Include the crashing function in the link to a crash log on Mac

Fixes <http://webkit.org/b/53780> Crash log links in results.html should include the
function that crashed on Mac

Reviewed by David Kilzer.

* Scripts/old-run-webkit-tests:
(crashLocation): Moved all the Windows code inside an if instead of using an early return,
but didn't otherwise change it. Added an if for Mac that parses the crashing function out of
the crash log.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use static_cast and other style cleanup in YarrInterpreter.cpp
<http://webkit.org/b/53772>

Reviewed by John Sullivan.

* yarr/YarrInterpreter.cpp:
(JSC::Yarr::Interpreter::InputStream::readChecked): Use
static_cast.
(JSC::Yarr::Interpreter::InputStream::checkInput): Remove
unnecessary else block.
(JSC::Yarr::Interpreter::matchAssertionEOL): Ditto.
(JSC::Yarr::Interpreter::backtrackBackReference): Ditto.
(JSC::Yarr::ByteCompiler::emitDisjunction): Use static_cast.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Remove more passing tests.

        * platform/chromium/test_expectations.txt: Even more passages removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Link to Mac crash logs from results.html

Fixes <http://webkit.org/b/14861> run-webkit-tests should link to Mac crash logs in
results.html

Reviewed by David Kilzer.

* Scripts/old-run-webkit-tests:
(captureSavedCrashLog): Added a case for Apple's Mac port. We get the crash log out of
~/Library/Logs/CrashReporter, and wait for ReportCrash to exit before trying to get the log.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Removed passing tests.

        * platform/chromium/test_expectations.txt: Passages removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Pavel Podivilov  <podivilov@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: evaluate on hover does not work on a breakpoint.
        https://bugs.webkit.org/show_bug.cgi?id=53768

        * inspector/front-end/SourceFrame.js:
        (WebInspector.SourceFrame.prototype._mouseHover):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r77625 and r77626.
        http://trac.webkit.org/changeset/77625
        http://trac.webkit.org/changeset/77626
        https://bugs.webkit.org/show_bug.cgi?id=53765

        It broke Windows builds (Requested by Ossy_ on #webkit).

        * Android.jscbindings.mk:
        * CMakeLists.txt:
        * ForwardingHeaders/pcre/pcre.h: Added.
        * ForwardingHeaders/yarr/Yarr.h: Removed.
        * ForwardingHeaders/yarr/YarrInterpreter.h: Removed.
        * ForwardingHeaders/yarr/YarrPattern.h: Removed.
        * WebCore.gyp/WebCore.gyp:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.vcproj/copyForwardingHeaders.cmd:
        * platform/text/RegularExpression.cpp:
        (WebCore::RegularExpression::Private::regexp):
        (WebCore::RegularExpression::Private::compile):
        (WebCore::RegularExpression::Private::Private):
        (WebCore::RegularExpression::Private::create):
        (WebCore::RegularExpression::Private::~Private):
        (WebCore::RegularExpression::match):

2011-02-04  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r77625 and r77626.
        http://trac.webkit.org/changeset/77625
        http://trac.webkit.org/changeset/77626
        https://bugs.webkit.org/show_bug.cgi?id=53765

        It broke Windows builds (Requested by Ossy_ on #webkit).

        * JavaScriptCore.exp:
        * JavaScriptCore.gyp/JavaScriptCore.gyp:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * create_regex_tables:
        * runtime/RegExp.cpp:
        * wtf/Platform.h:
        * yarr/Yarr.h:
        * yarr/YarrJIT.cpp:
        * yarr/YarrJIT.h:
        * yarr/YarrParser.h:
        * yarr/YarrPattern.h:
        * yarr/YarrSyntaxChecker.h:
        * yarr/yarr.pri: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Windows build fix. Unreviewed.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGenerated.make:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Windows 7 Release Tests] transitions/transition-end-event-multiple-04.html flakey
https://bugs.webkit.org/show_bug.cgi?id=53332

Unreviewed, attempt to get the Windows 7 Release bots uniformly green.

* platform/win/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Peter Varga  <pvarga@webkit.org>

        Rubber-stamped by Csaba Osztrogonác.

        Replace PCRE with Yarr in WebCore
        https://bugs.webkit.org/show_bug.cgi?id=53496

        Speculative windows build fix.

        No new tests needed.

        * platform/text/RegularExpression.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Peter Varga  <pvarga@webkit.org>

        Reviewed by Gavin Barraclough.

        Replace PCRE with Yarr in WebCore
        https://bugs.webkit.org/show_bug.cgi?id=53496

        * JavaScriptCore.exp:
        * JavaScriptCore.gyp/JavaScriptCore.gyp:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * create_regex_tables:
        * runtime/RegExp.cpp:
        * wtf/Platform.h:
        * yarr/Yarr.h:
        * yarr/YarrJIT.cpp:
        * yarr/YarrJIT.h:
        * yarr/YarrParser.h:
        * yarr/YarrPattern.h:
        * yarr/YarrSyntaxChecker.h:
        * yarr/yarr.pri: Added.
2011-02-04  Peter Varga  <pvarga@webkit.org>

        Reviewed by Gavin Barraclough.

        Replace PCRE with Yarr in WebCore
        https://bugs.webkit.org/show_bug.cgi?id=53496

        No new tests needed.

        * Android.jscbindings.mk:
        * CMakeLists.txt:
        * ForwardingHeaders/pcre/pcre.h: Removed.
        * ForwardingHeaders/yarr/Yarr.h: Added.
        * ForwardingHeaders/yarr/YarrInterpreter.h: Added.
        * ForwardingHeaders/yarr/YarrPattern.h: Added.
        * WebCore.gyp/WebCore.gyp:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.vcproj/copyForwardingHeaders.cmd:
        * platform/text/RegularExpression.cpp:
        (WebCore::RegularExpression::Private::create):
        (WebCore::RegularExpression::Private::Private):
        (WebCore::RegularExpression::Private::compile):
        (WebCore::RegularExpression::match):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Ilya Tikhonovsky  <loislo@chromium.org>

        Unreviewed.

        Rebaseline. The problem was introduced at r77601.

        * platform/chromium-mac/fast/repaint/shadow-multiple-horizontal-expected.checksum: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-horizontal-expected.png: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-strict-horizontal-expected.checksum: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-strict-horizontal-expected.png: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-strict-vertical-expected.checksum: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-strict-vertical-expected.png: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-vertical-expected.checksum: Added.
        * platform/chromium-mac/fast/repaint/shadow-multiple-vertical-expected.png: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Yury Semikhatsky.

        Web Inspector: Network panel filtering is broken.
        https://bugs.webkit.org/show_bug.cgi?id=53764

        * inspector/front-end/NetworkPanel.js:
        (WebInspector.NetworkPanel.prototype._sortItems):
        (WebInspector.NetworkPanel.prototype._sortByTimeline):
        (WebInspector.NetworkPanel.prototype._filter):
        (WebInspector.NetworkPanel.prototype._updateOffscreenRows):
        (WebInspector.NetworkDataGridNode.prototype.isFilteredOut):
        (WebInspector.NetworkDataGridNode.prototype.get selectable):
        (WebInspector.NetworkTotalGridNode.prototype.isFilteredOut):
        (WebInspector.NetworkTotalGridNode.prototype.get selectable):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Ilya Tikhonovsky  <loislo@chromium.org>

        Unreviewed.

        Rebaseline image for fast/box-shadow/shadow-tiling-artifact.html
        The problem was introduced at r77601.

        * platform/chromium-linux/fast/box-shadow/shadow-tiling-artifact-expected.checksum: Added.
        * platform/chromium-linux/fast/box-shadow/shadow-tiling-artifact-expected.png: Added.
        * platform/chromium-win/fast/box-shadow/shadow-tiling-artifact-expected.checksum: Added.
        * platform/chromium-win/fast/box-shadow/shadow-tiling-artifact-expected.png: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Sergio Villar Senin  <svillar@igalia.com>

        Unreviewed.

        New test expectations.

        * platform/gtk/fast/box-shadow/shadow-tiling-artifact-expected.txt: Added.
        * platform/gtk/fast/dom/HTMLInputElement/input-slider-update-expected.txt: Added.
        * platform/gtk/svg/text/font-size-below-point-five-2-expected.txt: Added.
        * platform/gtk/svg/text/font-size-below-point-five-expected.txt: Added.
        * platform/gtk/svg/text/scaled-font-expected.txt: Added.
        * platform/gtk/svg/text/small-fonts-2-expected.txt: Added.
        * platform/gtk/svg/text/small-fonts-3-expected.txt: Added.
        * platform/gtk/svg/text/small-fonts-expected.txt: Added.
        * platform/gtk/svg/text/small-fonts-in-html5-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Andrey Kosyakov  <caseq@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: support overriding user agent strings
        https://bugs.webkit.org/show_bug.cgi?id=51485

        Test: http/tests/inspector/extensions-useragent.html

        * inspector/Inspector.idl:
        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::disconnectFrontend):
        (WebCore::InspectorController::setUserAgentOverride):
        (WebCore::InspectorController::userAgentOverride):
        * inspector/InspectorController.h:
        * inspector/front-end/ExtensionAPI.js:
        (WebInspector.injectedExtensionAPI.InspectedWindow.prototype.reload):
        * inspector/front-end/ExtensionServer.js:
        (WebInspector.ExtensionServer.prototype._onReload):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::userAgent):
        (WebCore::FrameLoader::applyUserAgent):

2011-02-04  Andrey Kosyakov  <caseq@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: support overriding user agent strings
        https://bugs.webkit.org/show_bug.cgi?id=51485

        * http/tests/inspector/extensions-useragent-expected.txt: Added.
        * http/tests/inspector/extensions-useragent.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Ilya Tikhonovsky  <loislo@chromium.org>

        Unreviewed rollout two patches r77614 and r77612.

