GitLab

I miscommitted r170532 with unnecessary changes.
(r170532 was a sum of https://codereview.chromium.org/198673002 and https://codereview.chromium.org/216203006/,
but I should land only https://codereview.chromium.org/198673002.)
I'll reland those CLs separately.

> Rename DatabaseTaskSynchronizer to TaskSynchronizer and move it to platform/
> 
> - We want to use DatabaseTaskSynchronizer in the HTML parser thread in a follow-up CL. In preparation for that, this CL renames DatabaseTaskSynchronizer to TaskSynchronizer and moves it to platform/.
> 
> BUG=340522
> R=ager@chromium.org, tkent@chromium.org, wibling@chromium.org
> 
> Review URL: https://codereview.chromium.org/198673002

TBR=haraken@chromium.org

Review URL: https://codereview.chromium.org/213673004

git-svn-id: svn://svn.chromium.org/blink/trunk@170534 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=354445

Review URL: https://codereview.chromium.org/206503004

git-svn-id: svn://svn.chromium.org/blink/trunk@170533 bbb929c8-8fbe-4397-9dbb-9b2b20218538

- We want to use DatabaseTaskSynchronizer in the HTML parser thread in a follow-up CL. In preparation for that, this CL renames DatabaseTaskSynchronizer to TaskSynchronizer and moves it to platform/.

BUG=340522
R=ager@chromium.org, tkent@chromium.org, wibling@chromium.org

Review URL: https://codereview.chromium.org/198673002

git-svn-id: svn://svn.chromium.org/blink/trunk@170532 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL just copies a file just to make the diff of
https://codereview.chromium.org/198673002/ saner.

TBR=tkent
NOTRY=true

Review URL: https://codereview.chromium.org/220703002

git-svn-id: svn://svn.chromium.org/blink/trunk@170531 bbb929c8-8fbe-4397-9dbb-9b2b20218538

I was looking at how to port the BisonCSSParser and CSSParserObserver to the oilpan (GC) heap and noticed that we had a pointer from the BisonCSSParser back to the stack allocated CSSParserObserver/StyleSheetHandler. While it worked correctly making the BisonCSSParser stack allocated will IMO simplify the lifetime reasoning and make it less likely that we introduce errors.
This comes at the "expense" of always allocating a comment parser in the StyleSheetHandler. It is only used in the Inspector code so I am hoping that is okay.

R=ager@chromium.org, erik.corry@gmail.com, eseidel@chromium.org, haraken@chromium.org, oilpan-reviews@chromium.org, tkent@chromium.org, vegorov@chromium.org, zerny@chromium.org
BUG=

Review URL: https://codereview.chromium.org/216433003

git-svn-id: svn://svn.chromium.org/blink/trunk@170530 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=none

Review URL: https://codereview.chromium.org/211093005

git-svn-id: svn://svn.chromium.org/blink/trunk@170529 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=vegorov@chromium.org
BUG=356118

Review URL: https://codereview.chromium.org/218683007

git-svn-id: svn://svn.chromium.org/blink/trunk@170528 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Implements transform as defined in CSS Transforms[1]. Behavior is similar
to -webkit-transform, but the perspective() function is not allowed to
contain numbers without units. Like -webkit-transform and other user
agents, the resolved value of transform is not the computed value as 
currently defined by css-transitions[2]. 

The animation system is modified to avoid running animations and 
transitions on both the prefixed and unprefixed variants at the same 
time. Internally, animations are mapped to the unprefixed variant. 
As animations operate directly on the RenderStyle storage the result 
of specifying an animation/transition on either property is the same. 

Note, this property is guarded by the CSSTransformsUnprefixed runtime 
feature (enabled only for testing). 

[1] http://www.w3.org/TR/css3-transforms/
[2] http://lists.w3.org/Archives/Public/www-style/2013Dec/0317.html 

BUG=154772

Review URL: https://codereview.chromium.org/212543003

git-svn-id: svn://svn.chromium.org/blink/trunk@170527 bbb929c8-8fbe-4397-9dbb-9b2b20218538

We missed the compile assert in the VectorDestructor's clear method when having vectors where elements were only added.
This should fix that. It also caught a few places where we had to define a new trait.

R=ager@chromium.org, erik.corry@gmail.com, haraken@chromium.org, oilpan-reviews@chromium.org, tkent@chromium.org, vegorov@chromium.org, zerny@chromium.org
BUG=

Review URL: https://codereview.chromium.org/219243003

git-svn-id: svn://svn.chromium.org/blink/trunk@170526 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This avoids a "deprecated directive, use ‘%pure-parser’ [-Wdeprecated]"
warning in css/CSSGrammar.y and xml/XPathGrammar.y when building with
Bison 3.

