WebKit-https.git
10 years ago[wx] Build fixes for recent changes.
kevino@webkit.org [Tue, 28 Sep 2010 16:17:17 +0000 (16:17 +0000)]
[wx] Build fixes for recent changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68524 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Stephen White <senorblanco@chromium.org>
senorblanco@chromium.org [Tue, 28 Sep 2010 16:14:29 +0000 (16:14 +0000)]
2010-09-28  Stephen White  <senorblanco@chromium.org>

        Unreviewed; expectations fix.

        [CHROMIUM] Fix duplicate expectations.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68523 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd PluginControllerProxy class
andersca@apple.com [Tue, 28 Sep 2010 16:07:18 +0000 (16:07 +0000)]
Add PluginControllerProxy class
https://bugs.webkit.org/show_bug.cgi?id=46728

Reviewed by Adam Roben.

* PluginProcess/PluginControllerProxy.cpp: Added.
* PluginProcess/PluginControllerProxy.h: Added.
Add stubbed out PluginControllerProxy class.

* PluginProcess/WebProcessConnection.cpp:
(WebKit::WebProcessConnection::~WebProcessConnection):
Assert that we don't have any live plug-in controller proxies.

(WebKit::WebProcessConnection::addPluginControllerProxy):
Add the plug-in controller proxy to the map.

(WebKit::WebProcessConnection::removePluginControllerProxy):
Remove the plug-in controller proxy from the map. If the last plug-in controller
proxy went away, invalidate the connection.

* WebKit2.xcodeproj/project.pbxproj:
Add files.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68522 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Andras Becsi <abecsi@webkit.org>
abecsi@webkit.org [Tue, 28 Sep 2010 16:01:47 +0000 (16:01 +0000)]
2010-09-28  Andras Becsi  <abecsi@webkit.org>

        Reviewed by Csaba Osztrogonác.

        Undefined reference errors when linking due to gperf and inlining.
        webkit.org/b/29244

        EFL CMake changes by Leandro Pereira <leandro@profusion.mobi>

        Refactor gperf code generation and usage to fix the debug build with gcc>4.4.
        Hitherto gperf generated C code, these files were included in multiple C++ files across WebCore
        to access the functionality provided. This resulted in debug build failure with newer gcc versions
        because of a behaviour change of gcc, which disables C style inlining in debug mode.
        The make-hash-tools.pl script lets gperf generate C++ code for all gperf files now, which are compiled
        in their own compilation unit.
        The functionality provided by the generated code is wrapped behind HashTools.h, so there is no need
        for multiple inclusions of generated C files to access these functions.

        * WebCoreSupport/FrameLoaderClientQt.cpp:
        (WebCore::FrameLoaderClientQt::createPlugin):
2010-09-28  Andras Becsi  <abecsi@webkit.org>

        Reviewed by Csaba Osztrogonác.

        Undefined reference errors when linking due to gperf and inlining.
        webkit.org/b/29244

        EFL CMake changes by Leandro Pereira <leandro@profusion.mobi>

        Refactor gperf code generation and usage to fix the debug build with gcc>4.4.
        Hitherto gperf generated C code, these files were included in multiple C++ files across WebCore
        to access the functionality provided. This resulted in debug build failure with newer gcc versions
        because of a behaviour change of gcc, which disables C style inlining in debug mode.
        The make-hash-tools.pl script lets gperf generate C++ code for all gperf files now, which are compiled
        in their own compilation unit.
        The functionality provided by the generated code is wrapped behind HashTools.h, so there is no need
        for multiple inclusions of generated C files to access these functions.

        No new tests needed.

        * CMakeLists.txt:
        * WebCore.gyp/WebCore.gyp:
        * WebCore.pri:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * css/CSSParser.cpp:
        * css/makeprop.pl:
        * css/makevalues.pl:
        * html/DocTypeStrings.gperf:
        * html/HTMLDocument.cpp:
        * make-hash-tools.pl:
        * platform/ColorData.gperf:
        * platform/graphics/Color.cpp:
2010-09-28  Andras Becsi  <abecsi@webkit.org>

        Reviewed by Csaba Osztrogonác.

        Undefined reference errors when linking due to gperf and inlining.
        webkit.org/b/29244

        EFL CMake changes by Leandro Pereira <leandro@profusion.mobi>

        Refactor gperf code generation and usage to fix the debug build with gcc>4.4.
        Hitherto gperf generated C code, these files were included in multiple C++ files across WebCore
        to access the functionality provided. This resulted in debug build failure with newer gcc versions
        because of a behaviour change of gcc, which disables C style inlining in debug mode.
        The make-hash-tools.pl script lets gperf generate C++ code for all gperf files now, which are compiled
        in their own compilation unit.
        The functionality provided by the generated code is wrapped behind HashTools.h, so there is no need
        for multiple inclusions of generated C files to access these functions.

        * cmake/WebKitMacros.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68521 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoDon't call NPP_SetWindow until the plugin's HWND has been sized/positioned
aroben@apple.com [Tue, 28 Sep 2010 15:44:00 +0000 (15:44 +0000)]
Don't call NPP_SetWindow until the plugin's HWND has been sized/positioned

Test: platform/win/plugins/window-geometry-initialized-before-set-window.html

Reviewed by Anders Carlsson.

Fixes <http://webkit.org/b/46716> <rdar://problem/8482014> Full-page
Adobe Reader does not paint until window is resized

WebKit2:

* WebProcess/Plugins/Netscape/NetscapePlugin.cpp:
(WebKit::NetscapePlugin::geometryDidChange): Call NPP_SetWindow after
updating the platform (HWND) geometry.

WebKitTools:

Test that a plugin's HWND is sized/positioned before NPP_SetWindow is called

Reviewed by Anders Carlsson.

* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
(PluginTest::NPP_SetWindow): Added. Just returns NPERR_NO_ERROR at
this level.

* DumpRenderTree/TestNetscapePlugIn/PluginTest.h: Declared
NPP_SetWindow.

* DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowGeometryInitializedBeforeSetWindow.cpp: Added.
(WindowGeometryInitializedBeforeSetWindow::NPP_SetWindow): Checks that
the plugin's HWND has a non-zero size and that its size/position
matches that specified in the NPWindow.

* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NPP_SetWindow): Call through to the PluginTest.

* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
Added WindowGeometryInitializedBeforeSetWindow.cpp, and let VS reorder
the files as it saw fit.

LayoutTests:

Test that a plugin's HWND is sized/positioned before NPP_SetWindow is called

* platform/win/plugins/window-geometry-initialized-before-set-window-expected.txt: Added.
* platform/win/plugins/window-geometry-initialized-before-set-window.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68520 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoStart running plugins tests in WebKit2 on Windows
aroben@apple.com [Tue, 28 Sep 2010 15:43:30 +0000 (15:43 +0000)]
Start running plugins tests in WebKit2 on Windows

Fixes <http://webkit.org/b/46675> <rdar://problem/8484269>

Reviewed by John Sullivan.

* platform/win-wk2/Skipped: Removed the plugins and
platform/win/plugins directories and added individual failing tests.
plugins/geturlnotify-during-document-teardown.html and
plugins/mouse-events-fixedpos.html will also fail on SnowLeopard, but
they're already skipped there on both WebKit1 and WebKit2 thanks to
platform/mac-snowleopard/Skipped.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68519 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoOut of process plug-ins shouldn't be turned on just yet...
andersca@apple.com [Tue, 28 Sep 2010 15:29:45 +0000 (15:29 +0000)]
Out of process plug-ins shouldn't be turned on just yet...

* WebKit2Prefix.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68518 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Benjamin Poulain <benjamin.poulain@nokia.com>
andreas.kling@nokia.com [Tue, 28 Sep 2010 15:09:42 +0000 (15:09 +0000)]
2010-09-28  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Andreas Kling.

        [Qt] Remove support for Qt 4.5
        https://bugs.webkit.org/show_bug.cgi?id=46718

        Remove the code for versions of Qt prior to 4.6.

        * Api/qgraphicswebview.cpp:
        (QGraphicsWebViewPrivate::updateResizesToContentsForPage):
        (QGraphicsWebView::QGraphicsWebView):
        (QGraphicsWebView::sceneEvent):
        (QGraphicsWebViewPrivate::detachCurrentPage):
        * Api/qwebframe.cpp:
        (QWebFrame::load):
        * Api/qwebpage.cpp:
        (QWebPagePrivate::QWebPagePrivate):
        (QWebPagePrivate::handleSoftwareInputPanel):
        (QWebPagePrivate::inputMethodEvent):
        (QWebPagePrivate::touchEvent):
        (QWebPage::inputMethodQuery):
        (QWebPage::view):
        (QWebPage::event):
        * Api/qwebpage_p.h:
        * Api/qwebview.cpp:
        (QWebView::QWebView):
        (QWebViewPrivate::detachCurrentPage):
        (QWebView::event):
        * WebCoreSupport/EditorClientQt.cpp:
        (WebCore::EditorClientQt::setInputMethodState):
        * WebCoreSupport/FrameLoaderClientQt.cpp:
        (WebCore::FrameLoaderClientQt::createPlugin):
        * WebCoreSupport/PageClientQt.cpp:
        (WebCore::PageClientQWidget::setInputMethodHints):
        (WebCore::PageClientQGraphicsWidget::setInputMethodEnabled):
        (WebCore::PageClientQGraphicsWidget::inputMethodEnabled):
        (WebCore::PageClientQGraphicsWidget::setInputMethodHints):
        * WebCoreSupport/PageClientQt.h:
        (WebCore::QGraphicsItemOverlay::page):
        * tests/qwebframe/tst_qwebframe.cpp:
        * tests/qwebpage/tst_qwebpage.cpp:
        (tst_QWebPage::inputMethods_data):
        (inputMethodHints):
        (inputMethodEnabled):
        (tst_QWebPage::inputMethods):
        (tst_QWebPage::originatingObjectInNetworkRequests):
        * tests/qwebview/tst_qwebview.cpp:
        (tst_QWebView::reusePage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68517 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Philippe Normand <pnormand@igalia.com>
philn@webkit.org [Tue, 28 Sep 2010 14:57:37 +0000 (14:57 +0000)]
2010-09-28  Philippe Normand  <pnormand@igalia.com>

        Reviewed by Martin Robinson.

        Guard GRefPtr/GOwnPtr files with ENABLE(GLIB_SUPPORT)
        https://bugs.webkit.org/show_bug.cgi?id=46721

        Enable GOwnPtr/GRefPtr build only if glib support has been
        explicitly enabled using the WTF_ENABLE_GLIB_SUPPORT macro.

        * wtf/gobject/GOwnPtr.cpp:
        * wtf/gobject/GOwnPtr.h:
        * wtf/gobject/GRefPtr.cpp:
        * wtf/gobject/GRefPtr.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68516 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Andras Becsi <abecsi@webkit.org>
abecsi@webkit.org [Tue, 28 Sep 2010 14:51:04 +0000 (14:51 +0000)]
2010-09-28  Andras Becsi  <abecsi@webkit.org>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] ProcessLauncherHelper should not leave stray socket files in /tmp.
        https://bugs.webkit.org/show_bug.cgi?id=46722

        * UIProcess/Launcher/qt/ProcessLauncherQt.cpp:
        (WebKit::ProcessLauncherHelper::~ProcessLauncherHelper): close the socket.
        (WebKit::ProcessLauncherHelper::ProcessLauncherHelper): delete the helper on exit.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68515 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Benjamin Poulain <benjamin.poulain@nokia.com>
benjamin.poulain@nokia.com [Tue, 28 Sep 2010 14:40:39 +0000 (14:40 +0000)]
2010-09-28  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Andreas Kling.

        [Qt] Remove support for Qt 4.5
        https://bugs.webkit.org/show_bug.cgi?id=46718

        Remove the code for versions of Qt prior to 4.6.