        REGRESSION: Snow Leopard Intell Release anumber of failing tests.

        * runtime/Heap.cpp:
        (JSC::Heap::Heap):
        (JSC::Heap::reset):
        * runtime/Heap.h:
        * runtime/MarkedSpace.cpp:
        (JSC::MarkedSpace::allocate):
        (JSC::MarkedSpace::sweep):
        (JSC::MarkedSpace::reset):
        * runtime/MarkedSpace.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Pavel Podivilov  <podivilov@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: scripts panel displays wrong file name after reload.
        https://bugs.webkit.org/show_bug.cgi?id=53761

        * inspector/front-end/ScriptsPanel.js:
        (WebInspector.ScriptsPanel.prototype._resourceLoadingFinished):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebCore.exp.in:
        * inspector/CodeGeneratorInspector.pm:
        * inspector/InspectorClient.h:
        * inspector/InspectorFrontendClientLocal.cpp:
        (WebCore::InspectorFrontendClientLocal::InspectorFrontendClientLocal):
        (WebCore::InspectorFrontendClientLocal::changeAttachedWindowHeight):
        (WebCore::InspectorFrontendClientLocal::restoreAttachedWindowHeight):
        * inspector/InspectorFrontendClientLocal.h: ports that provide in-process implementation of the inspector front-end can
        provide platform-specific settings accessor.
        (WebCore::InspectorFrontendClientLocal::Settings::Settings):
        (WebCore::InspectorFrontendClientLocal::Settings::~Settings):
        (WebCore::InspectorFrontendClientLocal::Settings::inspectorAttachedHeight):
        (WebCore::InspectorFrontendClientLocal::Settings::storeInspectorAttachedHeight):
        * loader/EmptyClients.h:
2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebCoreSupport/WebInspectorClientCF.cpp:
        (populateSetting):
        (storeSetting):
        (WebInspectorClient::createFrontendSettings):
2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebCoreSupport/InspectorClientGtk.cpp:
        (WebKit::InspectorFrontendClient::InspectorFrontendClient):
        * WebCoreSupport/InspectorClientGtk.h:
2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebCoreSupport/WebInspectorClient.h:
        * WebCoreSupport/WebInspectorClient.mm:
        (WebInspectorClient::openInspectorFrontend):
        (WebInspectorFrontendClient::WebInspectorFrontendClient):
2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebCoreSupport/InspectorClientQt.cpp:
        (WebCore::InspectorFrontendClientQt::InspectorFrontendClientQt):
        * WebCoreSupport/InspectorClientQt.h:
2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebCoreSupport/WebInspectorClient.cpp:
        (WebInspectorClient::openInspectorFrontend):
        (WebInspectorFrontendClient::WebInspectorFrontendClient):
        * WebCoreSupport/WebInspectorClient.h:
2011-02-03  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: remove settings related methods from InspectorClient
        https://bugs.webkit.org/show_bug.cgi?id=53686

        * WebProcess/WebCoreSupport/WebInspectorFrontendClient.cpp:
        (WebKit::WebInspectorFrontendClient::WebInspectorFrontendClient):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Sergio Villar Senin  <svillar@igalia.com>

        Unreviewed.

        Skipping fast/events/pagehide-timeout.html that is failing on GTK
        bots. See https://bugs.webkit.org/show_bug.cgi?id=53733

        * platform/gtk/Skipped: skip fast/events/pagehide-timeout.html

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Ilya Tikhonovsky  <loislo@chromium.org>

        Unreviewed.

        Roll chromium to 73764.

        * DEPS:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Try to fix 32bit build.

* runtime/Heap.cpp:
(JSC::Heap::reset): Use an explicit cast to avoid shortening warnings,
since 1.5 is double (64bit), and the result is size_t (32bit).

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anton Muhin  <antonm@chromium.org>

        Reviewed by Adam Barth.

        [v8] frame several more JS code invocations into v8::TryCatch
        https://bugs.webkit.org/show_bug.cgi?id=53594

        This patch is preemptive and adjusts v8 bindings code to forthcoming small change
        in v8::ThrowException---currently sometimes exceptions thrown by this method
        do not reach surrounding v8::TryCatch handler (see
        http://code.google.com/p/v8/issues/detail?id=1072 and
        http://codereview.chromium.org/6397011/).  Therefore the goal of this patch
        is to make forthcoming v8 roll as smooth as possible (alas, we'll still need
        one rebaseline as of now.)

        * bindings/v8/V8Proxy.cpp:
        (WebCore::V8Proxy::runScript): Do not rely on empty handle as a signal of exception, wrap into v8::TryCatch instead
        * bindings/v8/V8WindowErrorHandler.cpp:
        (WebCore::V8WindowErrorHandler::callListenerFunction): Ditto

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Cameron Zwarich.

        Changed MarkedSpace to delegate grow/shrink decisions to Heap
        https://bugs.webkit.org/show_bug.cgi?id=53759

        SunSpider reports no change.

        * runtime/Heap.cpp:
        (JSC::Heap::Heap):
        (JSC::Heap::reset):
        * runtime/Heap.h: Reorganized a few data members for better cache locality.
        Added a grow policy.

        * runtime/MarkedSpace.cpp:
        (JSC::MarkedSpace::allocate):
        (JSC::MarkedSpace::sweep):
        (JSC::MarkedSpace::reset): Don't shrink automatically. Instead, wait for
        the heap to make an explicit sweep call.

        * runtime/MarkedSpace.h:
        (JSC::MarkedSpace::highWaterMark):
        (JSC::MarkedSpace::setHighWaterMark): Use a watermark to determine how
        many bytes to allocate before failing and giving the heap an opportunity
        to collect garbage. This also means that we allocate blocks on demand,
        instead of ahead of time.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        WebKit2: Need WebKit2 equivalent of WebResourceLoadDelegate::willSendRequest in the Bundle
        https://bugs.webkit.org/show_bug.cgi?id=52897
        <rdar://problem/8898294>

        * WebCore.exp.in: Add export now needed by WebKit2
2011-02-03  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        WebKit2: Need WebKit2 equivalent of WebResourceLoadDelegate::willSendRequest in the Bundle
        https://bugs.webkit.org/show_bug.cgi?id=52897
        <rdar://problem/8898294>

        * Shared/API/c/WKSharedAPICast.h:
        (WebKit::toAPI):
        * WebProcess/InjectedBundle/API/c/WKBundlePage.h:
        * WebProcess/InjectedBundle/InjectedBundlePageLoaderClient.cpp:
        (WebKit::InjectedBundlePageLoaderClient::willSendRequestForFrame):
        * WebProcess/InjectedBundle/InjectedBundlePageLoaderClient.h:
        * WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
        (WebKit::WebFrameLoaderClient::dispatchWillSendRequest):
2011-02-03  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        WebKit2: Need WebKit2 equivalent of WebResourceLoadDelegate::willSendRequest in the Bundle
        https://bugs.webkit.org/show_bug.cgi?id=52897
        <rdar://problem/8898294>

        * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
        (WTR::InjectedBundlePage::InjectedBundlePage):
        (WTR::InjectedBundlePage::willSendRequestForFrame):
        * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Ivan Krstić  <ike@apple.com>

        Reviewed by Maciej Stachowiak.

        Make WebProcess pass explicit homedir parameter to the sandbox
        <rdar://problem/8405760>
        <https://webkit.org/b/53558>

        * WebProcess/com.apple.WebProcess.sb:
        * WebProcess/mac/WebProcessMac.mm:
        (WebKit::initializeSandbox):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Ivan Krstić  <ike@apple.com>

        Unreviewed.

        Adding myself to committers.py.

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

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Victoria Kirst  <vrk@google.com>

        Reviewed by James Robinson.

        Replaces float literals with uniform values in shader code
        so that buggy drivers unable to parse float values in different
        locales will not produce a pink video.

        [chromium] Fix pink video bug with gpu-acceleration enabled
        https://bugs.webkit.org/show_bug.cgi?id=53568

        * platform/graphics/chromium/VideoLayerChromium.cpp:
        (WebCore::VideoLayerChromium::SharedValues::SharedValues):
        (WebCore::VideoLayerChromium::drawYUV):
        * platform/graphics/chromium/VideoLayerChromium.h:
        (WebCore::VideoLayerChromium::SharedValues::signAdjLocation):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add navigator.registerProtocolHandler behind a flag.
https://bugs.webkit.org/show_bug.cgi?id=52609

Patch by James Kozianski <koz@chromium.org> on 2011-02-03
Reviewed by Dimitri Glazkov.

Source/JavaScriptCore:

* Configurations/FeatureDefines.xcconfig:

Source/WebCore:

This method is described in the HTML5 specification here,
http://dev.w3.org/html5/spec/Overview.html#dom-navigator-registerprotocolhandler

This change is largely cribbed from B. Green's 29651 patches. It is
behind a flag so as not to break JS feature detection.

New layout test fast/dom/registerProtocolHandler.html.

