GitLab

Group navigation logs with the corresponding decisions message of the
navigation.

Push the group message to the top of the table when there's a new message in
that group. Since navigation logs could be called called much later compared to
decision logs, the message group need to float to the top of the table when new
messages are added to the group.

Bug: 791778
Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: Ic872fd155ca25a35286bafaaffde41c2b65a01d4
Reviewed-on: https://chromium-review.googlesource.com/809936
Commit-Queue: Thanh Le <thanhdle@chromium.org>
Reviewed-by: Ryan Sturm <ryansturm@chromium.org>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522230}

Skia upstream change keeps rounded rectangles
constructed from SVG convex. This alters
some existing SVG layout results. Disable
the new behavior pending the rebaseline.

TBR=djsollen@chromium.org

Bug: b:69622768
Change-Id: Ia32b9cffff4a011bc835eb014f3943e735d6775e
Reviewed-on: https://chromium-review.googlesource.com/811840Reviewed-by: Cary Clark <caryclark@chromium.org>
Commit-Queue: Cary Clark <caryclark@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522229}

This reverts commit 03ca6202.

Reason for revert: Broke scrolling in certain cases

Original change's description:
> [Home] Separate swipe detecting logic from the bottom sheet
> 
> This change moves all of the swipe detecting from the bottom sheet
> into its own class. A relatively simple interface has been introduced
> to prevent needing the entire bottom sheet to detect swipes.
> 
> BUG=
> 
> Change-Id: If8c61df4009320e016903d6f1ed0c231c9330e02
> Reviewed-on: https://chromium-review.googlesource.com/792337
> Commit-Queue: Matthew Jones <mdjones@chromium.org>
> Reviewed-by: Ted Choc <tedchoc@chromium.org>
> Reviewed-by: Theresa <twellington@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#522147}

TBR=tedchoc@chromium.org,twellington@chromium.org,mdjones@chromium.org

Change-Id: I22743cfe51fa3861d9bc2b44beef0f894fc0b30c
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/812266Reviewed-by: Matthew Jones <mdjones@chromium.org>
Commit-Queue: Matthew Jones <mdjones@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522228}

https://webrtc.googlesource.com/src.git/+log/c3da1e61bccf..2fb645091977

$ git log c3da1e61b..2fb645091 --date=short --no-merges --format='%ad %ae %s'

Created with:
  roll-dep src/third_party/webrtc
BUG=None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None,None


The AutoRoll server is located here: https://webrtc-chromium-roll.skia.org

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.chromium.linux:linux_chromium_archive_rel_ng;master.tryserver.chromium.mac:mac_chromium_archive_rel_ng;master.tryserver.chromium.win:win-msvc-dbg
TBR=webrtc-chromium-sheriffs-robots@google.com

Change-Id: I0103bd290c6b31c7df8b00f0ceebbc10ccde1348
Reviewed-on: https://chromium-review.googlesource.com/812146Reviewed-by: WebRTC Roll Bot <webrtc-autoroll@chromium.org>
Commit-Queue: WebRTC Roll Bot <webrtc-autoroll@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522227}

Recent Clang versions started taking into account that enums are signed
on Windows when emitting these warnings.

Bug: 792519
Change-Id: I5029cde81ac6f1777ec2317d431d55589b814465
Reviewed-on: https://chromium-review.googlesource.com/811685
Commit-Queue: Mark Mentovai <mark@chromium.org>
Reviewed-by: Mark Mentovai <mark@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522226}

This reverts commit 7df02c0d.

Reason for revert: Failed desktopui_ScreenLocker autotest

Original change's description:
> Modify Autotest API to support views based lockscreen
> 
> Bug: 781998
> Change-Id: I11a456b1341d044b1370652c337f105e0bef649a
> Reviewed-on: https://chromium-review.googlesource.com/803622
> Commit-Queue: Xiaoyin Hu <xiaoyinh@chromium.org>
> Reviewed-by: Jacob Dufault <jdufault@chromium.org>
> Reviewed-by: Achuith Bhandarkar <achuith@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#521137}

TBR=achuith@chromium.org,jdufault@chromium.org,xiaoyinh@chromium.org

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

Bug: 781998
Change-Id: I547a15a15d2cf3b95986e8cfbf548b95cd0b3087
Reviewed-on: https://chromium-review.googlesource.com/812264Reviewed-by: Xiaoyin Hu <xiaoyinh@chromium.org>
Commit-Queue: Xiaoyin Hu <xiaoyinh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522225}

If the load timer fires after an error, and there are no sources,
changing the network state can prevent cleanup from running on future
load attempts.

