WebKit-https.git
7 years ago[EFL][EWebLauncher] Use Ctrl +/- to zoom in EWebLauncher
commit-queue@webkit.org [Wed, 7 Nov 2012 04:39:00 +0000 (04:39 +0000)]
[EFL][EWebLauncher] Use Ctrl +/-  to zoom in EWebLauncher
https://bugs.webkit.org/show_bug.cgi?id=100207

Patch by KyungTae Kim <ktf.kim@samsung.com> on 2012-11-06
Reviewed by Gyuyoung Kim.

Use "Ctrl -" instead of "F7" to zoom out, and use "Ctrl +" instead of "F8" to zoom in,
and use "Ctrl 0" to set to default zoom level.
They are more common shortcuts on usual browsers.

* EWebLauncher/main.c:
(on_key_down):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133714 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRe-order variables in BidiRun and LayoutState
cevans@google.com [Wed, 7 Nov 2012 04:14:48 +0000 (04:14 +0000)]
Re-order variables in BidiRun and LayoutState
https://bugs.webkit.org/show_bug.cgi?id=100173

Reviewed by Eric Seidel.

The variable re-ordering and use of bitfields for bools has two benefits:
1) Size reduction. sizeof(BidiRun) goes down from 48 to 40 bytes on 64-bit. This is achieved by removing a bool member variable from BidiRun and packing it together with other bools in the BidiCharacterRun base class.
2) Security improvement. We have a lot of use-after-free in the RenderObject hierarchy, and the RenderArena class protects us from a lot of trouble by ensuring that objects of arbitrary type cannot be overlayed on top of freed RenderObjects. This change additionally makes sure that non-virtual RenderArena allocated objects do not have member variables which fully overlap the freed vtable pointer. This leaves re-used vtable pointers always pointing to either a valid vtable or an invalid address due to the freelist high-bit poisoning.

This change is exclusively about size savings; it is performance neutral as you would expect, including on Parser/html5-full-render.html

* platform/text/BidiResolver.h:
(WebCore::BidiCharacterRun::BidiCharacterRun): impact from re-ordering members.
(BidiCharacterRun): provide an efficiently packed bit of storage for BidiRun subclass to use, and re-order members to place bools adjacent.
* rendering/BidiRun.h:
(WebCore::BidiRun::BidiRun): use base class' efficiently packed bit storage for m_hasHyphen.
(BidiRun): m_hasHyphen is now stored in the base class.
* rendering/LayoutState.cpp:
(WebCore::LayoutState::LayoutState):
* rendering/LayoutState.h:
(WebCore::LayoutState::LayoutState): impact from re-ordering members.
(LayoutState): re-order members to place bools adjacently.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133713 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium-win] Refactor date/time format conversion code in LocaleWin
tkent@chromium.org [Wed, 7 Nov 2012 03:33:38 +0000 (03:33 +0000)]
[Chromium-win] Refactor date/time format conversion code in LocaleWin
https://bugs.webkit.org/show_bug.cgi?id=101329

Reviewed by Kentaro Hara.

convertWindowsDateFormatToLDML and convertWindowsTimeFormatToLDML did
similar jobs with different code. We can merge them into one function.

As for the date format conversion, we have tokenized a format string and
stored token objects into a vector. We skip token object generation and
generate a LDML-compliant pattern string during parsing.

No new tests. This should make no behavior changes and
WebKit/chromium/tests/LocaleWinTest.cpp has test cases.

* platform/text/win/LocaleWin.cpp:
(WebCore): Remove DateFormatToken, isFooSymbol, ensureShortDateTokens,
convertWindowsDateFormatToLDML, mapCharacterToDateTimeFieldType, and
convertWindowsTimeFormatToLDML.
(WebCore::commitLiteralToken): Store the result to the specified StringBuilder.
(WebCore::convertWindowsDateTimeFormat):
Renamed from parseDateFormat, and supports time format symbols.
(WebCore::LocaleWin::dateFormat): Use convertWindowsDateTimeFormat.
(WebCore::LocaleWin::monthFormat): Ditto.
(WebCore::LocaleWin::timeFormat): Ditto.
(WebCore::LocaleWin::shortTimeFormat): Ditto.
* platform/text/win/LocaleWin.h:
(LocaleWin): Remove ensureShortDateTokens and m_shortDateTokens.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133712 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMake Document::renderer faster by using the cached ptr for RenderView
commit-queue@webkit.org [Wed, 7 Nov 2012 03:31:32 +0000 (03:31 +0000)]
Make Document::renderer faster by using the cached ptr for RenderView
https://bugs.webkit.org/show_bug.cgi?id=101409

Patch by Elliott Sprehn <esprehn@chromium.org> on 2012-11-06
Reviewed by Eric Seidel.

In Bug 101277 I made Document::renderView() just return m_renderView instead
of casting the result of renderer(). While this made renderView() cheaper it
makes more sense to just make renderer() faster for Document pointers and
go back to the out of line method in RenderView.h because lots of places in
the code do document()->renderer().

No new tests, this is just a refactor.

* dom/Document.h:
(WebCore::Document::renderer):
* rendering/RenderView.h:
(WebCore::Document::renderView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133711 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[V8] neuterBinding should remember the type of its argument
abarth@webkit.org [Wed, 7 Nov 2012 02:31:55 +0000 (02:31 +0000)]
[V8] neuterBinding should remember the type of its argument
https://bugs.webkit.org/show_bug.cgi?id=101413

Reviewed by Kentaro Hara.