* Configurations/FeatureDefines.xcconfig:
* loader/EmptyClients.h:
(WebCore::EmptyChromeClient::registerProtocolHandler):
* page/Chrome.cpp:
(WebCore::Chrome::registerProtocolHandler):
* page/Chrome.h:
* page/ChromeClient.h:
* page/Navigator.cpp:
(WebCore::verifyCustomHandlerURL):
(WebCore::verifyProtocolHandlerScheme):
(WebCore::Navigator::registerProtocolHandler):
* page/Navigator.h:
* page/Navigator.idl:

Source/WebKit/chromium:

* features.gypi:
* public/WebViewClient.h:
* src/ChromeClientImpl.cpp:
(WebKit::ChromeClientImpl::registerProtocolHandler):
* src/ChromeClientImpl.h:

Source/WebKit/mac:

* Configurations/FeatureDefines.xcconfig:

Source/WebKit2:

* Configurations/FeatureDefines.xcconfig:

Tools:

* Scripts/build-webkit:

WebKitLibraries:

* win/tools/vsprops/FeatureDefines.vsprops:
* win/tools/vsprops/FeatureDefinesCairo.vsprops:

LayoutTests:

* fast/dom/register-protocol-handler-expected.txt: Added.
* fast/dom/register-protocol-handler.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Hayato Ito  <hayato@chromium.org>

        Reviewed by Eric Seidel.

        [NRWT] Remove TestArgs class, which is wrongly used.

        https://bugs.webkit.org/show_bug.cgi?id=53063

        * Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
        * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
        * Scripts/webkitpy/layout_tests/test_types/image_diff.py:
        * Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
        * Scripts/webkitpy/layout_tests/test_types/test_type_base_unittest.py:
        * Scripts/webkitpy/layout_tests/test_types/text_diff.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Andrew Wason  <rectalogic@rectalogic.com>

        Reviewed by Kenneth Russell.

        Change ENABLE_3D_CANVAS to ENABLE_WEBGL
        https://bugs.webkit.org/show_bug.cgi?id=53714

        * Scripts/build-webkit: change --3d-canvas build option
          to set ENABLE_WEBGL

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Brian Ryner  <bryner@chromium.org>

        Reviewed by Darin Fisher.

        Add a field to the ResourceResponse for tracking the socket address
        of the host that the resource was fetched from.  Patch was originally
        by Paul Marks.
        https://bugs.webkit.org/show_bug.cgi?id=53699

        * platform/network/chromium/ResourceResponse.cpp:
        (WebCore::ResourceResponse::doPlatformCopyData):
        (WebCore::ResourceResponse::doPlatformAdopt):
        * platform/network/chromium/ResourceResponse.h:
        (WebCore::ResourceResponse::socketAddress):
        (WebCore::ResourceResponse::setSocketAddress):
2011-02-03  Brian Ryner  <bryner@chromium.org>

        Reviewed by Darin Fisher.

        Add a field to the ResourceResponse for tracking the socket address
        of the host that the resource was fetched from.  Patch was originally
        by Paul Marks.
        https://bugs.webkit.org/show_bug.cgi?id=53699

        * public/WebURLResponse.h:
        * src/WebURLResponse.cpp:
        (WebKit::WebURLResponse::socketAddress):
        (WebKit::WebURLResponse::setSocketAddress):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Hayato Ito  <hayato@chromium.org>

        Reviewed by Tony Chang.

        [NRWT] Introduces Input/Output class used by base.Driver into port/base.py and
        move _run_single_test() and _process_output() functions from
        dump_render_tree_thread.py to a single_test_runner.py as an individual module.

        This is clean up and acts as a preparation for elimination of test_types/*
        classes. These classes will move to the single_test_runner module introduced
        in this patch.

        https://bugs.webkit.org/show_bug.cgi?id=53004

        * Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
        * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Added.
        * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
        * Scripts/webkitpy/layout_tests/layout_package/test_output.py: Removed.
        * Scripts/webkitpy/layout_tests/port/base.py:
        * Scripts/webkitpy/layout_tests/port/chromium.py:
        * Scripts/webkitpy/layout_tests/port/dryrun.py:
        * Scripts/webkitpy/layout_tests/port/test.py:
        * Scripts/webkitpy/layout_tests/port/webkit.py:
        * Scripts/webkitpy/layout_tests/test_types/image_diff.py:
        * Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
        * Scripts/webkitpy/layout_tests/test_types/text_diff.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * loader/EmptyClients.h:
        (WebCore::EmptyFrameLoaderClient::didRunInsecureContent):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::checkIfRunInsecureContent):
        * loader/FrameLoaderClient.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * public/WebFrameClient.h:
        (WebKit::WebFrameClient::didRunInsecureContent):
        * src/FrameLoaderClientImpl.cpp:
        (WebKit::FrameLoaderClientImpl::didRunInsecureContent):
        * src/FrameLoaderClientImpl.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebCoreSupport/FrameLoaderClientEfl.cpp:
        (WebCore::FrameLoaderClientEfl::didRunInsecureContent):
        * WebCoreSupport/FrameLoaderClientEfl.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::didRunInsecureContent):
        * WebCoreSupport/FrameLoaderClientGtk.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebCoreSupport/WebFrameLoaderClient.h:
        * WebCoreSupport/WebFrameLoaderClient.mm:
        (WebFrameLoaderClient::didRunInsecureContent):
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebCoreSupport/FrameLoaderClientQt.cpp:
        (WebCore::FrameLoaderClientQt::didRunInsecureContent):
        * WebCoreSupport/FrameLoaderClientQt.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebCoreSupport/WebFrameLoaderClient.cpp:
        (WebFrameLoaderClient::didRunInsecureContent):
        * WebCoreSupport/WebFrameLoaderClient.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebCoreSupport/FrameLoaderClientWinCE.cpp:
        (WebKit::FrameLoaderClientWinCE::didRunInsecureContent):
        * WebCoreSupport/FrameLoaderClientWinCE.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebKitSupport/FrameLoaderClientWx.cpp:
        (WebCore::FrameLoaderClientWx::didRunInsecureContent):
        * WebKitSupport/FrameLoaderClientWx.h:
2011-02-03  Adam Langley  <agl@chromium.org>

        Reviewed by Adam Barth.

        Plumb mixed script URL to FrameLoaderClient
        https://bugs.webkit.org/show_bug.cgi?id=52384

        Regressions covered by http/tests/security/mixedContent/*

        * WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
        (WebKit::WebFrameLoaderClient::didRunInsecureContent):
        * WebProcess/WebCoreSupport/WebFrameLoaderClient.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        REGRESSION: Artifacts on box-shadow corners in some cases
        https://bugs.webkit.org/show_bug.cgi?id=53731

        Fix overdrawing artifacts in ShadowBlur's tiling code path,
        which show up in shadows using a color with alpha.

        Test: fast/box-shadow/shadow-tiling-artifact.html

        * platform/graphics/ShadowBlur.cpp:
        (WebCore::ShadowBlur::drawRectShadowWithTiling): Ensure
        that the inner rect that gets filled does not overlap with any
        of the eight tiled areas by having the corner and side dimensions
        be the same for contiguous areas.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Alexey Proskuryakov.

        XSS Auditor is spinning inside decodeURLEscapeSequences() if there are
        percent signs in large posted data
        https://bugs.webkit.org/show_bug.cgi?id=53405

        If the input string contains many non-% characters followed by a %
        character that is not a valid URL escape sequence, then the old
        algorithm would only advance the initial search by one character
        (instead of jumping to just after the % character).  That would cause
        the algorithm to take N^2 time (in the number of characters before the
        first % character).  This patch just advances the search past the first
        % character so we can start looking for next % character sooner.

        * platform/KURL.cpp:
        (WebCore::decodeURLEscapeSequences):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Victoria Kirst  <vrk@google.com>

        Reviewed by James Robinson.

        [chromium] Fix redundant video frame paint on CSS LayerChromium for <video>
        https://bugs.webkit.org/show_bug.cgi?id=52868

        * src/WebMediaPlayerClientImpl.cpp:
        (WebKit::WebMediaPlayerClientImpl::paint):
        (WebKit::WebMediaPlayerClientImpl::paintCurrentFrameInContext):
        (WebKit::WebMediaPlayerClientImpl::acceleratedRenderingInUse):
        * src/WebMediaPlayerClientImpl.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Pavel Podivilov  <podivilov@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: click on a breakpoint highlights wrong line in source frame.
        https://bugs.webkit.org/show_bug.cgi?id=53692

        * inspector/front-end/BreakpointsSidebarPane.js:
        (WebInspector.JavaScriptBreakpointsSidebarPane.prototype._setupBreakpointElement):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anton Muhin  <antonm@chromium.org>

        Reviewed by Adam Barth.

        [v8] Bail out if to string conversion returned empty handle
        https://bugs.webkit.org/show_bug.cgi?id=53687

        This a temporary measure: actually one probably should never get empty handle
        if there was no exception.  The root cause is under investigation.
        The bailout though allows Chromium not to crash---attempt to convert an empty
        v8 hande into WebCore string crashes with invalid memory access.

        See http://code.google.com/p/chromium/issues/detail?id=71544

        There is no known reduction expressible as a layout test so far.  The crash found with automated testing tools.

        * bindings/v8/V8Binding.cpp:
        (WebCore::v8NonStringValueToWebCoreString): Bail out on empty handle
        * bindings/v8/V8Binding.h:
        (WebCore::V8ParameterBase::prepareBase): Ditto

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Joone Hur  <joone.hur@collabora.co.uk>

        Reviewed by Martin Robinson.

