Android消息机制不完全解析（下）

首先，看看在/frameworks/base/core/jni/android_os_MessageQueue.cpp文件中看看android.os.MessageQueue类中的四个原生函数的实现：

static void android_os_MessageQueue_nativeInit(JNIEnv* env, jobject obj) {

    NativeMessageQueue* nativeMessageQueue = new NativeMessageQueue();//构造NativeMessageQueue实例

    if (!nativeMessageQueue) {

        jniThrowRuntimeException(env, "Unable to allocate native queue");

        return;

    }

    nativeMessageQueue->incStrong(env);//强引用+1

    android_os_MessageQueue_setNativeMessageQueue(env, obj, nativeMessageQueue);

}

static void android_os_MessageQueue_nativeDestroy(JNIEnv* env, jobject obj) {

    NativeMessageQueue* nativeMessageQueue =

            android_os_MessageQueue_getNativeMessageQueue(env, obj);

    if (nativeMessageQueue) {

        android_os_MessageQueue_setNativeMessageQueue(env, obj, NULL);

        nativeMessageQueue->decStrong(env);//强引用-1，实际上会导致释放NativeMessageQueue实例

    }

}

static void android_os_MessageQueue_nativePollOnce(JNIEnv* env, jobject obj,

        jint ptr, jint timeoutMillis) {

    NativeMessageQueue* nativeMessageQueue = reinterpret_cast<NativeMessageQueue*>(ptr);//指针强制转换

    nativeMessageQueue->pollOnce(env, timeoutMillis);//调用nativeMessageQueue的pollonce函数

}

static void android_os_MessageQueue_nativeWake(JNIEnv* env, jobject obj, jint ptr) {

    NativeMessageQueue* nativeMessageQueue = reinterpret_cast<NativeMessageQueue*>(ptr);

    return nativeMessageQueue->wake();//调用nativeMessageQueue的wake函数

}

从代码中，可以看到这四个函数的实现都是依赖于NativeMessageQueue类。不过，在开始解析NativeMessageQueue之前，我们再看一些有意思的代码：

static void android_os_MessageQueue_setNativeMessageQueue(JNIEnv* env, jobject messageQueueObj,

        NativeMessageQueue* nativeMessageQueue) {

    env->SetIntField(messageQueueObj, gMessageQueueClassInfo.mPtr,

             reinterpret_cast<jint>(nativeMessageQueue));//把nativeMessageQueue的实例地址强转为java的int类型并保存到gMessageQueueClassInfo.mPtr中

}

那么gMessageQueueClassInfo.mPtr是什么呢？

static JNINativeMethod gMessageQueueMethods[] = {

    /* name, signature, funcPtr */

    { "nativeInit", "()V", (void*)android_os_MessageQueue_nativeInit },

    { "nativeDestroy", "()V", (void*)android_os_MessageQueue_nativeDestroy },

    { "nativePollOnce", "(II)V", (void*)android_os_MessageQueue_nativePollOnce },

    { "nativeWake", "(I)V", (void*)android_os_MessageQueue_nativeWake }

};

#define FIND_CLASS(var, className) \

        var = env->FindClass(className); \

        LOG_FATAL_IF(! var, "Unable to find class " className);

#define GET_FIELD_ID(var, clazz, fieldName, fieldDescriptor) \

        var = env->GetFieldID(clazz, fieldName, fieldDescriptor); \

        LOG_FATAL_IF(! var, "Unable to find field " fieldName);

//这个函数在Android启动的时候，会被系统调用

int register_android_os_MessageQueue(JNIEnv* env) {

    int res = jniRegisterNativeMethods(env, "android/os/MessageQueue",

            gMessageQueueMethods, NELEM(gMessageQueueMethods));//关联MessageQueueQueue的原生函数

    LOG_FATAL_IF(res < 0, "Unable to register native methods.");

    jclass clazz;

    FIND_CLASS(clazz, "android/os/MessageQueue");//获取MessageQueue的class

    GET_FIELD_ID(gMessageQueueClassInfo.mPtr, clazz,

            "mPtr", "I");//获取MessageQueue class的mPtr field的Id

    return 0;

}

上面的代码很像java的反射有木有？

    Class cls = Class.forName("android.os.MessageQueue");

