Chromium has now been updated, so we can remove the base headers from
rtc_media.
BUG=None
Review-Url: https://codereview.webrtc.org/2590813002
Cr-Commit-Position: refs/heads/master@{#15712}
The new name, OnCompleteReports rather than OnComplete, is needed
because in C++ method lookup, overriding a method hides all otherwise
inherited methods with the same name, even if they have a different
signature. And here, the intention is that each subclass should
override one or the other of the two methods, and inherit the method it
doesn't override.
This cl is a prerequisite for
https://codereview.webrtc.org/2567143003/, because the Chrome glue
code needs to retain the stats report after the OnComplete method has
returned.
Currently, Chrome makes a copy of the stats mapping (which breaks when
changing ValuePtr from an rtc::linked_ptr to an std::unique_ptr). After
this cl, Chrome can be fixed to take ownership and no longer needs to
copy anything, unblocking cl 2567143003.
BUG=webrtc:6424
Review-Url: https://codereview.webrtc.org/2584553002
Cr-Commit-Position: refs/heads/master@{#15708}
Documentation was also unclear, it seems it returned the RTP packet
size including RTP headers.
BUG=None.
Review-Url: https://codereview.webrtc.org/2588343002
Cr-Commit-Position: refs/heads/master@{#15707}
Rename variables and private functions to follow style,
replace remaining asserts with DCHECKs.
add 'ms' suffix to time variables derived from clock_
add 'ntp' suffix to time variables derived from ntp time.
No functional changes expected.
BUG=None
Review-Url: https://codereview.webrtc.org/2588753002
Cr-Commit-Position: refs/heads/master@{#15706}
Reason for revert:
This CL broke all Chromium WebRTC FYI bots. A roll+fix was attempted here: https://codereview.chromium.org/2590783003/, but failed to land. I'm reverting this CL now to make the tree green again. Make the API change gradual when you reland so that we can update Chromium between.
Original issue's description:
> Add disabled certificate check support to IceServer PeerConnection API.
>
> Refactor "OPT_SSLTCP" renaming it to "OPT_TLS_FAKE", making it clear
> that it's not actually some kind of SSL over TCP. Also making it clear
> that it's mutually exclusive with OPT_TLS.
>
> Add "OPT_TLS_INSECURE" that implements the new certificate-check
> disabled TLS mode, which is also mutually exclusive with the other
> TLS options.
>
> PortAllocator: Add a new TLS policy enum TlsCertPolicy which defines
> the new insecure mode and added it as a RelayCredentials member.
>
> TurnPort: Add new TLS policy member with appropriate getter and setter
> to avoid constructor bloat. Initialize it from the RelayCredentials
> after the TurnPort is created.
>
> Expose the new feature in the PeerConnection API via
> IceServer.tls_certificate_policy as well as via the Android JNI
> PeerConnection API.
>
> For security reasons we ensure that:
>
> 1) The policy is always explicitly initialized to secure.
> 2) API users have to explicitly integrate with the feature to
> use it, and will otherwise get no change in behavior.
> 3) The feature is not immediately exposed in non-native
> contexts. For example, disabling of certificate validation
> is not implemented via URI parsing since this would
> immediately allow it to be used from a web page.
>
> BUG=webrtc:6840
>
> Review-Url: https://codereview.webrtc.org/2557803002
> Cr-Commit-Position: refs/heads/master@{#15670}
> Committed: b0f04fdb9eTBR=pthatcher@webrtc.org,deadbeef@webrtc.org,hnsl@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6840
Review-Url: https://codereview.webrtc.org/2590153002
Cr-Commit-Position: refs/heads/master@{#15703}
The test verifies that the hysteresis window in the configuration works
as intended.
BUG=webrtc:6708
Review-Url: https://codereview.webrtc.org/2594563002
Cr-Commit-Position: refs/heads/master@{#15700}
No need to pass a whole struct around, when only one member is used.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2589833002
Cr-Commit-Position: refs/heads/master@{#15687}
Reason for revert:
breaks downstream project.
