Android 用户进程绘制过程.doc

Android 用户进程绘制过程下面我们来看下这个函数，它先调用了Surface的lockCanvas获取一个Canvas，然后再调用surface.unlockCanvasAndPost来表示绘制结束。cpp view plain copy 在CODE上查看代码片派生到我的代码片private boolean drawSoftware(Surface surface, AttachInfo attachInfo, int xoff, int yoff, boolean scalingRequired, Rect dirty) / Draw with software renderer. final Canvas canvas; try final int left = dirty.left; final int top = dirty.top; final int right = dirty.right; final int bottom = dirty.bottom; canvas = mSurface.lockCanvas(dirty);/获取一个Canvas . finally try surface.unlockCanvasAndPost(canvas);/绘制结束 catch (IllegalArgumentException e) Log.e(TAG, Could not unlock surface, e); mLayoutRequested = true; / ask wm for a new surface next time. /noinspection ReturnInsideFinallyBlock return false; if (LOCAL_LOGV) Log.v(TAG, Surface + surface + unlockCanvasAndPost); return true; 一、应用进程申请buffer我们先来看Surface的lockCanvas函数，其调用了native函数nativeLockCanvascpp view plain copy 在CODE上查看代码片派生到我的代码片public Canvas lockCanvas(Rect inOutDirty) throws Surface.OutOfResourcesException, IllegalArgumentException synchronized (mLock) checkNotReleasedLocked(); if (mLockedObject != 0) / Ideally, nativeLockCanvas() would throw in this situation and prevent the / double-lock, but that wont happen if mNativeObject was updated. We cant / abandon the old mLockedObject because it might still be in use, so instead / we just refuse to re-lock the Surface. throw new IllegalArgumentException(Surface was already locked); mLockedObject = nativeLockCanvas(mNativeObject, mCanvas, inOutDirty); return mCanvas; 这个函数在android_view_Surface.cpp中，这个函数内容很多，我们先调用了Surface的lock函数来申请buffer，然后新建了一个SkBitmap，设置了内存地址，并且把这个bitmap放入了Canvas中。cpp view plain copy 在CODE上查看代码片派生到我的代码片static jlong nativeLockCanvas(JNIEnv* env, jclass clazz, jlong nativeObject, jobject canvasObj, jobject dirtyRectObj) sp surface(reinterpret_cast(nativeObject); if (!isSurfaceValid(surface) doThrowIAE(env); return 0; Rect dirtyRect; Rect* dirtyRectPtr = NULL; if (dirtyRectObj) dirtyRect.left = env-GetIntField(dirtyRectObj, gRectClassInfo.left); dirtyRect.top = env-GetIntField(dirtyRectObj, gRectClassInfo.top); dirtyRect.right = env-GetIntField(dirtyRectObj, gRectClassInfo.right); dirtyRect.bottom = env-GetIntField(dirtyRectObj, gRectClassInfo.bottom); dirtyRectPtr = &dirtyRect; ANativeWindow_Buffer outBuffer; status_t err = surface-lock(&outBuffer, dirtyRectPtr);/从SurfaceFlinger中申请内存buffer if (err 0 & outBuffer.height 0) bitmap.setPixels(outBuffer.bits);/bitmap设置其内存地址 else / be safe with an empty bitmap. bitmap.setPixels(NULL); Canvas* nativeCanvas = GraphicsJNI:getNativeCanvas(env, canvasObj); nativeCanvas-setBitmap(bitmap);/设置Canvas的bitmap if (dirtyRectPtr) nativeCanvas-clipRect(dirtyRect.left, dirtyRect.top, dirtyRect.right, dirtyRect.bottom); if (dirtyRectObj) env-SetIntField(dirtyRectObj, gRectClassInfo.left, dirtyRect.left); env-SetIntField(dirtyRectObj, gRectClassInfo.top, dirtyRect.top); env-SetIntField(dirtyRectObj, gRectClassInfo.right, dirtyRect.right); env-SetIntField(dirtyRectObj, gRectClassInfo.bottom, dirtyRect.bottom); / Create another reference to the surface and return it. This reference / should be passed to nativeUnlockCanvasAndPost in place of mNativeObject, / because the latter could be replaced while the surface is locked. sp lockedSurface(surface); lockedSurface-incStrong(&sRefBaseOwner); return (jlong) lockedSurface.get(); 我们先来看下Surface的lock函数，先是调用了dequeueBuffer来申请内存，然后放入backBuffer。