GitLab

BUG=391593

Review URL: https://codereview.chromium.org/559423004

git-svn-id: svn://svn.chromium.org/blink/trunk@181903 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=413067
R=ojan@chromium.org

Review URL: https://codereview.chromium.org/559843006

git-svn-id: svn://svn.chromium.org/blink/trunk@181902 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=391593

Review URL: https://codereview.chromium.org/539353004

git-svn-id: svn://svn.chromium.org/blink/trunk@181901 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Gardening: mark large-inlined-script.html as Skip because it is really long, may fail, pass and timeout.

TBR= eustas@chromium.org, jianli@chromium.org
BUG=413249
NOTRY=true

Review URL: https://codereview.chromium.org/565923002

git-svn-id: svn://svn.chromium.org/blink/trunk@181900 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Add setDeviceColorProfile() to allow a render widget to set the
screen color profile of its blink::Page.

Also add a resetDeviceColorProfile() method used to restore the
blink::Page to default color profile state.

Implementation of each in Page are FIXME.

BUG=368663

Review URL: https://codereview.chromium.org/261633002

git-svn-id: svn://svn.chromium.org/blink/trunk@181899 bbb929c8-8fbe-4397-9dbb-9b2b20218538

A trivial conversion is defined as one that cannot throw an exception,
and thus need no v8::TryCatch or ExceptionState, and no "did it succeed?"
check after the conversion.

For such conversions, generate the code

  Type variable = convertedValue()

instead of

  TONATIVE_VOID(Type, variable, convertedValue())

Also include all such conversions in the set of argument conversions that
don't require a v8::TryCatch, since they then no longer do.

BUG=413257

Review URL: https://codereview.chromium.org/564063002

git-svn-id: svn://svn.chromium.org/blink/trunk@181897 bbb929c8-8fbe-4397-9dbb-9b2b20218538

String concatenation is used a lot to build error and exception
messages and it was larger than intended because of over eager
inlining.

Outlining template code saves 90 KB with clang and 65 KB with
gcc (lin64 content_shell). After this there is roughly 60 KB of
concatenation code left (clang).

clang:
Total change: -91866 bytes
==========================
  47 added, totalling +6264 bytes across 3 sources
  38 removed, totalling -18354 bytes across 2 sources
  79 grown, for a net change of +4604 bytes (98534 bytes before, 103138 bytes after) across 50 sources
  405 shrunk, for a net change of -84380 bytes (370130 bytes before, 285750 bytes after) across 134 sources

gcc:
Total change: -66849 bytes
==========================
  333 added, totalling +81561 bytes across 69 sources
  203 removed, totalling -65785 bytes across 76 sources
  211 grown, for a net change of +18666 bytes (228381 bytes before, 247047 bytes after) across 85 sources
  406 shrunk, for a net change of -101291 bytes (568083 bytes before, 466792 bytes after) across 160 sources

BUG=

Review URL: https://codereview.chromium.org/559133002

git-svn-id: svn://svn.chromium.org/blink/trunk@181896 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Resources which were already loaded and then revalidated were skewing the data:
load time was coming from the already loaded resource, compile time was the new
compilation time.

R=jochen@chromium.org
BUG=

Review URL: https://codereview.chromium.org/562653005

git-svn-id: svn://svn.chromium.org/blink/trunk@181895 bbb929c8-8fbe-4397-9dbb-9b2b20218538

After removing the temporary SCM directory change back to the
original cwd. This will avoid errors on future os.cwd() calls.

BUG=412774
TEST=test-webkitpy --all

Review URL: https://codereview.chromium.org/558103003

git-svn-id: svn://svn.chromium.org/blink/trunk@181894 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Remove Host() call from RebaselineServer. The host object is passed
to the RebaselineServer from outside.
This could help with the Mocking as the FIXME mentioned.

BUG=None
TEST=test-webkitpy

Review URL: https://codereview.chromium.org/561733003

git-svn-id: svn://svn.chromium.org/blink/trunk@181893 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=413604
TBR= pmeenan@chromium.org, jianli@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/551543003

git-svn-id: svn://svn.chromium.org/blink/trunk@181892 bbb929c8-8fbe-4397-9dbb-9b2b20218538

To follow the spec[1]. We should throw an type error when
a V8 value for IDL dictionary is neither an object, null, nor
undefined. The current implementation returns an
impl instance with default values.

[1] http://heycam.github.io/webidl/#es-dictionary

BUG=321462

Review URL: https://codereview.chromium.org/568703002

git-svn-id: svn://svn.chromium.org/blink/trunk@181891 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Replacing generic Dictionary. No changes in behavior.
Tests under fast/encoding cover this change.

