GitLab

Now that ICU was rolled to ICU 52, two layout tests marked as failure
pass beause the expectation was updated in advance to prepare for ICU 52 update.

Likewise, a test for shortStandAloneMonth names in ru_RU should be
re-enabled.

BUG=378919
TEST=1.LayoutTest: fast/forms/datetimelocal-multiple-fields/datetimelocal-multiple-fields-*
2. webkit_unit_tests: LocaleICU*
TBR=tkent@chromium.org

Review URL: https://codereview.chromium.org/403393004

git-svn-id: svn://svn.chromium.org/blink/trunk@178574 bbb929c8-8fbe-4397-9dbb-9b2b20218538

DevTools: Remove calls of WebInspector.timelineManager and WebInspector.cpuProfilerModel from ScreencastView

BUG=

Review URL: https://codereview.chromium.org/408853003

git-svn-id: svn://svn.chromium.org/blink/trunk@178572 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Rather than using `Resource::Type` to make decisions about what CSP
directive to apply to a resource request, we should be using
WebURLRequest::RequestContext. This patch teaches ContentSecurityPolicy
to correctly compare a URL against the right policy directive when given
a 'ResourceContext', and pipes the value through
'ResourceFetcher::canRequest' to make sure its available when it needs
to be.

BUG=390497

Review URL: https://codereview.chromium.org/398313002

git-svn-id: svn://svn.chromium.org/blink/trunk@178571 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The ASSERT checks that no flags cleared by clearPaintInvalidationState()
is still set up after calling repaint-after-compositing-update.

The ASSERT is skipped for SVG subtree as we have an optimization that
skips them. We could reproduce the exact condition but it seems not
very useful to do so.

The change caught several bugs (yay!):
- one in shouldCheckForPaintInvalidationAfterLayout()
that made the ASSERT trip on fast/repaint/table-row.html.
- one where setShouldDoFullPaintInvalidationAfterLayout didn't
mark the containing block chain for invalidation.

BUG=385169

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=178089

Review URL: https://codereview.chromium.org/389573008

git-svn-id: svn://svn.chromium.org/blink/trunk@178569 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Marker events are now displayed on the flame chart's timeline bar.

BUG=361045

Review URL: https://codereview.chromium.org/402113002

git-svn-id: svn://svn.chromium.org/blink/trunk@178568 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=

Review URL: https://codereview.chromium.org/397823003

git-svn-id: svn://svn.chromium.org/blink/trunk@178567 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The WebKitPoint interface implemented as DOMPoint in C++ sides.
Its name can conflict with a new implementation of DOMPoint.
So, we should rename DOMPoint to WebKitPoint in C++ sides.
We will have to remove it after a new implementation is shipped.

BUG=388780

Review URL: https://codereview.chromium.org/404473004

git-svn-id: svn://svn.chromium.org/blink/trunk@178566 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/400643002

Review URL: https://codereview.chromium.org/400643002

git-svn-id: svn://svn.chromium.org/blink/trunk@178565 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Because of https://codereview.chromium.org/401883007, webkit_tests was
failing with the stacktrace[1] below.

This patch attempts to fix that by adding in the extra (unused) argument to
android.py (which extends driver.py), mirroring the change to driver.py

[1]: http://build.chromium.org/p/chromium.webkit/builders/WebKit%20Android%20%28Nexus4%29/builds/16591

The same problem[2] happened when the patch first landed[3], until it was
reverted[4].

[2]: http://build.chromium.org/p/chromium.webkit/builders/WebKit%20Android%20%28Nexus4%29/builds/16610
[3]: https://codereview.chromium.org/378113003
[4]: https://codereview.chromium.org/398543005

worker/1: TypeError('start() takes exactly 3 arguments (4 given)') raised:
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/common/message_pool.py", line 254, in run
      worker.handle(message.name, message.src, *message.args)
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py", line 245, in handle
      device_failed = self._run_test(test_input, test_list_name)
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py", line 288, in _run_test
      self._name, self._driver, test_input, stop_when_done)
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/controllers/single_test_runner.py", line 48, in run_single_test
      return runner.run()
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/controllers/single_test_runner.py", line 126, in run
      return self._run_compare_test()
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/controllers/single_test_runner.py", line 139, in _run_compare_test
      driver_output = self._driver.run_test(self._driver_input(), self._stop_when_done)
    File "/b/build/slave/WebKit_Android__Nexus4_/build/src/third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/port/driver.py", line 154, in run_test
      self.start(driver_input.should_run_pixel_test, driver_input.args, stdin_deadline)

TBR=abarth@chromium.org,dpranke@chromium.org,thestig@chromium.org,ivandavid@chromium.org
NOTRY=true
NOTREECHECKS=true
BUG=388517

Review URL: https://codereview.chromium.org/401243005

git-svn-id: svn://svn.chromium.org/blink/trunk@178563 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Remove user-defined move constructor and assignment operator from WTF::String (https://codereview.chromium.org/367343003/)

Reason for revert:
This can be applied only when user-defined dtor is removed.