This CL prevents the network state from changing if a previous load
got far enough to have state that requires cleanup.

corrupt.* files have enough initial data to start playback, but fail
with a decode error shortly after. They were generating like this:
  $ cp test.mp4 corrupt.mp4
  $ dd conv=notrunc if=/dev/urandom of=corrupt.mp4 bs=1024 seek=20 count=20

Bug: 674716
Change-Id: I978af321921c7331829b9384772dc954c20dbc1f
Reviewed-on: https://chromium-review.googlesource.com/775891
Commit-Queue: Dan Sanders <sandersd@chromium.org>
Reviewed-by: Mike West <mkwst@chromium.org>
Reviewed-by: Mounir Lamouri <mlamouri@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522224}

Bug: 792630, 772252
Change-Id: I2cd4ea498cbdd614b3469cf92bd4d7b2ee994b39
Reviewed-on: https://chromium-review.googlesource.com/812209Reviewed-by: Scott Graham <scottmg@chromium.org>
Commit-Queue: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522223}

Reenable the FeatureEngagement testBadgedREadingListFeatureShouldShow
test by increasing the timeout.

Bug: 789943
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: Iace3e1e7c3beb1930ef7c85cc1184ac27665fcff
Reviewed-on: https://chromium-review.googlesource.com/812218Reviewed-by: edchin <edchin@chromium.org>
Commit-Queue: Gregory Chatzinoff <gchatz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522222}

AbstractHapticGamepad encapsulates the logic for translating haptic
effect descriptions into a sequence of vibration actuator commands.
Users of AbstractHapticGamepad should override SetVibration so that it
updates the vibration actuators on the device with the specified
intensities.

BUG=749295

Change-Id: I3744cd3b06efb7970bd87048996250e52c472919
Reviewed-on: https://chromium-review.googlesource.com/810006Reviewed-by: Brandon Jones <bajones@chromium.org>
Commit-Queue: Matt Reynolds <mattreynolds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522221}

Makes it easier to add a java_annotation_processor() template

This removes main_class from java_prebuilt() in favor of using
java_binary() with a jar_path.

Bug: 792170
Cq-Include-Trybots: master.tryserver.chromium.android:android_cronet_tester;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I8b92addae2228398f911d9456b9c55d5ebf48529
Reviewed-on: https://chromium-review.googlesource.com/810032Reviewed-by: Andrei Kapishnikov <kapishnikov@chromium.org>
Reviewed-by: Eric Stevenson <estevenson@chromium.org>
Commit-Queue: agrieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522220}

https://chromium.googlesource.com/catapult.git/+log/307c9f954dc5..a2299532b526

$ git log 307c9f954..a2299532b --date=short --no-merges --format='%ad %ae %s'
2017-12-06 simonhatch Dashboard - Fix /add_histograms from re-adding same diagnostic.
2017-12-06 rnephew [Telemetry/TADA] Make expectation parser only support raw data and not file path.
2017-12-06 perezju [Telemetry] Dump output of "ps" when failing to find PID

Created with:
  roll-dep src/third_party/catapult


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

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.chromium.android:android_optional_gpu_tests_rel
TBR=sullivan@chromium.org

Change-Id: I3ee30516264ef4ee24e5f45f6f2c5a9e03a7c31f
Reviewed-on: https://chromium-review.googlesource.com/812004
Reviewed-by: <catapult-deps-roller@chromium.org>
Commit-Queue: <catapult-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522219}

