Timeline


and

04/19/10:

23:39 Changeset [57874] by eric@webkit.org

2010-04-19 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Expand testatkroles to test ATK form roles.

  • tests/testatkroles.c: (test_webkit_atk_get_role_check_box): (test_webkit_atk_get_role_entry): (test_webkit_atk_get_role_label): (test_webkit_atk_get_role_listbox): (test_webkit_atk_get_role_password_text): (test_webkit_atk_get_role_push_button): (test_webkit_atk_get_role_radio_button): (main):
23:30 Changeset [57873] by eric@webkit.org

2010-04-19 Leandro Pereira <leandro@profusion.mobi>

Reviewed by David Hyatt.

Add missing dummy implementations in PluginPackageNone and PluginViewNone.
http://webkit.org/b/37478

  • plugins/PluginPackageNone.cpp: (WebCore::PluginPackage::NPVersion): Add dummy implementation.
  • plugins/PluginViewNone.cpp: (WebCore::PluginView::handleFocusInEvent): Add dummy implementation. (WebCore::PluginView::handleFocusOutEvent): Add dummy implementation.
23:18 Changeset [57872] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Unreviewed, build fix.

Turn off some unit tests for now - the new-run-webkit-tests dryrun
tests for chromium won't work if you don't have a chromium checkout.

https://bugs.webkit.org/show_bug.cgi?id=37841

  • Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
23:00 Changeset [57871] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Unreviewed, build fix.

(Re-)add dryrun.py; this was renamed from passing.py in the previous
CL but apparently somehow didn't get checked in.

https://bugs.webkit.org/show_bug.cgi?id=37841

  • Scripts/webkitpy/layout_tests/port/dryrun.py: Added.
22:41 Changeset [57870] by hamaji@chromium.org

2010-04-19 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by David Levin.

check-webkit-style: exits when encountering a deleted file
https://bugs.webkit.org/show_bug.cgi?id=37122

This reverts the quick fix done by r57119 and makes check_patch
not call check_file for deleted files.

Also this change fixes the behavior for "-", which should mean
stdin. Before this change, the style checker just ignored "-"
with a warning message.

  • Scripts/webkitpy/style/checker.py:
  • Scripts/webkitpy/style/checker_unittest.py:
22:10 Changeset [57869] by dbates@webkit.org

2010-04-19 Daniel Bates <dbates@rim.com>

No review, rolling out 57868.
http://trac.webkit.org/changeset/57868
https://bugs.webkit.org/show_bug.cgi?id=37748

Sheriffbot wasn't very inspirational after this patch.
Instead, he was silent when you said hi :-(. Rolling
out this patch so that I can debug/test this some more.

  • Scripts/webkitpy/common/net/bugzilla.py:
  • Scripts/webkitpy/common/net/bugzilla_unittest.py:
  • Scripts/webkitpy/tool/bot/irc_command.py:
  • Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
  • Scripts/webkitpy/tool/mocktool.py:
21:34 Changeset [57868] by dbates@webkit.org

2010-04-19 Daniel Bates <dbates@rim.com>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=37748

Make Sheriffbot more inspirational.

  • Scripts/webkitpy/common/net/bugzilla.py:
  • Scripts/webkitpy/common/net/bugzilla_unittest.py:
  • Scripts/webkitpy/tool/bot/irc_command.py:
  • Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
  • Scripts/webkitpy/tool/mocktool.py:
21:22 Changeset [57867] by simon.fraser@apple.com

2010-04-19 Simon Fraser <simon.fraser@apple.com>

Reviewed by Geoff Garen.

Clean up RenderPart/RenderPartObject/RenderFrame/RenderEmbeddedObject
https://bugs.webkit.org/show_bug.cgi?id=37741

Move m_hasFallbackContent from RenderPart to RenderEmbeddedObject,
since it's only used for <object> fallback.

  • rendering/RenderEmbeddedObject.cpp: (WebCore::RenderEmbeddedObject::RenderEmbeddedObject):
  • rendering/RenderEmbeddedObject.h: (WebCore::RenderEmbeddedObject::hasFallbackContent):
  • rendering/RenderPart.cpp: (WebCore::RenderPart::RenderPart):
  • rendering/RenderPart.h:
21:02 Changeset [57866] by simon.fraser@apple.com

2010-04-19 Simon Fraser <simon.fraser@apple.com>

Reviewed by Darin Adler.

Clean up RenderPart/RenderPartObject/RenderFrame/RenderEmbeddedObject
https://bugs.webkit.org/show_bug.cgi?id=37741

Add a new renderer for iframes, named RenderIFrame. Add a new shared base class
between this and RenderFrame, called RenderFrameBase (following the existing HTMLFrameElementBase),
and move code from RenderPart and RenderPartObject into these new classes.


There should be no functionality difference with this change, so no new tests.


Fixing up renderer names in the layout tests will be done in a later pass.


Fix build systems to include the new files.


  • Android.mk:
  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • html/HTMLIFrameElement.cpp: (WebCore::HTMLIFrameElement::createRenderer):
  • rendering/RenderEmbeddedObject.h:
  • rendering/RenderFrame.cpp: (WebCore::RenderFrame::RenderFrame):
  • rendering/RenderFrame.h:
  • rendering/RenderFrameBase.cpp: Added. (WebCore::RenderFrameBase::RenderFrameBase): (WebCore::RenderFrameBase::layoutWithFlattening):
  • rendering/RenderFrameBase.h: Added.
  • rendering/RenderIFrame.cpp: Added. (WebCore::RenderIFrame::RenderIFrame): (WebCore::RenderIFrame::calcHeight): (WebCore::RenderIFrame::calcWidth): (WebCore::RenderIFrame::flattenFrame): (WebCore::RenderIFrame::layout):
  • rendering/RenderIFrame.h: Added. (WebCore::RenderIFrame::renderName):
  • rendering/RenderPart.cpp:
  • rendering/RenderPart.h:
  • rendering/RenderPartObject.cpp:
  • rendering/RenderPartObject.h:
20:59 Changeset [57865] by dumi@chromium.org

Bindings clean-up.
https://bugs.webkit.org/show_bug.cgi?id=37833

Reviewed by Geoffrey Garen.

Move some WebSQLDatabases logic out of the bindings into
DOMWindow.cpp where it should be.

  • bindings/js/JSDOMWindowCustom.cpp:

(WebCore::JSDOMWindow::openDatabase):

  • bindings/v8/custom/V8DOMWindowCustom.cpp:

(WebCore::V8DOMWindow::openDatabaseCallback):

  • page/DOMWindow.cpp:

(WebCore::DOMWindow::openDatabase):

19:59 Changeset [57864] by kevino@webkit.org

Fix the Mac builders for now by restoring the keepAlive function.

19:10 Changeset [57863] by kevino@webkit.org

Build fixes for wx with latest trunk.

18:43 Changeset [57862] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed. Update chromium test expectations.

  • platform/chromium/test_expectations.txt:
17:56 Changeset [57861] by jberlin@webkit.org

Remove an extra space in the getComputedStyle-with-psuedo-element test that was causing it to fail on the Chromium bots, and remove the platform-specific results that were landed in the meantime.

Reviewed by Dan Bernstein.

  • fast/css/getComputedStyle/getComputedStyle-with-pseudo-element-expected.txt:
  • fast/css/getComputedStyle/getComputedStyle-with-pseudo-element.html:
  • platform/chromium-linux/fast/css/getComputedStyle/getComputedStyle-with-pseudo-element-expected.txt: Removed.
  • platform/chromium-win/fast/css/getComputedStyle/getComputedStyle-with-pseudo-element-expected.txt: Removed.
17:53 Changeset [57860] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Adam Barth.

new-run-webkit-tests - repurpose the "Passing" port as "Dryrun" port
that can be used to test platforms other than the one you are running
on. This can be useful for checking baselines and testing code
coverage.

Note that running the code on the "wrong" port requires each
port-specific implementation to actually not require any
platform-specific python code (e.g., the chromium-win port must
test for the existence of windows functions before calling them).

https://bugs.webkit.org/show_bug.cgi?id=37782

  • Scripts/webkitpy/layout_tests/port/chromium_win.py:
  • Scripts/webkitpy/layout_tests/port/dryrun.py: Renamed from WebKitTools/Scripts/webkitpy/layout_tests/port/passing.py.
  • Scripts/webkitpy/layout_tests/port/factory.py:
  • Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
17:34 Changeset [57859] by mrowe@apple.com

Touch WebKit.idl to force generated interfaces to be rebuilt. Boo.

17:33 Changeset [57858] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

new-run-webkit-tests: add --build (default) and --no-build
options to make that step optional. This flag modifies what happens
in port.check_build().

https://bugs.webkit.org/show_bug.cgi?id=37786

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
  • Scripts/webkitpy/layout_tests/port/chromium.py:
  • Scripts/webkitpy/layout_tests/port/webkit.py:
17:24 Changeset [57857] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed. Update chromium test expectations.

  • platform/chromium/test_expectations.txt:
17:16 Changeset [57856] by mrowe@apple.com

Roll out r57823.

It adds methods to IWebFrameLoadDelegatePrivate2, a delegate interface that is implemented
by shipping clients of WebKit on Windows. Adding methods to a delegate interface in this
manner breaks ABI compatibility and leads to crashes or other unexpected behavior when the
methods are used.

16:35 Changeset [57855] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Unreviewed, build fix.

new-run-webkit-tests - fix a typo in r57480 that caused us to stop
logging the actual list of unexpected results.

https://bugs.webkit.org/show_bug.cgi?id=37831

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
16:30 Changeset [57854] by andersca@apple.com

Fix build.

  • WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:

(WebKit::WebFrameLoaderClient::dispatchDidChangeIcons):

  • WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
16:26 Changeset [57853] by barraclough@apple.com

This broke windows.

Reviewed by NOBODY (rolling out r57829).

JavaScriptCore:

  • API/APIShims.h:

(JSC::APIEntryShimWithoutLock::APIEntryShimWithoutLock):
(JSC::APIEntryShimWithoutLock::~APIEntryShimWithoutLock):
(JSC::APICallbackShim::APICallbackShim):
(JSC::APICallbackShim::~APICallbackShim):

(JSC::checkSyntax):
(JSC::evaluate):

  • runtime/Identifier.cpp:

(JSC::Identifier::remove):
(JSC::Identifier::checkCurrentIdentifierTable):
(JSC::createIdentifierTableSpecificCallback):
(JSC::createIdentifierTableSpecific):

  • runtime/Identifier.h:

(JSC::ThreadIdentifierTableData::ThreadIdentifierTableData):
(JSC::defaultIdentifierTable):
(JSC::setDefaultIdentifierTable):
(JSC::currentIdentifierTable):
(JSC::setCurrentIdentifierTable):
(JSC::resetCurrentIdentifierTable):

  • runtime/InitializeThreading.cpp:

(JSC::initializeThreadingOnce):

  • runtime/JSGlobalData.cpp:

(JSC::JSGlobalData::create):

  • wtf/WTFThreadData.cpp: Removed.
  • wtf/WTFThreadData.h: Removed.

JavaScriptGlue:

  • ForwardingHeaders/wtf/WTFThreadData.h: Removed.
  • JSUtils.cpp:

(JSGlueAPIEntry::JSGlueAPIEntry):
(JSGlueAPIEntry::~JSGlueAPIEntry):
(JSGlueAPICallback::JSGlueAPICallback):
(JSGlueAPICallback::~JSGlueAPICallback):

WebCore:

  • ForwardingHeaders/wtf/WTFThreadData.h: Removed.
  • platform/ThreadGlobalData.cpp:

(WebCore::ThreadGlobalData::ThreadGlobalData):
(WebCore::ThreadGlobalData::~ThreadGlobalData):

  • platform/ThreadGlobalData.h:

(WebCore::ThreadGlobalData::atomicStringTable):

  • platform/text/AtomicString.cpp:

(WebCore::stringTable):

16:24 Changeset [57852] by mrowe@apple.com

Build fix.

  • platform/graphics/mac/GraphicsContext3DMac.cpp:
16:10 Changeset [57851] by andersca@apple.com

Reviewed and landed by Anders Carlsson.

Patch by Douglas Gregor <dgregor@apple.com> on 2010-04-19

  • runtime/UStringImpl.h:

Fix class/struct declaration mismatches.

16:05 Changeset [57850] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Adam Barth.

From a patch by Tor Arne Vestbo <tor.arne.vestbo@nokia.com>

new-run-webkit-tests: make the retry step more explicit
https://bugs.webkit.org/show_bug.cgi?id=37606

It might be confusing to see the test and percent counters
reset without any notice of what's going on, so we make the
message that a retry-run is started explicit.

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
15:59 Changeset [57849] by weinig@apple.com

Add support for opening a new window (File->New Window) to
Windows MiniBrowser. Accelerator doesn't work.

Reviewed by Anders Carlsson.

  • MiniBrowser/win/BrowserView.cpp:

(createNewPage): Use BrowserView::create.

  • MiniBrowser/win/BrowserWindow.cpp:

(BrowserWindow::onCommand): Respond to ID_FILE_NEW_WINDOW
by creating a new window.

  • MiniBrowser/win/BrowserWindow.h:

(BrowserWindow::create): Added. Don't allow creating
BrowserWindows on the stack by making constructor
private and exposing the create function.

  • MiniBrowser/win/MiniBrowser.cpp:

(MiniBrowser::createNewWindow): Move new window creation
logic here.

  • MiniBrowser/win/MiniBrowser.h:
  • MiniBrowser/win/MiniBrowser.rc:
  • MiniBrowser/win/main.cpp:

(_tWinMain): Use the new MiniBrowser::createNewWindow().

15:54 QtWebKitGraphics edited by noam.rosenthal@nokia.com
(diff)
15:27 Changeset [57848] by cfleizach@apple.com

2010-04-19 Chris Fleizach <cfleizach@apple.com>

Reviewed by Beth Dakin.

AX: aria-haspopup needs to be exposed
https://bugs.webkit.org/show_bug.cgi?id=37808

Test: platform/mac/accessibility/element-haspopup.html

  • accessibility/AccessibilityObject.h: (WebCore::AccessibilityObject::ariaHasPopup):
  • accessibility/AccessibilityRenderObject.cpp: (WebCore::AccessibilityRenderObject::ariaHasPopup): (WebCore::AccessibilityRenderObject::determineAriaRoleAttribute):
  • accessibility/AccessibilityRenderObject.h:
  • accessibility/mac/AccessibilityObjectWrapper.mm: (-[AccessibilityObjectWrapper additionalAccessibilityAttributeNames]): (-[AccessibilityObjectWrapper accessibilityAttributeValue:]):

2010-04-19 Chris Fleizach <cfleizach@apple.com>

Reviewed by Beth Dakin.

AX: aria-haspopup needs to be exposed
https://bugs.webkit.org/show_bug.cgi?id=37808

  • platform/mac/accessibility/element-haspopup-expected.txt: Added.
  • platform/mac/accessibility/element-haspopup.html: Added.

2010-04-19 Chris Fleizach <cfleizach@apple.com>

Reviewed by Beth Dakin.

AX: aria-haspopup needs to be exposed
https://bugs.webkit.org/show_bug.cgi?id=37808

  • DumpRenderTree/mac/AccessibilityUIElementMac.mm: (AccessibilityUIElement::hasPopup):
15:09 Changeset [57847] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

Submit a better workaround for r57806 than the one in r57831 - log
an error and exit if you try to run new-run-webkit-tests with --use-drt
on Windows.

https://bugs.webkit.org/show_bug.cgi?id=37822

  • Scripts/webkitpy/layout_tests/port/chromium.py:
15:03 Changeset [57846] by jesus@webkit.org

Unreviewed.

Just adding myself as a committer.

  • Scripts/webkitpy/common/config/committers.py:
14:59 Changeset [57845] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

Add slightly better logging to the websocket python wrapper script,
including a --verbose flag for debug output.

https://bugs.webkit.org/show_bug.cgi?id=37233

  • Scripts/webkitpy/layout_tests/port/websocket_server.py:
  • Scripts/new-run-webkit-websocketserver:
14:57 Changeset [57844] by andersca@apple.com

Fix a crash when rendering <select> elements with WebKit2.

Reviewed by Sam Weinig.

  • rendering/RenderThemeMac.mm:

(WebCore::RenderThemeMac::paintMenuList):
Set the current NSGraphicsContext before calling out to AppKit, otherwise the current graphics context
could point to a CGContext whose memory has been freed.

14:56 Changeset [57843] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

new-run-webkit-tests: add a way (--print-unexpected-results) to
(re-)print out the tests that produced unexpected results in the
last run. Also add a way (--retry-unexpected-results) to
automatically re-test them.

https://bugs.webkit.org/show_bug.cgi?id=37783

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14:48 Changeset [57842] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed, move chromium baselines for fast/url around.

The desired behavior is that chromium-win use one set of baselines and
chromium-mac/chromium-linux use another, however to do we have to
duplicate the mac+linux results into chromium-mac and chromium-linux.

  • platform/chromium-linux/fast/url/file-expected.txt: Added.
  • platform/chromium-linux/fast/url/relative-unix-expected.txt: Added.
  • platform/chromium-linux/fast/url/relative-win-expected.txt: Added.
  • platform/chromium-mac/fast/url/file-expected.txt: Added.
  • platform/chromium-mac/fast/url/relative-unix-expected.txt: Added.
  • platform/chromium-mac/fast/url/relative-win-expected.txt: Added.
  • platform/chromium/fast/url/file-expected.txt: Removed.
  • platform/chromium/fast/url/relative-unix-expected.txt: Removed.
  • platform/chromium/fast/url/relative-win-expected.txt: Removed.
14:41 Changeset [57841] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

Restructure the logging in new-run-webkit-tests so that many of log
messages that were logged to the MeteredStream also get logged in
--verbose mode.

https://bugs.webkit.org/show_bug.cgi?id=37780

  • Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14:39 Changeset [57840] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

new-run-webkit-tests: add 'summary' and 'unexpected-results' options
to the --log flag. Also add a 'progress' flag to enable the regular
progress bar (as opposed to 'detailed-progress', which enables the
dots). Also add a 'nothing' flag to allow you to be explicit that
you don't want any logging on the command line.

The default is
'--log detailed-progress,summary,unexpected,unexpected-results'

(The default logging is unchanged by this patch, this just makes things
properly configurable).

Note that 'nothing' doesn't work properly yet; I need a couple other
patches to land to avoid rewriting things several different ways.

https://bugs.webkit.org/show_bug.cgi?id=37785

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14:15 Changeset [57839] by dglazkov@chromium.org

Manipulating document fragment members while adding it to tree may result in loss of tree integrity.
https://bugs.webkit.org/show_bug.cgi?id=36031

Reviewed by Darin Adler.

WebCore:

Changes the logic of appending/inserting document fragment to first stashing all of its children
to a vector, then processing the vector. This avoids ghastliness that would be caused by mutation
events mucking with the document fragment while it's being appended/inserted.

Test: fast/dom/Node/fragment-mutation.html

  • dom/ContainerNode.cpp:

(WebCore::targetNodes): Added method to populate a vector of nodes (targets) to be used in

inserting/appending operation.

(WebCore::ContainerNode::insertBefore): Changed to use vector-based iteration.
(WebCore::ContainerNode::appendChild): Ditto.

  • dom/Node.cpp:

(WebCore::Node::checkReplaceChild): Cleaned up comments.
(WebCore::Node::checkAddChild): Ditto.

LayoutTests:

These tests study various side-effects of DOM mutation events fired during appending/inserting
document fragments.

  • fast/dom/Node/fragment-mutation-expected.txt: Added.
  • fast/dom/Node/fragment-mutation.html: Added.
14:10 Changeset [57838] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed - yet more rebaselines and expectation updates.

  • platform/chromium-win/fast/url/file-expected.txt: Added.
  • platform/chromium-win/fast/url/relative-unix-expected.txt: Added.
  • platform/chromium-win/fast/url/relative-win-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
14:07 Changeset [57837] by ggaren@apple.com

Checked in these tests I wrote becuase Balazs Kelemen wanted to use them.

Reviewed by Beth Dakin.

  • tests/perf: Added.
  • tests/perf/bench-allocate-nonretained.js: Added.
  • tests/perf/bench-allocate-retained.js: Added.
13:56 Changeset [57836] by kenneth@webkit.org

Make the URL change on committed load.

Reviewed by Simon Hausmann.

  • QtLauncher/mainwindow.cpp:

(MainWindow::buildUI):
(MainWindow::setAddressUrl):

  • QtLauncher/mainwindow.h:
13:51 Changeset [57835] by barraclough@apple.com

Reviewed by NOBODY (windows build fix).

13:50 Changeset [57834] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed. More chromium rebaselines+expectation updates.

  • platform/chromium-linux/fast/css/getComputedStyle: Added.
  • platform/chromium-linux/fast/css/getComputedStyle/getComputedStyle-with-pseudo-element-expected.txt: Added.
  • platform/chromium-mac/media/video-src-change-expected.txt:
  • platform/chromium-mac/media/video-src-invalid-remove-expected.txt: Removed.
  • platform/chromium-win/fast/css/getComputedStyle/getComputedStyle-with-pseudo-element-expected.txt: Added.
  • platform/chromium-win/media/video-src-change-expected.txt:
  • platform/chromium-win/media/video-src-invalid-remove-expected.txt: Removed.
  • platform/chromium/test_expectations.txt:
13:35 Changeset [57833] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Eric Seidel.

Fix a typo in the rebaselining tool that causes us to use "debug"
instead of "Debug" in a directory path, which fails on platforms with
case-sensitive filesystems.

https://bugs.webkit.org/show_bug.cgi?id=37819

  • Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
13:20 Changeset [57832] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed build fix.

Update chromium expectations

  • platform/chromium/test_expectations.txt:
13:17 Changeset [57831] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Unreviewed, build fix.

Fix breakage of chromium-win canary bots caused by r57806. That patch
introduced the option of using Chrome's new port of DumpRenderTree,
but unfortunately that port relies on the webkit.py class
implementation which uses non-blocking I/O that isn't available on
Windows. This patch turns off that option and doesn't import the
class if we're running on Windows.

https://bugs.webkit.org/show_bug.cgi?id=37817

  • Scripts/webkitpy/layout_tests/port/chromium.py:
13:12 Changeset [57830] by eric.carlson@apple.com

2010-04-19 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

Fix regression introduced in r57820.

  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::loadNextSourceChild): Create a new MediaPlayer instead of just setting a URL on the one used for the previous <source> element. This restores the behavior prior to the changes for https://bugs.webkit.org/show_bug.cgi?id=37728.
13:05 Changeset [57829] by barraclough@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37745
Move string uniquing tables to (new) WTFThreadData class.

Reviewed by Sam Weinig.

Remove AtomicString's dependency on ThreadGlobalData so that we can move
WebCore's string classes up to WTF.

JavaScriptCore:

WTFThreadData.cpp/.h are based on ThreadGlobalData from WebCore.
Moved JSC & WebCore's string uniquing tables to this class.

This patch introduces a temporary layering violation in providing forward
declarations of classes from JSC and WTF; this will be resolved as we move
more string code up to WTF.

  • API/APIShims.h:

(JSC::APIEntryShimWithoutLock::APIEntryShimWithoutLock):
(JSC::APIEntryShimWithoutLock::~APIEntryShimWithoutLock):
(JSC::APICallbackShim::APICallbackShim):
(JSC::APICallbackShim::~APICallbackShim):

(JSC::checkSyntax):
(JSC::evaluate):

  • runtime/Identifier.cpp:

(JSC::Identifier::remove):
(JSC::Identifier::checkCurrentIdentifierTable):

  • runtime/Identifier.h:
  • runtime/InitializeThreading.cpp:

(JSC::initializeThreadingOnce):

  • runtime/JSGlobalData.cpp:

(JSC::JSGlobalData::create):

  • wtf/WTFThreadData.cpp: Copied from WebCore/platform/ThreadGlobalData.cpp.

