深入浅出：Android屏幕刷新机制

if (status) {//初始化出现异常
String8 message;led to initialize display event receiver. status
message.appendFormat(“Fai=%d”, status);
jniThrowRuntimeException(env, message.string());
return 0;
}
receiver->incStrong(gDisplayEventReceiverClassInfo.clazz); // retain a reference for the object
return reinterpret_cast(receiver.get());
}

我们这里先看一下NativeDisplayEventReceiver的创建过程。

[NativeDisplayEventReceiver的创建]

NativeDisplayEventReceiver::NativeDisplayEventReceiver(JNIEnv* env,
jobject receiverWeak, const sp& messageQueue, jint vsyncSource) :
//继承了DisplayEventDispatcher，并传入了对应的messagequeue，将vsyncSource转化为了底层使用的变量
DisplayEventDispatcher(messageQueue->getLooper(),
static_castISurfaceComposer::VsyncSource(vsyncSource)),
mReceiverWeakGlobal(env->NewGlobalRef(receiverWeak)),
mMessageQueue(messageQueue) {
ALOGV(“receiver %p ~ Initializing display event receiver.”, this);
}

//DisplayEventDispatcher构造函数
DisplayEventDispatcher::DisplayEventDispatcher(const sp& looper,ISurfaceComposer::VsyncSource vsyncSource) :
//Vsync的来源传递给了mReceiver。这里相当于调用了mReceiver(DisplayEventReceiver)的构造函数
mLooper(looper), mReceiver(vsyncSource), mWaitingForVsync(false) {
ALOGV(“dispatcher %p ~ Initializing display event dispatcher.”, this);
}

这里会创建DisplayEventReceiver

//DisplayEventReceiver构造函数 frameworks\native\libs\gui\DisplayEventReceiver.cpp
DisplayEventReceiver::DisplayEventReceiver(ISurfaceComposer::VsyncSource vsyncSource,
ISurfaceComposer::ConfigChanged configChanged) {
//方法1 获取SurfaceFling服务,并保存在ComposerService中
sp sf(ComposerService::getComposerService());
if (sf != nullptr) {
//方法2 通过binder，最后跨进程调用surfaceFling的createDisplayEventConnection方法
//方法位置 ISurfaceComposer.cpp frameworks\native\libs\gui 66331 2020/3/22 1379
mEventConnection = sf->createDisplayEventConnection(vsyncSource, configChanged);
if (mEventConnection != nullptr) {
//方法3
mDataChannel = std::make_uniquegui::BitTube();
//方法4
mEventConnection->stealReceiveChannel(mDataChannel.get());
}
}
}

DisplayEventReceiver结构体是一个比较重要的类，其主要作用是建立与SurfaceFlinger**的连接。我们这里将对其每一个调用的方法都来进行一个自习的分析

• 方法1：获取SurfaceFlinger服务

sp sf(ComposerService::getComposerService());

ComposerService::getComposerService()

// frameworks\native\libs\gui\SurfaceComposerClient.cpp
/static/ sp ComposerService::getComposerService() {
ComposerService& instance = ComposerService::getInstance();
Mutex::Autolock _l(instance.mLock);//加锁
if (instance.mComposerService == nullptr) {
//获取SurfaceFling服务，并保存在ComposerService中
ComposerService::getInstance().connectLocked();
assert(instance.mComposerService != nullptr);
ALOGD(“ComposerService reconnected”);
}
return instance.mComposerService;
}

void ComposerService::connectLocked() {
const String16 name(“SurfaceFlinger”);
//通过getService方法获取SurfaceFlinger服务，并将获取到的服务保存到mComposerService变量中
while (getService(name, &mComposerService) != NO_ERROR) {
usleep(250000);
}
//创建死亡回调
…
mDeathObserver = new DeathObserver(const_cast<ComposerService>(this));
IInterface::asBinder(mComposerService)->linkToDeath(mDeathObserver);
}

通过getService方法来获取对应的SurfaceFlinger服务。这里会将获取到的服务保存到mComposerService变量中.

