GitLab

Change-Id: I7d0ebe844fe5396665e17e0202698a5ff25afbc8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1857299
Commit-Queue: Bryan McQuade <bmcquade@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Reviewed-by: Timothy Dresser <tdresser@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707038}

https://quiche.googlesource.com/quiche.git/+log/780eaa6043f2..13753e667da5

$ git log 780eaa604..13753e667 --date=short --no-merges --format='%ad %ae %s'
2019-10-17 bnc Use IETF QPACK error codes on the wire.
2019-10-16 rch gfe-relnote: Send QUIC connection close packets at all available encryption levels.
2019-10-16 fayang gfe-relnote: Deprecate gfe2_reloadable_flag_quic_fix_rto_retransmission3.
2019-10-16 wub gfe-relnote: (n/a) Always handle MSG_TOO_BIG when sending buffered packets. Protected by existing --gfe2_reloadable_flag_quic_treat_queued_packets_as_sent.
2019-10-16 fayang gfe-relnote: Remove QUIC v39.
2019-10-16 rch Remove broken code from ToyQuicClient which was landed in cr/275026885.

Update flag definition per the "Send QUIC connection..." change.

Remove QUIC_VERSION_39 references. Many seem to really be "the lowest
available/implemented/... QUIC version", and thus were changed to refer to V43.

Changed versions used in quic_http_utils_test.cc,
FilterSupportedAltSvcVersions to reflect removal of version 39.

Update components/domain_reliability/quic_error_mapping.cc to reflect
addition of QPACK error codes.

Created with:
  roll-dep src/net/third_party/quiche/src src/third_party/quic_trace/src

Change-Id: I04d14224215bf4efe5bd3ffa95c6d7e45ed1f8e0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865560
Commit-Queue: Zhongyi Shi <zhongyi@chromium.org>
Reviewed-by: Zhongyi Shi <zhongyi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707037}

Clamshell mode behavior will remain the same, i.e., tapping on the
overview background will return to its previous state, it can either be
in-app or home screen.

Bug: 997885

Change-Id: I2dae82657d4d92876c834c78dd62d53830fbe6cf
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865264
Commit-Queue: Xiaoqian Dai <xdai@chromium.org>
Reviewed-by: Xiaoqian Dai <xdai@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707036}

We introduced additional padding for the search engine icon experiment
focused state. This additional padding happens post-layout, so we don't
have to re-layout the view tree during the animation. As a result, the
UrlBar's width=match_parent makes the UrlBar too wide and the end of
the UrlBar's text is cut off by the next view.

Bug: 1012436
Change-Id: I5f46a2dd530783b002c529d3c4cdccbdd81e1b5d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1856868
Commit-Queue: Brandon Wylie <wylieb@chromium.org>
Reviewed-by: Theresa  <twellington@chromium.org>
Reviewed-by: Patrick Noland <pnoland@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707035}

Bug: 1015199
Change-Id: I73b009d4d4a20000a78ed15a11618adeec037800
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864525Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Ahmed Mehfooz <amehfooz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707034}

Moves 'Remove Access' next to 'Done', like in the bookmark bubble.
 - Use the dialog's built-in cancel button placement/handling
 - Avoid revoking access on the 'Esc' key by overriding Close()

Bug: 999756
Test: Button placement matches UX intent, no behavior change.
Change-Id: I91e3befbbef1a4d0bc80bb039b835bcfb7b0d504
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1862856
Commit-Queue: Michael Wasserman <msw@chromium.org>
Reviewed-by: Marijn Kruisselbrink <mek@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707033}

Include oobe.css for the newly added CSS variables.

Bug: 1012793
Test: Manual Test
Change-Id: If67e43f454b9ccf193029c8b144ef45062d057ba
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864364Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Yue Li <updowndota@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707032}

Without this change screen readers in browse mode [1] will incorrectly
read errors and warnings that have a tooltip of "42 errors, 99 warnings"
as "4299".

After this change screen readers will announce the same content as the
shown in the tooltip: "42 errors, 99 warnings".

Because errors and warnings are indicated with icons this change will
fix the screen reader bug and helps meet success criteria for
WCAG 1.1.1 Non-text Content [2].

