This CL reorders the update steps physically, to make them
easier to comprehend. It renames variables to be more verbose,
but also adds succinct mathematical descriptions (using Wikipedia
notation) to all steps.
No functional changes are intended with this change.
Bug: webrtc:14151
Change-Id: I6a4642e89e2b73639f0b4c928e07b317c14d5884
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271546
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37784}
* Update naming of data members.
* Start reordering code blocks in `PredictAndUpdate`.
(The "predict" step is done in this C:. The "update"
step will be improved in another cl.)
Bug: webrtc:14151
Change-Id: Idea1e8e786bd672dadedbcb3cd5598f4a033e81e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271023
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37767}
This CL simplifies and documents the interface of the Kalman filter better. A simple unit test verifying the filter's convergence is
added. No functional changes to the filter are intended.
Future CLs will improve the data member naming and code organization.
Bug: webrtc:14151
Change-Id: I01e60d4b783cabad3ccbdc711c5d746666dd16ca
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/265877
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37766}
Instead of showing individual byte differences, this CL detects
differences in the expected and actual byte streams of the evaluated
AEC dump and, if detected, parses the `audioproc::Event` proto lite
messages and calls EXPECT_EQ() for a subset of individual (sub-)fields.
Note that messages are parsed only if the byte streams of each message
pair do not match, so with no failures the test runs at no extra cost.
Plus, the the added funcionality can only be enabled for local
debugging by flipping the `kDumpWhenExpectMessageEqFails` flag - a
code change cannot land if the flag is set to true.
Note that `MessageDifferencer` (see [1]) could not be used because
it is not implemented for `MessageLite` protos.
[1] https://developers.google.com/protocol-buffers/docs/reference/cpp/google.protobuf.util.message_differencer
Bug: b/241923537
Change-Id: I8e0eda3b1ecfe06945b6dad5ee8871f8200d76d7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270922
Reviewed-by: Jesus de Vicente Pena <devicentepena@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37765}
When ScreencastPortal::OnStartRequestResponseSignal receives either a
non-zero response code or is missing the response data, it would
directly cast this to a RequestResponse. However, this direct cast is an
error. Per the documentation, the response signal returns the following
values with their corresponding meanings:
0 - Success
1 - User Cancelled
2 - Error
The RequestResponse enum however, has "kUnknown" as 0, and thus
"kSuccess" as 1 (with all other values also shifted up by 1 value). This
means that when the portal was cancelled, we were still receiving
RequestResponse::kSuccess. This fixes the issue by removing the improper
cast and adding a translation function. This function is local for now
since no where else attempted to cast values to a RequestResponse; but
can be moved if the need arises.
Fixed: chromium:1351824
Change-Id: I4cd44d90055147c9592d590c7969dcfc3297a3d9
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271240
Commit-Queue: Alexander Cooper <alcooper@chromium.org>
Auto-Submit: Alexander Cooper <alcooper@chromium.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37755}
This combines the below IPv6 fixes into the field trial
WebRTC-IPv6NetworkResolutionFixes:
1. Prefer global IPv6 address over link local
2. Use address family when resolving STUN hostname
WebRTC-PreferGlobalIPv6ToLinkLocal is currently in Dev but will be
rolled back temporarily.
Bug: webrtc:14334, webrtc:14131
Change-Id: I1fb3f55c4c5f3c5c0b441ece30e72cf393e074d0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271340
Reviewed-by: Diep Bui <diepbp@webrtc.org>
Commit-Queue: Sameer Vijaykar <samvi@google.com>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37754}
The current behaviour is to lookup using AF_UNSPEC, which leaves
the decision up to the getaddrinfo implementation, then filter to
resolved addresses matching the network family anyway.
Looking up using the network's family upfront avoids resolving to
an unusable address.
Bug: webrtc:14319, webrtc:14131
Change-Id: I4997452dc26aeb82e5d2890701721e7d477803a8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270625
Commit-Queue: Sameer Vijaykar <samvi@google.com>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37753}
The input SocketAddress for STUN host lookup is constructed with just
the hostname, so the family is AF_UNSPEC. So added an overload with a
target family to distinguish this from the family of the input addr.
Bug: webrtc:14319, webrtc:14131
Change-Id: Ia5ac5aa2e894e0c4dfb4417e3e8a76a6cec3ea71
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270624
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Sameer Vijaykar <samvi@google.com>
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Reviewed-by: Jonas Oreland <jonaso@google.com>
Cr-Commit-Position: refs/heads/main@{#37750}
that rtc::Location parameter was used only as extra information for the
RTC_CHECKs directly in the function, thus call stack of the crash should
provide all the information about the caller.
Bug: webrtc:11318
Change-Id: Iec6dd2c5de547f3e1601647a614be7ce57a55734
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270920
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37748}
Resolves an issue where, in Chrome, WebRTC event logs do not capture outgoing packets for video receivers because no reference to the event log was passed to the video receiver.
Bug: webrtc:14338
Change-Id: Ia33ce6f2d69a0e341530648b10a08516dc53abf3
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271080
Reviewed-by: Magnus Flodman <mflodman@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37746}
Remove check if `prev_estimate_` is less than 10 us since it can never
be less than 1 ms.
Bug: None
Change-Id: If151390d22fa0b6ecdc36af64168d3e2049c7b6b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271203
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Johannes Kron <kron@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37745}
Drops frames if the encoder has been configured with a new set of rtp
streams and a stray frame is returned from an encoder. This can happen with
hardware encoders that may deliver frames on a separate thread than were
they are configured.
This cl disable sending media on the RTP module a video layer is connected to and there by, old frames are dropped.
Bug: webrtc:1200, b/201798527
Change-Id: Id6bcfc3a846f6b8ed3b645cbbde571b819611a75
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/271122
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37744}
This behaviour has been fixed with the introduction of FrameBuffer3
Bug: webrtc:14033, webrtc:13343, webrtc:9974
Change-Id: Iba81c169706336e85194ed141324466e44a2c859
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/265867
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37742}
SendStunBindingRequest can cause a SocketAddress pointed to by the
iterator to be deleted asynchronously before returning, causing `it`
to be invalid when incrementing in the continuation step.
Bug: webrtc:7309, webrtc:14131
Change-Id: I3f7d3d7c12935d9592ef3642679a821c58826df9
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270744
Commit-Queue: Sameer Vijaykar <samvi@google.com>
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37741}
Also refactored a bit to support IPv6 networks and socket factories with
a mocked DNS resolver.
Bug: webrtc:14319, webrtc:14131
Change-Id: I32ac5beb9a72201bf83aac26aed6a670ed2d4955
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270741
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Commit-Queue: Sameer Vijaykar <samvi@google.com>
Cr-Commit-Position: refs/heads/main@{#37737}
Base port already keeps an AlwaysValidPointer to field_trials, so
TURN port's duplicate, private copy is redundant.
Bug: webrtc:14319, webrtc:14131
Change-Id: I94ee78ca5140c0b67826fbb94c35e28f30add943
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/270627
Reviewed-by: Jonas Oreland <jonaso@google.com>
Commit-Queue: Sameer Vijaykar <samvi@google.com>
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37735}
