GitLab

Bug: 868213
Change-Id: I1d5b577d46789cbcfd375423031b428362935320
Reviewed-on: https://chromium-review.googlesource.com/1214158Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Commit-Queue: Ryan Tseng <hinoka@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589624}

This was added in r524073 because code search didn't linkify bare crbug
links.  But now it does (see crbug.com/762061), so we don't need the
extra verbosity.

Change-Id: Ia0b2fd66b4717cb9dd87b6e49373918cec729c51
Reviewed-on: https://chromium-review.googlesource.com/1213512
Commit-Queue: Steve Kobes <skobes@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589623}

There are no flags to be updated

Leave data in QUIC buffers until everything has arrived.

Quartc cannot actually consume data until all the data for a stream has arrived.
Leaving stream data in QUIC's sequencer buffer until all the data for a stream
is ready to read allows QUIC flow control to account for data until Quartc is
ready to consume it.  This prevents a sender from running a Quartc client out of
memory by sending it an unbounded amount of data without terminating a stream.

QuartcStream supports a "deliver on complete" mode.  When using this mode, the
Quartc stream will not issue Delegate::OnReceived callbacks until the FIN bit
and all stream data has arrived.

QuartcStream sets its sequencer to level_triggered mode so that it will receive
OnDataAvailable() callbacks every time new data arrives.  Without this, the
sequencer only calls streams when the amount of readable data changes from zero
to non-zero.  This doesn't work for the deliver_on_complete mode, which requires
a callback when the last frame arrives.

Adds an accessor for QuicStreamSequencer::close_offset_.

n/a (Quartc only)

Merge internal change: 211145774

https://chromium-review.googlesource.com/c/1210166/
Bug: 112846671

--------------------------------------------------

Changes QuicStreamSequencer to handle ignore_read_data_ the same way
when level_triggered_ is true.

Currently, when level_triggered_ is true, QuicStreamSequencer calls
OnDataAvailable() each time new data may be read, regardless of
whether ignore_read_data_ has already been set.  With this change, it
does not call OnDataAvailable() if ignore_read_data_ has already been
set, and instead flushes buffered frames.  This is the same thing that
happens when new data is available if the sequencer is not
level_triggered_.

This makes QuicStreamSequencer::StopReading() work as described in the
documentation when the sequencer is level-triggered.  See
documentation here:
http://google3/third_party/quic/core/quic_stream_sequencer.h?l=83&rcl=208892429

Guarded by quic_stop_reading_when_level_triggered flag.

Merge internal change: 211145297

https://chromium-review.googlesource.com/c/1210622/

--------------------------------------------------

[QPACK] Remove QpackProgressiveEncode::string_offset_.

Totally unnecessary to keep an offset to a StringPiece, since changing
that StringPiece itself is cheap.

n/a (Prototype code not in production.)

Merge internal change: 211079590

https://chromium-review.googlesource.com/c/1210366/
Bug: 112770235

--------------------------------------------------

Add check for static streams when processing headers so that we won't try
getting dynamic streams for them. Protected by
--gfe2_restart_flag_quic_check_stream_nonstatic_on_header_list.

Merge internal change: 211017542

https://chromium-review.googlesource.com/c/1210266/
Bug: 113533504

--------------------------------------------------

Directly sample QUIC's BbrSender.

Use BandwidthSampler directly in QUIC's BbrSender instead of
referencing the interface.  No functional change, not flag protected.

Merge internal change: 211006022

https://chromium-review.googlesource.com/c/1210164/
Bug: 111468849

--------------------------------------------------

[QPACK] Reset ProgressiveDecoder::name_ and value_ before use, not after.

DCHECK(name_.empty()) and name_.clear() were kind of too far.  It makes more
sense to reset a member right before it is needed to be reset.  Note that the
total number of resets does not change: it is still called once per header
block.

n/a (Prototype code not in production.)

Merge internal change: 210957214

https://chromium-review.googlesource.com/c/1210363/
Bug: 112770235

--------------------------------------------------

Refactor QpackProgressiveEncoder.

Refactor QpackProgressiveEncoder states from Next() method into their own
private methods.  Previously there was an early return if and only if |state_|
was not changed, test this condition in Next() for breaking out of the loop.

n/a (Prototype code not in production.)

Merge internal change: 210831163

https://chromium-review.googlesource.com/c/1210362/

--------------------------------------------------

Move the flag count for FLAGS_quic_reloadable_flag_quic_use_get_cert_chain to the correct location.
The old location was not being hit because it was in the flag=false branch.

n/a - Flag count only

Merge internal change: 210794711

https://chromium-review.googlesource.com/c/1210264/

--------------------------------------------------

Add a FLAG_COUNT for FLAGS_quic_reloadable_flag_quic_new_lru_cache

Merge internal change: 210772455

https://chromium-review.googlesource.com/c/1210263/

--------------------------------------------------

Enable probing retransmissions and flexible app-limited handling in Quartc.

Probing retransmissions (fill_up_link_during_probing) drive QUIC out of STARTUP
when Quartc does not provide enough application data to reach QUIC's STARTUP
pacing rate.  QUIC retransmits existing data in order to probe for higher
bandwidth and eventually exit startup.

Flexible app-limited handling (quic_bbr_flexible_app_limited + BBR9) cause BBR
to ignore OnAppLimited() if there is already enough data in-flight for BBR to
make progress.  Under this mode, BBR only requests probing packets if it doesn't
have enough data in-flight.

Flexible app-limited handling makes it possible for Quartc to enable probing
retransmissions without starving itself, as this ensures there is some bandwidth
left over for Quartc to send new data.

QuicCallbacksTest needs to disable fill_up_link_during_probing in two test cases
in order to better control what data is written and when.  This applies to the
two cases for reporting buffered data.  These test cases require scenarios where
specific data is buffered or retransmitted.  Driving QUIC out of STARTUP to
avoid interference from probing retransmissions occasionally got stuck and timed
out, so simply disabling probing retransmissions makes it easier to cover these
cases.

n/a (Quartc only)

Merge internal change: 210728277

https://chromium-review.googlesource.com/c/1210163/
Bug: 112192494

--------------------------------------------------

Check handshaker is not null before calling
QuicCryptoServerHandshaker::CancelOutstandingCallbacks

Merge internal change: 210726104

https://chromium-review.googlesource.com/c/1210142/

--------------------------------------------------

Signal error on unknown opcodes in QpackDecoder.

