GitLab

Currently, extensions which are disabled in incognito mode are also loaded in an
incognito renderer. This is redundant and can cause non-required bindings to be
generated in the renderer. Also, it has potential security implications.

Prevent this by only loading incognito-enabled extensions in an incognito
renderer. However extensions which can't be enabled in the incognito mode (e.g.
platform apps etc.) still need to be loaded in incognito renderers, so that
incognito tabs can connect with them (issue 305394).

Also, add tests for the same.

BUG=527548

Review-Url: https://codereview.chromium.org/2766263003
Cr-Commit-Position: refs/heads/master@{#462707}

Also, re-enables switching from sw to accelerated canvas.

BUG=704571

Review-Url: https://codereview.chromium.org/2799033002
Cr-Commit-Position: refs/heads/master@{#462706}

This patch edits Element::rebuildPseudoElementLayoutTree() to make sure
it mirrors Element::updatePseudoElement() called in
Element::recaclStyle() by calling Element::updateFirstLetter() in the
case the PseudoElement is firstLetter as is done in
Element::updateFirstLetter() to ensure we don't call
Element::retachLayoutTree() uncoditionally.

Adds crash test and expectation as well.

BUG=700655

Review-Url: https://codereview.chromium.org/2752163003
Cr-Original-Commit-Position: refs/heads/master@{#462373}
Committed: https://chromium.googlesource.com/chromium/src/+/6d202487ff823def4672599f51e9f13880e6b877
Review-Url: https://codereview.chromium.org/2752163003
Cr-Commit-Position: refs/heads/master@{#462705}

It's possible for the app to make CORS requests from its guests without
any webview-accessible-resources, via injected content scripts.

BUG=613335
TEST=See bug comment 37 for a repro app.

Review-Url: https://codereview.chromium.org/2803963002
Cr-Commit-Position: refs/heads/master@{#462704}

BUG=597347
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2799563003
Cr-Commit-Position: refs/heads/master@{#462703}

Currently people-page uses listenOnce() from util.js, which it only got access to transitively from other deps and is fragile.

BUG=None
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2807513002
Cr-Commit-Position: refs/heads/master@{#462702}

Disabling the test, because it has been consistently failing
for the last 3 builds on Mac10.9 Tests (dbg).  OTOH, note that
FindIt says that it cannot find the culprit CL and thinks that
the test is just flaky.  In either case, disabling the test seems like
the right thing to do.

BUG=709231
TBR=hta@chromium.org

Review-Url: https://codereview.chromium.org/2799663006
Cr-Commit-Position: refs/heads/master@{#462701}

The HTML5 spec indicates that a load of an invalid URL via links or the
window.location API should be a no-op.  However, window.open calls
with an invalid URL should still create a new window, but pointed to
about:blank.

BUG=709059

Review-Url: https://codereview.chromium.org/2799953002
Cr-Commit-Position: refs/heads/master@{#462700}

TBR=jochen
BUG=709136

Review-Url: https://codereview.chromium.org/2802083002
Cr-Commit-Position: refs/heads/master@{#462699}

The last couple import jobs failed when uploading the second patchset;
I think that it failed at this point (and not when uploading the
CL initially) because when initially uploading the CL it uses -f.

We want to use -f in general to ignore presubmit warnings for things
like tabs.

Review-Url: https://codereview.chromium.org/2802093002
Cr-Commit-Position: refs/heads/master@{#462698}

In my last CL I added the serialization code but not deserialization.

TBR=danakj@chromium.org
BUG=704972

Review-Url: https://codereview.chromium.org/2801603009
Cr-Commit-Position: refs/heads/master@{#462697}

Reland of Changed CSSParserContext argument from * to & for null safety. (patchset #1 id:1 of https://codereview.chromium.org/2801893003/ )

Reason for revert:
This patch was reverted for breaking tests when it was landed at the same time as a clashing patch https://codereview.chromium.org/2783543002/. Both of these patches were reverted.

Landing them one at a time and rebasing should solve the problem.

