GitLab

This is a reland of 197dd2b5

This is an unedited reland. The old one was reverted because of a crash, but
this CL should fix the crash, making this reland safe:

https://chromium-review.googlesource.com/c/chromium/src/+/1955787

Original change's description:
> [omnibox] ZeroSuggestProvider: classify search results w/ input text
>
> Currently, when we get search suggest results back from
> ZeroSuggestProvider, nothing is bolded. This is inconsistent with
> as-you-type search suggestions.
>
> The root cause is that we purposely set AutocompleteInput::input_text
> to an empty string for all ZeroSuggest queries, to avoid leaking the
> user's current URL. (The URL is separately included in eligible).
>
> The lack of bolding is an unintended side effect.
>
> This CL fixes that, by re-classifying all the search suggest results
> that come back with the real input text on the client side.
>
> This is primarily intended to improve On Focus Query Refinements,
> but this will affect all ZeroSuggest modes that return search results
> from a non-empty omnibox. (Should not affect NTP).
>
> Note nothing is bolded at all when the input text is empty (on NTP).
>
> Navigational suggestions are not affected by this CL. They are
> currently not bolded at all.
>
> Bug: 963173
> Change-Id: I7982c780f4b03f3d2b4992051d291fe1cd86f4b7
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1947765
> Reviewed-by: manuk hovanesian <manukh@chromium.org>
> Commit-Queue: Tommy Li <tommycli@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#721114}

Bug: 963173
Change-Id: I4a6d1195a47cab19ef4cc9e1b3756e3f6ab9f20c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954250Reviewed-by: manuk hovanesian <manukh@chromium.org>
Commit-Queue: Tommy Li <tommycli@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722655}

For every cookie included in a request log metrics on the scheme that
the cookie was set by relative to the scheme of the request url.

Bug: 1028984
Change-Id: Icce3970a7cfe5212d4059e3e1256db633fb451c6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1938035
Commit-Queue: Steven Bingler <bingler@chromium.org>
Reviewed-by: Maksim Orlovich <morlovich@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Reviewed-by: Lily Chen <chlily@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722654}

Now integration tests will launch WebEngine with --enable-logging=stderr
and dup it's own stderr to WebEngine.

Bug: 1009099
Change-Id: I9b0e15f22fc7a758ccdc5e78a3548a34bca6651d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1956225
Commit-Queue: Sergey Ulanov <sergeyu@chromium.org>
Commit-Queue: Wez <wez@chromium.org>
Auto-Submit: Sergey Ulanov <sergeyu@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722653}

D3D11VideoDecoder is crashing periodically with a null |decode_cb_|.
It's unclear how this is happening.

This CL checks for the null callback and skips it, to prevent the
crash.  It also adds some debugging that hopefully will tell us
where the null callback is coming from:

 - Provided to Decode()
 - Popped off the buffer queue in DoDecode
 - Cleared elsewhere

Bug: 1012464
Change-Id: Idc4eb1842aa98989b9ceea30ed5aa769f0a63ce6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954735Reviewed-by: Ted Meyer <tmathmeyer@chromium.org>
Commit-Queue: Frank Liberato <liberato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722652}

6 pixel video tests are now expected to pass on SkiaRenderer Vulkan for
android after some bug fixes. Hence removing them from test expectation.

Bug: 991291
Change-Id: Ifc98a2b674dbd27c3980df6e69b139201eaa58dc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954258Reviewed-by: Brian Sheedy <bsheedy@chromium.org>
Commit-Queue: vikas soni <vikassoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722651}

These tests use properties that are not currently allowed by the spec.
Therefore they are not part of the basic ::marker functionality.

BUG=1031667

Change-Id: I79743233bdcee10715bf34bca56c49c7168832b9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1955729Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Commit-Queue: Oriol Brufau <obrufau@igalia.com>
Cr-Commit-Position: refs/heads/master@{#722650}

https://pdfium.googlesource.com/pdfium.git/+log/f3883e32a739..170f685582e1

git log f3883e32a739..170f685582e1 --date=short --first-parent --format='%ad %ae %s'
2019-12-06 kmoon@chromium.org Add utility for filter-encoding PDF streams
2019-12-06 patrik@pspdfkit.com Allow reading font name from Windows `name` table entry.