This is a small part of the patch in
https://bugs.webkit.org/show_bug.cgi?id=101110 that we can land
separately to try to tease out what is causing the memory regression.

* bindings/v8/SerializedScriptValue.cpp:
(WebCore):
(WebCore::neuterBinding):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133708 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoBuild fix.
andersca@apple.com [Wed, 7 Nov 2012 02:30:13 +0000 (02:30 +0000)]
Build fix.

* WebKitSystemInterface.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133707 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUpdate Java related WKSI function names
andersca@apple.com [Wed, 7 Nov 2012 01:58:54 +0000 (01:58 +0000)]
Update Java related WKSI function names
https://bugs.webkit.org/show_bug.cgi?id=101414

Reviewed by Sam Weinig.

Source/WebKit/mac:

* WebCoreSupport/WebChromeClient.mm:
(WebChromeClient::unavailablePluginButtonClicked):
* WebCoreSupport/WebFrameLoaderClient.mm:
(isPlugInInactive):

Source/WebKit2:

* UIProcess/Plugins/mac/PluginInfoStoreMac.mm:
(WebKit::PluginInfoStore::shouldUsePlugin):
(WebKit::PluginInfoStore::policyForPlugin):
(WebKit::PluginInfoStore::reactivateInactivePlugin):

WebKitLibraries:

Roll WebKitSystemInterface DEPS.

* WebKitSystemInterface.h:
* libWebKitSystemInterfaceLion.a:
* libWebKitSystemInterfaceMountainLion.a:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133702 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed gardening. Rebaseline the tests added in http://trac.webkit.org/changeset...
aboxhall@chromium.org [Wed, 7 Nov 2012 01:56:29 +0000 (01:56 +0000)]
Unreviewed gardening. Rebaseline the tests added in trac.webkit.org/changeset/133686 for different platforms.

* platform/chromium-linux/fast/backgrounds/transformed-body-background-expected.png: Added.
* platform/chromium-linux/fast/backgrounds/transformed-body-html-background-expected.png: Added.
* platform/chromium-linux/fast/backgrounds/transformed-html-body-background-expected.png: Added.
* platform/chromium-mac-lion/compositing/repaint/composited-document-element-expected.png:
* platform/chromium-mac-lion/platform/chromium/virtual/softwarecompositing/repaint/composited-document-element-expected.png: Added.
* platform/chromium-mac-snowleopard/compositing/repaint/composited-document-element-expected.png:
* platform/chromium-mac-snowleopard/platform/chromium/virtual/softwarecompositing/repaint/composited-document-element-expected.png: Added.
* platform/chromium-mac/compositing/repaint/composited-document-element-expected.png:
* platform/chromium-mac/fast/backgrounds/transformed-body-background-expected.png: Added.
* platform/chromium-mac/fast/backgrounds/transformed-body-html-background-expected.png: Added.
* platform/chromium-mac/fast/backgrounds/transformed-html-body-background-expected.png: Added.
* platform/chromium-mac/platform/chromium/virtual/softwarecompositing/repaint/composited-document-element-expected.png: Added.
* platform/efl-wk1/compositing/repaint/composited-document-element-expected.png: Renamed from LayoutTests/platform/efl/compositing/repaint/composited-document-element-expected.png.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133701 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd application occlusion criterion to enable/disable process suppression on Mac
commit-queue@webkit.org [Wed, 7 Nov 2012 01:17:53 +0000 (01:17 +0000)]
Add application occlusion criterion to enable/disable process suppression on Mac
https://bugs.webkit.org/show_bug.cgi?id=100997

Patch by Kiran Muppala <cmuppala@apple.com> on 2012-11-06
Reviewed by Alexey Proskuryakov.

Provide support for enabling and disabling process suppression when application is completely occluded on Mac.