All these opcodes will have to be implemented, so no reason to add test for
these errors now.

Before this CL, QpackDecoder just blindly assumed that the opcode was for
Literal Header Field Without Name Reference.

n/a (Prototype code not used in production.)

Merge internal change: 210716471

https://chromium-review.googlesource.com/c/1210063/

--------------------------------------------------

QPACK: factor out wire format binary constants.

n/a (Prototype code not used in production.)

Merge internal change: 210600833

https://chromium-review.googlesource.com/c/1210082/
Bug: 112770235

--------------------------------------------------

Change HpackVarintDecoder signature.

HpackVarintEncoder::StartEncoding() already takes |prefix_length| instead of
|prefix_mask|.  Sure if HpackVarintDecoder::Start* takes |prefix_mask| that
allows prefix masks to be baked in production code, cutting down on one CPU
instruction per Huffman encoding.  I argue that changing the interface to
|prefix_length| is cleaner and is worth the extra cost.

HpackVarintDecoder refactor, no functional change intended.

Merge internal change: 210596803

https://chromium-review.googlesource.com/c/1209984/
Bug: 112770235

--------------------------------------------------

Add a server_config_id argument to KeyExchangeSource::Create so that LetoKeyExchangeSource can use it.

n/a - unused code

Merge internal change: 210578371

https://chromium-review.googlesource.com/c/1209523/

--------------------------------------------------

Move HPACK integer encoder and decoder into a new common directory.

Refactor, no functional change.

Merge internal change: 210561940

https://chromium-review.googlesource.com/c/1208697/
Bug: 112770235

--------------------------------------------------

Add Huffman support in QpackEncoder.

Also remove silly DCHECK from http2::HuffmanEncode() to make life easier.

n/a (Prototype code not in production.)

Merge internal change: 210537018

https://chromium-review.googlesource.com/c/1209212/
Bug: 112770235

--------------------------------------------------

Add Huffman support in QpackDecoder.

n/a (Prototype code not in production.)

Merge internal change: 210454064

https://chromium-review.googlesource.com/c/1209211/
Bug: 112770235

--------------------------------------------------

Add QUIC_FLAG_COUNT for FLAGS_quic_reloadable_flag_quic_use_get_cert_chain.

n/a - Flag count only

Merge internal change: 210451731

https://chromium-review.googlesource.com/c/1209210/

--------------------------------------------------

When fill_up_link_during_probing_ is enabled by Quartc, check whether
more inflight is necessary after every probing packet is sent. Adds a
BBR9 connection option that only treats samples as app-limited if
bytes_in_flight is too low to observe a higher bandwidth.  Connection
option protected by
FLAGS_quic_reloadable_flag_quic_bbr_flexible_app_limited,
fill_up_link_during_probing_ not used in GFE.

Replaces cr/191675927

Merge internal change: 210409167

https://chromium-review.googlesource.com/c/1209170/
Bug: 77975811

--------------------------------------------------

Add QPACK fuzzers.

Add fuzzer for QpackDecoder.  Add another fuzzer for round-trips that should be
able to cover all code paths of QpackEncoder (but not QpackDecoder) and also
verify that we get back the original header list.  Add corpus for each fuzzer
that is minimized and covers almost all of respective source files.  Change
QpackDecoder to signal error instead of crashing on Huffman-encoded input.

Copy Chromium's FuzzedDataProvider to quic/platform.  Note to merge rotation:
in Chromium, quic/platform/impl/quic_fuzzed_data_provider.h will be trivial:
using QuicFuzzedDataProviderImpl = base::FuzzedDataProvider;

n/a (Prototype code not in production.)

Merge internal change: 210399075

https://chromium-review.googlesource.com/c/1208934/
Bug: 112770235

--------------------------------------------------

[QPACK] Introduce FragmentSizeGenerator.

This abstraction will be used for fuzzing.

n/a (Prototype code not in production.)

Merge internal change: 210394197

https://chromium-review.googlesource.com/c/1208677/
Bug: 112770235

--------------------------------------------------

Improve readability of QuicSentPacketManager::MarkForRetransmission

Improve the readability of QuicSentPacketManager::MarkForRetransmission
by using an early return and some helper methods.

n/a - reorganizing code

Merge internal change: 210193731

https://chromium-review.googlesource.com/c/1208676/

--------------------------------------------------

Pack QuicStreamFrame into QuicFrame more tightly to save 8 bytes in
QuicFrame.  Not flag protected, no functional change.

Merge internal change: 210144041

https://chromium-review.googlesource.com/c/1208437/
Bug: 112563066

--------------------------------------------------

Add BytesPendingRetransmission() to BufferedDataBytes() in QuartcCallbacks.

Retransmissions count against QUIC's pacing rate and congestion window exactly
as initial transmissions do.  Therefore, Quartc should treat data pending
retransmission exactly the same way it treats data that is buffered for initial
transmission.  In particular, both pending retransmissions and initial
transmissions should be accounted for in buffer-based back-pressure.

QUIC now calls OnBufferChanged(quic::QuartcStream*) whenever a stream frame is
lost or retransmitted.  QuartcStream::BytesPendingRetransmission() reports the
number of bytes waiting to be retransmitted for each stream.

Note: QuartcStream already handles the corner case of streams where
cancel_on_loss is set.  These streams will never retransmit, and always report
zero BytesPendingRetransmission().

QuartcCallbacks can account for both initial transmissions and retransmissions
by adding BytesPendingRetransmission() to BufferedDataBytes() before reporting
the value to RateControl.  (RateControl applies back-pressure to encoders.)

n/a (testonly, factors the packet filter out of QuartcSessionTest)

Merge internal change: 210140689

https://chromium-review.googlesource.com/c/1208433/

--------------------------------------------------

Log application-limited events and external parameters into QuicTrace

n/a (code not used in production)

Merge internal change: 210140046

https://chromium-review.googlesource.com/c/1208671/

--------------------------------------------------

Rename methods and member classes.

Rename member functions (methods) and member classes if part of the name is
redundant with the name of the enclosing class.