Original issue's description:
> Remove user-defined move constructor and assignment operator from WTF::String
> 
> WTF::String can rely on implicit move constructor and assignment
> operator as WTF::RefPtr class has them defined now (r177066).
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=178541

TBR=eseidel@chromium.org

Review URL: https://codereview.chromium.org/402163002

git-svn-id: svn://svn.chromium.org/blink/trunk@178561 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Performance::now() is defined[1] as number of milliseconds since
navigationStart. It works correctly for renderer-initiated navigations,
but for browser-initiated ones we override the navigationStart with the
recording taken on the browser side and this is not represented in
Performace::now().

Among other reasons to be sad about this, this bug makes Chromium's
pagecyclers not represent the renderer initialization times in their
plt measurements.

This patch fixes this by adjusting the fields that are supposed to
correspond to navigationStart when we change navigationStart from the
embedder.

[1] http://www.w3.org/TR/hr-time/#dom-performance-now

BUG=376004

Review URL: https://codereview.chromium.org/379873002

git-svn-id: svn://svn.chromium.org/blink/trunk@178558 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Minor cleanup: Font::codePath() already selects ComplexPath whenever
typesettingFeatures are present (our SimplePath doesn't support kerning
and ligatures), so there's no need to duplicate that test in width().

R=schenney@chromium.org,jbroman@chromium.org,eae@chromium.org

Review URL: https://codereview.chromium.org/401123005

git-svn-id: svn://svn.chromium.org/blink/trunk@178557 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://whatwg.org/html#dom-iframe-allowfullscreen

BUG=383813

Review URL: https://codereview.chromium.org/398843002

git-svn-id: svn://svn.chromium.org/blink/trunk@178555 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Otherwise http/tests/security/xss-DENIED-method-with-iframe-proto.html double-faults while trying to read those global variables that aren't declared.

BUG=

Review URL: https://codereview.chromium.org/405893003

git-svn-id: svn://svn.chromium.org/blink/trunk@178553 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Remove forward declaration from HTMLStyleElement.h - StyleSheet
Remove header inclusions from HTMLStyleElement.cpp - ContextFeatures.h, StyleEngine.h, StyleSheetContents.h

NOTRY=true

Review URL: https://codereview.chromium.org/405563002

git-svn-id: svn://svn.chromium.org/blink/trunk@178552 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Style sheet ID is required for the live location construction, and it is
always known at the point where CSSLocation instances are constructed.

R=vsevik

Review URL: https://codereview.chromium.org/404763002

git-svn-id: svn://svn.chromium.org/blink/trunk@178550 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Review URL: https://codereview.chromium.org/405093002

git-svn-id: svn://svn.chromium.org/blink/trunk@178546 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Patch by Mathias Bynens <mathiasb@opera.com>;.

BUG=391289
TBR=eustas

Review URL: https://codereview.chromium.org/404053003

git-svn-id: svn://svn.chromium.org/blink/trunk@178545 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=

Review URL: https://codereview.chromium.org/391413003

git-svn-id: svn://svn.chromium.org/blink/trunk@178543 bbb929c8-8fbe-4397-9dbb-9b2b20218538

WTF::String can rely on implicit move constructor and assignment
operator as WTF::RefPtr class has them defined now (r177066).

Review URL: https://codereview.chromium.org/367343003

git-svn-id: svn://svn.chromium.org/blink/trunk@178541 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL adds new hasTagName() functions that have the efficiency of
hasLocalName(), but are safe. We can always use hasTagName(), and we'll get a
compile time error if we try to use an SVG tag name with an HTML element.

This makes the code safer and no longer requires good understanding of
hasTagName() vs hasLocalName() to write efficient code. This CL drops support
for hasLocalName(QualifiedName) as calling hasTagName() is now as efficient.