    Field feild = cls.getField("mPtr");

到这里，我们就明白了android_os_MessageQueue_setNativeMessageQueue函数实际上把android.os.MessageQueue实例的mPtr值设置为nativeMessageQueue实例的地址。虽然Java语言没有指针的说法，但是，这里的mPtr却的的确确是作为一个指针使用的。现在，我们也就理解了，为什么mPtr可以被强制转换为nativeMessageQueue了。

小结：

android_os_MessageQueue_nativeInit和android_os_MessageQueue_nativeDestory两个函数做了些什么：

android_os_MessageQueue_nativeInit：构造NativeMessageQueue实例
android_os_MessageQueue_nativeDestory:销毁NativeMessageQeue实例

NativeMessageQueue

先来看看NativeMessageQueue的声明：

class NativeMessageQueue : public MessageQueue {

public:

    NativeMessageQueue();

    virtual ~NativeMessageQueue();

    virtual void raiseException(JNIEnv* env, const char* msg, jthrowable exceptionObj);

    void pollOnce(JNIEnv* env, int timeoutMillis);

    void wake();

private:

    bool mInCallback;

    jthrowable mExceptionObj;

};

NativeMessageQueue继承自MessageQueue（不是java中的android.os.MessageQueue哦），关于MessageQueue我们只需要了解，它包含了一个成员mLooper即可（有兴趣的同学可以查看/frameworks/base/core/jni/android_os_MessageQueue.h）

class MessageQueue  {

    ......

protected:

    sp<Looper> mLooper;

};

继续看NativeMessageQueue的代码：

NativeMessageQueue::NativeMessageQueue() : mInCallback(false), mExceptionObj(NULL) {

    mLooper = Looper::getForThread();

    if (mLooper == NULL) {

        mLooper = new Looper(false);//实例化mLooper

        Looper::setForThread(mLooper);

    }

}

NativeMessageQueue构造实例的时候，会实例化mLooper。

void NativeMessageQueue::pollOnce(JNIEnv* env, int timeoutMillis) {

    mInCallback = true;

    mLooper->pollOnce(timeoutMillis);

    mInCallback = false;

    if (mExceptionObj) {

        env->Throw(mExceptionObj);

        env->DeleteLocalRef(mExceptionObj);

        mExceptionObj = NULL;

    }

}

void NativeMessageQueue::wake() {

    mLooper->wake();

}

小结：

NativeMessageQueue的函数pollonce和wake实现相当简单，交给mLooper的同名函数。

Looper

先来看看Looper的声明/frameworks/native/include/utils/Looper.h：

class Looper : public ALooper, public RefBase {

protected:

    virtual ~Looper();

public:

    Looper(bool allowNonCallbacks);

    bool getAllowNonCallbacks() const;

    int pollOnce(int timeoutMillis, int* outFd, int* outEvents, void** outData);

    inline int pollOnce(int timeoutMillis) {

        return pollOnce(timeoutMillis, NULL, NULL, NULL);

    }

    void wake();

private:

    const bool mAllowNonCallbacks; // immutable

    int mWakeReadPipeFd;  // immutable

    int mWakeWritePipeFd; // immutable

    Mutex mLock;

    int mEpollFd; // immutable

    int pollInner(int timeoutMillis);

    void awoken();

};

因为代码有点多，所以上面的声明，已经被我精简了大部分，现在我们只关注我们关心的：pollonce和wake函数。

还是从构造函数开始（frameworks/native/utils/Looper.cpp）：

Looper::Looper(bool allowNonCallbacks) :

        mAllowNonCallbacks(allowNonCallbacks), mSendingMessage(false),

        mResponseIndex(0), mNextMessageUptime(LLONG_MAX) {

    int wakeFds[2];

    int result = pipe(wakeFds);//创建命名管道

    LOG_ALWAYS_FATAL_IF(result != 0, "Could not create wake pipe.  errno=%d", errno);