        * DumpRenderTree/qt/EventSenderQt.cpp:
        (EventSender::addTouchPoint):
        (EventSender::updateTouchPoint):
        (EventSender::setTouchModifier):
        (EventSender::touchStart):
        (EventSender::touchMove):
        (EventSender::touchEnd):
        (EventSender::clearTouchPoints):
        (EventSender::releaseTouchPoint):
        (EventSender::sendTouchEvent):
        * DumpRenderTree/qt/EventSenderQt.h:
        * DumpRenderTree/qt/TextInputControllerQt.cpp:
        (TextInputController::setMarkedText):
        * QtTestBrowser/launcherwindow.cpp:
        (LauncherWindow::LauncherWindow):
        (LauncherWindow::initializeView):
        (LauncherWindow::createChrome):
        (LauncherWindow::sendTouchEvent):
        (LauncherWindow::eventFilter):
        (LauncherWindow::applyZoom):
        (LauncherWindow::setTouchMocking):
        * QtTestBrowser/launcherwindow.h:
        * QtTestBrowser/utils.cpp:
        (urlFromUserInput):
        * QtTestBrowser/webview.cpp:
        (WebViewGraphicsBased::animatedFlip):
        (WebViewGraphicsBased::animatedYFlip):
        * QtTestBrowser/webview.h:
        (WebViewGraphicsBased::setYRotation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68514 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Andreas Kling <andreas.kling@nokia.com>
andreas.kling@nokia.com [Tue, 28 Sep 2010 14:24:30 +0000 (14:24 +0000)]
2010-09-28  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        Canvas: Crash when setting a font with size in 'ex' units
        https://bugs.webkit.org/show_bug.cgi?id=46538

        update() the style's font after setting the style's font description.
        Needed because CSSPrimitiveValue::computeLengthDouble() later assumes
        that the style's font is properly initialized (for xHeight().)

        Fixes crash on IE test center's canvas-text-font-002 test.

        * html/canvas/CanvasRenderingContext2D.cpp:
        (WebCore::CanvasRenderingContext2D::setFont):
2010-09-28  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        Canvas: Crash when setting a font with size in 'ex' units
        https://bugs.webkit.org/show_bug.cgi?id=46538

        * fast/canvas/canvas-font-ex-units-crash-expected.txt: Added.
        * fast/canvas/canvas-font-ex-units-crash.html: Added.
        * fast/canvas/script-tests/canvas-font-ex-units-crash.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68513 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Benjamin Poulain <benjamin.poulain@nokia.com>
benjamin.poulain@nokia.com [Tue, 28 Sep 2010 14:23:04 +0000 (14:23 +0000)]
2010-09-28  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Andreas Kling.

        [Qt] Remove support for Qt 4.5
        https://bugs.webkit.org/show_bug.cgi?id=46718

        Remove the code for versions of Qt prior to 4.6.

        * UIProcess/API/qt/qgraphicswkview.cpp:
        (QGraphicsWKView::event):
        * UIProcess/API/qt/qwkpage.cpp:
        * UIProcess/API/qt/qwkpage_p.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68512 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 İsmail Dönmez <ismail@namtrac.org>
andreas.kling@nokia.com [Tue, 28 Sep 2010 14:14:58 +0000 (14:14 +0000)]
2010-09-28  İsmail Dönmez  <ismail@namtrac.org>

        Reviewed by Andreas Kling.

        Test for WINCE instead of WINCEBASIC, compiler always defines WINCE.
        Remove reference to unexisting path JavaScriptCore/os-wince.

        * JavaScriptCore.pri:
        * wtf/Assertions.cpp:
2010-09-28  İsmail Dönmez  <ismail@namtrac.org>

        Reviewed by Andreas Kling.

        Fix DRT compilation on WinCE. Introduce a WCECOMPAT variable
        which should point to wcecompat installation. Needs wcecompat
        library from http://github.com/mauricek/wcecompat .

        * DumpRenderTree/qt/DumpRenderTree.pro:
        * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
        (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68511 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Benjamin Poulain <benjamin.poulain@nokia.com>
benjamin.poulain@nokia.com [Tue, 28 Sep 2010 14:09:17 +0000 (14:09 +0000)]
2010-09-28  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] Remove support for Qt 4.5
        https://bugs.webkit.org/show_bug.cgi?id=46718

        Remove the code for versions of Qt prior to 4.6.

        * WebCore.pro:
        * features.pri:
        * platform/network/qt/QNetworkReplyHandler.cpp:
        (WebCore::QNetworkReplyHandler::QNetworkReplyHandler):
        (WebCore::QNetworkReplyHandler::start):
        * platform/network/qt/ResourceRequestQt.cpp:
        (WebCore::ResourceRequest::toNetworkRequest):
        * platform/qt/QWebPageClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68510 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Satish Sampath <satish@chromium.org>
satish@chromium.org [Tue, 28 Sep 2010 12:52:39 +0000 (12:52 +0000)]
2010-09-28  Satish Sampath  <satish@chromium.org>

        Unreviewed, enable a couple of speech input layout tests in chromium as they run successfully now.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68509 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoClean up QWebPage initialization
vestbo@webkit.org [Tue, 28 Sep 2010 12:30:16 +0000 (12:30 +0000)]
Clean up QWebPage initialization

Reviewed by Simon Hausmann.

* Api/qwebpage.cpp:
* Api/qwebpage_p.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68508 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Benjamin Poulain <benjamin.poulain@nokia.com>
benjamin.poulain@nokia.com [Tue, 28 Sep 2010 12:18:41 +0000 (12:18 +0000)]
2010-09-28  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] tst_QWebPage::testStopScheduledPageRefresh() fails on MeeGo handset
        https://bugs.webkit.org/show_bug.cgi?id=46706

        The test require the network to succeed or fail in order to work.
        On MeeGo, the network can simply timeout the because of the netword access
        management. This patch makes the test use local URLs to avoid the
        problem.

        * tests/qwebpage/tst_qwebpage.cpp:
        (TestNetworkManager::TestNetworkManager):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68507 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoConvert verbatim unicode values in comitter names to characters
vestbo@webkit.org [Tue, 28 Sep 2010 11:40:02 +0000 (11:40 +0000)]
Convert verbatim unicode values in comitter names to characters

Reviewed by Csaba Osztrogonác.

* committers-autocomplete.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68506 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed.
ossy@webkit.org [Tue, 28 Sep 2010 11:13:37 +0000 (11:13 +0000)]
Unreviewed.

[Qt] 3 media tests fail with Qt version 4.6.3 and 4.7.0
https://bugs.webkit.org/show_bug.cgi?id=46249

* platform/qt/Skipped: 3 failing media tests added until fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68505 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Andrei Popescu <andreip@google.com>
andreip@google.com [Tue, 28 Sep 2010 10:52:35 +0000 (10:52 +0000)]
2010-09-28  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        [Chromium] Linking error due to wrong guard in WebIDBTransaction.h
        https://bugs.webkit.org/show_bug.cgi?id=46710

        Remove the wrong guard.

        * public/WebIDBTransaction.h:
        (WebKit::WebIDBTransaction::getIDBTransactionBackendInterface):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68504 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 10:52:14 +0000 (10:52 +0000)]
2010-09-28  MORITA Hajime  <morrita@google.com>

        Unreviewed, added an expectation.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68503 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Andras Becsi <abecsi@webkit.org>
abecsi@webkit.org [Tue, 28 Sep 2010 10:35:04 +0000 (10:35 +0000)]
2010-09-28  Andras Becsi  <abecsi@webkit.org>

        Reviewed by Csaba Osztrogonác.

        [Qt] Fix generation of WebPageMessages.h and WebPageMessageReceiver.cpp
        to avoid rebuilding them every time.

        * DerivedSources.pro: add target to extra compilers.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68502 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Alejandro G. Castro <alex@igalia.com>
alex@webkit.org [Tue, 28 Sep 2010 09:39:52 +0000 (09:39 +0000)]
2010-09-28  Alejandro G. Castro  <alex@igalia.com>

        Unreviewed, GTK+ rebaselines.

        * platform/gtk/fast/dom/Window/window-property-descriptors-expected.txt:
        * platform/gtk/fast/dom/prototype-inheritance-expected.txt:
        * platform/gtk/fast/js/global-constructors-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68501 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Tue, 28 Sep 2010 09:36:57 +0000 (09:36 +0000)]
2010-09-28  Kent Tamura  <tkent@chromium.org>

        Unreviewed, run sort-Xcode-project-file

        * WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68500 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Huahui Wu <mediadependent@gmail.com>
commit-queue@webkit.org [Tue, 28 Sep 2010 09:26:31 +0000 (09:26 +0000)]
2010-09-28  Huahui Wu  <mediadependent@gmail.com>

        Reviewed by Steve Block.

        Add two multi-touch tests for Android.
        https://bugs.webkit.org/show_bug.cgi?id=45221

        The original tests (touch-target.html and basic-multi-touch-events.html)
        requires the system being able to release one point while maintain another point.
        Current Android system can not provide this to browser, hence these tests are
        slightly modified for Android. Since these are limited version of the original ones,
        other platforms should pass these tests if they can pass the original ones.

        * fast/events/touch/basic-multi-touch-events-limited-expected.txt: Added.
        * fast/events/touch/basic-multi-touch-events-limited.html: Added.
        * fast/events/touch/script-tests/basic-multi-touch-events-limited.js: Added.
        (touchEventCallback):
        (verifyTouchEvent):
        (verifyTouchPoint):
        (verifyTouch):
        (multiTouchSequence):
        * fast/events/touch/script-tests/touch-target-limited.js: Added.
        (touchStartHandler):
        (touchMoveHandler):
        * fast/events/touch/touch-target-limited-expected.txt: Added.
        * fast/events/touch/touch-target-limited.html: Added.
2010-09-28  Huahui Wu  <mediadependent@gmail.com>

        Reviewed by Steve Block.

        Add multi-touch for Android.
        https://bugs.webkit.org/show_bug.cgi?id=45221

        The Android touch event used to take one point as the touch point,
        it's now changed to a vector of points to support multi-touch.

        Tests: fast/events/touch/basic-multi-touch-events-limited.html
               fast/events/touch/touch-target-limited.html

        * platform/PlatformTouchEvent.h:
        * platform/PlatformTouchPoint.h:
        * platform/android/PlatformTouchEventAndroid.cpp:
        (WebCore::PlatformTouchEvent::PlatformTouchEvent):
        * platform/android/PlatformTouchPointAndroid.cpp:
        (WebCore::PlatformTouchPoint::PlatformTouchPoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68499 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 09:24:09 +0000 (09:24 +0000)]
2010-09-28 MORITA Hajime  <morrita@google.com>

        Unreviewed, added an expectation.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68498 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 08:32:37 +0000 (08:32 +0000)]
2010-09-28  MORITA Hajime  <morrita@google.com>

        Unreviewed, removed some expectations.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68497 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Daniel Bates <dbates@rim.com>
dbates@webkit.org [Tue, 28 Sep 2010 07:25:40 +0000 (07:25 +0000)]
2010-09-28  Daniel Bates  <dbates@rim.com>

        Rollout changeset 68493 <http://trac.webkit.org/changeset/68493>
        <https://bugs.webkit.org/show_bug.cgi?id=39136>

        Rollout changeset 68493 because it broke Sheriffbot's rollout feature.
        In particular, this change caused Sheriffbot to raise an exception when
        trying to parse the bug id on a bug page. We need to look into this
        some more.

        * Scripts/webkitpy/common/net/bugzilla.py:
        * Scripts/webkitpy/common/net/bugzilla_unittest.py:
        * Scripts/webkitpy/tool/bot/sheriff.py:
        * Scripts/webkitpy/tool/commands/download.py:
        * Scripts/webkitpy/tool/commands/queues.py:
        * Scripts/webkitpy/tool/commands/upload.py:
        * Scripts/webkitpy/tool/steps/closebug.py:
        * Scripts/webkitpy/tool/steps/obsoletepatches.py:
        * Scripts/webkitpy/tool/steps/preparechangelog.py:
        * Scripts/webkitpy/tool/steps/updatechangelogswithreviewer.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68496 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-28 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Tue, 28 Sep 2010 07:21:32 +0000 (07:21 +0000)]
2010-09-28  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Move RegressionWindow further up the dependency chain
        https://bugs.webkit.org/show_bug.cgi?id=46698

        Baby steps towards pushing this information into bug posts.

