Android P 图形显示系统(六) SurfaceFlinger合成流程(一)

[TOC]

SurfaceFlinger合成流程(一)

通过前面的简单介绍,我们对HWC合成有大致的了解。下面我们根据实际代码进行讲解。前面章节,我们已经说过,Layer的创建,和BufferQueue,那么Buffer进入到BufferQueue队列中后,怎么进行合成显示的呢?我们继续来看。

Consumer端的FrameListener

你还记得Producer的frameAvailableListener吗?Buffer放入队列BufferQueue后,是不是通过frameAvailableListener->onFrameAvailable通知Consumer?大家可以再回望一下BufferQueueProducer::queueBuffer

frameAvailableListener是哪里来的?

我们先来看一下Consumer中Listener间的相互关系


消费者Consumer中的Listener
  • BufferLayer有专门的Consumer,BufferLayerConsumer;BufferLayerConsumer继承ConsumerBase。ConsumerBase通过IGraphicBufferConsumer和BufferQueue进行通信的。
  • BufferQueue中的frameAvailableListener,是一个IConsumerListener的接口,对应的Binder的Bn端实现为ProxyConsumerListener。
  • BufferLayer实现了ContentsChangedListener,ContentsChangedListener继承FrameAvailableListener。BufferLayer的Listener实现,被传给了ConsumerBase。
  • ConsumerBase实现ConsumerListener接口,构建ConsumerBase时,会创建ProxyConsumerListener,将ConsumerBase实现的Listener接口传给ProxyConsumerListener。
  • BufferQueue中Listener回调时,会回调到ConsumerBase中。ConsumerBase中再通过BufferLayer实现的,传下来的Listener回调到BufferLayer中。

层层回调,别弄混淆了。

关键代码:
BufferQueueProducer中通过frameAvailableListener->onFrameAvailable回调到ProxyConsumerListener中:

* frameworks/native/libs/gui/BufferQueue.cpp

void BufferQueue::ProxyConsumerListener::onFrameAvailable(
        const BufferItem& item) {
    sp<ConsumerListener> listener(mConsumerListener.promote());
    if (listener != NULL) {
        listener->onFrameAvailable(item);
    }
}

ProxyConsumerListener中的mConsumerListener是ConsumerBase中的实现。这里的listener->onFrameAvailable将回调到ConsumerBase中。

* frameworks/native/libs/gui/ConsumerBase.cpp

void ConsumerBase::onFrameAvailable(const BufferItem& item) {
    CB_LOGV("onFrameAvailable");

    sp<FrameAvailableListener> listener;
    { // scope for the lock
        Mutex::Autolock lock(mFrameAvailableMutex);
        listener = mFrameAvailableListener.promote();
    }

    if (listener != NULL) {
        CB_LOGV("actually calling onFrameAvailable");
        listener->onFrameAvailable(item);
    }
}

ConsumerBase中的mFrameAvailableListener是BufferLayer中的实现:

* frameworks/native/services/surfaceflinger/BufferLayer.cpp

void BufferLayer::onFrameAvailable(const BufferItem& item) {
    // Add this buffer from our internal queue tracker
    { // Autolock scope
        Mutex::Autolock lock(mQueueItemLock);
        mFlinger->mInterceptor.saveBufferUpdate(this, item.mGraphicBuffer->getWidth(),
                                                item.mGraphicBuffer->getHeight(),
                                                item.mFrameNumber);
        // Reset the frame number tracker when we receive the first buffer after
        // a frame number reset
        if (item.mFrameNumber == 1) {
            mLastFrameNumberReceived = 0;
        }

        // Ensure that callbacks are handled in order
        while (item.mFrameNumber != mLastFrameNumberReceived + 1) {
            status_t result = mQueueItemCondition.waitRelative(mQueueItemLock,
                                                               ms2ns(500));
            if (result != NO_ERROR) {
                ALOGE("[%s] Timed out waiting on callback", mName.string());
            }
        }

        mQueueItems.push_back(item);
        android_atomic_inc(&mQueuedFrames);

        // Wake up any pending callbacks
        mLastFrameNumberReceived = item.mFrameNumber;
        mQueueItemCondition.broadcast();
    }

    mFlinger->signalLayerUpdate();
}

BufferLayer中调用onFrameAvailable,去通知SurfaceFlinger进行合成。

到这里,应用端(Producer)生产完Buffer这件事,就通知到了SurfaceFlinger中了。

SurfaceFlinger的signalLayerUpdate,是通过MessageQueue来处理的,我们先来看看MessageQueue。

消息队列MessageQueue

MessageQueue是SurfaceFlinger中的消息队列,为什么需要消息队列?我们应用有一个主线程,专门进行UI的处理。SurfaceFlinger同样的,也有一个主线程,SurfaceFlinger的主线程主要进行显示数据的处理,也就是合成。

SurfaceFlinger中,mEventQueue是MessageQueue的一个栈对象,采用mutable修饰。SurfaceFlinger在初次引用时,会对mEventQueue进行初始化。

* frameworks/native/services/surfaceflinger/MessageQueue.cpp

void MessageQueue::init(const sp<SurfaceFlinger>& flinger)
{
    mFlinger = flinger;
    mLooper = new Looper(true);
    mHandler = new Handler(*this);
}