Can you make this change in a compatible way using anonymous union:
union {
bool is_first_packet_in_frame;
RTC_DEPRECATED bool isFirstPacket;
};
(unfortunetly this this treak breaks braced initialization in rtp_rtcp_impl_unittest.cc,
so that should be rewritting in a more classic way)
Original issue's description:
> Rename RTPVideoHeader.isFirstPacket to .is_first_packet_in_frame.
>
> Name should represent the actual meaning.
>
> BUG=None
>
> Review-Url: https://codereview.webrtc.org/2574943003
> Cr-Commit-Position: refs/heads/master@{#15684}
> Committed: efde908380TBR=stefan@webrtc.org,sprang@webrtc.org,johan@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=None
Review-Url: https://codereview.webrtc.org/2589783003
Cr-Commit-Position: refs/heads/master@{#15686}
"WebRTC.Call.NumberOfPauseEvents" -> "WebRTC.Video.NumberOfPauseEvents"
Recorded if a certain time has passed (10 sec) since the first media packet was sent.
Moved to per stream to know when media has started and to prevent logging stats for calls that was never in use.
Add histogram for percentage of paused video time for sent video streams:
"WebRTC.Video.PausedTimeInPercent"
BUG=b/32659204
Review-Url: https://codereview.webrtc.org/2530393003
Cr-Commit-Position: refs/heads/master@{#15681}
Reason for revert:
Bug affecting perf tests has been fixed. The issue was that I had accidentally disabled cpu overuse adaptation based on the encoders ScalingSettings, not just quality-based scaling.
Original issue's description:
> Revert of Properly report number of quality downscales in stats. (patchset #11 id:220001 of https://codereview.webrtc.org/2564373002/ )
>
> Reason for revert:
> Breaks perf tests
>
> Original issue's description:
> > Properly report number of quality downscales in stats.
> >
> > A regression was introduced in 876222f that caused these stats to
> > be reported incorrectly. This used to be only implemented for VP8
> > but should now be available for all codecs.
> >
> > BUG=webrtc:6860
> >
> > Review-Url: https://codereview.webrtc.org/2564373002
> > Cr-Commit-Position: refs/heads/master@{#15673}
> > Committed: 0c8c538835
>
> TBR=asapersson@webrtc.org,stefan@webrtc.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=webrtc:6860
>
> Review-Url: https://codereview.webrtc.org/2586783003
> Cr-Commit-Position: refs/heads/master@{#15678}
> Committed: fe04bd43ccTBR=asapersson@webrtc.org,stefan@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6860
Review-Url: https://codereview.webrtc.org/2588743002
Cr-Commit-Position: refs/heads/master@{#15680}
This effectively reverts commit c3e1cabc696240e4b5a128653264785292878205
(https://codereview.webrtc.org/2589703002/).
The reason the test was failing before was missing resource
dependencies in the GN file. This is now fixed.
Furthermore, the test did not trigger the complexity adaptation that
it was supposed to test, since the hysteresis window of the bitrate
was not taken into account. This is also fixed.
Finally, a percent label was added to a printout, to match the same
printout in the other test.
BUG=webrtc:6708
Review-Url: https://codereview.webrtc.org/2580383002
Cr-Commit-Position: refs/heads/master@{#15679}
Reason for revert:
Breaks perf tests
Original issue's description:
> Properly report number of quality downscales in stats.
>
> A regression was introduced in 876222f that caused these stats to
> be reported incorrectly. This used to be only implemented for VP8
> but should now be available for all codecs.
>
> BUG=webrtc:6860
>
> Review-Url: https://codereview.webrtc.org/2564373002
> Cr-Commit-Position: refs/heads/master@{#15673}
> Committed: 0c8c538835TBR=asapersson@webrtc.org,stefan@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6860
Review-Url: https://codereview.webrtc.org/2586783003
Cr-Commit-Position: refs/heads/master@{#15678}
I always find myself re-writing this function for debug purposes. It'd
save so much time if it already existed...