    // 保存命名管道

    mWakeReadPipeFd = wakeFds[0];

    mWakeWritePipeFd = wakeFds[1];

    result = fcntl(mWakeReadPipeFd, F_SETFL, O_NONBLOCK);//设置为非阻塞模式

    LOG_ALWAYS_FATAL_IF(result != 0, "Could not make wake read pipe non-blocking.  errno=%d",

            errno);

    result = fcntl(mWakeWritePipeFd, F_SETFL, O_NONBLOCK);//设置为非阻塞模式

    LOG_ALWAYS_FATAL_IF(result != 0, "Could not make wake write pipe non-blocking.  errno=%d",

            errno);

    // 开始使用epoll API，实现轮询

    // Allocate the epoll instance and register the wake pipe.

    mEpollFd = epoll_create(EPOLL_SIZE_HINT);//创建epoll文件描述符

    LOG_ALWAYS_FATAL_IF(mEpollFd < 0, "Could not create epoll instance.  errno=%d", errno);

    struct epoll_event eventItem;

    memset(& eventItem, 0, sizeof(epoll_event)); // zero out unused members of data field union

    eventItem.events = EPOLLIN;

    eventItem.data.fd = mWakeReadPipeFd;

    result = epoll_ctl(mEpollFd, EPOLL_CTL_ADD, mWakeReadPipeFd, & eventItem); // 把刚才创建的命名管道的读端加入的到epoll的监听队列中

    LOG_ALWAYS_FATAL_IF(result != 0, "Could not add wake read pipe to epoll instance.  errno=%d",

            errno);

}

Looper::~Looper() {

    close(mWakeReadPipeFd);//释放命名管道

    close(mWakeWritePipeFd);

    close(mEpollFd);//释放epoll文件描述符

}

Looper的构造函数中，出现了命名管道和epoll相关的代码，这是为什么呢？别急，接着看下去就知道了：

int Looper::pollOnce(int timeoutMillis, int* outFd, int* outEvents, void** outData) {

    int result = 0;

    for (;;) {

       // 这段代码暂时无视

       while (mResponseIndex < mResponses.size()) {

            const Response& response = mResponses.itemAt(mResponseIndex++);

            int ident = response.request.ident;

            if (ident >= 0) {//ident > 0, 即此response为noncallback，需要返回event，data等数据给调用者处理

                int fd = response.request.fd;

                int events = response.events;

                void* data = response.request.data;

#if DEBUG_POLL_AND_WAKE

                ALOGD("%p ~ pollOnce - returning signalled identifier %d: "

                        "fd=%d, events=0x%x, data=%p",

                        this, ident, fd, events, data);

#endif

                if (outFd != NULL) *outFd = fd;

                if (outEvents != NULL) *outEvents = events;

                if (outData != NULL) *outData = data;

                return ident;

            }

        }

        if (result != 0) {

#if DEBUG_POLL_AND_WAKE

            ALOGD("%p ~ pollOnce - returning result %d", this, result);

#endif

            if (outFd != NULL) *outFd = 0;

            if (outEvents != NULL) *outEvents = 0;

            if (outData != NULL) *outData = NULL;

            return result;

        }

        result = pollInner(timeoutMillis);//这一行才是重点！

    }

}

接着往下看，代码有些长，但请仔细看：

int Looper::pollInner(int timeoutMillis) {

#if DEBUG_POLL_AND_WAKE

    ALOGD("%p ~ pollOnce - waiting: timeoutMillis=%d", this, timeoutMillis);

#endif

    // 设置timeoutMillis的值为Math.min(timeoutMills, mNextMessageUptime)

    // Adjust the timeout based on when the next message is due.

    if (timeoutMillis != 0 && mNextMessageUptime != LLONG_MAX) {

        nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC);

        int messageTimeoutMillis = toMillisecondTimeoutDelay(now, mNextMessageUptime);

        if (messageTimeoutMillis >= 0

                && (timeoutMillis < 0 || messageTimeoutMillis < timeoutMillis)) {

            timeoutMillis = messageTimeoutMillis;

        }

#if DEBUG_POLL_AND_WAKE

        ALOGD("%p ~ pollOnce - next message in %lldns, adjusted timeout: timeoutMillis=%d",

                this, mNextMessageUptime - now, timeoutMillis);

#endif

    }

    // Poll.

    int result = ALOOPER_POLL_WAKE;

    mResponses.clear();

    mResponseIndex = 0;