Original issue's description:
> Revert of Changed CSSParserContext argument from * to & for null safety. (patchset #4 id:60001 of https://codereview.chromium.org/2786153004/ )
>
> Reason for revert:
>
> Findit identified CL at revision 462380 as the culprit for
> failures in the build cycles as shown on:
> https://findit-for-me.appspot.com/waterfall/culprit?key=ag9zfmZpbmRpdC1mb3ItbWVyRAsSDVdmU3VzcGVjdGVkQ0wiMWNocm9taXVtLzkxMmU1Zjc4N2VlNDE1Zjg3NDkyOGI5Njg0ZWE1MGRjNzVmODVkNDAM
>
> Original issue's description:
> > Changed CSSParserContext argument from * to & for null safety.
> >
> > Changed the CSSParserContext argument to the parseSingleValue method in
> > the property APIs from a pointer to a reference to ensure that nullptr
> > cannot be passed.
> >
> > Note that this patch changes the syntax inside parseSingleValue
> > implementations to get the address of CSSParserContext references to pass to
> > utility methods that still expect a pointer. Future patches should change these
> > utility methods to also take references instead of pointers, at which point the
> > syntax inside parseSingleValue implementations can be changed back to pass the
> > object.
> >
> > This patch:
> > - changed the CSSParserContext parameter to parseSingleValue in
> >   CSSPropertyAPIMethods.json5 so that the generated API header files (including
> >   the base class header file) would update
> > - changed the parameter in each of the .cpp files for the property APIs in
> >   core/css/properties
> > - updated the single call site in CSSPropertyParser.cpp
> > - updated the implementations of parseSingleValue in the property APIs to use
> >   reference syntax
> >
> > Diff of generated files:
> > https://gist.github.com/BugsNash/1850f85afc7313dd3dac1a890299ad07/revisions
> >
> > BUG=668012
> >
> > Review-Url: https://codereview.chromium.org/2786153004
> > Cr-Commit-Position: refs/heads/master@{#462380}
> > Committed: https://chromium.googlesource.com/chromium/src/+/912e5f787ee415f874928b9684ea50dc75f85d40
>
> TBR=shend@chromium.org,meade@chromium.org,bugsnash@chromium.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=668012
>
> Review-Url: https://codereview.chromium.org/2801893003
> Cr-Commit-Position: refs/heads/master@{#462383}
> Committed: https://chromium.googlesource.com/chromium/src/+/b7aae4a78d2902cb83671f7ef7792b14180ee462

TBR=meade@chromium.org,lukasza@google.com,jkrcal@google.com,foolip@google.com,ellyjones@google.com,shend@chromium.org,findit-for-me@appspot.gserviceaccount.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=668012

Review-Url: https://codereview.chromium.org/2805153002
Cr-Commit-Position: refs/heads/master@{#462696}

BUG=708808

Review-Url: https://codereview.chromium.org/2799843002
Cr-Commit-Position: refs/heads/master@{#462695}

Skip things mushrome doesn't have yet:
* Cursor manager
* DeviceDataManager

Also fix some dbus service initialization.

After this CL we boot to the login screen. You can't log in yet because
input events are still broken.

BUG=709097
TEST=manual, chrome --mash and --mus start on desktop and on device

Review-Url: https://codereview.chromium.org/2801023004
Cr-Commit-Position: refs/heads/master@{#462694}

|force_quic_port_| is a dead field.

BUG=475060

Review-Url: https://codereview.chromium.org/2802123002
Cr-Commit-Position: refs/heads/master@{#462693}

Now supporing "video/webm; codecs=vp09...." in calls to EME's
requestForMediaKeySystemAccess(). Must run with the following flag:
--enable-new-vp9-codec-string.

This CL does not modify existing support for "video/mp4". To enable
in MP4 run with --enable-vp9-in-mp4.

Runtime flags are checked in MimeUtilInternal ParseVp9CodecID().

BUG=687627

Review-Url: https://codereview.chromium.org/2793163003
Cr-Commit-Position: refs/heads/master@{#462692}

Add a feature to enable the new settings page with site breakdown. Also,
add support so that when Data Saver is enabled on Android, compression
stats collects site information. Don't collect site info when Data Saver
is disabled, and when data usage reporting is disabled (when the proxy
is turned off) keep the stats. The stats can be manually deleted using
a button on the settings page.

BUG=615560

Review-Url: https://codereview.chromium.org/2791563002
Cr-Commit-Position: refs/heads/master@{#462691}

Add support for load/remove persistent license at Android level.

Load:
1. Load eme session ID to key set ID map from persistent storage.
2. MediaDrm.restoreKeys.

Remove:
1. Generate key release request.
2. Update session with key release response.
3. Clear persistent storage.

