GitLab

R=vsevik
BUG=355459
NOTRY=true

Review URL: https://codereview.chromium.org/199633018

git-svn-id: svn://svn.chromium.org/blink/trunk@170649 bbb929c8-8fbe-4397-9dbb-9b2b20218538

GC is performed at the next return to the event loop.

Use this in svg animation tests to be able to reliably check that
objects die when expected.

R=erik.corry@gmail.com, haraken@chromium.org, kouhei@chromium.org, oilpan-reviews@chromium.org, vegorov@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/220203005

git-svn-id: svn://svn.chromium.org/blink/trunk@170648 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://dom.spec.whatwg.org/#dom-document-contenttype

BUG=279426

Review URL: https://codereview.chromium.org/151653004

git-svn-id: svn://svn.chromium.org/blink/trunk@170646 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The function should return an empty string in case no source has been
defined, compileShader() has not been called, or the translation has
failed for shader. It used to return null, after this patch it returns an
empty string like it is supposed to.

Also make getTranslatedShaderSourceANGLE pure virtual in
WebGraphicsContext3D.

BUG=358530
TEST=WebGL conformance tests with --enable-privileged-webgl-extensions
R=kbr@chromium.org

Review URL: https://codereview.chromium.org/216603004

git-svn-id: svn://svn.chromium.org/blink/trunk@170645 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch sets correct index to the multiple selections created with
ctrl-d.

Review URL: https://codereview.chromium.org/220203007

git-svn-id: svn://svn.chromium.org/blink/trunk@170643 bbb929c8-8fbe-4397-9dbb-9b2b20218538

unlockPixels is always called in the ImageExtractor destructor if
m_skiaImage is non-null. Clear m_skiaImage if extractImage fails and
lockPixels is never called so that lockPixels and unlockPixels calls
will be in balance.

Also optimize webgl-large-texture.html test, since some of the test
timeouts could be caused by the test running too long. Optimizations
include avoiding conversion between RGBA and RGB, checking the pixels in
32 bit chunks instead of 8 bits at a time, and loading the image from a
file instead of generating a JPEG data URL dynamically.

TEST=fast/canvas/webgl/webgl-large-texture.html
BUG=358313
R=kbr@chromium.org, skyostil@chromium.org

Review URL: https://codereview.chromium.org/221083002

git-svn-id: svn://svn.chromium.org/blink/trunk@170642 bbb929c8-8fbe-4397-9dbb-9b2b20218538

NavigatorServiceWorker's use of CallWith=ExecutionContext meant the
container was configured depending on how it was accessed: for
example, navigator.serviceWorker would associate it with the frame's
execution context; but iframe.contentWindow.navigator.serviceWorker,
window.opener.navigator.serviceWorker,
window.parent.navigator.serviceWorker, etc. would associate it with a
different execution context. This change always retrieves the
execution context from the frame associated with the Navigator object.

Separately, the ServiceWorkerContainerClient was a supplement to the
page, which meant a main frame and an iframe could end up sharing a
ServiceWorkerContainerClient, violating the assumption of a 1:1
relationship between provider and client. This change associates it
with the document.

BUG=358979

Review URL: https://codereview.chromium.org/213223005

git-svn-id: svn://svn.chromium.org/blink/trunk@170641 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch simplifies Selection.addRange()'s implementation by expanding
the selection based on DOM Ranges instead of VisiblePositions.

There is a slight behavior change caused by this shift of data structures
we use: ranges which are discrete in the DOM tree but visually adjacent will
be rejected after this patch is landed. However, I don't think this will be
a huge problem, since (1) API behavior relying on visual positions can be
tricky to users because they are Blink-internal only information; and
(2) users wishing to connect the regions can (and probably should) specify
a range which overlaps with the current selection.

BUG=346613

Review URL: https://codereview.chromium.org/205903003

git-svn-id: svn://svn.chromium.org/blink/trunk@170640 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Patches implementing all of the properties from CSS Transforms[1] have now
landed.