Created with:
  gclient setdep -r src/third_party/pdfium@170f685582e1

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
https://autoroll.skia.org/r/pdfium-autoroll
Please CC pdfium-deps-rolls@chromium.org 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

Bug: None
Tbr: pdfium-deps-rolls@chromium.org
Change-Id: I15c3c786e6b81d1c17a27aa69951c8aaee3adfef
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1956084Reviewed-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@{#722649}

This change adds TranslateInfobarMetricsRecorder, which will be used
by the Translate Messages logic to log Translate-specific Modal and Banner
events. This change also replicates existing Translate histogram metrics logging
in the TranslateInfobarController in InfobarTranslateCoordinator. Those
histograms are moved to the translate_constants file.

Bug: 1014959
Change-Id: I72fcb71569017a5d3d15f42bd19cf901a260411f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1949730
Commit-Queue: Chris Lu <thegreenfrog@chromium.org>
Reviewed-by: Alexei Svitkine <asvitkine@chromium.org>
Reviewed-by: Peter Lee <pkl@chromium.org>
Reviewed-by: Sergio Collazos <sczs@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722648}

Here is the overall plan of attack here:
Currently, there are three key classes:

* ExtensionInstalledBubble is a BubbleDelegate (bad)
* ExtensionInstalledBubbleUi is a BubbleUi (bad)
* ExtensionInstalledBubbleView is a BubbleDialogDelegateView (good)

The ideal end state is that all clients of ExtensionInstalledBubble
directly use ExtensionInstalledBubbleView instead, and then the former
two classes are deleted.

To get to that state, we:
1) Iteratively break every dependency from EIBV to EIB or EIBU, so that
   eventually an EIBV can be constructed without an EIB or EIBU - we do
   this by passing data from the EIB or EIBU into the EIVB's constructor
   directly so that it has it without asking the EIB.
2) Once EIBV does not depend on EIB or EIBU, make EIBV public (promote
   it into a header file), then have ShowExtensionInstalledBubble
   (the only non-test entry point to EIB) call EIBV directly
3) Delete EIB & EIBU :)

Bug: 496955
Change-Id: Ic013bf89badf507d13599ff92ebc50b6afa50f58
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954739Reviewed-by: Peter Boström <pbos@chromium.org>
Commit-Queue: Elly Fong-Jones <ellyjones@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722647}

This CL merges web_rtc_stats.h into the rtc_stats.h, and deletes it.
The former is a collection of adhoc testing and production
method declarations, as well as an alias, solely used within Blink.

It is the last web_rtc_xxx.h header in the Blink API layer,
finally being moved down to renderer/.

BUG=919392, 787254
R=guidou@chromium.org, haraken@chromium.org

Change-Id: I057fb72f474cecc79fb7d067deb2e351719f9116
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1953386
Commit-Queue: Antonio Gomes <tonikitoo@igalia.com>
Reviewed-by: Guido Urdaneta <guidou@chromium.org>
Reviewed-by: Kentaro Hara <haraken@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722646}

Currently the number of throughput UKM reporting is un-expectedly low on
canary channel. We suspect that the sampling logic could cause that.
This CL removes the sampling logic. Once we confirm that the reason is
the sampling, we will likely invent another sampling logic.

Bug: 1029964
Change-Id: I104de8f63ef261a72a9d93edeae2fd62a8235e3b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1946872Reviewed-by: Stephen Chenney <schenney@chromium.org>
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Commit-Queue: Sadrul Chowdhury <sadrul@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722645}

Remove the ICU iterator and use the precomputed graphemes from
RenderText.

The precomputed graphemes are landed here:
  https://chromium-review.googlesource.com/c/chromium/src/+/1945277/

Bug: 1025561
Change-Id: I9680250b24c00de9399e243b0d8a1ee9aca11a0b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1952026
Commit-Queue: Etienne Bergeron <etienneb@chromium.org>
Reviewed-by: Robert Liao <robliao@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722644}

This is a reland of 4430b4c3

Original change's description:
> display: Deduplicate display transform code with overlay transform code
>
> Display rotation transforms are a subset of overlay transforms, so we can
> reuse OverlayTransformToTransform to compute display rotation matrices.
>
> Bug: 1019015
>
> Change-Id: I95d099323cb866ccc4b9cea404e1e6f8f2dfff65
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1913604
> Commit-Queue: Michael Spang <spang@chromium.org>
> Reviewed-by: Khushal <khushalsagar@chromium.org>
> Reviewed-by: Ken Buchanan <kenrb@chromium.org>
> Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
> Reviewed-by: Mitsuru Oshima (OOO - back on Dec 2nd) <oshima@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#717832}