Root cause:
The icons do not have accessible names and screen readers will combine
the content of the spans:

   <span><error-icon/>42<span><span><warning-icon/>99</span>
   is announced as 4299

The fix:
* Set the accessible name of counter element (title alone is insufficient [3]).
* Make the children of the counter hidden otherwise NVDA (and JAWs)
  screen readers will ignore the aria-label [4].

[1]: https://www.nvaccess.org/files/nvda/documentation/userGuide.html#BrowseMode
[2]: https://www.w3.org/WAI/WCAG21/quickref/#non-text-content
[3]: https://crrev.com/18d41172/third_party/blink/renderer/devtools/front_end/ui/Tooltip.js#190
[4]: https://github.com/nvaccess/nvda/issues/1354#issuecomment-155287285

Screenshot (NVDA speech viewer): https://i.imgur.com/KznR7O7.png
      screen reader should announce the number of errors and warnings.

Test: Navigate to the counter in NVDA browse mode on Win10 and the
Bug: 963183
Change-Id: I46985e223a6f8bac5cd1dbdf597a4b0c9ab71567
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1861214Reviewed-by: Lorne Mitchell <lomitch@microsoft.com>
Commit-Queue: John Emau <John.Emau@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#707031}

Use the site engagement service to rank content index entries when
handing off offline items.

Change-Id: Ie9929f73e0fb52edff297fd63f21e0069eb3c8f5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866710
Commit-Queue: Rayan Kanso <rayankans@chromium.org>
Reviewed-by: Shakti Sahu <shaktisahu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707030}

When user tries to export password or view saved password by visiting
chrome://settings/passwords url and clicking on "Export passwords..."
or eye icon on password list respectively, OS Reauthentication is
triggered, and on successful authentication plain text password is
revealed. Currently OS authentication call is on UI Thread.
And if OS authentication takes time, it freezes/hangs
browser. This change has made authentication api async, and moved
windows authentication call to thread pool from UI thread. Since the
UI thread will be freed, browser will not freeze or hang.

Bug: 1007950
Change-Id: I2f35707fe458687a4ec22de1381d21ec9603733f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1828730
Commit-Queue: Ankit Jain <ankjain@microsoft.com>
Reviewed-by: Vasilii Sukhanov <vasilii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707029}

When a unittest uses multiple processes, it's subprocesses uses the host's
file system class, not the mock file system class. This is problematic
for the blinkpy unittest step running on many builders because artifacts
from prevoius runs are not deleted on the bot.

Bug: chromium:1015118
Change-Id: I7fdf01de5ea0f782ba3ca7d4aaffda71b8841da5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866975
Commit-Queue: Rakib Hasan <rmhasan@google.com>
Reviewed-by: Robert Ma <robertma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707028}

1. Ignore skipped tests that are not in the test_path and not in the same shard;
2. Set the expected test status to SKIP if the actual status is SKIP;
3. Fix the issue with skipped tests - pytest deselect flag only works with relative path and pytest separator :: between a test and its subtest;
4. Add support for isolated-script-test-filter


Bug: 992569
Change-Id: Ifa73e4c6372a12f63e7543c85d006cc32c848c25
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1777070
Commit-Queue: Kyle Ju <kyleju@chromium.org>
Reviewed-by: Robert Ma <robertma@chromium.org>
Reviewed-by: John Chen <johnchen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707027}

* Replace modulo operations with test and subtraction
* Remove clamping

Using spotify.github.io/web-audio-bench, without this CL we have:

TEST			μs	MIN	Q1	MEDIAN	Q3	MAX
Baseline		393	383	390	393	400	403
DelayAutomation-a-rate	1053	1018	1050	1053	1065	1102

With this CL:
Baseline		393	387	393	393	397	400
DelayAutomation-a-rate	485	437	443	485	487	513

The optimized version takes about half the time.

Bug: 1012198
Change-Id: I643cbf137038714aad5fa2fb66299a8412e667bd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864415Reviewed-by: Andrew MacPherson <andrew.macpherson@soundtrap.com>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707026}

When navigating between two pages, getTextWithAutocomplete() points
to the autocompleted text for the previous URL. This screws with the
behavior of StatusMediator, so I've added a workaround that checks
if the current autocomplete text contains the current url bar text.
If it does then we use the autocomplete text, otherwise we just use
the UrlBar text.