BUG=none
Review-Url: https://codereview.webrtc.org/2589773002
Cr-Commit-Position: refs/heads/master@{#15677}
Reason for revert:
Still hitting NOTREACHED.
Original issue's description:
> Revert of Disabling NOTREACHED which we're hitting flakily in browser tests. (patchset #1 id:1 of https://codereview.webrtc.org/2477663002/ )
>
> Reason for revert:
> To see if the NOTREACHED is still hit.
>
> Original issue's description:
> > Disabling NOTREACHED which we're hitting flakily in browser tests.
> >
> > I have no idea how bad it is that we're hitting this limit; I'm just
> > doing this to stop the tests from flaking.
> >
> > BUG=webrtc:6484
> >
> > Committed: https://crrev.com/6eaa55867b449df992752c1df540ec42f9d9b057
> > Cr-Commit-Position: refs/heads/master@{#14974}
>
> TBR=stefan@webrtc.org,phoglund@webrtc.org
> # Not skipping CQ checks because original CL landed more than 1 days ago.
> BUG=webrtc:6484
>
> Review-Url: https://codereview.webrtc.org/2585183002
> Cr-Commit-Position: refs/heads/master@{#15665}
> Committed: 9d7ea0920cTBR=stefan@webrtc.org,phoglund@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6484
Review-Url: https://codereview.webrtc.org/2585273002
Cr-Commit-Position: refs/heads/master@{#15676}
RTCStatsCollector relies on PeerConnection and its WebRtcSession. If the
PeerConnection is destroyed, reference counting keeps the
RTCStatsCollector alive until the request has completed. But the request
is using PeerConnection/WebRtcSession resources that are destroyed in
~PeerConnection().
To get around this problem, RTCStatsCollector::WaitForPendingRequest()
is added, which is invoked at ~PeerConnection().
Integration test added, it caused a segmentation fault before this
change / EXPECT failure.
BUG=chromium:627816
Review-Url: https://codereview.webrtc.org/2583613003
Cr-Commit-Position: refs/heads/master@{#15674}
A regression was introduced in 876222f that caused these stats to
be reported incorrectly. This used to be only implemented for VP8
but should now be available for all codecs.
BUG=webrtc:6860
Review-Url: https://codereview.webrtc.org/2564373002
Cr-Commit-Position: refs/heads/master@{#15673}
(This is a re-upload of https://codereview.webrtc.org/2567243003/, the
CQ stopped working there.)
The previously used WebRtcSession::GetTransportStats did a synchronous
invoke per channel (voice, video, data) on the signaling thread to the
network thread - e.g. 3 blocking invokes.
It is replaced by WebRtcSession::GetStats[_s] which can be invoked on
the signaling thread or on any thread if a ChannelNamePairs argument is
present (provided by WebRtcSession::GetChannelNamePairs on the signaling
thread).
With these changes, and changes allowing the getting of certificates
from any thread, the RTCStatsCollector can turn the 3 blocking thread
invokes into 1 non-blocking invoke.
BUG=webrtc:6875, chromium:627816
Review-Url: https://codereview.webrtc.org/2583883002
Cr-Commit-Position: refs/heads/master@{#15672}
The rtx streams were not included in the number of expected streams
but the test passed most of the time anyway due to how the checking was done.
Flake was caused when the number of registered streams jumped passed the
number of expected send streams excluding the number of rtx streams.
BUG=webrtc:6879
Review-Url: https://codereview.webrtc.org/2580343002
Cr-Commit-Position: refs/heads/master@{#15671}
Refactor "OPT_SSLTCP" renaming it to "OPT_TLS_FAKE", making it clear
that it's not actually some kind of SSL over TCP. Also making it clear
that it's mutually exclusive with OPT_TLS.