    // 开始轮询

    struct epoll_event eventItems[EPOLL_MAX_EVENTS];

    int eventCount = epoll_wait(mEpollFd, eventItems, EPOLL_MAX_EVENTS, timeoutMillis);

    // Acquire lock.

    mLock.lock();

    // Check for poll error.

    if (eventCount < 0) { //处理error

        if (errno == EINTR) {

            goto Done;

        }

        ALOGW("Poll failed with an unexpected error, errno=%d", errno);

        result = ALOOPER_POLL_ERROR;

        goto Done;

    }

    // Check for poll timeout.

    if (eventCount == 0) { // 未能等到event，故timeout

#if DEBUG_POLL_AND_WAKE

        ALOGD("%p ~ pollOnce - timeout", this);

#endif

        result = ALOOPER_POLL_TIMEOUT;

        goto Done;

    }

    // Handle all events.

#if DEBUG_POLL_AND_WAKE

    ALOGD("%p ~ pollOnce - handling events from %d fds", this, eventCount);

#endif

    for (int i = 0; i < eventCount; i++) {//有event，则处理

        int fd = eventItems[i].data.fd;

        uint32_t epollEvents = eventItems[i].events;

        if (fd == mWakeReadPipeFd) {

            if (epollEvents & EPOLLIN) {

                awoken();//读取命名管道内的数据

            } else {

                ALOGW("Ignoring unexpected epoll events 0x%x on wake read pipe.", epollEvents);

            }

        } else {

            ssize_t requestIndex = mRequests.indexOfKey(fd);

            if (requestIndex >= 0) {

                int events = 0;

                if (epollEvents & EPOLLIN) events |= ALOOPER_EVENT_INPUT;

                if (epollEvents & EPOLLOUT) events |= ALOOPER_EVENT_OUTPUT;

                if (epollEvents & EPOLLERR) events |= ALOOPER_EVENT_ERROR;

                if (epollEvents & EPOLLHUP) events |= ALOOPER_EVENT_HANGUP;

                pushResponse(events, mRequests.valueAt(requestIndex));//把event添加到mResponses中，等待后续处理

            } else {

                ALOGW("Ignoring unexpected epoll events 0x%x on fd %d that is "

                        "no longer registered.", epollEvents, fd);

            }

        }

    }

Done: ;

    // 处理C++层的Message

    // Invoke pending message callbacks.

    mNextMessageUptime = LLONG_MAX;

    while (mMessageEnvelopes.size() != 0) {

        nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC);

        const MessageEnvelope& messageEnvelope = mMessageEnvelopes.itemAt(0);

        if (messageEnvelope.uptime <= now) {

            // Remove the envelope from the list.

            // We keep a strong reference to the handler until the call to handleMessage

            // finishes.  Then we drop it so that the handler can be deleted *before*

            // we reacquire our lock.

            { // obtain handler

                sp<MessageHandler> handler = messageEnvelope.handler;

                Message message = messageEnvelope.message;

                mMessageEnvelopes.removeAt(0);

                mSendingMessage = true;

                mLock.unlock();

#if DEBUG_POLL_AND_WAKE || DEBUG_CALLBACKS

                ALOGD("%p ~ pollOnce - sending message: handler=%p, what=%d",

                        this, handler.get(), message.what);

#endif

                handler->handleMessage(message);//调用handler->handleMessage

            } // release handler

            mLock.lock();

            mSendingMessage = false;

            result = ALOOPER_POLL_CALLBACK;

        } else {

            // The last message left at the head of the queue determines the next wakeup time.

            mNextMessageUptime = messageEnvelope.uptime;//更新mNextMessageUptime

            break;

        }

    }

    // Release lock.

    mLock.unlock();

    // 处理mResponses

    // Invoke all response callbacks.

    for (size_t i = 0; i < mResponses.size(); i++) {

        Response& response = mResponses.editItemAt(i);

        if (response.request.ident == ALOOPER_POLL_CALLBACK) {

            int fd = response.request.fd;

            int events = response.events;

            void* data = response.request.data;

#if DEBUG_POLL_AND_WAKE || DEBUG_CALLBACKS

            ALOGD("%p ~ pollOnce - invoking fd event callback %p: fd=%d, events=0x%x, data=%p",

                    this, response.request.callback.get(), fd, events, data);