Bug: 1015147
Change-Id: Ie64cb4e1944005ab7273e8b66084033c2401b7ac
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864027
Commit-Queue: Brandon Wylie <wylieb@chromium.org>
Reviewed-by: Ted Choc <tedchoc@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707025}

https://chromium.googlesource.com/external/github.com/libexpat/libexpat.git/+log/a7bc26b69768..25d3c190c69e

$ git log a7bc26b69..25d3c190c --date=short --no-merges --format='%ad %ae %s'
2019-10-16 sebastian Changes: Fix name order for 2.2.8
2019-10-16 sebastian Changes: Give credit for #361 and #362
2019-10-14 mohammed.khajapasha Autotools: Add package name variable for expat.pc config file
2019-10-12 sebastian xmlparse.c: Limit declaration of rand_s to mingwrt <5.3.0
2019-10-12 sebastian Changes: Document improvements to CMake
2019-10-12 sebastian CMake: Install shipped xmlwf manpage in case of -DEXPAT_BUILD_DOCS=OFF
2019-10-11 sebastian CMake: Catch troublesome build combinations for -DEXPAT_CHAR_TYPE=wchar_t
2019-10-09 sebastian CMake: Catch troublesome build combinations for -DEXPAT_CHAR_TYPE=ushort
2019-10-09 sebastian CMake: Create libexpatw.{dll,so} and expatw.pc with -DEXPAT_CHAR_TYPE=(ushort|wchar_t)
2019-10-09 sebastian CMake: Fix visibility of EXPAT_WITH_LIBBSD and EXPAT_*RANDOM*
2019-10-12 sebastian Changes: Document #356
2019-10-09 sebastian Drop expat.spec RPM packaging (related to #330)
2019-10-06 sebastian Travis CI: Cover coverage.sh
2019-10-08 sebastian codepage.c: Define functions codepageMap and codepageConvert no more than once
2019-10-07 sebastian coverage.sh: Limit coverage logs to lib/ and xmlwf/ folders
2019-10-07 sebastian coverage.sh: Report more about calls to lcov (than their output)
2019-10-07 sebastian coverage.sh: Fix directory collisions for w/ and w/o attribute info
2019-10-06 sebastian coverage.sh: Re-use first xmlts.zip download
2019-10-06 sebastian coverage.sh: Move from GNU Autotools to CMake#
2019-10-06 sebastian coverage.sh: Copy libexpat.dll as well
2019-10-06 sebastian coverage.sh: Improve handling of absent libwinpthread-1.dll
2019-10-06 vadim Declare rand_s() for MinGW 32 ourselves
2019-10-01 sebastian xmltok.c: Address warnings about accidental non-ASCII dashes in comments (#355)
2019-10-01 sebastian Changes: Document #354
2019-09-29 sebastian Travis CI: Cover more non-default build parameters (in isolation)
2019-09-29 sebastian qa.sh: Exclude "make run-xmltest" for -DEXPAT_DTD=OFF
2019-09-29 sebastian runtests.c: Fix for -DEXPAT_DTD=OFF
2019-09-29 sebastian runtests.c: Fix for -DEXPAT_NS=OFF
2019-09-29 sebastian runtests.c: Fix compiler warning "unused function" with -DEXPAT_MIN_SIZE=ON
2019-09-29 sebastian xmlparse.c|xmlrole.c: Fix compiler warning "unused parameter" with -DEXPAT_DTD=OFF
2019-09-29 sebastian xmlparse.c: Address compiler warninging with -DEXPAT_CONTEXT_BYTES=OFF
2019-09-28 sebastian Changes: Fix spelling of "September"

Created with:
  roll-dep src/third_party/expat/src
R=bungeman@chromium.org,dcheng@chromium.org

Change-Id: I40469d12bc2581ed4db60b8ff028fd17e40cc743
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865752Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Commit-Queue: Ben Wagner <bungeman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707024}

This variable has always existed as |select_all|, but it's real meaning
is essentially |is_user_initiated|. This CL renames the parameter to
that new name, and plumbs it into OmniboxView.

It doesn't use that parameter yet. This CL has no behavior changes,
just an API change.