Bug: 778436
Cq-Include-Trybots: master.tryserver.chromium.mac:ios-simulator-cronet;master.tryserver.chromium.mac:ios-simulator-full-configs
Change-Id: I8ba2a695880a8594a145ecfa9724907253f5ad7f
Reviewed-on: https://chromium-review.googlesource.com/752074
Commit-Queue: Jared Saul <jsaul@google.com>
Reviewed-by: Moe Ahmadi <mahmadi@chromium.org>
Reviewed-by: Sebastien Seguin-Gagnon <sebsg@chromium.org>
Reviewed-by: Mathieu Perreault <mathp@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522218}

https://pdfium.googlesource.com/pdfium.git/+log/4bd2ad8cf901..fd9e0b8f5f92

$ git log 4bd2ad8cf..fd9e0b8f5 --date=short --no-merges --format='%ad %ae %s'
2017-12-06 thestig Fix nits on some font code.

Created with:
  roll-dep src/third_party/pdfium


The AutoRoll server is located here: https://pdfium-roll.skia.org

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: I4798f7891d9e82e538880fe193973022703f59cc
Reviewed-on: https://chromium-review.googlesource.com/811893
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522217}

This CL fixes a few warnings reported by the clang static analyzer:
* Remove the unnecessary `next_frame += bytes_to_copy` line from
  AudioPlayerBuffer.
* Call dictionaryWithDictionary with empty dictionary if data is nil.
* Add [super viewDidLoad] calls to view controllers.

Note that none of them causes actual problem in the official app.

Bug: 791222
Change-Id: Icaaf8a43481eb68950dbf6c25c9d7789237aed5e
Reviewed-on: https://chromium-review.googlesource.com/809928
Commit-Queue: Yuwei Huang <yuweih@chromium.org>
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522216}

warning: [ImmutableEnumChecker] enums should be immutable, and cannot have non-final fields
    private int mValue;

BUG=

Change-Id: Ib5753459af5ea967d052251f306612e747ade786
Reviewed-on: https://chromium-review.googlesource.com/811453Reviewed-by: Theresa <twellington@chromium.org>
Commit-Queue: Ted Choc <tedchoc@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522215}

NotificationDisplayService.

Test by commenting out most of OnPortalDetectionCompleted() (in addition
to unit tests).

Bug: 783018
Change-Id: I403967f28883c588d9c2e5fd43effe05e41f7eda
Reviewed-on: https://chromium-review.googlesource.com/807445Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Commit-Queue: Evan Stade <estade@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522214}

BUG=784183
TEST=test case in the bug in ASAN build
R=kbr@chromium.org

Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
Change-Id: Ie25ca328af99de7828e28e6a6e3d775f1bebc43f
Reviewed-on: https://chromium-review.googlesource.com/811826Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Zhenyao Mo <zmo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522213}

Previously these tests did not use the conventional TestRenderFrameHost
and TestRenderViewHost objects, but instead used RenderFrameHostImpl
and RenderViewHostImpl. This caused them to fail with CFI bad cast
errors when code in TestWebContents attempted to cast these objects
to the test classes.

This change eliminates the bad casts by creating
a RenderViewHostTestEnabler object in the scope of the test
fixture. This ensures that any Render*Host objects created for the
duration of the test are in fact of type TestRender*Host.

Bug: 469376
Change-Id: I9d4d048e7c8f1162fd4c8bc8e67ffc0932b6d220
Reviewed-on: https://chromium-review.googlesource.com/792450Reviewed-by: Cathy Li <chili@chromium.org>
Commit-Queue: Peter Collingbourne <pcc@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522212}

Bug: 759161
Change-Id: I81acfa8caf45d460ebfc940f8d182300623769e1
Reviewed-on: https://chromium-review.googlesource.com/811751Reviewed-by: Alexei Svitkine <asvitkine@chromium.org>
Commit-Queue: Alexei Svitkine <asvitkine@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522211}

https://pdfium.googlesource.com/pdfium.git/+log/4bd2ad8cf901..fd9e0b8f5f92

$ git log 4bd2ad8cf..fd9e0b8f5 --date=short --no-merges --format='%ad %ae %s'
2017-12-06 thestig Fix nits on some font code.