Review URL: https://codereview.chromium.org/217423011

git-svn-id: svn://svn.chromium.org/blink/trunk@170525 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The method's behavior is very similar to element.scrollIntoView().
This is just a another version about path on canvas.
('path' means both current default path and path object)
This is especially useful on devices with small screens,
where the whole canvas might not be visible at once.

Intent to Implement:
  https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/KLFSBeP42VQ

The spec reference is:
  http://www.w3.org/TR/2dcontext/#dom-context-2d-scrollpathintoview
  http://www.whatwg.org/specs/web-apps/current-work/#dom-context-2d-scrollpathintoview

BUG=351696

Review URL: https://codereview.chromium.org/196243007

git-svn-id: svn://svn.chromium.org/blink/trunk@170524 bbb929c8-8fbe-4397-9dbb-9b2b20218538

For a class this meager, it was a bit of a heavy-weight:

SVGPathStringBuilder::arcTo() clocked in at a whopping 7891 bytes
(Linux, x86-64, gcc-4.8.2), but even the much simpler lineToVertical()
used > .5k of text-space.
Sacrifice the StringOperators-based pattern and use a cascade of functions
instead.
Besides reducing footprint, this also eliminates some of the copying that
would occur between various temporary String objects.
With the change in place, arcTo() dropped to 640 bytes (and lineToVertical
got cut in half.)

Review URL: https://codereview.chromium.org/212603003

git-svn-id: svn://svn.chromium.org/blink/trunk@170523 bbb929c8-8fbe-4397-9dbb-9b2b20218538

These chicken/egg hacks appear to only be needed for scroll parts. This CL
moves them into RenderLayerScrollableArea so that we don't need to apply them
constantly for RenderLayers that aren't scrollable.

R=ojan@chromium.org

Review URL: https://codereview.chromium.org/220473002

git-svn-id: svn://svn.chromium.org/blink/trunk@170522 bbb929c8-8fbe-4397-9dbb-9b2b20218538

content_shell can not decode mov, so this CL makes canvas tests use mp4 file.
the desktop Perf bots (Windows, Mac, and Linux) do actually build Chrome-branded
binaries including all of the media codecs so H.264 encoded mp4 file can be
decoded.

bear-1280x720.mp4 is copied from chromium; src/media/test/data
Canvas video perf tests run on only the platform which can decode mp4 video.

BUG=242244, 358168

Review URL: https://codereview.chromium.org/219023009

git-svn-id: svn://svn.chromium.org/blink/trunk@170521 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Blink doent respect the form.submit() when called on focus event on invalid form.
    When an form submit button is clicked and if the form is invalid it will focus
    the field that shows error. If the user is calling the form.submit() on focus event
    on that field then it should submit the form. But currently the form does not get
    submitted. This patch removes this bug by completely removing the variable
    "m_isSubmittingOrPreparingForSubmission".

in API "submit" which is always called after API "prepareForSubmission". Also the variable can more appropriately be renamed to "m_isSubmitting".

Bug=357101

Review URL: https://codereview.chromium.org/199633017

git-svn-id: svn://svn.chromium.org/blink/trunk@170520 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Re-landing r170008, removing prefixed Shadow DOM APIs in Element. (https://codereview.chromium.org/218943003/)

Reason for revert:
Broke OpenAudioFiles/FileManagerBrowserTest again.

http://build.chromium.org/p/chromium.chromiumos/builders/Linux%20ChromiumOS%20Tests%20%282%29/builds/22744


Original issue's description:
> Re-landing r170008, removing prefixed Shadow DOM APIs in Element.
> 
> The previous attempt, https://codereview.chromium.org/208503004, was reverted in https://codereview.chromium.org/212303005/.
> The reason of reverting is that there is a JavaScript code uses prefixed Shadow DOM API in the chromium side.
> I've fixed that in https://codereview.chromium.org/214093002/.
> 
> Let me re-land r170008. I hope it won't break anything this time.
> 
> BUG=242153
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=170460

TBR=dglazkov@chromium.org,esprehn@chromium.org,hayato@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=242153

Review URL: https://codereview.chromium.org/220123005

git-svn-id: svn://svn.chromium.org/blink/trunk@170519 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Finish events occur when a sample occurs that causes a player to be newly limited or newly disassociated with a timeline. 
(From section 3.5.4 Player events 
http://dev.w3.org/fxtf/web-animations/#player-events
and section 5.19 The AnimationPlayerEvent interface
http://dev.w3.org/fxtf/web-animations/#the-animationplayerevent-interface )

Subsequent CL 216203003 dispatches the events declared here.

BUG=356120

Review URL: https://codereview.chromium.org/210363007

git-svn-id: svn://svn.chromium.org/blink/trunk@170518 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Implements perspective as defined in CSS Transforms[1]. Behavior is
similar to -webkit-perspective, but numbers without units are not
valid. In both variants we accept zero although Gecko and spec
require a positive value.