The followup CL will use this new parameter to prevent triggering
on-focus suggestions for renderer-initiated omnibox focuses.
Most notably, this prevents browser startup NTP load from triggering
on-focus suggestions.

TBR=aboxhall@chromium.org

Bug: 1013287, 996516
Change-Id: I9461ea6a9c925b163aed5708109f5f7caa00f0c1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865616
Commit-Queue: Tommy Li <tommycli@chromium.org>
Reviewed-by: Dan Beam <dbeam@chromium.org>
Reviewed-by: Justin Donnelly <jdonnelly@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707023}

The injected implementations of pushState and replaceState stringify
the given state using JSON.stringify. This behavior is non-standard,
and leaks to the web when JSON.stringify fails and throws a TypeError.
This happens when the given state object has a cycle.

To prevent leaking this implementation detail, this CL catches exceptions
thrown by JSON.stringify, and then throws a standard DataCloneError instead.

This fixes subtests in the following two Web Platform Tests that currently
fail in Chrome but pass in Safari:
html/browsers/histoy/the-history-interface/001.html
html/browsers/histoy/the-history-interface/002.html

Longer term, we should remove the stringification all together. This is
blocked on deleting legacy navigation and finding an alternate solution to
crbug.com/949305 that doesn't involve capturing and replaying the
state passed to pushState or replaceState.

Bug: 769945
Change-Id: I945251ffe09ecac79f086ca75030dc366ef99286
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866870Reviewed-by: Eugene But <eugenebut@chromium.org>
Commit-Queue: Ali Juma <ajuma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707022}

To be prepared for the new CDM which starts to support VP9 profile2,
mark expected support as EXPECT_ANY. After the new CDM is checked-in,
we'll update these tests with EXPECT_WV_SUCCESS.

An alternative would be to manually stop the src-internal DEPS roll, fix
the tests and manually roll the DEPS in the same CL. That'll be the
"more correct" way of doing it, but involves more work and is less
flexible, e.g. it's harder to revert to an older CDM.

Bug: 996372
Test: Update test expectations.
Change-Id: I3ce66c0ade6f1d6b1be050949f51d91895934d7d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866978Reviewed-by: John Rummell <jrummell@chromium.org>
Commit-Queue: Xiaohan Wang <xhwang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707021}

This is a less-aggressive version of
https://chromium-review.googlesource.com/c/chromium/src/+/1865756
in case we don't want to nuke everything related to key event
dispatch all at once.

      verify that key events are plumbed to JavaScript.

Bug: Internal b/142758140
Test: cast_shell_junit_tests
Test: cast_shell_unittests
Test: Use chrome://inspect to monitor key events in Cast app to
Change-Id: Ibe312778f74e76b834a80b871eda911adac18cd0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865146Reviewed-by: Luke Halliwell <halliwell@chromium.org>
Commit-Queue: Simeon Anfinrud <sanfin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707020}

The expectation is that the root surface fills the display space. This
ensures that sub-surfaces with negative origins don't cause the root
surface to be shifted and the host window resized.

Bug: 1015185
Test: Manually with test fullscreen_wayland_client
Test: Compare behavior while running with "weston --shell fullscreen-shell.so"
Change-Id: Id67f6e0584002329bf9941843ae6eb27c01c7225
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866608
Commit-Queue: Daniel Nicoara <dnicoara@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707019}

Implements https://github.com/w3c/wake-lock/pull/244.

The permission algorithms in the spec already assume it is, and so does our
implementation.

While here, make sure to catch the exception caught when converting JS to a
WakeLockPermissionDescriptor, as we will throw a TypeError when |type| is
not present or is invalid.

Bug: 257511
Change-Id: Ib20001d48296ea4af70368fa494b0d712348da6b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864659
Auto-Submit: Raphael Kubo da Costa <raphael.kubo.da.costa@intel.com>
Reviewed-by: Mounir Lamouri <mlamouri@chromium.org>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Commit-Queue: Raphael Kubo da Costa <raphael.kubo.da.costa@intel.com>
Cr-Commit-Position: refs/heads/master@{#707018}

Some LiteMode (aka Data saver) features such as lazyload, sending save-data in
request header, SaveData NetInfo API does not consider incognito profiles.
This CL fixes that.