#endif

            int callbackResult = response.request.callback->handleEvent(fd, events, data);//调用callback->handleEvent

            if (callbackResult == 0) {

                removeFd(fd);

            }

            // Clear the callback reference in the response structure promptly because we

            // will not clear the response vector itself until the next poll.

            response.request.callback.clear();

            result = ALOOPER_POLL_CALLBACK;

        }

    }

    return result;

}

代码比较长，所以分段分析：

    // 设置timeoutMillis的值为Math.min(timeoutMillis, mNextMessageUptime)

    // Adjust the timeout based on when the next message is due.

    if (timeoutMillis != 0 && mNextMessageUptime != LLONG_MAX) {

        nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC);

        int messageTimeoutMillis = toMillisecondTimeoutDelay(now, mNextMessageUptime);

        if (messageTimeoutMillis >= 0

                && (timeoutMillis < 0 || messageTimeoutMillis < timeoutMillis)) {

            timeoutMillis = messageTimeoutMillis;

        }#if DEBUG_POLL_AND_WAKE

        ALOGD("%p ~ pollOnce - next message in %lldns, adjusted timeout: timeoutMillis=%d",

                this, mNextMessageUptime - now, timeoutMillis);

#endif

    }

为了解析这段代码，需要先补充一些C++层的Message相关的代码：

struct Message {

    Message() : what(0) { }

    Message(int what) : what(what) { }

    /* The message type. (interpretation is left up to the handler) */

    int what;

};

class MessageHandler : public virtual RefBase {

    protected:

        virtual ~MessageHandler() { }

    public:

        /**

         * Handles a message.

         */

        virtual void handleMessage(const Message& message) = 0;

}

struct MessageEnvelope {

    MessageEnvelope() : uptime(0) { }

    MessageEnvelope(nsecs_t uptime, const sp<MessageHandler> handler,const Message& message) : uptime(uptime), handler(handler), message(message) {}

    nsecs_t uptime;

    MessageHandler> handler;

    Message message;

};

void Looper::sendMessage(const sp<MessageHandler>& handler, const Message& message) {

    nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC);

    sendMessageAtTime(now, handler, message);

}

void Looper::sendMessageDelayed(nsecs_t uptimeDelay, const sp<MessageHandler>& handler,

        const Message& message) {

    nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC);

    sendMessageAtTime(now + uptimeDelay, handler, message);

}

void Looper::sendMessageAtTime(nsecs_t uptime, const sp<MessageHandler>& handler,

        const Message& message) {

#if DEBUG_CALLBACKS

    ALOGD("%p ~ sendMessageAtTime - uptime=%lld, handler=%p, what=%d",

            this, uptime, handler.get(), message.what);

#endif

    size_t i = 0;

    { // acquire lock

        AutoMutex _l(mLock);

        size_t messageCount = mMessageEnvelopes.size();

        while (i < messageCount && uptime >= mMessageEnvelopes.itemAt(i).uptime) {

            i += 1;

        }

        MessageEnvelope messageEnvelope(uptime, handler, message);

        mMessageEnvelopes.insertAt(messageEnvelope, i, 1);

        // Optimization: If the Looper is currently sending a message, then we can skip

        // the call to wake() because the next thing the Looper will do after processing

        // messages is to decide when the next wakeup time should be.  In fact, it does

        // not even matter whether this code is running on the Looper thread.

        if (mSendingMessage) {

            return;

        }

    } // release lock

    // Wake the poll loop only when we enqueue a new message at the head.

    if (i == 0) {

        wake();

    }

}

void Looper::removeMessages(const sp<MessageHandler>& handler) {

#if DEBUG_CALLBACKS

    ALOGD("%p ~ removeMessages - handler=%p", this, handler.get());

#endif

    { // acquire lock

        AutoMutex _l(mLock);

        for (size_t i = mMessageEnvelopes.size(); i != 0; ) {

            const MessageEnvelope& messageEnvelope = mMessageEnvelopes.itemAt(--i);

            if (messageEnvelope.handler == handler) {

                mMessageEnvelopes.removeAt(i);

            }

        }

    } // release lock

}

void Looper::removeMessages(const sp<MessageHandler>& handler, int what) {

#if DEBUG_CALLBACKS

    ALOGD("%p ~ removeMessages - handler=%p, what=%d", this, handler.get(), what);

