replicant-frameworks_native/services/surfaceflinger/DisplayHardware/VirtualDisplaySurface.h
Jesse Hall 7414965606 Release virtual display buffer immediately after HWC set
Previously we only queued a virtual display buffer to the sink when
the next frame was about to be displayed. This may delay the "last"
frame of an animation indefinitely. Now we queue the buffer as soon as
HWC set() returns and gives us the release fence.

Bug: 8384764
Change-Id: I3844a188e0f6ef6ff28f3e11477cfa063a924b1a
2013-03-20 11:16:55 -07:00

84 lines
3.2 KiB
C++

/*
* Copyright 2013 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef ANDROID_SF_VIRTUAL_DISPLAY_SURFACE_H
#define ANDROID_SF_VIRTUAL_DISPLAY_SURFACE_H
#include "BufferQueueInterposer.h"
#include "DisplaySurface.h"
// ---------------------------------------------------------------------------
namespace android {
// ---------------------------------------------------------------------------
class HWComposer;
/* This DisplaySurface implementation uses a BufferQueueInterposer to pass
* partially- or fully-composited buffers from the OpenGL ES driver to
* HWComposer to use as the output buffer for virtual displays. Allowing HWC
* to compose into the same buffer that contains GLES results saves bandwidth
* compared to having two separate BufferQueues for frames with at least some
* GLES composition.
*
* The alternative would be to have two complete BufferQueues, one from GLES
* to HWC and one from HWC to the virtual display sink (e.g. video encoder).
* For GLES-only frames, the same bandwidth saving could be achieved if buffers
* could be acquired from the GLES->HWC queue and inserted into the HWC->sink
* queue. That would be complicated and doesn't help the mixed GLES+HWC case.
*
* On frames with no GLES composition, the VirtualDisplaySurface dequeues a
* buffer directly from the sink IGraphicBufferProducer and passes it to HWC,
* bypassing the GLES driver. This is only guaranteed to work if
* eglSwapBuffers doesn't immediately dequeue a buffer for the next frame,
* since we can't rely on being able to dequeue more than one buffer at a time.
*
* TODO(jessehall): Add a libgui test that ensures that EGL/GLES do lazy
* dequeBuffers; we've wanted to require that for other reasons anyway.
*/
class VirtualDisplaySurface : public DisplaySurface {
public:
VirtualDisplaySurface(HWComposer& hwc, int disp,
const sp<IGraphicBufferProducer>& sink,
const String8& name);
virtual sp<IGraphicBufferProducer> getIGraphicBufferProducer() const;
virtual status_t compositionComplete();
virtual status_t advanceFrame();
virtual void onFrameCommitted(int fenceFd);
virtual void dump(String8& result) const;
private:
virtual ~VirtualDisplaySurface();
// immutable after construction
HWComposer& mHwc;
int mDisplayId;
sp<BufferQueueInterposer> mSource;
String8 mName;
// mutable, must be synchronized with mMutex
Mutex mMutex;
sp<GraphicBuffer> mAcquiredBuffer;
};
// ---------------------------------------------------------------------------
} // namespace android
// ---------------------------------------------------------------------------
#endif // ANDROID_SF_VIRTUAL_DISPLAY_SURFACE_H