Bug: 1014144
Change-Id: I22de4bfc20d7791ab33971c2dcc2cdca1d5a8cda
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1863574
Commit-Queue: rajendrant <rajendrant@chromium.org>
Reviewed-by: Robert Ogden <robertogden@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707017}

Bug: internal b/140603369
Change-Id: I188c969198078afb251c488fb0574a628dc8a2be
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866973
Commit-Queue: Kenneth MacKay <kmackay@chromium.org>
Reviewed-by: Yuchen Liu <yucliu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707016}

The picture to show will be controlled by the Finch config. The default option
is no picture as today.

Bug: 1015447
Change-Id: If2157127277b243440d21aa5ef212f6119d65389
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866929Reviewed-by: Dominic Battré <battre@chromium.org>
Commit-Queue: Vasilii Sukhanov <vasilii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707015}

https://swiftshader.googlesource.com/SwiftShader.git/+log/6652f0b64287..a68a80a4dbf9

git log 6652f0b64287..a68a80a4dbf9 --date=short --no-merges --format='%ad %ae %s'
2019-10-17 sugoi@google.com Vulkan header update

Created with:
  gclient setdep -r src/third_party/swiftshader@a68a80a4dbf9

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
https://autoroll.skia.org/r/swiftshader-chromium-autoroll
Please CC swiftshader-team+autoroll@google.com on the revert to ensure that a human
is aware of the problem.

To report a problem with the AutoRoller itself, please file a bug:
https://bugs.chromium.org/p/skia/issues/entry?template=Autoroller+Bug

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

CQ_INCLUDE_TRYBOTS=luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_chromium_msan_rel_ng;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=swiftshader-team+autoroll@google.com

Bug: chromium:b/139800520
Change-Id: I5074253515186d087f8e97d983206823019c040a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866075Reviewed-by: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#707014}

This CL adds the boolean Android.Language.UiIsSystemLanguage histogram
to record whether the Chrome UI language is the same as the Android
system language.

Change-Id: Id40f6ede8e865c2dd1c0e01c9471a7cf6b4453f5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1863252
Auto-Submit: Trevor  Perrier <perrier@chromium.org>
Reviewed-by: Andrew Grieve <agrieve@chromium.org>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Reviewed-by: Robert Ogden <robertogden@chromium.org>
Commit-Queue: Trevor  Perrier <perrier@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707013}

This is a reland of 8e1e2bb6

Original change's description:
> Highlight text fragment directives on page reload
>
> Fix the behavior where fragment anchors were only kept alive if a scroll
> was needed, as is the behavior of element fragment anchors. This caused
> text fragment anchors not to be kept alive and we didn't highlight on
> reload.
>
> Note that *sometimes* we would highlight on reload, due to a race
> condition with fragment anchors, where we process the fragment before
> restoring scroll position (which is where should_scroll becomes false),
> see crbug/839292. This change makes it so we highlight on reload
> independent of whether a scroll is needed.
>
> Bug: 932551
> Change-Id: I38cc82c4920255d8bcae9f8b55ac80f24d1465da
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864714
> Commit-Queue: David Bokan <bokan@chromium.org>
> Reviewed-by: David Bokan <bokan@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#706847}

TBR=bokan@chromium.org,nburris@chromium.org

Bug: 932551
Change-Id: Ie633b2b4d14348f7452cbf778e914b401c9cd571
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866625
Commit-Queue: Lan Wei <lanwei@chromium.org>
Reviewed-by: Lan Wei <lanwei@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707012}

For easier test failure triage.

Bug: 990561
Change-Id: I4abeed8c2b97366316f7ed77f8df815f8f3a32ed
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864014
Commit-Queue: Hiroshige Hayashizaki <hiroshige@chromium.org>
Reviewed-by: Domenic Denicola <domenic@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707011}

Bug: 724598
Change-Id: If32e7ec6dda45f38c0d4b5e87d8251a1a60ed10a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1863568
Commit-Queue: Evan Liu <evliu@google.com>
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707010}

DoPeek() is driven on all post-handshake operations, so it's a common
point to figure out the early data reject.

