iNumRefFrame specifies total number of reference buffers to allocate.
For N temporal layers we need at least (N - 1) buffers to store last
encoded frames of all reference temporal layers.
There is no API in OpenH254 encoder to specify exact set of references
to be used to prediction of a given frame. Encoder can theoretically
use all available references.
Note that there is logic in OpenH264 which overrides iNumRefFrame to
max(iNumRefFrame, N - 1): https://source.chromium.org/chromium/chromium/src/+/main:third_party/openh264/src/codec/encoder/core/src/au_set.cpp;drc=8e90a2775c5b9448324fe8fef11d177cb65f36cc;l=122.
I.e., this change has no real effect. It only makes setup more clear.
Bug: none
Change-Id: If4b4970007e1cc55d8f052ea05213ab2e89a878f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/225480
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34445}
CL partially auto-generated with:
git grep -l "\bassert(" | grep "\.[c|h]" | \
xargs sed -i 's/\bassert(/RTC_DCHECK(/g'
And with:
git grep -l "RTC_DCHECK(false)" | \
xargs sed -i 's/RTC_DCHECK(false)/RTC_NOTREACHED()/g'
With some manual changes to include "rtc_base/checks.h" where
needed.
A follow-up CL will remove assert() from Obj-C code as well
and remove the #include of <assert.h>.
The choice to replace with RTC_DCHECK is because assert()
is because RTC_DCHECK has similar behavior as assert()
based on NDEBUG.
This CL also contains manual changes to switch from
basic RTC_DCHECK to other (preferred) versions like
RTC_DCHECK_GT (and similar).
Bug: webrtc:6779
Change-Id: I00bed8886e03d685a2f42324e34aef2c9b7a63b0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/224846
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34442}
If the RTP header extension playout-delay is used and set
to min=0, max>=0, frames are scheduled to be decoded as
soon as possible. There's a risk that too many frames are
sent to the decoder at once, which may cause problems
further down in the video pipeline.
This CL adds the fieldtrial WebRTC-ZeroPlayoutDelay with
the parameter min_pacing that determines the minimum
pacing interval between two frames scheduled for
decoding.
Bug: None
Change-Id: I471f7718761cfce9789b3aa8adea3e8a16ecb2fd
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/223742
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Johannes Kron <kron@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34387}
In cases where ToI420 fails it should be able to return null.
Bug: webrtc:12877
Change-Id: Ia13859c104d978a29712ae10f8e15acada8406ac
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/222613
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@google.com>
Cr-Commit-Position: refs/heads/master@{#34342}
Bilinear is faster but lesser quality, box is best quality. Our code
base has disagreed about which filter to use for quite some time,
causing aliasing bug reports. In an effort to avoid aliasing artifacts
and make our scaling filters more predictable, we're updating all uses
to kFilterBox.
WebRTC already uses kFilterBox everywhere except for these three
places. The main discrepency was between Chromium and WebRTC but that
has already been fixed. This CL fixes the last remaining bilinears.
This brings the WebRTC kFilterBox use count up from 11 to 14 and the
kFilterBilinear use count down from 3 to 0.
Bug: chromium:1212630
Change-Id: I5fe4aa92b9275d65b91ea97925533055d190d317
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/221372
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34248}
If at creation of a VP8 encoder there is not enough bitrate to enable a
given spatial layer - the configuration won't be updated to indicate
the correct temporal layer count. This means GetEncoderInfo() will
indicate lack of temporal layer support, which triggers issues with
rate allocation.
This CL fixes that by always setting an initial bitrate of 0bps.
Bug: webrtc:12788
Change-Id: I10974e85446b58e597d2ca415eaf2550306ce986
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/220929
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34198}
ffmpeg is going to be hiding the implementation of AVPacket, so we can't
allocate them on the stack anymore. av_init_packet is marked deprecated
on TOT ffmpeg, so remove its use everywhere in favor of av_packet_alloc
and av_packet_free.
Bug: chromium:1211508
Change-Id: I154311071123110dd749c71dec1ec2a0452b3908
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/217780
Commit-Queue: Ted Meyer <tmathmeyer@google.com>
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34106}
In CL https://webrtc-review.googlesource.com/c/src/+/216323 we fixed
the issue where I420 and I420A not being equal would result in dropping
frames in release builds.