TBR=oshima@chromium.org,sadrul@chromium.org,kenrb@chromium.org,kenrb@chromium.org

Bug: 1019015
Change-Id: I3eaa7819971e9ebd51329f22506cb6bb905d429a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1931131
Commit-Queue: Michael Spang <spang@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Reviewed-by: Michael Spang <spang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722643}

This change:
1) Removes BubbleController::GetName and its various implementations
2) Removes BubbleController::GetVisibleTime and its various implementations
3) Cleans up ExtensionInstalledBubbleBrowserTest to not depend on GetName

Bug: 496955
Change-Id: Iee271ea4d4dbae96f8dd77e2ce2be87bd36ecc3f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954260
Commit-Queue: Elly Fong-Jones <ellyjones@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722642}

Report animation smoothness for the transition into the hotseat in the
HomeLauncher.

The original change caused use-of-uninitialized-value flaky tests. These
were caused because the HotseatTransitionAnimator could outlive the
ShelfWidget in some tests.
This is a possible fix for the issue by avoiding the destruction of
|hotseat_transition_animator_| until the Shelf Widget is destroyed, thus
destroying the animator itself. Shelf Widget destruction occurs usually
after Shutdown().

Original CL: http://crrev.com/c/1917941

Bug: 1022178, 1022177, 1028255, 1030222
Change-Id: Ibf2030df7408704e8aa96ca863bcd55ba333174f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954355Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Ana Salazar <anasalazar@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722641}

This CL fixed two issues:
- The window jittering issue during dragging. It's because on Krane, we
can still get small scroll update events even the finger is still. To
avoid the window jittering, we only update the window if the scroll
update value is large enough.
- Overview never fade in issue. It's caused by the same reason. The
small scroll update event will restart the show-overview-timer each
time the event comes in thus prevent overview from showing up. We only
restart the time if the scroll update value is large enough.

Bug: 1031382, 1031381, 997885
Change-Id: Ie1f4ad0d6b504f12e2d9ccc9f73cc54c7169d716
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1955906
Commit-Queue: Xiaoqian Dai <xdai@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722640}

Bug: 1019521
Change-Id: I3a4e75f5d640760268435766bdd7efc2fa1c36f2
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1941072Reviewed-by: Luke Z <lpz@chromium.org>
Reviewed-by: Changwan Ryu <changwan@chromium.org>
Commit-Queue: Andrew Luo <aluo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722639}

Bug: b/144316700
Test: manual test
Change-Id: I5b0abdfe3358a2416a213bf0661cac560869b312
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1955803
Commit-Queue: Li Lin <llin@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722638}

This is provided alongside OverlayRequests added to request queues and
manages cancelling the request if its corresponding InfoBar is removed
from its InfoBarManager.

Bug: 1030357
Change-Id: Ic6644864cb7a5f6a89cca7b84baa090258415dea
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1951723
Commit-Queue: Kurt Horimoto <kkhorimoto@chromium.org>
Reviewed-by: Sergio Collazos <sczs@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722637}

Data frames will now be coalesced into one.

Change-Id: Ibd65d7098cbb81d96d1325333867892cd3ce2c4b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1956008Reviewed-by: Ryan Hamilton <rch@chromium.org>
Commit-Queue: Ryan Hamilton <rch@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722636}

crrev.com/722312 added the service to most .cmx files, but missed
tests-with-exec.cmx .

Bug: 1027048
Change-Id: Ic5a0dd85ea5b7b0c483f9466006f5d7ec6983fd9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1956004
Auto-Submit: Sergey Ulanov <sergeyu@chromium.org>
Reviewed-by: David Dorwin <ddorwin@chromium.org>
Commit-Queue: David Dorwin <ddorwin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722635}

Change-Id: I50943050017711479968fb1359af05d15f0199bb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1931538Reviewed-by: Demetrios Papadopoulos <dpapad@chromium.org>
Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Aran Gilman <gilmanmh@google.com>
Cr-Commit-Position: refs/heads/master@{#722634}

 - Moving ComponentsDomHandler to its own .h/cc files,
   and renaming to ComponentsHandler.
 - Moving some logic from ComponentsUI to ComponentsHandler so that proper
   lifetime methods (OnJavascriptAllowed/Disallowed) can be leveraged.

