         assert(THREAD->is_lock_owned((address)mark->locker()), "invariant") ;
      }
      if (mark->has_monitor()) {
         ObjectMonitor * m = mark->monitor() ;
         assert(((oop)(m->object()))->mark() == mark, "invariant") ;
         assert(m->is_entered(THREAD), "invariant") ;
      }
      return ;
   }

   mark = object->mark() ;

   // If the object is stack-locked by the current thread, try to
   // swing the displaced header from the box back to the mark.
   if (mark == (markOop) lock) {
      assert (dhw->is_neutral(), "invariant") ;
      if ((markOop) Atomic::cmpxchg_ptr (dhw, object->mark_addr(), mark) == mark) {//成功的释放了锁
         TEVENT (fast_exit: release stacklock) ;
         return;
      }
   }

   ObjectSynchronizer::inflate(THREAD, object)->exit (true, THREAD) ;//锁膨胀升级
 }

1、确保处于偏向锁状态时不会执行这段逻辑；
2、取出在获取轻量级锁时保存在BasicLock对象的mark数据dhw；
3、通过CAS尝试把dhw替换到当前的Mark Word，如果CAS成功，说明成功的释放了锁，否则执行步骤（4）；
4、如果CAS失败，说明有其它线程在尝试获取该锁，这时需要将该锁升级为重量级锁，并释放；

重量级锁

重量级锁通过对象内部的监视器（monitor）实现，其中monitor的本质是依赖于底层操作系统的Mutex Lock实现，操作系统实现线程之间的切换需要从用户态到内核态的切换，切换成本非常高。

锁膨胀过程

锁的膨胀过程通过ObjectSynchronizer::inflate函数实现

 ObjectMonitor * ATTR ObjectSynchronizer::inflate (Thread * Self, oop object) {
   // Inflate mutates the heap ...
   // Relaxing assertion for bug 6320749.
   assert (Universe::verify_in_progress() ||
           !SafepointSynchronize::is_at_safepoint(), "invariant") ;

   for (;;) {//自旋
       const markOop mark = object->mark() ;
       assert (!mark->has_bias_pattern(), "invariant") ;

       // The mark can be in one of the following states:
       // *  Inflated     - just return
       // *  Stack-locked - coerce it to inflated
       // *  INFLATING    - busy wait for conversion to complete
       // *  Neutral      - aggressively inflate the object.
       // *  BIASED       - Illegal.  We should never see this

       // CASE: inflated已膨胀，即重量级锁
       if (mark->has_monitor()) {//判断当前是否为重量级锁
           ObjectMonitor * inf = mark->monitor() ;//获取指向ObjectMonitor的指针
           assert (inf->header()->is_neutral(), "invariant");
           assert (inf->object() == object, "invariant") ;
           assert (ObjectSynchronizer::verify_objmon_isinpool(inf), "monitor is invalid");
           return inf ;
       }

       // CASE: inflation in progress - inflating over a stack-lock.膨胀等待（其他线程正在从轻量级锁转为膨胀锁）
       // Some other thread is converting from stack-locked to inflated.
       // Only that thread can complete inflation -- other threads must wait.
       // The INFLATING value is transient.
       // Currently, we spin/yield/park and poll the markword, waiting for inflation to finish.
       // We could always eliminate polling by parking the thread on some auxiliary list.
       if (mark == markOopDesc::INFLATING()) {
          TEVENT (Inflate: spin while INFLATING) ;
          ReadStableMark(object) ;
          continue ;
       }

       // CASE: stack-locked栈锁（轻量级锁）
       // Could be stack-locked either by this thread or by some other thread.
       //
       // Note that we allocate the objectmonitor speculatively, _before_ attempting
       // to install INFLATING into the mark word.  We originally installed INFLATING,
       // allocated the objectmonitor, and then finally STed the address of the
       // objectmonitor into the mark.  This was correct, but artificially lengthened
       // the interval in which INFLATED appeared in the mark, thus increasing
       // the odds of inflation contention.
       //
       // We now use per-thread private objectmonitor free lists.
       // These list are reprovisioned from the global free list outside the
       // critical INFLATING...ST interval.  A thread can transfer
       // multiple objectmonitors en-mass from the global free list to its local free list.
       // This reduces coherency traffic and lock contention on the global free list.
       // Using such local free lists, it doesn't matter if the omAlloc() call appears
       // before or after the CAS(INFLATING) operation.
       // See the comments in omAlloc().