BUG=403150

Review URL: https://codereview.chromium.org/564463002

git-svn-id: svn://svn.chromium.org/blink/trunk@181890 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Document for no reason.

This is important for Oilpan because the Ranges do not die until the
next GC. This gives a 10% improvement on DOM/textarea-edit in the
oilpan build.

R=yosin@chromium.org
BUG=388681

Review URL: https://codereview.chromium.org/565613002

git-svn-id: svn://svn.chromium.org/blink/trunk@181889 bbb929c8-8fbe-4397-9dbb-9b2b20218538

delete LayoutTests/platform/win-xp/svg/filters/feMerge-wrong-input-expected.txt

BUG=
TBR= jianli@chromium.org

Review URL: https://codereview.chromium.org/567983002

git-svn-id: svn://svn.chromium.org/blink/trunk@181888 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Month buttons should have role=gridcell, and YearListCell should not. So this CL
moves role=gridcell from ListCell to DayCell.
Month buttons should have their own IDs, and readable month label.

BUG=123896

Review URL: https://codereview.chromium.org/564053002

git-svn-id: svn://svn.chromium.org/blink/trunk@181887 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Code border became light gray; indents appeared.

BUG=391593

Review URL: https://codereview.chromium.org/551463003

git-svn-id: svn://svn.chromium.org/blink/trunk@181886 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch completes the transition away from MixedContentChecker as an
object with state, into a world where it just provides static methods.
This should clear up the interface, and make migration to RemoteFrame
a bit simpler.

After this patch, we no longer hang a MixedContentChecker object off
of FrameLoader, nor does MixedContentChecker need to hold a reference
to a frame. Everything we need to do the checks can be done by passing
values into the checker methods.

Note also that this patch deletes a number of expectation files: these
are testharness.js tests, which shouldn't have expectations
committed into the repository. Rather than modifying them, this patch
removes them.

As a drive-by, this also corrects the FormsSubmitted counter.
It ought to count all form submissions as "FormsSubmitted", and mixed
submissions as a subset thereof (e.g. 5 form submissions, 
with 2 of them mixed would result in FormsSubmitted: 5, 
MixedContentFormsSubmitted: 2, as opposed to the current 3 and 2).

BUG=411184

Review URL: https://codereview.chromium.org/566533003

git-svn-id: svn://svn.chromium.org/blink/trunk@181885 bbb929c8-8fbe-4397-9dbb-9b2b20218538

- bindings_core_dictionary_impl_generated target should have
  core_dictionary_idl_files as inputs, otherwise IDL dictionaries
  impl classes won't get re-generated when they are edited.
- impl_includes_for_type() should check idl_type.base_type
  (not idl_type.name), otherwise the CG won't include necessary
  header files for nullable interface members.

BUG=321462

Review URL: https://codereview.chromium.org/563163002

git-svn-id: svn://svn.chromium.org/blink/trunk@181884 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Since index is not part of the FormControl state for select, there is no
way to restore it to proper index if there are duplicate values. This
patch includes the index in the FormControl state for select.

BUG=401506

Review URL: https://codereview.chromium.org/541693003

git-svn-id: svn://svn.chromium.org/blink/trunk@181883 bbb929c8-8fbe-4397-9dbb-9b2b20218538

There was custom generated code specifically for handling arguments with
[Clamp], which is quite unnecessary since the toUInt*() have an integer
conversion configuration parameter that among other things can be used to
request clamping.

Related: Fix the somewhat broken handling the |Clamp| integer conversion
configuration in the toUInt*() helpers. An optimization code path for the
case that the V8 value is an int32 forgot to do clamping, there was no
clamping at all in toUInt64(), and +/- Infinity was converted to zero
instead of being clamped.

Clamping support is not added to toInt64(), since the clampTo<> helper in
Source/wtf/MathExtras.h explicitly doesn't support |long long int|, which
is int64_t on some platforms.

BUG=413257

Review URL: https://codereview.chromium.org/563793002

git-svn-id: svn://svn.chromium.org/blink/trunk@181882 bbb929c8-8fbe-4397-9dbb-9b2b20218538

To remove generic Dictionary from MutationObserver.
Added Vector<String> -> HashSet<AtomicString> conversion
because attributeFilter member in MutationObserverInit
is sequence<DOMString>.

Tests under fast/dom/MutationObserver cover this change.

Spec: http://www.w3.org/TR/dom/#interface-mutationobserver

BUG=403150

Review URL: https://codereview.chromium.org/560003002

git-svn-id: svn://svn.chromium.org/blink/trunk@181881 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=412088

Review URL: https://codereview.chromium.org/559733005

git-svn-id: svn://svn.chromium.org/blink/trunk@181880 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Use IDL dictionary (NotificationOptions) instead of generic
Dictionary. Tests under fast/notifications are passed.

