external/webrtc.git
2 hours agoRemove mouse cursor capturer from the ScreenCapturer interface master
sergeyu@chromium.org [Thu, 2 Oct 2014 01:47:10 +0000 (01:47 +0000)]
Remove mouse cursor capturer from the ScreenCapturer interface

Mouse can be captured using MouseCursorMonitor and all code in chromium
already uses it instead of ScreenCapturer.

R=jiayl@webrtc.org

Committed: https://code.google.com/p/webrtc/source/detail?r=7363

Review URL: https://webrtc-codereview.appspot.com/31529004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7365 4adac7df-926f-26a2-2b94-8c16560cd09d

2 hours agoRevert "Remove mouse cursor capturer from the ScreenCapturer interface"
sergeyu@chromium.org [Thu, 2 Oct 2014 01:36:43 +0000 (01:36 +0000)]
Revert "Remove mouse cursor capturer from the ScreenCapturer interface"

This reverts commit 0adc4953512ee0a57cf7f3c0591b024c2316554a. It broke
FYI bots

TBR=sergeyu@chromium.org

Review URL: https://webrtc-codereview.appspot.com/27649004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7364 4adac7df-926f-26a2-2b94-8c16560cd09d

4 hours agoRemove mouse cursor capturer from the ScreenCapturer interface
sergeyu@chromium.org [Thu, 2 Oct 2014 00:18:10 +0000 (00:18 +0000)]
Remove mouse cursor capturer from the ScreenCapturer interface

Mouse can be captured using MouseCursorMonitor and all code in chromium
already uses it instead of ScreenCapturer.

R=jiayl@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31529004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7363 4adac7df-926f-26a2-2b94-8c16560cd09d

5 hours agoAdd error trap for XFixesGetCursorImage()
sergeyu@chromium.org [Wed, 1 Oct 2014 23:07:12 +0000 (23:07 +0000)]
Add error trap for XFixesGetCursorImage()

BUG=https://code.google.com/p/webrtc/issues/detail?id=3245
R=jiayl@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31519004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7362 4adac7df-926f-26a2-2b94-8c16560cd09d

10 hours agoImport LappedTransform and friends.
andrew@webrtc.org [Wed, 1 Oct 2014 17:42:18 +0000 (17:42 +0000)]
Import LappedTransform and friends.

Add code for doing block-based frequency domain processing. Developed
and reviewed in isolation. Corresponding export CL:
https://chromereviews.googleplex.com/95187013/

R=bercic@google.com, kjellander@webrtc.org, turaj@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31539004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7359 4adac7df-926f-26a2-2b94-8c16560cd09d

11 hours agortc_unittest: turned sound's test gyp into gypi to speed up GYP generation.
henrike@webrtc.org [Wed, 1 Oct 2014 16:33:03 +0000 (16:33 +0000)]
rtc_unittest: turned sound's test gyp into gypi to speed up GYP generation.

BUG=N/A
R=andrew@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29629004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7358 4adac7df-926f-26a2-2b94-8c16560cd09d

12 hours agoRevert 7355 "Fix parallelization in libjingle_p2p_unittest."
henrike@webrtc.org [Wed, 1 Oct 2014 15:43:55 +0000 (15:43 +0000)]
Revert 7355 "Fix parallelization in libjingle_p2p_unittest."

Breaks waterfall.

TBR=pbos@webrtc.org
BUG=N/A

Review URL: https://webrtc-codereview.appspot.com/22909004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7357 4adac7df-926f-26a2-2b94-8c16560cd09d

13 hours agoAdds PRESUBMIT.py dispensation for depending on rtc_base.
henrike@webrtc.org [Wed, 1 Oct 2014 14:40:58 +0000 (14:40 +0000)]
Adds PRESUBMIT.py dispensation for depending on rtc_base.

Dispensation for: a few test suites, desktop capture and libjingle.

BUG=N/A
R=kjellander@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31549004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7356 4adac7df-926f-26a2-2b94-8c16560cd09d

15 hours agoFix parallelization in libjingle_p2p_unittest.
pbos@webrtc.org [Wed, 1 Oct 2014 12:31:31 +0000 (12:31 +0000)]
Fix parallelization in libjingle_p2p_unittest.

Adding VirtualSocketServers to SessionTest and RelayServerTest to avoid
contention on real ports.

R=juberti@webrtc.org
BUG=2597
TEST=third_party/gtest-parallel/gtest-parallel -w 64 out/Debug/libjingle_p2p_unittest

Review URL: https://webrtc-codereview.appspot.com/26679004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7355 4adac7df-926f-26a2-2b94-8c16560cd09d

18 hours agoFix parallelizability in modules_tests.
pbos@webrtc.org [Wed, 1 Oct 2014 10:05:40 +0000 (10:05 +0000)]
Fix parallelizability in modules_tests.

R=henrik.lundin@webrtc.org
BUG=3873
TEST=third_party/gtest-parallel/gtest-parallel -r 10 -w 64 out/Debug/modules_tests

Review URL: https://webrtc-codereview.appspot.com/24799004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7354 4adac7df-926f-26a2-2b94-8c16560cd09d

19 hours agoReland "Remove DTMF status methods from Voice Engine" r7276
henrik.lundin@webrtc.org [Wed, 1 Oct 2014 08:23:21 +0000 (08:23 +0000)]
Reland "Remove DTMF status methods from Voice Engine" r7276

This reverts r7277.

TBR=henrika@webrtc.org,pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29599004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7353 4adac7df-926f-26a2-2b94-8c16560cd09d

20 hours agoAdd support for MSan
kjellander@webrtc.org [Wed, 1 Oct 2014 08:03:19 +0000 (08:03 +0000)]
Add support for MSan

Add third_party/instrumented_libraries to setup_links.py
Add tools/msan/blacklist.txt which is the default location used
by MSan.

These changes are prerequisites to be able to use MSan with WebRTC.
To use it, one must also run:
sudo third_party/instrumented_libraries/install-build-deps.sh
to get the instrumented libraries installed (requires
/etc/apt/sources.list to be setup with deb-src entries).

NOTICE: Compilation is not yet working, but with this we can setup
a FYI bot to work with.

BUG=chromium:416871
TESTED=gclient sync + generate projects using:
GYP_DEFINES='msan=1 use_instrumented_libraries=1 instrumented_libraries_jobs=20' webrtc/build/gyp_webrtc
Built successfully in Release and ran a couple of tests (some crashed, some passed).

R=pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25649004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7352 4adac7df-926f-26a2-2b94-8c16560cd09d

22 hours agoUpdate checkdeps.py rules in DEPS
kjellander@webrtc.org [Wed, 1 Oct 2014 06:03:47 +0000 (06:03 +0000)]
Update checkdeps.py rules in DEPS

The initial rules didn't allow including
source from third_party, which is incorrect.
Cleanup irrelevant rules for directories that
are ignored, since WebRTC don't have any source
code in those locations.