(WTF::WTFThreadData::WTFThreadData):
(WTF::WTFThreadData::~WTFThreadData):

  • wtf/WTFThreadData.h: Copied from WebCore/platform/ThreadGlobalData.h.

(WTF::WTFThreadData::atomicStringTable):
(WTF::WTFThreadData::initializeIdentifierTable):
(WTF::WTFThreadData::currentIdentifierTable):
(WTF::WTFThreadData::setCurrentIdentifierTable):
(WTF::WTFThreadData::resetCurrentIdentifierTable):
(WTF::wtfThreadData):

JavaScriptGlue:

  • ForwardingHeaders/wtf/WTFThreadData.h: Added.
  • JSUtils.cpp: Update

(JSGlueAPIEntry::JSGlueAPIEntry):
(JSGlueAPIEntry::~JSGlueAPIEntry):
(JSGlueAPICallback::JSGlueAPICallback):
(JSGlueAPICallback::~JSGlueAPICallback):

WebCore:

  • ForwardingHeaders/wtf/WTFThreadData.h: Added.
  • platform/ThreadGlobalData.cpp: Remove m_atomicStringTable, all wtfThreadData() to ensure threadsafely initialized.

(WebCore::ThreadGlobalData::ThreadGlobalData):
(WebCore::ThreadGlobalData::~ThreadGlobalData):

  • platform/ThreadGlobalData.h: Remove m_atomicStringTable.

(WebCore::ThreadGlobalData::eventNames):

  • platform/text/AtomicString.cpp:

(WebCore::AtomicStringTable::create):
(WebCore::AtomicStringTable::table):
(WebCore::AtomicStringTable::destroy):
(WebCore::stringTable): Access the AtomicStringTable on wtfThreadData() rather then threadGlobalData().

13:00 CommitterTips edited by jesus@webkit.org
There is no more "--no-build" option anymore. (diff)
12:36 Changeset [57828] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Reviewed by abarth.

Fix a typo

  • Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
12:29 Changeset [57827] by jamesr@google.com

2010-04-19 James Robinson <jamesr@chromium.org>

Unreviewed build fix for Chromium

Add baselines for new platform-specific tests.

  • platform/chromium/fast/url: Added.
  • platform/chromium/fast/url/file-expected.txt: Added.
  • platform/chromium/fast/url/relative-expected.txt: Added.
  • platform/chromium/fast/url/relative-unix-expected.txt: Added.
  • platform/chromium/fast/url/relative-win-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
12:13 Changeset [57826] by adachan@apple.com

Build fix: wrap Settings::setLocalStorageQuota() and Settings::setSessionStorageQuota()
in #if ENABLE(DOM_STORAGE).

  • page/Settings.cpp:

(WebCore::Settings::Settings):

  • page/Settings.h:
12:13 Changeset [57825] by aroben@apple.com

Fix run-webkit-tests when there are spaces in the path

Fixes <http://webkit.org/b/37809>

Reviewed by Adam Barth.

  • Scripts/run-webkit-tests: Use an "indirect object" to specify the

path to the harness to exec(). According to perldoc, this usage will
prohibit perl from parsing the arguments to exec() via the shell,
which would incorrectly split paths with spaces in them, etc.

12:01 Changeset [57824] by dglazkov@chromium.org

Touch WebKit.idl to force a rebuild of Interfaces.vcproj

11:49 Changeset [57823] by dglazkov@chromium.org

WebCore: Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

Test: fast/dom/icon-url-property.html

  • dom/Document.cpp:

(WebCore::Document::setIconURL):

  • loader/DocumentLoader.cpp:

(WebCore::DocumentLoader::setIconURL):

  • loader/DocumentLoader.h:

(WebCore::DocumentLoader::iconURL):

  • loader/EmptyClients.h:

(WebCore::EmptyFrameLoaderClient::dispatchDidChangeIcons):

  • loader/FrameLoader.cpp:

(WebCore::FrameLoader::setIconURL):
(WebCore::FrameLoader::didChangeIcons):

  • loader/FrameLoader.h:
  • loader/FrameLoaderClient.h:

WebKit/chromium: Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • public/WebFrameClient.h:

(WebKit::WebFrameClient::didChangeIcons):

  • src/FrameLoaderClientImpl.cpp:

(WebKit::FrameLoaderClientImpl::dispatchDidChangeIcons):

  • src/FrameLoaderClientImpl.h:

WebKit/gtk: Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • WebCoreSupport/FrameLoaderClientGtk.cpp:

(WebKit::FrameLoaderClient::dispatchDidChangeIcons):

  • WebCoreSupport/FrameLoaderClientGtk.h:

WebKit/mac: Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • WebCoreSupport/WebFrameLoaderClient.h:
  • WebCoreSupport/WebFrameLoaderClient.mm:

(WebFrameLoaderClient::dispatchDidChangeIcons):

WebKit/qt: Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • WebCoreSupport/FrameLoaderClientQt.cpp:

(WebCore::FrameLoaderClientQt::dispatchDidChangeIcons):
(WebCore::FrameLoaderClientQt::didChangeTitle):

  • WebCoreSupport/FrameLoaderClientQt.h:

WebKit/win: Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • Interfaces/IWebFrameLoadDelegatePrivate2.idl:
  • WebCoreSupport/WebFrameLoaderClient.cpp:

(WebFrameLoaderClient::dispatchDidChangeIcons):

  • WebCoreSupport/WebFrameLoaderClient.h:
  • WebFrame.cpp:

(WebFrame::didChangeIcons):

  • WebFrame.h:

WebKitTools: Add test support for icon changes.

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • DumpRenderTree/LayoutTestController.cpp:

(LayoutTestController::LayoutTestController):
(dumpIconChangesCallback):
(LayoutTestController::staticFunctions):

  • DumpRenderTree/LayoutTestController.h:

(LayoutTestController::dumpIconChanges):
(LayoutTestController::setDumpIconChanges):

  • DumpRenderTree/win/FrameLoadDelegate.cpp:

(FrameLoadDelegate::didChangeIcons):

  • DumpRenderTree/win/FrameLoadDelegate.h:
  • WinLauncher/WinLauncher.h:

(WinLauncherWebHost::didChangeIcons):

LayoutTests: Add test for icon change notifications.

Patch by Dave Moore <davemoore@chromium.org> on 2010-04-19
Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • fast/dom/icon-url-property-expected.txt: Added.
  • fast/dom/icon-url-property.html: Added.
  • Skipped on all platforms but Win
11:37 Changeset [57822] by adachan@apple.com

WebCore: https://bugs.webkit.org/show_bug.cgi?id=37717
Allow clients concerned with memory consumption to set a quota on session storage
since the memory used won't be released until the Page is destroyed.
The default is noQuota, which matches the current behavior.

Reviewed by Jeremy Orlow.

  • WebCore.base.exp: Export Settings::setSessionStorageQuota().
  • page/Page.cpp:

(WebCore::Page::sessionStorage):

  • page/Settings.cpp:

(WebCore::Settings::Settings):
(WebCore::Settings::setSessionStorageQuota):

  • page/Settings.h:

(WebCore::Settings::sessionStorageQuota):

  • storage/StorageNamespace.cpp:

(WebCore::StorageNamespace::sessionStorageNamespace):

  • storage/StorageNamespace.h:
  • storage/StorageNamespaceImpl.cpp:

(WebCore::StorageNamespaceImpl::sessionStorageNamespace):

  • storage/StorageNamespaceImpl.h:

WebKit/chromium: https://bugs.webkit.org/show_bug.cgi?id=37717
Changes needed now that StorageNamespaceImpl::sessionStorageNamespace() and
StorageNamespace::sessionStorageNamespace() take in a quota parameter.

Reviewed by Jeremy Orlow.

  • src/StorageNamespaceProxy.cpp:

(WebCore::StorageNamespace::sessionStorageNamespace):

  • src/WebStorageNamespaceImpl.cpp:

(WebKit::WebStorageNamespace::createSessionStorageNamespace):

11:34 Changeset [57821] by hausmann@webkit.org

[Qt] Build fix for WinCE.

Patch by Jocelyn Turcotte <jocelyn.turcotte@nokia.com> on 2010-04-19
Reviewed by Simon Hausmann.

Moved the include of the non-existing errno.h header file inside
platform guard macros.

  • jit/ExecutableAllocatorFixedVMPool.cpp:
11:28 Changeset [57820] by eric.carlson@apple.com

2010-04-19 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

Setting media element 'src' attribute should trigger immediate load
https://bugs.webkit.org/show_bug.cgi?id=37728

  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::attributeChanged): Schedule load every time 'src' attribute changes unless it is missing. (WebCore::HTMLMediaElement::prepareForLoad): Include steps 3 to 6 from loadInternal. (WebCore::HTMLMediaElement::loadInternal): Steps 3 to 6 are now in prepareForLoad. (WebCore::HTMLMediaElement::loadResource): MediaPlayer is now allocated in prepareForLoad so the previously loading file, if any, is cancelled there.

2010-04-19 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

Setting media element 'src' attribute should trigger immediate load
https://bugs.webkit.org/show_bug.cgi?id=37728

  • media/event-attributes-expected.txt: Updated results.
  • media/event-attributes.html: Do not need to call load().
  • media/video-source-add-src-expected.txt: Removed.
  • media/video-source-add-src.html: Removed.
  • media/video-src-change-expected.txt: Updated results.
  • media/video-src-change.html: Update test to match current behavior.
  • media/video-src-invalid-remove-expected.txt: Updated results.
  • media/video-src-invalid-remove.html: Update test to match current behavior.
11:07 Changeset [57819] by hausmann@webkit.org

[Qt] Fix compilation against namespaced Qt.

Patch by Jocelyn Turcotte <jocelyn.turcotte@nokia.com> on 2010-04-19
Reviewed by Simon Hausmann.

WebCore:

  • platform/graphics/GraphicsLayer.h:
  • platform/graphics/Tile.h:
  • platform/graphics/qt/MediaPlayerPrivateQt.h:
  • platform/network/qt/NetworkStateNotifierPrivate.h:

WebKit/qt:

  • WebCoreSupport/ChromeClientQt.h:
  • WebCoreSupport/QtFallbackWebPopup.h:
11:00 Changeset [57818] by eric@webkit.org

2010-04-19 Balazs Kelemen <kb@inf.u-szeged.hu>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Destroy SharedTimerQt before destruction of QCoreApplication.

To avoid unsafe situations caused by running WebCore code (through firing timers) when destruction of QCoreApplication
has been started, we should explicitly destroy the SharedTimerQt instance on application exit.
We can achieve that through installing a self-destroying slot for the QCoreApplication::aboutToQuit() signal
into the SharedTimerQt instance.

https://bugs.webkit.org/show_bug.cgi?id=36832

No functional change so no new tests.

  • platform/qt/SharedTimerQt.cpp: (WebCore::SharedTimerQt::SharedTimerQt): (WebCore::SharedTimerQt::destroy): (WebCore::SharedTimerQt::inst):
10:48 Changeset [57817] by mitz@apple.com

Make the fix for <rdar://problem/7873647> from r57759 more robust.

Reviewed by Darin Adler.

  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::updateHoverActiveState): Use RefPtrs for the Nodes.

10:43 Changeset [57816] by dglazkov@chromium.org

2010-04-19 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Adam Roben.

WinLauncher.h should use LF line-endings and use native line-endings style.
https://bugs.webkit.org/show_bug.cgi?id=37807

  • WinLauncher/WinLauncher.h: Added property svn:eol-style, converted to LF line-endings.
10:38 Changeset [57815] by eric@webkit.org

2010-04-19 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Add and enable the build of testatkroles to test ATK non form roles.

  • GNUmakefile.am:

2010-04-19 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Add testatkroles to test ATK non form roles.

  • tests/testatkroles.c: Added. (finish_loading): (atk_roles_fixture_setup): (atk_roles_fixture_teardown): (get_child_and_test_role): (test_webkit_atk_get_role_document_frame): (test_webkit_atk_get_role_heading): (test_webkit_atk_get_role_image): (test_webkit_atk_get_role_link): (test_webkit_atk_get_role_list_and_item): (test_webkit_atk_get_role_paragraph): (test_webkit_atk_get_role_section): (test_webkit_atk_get_role_table): (main):
10:21 Changeset [57814] by mitz@apple.com

Finish exposing extremal shrink factors WebHTMLView uses when shrinking pages to fit in the
printing width as SPI.

Reviewed by Sam Weinig.

  • WebKit.exp: Export _WebHTMLViewPrintingMinimumShrinkFactor and _WebHTMLViewPrintingMaximumShrinkFactor.
10:11 Changeset [57813] by yurys@chromium.org

2010-04-19 Yury Semikhatsky <yurys@chromium.org>

Unreviewed. Chromium build fix.

  • bindings/v8/JavaScriptCallFrame.h:
10:01 Changeset [57812] by yurys@chromium.org

2010-04-19 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: implement JavaScriptCallFrame that works for v8.
Implementing this binding for v8 allows to make evaluations on
call frames and protects access to the debugger context from
inspected context.

https://bugs.webkit.org/show_bug.cgi?id=37755

  • WebCore.gyp/WebCore.gyp:
  • WebCore.gypi:
  • bindings/js/JSJavaScriptCallFrameCustom.cpp: (WebCore::JSJavaScriptCallFrame::scopeType):
  • bindings/v8/JavaScriptCallFrame.cpp: Added. (WebCore::JavaScriptCallFrame::JavaScriptCallFrame): (WebCore::JavaScriptCallFrame::~JavaScriptCallFrame): (WebCore::JavaScriptCallFrame::caller): (WebCore::JavaScriptCallFrame::sourceID): (WebCore::JavaScriptCallFrame::line): (WebCore::JavaScriptCallFrame::functionName): (WebCore::JavaScriptCallFrame::scopeChain): (WebCore::JavaScriptCallFrame::scopeType): (WebCore::JavaScriptCallFrame::thisObject): (WebCore::JavaScriptCallFrame::evaluate):
  • bindings/v8/JavaScriptCallFrame.h: Added. (WebCore::JavaScriptCallFrame::create):
  • bindings/v8/ScriptDebugServer.cpp: (WebCore::ScriptDebugServer::currentCallFrame):
  • bindings/v8/ScriptDebugServer.h:
  • bindings/v8/custom/V8InjectedScriptHostCustom.cpp: (WebCore::V8InjectedScriptHost::currentCallFrameCallback):
  • bindings/v8/custom/V8JavaScriptCallFrameCustom.cpp: Added. (WebCore::V8JavaScriptCallFrame::evaluateCallback): (WebCore::V8JavaScriptCallFrame::scopeChainAccessorGetter): (WebCore::V8JavaScriptCallFrame::scopeTypeCallback): (WebCore::V8JavaScriptCallFrame::thisObjectAccessorGetter): (WebCore::V8JavaScriptCallFrame::typeAccessorGetter):
  • inspector/JavaScriptCallFrame.idl:
  • inspector/front-end/InjectedScript.js: (injectedScriptConstructor.):
09:41 Changeset [57811] by ossy@webkit.org

[Qt] Unskip a bunch of now passing tests, add expected files and pixel results.
Additionally fix some minor pixel differences.

Rubber-stamped by Simon Hausmann.

Compared to Mac png files and Mac expected files.

  • platform/qt/Skipped:
  • platform/qt/fast/frames/frame-navigation-expected.checksum: Added.
  • platform/qt/fast/frames/frame-navigation-expected.png: Added.
  • platform/qt/fast/frames/frame-navigation-expected.txt: Updated.
  • platform/qt/fast/overflow/line-clamp-expected.checksum: Added.
  • platform/qt/fast/overflow/line-clamp-expected.png: Added.
  • platform/qt/fast/overflow/line-clamp-expected.txt: Added.
  • platform/qt/fast/reflections/opacity-reflection-transform-expected.checksum: Added.
  • platform/qt/fast/reflections/opacity-reflection-transform-expected.png: Added.
  • platform/qt/fast/reflections/opacity-reflection-transform-expected.txt: Added.
  • platform/qt/fast/repaint/fixed-tranformed-expected.checksum: Added.
  • platform/qt/fast/repaint/fixed-tranformed-expected.png: Added.
  • platform/qt/fast/repaint/fixed-tranformed-expected.txt: Added.
  • platform/qt/fast/repaint/repaint-during-scroll-expected.checksum: Added.
  • platform/qt/fast/repaint/repaint-during-scroll-expected.png: Added.
  • platform/qt/fast/repaint/repaint-during-scroll-expected.txt: Added.
  • platform/qt/http/tests/local/file-url-sent-as-referer-expected.checksum: Added.
  • platform/qt/http/tests/local/file-url-sent-as-referer-expected.png: Added.
  • platform/qt/http/tests/local/file-url-sent-as-referer-expected.txt: Updated.
  • platform/qt/http/tests/misc/error404-expected.checksum: Added.
  • platform/qt/http/tests/misc/error404-expected.png: Added.
  • platform/qt/http/tests/misc/error404-expected.txt: Added.
  • platform/qt/http/tests/misc/frame-access-during-load-expected.checksum: Added.
  • platform/qt/http/tests/misc/frame-access-during-load-expected.png: Added.
  • platform/qt/http/tests/misc/frame-access-during-load-expected.txt: Added.
  • platform/qt/http/tests/misc/generated-content-inside-table-expected.checksum: Added.
  • platform/qt/http/tests/misc/generated-content-inside-table-expected.png: Added.
  • platform/qt/http/tests/misc/generated-content-inside-table-expected.txt: Added.
  • platform/qt/http/tests/misc/iframe404-expected.checksum: Added.
  • platform/qt/http/tests/misc/iframe404-expected.png: Added.
  • platform/qt/http/tests/misc/iframe404-expected.txt: Added.
  • platform/qt/http/tests/misc/location-replace-crossdomain-expected.checksum: Added.
  • platform/qt/http/tests/misc/location-replace-crossdomain-expected.png: Added.
  • platform/qt/http/tests/misc/location-replace-crossdomain-expected.txt: Added.
  • platform/qt/http/tests/uri/css-href-expected.checksum: Added.
  • platform/qt/http/tests/uri/css-href-expected.png: Added.
  • platform/qt/http/tests/uri/css-href-expected.txt: Updated.
  • platform/qt/svg/custom/radial-gradient-with-outstanding-focalPoint-expected.checksum: Added.
  • platform/qt/svg/custom/radial-gradient-with-outstanding-focalPoint-expected.png: Added.
  • platform/qt/svg/custom/radial-gradient-with-outstanding-focalPoint-expected.txt: Added.
  • platform/qt/svg/text/foreignObject-text-clipping-bug-expected.checksum: Added.
  • platform/qt/svg/text/foreignObject-text-clipping-bug-expected.png: Added.
  • platform/qt/svg/text/foreignObject-text-clipping-bug-expected.txt: Added.
08:05 Changeset [57810] by jberlin@webkit.org

Rubber Stamped by Adam Roben

Chromium Release Build Fix.

  • css/CSSSelector.cpp:

(WebCore::CSSSelector::pseudoId):
In the case where the DATALIST is not enabled, fall through to NOPSEUDO instead of omitting PseudoInputListButton entirely from the switch.

06:28 Changeset [57809] by jberlin@webkit.org

First steps towards fixing bug 24021 - pseudo-element styles not accessible / retrievable via DOM methods.
https://bugs.webkit.org/show_bug.cgi?id=24021

Reviewed by Dave Hyatt.

WebCore:

Allows access to the computed styles for the pseudo-elements through the second argument to getComputedStyle.
This approach does not provide the correct values for 'length' properties and does not work for the ':selection' pseudo-element and will instead return results similiar to those returned by Firefox. This approach also requires waiting until at least one iteration of a hardware accelerated composited animation to return the correct values for the "opacity" and "transform" properties of a pseudo-element associated with the element being animated.
Those values need to be retrieved from the renderer for the pseudo-element as opposed to the cached RenderStyle for the element on which the pseudo-element is defined, which is further complicated by the fact that not all elements have renderers.

Test: fast/css/getComputedStyle/getComputedStyle-with-pseudo-element.html

  • WebCore.base.exp:
  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::CSSComputedStyleDeclaration::CSSComputedStyleDeclaration):
Parse the and store the pseudo-element specifier from the string provided by the user.
(WebCore::CSSComputedStyleDeclaration::getFontSizeCSSValuePreferringKeyword):
Get the computed style for the pseudo-element if it has been specified.
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):
Get the computed style for the pseudo-element if it has been specified, with a FIXME noting that the values returned for the "opacity" and "transform" properties of a pseudo-element associated with an element being animated and using hardware accelerated compositing will not be correct until after the first iteration of the animation.
(WebCore::CSSComputedStyleDeclaration::length):
Get the computed style for the pseudo-element if it has been specified.
(WebCore::CSSComputedStyleDeclaration::cssPropertyMatches):
Ditto.

  • css/CSSComputedStyleDeclaration.h:

(WebCore::computedStyle):
Take into consideration the pseudo-element.

  • css/CSSSelector.cpp:

(WebCore::CSSSelector::pseudoId):
Return the PseudoId that corresponds to the given PseudoType. If there is no corresponding PseudoId, returns NOPSEUDO.
(WebCore::nameToPseudoTypeMap):
Create and return the mapping between string names and PseudoTypes.
(WebCore::CSSSelector::parsePseudoType):
Parse and the given string into a PseudoType.
(WebCore::CSSSelector::extractPseudoType):
Refactored to use parsePseudoType.

  • css/CSSSelector.h:
  • css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::SelectorChecker::checkOneSelector):
Refactored to use pseudoId.

  • dom/Element.cpp:

(WebCore::Element::computedStyle):
If the pseudo-element is specified, then return the cached RenderStyle for that PseudoId. Added a FIXME to find the actual renders of the pseudo-elements instead of just the cached RenderStyle of the RenderStyle for the associated element.

  • dom/Element.h:

(WebCore::Element::virtualComputedStyle):
Because Element::computedStyle is used so often, don't make it virtual. Instead, provide a virtualComputedStyle method in the Node.h class andmake computedStyle non-virtual. That way the Element version and the Node version of computedStyle will have the same name and look the same at the call site, but the Element version will be more efficient.

  • dom/Node.h:

(WebCore::Node::computedStyle):
Ditto.

  • dom/Node.cpp:

(WebCore::Node::virtualComputedStyle):
Get the computed style for the pseudo-element if it has been specified.

  • page/DOMWindow.cpp:

(WebCore::DOMWindow::getComputedStyle):
Ditto.

LayoutTests:

  • fast/css/getComputedStyle/getComputedStyle-with-pseudo-element-expected.txt: Added.
  • fast/css/getComputedStyle/getComputedStyle-with-pseudo-element.html: Added.
04:54 Changeset [57808] by ossy@webkit.org

[Qt] lacks clipToImageBuffer()
https://bugs.webkit.org/show_bug.cgi?id=24289

Skip svg/clip-path tests introduced in r52511,
and move clip related tests here.

  • platform/qt/Skipped:
02:05 Changeset [57807] by tkent@chromium.org

2010-04-19 Kent Tamura <tkent@chromium.org>

Reviewed by Shinichiro Hamaji.

[DRT/Chromium] Fix a test initialization problem
https://bugs.webkit.org/show_bug.cgi?id=37791

  • DumpRenderTree/chromium/DumpRenderTree.cpp: (runTest): Call resetTestController() before runFileTest(). Some controllers initialize their fields in reset() and not in their constructors.
01:52 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
01:40 Changeset [57806] by tkent@chromium.org

2010-04-19 Kent Tamura <tkent@chromium.org>

Reviewed by Eric Seidel.

[Chromium] new-run-webkit-tests should use WebKitDriver for --use-drt
https://bugs.webkit.org/show_bug.cgi?id=37793

We need to use WebKitDriver instead of ChromiumDriver for Chromium
DRT because its interface is different from test_shell.

Chromium DRT has no UI. So we can't use it to show test results.