Bug: 1000659
Change-Id: Ibcb16048a8e804d98eafa574a371d563140f374a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1863380Reviewed-by: Steven Valdez <svaldez@chromium.org>
Commit-Queue: David Benjamin <davidben@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707009}

TBR=dtrainor@chromium.org

Bug: 1013374
Change-Id: Ie62c57c0eeed62915f381dd26c0c35ebda1bc044
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1863738Reviewed-by: Mustafa Emre Acer <meacer@chromium.org>
Reviewed-by: David Trainor <dtrainor@chromium.org>
Reviewed-by: Christopher Thompson <cthomp@chromium.org>
Reviewed-by: Joe DeBlasio <jdeblasio@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Commit-Queue: Mustafa Emre Acer <meacer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707008}

In addition to setting the Origin Trial feature name inside Blink we now
need to unconditionally enable access to the Serial service and the new
settings pages.

Bug: 967863
Change-Id: I6c21566526e1590d94a6055258f8b2c511f43110
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1863360Reviewed-by: Avi Drissman <avi@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Commit-Queue: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707007}

This eliminates a hairline that had been drawn in some PWA windows but
which had - for reasons that are unclear - not actually been rendered in
most cases. With the new 1dip rendering code, the line WAS rendered in
some cases, which caused a visual regression.

Bug: 1012979
Change-Id: I3cebc47cfde1b2e1205de8632fd92a8e43181a29
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866277
Auto-Submit: Dana Fried <dfried@chromium.org>
Commit-Queue: Charlene Yan <cyan@chromium.org>
Reviewed-by: Charlene Yan <cyan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707006}

This CL converts MediaPerception from media_perception.mojom
to new Mojo types using PendingReceiver.

Bug: 955171
Change-Id: Iba4548d03d13ba4fa6c426ce3830a14c8b3c5b7c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1851370Reviewed-by: Ken Rockot <rockot@google.com>
Reviewed-by: Sam McNally <sammc@chromium.org>
Reviewed-by: Luke Sorenson <lasoren@chromium.org>
Commit-Queue: Julie Kim <jkim@igalia.com>
Cr-Commit-Position: refs/heads/master@{#707005}

Instead of setting the rotation as part of the transform on
ash root window, call viz SetDisplayTransformHint so that
capturing code as a consumer of viz could get the correct
orientation.

Bug: 847693
Change-Id: Id70f91dc84fc9274145d5a160a51b8c7fa50deed
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1825892Reviewed-by: kylechar <kylechar@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Khushal <khushalsagar@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Commit-Queue: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707004}

https://dawn.googlesource.com/dawn.git/+log/919812ed1c25..8d000e0cc230

git log 919812ed1c25..8d000e0cc230 --date=short --no-merges --format='%ad %ae %s'
2019-10-17 cwallez@chromium.org Remove unneeded GN config removal after generated file cleanup

Created with:
  gclient setdep -r src/third_party/dawn@8d000e0cc230

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
https://autoroll.skia.org/r/dawn-chromium-autoroll
Please CC cwallez@google.com on the revert to ensure that a human
is aware of the problem.

To report a problem with the AutoRoller itself, please file a bug:
https://bugs.chromium.org/p/skia/issues/entry?template=Autoroller+Bug

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

CQ_INCLUDE_TRYBOTS=luci.chromium.try:dawn-linux-x64-deps-rel;luci.chromium.try:dawn-mac-x64-deps-rel;luci.chromium.try:dawn-win10-x64-deps-rel;luci.chromium.try:dawn-win10-x86-deps-rel
TBR=cwallez@google.com

Bug: None
Change-Id: Ia7a918cca3bd131f3fcaabd94683c685970b178d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865987Reviewed-by: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com>
Commit-Queue: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#707003}

While investigating opportunities where AXRange<>::{begin|end}() can
be improved, I found that AsLeafTextPosition was one of the higher
cost methods.

This is because it called AsTreePosition for each node until reaching
the leaf node, which roughly doubled the number of times MaxTextOffset
needed to be called.

For the scenario I was measuring, this CL reduced the overall amount
of time spent in AXRange::begin by ~50-60%, and the overall amount of
time spent in AXPosition<>::AsLeafTextPosition by ~80%!