        [Gtk] No need to set text encoding in the provisional phase
        https://bugs.webkit.org/show_bug.cgi?id=53487

        According to changeset 67253, setEncoding could be called multiple times from
        committedLoad, finishedLoading, dispatchDidFailLoading, and setMainDocumentError
        in FrameLoaderClient. To fix this, the relevant code was removed from
        FrameLoaderClient and moved to DocumentLoader::commitData. However, that
        code was not removed from FrameLoaderClient::finishedLoading in WebKitGtk+.

        Due to this reason, after loading a html document, other ports initialize the
        text encoding from FrameLoaderClient::finishedLoading, but WebKitGtk+ sets
        the same encoding again, even tries to set encoding in the provisional phase.
        This causes unnecessary encoding setting.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::FrameLoaderClient): Set m_hasRepresentation to false.
        (WebKit::FrameLoaderClient::makeRepresentation): Set m_hasRepresentation to true.
        (WebKit::FrameLoaderClient::revertToProvisionalState): Set m_hasRepresentation to true.
        (WebKit::FrameLoaderClient::finishedLoading): Skip the encoding setting when
        m_hasRepresentation is false.
        * WebCoreSupport/FrameLoaderClientGtk.h: Added m_hasRepresentation.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Unreviewed, build fix.

        Delete a line that was accidentally missed in r77586; without it
        there is still a race between the two threads.

        https://bugs.webkit.org/show_bug.cgi?id=51572

        * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

worker-importScriptsOnError needs to allow for the console error to
flush out before calling notifyDone.

Reviewed by Adam Barth.

* http/tests/workers/worker-importScriptsOnError-expected.txt:
* http/tests/workers/worker-importScriptsOnError.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77594 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Daniel Bates.

        Add another XSSFilter PerformanceTest
        https://bugs.webkit.org/show_bug.cgi?id=53750

        This PerformanceTest tests the case that was slow in
        https://bugs.webkit.org/show_bug.cgi?id=49845

        * XSSFilter/large-post-many-inline-scripts-and-events.html: Added.
        * XSSFilter/resources: Added.
        * XSSFilter/resources/target-for-large-post-many-inline-scripts-and-events.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77593 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Reviewed by Tony Chang.

        nrwt: handle "broken pipe" notifications from DRT more cleanly.

        https://bugs.webkit.org/show_bug.cgi?id=52927

        * Scripts/webkitpy/layout_tests/port/server_process.py:
        * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77592 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Attempt to fix Chromium build.

        * html/parser/XSSFilter.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77591 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Add PerformanceTest for XSSFilter
        https://bugs.webkit.org/show_bug.cgi?id=53741

        This performance tests covers the case of a large POST data and many
        small event handlers.

        * XSSFilter: Added.
        * XSSFilter/large-post-many-events.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77590 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Unreviewed, rolling out r77562.
        http://trac.webkit.org/changeset/77562
        https://bugs.webkit.org/show_bug.cgi?id=53630

        broke chromium mac build

        * WebCore.gyp/WebCore.gyp:
        * WebCore.gyp/mac/check_objc_rename.sh: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77589 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Daniel Bates.

        XSS Auditor severely affects loading performance after submitting a large form
        https://bugs.webkit.org/show_bug.cgi?id=49845

        The XSSFilter catches some more cases and has different console
        messages than the XSSAuditor.  We might want to improve these messages
        in the future.

        * http/tests/security/xssAuditor/base-href-control-char-expected.txt:
        * http/tests/security/xssAuditor/base-href-expected.txt:
        * http/tests/security/xssAuditor/base-href-null-char-expected.txt:
        * http/tests/security/xssAuditor/base-href-scheme-relative-expected.txt:
        * http/tests/security/xssAuditor/dom-write-innerHTML-expected.txt:
        * http/tests/security/xssAuditor/embed-tag-control-char-expected.txt:
        * http/tests/security/xssAuditor/embed-tag-expected.txt:
        * http/tests/security/xssAuditor/embed-tag-javascript-url-expected.txt:
        * http/tests/security/xssAuditor/embed-tag-null-char-expected.txt:
        * http/tests/security/xssAuditor/full-block-base-href-expected.txt:
        * http/tests/security/xssAuditor/full-block-object-tag-expected.txt:
        * http/tests/security/xssAuditor/link-onclick-entities-expected.txt:
        * http/tests/security/xssAuditor/object-embed-tag-control-char-expected.txt:
        * http/tests/security/xssAuditor/object-embed-tag-expected.txt:
        * http/tests/security/xssAuditor/object-embed-tag-null-char-expected.txt:
        * http/tests/security/xssAuditor/object-tag-expected.txt:
        * http/tests/security/xssAuditor/object-tag-javascript-url-expected.txt:
        * http/tests/security/xssAuditor/script-tag-with-source-entities-expected.txt:
2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Daniel Bates.

        XSS Auditor severely affects loading performance after submitting a large form
        https://bugs.webkit.org/show_bug.cgi?id=49845

        Switch over from the XSSAuditor to the XSSFilter, improving performance
        on this example.

        * html/parser/XSSFilter.cpp:
        (WebCore::XSSFilter::filterToken):
        * page/XSSAuditor.cpp:
        (WebCore::XSSAuditor::isEnabled):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77588 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Unreviewed, rolling out r77567.
        http://trac.webkit.org/changeset/77567
        https://bugs.webkit.org/show_bug.cgi?id=53468

        broke chromium linux svg, canvas tests, possibly win also?

        * platform/graphics/skia/ImageBufferSkia.cpp:
        (WebCore::getImageData):
        (WebCore::ImageBuffer::getUnmultipliedImageData):
        (WebCore::ImageBuffer::getPremultipliedImageData):
        (WebCore::putImageData):
        (WebCore::ImageBuffer::putUnmultipliedImageData):
        (WebCore::ImageBuffer::putPremultipliedImageData):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77587 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Reviewed by Kenneth Russell.

        Update, resubmit change to bug 51572. I think the race that was
        fixed in r76703 caused the hang that caused us to roll this
        patch out before, so I'd like to try again.

        https://bugs.webkit.org/show_bug.cgi?id=51572

        * Scripts/webkitpy/layout_tests/layout_package/message_broker.py:
        * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77586 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Download bundles should be moved to their final destination when they
finish.

https://bugs.webkit.org/show_bug.cgi?id=53668

Reviewed by Adam Roben.

* WebProcess/Downloads/Download.cpp:
(WebKit::Download::Download):
Initialize m_allowOverwrite.
(WebKit::Download::retrieveDestinationWithSuggestedFilename):
Return the result of requesting the destination filename from the UI
process.
(WebKit::Download::decideDestinationWithSuggestedFilename):
Call retrieveDestinationWithSuggestedFilename().
(WebKit::Download::didFinish):
Call platformDidFinish().

* WebProcess/Downloads/Download.h:
Declare retrieveDestinationWithSuggestedFilename() and
platformDidFinish(). Add a member variable to store whether the UI
process would like us to overwrite existing files.

* WebProcess/Downloads/cf/DownloadCFNet.cpp:
(WebKit::willResumeWithResponseCallback):
Remove some trailing whitespace.

* WebProcess/Downloads/curl/DownloadCurl.cpp:
(WebKit::Download::platformDidFinish):
Stubbed.

* WebProcess/Downloads/mac/DownloadMac.mm:
(WebKit::Download::platformDidFinish):
Stubbed.

* WebProcess/Downloads/qt/DownloadQt.cpp:
(WebKit::Download::platformDidFinish):
Stubbed.

* WebProcess/Downloads/cf/win/DownloadCFNetWin.cpp:
(WebKit::Download::didDecideDestination):
Store whether the UI process would like us to overwrite existing files.
(WebKit::Download::platformDidFinish):
Try to move the bundle to the final destination. If that fails, ask the
UI process for a new destination. If we receive one, try again to move
the bundle. Call didCreateDestination() to report the updated path of
file whether it's the bundle that we failed to move, or the new final
destination.

This is similar to what is done by WebKit 1's CFNetwork port.

* win/WebKit2.vcproj:
Add DownloadWin.cpp.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77585 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Small clean-up in DownloadCFNet.cpp.

Rubber-stamped by Mark Rowe.

* WebProcess/Downloads/cf/DownloadCFNet.cpp:
(WebKit::Download::didDecideDestination):
Moved out of the section labeled "CFURLDownload callbacks" and up near
the other member functions.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77584 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Daniel Bates.

        Skip a test for WebKit2 that uses missing DRT API
        https://bugs.webkit.org/show_bug.cgi?id=53748

        * platform/mac-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77583 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Abhishek Arya  <inferno@chromium.org>

        Unreviewed, simple qt test baseline.

        * platform/qt/fast/overflow/overflow-height-float-not-removed-crash-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77579 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-02  MORITA Hajime  <morrita@google.com>

        Reviewed by Dimitri Glazkov.

        Refactoring: <progress> should not use ShadowElement
        https://bugs.webkit.org/show_bug.cgi?id=53583

        Updated to catch the change on -webkit-progress-bar-value interpretation.

        * fast/dom/HTMLProgressElement/progress-bar-value-pseudo-element.html:
2011-02-02  MORITA Hajime  <morrita@google.com>

        Reviewed by Dimitri Glazkov.