  • Scripts/webkitpy/layout_tests/port/chromium.py:

04/18/10:

23:36 Changeset [57805] by eric@webkit.org

2010-04-18 Eric Seidel <eric@webkit.org>

Reviewed by Adam Roben.

Make failure-reason more forgiving
https://bugs.webkit.org/show_bug.cgi?id=37525

Removed search_limit, which wasn't very useful anyway.
Added a log about the long load time loading from the builders.
Prompt the user for what revision to start walking from (makes it easy to restart upon failure).

  • Scripts/webkitpy/common/net/buildbot.py:
  • Scripts/webkitpy/tool/commands/queries.py:
22:13 Changeset [57804] by eric@webkit.org

2010-04-18 Kwang Yul Seo <skyul@company100.net>

Reviewed by Laszlo Gombos.

[WINCE] Don't define WTF_CPU_MIDDLE_ENDIAN=1
https://bugs.webkit.org/show_bug.cgi?id=37434

Windows CE supports little-endian format only, so don't define
WTF_CPU_MIDDLE_ENDIAN=1.

  • wtf/Platform.h:
21:26 Changeset [57803] by cjerdonek@webkit.org

Created a class for check-webkit-style that encapsulates iterating
over text files and reading them.

Reviewed by Shinichiro Hamaji.

https://bugs.webkit.org/show_bug.cgi?id=37754

This revision is an intermediate step towards separating our
style-checking code from the logic of iterating over files and
reading them.

  • Scripts/webkitpy/common/system/logtesting.py:
    • Added a logMessages() method to the LoggingTestCase class. This method provides unit tests with access to the raw list of log messages in case the tester needs to do something more than simply assert the list of existing messages.
  • Scripts/webkitpy/style/checker.py:
    • Added a ProcessorBase class that processors of lists of lines should eventually inherit from.
    • Also added a FIXME to use the ProcessorBase class and the TextFileReader class added below.
  • Scripts/webkitpy/style/filereader.py: Added.
    • Created a TextFileReader class that encapsulates reading and iterating over text files.
  • Scripts/webkitpy/style/filereader_unittest.py: Added.
    • Added a TextFileReaderTest class to unit-test the new TextFileReader class.
21:22 Changeset [57802] by eric@webkit.org

2010-04-18 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Laszlo Gombos.

[Qt] Fix JavaScriptCore's include path for WinCE builds

https://bugs.webkit.org/show_bug.cgi?id=36751

2010-04-18 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Laszlo Gombos.

[Qt] Fix JavaScriptCore's include path for WinCE builds

https://bugs.webkit.org/show_bug.cgi?id=36751

  • WebCore.pro:
21:08 Changeset [57801] by eric@webkit.org

2010-04-18 Michael Forney <michael@mforney.org>

Reviewed by Laszlo Gombos.

https://bugs.webkit.org/show_bug.cgi?id=37762

Fixes the sandbox option in configure.

  • configure.ac: Fix cut and paste error in the sandbox option causing --{enable,disable}-sandbox to be ineffective, and it to be controlled by the ruby flag.
20:26 Changeset [57800] by tony@chromium.org

2010-04-15 Tony Chang <tony@chromium.org>

Reviewed by Adam Barth.

[chromium] new-run-webkit-tests should be able to use chromium DRT
https://bugs.webkit.org/show_bug.cgi?id=37645

Make sure that the lack of a chromium checkout doesn't cause the script to
fail.

  • Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
  • Scripts/webkitpy/layout_tests/port/chromium.py:
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py: Fix up a few paths

to be relative to an upstream output dir.

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py: Add --use-drt flag.
14:25 Changeset [57799] by eric@webkit.org

2010-04-18 Eric Seidel <eric@webkit.org>

Unreviewed, fixing the Qt bot.

Add a layer of indirection when calling run-webkit-tests to allow testing new-run-webkit-tests on various platforms
https://bugs.webkit.org/show_bug.cgi?id=37632

  • Scripts/run-webkit-tests:
    • Exit non-zero of launching the harness fails.
13:58 Changeset [57798] by eric@webkit.org

2010-04-18 Eric Seidel <eric@webkit.org>

Unreviewed, fixing the Qt bot.

Add a layer of indirection when calling run-webkit-tests to allow testing new-run-webkit-tests on various platforms
https://bugs.webkit.org/show_bug.cgi?id=37632

  • Scripts/run-webkit-tests:
    • Fix the wrapper to work for users who do not have WebKitTools/Scripts in their path.
13:56 Changeset [57797] by abarth@webkit.org

2010-04-18 Adam Barth <abarth@webkit.org>

Unreviewed. Update a label of the diagram to better match the
description in the text of the blog post.

  • blog-files/loader-diagram.svg:
13:34 Changeset [57796] by eric@webkit.org

2010-04-14 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Add a layer of indirection when calling run-webkit-tests to
allow testing new-run-webkit-tests on various platforms.
https://bugs.webkit.org/show_bug.cgi?id=37632

This will let us test and fix bugs in new-run-webkit-tests
without needing to restart the buildbot master between tests.

Currently this change leaves run-webkit-tests as-is, but once
its landed we will easily be able to turn on/off
new-run-webkit-tests for various ports/configurations.

I will send a note out to webkit-dev about how we will
be using this launcher script to test on the bots.

  • Scripts/old-run-webkit-tests: Copied from WebKitTools/Scripts/run-webkit-tests.
  • Scripts/run-webkit-tests:
    • A new script which decides whether to run new- or old- run-webkit-tests based on the platform.
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
    • Add some dummy argument handling for arguments which old-run-webkit-tests supports but new-run-webkit-tests does not yet.
10:43 Changeset [57795] by eric@webkit.org

2010-04-18 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Add Gtk bots to the list of "core builders" (builders which stop the commit-queue when they turn red)
https://bugs.webkit.org/show_bug.cgi?id=33295

The Gtk builders have been green every time I've looked
at them in the last 5 days or so. It would appear webkit
is now keeping them green and we should update the core
builder list to match reality.

  • Scripts/webkitpy/common/net/buildbot.py:
  • Scripts/webkitpy/common/net/buildbot_unittest.py:
04:01 Changeset [57794] by robert@webkit.org

[Qt] Add support for LayoutTestController commands

2010-04-18 Robert Hogan <robert@webkit.org>

Unreviewed fix to pixel differences in unskipped layout tests.

[Qt] Add support for LayoutTestController commands

Fix up minor pixel differences on unskipped tests from r57793.

https://bugs.webkit.org/show_bug.cgi?id=35844

  • platform/qt/editing/deleting/smart-delete-003-expected.txt:
  • platform/qt/editing/deleting/smart-delete-004-expected.txt:
  • platform/qt/editing/selection/5195166-1-expected.txt: Copied from LayoutTests/platform/mac/editing/selection/5195166-1-expected.txt.
  • platform/qt/editing/selection/5195166-2-expected.txt:
03:36 Changeset [57793] by robert@webkit.org

[Qt] Add support for LayoutTestController commands
2010-04-18 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Add support for LayoutTestController commands:

setSmartInsertDeleteEnabled
setSelectTrailingWhitespaceEnabled
execCommand
isCommandEnabled

https://bugs.webkit.org/show_bug.cgi?id=35844

  • Api/qwebpage.cpp: (QWebPagePrivate::QWebPagePrivate):
  • Api/qwebpage_p.h:
  • WebCoreSupport/DumpRenderTreeSupportQt.cpp: (DumpRenderTreeSupportQt::setSmartInsertDeleteEnabled): (DumpRenderTreeSupportQt::setSelectTrailingWhitespaceEnabled): (DumpRenderTreeSupportQt::executeCoreCommandByName): (DumpRenderTreeSupportQt::isCommandEnabled):
  • WebCoreSupport/DumpRenderTreeSupportQt.h:
  • WebCoreSupport/EditorClientQt.cpp: (WebCore::EditorClientQt::smartInsertDeleteEnabled): (WebCore::EditorClientQt::toggleSmartInsertDelete): (WebCore::EditorClientQt::isSelectTrailingWhitespaceEnabled):
  • WebCoreSupport/EditorClientQt.h:

2010-04-18 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Add support for LayoutTestController commands:

setSmartInsertDeleteEnabled
setSelectTrailingWhitespaceEnabled
execCommand
isCommandEnabled

Unskip tests:
editing/deleting/smart-editing-disabled.html
editing/execCommand/19089.html
editing/execCommand/delete-image-in-anchor.html
editing/execCommand/enabling-and-selection-2.html
editing/selection/doubleclick-whitespace-crash.html
editing/selection/doubleclick-whitespace-img-crash.html
editing/selection/doubleclick-whitespace.html
editing/selection/select-line.html
editing/deleting/5300379.html
editing/deleting/smart-delete-003.html
editing/deleting/smart-delete-004.html
editing/selection/5195166-1.html
editing/selection/5195166-2.html

https://bugs.webkit.org/show_bug.cgi?id=35844

  • platform/qt/Skipped:
  • platform/qt/editing/deleting/5300379-expected.txt: Update platform-specific results.
  • platform/qt/editing/deleting/smart-delete-003-expected.txt: Update platform-specific results.
  • platform/qt/editing/deleting/smart-delete-004-expected.txt: Update platform-specific results.
  • platform/qt/editing/selection/5195166-1-expected.txt: Copied from LayoutTests/platform/mac/editing/selection/5195166-1-expected.txt.
  • platform/qt/editing/selection/5195166-2-expected.txt: Update platform-specific results.

2010-04-18 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Add support for LayoutTestController commands:

setSmartInsertDeleteEnabled
setSelectTrailingWhitespaceEnabled
execCommand
isCommandEnabled

https://bugs.webkit.org/show_bug.cgi?id=35844

  • DumpRenderTree/qt/DumpRenderTreeQt.cpp: (WebCore::WebPage::resetSettings):
  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::setSmartInsertDeleteEnabled): (LayoutTestController::setSelectTrailingWhitespaceEnabled): (LayoutTestController::execCommand): (LayoutTestController::isCommandEnabled):
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
03:25 Changeset [57792] by tonikitoo@webkit.org

2010-04-18 Antonio Gomes <tonikitoo@webkit.org>

[Qt] Unreviewed fix: remove svg/custom/clip-path-display-none-child.svg from Skipped (non-existent test)

It was removed by r57511.

  • platform/qt/Skipped:

04/17/10:

23:27 Changeset [57791] by abarth@webkit.org

2010-04-17 Adam Barth <abarth@webkit.org>

Unreviewed. Added loader diagram for upcoming blog post.

  • blog-files/loader-diagram.svg: Added.
23:03 Changeset [57790] by ap@apple.com

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=37720
<rdar://problem/7873752> HTMLFrameSetElement-window-eventListener-attributes.html sometimes
crashes on SnowLeopard Release

Tests:

  • fast/dom/Window/HTMLBodyElement-window-eventListener-attributes.html:
  • fast/dom/Window/HTMLFrameSetElement-window-eventListener-attributes.html:
  • bindings/scripts/CodeGeneratorJS.pm: Use global object as a wrapper for window event listeners set via document.body. The body wrapper can go away if nothing else references it.
  • html/HTMLBodyElement.idl: Override listeners that exist on Element with custom implementations, because we need to use window object as a wrapper, not the element. Marked all window event listeners as such for code generator.
  • html/HTMLFrameSetElement.idl: Ditto.
22:00 Changeset [57789] by joepeck@webkit.org

2010-04-17 Juan C. Montemayor <jmonte03@cs.tufts.edu>

Reviewed by Joseph Pecoraro.

Databases pane doesn't recognize table creation/deletion
https://bugs.webkit.org/show_bug.cgi?id=20219

  • inspector/front-end/DatabaseQueryView.js: (WebInspector.DatabaseQueryView.prototype._queryFinished):
17:42 Changeset [57788] by eric@webkit.org

2010-04-17 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Remove steps_references and commands_references
https://bugs.webkit.org/show_bug.cgi?id=37758

We tried using the mumble_references convention to manage our
dependencies, but it doesn't seem to be providing much value for the
steps and commands module because these modules are small pieces of the
larger tool module. In this patch, I've removed the references file
for these modules.

I've left the style_references file for the style module because that
module seems better isolated from the rest of webkitpy and the
style_references file appears to be providing some value.

  • Scripts/webkitpy/tool/commands/commandtest.py:
  • Scripts/webkitpy/tool/commands/download_unittest.py:
  • Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
  • Scripts/webkitpy/tool/commands/queries_unittest.py:
  • Scripts/webkitpy/tool/commands/queues_unittest.py:
  • Scripts/webkitpy/tool/commands/queuestest.py:
  • Scripts/webkitpy/tool/commands/upload.py:
  • Scripts/webkitpy/tool/commands/upload_unittest.py:
  • Scripts/webkitpy/tool/commands_references.py: Removed.
  • Scripts/webkitpy/tool/steps/closebugforlanddiff_unittest.py:
  • Scripts/webkitpy/tool/steps/steps_unittest.py:
  • Scripts/webkitpy/tool/steps/updatechangelogswithreview_unittest.py:
  • Scripts/webkitpy/tool/steps/validatereviewer_unittest.py:
  • Scripts/webkitpy/tool/steps_references.py: Removed.
13:43 Changeset [57787] by eric@webkit.org

2010-04-17 Yaar Schnitman <yaar@chromium.org>

Reviewed by Adam Barth.

Auto-generate V8 bindings for canvas.* overloads
https://bugs.webkit.org/show_bug.cgi?id=37453

  • bindings/v8/custom/V8CanvasRenderingContext2DCustom.cpp: Deleted custom code that is now auto-generated.
  • bindings/scripts/CodeGeneratorV8.pm: Not generating custom signatures for overloaded methods and accepting Object as a string input.
  • html/canvas/CanvasRenderingContext2D.cpp: (WebCore::CanvasRenderingContext2D::drawImage): Added missing 9-arguments overloads for drawImage.
  • html/canvas/CanvasRenderingContext2D.h:
  • html/canvas/CanvasRenderingContext2D.idl: Overloaded methods defined (V8 only).
13:10 Changeset [57786] by eric@webkit.org

2010-04-17 Adam Barth <abarth@webkit.org>

Unreviewed.

Test file: URL canonicalization
https://bugs.webkit.org/show_bug.cgi?id=37700

Re-land http://trac.webkit.org/changeset/57719, hopefully not breaking
the Windows bot this time.

  • fast/url/file-expected.txt: Added.
  • fast/url/file.html: Added.
  • fast/url/mailto.html:
  • fast/url/script-tests/file.js: Added.
13:00 Changeset [57785] by eric@webkit.org

2010-04-17 Adam Barth <abarth@webkit.org>

Unreviewed, but I took Jeremy Orlow's suggestions on an earlier
iteration of this patch.

Add URL parsing tests for relative URL resolution
https://bugs.webkit.org/show_bug.cgi?id=37644

I've split the platform specific parts of this test into separate
files. On KURL-based ports these should have the same results on every
platform. We can decide whether to have platform-specific behavior at
a later time.

  • fast/url/relative-expected.txt: Added.
  • fast/url/relative-unix-expected.txt: Added.
  • fast/url/relative-unix.html: Added.
  • fast/url/relative-win-expected.txt: Added.
  • fast/url/relative-win.html: Added.
  • fast/url/relative.html: Added.
  • fast/url/script-tests/relative-unix.js: Added.
  • fast/url/script-tests/relative-win.js: Added.
  • fast/url/script-tests/relative.js: Added.
11:42 Changeset [57784] by eric@webkit.org

2010-04-17 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: move JavaScriptCallFrame.{h,cpp} to WebCore/bindings/js

https://bugs.webkit.org/show_bug.cgi?id=37740

  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • bindings/js/JSBindingsAllInOne.cpp:
  • bindings/js/JavaScriptCallFrame.cpp: Renamed from WebCore/inspector/JavaScriptCallFrame.cpp. (WebCore::JavaScriptCallFrame::JavaScriptCallFrame): (WebCore::JavaScriptCallFrame::caller): (WebCore::JavaScriptCallFrame::scopeChain): (WebCore::JavaScriptCallFrame::dynamicGlobalObject): (WebCore::JavaScriptCallFrame::functionName): (WebCore::JavaScriptCallFrame::type): (WebCore::JavaScriptCallFrame::thisObject): (WebCore::JavaScriptCallFrame::evaluate):
  • bindings/js/JavaScriptCallFrame.h: Renamed from WebCore/inspector/JavaScriptCallFrame.h. (WebCore::JavaScriptCallFrame::create): (WebCore::JavaScriptCallFrame::invalidate): (WebCore::JavaScriptCallFrame::isValid): (WebCore::JavaScriptCallFrame::sourceID): (WebCore::JavaScriptCallFrame::line): (WebCore::JavaScriptCallFrame::update):
  • inspector/InjectedScriptHost.h:
  • inspector/InspectorController.h:
09:00 Changeset [57783] by alex@webkit.org

2010-04-17 Alejandro G. Castro <alex@igalia.com>

Rubber-stamped by Xan Lopez.

Skipping media/audio-constructor-preload.html
https://bugs.webkit.org/show_bug.cgi?id=37753

  • platform/gtk/Skipped:
08:49 Changeset [57782] by alex@webkit.org

2010-04-17 Alejandro G. Castro <alex@igalia.com>

Rubber-stamped by Xan Lopez.

Skipping inspector/timeline-paint.html
https://bugs.webkit.org/show_bug.cgi?id=37752

  • platform/gtk/Skipped:
08:16 Changeset [57781] by alex@webkit.org

2010-04-17 Alejandro G. Castro <alex@igalia.com>

Reviewed by Xan Lopez.

We have to initialize the timer attribute after destroying it, not
doing it was causing crashes in some situations.

  • webkit/webkitdownload.cpp: (webkit_download_finalize):
07:55 Changeset [57780] by eric@webkit.org

2010-04-17 Vangelis Kokkevis <vangelis@chromium.org>

Reviewed by Dimitri Glazkov.

Adding definition of GLES2Context class.
https://bugs.webkit.org/show_bug.cgi?id=37541

No new functionality implemented yet, no tests.

  • WebCore.gypi: Added GLES2Context.h
  • platform/chromium/GLES2Context.h: Added.

2010-04-17 Vangelis Kokkevis <vangelis@chromium.org>

Reviewed by Dimitri Glazkov.

Adding the implementation of GLES2Context class which provides WebCore access to a GL ES context.
https://bugs.webkit.org/show_bug.cgi?id=37541

  • WebKit.gyp:
  • public/WebGLES2Context.h: Added. (WebKit::WebGLES2Context::~WebGLES2Context):
  • src/GLES2Context.cpp: Added. (WebCore::GLES2ContextInternal::GLES2ContextInternal): (WebCore::GLES2ContextInternal::~GLES2ContextInternal): (WebCore::GLES2ContextInternal::getWebGLES2Context): (WebCore::GLES2ContextInternal::initialize): (WebCore::GLES2Context::create): (WebCore::GLES2Context::~GLES2Context): (WebCore::GLES2Context::initialize): (WebCore::GLES2Context::makeCurrent): (WebCore::GLES2Context::destroy): (WebCore::GLES2Context::swapBuffers):
06:37 Changeset [57779] by jchaffraix@webkit.org

Crash while handling SVG font in the wrong namespace imported with @font-face
https://bugs.webkit.org/show_bug.cgi?id=18862

Reviewed by Darin Adler.

WebCore:

Test: fast/invalid/invalidSVGFont.html

  • loader/CachedFont.cpp:

(WebCore::CachedFont::getSVGFontById): Make sure we really get an SVGFontElement by using
getElementsByNameNS: the element factory chooses which element to create based on both
localName and namespace.

LayoutTests:

Test that load an SVG document not in the SVG namespace with @font-face does not
lead to a crash.

  • fast/invalid/invalidSVGFont-expected.txt: Added.
  • fast/invalid/invalidSVGFont.html: Added.
  • fast/invalid/resources/svgFontWrongNamespace.svg: Added.
01:13 Changeset [57778] by eric@webkit.org

2010-04-17 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

WebKit needs a Chromium Mac EWS Builder
https://bugs.webkit.org/show_bug.cgi?id=37742

  • Scripts/webkitpy/tool/commands/earlywarningsystem.py:
01:02 Changeset [57777] by abarth@webkit.org

2010-04-17 Adam Barth <abarth@webkit.org>

Fix expected results for unit test broken by
http://trac.webkit.org/changeset/57772

  • Scripts/webkitpy/common/net/buildbot_unittest.py:

04/16/10:

23:14 Changeset [57776] by ossy@webkit.org

Unreviewed, rolling out r57770.
http://trac.webkit.org/changeset/57770
https://bugs.webkit.org/show_bug.cgi?id=37746

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2010-04-16
8 test cases crashed (Requested by Ossy on #webkit).

WebCore:

  • platform/graphics/qt/GraphicsLayerQt.cpp:

(WebCore::AnimationQtBase::AnimationQtBase):
(WebCore::TransformAnimationQt::~TransformAnimationQt):
(WebCore::TransformAnimationQt::applyFrame):
(WebCore::OpacityAnimationQt::applyFrame):
(WebCore::GraphicsLayerQt::addAnimation):

LayoutTests:

  • platform/qt/Skipped:
21:40 Changeset [57775] by dbates@webkit.org

2010-04-16 Daniel Bates <dbates@rim.com>

Reviewed by Adam Treat.

https://bugs.webkit.org/show_bug.cgi?id=36312

Adds support for the viewport meta tag. The code is largely derived in whole
or in part from the WebCore-528.15 source published as part of the iPhone 3.1.3
source code <http://www.opensource.apple.com/source/WebCore/WebCore-528.15/>.

  • Android.mk: Added file ViewportArguments.cpp.
  • GNUmakefile.am: Added files ViewportArguments.cpp and ViewportArguments.h.
  • WebCore.gypi: Ditto.
  • WebCore.pro: Ditto.
  • WebCore.vcproj/WebCore.vcproj: Ditto.
  • WebCore.xcodeproj/project.pbxproj: Ditto.
  • dom/Document.cpp: (WebCore::isSeparator): Added. (WebCore::Document::processArguments): Added. (WebCore::Document::processViewport): Added.
  • dom/Document.h:
  • dom/ViewportArguments.cpp: Added. (WebCore::setViewportFeature): (WebCore::viewportErrorMessageTemplate): (WebCore::viewportErrorMessageLevel): (WebCore::reportViewportWarning):
  • dom/ViewportArguments.h: Added. (WebCore::): (WebCore::ViewportArguments::): (WebCore::ViewportArguments::ViewportArguments): (WebCore::ViewportArguments::hasCustomArgument):
  • html/HTMLMetaElement.cpp: (WebCore::HTMLMetaElement::process): Modified to call Document::processViewport.
  • page/ChromeClient.h: (WebCore::ChromeClient::didReceiveViewportArguments): Added.
20:23 Changeset [57774] by weinig@apple.com

Remove the need for a .defs file! Define WK_EXPORT.

Reviewed by Jon "The Belly" Honeycutt.

  • UIProcess/API/C/WKBase.h:
  • WebProcess/win/WebProcessMain.h:
  • win/WebKit2.def: Removed.
  • win/WebKit2.vcproj:
20:06 Changeset [57773] by weinig@apple.com

Teach windows MiniBrowser how to work with window.open()
and targeted links.

Reviewed by Adam Roben.

WebKit2:

Export WKPageSetPageUIClient.

  • win/WebKit2.def:

WebKitTools:

  • MiniBrowser/win/BrowserView.cpp:

(createNewPage): Create a new BrowserWindow and return its page.
(showPage): Show the page.
(closePage): Empty implementation.
(runJavaScriptAlert): Empty implementation.
(BrowserView::create): Register a UIClient.

  • MiniBrowser/win/BrowserView.h:

(BrowserView::webView): Added.
Change create to take a BrowserWindow instead of an HWND.

  • MiniBrowser/win/BrowserWindow.cpp:

(BrowserWindow::wndProc): Respond to WM_NCDESTROY.
(BrowserWindow::goToURL): Added. Forwards to BrowserView.
(BrowserWindow::onCreate): Don't always go to the default
url. Let the caller do this.
(BrowserWindow::onNCDestroy): Delete the window.