MessageQueue初始化时,创建了一个Looper和一个Handler。

此外,在SurfaceFlinger初始化时,创建了一个EventThread,并传给了MessageQueue。

void SurfaceFlinger::init() {
    ... ...
    sp<VSyncSource> sfVsyncSrc =
            new DispSyncSource(&mPrimaryDispSync, SurfaceFlinger::sfVsyncPhaseOffsetNs, true, "sf");
    mSFEventThread = new EventThread(sfVsyncSrc, *this, true);
    mEventQueue.setEventThread(mSFEventThread);

MessageQueue的setEventThread函数如下:

void MessageQueue::setEventThread(const sp<EventThread>& eventThread)
{
    if (mEventThread == eventThread) {
        return;
    }

    if (mEventTube.getFd() >= 0) {
        mLooper->removeFd(mEventTube.getFd());
    }

    mEventThread = eventThread;
    mEvents = eventThread->createEventConnection();
    mEvents->stealReceiveChannel(&mEventTube);
    mLooper->addFd(mEventTube.getFd(), 0, Looper::EVENT_INPUT,
            MessageQueue::cb_eventReceiver, this);
}

MessageQueue在setEventThread时,主要做以下几件事:

  • 创建一个BitTube对象mEventTube

  • 创建一个EventConnection

sp<EventThread::Connection> EventThread::createEventConnection() const {
    return new Connection(const_cast<EventThread*>(this));
}

Connection在第一次引用时,将会被注册到mEventThread中。

void EventThread::Connection::onFirstRef() {
    // NOTE: mEventThread doesn't hold a strong reference on us
    mEventThread->registerDisplayEventConnection(this);
}

在注册时,Connection将会被添加到mDisplayEventConnections 中。

status_t EventThread::registerDisplayEventConnection(
        const sp<EventThread::Connection>& connection) {
    Mutex::Autolock _l(mLock);
    mDisplayEventConnections.add(connection);
    mCondition.broadcast();
    return NO_ERROR;
}

mDisplayEventConnections是一个已经注册的Connection的集合。

  • 将mEventTube和EventConnection关联
status_t EventThread::Connection::stealReceiveChannel(gui::BitTube* outChannel) {
    outChannel->setReceiveFd(mChannel.moveReceiveFd());
    return NO_ERROR;
}

Connection创建时,将默认创建一个4k的BitTube,BitTube封装的是一对socket,一个发送,一个接收,可传输的Buffer大小为4K。

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;
        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);
        mReceiveFd.reset(sockets[0]);
        mSendFd.reset(sockets[1]);
    } else {
        mReceiveFd.reset();
        ALOGE("BitTube: pipe creation failed (%s)", strerror(errno));
    }
}

MessageQueue中的mEventTube,和mReceiveFd关联。

  • addFd函数,将fd添加到MessageQueue的Looper中。
    注意,Looper的callback为MessageQueue::cb_eventReceiver,data为MessageQueue本身。
int MessageQueue::cb_eventReceiver(int fd, int events, void* data) {
    MessageQueue* queue = reinterpret_cast<MessageQueue *>(data);
    return queue->eventReceiver(fd, events);
}

回到SurfaceFlinger的signalLayerUpdate函数:

void SurfaceFlinger::signalLayerUpdate() {
    mEventQueue.invalidate();
}

signalLayerUpdate中,调用MQ的invalidate。

void MessageQueue::invalidate() {
    mEvents->requestNextVsync();
}

MQ的invalidate的函数,将请求下一个Vsync。Vsync是一种同步机制,垂直同步,我们可以理解为SurfaceFlinger的工作节拍。

void EventThread::requestNextVsync(
        const sp<EventThread::Connection>& connection) {
    Mutex::Autolock _l(mLock);

    mFlinger.resyncWithRateLimit();

    if (connection->count < 0) {
        connection->count = 0;
        mCondition.broadcast();
    }
}

注意这里的count >= 0。

EventThread就是一个事件分发的线程,第一次引用时,线程启动。

void EventThread::onFirstRef() {
    run("EventThread", PRIORITY_URGENT_DISPLAY + PRIORITY_MORE_FAVORABLE);
}

EventThread的threadLoop函数体如下:

bool EventThread::threadLoop() {
    DisplayEventReceiver::Event event;
    Vector< sp<EventThread::Connection> > signalConnections;
    signalConnections = waitForEvent(&event);

    // dispatch events to listeners...
    const size_t count = signalConnections.size();
    for (size_t i=0 ; i<count ; i++) {
        const sp<Connection>& conn(signalConnections[i]);
        // now see if we still need to report this event
        status_t err = conn->postEvent(event);
        if (err == -EAGAIN || err == -EWOULDBLOCK) {
            ... ...
            ALOGW("EventThread: dropping event (%08x) for connection %p",
                    event.header.type, conn.get());
        } else if (err < 0) {
            removeDisplayEventConnection(signalConnections[i]);
        }
    }
    return true;
}

waitForEvent,等待事件Event。看看哪些Connection是被触发的,对于被触发的Connection,signalConnections,通过postEvent将事件Event分发出去。