后面调用GraphicBuffer的lockAsync来把Buffer中的handle的地址放到vaddr中，最后把vaddr放到outBuffer的bits中。cpp view plain copy 在CODE上查看代码片派生到我的代码片status_t Surface:lock( ANativeWindow_Buffer* outBuffer, ARect* inOutDirtyBounds) . ANativeWindowBuffer* out; int fenceFd = -1; status_t err = dequeueBuffer(&out, &fenceFd);/申请内存 if (err = NO_ERROR) sp backBuffer(GraphicBuffer:getSelf(out);/放到backBuffer中 const Rect bounds(backBuffer-width, backBuffer-height); . void* vaddr; status_t res = backBuffer-lockAsync(/这个函数就是把buffer的handle中的地址传到vaddr中 GRALLOC_USAGE_SW_READ_OFTEN | GRALLOC_USAGE_SW_WRITE_OFTEN, newDirtyRegion.bounds(), &vaddr, fenceFd); ALOGW_IF(res, failed locking buffer (handle = %p), backBuffer-handle); if (res != 0) err = INVALID_OPERATION; else mLockedBuffer = backBuffer; outBuffer-width = backBuffer-width; outBuffer-height = backBuffer-height; outBuffer-stride = backBuffer-stride; outBuffer-format = backBuffer-format; outBuffer-bits = vaddr;/buffer地址 return err; 1.1 申请bufferSurface的dequeueBuffer函数主要就是从SurfaceFlinger中申请buffer。cpp view plain copy 在CODE上查看代码片派生到我的代码片int Surface:dequeueBuffer(android_native_buffer_t* buffer, int* fenceFd) ATRACE_CALL(); ALOGV(Surface:dequeueBuffer); uint32_t reqWidth; uint32_t reqHeight; bool swapIntervalZero; PixelFormat reqFormat; uint32_t reqUsage; . int buf = -1; sp fence; status_t result = mGraphicBufferProducer-dequeueBuffer(&buf, &fence, swapIntervalZero,/在SurfaceFlinger中申请内存，只要返回mSlots中的序号 reqWidth, reqHeight, reqFormat, reqUsage); Mutex:Autolock lock(mMutex); sp& gbuf(mSlotsbuf.buffer); if (result & IGraphicBufferProducer:RELEASE_ALL_BUFFERS) freeAllBuffers(); if (result & IGraphicBufferProducer:BUFFER_NEEDS_REALLOCATION) | gbuf = 0) result = mGraphicBufferProducer-requestBuffer(buf, &gbuf);/根据需要拿到buffer . . *buffer = gbuf.get(); return OK; 下面我们再看下SurfaceFlinger中对应的dequeueBuffer和requestBuffer函数。我们先看BufferQueueProducer.cpp中dequeueBuffer关于分配buffer的一段代码：cpp view plain copy 在CODE上查看代码片派生到我的代码片. if (returnFlags & BUFFER_NEEDS_REALLOCATION) /需要分配内存 status_t error; BQ_LOGV(dequeueBuffer: allocating a new buffer for slot %d, *outSlot); sp graphicBuffer(mCore-mAllocator-createGraphicBuffer( width, height, format, usage, &error); if (graphicBuffer = NULL) BQ_LOGE(dequeueBuffer: createGraphicBuffer failed); return error; / Autolock scope Mutex:Autolock lock(mCore-mMutex); if (mCore-mIsAbandoned) BQ_LOGE(dequeueBuffer: BufferQueue has been abandoned); return NO_INIT; graphicBuffer-setGenerationNumber(mCore-mGenerationNumber); mSlots*outSlot.mGraphicBuffer = graphicBuffer;/将分配的内存放到mSlots中，outSlot就是给应用进程mSlots的序号 / Autolock scope . return returnFlags; 然后我们再来看其requestBuffer函数，就是根据序号，从mSlots拿到buffer。