在上一篇文章Android SurfaceFlinger服务代理对象获取过程源码分析中介绍了应用程序获取SurfaceFlinger中的各种远程Binder代理对象,SurfaceFlinger为每个应用程序创建Client对象来处理应用程序的Binder请求,同时返回Client的Binder代理对象给应用程序,这样应用程序就可以通过该Binder代理对象来请求SurfaceFlinger创建Surface了,本文在前面章节的基础上 分析Surface的创建过程。
SurfaceComposerClient mSession = new SurfaceComposerClient();sp<SurfaceControl> control= mSession->createSurface(0, dinfo.h, dinfo.w, PIXEL_FORMAT_RGB_565);首先在应用程序端构造SurfaceComposerClient对象,在构造过程中,通过ComposerService类查询到SurfaceFlinger服务的远程代理对象,并通过SurfaceFlinger的远程Binder代理对象连接SurfaceFlinger,获取SurfaceFlinger为应用程序创建的Client的远程代理对象,关于应用程序查询SurfaceFlinger服务代理对象,应用程序连接SurfaceFlinger的过程请查看Android SurfaceFlinger服务代理对象获取过程源码分析。得到SurfaceComposerClient对象后,就可以调用该对象的createSurface函数来创建Surface了,同时得到一个SurfaceControl对象。
sp<SurfaceControl> SurfaceComposerClient::createSurface( DisplayID display, //显示屏ID uint32_t w, //图像宽度 uint32_t h, //图像高度 PixelFormat format,//图形格式 uint32_t flags)//创建的Surface类型{ String8 name; const size_t SIZE = 128; char buffer[SIZE]; snprintf(buffer, SIZE, "<pid_%d>", getpid());//为当前创建的Surface格式化名称:<pid_%d> name.append(buffer); return SurfaceComposerClient::createSurface(name, display,w, h, format, flags);}将当前进程的ID号格式化为<pid_%d>的字符串来命名当前创建的Surface,然后调用SurfaceComposerClient的另一个重载函数createSurface来创建Surface
sp<SurfaceControl> SurfaceComposerClient::createSurface( const String8& name, DisplayID display, uint32_t w, uint32_t h, PixelFormat format, uint32_t flags){ sp<SurfaceControl> result; if (mStatus == NO_ERROR) { ISurfaceComposerClient::surface_data_t data; //mClient为服务端Client的远程代理对象BpSurfaceComposerClient,这里请求服务端Client创建Surface sp<ISurface> surface = mClient->createSurface(&data, name,display, w, h, format, flags); if (surface != 0) { //根据服务端返回来的ISurface,surface_data_t及当前SurfaceComposerClient对象来构造SurfaceControl对象 result = new SurfaceControl(this, surface, data); } } return result;}在Android SurfaceFlinger服务代理对象获取过程源码分析中我们知道,应用程序获得Client的远程代理对象BpSurfaceComposerClient后,保存到成员变量mClient中,这里就是通过Client的远程代理对象BpSurfaceComposerClient向服务端的Client请求创建Surface,并得到BSurface的远程Binder代理对象BpSurface,接着利用得到的BpSurface对象及SurfaceComposerClient对象为应用程序创建一个SurfaceControl对象。ISurface也是基于Binder进程通信框架设计的,其在Binder通信框架中的类关系图如下:
由于ISurfaceComposerClient也是基于Binder通信框架的,因此BpSurfaceComposerClient向服务端的Client请求创建Surface的过程分为客户端进程和服务端进程。
客户端进程:
frameworks\native\libs\gui\ISurfaceComposerClient.cpp
virtual sp<ISurface> createSurface( surface_data_t* params, const String8& name, DisplayID display, uint32_t w, uint32_t h, PixelFormat format, uint32_t flags){ Parcel data, reply; data.writeInterfaceToken(ISurfaceComposerClient::getInterfaceDescriptor()); data.writeString8(name); data.writeInt32(display); data.writeInt32(w); data.writeInt32(h); data.writeInt32(format); data.writeInt32(flags); remote()->transact(CREATE_SURFACE, data, &reply); params->readFromParcel(reply); return interface_cast<ISurface>(reply.readStrongBinder());}服务端进程:
frameworks\native\libs\gui\ISurfaceComposerClient.cpp