BUG=403150

Review URL: https://codereview.chromium.org/556193002

git-svn-id: svn://svn.chromium.org/blink/trunk@181879 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Revert of Service Worker: Add tests for same-scope new script registration (patchset #1 id:1 of https://codereview.chromium.org/512703002/)

Reason for revert:
This crash will be fixed by https://codereview.chromium.org/556003006.

Original issue's description:
> Revert of Service Worker: Add tests for same-scope new script registration (patchset #6 of https://codereview.chromium.org/480943002/)
> 
> Reason for revert:
> This is crashing on Mac.
> 
> Original issue's description:
> > Service Worker: Add tests for same-scope, new script registration
> > 
> > Tests for Chromium patch:
> > https://codereview.chromium.org/506043002/
> > 
> > BUG=398355
> > 
> > Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=180948
> 
> TBR=michaeln@chromium.org,nhiroki@chromium.org,falken@chromium.org
> NOTREECHECKS=true
> NOTRY=true
> BUG=398355
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=180956

TBR=michaeln@chromium.org,nhiroki@chromium.org,falken@chromium.org,mlamouri@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=398355

Review URL: https://codereview.chromium.org/547743003

git-svn-id: svn://svn.chromium.org/blink/trunk@181878 bbb929c8-8fbe-4397-9dbb-9b2b20218538

To remove generic Dictionary from a DOMPoint constructor.
No changes in behavior; fast/dom/geometry-interfaces-dom-point.html
should cover this change.

Spec: http://dev.w3.org/fxtf/geometry/#DOMPoint

BUG=403150

Review URL: https://codereview.chromium.org/558163002

git-svn-id: svn://svn.chromium.org/blink/trunk@181877 bbb929c8-8fbe-4397-9dbb-9b2b20218538

public/web/WebSettings.h does:
and uses some types from ICU in its definition. Thus, anything that
depends on the blink public API and includes this header must have the
icu include dirs on their include path. This forwards icu_config from
the blink_headers target so dependents do not have to do this themselves
(unless they are linking against ICU directly)

R=brettw@chromium.org

Review URL: https://codereview.chromium.org/563963003

git-svn-id: svn://svn.chromium.org/blink/trunk@181876 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Before applying this patch, run-bindings-tests uses 'tests' as component_dir.
So tests results had '#include "bindings/tests/.../V8Test....h".

However, for testing binding modularization, need to emluate 'Source/core' and 'Source/modules'.
So Source/bindings/tests/idls/core should be treated as 'core' (i.e. the tests should be treated as if they are placed under Source/core).
Souce/bindings/tests/idsl/modules should be treated as 'modules'.

BUG=358074

Review URL: https://codereview.chromium.org/563453002

git-svn-id: svn://svn.chromium.org/blink/trunk@181875 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The test suites are moving from being pulled in via src/DEPS to
LayoutTests/w3c to being checked in directly to LayoutTests/imported.

This import uses update-w3c-deps run manually and
imported csswg-test@6ed3bb6642b0d382ce8c3f7912172f2420eb7756 and
imported web-platform-tests@6ed3bb6642b0d382ce8c3f7912172f2420eb7756

In addition, we modify run-webkit-tests to manually skip the old
LayoutTests/w3c directory if it exists, and to start printing warnings
that the repos are stale and should be deleted.

This is patchset 3 of the 4 needed to complete the cutover.

TBR=eseidel@chromium.org
BUG=413454

Review URL: https://codereview.chromium.org/560893005

git-svn-id: svn://svn.chromium.org/blink/trunk@181874 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The compositor expects that the whole composited scrolling layer to be
invalidated in time and won't update the tiles that are newly visible
when scrolled if the tiles are up-to-date from the compositor side.

If we clip invalidations, the compositor will think the layer areas
outside of the clip are all valid and won't repaint them.

BUG=412516
TEST=compositing/repaint/should-not-clip-composited-viewport-scrolling-layer.html
TEST=compositing/repaint/should-not-clip-composited-overflow-scrolling-layer.html

Review URL: https://codereview.chromium.org/566433002

git-svn-id: svn://svn.chromium.org/blink/trunk@181873 bbb929c8-8fbe-4397-9dbb-9b2b20218538

scriptState->executionContext() returns NULL in the following situation.

