timestamp_ is only used in GenerateFrame() and its old value is
discarded. So it just needs to be a local variable in GenerateFrame().
As a result, we can remove the start_timestamp parameter from the
constructor and Init().
Also mark the GeneratePacket() method private because it is only used
internally.
R=stefan@webrtc.org
BUG=none
TEST=none
Review URL: https://webrtc-codereview.appspot.com/50149004
Cr-Commit-Position: refs/heads/master@{#9386}
Use the current parameter names in the comment for SetNackMode().
Add a warning comment about the lifetime of the return value of
GetNackList().
R=stefan@webrtc.org
BUG=none
TEST=none
Review URL: https://webrtc-codereview.appspot.com/52599004
Cr-Commit-Position: refs/heads/master@{#9321}
The most important change is to prevent a potential buffer overflow in
NackList(). It cannot happen if the |size| argument passed to NackList()
is consistent with the |max_nack_list_size| argument passed to
SetNackSettings(), and there is an assertion to check that. But it is
good to defend against this in the release build because assert() is
compiled away in the release build.
Remove the unused |master| parameter to the VCMReceiver constructor.
Remove the unused State() getter method and the corresponding state_
member.
Remove the declarations for the nonexistent GenerateReceiverId()
method and the receiver_id_counter_ member.
Remove the unneeded data_buffer_ member of TestVCMReceiver. It was
assigned to packet.dataPtr and then immediately overwritten by
stream_generator_->GetPacket() or stream_generator_->PopPacket().
R=stefan@webrtc.org
BUG=none
TEST=none
Review URL: https://webrtc-codereview.appspot.com/51119004
Cr-Commit-Position: refs/heads/master@{#9318}
uint32_t parameters don't need to be passed by reference. The
VCMJitterBuffer destructor doesn't need to be virtual because the
class has no virtual methods.
R=stefan@webrtc.org
BUG=none
Review URL: https://webrtc-codereview.appspot.com/55499004
Cr-Commit-Position: refs/heads/master@{#9288}
tl;dr - non-continuous frames (due to padding) would get stuck as incomplete if the previous complete frame arrived and was decoded before the padding arrived.This fix re-checks the incomplete frame list for continuous frames after old packets arrive.
When padding is enabled and RTX is not, padding is sent as empty RTP packets tacked onto the end of completed frames (meaning: same timestamp, but after a packet with the marker bit set). Given the following set of circumstances, codified in the new unit test method, a frame can get permanently stuck in the incomplete frames list:
- Frame A decoded (packets 94-95). Next expected sequence number is 96.
- Frame C arrives (packets 100-101) and is marked complete. It isn't continuous, since it starts at 100, so it's placed in the incomplete frame list.
- Frame B arrives (packets 96-97) and is complete, since 97 has a marker bit. Turns out that packets 98-99 are padding, but the receiver doesn't know that.
- Frame B is decoded, removed from the decodable frames list, and last decoded state is updated.
- Packets 98-99 arrive. They hit the IsOldPacket check and update the last decoded state, but they don't trigger FindAndInsertContinuousFrames.
- Further packets/frames arrive and complete, but FindAndInsertContinuousFrames only runs on frames that are newer than the newly completed frame.
In this state, Frame C is permanently stuck as incomplete, so the jitter buffer overall is stuck until max NACK age (default: 450 packets), the max NACK list size (default: 200 packets), or a keyframe arrives and IsContinuous returns true for the keyframe.
(Before the November refactoring, Frame B wouldn't have to have been decoded for the bug to trigger; just having a complete continuous frame at any time before the padding arrived would cause this state, as FindAndInsertContinuousFrames was only called when the frame originally became continuous and was inserted into the decodable frames list. Post refactoring, the frame is removed/re-added to the decodable list on every padding packet that arrives)
BUG=
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/50959004
Cr-Commit-Position: refs/heads/master@{#9264}
ViEChannels without default encoders doesn't register a receive codec by
default. This makes VideoReceiver::Decode return early, causing a
high-priority thread to effectively be busy looping. This would be
expected to wreck more havoc in a more cross-platform manner than it has
visibly done. On Windows XP however it manages to bring the whole
machine to a grinding halt forcing a reboot if CPU usage hits 100%.
BUG=chromium:470013
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/48049004
Cr-Commit-Position: refs/heads/master@{#8976}
CopyCodecSpecific nulls out the rtpheader pointer hence causing the crash downstream.