I crafted a document where there are many inline nodes that form
sentences with various levels of nesting, and used Narrator with
UIA enabled to navigate 5 sentences.

I used "UI for ETW" to capture perf traces, and WPA to view the data.

Before :
  * Weight for callees of AXRange<>::begin : 8651
  * Weight for callees of AXPosition<>::AsLeafTextPosition : 19412

After :
  * Weight for callees of AXRange<>::begin : 3336
  * Weight for callees of AXPosition<>::AsLeafTextPosition : 3523

Bug: 928948
Change-Id: I41966515f9ef91b86374d4885daeb67a71c50953
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1826007
Commit-Queue: Nektarios Paisios <nektar@chromium.org>
Reviewed-by: Kurt Catti-Schmidt <kschmi@microsoft.com>
Reviewed-by: Nektarios Paisios <nektar@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707002}

Favicon and title that happened in response to a user action, just
after a tab was backgrounded, were incorrectly tracked as signals
that the page is trying to communicate with the user in background.
Now that this issue is fixed, we are increasing the DB version to
automatically delete all the incorrect data.

No-Try: true
Bug: 1014976
Change-Id: If6bb9816d3a5f57d5daa12a53812c2f2d3a3b84f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865601
Commit-Queue: François Doray <fdoray@chromium.org>
Reviewed-by: Sébastien Marchand <sebmarchand@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707001}

This reverts commit 671afa99.

Reason for revert: Breaks proximity auth in the crash-and-restore
login flow (e.g. if chrome crashes during the session and logs you
back in automatically on restart). See crbug.com/1015215

Original change's description:
> Migrate the Chrome OS device_sync service off of the mojo pref service
> 
> DeviceSyncService runs in the browser process on the UI thread. It can
> directly use the Profile's PrefService*. This simplifies the code, and
> will reduce the amount of work rockot@ has to do for Service Manager
> cleanup.
> 
> Delete the DeviceSyncService mojo interface and implementation object.
> They primarily existed to support wiring up the mojo pref service.
> Instead, directly connect the DeviceSyncClientImpl to the backing
> DeviceSyncImpl.
> 
> Explicitly initialize the DeviceSyncImpl when the user's profile is
> ready. The old code had startup timing dependencies that assumed that
> DeviceSyncImpl was created after IdentityManager had information about
> the primary profile. The new code instantiates the DeviceSyncImpl
> object earlier, and IdentityManager may not be ready yet.
> 
> Bug: 977637, 1012941
> Test: bots
> Change-Id: Ia32b687404e5b68980cb3aecd3003e6f7e48ffcd
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1849217
> Commit-Queue: James Cook <jamescook@chromium.org>
> Reviewed-by: Greg Kerr <kerrnel@chromium.org>
> Reviewed-by: Josh Nohle <nohle@chromium.org>
> Reviewed-by: Scott Violet <sky@chromium.org>
> Reviewed-by: Ken Rockot <rockot@google.com>
> Cr-Commit-Position: refs/heads/master@{#706181}

TBR=jamescook@chromium.org,sky@chromium.org,rockot@google.com,kerrnel@chromium.org,hansberry@chromium.org,nohle@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug: 977637, 1012941
Change-Id: Id6cf5f43aac57415dae266a7871c08e4954fa24b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1865602Reviewed-by: James Cook <jamescook@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#707000}

...instead of the actual GUID of the bookmark node.
This fixes a crash, see linked bug.
The problem is that right now, before the GUID-aware merge algorithm is
rolled out, bookmark GUIDs are not actually guaranteed to be unique.
This can result in two separate bookmark nodes getting assigned the
same server ID, resulting in a CHECK() crash.

This is an immediate fix for the crash; longer-term we need to make
sure that bookmark node GUIDs are actually guaranteed to be unique.

Bug: 1004205
Change-Id: Ic2e5e17b025b65e0739e6cb82a58053ff11209a9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1864800
Commit-Queue: Marc Treib <treib@chromium.org>
Commit-Queue: Mohamed Amir Yosef <mamir@chromium.org>
Auto-Submit: Marc Treib <treib@chromium.org>
Reviewed-by: Mohamed Amir Yosef <mamir@chromium.org>
Cr-Commit-Position: refs/heads/master@{#706999}