BUG=
R=andrew@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30599004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7351 4adac7df-926f-26a2-2b94-8c16560cd09d

30 hours agoAdded presubmit protecting against inclusion of rtc_base, while allowing rtc_base_app...
henrike@webrtc.org [Tue, 30 Sep 2014 21:54:26 +0000 (21:54 +0000)]
Added presubmit protecting against inclusion of rtc_base, while allowing rtc_base_approved.

BUG=N/A
R=andrew@webrtc.org, kjellander@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29609004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7349 4adac7df-926f-26a2-2b94-8c16560cd09d

33 hours agoGN: Add common configs to tools and test.
kjellander@webrtc.org [Tue, 30 Sep 2014 19:07:58 +0000 (19:07 +0000)]
GN: Add common configs to tools and test.

Similar changes as in https://review.webrtc.org/28589004/
were missed in https://review.webrtc.org/25569004/.
This should fix the Chromium WebRTC FYI bots that currently
are broken due to lack of include paths.

BUG=3441
TBR=henrike@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25749004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7347 4adac7df-926f-26a2-2b94-8c16560cd09d

34 hours agoGN: Enable libvpx, add link target and convert some test targets
kjellander@webrtc.org [Tue, 30 Sep 2014 18:05:02 +0000 (18:05 +0000)]
GN: Enable libvpx, add link target and convert some test targets

Libvpx now supports GN and this CL turns on compiling it.
I also introduced an executable target 'webrtc_tests'
that depends on all in WeBRTC + tests in order to get a full
linking step executed (since we've seen link problems for GN
when rolling WebRTC into Chromium).

I also converted a few test targets and made a GN file for
third_party/gflags.

BUG=3441
TESTED=Trybots + full Chromium build with a symlinked src/third_party/webrtc
dir to a workspace wit this CL applied.

R=brettw@chromium.org
TBR=niklas.enbom@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25569004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7344 4adac7df-926f-26a2-2b94-8c16560cd09d

36 hours agoChanged mips_arch_variant variable value corresponding to Chromium code changes.
andrew@webrtc.org [Tue, 30 Sep 2014 15:53:24 +0000 (15:53 +0000)]
Changed mips_arch_variant variable value corresponding to Chromium code changes.

Chromium commit URL: https://crrev.com/c8a5da7455b57b2399e4a69e8100c098d9870052

R=andrew@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/23809004

Patch from Ljubomir Papuga <lpapuga@mips.com>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7343 4adac7df-926f-26a2-2b94-8c16560cd09d

36 hours agoRevert 7337 "Reland 28629004: adding new AEC dump start interfac..."
xians@webrtc.org [Tue, 30 Sep 2014 15:29:13 +0000 (15:29 +0000)]
Revert 7337 "Reland 28629004: adding new AEC dump start interfac..."

> Reland 28629004: adding new AEC dump start interface for chrome
>
> adding new AEC dump start interface for chrome.
>
> This is required because we are not allow to pass CRT objects across dll boundaries, that says, when we pass a file descriptor from chrome dll to libpeerconnection dll, the file descriptor will become invalid immediate, more information can be found here:
> http://msdn.microsoft.com/en-us/library/ms235460.aspx
>
> Chromium bug:crbug/415935
> TEST=bots
> R=bjornv@webrtc.org, kwiberg@webrtc.org
>
> Review URL: https://webrtc-codereview.appspot.com/27639004

TBR=xians@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22849004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7342 4adac7df-926f-26a2-2b94-8c16560cd09d

36 hours agoRevert 7338 "Fixed the android build by making the interface pur..."
xians@webrtc.org [Tue, 30 Sep 2014 15:26:15 +0000 (15:26 +0000)]
Revert 7338 "Fixed the android build by making the interface pur..."

> Fixed the android build by making the interface pure virtual.
>
> TBR=asapersson@webrtc.org, bjornv@webrtc.org,
>
> Review URL: https://webrtc-codereview.appspot.com/24789004

TBR=xians@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30579004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7341 4adac7df-926f-26a2-2b94-8c16560cd09d

37 hours agoCollecting stats every fixed time in webrtc_video_streaming.js test
houssainy@google.com [Tue, 30 Sep 2014 15:20:15 +0000 (15:20 +0000)]
Collecting stats every fixed time in webrtc_video_streaming.js test
and prepare the format these collected stats to be plotted using one of
external dev-tools.

R=andresp@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29589004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7340 4adac7df-926f-26a2-2b94-8c16560cd09d

37 hours agoMinor code change to fix some warnings in MIPS build.
andrew@webrtc.org [Tue, 30 Sep 2014 15:17:50 +0000 (15:17 +0000)]
Minor code change to fix some warnings in MIPS build.

R=andrew@webrtc.org, bjornv@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26619004

Patch from Ljubomir Papuga <lpapuga@mips.com>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7339 4adac7df-926f-26a2-2b94-8c16560cd09d

37 hours agoFixed the android build by making the interface pure virtual.
xians@webrtc.org [Tue, 30 Sep 2014 15:15:22 +0000 (15:15 +0000)]
Fixed the android build by making the interface pure virtual.

TBR=asapersson@webrtc.org, bjornv@webrtc.org,

Review URL: https://webrtc-codereview.appspot.com/24789004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7338 4adac7df-926f-26a2-2b94-8c16560cd09d

37 hours agoReland 28629004: adding new AEC dump start interface for chrome
xians@webrtc.org [Tue, 30 Sep 2014 14:35:15 +0000 (14:35 +0000)]
Reland 28629004: adding new AEC dump start interface for chrome

adding new AEC dump start interface for chrome.

This is required because we are not allow to pass CRT objects across dll boundaries, that says, when we pass a file descriptor from chrome dll to libpeerconnection dll, the file descriptor will become invalid immediate, more information can be found here:
http://msdn.microsoft.com/en-us/library/ms235460.aspx

Chromium bug:crbug/415935
TEST=bots
R=bjornv@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/27639004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7337 4adac7df-926f-26a2-2b94-8c16560cd09d

38 hours agoAdds isolate for rtc_unittests and moves sound's unittests to rtc_unittest.
henrike@webrtc.org [Tue, 30 Sep 2014 14:21:10 +0000 (14:21 +0000)]
Adds isolate for rtc_unittests and moves sound's unittests to rtc_unittest.

BUG=N/A
R=andrew@webrtc.org, kjellander@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31499004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7336 4adac7df-926f-26a2-2b94-8c16560cd09d

38 hours agoRevert 7334 "adding new AEC dump start interface for chrome."
xians@webrtc.org [Tue, 30 Sep 2014 13:30:05 +0000 (13:30 +0000)]
Revert 7334 "adding new AEC dump start interface for chrome."

