It was possible before if an input fps dropped due to cpu adaptation
Also, this CL removes occasional test failure (it could've happened if
input framerate got very low)
Bug: webrtc:11432,webrtc:11426
Change-Id: Id1a4df23302f7b8ab6781f1e7cca5112bfcfe9ea
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/170469
Reviewed-by: Artem Titov <titovartem@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#30802}
This class doesn't strictly follow rtc::TaskQueue semantics,
which makes it surprising and hard to use correctly.
Please use TaskQueueForTest instead.
This CL follows usual deprecation process:
1/ Rename.
% for i in `git ls-files` ; sed -i "s:SingleThreadedTaskQueueForTesting:DEPRECATED_SingleThreadedTaskQueueForTesting:" $i
2/ Annotate old name for downstream users and accidental new uses.
Bug: webrtc:10933
Change-Id: I80b4ee5a48df1f63f63a43ed0efdb50eb7fb156a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/150788
Reviewed-by: Elad Alon <eladalon@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Tommi <tommi@webrtc.org>
Commit-Queue: Yves Gerey <yvesg@google.com>
Cr-Commit-Position: refs/heads/master@{#29045}
This CL adds thread annotations and ensure that neither data races
nor deadlocks occur.
It prevents weird results and helps detecting other concurrency issues.
As a bonus, some dead code has been removed.
Bug: webrtc:10834
Change-Id: Ibd140db9e4dbf81b212044647e2d85bd18ef8d78
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/147278
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Commit-Queue: Yves Gerey <yvesg@google.com>
Cr-Commit-Position: refs/heads/master@{#28737}
Use the newly added total_decode_time_ms to get an accurate value
for the average decode time. The sparsely sampled decode_ms is
sensitive to the sampling instance.
Bug: chromium:980853
Change-Id: I9b63c8d1053fa95f74918807b83d1edb5cd726fb
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/147268
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Johannes Kron <kron@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28716}
The former became redundant and didn't guarantee
numerical stability for variance computation.
Bug: webrtc:10412
Change-Id: Idc291abe9add41bde9e7734f179e5d6c65f2630b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/132460
Commit-Queue: Yves Gerey <yvesg@google.com>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#27605}
This reverts commit 1f0a84a2ecea59f86adc1af70eed974a3c6d59ac.
Reason for revert: Partial Capture API is not needed, according to new info from the Chrome team.
Original change's description:
> Partial frame capture API part 5
>
> Wire up partial video frames in video quality tests
>
> Bug: webrtc:10152
> Change-Id: Ifa13bb308258c8d3930a6cfbcc97c95b132cecf3
> Reviewed-on: https://webrtc-review.googlesource.com/c/120410
> Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
> Reviewed-by: Stefan Holmer <stefan@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#26549}
TBR=ilnik@webrtc.org,sprang@webrtc.org,stefan@webrtc.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: webrtc:10152
Change-Id: I32017b1a7109a3615598a976f4b0e61edf4e8757
Reviewed-on: https://webrtc-review.googlesource.com/c/122088
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26628}
Wire up partial video frames in video quality tests
Bug: webrtc:10152
Change-Id: Ifa13bb308258c8d3930a6cfbcc97c95b132cecf3
Reviewed-on: https://webrtc-review.googlesource.com/c/120410
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26549}
Reland with fixes. Previous iteration affected media bitrate in bunch of tests.
Always use real VideoStreamsFactory in full stack tests
Because quality scaling is enabled now in full stack test, correct
factory should be used to compute actual resolution.
Also, since analyzed stream may be disabled completely now, change how
analyzer considers the test finished --- count captured frames and
stop if required amount of frames is captured and no new comparison were made.
Original Reviewed-on: https://webrtc-review.googlesource.com/c/118687
Bug: webrtc:10204
Change-Id: Id1d9066add185d56fe3cb6856b700d350576c6b2
Reviewed-on: https://webrtc-review.googlesource.com/c/119950
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26460}
This reverts commit 18cf2383aa2eb9de5778991c9d13b6b847143d37.
Reason for revert: Unexpected changes in webrtc_perf stats.
Original change's description:
> Always use real VideoStreamsFactory in full stack tests
>
> Because quality scaling is enabled now in full stack test, correct
> factory should be used to compute actual resolution.
>
> Also, since analyzed stream may be disabled completely now, change how
> analyzer considers the test finished --- count captured frames and
> stop if required amount of frames is captured and no new comparison were
> made.
>
> Bug: webrtc:10204
> Change-Id: I205ebc892969ec1cf2d83e054e5c95e089d32104
> Reviewed-on: https://webrtc-review.googlesource.com/c/118687
> Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
> Reviewed-by: Erik Språng <sprang@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#26358}
TBR=ilnik@webrtc.org,sprang@webrtc.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: webrtc:10204
Change-Id: Ia52fd55c9f68627166e0538d377003eae4ea518a
Reviewed-on: https://webrtc-review.googlesource.com/c/119946
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26405}
Because quality scaling is enabled now in full stack test, correct
factory should be used to compute actual resolution.
Also, since analyzed stream may be disabled completely now, change how
analyzer considers the test finished --- count captured frames and
stop if required amount of frames is captured and no new comparison were
made.
Bug: webrtc:10204
Change-Id: I205ebc892969ec1cf2d83e054e5c95e089d32104
Reviewed-on: https://webrtc-review.googlesource.com/c/118687
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26358}
Also, attach analyzer to the correct receive stream, instead of attaching
it only if there's one receive stream.
Bug: None
Change-Id: I34888b5bd09b61f0939d77b26cb0d10f9261d3cb
Reviewed-on: https://webrtc-review.googlesource.com/c/118688
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26357}
In particular, time_utils.h is currently pulled in via rtc_event.h
This CL is in preparation of moving parts of the RTC event log to api/.
Bug: webrtc:10206
Change-Id: Idd35aa9404afded4d29b1296344996c45b8c2e91
Reviewed-on: https://webrtc-review.googlesource.com/c/117921
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26326}
CallTest, VideoQualityTest and VideoAnalyzer used test::TestVideoCapturer
as an interface for video sources. Change to use VideoSourceInterface instead,
since that's all they need.
This is a preparation for making test::VcmCapturer usable as a
VideoTrackSource, and replace use of cricket::VideoCapturer in example code.
Bug: webrtc:6353
Change-Id: I445f5f6f9b7342230b89f53a5722df9c9e92834f
Reviewed-on: https://webrtc-review.googlesource.com/c/114881
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26047}
Allows enabling audio for RunWithAnalyzer method, and prints out audio jitterbuffer performance stats. Also fixes for RunWithRenderer when enabling audio (seg-faulted).
Bug: b/112299470
Change-Id: Ic7c0de1c455891f38cca317001c6c216e82f6ec3
Reviewed-on: https://webrtc-review.googlesource.com/92800
Commit-Queue: Christoffer Rodbro <crodbro@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#24208}
Replaced by a int64_t representing time in us.
Bug: webtrc:9584
Change-Id: I0505c020ef741ad940203ec300e8adb103856dda
Reviewed-on: https://webrtc-review.googlesource.com/91840
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#24204}