• 创建事件连接

sf->createDisplayEventConnection

virtual sp createDisplayEventConnection(VsyncSource vsyncSource,ConfigChanged configChanged) {
Parcel data, reply;
sp result;
//binder机制调用SurfaceFling的createDisplayEventConnection方法
//SurfaceFlinger.cpp frameworks\native\services\surfaceflinger
int err = data.writeInterfaceToken(ISurfaceComposer::getInterfaceDescriptor());
data.writeInt32(static_cast<int32_t>(vsyncSource));
data.writeInt32(static_cast<int32_t>(configChanged));
err = remote()->transact(
BnSurfaceComposer::CREATE_DISPLAY_EVENT_CONNECTION,
data, &reply);
…
result = interface_cast(reply.readStrongBinder());
return result;
}

可以看到，该方法使用的是Binder机制，而服务的提供方则是SurfaceFlinger。

//创建显示事件连接
sp SurfaceFlinger::createDisplayEventConnection(
ISurfaceComposer::VsyncSource vsyncSource, ISurfaceComposer::ConfigChanged configChanged) {
//makeResyncCallback是一个方法，定义在EventThread.h中。using ResyncCallback = std::function<void()>;
//创建一个resyncCallback
auto resyncCallback = mScheduler->makeResyncCallback([this] {
Mutex::Autolock lock(mStateLock);
return getVsyncPeriod();
});
//根据传入的Vsync类型，返回不同的Handler。如果是应用中注册的，则返回mAppConnectionHandle
const auto& handle = vsyncSource == eVsyncSourceSurfaceFlinger ? mSfConnectionHandle : mAppConnectionHandle;
//调用createDisplayEventConnection，传入了对应的handle,mScheduler是Scheduler.cpp结构体
return mScheduler->createDisplayEventConnection(handle, std::move(resyncCallback),
configChanged);
}

根据传入的vsyncSource类型来返回具体的Handler。因为我们这里使用过的应用类型，所以这里的handle是mAppConnectionHandle。

然后通过mScheduler创建对应的连接。

这里我们需要对handle进行一个补充说明

补充说明：

对于Handler的创建是在SurfaceFlinger的初始化方法init()中进行创建的

void SurfaceFlinger::init() {
…
mAppConnectionHandle =
mScheduler->createConnection(“app”, mVsyncModulator.getOffsets().app,
mPhaseOffsets->getOffsetThresholdForNextVsync(),
resyncCallback,
impl::EventThread::InterceptVSyncsCallback());
…
}

spScheduler::ConnectionHandle Scheduler::createConnection(
const char* connectionName, nsecs_t phaseOffsetNs, nsecs_t offsetThresholdForNextVsync,
ResyncCallback resyncCallback,
impl::EventThread::InterceptVSyncsCallback interceptCallback) {
//对应的id，累加的
const int64_t id = sNextId++;
//创建一个EventThread,名称为传入的connectionName
std::unique_ptr eventThread =
makeEventThread(connectionName, mPrimaryDispSync.get(), phaseOffsetNs,
offsetThresholdForNextVsync, std::move(interceptCallback));
//创建EventThreadConnection
auto eventThreadConnection = createConnectionInternal(eventThread.get(), std::move(resyncCallback),
ISurfaceComposer::eConfigChangedSuppress);
//创建ConnectionHandle,入参是id，
//然后将创建的connection并存入到map中。key是id。
mConnections.emplace(id,
std::make_unique(new ConnectionHandle(id),
eventThreadConnection,
std::move(eventThread)));
return mConnections[id]->handle;
}

这里创建的Handler，持有了对应的EventThread，而eventThreadConnection是通过EventThread来进行创建。创建eventThreadConnection以后，会将其保存到map中，对应的key则是id信息。