This is in preparation of migrating this page to JS modules.

Bug: 1028829
Change-Id: I530c439b3e659f3b1b5e5dc3ef05f2a3df88559e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1939893
Commit-Queue: Demetrios Papadopoulos <dpapad@chromium.org>
Reviewed-by: Sorin Jianu <sorin@chromium.org>
Reviewed-by: Rebekah Potter <rbpotter@chromium.org>
Auto-Submit: Demetrios Papadopoulos <dpapad@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722633}

Chrome OS SplitSettingsSync has a separate control for syncing OS data
types that isn't controlled by the browser's sync-the-feature. Change
ModelType::PRINTERS and WIFI_CONFIGURATIONS to run when the transport
layer is enabled when kSplitSettingsSync is enabled. They already have
checks for the OS sync feature being enabled.

Bug: 1013466, 1031549
Test: updated sync_integration_tests
Change-Id: If7fae05bda53bb68d2bcde0ce2fab6a9a7f7d416
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954361
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Marc Treib <treib@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722632}

Bug: 1027737
Change-Id: I6afe294f787a2450516e5a9728e2ec1798e5cd83
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1955823
Commit-Queue: Daniel Murphy <dmurph@chromium.org>
Commit-Queue: Ken Rockot <rockot@google.com>
Auto-Submit: Daniel Murphy <dmurph@chromium.org>
Reviewed-by: Ken Rockot <rockot@google.com>
Cr-Commit-Position: refs/heads/master@{#722631}

Overview and clamshell split view shall end if you double click the edge
of the split view window where it meets the overview grid.

Fixed: 1006950
Change-Id: I24c1901814724cf24275e947749baf5a408bb648
Bug: 1006950
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1951124
Commit-Queue: Avery Musbach <amusbach@chromium.org>
Reviewed-by: Xiaoqian Dai <xdai@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722630}

AppActivityRegistry class is responsible for:
* collecting information about app state and usage
* combining app usage information with restrictions from policy
* storing the data between the sessions

This changelist creates the skeleton of the class and will be
followed with remaining functionality implementation.

Bug: 1015658
Change-Id: I306830f756c974371551493379f6edb510f6e5c3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954448Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Aga Wronska <agawronska@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722629}

Add a function that'll return to WebUI the top 5 recently launched
installed apps sorted by site engagement) that have engagement within
the specified time period. Time period will be specified by the window
that the user selects in the clear browsing data dialog. This list of
apps will be used to warn users that they will be clearing data for
installed apps, similar to what is done on Android for important sites.

Bug: 997258
Change-Id: I7e06b6e000ffd0b2eec6173de22cc9b69c380659
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1906876
Commit-Queue: Jarryd Goodman <jarrydg@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Reviewed-by: Dominick Ng <dominickn@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722628}

No change to production logic.

This adds FlagOverrideHelper to help apply batches of commandline flags
and base::Features to the CommandLine singleton. It also is responsible
for mapping flag names to Flag instances, resolving a TODO. This class
is currently not used in production, but will be used in a future CL to
apply flag overrides indicated from the developer UI.

This adds robolectric unit tests to cover several corner cases.

Bug: 981143
Test: run_android_webview_junit_tests -f *FlagOverrideHelperTest#*
Change-Id: Ie982fc66aa62bf3d51ffe74c04c708b435157261
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1952032
Commit-Queue: Nate Fischer <ntfschr@chromium.org>
Reviewed-by: Richard Coles <torne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722627}

Summary
=======

This CL consolidates creation of network::mojom::URLLoaderFactoryParams
into a new class - content::URLLoaderFactoryParamsHelper.


Motivation
==========

This CL helps simplify the code by lowering the number of parameters
that need to be propagated by various methods.

This CL also makes it easier to audit security properties of the code,
by grouping all assignments into the helper class.  For example,
after the CL, it is hopefully much easier to see when |is_trusted| is
set.

This CL also avoids exposing unnecessary public //content APIs through
RenderProcessHost (e.g. by moving |is_trusted| and
|network_isolation_key| parameters into the new helper class).


No intended changes in behavior
===============================

In general, this CL intends to introduce no changes in behavior.  In
particular, the values that various scenarios put inside
network::mojom::URLLoaderFactoryParams should not be changed by this CL
(except in the dedicated worker scenario listed below).