* NetworkProcess/NetworkProcess.messages.in: Added message to update application occlusion status.
* NetworkProcess/mac/NetworkProcessMainMac.mm:
(WebKit::NetworkProcessMain): Removed call to take process suppression assertion since ChildProcess
takes one on constrution.
* PluginProcess/PluginProcess.messages.in: Added message to update application occlusion status.
* PluginProcess/mac/PluginProcessMainMac.mm:
(WebKit::PluginProcessMain): Removed call to take process suppression assertion since ChildProcess
takes one on construction.
* Shared/ChildProcess.cpp:
(WebKit::ChildProcess::ChildProcess): Take process suppression assertion on construction.
* Shared/ChildProcess.h:
(WebKit::ChildProcess::applicationIsOccluded):
* Shared/mac/ChildProcessMac.mm: Added.
(WebKit::ChildProcess::setApplicationIsOccluded): Release process suppression assertion if application
becomes occluded or take a new assertion if application becomes visible.
(WebKit::ChildProcess::disableProcessSuppression):
(WebKit::ChildProcess::enableProcessSuppression):
Helper methods to take and release process suppression assertion.
* SharedWorkerProcess/SharedWorkerProcess.messages.in: Added message to update application occlusion status.
* SharedWorkerProcess/mac/SharedWorkerProcessMainMac.mm:
(WebKit::SharedWorkerProcessMain): Removed call to take process suppression assertion since ChildProcess
takes one on construction.
* UIProcess/Network/NetworkProcessManager.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::didFinishLaunching): Update application occlusion status post launch.
* UIProcess/Network/NetworkProcessProxy.h:
(WebKit::NetworkProcessProxy::isValid):
* UIProcess/Network/mac/NetworkProcessManagerMac.mm: Added.
(WebKit::NetworkProcessManager::setApplicationIsOccluded):
* UIProcess/Network/mac/NetworkProcessProxyMac.mm:
(WebKit::NetworkProcessProxy::setApplicationIsOccluded):
* UIProcess/Plugins/PluginProcessManager.h:
* UIProcess/Plugins/PluginProcessProxy.cpp:
(WebKit::PluginProcessProxy::didFinishLaunching): Update application occlusion status post launch.
* UIProcess/Plugins/PluginProcessProxy.h:
(WebKit::PluginProcessProxy::isValid):
* UIProcess/Plugins/mac/PluginProcessManagerMac.mm: Added.
(WebKit::PluginProcessManager::setApplicationIsOccluded):
* UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::PluginProcessProxy::setApplicationIsOccluded):
* UIProcess/SharedWorkers/SharedWorkerProcessManager.h:
* UIProcess/SharedWorkers/SharedWorkerProcessProxy.cpp:
(WebKit::SharedWorkerProcessProxy::didFinishLaunching): Update application occlusion status post launch.
* UIProcess/SharedWorkers/SharedWorkerProcessProxy.h:
(WebKit::SharedWorkerProcessProxy::isValid):
* UIProcess/SharedWorkers/mac/SharedWorkerProcessManagerMac.mm: Added.
(WebKit::SharedWorkerProcessManager::setApplicationIsOccluded):
* UIProcess/SharedWorkers/mac/SharedWorkerProcessProxyMac.mm: Added.
(WebKit::SharedWorkerProcessProxy::setApplicationIsOccluded):
* UIProcess/WebContext.cpp:
(WebKit::WebContext::create): Added call to initialize process suppression support.
* UIProcess/WebContext.h:
(WebKit::WebContext::applicationIsOccluded):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didFinishLaunching): Update application occlusion status post launch.
* UIProcess/mac/WebContextMac.mm:
(WebKit::WebContext::applicationBecameVisible):
(WebKit::WebContext::applicationBecameOccluded):
Notification handlers for informing non UI WebKit2 processes of changes to application occlusion state.
(WebKit::WebContext::registerOcclusionNotificationHandlers):
(WebKit::WebContext::initializeProcessSuppressionSupport):
Helper method to register occlusion notification handlers if process suppression support is enabled.
* WebKit2.xcodeproj/project.pbxproj:
* WebProcess/WebProcess.messages.in: Added message to update application occlusion status.
* WebProcess/mac/WebProcessInitialization.mm:
(WebKit::InitializeWebProcess): Removed call to take process supression assertion since ChildProcess
takes one on construction.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133699 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agofast/forms/select-set-length-with-mutation-remove is slow on chromium debug builds
jsbell@chromium.org [Wed, 7 Nov 2012 01:14:01 +0000 (01:14 +0000)]
fast/forms/select-set-length-with-mutation-remove is slow on chromium debug builds
https://bugs.webkit.org/show_bug.cgi?id=60092

[Chromium] Unreviewed gardening - so slow it's timing out at 60s rather often.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133698 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[CSS Exclusions] Changeset r133384 defeated the rectilinear polygon tests
commit-queue@webkit.org [Wed, 7 Nov 2012 01:02:15 +0000 (01:02 +0000)]
[CSS Exclusions] Changeset r133384 defeated the rectilinear polygon tests
https://bugs.webkit.org/show_bug.cgi?id=101405

Patch by Hans Muller <hmuller@adobe.com> on 2012-11-06
Reviewed by Dirk Schulze.

When the exclusions/shape-inside subpixel layout JS code was refactored,
(see https://bugs.webkit.org/show_bug.cgi?id=100704) these two tests were not
properly updated. They now include subpixel-utils.js, which simple-polygon.js depends on.

* fast/exclusions/shape-inside/shape-inside-rectilinear-polygon-001-expected.html:
* fast/exclusions/shape-inside/shape-inside-rectilinear-polygon-001.html:
* fast/exclusions/shape-inside/shape-inside-rectilinear-polygon-002-expected.html:
* fast/exclusions/shape-inside/shape-inside-rectilinear-polygon-002.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133697 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDon't write pixels for crash-hw-sw-switch test; if it doesn't crash, it
senorblanco@chromium.org [Wed, 7 Nov 2012 00:48:19 +0000 (00:48 +0000)]
Don't write pixels for crash-hw-sw-switch test; if it doesn't crash, it
passes.  Unreviewed.

* css3/filters/crash-hw-sw-switch.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133696 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago(r125185) http/tests/images/jpg-img-partial-load.html timeout on Lion
noel.gordon@gmail.com [Wed, 7 Nov 2012 00:46:51 +0000 (00:46 +0000)]
(r125185) http/tests/images/jpg-img-partial-load.html timeout on Lion
https://bugs.webkit.org/show_bug.cgi?id=93636

Reviewed by Simon Fraser.