This is a follow-up to https://critique.corp.google.com/#review/209843547/depot/google3/third_party/quic/core/qpack/qpack_decoder.h&l=44R&vr=s9

Rename methods and classes.

Merge internal change: 210116798

https://chromium-review.googlesource.com/c/1208670/
Bug: 112770235

--------------------------------------------------

Add QPACK round trip tests; factor out QpackTestUtils.

Add QPACK round trip tests.  Factor out things common in QpackEncoderTest,
QpackDecoderTest, and QpackRoundTripTest into a new library.

n/a (Prototype code not in production.)

Merge internal change: 210042125

https://chromium-review.googlesource.com/c/1208413/
Bug: 112770235

--------------------------------------------------

Allow empty QPACK encoder/decoder input.

Allow decoding empty input in QpackDecoder::QpackProgressiveDecoder::Decode().
Allow encoding empty input in QpackEncoderTest::Encode() (test only).

n/a (Prototype code not in production.)

Merge internal change: 210039980

https://chromium-review.googlesource.com/c/1208430/
Bug: 112770235

--------------------------------------------------

Add QuicCrypter base class. No behavior change.

Merge internal change: 210002082

https://chromium-review.googlesource.com/c/1207614/

--------------------------------------------------

Move QpackVarintEncoder to third_party/http2

...and rename it to HpackVarintEncoder.  This CL buys us the following things:

* Free roundtrip tests, as HpackVarintDecoderTest uses
HpackBlockBuilder::AppendHighBitsAndVarint, which now relies on
HpackVarintEncoder.
* QPACK uses HPACK integer representation, so HpackVarintEncoder is a more
appropriate class name than QpackVarintEncoder, and third_party/http2 is a
more appropriate place than third_party/quic.
* Some reduction of code duplication in
HpackBlockBuilder::AppendHighBitsAndVarint().

The downside is that embedders that embed third_party/http2 but not
third_party/quic now have to include HpackVarintEncoder in production, resulting
in a slight increase in binary size.  Embedders of both third_party/http2 and
third_party/quic are not affected in the long run, because they need the encoder
anyway.  (This is not the case quite yet as third_party/quic/core/qpack is not
yet included in production code.)

Move class from third_party/quic to third_party/http2.

Merge internal change: 209978253

https://chromium-review.googlesource.com/c/1207895/
Bug: 112770235

--------------------------------------------------

Allow QUIC simulator alarms to be scheduled in the past

As discovered in b/112928831, this is something that does happen in real world
when write-blocking is involved, hence it would make sense to support it in
simulations too.

n/a (test-only change)

Merge internal change: 209975070

https://chromium-review.googlesource.com/c/1206434/
Bug: 112928831

--------------------------------------------------

Updates a DCHECK that fires in an end-to-end test when a client
disconnects immediately after sending a request.

This fixes the secure_quic test failures observed in the presubmits
for CL 209770890.

Modifies a debug-only assertion in QuicSpdyServerStreamBase::
CloseWriteSide(). No functional change for optimized builds.

Merge internal change: 209973796

https://chromium-review.googlesource.com/c/1206673/
Bug: 111649083

--------------------------------------------------

Refactor QuicLRUCache to use QuicLinkedHashMap instead of
SimpleLRUCache.  No functional change.  Protected by
FLAGS_quic_reloadable_flag_quic_new_lru_cache

This is a part of the effort to reduce QUIC platform dependencies, but this
would also have a nice effect of reducing the memory usage since SimpleLRUCache
claims it has at least 100 bytes of overhead per entry.

Merge internal change: 209969156

https://chromium-review.googlesource.com/c/1206432/

--------------------------------------------------

n/a, test only. Fix two typo in key exchange tests.

Merge internal change: 209901814

https://chromium-review.googlesource.com/c/1205837/

--------------------------------------------------

Fix bug in QpackVarintEncoder.

Turns out QpackVarintEncoder::varint_ is not a good way to keep track of whether
encoding is in progress.  If an integer just barely does not fit on the prefix,
then a zero byte needs to be emitted after the first byte.  However, since
|varint_| is zero, the encoder incorrectly omits this byte.

Note that this cannot happen with subsequent bytes.  If an integer does not fit
in the prefix and some bytes that follow, then the leftover part must be
non-zero.  If it wasn't, then the continuation bit on the last byte would not be
set.  This corner case can only happen if |varint == prefix_mask|, because the
byte that holds the prefix does not have a continuation bit as a mechanism to
signal continuation.

This CL fixes that by introducing a dedicated Boolean member to keep track of
whether encoding is in progress.  This Boolean is |encoding_in_progress_|
instead of |encoding_is_done_|, and consequently the accessor is renamed to
EncodingInProgress(), and every call site has to be inverted.  "Done" does not
make much sense for something that would need to be true upon construction
before any encoding is performed.

This CL also adds test coverage in QpackVarintEncoderTests, and also in
QpackEncoderTests.

n/a (Prototype code not in production.)

Merge internal change: 209872387

https://chromium-review.googlesource.com/c/1206696/
Bug: 112770235

--------------------------------------------------

Implement primitive QPACK decoder.

This primitive decoder only decodes non-Huffman encoded string literals without
name reference.

n/a (Prototype code not in production.)

Merge internal change: 209843547

https://chromium-review.googlesource.com/c/1206695/
Bug: 112770235

--------------------------------------------------

Factor out QpackVarintEncoder.

Factor out parts of QpackEncoder into its own QpackVarintEncoder class.  This is
a refactor only, improvements and bugfixes will come in subsequent CLs.

n/a (Prototype code not in production.)

Merge internal change: 209801638

https://chromium-review.googlesource.com/c/1206694/
Bug: 112770235

--------------------------------------------------

Add a BytesPendingRetransmission() method to QuartcStream.  Call
Delegate::OnBufferChanged() whenever a stream frame is lost or retransmitted.

This allows Quartc to account for bytes pending retransmission along with
buffered bytes when applying back-pressure on data sources.

Right now, QuartcStream sums the lengths of all intervals in the set of
pending retransmissions.  Since Quartc streams carry only a single audio or
video frame, there should generally be a small number of lost frames, so it
should not be an issue to add them up.  We can optimize later, if necessary.