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

BUG=154772

Review URL: https://codereview.chromium.org/221413004

git-svn-id: svn://svn.chromium.org/blink/trunk@170639 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Inline renderers have preferred logical widths that end up depending on
the autosizing font scale. This patch forces these preferred widths to
be recomputed. This recomputation is not necessary for text in blocks
because the block layout algorithm does not depend on the text width.
The parent null check (renderer->parent()...) is for list markers
which, interestingly, can not be in the render tree.

This patch fixes rendering on many sites, including hacker news.

BUG=338494

Review URL: https://codereview.chromium.org/218863002

git-svn-id: svn://svn.chromium.org/blink/trunk@170638 bbb929c8-8fbe-4397-9dbb-9b2b20218538

:any in selector components to the right of the rightmost
descendant relation.

This fixes a bug that caused this to be the case accidentally in cases such as:

:host(.foo) #bar

when #bar is not (yet) supported for the style invalidation
tree walk.

BUG=358704

Review URL: https://codereview.chromium.org/221643002

git-svn-id: svn://svn.chromium.org/blink/trunk@170637 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Make sure to measure the same input string as we're going to parse.

BUG=358667,352544

Review URL: https://codereview.chromium.org/220113006

git-svn-id: svn://svn.chromium.org/blink/trunk@170636 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=358964
TBR=leviw

Review URL: https://codereview.chromium.org/213833023

git-svn-id: svn://svn.chromium.org/blink/trunk@170635 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://src.chromium.org/viewvc/blink?view=revision&revision=170623

BUG=58719
TBR=ziran.sun@samsung.com

Review URL: https://codereview.chromium.org/221763003

git-svn-id: svn://svn.chromium.org/blink/trunk@170634 bbb929c8-8fbe-4397-9dbb-9b2b20218538

According to the specification, parenthesis, brackets and braces blocks inside media queries should be observed and ignored as a malformed media query.
This CL fixes the previous behavior that didn't take that into account.

BUG=358041

Review URL: https://codereview.chromium.org/217423005

git-svn-id: svn://svn.chromium.org/blink/trunk@170633 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This function doesn't do much anymore and can be inlined into its one caller.
This CL also removes some unnecessary bits in the UpdateLayerPositionsFlags
bitfield.

Review URL: https://codereview.chromium.org/220593008

git-svn-id: svn://svn.chromium.org/blink/trunk@170632 bbb929c8-8fbe-4397-9dbb-9b2b20218538

All the types of updates set this bit now. We can just trigger the geometry
update off the updateType enum.

Review URL: https://codereview.chromium.org/221563002

git-svn-id: svn://svn.chromium.org/blink/trunk@170631 bbb929c8-8fbe-4397-9dbb-9b2b20218538

After this CL, we don't update the graphics layer geometry incrementally.
Instead, the updates are deferred until its actually time to do a compositing
update. There's a lot of cleanup we can do after this CL lands, but I wanted to
keep this CL small. We can do the cleanup in subsequent CLs.

Review URL: https://codereview.chromium.org/218993017

git-svn-id: svn://svn.chromium.org/blink/trunk@170630 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Currently EventInterfaces.in generation is causing build problems
(fail to rebuild) because it does not specify dependencies correctly,
and it is not depended on correctly.
Fixing this is hard because of wonky circular dependency between core/
and bindings/.

This fixes it by moving EventInterfaces.in generation
to a separate file, core_bindings_generated.gyp,
and having a static list of event interfaces,
so EventInterfaces.in generation does not depend on any
other build steps, breaking the cycle.

The build action stays in bindings/,
because it depends on .idl files in modules/,
and we don't want build dependencies of core/ on modules/.

This also does some cleanup of generated_bindings.gyp
and bindings.gypi,
which incidentally moves
ServiceWorkerGlobalScopeConstructors.idl
into gen/blink (where it belongs), not gen/.

(For ease of reviewing it doesn't move the Python file,
but I can move it to build/scripts in a followup.)


== Details ==
The simple fix for EventInterfaces.in generation (adding dependencies)
creates a circular dependency between .gyp files, which causes Mac build bot
to fail. (This isn't a circular dependency between GYP *targets*, so builds
correctly on other platforms),
Breaking the circular dependency by omitting it breaks Android build.