        * Scripts/webkitpy/common/net/buildbot.py:
        * Scripts/webkitpy/common/net/buildbot_unittest.py:
        * Scripts/webkitpy/common/net/regressionwindow.py:
        * Scripts/webkitpy/tool/commands/queries.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68495 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoFix assertion failure in a case that the dirty flag of a textarea element is changed
tkent@chromium.org [Tue, 28 Sep 2010 07:09:56 +0000 (07:09 +0000)]
Fix assertion failure in a case that the dirty flag of a textarea element is changed
https://bugs.webkit.org/show_bug.cgi?id=45681

Reviewed by Alexey Proskuryakov.

WebCore:

We need to call setNeedsValidityCheck() after updating m_isDirty.

Test: fast/forms/textarea-checkValidity-crash.html

* html/HTMLTextAreaElement.cpp:
(WebCore::HTMLTextAreaElement::setValue):
(WebCore::HTMLTextAreaElement::setNonDirtyValue):
(WebCore::HTMLTextAreaElement::setValueCommon):
* html/HTMLTextAreaElement.h:

LayoutTests:

* fast/forms/textarea-checkValidity-crash-expected.txt: Added.
* fast/forms/textarea-checkValidity-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68494 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Daniel Bates <dbates@rim.com>
dbates@webkit.org [Tue, 28 Sep 2010 06:39:16 +0000 (06:39 +0000)]
2010-09-27  Daniel Bates  <dbates@rim.com>

        Reviewed by Adam Barth.

        sheriffbot can't roll out security patches
        https://bugs.webkit.org/show_bug.cgi?id=39136

        Make SheriffBot determine if it's authorized to view a bug
        whose change it wants to rollout before it tries to rollout
        the change.

        Moreover, make both webkit-patch and Sheriffbot provide human-
        readable error messages when they are not authorized to view
        a bug and when a bug number is invalid.

        Currently, Sheriffbot does not parse Bugzilla bugs for
        <bug error="...">, which indicates an error when retrieving
        a bug. In particular, error="NotPermitted" if a person (or bot)
        is not authorized to view a bug. For such error="NotPermitted" bugs,
        Sheriffbot raises an exception when parsing the bug report and
        this exception does not explicitly indicate Sheriffbot's lack
        of authorization. Instead, Sheriffbot should explicitly check
        for the presence <bug error="..."> before operating on a bug
        and error with a human-readable message if it's not permitted
        to view it.

        * Scripts/webkitpy/common/net/bugzilla.py: Added BugzillaError class.
        * Scripts/webkitpy/common/net/bugzilla_unittest.py:
          - Added unit test test_bug_parsing_for_bugzilla_not_permitted_error().
          - Added unit test test_bug_parsing_for_bugzilla_not_found_error().
          - Added unit test test_bug_parsing_for_bugzilla_invalid_bug_id_error().
        * Scripts/webkitpy/tool/bot/sheriff.py: Modified to catch BugzillaError.
        * Scripts/webkitpy/tool/commands/download.py: Ditto.
        * Scripts/webkitpy/tool/commands/queues.py: Ditto.
        * Scripts/webkitpy/tool/commands/upload.py: Ditto.
        * Scripts/webkitpy/tool/steps/closebug.py: Ditto.
        * Scripts/webkitpy/tool/steps/obsoletepatches.py: Ditto.
        * Scripts/webkitpy/tool/steps/preparechangelog.py: Ditto.
        * Scripts/webkitpy/tool/steps/updatechangelogswithreviewer.py: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68493 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Tue, 28 Sep 2010 06:22:36 +0000 (06:22 +0000)]
2010-09-27  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Expose more more failure information from Buildbot to SheriffBot
        https://bugs.webkit.org/show_bug.cgi?id=46697

        This patch moves the information about what tests failured closer to
        SheriffBot.  There are still a couple more patches to go before
        SheriffBot can post this information to bugs, but this is a step in
        that direction.  Yay for unit tests, which caught some bugs in earlier
        versions of this patch.

        * Scripts/webkitpy/common/net/buildbot.py:
        * Scripts/webkitpy/common/net/buildbot_unittest.py:
        * Scripts/webkitpy/common/net/regressionwindow.py: Added.
        * Scripts/webkitpy/tool/commands/queries.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68492 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 28 Sep 2010 06:04:40 +0000 (06:04 +0000)]
2010-09-27  Eric Seidel  <eric@webkit.org>

        Unreviewed.  Fixing 500 error seen in the status server.

        It turns out that a = b = []; b.append(1); then a[0] will be 1!
        This should have been obvious to me, I guess, but it was not what I was expecting.

        * QueueStatusServer/model/activeworkitems.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68491 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed fix after r68460.
ossy@webkit.org [Tue, 28 Sep 2010 06:04:39 +0000 (06:04 +0000)]
Unreviewed fix after r68460.

* platform/qt/Skipped: Add http/tests/canvas/webgl/origin-clean-conformance.html, because 3D canvas is disabled by default.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68490 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed fix after r68444. Add Qt specific expected files.
ossy@webkit.org [Tue, 28 Sep 2010 05:42:27 +0000 (05:42 +0000)]
Unreviewed fix after r68444. Add Qt specific expected files.

* platform/qt/fast/repaint/repaint-svg-after-style-change-expected.txt: Copied from LayoutTests/fast/repaint/repaint-svg-after-style-change-expected.txt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68489 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed fix after r68440. Update Qt specific expected files.
ossy@webkit.org [Tue, 28 Sep 2010 05:36:22 +0000 (05:36 +0000)]
Unreviewed fix after r68440. Update Qt specific expected files.

* platform/qt/fast/dom/Window/window-properties-expected.txt:
* platform/qt/fast/dom/Window/window-property-descriptors-expected.txt:
* platform/qt/fast/dom/prototype-inheritance-expected.txt:
* platform/qt/fast/js/global-constructors-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68488 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 05:29:56 +0000 (05:29 +0000)]
2010-09-27  MORITA Hajime  <morrita@google.com>

        Unreviewed, added an expectation.

        * platform/chromium/fast/dom/prototype-inheritance-expected.txt:
        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68487 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Abhishek Arya <inferno@chromium.org>
inferno@chromium.org [Tue, 28 Sep 2010 05:01:40 +0000 (05:01 +0000)]
2010-09-27  Abhishek Arya  <inferno@chromium.org>

        Reviewed by Dan Bernstein.

        When the block RunIn is destroyed, its line box tree is not deleted. As a result, it
        gets later used during dirtying of inline text boxes step where this deleted parent is
        accessed. The fix is to delete the line box tree before destroying the runin block.

        https://bugs.webkit.org/show_bug.cgi?id=46376

        Test: fast/text/dirty-inline-textbox-crash.html

        * rendering/RenderText.cpp:
        (WebCore::RenderText::dirtyLineBoxes):
2010-09-27  Abhishek Arya  <inferno@chromium.org>

        Reviewed by Dan Bernstein.

        Tests we do not crash on ASSERT(!m_hasBadParent) when dirtying
        the inline textboxes.
        https://bugs.webkit.org/show_bug.cgi?id=46376

        * fast/text/dirty-inline-textbox-crash-expected.txt: Added.
        * fast/text/dirty-inline-textbox-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68485 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Tue, 28 Sep 2010 04:45:22 +0000 (04:45 +0000)]
2010-09-27  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        commit-queue should reject patches that fail to land
        https://bugs.webkit.org/show_bug.cgi?id=46694

        This can happen, for example, if there's no reviewer.  Without this
        patch, we'll keep retrying the patch.

        * Scripts/webkitpy/tool/bot/commitqueuetask.py:
        * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68484 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Eric Uhrhane <ericu@chromium.org>
commit-queue@webkit.org [Tue, 28 Sep 2010 04:28:00 +0000 (04:28 +0000)]
2010-09-27  Eric Uhrhane  <ericu@chromium.org>

        Reviewed by David Levin.

        Fix event sequencing in FileWriter
        https://bugs.webkit.org/show_bug.cgi?id=46544

        Only set readyState to DONE when we're about to send the last progress
        event associated with an operation.  Make sure all progress events come
        from backend calls, and aren't ever fired synchronously in response to
        user JS method calls.

        No new tests, as none of this is fully implemented yet.

        * fileapi/FileWriter.cpp:
        (WebCore::FileWriter::write):
        (WebCore::FileWriter::truncate):
        (WebCore::FileWriter::abort):
        (WebCore::FileWriter::didWrite):
        (WebCore::FileWriter::didTruncate):
        (WebCore::FileWriter::didFail):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68483 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 03:49:53 +0000 (03:49 +0000)]
2010-09-27 MORITA Hajime  <morrita@google.com>

        Unreviewed, added an expectation.

        * platform/chromium-mac/fast/forms/input-button-sizes-expected.checksum: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68482 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Chris Rogers <crogers@google.com>
crogers@google.com [Tue, 28 Sep 2010 03:27:27 +0000 (03:27 +0000)]
2010-09-27  Chris Rogers  <crogers@google.com>

        Reviewed by James Robinson.

        Add HRTFDatabase files
        https://bugs.webkit.org/show_bug.cgi?id=46233

        No new tests since audio API is not yet implemented.

        * platform/audio/HRTFDatabase.cpp: Added.
        (WebCore::HRTFDatabase::create):
        (WebCore::HRTFDatabase::HRTFDatabase):
        (WebCore::HRTFDatabase::getKernelsFromAzimuthElevation):
        (WebCore::HRTFDatabase::indexFromElevationAngle):
        * platform/audio/HRTFDatabase.h: Added.
        (WebCore::HRTFDatabase::numberOfAzimuths):
        (WebCore::HRTFDatabase::sampleRate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68481 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 03:22:25 +0000 (03:22 +0000)]
2010-09-27 MORITA Hajime  <morrita@google.com>

        Unreviewed, added an expectation.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68480 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Tue, 28 Sep 2010 03:03:33 +0000 (03:03 +0000)]
2010-09-27  Simon Fraser  <simon.fraser@apple.com>

        Rubber-stamped by Adele Peterson.

        Add a harness for running the CSS 2.1 test suite.

        This harness assumes that you've got a local a copy of the suite
        adjacent to the 'harness' directory.

        The harness lets you go through the tests in chapter order,
        denoting pass/fail/skipped for each test.

        It uses a client-side database to store the results of testing.

        * CSSTestSuiteHarness/harness/harness.css: Added.
        * CSSTestSuiteHarness/harness/harness.html: Added.
        * CSSTestSuiteHarness/harness/harness.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68479 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoBuild fix. Not reviewed.
hclam@chromium.org [Tue, 28 Sep 2010 02:53:37 +0000 (02:53 +0000)]
Build fix. Not reviewed.

Use if instead of switch to avoid compilation error.

* src/VideoFrameChromiumImpl.cpp:
(WebKit::VideoFrameChromiumImpl::requiredTextureSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68478 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoBuild fix. Not reviewed.
hclam@chromium.org [Tue, 28 Sep 2010 02:37:38 +0000 (02:37 +0000)]
Build fix. Not reviewed.

WebCore:

Move constant definitions to VideoLayerChromium.cpp.
A proper fix should come later.

* platform/graphics/chromium/VideoLayerChromium.cpp:

WebKit/chromium:

Remove constant definitions from this file.

* src/VideoFrameChromiumImpl.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68477 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 02:37:29 +0000 (02:37 +0000)]
2010-09-27 MORITA Hajime  <morrita@google.com>

        Unreviewed, added an expectation.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68476 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoMake Document::nodesFromRect more flexible
tonikitoo@webkit.org [Tue, 28 Sep 2010 02:28:25 +0000 (02:28 +0000)]
Make Document::nodesFromRect more flexible
https://bugs.webkit.org/show_bug.cgi?id=46336

Patch by Antonio Gomes <agomes@rim.com> on 2010-09-23
Reviewed by Kenneth Christiansen.

