WebKit-https.git
9 years agoFix assert when foster parenting self-closing elements
abarth@webkit.org [Wed, 4 Jan 2012 07:03:47 +0000 (07:03 +0000)]
Fix assert when foster parenting self-closing elements
https://bugs.webkit.org/show_bug.cgi?id=75527

Unreviewed.

Source/WebCore:

* html/parser/HTMLConstructionSite.cpp:
(WebCore::HTMLConstructionSite::fosterParent):
    - We can't execute this task immediately because the code for
      self-closing elements wants to set the self-closing flag.
      Instead, we queue the task.

LayoutTests:

Add a test case that used to case an assert.

* html5lib/resources/webkit02.dat:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104014 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[BlackBerry] Need to support credentials in the URL
commit-queue@webkit.org [Wed, 4 Jan 2012 06:43:32 +0000 (06:43 +0000)]
[BlackBerry] Need to support credentials in the URL
https://bugs.webkit.org/show_bug.cgi?id=75341

Patch by Charles Wei <charles.wei@torchmobile.com.cn> on 2012-01-03
Reviewed by George Staikos.

No new tests required.

* platform/network/blackberry/NetworkJob.cpp:
(WebCore::NetworkJob::sendRequestWithCredentials):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104013 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoReplace clearLocalStorage.js with localStorage.clear() now that it's implemented
jochen@chromium.org [Wed, 4 Jan 2012 06:24:42 +0000 (06:24 +0000)]
Replace clearLocalStorage.js with localStorage.clear() now that it's implemented
https://bugs.webkit.org/show_bug.cgi?id=75484

Reviewed by Daniel Bates.

* storage/domstorage/localstorage/delete-removal.html:
* storage/domstorage/localstorage/enumerate-storage.html:
* storage/domstorage/localstorage/enumerate-with-length-and-key.html:
* storage/domstorage/localstorage/index-get-and-set.html:
* storage/domstorage/localstorage/missing-arguments.html:
* storage/domstorage/localstorage/private-browsing-affects-storage.html:
* storage/domstorage/localstorage/resources/clearLocalStorage.js: Removed.
* storage/domstorage/localstorage/simple-usage.html:
* storage/domstorage/localstorage/string-conversion.html:
* storage/domstorage/localstorage/window-open.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104012 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFix ChangeLog
abarth@webkit.org [Wed, 4 Jan 2012 06:22:18 +0000 (06:22 +0000)]
Fix ChangeLog

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Use data decoding swizzle for turbo JPEG grayscale images
noel.gordon@gmail.com [Wed, 4 Jan 2012 06:09:11 +0000 (06:09 +0000)]
[chromium] Use data decoding swizzle for turbo JPEG grayscale images
https://bugs.webkit.org/show_bug.cgi?id=75189

Reviewed by Adam Barth.

No new tests.  Covered by exiting tests, in particular for grayscale
images that were corrupted by earlier versions on libjpeg-turbo:

tables/mozilla/bugs/bug29314.html
tables/mozilla/bugs/bug13169.html
tables/mozilla/bugs/bug10565.html
tables/mozilla/bugs/bug11026.html
fast/repaint/backgroundSizeRepaint.html
fast/repaint/block-layout-inline-children-replaced.html
fast/repaint/clipped-relative.html
fast/repaint/selected-replaced.html
tables/mozilla/bugs/bug12908-1.html

* platform/image-decoders/jpeg/JPEGImageDecoder.cpp:
(WebCore::JPEGImageReader::decode):  Use a data decoding swizzle for
JCS_GRAYSCALE images.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoMinor speed improvement in HTML parser
abarth@webkit.org [Wed, 4 Jan 2012 05:51:02 +0000 (05:51 +0000)]
Minor speed improvement in HTML parser
https://bugs.webkit.org/show_bug.cgi?id=75517

Reviewed by Eric Seidel.

This improves the html-parser.html benchmark by 0.3%.  I don't think
that's actually anything to write home about, but this patch makes me
feel better about my life.

* html/parser/HTMLFormattingElementList.cpp:
(WebCore::HTMLFormattingElementList::tryToEnsureNoahsArkConditionQuickly):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoRemove temporary code that we've inserted to implement the [Supplemental] IDL increme...
haraken@chromium.org [Wed, 4 Jan 2012 05:40:45 +0000 (05:40 +0000)]
Remove temporary code that we've inserted to implement the [Supplemental] IDL incrementally
https://bugs.webkit.org/show_bug.cgi?id=75510

Reviewed by Adam Barth.

Now all build systems support the [Supplemental] IDL. This patch removes all
temporary code that we've inserted to implement the [Supplemental] IDL.

No tests. No change in behavior. Confirm that all builds pass.

* bindings/js/JSDOMWindowCustom.cpp:
* bindings/scripts/CodeGeneratorJS.pm:
(GenerateHeader):
* bindings/scripts/generate-bindings.pl:
* page/DOMWindow.idl: Removed all attributes with a temporal [Supplemented] IDL.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104008 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoMake fast/multicol/huge-column-count.html dumpAsText
noel.gordon@gmail.com [Wed, 4 Jan 2012 05:28:46 +0000 (05:28 +0000)]
Make fast/multicol/huge-column-count.html dumpAsText
https://bugs.webkit.org/show_bug.cgi?id=75513

Reviewed by Adam Barth.

Test added in r83470: convert it to dumpAsText, update expectations.

* fast/multicol/huge-column-count-expected.txt: Added.
* fast/multicol/huge-column-count.html:
* platform/chromium/test_expectations.txt:
* platform/efl/fast/multicol/huge-column-count-expected.png: Removed.
* platform/efl/fast/multicol/huge-column-count-expected.txt: Removed.
* platform/gtk/fast/multicol/huge-column-count-expected.txt: Removed.
* platform/mac/fast/multicol/huge-column-count-expected.txt: Removed.
* platform/qt/fast/multicol/huge-column-count-expected.png: Removed.
* platform/qt/fast/multicol/huge-column-count-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104007 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohtml-parser.html takes too long to run
abarth@webkit.org [Wed, 4 Jan 2012 04:39:04 +0000 (04:39 +0000)]
html-parser.html takes too long to run
https://bugs.webkit.org/show_bug.cgi?id=75515

Reviewed by Eric Seidel.

In working on runner.js, we introduced a factor of 10 increase to the
running time of this benchmark.  This patch dials down the number of
iterations to something more user-friendly.

* Parser/html-parser.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoGive embedders a chance to handle postMessage calls
commit-queue@webkit.org [Wed, 4 Jan 2012 04:38:42 +0000 (04:38 +0000)]
Give embedders a chance to handle postMessage calls
https://bugs.webkit.org/show_bug.cgi?id=73883

Source/WebCore:

To support cross-process postMessage calls in Chromium (bug 73337), we need to intercept
postMessage calls to proxy windows. Originally we were just going to add a native event
listener on the Chromium side, but that required more changes to WebKit and was a bit of
a hack. See bug 73359 for a discuss about moving to this approach.

Patch by Karl Koscher <supersat@chromium.org> on 2012-01-03
Reviewed by Adam Barth.

Test: platform/chromium/fast/events/intercept-postmessage.html

* loader/FrameLoaderClient.h:
(WebCore::FrameLoaderClient::willCheckAndDispatchPostMessage): new method to allow the
    embedder to intercept postMessage calls
* page/DOMWindow.cpp:
(WebCore::DOMWindow::postMessageTimerFired): add a call to
    FrameLoaderClient::willCheckAndDispatchPostMessage

Source/WebKit/chromium:

Patch by Karl Koscher <supersat@chromium.org> on 2012-01-03
Reviewed by Adam Barth.

* public/WebFrameClient.h:
(WebKit::WebFrameClient::willCheckAndDispatchMessageEvent):
    interface to give the embedder a chance to handle this postMessage call
* src/FrameLoaderClientImpl.cpp:
(WebKit::FrameLoaderClientImpl::willCheckAndDispatchMessageEvent): Call
    willCheckAndDispatchMessageEvent on WebFrameClient
* src/FrameLoaderClientImpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104005 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoscript-in-svg-in-HTML doesn't run
abarth@webkit.org [Wed, 4 Jan 2012 03:54:35 +0000 (03:54 +0000)]
script-in-svg-in-HTML doesn't run
https://bugs.webkit.org/show_bug.cgi?id=75233

Reviewed by Daniel Bates.

Add a test to make sure that we run scripts in SVG.

* fast/parser/script-in-svg-runs-expected.txt: Added.
* fast/parser/script-in-svg-runs.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Prevent crashing due to NULL texture updater.
commit-queue@webkit.org [Wed, 4 Jan 2012 03:39:11 +0000 (03:39 +0000)]
[chromium] Prevent crashing due to NULL texture updater.
https://bugs.webkit.org/show_bug.cgi?id=75288

Patch by Eric Penner <epenner@google.com> on 2012-01-03
Reviewed by James Robinson.

* platform/graphics/chromium/ContentLayerChromium.cpp:
* platform/graphics/chromium/ContentLayerChromium.h:
* platform/graphics/chromium/ImageLayerChromium.cpp:
* platform/graphics/chromium/ImageLayerChromium.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104003 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] CCLayerSorter accidentally reverses order of some layers.
commit-queue@webkit.org [Wed, 4 Jan 2012 03:36:46 +0000 (03:36 +0000)]
[chromium] CCLayerSorter accidentally reverses order of some layers.
https://bugs.webkit.org/show_bug.cgi?id=75046

Patch by Shawn Singh <shawnsingh@chromium.org> on 2012-01-03
Reviewed by James Robinson.

Source/WebCore:

Unit test added to CCLayerSorterTest.cpp

Fixes two related bugs that were causing z-ordering problems in
layers when preserves3D triggers the need for layer sorting.

The first problem was that CCLayerSorter accidentally reversed the
order of layers when there was no sorting dependency between them.

The second problem was that zDiff had numerical precision problems
that forced sorting dependencies that were sometimes incorrect,
when the dependencies should not have existed.

* platform/graphics/chromium/cc/CCLayerSorter.cpp:
(WebCore::CCLayerSorter::checkOverlap):
(WebCore::CCLayerSorter::sort):

Source/WebKit/chromium:

* tests/CCLayerSorterTest.cpp:
(WebCore::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[BlackBerry] Add support of blob form data to the BlackBerry port
leo.yang@torchmobile.com.cn [Wed, 4 Jan 2012 03:34:18 +0000 (03:34 +0000)]
[BlackBerry] Add support of blob form data to the BlackBerry port
https://bugs.webkit.org/show_bug.cgi?id=75218

Pass FormDataElement::encodedBlob type of form data to the BlackBerry
platform request to support blob data and blob file range.

Reviewed by George Staikos.

The port can't be built now, no new tests so far.

* platform/network/blackberry/ResourceRequestBlackBerry.cpp:
(WebCore::ResourceRequest::initializePlatformRequest):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoTwo null crashes in Treebuilder
abarth@webkit.org [Wed, 4 Jan 2012 02:57:43 +0000 (02:57 +0000)]
Two null crashes in Treebuilder
https://bugs.webkit.org/show_bug.cgi?id=66101

Reviewed by Eric Seidel.

Source/WebCore:

The underly issue with causing the crash is that we're re-entering the
tree builder.  We've done a bunch of point fixes around tree builder
re-entrancy, but neither the implementation nor the specification are
really designed to handle re-entrancy.

Firefox avoids this problem by putting the parser on its own thread.  I
don't think we're quite ready to do that yet (although we will
eventually, presumably, as computers become ever more parallel).  The
approach in this patch is to queue up the DOM mutations and actually
perform them on a shallower stack.  That's essentially the approach
we've used for executing <scripts>.

This patch queues up DOM modifications and executes them from a clean
call stack, stopping us from re-entering the tree builder.  We might
need to experiment with exactly where to kick off the queue, but this
location seems reasonable and fixes the crash.