The animation system is modified to avoid running animations and
transitions on both the prefixed and unprefixed variants at the same
time. Internally, animations are mapped to the unprefixed variant.
As animations operate directly on the RenderStyle storage the result
of specifying an animation/transition on either property is the same.

Transitions now track the specified transition-property such that
the end event may reference it. (eg. transition:
-webkit-perspective; will fire an end  event with
propertyName == '-webkit-perspective'). As transitions operate on
changes to values within a RenderStyle, 'transition: all;' only
ever starts transitions on the unprefixed variants.

Note, this property is guarded by the CSSTransformsUnprefixed runtime
feature (enabled only for testing).

[1] http://www.w3.org/TR/css3-transforms/

BUG=154772

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=170326

Review URL: https://codereview.chromium.org/211233003

git-svn-id: svn://svn.chromium.org/blink/trunk@170517 bbb929c8-8fbe-4397-9dbb-9b2b20218538

For consistency, have RepeatEvent 'upcall' to Event when tracing.

R=
BUG=340522

Review URL: https://codereview.chromium.org/219283007

git-svn-id: svn://svn.chromium.org/blink/trunk@170516 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Before this CL, if you changed the style on a RenderLayer that was directly
composited, we would trigger a rebuild of the entire compositing tree as part
of the "chicken-egg hacks." There's no reason for that. Instead, we simply need
to do a geometry update after changing style.

This makes the compositing update in the calculator faster by 7%.

R=ojan@chromium.org

Review URL: https://codereview.chromium.org/220193003

git-svn-id: svn://svn.chromium.org/blink/trunk@170515 bbb929c8-8fbe-4397-9dbb-9b2b20218538

We have similar comments for M33 and M34.

NOTRY=true
BUG=

Review URL: https://codereview.chromium.org/220513002

git-svn-id: svn://svn.chromium.org/blink/trunk@170514 bbb929c8-8fbe-4397-9dbb-9b2b20218538

AnimationTimeline has currentTime of type double - the time value for
this timeline.

Web Animations spec, section 5.1 The AnimationTimeline interface
http://dev.w3.org/fxtf/web-animations/#the-animationtimeline-interface


BUG=357432

Review URL: https://codereview.chromium.org/215043007

git-svn-id: svn://svn.chromium.org/blink/trunk@170513 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The currentTime() and duration() should simply use the media element, as
that is what paintMediaSlider() in RenderMediaControls uses, so the
existing code would not work if the media element and its controller had
different durations.

setCurrentTime() does require taking the controller into account, so add
some tests and expand that final instance of mediaControllerInterface().

BUG=341813, 341852
TEST=LayoutTests/media/controls-timeline.html
     LayoutTests/media/controls-timeline-with-controller.html

Review URL: https://codereview.chromium.org/219283004

git-svn-id: svn://svn.chromium.org/blink/trunk@170512 bbb929c8-8fbe-4397-9dbb-9b2b20218538

FocusController::dispatchEventsOnWindowAndFocusedNode now dispatches setFocus(false) when window loses focus and setFocus(true) when window focus is back. 

R=tkent, keishi1
BUG=102830
TEST=Window focus is lost and text is updated by user, dispatch both blur and change event

Review URL: https://codereview.chromium.org/205033004

git-svn-id: svn://svn.chromium.org/blink/trunk@170511 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=332630

Review URL: https://codereview.chromium.org/219083002

git-svn-id: svn://svn.chromium.org/blink/trunk@170510 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL merges the two update type fields in RenderLayer compositor. They were
both doing the same thing. In the future, we want to set more of the
RenderLayerCompositor dirty bits from the new m_updateType field, but this is a
start.

Also, remove the mutable keyword from a field where it wasn't needed.

Review URL: https://codereview.chromium.org/213833014

git-svn-id: svn://svn.chromium.org/blink/trunk@170509 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Revert of Put the squashing layer underneath the ancestor clipping layer if present. (https://codereview.chromium.org/218313002/)

Reason for revert:
The breakage is only with --enable-universal-accelerated-overflow-scroll true. Since that feature is not launched and not tied to squashing, going to re-lad the original CL.