Remove this test: fast/images/jpeg-partial-load.html is now used to test
partial jpeg image loads.

* http/tests/images/jpg-img-partial-load-expected.png: Removed.
* http/tests/images/jpg-img-partial-load-expected.txt: Removed.
* http/tests/images/jpg-img-partial-load.html: Removed.
* platform/mac/TestExpectations:
* platform/win/TestExpectations:
* platform/wincairo/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoClean up which storage cookie jar functions use
ap@apple.com [Wed, 7 Nov 2012 00:41:30 +0000 (00:41 +0000)]
Clean up which storage cookie jar functions use
https://bugs.webkit.org/show_bug.cgi?id=101395

Reviewed by Brady Eidson.

Source/WebCore:

* WebCore.exp.in:
* platform/mac/WebCoreSystemInterface.h:
* platform/mac/WebCoreSystemInterface.mm:
Updated for two new WKSI functions.

* platform/CookieJar.h: Added explanatory comments.

* platform/mac/CookieJar.mm:
(WebCore::cookies): Removed fallback on NSHTTPCookieStorage, the WKSI function
already implements it.
(WebCore::cookieRequestHeaderFieldValue): Ditto.
(WebCore::setCookies): Ditto.
(WebCore::cookiesEnabled): Ditto.
(WebCore::getRawCookies): Ditto.
(WebCore::deleteCookie): Ditto.
(WebCore::getHostnamesWithCookies): This function used to look at NSHTTPCookieStorage
exclusively. While it makes sense to bypass private storage session, default storage
session (currently only used for testing on Mac) is never OK to bypass. Changed to
use a WKSI function that supports both.
(WebCore::deleteCookiesForHostname): Ditto.
(WebCore::deleteAllCookies): Ditto.

* platform/network/cf/CookieJarCFNet.cpp:
(WebCore::getHostnamesWithCookies):
(WebCore::deleteCookiesForHostname):
(WebCore::deleteAllCookies):
Updated to match Foundation, so now these functions bypass private session.

* platform/network/cf/CookieStorageCFNet.cpp:
(WebCore::currentCFHTTPCookieStorage): Updated a comment.
(WebCore::defaultCFHTTPCookieStorage):  Added a function that returns cookie storage
for non-private session (or null if that would be an NSHTTPCookieStorage).
(WebCore::startObservingCookieChanges): We do not want to observe private cookies
here, and we certainly don't want a mismatch between start/stop due to private
browsing mode changes.
(WebCore::stopObservingCookieChanges): Ditto.

* platform/network/cf/CookieStorageCFNet.h: Added defaultCFHTTPCookieStorage().

Source/WebKit/mac:

* WebCoreSupport/WebSystemInterface.mm: (InitWebCoreSystemInterface): Updated for
two new functions.

Source/WebKit2:

* WebProcess/WebCoreSupport/mac/WebSystemInterface.mm: (InitWebCoreSystemInterface):
Updated for two new functions.

WebKitLibraries:

* WebKitSystemInterface.h:
* libWebKitSystemInterfaceLion.a:
* libWebKitSystemInterfaceMountainLion.a:
Added WKHTTPCookies() and WKDeleteAllHTTPCookies(). Updated ifdefs for currently
supported OS versions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL][WK2] Add ewk_database_manager APIs
commit-queue@webkit.org [Wed, 7 Nov 2012 00:32:54 +0000 (00:32 +0000)]
[EFL][WK2] Add ewk_database_manager APIs
https://bugs.webkit.org/show_bug.cgi?id=101193

Patch by Jihye Kang <jye.kang@samsung.com> on 2012-11-06
Reviewed by Gyuyoung Kim.

Add ewk_database_manager to manage web database and
ewk_database_manaager_origins_get to get origins list of web storage.
Add unit tests for ewk_database_manager APIs.

* PlatformEfl.cmake:
* UIProcess/API/efl/ewk_context.cpp:
(Ewk_Context::Ewk_Context):
(Ewk_Context::databaseManager):
(ewk_context_database_manager_get):
* UIProcess/API/efl/ewk_context.h:
* UIProcess/API/efl/ewk_context_private.h:
(Ewk_Context):
* UIProcess/API/efl/ewk_database_manager.cpp: Added.
(Ewk_Database_Manager::Ewk_Database_Manager):
(Ewk_Database_Manager::getDatabaseOrigins):
(Ewk_Database_Manager::createOriginList):
(Ewk_Database_Origins_Async_Get_Context):
(Ewk_Database_Origins_Async_Get_Context::Ewk_Database_Origins_Async_Get_Context):
(getDatabaseOriginsCallback):
(ewk_database_manager_origins_get):
* UIProcess/API/efl/ewk_database_manager.h: Added.
* UIProcess/API/efl/ewk_database_manager_private.h: Added.
(Ewk_Database_Manager):
(Ewk_Database_Manager::create):
* UIProcess/API/efl/tests/test_ewk2_context.cpp:
(TEST_F):
* UIProcess/API/efl/tests/test_ewk2_database_manager.cpp: Added.
(OriginData):
(OriginData::OriginData):
(getDatabaseOriginsCallback):
(timerCallback):
(TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133692 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Unreviewed gardening. A couple of expectation updates c/o dbarton@mathscri...
jsbell@chromium.org [Wed, 7 Nov 2012 00:29:48 +0000 (00:29 +0000)]
[Chromium] Unreviewed gardening. A couple of expectation updates c/o dbarton@mathscribe.com
and marking calendar-picker-key-operations.html flaky per webkit.org/b/101408

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133691 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRemoved useless image baselines for filter-empty-element-crash.
senorblanco@chromium.org [Wed, 7 Nov 2012 00:16:28 +0000 (00:16 +0000)]
Removed useless image baselines for filter-empty-element-crash.
Unreviewed.