This is based on WebKit r165699 by darin@apple.com: 
http://trac.webkit.org/changeset/165699

R=esprehn@chromium.org, adamk@chromium.org, eseidel@chromium.org

Review URL: https://codereview.chromium.org/406843002

git-svn-id: svn://svn.chromium.org/blink/trunk@178540 bbb929c8-8fbe-4397-9dbb-9b2b20218538

-Add back the dividers between failure cards.
The wall of text and buttons was overwhelming.
-Fix the heights of things so that the page contents
isn't all overflowing out of height:0 containers.
-Remove the useless divider in ct-builder-grid.
-Make the buttons look like esprehn's codereview tool.
-Add a new sheriff-friendly favicon.
-Have the results panel consist of two overflow:auto
regions. One for the menu and one for the results.
-Get rid of tabs on top. Instead have a header that
converts to a "go back" button when you are in the
results page.
-Have ct-tree-status be collapsed to 0 height when it's empty.
To make this work, have the default status be the empty
string so that the status attribute isn't removed when the value
is open.
-Fix ellipsizing CSS in ct-tree-status

NOTRY=true
R=abarth@chromium.org

Review URL: https://codereview.chromium.org/400353006

git-svn-id: svn://svn.chromium.org/blink/trunk@178539 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Also, fix the styling of FAIL failure types now that gtests use that
instead of having FAIL masquerade as TEXT.

BUG=395496
R=pdr@chromium.org

Review URL: https://codereview.chromium.org/409523002

git-svn-id: svn://svn.chromium.org/blink/trunk@178538 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This is so you can see the non-ellipsized tree status and
also to have a link to the page for opening the tree.

NOTRY=true

Review URL: https://codereview.chromium.org/398653005

git-svn-id: svn://svn.chromium.org/blink/trunk@178537 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This broke when I changed the results comparison to use polymer components.
This isn't 100% the same as the old pixelzoomer code. I took this as an
opportunity to simplify it.

Also, change the 4-space indent to 2-space indent and replace
endOfMicroTask calls with requestAnimationFrame to match the rest of
the codebase.

NOTRY=true

Review URL: https://codereview.chromium.org/403983002

git-svn-id: svn://svn.chromium.org/blink/trunk@178536 bbb929c8-8fbe-4397-9dbb-9b2b20218538

NOTRY=true

Review URL: https://codereview.chromium.org/407903002

git-svn-id: svn://svn.chromium.org/blink/trunk@178535 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The current distribution of files doesn't make sense. I have an idea for a
better split that I'd like to try, but that's speculative, so first we should
reset to a good state.

Review URL: https://codereview.chromium.org/408443004

git-svn-id: svn://svn.chromium.org/blink/trunk@178534 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=391662

Review URL: https://codereview.chromium.org/408543002

git-svn-id: svn://svn.chromium.org/blink/trunk@178532 bbb929c8-8fbe-4397-9dbb-9b2b20218538

As per recent spec changes, the start time should no longer factor in to
AnimationPlayer sort order. I've left in the SortInfo class instead of
simply holding on to an unsigned sequence number since we may in the
future add a mechanism for changing sort order, although this is easy
to remove later if won't be adding such functionality.

Review URL: https://codereview.chromium.org/399033002

git-svn-id: svn://svn.chromium.org/blink/trunk@178531 bbb929c8-8fbe-4397-9dbb-9b2b20218538

-Make NetworkSimulator error when runTest calls are nested.
-Harden NetworkSimulator against exceptions.
-Fix incorrect nested runTest call in treestatus_unittests.js.
-Fix network requests in ct-results-by-builder-tests.html.

This exposes that some of the tests were not overriding net
and thus the tests now do a ton of network requests.
This will go away when https://codereview.chromium.org/402603007
lands, so it doesn't seem worth the effort to fix all the
broken cases.

NOTRY=true
R=abarth@chromium.org

Review URL: https://codereview.chromium.org/401163002

git-svn-id: svn://svn.chromium.org/blink/trunk@178529 bbb929c8-8fbe-4397-9dbb-9b2b20218538

-Replace some methods with sugarjs equivalents (e.g. endsWith and unique).
-For methods that are only used in one place, move them to the module
they're used in. These and almost all the code left in the scripts
directory will all go away with https://codereview.chromium.org/402603007.

NOTRY=true

Review URL: https://codereview.chromium.org/400423002

git-svn-id: svn://svn.chromium.org/blink/trunk@178528 bbb929c8-8fbe-4397-9dbb-9b2b20218538

NOTRY=true

Review URL: https://codereview.chromium.org/405853002

git-svn-id: svn://svn.chromium.org/blink/trunk@178527 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The port to polymer is (basically) complete. Sheriff-o-matic
is replacing garden-o-matic.

Only functional change in this patch is to sort the scripts
and imports in ct-sheriff-o-matic.html. The rest of the patch
is just deleting unused files.

There's more code to delete with unused functions in some of
the remaining files, but that can be done in a followup.