#endif

    { // acquire lock

        AutoMutex _l(mLock);

        for (size_t i = mMessageEnvelopes.size(); i != 0; ) {

            const MessageEnvelope& messageEnvelope = mMessageEnvelopes.itemAt(--i);

            if (messageEnvelope.handler == handler

                    && messageEnvelope.message.what == what) {

                mMessageEnvelopes.removeAt(i);

            }

        }

    } // release lock

}

是不是觉得上面的代码似曾相识？和Java层的MessageQueue很像似有木有？和java层一样，C++层存在消息队列和消息处理机制，消息被保存到成员mMessageEvelopes中，并在pollInner函数中处理消息（调用MesageHandler的handleMesage函数处理）。

现在回到Looper：：pollonceh函数，我们就应该能够理解，pollOnce函数到timeOutMillis参数仅仅代表了Java层下一个Message的触发延迟，所以，我们还需要考虑C++层下一个Message的触发延迟，所以，代码设置timeoutMillis为timeoutMillis和mNextMessageUpTime中的较小值。

继续下一段代码：

    int result = ALOOPER_POLL_WAKE;

    mResponses.clear();

    mResponseIndex = 0;

    // 开始轮询

    struct epoll_event eventItems[EPOLL_MAX_EVENTS];

    int eventCount = epoll_wait(mEpollFd, eventItems, EPOLL_MAX_EVENTS, timeoutMillis);

    // Acquire lock.

    mLock.lock();

调用epoll函数，等待event发生，epoll_wait函数的返回值有三种可能：失败出错、没有event、有一个或多个event。

1. 失败的处理：

    // Check for poll error.

    if (eventCount < 0) { //处理error

        if (errno == EINTR) {

            goto Done;

        }

        ALOGW("Poll failed with an unexpected error, errno=%d", errno);

        result = ALOOPER_POLL_ERROR;

        goto Done;

    }

2. 没有event发生：

    // Check for poll timeout.

    if (eventCount == 0) { // 未能等到event，故timeout

#if DEBUG_POLL_AND_WAKE

        ALOGD("%p ~ pollOnce - timeout", this);

#endif

        result = ALOOPER_POLL_TIMEOUT;

        goto Done;

    }

3. 有event发生：

    // Handle all events.

#if DEBUG_POLL_AND_WAKE

    ALOGD("%p ~ pollOnce - handling events from %d fds", this, eventCount);

#endif

    for (int i = 0; i < eventCount; i++) {//有event，则处理

        int fd = eventItems[i].data.fd;

        uint32_t epollEvents = eventItems[i].events;

        if (fd == mWakeReadPipeFd) {//说明java层或者C++层有新的Message

            if (epollEvents & EPOLLIN) {

                awoken();//读取命名管道内的数据

            } else {

                ALOGW("Ignoring unexpected epoll events 0x%x on wake read pipe.", epollEvents);

            }

        } else {

            ssize_t requestIndex = mRequests.indexOfKey(fd);

            if (requestIndex >= 0) {

                int events = 0;

                if (epollEvents & EPOLLIN) events |= ALOOPER_EVENT_INPUT;

                if (epollEvents & EPOLLOUT) events |= ALOOPER_EVENT_OUTPUT;

                if (epollEvents & EPOLLERR) events |= ALOOPER_EVENT_ERROR;

                if (epollEvents & EPOLLHUP) events |= ALOOPER_EVENT_HANGUP;

                pushResponse(events, mRequests.valueAt(requestIndex));//把event添加到mResponses中，等待后续处理

            } else {

                ALOGW("Ignoring unexpected epoll events 0x%x on fd %d that is "

                        "no longer registered.", epollEvents, fd);

            }

        }

    }

处理event的时候，需要分为两类：fd==mWakeReadPipeFd和fd！=mWakeReadPipeFd

fd==mWakeReadPipeFd：说明C++层，或者java层有新的Message出现，需要处理。这种情况下，只需要读mWakeReadPipeFd内的数据即可

void Looper::awoken() {

#if DEBUG_POLL_AND_WAKE

    ALOGD("%p ~ awoken", this);

#endif

    char buffer[16];

    ssize_t nRead;

    do {

        nRead = read(mWakeReadPipeFd, buffer, sizeof(buffer));

    } while ((nRead == -1 && errno == EINTR) || nRead == sizeof(buffer));

}

fd！=mWakeReadPipeFd:首先需要搞明白fd是哪来的：

    struct Request {

        int fd;

        int ident;

        sp<LooperCallback> callback;

        void* data;

    };

    struct Response {

        int events;

        Request request;

    };

/**

 * A looper callback.

