This CL should reduce the number of timeouts for dequeueInputBuffer() which results in the log "MediaCodecVideo: dequeueInputBuffer error" followed by software fallback for VP8/VP9 and codec restart for H264.
A timeout always happen for dequeueInputBuffer() when frames_received_ > frames_decoded_ + num_input_buffers. The following code tries to drain the decoder before enqueuing more input buffers:
// Try to drain the decoder and wait until output is not too
// much behind the input.
if (frames_received_ > frames_decoded_ + max_pending_frames_) {
ALOGV("Received: %d. Decoded: %d. Wait for output...",
frames_received_, frames_decoded_);
if (!DeliverPendingOutputs(jni, kMediaCodecTimeoutMs,
true /* dropFrames */)) {
ALOGE << "DeliverPendingOutputs error";
return ProcessHWErrorOnCodecThread();
}
if (frames_received_ > frames_decoded_ + max_pending_frames_) {
ALOGE << "Output buffer dequeue timeout";
return ProcessHWErrorOnCodecThread();
}
...
}
However, for H264, |max_pending_frames_| can currently be larger than the number of input buffers so that the code above is never executed. This CL limits |max_pending_frames_| to the number of input buffers.
TBR=glaznev
BUG=b/24867188,b/24864151
Review URL: https://codereview.webrtc.org/1394303005
Cr-Commit-Position: refs/heads/master@{#10273}
Reason for revert:
Broke browser_tests on Mac. Still need to investigate the cause.
Original issue's description:
> Moving MediaStreamSignaling logic into PeerConnection.
>
> This needs to happen because in the future, m-lines will be offered
> based on the set of RtpSenders/RtpReceivers, rather than the set of
> tracks that MediaStreamSignaling knows about.
>
> Besides that, MediaStreamSignaling was a "glue class" without
> a clearly defined role, so it going away is good for other
> reasons as well.
>
> Committed: https://crrev.com/97c392935411398b506861601c82e31d95c591f0
> Cr-Commit-Position: refs/heads/master@{#10268}
TBR=pthatcher@webrtc.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.webrtc.org/1403633005
Cr-Commit-Position: refs/heads/master@{#10269}
This needs to happen because in the future, m-lines will be offered
based on the set of RtpSenders/RtpReceivers, rather than the set of
tracks that MediaStreamSignaling knows about.
Besides that, MediaStreamSignaling was a "glue class" without
a clearly defined role, so it going away is good for other
reasons as well.
Review URL: https://codereview.webrtc.org/1393563002
Cr-Commit-Position: refs/heads/master@{#10268}
SurfaceViewHelper requires EGL14 that was added in API level 17. Since the SurfaceViewHelper is only neeed when we capture to textures, this cl change back to not use it when we are capturing to byte buffers.
Also, thread.quitsafely was added in level 18. Instead a new ThreadUtil method has been added for this.
BUG=b/24782220
TEST = run
ninja -C out/Debug libjingle_peerconnection_android_unittest && CHECKOUT_SOURCE_ROOT=`pwd` build/android/adb_install_apk.py --debug out/Debug/apks/libjingle_peerconnection_android_unittest.apk && ./third_party/android_tools/sdk/platform-tools/adb shell am instrument -w -e class org.webrtc.VideoCapturerAndroidTest org.webrtc.test/android.test.InstrumentationTestRunner on a device running Android 4.1 (I tried Nexus 7, the first version)
Review URL: https://codereview.webrtc.org/1401023003
Cr-Commit-Position: refs/heads/master@{#10265}
This CL makes AddRef() and Release() const member methods and the refcount integer mutable. This is reasonable, because they only manage the lifetime of the object, and this is also how it's done in Chromium.
The purpose is to be able to capture a const pointer in a scoped_refptr, which is currenty impossible. The practial problem this CL solves is this:
void Foo::Bar() const {}
rtc::Callback0<void> Foo::MakeClosure() const {
return rtc::Bind(&Foo::Bar, this);
}
We currently capture |this| as const Foo*. With this CL, |this| will be captured as scoped_refptr<const Foo>.
A test is also added in bind_unittest to check this behaviour.
BUG=webrtc:5065
R=perkj@webrtc.org, tommi@webrtc.org
Review URL: https://codereview.webrtc.org/1403683004 .
Cr-Commit-Position: refs/heads/master@{#10253}
This make small refactorings to MediaVideoEncoder to prepare for adding support to encode from textures. The C++ layer does not have any functional changes.
- Moves ResetEncoder to always work on the codec thread
- Adds use of ThreadChecker.