The only change of URLLoaderFactoryParams is that WebPreferences
of the |ancestor_render_frame_host| will now be (correctly) taken into
account when creating a factory for a dedicated worker.

Another change in behavior introduced by this CL, is changing which
|origin| is passed to the WillCreateURLLoaderFactory method when
creating a factory for content scripts / isolated worlds:
- Before the CL, the isolated world origin would be passed
- After the CL, the main world origin is passed
This change should be insignificant:
- AwContentBrowserClient::WillCreateURLLoaderFactory and
  ShellContentBrowserClient::WillCreateURLLoaderFactory overrides do not
  dereference the |request_initiator| parameter.
- ChromeContentBrowserClient::WillCreateURLLoaderFactory passes
  the origin into signin::ProxyingURLLoaderFactory::MaybeProxyRequest
  and then into gaia::IsGaiaSignonRealm.  I believe that it is okay if
  content script requests are also proxied - if an extension has
  permission to inject content scripts then it effectively "owns" the
  origin anyway.

Bug: 1025303
Change-Id: I56c224dc67bcc637e77d4c9ffad91b87cb99d83c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1915299
Commit-Queue: Łukasz Anforowicz <lukasza@chromium.org>
Reviewed-by: Alex Moshchuk <alexmos@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722626}

Update cast_runner to delay connections for bindings and rewrite rules
until after the AppConfig is returned. This AppConfig will now contain
the agent_url, which will be used to contact the correct agent for
bindings, etc.

Bug: b/141573972
Test: cast_runner_integration_tests, Launch application

Change-Id: Ida6966fba6039fc2f9ea9d7c136711d13a0a5a2a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1895401Reviewed-by: Wez <wez@chromium.org>
Commit-Queue: Connor Lange <connorl@chromium.org>
Auto-Submit: Connor Lange <connorl@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722625}

The present CL addresses two blatantly broken things that together form
the behavior shown in the video linked on Monorail:
1. Even though ScreenOrientationController is only for tablet mode,
SplitViewController always gets display orientation from the
ScreenOrientationController, even in clamshell mode. If you change the
display orientation in clamshell mode, then the
ScreenOrientationController continues to indicate the display
orientation as it was at the onset of clamshell mode, and so the
SplitViewController continues to use that display orientation until you
switch to tablet mode and back to clamshell mode.
2. When the SplitViewController detects portrait orientation, clamshell
split view uses snap positions on the top and bottom, while WindowState
expects snap positions on the left and right. This disagreement between
the two classes disrupts their collaboration on clamshell split view.
SplitViewController's opinion is based on longstanding behavior of
tablet split view, and WindowState's opinion is based on longstanding
behavior of window snapping in clamshell mode. Ben McCanny agrees with
WindowState: the snap positions should be on the left and right because
that is how window snapping in clamshell mode has always worked.

Fixed: 1023845
Change-Id: Ie81346904552acae88cb36227f30de7d7fceed6b
Bug: 1023845
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1949453
Commit-Queue: Avery Musbach <amusbach@chromium.org>
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722624}

To avoid targeting many versions of Bison, we will now require editors
of xpath_grammar.y to manually regenerate the generated files. See
xpath_grammar.y for instructions on running Bison.

Since we're only targeting new versions of Bison, we can now replace
the %pure-parser directive with the equivalent %define. As a result,
we no longer need to silence the warning about %pure-parser.

Also removing //third_party/bison from DEPS, which contains a snapshot
of a Windows fork of Bison 2.4.1 (from 2012).

Bug: 1028421
Change-Id: I8deb986f0cafb8e7f093c5def71d427709a4a845
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1954609Reviewed-by: Chris Palmer <palmer@chromium.org>
Reviewed-by: Nico Weber <thakis@chromium.org>
Commit-Queue: Dan McArdle <dmcardle@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722623}

https://chrome-internal.googlesource.com/chrome/src-internal.git/+log/08b668ce35ac..7471ec58dc43


Created with:
  gclient setdep -r src-internal@7471ec58dc43

If this roll has caused a breakage, revert this CL and stop the roller
using the controls here:
https://skia-autoroll.corp.goog/r/src-internal-chromium-autoroll
Please CC jbudorick@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.chrome.try:linux-chromeos-chrome
Bug: None
Tbr: jbudorick@google.com
Change-Id: Icbdf9db9dff2b16a2d182f7cacd17dce83660f7c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1955863Reviewed-by: chromium-internal-autoroll <chromium-internal-autoroll@skia-corp.google.com.iam.gserviceaccount.com>
Commit-Queue: chromium-internal-autoroll <chromium-internal-autoroll@skia-corp.google.com.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#722622}

After MimeHandlerViewInCrossProcessFrame shipped, application/pdf can
be blocked by CORB in a //chrome-agnostic way (similarly to how
//services/network layer knows about and blocks application/zip).