 */

class LooperCallback : public virtual RefBase {

protected:

    virtual ~LooperCallback() { }

public:

    /**

     * Handles a poll event for the given file descriptor.

     * It is given the file descriptor it is associated with,

     * a bitmask of the poll events that were triggered (typically ALOOPER_EVENT_INPUT),

     * and the data pointer that was originally supplied.

     *

     * Implementations should return 1 to continue receiving callbacks, or 0

     * to have this file descriptor and callback unregistered from the looper.

     */

    virtual int handleEvent(int fd, int events, void* data) = 0;

};

int Looper::addFd(int fd, int ident, int events, const sp<LooperCallback>& callback, void* data) {

#if DEBUG_CALLBACKS

    ALOGD("%p ~ addFd - fd=%d, ident=%d, events=0x%x, callback=%p, data=%p", this, fd, ident,

            events, callback.get(), data);

#endif

    if (!callback.get()) {

        if (! mAllowNonCallbacks) {

            ALOGE("Invalid attempt to set NULL callback but not allowed for this looper.");

            return -1;

        }

        if (ident < 0) {//仅当Looper支持NonCallbacks，并且ident大于0时，允许添加callback为null的Fd

            ALOGE("Invalid attempt to set NULL callback with ident < 0.");

            return -1;

        }

    } else {

        ident = ALOOPER_POLL_CALLBACK;

    }

    int epollEvents = 0;

    if (events & ALOOPER_EVENT_INPUT) epollEvents |= EPOLLIN;

    if (events & ALOOPER_EVENT_OUTPUT) epollEvents |= EPOLLOUT;

    { // acquire lock

        AutoMutex _l(mLock);

        Request request;

        request.fd = fd;

        request.ident = ident;

        request.callback = callback;

        request.data = data;

        struct epoll_event eventItem;

        memset(& eventItem, 0, sizeof(epoll_event)); // zero out unused members of data field union

        eventItem.events = epollEvents;

        eventItem.data.fd = fd;

        ssize_t requestIndex = mRequests.indexOfKey(fd);

        if (requestIndex < 0) {

            int epollResult = epoll_ctl(mEpollFd, EPOLL_CTL_ADD, fd, & eventItem);

            if (epollResult < 0) {

                ALOGE("Error adding epoll events for fd %d, errno=%d", fd, errno);

                return -1;

            }

            mRequests.add(fd, request);

        } else {//存在则替换

            int epollResult = epoll_ctl(mEpollFd, EPOLL_CTL_MOD, fd, & eventItem);

            if (epollResult < 0) {

                ALOGE("Error modifying epoll events for fd %d, errno=%d", fd, errno);

                return -1;

            }

            mRequests.replaceValueAt(requestIndex, request);

        }

    } // release lock

    return 1;

}

从上面的代码，我们可知，Looper还支持添加Fd和自定义的callback，类似java层的Message.callback。

通过addFd函数，可以向Looper的mEpollFd添加指定的Fd，当Fd触发指定的event .e.i EPOLLIN or EPOLLOUT时，指定的相应的自定义callback就会得到执行。

另外，当Looper支持noncallback时，还可以向Looper添加callback为null的Fd，因为没有callback，所以Fd添加者需要调用int Looper::pollOnce(int timeoutMillis, int* outFd, int* outEvents, void** outData) 通过outFd、outEvent、outData参数获取数据并进行处理。

所以当Fd触发消息时，需要生成对应到reponse并添加到meResponses中，等待后续的处理。

void Looper::pushResponse(int events, const Request& request) {

    Response response;

    response.events = events;

    response.request = request;

    mResponses.push(response);

}

到这里为止，epoll_wait函数返回的三种结果的不同处理已经解析完毕，接下来代码进入共同的Done环节。

处理C++层的Message：

Done: ;

    // 处理C++层的Message

    // Invoke pending message callbacks.

    mNextMessageUptime = LLONG_MAX;

    while (mMessageEnvelopes.size() != 0) {

        nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC);

        const MessageEnvelope& messageEnvelope = mMessageEnvelopes.itemAt(0);

        if (messageEnvelope.uptime <= now) {

            // Remove the envelope from the list.