The patch changes the API for the Document::nodesFromRect() method as follows:

from
- RefPtf<NodeList> nodesFromRect(int x, int y, unsigned horizontalPadding, unsigned verticalPadding, bool ignoreClipping)

to
- RefPtf<NodeList> nodesFromRect(int x, int y, unsigned topPadding, unsigned rightPadding, unsigned bottomPadding,
                                 unsigned leftPadding, bool ignoreClipping)

Previously the method was receiving as parameter horizontal and vertical padding values.
In practice, the horizontal padding was being used to expand a given center point in both
right and left directions, while vertical padding was expanding it to the top and bottom directions.

Although this design makes it possible to improve the touch experience on mobile devices
(Document::nodesFromRect exposes the rect based hit testing system), empirical experiments
showed that, specially in capacitive touch screens, users tend to tap below elements. So
for even more accurate results (which means here the possibility of pleasant tapping experience),
it makes sense to use a region that is offset more above the touch point, favoring elements
above the touch point.

The patch basically implements that, increasing the API flexible and power by making it possible
to specify different padding values for each direction in a given orientation.

Changes were made accordingly to the HitTestResult, RenderLayer, RenderImage and RenderLineBoxList
classes.

WebCore:

* WebCore.exp.in:
* dom/Document.cpp:
(WebCore::Document::nodesFromRect):
* dom/Document.h:
* page/EventHandler.cpp:
(WebCore::EventHandler::hitTestResultAtPoint): For simplicity, I did not change the signature
    but the way the padding value passed in as parameter is being used to construct a HitTestResult.
* rendering/HitTestResult.cpp: Changed the rect-based bits from using IntSize (padding) to separated paddings
    for each direction.
(WebCore::HitTestResult::HitTestResult): Ditto.
(WebCore::HitTestResult::operator=): Ditto.
(WebCore::HitTestResult::rectFromPoint): Ditton.
* rendering/HitTestResult.h:
(WebCore::HitTestResult::topPadding): getter for the top padding.
(WebCore::HitTestResult::rightPadding): getter for the right padding.
(WebCore::HitTestResult::bottomPadding): getter for the bottom padding.
(WebCore::HitTestResult::leftPadding): getter for the left padding.
(WebCore::HitTestResult::rectFromPoint): Added a static method that given a center point and padding values for
    each direction, returns the rect that is going to be used in WebCore's hit testing processing.
* rendering/RenderImage.cpp: Changed the rect-based bits from using IntSize (padding) to separated paddings for each direction.
(WebCore::RenderImage::nodeAtPoint): Ditto.
* rendering/RenderLayer.cpp: Ditto.
(WebCore::RenderLayer::hitTestLayer): Ditto.
(WebCore::RenderLayer::hitTestList): Ditto.
* rendering/RenderLineBoxList.cpp: Ditto.
(WebCore::RenderLineBoxList::hitTest): Ditto.

LayoutTests:

* fast/dom/nodesFromRect-basic-expected.txt:
* fast/dom/nodesFromRect-basic.html:
* fast/dom/resources/nodesFromRect.js:
(check):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68475 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 28 Sep 2010 02:21:30 +0000 (02:21 +0000)]
2010-09-27  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Patch locks should expire if a patch is marked for retry
        https://bugs.webkit.org/show_bug.cgi?id=46682

        This was part Adam's original CommitQueueTask design,
        but support for it was missing from the server.
        I added the support, but triggering lock-release based on this
        special "retry" status feels a bit strange so I added a FIXME.

        I also changed the text in queuestatus.html to say "Lock Acquired"
        since "Last Activity" isn't really true.  We only update the lock
        date when the patch is started, not on every status update.

        I also noticed an exception in next-patch, which I fixed by re-writing
        the unzip logic in activeworkitems.py again.

        * QueueStatusServer/handlers/updatestatus.py:
        * QueueStatusServer/model/activeworkitems.py:
        * QueueStatusServer/templates/queuestatus.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68474 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 02:20:16 +0000 (02:20 +0000)]
2010-09-27  MORITA Hajime  <morrita@google.com>

        Unreviewed, added missing expectations.

        [Chromium] fast/forms/input-button-sizes.html failing since r68459
        https://bugs.webkit.org/show_bug.cgi?id=46686

        * platform/chromium-linux/fast/forms/input-button-sizes-expected.checksum: Added.
        * platform/chromium-linux/fast/forms/input-button-sizes-expected.png: Added.
        * platform/chromium-linux/fast/forms/input-button-sizes-expected.txt: Added.
        * platform/chromium-win/fast/forms/input-button-sizes-expected.checksum: Added.
        * platform/chromium-win/fast/forms/input-button-sizes-expected.png: Added.
        * platform/chromium-win/fast/forms/input-button-sizes-expected.txt: Added.
        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68473 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoBuild fix. Not reviewed.
hclam@chromium.org [Tue, 28 Sep 2010 02:17:58 +0000 (02:17 +0000)]
Build fix. Not reviewed.

Fix warnings again.

* src/VideoFrameChromiumImpl.cpp:
(WebKit::VideoFrameChromiumImpl::requiredTextureSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68472 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agohttps://bugs.webkit.org/show_bug.cgi?id=46688 - Add an API to get the WKBackForwardIt...
beidson@apple.com [Tue, 28 Sep 2010 02:15:50 +0000 (02:15 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=46688 - Add an API to get the WKBackForwardItem at an arbitrary index.

Reviewed by Sam Weinig.

* UIProcess/API/C/WKBackForwardList.cpp:
(WKBackForwardListGetItemAtIndex):
* UIProcess/API/C/WKBackForwardList.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68471 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Chris Rogers <crogers@google.com>
crogers@google.com [Tue, 28 Sep 2010 02:08:30 +0000 (02:08 +0000)]
2010-09-27  Chris Rogers  <crogers@google.com>

        Reviewed by James Robinson.

        Add HRTFDatabaseLoader files
        https://bugs.webkit.org/show_bug.cgi?id=46234

        No new tests since audio API is not yet implemented.

        * platform/audio/HRTFDatabaseLoader.cpp: Added.
        (WebCore::HRTFDatabaseLoader::createAndLoadAsynchronouslyIfNecessary):
        (WebCore::HRTFDatabaseLoader::HRTFDatabaseLoader):
        (WebCore::HRTFDatabaseLoader::~HRTFDatabaseLoader):
        (WebCore::databaseLoaderEntry):
        (WebCore::HRTFDatabaseLoader::load):
        (WebCore::HRTFDatabaseLoader::loadAsynchronously):
        (WebCore::HRTFDatabaseLoader::isLoaded):
        (WebCore::HRTFDatabaseLoader::defaultHRTFDatabase):
        * platform/audio/HRTFDatabaseLoader.h: Added.
        (WebCore::HRTFDatabaseLoader::database):
        (WebCore::HRTFDatabaseLoader::databaseSampleRate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68470 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Michael Saboff <msaboff@apple.com>
msaboff@apple.com [Tue, 28 Sep 2010 01:52:44 +0000 (01:52 +0000)]
2010-09-27  Michael Saboff  <msaboff@apple.com>

        Reviewed by Geoffrey Garen.

        Changed the initialization of JSArray objects to have space for
        3 elements for the constructor that takes a ArgList argument.
        This improves v8-deltablue performance by about 2.8% by reducing
        the number of realloc() calls.
        https://bugs.webkit.org/show_bug.cgi?id=46664

        * runtime/JSArray.cpp:
        (JSC::JSArray::JSArray):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68469 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoBuild fix. Not reviewed.
hclam@chromium.org [Tue, 28 Sep 2010 01:50:41 +0000 (01:50 +0000)]
Build fix. Not reviewed.

Add a default for unhandled cases for switch statement.

* src/VideoFrameChromiumImpl.cpp:
(WebKit::VideoFrameChromiumImpl::requiredTextureSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68468 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 MORITA Hajime <morrita@google.com>
morrita@google.com [Tue, 28 Sep 2010 01:44:05 +0000 (01:44 +0000)]
2010-09-27  MORITA Hajime  <morrita@google.com>

        Unreviewed, add an expectation.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68467 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Tony Chang <tony@chromium.org>
tony@chromium.org [Tue, 28 Sep 2010 01:40:31 +0000 (01:40 +0000)]
2010-09-27  Tony Chang  <tony@chromium.org>

        Reviewed by David Levin.

        [chromium] fix a warning when compiling DRT on 32-bit linux
        https://bugs.webkit.org/show_bug.cgi?id=46641

        * DumpRenderTree/chromium/TestEventPrinter.cpp:
        (DRTPrinter::handleImage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68466 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Ryosuke Niwa <rniwa@webkit.org>
rniwa@webkit.org [Tue, 28 Sep 2010 01:22:57 +0000 (01:22 +0000)]
2010-09-27  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Darin Adler.

        queryCommandState with justifyCenter, justifyLeft, and justifyRight always return false
        https://bugs.webkit.org/show_bug.cgi?id=45910

        Implemented queryCommandState for justifyCenter, justifyLeft, and justifyRight commands.
        Added getTextAlignment to convert -webkit-center, -webkit-left, and -webkit-right to
        center, left, and right respectively because they can be treated equally for editing purposes.

        Test: editing/execCommand/query-text-alignment.html

        * editing/ApplyStyleCommand.cpp:
        (WebCore::getTextAlignment): Added.
        (WebCore::getPropertiesNotIn): Uses getTextAlignment.
        * editing/EditorCommand.cpp:
        (WebCore::stateJustifyCenter): Added.
        (WebCore::stateJustifyLeft): Added.
        (WebCore::stateJustifyRight): Added.
        (WebCore::createCommandMap): Refers to stateJustifyCenter, stateJustifyLeft, and stateJustifyRight.
2010-09-27  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Darin Adler.

        queryCommandState with justifyCenter, justifyLeft, and justifyRight always return false
        https://bugs.webkit.org/show_bug.cgi?id=45910

        Added tests for queryCommandState with justifyCenter, justifyLeft, and justifyRight commands.

        * editing/execCommand/query-text-alignment-expected.txt: Added.
        * editing/execCommand/query-text-alignment.html: Added.
        * editing/execCommand/script-tests/query-text-alignment.js: Added.
        (queryTextAlignment):
        (selectFirstPosition):
        (selectMiddleOfHelloWorld):
        (runRangeTests):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68465 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed. Build fix.
hclam@chromium.org [Tue, 28 Sep 2010 01:18:59 +0000 (01:18 +0000)]
Unreviewed. Build fix.

Reorder member declarations and fix other warnings.

* platform/graphics/chromium/VideoLayerChromium.cpp:
(WebCore::VideoLayerChromium::VideoLayerChromium):
(WebCore::VideoLayerChromium::~VideoLayerChromium):
(WebCore::VideoLayerChromium::updateContents):
(WebCore::VideoLayerChromium::determineTextureFormat):
(WebCore::VideoLayerChromium::allocateTexturesIfNeeded):
* platform/graphics/chromium/VideoLayerChromium.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68464 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 James Robinson <jamesr@chromium.org>
jamesr@google.com [Tue, 28 Sep 2010 01:18:05 +0000 (01:18 +0000)]
2010-09-27  James Robinson  <jamesr@chromium.org>

        Update baselines for fast/repaint/repaint-svg-after-style-change.html

        * fast/repaint/repaint-svg-after-style-change-expected.txt:
        * platform/chromium/fast/repaint/repaint-svg-after-style-change-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68463 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Kenneth Russell <kbr@google.com>
kbr@google.com [Tue, 28 Sep 2010 01:00:21 +0000 (01:00 +0000)]
2010-09-27  Kenneth Russell  <kbr@google.com>

        Unreviewed, Qt build fix. Stub out CanvasRenderingContext::checkOrigin(const HTMLVideoElement*)
        on platforms that don't support video after r68460.

