Fill drops with last decoded frame to make them look like freeze at
playback and to keep decoded spatial layers in sync.
Bug: none
Change-Id: I65f7c21100985c22932a1edd441b6c724833c11e
Reviewed-on: https://webrtc-review.googlesource.com/73685
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23076}
If frame of current layer was dropped, pass base frame to decoder if
non_ref_for_inter_layer_pred is set to true.
Bug: none
Change-Id: If7bf5220b74f424106edf74867c9afa8cc2b1ec5
Reviewed-on: https://webrtc-review.googlesource.com/73440
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23074}
Cleanup based on review of move CL. Using separate CL to keep the
move clean.
Bug: webrtc:9155
Change-Id: Ie0b15c6d3efcd86e2fa5761f014d0daf72ccfa06
Reviewed-on: https://webrtc-review.googlesource.com/73367
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23068}
Functions to estimate pitch period and gain.
Bug: webrtc:9076
Change-Id: Icfe9430dcae11bdb96165c5bfe6e2b1d3bf848ab
Reviewed-on: https://webrtc-review.googlesource.com/70382
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23066}
This CL removes the updating of the buffered data used to to pad the
64 sample blocks to 128 samples FFTs. As that padding was used
incorrectly in one place this resolves an important issue.
Bug: webrtc:9159,chromium:833801,webrtc:9206
Change-Id: Ie6830878ebec6130b61d4e7e3169357f2e253073
Reviewed-on: https://webrtc-review.googlesource.com/73240
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23059}
This CL changes the way the suppressor gain is computed in AEC3 in that
the FFTs used are padded with data and windowed with a Hanning-style
window.
This gives better FFT accuracy, an behavior matching the suppressor
gain application, and also results in one less FFT operation.
Bug: webrtc:9204,chromium:837563
Change-Id: I612676c389cb76a3130966a9b596ff3f44d21863
Reviewed-on: https://webrtc-review.googlesource.com/73141
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23057}
The new API stores events gathered by event type. For example, it is
possible to ask for a list of all incoming RTCP messages or all audio
playout events.
The new API is experimental and may change over next few weeks. Once
it has stabilized and all unit tests and existing tools have been
ported to the new API, the old one will be removed.
This CL also updates the event_log_visualizer tool to use the new
parser API. This is not a funcional change except for:
- Incoming and outgoing audio level are now drawn in two separate plots.
- Incoming and outgoing timstamps are now drawn in two separate plots.
- RTCP count is no longer split into Video and Audio. It also counts
all RTCP packets rather than only specific message types.
- Slight timing difference in sendside BWE simulation due to only
iterating over transport feedbacks and not over all RTCP packets.
This timing changes are not visible in the plots.
Media type for RTCP messages might not be identified correctly by
rtc_event_log2text anymore. On the other hand, assigning a specific
media type to an RTCP packet was a bit hacky to begin with.
Bug: webrtc:8111
Change-Id: Ib244338c86a2c1a010c668a7aba440482023b512
Reviewed-on: https://webrtc-review.googlesource.com/73140
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Minyue Li <minyue@webrtc.org>
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23056}
If the adaptive gain is too low, we raise it slowly and only during
speech.
The CL gives better behavior at the start of a call. If the gain is too
high, the fixed-digital limits it. The gain is also quickly reduced by
the AdaptiveGainApplier.
Bug: webrtc:7494
Change-Id: I683f1e3e463cddec2d91f6c7f15c73e744430034
Reviewed-on: https://webrtc-review.googlesource.com/71484
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23053}
This allows to control inter-layer prediction at encoding VP9 SVC.
There are three options:
1. Disabled.
2. Enabled for all pictures.
3. Enabled for key pictures, disabled for others.
Inter-layer prediction is enabled for all pictures by default.
Bug: none
Change-Id: I49fe43d8744c92bec349d815100ba158519f0664
Reviewed-on: https://webrtc-review.googlesource.com/71500
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23049}
The MBP having both discrete and integrated graphic cards will do
automate graphics switching by default. When it switches from discrete to
integrated one, the current display ID of the built-in display will
change and this will cause screen capture stops.
So make screen capture of built-in display continuing even if its display
ID is changed.
Bug: chromium:836979
Change-Id: If4f2d04d99a2690ccd6f894d94e6f8ff58ba2ec8
Reviewed-on: https://webrtc-review.googlesource.com/72603
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Commit-Queue: Brave Yao <braveyao@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23048}
This reverts commit b04e5cae08b8a7bc27041c1606547f807aaa2fc1.
Reason for revert: The reason for the revert is that some scenarios were detected where this caused the delay estimation to occur too slowly.