        Refactoring: <progress> should not use ShadowElement
        https://bugs.webkit.org/show_bug.cgi?id=53583

        - Introduced RenderIndicatorPart and RenderProgressBarValuePart
          to be responsible for bar-part layout,
          which adopted layout logic from ShadowBlockElement.
        - ProgressBarValueElement is no longer a subclass of ShadowBlockElement.
        - Remove dependency from RenderProgress to HTMLProgressElement and
          ShadowBlockElement.
        - The shadow tree is no longer removed on detach(). It becomes persistent.
          This is now possible because the ShadowBlockElement dependency is gone.
        - ::-webkit-appearance for -webkit-progress-bar-value is no longer referred.
          That didn't make sense.

        * html/HTMLProgressElement.cpp:
        (WebCore::HTMLProgressElement::createShadowSubtreeIfNeeded):
        * html/HTMLProgressElement.h:
        * html/shadow/ProgressBarValueElement.h: Added.
        (WebCore::ProgressBarValueElement::ProgressBarValueElement):
        (WebCore::ProgressBarValueElement::shadowPseudoId):
        (WebCore::ProgressBarValueElement::createRenderer):
        (WebCore::ProgressBarValueElement::create):
        * rendering/RenderIndicator.cpp:
        (WebCore::RenderIndicatorPart::RenderIndicatorPart):
        (WebCore::RenderIndicatorPart::~RenderIndicatorPart):
        (WebCore::RenderIndicatorPart::layout):
        (WebCore::RenderIndicatorPart::styleDidChange):
        * rendering/RenderIndicator.h: Added RenderIndicatorPart class
        (WebCore::RenderIndicatorPart::originalVisibility):
        (WebCore::RenderIndicatorPart::requiresForcedStyleRecalcPropagation):
        (WebCore::RenderIndicatorPart::canHaveChildren):
        * rendering/RenderProgress.cpp:
        (WebCore::RenderProgressBarValuePart::preferredFrameRect):
        (WebCore::RenderProgressBarValuePart::shouldBeHidden):
        (WebCore::RenderProgress::updateFromElement):
        (WebCore::RenderProgress::layoutParts):
        (WebCore::RenderProgress::shouldHaveParts):
        * rendering/RenderProgress.h:
        (WebCore::RenderProgressBarValuePart::RenderProgressBarValuePart):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77578 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Reversion should not be marked as misspelled.
https://bugs.webkit.org/show_bug.cgi?id=53255

Source/WebCore:

Patch by Jia Pu <jpu@apple.com> on 2011-02-03
This patch includes fix for reported bug, and also some housekeeping changes.

To implement desired behavior, we need:
1. Add a new marker type, SpellCheckingExemption, since now we distingusish between text
   that shouldn't be spellchecked and text shouldn't be autocorrected.
2. Make sure that there is no pending correction panel when we enter markAllMisspellingsAndBadGrammarInRanges().
   Otherwise the spell checking code in that function may interfere with autocorrection. This
   is achieved by explicitly applying pending correction when user types space, line break or
   paragraph break.

Housekeeping code changes include:
1. Change manual-tests that were broken by relocated WebCore directory.
2. Use TextIterator in various DocumentMarkerController functions instead of using
   Node::traverseNextNode() directly.
3. Allow passing multiple marker types into DocumentMarkerController::removeMarkers() and
   DocumentMarkerController::hasMarkers() to improve clarity and efficiency.
4. Fixes of minor bugs that were exposed previously.

* WebCore.exp.in: Change signature of DocumentMarkerController::removeMarkers().

* dom/DocumentMarker.h: Added new marker type SpellCheckingExemption.

* dom/DocumentMarkerController.cpp:
(WebCore::DocumentMarkerController::removeMarkers): Use TextIterator to scan the range to be
   consistent with addMarker() function. Allow passing in multiple marker types in one call.
   Added a boolean argument to specify the behavior when removing markers that partially
   overlap the specified range.
(WebCore::DocumentMarkerController::removeMarkersFromMarkerMapVectorPair): Allow passing in
   multiple marker types in one call.
(WebCore::DocumentMarkerController::hasMarkers): Use TextIterator to scan the range to be
   consistent with addMarker() function. Allow passing in multiple marker types in one call.

* dom/DocumentMarkerController.h: Allow passing in multiple marker types to removeMarkers()
   and hasMarkers(). Added a boolean argument to removeMarkers() to specify the behavior when
   removing markers that partially overlap the specified range.

* editing/Editor.cpp:
(WebCore::markerTypesForAutocorrection): Add SpellCheckingExemption marker when apply correction.
(WebCore::markerTypesForReplacement): Ditto.
(WebCore::Editor::respondToChangedSelection): Reordered call to dismissCorrectionPanel() and
   setSelection() to make sure there is no pending correction when entering
   markAllMisspellingsAndBadGrammarInRanges().
(WebCore::Editor::appliedEditing): Only remove CorrectionIndicator markers when the command
   is a top level command to improve efficiency.
(WebCore::Editor::insertTextWithoutSendingTextEvent): Added code to applying pending correction.
(WebCore::Editor::insertLineBreak): Ditto.
(WebCore::Editor::insertParagraphSeparator): Ditto.
(WebCore::Editor::markAllMisspellingsAndBadGrammarInRanges): Don't mark mispelling if the
   text carries SpellCheckingExemption marker.
(WebCore::Editor::correctionPanelTimerFired): Reset correction panel if the returned suggestion
   from spellchecker is an empty string.
(WebCore::Editor::removeSpellAndCorrectionMarkersFromWordsToBeEdited):
   Use new DocumentMarkerController::removeMarkers() to replace custom implemenation to improve
   efficiency and readability.
(WebCore::Editor::applyCorrectionPanelInfo): Remove the code that set caret position after
   applying correction, since it's unnecessary. Also, store pre-correction string together with
   the marker for reversion panel to use.
(WebCore::Editor::applyAutocorrectionBeforeTypingIfAppropriate): Apply pending correction.
(WebCore::Editor::changeSelectionAfterCommand): Moved marker removal code to Editor::appliedEditing()
   where we have access to EditCommand object.

* editing/Editor.h: Added new function applyAutocorrectionAfterTypingIfAppropriate().

* manual-tests/autocorrection/autocorrection-cancelled-by-ESC.html: Change manual-tests that
   were broken by relocated WebCore directory.

* manual-tests/autocorrection/autocorrection-cancelled-by-typing-1.html: Ditto.

* manual-tests/autocorrection/autocorrection-contraction.html: Ditto.

* manual-tests/autocorrection/continue-typing-to-dismiss-reversion.html: Ditto.

* manual-tests/autocorrection/delete-to-dismiss-reversion.html: Ditto.

* manual-tests/autocorrection/delete-to-end-of-word-to-show-reversion.html: Ditto.

* manual-tests/autocorrection/dismiss-multiple-guesses.html: Ditto.

* manual-tests/autocorrection/move-to-end-of-word-to-show-reversion.html: Ditto.

* manual-tests/autocorrection/select-from-multiple-guesses.html: Ditto.

* manual-tests/autocorrection/spell-checking-after-reversion.html: Added.

* manual-tests/autocorrection/type-whitespace-to-dismiss-reversion.html: Change manual-tests that
   were broken by relocated WebCore directory.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDocumentMarkers): Code clean-up to be more concise.

LayoutTests:

Patch by Jia Pu <jpu@apple.com> on 2011-02-03
Rebaselined tests broken by https://bugs.webkit.org/show_bug.cgi?id=51389

* platform/mac/editing/spelling/autocorrection-contraction-expected.png:
* platform/mac/editing/spelling/autocorrection-contraction-expected.txt:
* platform/mac/editing/spelling/autocorrection-delete-expected.png:
* platform/mac/editing/spelling/autocorrection-delete-expected.txt:
* platform/mac/editing/spelling/autocorrection-removing-underline-after-paste-expected.png:
* platform/mac/editing/spelling/autocorrection-removing-underline-after-paste-expected.txt:
* platform/mac/editing/spelling/autocorrection-removing-underline-expected.png:
* platform/mac/editing/spelling/autocorrection-removing-underline-expected.txt:
* platform/mac/editing/spelling/autocorrection-simple-expected.checksum:
* platform/mac/editing/spelling/autocorrection-simple-expected.png:
* platform/mac/editing/spelling/autocorrection-simple-expected.txt:
* platform/mac/editing/spelling/autocorrection-simple.html:
* platform/mac/editing/spelling/delete-autocorrected-word-1-expected.png:
* platform/mac/editing/spelling/delete-autocorrected-word-1-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77577 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        Paint non-composited contents in a layer
        https://bugs.webkit.org/show_bug.cgi?id=53742

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::setNeedsDisplay):
        Call LayerTreeHost::setNonCompositedContentsNeedDisplayInRect.

        (WebKit::DrawingAreaImpl::scroll):
        Call LayerTreeHost::scrollNonCompositedContents.

        (WebKit::DrawingAreaImpl::setSize):
        Call LayerTreeHost::sizeDidChange.

        * WebProcess/WebPage/LayerTreeHost.h:
        Add new pure virtual member functions.

        * WebProcess/WebPage/mac/LayerTreeHostMac.mm:
        (WebKit::LayerTreeHostMac::LayerTreeHostMac):
        Create the non-composited content layer.

        (WebKit::LayerTreeHostMac::setNonCompositedContentsNeedDisplayInRect):
        Call setNeedsDisplay on the GraphicsLayer.

        (WebKit::LayerTreeHostMac::scrollNonCompositedContents):
        Just call setNonCompositedContentsNeedDisplayInRect.

        (WebKit::LayerTreeHostMac::sizeDidChange):
        Resize the layers and flush any changes.

        (WebKit::LayerTreeHostMac::paintContents):
        Call WebPage::drawRect.