> adding new AEC dump start interface for chrome.
>
> This is required because we are not allow to pass CRT objects across dll boundaries, that says, when we pass a file descriptor from chrome dll to libpeerconnection dll, the file descriptor will become invalid immediate, more information can be found here:
> http://msdn.microsoft.com/en-us/library/ms235460.aspx
>
> Chromium bug:crbug/415935
> TEST=bots
> R=bjornv@webrtc.org, kwiberg@webrtc.org
>
> Review URL: https://webrtc-codereview.appspot.com/28629004

TBR=xians@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26659004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7335 4adac7df-926f-26a2-2b94-8c16560cd09d

39 hours agoadding new AEC dump start interface for chrome.
xians@webrtc.org [Tue, 30 Sep 2014 13:11:27 +0000 (13:11 +0000)]
adding new AEC dump start interface for chrome.

This is required because we are not allow to pass CRT objects across dll boundaries, that says, when we pass a file descriptor from chrome dll to libpeerconnection dll, the file descriptor will become invalid immediate, more information can be found here:
http://msdn.microsoft.com/en-us/library/ms235460.aspx

Chromium bug:crbug/415935
TEST=bots
R=bjornv@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/28629004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7334 4adac7df-926f-26a2-2b94-8c16560cd09d

41 hours agoMinor modifications to test::RtpFileReader
henrik.lundin@webrtc.org [Tue, 30 Sep 2014 11:08:44 +0000 (11:08 +0000)]
Minor modifications to test::RtpFileReader

Adding original_length to the Packet struct. This is populated with
the plen value from the RTP dump file. In the case of reading a
pcap file, original_length will be equal to length.

Also increasing the maximum packet size to 3500 bytes. This is to
accomodate some test files that contain PCM16b audio encoding.

R=pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/28609004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7333 4adac7df-926f-26a2-2b94-8c16560cd09d

42 hours agoAdd default implementation of Add/RemoveObserver.
pbos@webrtc.org [Tue, 30 Sep 2014 09:45:25 +0000 (09:45 +0000)]
Add default implementation of Add/RemoveObserver.

Needed to remove Add/RemoveObserver from RTCVideoEncoderFactory in
Chromium before removing these completely. This is done to keep the
chromium.webrtc.fyi bots happy and to make rolling webrtc revisions
easier.

R=stefan@webrtc.org
BUG=3876

Review URL: https://webrtc-codereview.appspot.com/23839004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7332 4adac7df-926f-26a2-2b94-8c16560cd09d

42 hours agoaudio_processing/aecm: Added help function for calculating log of energy
bjornv@webrtc.org [Tue, 30 Sep 2014 09:31:28 +0000 (09:31 +0000)]
audio_processing/aecm: Added help function for calculating log of energy

The same operation of calculating log of the energy was executed four times. This CL adds a help function LogOfEnergyInQ8() for that.

BUG=N/A
TESTED=locally on linux and trybots
R=kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22819004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7331 4adac7df-926f-26a2-2b94-8c16560cd09d

42 hours agoaudio_processing: Removed usage of macro WEBRTC_SPL_MUL
bjornv@webrtc.org [Tue, 30 Sep 2014 09:29:28 +0000 (09:29 +0000)]
audio_processing: Removed usage of macro WEBRTC_SPL_MUL

WEBRTC_SPL_MUL is a trivial multiplication after casting to int32_t. This is already taken care of by the compiler which makes the macro unnecessary.

Affected components:
* AGC
* NSx

BUG=3348,3353
TESTED=locally on linux and trybots
R=kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25429004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7330 4adac7df-926f-26a2-2b94-8c16560cd09d

42 hours agoaudio_processing: Replaced trivial macro WEBRTC_SPL_LSHIFT_W32 with <<
bjornv@webrtc.org [Tue, 30 Sep 2014 09:26:36 +0000 (09:26 +0000)]
audio_processing: Replaced trivial macro WEBRTC_SPL_LSHIFT_W32 with <<

Affected components:
* AECM
* AGC
* HPF
* NSx

BUG=3348,3353
TESTED=locally on linux and trybots
R=kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/27629004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7329 4adac7df-926f-26a2-2b94-8c16560cd09d

43 hours agoRevert 7327 "Update isolate.gypi files + link to isolate_driver.py"
kjellander@webrtc.org [Tue, 30 Sep 2014 08:44:00 +0000 (08:44 +0000)]
Revert 7327 "Update isolate.gypi files + link to isolate_driver.py"

Breaks debug compilation (didn't run all trybots when testing this).

> Update isolate.gypi files + link to isolate_driver.py
>
> This updates the isolate.gypi copies we're forced to
> maintain in our code repo to Chromium revision c264a05.
>
> Since isolated testing is now using a new launch script
> in tools: isolate_driver.py, that is added to our links
> script.
>
> BUG=395700
> TESTED=Ran one of our tests with:
> ninja -C out/Release tools_unittests_run
> tools/isolate_driver.py run --isolated out/Release/tools_unittests.isolated --isolate webrtc/tools/tools_unittests.isolate
>
> R=henrika@webrtc.org, jam@chromium.org
>
> Review URL: https://webrtc-codereview.appspot.com/26649004

TBR=kjellander@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31509004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7328 4adac7df-926f-26a2-2b94-8c16560cd09d

43 hours agoUpdate isolate.gypi files + link to isolate_driver.py
kjellander@webrtc.org [Tue, 30 Sep 2014 08:34:57 +0000 (08:34 +0000)]
Update isolate.gypi files + link to isolate_driver.py

This updates the isolate.gypi copies we're forced to
maintain in our code repo to Chromium revision c264a05.

Since isolated testing is now using a new launch script
in tools: isolate_driver.py, that is added to our links
script.

BUG=395700
TESTED=Ran one of our tests with:
ninja -C out/Release tools_unittests_run
tools/isolate_driver.py run --isolated out/Release/tools_unittests.isolated --isolate webrtc/tools/tools_unittests.isolate

R=henrika@webrtc.org, jam@chromium.org

Review URL: https://webrtc-codereview.appspot.com/26649004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7327 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoAllow Android apps to set video renderer scaling type.
glaznev@webrtc.org [Mon, 29 Sep 2014 23:07:08 +0000 (23:07 +0000)]
Allow Android apps to set video renderer scaling type.
Also add type check for EGL context object received from apps and
switch to byte buffer video decoding if EGL context is incorrect

BUG=3851
R=tkchin@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25669004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7326 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoReland disallowing blocking calls on the worker thread.
jiayl@webrtc.org [Mon, 29 Sep 2014 22:45:55 +0000 (22:45 +0000)]
Reland disallowing blocking calls on the worker thread.
This fixed the issue that invoking the call when the thread is not started.

BUG=3559
R=juberti@webrtc.org, thorcarpenter@google.com

Review URL: https://webrtc-codereview.appspot.com/24769004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7325 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoSet thread scheduling parameters inside the new thread.
henrike@webrtc.org [Mon, 29 Sep 2014 18:25:27 +0000 (18:25 +0000)]
Set thread scheduling parameters inside the new thread.

This makes it possible to restrict threads from modifying scheduling
parameters of another thread in the Chrome Linux sandbox.