Created with:
  roll-dep src/third_party/pdfium


The AutoRoll server is located here: https://pdfium-roll.skia.org

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: I601a2c853d26536771a922d69c4a3a2ee45475ee
Reviewed-on: https://chromium-review.googlesource.com/811965
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522210}

This deals with the case where multiple calls are made to Execute, before
the store manages to initialize.

This patch adds a test for such case.

Bug: 792530
Change-Id: I270b0a035e55b6a66e8ada3b93d70ac0a1f207db
Reviewed-on: https://chromium-review.googlesource.com/811746
Commit-Queue: Filip Gorski <fgorski@chromium.org>
Commit-Queue: Peter Williamson <petewil@chromium.org>
Reviewed-by: Carlos Knippschild <carlosk@chromium.org>
Reviewed-by: Peter Williamson <petewil@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522209}

Bug: 775122
Change-Id: I3effce43273adbad46c0e085a01790c75427b489
Reviewed-on: https://chromium-review.googlesource.com/810057Reviewed-by: Ryan Hamilton <rch@chromium.org>
Reviewed-by: Shivani Sharma <shivanisha@chromium.org>
Commit-Queue: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522208}

This blocks the history service shutdown.
The issue was originally introduced in https://codereview.chromium.org/1960223002
However, when render process is gone, there is no need to check removed files.
We can do that next time when downloads page is loaded

BUG=477975

Change-Id: I468cabc7d4c564591dfeaf4b0af66290e1ac8f53
Reviewed-on: https://chromium-review.googlesource.com/810107
Commit-Queue: Min Qin(OOO 12/7-1/10) <qinmin@chromium.org>
Reviewed-by: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522207}

This is a second instance of the same issue and fix as reviewed in
https://chromium-review.googlesource.com/c/chromium/src/+/806442.

TBR=emircan@chromium.org,hajimehoshi@chromium.org

Bug: 789669
Test: third_party/WebKit/Tools/Scripts/run-webkit-tests --target gn_release --enable-leak-detection fast/mediastream/argument-types.html
Change-Id: If16c19395209a19a4548491f8b3e04893915adc2
Reviewed-on: https://chromium-review.googlesource.com/811865Reviewed-by: Christian Fremerey <chfremer@chromium.org>
Reviewed-by: Emircan Uysaler <emircan@chromium.org>
Commit-Queue: Christian Fremerey <chfremer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522206}

If a user hits "reload" on an unpacked extension, there's a chance that
reload will fail (if the local files were modified in a way that causes
a load error). Once this happens, successive calls to reload() will
fail, because ExtensionService::ReloadExtension() doesn't play nicely
with extensions that failed to reload.

Adjust devleoperPrivate.reload() to return a LoadError (similar to
loadUnpacked()) if the reload fails. The extensions page now checks for
this result, and shows the load error dialog if the reload fails. The
load error dialog can then trigger a retry of this load, which goes
through loadUnpacked() rather than reload() (thus circumventing
weirdness with ExtensionService::ReloadExtension()). This is comparable
to how the current extensions page handles this scenario.

This is *not* the ideal solution, as a) we shouldn't be going through
loadUnpacked() for a loaded extension and b) the user can still get into
a broken state (where the extension cannot be enabled, nor can it be
reloaded) by dismissing the load error without fixing the error, but it
brings us at least to parity with the old page.

Bug: 789659

Cq-Include-Trybots: master.tryserver.chromium.linux:closure_compilation
Change-Id: Ib0bd6eb4bfe74b1da588a9ea78e3396703fee973
Reviewed-on: https://chromium-review.googlesource.com/798599
Commit-Queue: Devlin <rdevlin.cronin@chromium.org>
Reviewed-by: Karan Bhatia <karandeepb@chromium.org>
Reviewed-by: Demetrios Papadopoulos <dpapad@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522205}

This change reduces createImageBitmap resize test tolerance from 10 to 0
for all image sources except for blob (which is set to 5).
It also refactors the tests to include pixelated, low and medium filter
qualities. Furthermore, it separates the test path for premul and unpremul
image bitmap sources when the filter quality is high, to match the way this
is implemented in chromium.

