Prevents bug where transmitted bitrate was reported as higher than what
was actually sent, since unused RTP modules weren't updated to say that
they sent zero.
BUG=
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/49979004
Cr-Commit-Position: refs/heads/master@{#9192}
The main purpose of this CL is to clean up RTCPSender::PrepareRTCP, but
it has quite a few ramifications. Notable changes:
* Removed the rtcpPacketTypeFlags bit vector and don't assume
RTCPPacketType values have a single unique bit set. This will allow
making this an enum class once rtcp_receiver has been overhauled.
* Flags are now stored in a map that is a member of the class. This
meant we could remove some bool flags (eg send_remb_) which was
previously masked into rtcpPacketTypeFlags and then masked out again
when testing if a remb packet should be sent.
* Make all build methods, eg. BuildREMB(), have the same signature.
An RtcpContext struct was introduced for this purpose. This allowed
the use of a map from RTCPPacketType to method pointer. Instead of
18 consecutive if-statements, there is now a single loop.
The context class also allowed some simplifications in the build
methods themselves.
* A few minor simplifications and cleanups.
The next step is to gradually replace the builder methods with the
builders from the new RtcpPacket classes.
BUG=2450
R=asapersson@webrtc.org, pbos@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/48329004
Cr-Commit-Position: refs/heads/master@{#9166}
Instantiates ProcessThread/ChannelGroup inside Call instead of using
VideoEngine or ViEBase. This removes the need for ViEChannelManager,
ViEInputManager and other ViESharedData completely.
Some interface headers are still referenced due to external interfaces
being defined there. Upon interface removal these will be moved to
implementation headers.
BUG=1695
R=mflodman@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/50849005
Cr-Commit-Position: refs/heads/master@{#9160}
Removes FixedSizeLockFreeQueue which isn't used anymore. This enabled
moving rtc::AtomicOps to webrtc/base/atomicops.h where they should be.
BUG=4330
R=tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/51789004
Cr-Commit-Position: refs/heads/master@{#9120}
Instead a ViECapturer object is allocated and directly operated on. This
additionally exposes ViESharedData to Call to access the module
ProcessThread, moving towards Call ownership of shared resources.
BUG=1695
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/45339004
Cr-Commit-Position: refs/heads/master@{#9119}
Replaces interface usage with direct calls on ViEEncoder removing a
layer of indirection. Also removing some methods from ViEImageProcess
that were only added for Video{Send,Receive}Stream usage.
BUG=1695
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/45319004
Cr-Commit-Position: refs/heads/master@{#9111}
This implementation registers RTX-APT map inside RTP sender and receiver.
While it only generates SDP with RTX associated with VP8 to make it
compatible with previous Chrome versions.
Should add following changes after reaches stable,
* Use RTX-APT map for building and restoring RTP packets.
* Add RTX support for RED or VP9 in Video engine.
* Set RTX payload type for RED inside FecConfig in EndToEndTest.
BUG=4024
R=mflodman@webrtc.org, pbos@webrtc.org, pthatcher@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/36889004
Cr-Commit-Position: refs/heads/master@{#9040}
Add separate functions for returning stats from send/receive stream and updated how functions are used.
Add test implementation for histogram methods in system_wrappers/interface/metrics.h.
BUG=4519
R=pbos@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/49639004
Cr-Commit-Position: refs/heads/master@{#9009}
For simulcast VideoSendStreams there are more than one entry in the SSRC
table causing RTCP to currently be delivered more than once per stream.
This messes up unique NACK stats as all NACK requests sent to such a
sender will be delivered multiple times and not look unique.
BUG=4544
R=asapersson@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/50709004
Cr-Commit-Position: refs/heads/master@{#8998}
It is just a pass through to webrtc/video_frame.h. Updated the callers
to include webrtc/video_frame.h instead and removed i420_video_frame.h.
This should fix pbos' TODO in i420_video_frame.h.
Tested on Linux with the following command lines:
$ rm -rf out/
$ ./webrtc/build/gyp_webrtc
$ ninja -C out/Debug
BUG=None
TEST=see above
R=magjed@webrtc.org, pbos@webrtc.org, tommi@webrtc.orgTBR=tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/46819004
Patch from Thiago Farina <tfarina@chromium.org>.
Cr-Commit-Position: refs/heads/master@{#8973}
Using the single stream bwe is really bad for the screenshare
test case in particular, but would probably help in other
cases as well so enabling it by default in CallTest setup.
BUG=
R=pbos@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/43089004
Cr-Commit-Position: refs/heads/master@{#8971}
All these tests crashed before r8811. These tests should've been with
that change but r8811 was pushed in before to make bots green.
BUG=1788, 1667
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/48669004
Cr-Commit-Position: refs/heads/master@{#8881}
These checks would help catching double-deletes, forgetting to destroy
streams and also catch if VideoEngine has held on to any stale
references.
BUG=1788
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/42929004
Cr-Commit-Position: refs/heads/master@{#8866}
This means all channels within the same group will share the same pacing queue and scheduler. It also means padding will be computed and sent by a single pacer. To accomplish this I also introduce a PacketRouter which finds the RTP module which owns the packet to be paced out.
BUG=4323
R=mflodman@webrtc.org, pbos@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/45549004
Cr-Commit-Position: refs/heads/master@{#8864}
Instead of setting bitrates based on codec target settings (which may
have previously been capped by a codec max bitrate), fetch the last
bandwidth allocated for this channel. This fixes broken low start bitrates
due to QCIF being set as default codec in WebRtcVideoEngine2 which caps
the max bitrate to 200kbps.
BUG=1788
R=mflodman@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/43789004
Cr-Commit-Position: refs/heads/master@{#8780}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8780 4adac7df-926f-26a2-2b94-8c16560cd09d
Original cl description:
This removes the none const pointer entry and SwapFrame.
Since frames delivered using VideoSendStream no longer use the external capture module, VideoSendStream will not get an incoming framerate callback. VideoSendStream now uses a rtc::RateTracker.
Also, the video engine must ensure that time stamps are always increasing.
With this, time stamps (ntp, render_time and rtp timestamps ) are checked and set in ViECapturer::OnIncomingCapturedFrame
This cl was previously reverted in https://webrtc-codereview.appspot.com/46549004/.
Patchset 1 contains the original patch after rebase.
Patshet 2 fix webrtc_perf_tests reported in chromium:465306
Note that chromium:465287 is being fixed in https://webrtc-codereview.appspot.com/43829004/
BUG=1128
R=magjed@webrtc.org, mflodman@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/47629004
Cr-Commit-Position: refs/heads/master@{#8776}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8776 4adac7df-926f-26a2-2b94-8c16560cd09d