Event这边的控制逻辑,基本都在waitForEvent中。waitForEvent中,采用while循环,条件是signalConnections为空。EventThread中主要控制两事件,Vsync事件和显示屏的HotPlug热插拔事件

    enum {
        DISPLAY_EVENT_VSYNC = fourcc('v', 's', 'y', 'n'),
        DISPLAY_EVENT_HOTPLUG = fourcc('p', 'l', 'u', 'g'),
    };

我们分段来看:

Vector< sp<EventThread::Connection> > EventThread::waitForEvent(
        DisplayEventReceiver::Event* event)
{
    Mutex::Autolock _l(mLock);
    Vector< sp<EventThread::Connection> > signalConnections;

    do {
        bool eventPending = false;
        bool waitForVSync = false;

        size_t vsyncCount = 0;
        nsecs_t timestamp = 0;
        for (int32_t i=0 ; i<DisplayDevice::NUM_BUILTIN_DISPLAY_TYPES ; i++) {
            timestamp = mVSyncEvent[i].header.timestamp;
            if (timestamp) {
                // 有一个Vsync事件要分发
                if (mInterceptVSyncs) {
                    mFlinger.mInterceptor.saveVSyncEvent(timestamp);
                }
                *event = mVSyncEvent[i];
                mVSyncEvent[i].header.timestamp = 0;
                vsyncCount = mVSyncEvent[i].vsync.count;
                break;
            }
        }

看看有没有Vsync事件要分发,timestamp不为0,表示有Vync事件要分发。

        if (!timestamp) {
            // no vsync event, see if there are some other event
            eventPending = !mPendingEvents.isEmpty();
            if (eventPending) {
                // we have some other event to dispatch
                *event = mPendingEvents[0];
                mPendingEvents.removeAt(0);
            }
        }

mPendingEvents,这里主要是是HotPlug事件。

找出在等待事件的Connection:

        size_t count = mDisplayEventConnections.size();
        for (size_t i=0 ; i<count ; ) {
            sp<Connection> connection(mDisplayEventConnections[i].promote());
            if (connection != NULL) {
                bool added = false;
                if (connection->count >= 0) {
                    // 需要Vsync,至少有一个Connection的count >= 0。
                    waitForVSync = true;
                    if (timestamp) {
                        // we consume the event only if it's time
                        // (ie: we received a vsync event)
                        if (connection->count == 0) {
                            // fired this time around
                            connection->count = -1;
                            signalConnections.add(connection);
                            added = true;
                        } else if (connection->count == 1 ||
                                (vsyncCount % connection->count) == 0) {
                            // continuous event, and time to report it
                            signalConnections.add(connection);
                            added = true;
                        }
                    }
                }

                if (eventPending && !timestamp && !added) {
                    // 没有Vsync事件要处理,但是有其他的事件要处理
                    signalConnections.add(connection);
                }
                ++i;
            } else {
                // Connection不存在了
                mDisplayEventConnections.removeAt(i);
                --count;
            }
        }

Connection的count >= 0表示需要Vsync。eventPending && !timestamp && !added表示没有Vsync事件要处理,但是有其他的事件要处理。

        if (timestamp && !waitForVSync) {
            disableVSyncLocked();
        } else if (!timestamp && waitForVSync) {
            enableVSyncLocked();
        }

timestamp && !waitForVSync如果有Vsync事件要分发,但是又没有Connection需要Vsync事件时,把Vsync给关掉。相反,如果有Connection需要Vsync,而此时又没有Vsync事件时,需要将Vsync打开。

        if (!timestamp && !eventPending) {
            if (waitForVSync) {
                bool softwareSync = mUseSoftwareVSync;
                nsecs_t timeout = softwareSync ? ms2ns(16) : ms2ns(1000);
                if (mCondition.waitRelative(mLock, timeout) == TIMED_OUT) {
                    if (!softwareSync) {
                        ALOGW("Timed out waiting for hw vsync; faking it");
                    }
                    // FIXME: how do we decide which display id the fake
                    // vsync came from ?
                    mVSyncEvent[0].header.type = DisplayEventReceiver::DISPLAY_EVENT_VSYNC;
                    mVSyncEvent[0].header.id = DisplayDevice::DISPLAY_PRIMARY;
                    mVSyncEvent[0].header.timestamp = systemTime(SYSTEM_TIME_MONOTONIC);
                    mVSyncEvent[0].vsync.count++;
                }
            } else {
                mCondition.wait(mLock);
            }
        }
    } while (signalConnections.isEmpty());

    return signalConnections;
}

如果此时没有Vsync事件,或其他的Event事件,那就处于等待中。如果是等待Vsync,那么通过mCondition.waitRelative进行等待,如果是硬件Vsync还不能用或者出现问题时,设置一个超时时间,进行屏幕的唤醒。
如果Connection需要Vsync,那么就进程sleep。

Vsync事件到来时,将回调到onVSyncEvent:

void EventThread::onVSyncEvent(nsecs_t timestamp) {
    Mutex::Autolock _l(mLock);
    mVSyncEvent[0].header.type = DisplayEventReceiver::DISPLAY_EVENT_VSYNC;
    mVSyncEvent[0].header.id = 0;
    mVSyncEvent[0].header.timestamp = timestamp;
    mVSyncEvent[0].vsync.count++;
    mCondition.broadcast();
}