BUG=
R=henrike@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/28539004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7324 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoDisable flaky tests:
asapersson@webrtc.org [Mon, 29 Sep 2014 14:30:07 +0000 (14:30 +0000)]
Disable flaky tests:
JsepPeerConnectionP2PTestClient.ReceivedBweStatsCombined
JsepPeerConnectionP2PTestClient.ReceivedBweStatsNotCombined

BUG=3871
R=henrike@webrtc.org, kjellander@webrtc.org, pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25709004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7323 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoFix parallel test execution for tools, testsupport and metrics tests.
kjellander@webrtc.org [Mon, 29 Sep 2014 11:47:28 +0000 (11:47 +0000)]
Fix parallel test execution for tools, testsupport and metrics tests.

BUG=2600
TESTED=Passing tests using:
python third_party/gtest-parallel/gtest-parallel -w 10 -r 20 out/Release/test_support_unittests
python third_party/gtest-parallel/gtest-parallel -w 10 -r 20 out/Release/tools_unittests
python third_party/gtest-parallel/gtest-parallel -w 10 -r 20 out/Release/video_engine_tests

R=pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31479004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7322 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoaudio_processing: Replaced macro WEBRTC_SPL_LSHIFT_W16 with <<
bjornv@webrtc.org [Mon, 29 Sep 2014 10:56:27 +0000 (10:56 +0000)]
audio_processing: Replaced macro WEBRTC_SPL_LSHIFT_W16 with <<

A trivial macro that serves no purpose. Affected components are:
* audio_processing/nsx
* audio_processing/agc
* audio_processing/aecm
* common_audio/LpcToReflCoef

BUG=3348,3353
TESTED=locally on linux
R=kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22539004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7321 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agocommon_audio refactoring: Removed macro WEBRTC_SPL_LSHIFT_U32
bjornv@webrtc.org [Mon, 29 Sep 2014 09:40:38 +0000 (09:40 +0000)]
common_audio refactoring: Removed macro WEBRTC_SPL_LSHIFT_U32

The macro is a trivial shift operator including a cast before shift. There is no guard against negative shifts. Replaced with << at place and added casts when necessary.

Affects both fixed and float point versions of iSAC

BUG=3348,3353
TESTED=locally on linux and trybots
R=kwiberg@webrtc.org, tina.legrand@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/27369004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7320 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoAdding getStats function to the exposed PeerConnection in RtcBot
houssainy@google.com [Mon, 29 Sep 2014 09:36:28 +0000 (09:36 +0000)]
Adding getStats function to the exposed PeerConnection in RtcBot

Exposed Peerconnection object has new function "getStats". This function
returns the stats as array of reports, and each report is RTCStatReport
with additional attributes names and stats.

names: array of all the stat names in current report.
Stats: dictionary and the key is the stat name, and value is the value
of this stat.

R=andresp@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/23779004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7319 4adac7df-926f-26a2-2b94-8c16560cd09d

2 days agoRemove callback from RtpDepacketizer::Parse().
pbos@webrtc.org [Mon, 29 Sep 2014 08:00:22 +0000 (08:00 +0000)]
Remove callback from RtpDepacketizer::Parse().

BUG=
R=pbos@webrtc.org, stefan@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30489004

Patch from Changbin Shao <changbin.shao@intel.com>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7318 4adac7df-926f-26a2-2b94-8c16560cd09d

3 days agoGN: Add common configs to all targets.
kjellander@webrtc.org [Sun, 28 Sep 2014 17:37:22 +0000 (17:37 +0000)]
GN: Add common configs to all targets.

This is needed to ensure we have the same build with GN
as with GYP, since GYP includes the common.gypi on a global level.
Several fixes has been needed in the past because some code have
been built without the right defines.

BUG=3441
R=brettw@chromium.org

Review URL: https://webrtc-codereview.appspot.com/28589004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7317 4adac7df-926f-26a2-2b94-8c16560cd09d

3 days agoInitialize SSL in unittest_main.cc.
pbos@webrtc.org [Sun, 28 Sep 2014 11:36:45 +0000 (11:36 +0000)]
Initialize SSL in unittest_main.cc.

Instead of having each test individually initialize and tear down SSL
move this to unittest_main.cc so that all tests are properly
initialized and new tests "don't have to think about it".

R=pthatcher@webrtc.org
BUG=

Review URL: https://webrtc-codereview.appspot.com/30549004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7316 4adac7df-926f-26a2-2b94-8c16560cd09d

3 days agoRoll chromium_revision deaf2f7e..c264a056 (295079:297113)
kjellander@webrtc.org [Sun, 28 Sep 2014 10:33:45 +0000 (10:33 +0000)]
Roll chromium_revision deaf2f7e..c264a056 (295079:297113)

Summary of changes (git diff deaf2f7e..c264a056 DEPS):
* buildtools ea4dc0e..56bc51a
* third_party/android_tools 7fc902d..bbafe51
* third_party/boringssl a70c75c..01fe820
* third_party/icu 8983113..d2abf6c1
* third_party/libsrtp dcc1fc62..98284c8
* third_party/libvpx 4947d55..efe9712d
* third_party/nss 7b5b6ec4..87b96db
* tools/swarming_client 14b5fd8..79940aee

BUG=
TBR=pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29579004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7315 4adac7df-926f-26a2-2b94-8c16560cd09d

4 days agoCleanup .gclient.bot_entries to avoid sync problems on bots.
kjellander@webrtc.org [Sat, 27 Sep 2014 18:41:03 +0000 (18:41 +0000)]
Cleanup .gclient.bot_entries to avoid sync problems on bots.

In https://webrtc-codereview.appspot.com/28509004 the buildbot
case was missed since they get a gclient entries file named
.gclient.bot_entries instead of the regular .gclient_entries
file due to the way the sync was implemented for buildbots.

This change makes the right file get wiped in the two cases.

BUG=
TBR=agable@chromium.org

Review URL: https://webrtc-codereview.appspot.com/28599004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7314 4adac7df-926f-26a2-2b94-8c16560cd09d

4 days agoRoll chromium_revision 6455c69..deaf2f7 (293954:295079)
kjellander@webrtc.org [Sat, 27 Sep 2014 18:10:30 +0000 (18:10 +0000)]
Roll chromium_revision 6455c69..deaf2f7 (293954:295079)

Mainly to pick up recent yasm changes needed for
turning on libvpx in GN.

Summary of changes (git diff 6455c69..deaf2f7 DEPS):
* third_party/boringssl 7bdec13..a70c75c
* third_party/libjpeg_turbo 3963fbc..034e9a9
* third_party/libvpx d95585f..4947d55
* tools/gyp 1972:1977

I had to add src/third_party/junit/src to be ignored
due to http://crbug.com/417292.

BUG=3855, chromium:417292
TBR=pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/27539004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7313 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoFix the duplicated candidate problem when using multiple STUN servers.
jiayl@webrtc.org [Fri, 26 Sep 2014 23:01:11 +0000 (23:01 +0000)]
Fix the duplicated candidate problem when using multiple STUN servers.