The specific dependency problem is that
core/core_generated.gyp:make_core_generated
...needs to depend on generation of EventInterfaces.in,
...which is currently in
bindings/generated_bindings.gyp:generated_bindings
but this depends on
core/core_generated.gyp:generated_testing_idls
(for generated IDL files),
which creates a cycle of *files*:
core/core_generated.gyp ->
bindings/generated_bindings.gyp ->
core/core_generated.gyp

We break this by having
core/core_generated.gyp:make_core_generated
depend on
bindings/core_bindings_generated.gyp:core_bindings_generated
...which has no dependency targets.

The dependency on bindings/ is because current EventInterfaces.in
generation dynamically computes the list of event interfaces (based on
ancestry: inherits from Event), which depends on the global
'compute_interfaces_info' step in bindings/, which computes ancestry and
a whole lot more.

This can be solved by instead having a static list of event interface IDL
files, which ｂreaks the dependency,
and also avoids unnecessary regenerations (only regen on event
.idl file changes, not any .idl file change) and simplifies the code.


Previous build failure/proposed fix:
Add RuntimeEnabled info to InterfaceInfo.pickle.
https://codereview.chromium.org/208953004/
Quick fix:
Fix dependencies for EventInterfaces.in
https://codereview.chromium.org/209723002/
...reverted in:
Revert of Fix dependencies for EventInterfaces.in
https://codereview.chromium.org/212713002/

R=abarth, haraken
BUG=341748
BUG=358074

Review URL: https://codereview.chromium.org/212983010

git-svn-id: svn://svn.chromium.org/blink/trunk@170628 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This reverts commit r167629 (f3f2415b3b74e995665bea03f4316e28424fceeb)
as it caused a secuirty regression.

BUG=353143
TBR=andersr@opera.com

Review URL: https://codereview.chromium.org/221683005

git-svn-id: svn://svn.chromium.org/blink/trunk@170627 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Android build hits this assertion all the time, rendering debug build useless.

This disabler isn't entirely correct, but assuming blitting-based scrolling
won't last long, it should be okay with a FIXME.

BUG=357345

Review URL: https://codereview.chromium.org/212223007

git-svn-id: svn://svn.chromium.org/blink/trunk@170626 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Previously, it was possible to paint a scroll corner that was larger
than the iframe it was attached to, effectively enabling it to paint
over parts of the parent frame. This change clips scrollbar painting
the same way we clip contents.

BUG=331168

Review URL: https://codereview.chromium.org/220243010

git-svn-id: svn://svn.chromium.org/blink/trunk@170625 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://src.chromium.org/viewvc/blink?view=revision&revision=170591

TBR=jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/220113011

git-svn-id: svn://svn.chromium.org/blink/trunk@170624 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=tkent@chromium.org
BUG=58719

Review URL: https://codereview.chromium.org/216923003

git-svn-id: svn://svn.chromium.org/blink/trunk@170623 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Follows Blink (crrev.com/164873003) and Chromium (crrev.com/164933002)

BUG=342555

Review URL: https://codereview.chromium.org/217133002

git-svn-id: svn://svn.chromium.org/blink/trunk@170621 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Instead of passing a boolean by reference we should just return one.

Review URL: https://codereview.chromium.org/221663002

git-svn-id: svn://svn.chromium.org/blink/trunk@170620 bbb929c8-8fbe-4397-9dbb-9b2b20218538

r170531 copied files to make the diff of r170536 saner.
Now that r170536 was reverted, we need to revert r170531 at the same time.
Otherwise, we'll leave meaningless files in the code base.

TBR=tkent

Review URL: https://codereview.chromium.org/221663003

git-svn-id: svn://svn.chromium.org/blink/trunk@170619 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This is a merge of http://trac.webkit.org/changeset/151623 by Simon Fraser <simon.fraser@apple.com>.

It was a mostly-mechanical merge. The reftest fixed-backgrounds.html required some tweaking,
since it relied on composited and non-composited output being identical. This required changing
the shadow from a soft shadow to a hard shadow.

