If a port is not used by any channel and if it has no connection for 30
seconds, it will be removed.
Note, as long as a port is used by a transport channel, it will be kept
even if it does not have any connection. This will be beneficial to
continual gathering because new connections can be created in the future
when network changes.
BUG=
R=pthatcher@webrtc.org, zhihuang@webrtc.org
Review URL: https://codereview.webrtc.org/2171183002 .
Cr-Commit-Position: refs/heads/master@{#13567}
This change makes WebRTC no longer stop sending video when we receive an
EWOULDBLOCK error from the operating system. This was previously
causing calls on a slow link (where the first hop is slow) to rapidly
oscillate between starting and stopping video.
We still do need to stop sending packets if there is no known good
connection we can use for that. We used to generate a synthetic
EWOULDBLOCK error in that case. This CL replaces it with a different
code (ENOTCONN); EWOULDBLOCK no longer stops the stream but ENOTCONN
does.
I've updated all the places where we seemed to be generating EWOULDBLOCK
for reasons other than some buffer been full; please give it a thorough
look in case I missed something.
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/2192963002 .
Cr-Commit-Position: refs/heads/master@{#13566}
If all connections on a port is destroyed, it will schedule an event
to check if it is dead after a timeout. Previously if a new connection
is created but destroyed before the event is fired, it will destroy the
port. With this change, we will not destoy it until it times out again
after the last created connection is destroyed.
BUG=
R=pthatcher@webrtc.org, zhihuang@webrtc.org
Review URL: https://codereview.webrtc.org/2184013003 .
Cr-Commit-Position: refs/heads/master@{#13563}
When the flag prune_turn_ports is set, When a high-priority turn port becomes available, it will prune low-priority ones. The pruned port will not be used for creating connections locally and its candidates will not be sent over to the remove side (unless they have been sent before being pruned).
This effectively reduces the number of TURN candidates and connections created by TURN ports.
BUG=
R=deadbeef@webrtc.org, pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/2093623004 .
Committed: https://crrev.com/17aac053f585e892114974d2eb248e05ad37f973
Cr-Original-Commit-Position: refs/heads/master@{#13335}
Cr-Commit-Position: refs/heads/master@{#13354}
Reason for revert:
Breaks Win32/Win64 Debug bots in client.webrtc waterfall
Original issue's description:
> Add config to prune low-priority TURN ports for creating connections
> When the flag prune_turn_ports is set, When a high-priority turn port becomes available, it will prune low-priority ones. The pruned port will not be used for creating connections locally and its candidates will not be sent over to the remove side (unless they have been sent before being pruned).
>
> This effectively reduces the number of TURN candidates and connections created by TURN ports.
>
> BUG=
> R=deadbeef@webrtc.org, pthatcher@webrtc.org
>
> Committed: https://crrev.com/17aac053f585e892114974d2eb248e05ad37f973
> Cr-Commit-Position: refs/heads/master@{#13335}
TBR=pthatcher@webrtc.org,deadbeef@webrtc.org,honghaiz@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=
Review-Url: https://codereview.webrtc.org/2111663003
Cr-Commit-Position: refs/heads/master@{#13342}
When the flag prune_turn_ports is set, When a high-priority turn port becomes available, it will prune low-priority ones. The pruned port will not be used for creating connections locally and its candidates will not be sent over to the remove side (unless they have been sent before being pruned).
This effectively reduces the number of TURN candidates and connections created by TURN ports.
BUG=
R=deadbeef@webrtc.org, pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/2093623004 .
Cr-Commit-Position: refs/heads/master@{#13335}
The "should I simulate EWOULDBLOCK?" determination now happens
solely in P2PTransportChannel. This also fixes a bug where the
"last packet id" was set even if no packet was sent.
R=honghaiz@webrtc.org, pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/2099783002 .
Cr-Commit-Position: refs/heads/master@{#13307}
Add sent_ping_requests, recv_ping_responses to ConnectionInfo.
recv_ping_responses_ will be incremented when OnConnectionRequestResponse() is called.
ent_ping_requests_ will be incremented when OnConnectionRequestSent() is called.
BUG=webrtc:5695
Review-Url: https://codereview.webrtc.org/1940493002
Cr-Commit-Position: refs/heads/master@{#13001}
If the remote side sends a candidate with the same address and port with an existing candidate,
but with a new ufrag and pwd, the local client will create a new connection from it
and destroy the old connection with the same remote address.
BUG=webrtc:5915
Review-Url: https://codereview.webrtc.org/2018693002
Cr-Commit-Position: refs/heads/master@{#13000}
This is the minumum allowed size, and will allow STUN pings to be smaller.