BUG=3723
R=pthatcher@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26469004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7312 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoGetting orientation is not working properly. VideoCaptureImpl::RotationFromDegrees...
braveyao@webrtc.org [Fri, 26 Sep 2014 22:50:06 +0000 (22:50 +0000)]
Getting orientation is not working properly. VideoCaptureImpl::RotationFromDegrees returns -1 in case fails not 0. So we need to change the if statement.

BUG=3869
TEST=Manual
R=glaznev@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25699004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7311 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoBuild one of NSS or BoringSSL but not both.
pthatcher@webrtc.org [Fri, 26 Sep 2014 18:53:40 +0000 (18:53 +0000)]
Build one of NSS or BoringSSL but not both.

The libraries have some common symbols. When both are linked I observed NSS
SHA1_Update called followed by BoringSSL SHA1_Final, which results in a
segfault. We should only link one of these.

Based off of https://review.webrtc.org/25689004/

BUG=3855
R=pthatcher@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31469004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7310 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoReverting part of
thorcarpenter@google.com [Fri, 26 Sep 2014 17:19:14 +0000 (17:19 +0000)]
Reverting part of
https://webrtc-codereview.appspot.com/15089004/diff/140001/talk/session/media/channelmanager.cc?context=10&column_width=80
because of a major regression hanging the executable on start.

R=jiayl@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22799004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7309 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoDo not assert for blocking call allowed in Thread::Join.
jiayl@webrtc.org [Fri, 26 Sep 2014 16:57:07 +0000 (16:57 +0000)]
Do not assert for blocking call allowed in Thread::Join.
We do not allow blocking call from the worker thread, but on Android the worker thread may stop/join a SignalThread, which hits the assert.
AssertBlockingIsAllowedOnCurrentThread is used to make sure a thread does not do Invoke, so check that in Thread::Join does not seem to add much value.

BUG=3857
R=juberti@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24709004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7308 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoRemove the different block lengths in ns_core
aluebs@webrtc.org [Fri, 26 Sep 2014 14:41:19 +0000 (14:41 +0000)]
Remove the different block lengths in ns_core

Relanding the CL: https://webrtc-codereview.appspot.com/30539004/
It had to be reverted because some development code was uploaded by mistake.

TBR=bjornv@webrtc.org

BUG=webrtc:3811

Review URL: https://webrtc-codereview.appspot.com/28589005

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7307 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoRevert 7297 "Remove the different block lengths in ns_core"
aluebs@webrtc.org [Fri, 26 Sep 2014 14:33:08 +0000 (14:33 +0000)]
Revert 7297 "Remove the different block lengths in ns_core"

> Remove the different block lengths in ns_core
>
> This CL has bit-exact output.
>
> What it does:
>   * Remove the blockLen10Ms, as it is hardcoded to be equal to blockLen.
>   * This makes outLen to be always zero, so it can be removed too.
>   * It also avoids the need to have an outBuf, because it is not used, so it is also removed
>   * Replaced blockLen10Ms by blockLen everywhere, since they were hardcoded to be equal.
>   * We don't need to check if outLen is zero, because it always is, so it was removed.
>   * Of course, the outBuf needs no initial set or copying around, because it is not used.
>
> BUG=webrtc:3811
> R=bjornv@webrtc.org, kwiberg@webrtc.org
>
> Review URL: https://webrtc-codereview.appspot.com/30539004

TBR=aluebs@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26629004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7306 4adac7df-926f-26a2-2b94-8c16560cd09d

5 days agoMark virtual overrides of ViENetwork and VoENetwork as such.
henrikg@webrtc.org [Fri, 26 Sep 2014 11:09:08 +0000 (11:09 +0000)]
Mark virtual overrides of ViENetwork and VoENetwork as such.

This will make further changes to these classes safer by ensuring that the
compile breaks if the base class changes and not all overrides are fixed.

BUG=none
TEST=none
R=henrikg@webrtc.org, mflodman@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26439004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7305 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoRevert 7302 "Roll chromium revision: 6455c69:2687a76"
marpan@webrtc.org [Thu, 25 Sep 2014 22:20:11 +0000 (22:20 +0000)]
Revert 7302 "Roll chromium revision: 6455c69:2687a76"

> Roll chromium revision: 6455c69:2687a76
>
> Pick up the libvpx roll: https://codereview.chromium.org/597703004/
>
> R=johannkoenig@google.com
> TBR=ajm@google.com
>
> Review URL: https://webrtc-codereview.appspot.com/23789004

TBR=marpan@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22789004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7304 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoAdd accessors for array of channel pointers in AudioBuffer. They are
claguna@google.com [Thu, 25 Sep 2014 20:52:08 +0000 (20:52 +0000)]
Add accessors for array of channel pointers in AudioBuffer. They are
needed as arguments to any multichannel audio processing unit.

R=andrew@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30499004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7303 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoRoll chromium revision: 6455c69:2687a76
marpan@webrtc.org [Thu, 25 Sep 2014 20:28:08 +0000 (20:28 +0000)]
Roll chromium revision: 6455c69:2687a76

Pick up the libvpx roll: https://codereview.chromium.org/597703004/

R=johannkoenig@google.com
TBR=ajm@google.com

Review URL: https://webrtc-codereview.appspot.com/23789004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7302 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoCall SSL_shutdown in OpenSSLStreamAdapter::Cleanup.
jiayl@webrtc.org [Thu, 25 Sep 2014 16:38:46 +0000 (16:38 +0000)]
Call SSL_shutdown in OpenSSLStreamAdapter::Cleanup.

BUG=crbug/414211
R=juberti@webrtc.org

Committed: https://code.google.com/p/webrtc/source/detail?r=7293

Review URL: https://webrtc-codereview.appspot.com/22739004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7301 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoExplicitly initialize SSL for tests.
pbos@webrtc.org [Thu, 25 Sep 2014 15:50:26 +0000 (15:50 +0000)]
Explicitly initialize SSL for tests.

Adding missing SSL initialization/cleanups in
TransportDescriptionFactoryTest and MediaSessionTest.

These being missing prevent these tests from being run individually
without other tests preceding them that initialize SSL.

BUG=3860
R=pthatcher@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24739004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7300 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoBump to version 39
tnakamura@webrtc.org [Thu, 25 Sep 2014 15:28:20 +0000 (15:28 +0000)]
Bump to version 39

TBR=niklas.enbom

Review URL: https://webrtc-codereview.appspot.com/24749004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7299 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoRemoving error triggered for disabling FEC on non-opus
minyue@webrtc.org [Thu, 25 Sep 2014 14:36:30 +0000 (14:36 +0000)]
Removing error triggered for disabling FEC on non-opus

A failure was triggered when one sets FEC status on a codec that does not support FEC. While it is definitely logical when one wants to enable it, it makes no good sense if one tries to disable it.