BUG=330852

--

Source/WebCore: Painting of fixed background images is wrong in composited layers
​https://bugs.webkit.org/show_bug.cgi?id=65793
Reviewed by Sam Weinig.
The code that computed background image geometry for background-attachment:fixed
images was unaware of compositing, so often painting the image at the wrong location.
Fix by having RenderBoxModelObject::calculateBackgroundImageGeometry() do the correct
math for fixed backgrounds in composited layer by offsetting the viewport rect by
the paint container's absolute position.
Tests: compositing/backgrounds/fixed-background-on-descendant.html
compositing/backgrounds/fixed-backgrounds.html
rendering/RenderBox.cpp:
(WebCore::RenderBox::getBackgroundPaintedExtent): Now returns a bool indicating
whether it is returning a reliable extent rect. It can return false in the case where
a background is fixed, since computing the correct extent would require finding
the appropriate composited ancestor to pass to calculateBackgroundImageGeometry().
This is OK since this function is used for "background opaque" optimizations.
(WebCore::RenderBox::computeBackgroundIsKnownToBeObscured): If getBackgroundPaintedExtent()
returns false, return false.
(WebCore::RenderBox::maskClipRect): We removed mask-attachment, so we never need to
compute the composited ancestor here and can pass null.
(WebCore::RenderBox::repaintLayerRectsForImage): Unwrap a comment.
If the changed image is related to a fixed background, geometry.hasNonLocalGeometry()
will be true. In that cause, just repaint the entire renderer rather than groveling
around for a composited ancestor.
rendering/RenderBox.h: Changed name and signature of backgroundPaintedExtent.
rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::paintFillLayerExtended): calculateBackgroundImageGeometry()
now needs to know the painting container.
(WebCore::RenderBoxModelObject::calculateBackgroundImageGeometry): Now takes a painting
container, that is required to correctly compute the viewport-relative offset for fixed
backgrounds. geometry.setHasNonLocalGeometry() is set for fixed backgrounds to indicate
to callers that, if they didn't pass a paint container, the destRect is not accurate.
The main bug fix is also here: we move the viewportRect by the absolute location of
paint container, which is equivalent to the composited layer offset.
(WebCore::RenderBoxModelObject::getGeometryForBackgroundImage): calculateBackgroundImageGeometry()
takes a paint container.
rendering/RenderBoxModelObject.h:
(WebCore::RenderBoxModelObject::BackgroundImageGeometry::BackgroundImageGeometry):
(WebCore::RenderBoxModelObject::BackgroundImageGeometry::setHasNonLocalGeometry):
(WebCore::RenderBoxModelObject::BackgroundImageGeometry::hasNonLocalGeometry):
rendering/RenderImage.cpp:
(WebCore::RenderImage::computeBackgroundIsKnownToBeObscured): If getBackgroundPaintedExtent()
can't cheaply give an accurate answer, return false.
rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateDirectlyCompositedBackgroundImage): Pass the paint container,
which is our own renderer.
LayoutTests: Fixed background images behave strangely with webkit transitions
​https://bugs.webkit.org/show_bug.cgi?id=65793
Reviewed by Sam Weinig.
Ref tests that compare fixed background rendering after a scroll, with and
without compositing, with a couple of layer configurations.
compositing/backgrounds/fixed-background-on-descendant-expected.html: Added.
compositing/backgrounds/fixed-background-on-descendant.html: Added.
compositing/backgrounds/fixed-backgrounds-expected.html: Added.
compositing/backgrounds/fixed-backgrounds.html: Added.

Review URL: https://codereview.chromium.org/102123013

git-svn-id: svn://svn.chromium.org/blink/trunk@170618 bbb929c8-8fbe-4397-9dbb-9b2b20218538

TBR=dpranke@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/219963018

git-svn-id: svn://svn.chromium.org/blink/trunk@170617 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This is a speculative revert to see if causes the WebKit Android N4 bot
to stop crashing.

TBR=haraken@chromium.org, rafaelw@chromium.org, tkent@chromium.org
BUG=

Review URL: https://codereview.chromium.org/219963019

git-svn-id: svn://svn.chromium.org/blink/trunk@170616 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Fix assertion in beginLayout on resize.