Hotplug事件到来时,将回调到onHotplugReceived:

void EventThread::onHotplugReceived(int type, bool connected) {
    ALOGE_IF(type >= DisplayDevice::NUM_BUILTIN_DISPLAY_TYPES,
            "received hotplug event for an invalid display (id=%d)", type);

    Mutex::Autolock _l(mLock);
    if (type < DisplayDevice::NUM_BUILTIN_DISPLAY_TYPES) {
        DisplayEventReceiver::Event event;
        event.header.type = DisplayEventReceiver::DISPLAY_EVENT_HOTPLUG;
        event.header.id = type;
        event.header.timestamp = systemTime();
        event.hotplug.connected = connected;
        mPendingEvents.add(event);
        mCondition.broadcast();
    }
}

注意,Event事件哪儿回调回来的我们先不管,我们先记住这里的逻辑。

Vsync事件是跟屏幕的刷新频率有关,比如60Hz的屏幕,两个Vsync事件间的时间为1/60s,也就是16.67ms左右。SurfaceFlinger每隔16.67ms进行一次合成,显示。

另外,需要注意的是,SurfaceFlinger和App的EventThread是分开的,不是同一个。

void SurfaceFlinger::init() {
    ... ...

    // start the EventThread
    sp<VSyncSource> vsyncSrc =
            new DispSyncSource(&mPrimaryDispSync, SurfaceFlinger::vsyncPhaseOffsetNs, true, "app");
    mEventThread = new EventThread(vsyncSrc, *this, false);
    sp<VSyncSource> sfVsyncSrc =
            new DispSyncSource(&mPrimaryDispSync, SurfaceFlinger::sfVsyncPhaseOffsetNs, true, "sf");
    mSFEventThread = new EventThread(sfVsyncSrc, *this, true);
    mEventQueue.setEventThread(mSFEventThread);

回到MessageQueue,Connection通过postEvent将Event抛出来后,通过sendEvents将事件发出去。

status_t EventThread::Connection::postEvent(
        const DisplayEventReceiver::Event& event) {
    ssize_t size = DisplayEventReceiver::sendEvents(&mChannel, &event, 1);
    return size < 0 ? status_t(size) : status_t(NO_ERROR);
}

DisplayEventReceiver中是通过BitTube将事件发出去,sendObjects注意这里的参数。

ssize_t DisplayEventReceiver::sendEvents(gui::BitTube* dataChannel,
        Event const* events, size_t count)
{
    return gui::BitTube::sendObjects(dataChannel, events, count);
}

数据是什么地方接受的呢?回到SurfaceFlinger

SurfaceFlinger线程run时,启动一个死循环,循环等待事件。

void SurfaceFlinger::run() {
    do {
        waitForEvent();
    } while (true);
}

waitForEvent中,调用MessageQueue的waitMessage函数:

void MessageQueue::waitMessage() {
    do {
        IPCThreadState::self()->flushCommands();
        int32_t ret = mLooper->pollOnce(-1);
        switch (ret) {
            case Looper::POLL_WAKE:
            case Looper::POLL_CALLBACK:
                continue;
            case Looper::POLL_ERROR:
                ALOGE("Looper::POLL_ERROR");
                continue;
            case Looper::POLL_TIMEOUT:
                // timeout (should not happen)
                continue;
            default:
                // should not happen
                ALOGE("Looper::pollOnce() returned unknown status %d", ret);
                continue;
        }
    } while (true);
}

waitMessage,通过采用一个死循环,处理Looper的pollOnce。Looper内部的逻辑就不看了,主要是采用epoll_wait对fd进行监听,BitTube发送Event对象后,epoll_wait结束,调用callback,处理事件

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

MessageQueue对应的callback为cb_eventReceiver:

int MessageQueue::cb_eventReceiver(int fd, int events, void* data) {
    MessageQueue* queue = reinterpret_cast<MessageQueue *>(data);
    return queue->eventReceiver(fd, events);
}

eventReceiver,处理事件:

int MessageQueue::eventReceiver(int /*fd*/, int /*events*/) {
    ssize_t n;
    DisplayEventReceiver::Event buffer[8];
    while ((n = DisplayEventReceiver::getEvents(&mEventTube, buffer, 8)) > 0) {
        for (int i=0 ; i<n ; i++) {
            if (buffer[i].header.type == DisplayEventReceiver::DISPLAY_EVENT_VSYNC) {
                mHandler->dispatchInvalidate();
                break;
            }
        }
    }
    return 1;
}

dispatchInvalidate,封装为MessageQueue::INVALIDATE

void MessageQueue::Handler::dispatchInvalidate() {
    if ((android_atomic_or(eventMaskInvalidate, &mEventMask) & eventMaskInvalidate) == 0) {
        mQueue.mLooper->sendMessage(this, Message(MessageQueue::INVALIDATE));
    }
}

MessageQueue中,两种Message,INVALIDATEREFRESH

    enum {
        INVALIDATE  = 0,
        REFRESH     = 1,
    };