* platform/chromium-linux/css3/filters/filter-empty-element-crash-expected.png: Removed.
* platform/chromium-mac-snowleopard/css3/filters/filter-empty-element-crash-expected.png: Removed.
* platform/chromium-mac/css3/filters/filter-empty-element-crash-expected.png: Removed.
* platform/chromium-win/css3/filters/filter-empty-element-crash-expected.png: Removed.
* platform/mac-snowleopard/css3/filters/filter-empty-element-crash-expected.png: Removed.
* platform/qt-5.0-wk2/css3/filters/filter-empty-element-crash-expected.png: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoBuild fix. r133601 broke the Windows build.
commit-queue@webkit.org [Wed, 7 Nov 2012 00:14:54 +0000 (00:14 +0000)]
Build fix. r133601 broke the Windows build.
https://bugs.webkit.org/show_bug.cgi?id=101404

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-06
Reviewed by Noam Rosenthal.

Only static const integral data members can be initialized within a class in the
Windows, so we initialize static const double date members out of the class.

* platform/graphics/texmap/TextureMapper.cpp:
(BitmapTexturePool):
(WebCore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoReduce parser overhead in JSC
oliver@apple.com [Wed, 7 Nov 2012 00:13:54 +0000 (00:13 +0000)]
Reduce parser overhead in JSC
https://bugs.webkit.org/show_bug.cgi?id=101127

Reviewed by Filip Pizlo.

An exciting journey into the world of architecture in which our hero
adds yet another layer to JSC codegeneration.

This patch adds a marginally more compact form of bytecode that is
free from any data specific to a given execution context, and that
does store any data structures necessary for execution.  To actually
execute this UnlinkedBytecode we still need to instantiate a real
CodeBlock, but this is a much faster linear time operation than any
of the earlier parsing or code generation passes.

As the unlinked code is context free we can then simply use a cache
from source to unlinked code mapping to completely avoid all of the
old parser overhead.  The cache is currently very simple and memory
heavy, using the complete source text as a key (rather than SourceCode
or equivalent), and a random eviction policy.

This seems to produce a substantial win when loading identical content
in different contexts.

* API/tests/testapi.c:
(main):
* CMakeLists.txt:
* GNUmakefile.list.am:
* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
* JavaScriptCore.xcodeproj/project.pbxproj:
* bytecode/CodeBlock.cpp:
* bytecode/CodeBlock.h:
    Moved a number of fields, and a bunch of logic to UnlinkedCodeBlock.h/cpp
* bytecode/Opcode.h:
    Added a global const init no op instruction needed to get correct
    behaviour without any associated semantics.
* bytecode/UnlinkedCodeBlock.cpp: Added.
* bytecode/UnlinkedCodeBlock.h: Added.
    A fairly shallow, GC allocated version of the old CodeBlock
    classes with a 32bit instruction size, and just metadata
    size tracking.
* bytecompiler/BytecodeGenerator.cpp:
* bytecompiler/BytecodeGenerator.h:
    Replace direct access to m_symbolTable with access through
    symbolTable().  ProgramCode no longer has a symbol table at
    all so some previously unconditional (and pointless) uses
    of symbolTable get null checks.
    A few other changes to deal with type changes due to us generating
    unlinked code (eg. pointer free, so profile indices rather than
    pointers).
* dfg/DFGByteCodeParser.cpp:
* dfg/DFGCapabilities.h:
    Support global_init_nop
* interpreter/Interpreter.cpp:
    Now get the ProgramExecutable to initialise new global properties
    before starting execution.
* jit/JIT.cpp:
* jit/JITDriver.h:
* jit/JITStubs.cpp:
* llint/LLIntData.cpp:
* llint/LLIntSlowPaths.cpp:
* llint/LowLevelInterpreter.asm:
* llint/LowLevelInterpreter32_64.asm:
* llint/LowLevelInterpreter64.asm:
    Adding init_global_const_nop everywhere else
* parser/Parser.h:
* parser/ParserModes.h: Added.
* parser/ParserTokens.h:
    Parser no longer needs a global object or callframe to function
* runtime/CodeCache.cpp: Added.
* runtime/CodeCache.h: Added.
    A simple, random eviction, Source->UnlinkedCode cache
* runtime/Executable.cpp:
* runtime/Executable.h:
    Executables now reference their unlinked counterparts, and
    request code specifically for the target global object.
* runtime/JSGlobalData.cpp:
* runtime/JSGlobalData.h:
    GlobalData now owns a CodeCache and a set of new structures
    for the unlinked code types.
* runtime/JSGlobalObject.cpp:
* runtime/JSGlobalObject.h:
    Utility functions used by executables to perform compilation

* runtime/JSType.h:
  Add new JSTypes for unlinked code

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] New baselines for some CSS3 filter tests. Unreviewed.
senorblanco@chromium.org [Wed, 7 Nov 2012 00:11:44 +0000 (00:11 +0000)]
[Chromium] New baselines for some CSS3 filter tests.  Unreviewed.