- Change Java MediaEncoder.Init to return true or false and introduce method getInputBuffers.
- Add simple unit test for Java MediaCodecVideoEncoder.
BUG=webrtc:4993
Review URL: https://codereview.webrtc.org/1396073003
Cr-Commit-Position: refs/heads/master@{#10250}
This CL is a small bug fix for "Android SurfaceViewRenderer: Allow to re-init after release() has been called" https://codereview.webrtc.org/1389203003/. It is only possible to clear the last image in release() if init() has been called beforehand.
TBR=hbos
BUG=webrtc:4742
Review URL: https://codereview.webrtc.org/1396573003 .
Cr-Commit-Position: refs/heads/master@{#10223}
These are the necessary changes in C++ related to the video capturer necessary to capture to a surface texture.
It does not handle scaling / cropping yet though.
BUG=
R=magjed@webrtc.org
Review URL: https://codereview.webrtc.org/1395673003 .
Cr-Commit-Position: refs/heads/master@{#10218}
This CL makes a thorough reset of all variables in release() and clears the last rendered image so that the SurfaceViewRenderer object can be reinitialized with init() and work properly. This CL also removes an implicit assumption that init() is called before surfaceCreated() - now they can be called in any order.
BUG=webrtc:4742
R=hbos@webrtc.org
Review URL: https://codereview.webrtc.org/1389203003 .
Cr-Commit-Position: refs/heads/master@{#10217}
What used to be the libpeerconnection library is now compiled
statically into the Chromium binary, so clean up references it.
BUG=chromium:482123
TBR=tommi@webrtc.org
Review URL: https://codereview.webrtc.org/1399513002 .
Cr-Commit-Position: refs/heads/master@{#10216}
This adds support for capturing to a texture in the Java part of VideoCapturerAndroid.
After this cl, the C++ also needs modification.
https://codereview.webrtc.org/1375953002/ contains the idea and have a working version where textures can be rendered in local preview.
BUG=webrtc:4993
R=magjed@webrtc.org
Review URL: https://codereview.webrtc.org/1383413002 .
Cr-Commit-Position: refs/heads/master@{#10213}
This CL should be the last one in a series to finally unblock camera texture capture.
The SurfaceTexture.updateTexImage() calls are moved from the video renderers into MediaCodecVideoDecoder, and the destructor of the texture frames will signal MediaCodecVideoDecoder that the frame has returned. This CL also removes the SurfaceTexture from the native handle and only exposes the texture matrix instead, because only the video source should access the SurfaceTexture.
BUG=webrtc:4993
R=glaznev@webrtc.org, perkj@webrtc.org
Review URL: https://codereview.webrtc.org/1378033003 .
Cr-Commit-Position: refs/heads/master@{#10203}
Every now and then we get CLs to codereview.webrtc.org
that are created from a Chromium checkout by editing
the code in third_party/webrtc or third_party/libjingle.
By editing these lower-level codereview.settings files,
we instead cause a crash during 'git cl upload', but the
contents of the file will also be printed, which can work
as an error message. The alternative would be to entirely
remove the files.
BUG=
R=andrew@webrtc.org, tommi@webrtc.org
Review URL: https://codereview.webrtc.org/1389963002 .
Cr-Commit-Position: refs/heads/master@{#10191}
The Java PeerConnection maintains a cached list of Java RtpSenders
and RtpReceivers so that the same objects are returned every time
getSenders() or getReceivers() is called. They are disposed of when
PeerConnection.dispose() is called, which will also dispose their
referenced MediaStreamTracks.
Review URL: https://codereview.webrtc.org/1368143002
Cr-Commit-Position: refs/heads/master@{#10189}
Every malformed URL should now produce an error message in JS, rather than
silently failing and possibly printing a warning message to the console (and
possibly crashing).
Also added some unit tests, and made "ParseIceServers" public.
BUG=445002
Review URL: https://codereview.webrtc.org/1344143002
Cr-Commit-Position: refs/heads/master@{#10186}
This allows to correctly report first frame event in applications which
use same remote video renderer for multiple calls.
R=wzh@webrtc.org
Review URL: https://codereview.webrtc.org/1384353002 .
Cr-Commit-Position: refs/heads/master@{#10176}
Fixed a problem where eglBase.makecurrent() could be called after the context had been released if SurfaceTextureHelper was first created and immedately disconnected.
Add the possibility to inject a thread to use instead of creating a new.
BUG= webrtc:4993
R=magjed@webrtc.org
Review URL: https://codereview.webrtc.org/1384923002 .
Cr-Commit-Position: refs/heads/master@{#10174}