       if (mark->has_locker()) {
           ObjectMonitor * m = omAlloc (Self) ;//获取一个可用的ObjectMonitor
           // Optimistically prepare the objectmonitor - anticipate successful CAS
           // We do this before the CAS in order to minimize the length of time
           // in which INFLATING appears in the mark.
           m->Recycle();
           m->_Responsible  = NULL ;
           m->OwnerIsThread =  ;
           m->_recursions   =  ;
           m->_SpinDuration = ObjectMonitor::Knob_SpinLimit ;   // Consider: maintain by type/class

           markOop cmp = (markOop) Atomic::cmpxchg_ptr (markOopDesc::INFLATING(), object->mark_addr(), mark) ;
           if (cmp != mark) {//CAS失败//CAS失败，说明冲突了，自旋等待//CAS失败，说明冲突了，自旋等待//CAS失败，说明冲突了，自旋等待
              omRelease (Self, m, true) ;//释放监视器锁
              continue ;       // Interference -- just retry
           }

           // We've successfully installed INFLATING (0) into the mark-word.
           // This is the only case where 0 will appear in a mark-work.
           // Only the singular thread that successfully swings the mark-word
           // to 0 can perform (or more precisely, complete) inflation.
           //
           // Why do we CAS a 0 into the mark-word instead of just CASing the
           // mark-word from the stack-locked value directly to the new inflated state?
           // Consider what happens when a thread unlocks a stack-locked object.
           // It attempts to use CAS to swing the displaced header value from the
           // on-stack basiclock back into the object header.  Recall also that the
           // header value (hashcode, etc) can reside in (a) the object header, or
           // (b) a displaced header associated with the stack-lock, or (c) a displaced
           // header in an objectMonitor.  The inflate() routine must copy the header
           // value from the basiclock on the owner's stack to the objectMonitor, all
           // the while preserving the hashCode stability invariants.  If the owner
           // decides to release the lock while the value is 0, the unlock will fail
           // and control will eventually pass from slow_exit() to inflate.  The owner
           // will then spin, waiting for the 0 value to disappear.   Put another way,
           // the 0 causes the owner to stall if the owner happens to try to
           // drop the lock (restoring the header from the basiclock to the object)
           // while inflation is in-progress.  This protocol avoids races that might
           // would otherwise permit hashCode values to change or "flicker" for an object.
           // Critically, while object->mark is 0 mark->displaced_mark_helper() is stable.
           // 0 serves as a "BUSY" inflate-in-progress indicator.

           // fetch the displaced mark from the owner's stack.
           // The owner can't die or unwind past the lock while our INFLATING
           // object is in the mark.  Furthermore the owner can't complete
           // an unlock on the object, either.
           markOop dmw = mark->displaced_mark_helper() ;
           assert (dmw->is_neutral(), "invariant") ;
           //CAS成功，设置ObjectMonitor的_header、_owner和_object等
           // Setup monitor fields to proper values -- prepare the monitor
           m->set_header(dmw) ;

           // Optimization: if the mark->locker stack address is associated
           // with this thread we could simply set m->_owner = Self and
           // m->OwnerIsThread = 1. Note that a thread can inflate an object
           // that it has stack-locked -- as might happen in wait() -- directly
           // with CAS.  That is, we can avoid the xchg-NULL .... ST idiom.
           m->set_owner(mark->locker());
           m->set_object(object);
           // TODO-FIXME: assert BasicLock->dhw != 0.