In particular, the current CL has been unblocked by the recent removal
of kMimeHandlerViewInCrossProcessFrame feature in r720036 - see:
https://chromium-review.googlesource.com/c/chromium/src/+/1869121/26/chrome/browser/net/system_network_context_manager.cc
and
https://chromium-review.googlesource.com/c/chromium/src/+/1869121/26/content/public/common/content_features.cc

Bug: 533069
Change-Id: I9af8db439fbe270cc3519fc5c6c4699c5842a7d9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1946846Reviewed-by: John Abd-El-Malek <jam@chromium.org>
Reviewed-by: Ken Buchanan <kenrb@chromium.org>
Auto-Submit: Łukasz Anforowicz <lukasza@chromium.org>
Commit-Queue: Ken Buchanan <kenrb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722621}

Currently, we have duplicate implementations of viz::ResourceFormat
to WGPUTextureFormat conversion. This CL exposes a new ToWGPUFormat
function in resource_format_utils and removes duplicate code.

Note: ToWGPUFormat calls ToDawnFormat under the hood and casts from
dawn::TextureFormat to WGPUTextureFormat. Looking at the
definitions [1][2], it seems like they are generated from the same
place, so this cast is safe?

[1] https://cs.chromium.org/chromium/src/out/Debug/gen/third_party/dawn/src/include/dawn/webgpu.h?type=cs&q=WGPUTextureFormat&sq=package:chromium&g=0&l=252
[2] https://cs.chromium.org/chromium/src/out/Debug/gen/third_party/dawn/src/include/dawn/webgpu_cpp.h?type=cs&q=BGRA8Unorm+f:dawn&sq=package:chromium&g=0&l=176

Change-Id: I6159240c40c084a6d9fd3c8f7b4ff66ab03063f7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1950750Reviewed-by: Eric Karl <ericrk@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Brian Ho <hob@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722620}

Bug: 907493
Change-Id: I3b61c48aaf2ac660f52bf797c2a392cb43abc140
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1902427
Commit-Queue: Robin Lewis <wrlewis@google.com>
Reviewed-by: Tien Mai <tienmai@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722619}

Bug: b/145738114
Test: loaded webpage on device
Change-Id: Ia65a07636ea4722a5467e0a9c37d9467ce7a5e28
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1956005
Commit-Queue: Albert Chaulk <achaulk@chromium.org>
Reviewed-by: Daniel Nicoara <dnicoara@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722618}

Now that blobs from IndexedDB are backed & read by the IndexedDB system,
they are no longer sent to the renderer as files. This means no content
policy changes or grants are necessary.

Tracking document:
https://docs.google.com/document/d/18suNOOzuEJbqgRJF0MB2VqdTyYqS4cvI2PGaCpyPXSw/edit?ts=5de6dc30&pli=1#heading=h.tw2iw2qfgduj

Bug: 1015212
Change-Id: If3d2fd80b3254f19a56ab16ed7fc39cf5f0cf684
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1955813Reviewed-by: Ken Rockot <rockot@google.com>
Reviewed-by: enne <enne@chromium.org>
Commit-Queue: Daniel Murphy <dmurph@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722617}

Split apart some ExtensionActionViewController related unittests that
were exercising multiple scenarios. Specifically:

ExtensionActionWantsToRunAppearance ->
    ExtensionActionWantsToRunAppearance +
    OverflowedPageActionAppearance

ExtensionActionBlockedActions ->
    BrowserActionBlockedActions +
    PageActionBlockedActions +
    PageActionBlockedActionsInOverflow

This will help with adapting these to the ExtensionsMenu behavior,
which will be done in a followup.

This does not change any behavior in production.

Bug: 984654

Change-Id: Iadd2374321ae8868a2b846049ad9886e4bf35e0f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1938307Reviewed-by: Peter Boström <pbos@chromium.org>
Commit-Queue: Devlin <rdevlin.cronin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#722616}