  • MiniBrowser/win/BrowserWindow.h:

(BrowserWindow::view): Added.
(BrowserWindow::window): Added.

  • MiniBrowser/win/main.cpp:

(_tWinMain):
Go to the default URL for the initial page. Allocate the initial
window on the heap for correctness.

19:53 Changeset [57772] by aroben@apple.com

Add the Windows Debug (Test) builder to the list of core builders

It's been green for a few days now, and all the known Windows
flakiness is Release-only.

Rubber-stamped by Mark Rowe.

  • Scripts/webkitpy/common/net/buildbot.py:

(BuildBot.core_builder_name_regexps): Added a regular expression to
match the "Windows Debug (Test)" builder.

19:30 Changeset [57771] by dumi@chromium.org

Get rid of the UNUSED_PARAM macro in GraphicsContext3D.cpp.
https://bugs.webkit.org/show_bug.cgi?id=37733

Reviewed by Eric Seidel.

  • src/GraphicsContext3D.cpp:

(WebCore::GraphicsContext3D::GraphicsContext3D):

19:04 Changeset [57770] by eric@webkit.org

2010-04-16 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Antti Koivisto.

[Qt] GraphicsLayer: support fill-modes
https://bugs.webkit.org/show_bug.cgi?id=36216

Remove LayoutTests/animations/fill-mode-transform.html from the skip list

  • platform/qt/Skipped:

2010-04-16 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Antti Koivisto.

[Qt] GraphicsLayer: support fill-modes
https://bugs.webkit.org/show_bug.cgi?id=36216
Implement the CSS-animation "fill mode" concept in GraphicsLayerQt. The concept
enables a key-frame animation to go to the animation's starting point before the delay,
and/or to stay at the animation's ending point after its ended, without reverting to the default
value.
We do that by manually setting the value to keyframe-0 before the delay if fill-mode is backwards/both,
and manually modifying the default value to the animated value as we animate, with fill-mode forwards/both.

  • platform/graphics/qt/GraphicsLayerQt.cpp: (WebCore::AnimationQtBase::AnimationQtBase): (WebCore::TransformAnimationQt::~TransformAnimationQt): (WebCore::TransformAnimationQt::applyFrame): (WebCore::GraphicsLayerQt::addAnimation):
17:24 Changeset [57769] by barraclough@apple.com
  • bindings/js/JSDesktopNotificationsCustom.cpp:

(WebCore::JSNotification::addEventListener):
(WebCore::JSNotification::removeEventListener):

Reviewed by NOBODY (arm build fix).

17:20 Changeset [57768] by dumi@chromium.org

Changing the return type of WebSecurityOrigin::createFromDatabaseIdentifier().
https://bugs.webkit.org/show_bug.cgi?id=34466

Reviewed by Jeremy Orlow.

Changing the return type of
WebSecurityOrigin::createFromDatabaseIdentifier() from
WebSecurityOrigin* to WebSecurityOrigin, to make it more
consistent with the other WebSecurityOrigin methods.

  • public/WebSecurityOrigin.h:
  • src/WebDatabase.cpp:

(WebKit::WebDatabase::closeDatabaseImmediately):

  • src/WebSecurityOrigin.cpp:

(WebKit::WebSecurityOrigin::createFromDatabaseIdentifier):

17:14 QtWebKitBackportingFixes edited by hausmann@webkit.org
(diff)
17:12 Changeset [57767] by dimich@chromium.org

WebCore: [v8] In Workers, script errors right after close() are not delivered to the Worker.onerror.
https://bugs.webkit.org/show_bug.cgi?id=37691

Reviewed by Yury Semikhatsky.

Existing worker-close.html will now work in Chromium.

  • bindings/v8/V8Utilities.cpp:

(WebCore::getScriptExecutionContext): Stop using proxy() to just retrieve WorkerContext which should be always available.

  • bindings/v8/WorkerContextExecutionProxy.h: removed workerContext() accessor which moved to WorkerScriptController.
  • bindings/v8/WorkerScriptController.h:

(WebCore::WorkerScriptController::workerContext): Added, to be able to pull WorkerContext out from the controller.

WebKit/chromium: [v8] In Workers, script errors right after close() are not delivered to the Worker.onerror
https://bugs.webkit.org/show_bug.cgi?id=37691

Reviewed by Yury Semikhatsky.

  • src/WebWorkerClientImpl.cpp:

(WebKit::WebWorkerClientImpl::createWorkerContextProxy): Don't need to pull V8 proxy to retrieve current WorkerContext.

LayoutTests: [v8] In Workers, script errors right after close() are not delivered to the Worker.onerror
https://bugs.webkit.org/show_bug.cgi?id=37691

Reviewed by Yury Semikhatsky.

Moved the expected result into platform-independent chromium subdirectory, since
the only chromium-specific difference is the way v8 formats an error message.
Also updated the expected result to match the passing now test.

  • platform/chromium-win/fast/workers/worker-close-expected.txt: Removed.
  • platform/chromium/fast/workers/worker-close-expected.txt: Added.
17:10 Changeset [57766] by barraclough@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37735
Remove JSC specific code from WebCore::AtomicString

Reviewed by Sam Weinig.

Add generic constructor/add/find methods that take a UChar* & length, along
with a known existing hash for the string.
This removes the remaining JSC specific code from platform/text.

  • bindings/js/JSAbstractWorkerCustom.cpp:

(WebCore::JSAbstractWorker::addEventListener):
(WebCore::JSAbstractWorker::removeEventListener):

  • bindings/js/JSDOMApplicationCacheCustom.cpp:

(WebCore::JSDOMApplicationCache::addEventListener):
(WebCore::JSDOMApplicationCache::removeEventListener):

  • bindings/js/JSDOMBinding.cpp:

(WebCore::ustringToAtomicString):
(WebCore::identifierToAtomicString):
(WebCore::findAtomicString):

  • bindings/js/JSDOMBinding.h:
  • bindings/js/JSDOMWindowCustom.cpp:

(WebCore::childFrameGetter):
(WebCore::JSDOMWindow::getOwnPropertySlot):
(WebCore::JSDOMWindow::getOwnPropertyDescriptor):
(WebCore::JSDOMWindow::open):
(WebCore::JSDOMWindow::addEventListener):
(WebCore::JSDOMWindow::removeEventListener):

  • bindings/js/JSElementCustom.cpp:

(WebCore::JSElement::setAttribute):
(WebCore::JSElement::setAttributeNS):

  • bindings/js/JSEventSourceCustom.cpp:

(WebCore::JSEventSource::addEventListener):
(WebCore::JSEventSource::removeEventListener):

  • bindings/js/JSHTMLAllCollectionCustom.cpp:

(WebCore::getNamedItems):
(WebCore::JSHTMLAllCollection::canGetItemsForName):

  • bindings/js/JSHTMLCollectionCustom.cpp:

(WebCore::getNamedItems):
(WebCore::JSHTMLCollection::canGetItemsForName):

  • bindings/js/JSHTMLDocumentCustom.cpp:

(WebCore::JSHTMLDocument::canGetItemsForName):

  • bindings/js/JSHTMLFormElementCustom.cpp:

(WebCore::JSHTMLFormElement::canGetItemsForName):
(WebCore::JSHTMLFormElement::nameGetter):

  • bindings/js/JSHTMLFrameSetElementCustom.cpp:

(WebCore::JSHTMLFrameSetElement::canGetItemsForName):
(WebCore::JSHTMLFrameSetElement::nameGetter):

  • bindings/js/JSMessageEventCustom.cpp:

(WebCore::JSMessageEvent::initMessageEvent):

  • bindings/js/JSMessagePortCustom.cpp:

(WebCore::JSMessagePort::addEventListener):
(WebCore::JSMessagePort::removeEventListener):

  • bindings/js/JSMimeTypeArrayCustom.cpp:

(WebCore::JSMimeTypeArray::canGetItemsForName):
(WebCore::JSMimeTypeArray::nameGetter):

  • bindings/js/JSNodeCustom.cpp:

(WebCore::JSNode::addEventListener):
(WebCore::JSNode::removeEventListener):

  • bindings/js/JSNodeListCustom.cpp:

(WebCore::JSNodeList::canGetItemsForName):
(WebCore::JSNodeList::nameGetter):

  • bindings/js/JSPluginArrayCustom.cpp:

(WebCore::JSPluginArray::canGetItemsForName):
(WebCore::JSPluginArray::nameGetter):

  • bindings/js/JSPluginCustom.cpp:

(WebCore::JSPlugin::canGetItemsForName):
(WebCore::JSPlugin::nameGetter):

  • bindings/js/JSPopStateEventCustom.cpp:

(WebCore::JSPopStateEvent::initPopStateEvent):

  • bindings/js/JSSVGElementInstanceCustom.cpp:

(WebCore::JSSVGElementInstance::addEventListener):
(WebCore::JSSVGElementInstance::removeEventListener):

  • bindings/js/JSWebSocketCustom.cpp:

(WebCore::JSWebSocket::addEventListener):
(WebCore::JSWebSocket::removeEventListener):

  • bindings/js/JSWorkerContextCustom.cpp:

(WebCore::JSWorkerContext::addEventListener):
(WebCore::JSWorkerContext::removeEventListener):

  • bindings/js/JSXMLHttpRequestCustom.cpp:

(WebCore::JSXMLHttpRequest::setRequestHeader):
(WebCore::JSXMLHttpRequest::getResponseHeader):
(WebCore::JSXMLHttpRequest::addEventListener):
(WebCore::JSXMLHttpRequest::removeEventListener):

  • bindings/js/JSXMLHttpRequestUploadCustom.cpp:

(WebCore::JSXMLHttpRequestUpload::addEventListener):
(WebCore::JSXMLHttpRequestUpload::removeEventListener):

  • platform/text/AtomicString.cpp:

(WebCore::AtomicString::add):
(WebCore::AtomicString::find):

  • platform/text/AtomicString.h:

(WebCore::AtomicString::AtomicString):

16:58 Changeset [57765] by aroben@apple.com

Fix typo from r55729

  • platform/win/Skipped:
16:51 Changeset [57764] by weinig@apple.com

Fix window.open() and targeted links.

Reviewed by Anders Carlsson.

WebKit2:

  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::didReceiveSyncMessage): Pass in the new pageID
instead of 0.

  • WebProcess/WebProcess.cpp:

(WebKit::WebProcess::createWebPage): Allow for the page to have already
been created, as is the case with programmatic window opening from within
WebCore (e.g. window.open() or <a target="_blank">).

WebKitTools:

  • MiniBrowser/mac/BrowserWindowController.m:

(_createNewPage): Use the correct initializer to and load the window.

16:45 Changeset [57763] by ajwong@chromium.org

Not reviewed. Fixing chromium build.

[chromium] Upstream test_expectations.txt changes from chromium.
Mark letter_spacing.html as consistently failing for mac.
Fixing a syntax error in the test_expectations.txt file.

  • platform/chromium/test_expectations.txt:
16:31 Changeset [57762] by ajwong@chromium.org

Not reviewed. Fixing chromiumg build.

[chromium] filterRes.svg is consistently failing now. Marking as such.

  • platform/chromium/test_expectations.txt:
16:23 Changeset [57761] by weinig@apple.com

Don't optimize debug builds.

Reviewed by Mark Rowe.

  • WebKit2.xcodeproj/project.pbxproj: Define GCC_OPTIMIZATION_LEVEL correctly.
16:13 Changeset [57760] by ukai@chromium.org

2010-04-16 Fumitoshi Ukai <ukai@chromium.org>

Reviewed by Alexey Proskuryakov.

WebSocket crash when it receives bad header.
https://bugs.webkit.org/show_bug.cgi?id=37682

  • websocket/tests/bad-handshake-crash-expected.txt: Added.
  • websocket/tests/bad-handshake-crash.html: Added.
  • websocket/tests/bad-handshake-crash_wsh.py: Added.
  • websocket/tests/script-tests/bad-handshake-crash.js: Added.

2010-04-16 Fumitoshi Ukai <ukai@chromium.org>

Reviewed by Alexey Proskuryakov.

WebSocket crash when it receives bad header.
https://bugs.webkit.org/show_bug.cgi?id=37682

If name or value is not valid UTF-8, nameStr or valueStr would be
null string, so crashed in headers->add(nameStr, valueStr).
Check both nameStr and valueStr are not null string.
Otherwise handshake will fail.

Test: websocket/tests/bad-handshake-crash.html

  • websockets/WebSocketHandshake.cpp: (WebCore::WebSocketHandshake::readHTTPHeaders): check nameStr and valueStr are not null string.
16:10 Changeset [57759] by mitz@apple.com

<rdar://problem/7873647> Crash when updating hover state

Reviewed by Simon Fraser.

WebCore:

Test: fast/dynamic/hover-style-recalc-crash.html

Updating the hover state of an element caused the document to need style
recalc, and then updating the hover state of a link caused style recalc,
which changed the render tree while updateHoverActiveState() was iterating
over it, leading to a crash.

  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::updateHoverActiveState): Collect the nodes to be
updated into vectors, then update their active and hover states.

LayoutTests:

  • fast/dynamic/hover-style-recalc-crash-expected.txt: Added.
  • fast/dynamic/hover-style-recalc-crash.html: Added.
16:07 Changeset [57758] by ajwong@chromium.org

Not reviewed. Build fix.

[chromium] Compile fix. Missing include header, and missing type conversion.

  • src/GraphicsContext3D.cpp:

(WebCore::GraphicsContext3DInternal::beginPaint):

15:42 Changeset [57757] by andersca@apple.com

Fix build.

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::runJavaScriptInMainFrame):

15:31 Changeset [57756] by barraclough@apple.com

Bug 37730 - Remove JSC::UString dependencies from WebCore::StringImpl
(Following on from bug #37675).

Reviewed by Oliver Hunt.

Make the argument ordering for UStringImpl's constructor & create
methods match, when passed a shared buffer.

(JSC::UStringImpl::create):

  • runtime/UStringImpl.h:
15:29 Changeset [57755] by scherkus@chromium.org

2010-04-16 Andrew Scherkus <scherkus@chromium.org>

Unreviewed, fixing Chromium test_expectations.txt due to removed tests in r57292.

  • platform/chromium/test_expectations.txt:
15:11 Changeset [57754] by dumi@chromium.org

Make Safari correctly allocate 5MB of DB storage to all new
origins.
https://bugs.webkit.org/show_bug.cgi?id=36671

Reviewed by Alexey Proskuryakov.

Eric fixed the same problem in
DatabaseTracker::fullPathForDatabaseNoLock() in r57128, but forgot
to fix it in DatabaseTracker::detailsForNameAndOrigin() too.

  • storage/DatabaseTracker.cpp:

(WebCore::DatabaseTracker::detailsForNameAndOrigin):

15:06 Changeset [57753] by scherkus@chromium.org

2010-04-16 Andrew Scherkus <scherkus@chromium.org>

Unreviewed, checking in Chromium baselines for media/video-seek-past-end-playing.html.

  • platform/chromium-mac/media/video-seek-past-end-playing-expected.txt: Added.
  • platform/chromium-win/media/video-seek-past-end-playing-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
15:05 Changeset [57752] by barraclough@apple.com

Bug 37730 - Remove JSC::UString dependencies from WebCore::StringImpl
(Following on from bug #37675).

Reviewed by Oliver Hunt.

Make the argument ordering for UStringImpl's constructor & create
methods match, when passed a shared buffer.

(JSC::UStringImpl::create):

  • runtime/UStringImpl.h:
14:57 Changeset [57751] by mrowe@apple.com

Merge r57615.

14:56 Changeset [57750] by mrowe@apple.com

Versioning.

14:55 Changeset [57749] by kinuko@chromium.org

2010-04-16 Kinuko Yasuda <kinuko@chromium.org>

Reviewed by Jian Li.

Implement FileStreamProxy that calls FileStream methods on FileThread for FileAPI
https://bugs.webkit.org/show_bug.cgi?id=37218

No new tests; tests will be added when we add upper layer implementations.

  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • html/FileStream.cpp: (WebCore::FileStream::stop):
  • html/FileStream.h:
  • html/FileStreamClient.h: (WebCore::FileStreamClient::didStop):
  • html/FileStreamProxy.cpp: Added
  • html/FileStreamProxy.h: Added
  • html/FileThreadTask.h: Added
14:50 Changeset [57748] by barraclough@apple.com

Bug 37730 - Remove JSC::UString dependencies from WebCore::StringImpl
(Following on from bug #37675).

Reviewed by Oliver Hunt.

JavaScriptCore:

Make the argument ordering for UStringImpl's constructor & create
methods match, when passed a shared buffer.

(JSC::UStringImpl::create):

  • runtime/UStringImpl.h:

WebCore:

Remove ustring() method, and constructor passed a UString.

  • WebCore.base.exp:
  • bindings/js/JSDOMBinding.cpp:

(WebCore::stringimplToUString):
(WebCore::jsStringSlowCase):
(WebCore::ustringToString):
(WebCore::stringToUString):
(WebCore::identifierToString):

  • bindings/js/JSDOMBinding.h:

(WebCore::jsString):

  • platform/text/AtomicString.cpp:

(WebCore::AtomicString::operator UString):

  • platform/text/StringImpl.cpp:

(WebCore::StringImpl::create):

  • platform/text/StringImpl.h:
14:39 Changeset [57747] by eric@webkit.org

2010-04-16 Jarkko Sakkinen <jarkko.j.sakkinen@gmail.com>

Reviewed by Simon Hausmann.

[Qt] WebGL is not visible when QGLWidget viewport is used
https://bugs.webkit.org/show_bug.cgi?id=37070

Added HostWindow parameter to the constructor of GraphicsContext3D.
Shared OpenGL context is initialized with parent QGLWidget.

  • html/canvas/WebGLRenderingContext.cpp: (WebCore::WebGLRenderingContext::create):
  • platform/graphics/GraphicsContext3D.h:
  • platform/graphics/mac/GraphicsContext3DMac.cpp: (WebCore::GraphicsContext3D::create): (WebCore::GraphicsContext3D::GraphicsContext3D):
  • platform/graphics/qt/GraphicsContext3DQt.cpp: (WebCore::GraphicsContext3DInternal::GraphicsContext3DInternal): (WebCore::GraphicsContext3DInternal::~GraphicsContext3DInternal): (WebCore::GraphicsContext3DInternal::getOwnerGLWidget): (WebCore::GraphicsContext3D::create): (WebCore::GraphicsContext3D::GraphicsContext3D):

2010-04-16 Jarkko Sakkinen <jarkko.j.sakkinen@gmail.com>

Reviewed by Simon Hausmann.

[Qt] WebGL is not visible when QGLWidget viewport is used
https://bugs.webkit.org/show_bug.cgi?id=37070

Added HostWindow parameter to the constructor of GraphicsContext3D.
Shared OpenGL context is initialized with parent QGLWidget.

  • src/GraphicsContext3D.cpp: (WebCore::GraphicsContext3D::GraphicsContext3D):
14:39 Changeset [57746] by mrowe@apple.com

New tag.

14:27 Changeset [57745] by eric@webkit.org

2010-04-16 Jarkko Sakkinen <jarkko.j.sakkinen@gmail.com>

Reviewed by Simon Hausmann.

[Qt] WebKit compilation fails with --3d-canvas
https://bugs.webkit.org/show_bug.cgi?id=37699

API for readPixels() has been changed.

  • platform/graphics/qt/GraphicsContext3DQt.cpp: (WebCore::GraphicsContext3D::readPixels):
14:16 Changeset [57744] by eric@webkit.org

2010-04-16 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Antti Koivisto.

[Qt]QtLauncher crash on page with CSS 3D transform
https://bugs.webkit.org/show_bug.cgi?id=36859

Added a neccessary null-pointer check, lack of which created the crash circumstances.

Tested by http://css-vfx.googlecode.com/svn/trunk/snowstack/snowstack.html

  • platform/graphics/qt/GraphicsLayerQt.cpp: (WebCore::GraphicsLayerQtImpl::flushChanges):
14:14 Changeset [57743] by scherkus@chromium.org

2010-04-16 Sergey Ulanov <sergeyu@chromium.org>

Unreviewed, checking in baselines due to changes to Chromium's video scaling algorithm.

  • platform/chromium-linux/media/controls-after-reload-expected.checksum:
  • platform/chromium-linux/media/controls-after-reload-expected.png:
  • platform/chromium-linux/media/controls-strict-expected.checksum:
  • platform/chromium-linux/media/controls-strict-expected.png:
  • platform/chromium-linux/media/video-aspect-ratio-expected.checksum:
  • platform/chromium-linux/media/video-aspect-ratio-expected.png:
  • platform/chromium-linux/media/video-controls-rendering-expected.checksum:
  • platform/chromium-linux/media/video-controls-rendering-expected.png:
  • platform/chromium-linux/media/video-display-toggle-expected.checksum:
  • platform/chromium-linux/media/video-display-toggle-expected.png:
  • platform/chromium-linux/media/video-layer-crash-expected.checksum:
  • platform/chromium-linux/media/video-layer-crash-expected.png:
  • platform/chromium-linux/media/video-no-audio-expected.checksum:
  • platform/chromium-linux/media/video-no-audio-expected.png:
  • platform/chromium-linux/media/video-transformed-expected.checksum:
  • platform/chromium-linux/media/video-transformed-expected.png:
  • platform/chromium-linux/media/video-volume-slider-expected.checksum:
  • platform/chromium-linux/media/video-volume-slider-expected.png:
  • platform/chromium-linux/media/video-zoom-expected.checksum:
  • platform/chromium-linux/media/video-zoom-expected.png:
  • platform/chromium-mac/media/controls-after-reload-expected.checksum:
  • platform/chromium-mac/media/controls-after-reload-expected.png:
  • platform/chromium-mac/media/controls-strict-expected.checksum:
  • platform/chromium-mac/media/controls-strict-expected.png:
  • platform/chromium-mac/media/video-aspect-ratio-expected.checksum:
  • platform/chromium-mac/media/video-aspect-ratio-expected.png:
  • platform/chromium-mac/media/video-controls-rendering-expected.checksum:
  • platform/chromium-mac/media/video-controls-rendering-expected.png:
  • platform/chromium-mac/media/video-display-toggle-expected.checksum:
  • platform/chromium-mac/media/video-display-toggle-expected.png:
  • platform/chromium-mac/media/video-layer-crash-expected.checksum:
  • platform/chromium-mac/media/video-layer-crash-expected.png:
  • platform/chromium-mac/media/video-no-audio-expected.checksum:
  • platform/chromium-mac/media/video-no-audio-expected.png:
  • platform/chromium-mac/media/video-transformed-expected.checksum:
  • platform/chromium-mac/media/video-transformed-expected.png:
  • platform/chromium-mac/media/video-volume-slider-expected.checksum:
  • platform/chromium-mac/media/video-volume-slider-expected.png:
  • platform/chromium-mac/media/video-zoom-expected.checksum:
  • platform/chromium-mac/media/video-zoom-expected.png:
  • platform/chromium-win/media/controls-after-reload-expected.checksum:
  • platform/chromium-win/media/controls-after-reload-expected.png:
  • platform/chromium-win/media/controls-strict-expected.checksum:
  • platform/chromium-win/media/controls-strict-expected.png:
  • platform/chromium-win/media/video-aspect-ratio-expected.checksum:
  • platform/chromium-win/media/video-aspect-ratio-expected.png:
  • platform/chromium-win/media/video-controls-rendering-expected.checksum:
  • platform/chromium-win/media/video-controls-rendering-expected.png:
  • platform/chromium-win/media/video-display-toggle-expected.checksum:
  • platform/chromium-win/media/video-display-toggle-expected.png:
  • platform/chromium-win/media/video-layer-crash-expected.checksum:
  • platform/chromium-win/media/video-layer-crash-expected.png:
  • platform/chromium-win/media/video-no-audio-expected.checksum:
  • platform/chromium-win/media/video-no-audio-expected.png:
  • platform/chromium-win/media/video-transformed-expected.checksum:
  • platform/chromium-win/media/video-transformed-expected.png:
  • platform/chromium-win/media/video-volume-slider-expected.checksum:
  • platform/chromium-win/media/video-volume-slider-expected.png:
14:08 Changeset [57742] by aroben@apple.com

Fix links to layout test results from build status pages

Reviewed by Mark Rowe.