Add "OPT_TLS_INSECURE" that implements the new certificate-check
disabled TLS mode, which is also mutually exclusive with the other
TLS options.
PortAllocator: Add a new TLS policy enum TlsCertPolicy which defines
the new insecure mode and added it as a RelayCredentials member.
TurnPort: Add new TLS policy member with appropriate getter and setter
to avoid constructor bloat. Initialize it from the RelayCredentials
after the TurnPort is created.
Expose the new feature in the PeerConnection API via
IceServer.tls_certificate_policy as well as via the Android JNI
PeerConnection API.
For security reasons we ensure that:
1) The policy is always explicitly initialized to secure.
2) API users have to explicitly integrate with the feature to
use it, and will otherwise get no change in behavior.
3) The feature is not immediately exposed in non-native
contexts. For example, disabling of certificate validation
is not implemented via URI parsing since this would
immediately allow it to be used from a web page.
BUG=webrtc:6840
Review-Url: https://codereview.webrtc.org/2557803002
Cr-Commit-Position: refs/heads/master@{#15670}
The following QP stats are added:
*avgQP for the current polling interval as a fraction of the current delta sum over the current delta of encoded frames
*total QPSum
*total number of encoded frames
BUG=NONE
Review-Url: https://codereview.webrtc.org/2578123002
Cr-Commit-Position: refs/heads/master@{#15669}
Reason for revert:
This fails on perf bots.
See for instance https://uberchromegw.corp.google.com/i/client.webrtc.perf/builders/Android32%20Tests%20%28K%20Nexus5%29/builds/105/steps/webrtc_perf_tests/logs/stdio:
I 643.319s run_tests_on_device(03832b22f0e944a0) Note: Google Test filter = AudioEncoderOpusComplexityAdaptationTest.AdaptationOff
I 643.319s run_tests_on_device(03832b22f0e944a0) [==========] Running 1 test from 1 test case.
I 643.319s run_tests_on_device(03832b22f0e944a0) [----------] Global test environment set-up.
I 643.319s run_tests_on_device(03832b22f0e944a0) [----------] 1 test from AudioEncoderOpusComplexityAdaptationTest
I 643.319s run_tests_on_device(03832b22f0e944a0) [ RUN ] AudioEncoderOpusComplexityAdaptationTest.AdaptationOff
I 643.319s run_tests_on_device(03832b22f0e944a0) ../../webrtc/modules/audio_coding/codecs/opus/opus_complexity_unittest.cc:36: Failure
I 643.319s run_tests_on_device(03832b22f0e944a0) Value of: audio_loop.Init(kInputFileName, kMaxLoopLengthSamples, kInputBlockSizeSamples)
I 643.320s run_tests_on_device(03832b22f0e944a0) Actual: false
I 643.320s run_tests_on_device(03832b22f0e944a0) Expected: true
I 643.320s run_tests_on_device(03832b22f0e944a0) [ CRASHED ]
I 643.405s run_tests_on_device(0384888cf0e945d1) >>ScopedMainEntryLogger
I 643.405s run_tests_on_device(0384888cf0e945d1) Note: Google Test filter = AudioEncoderOpusComplexityAdaptationTest.AdaptationOn
I 643.405s run_tests_on_device(0384888cf0e945d1) [==========] Running 1 test from 1 test case.
I 643.405s run_tests_on_device(0384888cf0e945d1) [----------] Global test environment set-up.