The unit tests on the the Gturn are also modified. A username with length of 16 bytes will be generated for Gturn only.
Review-Url: https://codereview.webrtc.org/1848083002
Cr-Commit-Position: refs/heads/master@{#12876}
if the network monitor detects it after the native code does.
Also set the network cost for ethernet, wifi, unknown, cellular network type to be 0, 10, 50, 900,
so that unknown networks will have lower precedence than known networks with low cost (like Wifi) but higher precedence than known networks with high cost.
And third, infer network type based on limited name matching in Android if there is no network monitor or network monitor did not find the type.
BUG=webrtc:5890
R=pthatcher@chromium.org, pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1976683003 .
Cr-Commit-Position: refs/heads/master@{#12833}
Depends on this CL in order to work in Chromium:
https://codereview.chromium.org/1976673002/
It works by creating pooled PortAllocatorSessions which can be picked up
by a P2PTransportChannel when needed (after a local description is set).
This can optimize candidate gathering time when there is some time between
creating a PeerConnection and setting a local description.
R=pthatcher@webrtc.org
Committed: 48e9d05f51
Review URL: https://codereview.webrtc.org/1956453003 .
Cr-Commit-Position: refs/heads/master@{#12729}
Reason for revert:
Breaks remoting_unittests. They defined their own operator== which conflicts with this one.
I'll remove the operator== in a roll CL. But until it's approved, I'm reverting this so the FYI bots will pass.
Original issue's description:
> Implement RTCConfiguration.iceCandidatePoolSize.
>
> It works by creating pooled PortAllocatorSessions which can be picked up
> by a P2PTransportChannel when needed (after a local description is set).
>
> This can optimize candidate gathering time when there is some time between
> creating a PeerConnection and setting a local description.
>
> R=pthatcher@webrtc.org
>
> Committed: 48e9d05f51TBR=pthatcher@webrtc.org,honghaiz@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review-Url: https://codereview.webrtc.org/1972043004
Cr-Commit-Position: refs/heads/master@{#12709}
It works by creating pooled PortAllocatorSessions which can be picked up
by a P2PTransportChannel when needed (after a local description is set).
This can optimize candidate gathering time when there is some time between
creating a PeerConnection and setting a local description.
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1956453003 .
Cr-Commit-Position: refs/heads/master@{#12708}
The problem with gmock is worked around by commenting out any other override declarations in classes using gmock.
NOPRESUBMIT=True
BUG=webrtc:3970
Review-Url: https://codereview.webrtc.org/1921653002
Cr-Commit-Position: refs/heads/master@{#12563}
But keep #including scoped_ptr.h in .h files, so as not to break
WebRTC users who expect those .h files to give them rtc::scoped_ptr.
BUG=webrtc:5520
Review URL: https://codereview.webrtc.org/1923163003
Cr-Commit-Position: refs/heads/master@{#12532}
This allows the reader to reference data, thus avoiding unnecessary
allocations and memory copies.
BUG=webrtc:5155,webrtc:5670
Review URL: https://codereview.webrtc.org/1821083002
Cr-Commit-Position: refs/heads/master@{#12160}
Also include that in the stun-ping request as part of the
network-info attribute.
Change the network cost to be 16 bits.
BUG=
Review URL: https://codereview.webrtc.org/1815473002
Cr-Commit-Position: refs/heads/master@{#12110}
Also changed from unsigned to signed integer per the style guide.
By the way, I kept all delta-times to be 32-bit int.
The only things left in the p2p dir are
1. proberprober/main.cc where Time() is used as the input for a random number.
2. pseudotcp.cc: where 32-bit time info is sent over the wire.
BUG=webrtc:5636
Review URL: https://codereview.webrtc.org/1793553002
Cr-Commit-Position: refs/heads/master@{#12019}
This is a good idea in general, because it makes ownership clearer,
but will also be very convenient when scoped_ptr is gone, since
unique_ptr doesn't have an .accept() method.
BUG=webrtc:5520
Review URL: https://codereview.webrtc.org/1800753003
Cr-Commit-Position: refs/heads/master@{#12002}
Chromium reported errors when building libjingle_nacl due to some methods used virtual instead of override when they were overriding the base class. My guess is that when one method starts using override, all other in the same class must too.
R=tommi@webrtc.orgTBR=pthatcher@webtrc.org
BUG=4173
Review URL: https://codereview.webrtc.org/1589563003 .
Cr-Commit-Position: refs/heads/master@{#11251}
To reduce the risk of future mistakes when connecting Ports, Port::OnSentPacket was made pure virtual to ensure that new implementations take care of it.
BUG=4173
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1577873003 .