        * html/canvas/CanvasRenderingContext.cpp:
        (WebCore::CanvasRenderingContext::checkOrigin):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68462 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoCrashTracer: 1,803 crashes in Safari at com.apple.WebCore: -[AccessibilityObjectWrapp...
cfleizach@apple.com [Tue, 28 Sep 2010 00:43:46 +0000 (00:43 +0000)]
CrashTracer: 1,803 crashes in Safari at com.apple.WebCore: -[AccessibilityObjectWrapper accessibilityIsIgnored] + 56
https://bugs.webkit.org/show_bug.cgi?id=46662

Reviewed by Beth Dakin.

WebCore:

Before calling updateBackingStore(), the wrapper object needs to be retained, lest it be invalidated during
the updateBackingStore call. This consolidates all calls to updateBackingStore().

Test: platform/mac/accessibility/removing-textarea-after-edit-crash.html

* accessibility/mac/AXObjectCacheMac.mm:
(WebCore::AXObjectCache::postPlatformNotification):
    Use an ASSERT here to catch the crash because DRT won't cause AppKit to post notifications.
* accessibility/mac/AccessibilityObjectWrapper.mm:
(-[AccessibilityObjectWrapper prepareAccessibilityMethod]):
(-[AccessibilityObjectWrapper accessibilityActionNames]):
(-[AccessibilityObjectWrapper accessibilityAttributeNames]):
(-[AccessibilityObjectWrapper accessibilityAttributeValue:]):
(-[AccessibilityObjectWrapper accessibilityFocusedUIElement]):
(-[AccessibilityObjectWrapper accessibilityHitTest:]):
(-[AccessibilityObjectWrapper accessibilityIsAttributeSettable:]):
(-[AccessibilityObjectWrapper accessibilityIsIgnored]):
(-[AccessibilityObjectWrapper accessibilityParameterizedAttributeNames]):
(-[AccessibilityObjectWrapper accessibilityPerformPressAction]):
(-[AccessibilityObjectWrapper accessibilityPerformIncrementAction]):
(-[AccessibilityObjectWrapper accessibilityPerformDecrementAction]):
(-[AccessibilityObjectWrapper accessibilityPerformAction:]):
(-[AccessibilityObjectWrapper accessibilitySetValue:forAttribute:]):
(-[AccessibilityObjectWrapper accessibilityAttributeValue:forParameter:]):
(-[AccessibilityObjectWrapper accessibilityIndexOfChild:]):
(-[AccessibilityObjectWrapper accessibilityArrayAttributeCount:]):
(-[AccessibilityObjectWrapper accessibilityArrayAttributeValues:index:maxCount:]):

LayoutTests:

* platform/mac/accessibility/removing-textarea-after-edit-crash-expected.txt: Added.
* platform/mac/accessibility/removing-textarea-after-edit-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68461 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Adrienne Walker <enne@google.com>
kbr@google.com [Tue, 28 Sep 2010 00:41:26 +0000 (00:41 +0000)]
2010-09-27  Adrienne Walker  <enne@google.com>

        Reviewed by Adam Barth.

        WebGL's readPixels doesn't respect origin-clean flag

        Push CanvasRenderingContext2D checkOrigin logic into base class so it can be shared between different kinds of canvas contexts.
        https://bugs.webkit.org/show_bug.cgi?id=45718

        Test: LayoutTests/http/tests/canvas/webgl/origin-clean-conformance.html

        * html/canvas/CanvasRenderingContext.cpp:
        (WebCore::CanvasRenderingContext::checkOrigin):
        * html/canvas/CanvasRenderingContext.h:
        (WebCore::CanvasRenderingContext::~CanvasRenderingContext):
        (WebCore::CanvasRenderingContext::canvas):
        (WebCore::CanvasRenderingContext::is2d):
        (WebCore::CanvasRenderingContext::is3d):
        (WebCore::CanvasRenderingContext::isAccelerated):
        (WebCore::CanvasRenderingContext::paintRenderingResultsToCanvas):
        (WebCore::CanvasRenderingContext::paintsIntoCanvasBuffer):
        (WebCore::CanvasRenderingContext::platformLayer):
        * html/canvas/CanvasRenderingContext2D.cpp:
        (WebCore::CanvasRenderingContext2D::setStrokeStyle):
        (WebCore::CanvasRenderingContext2D::setFillStyle):
        (WebCore::CanvasRenderingContext2D::drawImage):
        (WebCore::CanvasRenderingContext2D::drawImageFromRect):
        * html/canvas/CanvasRenderingContext2D.h:
        * html/canvas/WebGLRenderingContext.cpp:
        (WebCore::WebGLRenderingContext::readPixels):
        (WebCore::WebGLRenderingContext::texImage2D):
        (WebCore::WebGLRenderingContext::videoFrameToImage):
        (WebCore::WebGLRenderingContext::texSubImage2D):
        * html/canvas/WebGLRenderingContext.h:
        * html/canvas/WebGLRenderingContext.idl:
2010-09-27  Adrienne Walker  <enne@google.com>

        Reviewed by Adam Barth.

        WebGL's readPixels doesn't respect origin-clean flag

        Add new test by converting Khronos conformance test to be able to run within LayoutTests/http.
        https://bugs.webkit.org/show_bug.cgi?id=45718

        * http/tests/canvas/webgl/origin-clean-conformance-expected.txt: Added.
        * http/tests/canvas/webgl/origin-clean-conformance.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68460 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Darin Adler <darin@apple.com>
darin@apple.com [Tue, 28 Sep 2010 00:40:07 +0000 (00:40 +0000)]
2010-09-27  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        REGRESSION (r68054): Buttons no longer render at correct sizes
        https://bugs.webkit.org/show_bug.cgi?id=46513

        Test: fast/forms/input-button-sizes.html

        The setUpButtonCell function was running only once, but needed to run every time
        through the button function.

        * platform/mac/ThemeMac.mm:
        (WebCore::leakButtonCell): Added. Separate one-time initialization code.
        (WebCore::setUpButtonCell): Removed the first-time initialization part of this function.
        (WebCore::button): Merged the three functions back into one.
2010-09-27  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        REGRESSION (r68054): Buttons no longer render at correct sizes
        https://bugs.webkit.org/show_bug.cgi?id=46513

        * fast/forms/input-button-sizes.html: Copied from LayoutTests/fast/forms/button-sizes.html.
        Turned into a distinct test of <input type=button>; old test was for <button>.

        * platform/mac/fast/forms/input-button-sizes-expected.checksum: Added.
        * platform/mac/fast/forms/input-button-sizes-expected.png: Added.
        * platform/mac/fast/forms/input-button-sizes-expected.txt: Added.

        * platform/mac/fast/forms/disabled-select-change-index-expected.checksum: Added.
        * platform/mac/fast/forms/disabled-select-change-index-expected.png: Added.
        * platform/mac/fast/forms/file-input-disabled-expected.checksum: Added.
        * platform/mac/fast/forms/file-input-disabled-expected.png: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68459 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Zhenyao Mo <zmo@google.com>
zmo@google.com [Tue, 28 Sep 2010 00:31:36 +0000 (00:31 +0000)]
2010-09-27  Zhenyao Mo  <zmo@google.com>

        Unreviewed, fix a typo.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68458 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago[Qt] Remove setDeviceSize methods
kenneth@webkit.org [Tue, 28 Sep 2010 00:31:09 +0000 (00:31 +0000)]
[Qt] Remove setDeviceSize methods
https://bugs.webkit.org/show_bug.cgi?id=46347

Reviewed by Antonio Gomes.

Remove our setDeviceHeight() API and obtain it automatically from the
system. Also for testing purposes, make it possible to override the
values via two newly introduced environment variables.

* Api/qgraphicswebview.cpp:
* Api/qgraphicswebview.h:
* Api/qwebpage.cpp:
(getintenv):
(queryDeviceSizeForScreenContainingWidget):
(QWebPage::viewportConfigurationForSize):
* WebCoreSupport/PageClientQt.cpp:
(WebCore::PageClientQGraphicsWidget::windowRect):
* symbian/eabi/QtWebKitu.def:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68457 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoTurns out people other than me might want to build this thing.
andersca@apple.com [Tue, 28 Sep 2010 00:27:34 +0000 (00:27 +0000)]
Turns out people other than me might want to build this thing.

* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68456 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoBug 46680 - Inlining string concatenation can regress interpreter performance
barraclough@apple.com [Tue, 28 Sep 2010 00:25:06 +0000 (00:25 +0000)]
Bug 46680 - Inlining string concatenation can regress interpreter performance
<rdar://problem/8362752> REGRESSION: ~6.4% sunspider regression in interpreter
Do not inline calls to string concatenation in the interpret loop.

Reviewed by Darin Adler.

* interpreter/Interpreter.cpp:
(JSC::concatenateStrings):
(JSC::Interpreter::privateExecute):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68455 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago* Scripts/webkit2: Added property svn:ignore.
darin@apple.com [Tue, 28 Sep 2010 00:16:55 +0000 (00:16 +0000)]
* Scripts/webkit2: Added property svn:ignore.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68454 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 28 Sep 2010 00:14:21 +0000 (00:14 +0000)]
2010-09-27  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        queue-status should report when the patch was last locked to a queue
        https://bugs.webkit.org/show_bug.cgi?id=46674

        This isn't necessarily the best way to expose this information
        but having this accessible via the web interface is very
        useful until we come up with a nicer way to display this.

        I also cleaned up the code in activeworkitems.py a little
        to use list comprehensions and to have the code work with
        pairs instead of two lists at once.  Eventually I think those
        item/time pairs need to be their own little helper class.

        * QueueStatusServer/handlers/queuestatus.py:
        * QueueStatusServer/model/activeworkitems.py:
        * QueueStatusServer/templates/queuestatus.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68453 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago* Platform/CoreIPC/Connection.h:
andersca@apple.com [Tue, 28 Sep 2010 00:14:00 +0000 (00:14 +0000)]
* Platform/CoreIPC/Connection.h:
(CoreIPC::Connection::sendSync):
Add new sendSync overload that takes a message struct.

Reviewed by Sam Weinig.

* PluginProcess/WebProcessConnection.cpp:
(WebKit::WebProcessConnection::didReceiveSyncMessage):
Call didReceiveSyncWebProcessConnectionMessage.

* WebProcess/Plugins/PluginProcessConnection.cpp:
(WebKit::PluginProcessConnection::addPluginProxy):
Add the plug-in proxy to the map.

(WebKit::PluginProcessConnection::removePluginProxy):
Remove the plug-in proxy from the map. If the map is empty, disconnect from the
plug-in process.

* WebProcess/Plugins/PluginProxy.cpp:
(WebKit::generatePluginInstanceID):
Generate a unique plug-in instance ID.

(WebKit::PluginProxy::PluginProxy):
Add the plug-in proxy to the map.

(WebKit::PluginProxy::~PluginProxy):
Remove the plug-in proxy from the map.

(WebKit::PluginProxy::initialize):
Ask the plug-in process to create a plug-in.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68452 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Tony Chang <tony@chromium.org>
tony@chromium.org [Tue, 28 Sep 2010 00:04:44 +0000 (00:04 +0000)]
2010-09-27  Tony Chang  <tony@chromium.org>

        Reviewed by David Levin.

        [chromium] fix detection of missing fonts on DRT
        https://bugs.webkit.org/show_bug.cgi?id=46651

        * DumpRenderTree/chromium/TestShellGtk.cpp:
        (setupFontconfig):  Only print an error message if lohit isn't found in both locations.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68451 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Tony Chang <tony@chromium.org>
tony@chromium.org [Tue, 28 Sep 2010 00:01:10 +0000 (00:01 +0000)]
2010-09-27  Tony Chang  <tony@chromium.org>

        Reviewed by Dimitri Glazkov.

        add webkit unit tests to the chromium testers
        https://bugs.webkit.org/show_bug.cgi?id=46669

        * BuildSlaveSupport/build.webkit.org-config/master.cfg:
        * Scripts/run-chromium-webkit-unit-tests: Fix a bug where we weren't using the right configuration.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68450 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed.
kseo@webkit.org [Mon, 27 Sep 2010 23:55:02 +0000 (23:55 +0000)]
Unreviewed.