Test: fast/parser/re-enter-and-crash.html

* html/parser/HTMLConstructionSite.cpp:
(WebCore::executeTask):
    - Add a helper function for actually executing a queue DOM
      mutation.
(WebCore::HTMLConstructionSite::attach):
    - Instead of actually attaching the element to the DOM, this
      function now queues the element for attachment.  In a follow-up
      patch, I plan to change this function to return void (and I'll
      probably rename it to something that makes it clear that it only
      queues the attachment).
(WebCore::HTMLConstructionSite::executeQueuedTasks):
(WebCore::HTMLConstructionSite::insertDoctype):
    - Setting the compatmode from the Doctype requires the doctype to
      actually be in the DOM, so we need to execute our queued tasks
      synchronously.  We can likely improve this function by passing
      the Doctype element explicitly in a follow-up patch.
(WebCore::HTMLConstructionSite::insertSelfClosingHTMLElement):
(WebCore::HTMLConstructionSite::insertTextNode):
(WebCore::HTMLConstructionSite::findFosterSite):
(WebCore::HTMLConstructionSite::fosterParent):
* html/parser/HTMLConstructionSite.h:
(WebCore::HTMLConstructionSiteTask::HTMLConstructionSiteTask):
    - Add a task object that holds on to the relevant elements.  We
      define some vector traits for this object to match the traits on
      RefPtr (which make Vector operations faster by explaining that
      this type is moveable without having to churn reference counts).
* html/parser/HTMLElementStack.cpp:
(WebCore::HTMLElementStack::pushCommon):
    - We delay the "begin parsing" call until we actually attach the
      element to the DOM.  That splits the responsibility for calling
      begin/finished, which is less than ideal, but I didn't see
      another solution.
* html/parser/HTMLTreeBuilder.cpp:
(WebCore::HTMLTreeBuilder::constructTreeFromAtomicToken):
    - Kick off the queued mutations.

LayoutTests:

Test that we no longer crash on this test case.

* fast/parser/re-enter-and-crash-expected.txt: Added.
* fast/parser/re-enter-and-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@104000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoview-source doesn't colorize </script> correctly
abarth@webkit.org [Wed, 4 Jan 2012 02:41:57 +0000 (02:41 +0000)]
view-source doesn't colorize </script> correctly
https://bugs.webkit.org/show_bug.cgi?id=62971

Reviewed by Eric Seidel.

Source/WebCore:

The reason these tags weren't colorized correctly was because the
characters for the end tags where consumed by the tokenizer at the same
time as it consumed the text of the script tag.  These characters are
buffered internally by the tokenizer because the tokenizer is searching
for "an appropriate end tag", which is "</script>" in this case.

I tried a number of different approaches to fixing this bug.  The
approach in this patch adds an accessor to the tokenizer to read out
the buffered characters.  This approach makes it easier for the
HTMLSourceTracker to get exactly the right value for these buffered
characters without having to do a complicated simulation of the
buffering itself.

Tests: fast/frames/viewsource-plain-text-tags.html:

* html/parser/HTMLViewSourceParser.cpp:
(WebCore::HTMLViewSourceParser::pumpTokenizer):
* html/parser/HTMLDocumentParser.cpp:
(WebCore::HTMLDocumentParser::pumpTokenizer):
  - Give the HTMLSourceTracker a pointer to the tokenizer so it can ask
    for the buffered characters.
* html/parser/HTMLSourceTracker.cpp:
(WebCore::HTMLSourceTracker::start):
  - The idea here is to treat characters buffered internally by the
    tokenizer the same way we treat characters that were contained in a
    previous segment of source.  We copy them into our accumulation
    buffer and adjust the token base offset to account for the extra
    characters.
(WebCore::HTMLSourceTracker::end):
  - Don't consier the character buffered by the tokenizer to be part of
    the token we've just finished.
(WebCore::HTMLSourceTracker::sourceForToken):
  - Remove the assumption that all of the m_previousSource is contained
    in the source for the current token.  Because we now understand
    that the tokenizer can buffer some characters internally, we might
    not exhaust the m_previousSource.
* html/parser/HTMLSourceTracker.h:
  - Rename m_sourceFromPreviousSegments to m_previousSource and changed
    to a SegementedString to avoid extra mallocs we keep appending tiny
    segments.
  - Rename m_source to m_currentSource to contrast nicely with
    m_previousSource.
* html/parser/HTMLTokenizer.cpp:
(WebCore::HTMLTokenizer::flushBufferedEndTag):
(WebCore::HTMLTokenizer::nextToken):
  - Previously, we cleared the temporary buffer lazily when we needed
    to add new characters to it.  Now we clear it eagerly so that it's
    length tells us whether we're currently using it to store
    characters.
  - Previously, we weren't storing the character that we used to
    terminate the appropriate end tag (e.g., the > in </script>)
    because we didn't need to "unbuffer" that character into the
    HTMLToken::appendToCharacter.  Now, we do store that character in
    the temporary buffer so that the HTMLSourceTracker can see that
    character.
(WebCore::HTMLTokenizer::bufferedCharacters):
* html/parser/HTMLTokenizer.h:
(WebCore::HTMLTokenizer::numberOfBufferedCharacters):
  - Add accessor methods to copy out the buffered characters.
  - Also, unrelated to the rest of this patch, I made the constructor
    for HTMLTokenizer explicit because it takes only a single
    paramater, per the style guide.
* platform/text/SegmentedString.cpp:
(WebCore::SegmentedString::clear):
  - SegmentedString::clear wasn't clearing all of its state.  I don't
    think this issue was actually observable, but I noticed it when
    working on this patch, so I fixed it.

LayoutTests:

* fast/frames/viewsource-plain-text-tags.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoUnreviewed: fast/filesystem/workers/file-writer-events.html crash suppressed on Linux
dslomov@google.com [Wed, 4 Jan 2012 02:28:37 +0000 (02:28 +0000)]
Unreviewed: fast/filesystem/workers/file-writer-events.html crash suppressed on Linux
due to BUGCR108798.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoExclude Mac's PlatformEventFactory from iOS
commit-queue@webkit.org [Wed, 4 Jan 2012 02:26:11 +0000 (02:26 +0000)]
Exclude Mac's PlatformEventFactory from iOS
https://bugs.webkit.org/show_bug.cgi?id=75493

Patch by Benjamin Poulain <bpoulain@apple.com> on 2012-01-03
Reviewed by Sam Weinig.

Move PlatformEventFactory.(h|cpp) to PlatformEventFactoryMac(h|cpp) and exclude the file
from iOS build.

* Configurations/WebCore.xcconfig: Also exclude EventHandlerMac for consistency.
* WebCore.xcodeproj/project.pbxproj:
* page/mac/EventHandlerMac.mm:
* platform/mac/PlatformEventFactoryMac.h: Renamed from Source/WebCore/platform/mac/PlatformEventFactory.h.
* platform/mac/PlatformEventFactoryMac.mm: Renamed from Source/WebCore/platform/mac/PlatformEventFactory.mm.
(WebCore::globalPoint):
(WebCore::globalPointForEvent):
(WebCore::pointForEvent):
(WebCore::mouseButtonForEvent):
(WebCore::mouseEventTypeForEvent):
(WebCore::clickCountForEvent):
(WebCore::momentumPhaseForEvent):
(WebCore::phaseForEvent):
(WebCore::gestureEventTypeForEvent):
(WebCore::textFromEvent):
(WebCore::unmodifiedTextFromEvent):
(WebCore::keyIdentifierForKeyEvent):
(WebCore::isKeypadEvent):
(WebCore::windowsKeyCodeForKeyEvent):
(WebCore::isKeyUpEvent):
(WebCore::modifiersForEvent):
(WebCore::PlatformMouseEventBuilder::PlatformMouseEventBuilder):
(WebCore::PlatformEventFactory::createPlatformMouseEvent):
(WebCore::PlatformWheelEventBuilder::PlatformWheelEventBuilder):
(WebCore::PlatformEventFactory::createPlatformWheelEvent):
(WebCore::PlatformKeyboardEventBuilder::PlatformKeyboardEventBuilder):
(WebCore::PlatformEventFactory::createPlatformKeyboardEvent):
(WebCore::PlatformGestureEventBuilder::PlatformGestureEventBuilder):
(WebCore::PlatformEventFactory::createPlatformGestureEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] LocalFileSystemChromium needs some thread safety fixes.
levin@chromium.org [Wed, 4 Jan 2012 02:25:56 +0000 (02:25 +0000)]
[chromium] LocalFileSystemChromium needs some thread safety fixes.
https://bugs.webkit.org/show_bug.cgi?id=75494

Reviewed by Dmitry Titov.

This is essentially the same fix as r102894 except in slightly different code.

The important part of this fix is the removal of AllowCrossThreadAccess so
that ref counting happens appropriately.

Minor clean up throughout: Removed unnecessary WTF prefix in many
of these places and unnecessary String().

* src/LocalFileSystemChromium.cpp: See r102894 since this
is basically the same thing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoUnreviewed: filed CR109077 and updated expectations for http/tests/inspector/appcache...
dslomov@google.com [Wed, 4 Jan 2012 01:45:00 +0000 (01:45 +0000)]
Unreviewed: filed CR109077 and updated expectations for http/tests/inspector/appcache/appcache-swap.html.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoOptimize AudioBufferSourceNode process by avoiding interpolation when pitchRate==1
commit-queue@webkit.org [Wed, 4 Jan 2012 01:28:26 +0000 (01:28 +0000)]
Optimize AudioBufferSourceNode process by avoiding interpolation when pitchRate==1
https://bugs.webkit.org/show_bug.cgi?id=74592

Patch by Wei James <james.wei@intel.com> on 2012-01-03
Reviewed by Kenneth Russell.

Covered by existing webaudio layout tests, especially webaudio/audiobuffersource-playbackrate.html

* webaudio/AudioBufferSourceNode.cpp:
(WebCore::AudioBufferSourceNode::renderSilenceAndFinishIfNotLooping):
(WebCore::AudioBufferSourceNode::renderFromBuffer):
* webaudio/AudioBufferSourceNode.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago<http://webkit.org/b/75504> Web Inspector: Access to Node Highlighting info without...
joepeck@webkit.org [Wed, 4 Jan 2012 01:18:10 +0000 (01:18 +0000)]
<webkit.org/b/75504> Web Inspector: Access to Node Highlighting info without drawing

Provide a "getHighlight" method alongside "drawHighlight" that
gets most of the important node highlighting information. This
can be used by a port to do its own node highlighting or provide
extra content without interfering with the highlight rects.

Reviewed by Timothy Hatcher.

* WebCore.xcodeproj/project.pbxproj:
Privately expose the Highlight struct in DOMNodeHighlighter.h.

* inspector/DOMNodeHighlighter.cpp:
(WebCore::DOMNodeHighlighter::drawHighlight):
(WebCore::DOMNodeHighlighter::getHighlight):
* inspector/DOMNodeHighlighter.h:
Reuse the existing draw code to instead populate
the Highlight struct with the information it is
about to draw, and only draw if there is a context.

* inspector/InspectorController.cpp:
(WebCore::InspectorController::getHighlight):
* inspector/InspectorController.h:
* inspector/InspectorDOMAgent.cpp:
(WebCore::InspectorDOMAgent::getHighlight):
* inspector/InspectorDOMAgent.h:
Get highlight information by going through getHighlight
in InspectorController. This is alongside drawHighlight.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoClear localStorage before making assumption about its contents
jochen@chromium.org [Wed, 4 Jan 2012 01:10:54 +0000 (01:10 +0000)]
Clear localStorage before making assumption about its contents
https://bugs.webkit.org/show_bug.cgi?id=75469

Reviewed by Daniel Bates.

* platform/chromium/permissionclient/storage-permission.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Push drawsContent and contentsVisible into accelerated compositor
commit-queue@webkit.org [Wed, 4 Jan 2012 00:53:45 +0000 (00:53 +0000)]
[chromium] Push drawsContent and contentsVisible into accelerated compositor
https://bugs.webkit.org/show_bug.cgi?id=71209

Patch by Shawn Singh <shawnsingh@chromium.org> on 2012-01-03
Reviewed by James Robinson.