           // Must preserve store ordering. The monitor state must
           // be stable at the time of publishing the monitor address.
           guarantee (object->mark() == markOopDesc::INFLATING(), "invariant") ;
           object->release_set_mark(markOopDesc::encode(m));

           // Hopefully the performance counters are allocated on distinct cache lines
           // to avoid false sharing on MP systems ...
           if (ObjectMonitor::_sync_Inflations != NULL) ObjectMonitor::_sync_Inflations->inc() ;
           TEVENT(Inflate: overwrite stacklock) ;
           if (TraceMonitorInflation) {
             if (object->is_instance()) {
               ResourceMark rm;
               tty->print_cr("Inflating object " INTPTR_FORMAT " , mark " INTPTR_FORMAT " , type %s",
                 (void *) object, (intptr_t) object->mark(),
                 object->klass()->external_name());
             }
           }
           return m ;
       }

       // CASE: neutral 无锁
       // TODO-FIXME: for entry we currently inflate and then try to CAS _owner.
       // If we know we're inflating for entry it's better to inflate by swinging a
       // pre-locked objectMonitor pointer into the object header.   A successful
       // CAS inflates the object *and* confers ownership to the inflating thread.
       // In the current implementation we use a 2-step mechanism where we CAS()
       // to inflate and then CAS() again to try to swing _owner from NULL to Self.
       // An inflateTry() method that we could call from fast_enter() and slow_enter()
       // would be useful.

       assert (mark->is_neutral(), "invariant");
       ObjectMonitor * m = omAlloc (Self) ;
       // prepare m for installation - set monitor to initial state
       m->Recycle();
       m->set_header(mark);
       m->set_owner(NULL);
       m->set_object(object);
       m->OwnerIsThread =  ;
       m->_recursions   =  ;
       m->_Responsible  = NULL ;
       m->_SpinDuration = ObjectMonitor::Knob_SpinLimit ;       // consider: keep metastats by type/class

       if (Atomic::cmpxchg_ptr (markOopDesc::encode(m), object->mark_addr(), mark) != mark) {
           m->set_object (NULL) ;
           m->set_owner  (NULL) ;
           m->OwnerIsThread =  ;
           m->Recycle() ;
           omRelease (Self, m, true) ;
           m = NULL ;
           continue ;
           // interference - the markword changed - just retry.
           // The state-transitions are one-way, so there's no chance of
           // live-lock -- "Inflated" is an absorbing state.
       }

       // Hopefully the performance counters are allocated on distinct
       // cache lines to avoid false sharing on MP systems ...
       if (ObjectMonitor::_sync_Inflations != NULL) ObjectMonitor::_sync_Inflations->inc() ;
       TEVENT(Inflate: overwrite neutral) ;
       if (TraceMonitorInflation) {
         if (object->is_instance()) {
           ResourceMark rm;
           tty->print_cr("Inflating object " INTPTR_FORMAT " , mark " INTPTR_FORMAT " , type %s",
             (void *) object, (intptr_t) object->mark(),
             object->klass()->external_name());
         }
       }
       return m ;
   }
 }

膨胀过程的实现比较复杂，大概实现过程如下：
1、整个膨胀过程在自旋下完成；
2、mark->has_monitor()方法判断当前是否为重量级锁（上图18-25行），即Mark Word的锁标识位为 10，如果当前状态为重量级锁，执行步骤（3），否则执行步骤（4）；
3、mark->monitor()方法获取指向ObjectMonitor的指针，并返回，说明膨胀过程已经完成；
4、如果当前锁处于膨胀中（上图33-37行），说明该锁正在被其它线程执行膨胀操作，则当前线程就进行自旋等待锁膨胀完成，这里需要注意一点，虽然是自旋操作，但不会一直占用cpu资源，每隔一段时间会通过os::NakedYield方法放弃cpu资源，或通过park方法挂起；如果其他线程完成锁的膨胀操作，则退出自旋并返回；
5、如果当前是轻量级锁状态（上图58-138行），即锁标识位为 00，膨胀过程如下：

1. 通过omAlloc方法，获取一个可用的ObjectMonitor monitor，并重置monitor数据；
2. 通过CAS尝试将Mark Word设置为markOopDesc:INFLATING，标识当前锁正在膨胀中，如果CAS失败，说明同一时刻其它线程已经将Mark Word设置为markOopDesc:INFLATING，当前线程进行自旋等待膨胀完成；
3. 如果CAS成功，设置monitor的各个字段：_header、_owner和_object等，并返回；