Adding myself to the committers list.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68449 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Zhenyao Mo <zmo@google.com>
zmo@google.com [Mon, 27 Sep 2010 23:52:59 +0000 (23:52 +0000)]
2010-09-27  Zhenyao Mo  <zmo@google.com>

        Unreviewed.  Three webgl tests TIMEOUT after hooking up ANGLE with in-process-webgl port.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68448 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Victoria Kirst <vrk@google.com>
commit-queue@webkit.org [Mon, 27 Sep 2010 23:47:54 +0000 (23:47 +0000)]
2010-09-27  Victoria Kirst  <vrk@google.com>

        Reviewed by James Robinson.

        Modifies VideoLayerChromium to do YUV to RGB color conversion in the
        GPU. Also adds support for RGBA video frame formats, though this code
        path is currently never run because the video frames are in YV12
        format.
        https://bugs.webkit.org/show_bug.cgi?id=45069

        * platform/graphics/chromium/LayerRendererChromium.cpp:
        (WebCore::LayerRendererChromium::initializeSharedObjects):
        (WebCore::LayerRendererChromium::cleanupSharedObjects):
        * platform/graphics/chromium/LayerRendererChromium.h:
        (WebCore::LayerRendererChromium::videoLayerSharedValues):
        * platform/graphics/chromium/VideoLayerChromium.cpp:
        (WebCore::VideoLayerChromium::SharedValues::SharedValues):
        (WebCore::VideoLayerChromium::SharedValues::~SharedValues):
        (WebCore::VideoLayerChromium::VideoLayerChromium):
        (WebCore::VideoLayerChromium::~VideoLayerChromium):
        (WebCore::VideoLayerChromium::updateContents):
        (WebCore::VideoLayerChromium::updateYUVContents):
        (WebCore::VideoLayerChromium::allocateYUVTextures):
        (WebCore::VideoLayerChromium::updateYUVTextures):
        (WebCore::VideoLayerChromium::updateRGBAContents):
        (WebCore::VideoLayerChromium::allocateTexture):
        (WebCore::VideoLayerChromium::updateTexture):
        (WebCore::VideoLayerChromium::updateContentsCompleted):
        (WebCore::VideoLayerChromium::draw):
        (WebCore::VideoLayerChromium::drawYUV):
        (WebCore::VideoLayerChromium::drawRGBA):
        * platform/graphics/chromium/VideoLayerChromium.h:
        (WebCore::VideoLayerChromium::drawsContent):
        (WebCore::VideoLayerChromium::SharedValues::yuvShaderProgram):
        (WebCore::VideoLayerChromium::SharedValues::rgbaShaderProgram):
        (WebCore::VideoLayerChromium::SharedValues::yuvShaderMatrixLocation):
        (WebCore::VideoLayerChromium::SharedValues::rgbaShaderMatrixLocation):
        (WebCore::VideoLayerChromium::SharedValues::yuvWidthScaleFactorLocation):
        (WebCore::VideoLayerChromium::SharedValues::rgbaWidthScaleFactorLocation):
        (WebCore::VideoLayerChromium::SharedValues::yTextureLocation):
        (WebCore::VideoLayerChromium::SharedValues::uTextureLocation):
        (WebCore::VideoLayerChromium::SharedValues::vTextureLocation):
        (WebCore::VideoLayerChromium::SharedValues::yuvAlphaLocation):
        (WebCore::VideoLayerChromium::SharedValues::rgbaTextureLocation):
        (WebCore::VideoLayerChromium::SharedValues::ccMatrixLocation):
        (WebCore::VideoLayerChromium::SharedValues::initialized):
2010-09-27  Victoria Kirst  <vrk@google.com>

        Reviewed by James Robinson.

        Fixing constants in VideoFrameChromiumImpl so that it overloads the
        VideoFrameChromium const declarations. Also adding logic such that a
        VideoFrameChromium understands what size its texture should be based
        on frame format.
        https://bugs.webkit.org/show_bug.cgi?id=45069

        * src/VideoFrameChromiumImpl.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68447 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWebCore: ImageDecoderSkia.cpp needs to check for allocator failure when copying
pkasting@chromium.org [Mon, 27 Sep 2010 23:44:36 +0000 (23:44 +0000)]
WebCore: ImageDecoderSkia.cpp needs to check for allocator failure when copying
bitmaps.
https://bugs.webkit.org/show_bug.cgi?id=46437

Reviewed by James Robinson.

* manual-tests/large-size-image-crash.html: Added.
* manual-tests/resources/large-size-image-crash.gif: Added.
* platform/image-decoders/ImageDecoder.cpp:
(WebCore::RGBA32Buffer::copyBitmapData):
* platform/image-decoders/ImageDecoder.h:
* platform/image-decoders/gif/GIFImageDecoder.cpp:
(WebCore::GIFImageDecoder::initFrameBuffer):
* platform/image-decoders/qt/RGBA32BufferQt.cpp:
(WebCore::RGBA32Buffer::copyBitmapData):
* platform/image-decoders/skia/ImageDecoderSkia.cpp:
(WebCore::RGBA32Buffer::copyBitmapData):

LayoutTests: This resource should have been in r62399; without it the test no-ops.
https://bugs.webkit.org/show_bug.cgi?id=41487

Reviewed by James Robinson.

* fast/images/resources/large-size-image-crash.jpeg: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68446 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd WebProcessConnection CreatePlugin message
andersca@apple.com [Mon, 27 Sep 2010 23:38:21 +0000 (23:38 +0000)]
Add WebProcessConnection CreatePlugin message
https://bugs.webkit.org/show_bug.cgi?id=46668

Reviewed by Adam Roben.

* DerivedSources.make:
Add WebProcessConnection.

* Platform/CoreIPC/HandleMessage.h:
Add handleMessage overload for a sync message with two input parameters
and one output parameter.

(CoreIPC::handleMessage):
* Platform/CoreIPC/MessageID.h:
Add MessageClassWebProcessConnection.

* PluginProcess/WebProcessConnection.cpp:
(WebKit::WebProcessConnection::createPlugin):
Add stub.

* PluginProcess/WebProcessConnection.messages.in: Added.
* Scripts/webkit2/messages.py:
Include headers directly for types that we believe are nested structs.

* Scripts/webkit2/messages_unittest.py:
Update expected results.

* WebKit2.xcodeproj/project.pbxproj:
Add new files.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68445 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 James Robinson <jamesr@chromium.org>
jamesr@google.com [Mon, 27 Sep 2010 23:38:13 +0000 (23:38 +0000)]
2010-09-27  James Robinson  <jamesr@chromium.org>

        Reviewed by Simon Fraser.

        Repaint test for dynamically changing an SVG element's style attribute
        https://bugs.webkit.org/show_bug.cgi?id=46499

        * fast/repaint/repaint-svg-after-style-change-expected.checksum: Added.
        * fast/repaint/repaint-svg-after-style-change-expected.png: Added.
        * fast/repaint/repaint-svg-after-style-change-expected.txt: Added.
        * fast/repaint/repaint-svg-after-style-change.html: Added.
2010-09-27  James Robinson  <jamesr@chromium.org>

        Reviewed by Simon Fraser.

        Revert r63307, it broke repainting dynamically changed SVG content
        https://bugs.webkit.org/show_bug.cgi?id=46499

        This patch reverts r63307 and adds a regression test for what broke.

        Test: fast/repaint/repaint-svg-after-style-change.html

        * rendering/RenderForeignObject.cpp:
        (WebCore::RenderForeignObject::layout):
        * rendering/RenderPath.cpp:
        (WebCore::RenderPath::layout):
        * rendering/RenderSVGContainer.cpp:
        (WebCore::RenderSVGContainer::layout):
        * rendering/RenderSVGImage.cpp:
        (WebCore::RenderSVGImage::layout):
        * rendering/RenderSVGRoot.cpp:
        (WebCore::RenderSVGRoot::layout):
        * rendering/RenderSVGText.cpp:
        (WebCore::RenderSVGText::layout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68444 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Mon, 27 Sep 2010 23:35:53 +0000 (23:35 +0000)]
2010-09-27  Martin Robinson  <mrobinson@igalia.com>

        Unskip a test that just needed a platform-specific baseline and add a baseline for it.

        * platform/gtk/Skipped: Unskip test with new baseline.
        * platform/gtk/editing/selection/extend-selection-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68443 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Vangelis Kokkevis <vangelis@chromium.org>
vangelis@chromium.org [Mon, 27 Sep 2010 23:35:40 +0000 (23:35 +0000)]
2010-09-27  Vangelis Kokkevis  <vangelis@chromium.org>

        Reviewed by James Robinson.

        [chromium] WebViewImpl now holds a ref counted pointer to the LayerRendererChromium
        to ensure that the compositor does not get destroyed before the layers used
        by it do. This was done to ensure that layers properly clean up their texture
        resources on destruction.
        https://bugs.webkit.org/show_bug.cgi?id=46139

        * src/WebViewImpl.cpp:
        (WebKit::WebViewImpl::setIsAcceleratedCompositingActive):
        * src/WebViewImpl.h:
2010-09-27  Vangelis Kokkevis  <vangelis@chromium.org>

        Reviewed by James Robinson.

        [chromium] Making LayerChromium's destructor virtual so that the appropriate
        derived class destructors get called when the layer tree is taken down.
        ContentLayerChromium and its derived types (ImageLayerChromium for now) will
        free up the texture allocated for their contents upon destruction. To avoid
        having the LayerRendererChromium destructor getting called before the layers
        are destroyed and thus leaving their textures orphaned, the layers now hold
        a ref counted pointer to the renderer that uses them.
        https://bugs.webkit.org/show_bug.cgi?id=46139

        * platform/graphics/chromium/ContentLayerChromium.cpp:
        (WebCore::ContentLayerChromium::~ContentLayerChromium):
        (WebCore::ContentLayerChromium::setLayerRenderer):
        (WebCore::ContentLayerChromium::cleanupResources):
        * platform/graphics/chromium/ContentLayerChromium.h:
        * platform/graphics/chromium/LayerChromium.cpp:
        (WebCore::LayerChromium::setLayerRenderer):
        * platform/graphics/chromium/LayerChromium.h:
        (WebCore::LayerChromium::layerRenderer):
        * platform/graphics/chromium/LayerRendererChromium.cpp:
        (WebCore::LayerRendererChromium::create):
        (WebCore::LayerRendererChromium::deleteLayerTexture):
        (WebCore::LayerRendererChromium::cleanupSharedObjects):
        * platform/graphics/chromium/LayerRendererChromium.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68442 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Alexey Marinichev <amarinichev@chromium.org>
commit-queue@webkit.org [Mon, 27 Sep 2010 23:28:20 +0000 (23:28 +0000)]
2010-09-27  Alexey Marinichev  <amarinichev@chromium.org>

        Reviewed by David Levin.

        Test that scrolls a checkerboard pattern.  If the hardware compositor
        miscalculates texture coordinates, corruption is very likely to occur.

        * compositing/checkerboard.html: Added.
        * compositing/checkerboard-expected.checksum: Added.
        * compositing/checkerboard-expected.png: Added.
        * compositing/checkerboard-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68441 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Erik Arvidsson <arv@chromium.org>
arv@chromium.org [Mon, 27 Sep 2010 23:12:11 +0000 (23:12 +0000)]
2010-09-27  Erik Arvidsson  <arv@chromium.org>

        Reviewed by Darin Adler.