Source/WebCore:

Tests: compositing/visibility/visibility-simple-canvas2d-layer.html
       compositing/visibility/visibility-simple-video-layer.html
       compositing/visibility/visibility-simple-webgl-layer.html

Rebaselined compositing/visibility/visibility-image-layers.html

The CSS visibility property was not being properly obeyed by many
specialized layer types in the accelerated compositor. The root of
the problem was that drawsContent and contentsVisible flags were
not properly propagated into the compositor. This patch removes
the drawsContent() callback, and instead makes sure that
drawsContent and contentsVisible are "pushed" through the
accelerated compositor.

* platform/graphics/chromium/Canvas2DLayerChromium.cpp:
(WebCore::Canvas2DLayerChromium::drawsContent):
* platform/graphics/chromium/ContentLayerChromium.cpp:
* platform/graphics/chromium/ContentLayerChromium.h:
* platform/graphics/chromium/GraphicsLayerChromium.cpp:
(WebCore::GraphicsLayerChromium::setDrawsContent):
(WebCore::GraphicsLayerChromium::setContentsVisible):
(WebCore::GraphicsLayerChromium::setContentsToImage):
(WebCore::GraphicsLayerChromium::setContentsToCanvas):
(WebCore::GraphicsLayerChromium::setContentsToMedia):
(WebCore::GraphicsLayerChromium::updateLayerIsDrawable):
(WebCore::GraphicsLayerChromium::setupContentsLayer):
* platform/graphics/chromium/GraphicsLayerChromium.h:
* platform/graphics/chromium/LayerChromium.cpp:
(WebCore::LayerChromium::LayerChromium):
(WebCore::LayerChromium::setIsDrawable):
* platform/graphics/chromium/LayerChromium.h:
(WebCore::LayerChromium::drawsContent):
* platform/graphics/chromium/PluginLayerChromium.h:
* platform/graphics/chromium/TiledLayerChromium.cpp:
(WebCore::TiledLayerChromium::drawsContent):
* platform/graphics/chromium/VideoLayerChromium.h:
* platform/graphics/chromium/WebGLLayerChromium.cpp:
(WebCore::WebGLLayerChromium::drawsContent):
* platform/graphics/chromium/cc/CCLayerImpl.cpp:
(WebCore::CCLayerImpl::dumpLayerProperties):

Source/WebKit/chromium:

* src/WebContentLayerImpl.cpp:
(WebKit::WebContentLayerImpl::WebContentLayerImpl):
(WebKit::WebContentLayerImpl::setDrawsContent):
* src/WebContentLayerImpl.h:
* src/WebExternalTextureLayerImpl.cpp:
(WebKit::WebExternalTextureLayerImpl::drawsContent):
* src/WebLayerImpl.cpp:
(WebKit::WebLayerImpl::WebLayerImpl):
* src/WebLayerImpl.h:
* tests/CCLayerTreeHostTest.cpp:
(WTF::ContentLayerChromiumWithUpdateTracking::ContentLayerChromiumWithUpdateTracking):
* tests/Canvas2DLayerChromiumTest.cpp:
* tests/LayerChromiumTest.cpp:
* tests/TiledLayerChromiumTest.cpp:
(WTF::FakeTiledLayerChromium::FakeTiledLayerChromium):

LayoutTests:

* compositing/visibility/visibility-image-layers-expected.png:
* compositing/visibility/visibility-image-layers-expected.txt: Renamed from LayoutTests/platform/mac/compositing/visibility/visibility-image-layers-expected.txt.
* compositing/visibility/visibility-simple-canvas2d-layer-expected.png: Added.
* compositing/visibility/visibility-simple-canvas2d-layer-expected.txt: Added.
* compositing/visibility/visibility-simple-canvas2d-layer.html: Added.
* compositing/visibility/visibility-simple-video-layer-expected.png: Added.
* compositing/visibility/visibility-simple-video-layer-expected.txt: Added.
* compositing/visibility/visibility-simple-video-layer.html: Added.
* compositing/visibility/visibility-simple-webgl-layer-expected.png: Added.
* compositing/visibility/visibility-simple-webgl-layer-expected.txt: Added.
* compositing/visibility/visibility-simple-webgl-layer.html: Added.
* platform/chromium-mac-snowleopard/compositing/visibility/visibility-image-layers-expected.png: Removed.
* platform/chromium-win/compositing/visibility/visibility-image-layers-expected.png: Removed.
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Make sure root damage rect gets passed to renderer
commit-queue@webkit.org [Wed, 4 Jan 2012 00:34:05 +0000 (00:34 +0000)]
[chromium] Make sure root damage rect gets passed to renderer
https://bugs.webkit.org/show_bug.cgi?id=74893

Patch by Shawn Singh <shawnsingh@chromium.org> on 2012-01-03
Reviewed by James Robinson.

Source/WebCore:

Unit test added to CCLayerTreeHostImplTest.cpp

* platform/graphics/chromium/LayerRendererChromium.cpp:
(WebCore::LayerRendererChromium::swapBuffers):
* platform/graphics/chromium/LayerRendererChromium.h:
* platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
(WebCore::CCLayerTreeHostImpl::calculateRenderPasses):
(WebCore::CCLayerTreeHostImpl::swapBuffers):
* platform/graphics/chromium/cc/CCLayerTreeHostImpl.h:

Source/WebKit/chromium:

* tests/CCLayerTreeHostImplTest.cpp:
(WebKit::TEST_F):
(WebKit::PartialSwapTrackerContext::initialize):
(WebKit::PartialSwapTrackerContext::postSubBufferCHROMIUM):
(WebKit::PartialSwapTrackerContext::getString):
(WebKit::PartialSwapTrackerContext::partialSwapRect):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Unreviewed, svg/carto.net/selectionlist.svg appears to
tony@chromium.org [Wed, 4 Jan 2012 00:29:02 +0000 (00:29 +0000)]
[chromium] Unreviewed, svg/carto.net/selectionlist.svg appears to
just need a rebaseline on Linux.

* platform/chromium-linux/svg/carto.net/selectionlist-expected.png:
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoUnreviewed: files bug CR109068 for appcache/interrupted-update.html crash.
dslomov@google.com [Wed, 4 Jan 2012 00:25:44 +0000 (00:25 +0000)]
Unreviewed: files bug CR109068 for appcache/interrupted-update.html crash.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoIncrement the code-review.js version number to bust browser
ojan@chromium.org [Wed, 4 Jan 2012 00:23:55 +0000 (00:23 +0000)]
Increment the code-review.js version number to bust browser
caches.

* PrettyPatch/PrettyPatch.rb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoWhen showing results.html pass a correct URL to webbrowser.open
jochen@chromium.org [Wed, 4 Jan 2012 00:18:27 +0000 (00:18 +0000)]
When showing results.html pass a correct URL to webbrowser.open
https://bugs.webkit.org/show_bug.cgi?id=75438

Reviewed by Adam Barth.

* Scripts/webkitpy/common/system/filesystem.py:
(file_path_as_url):
* Scripts/webkitpy/common/system/filesystem_mock.py:
(file_path_as_url):
* Scripts/webkitpy/common/system/filesystem_unittest.py:
(test_file_path_as_url):
* Scripts/webkitpy/layout_tests/port/base.py:
(show_results_html_file):
* Scripts/webkitpy/layout_tests/port/win_unittest.py:
(test_show_results_html_file):
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
(test_results_directory_relative):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoshrink the header on the flakiness dashboard
ojan@chromium.org [Wed, 4 Jan 2012 00:16:42 +0000 (00:16 +0000)]
shrink the header on the flakiness dashboard
https://bugs.webkit.org/show_bug.cgi?id=65153

Reviewed by Adam Roben.

Move some of the "show" links to be to the right of the header
instead of below the header when they content they're showing
is hidden. That way we don't waste vertical space on them.

No new tests. This code is covered by existing tests.

* TestResultServer/static-dashboards/flakiness_dashboard.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoREGRESSION (r98196-98236): Incorrect layout of iGoogle with RSS feeds
fpizlo@apple.com [Wed, 4 Jan 2012 00:06:42 +0000 (00:06 +0000)]
REGRESSION (r98196-98236): Incorrect layout of iGoogle with RSS feeds
https://bugs.webkit.org/show_bug.cgi?id=75303
<rdar://problem/10633533>

Source/JavaScriptCore:

Reviewed by Gavin Barraclough.

The this argument was not being kept alive in some cases during inlining and intrinsic
optimizations.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::handleCall):
(JSC::DFG::ByteCodeParser::emitFunctionCheck):
(JSC::DFG::ByteCodeParser::handleInlining):

LayoutTests:

Reviewed by Gavin Barraclough.

* fast/js/dfg-inline-unused-this-expected.txt: Added.
* fast/js/dfg-inline-unused-this-method-check-expected.txt: Added.
* fast/js/dfg-inline-unused-this-method-check.html: Added.
* fast/js/dfg-inline-unused-this.html: Added.
* fast/js/dfg-intrinsic-unused-this-expected.txt: Added.
* fast/js/dfg-intrinsic-unused-this-method-check-expected.txt: Added.
* fast/js/dfg-intrinsic-unused-this-method-check.html: Added.
* fast/js/dfg-intrinsic-unused-this.html: Added.
* fast/js/script-tests/dfg-inline-unused-this-method-check.js: Added.
(foo):
(bar):
(baz):
* fast/js/script-tests/dfg-inline-unused-this.js: Added.
(foo):
(bar):
(baz):
* fast/js/script-tests/dfg-intrinsic-unused-this-method-check.js: Added.
(bar):
(baz):
* fast/js/script-tests/dfg-intrinsic-unused-this.js: Added.
(bar):
(baz):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Fix compilation with clang and the shared build
tony@chromium.org [Tue, 3 Jan 2012 23:59:01 +0000 (23:59 +0000)]
[chromium] Fix compilation with clang and the shared build
https://bugs.webkit.org/show_bug.cgi?id=75492

Reviewed by Ojan Vafai.

* WebKit.gyp: Turn off -Wglobal-constructors because test files using
GTEST have global constructors.
* tests/InnerGestureRecognizerTest.cpp:
(TEST_F): Move global into test function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agov8 binding: npCreateV8ScriptObject() should not returned an existing V8NPObject if...
commit-queue@webkit.org [Tue, 3 Jan 2012 23:40:09 +0000 (23:40 +0000)]
v8 binding: npCreateV8ScriptObject() should not returned an existing V8NPObject if the rootObject doesn't match
https://bugs.webkit.org/show_bug.cgi?id=74515

Patch by Yuzhu Shen <yzshen@chromium.org> on 2012-01-03
Reviewed by Nate Chapin.

Source/WebCore:

Test: http/tests/plugins/create-v8-script-objects.html

* bindings/v8/NPV8Object.cpp:
(WebCore::freeV8NPObject):
(WebCore::npCreateV8ScriptObject):

LayoutTests:

* http/tests/plugins/create-v8-script-objects-expected.txt: Added.
* http/tests/plugins/create-v8-script-objects.html: Added.
* http/tests/plugins/resources/create-v8-script-objects-iframe.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Revert accidentally removed chrome client compositing check
enne@google.com [Tue, 3 Jan 2012 23:39:13 +0000 (23:39 +0000)]
[chromium] Revert accidentally removed chrome client compositing check
https://bugs.webkit.org/show_bug.cgi?id=75500

Reviewed by James Robinson.

This check was overzealously removed in r103968 as a part of
re-enabling compositing for RTL pages. Add it back.

* src/ChromeClientImpl.cpp:
(WebKit::ChromeClientImpl::allowedCompositingTriggers):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoWindows build fix.
barraclough@apple.com [Tue, 3 Jan 2012 23:31:25 +0000 (23:31 +0000)]
Windows build fix.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoLeak of WebNotificationClient when page is destroyed
jonlee@apple.com [Tue, 3 Jan 2012 23:27:26 +0000 (23:27 +0000)]
Leak of WebNotificationClient when page is destroyed
https://bugs.webkit.org/show_bug.cgi?id=74980
<rdar://problem/10611231>

Reviewed by Mark Rowe.