Bug: 744647, 792607
Change-Id: I4e0b8af71d88f5885ff56540a17ec8b716cd58d7
Reviewed-on: https://chromium-review.googlesource.com/809808
Commit-Queue: Mohammad Reza Zakerinasab <zakerinasab@chromium.org>
Reviewed-by: Justin Novosad <junov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522204}

We check that the input has at least 5 characters before reading
characters[4].

An rgb(a) color contains at least 3 numbers or percentages,
plus separator characters, so the length of a valid rgb(a) color
will be well over 5.

TBR=nainar@chromium.org
BUG=792306

Change-Id: Iceb49992ee7c5da346e2f64cbc7ccb82ad619f81
Reviewed-on: https://chromium-review.googlesource.com/811784Reviewed-by: Eric Willigers <ericwilligers@chromium.org>
Commit-Queue: Eric Willigers <ericwilligers@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522203}

https://chromium.googlesource.com/angle/angle.git/+log/10a4d43462fa..378c3a51d5f0

$ git log 10a4d4346..378c3a51d --date=short --no-merges --format='%ad %ae %s'
2017-12-04 oetuaho Clean up storing interface blocks in the symbol table

Created with:
  roll-dep src/third_party/angle


The AutoRoll server is located here: https://angle-chromium-roll.skia.org

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.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
TBR=geofflang@chromium.org

Change-Id: Ief1e722e35262c8e857382e9d1c4f57af919fcd3
Reviewed-on: https://chromium-review.googlesource.com/811894Reviewed-by: angle-deps-roller . <angle-deps-roller@chromium.org>
Commit-Queue: angle-deps-roller . <angle-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522202}

https://pdfium.googlesource.com/pdfium.git/+log/4bd2ad8cf901..fd9e0b8f5f92

$ git log 4bd2ad8cf..fd9e0b8f5 --date=short --no-merges --format='%ad %ae %s'
2017-12-06 thestig Fix nits on some font code.

Created with:
  roll-dep src/third_party/pdfium


The AutoRoll server is located here: https://pdfium-roll.skia.org

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: Ie542801610893c18ffcf81d0df84f40238099403
Reviewed-on: https://chromium-review.googlesource.com/811888
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522201}

Add a reference to the returned invisible cursor when we try to create a
cursor from an invalid bitmap.

X11CursorFactoryOzone is only used in the chromeos-on-linux development
environment. I've double checked that this doesn't occur in
CursorDataFactoryOzone and BitmapCursorFactoryOzone.

Bug: 791243
Change-Id: Ia12edf1638a420b61eb7f72f3696dabf506a44e0
Reviewed-on: https://chromium-review.googlesource.com/809492
Commit-Queue: Elliot Glaysher <erg@chromium.org>
Reviewed-by: kylechar <kylechar@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522200}

Many experience use the gamepad id to choose a hardcoded mapping. The
defacto standard id to use for OpenVR is "OpenVR Gamepad", and we should
use that as well.

Change-Id: I7717516c76a4fb7f5bccf16bbd8d197470de71e1
Reviewed-on: https://chromium-review.googlesource.com/810038Reviewed-by: Brandon Jones <bajones@chromium.org>
Commit-Queue: Bill Orr <billorr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522199}

Every RenderFrameHost exposes an InterfaceProvider to its corresponding
RenderFrame, which is used by content/renderer/, Blink and WebUI JS code
to access frame-scoped services implemented by the RenderFrameHost.

Previously, for each frame, there was just a single InterfaceProvider
interface connection that was used throughout the entire lifetime of the
RenderFrame. The client end of this connection was passed in to the
RenderFrame synchronously at construction time, and was then bound to
|RenderFrameImpl::remote_interfaces_|.