  • BuildSlaveSupport/build.webkit.org-config/master.cfg:

(ExtractTestResults.finished): Prepend "/" on the URL of the test
results page so that it is treated as an absolute URL.

13:52 Changeset [57741] by andersca@apple.com

Make sure to update the current graphics context when calling out to AppKit.

Reviewed by Sam Weinig.

  • platform/mac/ThemeMac.mm:

(WebCore::paintCheckbox):
(WebCore::paintRadio):

13:34 Changeset [57740] by barraclough@apple.com
  • WebFrame.cpp:

(WebFrame::stringByEvaluatingJavaScriptInScriptWorld):

  • WebView.cpp:

(WebView::stringByEvaluatingJavaScriptFromString):

Reviewed by NOBODY (Windows build fix).

13:26 Changeset [57739] by andersca@apple.com

Fix windows build.

Reviewed by Sam Weinig.

  • Platform/win/RunLoopWin.cpp:

(RunLoop::run):

  • UIProcess/Launcher/win/WebProcessLauncher.cpp:

(WebKit::webThreadBody):
(WebKit::launchWebProcess):

  • WebProcess/win/WebProcessMain.cpp:

(WebKit::WebProcessMain):

13:21 Changeset [57738] by barraclough@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37675
Remove casts/constructors to/from JSC::UString type from WebCore::String

Reviewed by Sam Weinig & Oliver Hunt.

WebCore's strings should not know about JSC::UString, this should be abstracted
away in the bindings. Add explicit conversion methods rather than relying on
overloaded cast operators / constructors being implicitly called.

This patch only changes the class String, once this has landed StringImpl, and
hopefully AtomicString too, should follow suit.

WebCore:

This patch adds:

WebCore::identifierToString
WebCore::ustringToString
WebCore::stringToUString

  • to JSDOMBindings.h, and updates code to call these methods.
  • WebCore.base.exp:
  • WebCore.order:
  • bindings/js/CachedScriptSourceProvider.h:

(WebCore::CachedScriptSourceProvider::CachedScriptSourceProvider):

  • bindings/js/JSAudioConstructor.cpp:

(WebCore::constructAudio):

  • bindings/js/JSCSSStyleDeclarationCustom.cpp:

(WebCore::JSCSSStyleDeclaration::nameGetter):

  • bindings/js/JSCanvasRenderingContext2DCustom.cpp:

(WebCore::toHTMLCanvasStyle):
(WebCore::JSCanvasRenderingContext2D::setFillColor):
(WebCore::JSCanvasRenderingContext2D::setStrokeColor):
(WebCore::JSCanvasRenderingContext2D::drawImageFromRect):
(WebCore::JSCanvasRenderingContext2D::setShadow):
(WebCore::JSCanvasRenderingContext2D::fillText):
(WebCore::JSCanvasRenderingContext2D::strokeText):

  • bindings/js/JSClipboardCustom.cpp:

(WebCore::JSClipboard::types):
(WebCore::JSClipboard::clearData):
(WebCore::JSClipboard::getData):
(WebCore::JSClipboard::setData):

  • bindings/js/JSCustomXPathNSResolver.cpp:

(WebCore::JSCustomXPathNSResolver::lookupNamespaceURI):

  • bindings/js/JSDOMBinding.cpp:

(WebCore::identifierToString):
(WebCore::ustringToString):
(WebCore::stringToUString):
(WebCore::valueToStringWithNullCheck):
(WebCore::valueToStringWithUndefinedOrNullCheck):
(WebCore::reportException):

  • bindings/js/JSDOMBinding.h:
  • bindings/js/JSDOMFormDataCustom.cpp:

(WebCore::JSDOMFormData::append):

  • bindings/js/JSDOMWindowCustom.cpp:

(WebCore::namedItemGetter):
(WebCore::JSDOMWindow::setLocation):
(WebCore::JSDOMWindow::openDatabase):

  • bindings/js/JSDatabaseCustom.cpp:

(WebCore::JSDatabase::changeVersion):

  • bindings/js/JSDocumentCustom.cpp:

(WebCore::JSDocument::setLocation):

  • bindings/js/JSEventListener.cpp:

(WebCore::JSEventListener::handleEvent):

  • bindings/js/JSEventSourceConstructor.cpp:

(WebCore::constructEventSource):

  • bindings/js/JSHTMLAllCollectionCustom.cpp:

(WebCore::callHTMLAllCollection):

  • bindings/js/JSHTMLCanvasElementCustom.cpp:

(WebCore::JSHTMLCanvasElement::getContext):

  • bindings/js/JSHTMLCollectionCustom.cpp:

(WebCore::callHTMLCollection):

  • bindings/js/JSHTMLDocumentCustom.cpp:

(WebCore::JSHTMLDocument::nameGetter):
(WebCore::documentWrite):

  • bindings/js/JSInjectedScriptHostCustom.cpp:

(WebCore::InjectedScriptHost::createInjectedScript):

  • bindings/js/JSInspectorFrontendHostCustom.cpp:

(WebCore::JSInspectorFrontendHost::showContextMenu):

  • bindings/js/JSLazyEventListener.cpp:

(WebCore::JSLazyEventListener::initializeJSFunction):

  • bindings/js/JSLocationCustom.cpp:

(WebCore::JSLocation::setHref):
(WebCore::JSLocation::setProtocol):
(WebCore::JSLocation::setHost):
(WebCore::JSLocation::setHostname):
(WebCore::JSLocation::setPathname):
(WebCore::JSLocation::setSearch):
(WebCore::JSLocation::setHash):
(WebCore::JSLocation::replace):
(WebCore::JSLocation::assign):

  • bindings/js/JSMessageEventCustom.cpp:

(WebCore::JSMessageEvent::initMessageEvent):

  • bindings/js/JSNamedNodeMapCustom.cpp:

(WebCore::JSNamedNodeMap::canGetItemsForName):
(WebCore::JSNamedNodeMap::nameGetter):

  • bindings/js/JSOptionConstructor.cpp:

(WebCore::constructHTMLOptionElement):

  • bindings/js/JSSQLResultSetRowListCustom.cpp:

(WebCore::JSSQLResultSetRowList::item):

  • bindings/js/JSSQLTransactionCustom.cpp:

(WebCore::JSSQLTransaction::executeSql):

  • bindings/js/JSSharedWorkerConstructor.cpp:

(WebCore::constructSharedWorker):

  • bindings/js/JSStorageCustom.cpp:

(WebCore::JSStorage::canGetItemsForName):
(WebCore::JSStorage::nameGetter):
(WebCore::JSStorage::deleteProperty):
(WebCore::JSStorage::getOwnPropertyNames):
(WebCore::JSStorage::putDelegate):

  • bindings/js/JSStyleSheetListCustom.cpp:

(WebCore::JSStyleSheetList::canGetItemsForName):
(WebCore::JSStyleSheetList::nameGetter):

  • bindings/js/JSWebKitCSSMatrixConstructor.cpp:

(WebCore::constructWebKitCSSMatrix):

  • bindings/js/JSWebSocketConstructor.cpp:

(WebCore::constructWebSocket):

  • bindings/js/JSWebSocketCustom.cpp:

(WebCore::JSWebSocket::send):

  • bindings/js/JSWorkerConstructor.cpp:

(WebCore::constructWorker):

  • bindings/js/JSWorkerContextCustom.cpp:

(WebCore::JSWorkerContext::importScripts):

  • bindings/js/JSXMLHttpRequestCustom.cpp:

(WebCore::JSXMLHttpRequest::open):
(WebCore::JSXMLHttpRequest::setRequestHeader):
(WebCore::JSXMLHttpRequest::send):
(WebCore::JSXMLHttpRequest::overrideMimeType):

  • bindings/js/JSXSLTProcessorCustom.cpp:

(WebCore::JSXSLTProcessor::setParameter):
(WebCore::JSXSLTProcessor::getParameter):
(WebCore::JSXSLTProcessor::removeParameter):

  • bindings/js/ScheduledAction.cpp:

(WebCore::ScheduledAction::create):

  • bindings/js/ScriptCallFrame.cpp:

(WebCore::ScriptCallFrame::ScriptCallFrame):

  • bindings/js/ScriptController.cpp:

(WebCore::ScriptController::evaluateInWorld):

  • bindings/js/ScriptDebugServer.cpp:

(WebCore::ScriptDebugServer::hasBreakpoint):
(WebCore::ScriptDebugServer::dispatchDidParseSource):
(WebCore::ScriptDebugServer::dispatchFailedToParseSource):
(WebCore::ScriptDebugServer::sourceParsed):

  • bindings/js/ScriptEventListener.cpp:

(WebCore::getEventListenerHandlerBody):

  • bindings/js/ScriptFunctionCall.cpp:

(WebCore::ScriptFunctionCall::appendArgument):
(WebCore::ScriptFunctionCall::call):
(WebCore::ScriptFunctionCall::construct):

  • bindings/js/ScriptObject.cpp:

(WebCore::ScriptObject::set):

  • bindings/js/ScriptProfiler.cpp:

(WebCore::ScriptProfiler::start):
(WebCore::ScriptProfiler::stop):

  • bindings/js/ScriptString.h:

(WebCore::ScriptString::operator String):
(WebCore::ScriptString::ustring):
(WebCore::ScriptString::operator+=):

  • bindings/js/ScriptValue.cpp:

(WebCore::ScriptValue::getString):

  • bindings/js/ScriptValue.h:

(WebCore::ScriptValue::toString):

  • bindings/js/SerializedScriptValue.cpp:

(WebCore::SerializedObject::set):
(WebCore::SerializingTreeWalker::convertIfTerminal):
(WebCore::DeserializingTreeWalker::putProperty):

  • bindings/js/StringSourceProvider.h:

(WebCore::StringSourceProvider::StringSourceProvider):

  • bindings/objc/WebScriptObject.mm:

(-[WebScriptObject callWebScriptMethod:withArguments:]):
(-[WebScriptObject setValue:forKey:]):
(-[WebScriptObject valueForKey:]):
(-[WebScriptObject removeWebScriptKey:]):

  • bindings/scripts/CodeGeneratorJS.pm:
  • bridge/IdentifierRep.cpp:

(WebCore::IdentifierRep::get):

  • bridge/c/c_utility.cpp:

(JSC::Bindings::identifierFromNPIdentifier):

  • inspector/InspectorController.cpp:

(WebCore::InspectorController::scriptImported):
(WebCore::InspectorController::addProfileFinishedMessageToConsole):
(WebCore::InspectorController::createProfileHeader):

  • inspector/InspectorResource.cpp:

(WebCore::InspectorResource::sourceString):

  • inspector/JavaScriptCallFrame.cpp:

(WebCore::JavaScriptCallFrame::functionName):

  • platform/KURL.h:

(WebCore::KURL::operator const String&):

  • platform/text/AtomicString.cpp:

(WebCore::AtomicString::operator UString):

  • platform/text/AtomicString.h:
  • platform/text/PlatformString.h:
  • platform/text/String.cpp:

WebKit/mac:

  • Plugins/Hosted/NetscapePluginHostProxy.mm:

(identifierFromIdentifierRep):

  • Plugins/Hosted/NetscapePluginInstanceProxy.mm:

(WebKit::NetscapePluginInstanceProxy::addValueToArray):
(WebKit::NetscapePluginInstanceProxy::moveGlobalExceptionToExecState):

  • Plugins/Hosted/ProxyInstance.mm:

(WebKit::ProxyInstance::getPropertyNames):

  • WebView/WebFrame.mm:

(-[WebFrame _stringByEvaluatingJavaScriptFromString:forceUserGesture:]):
(-[WebFrame _stringByEvaluatingJavaScriptFromString:withGlobalObject:inScriptWorld:]):

  • WebView/WebScriptDebugDelegate.mm:

(-[WebScriptCallFrame evaluateWebScript:]):

  • WebView/WebScriptDebugger.mm:

(toNSURL):

  • WebView/WebView.mm:

(aeDescFromJSValue):

13:15 Changeset [57737] by weinig@apple.com

Make resizing responsive on Windows.

Reviewed by Anders Carlsson.

  • Use the same waitFor logic as do for the Mac resizing DrawingAreaUpdateChunk code.
  • Shared/win/UpdateChunk.cpp:

(WebKit::UpdateChunk::UpdateChunk):

  • Shared/win/UpdateChunk.h: Add a constructor that only takes an IntRect and allocates the shared memory mapping for you.
  • UIProcess/win/DrawingAreaProxy.cpp:

(WebKit::DrawingAreaProxy::DrawingAreaProxy):
(WebKit::DrawingAreaProxy::ensureBackingStore):
(WebKit::DrawingAreaProxy::paint):
(WebKit::DrawingAreaProxy::drawUpdateChunkIntoBackingStore):
(WebKit::DrawingAreaProxy::setSize):
(WebKit::DrawingAreaProxy::didSetSize):
(WebKit::DrawingAreaProxy::didReceiveMessage):

  • UIProcess/win/DrawingAreaProxy.h:

Perform wait in paint as we do on the mac.

  • UIProcess/win/WebView.cpp:

(WebKit::WebView::onSizeEvent):
Change to use an IntSize.

  • WebProcess/WebPage/win/DrawingAreaUpdateChunk.cpp:

(WebKit::DrawingAreaUpdateChunk::paintIntoUpdateChunk):
(WebKit::DrawingAreaUpdateChunk::display):
(WebKit::DrawingAreaUpdateChunk::setSize):

  • WebProcess/WebPage/win/DrawingAreaUpdateChunk.h:

Specialize setSize() drawing and factor out painting
into a helper function.

13:13 Changeset [57736] by ajwong@chromium.org

Not reviewed. Build fix for chromium.

[chromium] Add in new baselines for chromium-mac and chromium-win
also since this test there as well.

  • platform/chromium-mac/fast/url/host-expected.txt: Copied from LayoutTests/platform/chromium-linux/fast/url/host-expected.txt.
  • platform/chromium-mac/fast/url/path-expected.txt: Copied from LayoutTests/platform/chromium-linux/fast/url/path-expected.txt.
  • platform/chromium-mac/fast/url/standard-url-expected.txt:
  • platform/chromium-win/fast/url/host-expected.txt: Copied from LayoutTests/platform/chromium-linux/fast/url/host-expected.txt.
  • platform/chromium-win/fast/url/path-expected.txt: Copied from LayoutTests/platform/chromium-linux/fast/url/path-expected.txt.
  • platform/chromium-win/fast/url/standard-url-expected.txt:
13:05 Changeset [57735] by ajwong@chromium.org

Not reviewed. Build fix for chromium.

[chromium] Add in new baselines for chromium-linux since this test
fully passes on this platform.

  • platform/chromium-linux/fast/url/host-expected.txt: Added.
  • platform/chromium-linux/fast/url/path-expected.txt: Added.
  • platform/chromium-linux/fast/url/standard-url-expected.txt: Added.
12:58 Changeset [57734] by andersca@apple.com

2010-04-16 Anders Carlsson <andersca@apple.com>

Reviewed by David Hyatt.

Always pass a view to the AppKit drawing functions.
https://bugs.webkit.org/show_bug.cgi?id=37724

  • platform/mac/ThemeMac.mm: (WebCore::paintCheckbox): (WebCore::paintRadio):
  • rendering/RenderThemeMac.mm: (WebCore::RenderThemeMac::paintMenuList): (WebCore::RenderThemeMac::paintSliderThumb): (WebCore::RenderThemeMac::paintSearchField): (WebCore::RenderThemeMac::paintSearchFieldCancelButton): (WebCore::RenderThemeMac::paintSearchFieldResultsDecoration): (WebCore::RenderThemeMac::paintSearchFieldResultsButton):
12:45 Changeset [57733] by ap@apple.com

Reviewed by Dan Bernstein.

https://bugs.webkit.org/show_bug.cgi?id=37718
Safari crashes with certain JavaScript charCode events in EventHandler::needsKeyboardEventDisambiguationQuirks

Test: fast/events/recorded-keydown-event.html

  • dom/KeyboardEvent.cpp: (WebCore::KeyboardEvent::charCode): Check if the view (window) is frameless.
12:36 Changeset [57732] by andersca@apple.com

2010-04-16 Anders Carlsson <andersca@apple.com>

Reviewed by David Hyatt.

Make run loops be allocated as thread specific data.
https://bugs.webkit.org/show_bug.cgi?id=37723

  • Platform/RunLoop.cpp: (RunLoop::initializeMainRunLoop): (RunLoop::current): (RunLoop::main):
  • Platform/RunLoop.h:
  • Platform/mac/RunLoopMac.mm: (RunLoop::run): (RunLoop::stop):
  • UIProcess/Launcher/mac/WebProcessLauncher.mm: (WebKit::webThreadBody): (WebKit::launchWebProcess):
  • UIProcess/ResponsivenessTimer.cpp: (WebKit::ResponsivenessTimer::ResponsivenessTimer):
  • WebProcess/Launching/mac/WebProcessMain.mm: (main):
  • WebProcess/WebProcess.cpp: (WebKit::WebProcess::isSeparateProcess):
11:52 Changeset [57731] by aroben@apple.com

Skip HTMLFrameSetElement-window-eventListener-attributes.html on SnowLeopard

It is crashing sometimes. <http://webkit.org/b/37720> tracks investigating this.

Rubber-stamped by Sam Weinig.

  • platform/mac-snowleopard/Skipped:
11:40 Changeset [57730] by aroben@apple.com

Make it possible for clients to instantiate a WebUserContentURLPattern

Reviewed by Tim Hatcher.

  • ForEachCoClass.h: Added WebUserContentURLPattern to the FOR_EACH_COCLASS macro, which

lists all our instantiatable classes.

  • WebKitClassFactory.cpp: Added a now-required #include.
11:19 Changeset [57729] by aroben@apple.com

SKip xss-DENIED-mime-type-execute-as-html.html to see if it makes the http test stop timing out on Windows

<http://webkit.org/b/37714> tracks this investigation.

Rubber-stamped by Sam Weinig.

  • platform/win/Skipped:
11:19 Changeset [57728] by ukai@chromium.org

2010-04-16 Fumitoshi Ukai <ukai@chromium.org>

Reviewed by Jian Li.

[chromium] WebKit::WebURLResponse::addHTTPHeaderField will crash if response is invalid UTF-8
https://bugs.webkit.org/show_bug.cgi?id=37687

  • src/WebURLResponse.cpp: (WebKit::WebURLResponse::addHTTPHeaderField): ignore if name or value is null string.
11:07 Changeset [57727] by dglazkov@chromium.org

2010-04-16 Jian Li <jianli@chromium.org>

Reviewed by Dmitry Titov.

Update FormDataList to fix style violations in old code.
https://bugs.webkit.org/show_bug.cgi?id=37689

  • html/FormDataList.h: (WebCore::FormDataList::appendData): (WebCore::FormDataList::appendBlob):
11:00 Changeset [57726] by jianli@chromium.org

Update FormDataList to fix style violations in old code.
https://bugs.webkit.org/show_bug.cgi?id=37689

Reviewed by Dmitry Titov.

  • html/FormDataList.h:

(WebCore::FormDataList::appendData):
(WebCore::FormDataList::appendBlob):

10:59 Changeset [57725] by andersca@apple.com

Fix WebKit2 build.

  • WebCore.base.exp:
10:18 Changeset [57724] by eric@webkit.org

2010-04-16 Jay Civelli <jcivelli@chromium.org>

Reviewed by Dimitri Glazkov.

[chromium] Select popups would assert when destroyed.
https://bugs.webkit.org/show_bug.cgi?id=37436

  • platform/chromium/PopupMenuChromium.cpp: (WebCore::PopupContainer::PopupContainer): (WebCore::PopupContainer::showPopup): (WebCore::PopupContainer::notifyPopupHidden):
  • platform/chromium/PopupMenuChromium.h:

2010-04-16 Jay Civelli <jcivelli@chromium.org>

Reviewed by Dimitri Glazkov.

[chromium] Select popups would assert when destroyed.
Also adding unit-tests for the select popup code.
https://bugs.webkit.org/show_bug.cgi?id=37436

  • WebKit.gyp:
  • src/WebViewImpl.h: (WebKit::WebViewImpl::selectPopup):
  • tests/PopupMenuTest.cpp: Added.
10:09 Changeset [57723] by mitz@apple.com

Expose the extremal shrink factors WebHTMLView uses when shrinking pages to fit in the
printing width as SPI.

Reviewed by John Sullivan.

  • WebView/WebHTMLView.mm:

Replaced two macros with constants.
(-[WebHTMLView _beginPrintModeWithPageWidth:shrinkToFit:]): Changed to use the constants.
(-[WebHTMLView _scaleFactorForPrintOperation:]): Ditto.

  • WebView/WebHTMLViewPrivate.h: Declared _WebHTMLViewPrintingMinimumShrinkFactor and

_WebHTMLViewPrintingMaximumShrinkFactor.

09:18 Changeset [57722] by eric@webkit.org

2010-04-16 Sheriff Bot <webkit.review.bot@gmail.com>

Unreviewed, rolling out r57719.
http://trac.webkit.org/changeset/57719
https://bugs.webkit.org/show_bug.cgi?id=37708

Broke 8 URL tests on Windows due to use of file:///tmp/ URLs,
which don't work on Windows. (Requested by aroben on #webkit).

  • fast/url/anchor.html:
  • fast/url/file-expected.txt: Removed.
  • fast/url/file.html: Removed.
  • fast/url/ipv4.html:
  • fast/url/path-url.html:
  • fast/url/port.html:
  • fast/url/query.html:
  • fast/url/script-tests/TEMPLATE.html:
  • fast/url/script-tests/file.js: Removed.
  • fast/url/standard-url.html:
  • fast/url/trivial.html:
07:11 Changeset [57721] by tonikitoo@webkit.org

2010-04-16 Antonio Gomes <tonikitoo@webkit.org>

Unreviewed QtWebKit (with Qt 4.7) build fix.

Bug 37683 moved code from FontQt.cpp to FontPlatformDataQt.cpp but did not
renamed the variable used.

  • platform/graphics/qt/FontPlatformDataQt.cpp: (WebCore::FontPlatformData::FontPlatformData):
01:06 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
00:36 Changeset [57720] by eric@webkit.org

2010-04-16 Tony Chang <tony@chromium.org>

Reviewed by Dimitri Glazkov.

[chromium] build DRT on Linux
https://bugs.webkit.org/show_bug.cgi?id=37690

  • DEPS: Need to roll deps to remove a dependency on src/chrome
  • gyp_webkit: generate makefiles for DRT.gyp on Linux

2010-04-16 Tony Chang <tony@chromium.org>

Reviewed by Dimitri Glazkov.

[chromium] build DRT on Linux
https://bugs.webkit.org/show_bug.cgi?id=37690

  • Scripts/build-dumprendertree: Add support for win and linux
00:27 Changeset [57719] by mjs@apple.com

2010-04-16 Maciej Stachowiak <mjs@apple.com>

Reviewed by Adam Barth.

Test file: URL canonicalization
https://bugs.webkit.org/show_bug.cgi?id=37700

  • fast/url/script-tests/file.js: Added new test.
  • fast/url/file-expected.txt: Added (expected results).
  • fast/url/file.html: Added (wrapper).
  • fast/url/script-tests/TEMPLATE.html: Add an explicit dependable <base> URL to the template, to avoid making test expectations system dependent.