Source/WebCore:

* notifications/NotificationController.cpp:
(WebCore::NotificationController::~NotificationController): Notify the client
that the controller has been destroyed.
* notifications/NotificationPresenter.h: Add notificationControllerDestroyed()
callback to the client interface.

Source/WebKit:

* blackberry/WebCoreSupport/NotificationPresenterImpl.cpp:
(WebKit::NotificationPresenterImpl::notificationControllerDestroyed): Empty implementation, because ownership
model differs.
* blackberry/WebCoreSupport/NotificationPresenterImpl.h:

Source/WebKit/chromium:

* src/NotificationPresenterImpl.cpp:
(WebKit::NotificationPresenterImpl::notificationControllerDestroyed): Empty implementation because ownership
model differs.
* src/NotificationPresenterImpl.h:

Source/WebKit/efl:

* WebCoreSupport/NotificationPresenterClientEfl.cpp:
(WebCore::NotificationPresenterClientEfl::notificationControllerDestroyed): Stub implementation, because
the notification presenter is unimplemented.
* WebCoreSupport/NotificationPresenterClientEfl.h:

Source/WebKit/mac:

* WebCoreSupport/WebNotificationClient.h:
* WebCoreSupport/WebNotificationClient.mm:
(WebNotificationClient::notificationControllerDestroyed): Remove the instance.

Source/WebKit/qt:

* WebCoreSupport/NotificationPresenterClientQt.cpp:
(WebCore::NotificationPresenterClientQt::notificationControllerDestroyed): Empty implementation, because
ownership model differs.
* WebCoreSupport/NotificationPresenterClientQt.h:

Source/WebKit2:

* WebProcess/WebCoreSupport/WebNotificationClient.cpp:
(WebKit::WebNotificationClient::notificationControllerDestroyed): Delete the instance.
* WebProcess/WebCoreSupport/WebNotificationClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agogetComputedStyle for list-style is not implemented.
alexis.menard@openbossa.org [Tue, 3 Jan 2012 23:20:43 +0000 (23:20 +0000)]
getComputedStyle for list-style is not implemented.
https://bugs.webkit.org/show_bug.cgi?id=75443

Reviewed by Tony Chang.

Source/WebCore:

Implement getComputedStyle for list-style.

Test: fast/css/getComputedStyle/getComputedStyle-list-style-shorthand.html

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

LayoutTests:

Add new tests to make sure we return the correct values.

* fast/css/getComputedStyle/getComputedStyle-list-style-shorthand-expected.txt: Added.
* fast/css/getComputedStyle/getComputedStyle-list-style-shorthand.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Unreviewed, rebaseline of svg/hixie/text/003.html. It
tony@chromium.org [Tue, 3 Jan 2012 23:13:52 +0000 (23:13 +0000)]
[chromium] Unreviewed, rebaseline of svg/hixie/text/003.html.  It
doesn't look flaky on the flakiness dashboard.

* platform/chromium-linux/svg/hixie/text/003-expected.png: Added.
* platform/chromium-mac-leopard/svg/hixie/text/003-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/hixie/text/003-expected.png: Added.
* platform/chromium-mac/svg/hixie/text/003-expected.png: Removed.
* platform/chromium-win/svg/hixie/text/003-expected.png: Added.
* platform/chromium-win/svg/hixie/text/003-expected.txt: Added.
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Remove unneeded content vs. layer space distinction in tiled layers
enne@google.com [Tue, 3 Jan 2012 23:07:32 +0000 (23:07 +0000)]
[chromium] Remove unneeded content vs. layer space distinction in tiled layers
https://bugs.webkit.org/show_bug.cgi?id=75498

Reviewed by James Robinson.

Long ago, before scrollbars became their own layers, tiled layers had
a distinction between content space (what's mostly called layer space
elsewhere) and layer space (the transformed content space). It seemed
like this transform could get reused for supporting RTL pages, but as
that's not the case, it should be eradicated.

Tiled layers now deal with everything in "layer space", which is a 2d
rectangle starting at (0, 0) in the top left of the layer and
extending to contentBounds() in the bottom right.

As no code actually set the tiler position, this change should be a
no-op.

Test: LayoutTests/compositing/

* platform/graphics/chromium/TiledLayerChromium.cpp:
(WebCore::UpdatableTile::isDirty):
(WebCore::UpdatableTile::clearDirty):
(WebCore::TiledLayerChromium::updateCompositorResources):
(WebCore::TiledLayerChromium::createTile):
(WebCore::TiledLayerChromium::invalidateRect):
(WebCore::TiledLayerChromium::protectTileTextures):
(WebCore::TiledLayerChromium::prepareToUpdateTiles):
(WebCore::TiledLayerChromium::prepareToUpdate):
(WebCore::TiledLayerChromium::prepareToUpdateIdle):
(WebCore::TiledLayerChromium::needsIdlePaint):
(WebCore::TiledLayerChromium::idlePaintRect):
* platform/graphics/chromium/TiledLayerChromium.h:
* platform/graphics/chromium/cc/CCLayerTilingData.cpp:
(WebCore::CCLayerTilingData::operator=):
(WebCore::CCLayerTilingData::layerRectToTileIndices):
(WebCore::CCLayerTilingData::tileRect):
* platform/graphics/chromium/cc/CCLayerTilingData.h:
* platform/graphics/chromium/cc/CCTiledLayerImpl.cpp:
(WebCore::CCTiledLayerImpl::appendQuads):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Remove WebLayerClient
jamesr@google.com [Tue, 3 Jan 2012 22:33:16 +0000 (22:33 +0000)]
[chromium] Remove WebLayerClient
https://bugs.webkit.org/show_bug.cgi?id=75481

Reviewed by Darin Fisher.

* WebKit.gyp:
* public/platform/WebContentLayer.h:
* public/platform/WebExternalTextureLayer.h:
* public/platform/WebLayer.h:
* public/platform/WebLayerClient.h: Removed.
* src/WebContentLayer.cpp:
* src/WebExternalTextureLayer.cpp:
* src/WebLayer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoWindows build fix.
barraclough@apple.com [Tue, 3 Jan 2012 22:33:02 +0000 (22:33 +0000)]
Windows build fix.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Support RTL pages and elements in the compositor
enne@google.com [Tue, 3 Jan 2012 22:13:15 +0000 (22:13 +0000)]
[chromium] Support RTL pages and elements in the compositor
https://bugs.webkit.org/show_bug.cgi?id=56591

Reviewed by James Robinson.

Source/WebKit/chromium:

As the noncomposited content host layer is created and managed by the
Chromium compositor, adjust its transform appropriately in the layer
tree and at paint time to handle the offset required due to the
initial containing block of RTL pages.

* src/ChromeClientImpl.cpp:
(WebKit::ChromeClientImpl::allowedCompositingTriggers):
* src/NonCompositedContentHost.cpp:
(WebKit::NonCompositedContentHost::setViewport):
(WebKit::NonCompositedContentHost::paintContents):
* src/NonCompositedContentHost.h:
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::updateLayerTreeViewport):

LayoutTests:

Re-enable RTL tests for Chromium.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohttps://bugs.webkit.org/show_bug.cgi?id=75140
barraclough@apple.com [Tue, 3 Jan 2012 21:45:56 +0000 (21:45 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=75140

Reviewed by Sam Weinig.

Rewrite JSArray::putSlowCase to be much cleaner & simpler.

This rewrite only significantly changes behaviour for sparse array, specifically
in how sparse arrays are reified back to vector form. This does not affect arrays
with less than 10000 entries (since these always use a vector). The more common
cases of sparse array behavior (though large sparse arrays are rare) - arrays that
always remain sparse, and arrays that are filled in reverse sequential order -
should be just as fast or faster (since reification is simpler & no longer
requires map lookups) after these changes.

Simplifying this code allows all cases of putByIndex that need to grow the vector
to do so via increaseVectorLength, which means that this method can encapsulate
the policy of determining how the vector should be grown.

No performance impact.

* runtime/JSArray.cpp:
(JSC::isDenseEnoughForVector):
    - any array of length <= MIN_SPARSE_ARRAY_INDEX is dense enough for a vector.
(JSC::JSArray::putByIndex):
    - simplify & comment.
(JSC::JSArray::putByIndexBeyondVectorLength):
    - Re-written to be much clearer & simpler.
(JSC::JSArray::increaseVectorLength):
(JSC::JSArray::increaseVectorPrefixLength):
    - add explicit checks against MAX_STORAGE_VECTOR_LENGTH, so clients do not need do so.
(JSC::JSArray::push):
    - simplify & comment.
* runtime/JSArray.h:
    - removed SparseArrayValueMap::take.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoWindows build fix.
barraclough@apple.com [Tue, 3 Jan 2012 21:37:56 +0000 (21:37 +0000)]
Windows build fix.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohttps://bugs.webkit.org/show_bug.cgi?id=75140
barraclough@apple.com [Tue, 3 Jan 2012 21:33:20 +0000 (21:33 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=75140

Reviewed by Sam Weinig.

Simplify JSArray creation - remove ArgsList/JSValue* create methods
(this functionality can be implemented in terms of tryCreateUninitialized).

* JavaScriptCore.exp:
* runtime/ArrayConstructor.cpp:
    - use constructArray/constructEmptyArray instead of calling JSArray::create directly
(JSC::constructArrayWithSizeQuirk):
* runtime/JSArray.cpp:
* runtime/JSArray.h:
    - removed ArgsList/JSValue* create methods
* runtime/JSGlobalObject.h:
(JSC::constructEmptyArray):
(JSC::constructArray):
    - changed to be implemented in terms of JSArray::tryCreateUninitialized

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoAdd a tool to analyze change logs
rniwa@webkit.org [Tue, 3 Jan 2012 21:24:49 +0000 (21:24 +0000)]
Add a tool to analyze change logs
https://bugs.webkit.org/show_bug.cgi?id=72243

Reviewed by Eric Seidel.

Add the initial implementation of analyze-changelog command. This version supports parsing multiple changelogs
in a single directory and summarizes the result in json files. It also copies summary.html to the output directory
for visualizing json files.

Usage: webkit-patch analyze-changelog [--changelog-count N] <CHANGELOG_DIR> <OUTPUT_DIR>
where CHANGELOG_DIR is the directory to look for changelogs, OUTPUT_DIR is the directory into which the tool
generates json files, and N is the number of changelogs to parse.

* Scripts/webkitpy/common/config/contributionareas.py: Added names().
* Scripts/webkitpy/common/system/filesystem_mock.py: Added readline and iterator support to ReadableTextFileObject.
* Scripts/webkitpy/tool/commands/__init__.py:
* Scripts/webkitpy/tool/commands/analyzechangelog.py: Added.
* Scripts/webkitpy/tool/commands/analyzechangelog_unittest.py: Added.
* Scripts/webkitpy/tool/commands/data/summary.html: Added.
* Scripts/webkitpy/tool/steps/options.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohttps://bugs.webkit.org/show_bug.cgi?id=75429
barraclough@apple.com [Tue, 3 Jan 2012 21:05:46 +0000 (21:05 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=75429
ThrowTypeError should be a singleton object

Reviewed by Sam Weinig.

Source/JavaScriptCore:

Per section 13.2.3 of the spec.
We could change setAccessorDescriptor to be able to share the global
GetterSetter object, rather than storing the accessor functions and
creating a new GetterSetter in defineProperty - but this won't be a
small change to PropertyDescriptors (and would probably mean making
GetterSetter objects immutable?) - so I'll leave that for another
patch.