BUG=
R=tina.legrand@webrtc.org, xians@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24729004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7298 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoRemove the different block lengths in ns_core
aluebs@webrtc.org [Thu, 25 Sep 2014 13:53:43 +0000 (13:53 +0000)]
Remove the different block lengths in ns_core

This CL has bit-exact output.

What it does:
  * Remove the blockLen10Ms, as it is hardcoded to be equal to blockLen.
  * This makes outLen to be always zero, so it can be removed too.
  * It also avoids the need to have an outBuf, because it is not used, so it is also removed
  * Replaced blockLen10Ms by blockLen everywhere, since they were hardcoded to be equal.
  * We don't need to check if outLen is zero, because it always is, so it was removed.
  * Of course, the outBuf needs no initial set or copying around, because it is not used.

BUG=webrtc:3811
R=bjornv@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30539004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7297 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoRevert r7049/r7123, which added unnecessary "u"s to "return 0"s.
henrik.lundin@webrtc.org [Thu, 25 Sep 2014 07:38:14 +0000 (07:38 +0000)]
Revert r7049/r7123, which added unnecessary "u"s to "return 0"s.

r7049 added some unnecessary casts ("return 0" -> "return static_cast<uint16_t>(0)").  r7123 converted these to "return 0u".  The original impetus for this was to eliminate type conversion warnings.  However, the 'u's are unnecessary; Visual Studio can return "0" from a function returning an unsigned value without producing a warning.  The real reason for the original warnings was that the code was returning -1 from a function returning an unsigned value, which does need a cast; the -1s were eliminated before the above two revisions landed.

Also reverse the order of some conditionals to prevent possible underflow.
While the underflow wouldn't have changed the behavior of the code, it's easier
to reason about the code when such underflow can't happen, and possibly safer
against future modifications as well.

BUG=3663
TEST=none
R=tina.legrand@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22599004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7296 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoFix typo from RtpPacketizerH264.
pbos@webrtc.org [Thu, 25 Sep 2014 07:31:42 +0000 (07:31 +0000)]
Fix typo from RtpPacketizerH264.

BUG=
R=pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/27609004

Patch from Changbin Shao <changbin.shao@intel.com>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7295 4adac7df-926f-26a2-2b94-8c16560cd09d

6 days agoRevert "Call SSL_shutdown in OpenSSLStreamAdapter::Cleanup." (rev 7293).
andresp@webrtc.org [Thu, 25 Sep 2014 07:30:14 +0000 (07:30 +0000)]
Revert "Call SSL_shutdown in OpenSSLStreamAdapter::Cleanup." (rev 7293).

Breaks windows bot as it was already showing on the try jobs on the

BUG=crbug/414211
R=jiayl@webrtc.org,juberti@webrtc.org
TBR=jiayl@webrtc.org,juberti@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26599004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7294 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoCall SSL_shutdown in OpenSSLStreamAdapter::Cleanup.
jiayl@webrtc.org [Wed, 24 Sep 2014 21:13:39 +0000 (21:13 +0000)]
Call SSL_shutdown in OpenSSLStreamAdapter::Cleanup.

BUG=crbug/414211
R=juberti@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22739004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7293 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoEnable render downmixing to mono in AudioProcessing.
andrew@webrtc.org [Wed, 24 Sep 2014 20:06:23 +0000 (20:06 +0000)]
Enable render downmixing to mono in AudioProcessing.

In practice, we have been doing this since time immemorial, but have
relied on the user to do the downmixing (first voice engine then
Chromium). It's more logical for this burden to fall on AudioProcessing,
however, who can be expected to know that this is a reasonable approach
for AEC. Permitting two render channels results in running two AECs
serially.

Critically, in my recent change to have Chromium adopt the float
interface:
https://codereview.chromium.org/420603004
I removed the downmixing by Chromium, forgetting that we hadn't yet
enabled this feature in AudioProcessing. This corrects that oversight.

The change in paths hit by production users is very minor. As commented
it required adding downmixing to the int16_t path to satisfy
bit-exactness tests.

For reference, find the ApmTest.Process errors here:
https://paste.googleplex.com/6372007910309888

BUG=webrtc:3853
TESTED=listened to the files output from the Process test, and verified
that they sound as expected: higher echo while the AEC is adapting, but
afterwards very close.

R=aluebs@webrtc.org, bjornv@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31459004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7292 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoAdd missing DesktopConfigurationMonitor Unlock in webrtc::ScreenCapturerMac
jiayl@webrtc.org [Wed, 24 Sep 2014 17:23:46 +0000 (17:23 +0000)]
Add missing DesktopConfigurationMonitor Unlock in webrtc::ScreenCapturerMac

BUG=3837
R=jiayl@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30469004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7291 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoFix a problem in Thread::Send.
jiayl@webrtc.org [Wed, 24 Sep 2014 17:14:05 +0000 (17:14 +0000)]
Fix a problem in Thread::Send.
Previously if thread A->Send is called on thread B, B->ReceiveSends will be called, which enables an arbitrary thread to invoke calls on B while B is wait for A->Send to return. This caused mutliple problems like issue 3559, 3579.
The fix is to limit B->ReceiveSends to only process requests from A.
Also disallow the worker thread invoking other threads.

BUG=3559
R=juberti@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/15089004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7290 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoCall NS AnalyzeCaptureAudio before AEC
aluebs@webrtc.org [Wed, 24 Sep 2014 14:18:03 +0000 (14:18 +0000)]
Call NS AnalyzeCaptureAudio before AEC

This attenuates the noise pumping generated from the NS adapting to the AEC comfort noise.

When there is echo present the AEC suppresses it and adds comfort noise. This is underestimated on purpose to avoid adding more than the original background noise. The NS has to be called after the AEC, because every non-linear processing before it can ruin its performance. Therefore the noise estimation can adapt to this comfort noise, making it less aggressive and generating noise pumping.

By putting the noise estimation analysis stage from the NS before the AEC, this effect can be avoided. This has been tested manually on recordings where noise pumping was present: Two long recordings done in our team by bjornv and kwiberg plus the most noisy (5) recordings in the QA set.

On the other hand, one risk of doing this is to not adapt to the comfort noise and therefore suppress too much. As verified in the tested files, this is not a problem in practice.

BUG=webrtc:3763
R=andrew@webrtc.org, bjornv@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24679004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7289 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoReduce jitter delay for low fps streams.
sprang@webrtc.org [Wed, 24 Sep 2014 14:06:56 +0000 (14:06 +0000)]
Reduce jitter delay for low fps streams.
Enabled by finch flag.

BUG=
R=stefan@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31389005

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7288 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoMoved the filter calculation from analyze to process in ns_core
aluebs@webrtc.org [Wed, 24 Sep 2014 13:23:49 +0000 (13:23 +0000)]
Moved the filter calculation from analyze to process in ns_core

It makes sense to have it there if the analyze and process methods are called in different stages.
Tested over the entire QA set for bit exactness.