More details about the codec type enums:
There are 2 enums defined. webrtc::VideoCodecType webrtc::RtpCodecTypes and they don't match. Inside CopyCodecSpecific in generic_encoder.cc, it was converted from the first to the 2nd type. At that point, it'll be kRtpVideoNone (as the effect of memset to 0). kRtpVideoNone is a bad value as it could cause assert. Later, it'll be reset to kRtpVideoGeneric in RTPSender::SendOutgoingData so it's not a concern.
BUG=4511
R=pbos@webrtc.org, pthatcher@webrtc.org, stefan@webrtc.org
Committed: https://crrev.com/29b1a1c0c7c6f4b1ae4d63844b1dfaa7a72530a0
Cr-Commit-Position: refs/heads/master@{#8951}
Review URL: https://webrtc-codereview.appspot.com/47999004
Cr-Commit-Position: refs/heads/master@{#8955}
CopyCodecSpecific nulls out the rtpheader pointer hence causing the crash downstream.
More details about the codec type enums:
There are 2 enums defined. webrtc::VideoCodecType webrtc::RtpCodecTypes and they don't match. Inside CopyCodecSpecific in generic_encoder.cc, it was converted from the first to the 2nd type. At that point, it'll be kRtpVideoNone (as the effect of memset to 0). kRtpVideoNone is a bad value as it could cause assert. Later, it'll be reset to kRtpVideoGeneric in RTPSender::SendOutgoingData so it's not a concern.
BUG=4511
R=pbos@webrtc.org, pthatcher@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/47999004
Cr-Commit-Position: refs/heads/master@{#8951}
All RTP packets from sender side will carry the rotation info. (will file a bug to track this) On the receiving side, only packets with marker bit set will be examined.
Tests completed:
1. android standalone to android standalone
2. android standalone to chrome (with and without this change)
3. android on chrome
BUG=4145
R=glaznev@webrtc.org, mflodman@webrtc.org, perkj@webrtc.org, pthatcher@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/47399004
Cr-Commit-Position: refs/heads/master@{#8905}
This CL avoids changing the mentioned callbacks during a call, to avoid
a potential deadlock when acquiring _sendCritSect and calling
_mediaOpt.SetTargetRates.
Moving the critsect revealed a race for the FEC parameters in RtpVideoSender, so the CL grew a bit to avoid this. I also cleaned up some code here at the same time, but tried to keep it at a minimum since this CL had already increased a lot in size.
BUG=769
R=pbos@webrtc.org, stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/42939004
Cr-Commit-Position: refs/heads/master@{#8899}
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}
Removes a lot of code that prevents refactoring VideoCodingModule. Tests
covering the module should be TEST_Fs, and this looks like like fairly
unused code in general.
Adds a 'rtp_player' binary which performs a small subset.
BUG=4391
R=stefan@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/44559004
Cr-Commit-Position: refs/heads/master@{#8787}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8787 4adac7df-926f-26a2-2b94-8c16560cd09d
CVO is, instead of rotating frame on the capture side, to have renderer rotate the frame based on a new rtp header extension.
The change includes
1. encoder side needs to pass this from raw frame to the encoded frame.
2. decoder needs to copy it from rtp packet (only the last packet of a frame has this info) to decoded frame.
R=mflodman@webrtc.orgTBR=stefan@webrtc.org
BUG=4145
Review URL: https://webrtc-codereview.appspot.com/46429006
Cr-Commit-Position: refs/heads/master@{#8767}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8767 4adac7df-926f-26a2-2b94-8c16560cd09d
> Speculative revert of 8631 "Remove lock from Bitrate() and FrameRate() in Video..."
>
> We ran into the alignment problem on Mac 10.9 debug again. This is the only CL I see in the range that adds an rtc::CriticalSection, so I'm trying out reverting it before attempting another roll.
>
> > Remove lock from Bitrate() and FrameRate() in VideoSender.
> > These methods are called on the VideoSender's construction thread, which is the same thread as modifies the value of _encoder. It's therefore safe to not require a lock to access _encoder on this thread.
> >
> > I'm making access to the rate variables from VCMGenericEncoder, thread safe, by using a lock that's not associated with the encoder. There should be little to no contention there. While modifying VCMGenericEncoder, I noticed that a couple of member variables weren't needed, so I removed them.
> >
> > The reason for this change is that getStats is currently contending with the encoder when Bitrate() is called. On my machine, this means that getStats can take about 25-30ms instead of ~1ms.
> >
> > Also adding some documentation for other methods and a suggestion for how we could avoid contention between the encoder and the network thread.
