GitLab

R=jochen@chromium.org
BUG=363103

Review URL: https://codereview.chromium.org/237083002

git-svn-id: svn://svn.chromium.org/blink/trunk@171468 bbb929c8-8fbe-4397-9dbb-9b2b20218538

When text is selected inside input element, it should change the
color and background color according to the ::selection pseudo
element.

BUG=167676

Review URL: https://codereview.chromium.org/220343006

git-svn-id: svn://svn.chromium.org/blink/trunk@171467 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch fixes heap-use-after-free error (regression caused by r171246)
in the 'MutableStylePropertySet::setProperty' method. This error turned
up as the 'setProperty' method argument contained a pointer from the vector
buffer which had been previously freed (as we started to use 'Vector::swap'
instead of assignment operator).

BUG=362310

Review URL: https://codereview.chromium.org/234823004

git-svn-id: svn://svn.chromium.org/blink/trunk@171466 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=344583

Review URL: https://codereview.chromium.org/171263004

git-svn-id: svn://svn.chromium.org/blink/trunk@171465 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=332630
R=dpranke@chromium.org

Review URL: https://codereview.chromium.org/230533002

git-svn-id: svn://svn.chromium.org/blink/trunk@171464 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=pfeldman, vsevik
BUG=160510

Review URL: https://codereview.chromium.org/225873004

git-svn-id: svn://svn.chromium.org/blink/trunk@171463 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Currently we have a lot of DOMRequestStates in IndexedDB code. However, the only object where we really need a script state is IDBRequest. Other objects just need v8::Isolate. Thus:

- This CL replaces DOMRequestState in IDBRequest with NewScriptState.
- This CL replaces other DOMRequestState with v8::Isolate.

BUG=357144

Review URL: https://codereview.chromium.org/236773003

git-svn-id: svn://svn.chromium.org/blink/trunk@171462 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/236973002

Review URL: https://codereview.chromium.org/236973002

git-svn-id: svn://svn.chromium.org/blink/trunk@171461 bbb929c8-8fbe-4397-9dbb-9b2b20218538

we don't do actions that could modify v8 heap so we can report GC events synchronously. Otherwise these GC events would get call stacks of the next event. We will not collect call stacks for gc event at all because it could be triggered even a single small allocation and the call stack will mislead the developer. If someone would like to know what code allocates memory then he needs to use our allocation profiler with allocation stacks.

BUG=361045
R=yurys@chromium.org

Review URL: https://codereview.chromium.org/235043002

git-svn-id: svn://svn.chromium.org/blink/trunk@171460 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=ager@chromium.org, haraken@chromium.org, jochen@chromium.org, mikhail.pozdnyakov@intel.com, vegorov@chromium.org, wibling@chromium.org, zerny@chromium.org
BUG=

Review URL: https://codereview.chromium.org/223373002

git-svn-id: svn://svn.chromium.org/blink/trunk@171459 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=359977

Review URL: https://codereview.chromium.org/235703007

git-svn-id: svn://svn.chromium.org/blink/trunk@171458 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The <select> element does not ignore the <option>/<optgroup> children that have "display:none". Conseqently, these children are displayed on the list. Code has been added to detect the change in display property and also to drop <option>/<optgroup> children whenever required.

BUG=81415

Review URL: https://codereview.chromium.org/189543012

git-svn-id: svn://svn.chromium.org/blink/trunk@171457 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=357096
R=pfeldman@chromium.org

Review URL: https://codereview.chromium.org/236023006

git-svn-id: svn://svn.chromium.org/blink/trunk@171456 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Adds dispatch change event at touchEnd.

R=tkent@chromium.org, keishi1@chromium.org
BUG=361335
TEST=On touch start check input event is dispatched and on touch end check change event is dispatched

Review URL: https://codereview.chromium.org/234933002

git-svn-id: svn://svn.chromium.org/blink/trunk@171455 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Old version navigator content utils APIs have been kept since r165458. Because
blink bots might ignore dependency with new content module. However, blink bots
is using updated content module which supports new navigator content utils APIs. Thus,
this cl removes the old navigator content utils APIs.