Original change's description:
> Making the delay estimator more robust to noisy nearends and low echoes
>
> This CL reduces the delay estimator step size to make it react better in
> scenarios where the environment is noisy, or the echo level is fairly
> low.
>
> Bug: webrtc:9177,chromium:835281
> Change-Id: I482d898c91eddc497e1284ee500d26df21a0574a
> Reviewed-on: https://webrtc-review.googlesource.com/71486
> Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
> Commit-Queue: Per Åhgren <peah@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22990}
TBR=gustaf@webrtc.org,peah@webrtc.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: webrtc:9177, chromium:835281
Change-Id: I33e09ebfed8ad8330419e554f482c956608befce
Reviewed-on: https://webrtc-review.googlesource.com/72843
Reviewed-by: Per Åhgren <peah@webrtc.org>
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Oleh Prypin <oprypin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23042}
For consistency with the VP9 RTP spec which uses term "picture" for set
of frames which belong to the same time instance.
Bug: none
Change-Id: I30e92d5debb008feb58f770b63fe10c2e0029267
Reviewed-on: https://webrtc-review.googlesource.com/72180
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23040}
This CL makes sure that the coherence-based gains are affected by the
upper gain limit during call start-up and after resets.
Bug: webrtc:9159,chromium:833801
Change-Id: I93fdd173b6e11ea861d0e01e12c048ec0a91db70
Reviewed-on: https://webrtc-review.googlesource.com/72841
Commit-Queue: Per Åhgren <peah@webrtc.org>
Reviewed-by: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23039}
It's safe to ignore this overflow since it only affects audio data,
not indices or anything like that.
Bug: chromium:835637
Change-Id: I60162e4627b08d5e3ba3a21fdae8087f098c7e46
Reviewed-on: https://webrtc-review.googlesource.com/72701
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23030}
This reverts commit 9e336ec0b8a77c3461d13677cff3563c11c88daa.
Reason for revert: Code can accidentally include the deprecated parser but link with the new one, or vice versa. Reverting to fix naming.
Original change's description:
> Create new API for RtcEventLogParser.
>
> The new API stores events gathered by event type. For example, it is
> possible to ask fo a list of all incoming RTCP messages or all audio
> playout events.
>
> The new API is experimental and may change over next few weeks. Once
> it has stabilized and all unit tests and existing tools have been
> ported to the new API, the old one will be removed.
>
> This CL also updates the event_log_visualizer tool to use the new
> parser API. This is not a funcional change except for:
> - Incoming and outgoing audio level are now drawn in two separate plots.
> - Incoming and outgoing timstamps are now drawn in two separate plots.
> - RTCP count is no longer split into Video and Audio. It also counts
> all RTCP packets rather than only specific message types.
> - Slight timing difference in sendside BWE simulation due to only
> iterating over transport feedbacks and not over all RTCP packets.
> This timing changes are not visible in the plots.
>
>
> Media type for RTCP messages might not be identified correctly by
> rtc_event_log2text anymore. On the other hand, assigning a specific
> media type to an RTCP packet was a bit hacky to begin with.
>
> Bug: webrtc:8111
> Change-Id: I8e7168302beb69b2e163a097a2a142b86dd4a26b
> Reviewed-on: https://webrtc-review.googlesource.com/60865
> Reviewed-by: Minyue Li <minyue@webrtc.org>
> Reviewed-by: Sebastian Jansson <srte@webrtc.org>
> Commit-Queue: Björn Terelius <terelius@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#23015}
TBR=terelius@webrtc.org,srte@webrtc.org,minyue@webrtc.org
Change-Id: Ib4bbcf0563423675a3cc1dce59ebf665e0c5dae9
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:8111
Reviewed-on: https://webrtc-review.googlesource.com/72500
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23026}
This deletes the resilienceOn flag in VideoCodecVP8 and VideoCodecVP9.
Instead, the implementations of VP8 and VP9 set resilience mode
internally, based on the configuration of temporal and spatial layers.
The nack_enabled argument to VideoCodecInitializer::SetupCodec becomes
unused with this cl. In a followup, it will be deleted, together with
the corresponding argument to VideoStreamEncoder methods.
An applications which really wants to configure resilience differently
can do that by injecting an EncoderFactory with encoders behaving
as desired.
Bug: webrtc:8830
Change-Id: I9990faf07d3e95c0fb4a56fcc9a56c2005b4a6fa
Reviewed-on: https://webrtc-review.googlesource.com/71380
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23025}
This CL is created from a work initiated at https://webrtc-review.googlesource.com/c/src/+/61160
The purpose of this work is to improve the performance of the echo canceler (AEC3) when the farend signal contains stationary noises:
- An stationarity estimator of the farend signal has been added for detecting the portions of the farend signal that are pure noise.