* JavaScriptCore.exp:
    - don't export setAccessorDescriptor
* runtime/Arguments.cpp:
(JSC::Arguments::createStrictModeCallerIfNecessary):
(JSC::Arguments::createStrictModeCalleeIfNecessary):
    - call throwTypeErrorGetterSetter instead of createTypeErrorFunction
* runtime/Error.cpp:
* runtime/Error.h:
    - remove createTypeErrorFunction
* runtime/JSFunction.cpp:
* runtime/JSFunction.h:
    - remove unused createDescriptorForThrowingProperty
* runtime/JSGlobalObject.cpp:
(JSC::JSGlobalObject::reset):
(JSC::JSGlobalObject::visitChildren):
    - removed m_strictModeTypeErrorFunctionStructure.
* runtime/JSGlobalObject.h:
(JSC::JSGlobalObject::internalFunctionStructure):
    - removed m_strictModeTypeErrorFunctionStructure.
* runtime/PropertyDescriptor.cpp:
(JSC::PropertyDescriptor::setAccessorDescriptor):
    - changed to take a GetterSetter
* runtime/PropertyDescriptor.h:
    - changed to take a GetterSetter

LayoutTests:

Per section 13.2.3 of the spec.

* fast/js/basic-strict-mode-expected.txt:
    - ThrowTypeError is a singleton, so cannot generate
      property-specific error messages.
* fast/js/script-tests/strict-throw-type-error.js: Added.
* fast/js/strict-throw-type-error-expected.txt: Added.
* fast/js/strict-throw-type-error.html: Added.
    - added test case that ThrowTypeError is a singleton

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago<http://webkit.org/b/59951> Implement Date and Time Input Value Sanitization
joepeck@webkit.org [Tue, 3 Jan 2012 20:21:44 +0000 (20:21 +0000)]
<webkit.org/b/59951> Implement Date and Time Input Value Sanitization

Source/WebCore:

Sanitize non-parsable incoming strings to the empty string.

Reviewed by Kent Tamura.

Covered by existing tests.

* html/BaseDateAndTimeInputType.h:
* html/BaseDateAndTimeInputType.cpp:
(WebCore::BaseDateAndTimeInputType::sanitizeValue): check if the
incoming string is valid. If not just return the empty string.

LayoutTests:

Update ValidityState typeMismatch tests for date types, since
setting an invalid value now sanitized to the empty string.
So, what used to be an invalid value with a typeMismatch is
now a valid empty string and there is never a typeMismatch.

  - for rangeOverflow / rangeUnderflow tests update the
    test to point out when a value was sanitized.
  - for typeMismatch tests, typeMismatch can no longer
    happen for sanitized values. Change the test to output
    if the value was sanitized or not and ensure that
    we are never in a typeMismatch state. Disabled inputs
    are also sanitized when their value changes.

Reviewed by Kent Tamura.

* fast/forms/date/ValidityState-rangeOverflow-date-expected.txt:
* fast/forms/date/ValidityState-rangeOverflow-date.html:
* fast/forms/date/ValidityState-rangeUnderflow-date-expected.txt:
* fast/forms/date/ValidityState-rangeUnderflow-date.html:
* fast/forms/date/ValidityState-typeMismatch-date-expected.txt:
* fast/forms/date/ValidityState-typeMismatch-date.html:
* fast/forms/datetime/ValidityState-rangeOverflow-datetime-expected.txt:
* fast/forms/datetime/ValidityState-rangeOverflow-datetime.html:
* fast/forms/datetime/ValidityState-rangeUnderflow-datetime-expected.txt:
* fast/forms/datetime/ValidityState-rangeUnderflow-datetime.html:
* fast/forms/datetime/ValidityState-typeMismatch-datetime-expected.txt:
* fast/forms/datetime/ValidityState-typeMismatch-datetime.html:
* fast/forms/datetimelocal/ValidityState-rangeOverflow-datetimelocal-expected.txt:
* fast/forms/datetimelocal/ValidityState-rangeOverflow-datetimelocal.html:
* fast/forms/datetimelocal/ValidityState-rangeUnderflow-datetimelocal-expected.txt:
* fast/forms/datetimelocal/ValidityState-rangeUnderflow-datetimelocal.html:
* fast/forms/datetimelocal/ValidityState-typeMismatch-datetimelocal-expected.txt:
* fast/forms/datetimelocal/ValidityState-typeMismatch-datetimelocal.html:
* fast/forms/month/ValidityState-rangeOverflow-month-expected.txt:
* fast/forms/month/ValidityState-rangeOverflow-month.html:
* fast/forms/month/ValidityState-rangeUnderflow-month-expected.txt:
* fast/forms/month/ValidityState-rangeUnderflow-month.html:
* fast/forms/month/ValidityState-typeMismatch-month-expected.txt:
* fast/forms/month/ValidityState-typeMismatch-month.html:
* fast/forms/time/ValidityState-rangeOverflow-time-expected.txt:
* fast/forms/time/ValidityState-rangeOverflow-time.html:
* fast/forms/time/ValidityState-rangeUnderflow-time-expected.txt:
* fast/forms/time/ValidityState-rangeUnderflow-time.html:
* fast/forms/time/ValidityState-typeMismatch-time-expected.txt:
* fast/forms/time/ValidityState-typeMismatch-time.html:
* fast/forms/week/ValidityState-rangeOverflow-week-expected.txt:
* fast/forms/week/ValidityState-rangeOverflow-week.html:
* fast/forms/week/ValidityState-rangeUnderflow-week-expected.txt:
* fast/forms/week/ValidityState-rangeUnderflow-week.html:
* fast/forms/week/ValidityState-typeMismatch-week-expected.txt:
* fast/forms/week/ValidityState-typeMismatch-week.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoRemove unnecessary forward declaration of Function
weinig@apple.com [Tue, 3 Jan 2012 20:19:48 +0000 (20:19 +0000)]
Remove unnecessary forward declaration of Function
https://bugs.webkit.org/show_bug.cgi?id=75485

Reviewed by Dan Bernstein.

* Platform/WorkQueue.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoRenderLayer::backgroundClipRect should not check parent()
jchaffraix@webkit.org [Tue, 3 Jan 2012 20:08:14 +0000 (20:08 +0000)]
RenderLayer::backgroundClipRect should not check parent()
https://bugs.webkit.org/show_bug.cgi?id=73731

Reviewed by Simon Fraser.

Clean-up only, no expected change in behavior.

* rendering/RenderLayer.cpp:
(WebCore::backgroundClipRectForPosition): Changed RenderObject::isPositioned() to
a check for AbsolutePosition for consistency but also as this is equivalent due to:
- the previous check for FixedPosition.
- RenderView, which is positioned, will never goes to this code as it has no parent().

(WebCore::RenderLayer::backgroundClipRect): Removed the parent() check. While at
it, also moved the inline initialization of |backgroundClipRect| to its own function
and removed a |view| check as the associated ASSERT seems to never have been reached.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago <rdar://problem/10637779> REGRESSION (r102247): Focus ring is not drawn aroun...
ap@apple.com [Tue, 3 Jan 2012 20:02:51 +0000 (20:02 +0000)]
    <rdar://problem/10637779> REGRESSION (r102247): Focus ring is not drawn around a button

        Rubber-stamped by Dan Bernstein.

        * platform/mac/ThemeMac.mm: (WebCore::updateStates): Flip an incorrect conditional, this code
        should run on Snow Leopard and Lion.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoSource/WebCore: getComputedStyle should return shorthands property with the minimum...
alexis.menard@openbossa.org [Tue, 3 Jan 2012 19:56:48 +0000 (19:56 +0000)]
Source/WebCore: getComputedStyle should return shorthands property with the minimum number of sides possible.
https://bugs.webkit.org/show_bug.cgi?id=75434

Reviewed by Tony Chang.

As stated in http://www.w3.org/TR/DOM-Level-2-Style/css.html#CSS-CSS2Properties we should
return shorthands properties with the minimum number of sides possible.

Existing tests have been updated accordingly.

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):
(WebCore::CSSComputedStyleDeclaration::getCSSPropertyValuesForSidesShorthand):
* css/CSSComputedStyleDeclaration.h:

LayoutTests: getComputedStyle should return shorthands properties with the minimum number of sides possible.
https://bugs.webkit.org/show_bug.cgi?id=75434

Reviewed by Tony Chang.

As stated in http://www.w3.org/TR/DOM-Level-2-Style/css.html#CSS-CSS2Properties we should
return shorthands properties with the minimum number of sides possible.

* fast/css/getComputedStyle/getComputedStyle-border-color-shorthand-expected.txt:
* fast/css/getComputedStyle/getComputedStyle-border-color-shorthand.html:
* fast/css/getComputedStyle/getComputedStyle-border-style-shorthand-expected.txt:
* fast/css/getComputedStyle/getComputedStyle-border-style-shorthand.html:
* fast/css/getComputedStyle/getComputedStyle-border-width-expected.txt:
* fast/css/getComputedStyle/getComputedStyle-border-width.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoMake Port subclasses override the operating_system method rather than setting an...
aroben@apple.com [Tue, 3 Jan 2012 19:53:50 +0000 (19:53 +0000)]
Make Port subclasses override the operating_system method rather than setting an attribute

Without this patch, any use of Port.operating_system() within a port's __init__ method would
return the default value ("mac"). This was confusing the version-checking logic inside
ApplePort.__init__ on Windows.

Fixes <http://webkit.org/b/75479> <rdar://problem/10637385> REGRESSION (r102161): NRWT
crashes on launch on Windows 7 SP1

Reviewed by Eric Seidel.

* Scripts/webkitpy/layout_tests/port/apple.py:
(ApplePort.__init__): Replaced uses of self._operating_system with self.operating_system().

* Scripts/webkitpy/layout_tests/port/base.py:
(Port.__init__): Removed code that set the self._operating_system attribute. Subclasses
should now override the operating_system method instead.
(Port.operating_system): Changed to always return "mac". Subclasses should override to
return something else.

* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
(ChromiumLinuxPort.__init__):
(ChromiumLinuxPort.operating_system):
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
(ChromiumMacPort.__init__):
(ChromiumMacPort.operating_system):
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
(ChromiumWinPort.__init__):
(ChromiumWinPort.operating_system):
* Scripts/webkitpy/layout_tests/port/mac.py:
(MacPort.__init__):
(MacPort.operating_system):
* Scripts/webkitpy/layout_tests/port/win.py:
(WinPort.__init__):
(WinPort.operating_system):
Changed to override the operating_system method rather than setting the _operating_system
attribute, since setting the attribute isn't sufficient for code inside __init__ methods.

* Scripts/webkitpy/layout_tests/port/qt.py:
(QtPort.__init__):
(QtPort._path_to_webcore_library):
Changed to use self.operating_system() instead of self._operating_system, for consistency
with other Port-related code.

(QtPort.operating_system): Added this override of Port.operating_system that returns the
value we determined in our __init__ method.

* Scripts/webkitpy/layout_tests/port/test.py:
(TestPort.operating_system): Added this override of Port.operating_system that returns the
value we determined in our __init__ method.

* Scripts/webkitpy/layout_tests/port/base_unittest.py:
(PortTest.test_operating_system):
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
(ChromiumLinuxPortTest.test_operating_system):
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
(ChromiumMacPortTest.test_operating_system):
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
(ChromiumWinPortTest.test_operating_system):
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
(MacPortTest.test_operating_system):
* Scripts/webkitpy/layout_tests/port/qt_unittest.py:
(QtPortTest.test_operating_system):
* Scripts/webkitpy/layout_tests/port/win_unittest.py:
(WinPortTest.test_operating_system):
Test that all ports return the expected operating system string.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] GTK+ 3 theming code does not use the GtkOrientable style class
commit-queue@webkit.org [Tue, 3 Jan 2012 19:51:30 +0000 (19:51 +0000)]
[GTK] GTK+ 3 theming code does not use the GtkOrientable style class
https://bugs.webkit.org/show_bug.cgi?id=72789

Patch by Martin Robinson <mrobinson@igalia.com> on 2012-01-03
Reviewed by Philippe Normand.

No new tests. Testing GTK+ theming is quite difficult
because no two distributions have the same set of themes.
Current tests prevent regressions.