* platform/chromium-mac/css3/filters/crash-hw-sw-switch-expected.png:
* platform/chromium-mac/css3/filters/effect-reference-expected.png:
* platform/chromium-mac/css3/filters/effect-reference-expected.txt:
* platform/chromium-win/css3/filters/effect-reference-expected.png:
* platform/chromium-win/css3/filters/effect-reference-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago-webkit-background-clip:text produces artifacts when applied to the body and the...
simon.fraser@apple.com [Wed, 7 Nov 2012 00:03:53 +0000 (00:03 +0000)]
-webkit-background-clip:text produces artifacts when applied to the body and the browser is resized
https://bugs.webkit.org/show_bug.cgi?id=89287

Reviewed by Beth Dakin.

Source/WebCore:

If the body had -webkit-background-clip: text, we'd fail to paint
anything behind it, so would see garbage pixels.

Fix by having RenderView::paintBoxDecorations() check for a background-clip of "text"
on the renderer that paints the root background.

Added some new pixel tests for combinations of html and body transform
and backgrounds; earlier versions of the patch broke some of these tests.

Tests: fast/backgrounds/background-clip-text-on-body.html
       fast/backgrounds/transformed-body-background.html
       fast/backgrounds/transformed-body-html-background.html
       fast/backgrounds/transformed-html-body-background.html

* rendering/RenderView.cpp:
(WebCore::rendererObscuresBackground): Broke up the single condition
into early 'false' returns when possible. We need to also check whether
the renderer that actually paints the background (which might be the body)
will fill it; background-clip: text does not.
(WebCore::RenderView::paintBoxDecorations): Rather than checking firstChild(),
actually check the root renderer, so that we can reliably get to the renderer
that mains the root background.

LayoutTests:

Test with -webkit-background-clip: text on the body, which scrolls.

Some additional tests that exercise different combinations of html and
body transform and background.

* compositing/repaint/composited-document-element.html:
* fast/backgrounds/background-clip-text-on-body-expected.png: Added.
* fast/backgrounds/background-clip-text-on-body-expected.txt: Added.
* fast/backgrounds/background-clip-text-on-body.html: Added.
* fast/backgrounds/transformed-body-background-expected.png: Added.
* fast/backgrounds/transformed-body-background-expected.txt: Added.
* fast/backgrounds/transformed-body-background.html: Added.
* fast/backgrounds/transformed-body-html-background-expected.png: Added.
* fast/backgrounds/transformed-body-html-background-expected.txt: Added.
* fast/backgrounds/transformed-body-html-background.html: Added.
* fast/backgrounds/transformed-html-body-background-expected.png: Added.
* fast/backgrounds/transformed-html-body-background-expected.txt: Added.
* fast/backgrounds/transformed-html-body-background.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDon't warn about RefPtr* parameters
oliver@apple.com [Tue, 6 Nov 2012 23:52:34 +0000 (23:52 +0000)]
Don't warn about RefPtr* parameters
https://bugs.webkit.org/show_bug.cgi?id=101403

Reviewed by Gavin Barraclough.

RefPtr<>* is a valid parameter type.

* Scripts/webkitpy/style/checkers/cpp.py:
(check_function_definition_and_pass_ptr):
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
(PassPtrTest.test_ref_ptr_parameter_value):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoImplement save and restore for a bunch of InternalSettings state
senorblanco@chromium.org [Tue, 6 Nov 2012 23:51:53 +0000 (23:51 +0000)]
Implement save and restore for a bunch of InternalSettings state
https://bugs.webkit.org/show_bug.cgi?id=101394

Reviewed by Dirk Pranke.

In http://trac.webkit.org/changeset/133399, I resolved a bunch of
flaky tests by saving and restoring the InternalSettings state
EnableCompositingForFixedPosition.  On a hunch, I git grepped for all
internal settings state which is being set by layout tests but not
saved and restored.  It turns out there about 15 calls.

This patch takes care of five of them, mostly compositing-related.
The others I was either not sure about, or were already at least
being reset to known values by the testing harness.

Hopefully this will alleviate some further test flakiness.

* testing/InternalSettings.cpp:
* testing/InternalSettings.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoFix build warning in RenderLayer.cpp caused by r133628
commit-queue@webkit.org [Tue, 6 Nov 2012 23:50:36 +0000 (23:50 +0000)]
Fix build warning in RenderLayer.cpp caused by r133628
https://bugs.webkit.org/show_bug.cgi?id=101401

Patch by KyungTae Kim <ktf.kim@samsung.com> on 2012-11-06
Reviewed by Simon Fraser.

The static function WebCore::frameViewFromLayer is only used when USE(ACCELERATED_COMPOSITING),
so define it only that case for removing -Wunused-function warning

* rendering/RenderLayer.cpp:
(WebCore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[CSS Exclusions] Store ExclusionPolygonEdge vertices in clockwise order
commit-queue@webkit.org [Tue, 6 Nov 2012 23:47:04 +0000 (23:47 +0000)]
[CSS Exclusions] Store ExclusionPolygonEdge vertices in clockwise order
https://bugs.webkit.org/show_bug.cgi?id=100763

Patch by Hans Muller <hmuller@adobe.com> on 2012-11-06
Reviewed by Darin Adler.

