GitLab

BUG=693226
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2733273002
Cr-Commit-Position: refs/heads/master@{#455956}

Disable sina.com.cn as if fails as part of v8.runtime_stats.top_25 on Linux

BUG=699579

Review-Url: https://codereview.chromium.org/2744733002
Cr-Commit-Position: refs/heads/master@{#455955}

In root layer scrolling mode, the main GraphicsLayer in the LayoutView's
CompositedLayerMapping is now connected directly to the VisualViewport scroll
layer (in the main frame), or to the layer of an owning LayoutIFrame.

BUG=698464
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2728273002
Cr-Commit-Position: refs/heads/master@{#455954}

We are adding COMPONENT/TEAM information into OWNERS file
(offline_pages, etc).

Proposal to add TEAM/COMPONENT information into OWNERS files
http://bit.ly/add-team-component-proposal
Proposal about how to get suggested component for directory.
http://bit.ly/directory-mapping-proposal
TEAM-COMPONENT mapping
http://bit.ly/component-team-mapping

Additional Information:
Component lists
https://bugs.chromium.org/p/chromium/adminComponents

BUG=679905
R=dewittj@chromium.org

Review-Url: https://codereview.chromium.org/2744723002
Cr-Commit-Position: refs/heads/master@{#455953}

BUG=681232

Review-Url: https://codereview.chromium.org/2736403005
Cr-Commit-Position: refs/heads/master@{#455952}

DCHECK(m_globalProxy != context()->Global()) sometimes fail. In builds
with DCHECKs turned off this issue manifests with "Type Error: no
access" errors in console and scripts not working.

Started happening approx. in M56.
Repro steps:
1. Install several extensions;
2. Visit a number of pages in a tab;
3. Do a lot of back and forward navigations quickly (not waiting till
pages load).

This is caused by setting the global object in WindowProxy after it was
initialized. This happens when WindowProxyManagerBase::setGlobals
processes a non-main world first. It sends out a didCreateScriptContext
call. ExtensionFrameHelper::DidCreateScriptContext wants to know whether
it is the main world script context, and compares with
WebLocalFrameImpl::mainWorldScriptContext(). As a result, main world
script context is initialized. At this point the call stack is like this:
1  blink::LocalWindowProxy::initialize
2  blink::ScriptController::windowProxy
3  blink::toV8Context
4  blink::ScriptState::forMainWorld
5  blink::WebLocalFrameImpl::mainWorldScriptContext
6  extensions::ExtensionFrameHelper::DidCreateScriptContext
7  content::RenderFrameImpl::didCreateScriptContext
8  blink::LocalWindowProxy::initialize
9  blink::WindowProxy::setGlobal
10 blink::WindowProxyManagerBase::setGlobals
11 blink::WebFrame::swap
12 content::RenderFrameImpl::SwapIn
13 content::RenderFrameImpl::didCommitProvisionalLoad
14 blink::FrameLoaderClientImpl::dispatchDidCommitLoad
15 blink::FrameLoader::receivedFirstData
16 blink::DocumentLoader::createWriterFor
17 blink::DocumentLoader::ensureWriter
18 blink::DocumentLoader::commitData

After that WindowProxyManagerBase::setGlobals continues its work and
sets the global object for the main world WindowProxy, and it becomes
broken.

BUG=700077

Review-Url: https://codereview.chromium.org/2743653002
Cr-Commit-Position: refs/heads/master@{#455951}

BUG=697351

Review-Url: https://codereview.chromium.org/2735293002
Cr-Commit-Position: refs/heads/master@{#455950}

Since many tests depend on just these constants, I think is better to
create a separate file for these since the Model classes might change
again and we don't wan't to keep updating these tests.

This CL removes the model code from ToolsMenuVC and fixes a bug in
ToolsMenuModel

Downstream tests compile and run successfully against these changes.

BUG=682880

Review-Url: https://codereview.chromium.org/2722693002
Cr-Commit-Position: refs/heads/master@{#455949}

Disabling it for ElCapitan does not seem to work, disabling it for all mac platforms instead as the orignal comment suggest.

Filed a catapult isuse at https://github.com/catapult-project/catapult/issues/3359

BUG=657665

Review-Url: https://codereview.chromium.org/2744513004
Cr-Commit-Position: refs/heads/master@{#455948}

Added a new utility function to clarify and simplify checking whether
hardware acceleration is currently being used.