6、如果是无锁（中立，上图150-186行），重置监视器值；

monitor竞争

当锁膨胀完成并返回对应的monitor时，并不表示该线程竞争到了锁，真正的锁竞争发生在ObjectMonitor::enter方法中。

 void ATTR ObjectMonitor::enter(TRAPS) {
   // The following code is ordered to check the most common cases first
   // and to reduce RTS->RTO cache line upgrades on SPARC and IA32 processors.
   Thread * const Self = THREAD ;
   void * cur ;

   cur = Atomic::cmpxchg_ptr (Self, &_owner, NULL) ;
   if (cur == NULL) {//CAS成功
      // Either ASSERT _recursions == 0 or explicitly set _recursions = 0.
      assert (_recursions ==    , "invariant") ;
      assert (_owner      == Self, "invariant") ;
      // CONSIDER: set or assert OwnerIsThread == 1
      return ;
   }

   if (cur == Self) {//重入锁
      // TODO-FIXME: check for integer overflow!  BUGID 6557169.
      _recursions ++ ;
      return ;
   }

   if (Self->is_lock_owned ((address)cur)) {
     assert (_recursions == , "internal state error");
     _recursions =  ;
     // Commute owner from a thread-specific on-stack BasicLockObject address to
     // a full-fledged "Thread *".
     _owner = Self ;
     OwnerIsThread =  ;
     return ;
   }

   // We've encountered genuine contention.
   assert (Self->_Stalled == , "invariant") ;
   Self->_Stalled = intptr_t(this) ;

   // Try one round of spinning *before* enqueueing Self
   // and before going through the awkward and expensive state
   // transitions.  The following spin is strictly optional ...
   // Note that if we acquire the monitor from an initial spin
   // we forgo posting JVMTI events and firing DTRACE probes.
   if (Knob_SpinEarly && TrySpin (Self) > ) {
      assert (_owner == Self      , "invariant") ;
      assert (_recursions ==     , "invariant") ;
      assert (((oop)(object()))->mark() == markOopDesc::encode(this), "invariant") ;
      Self->_Stalled =  ;
      return ;
   }

   assert (_owner != Self          , "invariant") ;
   assert (_succ  != Self          , "invariant") ;
   assert (Self->is_Java_thread()  , "invariant") ;
   JavaThread * jt = (JavaThread *) Self ;
   assert (!SafepointSynchronize::is_at_safepoint(), "invariant") ;
   assert (jt->thread_state() != _thread_blocked   , "invariant") ;
   assert (this->object() != NULL  , "invariant") ;
   assert (_count >= , "invariant") ;

   // Prevent deflation at STW-time.  See deflate_idle_monitors() and is_busy().
   // Ensure the object-monitor relationship remains stable while there's contention.
   Atomic::inc_ptr(&_count);

   EventJavaMonitorEnter event;

   { // Change java thread status to indicate blocked on monitor enter.
     JavaThreadBlockedOnMonitorEnterState jtbmes(jt, this);

     DTRACE_MONITOR_PROBE(contended__enter, this, object(), jt);
     if (JvmtiExport::should_post_monitor_contended_enter()) {
       JvmtiExport::post_monitor_contended_enter(jt, this);
     }

     OSThreadContendState osts(Self->osthread());
     ThreadBlockInVM tbivm(jt);

     Self->set_current_pending_monitor(this);

     // TODO-FIXME: change the following for(;;) loop to straight-line code.
     for (;;) {
       jt->set_suspend_equivalent();
       // cleared by handle_special_suspend_equivalent_condition()
       // or java_suspend_self()

       EnterI (THREAD) ;

...省略... }

1、通过CAS尝试把monitor的_owner字段设置为当前线程；
2、如果设置之前的_owner指向当前线程，说明当前线程再次进入monitor，即重入锁，执行_recursions ++ ，记录重入的次数；
3、如果之前的_owner指向的地址在当前线程中，这种描述有点拗口，换一种说法：之前_owner指向的BasicLock在当前线程栈上，说明当前线程是第一次进入该monitor，设置_recursions为1，_owner为当前线程，该线程成功获得锁并返回；
4、如果获取锁失败，则等待锁的释放；

monitor等待

monitor竞争失败的线程，通过自旋执行ObjectMonitor::EnterI方法等待锁的释放，EnterI方法的部分逻辑实现如下：

 ObjectWaiter node(Self) ;
     Self->_ParkEvent->reset() ;
     node._prev   = (ObjectWaiter *) 0xBAD ;
     node.TState  = ObjectWaiter::TS_CXQ ;