But we forgot to update the corresponding DCHECK, meaning the I420 not
being the same as I420A issue still causes crashes on debug builds.
(I must have been running a release build not to catch this before?)
This CL replaces the DCHECK_EQ with an RTC_NOTREACHED inside the
IsCompatibleVideoFrameBufferType check.
Because this only affects debug builds, this CL does not need to be
backmerged anywhere.
Bug: chromium:1203206
Change-Id: I101823e8bca293e94d0f7ce507fe78cedff3ea1b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/219281
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34048}
Specifically, use reference instead of pointer for out parameter
and place the out parameter last, for the following methods
ReadUInt8
ReadUInt16
ReadUInt32
ReadBits
PeekBits
ReadNonSymmetric
ReadSignedExponentialGolomb
ReadExponentialGolomb
Bug: webrtc:11933
Change-Id: I3f1efe3e29155985277b0cd18700ddea25fe7914
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218504
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#34037}
When VideoFrameType for svc upper layer is kVideoFrameDelta for key pic,
the svc unittest will fail due to the wrong frame type for the super
frame of first key picture.
Bug: None
Change-Id: Iff026aaecb73890d3c45d2c88c9654a12d6fe3bf
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/216461
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Zhaoliang Ma <zhaoliang.ma@intel.com>
Cr-Commit-Position: refs/heads/master@{#33986}
Historically the PacketBuffer used a callback for assembled frames, and because of that RtpPacketInfos were piped through it even though they didn't have anything to do with the PacketBuffer.
With this CL RtpPacketInfos are stored in the RtpVideoStreamReceiver(2) instead.
Bug: webrtc:12579
Change-Id: Ia6285b59e135910eee7234b89b23425bb0fc0d2b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215320
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33980}
Before this CL, timestamps of received packets were rounded
to the nearest millisecond and stored as int64_t. Due to the
rounding it sometimes happened that timestamps later in the
pipeline that are not rounded seem to occur even before the
video frame was received.
Change-Id: I92d8f3540b23baae2d4a1dc6a7cb3f58bcdaad18
Bug: webrtc:12722
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/216398
Reviewed-by: Chen Xing <chxg@google.com>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Johannes Kron <kron@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33916}
VideoCodecInitializer::SetupCodec never sets startBitrate,
so SetAv1SvcConfig shouldn't use it.
Bug: webrtc:12720
Change-Id: I04835dc27368f32c19132d93c72364173d7050fc
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/217382
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33915}
Also fix similar annotation on NackModule to have effect
(when defining an alias with C++ using, ABSL_DEPRECATED should appear
on the left hand side).
Bug: webrtc:12339
Change-Id: Id80a20bf2c56a826777b8a40e06ac5c65e4f8db7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/217242
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33905}
This is essentially replacing `new rtc::RefCountedObject` with
`rtc::make_ref_counted` in many files. In a couple of places I
made minor tweaks to make things compile such as adding parenthesis
when they were missing.
Bug: webrtc:12701
Change-Id: I3828dbf3ee0eb0232f3a47067474484ac2f4aed2
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215973
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Tommi <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33852}
In Chromium, the I420ABufferInterface implementation uses the default
CropAndScale() implementation which converts to I420 in the process.
This should be OK, because we do not encode the alpha channel anyway,
so having WebRTC scaling ignore the alpha channel might even be a good
thing. Unfortunatety, an if statement in the LibvpxVp8Encoder did not
consider I420A and I420 to be the same, resulting in dropping perfectly
valid frames.
This CL fixes that by considering I420A and I420 "compatible" in a
comparison helper function. The problem only happens in this encoder,
so only this encoder needs to be fixed.
Bug: chromium:1203206
Change-Id: Iec434d4ada897c79e09914cac823148fd5b05e57
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/216323
Reviewed-by: Evan Shrubsole <eshr@google.com>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33845}
Fix simulcast svc controller to reuse dropped frame configuration,
same as full svc and k-svc controllers do.
This fuzzer reminded the issue was still there.