I 643.405s run_tests_on_device(0384888cf0e945d1) [----------] 1 test from AudioEncoderOpusComplexityAdaptationTest
I 643.405s run_tests_on_device(0384888cf0e945d1) [ RUN ] AudioEncoderOpusComplexityAdaptationTest.AdaptationOn
I 643.405s run_tests_on_device(0384888cf0e945d1) ../../webrtc/modules/audio_coding/codecs/opus/opus_complexity_unittest.cc:36: Failure
I 643.405s run_tests_on_device(0384888cf0e945d1) Value of: audio_loop.Init(kInputFileName, kMaxLoopLengthSamples, kInputBlockSizeSamples)
I 643.405s run_tests_on_device(0384888cf0e945d1) Actual: false
I 643.405s run_tests_on_device(0384888cf0e945d1) Expected: true
I 643.405s run_tests_on_device(0384888cf0e945d1) [ CRASHED ]
Original issue's description:
> Re-enable Opus complexity tests on Android
>
> BUG=webrtc:6708
> TBR=minyue@webrtc.org
>
> Review-Url: https://codereview.webrtc.org/2589673002
> Cr-Commit-Position: refs/heads/master@{#15667}
> Committed: 14479ddfd4
TBR=
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6708
Review-Url: https://codereview.webrtc.org/2589703002
Cr-Commit-Position: refs/heads/master@{#15668}
Also rename internal::FlexfecReceiveStream to FlexfecReceiveStreamImpl.
BUG=webrtc:6849
Review-Url: https://codereview.webrtc.org/2561123002
Cr-Commit-Position: refs/heads/master@{#15666}
Reason for revert:
To see if the NOTREACHED is still hit.
Original issue's description:
> Disabling NOTREACHED which we're hitting flakily in browser tests.
>
> I have no idea how bad it is that we're hitting this limit; I'm just
> doing this to stop the tests from flaking.
>
> BUG=webrtc:6484
>
> Committed: https://crrev.com/6eaa55867b449df992752c1df540ec42f9d9b057
> Cr-Commit-Position: refs/heads/master@{#14974}
TBR=stefan@webrtc.org,phoglund@webrtc.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=webrtc:6484
Review-Url: https://codereview.webrtc.org/2585183002
Cr-Commit-Position: refs/heads/master@{#15665}
In addition to moving tools/mb -> tools-webrtc/mb, also
move webrtc/build/mb_config.pyl into tools-webrtc/mb
to match the default location better.
Remove Chromium-specific check for 'mb validate' that failed
due to doing this (we never cleaned that when we forked the code).
BUG=webrtc:5006
TBR=ehmaldonado@webrtc.org
NOTRY=True
TESTED=Manually ran:
tools-webrtc/mb/mb.py gen -m client.webrtc -b 'iOS64 Release' --config-file tools-webrtc/mb/mb_config.pyl //out/Release-iphoneos
tools-webrtc/mb/mb.py gen -m client.webrtc -b 'Mac64 Release' --config-file tools-webrtc/mb/mb_config.pyl //out/Release-mac
Review-Url: https://codereview.webrtc.org/2585743002 .
Cr-Commit-Position: refs/heads/master@{#15664}
H264 High profile is causing problems for some downstream apps. Therefore,
put it behind a field trial instead.
BUG=webrtc:6337
Review-Url: https://codereview.webrtc.org/2580963004
Cr-Commit-Position: refs/heads/master@{#15663}
The WebRTC valgrind wrapper scripts needs to be moved in order to
unlock us from depending on a Chromium checkout.
BUG=webrtc:5006
TBR=ehmaldonado@webrtc.org
NOTRY=True
Review-Url: https://codereview.webrtc.org/2578093002
Cr-Commit-Position: refs/heads/master@{#15662}
Permits overriding the source-default is_screencast option to be able to
treat screencast sources as fluid video, preserving motion at the loss
of individual frame quality (or vice versa).
BUG=chromium:653531
R=deadbeef@webrtc.org
Review-Url: https://codereview.webrtc.org/2579993003
Cr-Commit-Position: refs/heads/master@{#15659}
Having that dependency require user of RtpPacket to ensure
RtpHeaderExtensionMap always outlive packet and that RtpPacket's access
to RtpHeaderExtensionMap is properly syncrhonized.
Dropping this dependencies make use of RtpPacket less error-prone.
BUG=webrtc:5261
Review-Url: https://codereview.webrtc.org/2576653003
Cr-Commit-Position: refs/heads/master@{#15653}