而连接处理器：ConnectionHandle则是一个持有id的对象。

我们回到主线。。。。

mScheduler->createDisplayEventConnection

// frameworks\native\services\surfaceflinger\Scheduler\Scheduler.cpp

sp Scheduler::createDisplayEventConnection(
const spScheduler::ConnectionHandle& handle, ResyncCallback resyncCallback,
ISurfaceComposer::ConfigChanged configChanged) {
RETURN_VALUE_IF_INVALID(nullptr);
//传入了handle.id。能够表明连接是app还是surfaceFlinger
return createConnectionInternal(mConnections[handle->id]->thread.get(),
std::move(resyncCallback), configChanged);
}

sp Scheduler::createConnectionInternal(
EventThread* eventThread, ResyncCallback&& resyncCallback,
ISurfaceComposer::ConfigChanged configChanged) {
//调用EventThread的方法,创建事件连接器
return eventThread->createEventConnection(std::move(resyncCallback), configChanged);
}

我们看看事件连接器EventThreadConnection的创建过程

sp EventThread::createEventConnection(
ResyncCallback resyncCallback, ISurfaceComposer::ConfigChanged configChanged) const {
return new EventThreadConnection(const_cast<EventThread*>(this), std::move(resyncCallback),
configChanged);
}

EventThreadConnection::EventThreadConnection(EventThread* eventThread,
ResyncCallback resyncCallback,
ISurfaceComposer::ConfigChanged configChanged)
: resyncCallback(std::move(resyncCallback)),
configChanged(configChanged),
mEventThread(eventThread),
mChannel(gui::BitTube::DefaultSize) {}

EventThreadConnection的构造方法中最重要的是创建了mChannel，而它是gui::BitTube类型的

// frameworks\native\libs\gui\BitTube.cpp
BitTube::BitTube(size_t bufsize) {
init(bufsize, bufsize);
}

void BitTube::init(size_t rcvbuf, size_t sndbuf) {
int sockets[2];
if (socketpair(AF_UNIX, SOCK_SEQPACKET, 0, sockets) == 0) {
size_t size = DEFAULT_SOCKET_BUFFER_SIZE;
//创建对应一对socket：0和1，一个用来读，一个用来写。
setsockopt(sockets[0], SOL_SOCKET, SO_RCVBUF, &rcvbuf, sizeof(rcvbuf));
setsockopt(sockets[1], SOL_SOCKET, SO_SNDBUF, &sndbuf, sizeof(sndbuf));
// since we don’t use the “return channel”, we keep it small…
setsockopt(sockets[0], SOL_SOCKET, SO_SNDBUF, &size, sizeof(size));
setsockopt(sockets[1], SOL_SOCKET, SO_RCVBUF, &size, sizeof(size));
fcntl(sockets[0], F_SETFL, O_NONBLOCK);
fcntl(sockets[1], F_SETFL, O_NONBLOCK);
//将mReceiveFd文件和socket进行绑定。当Vsync到来的时候，会通过mSendFd文件来写入消息，通过对文件的消息写入监听，完成了对Vsync信号的监听
mReceiveFd.reset(sockets[0]);
mSendFd.reset(sockets[1]);
} else {
mReceiveFd.reset();
}
}

在初始化方法中，创建了一对socket，然后将mReceiveFd和mSendFd进行了绑定。当Vsync到来的时候通过mSendFd写入消息，然后APP就可以监听文件的变化。

在创建EventThreadConnection对象的时候，会自动调用onFirstRef方法。

void EventThreadConnection::onFirstRef() {
mEventThread->registerDisplayEventConnection(this);
}

status_t EventThread::registerDisplayEventConnection(const sp& connection) {
std::lock_guardstd::mutex lock(mMutex);

// this should never happen
auto it = std::find(mDisplayEventConnections.cbegin(),
mDisplayEventConnections.cend(), connection);
if (it != mDisplayEventConnections.cend()) {
ALOGW(“DisplayEventConnection %p already exists”, connection.get());
mCondition.notify_all();
return ALREADY_EXISTS;
}
//将连接放入到需要通知的列表中。
mDisplayEventConnections.push_back(connection);
//有新的连接了，就需要唤醒AppEventThread线程使能Vsync信号了。
mCondition.notify_all();
return NO_ERROR;
}

会将我们创建的连接放入到EventThread管理的mDisplayEventConnections中，然后唤醒AppEventThread线程使能Vsync信号