status_t BnSurfaceComposerClient::onTransact( uint32_t code, const Parcel& data, Parcel* reply, uint32_t flags){ switch(code) { case CREATE_SURFACE: { CHECK_INTERFACE(ISurfaceComposerClient, data, reply); surface_data_t params; String8 name = data.readString8(); DisplayID display = data.readInt32(); uint32_t w = data.readInt32(); uint32_t h = data.readInt32(); PixelFormat format = data.readInt32(); uint32_t flags = data.readInt32(); sp<ISurface> s = createSurface(?ms, name, display, w, h,format, flags); params.writeToParcel(reply); reply->writeStrongBinder(s->asBinder()); return NO_ERROR; } break; default: return BBinder::onTransact(code, data, reply, flags); }}createSurface函数在BnSurfaceComposerClient的子类Client中实现
frameworks\native\services\surfaceflinger\SurfaceFlinger.cpp
sp<ISurface> Client::createSurface( ISurfaceComposerClient::surface_data_t* params, const String8& name, DisplayID display, uint32_t w, uint32_t h, PixelFormat format, uint32_t flags){ //构造一个MessageCreateSurface消息 sp<MessageBase> msg = new MessageCreateSurface(mFlinger.get(),params, name, this, display, w, h, format, flags); //将消息投递到SurfaceFlinger的mEventQueue队列中,并睡眠等待消息处理 mFlinger->postMessageSync(msg); //当消息得到处理后,线程唤醒,读取处理结果 return static_cast<MessageCreateSurface*>( msg.get() )->getResult();}关于SurfaceFlinger的消息队列这里不做分析。MessageCreateSurface消息被投递到SurfaceFlinger的mEventQueue队列中后,MessageCreateSurface类的handler函数被调用,用来处理MessageCreateSurface消息。
virtual bool handler() { result = flinger->createSurface(params, name, client,display, w, h, format, flags); return true;}这里又调用SurfaceFlinger的createSurface函数来创建Surface
sp<ISurface> SurfaceFlinger::createSurface( ISurfaceComposerClient::surface_data_t* params, const String8& name, const sp<Client>& client, DisplayID d, uint32_t w, uint32_t h, PixelFormat format, uint32_t flags){ sp<LayerBaseClient> layer; sp<ISurface> surfaceHandle; //图像的宽高必须大于0 if (int32_t(w|h) < 0) { ALOGE("createSurface() failed, w or h is negative (w=%d, h=%d)",int(w), int(h)); return surfaceHandle; } //根据flag来判断请求创建的Surface类型,实际创建的是LayerBaseClient对象 sp<Layer> normalLayer; switch (flags & eFXSurfaceMask) { case eFXSurfaceNormal: normalLayer = createNormalSurface(client, d, w, h, flags, format); layer = normalLayer; break; case eFXSurfaceBlur: // for now we treat Blur as Dim, until we can implement it // efficiently. case eFXSurfaceDim: layer = createDimSurface(client, d, w, h, flags); break; case eFXSurfaceScreenshot: layer = createScreenshotSurface(client, d, w, h, flags); break; } if (layer != 0) { //初始化LayerBaseClient对象 layer->initStates(w, h, flags); layer->setName(name); ssize_t token = addClientLayer(client, layer); //从LayerBaseClient对象中取出BSurface本地对象 surfaceHandle = layer->getSurface(); if (surfaceHandle != 0) { //初始化参数params params->token = token; params->identity = layer->getIdentity(); if (normalLayer != 0) { Mutex::Autolock _l(mStateLock); mLayerMap.add(layer->getSurfaceBinder(), normalLayer); } } setTransactionFlags(eTransactionNeeded); } return surfaceHandle;}该函数首先根据应用程序发送过来的参数构造LayerBaseClient对象,并初始化该对象,最后从LayerBaseClient对象中取得BSurface本地对象,然后将BSurface对象的远程Binder代理对象返回给应用程序。