This is a reland of https://webrtc-review.googlesource.com/c/src/+/212281
Bug: webrtc:11999
Change-Id: Id3b2cd6c7e0923adfffb4e04c35ed2d6faca6704
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215921
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33802}
This is a refactor to simplify a follow-up CL of adding
SdpVideoFormat::IsSameCodec.
The original files media/base/h264_profile_level_id.* and
media/base/vp9_profile.h must be kept until downstream projects
stop using them.
Bug: chroimium:1187565
Change-Id: Ib39eca095a3d61939a914d9bffaf4b891ddd222f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215236
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Johannes Kron <kron@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33782}
Biggest change is a new helper class used to read data from the
bitstream and then pass the result to a function if reading was
successful. There's also helper to do if/else flow based on the read
values. This avoids a bunch of temporaries and in my view makes the
code esaier to read.
For example, this block:
uint32_t bit;
RETURN_FALSE_IF_ERROR(br->ReadBits(&bit, 1));
if (bit) {
RETURN_FALSE_IF_ERROR(br->ConsumeBits(7));
}
...is now written as:
RETURN_IF_FALSE(
br->IfNextBoolean([br] { return br->ConsumeBits(7); }));
In addition, we parse and put a few extra things in FrameInfo:
show_existing_frame, is_keyframe, and base_qp.
Bug: webrtc:12354
Change-Id: Ia0b707b223a1afe0a4521ce2b995437d41243c06
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215239
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33776}
Like aom and openh264, VP9 can be disabled with the gn argument.
Bug: None
Change-Id: I7d67e3946afae0bb4cac8a7e591445604dda9ce1
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215260
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33737}
In RtpVideoStreamReceiver2 it can be protected by the `worker_task_checker_` instead.
Bug: webrtc:12579
Change-Id: I4f7d64f16172139eddc7a3e07d1dbbf338beaf2e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215224
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33734}
VideoStreamEncoderTest: Remove unneeded set_timestamp_rtp in CreateFrame methods (the timestamp is set based on ntp_time_ms in VideoStreamEncoder::OnFrame).
Bug: none
Change-Id: I6b5531a9ac21cde5dac54df6de9b9d43261e90c6
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214488
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Commit-Queue: Åsa Persson <asapersson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33683}
There's been reports of dropped frames that are not counted and
correctly reported by getStats().
If a HW decoder is used and the system is provoked by stressing
the system, I've been able to reproduce this problem. It turns out
that we've missed frames that are dropped because there is no
callback to the Decoded() function.
This CL restructures the code so that dropped frames are counted
even in cases where there's no corresponding callback for some frames.
Bug: webrtc:11229
Change-Id: I0216edba3733399c188649908d459ee86a9093d0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214783
Commit-Queue: Johannes Kron <kron@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33671}
...though the big issue was probably that pending frames weren't being
culled properly in the case of frame dropping.
Bug: webrtc:12596
Change-Id: I9a03282b2a99087aa7c5650e57ce30fe0f0d3036
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214127
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33638}
While debugging https://crbug.com/1195144 I found it useful to clarify
this log statement.
The log would say "When scaling [kNative], the image was unexpectedly
converted to [kI420]..." but not saying what it was trying to convert
it to. This CL adds: "... instead of [kNV12]."
Bug: chromium:1195144
Change-Id: I13e0040edf5d7d98d80ce674812f67dfb73be36e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214040
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33634}
libaom is compiled with REALTIME_ONLY option. Soon it will be impossible
to create encoder or request default config with usage other than
AOM_USAGE_REALTIME. Fixing the wrapper to use proper usage parameter
Bug: None
Change-Id: I862741a724e4a8524f22ae79700b3da6517dbfb2
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214100
Commit-Queue: Fyodor Kyslov <kyslov@google.com>
Reviewed-by: Marco Paniconi <marpan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33624}
This change adds basic support for setting codecType kVideoCodecAV1 in
VCMEncodedFrames.
Bug: chromium:1191972
Change-Id: I258b39ff89c8b92ebbb288ef32c88b900a35d10e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/213182
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Markus Handell <handellm@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33594}
This reverts commit c184047fef.
Reason for revert: Breaks the WebRTC->Chromium roll:
ERROR Unresolved dependencies.