* platform/gtk/RenderThemeGtk3.cpp:
(WebCore::applySliderStyleContextClasses): Added this helper that
properly handles the orientable style class.
(WebCore::RenderThemeGtk::paintSliderTrack): Use the new helper.
(WebCore::RenderThemeGtk::paintSliderThumb): Ditto.
* platform/gtk/ScrollbarThemeGtk3.cpp:
(WebCore::applyScrollbarStyleContextClasses): Added this helper that
properly handles the orientable style class.
(WebCore::ScrollbarThemeGtk::paintTrackBackground): Use the new helper.
(WebCore::ScrollbarThemeGtk::paintScrollbarBackground): Ditto.
(WebCore::ScrollbarThemeGtk::paintThumb): Ditto.
(WebCore::ScrollbarThemeGtk::paintButton): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoNULL ptr in WebCore::SVGStyledTransformableElement::animatedLocalTransform
commit-queue@webkit.org [Tue, 3 Jan 2012 19:33:09 +0000 (19:33 +0000)]
NULL ptr in WebCore::SVGStyledTransformableElement::animatedLocalTransform
https://bugs.webkit.org/show_bug.cgi?id=75227

Patch by Florin Malita <fmalita@google.com> on 2012-01-03
Reviewed by Dirk Schulze.

Source/WebCore:

Test: svg/custom/webkit-transform-crash.html

* svg/SVGStyledTransformableElement.cpp:
(WebCore::SVGStyledTransformableElement::animatedLocalTransform):
Add a null-renderer check.

LayoutTests:

* svg/custom/webkit-transform-crash-expected.txt: Added.
* svg/custom/webkit-transform-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFixed a typo in the change log
mitz@apple.com [Tue, 3 Jan 2012 19:29:09 +0000 (19:29 +0000)]
Fixed a typo in the change log

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agogetComputedStyle for border is not implemented.
alexis.menard@openbossa.org [Tue, 3 Jan 2012 19:13:48 +0000 (19:13 +0000)]
getComputedStyle for border is not implemented.
https://bugs.webkit.org/show_bug.cgi?id=75319

Reviewed by Tony Chang.

Source/WebCore:

Implement getComputedStyle for border.

fast/css/getComputedStyle/getComputedStyle-border-shorthand.html was extented.

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

LayoutTests:

Extend existing test to cover that we return correct values.

* fast/css/getComputedStyle/getComputedStyle-border-shorthand-expected.txt:
* fast/css/getComputedStyle/getComputedStyle-border-shorthand.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103948 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoSort the WebCore XCode project file
alexis.menard@openbossa.org [Tue, 3 Jan 2012 19:03:37 +0000 (19:03 +0000)]
Sort the WebCore XCode project file

Reviewed by Dimitri Glazkov.

* WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Set tiler bounds explicitly rather than growing them
enne@google.com [Tue, 3 Jan 2012 18:59:13 +0000 (18:59 +0000)]
[chromium] Set tiler bounds explicitly rather than growing them
https://bugs.webkit.org/show_bug.cgi?id=75331

Reviewed by James Robinson.

Rather than only growing bounds, make the bounds on the tiler be set
explicitly by TiledLayerChromium. This patch now properly invalidates
regions when the bounds change.

* platform/graphics/chromium/TiledLayerChromium.cpp:
(WebCore::TiledLayerChromium::updateBounds):
(WebCore::TiledLayerChromium::prepareToUpdate):
(WebCore::TiledLayerChromium::prepareToUpdateIdle):
* platform/graphics/chromium/TiledLayerChromium.h:
* platform/graphics/chromium/cc/CCLayerTilingData.cpp:
(WebCore::CCLayerTilingData::setBounds):
(WebCore::CCLayerTilingData::bounds):
* platform/graphics/chromium/cc/CCLayerTilingData.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] Slider thumb is not centered on the track with the unico theme
commit-queue@webkit.org [Tue, 3 Jan 2012 18:55:11 +0000 (18:55 +0000)]
[GTK] Slider thumb is not centered on the track with the unico theme
https://bugs.webkit.org/show_bug.cgi?id=72802

Patch by Martin Robinson <mrobinson@igalia.com> on 2012-01-03
Reviewed by Philippe Normand.

No new tests. This only applies to certain GTK+ themes.
Existing tests cover the default GTK+ theme.

* platform/gtk/RenderThemeGtk3.cpp:
(WebCore::RenderThemeGtk::paintSliderThumb): Do not resize the
slider thumb by the trough border.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[chromium] Make tiled layer's tiler always exist
enne@google.com [Tue, 3 Jan 2012 18:48:06 +0000 (18:48 +0000)]
[chromium] Make tiled layer's tiler always exist
https://bugs.webkit.org/show_bug.cgi?id=75329

Reviewed by James Robinson.

Source/WebCore:

TiledLayerChromium has some awkward code to handle lazily creating the
tiler, including lots of null checks and shadowing the tile size.
Instead, always create the tiler up front. As values change that
invalidate all tiles, reset it rather than clear the pointer.

* platform/graphics/chromium/LayerChromium.cpp:
(WebCore::LayerChromium::setIsNonCompositedContent):
* platform/graphics/chromium/LayerChromium.h:
* platform/graphics/chromium/TiledLayerChromium.cpp:
(WebCore::TiledLayerChromium::TiledLayerChromium):
(WebCore::TiledLayerChromium::cleanupResources):
(WebCore::TiledLayerChromium::setTileSize):
(WebCore::TiledLayerChromium::setBorderTexelOption):
(WebCore::TiledLayerChromium::drawsContent):
(WebCore::TiledLayerChromium::updateCompositorResources):
(WebCore::TiledLayerChromium::setTilingOption):
(WebCore::TiledLayerChromium::pushPropertiesTo):
(WebCore::TiledLayerChromium::setIsNonCompositedContent):
(WebCore::TiledLayerChromium::invalidateRect):
(WebCore::TiledLayerChromium::protectTileTextures):
(WebCore::TiledLayerChromium::prepareToUpdate):
(WebCore::TiledLayerChromium::prepareToUpdateIdle):
(WebCore::TiledLayerChromium::needsIdlePaint):
(WebCore::TiledLayerChromium::idlePaintRect):
* platform/graphics/chromium/TiledLayerChromium.h:
* platform/graphics/chromium/cc/CCLayerTilingData.cpp:
(WebCore::CCLayerTilingData::setBorderTexelOption):
* platform/graphics/chromium/cc/CCLayerTilingData.h:
* platform/graphics/gpu/TilingData.cpp:
(WebCore::TilingData::setHasBorderTexels):
* platform/graphics/gpu/TilingData.h:

Source/WebKit/chromium:

Update test due to changing API. Add a test for changing border
texel option on TilingData.

* tests/TiledLayerChromiumTest.cpp:
(WTF::FakeTiledLayerChromium::FakeTiledLayerChromium):
* tests/TilingDataTest.cpp:
(WebCore::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] run-gtk-test should ignore files with names without a "test" or "Test" prefix
mario@webkit.org [Tue, 3 Jan 2012 17:00:13 +0000 (17:00 +0000)]
[GTK] run-gtk-test should ignore files with names without a "test" or "Test" prefix
https://bugs.webkit.org/show_bug.cgi?id=75474

Reviewed by Philippe Normand.

* Scripts/run-gtk-tests: Ignore files whose name does not start
with a "test" (for WK1's API tests) or a "Test" prefix (for WK2).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[Qt] Add UI for JavaScript Confirmation dialog in the Qt MiniBrowser
commit-queue@webkit.org [Tue, 3 Jan 2012 16:52:30 +0000 (16:52 +0000)]
[Qt] Add UI for JavaScript Confirmation dialog in the Qt MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=75472

Patch by Alexander Færøy <alexander.faeroy@nokia.com> on 2012-01-03
Reviewed by Tor Arne Vestbø.

* MiniBrowser/qt/MiniBrowser.pro:
* MiniBrowser/qt/MiniBrowser.qrc:
* MiniBrowser/qt/qml/BrowserWindow.qml:
* MiniBrowser/qt/qml/ConfirmDialog.qml: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoAdd Production configuration to gtest Xcode project
ddkilzer@apple.com [Tue, 3 Jan 2012 16:25:10 +0000 (16:25 +0000)]
Add Production configuration to gtest Xcode project
<http://webkit.org/b/75153>

Reviewed by Adam Roben.

* gtest/README.WebKit: Updated.
* gtest/xcode/Config/ProductionProject.xcconfig: Added.
* gtest/xcode/gtest.xcodeproj/project.pbxproj:
- Added Production configuration based on Release configuration.
- Made Production the default configuration.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[Qt] Add UI for JavaScript Alert dialog in the Qt MiniBrowser
commit-queue@webkit.org [Tue, 3 Jan 2012 16:20:29 +0000 (16:20 +0000)]
[Qt] Add UI for JavaScript Alert dialog in the Qt MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=75471

Patch by Alexander Færøy <alexander.faeroy@nokia.com> on 2012-01-03
Reviewed by Tor Arne Vestbø.

Add simple UI for JavaScript Alert dialogs. This patch also adds a
simple DialogButton QML type and a Dialog QML type which will be
reusable for other JavaScript dialogs.

* MiniBrowser/qt/MiniBrowser.qrc:
* MiniBrowser/qt/qml/AlertDialog.qml: Added.
* MiniBrowser/qt/qml/BrowserWindow.qml:
* MiniBrowser/qt/qml/Dialog.qml: Added.
* MiniBrowser/qt/qml/DialogButton.qml: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[Qt][WK2] MiniBrowser doesn't show url bar
commit-queue@webkit.org [Tue, 3 Jan 2012 16:06:56 +0000 (16:06 +0000)]
[Qt][WK2] MiniBrowser doesn't show url bar
https://bugs.webkit.org/show_bug.cgi?id=75470

Patch by Rafael Brandao <rafael.lobo@openbossa.org> on 2012-01-03
Reviewed by Tor Arne Vestbø.

* MiniBrowser/qt/qml/BrowserWindow.qml: Remove clipping until it is fixed
on QtDeclarative (QTBUG-23422).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] Simplify loader client WebKit2 GTK+ API
carlosgc@webkit.org [Tue, 3 Jan 2012 16:04:59 +0000 (16:04 +0000)]
[GTK] Simplify loader client WebKit2 GTK+ API
https://bugs.webkit.org/show_bug.cgi?id=74605

Reviewed by Gustavo Noronha Silva.

Source/WebKit2:

Remove WebKitWebLoaderClient from public API keeping it as a
private object. Loading API has been reduced to two signals in
WebKitWebView.