BUG=353309

Review URL: https://codereview.chromium.org/221193004

git-svn-id: svn://svn.chromium.org/blink/trunk@170615 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch fixes a bug where <use> elements did not work in author shadow
trees due to a restriction that <use> trees are not expanded inside
shadow trees. This restriction was in place so that we don't nest <use>
shadow trees but instead expand nested <use> elements manually using
the top-level <use> element.

For example:
<defs>
__<use id="a">
____<rect>
__<use id="b" xlink:href="#a">
</defs>
<use id="c" xlink:href="#b">

When creating the "c" shadow tree, "b" is first added:
<use id="c">
__#user-agent-shadow-root
____<use id="b">

And then "b" is "expanded" (see: expandUseElementsInShadowTree):
<use id="c">
__#user-agent-shadow-root
____<g id="b">
______<use id="a">

Finally "a" is "expanded":
<use id="c">
__#user-agent-shadow-root
____<g id="b">
______<g id="a">
________<rect>

The checks preventing <use> tree expansion in shadow trees were
intended to prevent the nested <use> elements (or <symbol>) from
creating nested shadow roots--notice there is only one shadow root
in the above example.

This patch switches the isInShadowTree checks with isInUserAgentShadowTree
which will prevent <use> elements from cloning their targets when nested
inside another <use> shadow tree, but allow <use> elements with author
shadow roots.

BUG=356891

Review URL: https://codereview.chromium.org/216463003

git-svn-id: svn://svn.chromium.org/blink/trunk@170614 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Element::supportsStyleSharing() always returns false for elements that
are hovered, focused or active or are the :target so we don't need to
check those in the SharedStyleFinder.

Review URL: https://codereview.chromium.org/221463004

git-svn-id: svn://svn.chromium.org/blink/trunk@170613 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=357672

Review URL: https://codereview.chromium.org/220473008

git-svn-id: svn://svn.chromium.org/blink/trunk@170612 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This is necessary because a relation() of SubSelector does not in all cases guarantee that tagHistory() is non-NULL.

This happens in particular for this selector:

af|inputtext:focus::content { ... }

BUG=358475

Review URL: https://codereview.chromium.org/218403012

git-svn-id: svn://svn.chromium.org/blink/trunk@170611 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL unpacks the arguments to updateAfterLayout from being a bit field to
being separate Boolean values. As part of that unpacking, I learned that one of
the bits is always set, which let me remove an argument to
updateCompositingDescendantGeometry because it was always the same.

Review URL: https://codereview.chromium.org/220163006

git-svn-id: svn://svn.chromium.org/blink/trunk@170610 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This was used to find frames by xpath, but it has no remaining callers.

BUG=304341
TEST=no change

Review URL: https://codereview.chromium.org/215473002

git-svn-id: svn://svn.chromium.org/blink/trunk@170609 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Change FontCacheSkiaWin to set the minimum font size for anti aliasing
on a per script basis and assign min sizes for Han, Hiragana, Katakana,
Hangul, and Bengali. The current values where chosen by picking the
minimum value for which Wikipedia in said script was legible and will
likely require further tweaking based on user input.

R=dglazkov@chromium.org
BUG=357864

Review URL: https://codereview.chromium.org/221153002

git-svn-id: svn://svn.chromium.org/blink/trunk@170608 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Speculative rollout of r170564 to see if unit_tests and content_unittests start working again.

TBR=haraken
BUG=340522

Review URL: https://codereview.chromium.org/219243014

git-svn-id: svn://svn.chromium.org/blink/trunk@170602 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The pad-gesture-fling layout tests assumes the fling curve will generate a
minimum number of events.  This may not always hold, depending on the
fling curve parameters, the fling curve update frequency and whether the main
thread is slow.  Instead, impose a maximum limit on the number of fling updates
it takes to reach the minimum offset target, and use preventDefault() on the
mousewheel event to prevent the fling from terminating early.

Review URL: https://codereview.chromium.org/219243010

git-svn-id: svn://svn.chromium.org/blink/trunk@170597 bbb929c8-8fbe-4397-9dbb-9b2b20218538