Review URL: https://codereview.chromium.org/232263003

git-svn-id: svn://svn.chromium.org/blink/trunk@171454 bbb929c8-8fbe-4397-9dbb-9b2b20218538

We removed experimental flag -webkit-visual-word 2 years ago:
http://src.chromium.org/viewvc/blink/trunk/Source/core/page/DOMSelection.cpp?r1=89440&r2=102252
Selection.modify("-webkit-visual-word") does nothing.
However, there are still tests using the flag.
They confirm that scripts don't crash when they call Selection.modify("-webkit-visual-word").
It doesn't make sense.

BUG=

Review URL: https://codereview.chromium.org/236733005

git-svn-id: svn://svn.chromium.org/blink/trunk@171453 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/236893002

git-svn-id: svn://svn.chromium.org/blink/trunk@171452 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Calling a navigator method leads to a crash when:
 - The navigator is associated with an iframe, 
 - The iframe has some contents and
 - The iframe is detached.

This CL fixes the crash.

BUG=358800
R=abarth

Review URL: https://codereview.chromium.org/234583004

git-svn-id: svn://svn.chromium.org/blink/trunk@171451 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://heycam.github.io/webidl/#es-operations states that an API returning
a Promise should return a rejected Promise when an error occurs rather than
throwing an exception.

BUG=359386, 362992
R=nbarth@chromium.org, eroman@chromium.org

Review URL: https://codereview.chromium.org/232563003

git-svn-id: svn://svn.chromium.org/blink/trunk@171450 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/236203006

git-svn-id: svn://svn.chromium.org/blink/trunk@171449 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL enables Oilpan for WorkerPerformance by default. This CL doesn't enable Oilpan for SharedWorkerPerformance and WorkerGlobalScopePerformance because they are supplementing off-heap objects.

BUG=340522

Review URL: https://codereview.chromium.org/236613005

git-svn-id: svn://svn.chromium.org/blink/trunk@171448 bbb929c8-8fbe-4397-9dbb-9b2b20218538

- This CL is one of incremental steps to remove DOMRequestState from IDB code base.

- The essence of this CL is to remove DOMRequestState from idbAnyToScriptValue() and idbKeyToScriptValue(). For that goal, this CL passes NewScriptState around in a bunch of places.

- This CL implements [CallWith=NewScriptState] for DOM attributes.

- This CL caches NewScriptState onto ExecutableWithDatabase objects, IndexPopulator and OpenDatabaseCallback.

BUG=357144

Review URL: https://codereview.chromium.org/236783002

git-svn-id: svn://svn.chromium.org/blink/trunk@171447 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Overriding overflow values to VISIBLE if they are different.

BUG=362510

Review URL: https://codereview.chromium.org/235043003

git-svn-id: svn://svn.chromium.org/blink/trunk@171446 bbb929c8-8fbe-4397-9dbb-9b2b20218538

CPUFlameChartDataProvider is made to not depend on CPUProfileView.

Review URL: https://codereview.chromium.org/235453008

git-svn-id: svn://svn.chromium.org/blink/trunk@171445 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/236203009

Review URL: https://codereview.chromium.org/236203009

git-svn-id: svn://svn.chromium.org/blink/trunk@171444 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Some of the StyleBuilderFunctions handlers were being defined in
StyleBuilder.cpp. This patch cleans moves them to the correct place and
makes some minor clean-ups inside (removing the unused condition check
and a redundant isPrimitiveValue check).

Review URL: https://codereview.chromium.org/236763002

git-svn-id: svn://svn.chromium.org/blink/trunk@171443 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Flakiness dashboard link:
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20Blink&tests=http%2Ftests%2Fxmlhttprequest%2Fxmlhttprequest-responseXML-exception.html

TBR=leviw@chromium.org
BUG=362503

Review URL: https://codereview.chromium.org/228663015

git-svn-id: svn://svn.chromium.org/blink/trunk@171442 bbb929c8-8fbe-4397-9dbb-9b2b20218538

HTMLHtmlElement's size might not be the size of the document, such as
HTMLBodyElement's size, so when the selection root in the html element,
the best thing we can do is get the size of the content because <html>
and <body> might have a different size.