R=emaxx@chromium.org, piman@chromium.org, jbauman@chromium.org
BUG=630728

Review-Url: https://codereview.chromium.org/2737273003
Cr-Commit-Position: refs/heads/master@{#455947}

Bug: internal b/35396333
Test: RPI3
Change-Id: Ic8803d1d8fa0571e9c73b6f294c9b45239ce5a2c
Review-Url: https://codereview.chromium.org/2745653002
Cr-Commit-Position: refs/heads/master@{#455946}

This patch makes |adjustedSelectionStartForStyleComputation()| to take
|Position| instead of |VisibleSeleciton| as a preparation of usage of
|VisibleSeleciton| for improving code health.

BUG=657237
TEST=n/a; no behavior changes

Review-Url: https://codereview.chromium.org/2739023003
Cr-Commit-Position: refs/heads/master@{#455945}

We are adding COMPONENT/TEAM information into OWNERS file
(wft/text).

Proposal to add TEAM/COMPONENT information into OWNERS files
http://bit.ly/add-team-component-proposal
Proposal about how to get suggested component for directory.
http://bit.ly/directory-mapping-proposal
TEAM-COMPONENT mapping
http://bit.ly/component-team-mapping

Additional Information:
Component lists
https://bugs.chromium.org/p/chromium/adminComponents

BUG=679905
R=yutak@chromium.org

Review-Url: https://codereview.chromium.org/2739233002
Cr-Commit-Position: refs/heads/master@{#455944}

BUG=689289

Review-Url: https://codereview.chromium.org/2678233002
Cr-Commit-Position: refs/heads/master@{#455943}

There is a bug that, when a document is attached, the idle time spell
checker is requested for cold mode invocation even without the
IdleTimeSpellChecking flag.

This patch fixes the bug. To ensure that idle time spell checker is
only invoked with the flag, a DCHECK is also added to the handleEvnet()
function.

Related unit tests are disabled when there is no flag.

BUG=700081,517298

Review-Url: https://codereview.chromium.org/2744603004
Cr-Commit-Position: refs/heads/master@{#455942}

https://dom.spec.whatwg.org/#dom-range-surroundcontents
According to the DOM standard, we should not check it before starting DOM
mutation. Step 5 throws HierarchyRequestError instead.

This CL fixes 44 tests in wpt/dom/ranges/Range-surroundContents.html.

* images/element-gcd-while-generating-alt-content.html: Removed.
  This CL changes behavior of the test. The test was for a crash bug in
  pre-Olilpan era, and we don't have code for the crash fix any longer.

* fast/dom/Range/surroundContents-1.html: Removed.
  This is a pixel test to verify HierarchyRequestError which this CL removes.
  This behavior is tested well in WPT.

* shadow-dom/range-surround-contents.html: Updated
  DOM trees are modified because this CL removes a precondition check.

BUG=699861

Review-Url: https://codereview.chromium.org/2745443003
Cr-Commit-Position: refs/heads/master@{#455941}

This patch makes |SpellChecker::clearMisspellingsForMovingParagraphs| to take
|SelectionInDOMTree| instead of |VisibleSelection| as a preparation of reducing
usage of |VisibleSelection| for improving code health.

BUG=657237
TEST=n/a; no behavior changes

Review-Url: https://codereview.chromium.org/2742753002
Cr-Commit-Position: refs/heads/master@{#455940}

https://skia.googlesource.com/skia.git/+log/f1b61afbe971..3304c447b953

$ git log f1b61afbe..3304c447b --date=short --no-merges --format='%ad %ae %s'
2017-03-09 fmalita Revert "Hold sk_sp<GrAtlasTextBlob> refs in GrTextBlobCache instead of raw ptrs"
2017-03-09 halcanary FuzzCanvas: ImageFilter
2017-03-09 reed take fast case in swap() if we're using malloc OR we're empty

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
TBR=djsollen@chromium.org

Change-Id: Iaf67dbbf51d0998c930a406c9aee70ccfdedf790
Reviewed-on: https://chromium-review.googlesource.com/452790Reviewed-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@{#455939}

Utilize EphemeralRange::commonAncestorContainer() in EditingStyleUtilities::createStyleAtSelectionStart()

This patch changes EditingStyleUtilities::createStyleAtSelectionStart()| to use
|EphemeralRange::commonAncestorContainer()| instead of |Range|'s member function
to reduce source code size for improving code health.