     // Push "Self" onto the front of the _cxq.
     // Once on cxq/EntryList, Self stays on-queue until it acquires the lock.
     // Note that spinning tends to reduce the rate at which threads
     // enqueue and dequeue on EntryList|cxq.
     ObjectWaiter * nxt ;
     for (;;) {
         node._next = nxt = _cxq ;
         if (Atomic::cmpxchg_ptr (&node, &_cxq, nxt) == nxt) break ;

         // Interference - the CAS failed because _cxq changed.  Just retry.
         // As an optional optimization we retry the lock.
         if (TryLock (Self) > ) {
             assert (_succ != Self         , "invariant") ;
             assert (_owner == Self        , "invariant") ;
             assert (_Responsible != Self  , "invariant") ;
             return ;
         }
     }

1、当前线程被封装成ObjectWaiter对象node，状态设置成ObjectWaiter::TS_CXQ；
2、在for循环中，通过CAS把node节点push到_cxq列表中，同一时刻可能有多个线程把自己的node节点push到_cxq列表中；
3、node节点push到_cxq列表之后，通过自旋尝试获取锁，如果还是没有获取到锁，则通过park将当前线程挂起，等待被唤醒，实现如下：

 for (;;) {

         if (TryLock (Self) > ) break ;
         assert (_owner != Self, "invariant") ;

         if ((SyncFlags & ) && _Responsible == NULL) {
            Atomic::cmpxchg_ptr (Self, &_Responsible, NULL) ;
         }

         // park self
         if (_Responsible == Self || (SyncFlags & )) {
             TEVENT (Inflated enter - park TIMED) ;
             Self->_ParkEvent->park ((jlong) RecheckInterval) ;
             // Increase the RecheckInterval, but clamp the value.
             RecheckInterval *=  ;
             if (RecheckInterval > ) RecheckInterval =  ;
         } else {
             TEVENT (Inflated enter - park UNTIMED) ;
             Self->_ParkEvent->park() ;//当前线程挂起
         }

         if (TryLock(Self) > 0) break ;

         // The lock is still contested.
         // Keep a tally of the # of futile wakeups.
         // Note that the counter is not protected by a lock or updated by atomics.
         // That is by design - we trade "lossy" counters which are exposed to
         // races during updates for a lower probe effect.
         TEVENT (Inflated enter - Futile wakeup) ;
         if (ObjectMonitor::_sync_FutileWakeups != NULL) {
            ObjectMonitor::_sync_FutileWakeups->inc() ;
         }
         ++ nWakeups ;

         // Assuming this is not a spurious wakeup we'll normally find _succ == Self.
         // We can defer clearing _succ until after the spin completes
         // TrySpin() must tolerate being called with _succ == Self.
         // Try yet another round of adaptive spinning.
         if ((Knob_SpinAfterFutile & ) && TrySpin (Self) > ) break ;

         // We can find that we were unpark()ed and redesignated _succ while
         // we were spinning.  That's harmless.  If we iterate and call park(),
         // park() will consume the event and return immediately and we'll
         // just spin again.  This pattern can repeat, leaving _succ to simply
         // spin on a CPU.  Enable Knob_ResetEvent to clear pending unparks().
         // Alternately, we can sample fired() here, and if set, forgo spinning
         // in the next iteration.

         if ((Knob_ResetEvent & ) && Self->_ParkEvent->fired()) {
            Self->_ParkEvent->reset() ;
            OrderAccess::fence() ;
         }
         if (_succ == Self) _succ = NULL ;

         // Invariant: after clearing _succ a thread *must* retry _owner before parking.
         OrderAccess::fence() ;
     }