Makes QuicStream::OnStreamFrameLost() virtual so that Quartc can be
notified when it is called.  Adds an accessor for
QuicStreamSendBuffer::pending_retransmission_.  These are only used in Quartc.

Merge internal change: 209786500

https://chromium-review.googlesource.com/c/1206693/

--------------------------------------------------

Fix QPACK encoder bug with zero length names and values.

Encoder was not working properly with small fragments on state transitions,
because kNameString and kValueString states would have needed to be run even if
there was no space in buffer.  Change logic so that every state actually writes
to the buffer, by skipping these states if name or value is empty.

Parametrize all QPACK encoder tests to write the whole buffer at once or to
write octet by octet.

n/a (Prototype code not in production.)

Merge internal change: 209759976

https://chromium-review.googlesource.com/c/1206671/
Bug: 112770235

--------------------------------------------------

Deprecate FLAGS_quic_reloadable_flag_quic_retransmissions_app_limited.

n/a (Flag deprecation)

Merge internal change: 209746153

https://chromium-review.googlesource.com/c/1206690/

--------------------------------------------------

Implement primivite QPACK encoder.

This primitive encoder only emits non-Huffman encoded string literals.
Tests are very rudimentary.

n/a (Prototype code not in production.)

Merge internal change: 209742299

https://chromium-review.googlesource.com/c/1206670/
Bug: 112770235

--------------------------------------------------

Update and expand comments in quic_versions.h

n/a (comments-only change)

Merge internal change: 209720017

https://chromium-review.googlesource.com/c/1205531/

--------------------------------------------------

Deprecate FLAGS_quic_reloadable_flag_quic_ack_reordered_packets.

n/a (Flag deprecation)

Merge internal change: 209699826

https://chromium-review.googlesource.com/c/1205198/

--------------------------------------------------

Modify CryptoMessageParser::ProcessInput to take an EncryptionLevel

This is in preparation for the new interface between TLS and QUIC, where when
the TLS stack is provided with a handshake message from QUIC, it also needs to
be provided with the EncryptionLevel at which the message was received.

Refactor QUIC's CryptoMessageParser. No behavior change.

Merge internal change: 209514195

https://chromium-review.googlesource.com/c/1205314/
Bug: 112643533

--------------------------------------------------

QUIC: remove unused Perspective params from CryptoFramer

Merge internal change: 209510531

https://chromium-review.googlesource.com/c/1205196/
Bug: 112643533

--------------------------------------------------

Add a basic HTTP over QUIC framer which is capable of parsing HTTP over QUIC frames.
Serialization is still lacking, as are some performance optimizations.

n/a - Unused code

Merge internal change: 209504683

https://chromium-review.googlesource.com/c/1204672/

R=rch@chromium.org

Change-Id: Ibeb5bee8cc6a41c7d7064b896f5554193f791152
Reviewed-on: https://chromium-review.googlesource.com/1213936Reviewed-by: Ryan Hamilton <rch@chromium.org>
Commit-Queue: Frank Kastenholz <fkastenholz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589622}

The current ChromeOS's camera stack (cros-camera) assumes that
capture_unittests would run in a single process.

Bug: 881682
Change-Id: I7a89f63ee16c7eb97b83b15ed46ee8061d1d601d
Reviewed-on: https://chromium-review.googlesource.com/1212433
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Ben Pastene <bpastene@chromium.org>
Reviewed-by: Achuith Bhandarkar <achuith@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589621}

This reverts commit 41aadb51.

Reason for revert: Introduces static initializers according to https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/mac-rel/552.

Original change's description:
> [jumbo] Make it one g_id_map instead of two.
> 
> A commit just introduced a second g_id_map in ui/views/cocoa,
> which broke jumbo builds. This renames the g_id_map
> globals to the longer names g_id_to_impl_map and
> g_id_to_host_impl_map.
> 
> Bug: 859152
> 
> TBR=ccameron@chromium.org,ellyjones@chromium.org
> 
> Change-Id: Id30d71b67f4bfb7c3ddc3ec8d03c759b5b7df14f
> Reviewed-on: https://chromium-review.googlesource.com/1212862
> Reviewed-by: Daniel Bratell <bratell@opera.com>
> Commit-Queue: Daniel Bratell <bratell@opera.com>
> Cr-Commit-Position: refs/heads/master@{#589465}

TBR=ellyjones@chromium.org,ccameron@chromium.org,bratell@opera.com

Change-Id: I70fdf14f4759f74e9572e0b69884febb2607adcc
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 859152
Reviewed-on: https://chromium-review.googlesource.com/1213705Reviewed-by: Justin Donnelly <jdonnelly@chromium.org>
Commit-Queue: Justin Donnelly <jdonnelly@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589620}

This reverts commit 99e4e0f3.

Reason for revert: Looks like this broke compile, possibly due to conflicts w/ https://chromium.googlesource.com/chromium/src/+/19e91cc9aefac767e9482351123b11f1bb6e2c70

Original change's description:
> [AF] Make IsCreditCardUploadEnabled use ActiveAccount email if in Butter
> 
> It used to get the Primary Account's email but should get the active
> account's one instead if in Butter state.
> 
> Bug: 881730
> Change-Id: I1e48d39bd1c908bee82034e1eb7b40231f6c788d
> Reviewed-on: https://chromium-review.googlesource.com/1213444
> Reviewed-by: Mathieu Perreault <mathp@chromium.org>
> Commit-Queue: Sebastien Seguin-Gagnon <sebsg@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#589617}

TBR=mathp@chromium.org,sebsg@chromium.org,feuunk@chromium.org

Change-Id: Ida5dc715c7eb6618d213a676f00a0476b0cd2cad
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 881730
Reviewed-on: https://chromium-review.googlesource.com/1214105Reviewed-by: John Budorick <jbudorick@chromium.org>
Commit-Queue: John Budorick <jbudorick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589619}

ES team would like to use this infrastructure.

Change-Id: Ia997c91d1525c36690aadf8f47f0b0cf193201df
Reviewed-on: https://chromium-review.googlesource.com/1213312Reviewed-by: Cathy Li <chili@chromium.org>
Commit-Queue: Justin DeWitt <dewittj@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589618}

It used to get the Primary Account's email but should get the active
account's one instead if in Butter state.