整个步骤二，其实是根据传入的vsyncSource，指导对应的监听者是来自APP，然后创建一对socket连接，来进行进程间的通信。

我们继续回到主线进行跟踪处理

DisplayEventReceiver::DisplayEventReceiver(ISurfaceComposer::VsyncSource vsyncSource,
ISurfaceComposer::ConfigChanged configChanged) {
//方法1 获取SurfaceFling服务,并保存在ComposerService中
sp sf(ComposerService::getComposerService());
if (sf != nullptr) {
//方法2 通过binder，最后跨进程调用surfaceFling的createDisplayEventConnection方法
//方法位置 ISurfaceComposer.cpp frameworks\native\libs\gui
mEventConnection = sf->createDisplayEventConnection(vsyncSource, configChanged);
if (mEventConnection != nullptr) {
//方法3 获取方法二中创建的gui::BitTube对象
mDataChannel = std::make_uniquegui::BitTube();
//方法4
mEventConnection->stealReceiveChannel(mDataChannel.get());
}
}
}

方法3是获取了对应的gui::BitTube对象。我们主要来分析一下方法四。

方法四调用了EventThreadConnect的stealReceiveChannel

status_t EventThreadConnection::stealReceiveChannel(gui::BitTube* outChannel) {
outChannel->setReceiveFd(mChannel.moveReceiveFd());
return NO_ERROR;
}

这的mChannel是gui::BitTube。这里将事件连接器EventThreadConnection中创建的Fd复制给了outChannel。也就是DisplayEventReceiver的mDataChannel。

所以这时候app进程就有了mReceivedFd，surfaceFlinger进程有了mSendFd。这时候通过socket就能够进行通信了。

整个DisplayEventReceiver的作用是创建一个socket以及对应的文件，然后实现和SurfaceFlinger的双向通讯。

这里我们为止，我们只是创建NativeDisplayEventReceiver。

那么后续还有

receiver->initialize()

status_t DisplayEventDispatcher::initialize() {
//异常检测
status_t result = mReceiver.initCheck();
if (result) {
ALOGW(“Failed to initialize display event receiver, status=%d”, result);
return result;
}
//这里的Looper就是应用app进程的主线程Looper，这一步就是将创建的BitTube信道的
//fd添加到Looper的监听。
int rc = mLooper->addFd(mReceiver.getFd(), 0, Looper::EVENT_INPUT,
this, NULL);
if (rc < 0) {
return UNKNOWN_ERROR;
}
return OK;
}

这里之所以能够加入到监听，是因为我们的

这里整个方法比较简单，就是进行异常的检测，让后将在步骤一中创建的fd文件加入到Looper的监听中。

到这里为止，整个流程算是打通了。

java层通过DisplayEventReceive的nativeInit函数，创建了应用层和SurfaceFlinger的连接，通过一对socket，对应mReceiveFd和mSendFd，应用层通过native层Looper将mReceiveFd加入监听，等待mSendFd的写入。

那么mSendFd什么时候写入，又是如何传递到应用层的呢？

当我们进行页面刷新绘制的时候，看一下如何注册对于Vsync的监听的

@UnsupportedAppUsage
void scheduleTraversals() {
…
mChoreographer.postCallback(Choreographer.CALLBACK_TRAVERSAL, mTraversalRunnable, null);
…
}

public void postCallback(int callbackType, Runnable action, Object token) {
postCallbackDelayed(callbackType, action, token, 0);
}

public void postCallbackDelayed(int callbackType,Runnable action, Object token, long delayMillis) {
postCallbackDelayedInternal(callbackType, action, token, delayMillis);
}

private void postCallbackDelayedInternal(int callbackType,Object action, Object token, long delayMillis) {
…
//需要立即进行绘制
scheduleFrameLocked(now);
…
}

private void scheduleFrameLocked(long now) {
…
scheduleVsyncLocked();
…
}

private void scheduleVsyncLocked() {
//执行同步功能，进行一次绘制。这里会进行一个VSYNC事件的监听注册，如果有有
mDisplayEventReceiver.scheduleVsync();
}

public void scheduleVsync() {
…
nativeScheduleVsync(mReceiverPtr);
…
}

这里的**nativeScheduleVsync()**就是应用层向native层注册监听下一次Vsync信号的方法。

nativeScheduleVsync

//base\core\jni\android_view_DisplayEventReceiver.cpp 8492 2020/9/14 96
static void nativeScheduleVsync(JNIEnv* env, jclass clazz, jlong receiverPtr) {
sp receiver =
reinterpret_cast<NativeDisplayEventReceiver*>(receiverPtr);
//调用Recivier的调度方法
status_t status = receiver->scheduleVsync();
}

