When a large queue of frames builds up due to a lost frame, the decode
delay can sometimes become quite large. In this case the stream may
signal as timed out when in fact it is not. Instead, the delay should
be capped at the timeout limit.
Bug: webrtc:14168
Change-Id: I5b4e8851b2c6d7d27a698627dc1633931d7fc00e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/265404
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37199}
Adds new class DecodeSynchronizer that will coalesce the decoding
of received streams on the metronome. This feature is experimental and
is backed by a field trial WebRTC-FrameBuffer3.
This experiment now has 3 arms to it,
"WebRTC-FrameBuffer3/arm:FrameBuffer2/": Default, uses old frame buffer.
"WebRTC-FrameBuffer3/arm:FrameBuffer3/": Uses new frame buffer.
"WebRTC-FrameBuffer3/arm:SyncDecoding/": Uses new frame buffer with
frame scheduled on the metronome.
The SyncDecoding arm will not work until it is wired up in the follow-up
CL.
This change also makes the following modifications,
* Adds FakeMetronome utilities for tests using a metronome.
* Makes FrameDecodeScheduler an interface. The default implementation is
TaskQueueFrameDecodeScheduler.
* FrameDecodeScheduler now has a Stop() method, which must be called
before destruction.
TBR=philipel@webrtc.org
Change-Id: I58a306bb883604b0be3eb2a04b3d07dbdf185c71
Bug: webrtc:13658
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/250665
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Stefan Holmer <holmer@google.com>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35988}
This emulates behaviour from frame buffer 2, but does not handle stats.
In contrast to frame buffer 2, all work happens on the same task queue.
FrameBuffer3Proxy encapsulates FrameBuffer3 and scheduler behind
a field trial WebRTC-FrameBuffer3.
This separates frame scheduling behaviour into a few components,
VideoReceiveStreamTimeoutTracker
* Handles the stream timeouts.
FrameDecodeScheduler
* Manages the scheduling and cancelling of frames being sent to the
decoder.
FrameDecodeTiming
* Handles the timing and ordering of frames to be decoded.
Other changes
* Adds CurrentSize() method to FrameBuffer3
* Move timing to a separate library
* Does a thread check for Receive statistics as this is now
on the worker thread.
* Adds `FlushImmediate` method to RunLoop so that
video_receive_stream2_unittest can pass when scheduling is happening
on the worker thread.
Change-Id: Ia8d2e5650d1708cdc1be3631a5214134583a0721
Bug: webrtc:13343
Tested: Ran webrtc_perf_tests, video_engine_tests, rtc_unittests forcing frame buffer3
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/241603
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Markus Handell <handellm@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35847}