Message的分发略过,Handler对Message的处理如下:

void MessageQueue::Handler::handleMessage(const Message& message) {
    switch (message.what) {
        case INVALIDATE:
            android_atomic_and(~eventMaskInvalidate, &mEventMask);
            mQueue.mFlinger->onMessageReceived(message.what);
            break;
        case REFRESH:
            android_atomic_and(~eventMaskRefresh, &mEventMask);
            mQueue.mFlinger->onMessageReceived(message.what);
            break;
    }
}

收到消息后,再调回SurfaceFlinger在onMessageReceived中处理。

再看SurfaceFlinger的处理之前,我们先稍微整理一下MessageQueue,MessageQueue的类图如下:


MessageQueue的类图

显示设备DispalyDevice

SurfaceFlinger中,每个显示屏我们用DisplayDevice进行描述,它除了描述了Display的信息,还包括很多和合成相关的逻辑。相比于native层,Display信息是在Android的Framework层管理的,提供了专门的服务DisplayManagerService(DMS),DMS后续再介绍。

Display接口服务

。从Android 8.0开始,Vsync和hotplug的接收接口IDisplayEventReceiver作为一个单独的库从SurfaceFlinger中独立出来,设计为3层模式,JAVA层,Native层和HAL层。编译为libdisplayservicehidl库,代码在如下位置:

frameworks/native/services/displayservice

Display 接口 Android.bp如下:

cc_library_shared {
    name: "libdisplayservicehidl",

    srcs: [
        "DisplayService.cpp",
        "DisplayEventReceiver.cpp",
    ],

    shared_libs: [
        "libbase",
        "liblog",
        "libgui",
        "libhidlbase",
        "libhidltransport",
        "libutils",
        "android.frameworks.displayservice@1.0",
    ],

    export_include_dirs: ["include"],
    export_shared_lib_headers: [
        "android.frameworks.displayservice@1.0",
        "libgui",
        "libutils",
    ],

    cflags: [
        "-Werror",
        "-Wall",
    ]
}

hal层也抽象出Display的单独模块displayservice。代码为在:

frameworks/hardware/interfaces/displayservice

displayservice的Android.bp如下:

hidl_interface {
    name: "android.frameworks.displayservice@1.0",
    root: "android.frameworks",
    vndk: {
        enabled: true,
    },
    srcs: [
        "types.hal",
        "IDisplayEventReceiver.hal",
        "IDisplayService.hal",
        "IEventCallback.hal",
    ],
    interfaces: [
        "android.hidl.base@1.0",
    ],
    types: [
        "Status",
    ],
    gen_java: true,
}

displayservice还比较简单,没有太多接口:

  • types中只定义了一个状态
* frameworks/hardware/interfaces/displayservice/1.0/types.hal

package android.frameworks.displayservice@1.0;

enum Status : uint32_t {
    SUCCESS,
    BAD_VALUE,
    UNKNOWN,
};
  • IDisplayEventReceiver.hal中定义了Receiver的接口
package android.frameworks.displayservice@1.0;

import IEventCallback;

interface IDisplayEventReceiver {
    /**
     * 添加callback,开始接收Events事件,热插拔是默认打开的,Vysnc需要通过setVsyncRate打开
     */
    init(IEventCallback callback) generates (Status status);

    /**
     * 开始或停止发送callback
     */
    setVsyncRate(int32_t count) generates (Status status);

    /**
     * 请求一个Vsync,如果setVsyncRate是0,这不起作用
     */
    requestNextVsync() generates (Status status);

    /**
     * Server端丢弃所以的callback,停止发送
     */
    close() generates (Status status);
};
  • IDisplayService.hal
package android.frameworks.displayservice@1.0;

import IDisplayEventReceiver;

interface IDisplayService {
    /**
     * 创建新的receiver.
     */
    getEventReceiver() generates(IDisplayEventReceiver receiver);
};
  • IEventCallback.hal
package android.frameworks.displayservice@1.0;

interface IEventCallback {
    /**
     * Vsync事件
     */
    oneway onVsync(uint64_t timestamp, uint32_t count);

    /**
     * hotplug事件
     */
    oneway onHotplug(uint64_t timestamp, bool connected);
};

displayservice的接口,主要是提供给Vendor的HAL使用,让Vendor的HAL也能够接收Vsync数据。libdisplayservicehidl中也主要是DisplayEventReceiver。所以,这里的IDisplayEventReceiver接口 这么设计的 主要是提供给Vendor用。

显示屏的类型

显示屏幕什么时候创建?各类型的显示屏不一样,Android支持3中类型的显示屏:主显外显虚显

* frameworks/native/services/surfaceflinger/DisplayDevice.h

    enum DisplayType {
        DISPLAY_ID_INVALID = -1,
        DISPLAY_PRIMARY     = HWC_DISPLAY_PRIMARY,
        DISPLAY_EXTERNAL    = HWC_DISPLAY_EXTERNAL,
        DISPLAY_VIRTUAL     = HWC_DISPLAY_VIRTUAL,
        NUM_BUILTIN_DISPLAY_TYPES = HWC_NUM_PHYSICAL_DISPLAY_TYPES,
    };