4、当该线程被唤醒时，会从挂起的点继续执行，通过ObjectMonitor::TryLock尝试获取锁，TryLock方法实现如下：

 int ObjectMonitor::TryLock (Thread * Self) {
    for (;;) {
       void * own = _owner ;
       if (own != NULL) return  ;
       if (Atomic::cmpxchg_ptr (Self, &_owner, NULL) == NULL) {//CAS成功，获取锁
          // Either guarantee _recursions == 0 or set _recursions = 0.
          assert (_recursions == , "invariant") ;
          assert (_owner == Self, "invariant") ;
          // CONSIDER: set or assert that OwnerIsThread == 1
          return  ;
       }
       // The lock had been free momentarily, but we lost the race to the lock.
       // Interference -- the CAS failed.
       // We can either return -1 or retry.
       // Retry doesn't make as much sense because the lock was just acquired.
       if (true) return - ;
    }
 }

其本质就是通过CAS设置monitor的_owner字段为当前线程，如果CAS成功，则表示该线程获取了锁，跳出自旋操作，执行同步代码，否则继续被挂起；

monitor释放

当某个持有锁的线程执行完同步代码块时，会进行锁的释放，给其它线程机会执行同步代码，在HotSpot中，通过退出monitor的方式实现锁的释放，并通知被阻塞的线程，具体实现位于ObjectMonitor::exit方法中。

 void ATTR ObjectMonitor::exit(bool not_suspended, TRAPS) {
    Thread * Self = THREAD ;
    if (THREAD != _owner) {
      if (THREAD->is_lock_owned((address) _owner)) {
        // Transmute _owner from a BasicLock pointer to a Thread address.
        // We don't need to hold _mutex for this transition.
        // Non-null to Non-null is safe as long as all readers can
        // tolerate either flavor.
        assert (_recursions == , "invariant") ;
        _owner = THREAD ;
        _recursions =  ;
        OwnerIsThread =  ;
      } else {
        // NOTE: we need to handle unbalanced monitor enter/exit
        // in native code by throwing an exception.
        // TODO: Throw an IllegalMonitorStateException ?
        TEVENT (Exit - Throw IMSX) ;
        assert(false, "Non-balanced monitor enter/exit!");
        if (false) {
           THROW(vmSymbols::java_lang_IllegalMonitorStateException());
        }
        return;
      }
    }

    if (_recursions != ) {
      _recursions--;        // this is simple recursive enter
      TEVENT (Inflated exit - recursive) ;
      return ;
    }
...省略...

1、如果是重量级锁的释放，monitor中的_owner指向当前线程，即THREAD == _owner；
2、根据不同的策略（由QMode指定），从cxq或EntryList中获取头节点，通过ObjectMonitor::ExitEpilog方法唤醒该节点封装的线程，唤醒操作最终由unpark完成，实现如下：

 void ObjectMonitor::ExitEpilog (Thread * Self, ObjectWaiter * Wakee) {
    assert (_owner == Self, "invariant") ;

    // Exit protocol:
    // 1. ST _succ = wakee
    // 2. membar #loadstore|#storestore;
    // 2. ST _owner = NULL
    // 3. unpark(wakee)

    _succ = Knob_SuccEnabled ? Wakee->_thread : NULL ;
    ParkEvent * Trigger = Wakee->_event ;

    // Hygiene -- once we've set _owner = NULL we can't safely dereference Wakee again.
    // The thread associated with Wakee may have grabbed the lock and "Wakee" may be
    // out-of-scope (non-extant).
    Wakee  = NULL ;

    // Drop the lock
    OrderAccess::release_store_ptr (&_owner, NULL) ;
    OrderAccess::fence() ;                               // ST _owner vs LD in unpark()

    if (SafepointSynchronize::do_call_back()) {
       TEVENT (unpark before SAFEPOINT) ;
    }

    DTRACE_MONITOR_PROBE(contended__exit, this, object(), Self);
    Trigger->unpark() ;

    // Maintain stats and report events to JVMTI
    if (ObjectMonitor::_sync_Parks != NULL) {
       ObjectMonitor::_sync_Parks->inc() ;
    }
 }

3、被唤醒的线程，继续执行monitor的竞争；

四.总结

本文重点介绍了Synchronized原理以及JVM对Synchronized的优化。简单来说解决三种场景：

1）只有一个线程进入临界区，偏向锁

2）多个线程交替进入临界区，轻量级锁

3）多线程同时进入临界区，重量级锁

========================================================

参考：

《深入理解 Java 虚拟机》

JVM源码分析之synchronized实现