Bug: 881730
Change-Id: I1e48d39bd1c908bee82034e1eb7b40231f6c788d
Reviewed-on: https://chromium-review.googlesource.com/1213444Reviewed-by: Mathieu Perreault <mathp@chromium.org>
Commit-Queue: Sebastien Seguin-Gagnon <sebsg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589617}

The IceTransportAdapter splits P2PTransportChannel-specific logic
out of IceTransportHost and puts it behind an interface to
facilitate future testing. The API methods are changed to map to
the requirements of the RTCIceTransport rather than the
implementation details of the P2PTransportChannel.

Bug: 864871
Change-Id: I2a7036f675a11460e9721cc937beec17b7f03cc9
Reviewed-on: https://chromium-review.googlesource.com/1200127
Commit-Queue: Steve Anton <steveanton@chromium.org>
Reviewed-by: Henrik Boström <hbos@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589616}

Bug: 879147
Change-Id: I861ceb38370a0de959c6881735819097a030cb09
Reviewed-on: https://chromium-review.googlesource.com/1202542
Commit-Queue: Martin Robinson <mrobinson@igalia.com>
Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589615}

It looks like there is bug in GCC 6, which cannot go through
structure initialization normally.

Thus, instead of a default initialization of one of the members,
explicitly initialize it to a default value.

Change-Id: Ia55cc6658e6b6b2f8a80c2582dd28f001c9e648c
Reviewed-on: https://chromium-review.googlesource.com/1213181Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Maksim Sisov <msisov@igalia.com>
Cr-Commit-Position: refs/heads/master@{#589614}

This adds TestBrowserThreadBundles to unittests that will need it when
we replace BrowserThread::PostTask() and friends with base/post_task.h.

The TestBrowserThreadBundle initializes globals that are required for
things like posting a task to or obtaining a TaskRunner for a BrowserThread.

I'm making this change in a separate patch to make review of the
future automated refactoring patches easier.

Depends on https://crrev.com/c/1185887.

Bug: 878356
Change-Id: Ifd332b995e3449b26d9c93283fb708792b2e66eb
Reviewed-on: https://chromium-review.googlesource.com/1208050Reviewed-by: Derek Cheng <imcheng@chromium.org>
Reviewed-by: Jialiu Lin <jialiul@chromium.org>
Commit-Queue: Eric Seckler <eseckler@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589613}

We added a call to sample() to get the initial counters in the Initialize
method to avoid loosing metrics for the first cycle. Turn out this was not
the corect fix.

For a new process, the function Initialize and Sample can get called in
the sample thread quantum which may lead to a zero interval.

The appropriate fix to get the initial counters will be to trigger a
collecting cycle at the same site we register the one-shot timer. That way
the first cycle is not lost.

This CL only intent to fix the crash.

0x503f589d (chrome.dll - logging.cc: 863)	logging::LogMessage::~LogMessage()
0x516404dd (chrome.dll - process_metrics.cc: 117)	base::ProcessMetrics::GetPlatformIndependentCPUUsage()
0x529074a5 (chrome.dll - process_metrics_history.cc: 51)	performance_monitor::ProcessMetricsHistory::SampleMetrics()
0x5247aafb (chrome.dll - performance_monitor.cc: 199)	performance_monitor::PerformanceMonitor::UpdateMetricsOnIOThread(int)
0x50a78a4e (chrome.dll - bind_internal.h: 658)	base::internal::Invoker<base::internal::BindState<content::ServiceWorkerDatabase::Status (content::ServiceWorkerDatabase::*)(std::vector<content::ServiceWorkerDatabase::RegistrationData,std::allocator<content::ServiceWorkerDatabase::RegistrationData> > *) __attribute__((thiscall)),base::internal::UnretainedWrapper<content::ServiceWorkerDatabase>,std::vector<content::ServiceWorkerDatabase::RegistrationData,std::allocator<content::ServiceWorkerDatabase::RegistrationData> > *>,content::ServiceWorkerDatabase::Status ()>::RunOnce
0x502c65b5 (chrome.dll - task_annotator.cc: 101)	base::debug::TaskAnnotator::RunTask(char const *,base::PendingTask *)
0x502c9585 (chrome.dll - message_loop.cc: 434)	base::MessageLoop::RunTask(base::PendingTask *)
0x502c9462 (chrome.dll - message_loop.cc: 445)	base::MessageLoop::DeferOrRunPendingTask(base::PendingTask)
0x5023e22e (chrome.dll - message_loop.cc: 517)	base::MessageLoop::DoWork()
0x5023e144 (chrome.dll - message_pump_win.cc: 494)	base::MessagePumpForIO::DoRunLoop()
0x5023e0a0 (chrome.dll - message_pump_win.cc: 52)	base::MessagePumpWin::Run(base::MessagePump::Delegate *)
0x5023dff7 (chrome.dll - message_loop.cc: 386)	base::MessageLoop::Run(bool)
0x5023dcb3 (chrome.dll - run_loop.cc: 102)	base::RunLoop::Run()
0x5023dbbb (chrome.dll - thread.cc: 262)	base::Thread::Run(base::RunLoop *)
0x5023daf3 (chrome.dll - browser_process_sub_thread.cc: 175)	content::BrowserProcessSubThread::IOThreadRun(base::RunLoop *)
0x5023dab8 (chrome.dll - browser_process_sub_thread.cc: 127)	content::BrowserProcessSubThread::Run(base::RunLoop *)
0x5023ad44 (chrome.dll - thread.cc: 357)	base::Thread::ThreadMain()
0x5164f0e4 (chrome.dll - platform_thread_win.cc: 100)	base::`anonymous namespace'::ThreadFunc
0x778aefab (kernel32.dll + 0x0004efab)	BaseThreadInitThunk
0x77753627 (ntdll.dll + 0x00063627)
0x777535fa (ntdll.dll + 0x000635fa)

R=oysteine@chromium.org

Bug: 880071
Change-Id: I33acf14f4eaae007edd265c0230def783ca35393
Reviewed-on: https://chromium-review.googlesource.com/1213870Reviewed-by: oysteine <oysteine@chromium.org>
Commit-Queue: Etienne Bergeron <etienneb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589612}

BUG=chromium:879737
TEST=Manually tested on an eve device with YAPS.