Source/WebCore:

The ExclusionPolygon's internal logic, notably getVertexIntersectionVertices(),
assumes that ExclusionPolygonEdge vertices are stored in clockwise order.
If this is not true, then we construct the polygon's edges in reverse.

To determine if the vertices are in clockwise order, we compare the slope
of the line between the top,left vertex and its previous vertex, with the
the slope of the line of vertices before and after the top,left vertex.
If the slope of the latter is greater, then top,left vertex is located
to its left, and the vertices are clockwise.

Test: fast/exclusions/shape-inside/shape-inside-counterclockwise-polygon.html

* rendering/ExclusionPolygon.cpp:
(WebCore::determinant): Used to decide if the min vertex is on the left side of the line that passes through its adjacent vertices.
(WebCore::ExclusionPolygon::ExclusionPolygon): If the polygon's vertices are specified in counterclockwise order, construct the edges in reverse.
(WebCore::appendIntervalX): Made this an inline since it was only defined as a function for the sake of clarity.

LayoutTests:

Verify that a shape-inside polygon whose vertices are specified in counterclockwise
order is handled correctly.

* fast/exclusions/shape-inside/shape-inside-counterclockwise-polygon-expected.html: Added.
* fast/exclusions/shape-inside/shape-inside-counterclockwise-polygon.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, remove a couple of passing tests from chromium expectations.
dpranke@chromium.org [Tue, 6 Nov 2012 23:44:28 +0000 (23:44 +0000)]
Unreviewed, remove a couple of passing tests from chromium expectations.

* platform/chromium-mac-mountainlion/TestExpectations:
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Unreviewed gardening after r133676.
rakuco@webkit.org [Tue, 6 Nov 2012 23:38:23 +0000 (23:38 +0000)]
[EFL] Unreviewed gardening after r133676.

* platform/efl-wk2/TestExpectations: Unskip a few passing tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Crash in InRegionScrollerPrivate.
commit-queue@webkit.org [Tue, 6 Nov 2012 23:30:04 +0000 (23:30 +0000)]
[BlackBerry] Crash in InRegionScrollerPrivate.
https://bugs.webkit.org/show_bug.cgi?id=101399

Patch by Genevieve Mak <gmak@rim.com> on 2012-11-06
Reviewed by Rob Buis.

Add a pointer to the enclosing document as the pointer to layer
or node along the way may become invalid.
Use document pointer in clearDocumentData.
PR #238469

* Api/InRegionScroller.cpp:
(BlackBerry::WebKit::InRegionScrollerPrivate::clearDocumentData):
* WebKitSupport/InRegionScrollableArea.cpp:
(BlackBerry::WebKit::InRegionScrollableArea::InRegionScrollableArea):
(BlackBerry::WebKit::InRegionScrollableArea::document):
(WebKit):
* WebKitSupport/InRegionScrollableArea.h:
(WebCore):
(InRegionScrollableArea):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Simplify the build system
commit-queue@webkit.org [Tue, 6 Nov 2012 23:25:20 +0000 (23:25 +0000)]
[EFL] Simplify the build system
https://bugs.webkit.org/show_bug.cgi?id=101392

Patch by Laszlo Gombos <l.gombos@samsung.com> on 2012-11-06
Reviewed by Kenneth Rohde Christiansen.

.:

Remove unused cmake variables. In addition there is no longer a need to define
WTF_USE_TEXTURE_MAPPER_GL in the build system as that is now handled in Platform.h
(see r133623).

* Source/cmake/OptionsEfl.cmake:

Source/WebKit:

Use WTF_USE_3D_GRAPHICS to guard the inclusion of surfaces and textmap
directories under WebCore/graphics to be consistent with
WebCore/PlatformEfl.cmake.

* PlatformEfl.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agohttp/tests/inspector-enabled/dynamic-scripts.html is flakey
jsbell@chromium.org [Tue, 6 Nov 2012 23:06:06 +0000 (23:06 +0000)]
http/tests/inspector-enabled/dynamic-scripts.html is flakey
https://bugs.webkit.org/show_bug.cgi?id=100926

Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[WTR] Reset EventSender before running each test.
rakuco@webkit.org [Tue, 6 Nov 2012 23:04:50 +0000 (23:04 +0000)]
[WTR] Reset EventSender before running each test.
https://bugs.webkit.org/show_bug.cgi?id=101353

Reviewed by Simon Fraser.

Contrary to what was done to at least some ports in DumpRenderTree,
EventSender did not currently reset its values in WTR. This caused
flakiness if a test such as editing/selection/fake-doubleclick.html was
run twice in a row, since the click count was not reset across each
run.