Cr-Commit-Position: refs/heads/master@{#11247}
hasn't received anything yet. Deleting an ICE connection before it is pruned or timed out
when it hasn't received anything yet leads to ICE connections being deleted
before they have a chance to send a ping and receive a response.
BUG=
Review URL: https://codereview.webrtc.org/1544003002
Cr-Commit-Position: refs/heads/master@{#11151}
If the local port and the remote candidate's protocols do not match,
do not even try to pair them.
This avoids printing out confusing logs like
"Attempt to change a remote candidate..." in p2ptransportchannel
when two remote candidates have the same port number but different
protocols.
BUG=
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1516613002 .
Cr-Commit-Position: refs/heads/master@{#11034}
ARRAY_SIZE is the old version of arraysize and does not cover
all the cases in C++, arraysize is a copy of Chromium's
version and thus have wider coverage.
BUG=None
R=tommi@webrtc.org
Review URL: https://codereview.webrtc.org/1405023016
Cr-Commit-Position: refs/heads/master@{#10594}
Reason for revert:
Broke the Windows build:
[226/365] LINK_EMBED cc_perftests.exe
FAILED: ninja -t msvc -e environment.x86 -- E:\b\build\goma/gomacc "E:\b\depot_tools\win_toolchain\vs2013_files\VC\bin\amd64_x86\cl.exe" /nologo /showIncludes /FC @obj\remoting\protocol\remoting_unittests.channel_socket_adapter_unittest.obj.rsp /c ..\..\remoting\protocol\channel_socket_adapter_unittest.cc /Foobj\remoting\protocol\remoting_unittests.channel_socket_adapter_unittest.obj /Fdobj\remoting\remoting_unittests.cc.pdb
e:\b\build\slave\win\build\src\remoting\protocol\channel_socket_adapter_unittest.cc(36) : error C3861: 'set_readable': identifier not found
ninja: build stopped: subcommand failed.
Original issue's description:
> Replace readable with receiving where receiving means receiving anything (stun ping, response or data packet).
> If a connection does not receive for 30 seconds, it will be deleted.
> BUG=
>
> Committed: https://crrev.com/ae16f8547d3b447f62f6660f13688585c6c3de15
> Cr-Commit-Position: refs/heads/master@{#10001}
TBR=pthatcher@webrtc.org,honghaiz@webrtc.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=
Review URL: https://codereview.webrtc.org/1356103002
Cr-Commit-Position: refs/heads/master@{#10002}
If a connection does not receive for 30 seconds, it will be deleted.
BUG=
Review URL: https://codereview.webrtc.org/1351673003
Cr-Commit-Position: refs/heads/master@{#10001}
Since the TCPConnection has never been connected, they are not scheduled for ping hence will never be detected.
Also fix the case when reconnect fails, as it has become READABLE before, it also will not be deleted.
BUG=webrtc:4936
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1307083002 .
Cr-Commit-Position: refs/heads/master@{#9782}
During the reconnection phase, EWOULDBLOCK has been returned to upper layer which stops the sending of video stream.
BUG=webrtc:4930
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1288553010 .
Cr-Commit-Position: refs/heads/master@{#9767}
Migrated from https://codereview.webrtc.org/1275703006/ which causes test failures for android. On android, loopback interface was used as local interface to generate candidates. Add a test case to make sure this won't be broken in the future.
Also observed some failures under content_browsertests in chromium.fyi bot but can't repro locally. Might just be temporary test issue.
BUG=webrtc:4517
TBR=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1299333003 .
Cr-Commit-Position: refs/heads/master@{#9746}
The IceEndpointType has the format of <local_endpoint>_<remote_endpoint>. It is recorded on the BestConnection when we have the first OnTransportCompleted signaled.
BUG=webrtc:4918
R=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1277263002 .
Cr-Commit-Position: refs/heads/master@{#9737}
This reverts commit 0a2955f227666efd87b2a303a69c083ef801c528.
Revert "In the past, P2PPortAllocator.enable_multiple_routes is the indicator whether we should bind to the any address. It's easy to translate that into a port allocator flag in P2PPortAllocator's ctor. Going forward, we have to depend on an asynchronous permission check to determine whether gathering local address is allowed or not, hence the current way of passing it through constructor approach won't work any more. The asynchronous check will trigger SignalNetowrksChanged so we could only check that inside DoAllocate."
This reverts commit ba9ab4cd8d2e8fbc068dc36b5e6f6331d7deeccf.
TBR=pthatcher@webrtc.org
Review URL: https://codereview.webrtc.org/1288843003 .
Cr-Commit-Position: refs/heads/master@{#9729}