cpp view plain copy 在CODE上查看代码片派生到我的代码片status_t BufferQueueProducer:requestBuffer(int slot, sp* buf) ATRACE_CALL(); BQ_LOGV(requestBuffer: slot %d, slot); Mutex:Autolock lock(mCore-mMutex); if (mCore-mIsAbandoned) BQ_LOGE(requestBuffer: BufferQueue has been abandoned); return NO_INIT; if (slot = BufferQueueDefs:NUM_BUFFER_SLOTS) BQ_LOGE(requestBuffer: slot index %d out of range 0, %d), slot, BufferQueueDefs:NUM_BUFFER_SLOTS); return BAD_VALUE; else if (mSlotsslot.mBufferState != BufferSlot:DEQUEUED) BQ_LOGE(requestBuffer: slot %d is not owned by the producer (state = %d), slot, mSlotsslot.mBufferState); return BAD_VALUE; mSlotsslot.mRequestBufferCalled = true; *buf = mSlotsslot.mGraphicBuffer; return NO_ERROR; 1.2 将buffer的地址放到vaddr中我们继续分析Surface的lock函数，来看下GraphicBuffer的lockAsyc，其调用了getBufferMapper().lockAsync，而这个getBufferMapper返回的是GraphicBufferMappercpp view plain copy 在CODE上查看代码片派生到我的代码片status_t GraphicBuffer:lockAsync(uint32_t inUsage, const Rect& rect, void* vaddr, int fenceFd) if (rect.left width | rect.top height) ALOGE(locking pixels (%d,%d,%d,%d) outside of buffer (w=%d, h=%d), rect.left, rect.top, rect.right, rect.bottom, width, height); return BAD_VALUE; status_t res = getBufferMapper().lockAsync(handle, inUsage, rect, vaddr, fenceFd); return res; 这里的GraphicBufferMapper:lockAsync调用的是mAllocMod的lock函数cpp view plain copy 在CODE上查看代码片派生到我的代码片status_t GraphicBufferMapper:lockAsync(buffer_handle_t handle, uint32_t usage, const Rect& bounds, void* vaddr, int fenceFd) ATRACE_CALL(); status_t err; if (mAllocMod-common.module_api_version = GRALLOC_MODULE_API_VERSION_0_3) err = mAllocMod-lockAsync(mAllocMod, handle, static_cast(usage), bounds.left, bounds.top, bounds.width(), bounds.height(), vaddr, fenceFd); else if (fenceFd = 0) sync_wait(fenceFd, -1); close(fenceFd); err = mAllocMod-lock(mAllocMod, handle, static_cast(usage), bounds.left, bounds.top, bounds.width(), bounds.height(), vaddr); ALOGW_IF(err, lockAsync(.) failed %d (%s), err, strerror(-err); return err; mAllocMod也是Gralloc模块。cpp view plain copy 在CODE上查看代码片派生到我的代码片GraphicBufferMapper:GraphicBufferMapper() : mAllocMod(0) hw_module_t const* module; int err = hw_get_module(GRALLOC_HARDWARE_MODULE_ID, &module); ALOGE_IF(err, FATAL: cant find the %s module, GRALLOC_HARDWARE_MODULE_ID); if (err = 0) mAllocMod = reinterpret_cast(module); 最后我们来看下Gralloc模块的lock函数，其实也就是将handle的base（buffer地址）放到vaddr中。cpp view plain copy 在CODE上查看代码片派生到我的代码片static int gralloc_lock(gralloc_module_t const* module, buffer_handle_t handle, int usage, int l, int t, int w, int h, void* vaddr) if (private_handle_t:validate(handle) flags & private_handle_t:PRIV_FLAGS_USES_UMP | hnd-flags & private_handle_t:PRIV_FLAGS_USES_ION) hnd-writeOwner = usage & GRALLOC_USAGE_SW_WRITE_MASK; if (usage & (GRALLOC_USAGE_SW_READ_MASK | GRALLOC_USAGE_SW_WRITE_MASK) *vaddr = (void*)hnd-base;/将handle的base（buffer地址）放到vaddr中 return 0; 1.3 Canvas的获取这样我们再回过头来看nativeLockCanvas函数，buffer也申请了，buffer地址也放到了Canvas的SkBitmap中了（意味着写到canvas，可以写到buffer的地址中了）。最后我们再来看Canvas的获取。在nativeLockCanvas中通过GraphicsJNI:getNativeCanvas来获取到canvasObj，这个就是上层在ViewRootImpl的Canvas的对象。cpp view plain copy 在CODE上查看代码片派生到我的代码片Canvas* nativeCanvas = GraphicsJNI:getNativeCanvas(env, canvasObj); 我们来看下这个函数，这个就是获取Java层。