这里的receiver，是NativeDisplayEventReceiver。而NativeDisplayEventReceiver是继承自DisplayEventDispatcher

DisplayEventDispatcher->scheduleVsync();

//调度Vsync
status_t DisplayEventDispatcher::scheduleVsync() {
//如果当前正在等待Vsync信号，那么直接返回
if (!mWaitingForVsync) {
nsecs_t vsyncTimestamp;
PhysicalDisplayId vsyncDisplayId;
uint32_t vsyncCount;
//重点方法1 处理对应的准备事件，如果获取到了Vsync信号的话，这里会返回true
if (processPendingEvents(&vsyncTimestamp, &vsyncDisplayId, &vsyncCount)) {
ALOGE(“dispatcher %p ~ last event processed while scheduling was for %” PRId64 “”,
this, ns2ms(static_cast<nsecs_t>(vsyncTimestamp)));
}
//重点方法2 请求下一个Vsync信号
status_t status = mReceiver.requestNextVsync();
…
//设置正在等待Vsync信号
mWaitingForVsync = true;
}
return OK;
}

这里我们跟踪一下方法1

DisplayEventDispatcher::processPendingEvents

bool DisplayEventDispatcher::processPendingEvents(
nsecs_t* outTimestamp, PhysicalDisplayId* outDisplayId, uint32_t* outCount) {
bool gotVsync = false;
DisplayEventReceiver::Event buf[EVENT_BUFFER_SIZE];
ssize_t n;
//获取对应的事件
while ((n = mReceiver.getEvents(buf, EVENT_BUFFER_SIZE)) > 0) {
ALOGV(“dispatcher %p ~ Read %d events.”, this, int(n));
for (ssize_t i = 0; i < n; i++) {
const DisplayEventReceiver::Event& ev = buf[i];
switch (ev.header.type) {
case DisplayEventReceiver::DISPLAY_EVENT_VSYNC://Vsync类型
//获取到最新的Vsync信号，然后将时间戳等信息保存下来
gotVsync = true;
*outTimestamp = ev.header.timestamp;
*outDisplayId = ev.header.displayId;
*outCount = ev.vsync.count;
break;
…
return gotVsync;
}

会通过getEvents方法获取到对应的事件类型，然后返回是否为Vsync信号。

DisplayEventReceiver::getEvents

// native\libs\gui\DisplayEventReceiver.cpp

ssize_t DisplayEventReceiver::getEvents(DisplayEventReceiver::Event* events,size_t count) {
//这里的mDataChannel是在init中创建的，用来接收Vsync信号
return DisplayEventReceiver::getEvents(mDataChannel.get(), events, count);
}
ssize_t DisplayEventReceiver::getEvents(gui::BitTube* dataChannel,
Event* events, size_t count)
{
return gui::BitTube::recvObjects(dataChannel, events, count);
}

//native\libs\gui\BitTube.cpp
static ssize_t recvObjects(BitTube* tube, T* events, size_t count) {
return recvObjects(tube, events, count, sizeof(T));
}

ssize_t BitTube::recvObjects(BitTube* tube, void* events, size_t count, size_t objSize) {
char* vaddr = reinterpret_cast<char*>(events);
//通过socket读取数据
ssize_t size = tube->read(vaddr, count * objSize);
return size < 0 ? size : size / static_cast<ssize_t>(objSize);
}
//读取数据
ssize_t BitTube::read(void* vaddr, size_t size) {
ssize_t err, len;
do {
//将mReceiveFd接收到的数据，放入到size大小的vaddr缓冲区。并返回实际接收到的数据大小len
len = ::recv(mReceiveFd, vaddr, size, MSG_DONTWAIT);
err = len < 0 ? errno : 0;
} while (err == EINTR);
if (err == EAGAIN || err == EWOULDBLOCK) {
//如果接收出现异常，返回0
return 0;
}
return err == 0 ? len : -err;
}