BUG=n/a
TEST=n/a; no behavior changes

Review-Url: https://codereview.chromium.org/2739123002
Cr-Commit-Position: refs/heads/master@{#455938}

BUG=697448

Review-Url: https://codereview.chromium.org/2736003003
Cr-Commit-Position: refs/heads/master@{#455937}

Chrome has a direct dependency on a custom libprotobuf_lite in
third_party, but Xenial and Yakkety add a dependency on the system
libprotobuf-lite via Mir (which gets loaded from Gtk).

Our third_party protobuf is not compatible with the upstream protobuf.
Combine this with the fact that Xenial uses version 2 while Yakkety
uses version 3, and it's basically impossible to make our third_party
protobuf cooperate with the system one.  The solution is to always
statically link protobuf on Linux.

BUG=79722,700120
R=pkasting@chromium.org
TBR=bengr@chromium.org
CC=dpranke@chromium.org,rsimha@chromium.org

Review-Url: https://codereview.chromium.org/2746493002
Cr-Commit-Position: refs/heads/master@{#455936}

This change moves the MHTML generation step of writing the footer from
the renderer process to the browser process. Even though this spreads
MHTML code knowledge between Blink and content (a drawback) it does make
the generation process more robust and allows MHTML save operations to
become more robust to DOM changes.

The main issue we are facing with the Offline efforts is that frames can
be deleted after a MHTML save operation started and that will likely
cause it to fail completely. With this modification we can allow these
operations to optionally ignore missing frames.

I chose to still keep the footer code generation in MHTMLArchive but
make it explicit it is only used for testing purposes now. It seemed
more consistent and understandable for those tests to keep on generating
and reading fully valid MHTML files.

BUG=672292
TBR=jam@chromium.org

Review-Url: https://codereview.chromium.org/2731293004
Cr-Commit-Position: refs/heads/master@{#455935}

This patch gets rid of include-directive for "VisibleSelection.h" from
"EditCommand.h" to reduce source code size and faster compilation for
improving code health.

BUG=n/a
TEST=n/a; no behavior changes

Review-Url: https://codereview.chromium.org/2744583002
Cr-Commit-Position: refs/heads/master@{#455934}

BUG=internal b/36023419
TEST=None

Review-Url: https://codereview.chromium.org/2745613003
Cr-Commit-Position: refs/heads/master@{#455933}

BUG=454984

Review-Url: https://codereview.chromium.org/2742783003
Cr-Commit-Position: refs/heads/master@{#455932}

The error messages returned by the compiler are hard to understand and
most singleton implementations live outside base/.

Also adds an example for leaky singletons because it's not clear that
DefaultSingletonTraits must still be friended.

Review-Url: https://codereview.chromium.org/2741493002
Cr-Commit-Position: refs/heads/master@{#455931}

This fixes
ClientOnPageFinishedTest_testCalledOnCancelingProvisionalLoad
ClientOnPageFinishedTest_testCalledOnCancelingProvisionalLoad_with__--webview-sandboxed-renderer_
ClientOnPageFinishedTest_testNotCalledOnDomModificationWithJavascriptUrlAfterNonCommittedLoadFromApi
ClientOnPageFinishedTest_testNotCalledOnDomModificationWithJavascriptUrlAfterNonCommittedLoadFromApi_with__--webview-sandboxed-renderer_
ClientOnReceivedError2Test#testNotCalledOnStopLoading
ClientOnReceivedError2Test#testNotCalledOnStopLoading with {--webview-sandboxed-renderer}
with PlzNavigate.

BUG=645983
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2742853002
Cr-Commit-Position: refs/heads/master@{#455930}

Changes characteristic.getDescriptor(s) to return a more actionable error.

BUG=684445

Review-Url: https://codereview.chromium.org/2728813002
Cr-Commit-Position: refs/heads/master@{#455929}

RendererCompositorFrameSink, DelegatedFrameHost will evict the renderer's
surfaces without the renderer knowing when we are under memory pressure.
This means that the renderer might continue sending frames using the same
local surface id that is destroyed, and we need to make sure we can handle
this situation properly.

1) If the renderer's surface is in the garbage collector's queue, we just
take it out of there and use it again.