主显示屏幕和外显,都采用热插拔的形式,连接断开时从底层驱动上报热插拔事件,这是在EventThread中处理的。

  • 主显示屏 DISPLAY_PRIMARY
    主显示屏幕默认是必现支持的,也就是说,开机时就应该上报 * 连接* 事件,知道屏幕关闭时,才断开。这里说的屏幕关闭是真正的关闭,休眠,锁屏等状态屏幕还是开着的。基本也就是关机的时候。

  • 外显示屏 DISPLAY_EXTERNAL
    外显示屏幕,一般是有线连接的屏幕,比如HDMI,MHL或者是其他连接标准连接的屏幕,外显一般经常进行热插拔。

  • 虚拟显示屏 DISPLAY_VIRTUAL
    虚拟显示屏,也就是说这个显示屏是不存在物理设备的,是个虚拟的。

我们先看看DisplayDevice相关类之间的关系:


DisplayDevice和Surface相关的类图

我们可以这么来理解:

  • Android支持多个显示屏幕,每一个显示屏幕用DisplayDevice进行描述,SurfaceFlinger中有这些显示屏的信息即mDisplay
  • 每个显示屏幕,都一个相关联的Buffer,这个Buffer用 DisplaySurface进行描述。
  • 每种类型的显示屏,具体采用的Buffer不尽相同,主屏和外显采用FramebufferSurface,而虚拟显示屏采用VirtualDisplaySurface。
  • DisplaySurface有自己的BufferQueue,都继承ConsumerBase,所以这里DisplaySurface比较特殊,都是消费者。VirtualDisplaySurface更猛,它还继承BnGraphicBufferProducer,也是生产者。
  • RE_Surface是RE命名空间的Surface,和BufferQueue中的Surface同名,但是作用不一样。主要是给RenderEngine用,RenderEngine是一个抽象类,渲染引擎,用于Client端的合成,一般用OpenGL进程合成。

DisplayDevice的热插拔处理

SurfaceFlinger在初始化的时候,注册Callback接口后。显示屏幕插上和断开时,将通过HAL回调回来。HAL回调的过程先不关注,从SurfaceFlinger中开始看。

void SurfaceFlinger::onHotplugReceived(int32_t sequenceId,
        hwc2_display_t display, HWC2::Connection connection,
        bool primaryDisplay) {
    ALOGV("onHotplugReceived(%d, %" PRIu64 ", %s, %s)",
          sequenceId, display,
          connection == HWC2::Connection::Connected ?
                  "connected" : "disconnected",
          primaryDisplay ? "primary" : "external");

    ConditionalLock lock(mStateLock,
            std::this_thread::get_id() != mMainThreadId);

    if (primaryDisplay) {
        getBE().mHwc->onHotplug(display, connection);
        if (!mBuiltinDisplays[DisplayDevice::DISPLAY_PRIMARY].get()) {
            createBuiltinDisplayLocked(DisplayDevice::DISPLAY_PRIMARY);
        }
        createDefaultDisplayDevice();
    } else {
        if (sequenceId != getBE().mComposerSequenceId) {
            return;
        }
        if (getBE().mHwc->isUsingVrComposer()) {
            ALOGE("External displays are not supported by the vr hardware composer.");
            return;
        }
        getBE().mHwc->onHotplug(display, connection);
        auto type = DisplayDevice::DISPLAY_EXTERNAL;
        if (connection == HWC2::Connection::Connected) {
            createBuiltinDisplayLocked(type);
        } else {
            mCurrentState.displays.removeItem(mBuiltinDisplays[type]);
            mBuiltinDisplays[type].clear();
        }
        setTransactionFlags(eDisplayTransactionNeeded);
    }
}

接收到屏幕插拔事件后,主要做了如下的处理:

  • 通知HWC onHotplug

通过onHotplug通知HWC;如果是连接,HWC将去获取新添加Display的config信息,如果是断开,将HWC中的Display同步断开。

  • 创建Display的Token
    createBuiltinDisplayLocked如果是连接状态,都将通过createBuiltinDisplayLocked创建Display的Token。添加到mBuiltinDisplays中,mBuiltinDisplays它只是Display的IBinder列表,我们也称之为Token列表。
sp<IBinder> mBuiltinDisplays[DisplayDevice::NUM_BUILTIN_DISPLAY_TYPES];

这个时候还没有创建DisplayDevice,只是创建了一个Token而已。Display的Token通知添加到mCurrentState的displays中。

  • 创建主屏幕对应的DisplayDevice

如果是主屏,还会通过createDefaultDisplayDevice创建默认的DisplayDevice。

void SurfaceFlinger::createDefaultDisplayDevice() {
    const DisplayDevice::DisplayType type = DisplayDevice::DISPLAY_PRIMARY;
    wp<IBinder> token = mBuiltinDisplays[type];