Change-Id: I9a99d5a76178ba71fcd4017c6e43f1fad0fc77b8
Reviewed-on: https://chromium-review.googlesource.com/1208827
Commit-Queue: Tim Zheng <timzheng@chromium.org>
Reviewed-by: Nicholas Verne <nverne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589611}

https://chromium.googlesource.com/catapult.git/+log/bcc3e491fb2e..32654a45d3da


git log bcc3e491fb2e..32654a45d3da --date=short --no-merges --format='%ad %ae %s'
2018-09-07 nednguyen@google.com Remove common/battor/
2018-09-07 simonhatch@chromium.org Dashboard - Simplify find_anomalies by removing benchmark_duration
2018-09-07 simonhatch@chromium.org Dashboard - Generate stats on alert #'s
2018-09-07 nednguyen@google.com Fix BUILD.gn by running generate_telemetry_build.py
2018-09-07 ulan@chromium.org Add a metric that tracks the number of console error messages
2018-09-07 wangge@google.com Reformat the `run_benchmark` help message.
2018-09-06 benjhayden@chromium.org Fix flaky dashboard tests: timeseries2_test, file_bug_test, others.
2018-09-06 wangge@google.com Update help message for `run_benchmark.py`.
2018-09-06 benjhayden@chromium.org dashboard/bin/run_wct_tests
2018-09-06 nastasoiuf@google.com Stacked bar plot integration
2018-09-06 simonhatch@chromium.org Dashboard - Fix stats generation for chrome health metrics
2018-09-06 rmcilroy@chromium.org [Tracing] Add support for Optimize-Background bucket to V8's runtime call stats table
2018-09-06 anthonyalridge@google.com Add back bar chart interaction
2018-09-06 mattm@chromium.org netlog_viewer: rename |global| in cr.js to |cr_global|
2018-09-05 mattm@chromium.org netlog_viewer: Apply chromium changes up to 940b87bb
2018-09-05 jbudorick@chromium.org Fix systrace/bin/OWNERS.


Created with:
  gclient setdep -r src/third_party/catapult@32654a45d3da

The AutoRoll server is located here: https://autoroll.skia.org/r/catapult-autoroll

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.

CQ_INCLUDE_TRYBOTS=luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel

BUG=chromium:881543,chromium:881379,chromium:881361,chromium:881543,chromium:880432,chromium:863390,chromium:877660,chromium:866423,chromium:879229,chromium:866423
TBR=sullivan@chromium.org

Change-Id: I2206a43b38a09c7427d4d4ef02fe4c820dbaee98
Reviewed-on: https://chromium-review.googlesource.com/1213476Reviewed-by: catapult-chromium-autoroll <catapult-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: catapult-chromium-autoroll <catapult-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#589610}

Use a ScrollView instead of the Stream for the NTP if supervised
user or enterprise policy is enabled.

Bug: 862832
Change-Id: I8987db14dac17ef525860817fcc341b254fea358
Reviewed-on: https://chromium-review.googlesource.com/1208814
Commit-Queue: Becky Zhou <huayinz@chromium.org>
Reviewed-by: Theresa <twellington@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589609}

WT_EXECUTEINWAITTHREAD is theoretically slightly more efficient than
WT_EXECUTEDEFAULT when running callbacks registered by
RegisterWaitForSingleObject but it comes with an ill-defined risk of
deadlocks. We seem to be hitting those, so this change switches to
the safer flag.

If no improvement is seen then this change will be reverted.

Bug: 877012
Change-Id: I773cad8cd0621fb487770306bee4cc9812d4f243
Reviewed-on: https://chromium-review.googlesource.com/1200484
Commit-Queue: Bruce Dawson <brucedawson@chromium.org>
Reviewed-by: Greg Thompson <grt@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589608}

SigninManagerAndroid::SignOut() has historically hard coded the reason
as signin_metrics::USER_CLICKED_SIGNOUT_SETTINGS. This CL begins to
address the TODO there by plumbing the reason from the java layer and
various callers.

For now the CL preserves the existing "USER_CLICKED" everywhere. Follow
up work should change this reason to reflect the real trigger.

Change-Id: Id225b9bf5e9b9fd5771c2353616c6a1a82e60af8
Bug: 873116
Reviewed-on: https://chromium-review.googlesource.com/1170763Reviewed-by: David Roger <droger@chromium.org>
Reviewed-by: Theresa <twellington@chromium.org>
Reviewed-by: Bernhard Bauer <bauerb@chromium.org>
Reviewed-by: Boris Sazonov <bsazonov@chromium.org>
Commit-Queue: Chrome Cunningham <chcunningham@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589607}

Integers should be parsed per the <integer> production, and thus should
not accept units.

In whitespace-length.html there was a bug in the test-setup that made
some tests fail that ought to pass (incorrect parameter.)

Also take the opportunity to clean up the tests a bit.

Change-Id: Id7b8964ec52f484dad424b0a1b7be6d70d062eb6
Reviewed-on: https://chromium-review.googlesource.com/1213149Reviewed-by: Philip Rogers <pdr@chromium.org>
Commit-Queue: Fredrik Söderquist <fs@opera.com>
Cr-Commit-Position: refs/heads/master@{#589606}

This reverts commit 730fa6b3.

Reason for reland: Added missing dep.

TBR=jdonnelly@chromium.org,agrieve@chromium.org,tiborg@chromium.org

Change-Id: If982cfe9cb032bfcb8dfa15f8d5b618511ec824a
Reviewed-on: https://chromium-review.googlesource.com/1213306
Commit-Queue: agrieve <agrieve@chromium.org>
Reviewed-by: agrieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589605}

Changes:
Hide the search results page when search box is closed and reshow it
when search box is opened, so the focus will not move onto the suggested
apps in search results page when zero state is enabled.

Bug: 880980
Change-Id: I6f32f0841fba094bc731bb0447f0f2cb055557b7
Reviewed-on: https://chromium-review.googlesource.com/1213926Reviewed-by: Alex Newcomer <newcomer@chromium.org>
Commit-Queue: Weidong Guo <weidongg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589604}

It no longer fails.