Other test wrappers regenerated for the above change:

  • fast/url/anchor.html:
  • fast/url/ipv4.html:
  • fast/url/path-url.html:
  • fast/url/port.html:
  • fast/url/query.html:
  • fast/url/standard-url.html:
  • fast/url/trivial.html:
00:25 Changeset [57718] by eric@webkit.org

2010-04-16 Noam Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

Make GraphicsLayerQt always use ItemCoordinateCache, and remove ItemUsesExtendedStyleOption.
This aligns our implementation with the Safari implementation - layers are always uploaded
to textures in item units, and WebCore is responsible for the heuristics.

[Qt] GraphicsLayer: performance optimizations
https://bugs.webkit.org/show_bug.cgi?id=35393

No new tests. Still no FPS benchmarks available (on any platform)
but animations are noticably better.

  • platform/graphics/qt/GraphicsLayerQt.cpp: (WebCore::GraphicsLayerQtImpl::GraphicsLayerQtImpl): (WebCore::GraphicsLayerQtImpl::paint): (WebCore::GraphicsLayerQtImpl::flushChanges): (WebCore::TransformAnimationQt::updateState):
00:20 QtWebKitRelease20 edited by henry.haverinen@nokia.com
(diff)
00:19 QtWebKit edited by henry.haverinen@nokia.com
(diff)
00:14 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
00:14 Changeset [57717] by eric@webkit.org

2010-04-16 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] WebCore::Font::font() causes a QFont detach
https://bugs.webkit.org/show_bug.cgi?id=37683

Moved the setStyleStrategy call to FontPlatformData
to avoid the detach.

Thanks to Holger for spotting this.

  • platform/graphics/qt/FontPlatformDataQt.cpp: (WebCore::FontPlatformData::FontPlatformData):
  • platform/graphics/qt/FontQt.cpp: (WebCore::Font::font):
00:13 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
00:10 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
00:03 Changeset [57716] by eric@webkit.org

2010-04-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add EFL port files to efl/WebCoreSupport.
http://webkit.org/b/35915

  • efl/WebCoreSupport/FrameLoaderClientEfl.cpp: Added.
  • efl/WebCoreSupport/FrameLoaderClientEfl.h: Added.

04/15/10:

23:20 Changeset [57715] by eric@webkit.org

2010-04-15 Matt Perry <mpcomplete@chromium.org>

Reviewed by Adam Barth.

Remove the check for the main frame's scheme when deciding which
v8 extensions to add to a script context. Instead, Chromium will
handle that check elsewhere to allow finer-grained control over
what APIs we expose to web pages.
https://bugs.webkit.org/show_bug.cgi?id=37681

  • bindings/v8/V8DOMWindowShell.cpp: (WebCore::V8DOMWindowShell::createNewContext):
22:59 Changeset [57714] by eric@webkit.org

2010-04-15 Luiz Agostini <luiz.agostini@openbossa.org>

Reviewed by Antti Koivisto.

No default selection for <select multiple> menu lists.
https://bugs.webkit.org/show_bug.cgi?id=37530

Manual test: manual-tests/no-listbox-rendering.html

For menu lists, if the selection is not indicated by the html file, the first <option> will be
selected after loading the page or reseting the form. On the other hand listboxes may have no
element selected after loading the page or reseting the form.

When NO_LISTBOX_RENDERING is enabled listboxes becomes menu lists. Those <select multiple>
that did not have selected elements, now being menu lists, will have the first <option>
selected. That is the behavior difference that this patch corrects.

When NO_LISTBOX_RENDERING is enabled usesMenuList() always returns true then usesMenuList() cannot
be used to decide about initial selection of the elements. This patch replaces (usesMenuLists())
by (!multiple && size <= 1) where initial selection is considered.

  • dom/SelectElement.cpp: (WebCore::SelectElement::recalcListItems): (WebCore::SelectElement::reset):
  • manual-tests/no-listbox-rendering.html: Added.
22:36 Changeset [57713] by tony@chromium.org

2010-04-15 Tony Chang <tony@chromium.org>

Not reviewed, trying to make the tree green.

[qt] fast/url/hosts.html is failing, skip until someone can
investigate.
https://bugs.webkit.org/show_bug.cgi?id=37697

  • platform/qt/Skipped:
19:50 Changeset [57712] by eric@webkit.org

2010-04-15 Adam Barth <abarth@webkit.org>

Reviewed by Jeremy Orlow.

Clean up unicode URL tests cases
https://bugs.webkit.org/show_bug.cgi?id=37678

When we started writing these URL tests, I didn't understand how to
handle the unicodes test cases properly. With a bunch of help (and a
pointer to http://rishida.net/tools/conversion/), I've fixed these to
be more or less faithful to the original unit tests.

  • fast/url/anchor-expected.txt:
  • fast/url/query-expected.txt:
  • fast/url/script-tests/anchor.js:
  • fast/url/script-tests/query.js:
  • fast/url/script-tests/standard-url.js:
  • fast/url/script-tests/trivial.js:
    • Updated this test to use the shared implementations
    • Also, removed assumptions about the base URL.
  • fast/url/standard-url-expected.txt:
  • fast/url/trivial-expected.txt:
19:38 Changeset [57711] by eric@webkit.org

2010-04-15 Adam Barth <abarth@webkit.org>

Reviewed by Jeremy Orlow.

Add URL parsing tests for host names
https://bugs.webkit.org/show_bug.cgi?id=37666

  • fast/url/host-expected.txt: Added.
  • fast/url/host.html: Added.
  • fast/url/script-tests/host.js: Added.
19:27 Changeset [57710] by eric@webkit.org

2010-04-15 Adam Barth <abarth@webkit.org>

Reviewed by Jeremy Orlow.

Add URL parsing tests for paths
https://bugs.webkit.org/show_bug.cgi?id=37672

  • fast/url/path-expected.txt: Added.
  • fast/url/path.html: Added.
  • fast/url/script-tests/path.js: Added.
19:01 Changeset [57709] by eric@webkit.org

2010-04-15 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Index validation code validates too many vertex attributes
https://bugs.webkit.org/show_bug.cgi?id=31892

  • fast/canvas/webgl/index-validation-expected.txt: Add new test cases for index validations.
  • fast/canvas/webgl/script-tests/index-validation.js: Ditto.

2010-04-15 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Index validation code validates too many vertex attributes
https://bugs.webkit.org/show_bug.cgi?id=31892

  • html/canvas/WebGLProgram.cpp: (WebCore::WebGLProgram::cacheActiveAttribLocations): Cache active attribute locations for a program at linkProgram time. (WebCore::WebGLProgram::getActiveAttribLocation): Get the cached attribute location. (WebCore::WebGLProgram::numActiveAttribLocations): Get the number of cached attribute locations.
  • html/canvas/WebGLProgram.h: Add attribute locations member.
  • html/canvas/WebGLRenderingContext.cpp: (WebCore::WebGLRenderingContext::validateRenderingState): Add logic not to validate those attributes that do not belong to the current program. (WebCore::WebGLRenderingContext::linkProgram): Call cacheActiveAttribLocations().
18:38 Changeset [57708] by jianli@chromium.org

Unreviewed.

Mark send-form-data-with-sliced-file.html as failed on Chromium Linux.

  • platform/chromium/test_expectations.txt:
18:34 Changeset [57707] by weinig@apple.com

Use GDI text rendering on Windows by default.

Reviewed by Adam Roben.

  • WebProcess/WebPage/win/WebPageWin.cpp:

(WebKit::WebPage::platformInitialize): Use the AlternateRenderingMode
setting.

18:24 Changeset [57706] by weinig@apple.com

Use the threaded process model for MiniBrowser if holding down
the shift key on startup.

Reviewed by Adam Roben.

  • MiniBrowser/win/BrowserView.cpp:

(BrowserView::create):

18:20 Changeset [57705] by weinig@apple.com

Don't assert when soft-linked libraries can't be found

Patch by Adam Roben <aroben@apple.com> on 2010-04-16
In some situations (e.g., when using SOFT_LINK_OPTIONAL), we expect soft-link libraries not
to be able to be found in all cases. So we shouldn't assert that they're always found.

Reviewed by Sam Weinig.

  • platform/win/SoftLinking.h:

(SOFT_LINK): Don't assert when LoadLibrary fails.

18:17 Changeset [57704] by dimich@chromium.org

2010-04-15 Dmitry Titov <dimich@chromium.org>

Unreviewed, rolling out r57688.
http://trac.webkit.org/changeset/57688
https://bugs.webkit.org/show_bug.cgi?id=34992

Makes fast/workers/dedicated-worker-lifecycle.html crashing on all GTK bots

  • fast/workers/change-version-handle-reuse-worker-expected.txt: Removed.
  • fast/workers/change-version-handle-reuse-worker.html: Removed.
  • fast/workers/execute-sql-args-worker-expected.txt: Removed.
  • fast/workers/execute-sql-args-worker.html: Removed.
  • fast/workers/resources/database-worker-controller.js: Removed.
  • fast/workers/resources/database-worker.js: Removed.
  • storage/change-version-handle-reuse-expected.txt:
  • storage/change-version-handle-reuse.html:
  • storage/change-version-handle-reuse.js: Removed.
  • storage/execute-sql-args.html:
  • storage/execute-sql-args.js: Removed.

2010-04-15 Dmitry Titov <dimich@chromium.org>

Unreviewed, rolling out r57688.
http://trac.webkit.org/changeset/57688
https://bugs.webkit.org/show_bug.cgi?id=34992

Makes fast/workers/dedicated-worker-lifecycle.html crashing on all GTK bots

  • bindings/js/JSWorkerContextCustom.cpp:
  • bindings/v8/custom/V8WorkerContextCustom.cpp:
  • storage/Database.idl:
  • storage/SQLError.idl:
  • storage/SQLResultSet.idl:
  • storage/SQLResultSetRowList.idl:
  • storage/SQLTransaction.idl:
  • workers/WorkerContext.cpp: (WebCore::WorkerContext::openDatabase):
  • workers/WorkerContext.h: (WebCore::WorkerContext::databaseExceededQuota):
  • workers/WorkerContext.idl:
18:16 Changeset [57703] by tony@chromium.org

2010-04-15 Tony Chang <tony@chromium.org>

Reviewed by Dimitri Glazkov.

[chromium] Add TestShellGtk.cpp so we can link on Linux
https://bugs.webkit.org/show_bug.cgi?id=37561

  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Add new file and

add platform file exceptions.

  • DumpRenderTree/chromium/TestShellGtk.cpp: Added. (AlarmHandler): (TestShell::waitTestFinished):
18:13 Changeset [57702] by tony@chromium.org

2010-04-15 Tony Chang <tony@chromium.org>

Reviewed by Dimitri Glazkov.

build DRT on chromium mac
https://bugs.webkit.org/show_bug.cgi?id=37639

  • gyp_webkit: generate build files for DRT on mac

2010-04-15 Tony Chang <tony@chromium.org>

Reviewed by Dimitri Glazkov.

build DRT on chromium mac
https://bugs.webkit.org/show_bug.cgi?id=37639

  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
  • Scripts/build-dumprendertree: enable build-dumprendertree --chromium
18:08 Changeset [57701] by yurys@chromium.org

2010-04-15 Yury Semikhatsky <yurys@google.com>

Reviewed by Pavel Feldman.


Support basic debugging capabilities including step in/over/out in v8
implementation of ScriptDebugServer.

https://bugs.webkit.org/show_bug.cgi?id=37604

  • bindings/js/JSInjectedScriptHostCustom.cpp: (WebCore::InjectedScriptHost::createInjectedScript):
  • bindings/v8/ScriptDebugServer.cpp: (WebCore::ScriptDebugServer::ScriptDebugServer): (WebCore::ScriptDebugServer::setDebuggerScriptSource): (WebCore::ScriptDebugServer::addListener): (WebCore::ScriptDebugServer::removeListener): (WebCore::ScriptDebugServer::setBreakpoint): (WebCore::ScriptDebugServer::removeBreakpoint): (WebCore::ScriptDebugServer::clearBreakpoints): (WebCore::ScriptDebugServer::setBreakpointsActivated): (WebCore::ScriptDebugServer::continueProgram): (WebCore::ScriptDebugServer::stepIntoStatement): (WebCore::ScriptDebugServer::stepOverStatement): (WebCore::ScriptDebugServer::stepOutOfFunction): (WebCore::ScriptDebugServer::currentCallFrameState): (WebCore::ScriptDebugServer::currentCallFrameV8): (WebCore::ScriptDebugServer::onV8DebugMessage): (WebCore::ScriptDebugServer::onV8DebugHostDispatch): (WebCore::ScriptDebugServer::handleV8DebugHostDispatch): (WebCore::ScriptDebugServer::handleV8DebugMessage): (WebCore::ScriptDebugServer::dispatchDidParseSource): (WebCore::ScriptDebugServer::ensureDebuggerScriptCompiled): (WebCore::ScriptDebugServer::didResume):
  • bindings/v8/ScriptDebugServer.h: (WebCore::ScriptDebugServer::pauseOnExceptionsState): (WebCore::ScriptDebugServer::setPauseOnExceptionsState): (WebCore::ScriptDebugServer::setMessageLoopDispatchHandler):
  • bindings/v8/custom/V8InjectedScriptHostCustom.cpp: (WebCore::InjectedScriptHost::createInjectedScript): (WebCore::V8InjectedScriptHost::currentCallFrameCallback): (WebCore::V8InjectedScriptHost::isActivationCallback):
  • inspector/front-end/InjectedScript.js: (injectedScriptConstructor): (injectedScriptConstructor.):
  • inspector/front-end/ScriptsPanel.js: (WebInspector.ScriptsPanel):

2010-04-15 Yury Semikhatsky <yurys@google.com>

Reviewed by Pavel Feldman.

Support basic debugging capabilities including step in/over/out in v8
implementation of ScriptDebugServer.

https://bugs.webkit.org/show_bug.cgi?id=37604

  • WebKit.gypi:
  • src/DebuggerAgent.h:
  • src/DebuggerAgentImpl.cpp: (WebKit::DebuggerAgentImpl::setDebuggerScriptSource):
  • src/DebuggerAgentImpl.h:
  • src/DebuggerAgentManager.cpp: (WebKit::DebuggerAgentManager::hostDispatchHandler): (WebKit::DebuggerAgentManager::debugAttach): (WebKit::DebuggerAgentManager::debugDetach): (WebKit::DebuggerAgentManager::setMessageLoopDispatchHandler):
  • src/DebuggerAgentManager.h:
  • src/InspectorFrontendClientImpl.cpp: (WebKit::InspectorFrontendClientImpl::windowObjectCleared):
  • src/js/DebuggerScript.js: Added. (debuggerScriptConstructor.DebuggerScript.getAfterCompileScript): (debuggerScriptConstructor.DebuggerScript.getScripts): (debuggerScriptConstructor.DebuggerScript._formatScript): (debuggerScriptConstructor.DebuggerScript.setBreakpoint): (debuggerScriptConstructor.DebuggerScript.removeBreakpoint): (debuggerScriptConstructor.DebuggerScript.currentCallFrame): (debuggerScriptConstructor.DebuggerScript.stepIntoStatement): (debuggerScriptConstructor.DebuggerScript.stepOverStatement): (debuggerScriptConstructor.DebuggerScript.stepOutOfFunction): (debuggerScriptConstructor.DebuggerScript.clearBreakpoints): (debuggerScriptConstructor.DebuggerScript.setBreakpointsActivated): (debuggerScriptConstructor.DebuggerScript._frameMirrorToJSCallFrame): (debuggerScriptConstructor.DebuggerScript._webkitToV8LineNumber): (debuggerScriptConstructor.DebuggerScript._v8ToWwebkitLineNumber): (debuggerScriptConstructor):
  • src/js/DevTools.js: (WebInspector.loaded): (.): ():
  • src/js/DevToolsHostStub.js: (.RemoteDebuggerAgentStub.prototype.setDebuggerScriptSource):
  • src/js/InspectorControllerImpl.js: (devtools.InspectorBackendImpl):
18:00 Changeset [57700] by mrowe@apple.com

Merge r57692.

18:00 Changeset [57699] by eric@webkit.org

2010-04-15 Matt Perry <mpcomplete@chromium.org>

Reviewed by Dimitri Glazkov.

Implement layoutTestController's addUserStyleSheet and
setAuthorAndUserStylesEnabled on Chromium port.
https://bugs.webkit.org/show_bug.cgi?id=37595

  • public/WebSettings.h:
  • public/WebView.h:
  • src/WebSettingsImpl.cpp: (WebKit::WebSettingsImpl::setAuthorAndUserStylesEnabled):
  • src/WebSettingsImpl.h:
  • src/WebViewImpl.cpp: (WebKit::WebViewImpl::addUserStyleSheet):
  • src/WebViewImpl.h:
17:58 Changeset [57698] by mrowe@apple.com

Versioning.

17:58 Changeset [57697] by jianli@chromium.org

Unreviewed.

Updated Skipped list to use the right directory due to directory
shuffling at r57695.

  • platform/gtk/Skipped:
  • platform/qt/Skipped:
  • platform/win/Skipped:
17:54 Changeset [57696] by ajwong@chromium.org

Not reviewed. Sheriffing.

[Chromium] Build fix. Rebaseline these again since the fonts changed again.

  • platform/chromium-win/fast/backgrounds/svg-as-background-1-expected.checksum:
  • platform/chromium-win/fast/backgrounds/svg-as-background-1-expected.png:
  • platform/chromium-win/fast/backgrounds/svg-as-background-3-expected.checksum:
  • platform/chromium-win/fast/backgrounds/svg-as-background-3-expected.png:
17:50 Changeset [57695] by jianli@chromium.org

Support using FormData to send a sliced file via XHR.
https://bugs.webkit.org/show_bug.cgi?id=36678

Reviewed by Dmitry Titov.

WebCore:

Tests: http/tests/local/formdata/send-form-data-with-sliced-file.html

  • html/Blob.h:

(WebCore::Blob::isFile):

  • html/DOMFormData.cpp:

(WebCore::DOMFormData::append):

  • html/File.h:

(WebCore::File::isFile):

  • html/FormDataList.h:

(WebCore::FormDataList::appendBlob):
(WebCore::FormDataList::Item::Item):
(WebCore::FormDataList::Item::blob):

  • html/HTMLInputElement.cpp:

(WebCore::HTMLInputElement::appendFormData):

  • platform/network/FormData.cpp:

(WebCore::FormData::appendDOMFormData):

  • platform/network/mac/FormDataStreamMac.mm:

(WebCore::closeCurrentStream):

LayoutTests:

Move common functionality to test sending FormData into a helper file
so that it can be shared by FormData layout tests.
Also move the form data testing files to a new subdirectory formdata/.

  • http/tests/local/formdata/resources/send-form-data-common.js: Added.
  • http/tests/local/formdata/script-tests/TEMPLATE.html: Added.
  • http/tests/local/formdata/script-tests/send-form-data-with-sliced-file.js: Added.
  • http/tests/local/formdata/script-tests/send-form-data.js: Added.
  • http/tests/local/formdata/send-form-data-expected.txt: Renamed from LayoutTests/http/tests/local/send-form-data-expected.txt.
  • http/tests/local/formdata/send-form-data-with-sliced-file-expected.txt: Added.
  • http/tests/local/formdata/send-form-data-with-sliced-file.html: Added.
  • http/tests/local/formdata/send-form-data.html: Added.
  • http/tests/local/resources/send-form-data.js: Removed.
  • http/tests/local/send-form-data.html: Removed.
  • platform/gtk/Skipped:
  • platform/qt/Skipped:
  • platform/win/Skipped:
17:43 Changeset [57694] by eric@webkit.org

2010-04-15 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Must enable GL_VERTEX_PROGRAM_POINT_SIZE during initialization
https://bugs.webkit.org/show_bug.cgi?id=37178

  • fast/canvas/webgl/point-size-expected.txt: Added.
  • fast/canvas/webgl/point-size.html: Added.

2010-04-15 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Must enable GL_VERTEX_PROGRAM_POINT_SIZE during initialization
https://bugs.webkit.org/show_bug.cgi?id=37178

Test: fast/canvas/webgl/point-size.html

  • platform/graphics/mac/GraphicsContext3DMac.cpp: Enable GL_VERTEX_PROGRAM_POINT_SIZE during initialization. (WebCore::GraphicsContext3D::GraphicsContext3D):

2010-04-15 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Must enable GL_VERTEX_PROGRAM_POINT_SIZE during initialization
https://bugs.webkit.org/show_bug.cgi?id=37178

  • src/WebGraphicsContext3DDefaultImpl.cpp: Enable GL_VERTEX_PROGRAM_POINT_SIZE during initialization. (WebKit::WebGraphicsContext3DDefaultImpl::initialize):
17:42 Changeset [57693] by mrowe@apple.com

New tag.

17:40 Changeset [57692] by mrowe@apple.com

<rdar://problem/7870651> WebDynamicScrollBarsView.h generates compile errors when included in plain Objective-C files.

Reviewed by Sam Weinig.

  • WebView/WebDynamicScrollBarsView.h:
17:36 Changeset [57691] by weinig@apple.com

Fix crash when trying to load an invalid URL.

Reviewed by Adam Roben.

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::loadURL): Use constructor for ResourceRequest
that takes a KURL instead of the one that takes a String. The one
that takes a string expects a valid URL.

17:31 Changeset [57690] by eric@webkit.org

2010-04-15 Kent Tamura <tkent@chromium.org>

Reviewed by Dimitri Glazkov.

[DRT/Chromium] Add LayoutTestHelper for Mac
https://bugs.webkit.org/show_bug.cgi?id=37668

LayouTestHelper.mm is based on webkit/tools/test_shell/mac/layout_test_helper.mm
of Chromium.

  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
  • DumpRenderTree/chromium/LayoutTestHelper.mm: Added.
17:22 Changeset [57689] by weinig@apple.com

Make tooltips work. Thanks Adam!

Reviewed by Adam Roben.

  • UIProcess/win/WebView.cpp:

(WebKit::WebView::toolTipChanged): Pass the WebView's HWND, not the tooltip's.

17:10 Changeset [57688] by eric@webkit.org

2010-04-15 Eric Uhrhane <ericu@chromium.org>

Reviewed by Dmitry Titov.

Add bindings for async DB API in Workers.
https://bugs.webkit.org/show_bug.cgi?id=34992

  • storage/change-version-handle-reuse.html: Refactored to share code with the worker version of the test.
  • storage/change-version-handle-reuse-expected.html: Removed reference to reloading, which this test doesn't do.
  • storage/change-version-handle-reuse.js: This is the extracted shared core of the test.
  • storage/execute-sql-args.html: Refactored to share code with the worker version of the test.
  • storage/execute-sql-args.js: This is the extracted shared core of the test.

These are the worker versions of the tests; they're tiny shims around shared code.

  • fast/workers/execute-sql-args-worker-expected.txt: Added.
  • fast/workers/execute-sql-args-worker.html: Added.
  • fast/workers/change-version-handle-reuse-worker-expected.txt: Added.
  • fast/workers/change-version-handle-reuse-worker.html: Added. This file gets included from worker database tests; it does all the boilerplate that's needed on the DOM side.
  • fast/workers/resources/database-worker-controller.js: Added. This file is the boilerplate for the worker side of the database tests. It will import and run the actual test on command, and will report back completion and exceptions.
  • fast/workers/resources/database-worker.js: Added.

2010-04-15 Eric Uhrhane <ericu@chromium.org>

Reviewed by Dmitry Titov.

Add bindings for async DB API in Workers.
https://bugs.webkit.org/show_bug.cgi?id=34992

Tests: storage/change-version-handle-reuse-worker.html

storage/execute-sql-args-worker.html

  • bindings/js/JSWorkerContextCustom.cpp: Add openDatabase binding. (WebCore::JSWorkerContext::openDatabase):
  • bindings/v8/custom/V8WorkerContextCustom.cpp: Add openDatabase stub; Chromium will need work both in V8 and in the browser process before we can turn this on there. (WebCore::V8WorkerContext::openDatabaseCallback):

Add NoStaticTables flags to all objects now shared with workers.