BUG=361927

Review URL: https://codereview.chromium.org/236083003

git-svn-id: svn://svn.chromium.org/blink/trunk@171441 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch makes sure we pass |RenderObject| belong to RenderView in
|RenderView::setSelection|, which takes two |RenderObject|s for start and end of
selection, in |FrameSeleciton::updateAppearance|.

The bug is caused by |VisibleSelection::base| and |VisibleSelection::start|
are in different document, |base| points to IFRAME and |start| points |TextNode|
in IFRAME. This causes |RenderView|, which holds |RenderObject|s of selection
start points and end points, have dangling |RenderObject|'s. Because,
|RenderView| doesn't know destructed |RenderObject| belongs to another
|RenderView|.

BUG=356690
TEST=LayoutTests/undo/execCommand/crash-redo-with-iframes.html

Review URL: https://codereview.chromium.org/234463003

git-svn-id: svn://svn.chromium.org/blink/trunk@171440 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/235503010

git-svn-id: svn://svn.chromium.org/blink/trunk@171439 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Since the bindings already check the enumeration, there's no need to
check that the value passed is valid. Make the parseWinding helper function
just return the translated value instead of a status boolean.

Review URL: https://codereview.chromium.org/226733004

git-svn-id: svn://svn.chromium.org/blink/trunk@171438 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Mark virtual/gpu/fast/canvas/canvas-composite-stroke-alpha.html as flaky in WebKit Win Oilpan (dbg), WebKit Win7 (dbg)

This test looks flaky for a long time in [WebKit Win Oilpan (dbg), WebKit Win7 (dbg)].

I can't tell the exact possible revision from the dashboard:
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=virtual%2Fgpu%2Ffast%2Fcanvas%2Fcanvas-composite-stroke-alpha.html

TBR=leviw@chromium.org
BUG=363029

Review URL: https://codereview.chromium.org/236633005

git-svn-id: svn://svn.chromium.org/blink/trunk@171437 bbb929c8-8fbe-4397-9dbb-9b2b20218538

There are some tests calling Selection.collapse(null).
Selection.collapse(null) clears the current selection and its behavior is essentially same as Selection.removeAllRanges().
However, the Selection.collapse method is intended to create a range forming a caret:
https://dvcs.w3.org/hg/editing/raw-file/tip/editing.html#dom-selection-collapse
Calling Selection.collapse(null) to clear the selection is non-standard and I'm going to change Selection.idl to restrict such a usage.

BUG=361941

Review URL: https://codereview.chromium.org/236693002

git-svn-id: svn://svn.chromium.org/blink/trunk@171436 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Since we check all the different states of being a user action element we
should just check the primary bit instead. This function is super hot so
avoiding some branches helps.

This will make us not share for elements in the "active chain", but in the 
common case that's the same as the regular :active elements unless you mutate
the elements in the :active state inside an action (ex. click)

Review URL: https://codereview.chromium.org/236153005

git-svn-id: svn://svn.chromium.org/blink/trunk@171435 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This error was introduced in: https://codereview.chromium.org/220963002

BUG=

Review URL: https://codereview.chromium.org/236083009

git-svn-id: svn://svn.chromium.org/blink/trunk@171434 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=

Review URL: https://codereview.chromium.org/224843014

git-svn-id: svn://svn.chromium.org/blink/trunk@171433 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=90115

Review URL: https://codereview.chromium.org/235503008

git-svn-id: svn://svn.chromium.org/blink/trunk@171432 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Element::parseAttribute does nothing so there's no reason to have a
special case to call it for id, class and style attributes.

Review URL: https://codereview.chromium.org/236363005

git-svn-id: svn://svn.chromium.org/blink/trunk@171431 bbb929c8-8fbe-4397-9dbb-9b2b20218538

TBR=apavlov,pfeldman

Review URL: https://codereview.chromium.org/236613003

git-svn-id: svn://svn.chromium.org/blink/trunk@171430 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This test looks flaky also in linux.

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=http%2Ftests%2Fimages%2Fwebp-partial-load.html

TBR=leviw@chromium.org
BUG=313846

Review URL: https://codereview.chromium.org/236793002

git-svn-id: svn://svn.chromium.org/blink/trunk@171429 bbb929c8-8fbe-4397-9dbb-9b2b20218538