As of this CL, this primordial InterfaceProvider interface connection
will only service interface requests originating from the initial empty
document in the frame. For each subsequent cross-document navigation, a
new message pipe is created, with its client end being bound to the
RenderFrame's |remote_interfaces_| when the navigation commits, and its
request end sent to the browser process as part of the
DidCommitProvisionalLoad message.

The one and only exception to this rule is when a browsing context is
navigated from the initial empty document to another same-origin
document, in which case the global object (i.e. the Window instance)
associated with the initial document is reused for the new Document
corresponding to the first real navigation committed. This is to
support the following use-case:
  1) Parent frame dynamically injects an <iframe>.
  2) The parent frame calls `child.contentDocument.write(...)` to inject
     script that may stash objects on the child frame's global object
     (LocalDOMWindow). Internally, these objects may be using Mojo
     services exposed by the RenderFrameHost. The InterfaceRequests for
     these will be en-route to the RenderFrameHost for some time.
  3) The `child` frame commits a first real load that is same-origin.
  4) The global object in the child frame's browsing context is re-used.
  5) When the InterfaceRequests arrive to the browser process, the
     RenderFrameHost should not discard these requests, so that JS
     objects stashed on the global object will continue to work.

On the browser process side, when the RenderFrameHost receives an
InterfaceProvider request along with DidCommitProvisionalLoad, it
immediately breaks the old InterfaceProvider connection. Pending
interface requests on the old connection are dropped, if any. Interface
requests possibly pending on the new InterfaceProvider connection are
only dispatched after WebContentsObserver::DidFinishNavigation fires.

This ensures that interface requests racing with navigation commit will
be either ignored, or correctly serviced in the security context of the
document they originated from. In other words, when an interface request
is dispatched, the invariant will hold that calling GetLastCommittedURL
on the RenderFrameHost will return a URL that matches that of the
document from which the interface request actually originated from.

Note that |remote_interfaces_| remains a non-Channel-associated
interface. Also note that it remains the responsibility of individual
features using |remote_interfaces_| to close interface connections
already established through the InterfaceProvider when a cross-document
navigation commits, if needed in particular use-case.

Bug: 729021
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_site_isolation
Change-Id: I1eb4aa78be9627f70fdf6e18af570f083b9e306d
Reviewed-on: https://chromium-review.googlesource.com/735686Reviewed-by: Chrome Cunningham <chcunningham@chromium.org>
Reviewed-by: Nasko Oskov <nasko@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Jeremy Roman <jbroman@chromium.org>
Reviewed-by: Ken Rockot <rockot@chromium.org>
Reviewed-by: Max Morin <maxmorin@chromium.org>
Commit-Queue: Balazs Engedy <engedy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522198}

Tests multipart form POSTs with file inputs where the selected "file"
was constructed using the `File` constructor and added to a
`DataTransferItemList` (this avoids the user gesture requirement which
otherwise would consign this to manual testing.) For the non-ASCII
filenames with non-UTF-8 accept-charsets this also verifies fallback
encoding/replacement of unrepresentable characters using numeric
character references. https://github.com/whatwg/html/issues/2861

Coverage for fallback encoding is still tentative because filename
fallback encoding is not yet standardized.
https://github.com/whatwg/html/issues/3223

Bug: 661819
Change-Id: Ic646f76b0c8a0792d1214a7848d2238bcc3a76e7
Reviewed-on: https://chromium-review.googlesource.com/811625Reviewed-by: Victor Costan <pwnall@chromium.org>
Reviewed-by: Joshua Bell <jsbell@chromium.org>
Commit-Queue: Benjamin Wiley Sittler <bsittler@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522197}

In this CL, we use Notification.Builder.recoverBuilder() to create a Builder
from the notification sent by Chrome, and reset the notification's channel
ID when necessary.