        (WebKit::LayerTreeHostMac::flushPendingLayerChangesRunLoopObserverCallback):
        Lay out the web page.

        (WebKit::LayerTreeHostMac::flushPendingLayerChanges):
        Call syncCompositingStateForThisLayerOnly on the non-composited content layer.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77575 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed.

Adding myself to committers.py.

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

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77574 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Abhishek Arya  <inferno@chromium.org>

        Unreviewed, qt build fix.

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

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77573 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        Don't clear the callbacks ref in the proxy layer for IndexedDB
        https://bugs.webkit.org/show_bug.cgi?id=53535

        Clearing the callbacks doesn't actually solve any problems and makes
        this code behave subtly differently than how it would in Safari and
        other single process environments. Let's remove the difference.

        * src/IDBCallbacksProxy.cpp:
        (WebCore::IDBCallbacksProxy::onError):
        (WebCore::IDBCallbacksProxy::onSuccess):
        * src/IDBTransactionCallbacksProxy.cpp:
        (WebCore::IDBTransactionCallbacksProxy::onAbort):
        (WebCore::IDBTransactionCallbacksProxy::onComplete):
        (WebCore::IDBTransactionCallbacksProxy::onTimeout):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77572 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Include the crashing function in the link to a crash log

Fixes <http://webkit.org/b/53739> Crash log links in results.html should include the
function that crashed

Reviewed by David Kilzer.

* Scripts/old-run-webkit-tests:
(crashLocation): Added. Returns the location of the crash.
(linksForErrorTest): Include the crash location in the link text for the crash log, if one
could be determined.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77571 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Versioning.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77570 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        When in accelerated compositing mode, wait for DidSetSize messages
        https://bugs.webkit.org/show_bug.cgi?id=53736

        Normally we wait for DidSetSize messages from ::paint, but that's never called in
        accelerated compositing mode so we just do it right after sending the SetSize message.

        * UIProcess/DrawingAreaProxyImpl.cpp:
        (WebKit::DrawingAreaProxyImpl::paint):
        Call waitForAndDispatchDidSetSize.

        (WebKit::DrawingAreaProxyImpl::sendSetSize):
        Call waitForAndDispatchDidSetSize when in accelerated compositing mode.

        (WebKit::DrawingAreaProxyImpl::waitForAndDispatchDidSetSize):
        New utility function which waits for and dispatches a DidSetSize message.

        * UIProcess/DrawingAreaProxyImpl.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77568 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Brian Salomon  <bsalomon@google.com>

        Reviewed by James Robinson.

        NHandle non-raster backed images in getUnmultipliedImageData()
        https://bugs.webkit.org/show_bug.cgi?id=53468

        No new tests. Existing canvas tests sufficient
        LayoutTests/canvas/philip/...

        * platform/graphics/skia/ImageBufferSkia.cpp:
        (WebCore::getImageData):
        (WebCore::ImageBuffer::getUnmultipliedImageData):
        (WebCore::ImageBuffer::getPremultipliedImageData):
        (WebCore::putImageData):
        (WebCore::ImageBuffer::putUnmultipliedImageData):
        (WebCore::ImageBuffer::putPremultipliedImageData):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77567 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Hans Wennborg  <hans@chromium.org>

        Reviewed by Jeremy Orlow.

        Chromium: Remove deprecated WebIDBObjectStore::put() function
        https://bugs.webkit.org/show_bug.cgi?id=53669

        Remove deprecated put() function. Chromium side code now overrides
        the new version.

        * public/WebIDBObjectStore.h:
        (WebKit::WebIDBObjectStore::put):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77566 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Abhishek Arya  <inferno@chromium.org>

        Reviewed by James Robinson.

        Tests that we do not crash when overflowing the height of a root
        inline box.
        https://bugs.webkit.org/show_bug.cgi?id=53729

        * fast/overflow/overflow-height-float-not-removed-crash-expected.txt: Added.
        * fast/overflow/overflow-height-float-not-removed-crash.html: Added.
2011-02-03  Abhishek Arya  <inferno@chromium.org>

        Reviewed by James Robinson.

        Enforce more limits on root inline boxes height calculations.
        https://bugs.webkit.org/show_bug.cgi?id=53729

        Test: fast/overflow/overflow-height-float-not-removed-crash.html

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::removeFloatingObject): prevent logicalBottom to
        become negative when logicalTop is INT_MAX.
        (WebCore::RenderBlock::markLinesDirtyInBlockRange): when logicalBottom
        is INT_MAX, we should dirty everything. So, we bail out to make
        afterLowest equal to the lastRootBox() or lowestDirstLine.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77565 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Merge PerformanceTests/Parser/ChangeLog and
PerformanceTests/PageLoad/ChangeLog into PerformanceTests/ChangeLog.

Rubber-stamped by Eric Seidel.

It's silly to have ChangeLogs for each of these directories separately.
I've left SunSpider with its own ChangeLog because it's more of an
independent entity.

* ChangeLog: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77564 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Worker.importScript() should clean errors for cross origin imports.
https://bugs.webkit.org/show_bug.cgi?id=52871

Reviewed by Adam Barth and Oliver Hunt.

Source/WebCore:

Test: http/tests/workers/worker-importScriptsOnError.html

* bindings/js/WorkerScriptController.cpp:
(WebCore::WorkerScriptController::evaluate): Use sanitizeScriptError
to determine when to create a clean exception.
* bindings/v8/WorkerContextExecutionProxy.cpp:
(WebCore::WorkerContextExecutionProxy::evaluate): Ditto.
* dom/ScriptExecutionContext.cpp:
(WebCore::ScriptExecutionContext::sanitizeScriptError): Figure out
if the error needs to be cleaned up.
(WebCore::ScriptExecutionContext::dispatchErrorEvent): Extracted
sanitizeScriptError for use by other places.
* dom/ScriptExecutionContext.h:
* workers/WorkerContext.cpp:
(WebCore::WorkerContext::importScripts): Use the reponse url when
telling the evaluate where the script came fro.
* workers/WorkerScriptLoader.cpp:
(WebCore::WorkerScriptLoader::responseURL): Expose the url that
the script was loaded from (which may be different from url() due
to redirects).
(WebCore::WorkerScriptLoader::didReceiveResponse): Capture the reponse url.
* workers/WorkerScriptLoader.h:

LayoutTests:

* http/tests/workers/resources/worker-importScripts-error.js: Added.
* http/tests/workers/resources/worker-importScripts-throw.js: Added.
* http/tests/workers/worker-importScriptsOnError.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77563 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Mark Mentovai  <mark@chromium.org>

        Reviewed by Dimitri Glazkov.

        Chromium GYP build fix.

        When various settings were moved to webcore_prerequisites in r66364,
        things that should have been direct_dependent_settings were not marked
        as such. GYP 'defines', for example, make no sense on a 'none'-type
        target such as webcore_prerequisites. It appears that it was intended
        for these settings to be pushed to direct dependents, which would make
        direct_dependent_settings correct.

        Losing the ChromiumWebCoreObjC defines on the Mac, for example, caused
        http://crbug.com/71537, which at best causes Mac console log spew, and
        at worst may result in Chromium's copy of WebCore using system
        definitions of certain Objective-C classes at runtime, or vice-versa.

        The build now includes a postbuild step to prevent
        http://crbug.com/71537 from regressing again. The build will fail upon
        regression.

        https://bugs.webkit.org/show_bug.cgi?id=53630

        * WebCore.gyp/WebCore.gyp: Move things in webcore_prerequisites into
          direct_dependent_settings as needed, add the check_objc_rename
          postbuild step.
        * WebCore.gyp/mac/check_objc_rename.sh: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77562 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        Pass along an UpdateInfo object to the ExitAcceleratedCompositingMode message
        https://bugs.webkit.org/show_bug.cgi?id=53734

        * UIProcess/DrawingAreaProxy.h:
        (WebKit::DrawingAreaProxy::exitAcceleratedCompositingMode):
        Add an UpdateInfo parameter.

        * UIProcess/DrawingAreaProxy.messages.in:
        Add an UpdateInfo parameter to the ExitAcceleratedCompositingMode message.

        * UIProcess/DrawingAreaProxyImpl.cpp:
        (WebKit::DrawingAreaProxyImpl::exitAcceleratedCompositingMode):
        Incorporate the passed in update.

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::exitAcceleratedCompositingMode):
        Create a new UpdateInfo object, paint into it and pass it along as a parameter
        to the ExitAcceleratedCompositingMode message.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77561 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Make XSSFilter go fast by adding a SuffixTree
        https://bugs.webkit.org/show_bug.cgi?id=53665

        The SuffixTree lets us quickly reject snippets if the POST data is
        large (because we can avoid a linear scan over the POST data).

        * html/parser/XSSFilter.cpp:
        (WebCore::XSSFilter::init):
        (WebCore::XSSFilter::isContainedInRequest):
        * html/parser/XSSFilter.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77560 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Mihai Parparita  <mihaip@chromium.org>

        Reviewed by Alexey Proskuryakov.

        REGRESSION (r77355): Page cache layout tests crash
        https://bugs.webkit.org/show_bug.cgi?id=53648

        Add test that calls setTimeout during a pagehide event handler (the
        timeout should not fire while the document is in the page cache).

        Skipped for Chromium since it doesn't use the page cache.