    // All non-virtual displays are currently considered secure.
    const bool isSecure = true;

    sp<IGraphicBufferProducer> producer;
    sp<IGraphicBufferConsumer> consumer;
    BufferQueue::createBufferQueue(&producer, &consumer);

    sp<FramebufferSurface> fbs = new FramebufferSurface(*getBE().mHwc, type, consumer);

    bool hasWideColorModes = false;
    std::vector<android_color_mode_t> modes = getHwComposer().getColorModes(type);
    for (android_color_mode_t colorMode : modes) {
        switch (colorMode) {
            case HAL_COLOR_MODE_DISPLAY_P3:
            case HAL_COLOR_MODE_ADOBE_RGB:
            case HAL_COLOR_MODE_DCI_P3:
                hasWideColorModes = true;
                break;
            default:
                break;
        }
    }
    bool useWideColorMode = hasWideColorModes && hasWideColorDisplay && !mForceNativeColorMode;
    sp<DisplayDevice> hw = new DisplayDevice(this, DisplayDevice::DISPLAY_PRIMARY, type, isSecure,
                                             token, fbs, producer, useWideColorMode);
    mDisplays.add(token, hw);
    android_color_mode defaultColorMode = HAL_COLOR_MODE_NATIVE;
    if (useWideColorMode) {
        defaultColorMode = HAL_COLOR_MODE_SRGB;
    }
    setActiveColorModeInternal(hw, defaultColorMode);
    hw->setCompositionDataSpace(HAL_DATASPACE_UNKNOWN);

    // Add the primary display token to mDrawingState so we don't try to
    // recreate the DisplayDevice for the primary display.
    mDrawingState.displays.add(token, DisplayDeviceState(type, true));

    // make the GLContext current so that we can create textures when creating
    // Layers (which may happens before we render something)
    hw->makeCurrent();
}
  • 看到没有,主Display的BufferQueue是就是此时创建的。并且创建了前面所说的FramebufferSurface,这就是FBTarget。
  • DisplayDevice对应的类型为DISPLAY_PRIMARY,创建了DisplayDevice后,添加到mDisplays中。而Display的Token被添加到mDrawingState的displays中,注意,是mDrawingState。
  • 这里还有颜色模式Colormode和数据空间DataSpace的设置,这两个概念后续介绍。
  • 最后,调用makeCurrent,表示该DisplayDevice进入可用状态
bool DisplayDevice::makeCurrent() const {
    bool success = mFlinger->getRenderEngine().setCurrentSurface(mSurface);
    setViewportAndProjection();
    return success;
}

makeCurrent主要做了两件事:其一,设置RenderEngine的Surface,这个Surface封装了前面BufferQueue中创建的Producer,以及对应的NativeWindow;其二,设置Display的Viewport和Projection,做过OpengGL开发的,对这个应该没有什么难题,视窗大小和投影矩阵。也是设置到RenderEngine中,GPU合成用。

  • 非主屏删除处理
    非主屏时删除时,将Display的Token从mBuiltinDisplays中删掉,且Token也从mCurrentState中删掉。注意,这里的是mCurrentState。主屏一般只会添加一次,没有断开操纵,断开时系统已经关了。

  • Transaction处理
    连接或断开显示屏,也算是一种Transaction。通过setTransactionFlags,设置处理的flag eDisplayTransactionNeeded。

到此,连接时,主显的Token是添加到mDrawingState中的,已经创建对应的DisplayDevice,且没有断开处理。而非主显只创建了Display的Token,添加到这里的是mCurrentState中,还没有创建对应的DisplayDevice,断开的屏幕,Token从mCurrentState删除。简单点来说,mDrawingState中的Token都创建了DisplayDevice;在mCurrentState中的不再mDrawingState中的,都是添加的;在mDrawingState中的,不在mCurrentState中的都是断开的。

创建DisplayDevice

上面设置的setTransactionFlags什么时候处理?Vsync来的时候,Vsync来后,通过INVALIDATE消息,又回到SurfaceFlinger处理。中间的过程稍后介绍,我们直接看对这里设置的eDisplayTransactionNeeded的处理流程。

处理eDisplayTransactionNeeded时,其实就是同步 mDrawingState 和 mCurrentState 中displays。

  • 屏幕断开时的处理
    先处理删除的Display,屏幕的Token在mDrawingState中的,不在mCurrentState中的都是断开的。逻辑如下:
void SurfaceFlinger::handleTransactionLocked(uint32_t transactionFlags)
{
            ``` ```
    if (transactionFlags & eDisplayTransactionNeeded) {
        const KeyedVector<  wp<IBinder>, DisplayDeviceState>& curr(mCurrentState.displays);
        const KeyedVector<  wp<IBinder>, DisplayDeviceState>& draw(mDrawingState.displays);
        if (!curr.isIdenticalTo(draw)) {
            mVisibleRegionsDirty = true;
            const size_t cc = curr.size();
                  size_t dc = draw.size();