BUG=664261
TEST=test only change

Change-Id: Id17c9b0950fd20e001b27f104912542daaf8bc8b
Reviewed-on: https://chromium-review.googlesource.com/1213931Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589603}

https://skia.googlesource.com/skia.git/+log/459c914be7a2..1fc5fbd80451


git log 459c914be7a2..1fc5fbd80451 --date=short --no-merges --format='%ad %ae %s'
2018-09-07 michaelludwig@google.com Reland "Redo how extra emit code flushing operates" with type fix.


Created with:
  gclient setdep -r src/third_party/skia@1fc5fbd80451

The AutoRoll server is located here: https://autoroll.skia.org/r/skia-autoroll

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel
TBR=jvanverth@chromium.org

Change-Id: I9d1b5f62e000198b09d38c713fddb9123e8323db
Reviewed-on: https://chromium-review.googlesource.com/1213479Reviewed-by: skia-chromium-autoroll <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: skia-chromium-autoroll <skia-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#589602}

TBR=tikuta@chromium.org

Bug: 330263
Change-Id: I45e4e7597eeaf602c4cb3daafe66e73f33574cc9
Reviewed-on: https://chromium-review.googlesource.com/1213536Reviewed-by: Nico Weber <thakis@chromium.org>
Commit-Queue: Nico Weber <thakis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589601}

Approved by UX.

Bug: 877111
Change-Id: Ib890a03e22bbb017848cf413fbbc0e9e18a81e5d
Reviewed-on: https://chromium-review.googlesource.com/1196916Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Reviewed-by: Alex Newcomer <newcomer@chromium.org>
Commit-Queue: Daniel Zhang <oxyflush@google.com>
Cr-Commit-Position: refs/heads/master@{#589600}

BUG=881237

Change-Id: I7dc76e78f958996d1f61ccb45209e7afb8854f67
Reviewed-on: https://chromium-review.googlesource.com/1213927Reviewed-by: Yusuf Ozuysal <yusufo@chromium.org>
Commit-Queue: Theresa <twellington@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589599}

Create a place for Chrome App APIs to go, and move the chrome.browser
API there. This is the first API moved from Chrome Apps into the apps-
specific directory. This involves a number of steps:
- Create a chrome/common/apps/platform_apps/ directory, and create
  ChromeAppsAPIProvider and ChromeAppsAPIPermissions. The APIProvider
  is then registered with the singleton ExtensionsClient to provide
  information about the Chrome Apps APIs to the extension system.
- Create chrome/common/apps/platform_apps/api. This includes the API
  schema definitions (right now, just browser.idl), the features files
  (_api_features.json and _permission_features.json), and the BUILD.gn
  to generate the C++ types and JSON API strings. Types are generated
  in the apps:: namespace, instead of the extensions:: namespace. They
  are similarly generated into chrome/common/apps/platform_apps/api
  instead of chrome/common/extensions/api.
- Add a ChromeAppsBrowserAPIProvider to register the app-related
  ExtensionFunctions with the Extensions system.
- Move the browser API implementation (browser_api.*) to
  chrome/browser/apps/platform_apps/api/browser/, and consume the
  generated types.
- Register the new API location with the docserver so that it can look
  up the API schemas, and bump the docserver version.

With this, we have all the pieces necessary to continue moving App-
specific APIs en masse to the App-specific directories, and begin to
carve out the dependency from the extensions system.

Bug: 877240

Change-Id: If3eba512c24ef3d1c27ed5ff3f215f56eb13a818
Reviewed-on: https://chromium-review.googlesource.com/1187129
Commit-Queue: Devlin <rdevlin.cronin@chromium.org>
Reviewed-by: Dominick Ng <dominickn@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Reviewed-by: Istiaque Ahmed <lazyboy@chromium.org>
Reviewed-by: Karan Bhatia <karandeepb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589598}

This change adds a getTransports() method to
AuthenticatorAttestationResponse[1]. This returns a non-empty sequence
of AuthenticatorTranports[2] (without duplicates) where the first value
is the transport used by the registration and the following are
additional transports that the authenticator supports.

This is behind a flag for now while under discussion.
See https://github.com/w3c/webauthn/pull/1050.

TBRs:
  mnissler: trivial change to keep up with APIs in fake_certificate.cc
  sergeyu: trivial change to keep up with APIs in rsa_key_pair.cc

TBR=sergeyu,mnissler

[1] https://www.w3.org/TR/webauthn/#authenticatorattestationresponse
[2] https://www.w3.org/TR/webauthn/#transport

Change-Id: I63738f2aa6243c3f2ce2ebe684b70f1caa1c7ef3
Reviewed-on: https://chromium-review.googlesource.com/1180083
Commit-Queue: Adam Langley <agl@chromium.org>
Reviewed-by: Philip Jägenstedt <foolip@chromium.org>
Reviewed-by: Nasko Oskov <nasko@chromium.org>
Reviewed-by: Robert Sesek <rsesek@chromium.org>
Reviewed-by: Balazs Engedy <engedy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589597}

This reverts commit ca672c42.

Reason for revert: Root cause of trybot failures is still unclear.  Creating this revert to run trybots.

Original change's description:
> Revert "[Autofill] Populate ChromeUserContext in Payments requests"
> 
> This reverts commit d9183af5.
> 
> Reason for revert: suspected of causing https://bugs.chromium.org/p/chromium/issues/detail?id=881420
> 
> Original change's description:
> > [Autofill] Populate ChromeUserContext in Payments requests
> > 
> > This change adds |chrome_user_context| and |full_sync_enabled| values to
> > Unmask, GetDetails, Upload, and MigrateCards requests, as long as either
> > of the appropriate flags are enabled.
> > 
> > Bug: 840703
> > Change-Id: Ieb3baf980b68add43a3111c0902a03b574b1adcf
> > Reviewed-on: https://chromium-review.googlesource.com/1208434
> > Commit-Queue: Jared Saul <jsaul@google.com>
> > Commit-Queue: Sebastien Seguin-Gagnon <sebsg@chromium.org>
> > Reviewed-by: Sebastien Seguin-Gagnon <sebsg@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#589060}
> 
> TBR=sebsg@chromium.org,jsaul@google.com,dlkumar@google.com
> 
> Change-Id: Id88b0a194dad400be5ba95a36f11d8ebb5ccd1ec
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: 840703
> Reviewed-on: https://chromium-review.googlesource.com/1211282
> Reviewed-by: John Budorick <jbudorick@chromium.org>
> Commit-Queue: John Budorick <jbudorick@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#589215}