  • storage/Database.idl:
  • storage/SQLError.idl:
  • storage/SQLResultSet.idl:
  • storage/SQLResultSetRowList.idl:
  • storage/SQLTransaction.idl:
  • workers/WorkerContext.h: Add databaseExceededQuota.
  • workers/WorkerContext.cpp: (WebCore::WorkerContext::databaseExceededQuota): Add stub implementation for testing; you just get 5MB for now. (WebCore::WorkerContext::openDatabase): Remove invalid assertion.

Add the IDL for the call to openDatabase.

  • workers/WorkerContext.idl:
16:58 Changeset [57687] by eric@webkit.org

2010-04-15 Nicolas Weber <thakis@chromium.org>

Reviewed by Dimitri Glazkov.

Fix drag image thumbnails for indexed images.
https://bugs.webkit.org/show_bug.cgi?id=37621

  • platform/chromium/DragImageChromiumMac.cpp: (WebCore::scaleDragImage): Always use RGB color space. (WebCore::dissolveDragImageToFraction): Always use RGB color space.
16:56 Changeset [57686] by andersca@apple.com

Fix build dependencies.

Reviewed by Adam Roben.

  • WebKit2.sln:
16:52 Changeset [57685] by ap@apple.com

Reviewed by Geoff Garen.

https://bugs.webkit.org/show_bug.cgi?id=37494
<rdar://problem/7857060> REGRESSION (r57340): fast/events/mouse-drag-from-frame-to-other-frame.html
fails on Windows

  • DumpRenderTree/win/EventSender.cpp: (makeEventSender):
  • DumpRenderTree/win/EventSender.h: Tell EventSender if it's being created for a top frame.


  • DumpRenderTree/win/FrameLoadDelegate.cpp: (FrameLoadDelegate::didClearWindowObjectForFrameInStandardWorld): We only want to reset EventSender machinery when a new test is loaded, not when an iframe (or just its global object) is created.
16:51 Changeset [57684] by weinig@apple.com

Fix Windows WebKit2 build.

WebKit2:

Patch by Adam Roben <aroben@apple.com> on 2010-04-15

  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::processDidExit):

  • win/WebKit2Generated.make:

WebKitTools:

Patch by Adam Roben <aroben@apple.com> on 2010-04-15

  • MiniBrowser/win/MiniBrowser.cpp:
16:37 WebKit Team edited by joepeck@webkit.org
Italics look like quotes in WikiFormatting. But its Magic Quotes. (diff)
16:34 WebKit Team edited by joepeck@webkit.org
Moving up in the world ;) (diff)
15:53 Changeset [57683] by aroben@apple.com

Export WebUserContentURLPattern from WebKit

Rubber-stamped by Mark Rowe.

  • WebKit.exp:
15:33 Changeset [57682] by mitz@apple.com

Made consecutive calls to -[WebHTMLView _beginPrintModeWithPageWidth:shrinkToFit:] work
without intermediate calls -[WebHTMLView _endPrintMode].

Reviewed by Simon Fraser.

  • WebView/WebHTMLView.mm:

(-[WebHTMLView _setPrinting:minimumPageWidth:maximumPageWidth:adjustViewSize:]):

15:30 Changeset [57681] by cfleizach@apple.com

Unreviewed.

Layout test fix for new test that works differently on leopard v. snow leopard.

AXHelp is being appended from ancestors incorrectly
https://bugs.webkit.org/show_bug.cgi?id=37659

  • platform/mac/accessibility/unexpected-help-text-expected.txt:
  • platform/mac/accessibility/unexpected-help-text.html:
15:19 Changeset [57680] by aroben@apple.com

Expose UserContentURLPattern as WebKit SPI

Fixes <http://webkit.org/b/37354>.

Reviewed by Tim Hatcher.

WebCore:

  • WebCore.base.exp: Export UserContentURLPattern::parse, and sorted

the file.

  • WebCore.xcodeproj/project.pbxproj: Marked UserContentURLPattern.h as

"Private".

  • page/UserContentURLPattern.h:

(WebCore::UserContentURLPattern::UserContentURLPattern): Added a
default constructor.
(WebCore::UserContentURLPattern::isValid): Added this getter.

WebKit:

  • WebKit.xcodeproj/project.pbxproj: Add WebUserContentURLPattern.

WebKit/mac:

  • Misc/WebUserContentURLPattern.h: Added.
  • Misc/WebUserContentURLPattern.mm: Added.

(-[WebUserContentURLPattern initWithPatternString:]): Initialize
_private and then parse the passed-in string into a
UserContentURLPattern.
(-[WebUserContentURLPattern dealloc]): Release _private.

(-[WebUserContentURLPattern isValid]):
(-[WebUserContentURLPattern scheme]):
(-[WebUserContentURLPattern host]):
(-[WebUserContentURLPattern matchesSubdomains]):
Call through to UserContentURLPattern.

WebKit/win:

  • Interfaces/IWebUserContentURLPattern.idl: Added.
  • Interfaces/WebKit.idl: Added WebUserContentURLPattern.
  • WebKit.vcproj/Interfaces.vcproj: Added IWebUserContentURLPattern.
  • WebKit.vcproj/WebKit.vcproj: Added WebUserContentURLPattern.
  • WebUserContentURLPattern.cpp: Added.

(WebUserContentURLPattern::WebUserContentURLPattern):
(WebUserContentURLPattern::~WebUserContentURLPattern):
(WebUserContentURLPattern::createInstance):
(WebUserContentURLPattern::AddRef):
(WebUserContentURLPattern::Release):
(WebUserContentURLPattern::QueryInterface):
Standard COM implementations.

(WebUserContentURLPattern::parse): Parse the string into a
UserContentURLPattern and store it.

(WebUserContentURLPattern::isValid):
(WebUserContentURLPattern::scheme):
(WebUserContentURLPattern::host):
(WebUserContentURLPattern::matchesSubdomains):
Call through to UserContentURLPattern.

  • WebUserContentURLPattern.h: Added.
15:18 Changeset [57679] by ojan@chromium.org

2010-04-15 Ojan Vafai <ojan@chromium.org>

Reviewed by Adam Barth.

Include codereview issue number in patch description
https://bugs.webkit.org/show_bug.cgi?id=37677

This lets us know which rietveld issue this patch is tied to.

Also, make it so that --fancy-review overrides --no-review.

  • Scripts/webkitpy/tool/steps/postcodereview.py:
  • Scripts/webkitpy/tool/steps/postdiff.py:
15:11 Changeset [57678] by aroben@apple.com

Make --exit-after-n-failures work when all tests are timing out or crashing

Fixes <http://webkit.org/b/37679>.

Reviewed by Jon Honeycutt.

  • Scripts/run-webkit-tests:

(top level): When a test crashes or times out, break out of the main loop if
stopRunningTestsEarlyIfNeeded returns true. Moved some code from the bottom of the main loop
from here...
(stopRunningTestsEarlyIfNeeded): ...to here.

14:50 Changeset [57677] by ajwong@chromium.org

Not reviewed, sheriffing.

[Chromium] Update chromium layout tests expecataions with new baselines. Remove expectations for some tests that seem to be passing now.

  • platform/chromium-win/fast/backgrounds/svg-as-background-1-expected.checksum:
  • platform/chromium-win/fast/backgrounds/svg-as-background-1-expected.png:
  • platform/chromium-win/fast/backgrounds/svg-as-background-3-expected.checksum:
  • platform/chromium-win/fast/backgrounds/svg-as-background-3-expected.png:
  • platform/chromium/test_expectations.txt:
14:23 Changeset [57676] by weinig@apple.com

Add WebHistoryClient support.
https://bugs.webkit.org/show_bug.cgi?id=37671

Reviewed by Anders Carlsson.

WebKit2:

Adds the following callbacks:

didNavigateWithNavigationData
didPerformClientRedirect
didPerformServerRedirect
didUpdateHistoryTitle

  • Shared/CoreIPCSupport/WebPageProxyMessageKinds.h:

(WebPageProxyMessage::):

  • Shared/WebNavigationDataStore.h: Added.
  • UIProcess/API/C/WKAPICast.h:
  • UIProcess/API/C/WKBase.h:
  • UIProcess/API/C/WKNavigationData.cpp: Added.
  • UIProcess/API/C/WKNavigationData.h: Added.
  • UIProcess/API/C/WKPage.cpp:
  • UIProcess/API/C/WKPage.h:
  • UIProcess/API/C/WebKit2.h:
  • UIProcess/WebHistoryClient.cpp: Copied from UIProcess/WebUIClient.cpp.
  • UIProcess/WebHistoryClient.h: Copied from UIProcess/WebUIClient.h.
  • UIProcess/WebNavigationData.cpp: Added.
  • UIProcess/WebNavigationData.h: Added.
  • UIProcess/WebPageProxy.cpp:
  • UIProcess/WebPageProxy.h:
  • WebKit2.xcodeproj/project.pbxproj:
  • WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
  • win/WebKit2.vcproj:

WebKitTools:

Add HistoryClient logging.

  • MiniBrowser/mac/BrowserWindowController.m:

(_didNavigateWithNavigationData):
(_didPerformClientRedirect):
(_didPerformServerRedirect):
(_didUpdateHistoryTitle):
(-[BrowserWindowController awakeFromNib]):

14:15 Changeset [57675] by andersca@apple.com

2010-04-15 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Start the mini browser in threaded mode if shift is pressed during startup.
https://bugs.webkit.org/show_bug.cgi?id=37670

  • MiniBrowser/mac/AppDelegate.m: (-[BrowserAppDelegate init]):
14:05 Changeset [57674] by cfleizach@apple.com

AXHelp is being appended from ancestors incorrectly
https://bugs.webkit.org/show_bug.cgi?id=37659

Reviewed by Beth Dakin.

WebCore:

Test: platform/mac/accessibility/unexpected-help-text.html

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::helpText):

WebKitTools:

  • DumpRenderTree/AccessibilityUIElement.cpp:

(getHelpTextCallback):
(AccessibilityUIElement::getJSClass):

  • DumpRenderTree/AccessibilityUIElement.h:
  • DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:

(AccessibilityUIElement::helpText):

  • DumpRenderTree/mac/AccessibilityUIElementMac.mm:

(AccessibilityUIElement::helpText):

  • DumpRenderTree/win/AccessibilityUIElementWin.cpp:

(AccessibilityUIElement::helpText):

LayoutTests:

  • platform/mac/accessibility/table-attributes-expected.txt:
  • platform/mac/accessibility/table-sections-expected.txt:

Update tests because they were wrong. AXCells should not have their AXHelp come from the AXTable.

  • platform/mac/accessibility/unexpected-help-text-expected.txt: Added.
  • platform/mac/accessibility/unexpected-help-text.html: Added.
14:01 Changeset [57673] by dpranke@chromium.org

2010-04-15 Dirk Pranke <dpranke@chromium.org>

Reviewed by Adam Barth.

Re-format run-webkit-tests to fit in 80-columns for PEP-8 compliance.
(broken by r57463 and r57381, at least). I've also filed bug 37477
to fix check-webkit-style to catch these things.

https://bugs.webkit.org/show_bug.cgi?id=38586

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14:01 Changeset [57672] by dpranke@chromium.org

2010-04-15 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Add a way to override the user-visible name for the test binary since
some ports don't call it DumpRenderTree (e.g., Chromium Win uses
test_shell, Chromium Mac uses TestShell) by adding a driver_name()
method to the Port interface.

https://bugs.webkit.org/show_bug.cgi?id=37631

  • Scripts/webkitpy/layout_tests/port/base.py:
  • Scripts/webkitpy/layout_tests/port/chromium.py:
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py:
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
13:55 Changeset [57671] by hyatt@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37669, REGRESSION: visited styles don't work right when only the visited path specifies
a pseudoelement.

Reviewed by Anders Carlsson.

Rework the pseudo cache on RenderStyles to support nesting, i.e., a pseudo hanging off a pseudo. The existing model gets
confused by this concept, since it relies on a singly linked list of chained pseudo styles (instead of a Vector owned by a primary
style). I changed the style cache to be a Vector instead.

Reworked both styleForElement and pseudoStyleForElement to resolve visited styles first, since in the pseudoStyleForElement case
you need to do this in order to know to allocate an unvisited pseudo style even if one shouldn't normally exist.

  • css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::styleForElement):
(WebCore::CSSStyleSelector::pseudoStyleForElement):

  • dom/Element.cpp:

(WebCore::Element::pseudoStyleCacheIsInvalid):

  • rendering/style/RenderStyle.cpp:

(WebCore::RenderStyle::getCachedPseudoStyle):
(WebCore::RenderStyle::addCachedPseudoStyle):

  • rendering/style/RenderStyle.h:

(WebCore::):
(WebCore::InheritedFlags::cachedPseudoStyles):

13:45 Changeset [57670] by ajwong@chromium.org

2010-04-15 Albert J. Wong <ajwong@chromium.org>

Unreviewed, rolling out r57660.
http://trac.webkit.org/changeset/57660
https://bugs.webkit.org/show_bug.cgi?id=37604

Broke a large number of inspector layout tests in chromium.

  • bindings/js/JSInjectedScriptHostCustom.cpp: (WebCore::InjectedScriptHost::createInjectedScript):
  • bindings/v8/ScriptDebugServer.cpp:
  • bindings/v8/ScriptDebugServer.h: (WebCore::ScriptDebugServer::addListener): (WebCore::ScriptDebugServer::removeListener): (WebCore::ScriptDebugServer::setBreakpoint): (WebCore::ScriptDebugServer::removeBreakpoint): (WebCore::ScriptDebugServer::clearBreakpoints): (WebCore::ScriptDebugServer::setBreakpointsActivated): (WebCore::ScriptDebugServer::pauseOnExceptionsState): (WebCore::ScriptDebugServer::setPauseOnExceptionsState): (WebCore::ScriptDebugServer::continueProgram): (WebCore::ScriptDebugServer::stepIntoStatement): (WebCore::ScriptDebugServer::stepOverStatement): (WebCore::ScriptDebugServer::stepOutOfFunction): (WebCore::ScriptDebugServer::currentCallFrameState): (WebCore::ScriptDebugServer::ScriptDebugServer):
  • bindings/v8/custom/V8InjectedScriptHostCustom.cpp: (WebCore::InjectedScriptHost::createInjectedScript): (WebCore::V8InjectedScriptHost::currentCallFrameCallback): (WebCore::V8InjectedScriptHost::isActivationCallback):
  • inspector/front-end/InjectedScript.js: (injectedScriptConstructor):
  • inspector/front-end/ScriptsPanel.js: (WebInspector.ScriptsPanel):

2010-04-15 Albert J. Wong <ajwong@chromium.org>

Unreviewed, rolling out r57660.
http://trac.webkit.org/changeset/57660
https://bugs.webkit.org/show_bug.cgi?id=37604

Broke a large number of inspector layout tests in chromium.

  • WebKit.gypi:
  • src/DebuggerAgent.h:
  • src/DebuggerAgentImpl.cpp:
  • src/DebuggerAgentImpl.h:
  • src/DebuggerAgentManager.cpp: (WebKit::DebuggerAgentManager::debugAttach): (WebKit::DebuggerAgentManager::debugDetach): (WebKit::DebuggerAgentManager::setMessageLoopDispatchHandler):
  • src/DebuggerAgentManager.h:
  • src/InspectorFrontendClientImpl.cpp: (WebKit::InspectorFrontendClientImpl::windowObjectCleared):
  • src/js/DebuggerScript.js: Removed.
  • src/js/DevTools.js: (WebInspector.loaded): (WebInspector.UnresolvedPropertyValue): ():
  • src/js/DevToolsHostStub.js:
  • src/js/InspectorControllerImpl.js: (devtools.InspectorBackendImpl):
13:44 Changeset [57669] by eric@webkit.org

2010-04-15 Kent Hansen <kent.hansen@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Mark QWebFrame::overloadedSlots autotest as expected failure

https://bugs.webkit.org/show_bug.cgi?id=37319

  • tests/qwebframe/tst_qwebframe.cpp:
13:40 Changeset [57668] by dumi@chromium.org

DatabaseTracker refactoring: remove the dependency on
OriginQuotaManager from DatabaseTracker.h
https://bugs.webkit.org/show_bug.cgi?id=31482

Reviewed by Dimitri Glazkov.

  • storage/DatabaseTracker.cpp:

(WebCore::DatabaseTracker::DatabaseTracker):
(WebCore::DatabaseTracker::~DatabaseTracker):
(WebCore::DatabaseTracker::canEstablishDatabase):
(WebCore::DatabaseTracker::hasEntryForOrigin):
(WebCore::DatabaseTracker::getMaxSizeForDatabase):
(WebCore::DatabaseTracker::databaseChanged):
(WebCore::DatabaseTracker::fullPathForDatabaseNoLock):
(WebCore::DatabaseTracker::fullPathForDatabase):
(WebCore::DatabaseTracker::populateOrigins):
(WebCore::DatabaseTracker::origins):
(WebCore::DatabaseTracker::databaseNamesForOrigin):
(WebCore::DatabaseTracker::addOpenDatabase):
(WebCore::DatabaseTracker::removeOpenDatabase):
(WebCore::DatabaseTracker::usageForOriginNoLock):
(WebCore::DatabaseTracker::usageForOrigin):
(WebCore::DatabaseTracker::quotaForOrigin):
(WebCore::DatabaseTracker::setQuota):
(WebCore::DatabaseTracker::deleteOrigin):
(WebCore::DatabaseTracker::deleteDatabase):
(WebCore::DatabaseTracker::deleteDatabaseFile):

  • storage/DatabaseTracker.h:
  • storage/SQLTransactionClient.cpp:

(WebCore::SQLTransactionClient::didExecuteStatement):

13:36 Changeset [57667] by weinig@apple.com

Remove empty file configurations.

Reviewed by Adam Roben.

  • win/WebKit2.vcproj:
13:28 Changeset [57666] by eric@webkit.org

2010-04-15 Anton Muhin <antonm@chromium.org>

Reviewed by David Levin.

[v8] mark Array.prototype.sort tests which verify if sorting is stable as passing for Chromium
https://bugs.webkit.org/show_bug.cgi?id=37655

  • platform/chromium/test_expectations.txt:
13:18 Changeset [57665] by eric@webkit.org

2010-04-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add EFL port files to efl/WebCoreSupport.
http://webkit.org/b/35918

  • efl/WebCoreSupport/EditorClientEfl.cpp: Added.
  • efl/WebCoreSupport/EditorClientEfl.h: Added.
13:07 Changeset [57664] by eric@webkit.org

2010-04-15 Zhenyao Mo <zmo@google.com>

Reviewed by Adam Barth.

Several tests in fast/canvas/webgl/ failed randomly on Leopard Commit Bot
This fixes an uninitialized variable bug and restores a glFinish call that used to be present.
https://bugs.webkit.org/show_bug.cgi?id=36908

  • platform/graphics/mac/GraphicsContext3DMac.cpp: (WebCore::GraphicsContext3D::GraphicsContext3D): Initialize width/height to 0/0. (WebCore::GraphicsContext3D::prepareTexture): Restore glFinish() in every path.
12:33 Changeset [57663] by tonikitoo@webkit.org

Spatial Navigation: make hasOffscreenRect() to earlier return 'true' if absoluteClippedOverflowRect() gives an empty rect
https://bugs.webkit.org/show_bug.cgi?id=37635

Reviewed by Simon Fraser.
Patch by Antonio Gomes <tonikitoo@webkit.org>

absoluteClippedOverflowRect method of RenderObject does return an empty IntRect for offscreen nodes.
So hasOffscreenRect method (SpatialNavigation.cpp) can safily bail out earlier in such cases.

  • page/SpatialNavigation.cpp:

(WebCore::hasOffscreenRect):

12:26 Changeset [57662] by tonikitoo@webkit.org

Spatial Navigation: remove unnecessery assignment in updateFocusCandidateIfCloser method https://bugs.webkit.org/show_bug.cgi?id=37634

Reviewed by Simon Fraser.
Patch by Antonio Gomes <tonikitoo@webkit.org>

This assignment line is not longer needed after r57061, bug that refactored all
assignment logic to happen lines below in the method.

  • page/FocusController.cpp:

(WebCore::updateFocusCandidateIfCloser):

11:47 Changeset [57661] by weinig@apple.com

Fix WebKit2s build. Don't return temporaries.

Reviewed by Anders Carlsson.

  • UIProcess/API/C/cf/WKStringCF.cpp:

(WKStringCreateWithCFString):

  • UIProcess/API/C/cf/WKURLCF.cpp:

(WKURLCreateWithCFURL):
(WKURLCopyCFURL):

11:40 Changeset [57660] by yurys@chromium.org

2010-04-15 Yury Semikhatsky <yurys@google.com>

Reviewed by Pavel Feldman.


Support basic debugging capabilities including step in/over/out in v8
implementation of ScriptDebugServer.

https://bugs.webkit.org/show_bug.cgi?id=37604

  • bindings/js/JSInjectedScriptHostCustom.cpp: (WebCore::InjectedScriptHost::createInjectedScript):
  • bindings/v8/ScriptDebugServer.cpp: (WebCore::ScriptDebugServer::ScriptDebugServer): (WebCore::ScriptDebugServer::setDebuggerScriptSource): (WebCore::ScriptDebugServer::addListener): (WebCore::ScriptDebugServer::removeListener): (WebCore::ScriptDebugServer::setBreakpoint): (WebCore::ScriptDebugServer::removeBreakpoint): (WebCore::ScriptDebugServer::clearBreakpoints): (WebCore::ScriptDebugServer::setBreakpointsActivated): (WebCore::ScriptDebugServer::continueProgram): (WebCore::ScriptDebugServer::stepIntoStatement): (WebCore::ScriptDebugServer::stepOverStatement): (WebCore::ScriptDebugServer::stepOutOfFunction): (WebCore::ScriptDebugServer::currentCallFrameState): (WebCore::ScriptDebugServer::currentCallFrameV8): (WebCore::ScriptDebugServer::onV8DebugMessage): (WebCore::ScriptDebugServer::onV8DebugHostDispatch): (WebCore::ScriptDebugServer::handleV8DebugHostDispatch): (WebCore::ScriptDebugServer::handleV8DebugMessage): (WebCore::ScriptDebugServer::dispatchDidParseSource): (WebCore::ScriptDebugServer::ensureDebuggerScriptCompiled): (WebCore::ScriptDebugServer::didResume):
  • bindings/v8/ScriptDebugServer.h: (WebCore::ScriptDebugServer::pauseOnExceptionsState): (WebCore::ScriptDebugServer::setPauseOnExceptionsState): (WebCore::ScriptDebugServer::setMessageLoopDispatchHandler):
  • bindings/v8/custom/V8InjectedScriptHostCustom.cpp: (WebCore::InjectedScriptHost::createInjectedScript): (WebCore::V8InjectedScriptHost::currentCallFrameCallback): (WebCore::V8InjectedScriptHost::isActivationCallback):
  • inspector/front-end/InjectedScript.js: (injectedScriptConstructor): (injectedScriptConstructor.):
  • inspector/front-end/ScriptsPanel.js: (WebInspector.ScriptsPanel):

2010-04-15 Yury Semikhatsky <yurys@google.com>

Reviewed by Pavel Feldman.