sp<Layer> SurfaceFlinger::createNormalSurface( const sp<Client>& client, DisplayID display, uint32_t w, uint32_t h, uint32_t flags, PixelFormat& format){ // initialize the surfaces switch (format) { // TODO: take h/w into account case PIXEL_FORMAT_TRANSPARENT: case PIXEL_FORMAT_TRANSLUCENT: format = PIXEL_FORMAT_RGBA_8888; break; case PIXEL_FORMAT_OPAQUE:#ifdef NO_RGBX_8888 format = PIXEL_FORMAT_RGB_565;#else format = PIXEL_FORMAT_RGBX_8888;#endif break; }#ifdef NO_RGBX_8888 if (format == PIXEL_FORMAT_RGBX_8888) format = PIXEL_FORMAT_RGBA_8888;#endif //构造Layer对象 sp<Layer> layer = new Layer(this, display, client); status_t err = layer->setBuffers(w, h, format, flags); if (CC_LIKELY(err != NO_ERROR)) { ALOGE("createNormalSurfaceLocked() failed (%s)", strerror(-err)); layer.clear(); } return layer;}该函数首先判断图像格式,然后构造一个Layer对象,接着为该Layer对象设置缓冲区buffer,上图显示了Layer类的继承关系,在构造Layer对象时,其父类的构造函数会被依次调用。
Layer::Layer(SurfaceFlinger* flinger, DisplayID display, const sp<Client>& client) : LayerBaseClient(flinger, display, client), mTextureName(-1U), mQueuedFrames(0), mCurrentTransform(0), mCurrentScalingMode(NATIVE_WINDOW_SCALING_MODE_FREEZE), mCurrentOpacity(true), mRefreshPending(false), mFrameLatencyNeeded(false), mFrameLatencyOffset(0), mFormat(PIXEL_FORMAT_NONE), mGLExtensions(GLExtensions::getInstance()), mOpaqueLayer(true), mNeedsDithering(false), mSecure(false), mProtectedByApp(false){ mCurrentCrop.makeInvalid(); glGenTextures(1, &mTextureName);}Layer的父类LayerBaseClient构造过程:
LayerBaseClient::LayerBaseClient(SurfaceFlinger* flinger, DisplayID display, const sp<Client>& client) : LayerBase(flinger, display), mHasSurface(false), mClientRef(client), mIdentity(uint32_t(android_atomic_inc(&sIdentity))){}LayerBaseClient的父类LayerBase的构造过程:
LayerBase::LayerBase(SurfaceFlinger* flinger, DisplayID display) : dpy(display), contentDirty(false), sequence(uint32_t(android_atomic_inc(&sSequence))), mFlinger(flinger), mFiltering(false), mNeedsFiltering(false), mOrientation(0), mPlaneOrientation(0), mTransactionFlags(0), mPremultipliedAlpha(true), mName("unnamed"), mDebug(false){ const DisplayHardware& hw(flinger->graphicPlane(0).displayHardware()); mFlags = hw.getFlags();}由于Layer继承于RefBase类,同时实现了该类的onFirstRef函数,在前面就介绍了,当第一次强引用RefBase的子类对象,并且该子类对象实现了onFirstRef函数,onFirstRef函数会被自动调用,Layer对象也不例外:
void Layer::onFirstRef(){ LayerBaseClient::onFirstRef(); // Creates a custom BufferQueue for SurfaceTexture to use sp<BufferQueue> bq = new SurfaceTextureLayer(); mSurfaceTexture = new SurfaceTexture(mTextureName, true,GL_TEXTURE_EXTERNAL_OES, false, bq); mSurfaceTexture->setConsumerUsageBits(getEffectiveUsage(0)); mSurfaceTexture->setFrameAvailableListener(new FrameQueuedListener(this)); mSurfaceTexture->setSynchronousMode(true);#ifdef TARGET_DISABLE_TRIPLE_BUFFERING#warning "disabling triple buffering" mSurfaceTexture->setBufferCountServer(2);#else mSurfaceTexture->setBufferCountServer(3);#endif}
SurfaceTextureLayer也是基于Binder通信框架设计的:
这里就是为当前创建的Layer构造并初始化SurfaceTexture对象,该SurfaceTexture对象用来管理BufferQueue。回到SurfaceFlinger的创建 普通Surface函数createNormalSurface,构造完Layer对象后,还需根据图像大小,格式来设置Layer对象中的buffer