        Implement HTML 5's HTMLElement.classList property
        https://bugs.webkit.org/show_bug.cgi?id=20709

        * fast/dom/HTMLElement/class-list-expected.txt: Added.
        * fast/dom/HTMLElement/class-list-gc-expected.txt: Added.
        * fast/dom/HTMLElement/class-list-gc.html: Added.
        * fast/dom/HTMLElement/class-list-quirks-expected.txt: Added.
        * fast/dom/HTMLElement/class-list-quirks.html: Added.
        * fast/dom/HTMLElement/class-list.html: Added.
        * fast/dom/HTMLElement/script-tests/class-list-gc.js: Added.
        (gc):
        * fast/dom/HTMLElement/script-tests/class-list.js: Added.
        (createElement):
        * fast/dom/Window/window-properties-expected.txt:
        * fast/dom/Window/window-property-descriptors-expected.txt:
        * fast/dom/prototype-inheritance-2-expected.txt:
        * fast/dom/prototype-inheritance-expected.txt:
        * fast/js/global-constructors-expected.txt:
        * perf/class-list-remove-expected.txt: Added.
        * perf/class-list-remove.html: Added.
2010-09-27  Erik Arvidsson  <arv@chromium.org>

        Reviewed by Darin Adler.

        Implement HTML 5's HTMLElement.classList property
        https://bugs.webkit.org/show_bug.cgi?id=20709

        This adds a DOMTokenList class that is used for the classList property. DOMTokenList uses a SpaceSplitString for fast
        contains check. In standards mode the existing classNames is used but in quirks mode we use an internal SpaceSplitString
        because classList is always case sensitive.

        Tests: fast/dom/HTMLElement/class-list-gc.html
               fast/dom/HTMLElement/class-list-quirks.html
               fast/dom/HTMLElement/class-list.html
               perf/class-list-remove.html

        * Android.derived.jscbindings.mk:
        * Android.derived.v8bindings.mk:
        * Android.jscbindings.mk:
        * CMakeLists.txt:
        * DerivedSources.cpp:
        * DerivedSources.make:
        * GNUmakefile.am:
        * WebCore.gypi:
        * WebCore.pri:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * bindings/gobject/GNUmakefile.am:
        * bindings/js/JSElementCustom.cpp:
        (WebCore::JSElement::markChildren): Make sure that we keep the wrapper while the element is alive.
        * bindings/scripts/CodeGeneratorJS.pm: DOMTokenList has a string indexed getter.
        * bindings/scripts/CodeGeneratorV8.pm:
        * bindings/v8/custom/V8DOMTokenListCustom.cpp: Added.
        (WebCore::toV8):
        * dom/Element.cpp:
        (WebCore::Element::classList):
        (WebCore::Element::optionalClassList):
        * dom/Element.h:
        * dom/ElementRareData.h: This now has an OwnPtr to a ClassList if the Element::classList is ever called.
        * dom/SpaceSplitString.h:
        (WebCore::SpaceSplitString::isNull):
        * dom/StyledElement.cpp:
        (WebCore::StyledElement::classAttributeChanged): Update the classList if it exists.
        * dom/StyledElement.h:
        * html/DOMTokenList.cpp: Added.
        (WebCore::validateToken):
        (WebCore::DOMTokenList::DOMTokenList):
        (WebCore::DOMTokenList::ref):
        (WebCore::DOMTokenList::deref):
        (WebCore::DOMTokenList::length):
        (WebCore::DOMTokenList::item):
        (WebCore::DOMTokenList::contains):
        (WebCore::DOMTokenList::containsInternal): The internal methods do no validation of the token.
        (WebCore::DOMTokenList::add):
        (WebCore::DOMTokenList::addInternal):
        (WebCore::DOMTokenList::remove):
        (WebCore::DOMTokenList::removeInternal):
        (WebCore::DOMTokenList::toggle):
        (WebCore::DOMTokenList::toString):
        (WebCore::DOMTokenList::reset):
        (WebCore::DOMTokenList::classNames):
        * html/DOMTokenList.h: Added.
        (WebCore::DOMTokenList::create):
        (WebCore::DOMTokenList::element):
        * html/DOMTokenList.idl: Added.
        * html/HTMLElement.idl:
        * page/DOMWindow.cpp:
        * page/DOMWindow.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68440 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Kenneth Russell <kbr@google.com>
kbr@google.com [Mon, 27 Sep 2010 23:07:40 +0000 (23:07 +0000)]
2010-09-27  Kenneth Russell  <kbr@google.com>

        Reviewed by James Robinson.

        Add local triangulation of cubic curve control points
        https://bugs.webkit.org/show_bug.cgi?id=45252

        (Re-landing; I neglected to actually add the new files in
        http://trac.webkit.org/changeset/68045 .)

        Adding a localized triangulation algorithm which takes as input
        the four control points of a cubic curve segment and provides both
        triangles as well as the ability to walk the interior edges. The
        latter will be used later to fill the interior of shapes bounded
        by these cubic curves, quadratic curves and line segments.

        * platform/graphics/gpu/LoopBlinnLocalTriangulator.cpp: Added.
        (WebCore::LoopBlinnLocalTriangulator::Triangle::contains):
        (WebCore::LoopBlinnLocalTriangulator::Triangle::nextVertex):
        (WebCore::LoopBlinnLocalTriangulator::Triangle::indexForVertex):
        (WebCore::LoopBlinnLocalTriangulator::Triangle::makeCounterClockwise):
        (WebCore::LoopBlinnLocalTriangulator::LoopBlinnLocalTriangulator):
        (WebCore::LoopBlinnLocalTriangulator::reset):
        (WebCore::LoopBlinnLocalTriangulator::triangulate):
        (WebCore::LoopBlinnLocalTriangulator::triangulateHelper):
        (WebCore::LoopBlinnLocalTriangulator::addTriangle):
        (WebCore::LoopBlinnLocalTriangulator::addInteriorVertex):
        (WebCore::LoopBlinnLocalTriangulator::isSharedEdge):
        * platform/graphics/gpu/LoopBlinnLocalTriangulator.h: Added.
        (WebCore::LoopBlinnLocalTriangulator::Vertex::Vertex):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::xyCoordinates):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::klmCoordinates):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::set):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::end):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::setEnd):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::marked):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::setMarked):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::interior):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::setInterior):
        (WebCore::LoopBlinnLocalTriangulator::Vertex::resetFlags):
        (WebCore::LoopBlinnLocalTriangulator::Triangle::Triangle):
        (WebCore::LoopBlinnLocalTriangulator::Triangle::getVertex):
        (WebCore::LoopBlinnLocalTriangulator::Triangle::setVertices):
        (WebCore::LoopBlinnLocalTriangulator::getVertex):
        (WebCore::LoopBlinnLocalTriangulator::numberOfTriangles):
        (WebCore::LoopBlinnLocalTriangulator::getTriangle):
        (WebCore::LoopBlinnLocalTriangulator::numberOfInteriorVertices):
        (WebCore::LoopBlinnLocalTriangulator::getInteriorVertex):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68439 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Chris Rogers <crogers@google.com>
crogers@google.com [Mon, 27 Sep 2010 23:00:58 +0000 (23:00 +0000)]
2010-09-27  Chris Rogers  <crogers@google.com>

        Reviewed by Kenneth Russell.

        Add AudioNode files
        https://bugs.webkit.org/show_bug.cgi?id=45571

        No new tests since audio API is not yet implemented.

        * webaudio/AudioNode.cpp: Added.
        (WebCore::AudioNode::AudioNode):
        (WebCore::AudioNode::~AudioNode):
        (WebCore::AudioNode::setType):
        (WebCore::AudioNode::lazyInitialize):
        (WebCore::AudioNode::addInput):
        (WebCore::AudioNode::addOutput):
        (WebCore::AudioNode::input):
        (WebCore::AudioNode::output):
        (WebCore::AudioNode::connect):
        (WebCore::AudioNode::disconnect):
        (WebCore::AudioNode::processIfNecessary):
        (WebCore::AudioNode::pullInputs):
        (WebCore::AudioNode::ref):
        (WebCore::AudioNode::deref):
        (WebCore::AudioNode::finishDeref):
        (WebCore::AudioNode::printNodeCounts):
        * webaudio/AudioNode.h: Added.
        (WebCore::AudioNode::context):
        (WebCore::AudioNode::type):
        (WebCore::AudioNode::isInitialized):
        (WebCore::AudioNode::numberOfInputs):
        (WebCore::AudioNode::numberOfOutputs):
        (WebCore::AudioNode::sampleRate):
        (WebCore::AudioNode::checkNumberOfChannelsForInput):
        * webaudio/AudioNode.idl: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68438 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agohttps://bugs.webkit.org/show_bug.cgi?id=46665
hyatt@apple.com [Mon, 27 Sep 2010 22:58:34 +0000 (22:58 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=46665

Reviewed by Sam Weinig.

Patch RenderView's layout and computeLogicalWidth/Height methods to be block-flow-aware.

Added fast/root-lr-basic.html

WebCore:

* rendering/RenderView.cpp:
(WebCore::RenderView::computeLogicalHeight):
(WebCore::RenderView::computeLogicalWidth):
(WebCore::RenderView::layout):
* rendering/RenderView.h:
(WebCore::RenderView::viewLogicalWidth):
(WebCore::RenderView::viewLogicalHeight):

LayoutTests:

* fast/blockflow/root-lr-basic.html: Added.
* platform/mac/fast/blockflow: Added.
* platform/mac/fast/blockflow/root-lr-basic-expected.checksum: Added.
* platform/mac/fast/blockflow/root-lr-basic-expected.png: Added.
* platform/mac/fast/blockflow/root-lr-basic-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68436 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd remaining event handlers to WKView.
weinig@apple.com [Mon, 27 Sep 2010 22:55:14 +0000 (22:55 +0000)]
Add remaining event handlers to WKView.

Reviewed by Anders Carlsson.

Also fixes <rdar://problem/8467058>
Does not get decidePolicyForNavigationAction callback when middle clicking a link

- Adds otherButton and rightButton NSResponder event handlers to WKView.
- Renames WebPageProxy event handlers to use the handle prefix.
- Uses a macro to reduce duplicate code for WKView event handlers.
- Add view parameter to WebEventFactory::createWebKeyboardEvent on the mac
  for uniformity.

* Shared/mac/WebEventFactory.h:
* Shared/mac/WebEventFactory.mm:
(WebKit::WebEventFactory::createWebKeyboardEvent):
* UIProcess/API/mac/WKView.mm:
* UIProcess/API/qt/qwkpage.cpp:
(QWKPagePrivate::keyPressEvent):
(QWKPagePrivate::keyReleaseEvent):
(QWKPagePrivate::mouseMoveEvent):
(QWKPagePrivate::mousePressEvent):
(QWKPagePrivate::mouseReleaseEvent):
(QWKPagePrivate::mouseDoubleClickEvent):
(QWKPagePrivate::wheelEvent):
(QWKPagePrivate::touchEvent):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::handleMouseEvent):
(WebKit::WebPageProxy::handleWheelEvent):
(WebKit::WebPageProxy::handleKeyboardEvent):
(WebKit::WebPageProxy::handleTouchEvent):
* UIProcess/WebPageProxy.h:
* UIProcess/win/WebView.cpp:
(WebKit::WebView::onMouseEvent):
(WebKit::WebView::onWheelEvent):
(WebKit::WebView::onKeyEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68435 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-24 Zhenyao Mo <zmo@google.com>
zmo@google.com [Mon, 27 Sep 2010 22:51:47 +0000 (22:51 +0000)]
2010-09-24  Zhenyao Mo  <zmo@google.com>

        Reviewed by Kenneth Russell.