BUG=493521
TEST=test page

Review-Url: https://codereview.chromium.org/2796843002
Cr-Commit-Position: refs/heads/master@{#462690}

Follow up to https://codereview.chromium.org/2639303004

BUG=668012

Review-Url: https://codereview.chromium.org/2752563002
Cr-Commit-Position: refs/heads/master@{#462689}

https://skia.googlesource.com/skia.git/+log/2e2b27fcc150..9a121cc6ad74

$ git log 2e2b27fcc..9a121cc6a --date=short --no-merges --format='%ad %ae %s'
2017-04-06 robertphillips Revert "Make GrCopySurfaceOp friendlier to GrSurfaceProxy"
2017-04-05 hans Specialize boxBlur(Interp) based on 'transpose' parameter
2017-04-06 msarett Revert "Manual revert: Add support for writing ICC profiles to webp encoder"
2017-04-06 mtklein jumper, gather_8888
2017-04-06 robertphillips Make GrCopySurfaceOp friendlier to GrSurfaceProxy
2017-04-06 mtklein tweaks to make gather_* easier in SkJumper
2017-04-06 mtklein jumper, turn off a few fancy features
2017-04-06 reed make copyTo_with_allocator android only

Created with:
  roll-dep src/third_party/skia


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

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_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;master.tryserver.chromium.android:android_optional_gpu_tests_rel
TBR=jvanverth@chromium.org

Change-Id: I1cc80e87fd3411abe7b0e9e40a9d98b20045280c
Reviewed-on: https://chromium-review.googlesource.com/470527Reviewed-by: Skia Deps Roller <skia-deps-roller@chromium.org>
Commit-Queue: Skia Deps Roller <skia-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#462688}

[CrOS Tether] Update SecureChannel and BleConnectionManager to use CryptAuthService instead of a Delegate.

BUG=672263

Review-Url: https://codereview.chromium.org/2800883003
Cr-Commit-Position: refs/heads/master@{#462687}

Before this CL, if a LoFi image was somehow loaded without the
SERVER_LOFI_ON previews state bit set on the respective request, Blink
would not reload that image if the user opts out of the LoFi experience
on that page.

This CL causes Blink to recognize any loaded image with the LoFi
response headers as a LoFi image, regardless of its previews state bits.

BUG=707916

Review-Url: https://codereview.chromium.org/2795073003
Cr-Commit-Position: refs/heads/master@{#462686}

This CL wraps done() method in the global scope, which is defined by
testharness.js to prevent a test from finishing prematurely.

2 layout tests are found from this check and fixed accordingly.

BUG=708817
TESTS=
webaudio/AudioParam/audioparam-update-value-attribute.html
webaudio/DynamicsCompressor/dynamicscompressor-basic.html

Review-Url: https://codereview.chromium.org/2804103002
Cr-Commit-Position: refs/heads/master@{#462685}

This avoids updating the surface contents scale based on the current
transform if layer_transforms_should_scale_layer_contents is false. The
result is that render surfaces size doesn't need to be adjusted during
transform animations in the Chrome OS UI. This is a prerequisite to
caching of render surfaces and is the preferred behavior as UI raster
scale should not change during these animations.

BUG=708513
TEST=cc_unittests --gtest_filter=LayerTreeHostCommonTest.CanAdjustRasterScaleTest
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2795963002
Cr-Commit-Position: refs/heads/master@{#462684}

Preparing the MultiprocessTestClientLauncher for transitioning to the
new base ChildProcessLauncher. More specifically, making ITestClient
similar to IChildProcessService. This requires moving extra methods to
a new interface ITestController, and creating a callback interface
ITestCallback that lets the service sets the ITestController interface
on the client.

BUG=702316

Review-Url: https://codereview.chromium.org/2781913002
Cr-Commit-Position: refs/heads/master@{#462683}

On certain platforms, the icons in our toolbar were tinted pink instead
of dark grey.  I tracked this down to the API we use to retrieve the
icon color.  We retrieve the color by using a resourceID from the current
context and pass it to a Chromium compat API.  The problem is that our
resourceID is based on the current Theme and the compat getColor method
is not Theme aware.  Therefore the ID retrieves a different color
resource than what we expected.  Note that this is only a problem on M
and later releases.

The fix for us is to just use the Android ContextCompat class for the
color lookup instead of the Chromium version.

BUG=704150

Review-Url: https://codereview.chromium.org/2799803004
Cr-Commit-Position: refs/heads/master@{#462682}

BUG=672263

Review-Url: https://codereview.chromium.org/2803913002
Cr-Commit-Position: refs/heads/master@{#462681}

This suppresses unnecessary error message (due to the local hardware
configuration) from the exception from should.throw().