Original issue's description:
> Revert of Put the squashing layer underneath the ancestor clipping layer if present. (https://codereview.chromium.org/215823002/)
> 
> Reason for revert:
> This appears to have unexpected effects on rendering. Bits of gmail are missing for example.
> 
> Original issue's description:
> > Put the squashing layer underneath the ancestor clipping layer if present.
> > 
> > Previously, this was the case for m_graphicsLayer but not m_squashingLayer.
> > Since they now are guaranteed to have the same clipping ancestor, this is both
> > safe and correct.
> > 
> > BUG=336563
> > 
> > Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=170348
> 
> TBR=hartmanng@chromium.org,chrishtr@chromium.org
> NOTREECHECKS=true
> NOTRY=true
> BUG=336563
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=170413

TBR=hartmanng@chromium.org,vollick@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=336563

Review URL: https://codereview.chromium.org/219223008

git-svn-id: svn://svn.chromium.org/blink/trunk@170507 bbb929c8-8fbe-4397-9dbb-9b2b20218538

According to specification, onaudioprocess() evenhandler should delever playbackTime attribute to JS.
The time when the audio will be played in the same time coordinate system as AudioContext.currentTime.
playbackTime allows for very tight synchronization between processing directly in JavaScript
with the other events in the context's rendering graph.

BUG=332782

Review URL: https://codereview.chromium.org/210973002

git-svn-id: svn://svn.chromium.org/blink/trunk@170505 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=358226

Review URL: https://codereview.chromium.org/219523004

git-svn-id: svn://svn.chromium.org/blink/trunk@170504 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This reverts: https://src.chromium.org/viewvc/blink?revision=170491&view=revision

Failed attempt to land two-sided patch. The chromium side broke the blink build and then this patch further broke it after the chromium side was reverted.

In the future, please do a three-step commit if that's required to ensure no build breakage.

TBR=a.renevier@samsung.com
BUG=348071

Review URL: https://codereview.chromium.org/220503002

git-svn-id: svn://svn.chromium.org/blink/trunk@170503 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This fixes two possible crash:
1. document->frame() could be null when ServiceWorkerContainerClient::from is called
2. NavigatorServiceWorker.m_serviceWorker could be null in ServiceWorkerContainer::detachClient

BUG=355998,356017
TEST=fast/serviceworker/access-container-with-invalid-context.html

Review URL: https://codereview.chromium.org/217023003

git-svn-id: svn://svn.chromium.org/blink/trunk@170501 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Instead of having one scheduling function per callback type, we can use
a single shared one except for the attribute callback case.

Review URL: https://codereview.chromium.org/220013003

git-svn-id: svn://svn.chromium.org/blink/trunk@170500 bbb929c8-8fbe-4397-9dbb-9b2b20218538

NotifyPositionChange pass on stored m_currentPos, and didn't use its
input variable. Removing its input variable from signature to reflect
this.

BUG=

Review URL: https://codereview.chromium.org/212063004

git-svn-id: svn://svn.chromium.org/blink/trunk@170499 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Styles generated by animations and transitions are currently applied
following adjustment (to avoid triggering transitions on adjustment).

Special handling existed for mirroring the adjustment behavior of
auto z-index when animating transform, however this failed to account
for animations of perspective. This has been broken as far back as
I am able to bisect and remains broken in WebKit for the same reason.

Review URL: https://codereview.chromium.org/219013002

git-svn-id: svn://svn.chromium.org/blink/trunk@170498 bbb929c8-8fbe-4397-9dbb-9b2b20218538

There is an object LineFragmentationData that is used for pagination
that is created on demand. That object was created for normal documents
and cost roughly 32 bytes per line, adding up to 3-5 MB on a big text
document (5% of all memory usage).

Testcase: http://norvig.com/big.txt - 3-10 MB RAM savings.

I don't know what I'm doing but this patch prevents the allocation and
saves the memory.

BUG=353993

Review URL: https://codereview.chromium.org/200053007

git-svn-id: svn://svn.chromium.org/blink/trunk@170497 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Rename LayoutBox to CSSBoxType. Also rename the functions and methods that mention "LayoutBox" to something that makes sense for each specific situation.

This is a merge of http://trac.webkit.org/changeset/165843 by Bem Jones-Bey <bjonesbe@adobe.com>.

BUG=358377

Review URL: https://codereview.chromium.org/220023003

git-svn-id: svn://svn.chromium.org/blink/trunk@170496 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=357707

Review URL: https://codereview.chromium.org/217053009

git-svn-id: svn://svn.chromium.org/blink/trunk@170495 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This reverts commit d1eea2e66e64c3c019885ad7e94933aa0fc05576.
On top of that, this CL makes sure that only the preloader uses the thread safe MQ parser. That's in order to reduce the risk of breaking content when introducing the new MQ parser.

BUG=358078

Review URL: https://codereview.chromium.org/219183002

git-svn-id: svn://svn.chromium.org/blink/trunk@170494 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=ojan
BUG=

Review URL: https://codereview.chromium.org/216443004

git-svn-id: svn://svn.chromium.org/blink/trunk@170493 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=357782
TEST=no change

Review URL: https://codereview.chromium.org/219163005

git-svn-id: svn://svn.chromium.org/blink/trunk@170492 bbb929c8-8fbe-4397-9dbb-9b2b20218538