Support basic debugging capabilities including step in/over/out in v8
implementation of ScriptDebugServer.

https://bugs.webkit.org/show_bug.cgi?id=37604

  • WebKit.gypi:
  • src/DebuggerAgent.h:
  • src/DebuggerAgentImpl.cpp: (WebKit::DebuggerAgentImpl::setDebuggerScriptSource):
  • src/DebuggerAgentImpl.h:
  • src/DebuggerAgentManager.cpp: (WebKit::DebuggerAgentManager::hostDispatchHandler): (WebKit::DebuggerAgentManager::debugAttach): (WebKit::DebuggerAgentManager::debugDetach): (WebKit::DebuggerAgentManager::setMessageLoopDispatchHandler):
  • src/DebuggerAgentManager.h:
  • src/InspectorFrontendClientImpl.cpp: (WebKit::InspectorFrontendClientImpl::windowObjectCleared):
  • src/js/DebuggerScript.js: Added. (debuggerScriptConstructor.DebuggerScript.getAfterCompileScript): (debuggerScriptConstructor.DebuggerScript.getScripts): (debuggerScriptConstructor.DebuggerScript._formatScript): (debuggerScriptConstructor.DebuggerScript.setBreakpoint): (debuggerScriptConstructor.DebuggerScript.removeBreakpoint): (debuggerScriptConstructor.DebuggerScript.currentCallFrame): (debuggerScriptConstructor.DebuggerScript.stepIntoStatement): (debuggerScriptConstructor.DebuggerScript.stepOverStatement): (debuggerScriptConstructor.DebuggerScript.stepOutOfFunction): (debuggerScriptConstructor.DebuggerScript.clearBreakpoints): (debuggerScriptConstructor.DebuggerScript.setBreakpointsActivated): (debuggerScriptConstructor.DebuggerScript._frameMirrorToJSCallFrame): (debuggerScriptConstructor.DebuggerScript._webkitToV8LineNumber): (debuggerScriptConstructor.DebuggerScript._v8ToWwebkitLineNumber): (debuggerScriptConstructor):
  • src/js/DevTools.js: (WebInspector.loaded): (.): ():
  • src/js/DevToolsHostStub.js: (.RemoteDebuggerAgentStub.prototype.setDebuggerScriptSource):
  • src/js/InspectorControllerImpl.js: (devtools.InspectorBackendImpl):
11:32 Changeset [57659] by zimmermann@webkit.org

2010-04-15 Nikolas Zimmermann <nzimmermann@rim.com>

Reviewed by Beth Dakin.

RenderSVGResource <-> id mapping should be cached
https://bugs.webkit.org/show_bug.cgi?id=37575

Test: svg/custom/clip-path-id-changes.svg

  • rendering/RenderSVGResource.h: (WebCore::RenderSVGResource::RenderSVGResource): (WebCore::RenderSVGResource::~RenderSVGResource): (WebCore::RenderSVGResource::idChanged): (WebCore::getRenderSVGResourceById):
  • svg/SVGDocumentExtensions.cpp: (WebCore::SVGDocumentExtensions::addResource): (WebCore::SVGDocumentExtensions::removeResource): (WebCore::SVGDocumentExtensions::resourceById):
  • svg/SVGDocumentExtensions.h:
  • svg/SVGStyledElement.cpp: (WebCore::SVGStyledElement::svgAttributeChanged):

2010-04-15 Nikolas Zimmermann <nzimmermann@rim.com>

Reviewed by Beth Dakin.

RenderSVGResource <-> id mapping should be cached
https://bugs.webkit.org/show_bug.cgi?id=37575

  • platform/mac/svg/custom/clip-path-id-changes-expected.checksum: Added.
  • platform/mac/svg/custom/clip-path-id-changes-expected.png: Added.
  • platform/mac/svg/custom/clip-path-id-changes-expected.txt: Added.
  • svg/custom/clip-path-id-changes.svg: Added.
11:24 Changeset [57658] by ajwong@chromium.org

2010-04-15 Justin Schuh <jschuh@chromium.org>

Reviewed by Dimitri Glazkov.

LayoutTest breakage in V8 bindings after r57627
https://bugs.webkit.org/show_bug.cgi?id=37660

Fixes breakage of the following tests due to an error in V8 bindings for
NamedNodeMap:
hc_namednodemapinuseattributeerr.html
hc_namednodemapsetnameditemreturnvalue.html
hc_namednodemapwrongdocumenterr.html
hc_namednodemapinvalidtype1.html
NamedNodeMap-setNamedItem-crash.html

  • bindings/v8/custom/V8NamedNodeMapCustom.cpp: (WebCore::V8NamedNodeMap::setNamedItemNSCallback): (WebCore::V8NamedNodeMap::setNamedItemCallback):
11:22 Changeset [57657] by hyatt@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37567, :first-letter inside a :visited link is wrong color. Make sure
that the pseudo style caching allows visited link styles to hang off other pseudo styles.

Reviewed by Anders Carlsson.

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::updateFirstLetter):

  • rendering/style/RenderStyle.cpp:

(WebCore::RenderStyle::getCachedPseudoStyle):

11:21 Changeset [57656] by tonikitoo@webkit.org

REGRESSION(r56552): Broken scrollbars size https://bugs.webkit.org/show_bug.cgi?id=36853

Reviewed by Kenneth Christiansen and Tor Arne Vestbø.
Patch by Antonio Gomes <tonikitoo@webkit.org>

The regression was caused by r56552, which introduced a fix to bug
webkit.org/b/21300. The bug solved an issue with the resize handle on mac,
but did it in a way that affected all Qt platforms and thus broke the behavior
on non-mac platforms.

This patch makes the mac specific change ifdef'ed and only applied for the mac
platform.

  • WebCoreSupport/ChromeClientQt.cpp:

(WebCore::ChromeClientQt::windowResizerRect):

11:13 Changeset [57655] by benm@google.com

2010-04-15 Ben Murdoch <benm@google.com>

Unreviewed, rolling out r57652.
http://trac.webkit.org/changeset/57652
https://bugs.webkit.org/show_bug.cgi?id=37609

Caused a build break on Chromium Mac and Layout Test fail on
Qt

  • src/WebInputEventConversion.cpp: (WebKit::toPlatformTouchPointState):

2010-04-15 Ben Murdoch <benm@google.com>

Unreviewed, rolling out r57652.
http://trac.webkit.org/changeset/57652
https://bugs.webkit.org/show_bug.cgi?id=37609

Caused a build break on Chromium Mac and Layout Test fail on
Qt

  • platform/PlatformTouchPoint.h: (WebCore::PlatformTouchPoint::):
  • platform/qt/PlatformTouchPointQt.cpp: (WebCore::PlatformTouchPoint::PlatformTouchPoint):
10:51 Changeset [57654] by andersca@apple.com

2010-04-14 Anders Carlsson <andersca@apple.com>

Reviewed by Adam Roben.

Add "Force Repaint" to debug menu.
https://bugs.webkit.org/show_bug.cgi?id=37627

  • MiniBrowser/mac/BrowserWindowController.h:
  • MiniBrowser/mac/BrowserWindowController.m: (-[BrowserWindowController forceRepaint:]):
  • MiniBrowser/mac/English.lproj/MainMenu.xib:
10:33 Changeset [57653] by eric@webkit.org

2010-04-15 Yaar Schnitman <yaar@chromium.org>

Reviewed by Nate Chapin.

Overloads auto-generation in V8
https://bugs.webkit.org/show_bug.cgi?id=37373

This will be used by XHR.send/open, Canvas.*, WebGL.* methods that are currently custom. When more than a single overload exists for a method, the correct overload is chosen based on the total number of arguments passed as well as the values passed to non-primitive arguments.

Overload dispatch order depends on the order the functions are defined in the IDL. Overloads must be specified from the most precise (overloads with wrapper type arguments) to the least precise (overloads with only primitive type arguments).

  • bindings/scripts/CodeGeneratorV8.pm: Identify and output overloads callbacks and dispatch code.
  • bindings/v8/test/TestObj.idl: Overloads sample.
  • bindings/v8/test/V8TestObj.cpp: Output change.
10:18 Changeset [57652] by benm@google.com

2010-04-14 Ben Murdoch <benm@google.com>

Reviewed by Kenneth Rohde Christiansen.

The TouchStationary state of WebCore::PlatformTouchPoint is not
handled inside the touch event handler.
https://bugs.webkit.org/show_bug.cgi?id=37609

After discussions at the WebKit contributors meeting, we decided that
this is a currently unused state without a good future use case in the
Touch API and thus decided to remove it. This patch actions that decision.

  • src/WebInputEventConversion.cpp: (WebKit::toPlatformTouchPointState): Remove TouchStationary.

2010-04-15 Ben Murdoch <benm@google.com>

Reviewed by Kenneth Rohde Christiansen.

The TouchStationary state of WebCore::PlatformTouchPoint
is not handled inside the touch event handler.
https://bugs.webkit.org/show_bug.cgi?id=37609

After discussions at the WebKit contributors meeting, we decided
that this is a currently unused state without a good future use
case in the Touch API and thus decided to remove it. This patch
actions that decision.

As the TouchStationary state is not handled in the EventHandler,
there is no change in functionality so no new tests are required.

  • platform/PlatformTouchPoint.h: (WebCore::PlatformTouchPoint::): Remove TouchStationary.
  • platform/qt/PlatformTouchPointQt.cpp: (WebCore::PlatformTouchPoint::PlatformTouchPoint): Remove TouchStationary.
10:03 Changeset [57651] by weinig@apple.com

Add debug-minibrowser script.

Reviewed by Adam Roben.

  • Scripts/debug-minibrowser: Copied from Scripts/run-minibrowser.
  • Scripts/webkitdirs.pm:
09:57 Changeset [57650] by sfalken@apple.com

2010-04-15 Steve Falkenburg <sfalken@apple.com>

Reviewed by Adam Roben.

Use a lower-overhead mechanism for plug-in message throttling
https://bugs.webkit.org/show_bug.cgi?id=37642
<rdar://problem/7418285> Very high CPU usage idling in gmail under Windows 7 (plug-in related)


GMail has an instance of Flash that loads on the main mail page.
This Flash content sends us a constant stream of WM_USER+1 messsages.


There was already code in PluginMessageThrottlerWin to queue and process these
excess WM_USER+1 messages from Flash. Unfortunately, there were a couple of
problems with this code:


  • The timer used to process the excess messages had a very low timeout (1ms). Chrome uses a value of 5ms for this delay, and doesn't use excess CPU here, while still maintaining good Flash frame rate.


  • The overhead involved in generating a constant stream of 5ms timers still swamped the CPU, resulting in continued high CPU utilization.


To fix this, I changed the throttling code to:


  • Process a queued WM_USER+1 message directly if none has been processed in 5ms. This allows us to avoid the overhead of a timer.


  • Process remaining delayed WM_USER+1 messages on a 16ms timer.


This reduces our CPU utilization idling in GMail from ~20% to ~2-3% on my system.
I also verified the frame rate for Flash content wasn't reduced with this change.

  • plugins/win/PluginMessageThrottlerWin.cpp: (WebCore::PluginMessageThrottlerWin::PluginMessageThrottlerWin): Initialize m_lastMessageTime. (WebCore::PluginMessageThrottlerWin::appendMessage): Process a queued message directly if >5ms have passed. (WebCore::PluginMessageThrottlerWin::processQueuedMessage): Split out from messageThrottleTimerFired. (WebCore::PluginMessageThrottlerWin::messageThrottleTimerFired): Call through to processQueuedMessage.
  • plugins/win/PluginMessageThrottlerWin.h: Add processQueuedMessage, m_lastMessageTime.
08:59 Changeset [57649] by eric@webkit.org

2010-04-15 Roland Steiner <rolandsteiner@chromium.org>

Reviewed by Dimitri Glazkov.

Bug 37636 - [DRT/Chromium] Implement DRT/Chromium for Windows
https://bugs.webkit.org/show_bug.cgi?id=37636

Second patch: add Windows-specific implementation parts
of TestShell.

  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
  • DumpRenderTree/chromium/TestShell.h: (TestShell::finishedEvent):
  • DumpRenderTree/chromium/TestShellWin.cpp: (watchDogThread): (TestShell::waitTestFinished):
08:50 Changeset [57648] by eric@webkit.org

2010-04-15 Roland Steiner <rolandsteiner@chromium.org>

Reviewed by Dimitri Glazkov.

Bug 37636 - [DRT/Chromium] Implement DRT/Chromium for Windows
https://bugs.webkit.org/show_bug.cgi?id=37636

First patch: fix compiler errors.

  • DumpRenderTree/chromium/EventSender.cpp: (EventSender::reset): (EventSender::dispatchMessage):
  • DumpRenderTree/chromium/LayoutTestController.cpp: (LayoutTestController::pathToLocalResource):
  • DumpRenderTree/chromium/TestWebWorker.h:
  • DumpRenderTree/chromium/TextInputController.cpp:
  • DumpRenderTree/chromium/WebViewHost.h:
  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
08:20 Changeset [57647] by antonm@chromium.org

2010-04-15 Anton Muhin <antonm@chromium.org>

Not reviewed, sheriffing.

[Chromium] rebaseline URL related tests: update expectations for Win and Mac.

https://bugs.webkit.org/show_bug.cgi?id=37654

  • platform/chromium-mac/fast/url/mailto-expected.txt: Added.
  • platform/chromium-mac/fast/url/scheme-expected.txt: Added.
  • platform/chromium-win/fast/url/mailto-expected.txt: Added.
  • platform/chromium-win/fast/url/scheme-expected.txt: Added.
07:52 Changeset [57646] by hamaji@chromium.org

2010-04-15 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by David Levin.

LEAK: in ThreadableWebSocketChannel::create()
https://bugs.webkit.org/show_bug.cgi?id=37584

No new tests because this change just fixes a leak.

  • websockets/WorkerThreadableWebSocketChannel.cpp: (WebCore::WorkerThreadableWebSocketChannel::WorkerThreadableWebSocketChannel):
  • websockets/WorkerThreadableWebSocketChannel.h: (WebCore::WorkerThreadableWebSocketChannel::Bridge::create):
07:41 Changeset [57645] by ggaren@apple.com

Fixed complexity and performance FIXME created by using KURL in the UI
process -- it turned out that everywhere we were using KURL, we could
have just used String instead. (That's how Windows WebKit works, too.)

Reviewed by Anders Carlsson.

I kept WKURLRef and WKStringRef distinct opaque types in the API for now,
though, since there may be profit in changing their backing stores in the
future, and it's nice for the API to encode a difference between generic
strings and strings that are valid, canonical URLs.

  • Shared/KURLWrapper.h: Removed. Yay!
  • Shared/WebCoreTypeArgumentMarshalling.h: Nixed KURL marshalling functions.

Old callers marshal Strings now, instead. (This is what KURL was doing
under the covers, anyway.)

  • UIProcess/API/C/WKAPICast.h:

(toWK): Backed by StringImpl* now.
(toURLRef): Added a disambiguating function for specifying that you want
a WKURLRef, since StringImpl* converts to WKStringRef by default.

  • UIProcess/API/C/WKFrame.cpp:

(WKFrameGetProvisionalURL):
(WKFrameGetURL):

  • UIProcess/API/C/WKPage.cpp:

(WKPageLoadURL):

  • UIProcess/API/C/WKURL.cpp:
  • UIProcess/API/C/cf/WKURLCF.cpp:

(WKURLCreateWithCFURL):
(WKURLCopyCFURL):

  • UIProcess/WebFrameProxy.cpp:

(WebKit::WebFrameProxy::didStartProvisionalLoad):
(WebKit::WebFrameProxy::didCommitLoad):

  • UIProcess/WebFrameProxy.h:

(WebKit::WebFrameProxy::url):
(WebKit::WebFrameProxy::provisionalURL):

  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::loadURL):
(WebKit::WebPageProxy::didReceiveMessage):
(WebKit::WebPageProxy::didStartProvisionalLoadForFrame):
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::WebPageProxy::decidePolicyForNewWindowAction):
(WebKit::WebPageProxy::decidePolicyForMIMEType):
(WebKit::WebPageProxy::processDidExit):

  • UIProcess/WebPageProxy.h:

(WebKit::WebPageProxy::urlAtProcessExit):

  • UIProcess/WebPolicyClient.cpp:

(WebKit::WebPolicyClient::decidePolicyForNavigationAction):
(WebKit::WebPolicyClient::decidePolicyForNewWindowAction):
(WebKit::WebPolicyClient::decidePolicyForMIMEType):

  • UIProcess/WebPolicyClient.h:
  • WebKit2.xcodeproj/project.pbxproj:
  • WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:

(WebKit::WebFrameLoaderClient::dispatchDidStartProvisionalLoad):
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForMIMEType):
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNewWindowAction):
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::loadURL):
(WebKit::WebPage::didReceiveMessage):

  • WebProcess/WebPage/WebPage.h: Replaced KURL / KURLWrapper with String.
07:40 Changeset [57644] by antonm@chromium.org

2010-04-15 Anton Muhin <antonm@chromium.org>

Not reviewed, sheriffing.

[Chromium] rebaseline svg/custom/path-bad-data.svg after http://trac.webkit.org/changeset/57591/

http://trac.webkit.org/changeset/57591://bugs.webkit.org/show_bug.cgi?id=37657

  • platform/chromium-linux/svg/custom/path-bad-data-expected.checksum:
  • platform/chromium-linux/svg/custom/path-bad-data-expected.png:
  • platform/chromium-win/svg/custom/path-bad-data-expected.checksum:
  • platform/chromium-win/svg/custom/path-bad-data-expected.png:
  • platform/chromium-win/svg/custom/path-bad-data-expected.txt:
07:18 Changeset [57643] by antonm@chromium.org

2010-04-15 Anton Muhin <antonm@chromium.org>

Not reviewed, sheriffing.

[Chromium] rebaseline fast/repaint/fixed-tranformed.html after
http://trac.webkit.org/changeset/57633/

https://bugs.webkit.org/show_bug.cgi?id=37656

  • platform/chromium-linux/fast/repaint/fixed-tranformed-expected.checksum: Added.
  • platform/chromium-linux/fast/repaint/fixed-tranformed-expected.png: Added.
  • platform/chromium-mac/fast/repaint/fixed-tranformed-expected.checksum: Added.
  • platform/chromium-mac/fast/repaint/fixed-tranformed-expected.png: Added.
  • platform/chromium-win-xp/fast/repaint: Added.
  • platform/chromium-win-xp/fast/repaint/fixed-tranformed-expected.png: Added.
  • platform/chromium-win/fast/repaint/fixed-tranformed-expected.checksum: Added.
  • platform/chromium-win/fast/repaint/fixed-tranformed-expected.png: Added.
06:41 Changeset [57642] by wsiegrist@apple.com

2010-04-15 William Siegrist <wsiegrist@apple.com>

Give a name to new KillOldProcesss step so it does not collide with the other compile step.

06:39 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
06:19 Changeset [57641] by wsiegrist@apple.com

2010-04-15 William Siegrist <wsiegrist@apple.com>

Add core/noncore categories to builders.

06:15 Changeset [57640] by antonm@chromium.org

2010-04-15 Anton Muhin <antonm@chromium.org>

Not reviewed, sheriffing.

Rebaseline some recently added LayoutTests under fast/url.

  • platform/chromium-linux/fast/url: Added.
  • platform/chromium-linux/fast/url/mailto-expected.txt: Added.
  • platform/chromium-linux/fast/url/scheme-expected.txt: Added.
  • platform/chromium-mac/fast/url: Added.
  • platform/chromium-mac/fast/url/ipv6-expected.txt: Added.
  • platform/chromium-mac/fast/url/path-url-expected.txt: Added.
  • platform/chromium-mac/fast/url/port-expected.txt: Added.
  • platform/chromium-mac/fast/url/standard-url-expected.txt: Added.
  • platform/chromium-win/fast/url: Added.
  • platform/chromium-win/fast/url/ipv6-expected.txt: Added.
  • platform/chromium-win/fast/url/path-url-expected.txt: Added.
  • platform/chromium-win/fast/url/port-expected.txt: Added.
  • platform/chromium-win/fast/url/standard-url-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
06:03 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
05:54 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
05:32 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
04:51 Changeset [57639] by eric@webkit.org

2010-04-15 Gyuyoung Kim <gyuyoung.kim@samsung.com>

Reviewed by Eric Seidel.

Duplicated patches related to wml were pushed to trunk. So, one of
the patches should be reverted.
https://bugs.webkit.org/show_bug.cgi?id=37542

  • wml/WMLOptionElement.h:
  • wml/WMLSelectElement.h:
04:01 QtWebKitTableOfFeatures20 edited by henry.haverinen@nokia.com
(diff)
03:32 BuildingQtOnLinux edited by jocelyn.turcotte@nokia.com
(diff)
03:03 Changeset [57638] by eric@webkit.org

2010-04-15 Bruno Schmidt <bruno.schmidt@gmail.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Null QObjects properties cause Segmentation Fault
https://bugs.webkit.org/show_bug.cgi?id=34730

QObjects exported to the QWebkit javascript with properties that are
a null "QObject*" cause Segmentation Fault.

If an QObject is added to the javascript context and it contains
properties of the type QObject* with NULL value, calling the property
causes Segmentation Fault.
So now the code below properly checks for null pointers:

  • bridge/qt/qt_instance.cpp: (JSC::Bindings::QtInstance::getClass): may return NULL (JSC::Bindings::QtInstance::getMethod): may return jsNull() (JSC::Bindings::QtInstance::stringValue): may return jsNull() (JSC::Bindings::QtInstance::booleanValue): may return false
  • bridge/qt/qt_runtime.cpp: (JSC::Bindings::convertValueToQVariant): (JSC::Bindings::convertQVariantToValue): May return jsNull on QObjectStar

2010-04-15 Bruno Schmidt <bruno.schmidt@gmail.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Null QObjects properties cause Segmentation Fault
https://bugs.webkit.org/show_bug.cgi?id=34730

QObjects exported to the QWebkit javascript with properties that are
a null "QObject*" cause Segmentation Fault.

If an QObject is added to the javascript context and it contains
properties of the type QObject* with NULL value, calling the property
causes Segmentation Fault.

Follow the tests for the corrections done over WebCore.

  • tests/qwebframe/tst_qwebframe.cpp: (MyQObject::MyQObject): init the field m_objectStar (MyQObject::objectStarProperty): read the Object* prop (MyQObject::setObjectStarProperty): write the Object* prop (tst_QWebFrame::getSetStaticProperty): new tests for the new prop
02:53 Changeset [57637] by eric@webkit.org

2010-04-15 Jedrzej Nowacki <jedrzej.nowacki@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

Fix memory leak in QScriptEngine::evaluate().

QScriptEnginePrivate::evaluate should release temporary variables.

[Qt] QScriptEngine::evaluate has memory leak.
https://bugs.webkit.org/show_bug.cgi?id=37596

  • qt/api/qscriptengine_p.cpp: (QScriptEnginePrivate::evaluate):
  • qt/api/qscriptengine_p.h:
01:14 Changeset [57636] by abarth@webkit.org

2010-04-15 Jonathan Dixon <joth@chromium.org>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=37492

Add mailto: canonicalization tests to the URL regressions tests

  • fast/url/mailto-expected.txt: Added.
  • fast/url/mailto.html: Added.
  • fast/url/script-tests/mailto.js: Added.
01:07 Changeset [57635] by abarth@webkit.org

2010-04-15 Adam Barth <abarth@webkit.org>

Reviewed by Maciej Stachowiak.

Add URL scheme parsing tests
https://bugs.webkit.org/show_bug.cgi?id=37496

I'm not super happy with how these test turned out. It's difficult to
test these cases with this methodology because most of the cases get
treated as relative URLs. I tired using the "protocol" property of the
HTMLAnchorElement, but that often caused the test case to be ignored
for reasons I don't fully understand.

  • fast/url/resources/utilities.js: (setBaseURL):
    • In the course of trying various variations of this test, I wrote this function. It seemed useful, so I left it here for the benefit of future test writing.
  • fast/url/scheme-expected.txt: Added.
  • fast/url/scheme.html: Added.
  • fast/url/script-tests/scheme.js: Added.
00:55 Changeset [57634] by abarth@webkit.org

2010-04-15 Adam Barth <abarth@webkit.org>

Reviewed by Darin Fisher.

Add comments back to IPv4 URL parsing test
https://bugs.webkit.org/show_bug.cgi?id=37620

I removed these comments when importing this test. The comments are
likley to be helpful to us in the future, so I've restored them.

  • fast/url/script-tests/ipv4.js:
Note: See TracTimeline for information about the timeline view.