TBR=sebsg@chromium.org,jsaul@google.com,jbudorick@chromium.org,dlkumar@google.com

Change-Id: Ibaab351513bef9c23a4da09cfd8c72514d49731a
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 840703
Reviewed-on: https://chromium-review.googlesource.com/1211728Reviewed-by: Moe Ahmadi <mahmadi@chromium.org>
Reviewed-by: Sebastien Seguin-Gagnon <sebsg@chromium.org>
Reviewed-by: Jared Saul <jsaul@google.com>
Commit-Queue: Jared Saul <jsaul@google.com>
Cr-Commit-Position: refs/heads/master@{#589596}

No behaviorial changes, just moves the lite page tests to a separate
file.

Bug: 864635
Change-Id: I77a2b542febf1b7052085b401d3af13bbb41c825
Reviewed-on: https://chromium-review.googlesource.com/1213472
Commit-Queue: Robert Ogden <robertogden@chromium.org>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589595}

For windows with opener |scriptMessage.frameInfo.securityOrigin| returns
opener's security origin, while document URL can be about:blank.
This change allows all the messages coming from the main frame, which
was the orginal intention of the code.

This CL does not have a test because code that filters cross-origin
messages will be removed anyway.

Bug: 879464
Change-Id: I15dac35bb3911ad93fc6bd2fe18d1b3c2385e649
Reviewed-on: https://chromium-review.googlesource.com/1212402Reviewed-by: Mike Dougherty <michaeldo@chromium.org>
Commit-Queue: Eugene But <eugenebut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589594}

As we may have two envs, WidgetFocusManager needs some context to know which
WidgetFocusManager to return.

BUG=874554
TEST=covered by tests

Change-Id: I397e14496351fc9839273aa7284c55cdb35555c8
Reviewed-on: https://chromium-review.googlesource.com/1212088Reviewed-by: Jun Mukai <mukai@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589593}

Bug: 878755
Cq-Include-Trybots: luci.chromium.try:ios-simulator-cronet;luci.chromium.try:ios-simulator-full-configs
Change-Id: I00f53e806094ecda6449def21b12b54fe386fc47
Reviewed-on: https://chromium-review.googlesource.com/1212271Reviewed-by: Peter Lee <pkl@chromium.org>
Commit-Queue: Eugene But <eugenebut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589592}

Cast extensions and embedded frames should not navigate by default.

Bug: internal b/74124275
Bug: internal b/112805642

Test: Try to navigate extension page, verify the navigation is blocked.
Change-Id: I84174d9ed2d48038edfef0fcb17511694dbdbe3c
Reviewed-on: https://chromium-review.googlesource.com/1212344Reviewed-by: Luke Halliwell <halliwell@chromium.org>
Commit-Queue: Sean Topping <seantopping@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589591}

Bug: 754104
Change-Id: I8f2f41bcc2c41310dbdd5f88083f5818373f2613
Reviewed-on: https://chromium-review.googlesource.com/1213473Reviewed-by: Ben Pastene <bpastene@chromium.org>
Commit-Queue: John Budorick <jbudorick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589590}

https://chromium.googlesource.com/external/github.com/KhronosGroup/SPIRV-Tools.git/+log/3e38c99f5d75..40a68547dcda


git log 3e38c99f5d75..40a68547dcda --date=short --no-merges --format='%ad %ae %s'
2018-09-07 dj2@everburning.com Roll spirv-headers in DEPS (#1884)


Created with:
  gclient setdep -r src/third_party/SPIRV-Tools/src@40a68547dcda

The AutoRoll server is located here: https://autoroll.skia.org/r/spirv-tools-chromium-autoroll

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, please contact the current sheriff, who should
be CC'd on the roll, and stop the roller if necessary.


TBR=dsinclair@chromium.org

Change-Id: I3c2b487c20eb1d270cc12d58adcd938deeb90dec
Reviewed-on: https://chromium-review.googlesource.com/1213770Reviewed-by: spirv-chromium-autoroll <spirv-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Commit-Queue: spirv-chromium-autoroll <spirv-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#589589}

This was missed in a previous CL crrev.com/c/1200524.

BUG=chromium:879737
TEST=Browsertests passed with code that uses this setting.

Change-Id: I6ca354c0a781a16d17fb2be5c12203209992e25b
Reviewed-on: https://chromium-review.googlesource.com/1211731Reviewed-by: Dan Erat <derat@chromium.org>
Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Commit-Queue: Tim Zheng <timzheng@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589588}

Bug: 881642
Change-Id: I3c1f72a6a2c39c86de60ab7132b2384ddcfa1ca4
Reviewed-on: https://chromium-review.googlesource.com/1211371
Commit-Queue: Xiaoqian Dai <xdai@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589587}

Verify that an empty blacklist and whitelist are handled properly by
adding a unit test for the behavior.

Bug: None
Change-Id: I3eda6222b211f4306acad7a9c0030b834d6e694e
Reviewed-on: https://chromium-review.googlesource.com/1211980Reviewed-by: Xiaoqian Dai <xdai@chromium.org>
Commit-Queue: Sean Kau <skau@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589586}

This CL eliminates the "Autofill and payments" page and move "Addresses
and more" and "Payment methods" into the main Settings page.

A follow-up CL will fix allow access to the "Payment methods" page even
if the corresponding toggle is enabled by policy, to allow users to
access payment apps.

Screenshots (Googlers only): https://drive.google.com/open?id=17VDj8kUArJVxilLaT9CvzYYepewsUGMF

Bug: 860526
Change-Id: I9120bd7ca21607485b2b6e144aaf8a5e31abc3c3
Reviewed-on: https://chromium-review.googlesource.com/1208191Reviewed-by: Theresa <twellington@chromium.org>
Reviewed-by: Mathieu Perreault <mathp@chromium.org>
Commit-Queue: Fabio Tirelo <ftirelo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#589585}