        * fast/events/pagehide-timeout-expected.txt: Added.
        * fast/events/pagehide-timeout.html: Added.
        * fast/events/resources/pagehide-timeout-go-back.html: Added.
        * platform/chromium/test_expectations.txt:
2011-02-03  Mihai Parparita  <mihaip@chromium.org>

        Reviewed by Alexey Proskuryakov.

        REGRESSION (r77355): Page cache layout tests crash
        https://bugs.webkit.org/show_bug.cgi?id=53648

        Test: fast/events/pagehide-timeout.html

        Suspend active DOM objects after all pagehide event handlers have run,
        otherwise it's possible for them to create more objects that weren't
        getting suspended.

        * history/CachedFrame.cpp:
        (WebCore::CachedFrame::CachedFrame):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77559 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        SerializedScriptValue should not require v8 to create undefined and null values
        https://bugs.webkit.org/show_bug.cgi?id=53730

        Instead of creating a v8 type and passing that into the constructor, just use
        the writer class directly. While I was at it, I cleaned up the code a bit too
        by getting rid of the WireData/StringValue enum as I found that personally
        confusing.

        This is necessary because these methods are called by IndexedDB in the browser
        process where v8 is not spun up.

        No functionality changed and not possible to test.

        * bindings/v8/SerializedScriptValue.cpp:
        (WebCore::SerializedScriptValue::createFromWire):
        (WebCore::SerializedScriptValue::create):
        (WebCore::SerializedScriptValue::nullValue):
        (WebCore::SerializedScriptValue::undefinedValue):
        (WebCore::SerializedScriptValue::release):
        (WebCore::SerializedScriptValue::SerializedScriptValue):
        * bindings/v8/SerializedScriptValue.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77558 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Oliver Hunt.

        Not all blocks are freed when the heap is freed (counting is hard!)
        https://bugs.webkit.org/show_bug.cgi?id=53732

        * runtime/MarkedSpace.cpp:
        (JSC::MarkedSpace::destroy): Freeing a block compacts the list, so just
        keep freeing block 0 until there are no blocks left.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77557 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Unreviewed, build fix.

        Actually make the changes suggested by Mihai in his review
        of bug 53720 :(

        https://bugs.webkit.org/show_bug.cgi?id=53720

        * Scripts/webkitpy/layout_tests/port/test_files.py:
        * Scripts/webkitpy/layout_tests/port/test_files_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77556 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix for <rdar://problem/8944544> Ability to animate track
for WKPainter scrollers

Reviewed by Sam Weinig.

Source/WebCore:

Two new WebKitSystemInterface functions.
* WebCore.exp.in:
* platform/mac/WebCoreSystemInterface.h:
* platform/mac/WebCoreSystemInterface.mm:

Use Scrollbar::convertFromContainingView() to return the right point.
* platform/mac/ScrollAnimatorMac.mm:
(-[ScrollbarPainterControllerDelegate scrollerImpPair:convertContentPoint:toScrollerImp:]):

ScrollKnobAnimation is now ScrollbarPartAnimation. It can
now be used to animate the knob or the track.
(-[ScrollbarPartAnimation initWithScrollbarPainter:part:WebCore::scrollAnimator:WebCore::animateAlphaTo:duration:]):
(-[ScrollbarPartAnimation setCurrentProgress:]):
(-[ScrollbarPainterDelegate setUpAnimation:scrollerPainter:part:WebCore::animateAlphaTo:duration:]):
(-[ScrollbarPainterDelegate scrollerImp:animateKnobAlphaTo:duration:]):
(-[ScrollbarPainterDelegate scrollerImp:animateTrackAlphaTo:duration:]):

Scrollbars need invalodating after the overlay state changes.
(-[ScrollbarPainterDelegate scrollerImp:overlayScrollerStateChangedTo:]):

Source/WebKit/mac:

* WebCoreSupport/WebSystemInterface.mm:
(InitWebCoreSystemInterface):

Source/WebKit2:

* WebCoreSupport/WebSystemInterface.mm:
(InitWebCoreSystemInterface):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77555 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        Added SL-specific result for the test and put back the original result that
        I've neglected to do in r77539.

        * platform/mac-snowleopard/svg/W3C-SVG-1.1-SE/types-dom-01-b-expected.txt: Added.
        * platform/mac/svg/text/font-size-below-point-five-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77554 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Reviewed by Mihai Parparita.

        new-run-webkit-tests: Fix bug introduced in r77434 that was
        causing us to run the canvas GPU tests on the Mac GPU port. This
        bug revealed that we were not setting the port.name() field
        properly in many cases, so I've cleaned up all of that code, and
        removed a few comments about "version-specific" GPU ports that
        don't exist and just confused things.

        Testing also revealed that port.abspath_for_test() wasn't
        normalizing paths on Windows properly, so I fixed that as well.

        https://bugs.webkit.org/show_bug.cgi?id=53719

        * Scripts/webkitpy/layout_tests/port/base.py:
        * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
        * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77551 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dirk Pranke  <dpranke@chromium.org>

        Reviewed by Mihai Parparita.

        new-run-webkit-tests: fix normalization of paths on windows when gathering files

        r77434 introduced a bug that was causing the test expectations
        for the GPU ports to not be treated properly. It turns out that
        when we gathered the list of test files on Windows, we would
        return paths of the form "c:\LayoutTests/fast/canvas", and the
        mixture of backslashes and forward slashes was confusing things.

        This patch normalizes all of the filenames returned from
        test_files.find(), and adds better tests for this (fixing a
        couple of other bugs found in the meantime).

        https://bugs.webkit.org/show_bug.cgi?id=53720

        * Scripts/webkitpy/layout_tests/port/test.py:
        * Scripts/webkitpy/layout_tests/port/test_files.py:
        * Scripts/webkitpy/layout_tests/port/test_files_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77550 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Scroll thumb jumps to top when resizing horizontally.

Reviewed by Beth Dakin.

* platform/ScrollView.cpp:
(WebCore::ScrollView::updateScrollbars): Add call to update
the scrollbar's offset in the case where we may have created
a new scrollbar but have not changed the current position.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77549 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Justin Schuh  <jschuh@chromium.org>

        Reviewed by Dirk Schulze.

        startAnimations should use a local, RefCounted Vector.
        https://bugs.webkit.org/show_bug.cgi?id=53458

        Test: svg/custom/use-animation-in-fill.html

        * svg/SVGDocumentExtensions.cpp:
        (WebCore::SVGDocumentExtensions::startAnimations):
2011-02-03  Justin Schuh  <jschuh@chromium.org>

        Reviewed by Dirk Schulze.

        startAnimations should use a local, RefCounted Vector.
        https://bugs.webkit.org/show_bug.cgi?id=53458

        * svg/custom/use-animation-in-fill-expected.txt: Added.
        * svg/custom/use-animation-in-fill.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77548 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed.

small text which is scaled to be large renders pixelated
https://bugs.webkit.org/show_bug.cgi?id=12448

* platform/qt/Skipped: Add failing tests until fix.
 - css3/zoom-coords.xhtml
 - svg/zoom/page/zoom-zoom-coords.xhtml

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77547 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        Change DidSetSize to take a LayerTreeContext parameter
        https://bugs.webkit.org/show_bug.cgi?id=53724

        Since SetSize can cause layout to happen, we need to pass along a layer tree context to
        the DidSetSize message so the UI process can enter/exit accelerated compositing mode appropriately.

        * Shared/LayerTreeContext.h:
        * Shared/mac/LayerTreeContextMac.mm:
        Add equality operators to LayerTreeContext.

        * UIProcess/DrawingAreaProxy.h:
        (WebKit::DrawingAreaProxy::didSetSize):
        This now takes a LayerTreeContext parameter.

        * UIProcess/DrawingAreaProxy.messages.in:
        Add LayerTreeContext parameter.

        * UIProcess/DrawingAreaProxyImpl.cpp:
        Get rid of m_isInAcceleratedCompositingMode and use the current layer context instead.

        (WebKit::DrawingAreaProxyImpl::paint):
        Bail if the call to DidSetSize ended up entering accelerated compositing mode.

        (WebKit::DrawingAreaProxyImpl::didSetSize):
        If the new layer tree context is different, enter/exit accelerated compositing mode accordingly.

        (WebKit::DrawingAreaProxyImpl::enterAcceleratedCompositingMode):
        (WebKit::DrawingAreaProxyImpl::exitAcceleratedCompositingMode):
        Update the layer tree context.

        * UIProcess/DrawingAreaProxyImpl.h:
        (WebKit::DrawingAreaProxyImpl::isInAcceleratedCompositingMode):
        Add simple getter.

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::setSize):
        Pass along the layer tree context.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77546 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Adam Barth  <abarth@webkit.org>

        Reviewed by Daniel Bates.

        XSSFilter shouldn't bother to analyze pages without "injection"
        characters in the request
        https://bugs.webkit.org/show_bug.cgi?id=53664

        If the request lacks these "injection" characters, then it's unlikely
        that there's a reflective XSS attack happening.  This hueristic lets us
        avoid analyzing the vast majority of responses for XSS.  Of course, the
        hueristic isn't perfect.  Because of this huerstic, we miss out on
        injections into unquoted attributes.  However, it's a trade-off that's
        worked well in the XSSAuditor.

        * html/parser/XSSFilter.cpp:
        (WebCore::HTMLNames::isRequiredForInjection):
        (WebCore::XSSFilter::XSSFilter):
        (WebCore::XSSFilter::init):
        (WebCore::XSSFilter::filterToken):
        (WebCore::XSSFilter::isContainedInRequest):
        * html/parser/XSSFilter.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77545 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        Add a failure until I investigate its cause.