- When the echo canceler deals with a portion of the signal that contains basically noise, the echo suppressor is able to back-off and avoid the fading of the nearend speech.
Change-Id: Id4b87fc59f4765bf1fca36d1cab39a49aabe104a
Bug: webrtc:9193,chromium:836790
Reviewed-on: https://webrtc-review.googlesource.com/64141
Reviewed-by: Per Åhgren <peah@webrtc.org>
Commit-Queue: Jesus de Vicente Pena <devicentepena@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23024}
The code that attenuates narrow banded echo peaks in low frequencies
is removed as it affects transparency negatively.
Bug: webrtc:9192,chromium:836729
Change-Id: Ib90ce6a3db0a75e8d69bdca432e1f8f8bfbbd988
Reviewed-on: https://webrtc-review.googlesource.com/72380
Reviewed-by: Per Åhgren <peah@webrtc.org>
Commit-Queue: Gustaf Ullberg <gustaf@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23022}
The goal is to make these injectable, and only VP8 and VP9 specific
targets should depend on them.
Bug: webrtc:7925
Change-Id: Ie9239a54d197fe70c93de0582797211fef6997a2
Reviewed-on: https://webrtc-review.googlesource.com/72082
Commit-Queue: Anders Carlsson <andersc@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23021}
The new API stores events gathered by event type. For example, it is
possible to ask fo a list of all incoming RTCP messages or all audio
playout events.
The new API is experimental and may change over next few weeks. Once
it has stabilized and all unit tests and existing tools have been
ported to the new API, the old one will be removed.
This CL also updates the event_log_visualizer tool to use the new
parser API. This is not a funcional change except for:
- Incoming and outgoing audio level are now drawn in two separate plots.
- Incoming and outgoing timstamps are now drawn in two separate plots.
- RTCP count is no longer split into Video and Audio. It also counts
all RTCP packets rather than only specific message types.
- Slight timing difference in sendside BWE simulation due to only
iterating over transport feedbacks and not over all RTCP packets.
This timing changes are not visible in the plots.
Media type for RTCP messages might not be identified correctly by
rtc_event_log2text anymore. On the other hand, assigning a specific
media type to an RTCP packet was a bit hacky to begin with.
Bug: webrtc:8111
Change-Id: I8e7168302beb69b2e163a097a2a142b86dd4a26b
Reviewed-on: https://webrtc-review.googlesource.com/60865
Reviewed-by: Minyue Li <minyue@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23015}
Include dropped frames by the encoder in the frame drop percentage.
To react faster at low framerates:
- Use ExpFilter instead of MovingAverage to filter QP values.
- Reduce sampling interval while waiting for minimum number of needed frames (when not in fast rampup mode).
A separate slower ExpFilter is used for upscaling.
Bug: webrtc:9169
Change-Id: If7ff6c3bd4201fda2da67125889838fe96ce7061
Reviewed-on: https://webrtc-review.googlesource.com/70761
Commit-Queue: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23014}
This converts the reserved bit in VP9 RTP payload descriptor into the
flag which indicates whether current frame can be used for prediction
of next spatial layer or not.
VP9 encoder wrapper sets non_ref_for_inter_layer_pred=false for all
frames for now.
Bug: none
Change-Id: I32f68868686475905fb09173cffd2b6e1bedcb7c
Reviewed-on: https://webrtc-review.googlesource.com/71080
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23010}
If the packets are ordered by their ack time, their send
time can decrease in the case they have been reordered.
Bug: webrtc:8415
Change-Id: Ic9fca8d47de37b931085aeefcd62bbddd8869db9
Reviewed-on: https://webrtc-review.googlesource.com/72003
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22999}
...continuation of review in https://webrtc-review.googlesource.com/c/src/+/70781
This CL ensures that the FineAudioBuffer can support stereo and also adapts
all classes which uses the FineAudioBuffer.
Note that, this CL does NOT enable stereo on mobile platforms by default. All it does is to ensure
that we *can*. As is, the only functional change is that all clients
will now use a FineAudioBuffer implementation which supports stereo (see
separate unittest).
The FineAudioBuffer constructor has been modified since it is better to
utilize the information provided in the injected AudioDeviceBuffer pointer
instead of forcing the user to supply redundant parameters.
The capacity parameter was also removed since it adds no value now when the
more flexible rtc::BufferT is used.
I have also done local changes (not included in the CL) where I switch
all affected audio backends to stereo and verified that it works in real-time
on all affected platforms (Androiod:OpenSL ES, Android:AAudio and iOS).