//third_party/webrtc/test/fuzzers:vp9_encoder_references_fuzzer(//build/toolchain/win:win_clang_x64)
needs //third_party/webrtc/modules/video_coding:mock_libvpx_interface(//build/toolchain/win:win_clang_x64)
We need to add tryjob to catch these. The fix is to make
//third_party/webrtc/modules/video_coding:mock_libvpx_interface
visible in built_with_chromium builds by moving the target
out of this "if" https://source.chromium.org/chromium/chromium/src/+/master:third_party/webrtc/modules/video_coding/BUILD.gn;l=615;drc=3889de1c4c7ae56ec742fb9ee0ad89657f638169.
Original change's description:
> Add fuzzer to validate libvpx vp9 encoder wrapper
>
> Fix simulcast svc controller to reuse dropped frame configuration,
> same as full svc and k-svc controllers do.
> This fuzzer reminded the issue was still there.
>
> Bug: webrtc:11999
> Change-Id: I74156bd743124723562e99deb48de5b5018a81d0
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/212281
> Reviewed-by: Erik Språng <sprang@webrtc.org>
> Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#33568}
TBR=danilchap@webrtc.org,sprang@webrtc.org
Change-Id: I1676986308c6d37ff168467ff2099155e8895452
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:11999
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/212973
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33573}
Fix simulcast svc controller to reuse dropped frame configuration,
same as full svc and k-svc controllers do.
This fuzzer reminded the issue was still there.
Bug: webrtc:11999
Change-Id: I74156bd743124723562e99deb48de5b5018a81d0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/212281
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33568}
Follow-up CL to VP8 and VP9 encoders taking care of mapping.
Context again:
This CL is part of Optimized Scaling efforts. In Chromium, the native
frame buffer is getting an optimized CropAndScale() implementation. To
support HW accelerated scaling, returning pre-scaled images and skipping
unnecessary intermediate downscales, WebRTC needs to 1) use CropAndScale
instead of libyuv::XXXXScale and 2) only map buffers it actually intends
to encode.
In this CL, VideoStreamEncoder no longer calls GetMappedFrameBuffer() on
behalf of the encoders, since the encoders are now able to either do the
mapping or performs ToI420() anyway.
- Tests for old VSE behaviors are updated to test the new behavior (i.e.
that native frames are pretty much always forwarded).
- The "having to call ToI420() twice" workaround to Android bug
https://crbug.com/webrtc/12602 is added to H264 and AV1 encoders.
Bug: webrtc:12469
Change-Id: Ibdc2e138d4782a140f433c8330950e61b9829f43
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/211940
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Evan Shrubsole <eshr@google.com>
Cr-Commit-Position: refs/heads/master@{#33548}
This will further speed up intra frame encoding
Bug: None
Change-Id: I3c836502cdcb1037e3128850a085b92acd8fc7ad
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/212821
Reviewed-by: Marco Paniconi <marpan@webrtc.org>
Commit-Queue: Fyodor Kyslov <kyslov@google.com>
Cr-Commit-Position: refs/heads/master@{#33544}
This is a follow-up to the VP9, fixing VP8 this time. Context again:
This CL is part of Optimized Scaling efforts. In Chromium, the native
frame buffer is getting an optimized CropAndScale() implementation. To
support HW accelerated scaling, returning pre-scaled images and skipping
unnecessary intermediate downscales, WebRTC needs to 1) use CropAndScale
instead of libyuv::XXXXScale and 2) only map buffers it actually intends
to encode.
- To achieve this, WebRTC encoders are updated to map kNative video
buffers so that in a follow-up CL VideoStreamEncoder can stop mapping
intermediate buffer sizes.
Bug: webrtc:12469, chromium:1157072
Change-Id: I026527ae77e36f66d02e149ad6fe304f6a8ccb05
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/212600
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Evan Shrubsole <eshr@google.com>
Cr-Commit-Position: refs/heads/master@{#33537}
Namespace used because of copy-pasting an old pattern, should never have been used in the first place. Removing it now to make followup refactoring prettier.
Bug: webrtc:12579
Change-Id: I00a80958401cfa368769dc0a1d8bbdd76aaa4ef5
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/212603
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33536}