        * platform/chromium/test_expectations.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77544 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        Updated platform/mac-leopard baselines for fast/backgrounds after r77485.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77540 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        Move Leopard result to the right spot, put back the original result.

        * platform/mac-leopard/svg/text/font-size-below-point-five-expected.txt:
            Renamed from LayoutTests/platform/mac/svg/text/font-size-below-point-five-expected.txt.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77539 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Skip an SVG test that fails depending on what other tests came before it

Running any other SVG test before this one causes it to fail. If we were to check in new
failing results, it might start passing again if some more tests are added before it that
make it the first test that is run for a particular invocation of DumpRenderTree. So the
safest thing to do for now is to skip it. <http://webkit.org/b/35700> tracks this failure.

* platform/win/Skipped: Added svg/W3C-SVG-1.1-SE/types-dom-01-b.svg.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77538 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add links to Windows crash logs in results.html

Fixes <http://webkit.org/b/53718> run-webkit-tests should link to Windows crash logs in
results.html

Reviewed by David Kilzer.

* Scripts/old-run-webkit-tests:
(top level): Added $crashLogTag and $windowsCrashLogFilePrefix (which came from
createDebuggerCommandFile).
(testCrashedOrTimedOut): If the test crashed, capture any saved crash log after the dump
tool has exited.
(captureSavedCrashLog): Added. Finds the crash log for the test that just crashed and moves
it into the test results directory tree. Only implemented for Cygwin currently.
(findNewestFileMatchingGlob): Added. Does what it says.
(htmlForResultsSection): Only link to files that exist.
(linksForErrorTest): Add a link to the crash log.
(deleteExpectedAndActualResults): Delete any old crash log for this test.
(createDebuggerCommandFile): Use the new $windowsCrashLogFilePrefix constant instead of
hardcoding it here.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77537 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        Updated platform/mac and platform/mac-leopard baselines for svg/ after r77485.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77536 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Martin Robinson  <mrobinson@igalia.com>

        Fix GTK+ WebGL build after r76600.

        * configure.ac: Make sure to respect configure.ac conventions
        consistently when it comes to capitalization. Remove an old
        reference to enable_3d_canvas.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77535 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Get rid of LayerTreeContext::seed on Mac.
Add sequence checks to enterAcceleratedCompositingMode and exitAcceleratedCompositingMode and
have them call newly added overloads that don't have the checks.

Reviewed by Dan Bernstein.

* Shared/LayerTreeContext.h:
* Shared/mac/LayerTreeContextMac.mm:
(WebKit::LayerTreeContext::LayerTreeContext):
(WebKit::LayerTreeContext::encode):
(WebKit::LayerTreeContext::decode):
* UIProcess/DrawingAreaProxyImpl.cpp:
(WebKit::DrawingAreaProxyImpl::enterAcceleratedCompositingMode):
(WebKit::DrawingAreaProxyImpl::exitAcceleratedCompositingMode):
* UIProcess/DrawingAreaProxyImpl.h:
* WebProcess/WebPage/mac/LayerTreeHostMac.mm:
(WebKit::LayerTreeHostMac::LayerTreeHostMac):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77534 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        Get rid of timestamps in UpdateInfo
        https://bugs.webkit.org/show_bug.cgi?id=53713

        Add a 64-bit sequence number to all DrawingAreaProxy messages, and use it to
        check whether Update messages should be handled or not.

        * Shared/UpdateInfo.cpp:
        (WebKit::UpdateInfo::encode):
        (WebKit::UpdateInfo::decode):
        * Shared/UpdateInfo.h:
        (WebKit::UpdateInfo::UpdateInfo):
        * UIProcess/BackingStore.cpp:
        (WebKit::BackingStore::BackingStore):
        (WebKit::BackingStore::incorporateUpdate):
        * UIProcess/BackingStore.h:
        * UIProcess/DrawingAreaProxy.h:
        (WebKit::DrawingAreaProxy::update):
        (WebKit::DrawingAreaProxy::didSetSize):
        (WebKit::DrawingAreaProxy::enterAcceleratedCompositingMode):
        (WebKit::DrawingAreaProxy::exitAcceleratedCompositingMode):
        * UIProcess/DrawingAreaProxy.messages.in:
        * UIProcess/DrawingAreaProxyImpl.cpp:
        (WebKit::DrawingAreaProxyImpl::DrawingAreaProxyImpl):
        (WebKit::DrawingAreaProxyImpl::update):
        (WebKit::DrawingAreaProxyImpl::didSetSize):
        (WebKit::DrawingAreaProxyImpl::enterAcceleratedCompositingMode):
        (WebKit::DrawingAreaProxyImpl::exitAcceleratedCompositingMode):
        * UIProcess/DrawingAreaProxyImpl.h:
        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::generateSequenceNumber):
        (WebKit::DrawingAreaImpl::setSize):
        (WebKit::DrawingAreaImpl::enterAcceleratedCompositingMode):
        (WebKit::DrawingAreaImpl::exitAcceleratedCompositingMode):
        (WebKit::DrawingAreaImpl::display):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77533 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Vangelis Kokkevis  <vangelis@chromium.org>

        Reviewed by Kenneth Russell.

        [chromium] Fixing a compositor crash occurring on layers
        without an associated RenderSurface.
        https://bugs.webkit.org/show_bug.cgi?id=53679
        Regression was introduced by in r77425

        Test: http://webkit.org/blog/386/3d-transforms/ doesn't crash
        anymore.

        * platform/graphics/chromium/LayerRendererChromium.cpp:
        (WebCore::LayerRendererChromium::drawLayer):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77532 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Updated baselines that are specific to Leopard.

        * platform/chromium-mac-leopard/svg/custom/linking-a-03-b-preserveAspectRatio-expected.txt:
        * platform/chromium-mac-leopard/svg/custom/linking-a-03-b-viewTarget-expected.txt:
        * platform/chromium-mac-leopard/svg/custom/linking-a-03-b-zoomAndPan-expected.txt:
        * platform/chromium-mac-leopard/svg/custom/stroked-pattern-expected.png:
        * platform/chromium-mac-leopard/svg/custom/stroked-pattern-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77531 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/8948788> Text emphasis marks have wrong orientation for vertical text
https://bugs.webkit.org/show_bug.cgi?id=53709

Reviewed by Anders Carlsson.

Source/WebCore:

Covered by rendering of fast/text/emphasis-vertical.html

* platform/graphics/mac/SimpleFontDataMac.mm:
(WebCore::SimpleFontData::scaledFontData): Give the scaled font the same orientation this font
has.

LayoutTests:

* platform/mac/fast/text/emphasis-vertical-expected.checksum:
* platform/mac/fast/text/emphasis-vertical-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77530 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Anders Carlsson  <andersca@apple.com>

        Reviewed by Darin Adler.

        Don't send enter/exit accelerated compositing mode messages when handling setSize
        https://bugs.webkit.org/show_bug.cgi?id=53706

        Don't send EnterAcceleratedCompositingMode/ExitAcceleratedCompositing mode when
        the mode changes happen as a result of relayout in setSize, since an updated state
        is going to be sent back in the DidSetSize message.

        * UIProcess/DrawingAreaProxyImpl.cpp:
        (WebKit::DrawingAreaProxyImpl::didSetSize):
        Add an assert.

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::DrawingAreaImpl):
        Initialize m_inSetSize.

        (WebKit::DrawingAreaImpl::setRootCompositingLayer):
        Factor out code into enterAcceleratedCompositingMode and exitAcceleratedCompositingMode.

        (WebKit::DrawingAreaImpl::setSize):
        Set m_inSetSize to true while handling setSize.

        (WebKit::DrawingAreaImpl::enterAcceleratedCompositingMode):
        (WebKit::DrawingAreaImpl::exitAcceleratedCompositingMode):
        Take over the message sending responsibility from LayerTreeHostMac, and don't send messages
        if we're currently handling a SetSize message.

        * WebProcess/WebPage/DrawingAreaImpl.h:
        Add m_inSetSize.

        * WebProcess/WebPage/LayerTreeHost.h:
        Add layerTreeHost getter.

        * WebProcess/WebPage/mac/LayerTreeHostMac.h:
        * WebProcess/WebPage/mac/LayerTreeHostMac.mm:
        (WebKit::LayerTreeHostMac::LayerTreeHostMac):
        Don't send EnterAcceleratedCompositingMode messages.

        (WebKit::LayerTreeHostMac::layerTreeContext):
        Add getter for the current layer tree context.

        (WebKit::LayerTreeHostMac::invalidate):
        Don't send ExitAcceleratedCompositingMode messages.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77529 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Martin Robinson  <mrobinson@igalia.com>

        Fix one GTK+ skipped list entry that referred to the wrong test file.

        * platform/gtk/Skipped: Fix skipped list entry.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77528 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Removed the checksum that is out of place.

        * platform/chromium-win-xp/svg/text/text-fonts-01-t-expected.checksum: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77527 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-03  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Updated SVG baselines after r77485 for Mac and added expectations
        of failure while waiting for platform/mac pixel results to land.

        * platform/chromium-mac/svg/custom/linking-a-03-b-transform-expected.txt: Removed.
        * platform/chromium-mac/svg/custom/stroked-pattern-expected.checksum: Added.
        * platform/chromium-mac/svg/custom/stroked-pattern-expected.png: Added.
        * platform/chromium/test_expectations.txt: Added fails.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77526 268f45cc-cd09-0410-ab3c-d52691b4dbfc