            for (size_t i=0 ; i<dc ;) {
                const ssize_t j = curr.indexOfKey(draw.keyAt(i));
                if (j < 0) {
                    if (!draw[i].isMainDisplay()) {
                        const sp<const DisplayDevice> defaultDisplay(getDefaultDisplayDeviceLocked());
                        defaultDisplay->makeCurrent();
                        sp<DisplayDevice> hw(getDisplayDeviceLocked(draw.keyAt(i)));
                        if (hw != NULL)
                            hw->disconnect(getHwComposer());
                        if (draw[i].type < DisplayDevice::NUM_BUILTIN_DISPLAY_TYPES)
                            mEventThread->onHotplugReceived(draw[i].type, false);
                        mDisplays.removeItem(draw.keyAt(i));
                    } else {
                        ALOGW("trying to remove the main display");
                    }
                } else {
                    // this display is in both lists. see if something changed.
                    const DisplayDeviceState& state(curr[j]);
                    const wp<IBinder>& display(curr.keyAt(j));
                    const sp<IBinder> state_binder = IInterface::asBinder(state.surface);
                    const sp<IBinder> draw_binder = IInterface::asBinder(draw[i].surface);
                    if (state_binder != draw_binder) {
                        sp<DisplayDevice> hw(getDisplayDeviceLocked(display));
                        if (hw != NULL)
                            hw->disconnect(getHwComposer());
                        mDisplays.removeItem(display);
                        mDrawingState.displays.removeItemsAt(i);
                        dc--;
                        // at this point we must loop to the next item
                        continue;
                    }

                    const sp<DisplayDevice> disp(getDisplayDeviceLocked(display));
                    if (disp != NULL) {
                        if (state.layerStack != draw[i].layerStack) {
                            disp->setLayerStack(state.layerStack);
                        }
                        if ((state.orientation != draw[i].orientation)
                                || (state.viewport != draw[i].viewport)
                                || (state.frame != draw[i].frame))
                        {
                            disp->setProjection(state.orientation,
                                    state.viewport, state.frame);
                        }
                        if (state.width != draw[i].width || state.height != draw[i].height) {
                            disp->setDisplaySize(state.width, state.height);
                        }
                    }
                }
                ++i;
            }

如果Token在mDrawingState中,而没有在mCurrentState中,说明这个屏已经被断开了,需要删掉DisplayDevice。如果Token在两个状态中都存在,有修改,暂时将Token中mDrawingState删掉。注意两个状态啊,要不然理解不对。

断开时,调用DisplayDevice的disconnect,这其中将调用HWC中,HWC中创建的对应的Device也将被删除。同时将DisplayDevice从mDisplays中删除。

  • 处理添加的Display
    屏幕的Token在mCurrentState中的不再mDrawingState中的,都是添加的;处理逻辑如下。
void SurfaceFlinger::handleTransactionLocked(uint32_t transactionFlags)
{
            ``` ```
            for (size_t i=0 ; i<cc ; i++) {
                if (draw.indexOfKey(curr.keyAt(i)) < 0) {
                    const DisplayDeviceState& state(curr[i]);

                    sp<DisplaySurface> dispSurface;
                    sp<IGraphicBufferProducer> producer;
                    sp<IGraphicBufferProducer> bqProducer;
                    sp<IGraphicBufferConsumer> bqConsumer;
                    BufferQueue::createBufferQueue(&bqProducer, &bqConsumer);

                    int32_t hwcId = -1;
                    if (state.isVirtualDisplay()) {
                        // Virtual displays without a surface are dormant:
                        // they have external state (layer stack, projection,
                        // etc.) but no internal state (i.e. a DisplayDevice).
                        if (state.surface != NULL) {

                            // 虚拟显示用硬件
                            ... ...

                            sp<VirtualDisplaySurface> vds =
                                    new VirtualDisplaySurface(*getBE().mHwc,
                                            hwcId, state.surface, bqProducer,
                                            bqConsumer, state.displayName);

                            dispSurface = vds;
                            producer = vds;
                        }
                    } else {
                        ALOGE_IF(state.surface!=NULL,
                                "adding a supported display, but rendering "
                                "surface is provided (%p), ignoring it",
                                state.surface.get());

                        hwcId = state.type;
                        dispSurface = new FramebufferSurface(*getBE().mHwc, hwcId, bqConsumer);
                        producer = bqProducer;
                    }

                    const wp<IBinder>& display(curr.keyAt(i));
                    if (dispSurface != NULL) {
                        sp<DisplayDevice> hw =
                                new DisplayDevice(this, state.type, hwcId, state.isSecure, display,
                                                  dispSurface, producer, hasWideColorDisplay);
                        hw->setLayerStack(state.layerStack);
                        hw->setProjection(state.orientation,
                                state.viewport, state.frame);
                        hw->setDisplayName(state.displayName);
                        mDisplays.add(display, hw);
                        if (!state.isVirtualDisplay()) {
                            mEventThread->onHotplugReceived(state.type, true);
                        }
                    }
                }
            }
        }
    }
            ... ...

添加屏幕时,根据前面已经创建的BufferQueue,创建对应的DisplaySurface,外显和虚显的不一样。创建DisplaySurface后,再创建DisplayDevice;设置DisplayDevice的stack,投影矩阵Projection;将创建的DisplayDevice添加到mDisplays中;最后,对外显,调用EventThread的onHotplugReceived。

EventThread的onHotplugReceived函数中,将封装一个hotplug的Event事件DISPLAY_EVENT_HOTPLUG,EventThread再将事件分发出去。这里对hotplug感兴趣的主要就是框架层了,回调给DisplayManagerService。

hotplug的流程

这里主要将了Display相关的逻辑,主要是热插拔的处理。下面是整个Android系统添加屏幕时的处理流程。


Display热插拔的流程

Display相关的就介绍到这里,后续讲合成时,还会有很多相关的流程。

推荐阅读更多精彩内容