* GNUmakefile.am: Remove WebKitWebLoaderClientPrivate.h.
* UIProcess/API/gtk/WebKitDefines.h: Remove forward delcarations
that are no longer needed.
* UIProcess/API/gtk/WebKitWebLoaderClient.cpp:
(didStartProvisionalLoadForFrame): Emit
WebKitWebView::load-changed signal with Started event.
(didReceiveServerRedirectForProvisionalLoadForFrame): Emit
WebKitWebView::load-changed signal with Redirected event.
(didFailProvisionalLoadWithErrorForFrame): Emit
WebKitWebView::load-failed signal with Started event.
(didCommitLoadForFrame): Emit WebKitWebView::load-changed signal
with Committed event.
(didFinishLoadForFrame): Emit WebKitWebView::load-changed signal
with Finished event.
(didFailLoadWithErrorForFrame): Emit WebKitWebView::load-failed
signal with Committed event.
(webkit_web_loader_client_init):
(webkit_web_loader_client_class_init):
* UIProcess/API/gtk/WebKitWebLoaderClient.h: Remove public API and
move private API here.
* UIProcess/API/gtk/WebKitWebLoaderClientPrivate.h: Removed.
* UIProcess/API/gtk/WebKitWebView.cpp:
(webkitWebViewLoadFail): Default implementation of
WebKitWebView::load-failed signal to show a custom error page.
(webkit_web_view_class_init): Add load-changed and load-failed
signals.
(webkitWebViewLoadChanged): Emit WebKitWebView::load-changed with
the given event.
(webkitWebViewLoadFailed): Emit WebKitWebView::load-failed signal
with the given event, failing uri and error.
* UIProcess/API/gtk/WebKitWebView.h:
* UIProcess/API/gtk/WebKitWebViewPrivate.h:
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml: Remove loader
client section.
* UIProcess/API/gtk/docs/webkit2gtk-sections.txt: Remove loader
client symbols.
* UIProcess/API/gtk/docs/webkit2gtk.types: Remove
webkit_web_loader_client_get_type.
* UIProcess/API/gtk/tests/LoadTrackingTest.cpp:
(loadChangedCallback):
(loadFailedCallback):
(LoadTrackingTest::LoadTrackingTest):
(LoadTrackingTest::~LoadTrackingTest):
(LoadTrackingTest::provisionalLoadStarted):
(LoadTrackingTest::provisionalLoadReceivedServerRedirect):
(LoadTrackingTest::provisionalLoadFailed):
(LoadTrackingTest::loadCommitted):
(LoadTrackingTest::loadFinished):
(LoadTrackingTest::loadFailed):
* UIProcess/API/gtk/tests/LoadTrackingTest.h:
* UIProcess/API/gtk/tests/TestWebKitWebLoaderClient.cpp:
(testLoadingError):
(LoadStopTrackingTest::loadCommitted):
(LoadStopTrackingTest::loadFailed):
(testLoadCancelled):
* UIProcess/API/gtk/tests/TestWebKitWebView.cpp:
(beforeAll):
* UIProcess/API/gtk/webkit2.h: Remove loader client header.
* UIProcess/API/gtk/webkit2marshal.list:

Tools:

* gtk/generate-gtkdoc: Ignore WebKitWebLoaderClient since it's now
private API.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoUnreviewed. Rebaselined run-bindings-tests results.
haraken@chromium.org [Tue, 3 Jan 2012 15:46:42 +0000 (15:46 +0000)]
Unreviewed. Rebaselined run-bindings-tests results.

* bindings/scripts/test/JS/JSTestEventConstructor.cpp:
(WebCore::JSTestEventConstructor::destroy):
* bindings/scripts/test/JS/JSTestEventConstructor.h:
(WebCore::JSTestEventConstructor::releaseImplIfNotNull):
* bindings/scripts/test/JS/JSTestInterface.cpp:
(WebCore::JSTestInterface::destroy):
* bindings/scripts/test/JS/JSTestInterface.h:
(WebCore::JSTestInterface::releaseImplIfNotNull):
* bindings/scripts/test/JS/JSTestMediaQueryListListener.cpp:
(WebCore::JSTestMediaQueryListListener::destroy):
* bindings/scripts/test/JS/JSTestMediaQueryListListener.h:
(WebCore::JSTestMediaQueryListListener::releaseImplIfNotNull):
* bindings/scripts/test/JS/JSTestNamedConstructor.cpp:
(WebCore::JSTestNamedConstructor::destroy):
* bindings/scripts/test/JS/JSTestNamedConstructor.h:
(WebCore::JSTestNamedConstructor::releaseImplIfNotNull):
* bindings/scripts/test/JS/JSTestObj.cpp:
(WebCore::JSTestObj::destroy):
* bindings/scripts/test/JS/JSTestObj.h:
(WebCore::JSTestObj::releaseImplIfNotNull):
* bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.cpp:
(WebCore::JSTestSerializedScriptValueInterface::destroy):
* bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.h:
(WebCore::JSTestSerializedScriptValueInterface::releaseImplIfNotNull):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] Fix scrollbars size with GTK+ 3.x
carlosgc@webkit.org [Tue, 3 Jan 2012 14:15:42 +0000 (14:15 +0000)]
[GTK] Fix scrollbars size with GTK+ 3.x
https://bugs.webkit.org/show_bug.cgi?id=75467

Reviewed by Xan Lopez.

* platform/gtk/RenderThemeGtk3.cpp:
(WebCore::getStyleContext): Add the style class for the given
widget type to the GtkWidgetPath when creating GtkStyleContext
objects.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[Qt] Clean up detection of fontconfig
vestbo@webkit.org [Tue, 3 Jan 2012 12:45:17 +0000 (12:45 +0000)]
[Qt] Clean up detection of fontconfig

We decide in features.prf whether or not fontconfig is available, and
set the HAVE_FONTCONFIG define accordingly. All leaf project files
should use this define to decide whether or not to link against
fontconfig, not do their own detection.

We don't need to add link_pkgconfig to CONFIG in each project file,
as we selectivly enable that configuration in defaults_post when
needed.

Reviewed by Ossy.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agogetComputedStyle for outline is not implemented.
alexis.menard@openbossa.org [Tue, 3 Jan 2012 12:39:01 +0000 (12:39 +0000)]
getComputedStyle for outline is not implemented.
https://bugs.webkit.org/show_bug.cgi?id=75441

Reviewed by Antonio Gomes.

Source/WebCore:

Implement getComputedStyle for outline.

Test: fast/css/getComputedStyle/getComputedStyle-outline-shorthand.html

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

LayoutTests:

Add new tests to make sure we return the correct values.

* fast/css/getComputedStyle/getComputedStyle-outline-shorthand-expected.txt: Added.
* fast/css/getComputedStyle/getComputedStyle-outline-shorthand.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohttps://bugs.webkit.org/show_bug.cgi?id=75468
hclam@chromium.org [Tue, 3 Jan 2012 12:37:19 +0000 (12:37 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=75468
Suppress layout test failure due to a new test.

Unreviewed, build fix.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[Qt] Introduce qmake replace function for the active build config
vestbo@webkit.org [Tue, 3 Jan 2012 11:41:49 +0000 (11:41 +0000)]
[Qt] Introduce qmake replace function for the active build config

Reviewed by Ossy.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[Qt][WK2] tst_favIconLoad.qml crashes on debug mode
commit-queue@webkit.org [Tue, 3 Jan 2012 10:03:54 +0000 (10:03 +0000)]
[Qt][WK2] tst_favIconLoad.qml crashes on debug mode
https://bugs.webkit.org/show_bug.cgi?id=75448

Patch by Rafael Brandao <rafael.lobo@openbossa.org> on 2012-01-03
Reviewed by Kenneth Rohde Christiansen.

* UIProcess/API/qt/tests/qmltests/WebView/tst_favIconLoad.qml: Loads
"about:blank" instead of an empty url between tests to clean favicon.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] Use gdk_screen_get_monitor_workarea() when available for screenAvailableRect()
carlosgc@webkit.org [Tue, 3 Jan 2012 09:17:04 +0000 (09:17 +0000)]
[GTK] Use gdk_screen_get_monitor_workarea() when available for screenAvailableRect()
https://bugs.webkit.org/show_bug.cgi?id=75435

Reviewed by Martin Robinson.

Source/WebCore:

* platform/gtk/GtkVersioning.c:
(getScreenCurrentDesktop):
(getScreenWorkArea):
(gdk_screen_get_monitor_workarea): Implement it when GTK+ < 3.3.6.
* platform/gtk/GtkVersioning.h:
* platform/gtk/PlatformScreenGtk.cpp:
(WebCore::screenAvailableRect): Use
gdk_screen_get_monitor_workarea() instead of our own
implementation.

Source/WebKit/gtk:

* GNUmakefile.am: Make sure unit tests link to X11.

Tools:

* GNUmakefile.am: Make sure DRT links to X11.
* WebKitTestRunner/GNUmakefile.am: Make sure WTR links to X11.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[WK2][GTK] WebSettings support in MiniBrowser
philn@webkit.org [Tue, 3 Jan 2012 08:10:30 +0000 (08:10 +0000)]
[WK2][GTK] WebSettings support in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=73773

Reviewed by Martin Robinson.

Patch heavily inspired from the WebKit GtkLauncher to support
command-line WebSettings. Example: --enable-fullscreen=TRUE.
The full list of options can be consulted with --help-websettings.

* MiniBrowser/gtk/main.c:
(loadURI):
(parseOptionEntryCallback):
(isValidParameterType):
(getOptionEntriesFromWebKitSettings):
(addSettingsGroupToContext):
(main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] GtkLauncher settings improvements
philn@webkit.org [Tue, 3 Jan 2012 08:04:49 +0000 (08:04 +0000)]
[GTK] GtkLauncher settings improvements
https://bugs.webkit.org/show_bug.cgi?id=75439

Reviewed by Martin Robinson.

* GtkLauncher/main.c:
(filenameToURL): We can use NULL in C.
(isValidParameterType): Utility function to filter GParamTypes
that can be edited.
(getOptionEntriesFromWebKitWebSettings): Avoid handling
construct-only properties and use the isValidParameterType
function. Makes the code a bit cleaner.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoCheck in fixes for jsc tests following bug #75455.
barraclough@apple.com [Tue, 3 Jan 2012 07:00:20 +0000 (07:00 +0000)]
Check in fixes for jsc tests following bug #75455.

* tests/mozilla/ecma/GlobalObject/15.1.2.2-1.js:
* tests/mozilla/ecma/GlobalObject/15.1.2.2-2.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFixed a typo I made in r103871.
mitz@apple.com [Tue, 3 Jan 2012 06:56:26 +0000 (06:56 +0000)]
Fixed a typo I made in r103871.

Fixes <http://webkit.org/b/75450> Inline PDF doesn't preview correctly when attempting to print its frame

* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::drawRectToPDF): Changed USE(GC) to USE(CG).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohttps://bugs.webkit.org/show_bug.cgi?id=75452
barraclough@apple.com [Tue, 3 Jan 2012 06:56:22 +0000 (06:56 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=75452
If argument to Error is undefined, message is not set

Reviewed by Sam Weinig.

Source/JavaScriptCore:

Per section 15.11.1.1 of the spec.

* runtime/ErrorInstance.h:
(JSC::ErrorInstance::create):
(JSC::ErrorInstance::finishCreation):

LayoutTests:

Added test case.

* fast/js/native-error-prototype-expected.txt:
* fast/js/script-tests/native-error-prototype.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoREGRESSION(r103919): Remove resolve-supplemental.pl from the WebCore target in Xcode
haraken@chromium.org [Tue, 3 Jan 2012 05:45:13 +0000 (05:45 +0000)]
REGRESSION(r103919): Remove resolve-supplemental.pl from the WebCore target in Xcode
https://bugs.webkit.org/show_bug.cgi?id=75458

Reviewed by Adam Barth.

At r103919, we added resolve-supplemental.pl to WebCore.framework/Resources,
but we should have added it to the project without adding it to any target.
This patch fixes it.

No tests. No change in behavior.

* WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoES5 prohibits parseInt from supporting octal
barraclough@apple.com [Tue, 3 Jan 2012 04:32:00 +0000 (04:32 +0000)]
ES5 prohibits parseInt from supporting octal
https://bugs.webkit.org/show_bug.cgi?id=75455

Reviewed by Sam Weinig.

See sections 15.1.2.2 and annex E.

Source/JavaScriptCore:

* runtime/JSGlobalObjectFunctions.cpp:
(JSC::parseInt):

LayoutTests:

* fast/js/kde/GlobalObject-expected.txt:
* fast/js/kde/script-tests/GlobalObject.js:
    - Change test to check for correct result.
* sputnik/Implementation_Diagnostics/S15.1.2.2_D1.1-expected.txt:
* sputnik/Implementation_Diagnostics/S15.1.2.2_D1.2-expected.txt:
    - One test asserts parseInt handles octal, one asserts it does not!
      This patch switches which one passes & which fails.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agohttps://bugs.webkit.org/show_bug.cgi?id=55343
barraclough@apple.com [Tue, 3 Jan 2012 02:36:34 +0000 (02:36 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=55343
Global JSON should be configurable but isn't

Reviewed by Sam Weinig.