cpp view plain copy 在CODE上查看代码片派生到我的代码片android:Canvas* GraphicsJNI:getNativeCanvas(JNIEnv* env, jobject canvas) SkASSERT(env); SkASSERT(canvas); SkASSERT(env-IsInstanceOf(canvas, gCanvas_class); jlong canvasHandle = env-GetLongField(canvas, gCanvas_nativeInstanceID); if (!canvasHandle) return NULL; return reinterpret_cast(canvasHandle); 最后通过如下没我们发现是Canvas的mNativeCanvasWrapper对象。cpp view plain copy 在CODE上查看代码片派生到我的代码片gCanvas_class = make_globalref(env, android/graphics/Canvas); gCanvas_nativeInstanceID = getFieldIDCheck(env, gCanvas_class, mNativeCanvasWrapper, J); 而在Canvas中如果是通过mNativeCanvasWrapper复制Canvas对象，使用如下构造函数。cpp view plain copy 在CODE上查看代码片派生到我的代码片public Canvas(long nativeCanvas) if (nativeCanvas = 0) throw new IllegalStateException(); mNativeCanvasWrapper = nativeCanvas; mFinalizer = new CanvasFinalizer(mNativeCanvasWrapper); mDensity = Bitmap.getDefaultDensity(); 那么必然会有一个地方创建了Canvas对象，最后我们发现会在View中创建Canvas.cpp view plain copy 在CODE上查看代码片派生到我的代码片. Canvas canvas; if (attachInfo != null) canvas = attachInfo.mCanvas; if (canvas = null) canvas = new Canvas(); canvas.setBitmap(bitmap); . 最后都是通过SkBitmap来写数据到buffer中去。二、应用进程绘制绘制的话，我们举个简单的例子canvas.drawRectcpp view plain copy 在CODE上查看代码片派生到我的代码片public void drawRect(NonNull RectF rect, NonNull Paint paint) native_drawRect(mNativeCanvasWrapper, rect.left, rect.top, rect.right, rect.bottom, paint.getNativeInstance(); 我们来看下native_drawRect函数cpp view plain copy 在CODE上查看代码片派生到我的代码片static void drawRect(JNIEnv* env, jobject, jlong canvasHandle, jfloat left, jfloat top, jfloat right, jfloat bottom, jlong paintHandle) const Paint* paint = reinterpret_cast(paintHandle); get_canvas(canvasHandle)-drawRect(left, top, right, bottom, *paint); 最后是调用了SkiaCanvas的drawRectCoordscpp view plain copy 在CODE上查看代码片派生到我的代码片void SkiaCanvas:drawRect(float left, float top, float right, float bottom, const SkPaint& int) mCanvas-drawRectCoords(left, top, right, bottom, paint); SkCanvas的drawRectCoordscpp view plain copy 在CODE上查看代码片派生到我的代码片void SkCanvas:drawRectCoords(SkScalar left, SkScalar top, SkScalar right, SkScalar bottom, const SkPaint& paint) TRACE_EVENT0(disabled-by-default-skia, SkCanvas:drawRectCoords(); SkRect r; r.set(left, top, right, bottom); this-drawRect(r, paint); SkCanvas:DrawRect函数，这里的fDevice就是SkBitmap。cpp view plain copy 在CODE上查看代码片派生到我的代码片void SkCanvas:DrawRect(const SkDraw& draw, const SkPaint& paint, const SkRect& r, SkScalar textSize) if (paint.getStyle() = SkPaint:kFill_Style) draw.fDevice-drawRect(draw, r, paint); else SkPaint p(paint); p.setStrokeWidth(SkScalarMul(textSize, paint.getStrokeWidth(); draw.fDevice-drawRect(draw, r, p); SkBitmap之前我们把buffer的地址传进去了，这样就可以将数据写入buffer了。三、绘制完成在ViewRootImpl中绘制完成后，最后会调用surface.unlockCanvasAndPost(canvas)cpp view plain copy 在CODE上查看代码片派生到我的代码片public void unlockCanvasAndPost(Canvas canvas) ed (mLock) checkNotReleasedLocked(); if (mHwuiContext != null) mHwuiContext.unlockAndPost(canvas); else unlockSwCanvasAndPost(canvas); 正常是调用unlockSwCanvasAndPost函数，这个函数主要是调用了nativeUnlockCanvasAndPost native函数。这里仙剑SkBitmap重新设置了一个空的，然后调用了surface的unlockAndPost函数cpp view plain copy 在CODE上查看代码片派生到我的代码片static void nativeUnlockCanvasAndPost(JNIEnv* env, jclass clazz, jlong nativeObject, jobjec