status_t Layer::setBuffers( uint32_t w, uint32_t h, PixelFormat format, uint32_t flags){ // this surfaces pixel format PixelFormatInfo info; status_t err = getPixelFormatInfo(format, &info); if (err) { ALOGE("unsupported pixelformat %d", format); return err; } // the display's pixel format const DisplayHardware& hw(graphicPlane(0).displayHardware()); uint32_t const maxSurfaceDims = min(hw.getMaxTextureSize(), hw.getMaxViewportDims()); // never allow a surface larger than what our underlying GL implementation // can handle. if ((uint32_t(w)>maxSurfaceDims) || (uint32_t(h)>maxSurfaceDims)) { ALOGE("dimensions too large %u x %u", uint32_t(w), uint32_t(h)); return BAD_VALUE; } PixelFormatInfo displayInfo; getPixelFormatInfo(hw.getFormat(), &displayInfo); const uint32_t hwFlags = hw.getFlags(); mFormat = format; mSecure = (flags & ISurfaceComposer::eSecure) ? true : false; mProtectedByApp = (flags & ISurfaceComposer::eProtectedByApp) ? true : false; mOpaqueLayer = (flags & ISurfaceComposer::eOpaque); mCurrentOpacity = getOpacityForFormat(format); mSurfaceTexture->setDefaultBufferSize(w, h); mSurfaceTexture->setDefaultBufferFormat(format); mSurfaceTexture->setConsumerUsageBits(getEffectiveUsage(0)); // we use the red index int displayRedSize = displayInfo.getSize(PixelFormatInfo::INDEX_RED); int layerRedsize = info.getSize(PixelFormatInfo::INDEX_RED); mNeedsDithering = layerRedsize > displayRedSize; return NO_ERROR;}最后返回到SurfaceFlinger的createSurface函数设置Layer的当前状态及名称,同时将当前创建的Layer添加到Client的窗口列表及SurfaceFlinger维护的Z秩序列表中
layer->initStates(w, h, flags);layer->setName(name);ssize_t token = addClientLayer(client, layer);将当前Layer添加的Client及SurfaceFlinger维护的列表中
ssize_t SurfaceFlinger::addClientLayer(const sp<Client>& client,const sp<LayerBaseClient>& lbc){ // attach this layer to the client size_t name = client->attachLayer(lbc); Mutex::Autolock _l(mStateLock); // add this layer to the current state list addLayer_l(lbc); return ssize_t(name);}函数调用attachLayer将当前Layer添加到当前应用程序在SurfaceFlinger中的Client的mLayers列表中
DefaultKeyedVector< size_t, wp<LayerBaseClient> > mLayers;size_t Client::attachLayer(const sp<LayerBaseClient>& layer){ Mutex::Autolock _l(mLock); size_t name = mNameGenerator++; mLayers.add(name, layer); return name;}接着调用函数addLayer_l将当前创建的Layer添加到SurfaceFlinger中
status_t SurfaceFlinger::addLayer_l(const sp<LayerBase>& layer){ ssize_t i = mCurrentState.layersSortedByZ.add(layer); return (i < 0) ? status_t(i) : status_t(NO_ERROR);}
sp<ISurface> LayerBaseClient::getSurface(){ sp<ISurface> s; Mutex::Autolock _l(mLock); LOG_ALWAYS_FATAL_IF(mHasSurface,"LayerBaseClient::getSurface() has already been called"); mHasSurface = true; s = createSurface(); mClientSurfaceBinder = s->asBinder(); return s;}createSurface()用于创建一个BSurface对象,同时将创建的BSurface的Binder本地对象保存到LayerBaseClient的父类LayerBase的成员变量mClientSurfaceBinder中。
sp<ISurface> LayerBaseClient::createSurface(){ sp<ISurface> sur(new BSurface(mFlinger, this)); return sur;}层层返回到SurfaceFlinger的createSurface函数,该函数最终返回一个BSurface对象,最后在BnSurfaceComposerClient的onTransact函数中通过writeStrongBinder(s->asBinder())将BSurface的Binder本地对象写入到Binder驱动中,应用程序端BpSurfaceComposerClient通过reply.readStrongBinder()从Binder驱动中读取BSurface的Binder远程代理对象,并通过interface_cast<ISurface>创建BSurface的远程代理对象BpSurface,到此Surface就创建完成了,现在来总结一下Surface的整个创建过程:
1.应用程序进程通过SurfaceComposerClient对象向SurfaceFlinger请求创建Surface;
2.SurfaceComposerClient对象借助Client的远程代理对象BpSurfaceComposerClient向Client请求创建Surface;
3.Client接收到BpSurfaceComposerClient的CREATE_SURFACE请求后,向SurfaceFlinger的事件队列发送MessageCreateSurface消息;
4.MessageCreateSurface消息处理过程中,使用SurfaceFlinger来创建Surface;
5.SurfaceFlinger根据参数创建并初始化Layer对象;
6.创建BSurface对象,并将该对象的Binder远程对象返回给应用程序进程。