        Hook up ANGLE with chromium --in-process-webgl port
        https://bugs.webkit.org/show_bug.cgi?id=44309

        * WebKit.gyp: Add angle dependency.
        * src/WebGraphicsContext3DDefaultImpl.cpp: Using angle for shader validation and translation.
        (WebKit::WebGraphicsContext3DDefaultImpl::WebGraphicsContext3DDefaultImpl):
        (WebKit::WebGraphicsContext3DDefaultImpl::~WebGraphicsContext3DDefaultImpl):
        (WebKit::WebGraphicsContext3DDefaultImpl::initialize):
        (WebKit::WebGraphicsContext3DDefaultImpl::compileShader):
        (WebKit::WebGraphicsContext3DDefaultImpl::getIntegerv): Using defined const instead.
        (WebKit::WebGraphicsContext3DDefaultImpl::getShaderiv):
        (WebKit::WebGraphicsContext3DDefaultImpl::getShaderInfoLog):
        (WebKit::WebGraphicsContext3DDefaultImpl::getShaderSource):
        (WebKit::WebGraphicsContext3DDefaultImpl::shaderSource):
        (WebKit::WebGraphicsContext3DDefaultImpl::createShader):
        (WebKit::WebGraphicsContext3DDefaultImpl::deleteShader):
        (WebKit::WebGraphicsContext3DDefaultImpl::angleCreateCompilers):
        (WebKit::WebGraphicsContext3DDefaultImpl::angleDestroyCompilers):
        (WebKit::WebGraphicsContext3DDefaultImpl::angleValidateShaderSource):
        * src/WebGraphicsContext3DDefaultImpl.h: Add necessary data and functions for angle stuff.
        (WebKit::WebGraphicsContext3DDefaultImpl::ShaderSourceEntry::ShaderSourceEntry):
        (WebKit::WebGraphicsContext3DDefaultImpl::ShaderSourceEntry::~ShaderSourceEntry):
2010-09-24  Zhenyao Mo  <zmo@google.com>

        Reviewed by Kenneth Russell.

        Hook up ANGLE with chromium --in-process-webgl port
        https://bugs.webkit.org/show_bug.cgi?id=44309

        * platform/graphics/GraphicsContext3D.h: Move Angle stuff inside MAC only section.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68434 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Mihai Parparita <mihaip@chromium.org>
mihaip@chromium.org [Mon, 27 Sep 2010 22:42:42 +0000 (22:42 +0000)]
2010-09-27  Mihai Parparita  <mihaip@chromium.org>

        Reviewed by Eric Seidel.

        webkitpy.common.system.user_unittest.UserTest.test_prompt_with_list spams the console
        https://bugs.webkit.org/show_bug.cgi?id=46634

        Use OutputCapture to prevent console spam (and also check for expected
        output).

        * Scripts/webkitpy/common/system/user_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68432 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Zhenyao Mo <zmo@google.com>
zmo@google.com [Mon, 27 Sep 2010 22:41:02 +0000 (22:41 +0000)]
2010-09-27  Zhenyao Mo  <zmo@google.com>

        Unreviewed, add fast/canvas/webgl/object-deletion-behaviour.html to test_expectations.txt (failing on Win/Linux).

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68431 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agohttps://bugs.webkit.org/show_bug.cgi?id=46659
hyatt@apple.com [Mon, 27 Sep 2010 22:32:34 +0000 (22:32 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=46659

Reviewed by Sam Weinig.

Patch the width/height/x/y calls in layoutBlock() to be their logical equivalents instead.

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::layoutBlock):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68430 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago<rdar://problem/8348990> / <http://webkit.org/b/46633> Make it possible to disable...
mrowe@apple.com [Mon, 27 Sep 2010 22:10:10 +0000 (22:10 +0000)]
<rdar://problem/8348990> / <webkit.org/b/46633> Make it possible to disable sandboxing of
web process via runtime flag.

Patch by Ivan Krstić <ike@apple.com> on 2010-09-27
Reviewed by Mark Rowe.

Enable sandboxing of the web process only if the DisableSandbox user default is not set for the
web process.

* WebProcess/mac/WebProcessMainMac.mm:
(WebKit::WebProcessMain):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68429 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd support for autogenerating synchronous message handlers
andersca@apple.com [Mon, 27 Sep 2010 22:01:06 +0000 (22:01 +0000)]
Add support for autogenerating synchronous message handlers
https://bugs.webkit.org/show_bug.cgi?id=46654

Reviewed by Adam Roben.

* Scripts/webkit2/messages.py:
(reply_type): Renamed from reply_base_class.
(delayed_reply_type): Renamed from delayed_reply_base_class.
(message_to_struct_declaration): Add a Reply type typedef.
(async_case_statement): Rename from case_statement.
(generate_message_handler): Generate a sync message handler as well.

* Scripts/webkit2/messages_unittest.py:
Update expected results.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68428 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agohttps://bugs.webkit.org/show_bug.cgi?id=46649, fix failing layout tests.
hyatt@apple.com [Mon, 27 Sep 2010 21:56:19 +0000 (21:56 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=46649, fix failing layout tests.

Reviewed by Sam Weinig.

The implementation of setting the before/after margins was wrong and poking the wrong margin values.

Once I made them set the correct values, it revealed that one of the new block flow tests wasn't
actually working.  In order to fix it, I had to make the block direction margin computation actually
use the containing block's block-flow in order to return the right answer.  This involved cleaning
up computeBlockDirectionMargins to be more like computeInlineDirectionMargins.

fast/css/logical-property-resolution.html also caught the bug.  Its results are now correct.

WebCore:

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::adjustPositionedBlock):
(WebCore::RenderBlock::determineHorizontalPosition):
(WebCore::RenderBlock::layoutBlockChild):
(WebCore::RenderBlock::insertFloatingObject):
* rendering/RenderBox.cpp:
(WebCore::RenderBox::setMarginBeforeUsing):
(WebCore::RenderBox::setMarginAfterUsing):
(WebCore::RenderBox::computeLogicalWidth):
(WebCore::RenderBox::computeInlineDirectionMargins):
(WebCore::RenderBox::computeLogicalHeight):
(WebCore::RenderBox::computeBlockDirectionMargins):
* rendering/RenderBox.h:
* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::layoutHorizontalBox):
(WebCore::RenderFlexibleBox::layoutVerticalBox):
* rendering/RenderObject.h:
* rendering/RenderTable.cpp:
(WebCore::RenderTable::computeLogicalWidth):
* rendering/RenderTableRow.cpp:
(WebCore::RenderTableRow::layout):
* rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::marginBeforeUsing):
(WebCore::RenderStyle::marginAfterUsing):
* rendering/style/RenderStyle.h:

LayoutTests:

* fast/css/logical-property-resolution-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68427 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoFix thinko.
andersca@apple.com [Mon, 27 Sep 2010 21:54:59 +0000 (21:54 +0000)]
Fix thinko.

* runtime/JSCell.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68426 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoTry to fix Windows build.
andersca@apple.com [Mon, 27 Sep 2010 21:48:35 +0000 (21:48 +0000)]
Try to fix Windows build.

Reviewed by Adam Roben.

* runtime/JSCell.h:
(JSC::MSVCBugWorkaround::MSVCBugWorkaround):
(JSC::MSVCBugWorkaround::~MSVCBugWorkaround):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68425 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-14 Zhenyao Mo <zmo@google.com>
zmo@google.com [Mon, 27 Sep 2010 21:27:54 +0000 (21:27 +0000)]
2010-09-14  Zhenyao Mo  <zmo@google.com>

        Reviewed by Kenneth Russell.

        Textures and renderbuffers should be detached first before deletion if they are attached to framebuffers
        https://bugs.webkit.org/show_bug.cgi?id=43942

        Test: fast/canvas/webgl/object-deletion-behaviour.html

        * html/canvas/WebGLBuffer.cpp:
        (WebCore::WebGLBuffer::deleteObjectImpl): Delete the object only once.
        * html/canvas/WebGLFramebuffer.cpp: Use RefPtr for attachment WebGLObjects.
        (WebCore::WebGLFramebuffer::WebGLFramebuffer):
        (WebCore::WebGLFramebuffer::removeAttachment): Remove an attachment data ref.
        (WebCore::WebGLFramebuffer::onAttachedObjectChange):
        (WebCore::WebGLFramebuffer::getColorBufferFormat):
        (WebCore::WebGLFramebuffer::deleteObjectImpl):
        (WebCore::WebGLFramebuffer::initializeRenderbuffers):
        * html/canvas/WebGLFramebuffer.h: Declare removeAttachment().
        * html/canvas/WebGLObject.cpp:
        (WebCore::WebGLObject::setObject): Make it protected and can only be called when initializing.
        * html/canvas/WebGLObject.h:
        (WebCore::WebGLObject::onDetached): Fix a bug.
        (WebCore::WebGLObject::isDeleted): Expose the delete flag.
        * html/canvas/WebGLProgram.cpp:
        (WebCore::WebGLProgram::deleteObjectImpl): Detach shaders when deleting a program.
        * html/canvas/WebGLRenderbuffer.cpp:
        (WebCore::WebGLRenderbuffer::deleteObjectImpl): Delete the object only once.
        * html/canvas/WebGLRenderingContext.cpp:
        (WebCore::WebGLRenderingContext::deleteFramebuffer): Bind to internal fbo when deleting the currently-bound fbo.
        (WebCore::WebGLRenderingContext::deleteRenderbuffer): Remove renderbuffer from currently-bound fbo upon deletion.
        (WebCore::WebGLRenderingContext::deleteTexture): Remove texture from currently-bound fbo upon deletion.
        (WebCore::WebGLRenderingContext::useProgram): Fix a condition.
        * html/canvas/WebGLShader.cpp:
        (WebCore::WebGLShader::deleteObjectImpl): Delete the object only once.
        * html/canvas/WebGLTexture.cpp:
        (WebCore::WebGLTexture::deleteObjectImpl): Ditto.
2010-09-14  Zhenyao Mo  <zmo@google.com>

        Reviewed by Kenneth Russell.

        Textures and renderbuffers should be detached first before deletion if they are attached to framebuffers
        https://bugs.webkit.org/show_bug.cgi?id=43942

        * fast/canvas/webgl/object-deletion-behaviour-expected.txt: Added.
        * fast/canvas/webgl/object-deletion-behaviour.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68424 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Ryosuke Niwa <rniwa@webkit.org>
rniwa@webkit.org [Mon, 27 Sep 2010 21:10:54 +0000 (21:10 +0000)]
2010-09-27  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Antonio Gomes.

        stateStyle (@EditorCommand.cpp) should ask EditingBehavior for platform specific behavior
        https://bugs.webkit.org/show_bug.cgi?id=41989

        queryCommandState returns true even when in the mixed state
        https://bugs.webkit.org/show_bug.cgi?id=46382

        Fixed stateStyle so that it calls selectionStartHasStyle when shouldToggleStyleBasedOnStartOfSelection
        is true (Mac platforms) and calls selectionHasStyle otherwise (non-Mac platforms).

        Also fixed queryCommandState so that it only returns true when the state is TrueTriState
        as supposed to returning true whenever the state is not FalseTriState. New behavior matches that of
        other browsers such as Firefox and Internet Explorer.

        * dom/Document.cpp:
        (WebCore::Document::queryCommandState):
        * editing/EditorCommand.cpp:
        (WebCore::executeToggleStyle):
        (WebCore::stateStyle):
2010-09-27  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Antonio Gomes.

        stateStyle (@EditorCommand.cpp) should ask EditingBehavior for platform specific behavior
        https://bugs.webkit.org/show_bug.cgi?id=41989

        Added a test to ensure queryCommandState returns correct values on Mac and Windows.

        * editing/execCommand/query-command-state-expected.txt: Added.
        * editing/execCommand/script-tests/query-command-state.js: Added.
        (testQueryCommandState):
        (selectAll):
        (selectSecondWord):
        (selectFirstTwoWords):
        (runTests):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68423 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-09-27 Erik Arvidsson <arv@chromium.org>
arv@chromium.org [Mon, 27 Sep 2010 20:43:48 +0000 (20:43 +0000)]
2010-09-27  Erik Arvidsson  <arv@chromium.org>

        Reviewed by Darin Adler.

        Add operator == for AtomicString and Vector<Uchar>
        https://bugs.webkit.org/show_bug.cgi?id=46509

        * JavaScriptCore.exp:
        * wtf/text/AtomicString.cpp:
        (WTF::operator==):
        * wtf/text/AtomicString.h:
        (WTF::operator==):
        (WTF::operator!=):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@68422 268f45cc-cd09-0410-ab3c-d52691b4dbfc