Bug: 700228
Change-Id: I34cfa00ee2cb93935053020da487242709fbb36e
Reviewed-on: https://chromium-review.googlesource.com/810984
Commit-Queue: Xi Han <hanxi@chromium.org>
Reviewed-by: Anita Woodruff <awdf@chromium.org>
Reviewed-by: Yaron Friedman <yfriedman@chromium.org>
Reviewed-by: Peter Kotwicz <pkotwicz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522196}

https://chromium.googlesource.com/chromiumos/platform/system_api.git/+log/5c4b603461a7..304b21ef6d91

$ git log 5c4b60346..304b21ef6 --date=short --no-merges --format='%ad %ae %s'
2017-12-05 xiaochu add UnmountComponent method constant
2017-12-01 ljusten authpolicy: Add GetUserStatusRequest and OU errors

Created with:
  roll-dep src/third_party/cros_system_api chromiumos/platform/system_api.git

Bug=chromium:769970
TEST=None

Change-Id: Id85f7273a891b93370d1d4c2e79a73715242dbc8
Reviewed-on: https://chromium-review.googlesource.com/811628Reviewed-by: Dan Erat <derat@chromium.org>
Commit-Queue: Xiaochu Liu <xiaochu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522195}

Set the provider ID for Cast/Dial sinks discovered in browser, so that it's not UNKNOWN.
We will remove UNKNOWN as the default argument for the ctor param in a separate patch to avoid bugs like this.

Bug: 792586
Change-Id: Ia43a3e451908f8a8e3c902b7869196f2ff6a7b2a
Reviewed-on: https://chromium-review.googlesource.com/811666Reviewed-by: Derek Cheng <imcheng@chromium.org>
Commit-Queue: Takumi Fujimoto <takumif@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522194}

Bug: 699569, 718047
Cq-Include-Trybots: master.tryserver.chromium.android:android_cronet_tester;master.tryserver.chromium.mac:ios-simulator-cronet
Change-Id: I3e2d9e55cac52b392bd31c96c51cc94a361b618b
Reviewed-on: https://chromium-review.googlesource.com/809545Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Commit-Queue: Yuzhu Shen <yzshen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522193}

Stop using SpdySession instances in Http2PushPromiseIndexTests.  This
makes tests much simpler.

Rename SpdySessionKey::Equals() to SpdySessionKey::operator==() so that
it can be tested by gMock.  This conforms to the style guide "prefer to
define == [...] rather than Equals()", see
https://google.github.io/styleguide/cppguide.html#Operator_Overloading

Define struct UnclaimedPushedStream instead of using std::pair or
std::tuple.  Named members are much nicer than first, second, or get<>.

Define CompareByUrl in conformance with the same section of the style
guide "if your type doesn't have a natural ordering [...] use a custom
comparator rather than overloading <".

The next CL will merge SpdySession::UnclaimedPushedStreamContainer into
Http2PushPromiseIndex.  For this, two sets of UnclaimedPushedStream
entries will need to be maintained with identical
contents but different ordering.  The other will be done by
CompareByDelegate and will allow efficient lookup of entries for a given
Delegate.

Bug: 776415, 791054
Change-Id: I0c1912e9932b102e17528f9319163232860a8813
Reviewed-on: https://chromium-review.googlesource.com/809187Reviewed-by: Helen Li <xunjieli@chromium.org>
Commit-Queue: Bence Béky <bnc@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522192}

Adds LatencyInfo entries for 5 frame sources:
1) Renderer main.
2) Renderer compositor.
3) UI main.
4) UI compositor.
5) Display compositor.

These will be used to calculate metrics in followup patches.

This patch also:
* Discards LatencyInfos that don't result in damage in the Display,
    which is consistent with what we do earlier in the pipeline.
    The exception being snapshot requests to keep them working.
* Adds a LatencyInfo::TraceIntermediateFlowEvents helper.
* Fixes LayerTreeHostImpl::DrawLayers so it only adds
    INPUT_EVENT_LATENCY_RENDERER_SWAP_COMPONENT in the Renderer.

Bug: 790759
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel
Change-Id: I478db123d536654acf3633ca5efb4e4d20132e78
Reviewed-on: https://chromium-review.googlesource.com/783849
Commit-Queue: Brian Anderson <brianderson@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: vmpstr <vmpstr@chromium.org>
Reviewed-by: Timothy Dresser <tdresser@chromium.org>
Cr-Commit-Position: refs/heads/master@{#522191}