BUG=webrtc:3811
R=bjornv@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29549004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7287 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoaudioproc: Now also writes to output file in simulation mode
bjornv@webrtc.org [Wed, 24 Sep 2014 12:21:51 +0000 (12:21 +0000)]
audioproc: Now also writes to output file in simulation mode

After changing to use wav as default file format no output was written in simulation mode.

BUG=3359
TESTED=locally
R=aluebs@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25639004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7286 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoWebRtcIsac_Encode and WebRtcIsacfix_Encode: Type encoded stream as uint8_t
kwiberg@webrtc.org [Wed, 24 Sep 2014 10:31:02 +0000 (10:31 +0000)]
WebRtcIsac_Encode and WebRtcIsacfix_Encode: Type encoded stream as uint8_t

We have to fix both at once, since there's a macro that calls one of
them or the other.

BUG=909
R=andrew@webrtc.org, bjornv@webrtc.org, henrik.lundin@webrtc.org, minyue@webrtc.org

Committed: https://code.google.com/p/webrtc/source/detail?r=7266

Review URL: https://webrtc-codereview.appspot.com/19229004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7285 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoThread annotation of rtc::CriticalSection.
pbos@webrtc.org [Wed, 24 Sep 2014 07:10:57 +0000 (07:10 +0000)]
Thread annotation of rtc::CriticalSection.

Effectively re-lands r5516 which was reverted because talk/-only
checkouts existed. This now resides in webrtc/base/, so no talk/-only
checkouts should be possible.

This change also enables -Wthread-safety for talk/ and fixes a bug in
talk/media/webrtc/webrtcvideoengine2.cc where a guarded variable was
read without taking the corresponding lock.

R=andresp@webrtc.org, mflodman@webrtc.org, pthatcher@webrtc.org
BUG=

Review URL: https://webrtc-codereview.appspot.com/27569004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7284 4adac7df-926f-26a2-2b94-8c16560cd09d

7 days agoMove thread_annotations.h to webrtc/base/.
pbos@webrtc.org [Wed, 24 Sep 2014 06:05:00 +0000 (06:05 +0000)]
Move thread_annotations.h to webrtc/base/.

R=andresp@webrtc.org, mflodman@webrtc.org
BUG=

Review URL: https://webrtc-codereview.appspot.com/27579004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7283 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoChange Android video renderer to maintain video aspect
glaznev@webrtc.org [Tue, 23 Sep 2014 23:58:52 +0000 (23:58 +0000)]
Change Android video renderer to maintain video aspect
ratio when displaying camera or decoded video frames.

-

R=tkchin@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30509004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7282 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoSwitch HW video decoder to output byte buffers if video
glaznev@webrtc.org [Tue, 23 Sep 2014 21:42:15 +0000 (21:42 +0000)]
Switch HW video decoder to output byte buffers if video
renderer EGL context is not provided by app.

R=tkchin@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24699004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7281 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days ago(Auto)update libjingle 76169599-> 76176062
buildbot@webrtc.org [Tue, 23 Sep 2014 17:41:48 +0000 (17:41 +0000)]
(Auto)update libjingle 76169599-> 76176062

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7280 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoUse VPX_IMG_FMT_*/VPX_PLANE_* defines
johannkoenig@google.com [Tue, 23 Sep 2014 17:31:47 +0000 (17:31 +0000)]
Use VPX_IMG_FMT_*/VPX_PLANE_* defines

The compatibility layer has been removed upstream:
https://gerrit.chromium.org/gerrit/gitweb?p=webm%2Flibvpx.git;a=commit;h=9cdaa3d72eade9ad162ef8f78a93bd8f85c6de10

BUG=webrtc:3839
R=marpan@google.com, marpan@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22729004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7279 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoEnable ipv6 by default for webrtc under a Finch experiment.
guoweis@webrtc.org [Tue, 23 Sep 2014 16:23:02 +0000 (16:23 +0000)]
Enable ipv6 by default for webrtc under a Finch experiment.

Reapply 23529005 after fixing the build break issue (Chromium:582133002)

Committed: https://code.google.com/p/webrtc/source/detail?r=7253

Review URL: https://webrtc-codereview.appspot.com/23529005

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7278 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoRevert "Remove DTMF status methods from Voice Engine" r7276
henrik.lundin@webrtc.org [Tue, 23 Sep 2014 13:15:14 +0000 (13:15 +0000)]
Revert "Remove DTMF status methods from Voice Engine" r7276

This change caused some trouble.

TBR=henrika@webrtc.org,pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/29569004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7277 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoRemove DTMF status methods from Voice Engine
henrik.lundin@webrtc.org [Tue, 23 Sep 2014 12:54:04 +0000 (12:54 +0000)]
Remove DTMF status methods from Voice Engine

These methods are not used.

R=henrika@webrtc.org, pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24689004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7276 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoRevert "Set minimum SDK level to 10.7 for Mac and iOS" (r7175)
kjellander@webrtc.org [Tue, 23 Sep 2014 12:43:14 +0000 (12:43 +0000)]
Revert "Set minimum SDK level to 10.7 for Mac and iOS" (r7175)

Reverting this since it didn't fix the build failures.
We ended up passing mac_sdk=10.9 in GYP_DEFINES on the bots
to to make the build pass again
(https://codereview.chromium.org/573673002).

BUG=3120
R=mcasas@webrtc.org, tommi@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/24669004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7275 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agogn: Hide modules/video_capture:video_capture_internal_impl behind an arg
pbos@webrtc.org [Tue, 23 Sep 2014 12:37:06 +0000 (12:37 +0000)]
gn: Hide modules/video_capture:video_capture_internal_impl behind an arg

R=andresp@webrtc.org, brettw@chromium.org, kjellander@webrtc.org, pbos@webrtc.org, brettw

Review URL: https://webrtc-codereview.appspot.com/30479004

Patch from Cem Kocagil <ckocagil@chromium.org>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7274 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoReland "Converting five tests to use new AudioCoding interface" (r7258)
henrik.lundin@webrtc.org [Tue, 23 Sep 2014 12:05:34 +0000 (12:05 +0000)]
Reland "Converting five tests to use new AudioCoding interface" (r7258)

This CL reverts r7264. The problem was that iSAC-SWB and iSAC-FB are
not supported on android. These are now disabled.

BUG=3520
R=kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/23739004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7273 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoReland (rev 7259) "Convert AcmReceiverTest to new AudioCoding interface"
andresp@webrtc.org [Tue, 23 Sep 2014 11:37:57 +0000 (11:37 +0000)]
Reland (rev 7259) "Convert AcmReceiverTest to new AudioCoding interface"

Was reverted by mistake in 7260. Actual culprit was 7258.

BUG=3520
R=henrik.lundin@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/22719004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7272 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoaudio_processing/agc: Solved building with AGC_DEBUG + few style changes
bjornv@webrtc.org [Tue, 23 Sep 2014 11:21:39 +0000 (11:21 +0000)]
audio_processing/agc: Solved building with AGC_DEBUG + few style changes

webrtc did not build if AGC_DEBUG was turned on. This CL fixes that. Has no impact on performance since it is development/debug code.