2) If the renderer's surface is actually garbage collected, we create the
surface again.

I have added unit tests and also tested this CL together with
https://codereview.chromium.org/2731793002/. Without this CL, if the max
number of saved frames is 1 and we switch between two tabs very fast, a
DCHECK fails in SurfaceManager saying that the surface we are trying to
create already exists, but using this CL this issue doesn't happen.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2736053004
Cr-Commit-Position: refs/heads/master@{#455928}

This is a follow-up CL of crrev.com/2718963002. We don't need this.

BUG=696844

Review-Url: https://codereview.chromium.org/2739543002
Cr-Commit-Position: refs/heads/master@{#455927}

Adds a DOMTokenList backed controlsList/controlslist attribute to HTMLMediaElement with three keywords: nodownload, nofullscreen and noremoteplayback.

Spec change is discussed here: https://github.com/whatwg/html/issues/2293
Spec change PR is here: https://github.com/whatwg/html/pull/2426
WICG repo for the API is here: https://github.com/WICG/controls-list
Intent to ship is here: https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/tFuQd3AcsIQ/discussion

BUG=650174,685018
TEST=manual+layout tests

Review-Url: https://codereview.chromium.org/2657723002
Cr-Commit-Position: refs/heads/master@{#455926}

This patch prepares to remove dependency on //mojo/edk/js to run
js-bindings unittets.

BUG=647036

Review-Url: https://codereview.chromium.org/2743623003
Cr-Commit-Position: refs/heads/master@{#455925}

This is the chromium side version of https://codereview.chromium.org/2736903005

Review-Url: https://codereview.chromium.org/2741533003
Cr-Commit-Position: refs/heads/master@{#455924}

Since all clients of CRWSessionController have been updated to use
NavigationItems, it is now safe to remove CRWSessionEntry.  This CL
updates CRWSessionController implementation to directly own
NavigationItems rather than owning them transitively through
CRWSessionEntry.

BUG=454984

Review-Url: https://codereview.chromium.org/2737203002
Cr-Commit-Position: refs/heads/master@{#455923}

BUG=700207
TBR=zqzhang
NOTRY=true

Review-Url: https://codereview.chromium.org/2738343002
Cr-Commit-Position: refs/heads/master@{#455922}

ConnectionParams will be augmented in a follow up patch to take extra
objects on Android (channels to support sending Parcelable instances).

BUG=699311
TBR=jam@chromium.org,sky@chromium.org,lhchavez@chromium.org,bradnelson@chromium.org

Review-Url: https://codereview.chromium.org/2738853002
Cr-Commit-Position: refs/heads/master@{#455921}

BUG=none

Review-Url: https://codereview.chromium.org/2738333002
Cr-Commit-Position: refs/heads/master@{#455920}

This will eventually replace the existing preferences service
implementation.

High-level overview:

PrefStoreImpl: Wraps a PrefStore to expose its Observer interface as a Mojo
interface.

PrefStoreClient: Implements the PrefStore interface and acts as a cache,
providing synchronous access to prefs stored in a PrefStoreImpl.

PrefStoreManagerImpl: Facilitates connections between PrefStoreClients and
PrefStoreImpls.

PrefStoreAdapter: Helper to tie the lifetime of the PrefStoreImpl
to the underlying PrefStore (which is indirectly held by the
PrefService).

BUG=654988

Review-Url: https://codereview.chromium.org/2635153002
Cr-Commit-Position: refs/heads/master@{#455919}

This method only resets the base score for a URL. It does not affect the
bonus score that an origin might receive, e.g. from being installed to
the user's desktop or homescreen. This means that getting the score of
an origin immediately after resetting it might return an unexpected
number.

This CL renames the method and adds a comment explicitly calling out the
potential difference.

BUG=699381
TBR=bauerb@chromium.org
TBR=dbeam@chromium.org

Review-Url: https://codereview.chromium.org/2735333002
Cr-Commit-Position: refs/heads/master@{#455918}

After an observation that we don't actually need an rtree or a scale
to gather discardable images (we process every item at scale 1.f), we
decided that it's best to replicate a few lines of the raster code at
the benefit of faster implementation and less peak memory.

This patch does this. This path is already tested by cc_unittests.

R=danakj@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2743793002
Cr-Commit-Position: refs/heads/master@{#455917}