[Main thread]
- WorkerThread::stop()
 - m_workerGlobalScope->script()->scheduleExecutionTermination();
  - v8::V8::TerminateExecution(m_isolate);

[Worker thread]
- FetchBodyStream::readAsync()
 - ScriptPromiseResolver::create()
  ---> v8::Promise::Resolver::New()
     - v8::internal::Execution::Call
      - v8::internal::Invoke
       - JS
        - v8::internal::__RT_impl_Runtime_StackGuard
         - v8::internal::Isolate::TerminateExecution()
            Sets TerminationException.
 - scriptState->executionContext()
  - toExecutionContext()
   - V8WorkerGlobalScope::findInstanceInPrototypeChain()
    - V8PerIsolateData::findInstanceInPrototypeChain()
     - V8PerIsolateData::findInstanceInPrototypeChain()
      - v8::Object::FindInstanceInPrototypeChain()
       - ON_BAILOUT(isolate,
                    "v8::Object::FindInstanceInPrototypeChain()",
                    return Local<v8::Object>());
        - IsExecutionTerminatingCheck()
          Returns true.

BUG=409755
TEST=run_webkit_tests http/tests/serviceworker/request.html

Review URL: https://codereview.chromium.org/535193002

git-svn-id: svn://svn.chromium.org/blink/trunk@181872 bbb929c8-8fbe-4397-9dbb-9b2b20218538

It will be less error-prone to pass references instead of pointers.
Customers don't need to add ASSERT().

This is a follow-up CL of https://codereview.chromium.org/547383004/,
and no changes in behavior.

BUG=321462

Review URL: https://codereview.chromium.org/558283004

git-svn-id: svn://svn.chromium.org/blink/trunk@181871 bbb929c8-8fbe-4397-9dbb-9b2b20218538

TBR=oilpan-reviews,adamk@chromium.org
BUG=
NOTRY=true

Review URL: https://codereview.chromium.org/570453003

git-svn-id: svn://svn.chromium.org/blink/trunk@181870 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Just some styling and import boogs, but they meant buttons
didn't render at all.

NOTRY=true
TBR=leviw@chromium.org

Review URL: https://codereview.chromium.org/564023002

git-svn-id: svn://svn.chromium.org/blink/trunk@181869 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Instead of going through string(), use
the AtomicString API directly.

Review URL: https://codereview.chromium.org/560273004

git-svn-id: svn://svn.chromium.org/blink/trunk@181868 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL caused use-after-free in :LocalDOMWindow::willDetachDocumentFromFrame.
https://code.google.com/p/chromium/issues/detail?id=413316

> Simplify DOMWindowProperty and its unregistration.
> 
> As LocalDOMWindow controls the registration lifetime of its
> DOMWindowProperties, there's no need to offer a separate unregistration
> mechanism. Remove it, simplify notification code + retire now unused
> LocalDOMWindow back reference on DOMWindowProperty.
> 
> R=haraken,ager
> BUG=
> 
> Review URL: https://codereview.chromium.org/558213002

TBR=sigbjornf@opera.com

Review URL: https://codereview.chromium.org/564013002

git-svn-id: svn://svn.chromium.org/blink/trunk@181867 bbb929c8-8fbe-4397-9dbb-9b2b20218538

We have basic support for IDL dictionary so we can use it for
MIDIOptions. Renamed m_sysexEnabled and setSysexEnabled()
because their role have changed.

MIDIOptions is the first IDL dictionary in modules/ so this CL
also contains GYP/GN changes.

Existing tests (LayoutTests/webmidi) should cover this change.

BUG=403150

Review URL: https://codereview.chromium.org/547383004

git-svn-id: svn://svn.chromium.org/blink/trunk@181866 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch adds trace events for setNeedsLayout calls so that we can
do end-to-end tracking of what causes repaints to occur. The devtools
side isn't yet hooked up so this patch simply adds the trace event.

As a cleanup, "core/inspector/InspectorTraceEvents.h" has been moved
up into RenderObject and unneeded includes have been removed.

R=kouhei
BUG=410701

Review URL: https://codereview.chromium.org/561413002

git-svn-id: svn://svn.chromium.org/blink/trunk@181865 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Directly embed the failure cards into the failure stream, and abstract
the button UI out into ct-failure-card-buttons.

NOTRY=true
R=ojan@chromium.org

Review URL: https://codereview.chromium.org/565523002

git-svn-id: svn://svn.chromium.org/blink/trunk@181864 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Instead of passing a String url and having ResourceRequest
create a KURL out of it, pass in the readily available
KURL object.

Review URL: https://codereview.chromium.org/561383002

git-svn-id: svn://svn.chromium.org/blink/trunk@181862 bbb929c8-8fbe-4397-9dbb-9b2b20218538