* Name change to WEBRT_AGC_DEBUG_DUMP
* Added build flag agc_debug_dump to .gypi
* Added missing "%d" in printf at two places
* Some line length related style changes

Tested audioproc and modules_unittests with GYP_DEFINES=agc_debug_dump=1 webrtc/build/gyp_webrtc

BUG=N/A
TESTED=locally and trybots
R=aluebs@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31429004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7271 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoSkeleton for registering external encoders/decoders.
pbos@webrtc.org [Tue, 23 Sep 2014 09:40:22 +0000 (09:40 +0000)]
Skeleton for registering external encoders/decoders.

R=pthatcher@webrtc.org
BUG=1788

Review URL: https://webrtc-codereview.appspot.com/31429005

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7270 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agoUnit tests for SSLAdapter
tkchin@webrtc.org [Tue, 23 Sep 2014 05:56:44 +0000 (05:56 +0000)]
Unit tests for SSLAdapter

R=juberti@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/17309004

Patch from Manish Jethani <manish.jethani@gmail.com>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7269 4adac7df-926f-26a2-2b94-8c16560cd09d

8 days agomodules_unittests: Turned on ApmTest.Process test for Android
bjornv@webrtc.org [Tue, 23 Sep 2014 05:03:44 +0000 (05:03 +0000)]
modules_unittests: Turned on ApmTest.Process test for Android

The reason why ApmTest.Process breaks on Android is that two metrics over counts. I decided to add an offset and a different slack to the EXPECT_NEAR() calls that are affected. I think this is a reasonable approach since we have no more than two failing metrics. If any feature change that will make another metric fail, we should go back to the desk and find another way of solving this.

BUG=114
TESTED=locally on Nexus 7 and trybots
R=aluebs@webrtc.org, andrew@webrtc.org, kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26509004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7268 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoRevert 7266 "WebRtcIsac_Encode and WebRtcIsacfix_Encode: Type en..."
andrew@webrtc.org [Tue, 23 Sep 2014 01:32:57 +0000 (01:32 +0000)]
Revert 7266 "WebRtcIsac_Encode and WebRtcIsacfix_Encode: Type en..."

This was causing apparently legitimate failures on the following bots:
http://chromegw/i/client.webrtc/builders/Linux64%20Release%20%5Blarge%20tests%5D/builds/2599
http://chromegw/i/client.webrtc/builders/Android%20Tests%20%28KK%20Nexus5%29%28dbg%29/builds/2023
http://chromegw/i/client.webrtc/builders/Android%20Tests%20%28JB%20Nexus7.2%29%28dbg%29/builds/1825
http://chromegw/i/client.webrtc/builders/Android%20Tests%20%28KK%20Nexus5%29/builds/2013
http://chromegw/i/client.webrtc/builders/Android%20Tests%20%28JB%20Nexus7.2%29/builds/1795

> WebRtcIsac_Encode and WebRtcIsacfix_Encode: Type encoded stream as uint8_t
>
> We have to fix both at once, since there's a macro that calls one of
> them or the other.
>
> BUG=909
> R=andrew@webrtc.org, bjornv@webrtc.org, henrik.lundin@webrtc.org, minyue@webrtc.org
>
> Review URL: https://webrtc-codereview.appspot.com/19229004

TBR=kwiberg@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/30519004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7267 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoWebRtcIsac_Encode and WebRtcIsacfix_Encode: Type encoded stream as uint8_t
kwiberg@webrtc.org [Mon, 22 Sep 2014 23:04:14 +0000 (23:04 +0000)]
WebRtcIsac_Encode and WebRtcIsacfix_Encode: Type encoded stream as uint8_t

We have to fix both at once, since there's a macro that calls one of
them or the other.

BUG=909
R=andrew@webrtc.org, bjornv@webrtc.org, henrik.lundin@webrtc.org, minyue@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/19229004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7266 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoRemove engine-level SetOptions.
pbos@webrtc.org [Mon, 22 Sep 2014 16:07:18 +0000 (16:07 +0000)]
Remove engine-level SetOptions.

Already removed in WebRtcVideoEngine.

R=andresp@webrtc.org
BUG=1788

Review URL: https://webrtc-codereview.appspot.com/29549005

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7265 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoRevert "Converting five tests to use new AudioCoding interface" (rev 7258).
andresp@webrtc.org [Mon, 22 Sep 2014 15:49:56 +0000 (15:49 +0000)]
Revert "Converting five tests to use new AudioCoding interface" (rev 7258).

This time reverts the Cl that actually broke the tests. Got the wrong rev before. :/

BUG=3520
TESTED=Locally with CHECKOUT_SOURCE_ROOT=`pwd` build/android/test_runner.py gtest -s modules_unittests --gtest_filter=AcmReceiverBitExactness.8kHzOutput --verbose --isolate-file-path=webrtc/modules/modules_unittests.isolate
TBR=henrik.lundin@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26579004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7264 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoAdding test file path as argument of the rtcBot run command's arguments.
houssainy@google.com [Mon, 22 Sep 2014 15:24:56 +0000 (15:24 +0000)]
Adding test file path as argument of the rtcBot run command's arguments.

The new command to run rtcBot is:-
node test.js <bot_type> <test_file_path>

R=andresp@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/31419004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7263 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoRemove Get/SetNetEQPlayoutMode APIs
henrik.lundin@webrtc.org [Mon, 22 Sep 2014 14:30:10 +0000 (14:30 +0000)]
Remove Get/SetNetEQPlayoutMode APIs

These are not used anymore.

R=henrika@webrtc.org, pbos@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/26569004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7262 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoAdding webrtc_video_streaming test
houssainy@google.com [Mon, 22 Sep 2014 13:52:39 +0000 (13:52 +0000)]
Adding webrtc_video_streaming test
This test is streaming video and audio between two bots using webrtc js api.

R=andresp@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/28469004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7261 4adac7df-926f-26a2-2b94-8c16560cd09d

9 days agoRevert "Convert AcmReceiverTest to new AudioCoding interface" (rev 7258).
andresp@webrtc.org [Mon, 22 Sep 2014 13:18:34 +0000 (13:18 +0000)]
Revert "Convert AcmReceiverTest to new AudioCoding interface" (rev 7258).

Breaks android modules_unittests tests by crashing on AcmReceiverBitExactness.8kHzOutput
Was already visible on "git cl try" before submitting on https://webrtc-codereview.appspot.com/23719004/#

BUG=3520
R=kwiberg@webrtc.org, henrik.lundin@webrtc.org
TBR=kwiberg@webrtc.org, henrik.lundin@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/25629004

git-svn-id: http://webrtc.googlecode.com/svn/trunk@7260 4adac7df-926f-26a2-2b94-8c16560cd09d