NOTRY=true

Review URL: https://codereview.chromium.org/405843002

git-svn-id: svn://svn.chromium.org/blink/trunk@178526 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Tidy up after r178426 and remove now-unused includes.

R=haraken@chromium.org
BUG=394828

Review URL: https://codereview.chromium.org/409433002

git-svn-id: svn://svn.chromium.org/blink/trunk@178525 bbb929c8-8fbe-4397-9dbb-9b2b20218538

In the spirit of sheriff-o-matic only showing problems that
need solving, there's no point in showing tree open status.
We might want to reconsider this in the future where we
show you the actual status message, but right now the open
state is hard-coded to say "Tree is open", which is not
information I need to see.

Also, cleaned up a couple minor polymer things:
-No need to set the attributes attribute if we're using publish.
-Initialize all the member variables on ct-tree-status.
-Don't use setAttribute. Should basically always set element
values by property.

NOTRY=true
R=abarth@chromium.org

Review URL: https://codereview.chromium.org/403723005

git-svn-id: svn://svn.chromium.org/blink/trunk@178524 bbb929c8-8fbe-4397-9dbb-9b2b20218538

[Blink] XMLTreeViewer should load resources from blink_resources.grd instead of relying upon xxd.py (Part-2)

Blink should make use of blink_resources.grd for the inline resources for displaying XML tree view.
This removes the dependency upon using xxd.py which embeds the resources as strings.

BUG=312586

Review URL: https://codereview.chromium.org/391963002

git-svn-id: svn://svn.chromium.org/blink/trunk@178522 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Oilpan: WebAudio: Apply the weak HashMap pattern to remove an entry from AudioContext::m_dirtyAudioSummingJunctions.

r178238, which made AudioSummingJunction::m_context a weak member, didn't
work. If the AudioContext and an AudioSummingJunction object become unreachable
togeter, AudioSummingJunction::m_context is not cleared and their destuction
order is not deterministic. We must not touch weak members in destructors.

With this CL, we register all of AudioSummingJunction objects to AudioContext,
and applies the weak HashMap pattern to clear m_dirtyAudioSummingJunctions.

BUG=392788

Review URL: https://codereview.chromium.org/393133003

git-svn-id: svn://svn.chromium.org/blink/trunk@178521 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This blink side patch has to be preceded to delete below reduntdant
19 header files in ui/gfx of chromium.
(Please refer to https://codereview.chromium.org/392013005/)
These header files need to be deleted 
since they were already moved from ui/gfx to ui/gfx/geometry.

[Rendunt 19 header files to be deleted on chromium-side]
ui/gfx/box_f.h
ui/gfx/point.h
ui/gfx/point3_f.h
ui/gfx/quad_f.h
ui/gfx/insets.h
ui/gfx/point_f.h
ui/gfx/vector2d.h
ui/gfx/matrix3_f.h
ui/gfx/size.h
ui/gfx/size_f.h
ui/gfx/rect.h
ui/gfx/rect_f.h
ui/gfx/rect_conversions.h
ui/gfx/size_conversions.h
ui/gfx/point_conversions.h
ui/gfx/vector2d_conversions.h
ui/gfx/safe_interget_conversions.h
ui/gfx/vector3d_f.h
ui/gfx/vector2d_f.h

BUG=395370

Review URL: https://codereview.chromium.org/400403002

git-svn-id: svn://svn.chromium.org/blink/trunk@178520 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Text wrapping is a little funny in an absolute-positioned element: the
available width of an off-screen element is zero which causes wrapping.
I created a minimal example of this is at http://pr.gg/wrapgenius.html
This behavior is well-known and there are workarounds [1].

The flakiness dashboard popup is similarly affected (see:
http://pr.gg/flakywrapping.png). Because the flakiness dashboard
revision popup has breaks already, we can just use
white-space: nowrap to fix this without any involved workarounds.

[1] http://stackoverflow.com/questions/14819574/preserve-normal-word-wrapping-inside-absolutely-positioned-container

NOTRY=true

Review URL: https://codereview.chromium.org/403553002

git-svn-id: svn://svn.chromium.org/blink/trunk@178519 bbb929c8-8fbe-4397-9dbb-9b2b20218538

r178351 enabled ScriptForbiddenScope in release builds, then we hit a lot of crashes from real-world plugins (see the below bugs). We should suppress the ScriptForbiddenScopes in Document.cpp and FrameView.cpp in release builds.

BUG=395299,395081,395164

Review URL: https://codereview.chromium.org/403133002

git-svn-id: svn://svn.chromium.org/blink/trunk@178518 bbb929c8-8fbe-4397-9dbb-9b2b20218538