Source/JavaScriptCore:

* runtime/JSGlobalObject.cpp:
(JSC::JSGlobalObject::reset):
    - make JSON configurable

LayoutTests:

* ietestcenter/Javascript/11.4.1-4.a-8-expected.txt:
    - check in passing result

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoEnable the [Supplemental] IDL on GTK/GObject
haraken@chromium.org [Tue, 3 Jan 2012 01:24:50 +0000 (01:24 +0000)]
Enable the [Supplemental] IDL on GTK/GObject
https://bugs.webkit.org/show_bug.cgi?id=75411

Reviewed by Adam Barth.

This patch enables the [Supplemental] IDL on GObject bindings of GTK
by changing the build flow of bindings/gobject/GNUmakefile.am as follows.

- Previous build flow:
    foreach $idl (all IDL files) {
        generate-bindings.pl depends on $idl;
        generate-bindings.pl reads $idl;
        generate-bindings.pl generates .h and .cpp files for $idl;
    }

- New build flow (See the discussions in bug 72138 for more details):
    resolve-supplemental.pl depends on all IDL files;
    resolve-supplemental.pl reads all IDL files;
    resolve-supplemental.pl resolves the dependency of [Supplemental=XXXX];
    resolve-supplemental.pl outputs supplemental_dependency.tmp;
    foreach $idl (all IDL files) {
        generate-bindings.pl depends on $idl and supplemental_dependency.tmp;
        generate-bindings.pl reads $idl;
        generate-bindings.pl reads supplemental_dependency.tmp;
        generate-bindings.pl generates .h and .cpp files for $idl,
            including all attributes in the IDL files that are implementing $idl;
    }

Tests: Confirm that build succeeds.
       http/tests/websocket/tests/*

* GNUmakefile.am:
* bindings/gobject/GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoAdd resolve-supplemental.pl to project.pbxproj
haraken@chromium.org [Tue, 3 Jan 2012 00:57:21 +0000 (00:57 +0000)]
Add resolve-supplemental.pl to project.pbxproj
https://bugs.webkit.org/show_bug.cgi?id=75426

Reviewed by Adam Barth.

We should have added "resolve-supplemental.pl" to build scripts of AppleWebKit
when we enabled the [Supplemental] IDL on AppleWebKit at r103519.
This is a follow-up patch for it.

No tests. Just confirm that the AppleWebKit build passes.

* WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoAdd resolve-supplemental.pl to build scripts of AppleWin
haraken@chromium.org [Tue, 3 Jan 2012 00:44:29 +0000 (00:44 +0000)]
Add resolve-supplemental.pl to build scripts of AppleWin
https://bugs.webkit.org/show_bug.cgi?id=75412

Reviewed by Adam Barth.

We should have added "resolve-supplemental.pl" to build scripts of AppleWin
when we enabled the [Supplemental] IDL on AppleWin at r103519.
This is a follow-up patch for it.

No tests. Just confirm that the AppleWin build passes.

* WebCore.vcproj/MigrateScripts:
* WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFix the Windows build (again).
aestes@apple.com [Tue, 3 Jan 2012 00:39:40 +0000 (00:39 +0000)]
Fix the Windows build (again).

* Platform/CoreIPC/Connection.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFix windows build.
weinig@apple.com [Mon, 2 Jan 2012 23:59:48 +0000 (23:59 +0000)]
Fix windows build.

* Platform/CoreIPC/Connection.h:
* Platform/CoreIPC/win/ConnectionWin.cpp:
* Platform/RunLoop.h:
Missing declarations/includes of BinarySemaphore.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFix the build.
weinig@apple.com [Mon, 2 Jan 2012 23:55:40 +0000 (23:55 +0000)]
Fix the build.

* bindings/scripts/CodeGeneratorJS.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFix the Windows build after r103888.
aestes@apple.com [Mon, 2 Jan 2012 23:41:38 +0000 (23:41 +0000)]
Fix the Windows build after r103888.

* WebCore.vcproj/WebCore.vcproj: Remove CollectionCache.{cpp, h} from
the WebCore vcproj.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoREGRESSION(r100517): We're leaking many, many DOM objects!
weinig@apple.com [Mon, 2 Jan 2012 23:31:57 +0000 (23:31 +0000)]
REGRESSION(r100517): We're leaking many, many DOM objects!
https://bugs.webkit.org/show_bug.cgi?id=75451

Reviewed by Mark Rowe.

* bindings/scripts/CodeGeneratorJS.pm:
Add a temporary workaround to the problem of handle finalizers
not getting called by adding back the destructors (or rather
their replacement, destroy() functions).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoMove dispatchSentMessagesUntil out of the RunLoop class.
weinig@apple.com [Mon, 2 Jan 2012 23:08:10 +0000 (23:08 +0000)]
Move dispatchSentMessagesUntil out of the RunLoop class.
https://bugs.webkit.org/show_bug.cgi?id=75320

Reviewed by Dan Bernstein.

* Platform/CoreIPC/Connection.cpp:
(CoreIPC::Connection::SyncMessageState::waitWhileDispatchingSentWin32Messages):
* Platform/CoreIPC/Connection.h:
* Platform/CoreIPC/win/ConnectionWin.cpp:
(CoreIPC::Connection::dispatchSentMessagesUntil):
* Platform/win/RunLoopWin.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[EFL] Remove whitespace/declaration style exception
gyuyoung.kim@samsung.com [Mon, 2 Jan 2012 21:09:32 +0000 (21:09 +0000)]
[EFL] Remove whitespace/declaration style exception
https://bugs.webkit.org/show_bug.cgi?id=75424

Reviewed by David Levin.

Source/WebKit/efl:

EFL port doesn't need to avoid *whitespace/declaration* exception because of coding
style change. However, EWebLauncher and MiniBrowser need to use EFL coding style because
they are EFL application.

* WebCoreSupport/FrameLoaderClientEfl.cpp:
* ewk/ewk_frame.cpp:
(ewk_frame_feed_touch_event):
* ewk/ewk_tiled_matrix.cpp:
(ewk_tile_matrix_invalidate):
* ewk/ewk_view.cpp:

Tools:

Remove *whitespace/eclaration* style rule. However, EWebLauncher and MiniBrowser are EFL simple
application. So, they still need to use EFL coding style.

* Scripts/webkitpy/style/checker.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[mac] Remove -[WebView setHoverFeedbackSuspended:]
mitz@apple.com [Mon, 2 Jan 2012 19:18:01 +0000 (19:18 +0000)]
[mac] Remove -[WebView setHoverFeedbackSuspended:]
https://bugs.webkit.org/show_bug.cgi?id=75406

Reviewed by Tim Hatcher.

* WebView/WebHTMLView.mm:
(-[WebHTMLView _updateMouseoverWithEvent:]): Removed check for suspended hover updates.
* WebView/WebHTMLViewInternal.h: Removed -_hoverFeedbackSuspendedChanged.
* WebView/WebView.mm:
(-[WebView setHoverFeedbackSuspended:]): Left empty implementation to support versions of
Safari that call this.
(-[WebView isHoverFeedbackSuspended]): Left implementation that always returns NO to support
versions of Safari that call this.
* WebView/WebViewData.h: Removed hoverFeedbackSuspended ivar.
* WebView/WebViewPrivate.h: Removed method definitions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoBuild fix.
mitz@apple.com [Mon, 2 Jan 2012 18:04:52 +0000 (18:04 +0000)]
Build fix.

* Platform/CoreIPC/mac/ConnectionMac.cpp:
* Platform/mac/SharedMemoryMac.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago2012-01-02 Nikolas Zimmermann <nzimmermann@rim.com>
zimmermann@webkit.org [Mon, 2 Jan 2012 16:26:41 +0000 (16:26 +0000)]
2012-01-02  Nikolas Zimmermann  <nzimmermann@rim.com>

        Not reviewed. Fix build on Lion, by including mach_error.h in two places that need it.

        * PluginProcess/mac/PluginProcessMainMac.mm:
        * WebProcess/mac/WebProcessMainMac.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoFix chromium mac build after r103905
caio.oliveira@openbossa.org [Mon, 2 Jan 2012 15:39:10 +0000 (15:39 +0000)]
Fix chromium mac build after r103905
https://bugs.webkit.org/show_bug.cgi?id=75436

Reviewed by Csaba Osztrogonác.

Adding parenthesis to disambiguate because clang gives error
"expression is not assignable".

* rendering/svg/RenderSVGResourceGradient.cpp:
(WebCore::RenderSVGResourceGradient::applyResource):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] Add a way to change web view settings in MiniBrowser
carlosgc@webkit.org [Mon, 2 Jan 2012 14:56:35 +0000 (14:56 +0000)]
[GTK] Add a way to change web view settings in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=71568

Reviewed by Philippe Normand.

.:

* GNUmakefile.am: Initialize minibrowser_built_sources.

Tools:

Add a preferences button to the toolbar that shows a dialog to
change settings of the current web view. It uses a GtkTreeView
with a custom cell renderer to support multiple types of values.

* MiniBrowser/gtk/BrowserCellRendererVariant.c: Added.
(browserCellRendererVariantFinalize):
(browserCellRendererVariantGetProperty):
(browserCellRendererVariantSetModeForValue):
(browserCellRendererVariantSetProperty):
(browserCellRendererVariantGetRendererForValue):
(browserCellRendererVariantCellRendererTextEdited):
(browserCellRendererVariantCellRendererSpinEdited):
(browserCellRendererVariantCellRendererActivate):
(browserCellRendererVariantCellRendererRender):
(browserCellRendererVariantCellRendererStartEditing):
(browserCellRendererVariantCellRendererGetPreferredWidth):
(browserCellRendererVariantCellRendererGetPreferredHeight):
(browserCellRendererVariantCellRendererGetPreferredWidthForHeight):
(browserCellRendererVariantCellRendererGetPreferredHeightForWidth):
(browserCellRendererVariantCellRendererGetAlignedArea):
(browser_cell_renderer_variant_init):
(browser_cell_renderer_variant_class_init):
(browser_cell_renderer_variant_new):
* MiniBrowser/gtk/BrowserCellRendererVariant.h: Added.
* MiniBrowser/gtk/BrowserSettingsDialog.c: Added.
(cellRendererChanged):
(browserSettingsDialogSetProperty):
(browser_settings_dialog_init):
(browserSettingsDialogConstructed):
(browser_settings_dialog_class_init):
(browser_settings_dialog_new):
* MiniBrowser/gtk/BrowserSettingsDialog.h: Added.
* MiniBrowser/gtk/BrowserWindow.c:
(settingsCallback):
(browser_window_init):
* MiniBrowser/gtk/GNUmakefile.am:
* MiniBrowser/gtk/browser-marshal.list: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years agoUse HashMap<OwnPtr> in RenderSVGResourceGradient
caio.oliveira@openbossa.org [Mon, 2 Jan 2012 14:07:53 +0000 (14:07 +0000)]
Use HashMap<OwnPtr> in RenderSVGResourceGradient
https://bugs.webkit.org/show_bug.cgi?id=75364

Reviewed by Daniel Bates.

* rendering/svg/RenderSVGResourceGradient.cpp:
(WebCore::RenderSVGResourceGradient::removeAllClientsFromCache):
(WebCore::RenderSVGResourceGradient::removeClientFromCache):
(WebCore::RenderSVGResourceGradient::applyResource): avoid looking
up the hash twice by using HashMap::add().
* rendering/svg/RenderSVGResourceGradient.h: rename m_gradient to m_gradientMap.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

9 years ago[GTK] Implement create/ready-to-show/close signals in MiniBrowser
carlosgc@webkit.org [Mon, 2 Jan 2012 13:39:21 +0000 (13:39 +0000)]
[GTK] Implement create/ready-to-show/close signals in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=74711

Reviewed by Philippe Normand.

* MiniBrowser/gtk/BrowserWindow.c:
(webViewClose): Destroy the window.
(webViewReadyToShow): Use WindowProperties of the new view to
setup and show the window.
(webViewCreate): Create a new view and add it to a browser window.
(browser_window_init):
(browserWindowConstructed): Connect to WebKitWebView::create signal.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@103904 268f45cc-cd09-0410-ab3c-d52691b4dbfc