BUG=709167
TEST=webaudio/dom-exceptions.html

Review-Url: https://codereview.chromium.org/2803913003
Cr-Commit-Position: refs/heads/master@{#462680}

The new rule is that coordinators don't decide how they are presented
(presentation, containment, etc.). The parent can decide in
-childCoordinatorDidStart|WillStop.
That also means that the CoordinationContext is unnecessary.

BUG=none
R=marq@chromium.org,rohitrao@chromium.org,edchin@chromium.org,sczs@chromium.org

Review-Url: https://codereview.chromium.org/2798833003
Cr-Commit-Position: refs/heads/master@{#462679}

CommitSurfaceHierarchy() moves pending_frame_callbacks_ into
frame_callbacks_ before calling UpdateSurface() so that the ack doesn't
arrive too early.

exo::Surface is the only class which needs
MojoCompositorFrameSinkClient::WillDrawSurface(). With this change, we
can remove WillDrawSurface from the mojo interface and reduce IPC overhead.

BUG=671202, 709076

Review-Url: https://codereview.chromium.org/2786643002
Cr-Commit-Position: refs/heads/master@{#462678}

For the 1st phase of LayoutNG's multicol implementation we want
to utilize the existing ColumnBalancer class. That's why multicol blocks
need to processed by Legacy Layout engine.

This patch also fixes some DCHECK(NOT_IMPLEMENTED) crashes.

BUG=635619

Review-Url: https://codereview.chromium.org/2803093003
Cr-Commit-Position: refs/heads/master@{#462677}

Currently does font size and line height.

Color is next. And there is still plumbing required to ensure more
controls/dialogs have access to views::style (and stop using
SetFontList). This CL establishes a core requirement to let us iterate
on layout for specific dialogs.

Example screenshots at https://crbug.com/691891#c10

BUG=691891

Review-Url: https://codereview.chromium.org/2765883004
Cr-Commit-Position: refs/heads/master@{#462676}

This is so all users must confirm sign in via "OK, GOT IT"
instead of clicking "Settings" link (which closes the dialog
without either of the "UNDO" or "OK, GOT IT" buttons being
clicked).

R=rogerta@chromium.org
BUG=709113
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2800683008
Cr-Commit-Position: refs/heads/master@{#462675}

BUG=703975
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2801453002
Cr-Commit-Position: refs/heads/master@{#462674}

BUG=669108

Review-Url: https://codereview.chromium.org/2802113002
Cr-Commit-Position: refs/heads/master@{#462673}

When the available screen height is small the BottomSheet skips the half
state, so using #onTransitionPeekToHalf() is an unreliable way to make
the bottom nav visible. Use @onSheetOFfsetChanged() instead

BUG=

Review-Url: https://codereview.chromium.org/2800973003
Cr-Commit-Position: refs/heads/master@{#462672}

return a boolean so that synchronous error can be detected.

chromium: Remove QuicChromiumClientSession::ResumeCryptoConnect as well,
which is no longer used.

Merge internal change: 152398936

BUG=704949

Review-Url: https://codereview.chromium.org/2806553004
Cr-Commit-Position: refs/heads/master@{#462671}

https://pdfium.googlesource.com/pdfium.git/+log/deff61fe7368..c758d9dd3f94

$ git log deff61fe7..c758d9dd3 --date=short --no-merges --format='%ad %ae %s'
2017-04-06 dalecurtis Remove stale gn arg to avoid undeclared argument warning.
2017-04-06 npm Cleanup CFX_BitmapComposer
2017-04-06 dsinclair Cleanup nits from previous XML Cl

Created with:
  roll-dep src/third_party/pdfium
BUG=645890

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

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls

TBR=dsinclair@chromium.org

Review-Url: https://codereview.chromium.org/2798063005
Cr-Commit-Position: refs/heads/master@{#462670}

Adds a new dialog class based on our current UX mocks. The intention
is to iterate with the UX team to come up with a UI for the Chrome
Cleanup tool.

Some functionality is currently missing and will be added in
subsequent CLs:
- Checkbox to ask for logs upload permissions
- Animation when the details section is expanded/folded

BUG=690020

Review-Url: https://codereview.chromium.org/2795133002
Cr-Commit-Position: refs/heads/master@{#462669}

BleConnectionManager: Check when receiving BLE advertisements if a SecureChannel already exists. Not doing so results in a crash.

BUG=706640

Review-Url: https://codereview.chromium.org/2804123002
Cr-Commit-Position: refs/heads/master@{#462668}