Second Patch for async RenderEngine
- pass a vector of <LayerSettings> instead of a vector of pointers to
drawLayers function to ensure lifecycle safety.
- capture all local variables to call drawLayerInternals in the render
engine thread to avoid of pointers being invalidated if it takes long
time to pop mFunctions calls.
- change renderScreenImplLocked return type as a shared_future
object to unblock SF main thread for screen capture events.
- block region sampling thread only when SF main thread hasn't
completed capture screen event.
Bug: 180657548
Test: SurfaceFlinger_test, android.hardware.graphics.composer@2.2-vts, libcompositionengine_test, librenderengine_test, libsurfaceflinger_unittest pass
Change-Id: I615f2927d30524988fb12df22fe331e7217c3058
diff --git a/services/surfaceflinger/BufferQueueLayer.cpp b/services/surfaceflinger/BufferQueueLayer.cpp
index f98681e..6c60e53 100644
--- a/services/surfaceflinger/BufferQueueLayer.cpp
+++ b/services/surfaceflinger/BufferQueueLayer.cpp
@@ -48,7 +48,9 @@
// Interface implementation for Layer
// -----------------------------------------------------------------------
-void BufferQueueLayer::onLayerDisplayed(const sp<Fence>& releaseFence) {
+void BufferQueueLayer::onLayerDisplayed(
+ std::shared_future<renderengine::RenderEngineResult> futureRenderEngineResult) {
+ sp<Fence> releaseFence = new Fence(dup(futureRenderEngineResult.get().drawFence));
mConsumer->setReleaseFence(releaseFence);
// Prevent tracing the same release multiple times.