> >
> > BUG=2822
> > R=mflodman@webrtc.org
> >
> > Review URL: https://webrtc-codereview.appspot.com/43479004
>
> TBR=tommi@webrtc.org
>
> Review URL: https://webrtc-codereview.appspot.com/45529004TBR=tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/46519004
Cr-Commit-Position: refs/heads/master@{#8645}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8645 4adac7df-926f-26a2-2b94-8c16560cd09d
We ran into the alignment problem on Mac 10.9 debug again. This is the only CL I see in the range that adds an rtc::CriticalSection, so I'm trying out reverting it before attempting another roll.
> Remove lock from Bitrate() and FrameRate() in VideoSender.
> These methods are called on the VideoSender's construction thread, which is the same thread as modifies the value of _encoder. It's therefore safe to not require a lock to access _encoder on this thread.
>
> I'm making access to the rate variables from VCMGenericEncoder, thread safe, by using a lock that's not associated with the encoder. There should be little to no contention there. While modifying VCMGenericEncoder, I noticed that a couple of member variables weren't needed, so I removed them.
>
> The reason for this change is that getStats is currently contending with the encoder when Bitrate() is called. On my machine, this means that getStats can take about 25-30ms instead of ~1ms.
>
> Also adding some documentation for other methods and a suggestion for how we could avoid contention between the encoder and the network thread.
>
> BUG=2822
> R=mflodman@webrtc.org
>
> Review URL: https://webrtc-codereview.appspot.com/43479004TBR=tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/45529004
Cr-Commit-Position: refs/heads/master@{#8640}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8640 4adac7df-926f-26a2-2b94-8c16560cd09d
These methods are called on the VideoSender's construction thread, which is the same thread as modifies the value of _encoder. It's therefore safe to not require a lock to access _encoder on this thread.
I'm making access to the rate variables from VCMGenericEncoder, thread safe, by using a lock that's not associated with the encoder. There should be little to no contention there. While modifying VCMGenericEncoder, I noticed that a couple of member variables weren't needed, so I removed them.
The reason for this change is that getStats is currently contending with the encoder when Bitrate() is called. On my machine, this means that getStats can take about 25-30ms instead of ~1ms.
Also adding some documentation for other methods and a suggestion for how we could avoid contention between the encoder and the network thread.
BUG=2822
R=mflodman@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/43479004
Cr-Commit-Position: refs/heads/master@{#8631}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8631 4adac7df-926f-26a2-2b94-8c16560cd09d
Clang version changed 223108:230914
Details: e144d30..6fdb142/tools/clang/scripts/update.sh
Removes the OVERRIDE macro defined in:
* webrtc/base/common.h
* webrtc/typedefs.h
The majority of the source changes were done by running this in src/:
perl -0pi -e "s/virtual\s([^({;]*(\([^({;]*\)[^({;]*))(OVERRIDE|override)/\1override/sg" `find {talk,webrtc} -name "*.h" -o -name "*.cc*" -o -name "*.mm*"`
which converted all:
virtual Foo() OVERRIDE
functions to:
Foo() override
Then I manually edited:
* talk/media/webrtc/fakewebrtccommon.h
* webrtc/test/fake_common.h
Remaining uses of OVERRIDE was fixed by search+replace.
Manual edits were done to fix virtual destructors that were
overriding inherited ones.
Finally a build error related to the pure virtual definitions of
Read, Write and Rewind in common_types.h required a bit of
refactoring in:
* webrtc/common_types.cc
* webrtc/common_types.h
* webrtc/system_wrappers/interface/file_wrapper.h
* webrtc/system_wrappers/source/file_impl.cc
This roll should make it possible for us to finally re-enable deadlock
detection for TSan on the buildbots.
BUG=4106
R=pbos@webrtc.org, tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/41069004
Cr-Commit-Position: refs/heads/master@{#8596}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8596 4adac7df-926f-26a2-2b94-8c16560cd09d
This reverts commit r8434.
Reason for revert: Introduced a race condition. If ViECaptureProcess() -> SwapCapturedAndDeliverFrameIfAvailable() is called twice without a call to OnIncomingCapturedFrame() in between (with both captured_frame_ and deliver_frame_ populated), an old frame will be delivered again, since captured_frame_->IsZeroSize() will never be true.
BUG=4352
TBR=perkj@webrtc.org, stefan@webrtc.org, tommi@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/40129004
Cr-Commit-Position: refs/heads/master@{#8530}
git-svn-id: http://webrtc.googlecode.com/svn/trunk@8530 4adac7df-926f-26a2-2b94-8c16560cd09d