Also note that, changes in:
sdk/android/src/jni/audio_device/aaudio_player.cc
sdk/android/src/jni/audio_device/aaudio_recorder.cc
sdk/android/src/jni/audio_device/opensles_player.cc
sdk/android/src/jni/audio_device/opensles_recorder.cc
are simply copies of the changes done under modules/audio_device/android since we currently
have two versions of the ADM for Android.
Bug: webrtc:9172
Change-Id: I1ed3798bd1925381d68f0f9492af921f515b9053
Reviewed-on: https://webrtc-review.googlesource.com/71201
Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22998}
This CL overrides the power-based suppressor gain decision with
a coherence based descision for the cases when that indicates a
higher suppressor gain.
Bug: webrtc:9159,chromium:833801
Change-Id: I0e7d82ac1b8c70ffe9d45907559bb14b1b849d71
Reviewed-on: https://webrtc-review.googlesource.com/71660
Commit-Queue: Per Åhgren <peah@webrtc.org>
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22997}
This CL reduces the delay estimator step size to make it react better in
scenarios where the environment is noisy, or the echo level is fairly
low.
Bug: webrtc:9177,chromium:835281
Change-Id: I482d898c91eddc497e1284ee500d26df21a0574a
Reviewed-on: https://webrtc-review.googlesource.com/71486
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22990}
This moves the PostUpdates function from SendSideCongestionController
to the ControlHandler class.
Bug: None
Change-Id: I4000484a1df9d5fae02573196153c24f4f940219
Reviewed-on: https://webrtc-review.googlesource.com/70223
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22987}
Since the webrtc_common build target does not have visibility set, we
cannot easily use BitrateAllocation in other parts of Chromium.
This is currently blocking parts of chromium:794608, and I know of other
usage outside webrtc already, so moving it to api/ should be warranted.
Also, since there's some naming confusion and this class is video
specific rename it VideoBitrateAllocation. This also fits with the
standard interface for producing these: VideoBitrateAllocator.
Bug: chromium:794608
Change-Id: I4c0fae40f9365e860c605a76a4f67ecc9b9cf9fe
Reviewed-on: https://webrtc-review.googlesource.com/70783
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22986}
We only support on (formely kResilientStream) and off (formely
kResilienceOff). The third mode, kResilientFrames, was not
implemented.
Bug: None
Change-Id: Ida82f6a33eda9d943ea70bc8ae4e6bddb720b0e8
Reviewed-on: https://webrtc-review.googlesource.com/71481
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22984}
Only specially taggged targets may transitively depend on poisonous
targets. We first apply it to audio codecs.
This makes it much clearer exactly what parts of the code still have
dependencies on the audio codecs (and we want to eventually get rid of
pretty much all of them).
Bug: webrtc:8396, webrtc:9121
Change-Id: Iba5c2e806c702b5cfe881022674705f647896d43
Reviewed-on: https://webrtc-review.googlesource.com/69520
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22979}
This fixes misprint in the code which calculates target bitrate of a
VP9 spatial layer where "-" was used instead of "+".
Bug: none
Change-Id: I17d76a84d00e453c055c068968d7b276e9c23f51
Reviewed-on: https://webrtc-review.googlesource.com/71663
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22974}
NetEq was (up until this CL) capable of fading over to generating a
constant background noise when voice expansion had lasted too long.
However, the code has for a really long time only ever used the "off"
mode, which meant that long expansions are faded down to complete
silence (only zeros), i.e., background noise fill was not used.
Removing the other two modes ("on" and "fade") simplifies the code.
Bug: webrtc:9180
Change-Id: Ia2d46960208f3d75c9659ad3f027c52e5ecfb6b0
Reviewed-on: https://webrtc-review.googlesource.com/71485
Commit-Queue: Henrik Lundin <henrik.lundin@webrtc.org>
Reviewed-by: Minyue Li <minyue@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22969}
The target modules/audio_coding:isac_neon needs to link with
transform_tables.c but adding a dependency between isac_neon and
isac_fix_c creates a circular dependency.
This CL moves transform_tables.c to isac_fix_common (which is already a
dependency of isac_neon).
Bug: None
Change-Id: I4135ec772b0017e77f1411e9a8093b495220c636
Reviewed-on: https://webrtc-review.googlesource.com/71581
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22968}
Proper header include is missing for this file causing clang to complain about missing prototype for function `WebRtcIsacfix_AllpassFilter2FixDec16Neon`
Bug: None
Change-Id: Idb32e9fab6760a9a56f1db2d43e7c8e2e1fe5359
Reviewed-on: https://webrtc-review.googlesource.com/70370
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22967}