            // We keep a strong reference to the handler until the call to handleMessage

            // finishes.  Then we drop it so that the handler can be deleted *before*

            // we reacquire our lock.

            { // obtain handler

                sp<MessageHandler> handler = messageEnvelope.handler;

                Message message = messageEnvelope.message;

                mMessageEnvelopes.removeAt(0);

                mSendingMessage = true;

                mLock.unlock();

#if DEBUG_POLL_AND_WAKE || DEBUG_CALLBACKS

                ALOGD("%p ~ pollOnce - sending message: handler=%p, what=%d",

                        this, handler.get(), message.what);

#endif

                handler->handleMessage(message);//调用handler->handleMessage

            } // release handler

            mLock.lock();

            mSendingMessage = false;

            result = ALOOPER_POLL_CALLBACK;

        } else {

            // The last message left at the head of the queue determines the next wakeup time.

            mNextMessageUptime = messageEnvelope.uptime;//更新mNextMessageUptime

            break;

        }

    }

处理Response：

    // 处理mResponses

    // Invoke all response callbacks.

    for (size_t i = 0; i < mResponses.size(); i++) {

        Response& response = mResponses.editItemAt(i);

        if (response.request.ident == ALOOPER_POLL_CALLBACK) {

            int fd = response.request.fd;

            int events = response.events;

            void* data = response.request.data;

#if DEBUG_POLL_AND_WAKE || DEBUG_CALLBACKS

            ALOGD("%p ~ pollOnce - invoking fd event callback %p: fd=%d, events=0x%x, data=%p",

                    this, response.request.callback.get(), fd, events, data);

#endif

            int callbackResult = response.request.callback->handleEvent(fd, events, data);//调用callback->handleEvent

            if (callbackResult == 0) {

                removeFd(fd);

            }

            // Clear the callback reference in the response structure promptly because we

            // will not clear the response vector itself until the next poll.

            response.request.callback.clear();

            result = ALOOPER_POLL_CALLBACK;

        }

    }

搞懂了上面的代码，我们就很容易明白wake函数做了些什么：

void Looper::wake() {

#if DEBUG_POLL_AND_WAKE

    ALOGD("%p ~ wake", this);

#endif

    ssize_t nWrite;

    do {

        nWrite = write(mWakeWritePipeFd, "W", 1);//向mWakeWritePipeFd中写入一个字符，所以mWakeReadPipeFd就会触发EPOLLIN event    } while (nWrite == -1 && errno == EINTR);

    if (nWrite != 1) {

        if (errno != EAGAIN) {

            ALOGW("Could not write wake signal, errno=%d", errno);

        }

    }

}

小结：

Looper通过epoll函数组实现了一个可以支持随时唤醒的阻塞机制
Looper支持两种不同的方式处理消息：Message + MessageHandler 和 LooperCallback。
Looper的阻塞在如下四种条件下会被唤醒：
- 发生错误
- 等待超时
- 出现需要处理的新Message（包括C++层和Java层）
- 由addFd函数添加的Fd触发event

总结：

在哪个线程调用JAVA层的Looper.loop()，Mesage和callback（包括Java层和C++层）就在哪个线程被处理，上图为Looper.loop函数的时序图。
C++层的NativeMesasgeQueue不应该是Java层的MesageQueue的内部实现，而更接近于“栾生兄弟”的关系。MessageQueue负责处理java层上到消息，NativeMessageQueue负责处理C++层上的消息。其中Java层是在android.os.Looper.looper函数中调用android.os.Handler.dispatchMessage处理，而C++层是在android::Looper::pollInner函数中调用android::MessageHandler::handleMessage & android:LooperCallback::handleEvent函数处理。
NativeMessageQueue利用Looper类实现了一个基于epoll函数和文件描述符（Fd）的可唤醒的阻塞机制。