这里将接收到的数据放入到对应的缓冲区，并返回数据之后，会校验返回的具体的数据类型。

status_t DisplayEventReceiver::requestNextVsync() {
//校验当前连接存在
if (mEventConnection != nullptr) {
//通过连接请求下一个Vsync信号。这个mEventConnection。是在DisplayEventReceiver初始化的时候创建的
//具体的是EventThreadConnection（位于EventThread中）
mEventConnection->requestNextVsync();
return NO_ERROR;
}
return NO_INIT;
}

void EventThreadConnection::requestNextVsync() {
ATRACE_NAME(“requestNextVsync”);
mEventThread->requestNextVsync(this);
}

void EventThread::requestNextVsync(const sp& connection) {
if (connection->resyncCallback) {
connection->resyncCallback();
}
//线程锁机制
std::lock_guardstd::mutex lock(mMutex);
//vsyncRequest默认值是None.定义在EventThread.h文件中
if (connection->vsyncRequest == VSyncRequest::None) {
//之所以Vsync是一次性的，是因为，当我们当前是None之后，会将这个字段设置为Single。
//后续硬件再有Vsync信号过来的时候，不会再执行这个方法
connection->vsyncRequest = VSyncRequest::Single;
mCondition.notify_all();
}
}

这里当有Vsync的信号过来的时候，会调用一个notify_all()。这个方法会唤醒所有执行了**wait()**方法的线程。

那么这个到底会唤醒谁呢？

这里就不得不说一下EventThread创建过程中了。

EventThread::EventThread(VSyncSource* src, std::unique_ptr uniqueSrc,
InterceptVSyncsCallback interceptVSyncsCallback, const char* threadName)
: mVSyncSource(src),
mVSyncSourceUnique(std::move(uniqueSrc)),
mInterceptVSyncsCallback(std::move(interceptVSyncsCallback)),
mThreadName(threadName) {
…
//创建了mThread线程
mThread = std::thread(this NO_THREAD_SAFETY_ANALYSIS {
std::unique_lockstd::mutex lock(mMutex);
//创建线程的时候调用了threadMain函数
threadMain(lock);
});
…
}

在EventThread创建时，会创建一个线程，然后调用threadMain方法。

//在创建EventThread的时候会调用该方法。会不断的遍历
void EventThread::threadMain(std::unique_lockstd::mutex& lock) {
DisplayEventConsumers consumers;
//只要没有退出，则一直遍历循环
while (mState != State::Quit) {
std::optionalDisplayEventReceiver::Event event;
…
//是否有Vsync请求
bool vsyncRequested = false;
…
//查询所有的连接，其实这里一个连接就是一个监听
auto it = mDisplayEventConnections.begin();
while (it != mDisplayEventConnections.end()) {
if (const auto connection = it->promote()) {
vsyncRequested |= connection->vsyncRequest != VSyncRequest::None;
//遍历，将需要通知的监听放入到consumers中
if (event && shouldConsumeEvent(*event, connection)) {
consumers.push_back(connection);
}

++it;
} else {
it = mDisplayEventConnections.erase(it);
}
}

if (!consumers.empty()) {
//进行事件的分发。最终会调用gui::BitTube::sendObjects函数
dispatchEvent(*event, consumers);
consumers.clear();
}

State nextState;
if (mVSyncState && vsyncRequested) {
nextState = mVSyncState->synthetic ? State::SyntheticVSync : State::VSync;
} else {
ALOGW_IF(!mVSyncState, “Ignoring VSYNC request while display is disconnected”);
nextState = State::Idle;
}

if (mState != nextState) {
if (mState == State::VSync) {
mVSyncSource->setVSyncEnabled(false);
} else if (nextState == State::VSync) {