muduo::EventLoopThread、EventLoopThreadPool分析

muduo的并发模型为one loop per thread+ threadpool。为了方便使用，muduo封装了EventLoop和Thread为EventLoopThread，为了方便使用线程池，又把EventLoopThread封装为EventLoopThreadPool。

EventLoopThread

任何一个线程，只要创建并运行了EventLoop，就是一个IO线程。
EventLoopThread类就是一个封装了的IO线程。
EventLoopThread的工作流程为：
1、在主线程（暂且这么称呼）创建EventLoopThread对象。
2、主线程调用EventLoopThread.start()，启动EventLoopThread中的线程（称为IO线程），这是主线程要等待IO线程创建完成EventLoop对象。
3、IO线程调用threadFunc创建EventLoop对象。通知主线程已经创建完成。
4、主线程返回创建的EventLoop对象。

EventLoopThread.h

class EventLoopThread : boost::noncopyable
{
 public:
  typedef boost::function<void(EventLoop*)> ThreadInitCallback;

  EventLoopThread(const ThreadInitCallback& cb = ThreadInitCallback(),
                  const string& name = string());
  ~EventLoopThread();
  EventLoop* startLoop();//启动本线程，返回本线程中的EventLoop

 private:
  void threadFunc();

  EventLoop* loop_;//本线程持有的EventLoop对象指针
  bool exiting_;//是否已经退出
  Thread thread_;//本线程
  MutexLock mutex_;
  Condition cond_;
  ThreadInitCallback callback_;//回调函数
};

EventLoopThread.cc

EventLoopThread::EventLoopThread(const ThreadInitCallback& cb,
                                 const string& name)
  : loop_(NULL),
    exiting_(false),
    thread_(boost::bind(&EventLoopThread::threadFunc, this), name),//创建线程，在回调函数创建EventLoop
    mutex_(),
    cond_(mutex_),
    callback_(cb)
{
}

EventLoopThread::~EventLoopThread()
{
  exiting_ = true;
  if (loop_ != NULL) // not 100% race-free, eg. threadFunc could be running callback_.
  {
    // still a tiny chance to call destructed object, if threadFunc exits just now.
    // but when EventLoopThread destructs, usually programming is exiting anyway.
    loop_->quit();//退出loop循环
    thread_.join();//等待线程退出
  }
}

EventLoop* EventLoopThread::startLoop()//另一个线程在调用这个函数
{
  assert(!thread_.started());
  thread_.start();//当前线程启动，调用threadFunc()

  {
    MutexLockGuard lock(mutex_);
    while (loop_ == NULL)
    {
      cond_.wait();//等待创建好当前IO线程
    }
  }

  return loop_;
}

void EventLoopThread::threadFunc()
{
  EventLoop loop;//创建EventLoop对象。注意，在栈上

  if (callback_)
  {
    callback_(&loop);
  }

  {
    MutexLockGuard lock(mutex_);
    loop_ = &loop;
    cond_.notify();//通知startLoop
  }

  loop.loop();//会在这里循环，直到EventLoopThread析构。此后不再使用loop_访问EventLoop了
  //assert(exiting_);
  loop_ = NULL;
}

EventLoopThreadPool

muduo的思想时eventLoop+thread pool，为了更方便使用，将EventLoopThread做了封装。main reactor可以创建sub reactor，并发一些任务分发到sub reactor中去。

EventLoopThreadPool的思想比较简单，用一个main reactor创建EventLoopThreadPool。在EventLoopThreadPool中将EventLoop和Thread绑定，可以返回EventLoop对象来使用EventLoopThreadPool中的Thread。

EventLoopThreadPool.h

class EventLoop;

class EventLoopThread : boost::noncopyable
{
 public:
  typedef boost::function<void(EventLoop*)> ThreadInitCallback;

  EventLoopThread(const ThreadInitCallback& cb = ThreadInitCallback(),
                  const string& name = string());
  ~EventLoopThread();
  EventLoop* startLoop();//启动本线程，返回本线程中的EventLoop

 private:
  void threadFunc();

  EventLoop* loop_;//本线程持有的EventLoop对象指针
  bool exiting_;//是否已经退出
  Thread thread_;//本线程
  MutexLock mutex_;
  Condition cond_;
  ThreadInitCallback callback_;//回调函数
};

EventLoopThreadPool.cc

EventLoopThreadPool::EventLoopThreadPool(EventLoop* baseLoop, const string& nameArg)
  : baseLoop_(baseLoop),
    name_(nameArg),
    started_(false),
    numThreads_(0),
    next_(0)
{
}

EventLoopThreadPool::~EventLoopThreadPool()
{
  // Don't delete loop, it's stack variable
}

void EventLoopThreadPool::start(const ThreadInitCallback& cb)
{
  assert(!started_);
  baseLoop_->assertInLoopThread();

  started_ = true;

  for (int i = 0; i < numThreads_; ++i)
  {
    char buf[name_.size() + 32];
    snprintf(buf, sizeof buf, "%s%d", name_.c_str(), i);
    EventLoopThread* t = new EventLoopThread(cb, buf);
    threads_.push_back(t);
    loops_.push_back(t->startLoop());
  }
  if (numThreads_ == 0 && cb)
  {
    cb(baseLoop_);
  }
}

EventLoop* EventLoopThreadPool::getNextLoop()
{
  baseLoop_->assertInLoopThread();
  assert(started_);
  EventLoop* loop = baseLoop_;//loops_为空，则返回baseloop

  if (!loops_.empty())//循环分配
  {
    // round-robin
    loop = loops_[next_];
    ++next_;
    if (implicit_cast<size_t>(next_) >= loops_.size())
    {
      next_ = 0;
    }
  }
  return loop;
}

EventLoop* EventLoopThreadPool::getLoopForHash(size_t hashCode)
{
  baseLoop_->assertInLoopThread();
  EventLoop* loop = baseLoop_;

  if (!loops_.empty())
  {
    loop = loops_[hashCode % loops_.size()];//根据hashCode分配
  }
  return loop;
}

std::vector<EventLoop*> EventLoopThreadPool::getAllLoops()
{
  baseLoop_->assertInLoopThread();
  assert(started_);
  if (loops_.empty())
  {
    return std::vector<EventLoop*>(1, baseLoop_);
  }
  else
  {
    return loops_;
  }
}

可以写个简单的测试程序，创建一个EventLoopThreadPool，打印其中线程的ID和name。

#include <muduo/net/EventLoop.h>
#include <muduo/net/EventLoopThread.h>
#include <muduo/net/EventLoopThreadPool.h>

#include <stdio.h>

using namespace muduo;
using namespace muduo::net;

void runInThread()
{
    printf("runInThread(): name = %s, tid = %d\n",
           CurrentThread::name(), CurrentThread::tid());
}

int main()
{
    printf("main(): pid = %d, tid = %d\n",
           getpid(), CurrentThread::tid());

    runInThread();
    EventLoop loop;
    EventLoopThreadPool loopThreadPool(&loop, "sub Reactor");
    loopThreadPool.setThreadNum(5);
    loopThreadPool.start();
    for(int i=0; i<10; ++i)
    {
        EventLoop* loopFromPool=loopThreadPool.getNextLoop();
        loopFromPool->runInLoop(runInThread);
    }
    sleep(3);
    printf("exit main().\n");
}