* WebKitTestRunner/TestController.cpp:
(WTR::TestController::TestController): Do not create m_eventSenderProxy
in the constructor anymore, it is created in
resetStateToConsistentValues now.
(WTR::TestController::resetStateToConsistentValues): Recreate
m_eventSenderProxy every time this method is run.
* WebKitTestRunner/TestController.h: Make m_eventSenderProxy an OwnPtr
and guard it with a platform #ifdef.
* WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
(WTR::EventSenderProxy::~EventSenderProxy): Add an empty
implementation.
* WebKitTestRunner/mac/EventSenderProxy.mm:
(WTR::EventSenderProxy::~EventSenderProxy): Ditto.
* WebKitTestRunner/qt/EventSenderProxyQt.cpp:
(WTR::EventSenderProxy::~EventSenderProxy): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] platform/chromium-linux/fast/text/international/complex-joining-using...
jsbell@chromium.org [Tue, 6 Nov 2012 22:58:24 +0000 (22:58 +0000)]
[Chromium] platform/chromium-linux/fast/text/international/complex-joining-using-gpos.html flaky on linux debug
https://bugs.webkit.org/show_bug.cgi?id=101377

Unreviewed gardening. Add ImageOnlyFailure expectation, for extra fun.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Update BB10 form theme.
commit-queue@webkit.org [Tue, 6 Nov 2012 22:50:30 +0000 (22:50 +0000)]
[BlackBerry] Update BB10 form theme.
https://bugs.webkit.org/show_bug.cgi?id=100760

Patch by Tiancheng Jiang <tijiang@rim.com> on 2012-11-06
Reviewed by Rob Buis.

RIM PR 238123
Internal Reviewed by Chris Hutten-Czapski
BB10 selection theme use slice image function to draw selection button,
any border-radius setting other than 0 will make the button corner
clipped. Always set button border-radius: 0 for BB10 theme.

* css/themeBlackBerry.css:
(select[size][multiple]):
(select[size="1"]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL][WK2] Refactor Ewk_Auth_Request, Ewk_Url_Scheme_Request and Ewk_Download_Job...
commit-queue@webkit.org [Tue, 6 Nov 2012 22:21:23 +0000 (22:21 +0000)]
[EFL][WK2] Refactor Ewk_Auth_Request, Ewk_Url_Scheme_Request and Ewk_Download_Job to be Ewk_Objects
https://bugs.webkit.org/show_bug.cgi?id=101383

Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-11-06
Reviewed by Kenneth Rohde Christiansen.

Ewk_Auth_Request, Ewk_Url_Scheme_Request and Ewk_Download_Job are Ewk_Objects now.

Source/WebKit2:

* UIProcess/API/efl/EwkViewCallbacks.h:
* UIProcess/API/efl/EwkViewImpl.h:
* UIProcess/API/efl/ewk_auth_request.cpp:
(EwkAuthRequest::EwkAuthRequest):
(EwkAuthRequest::~EwkAuthRequest):
(EwkAuthRequest::suggestedUsername):
(EwkAuthRequest::realm):
(EwkAuthRequest::host):
(EwkAuthRequest::continueWithoutCredential):
(EwkAuthRequest::authenticate):
(EwkAuthRequest::isRetrying):
(ewk_auth_request_suggested_username_get):
(ewk_auth_request_cancel):
(ewk_auth_request_authenticate):
(ewk_auth_request_retrying_get):
(ewk_auth_request_realm_get):
(ewk_auth_request_host_get):
* UIProcess/API/efl/ewk_auth_request.h:
* UIProcess/API/efl/ewk_auth_request_private.h:
(EwkAuthRequest):
(EwkAuthRequest::create):
* UIProcess/API/efl/ewk_back_forward_list.h:
* UIProcess/API/efl/ewk_context_private.h:
(EwkContext):
* UIProcess/API/efl/ewk_download_job.cpp:
(EwkDownloadJob::EwkDownloadJob):
(EwkDownloadJob::id):
(EwkDownloadJob::viewImpl):
(ewk_download_job_state_get):
(EwkDownloadJob::state):
(ewk_download_job_request_get):
(EwkDownloadJob::request):
(ewk_download_job_response_get):
(EwkDownloadJob::response):
(ewk_download_job_destination_get):
(EwkDownloadJob::destination):
(ewk_download_job_destination_set):
(EwkDownloadJob::setDestination):
(ewk_download_job_suggested_filename_get):
(EwkDownloadJob::suggestedFileName):
(ewk_download_job_cancel):
(EwkDownloadJob::cancel):
(ewk_download_job_estimated_progress_get):
(EwkDownloadJob::estimatedProgress):
(ewk_download_job_elapsed_time_get):
(EwkDownloadJob::elapsedTime):
(EwkDownloadJob::setResponse):
(EwkDownloadJob::setSuggestedFileName):
(EwkDownloadJob::incrementReceivedData):
(EwkDownloadJob::setState):
* UIProcess/API/efl/ewk_download_job.h:
* UIProcess/API/efl/ewk_download_job_private.h:
(EwkDownloadJob):
(EwkDownloadJob::create):
* UIProcess/API/efl/ewk_url_scheme_request.cpp:
(EwkUrlSchemeRequest::EwkUrlSchemeRequest):
(EwkUrlSchemeRequest::id):
(EwkUrlSchemeRequest::url):
(EwkUrlSchemeRequest::scheme):
(EwkUrlSchemeRequest::path):
(EwkUrlSchemeRequest::finish):
(ewk_url_scheme_request_scheme_get):
(ewk_url_scheme_request_url_get):
(ewk_url_scheme_request_path_get):
(ewk_url_scheme_request_finish):
* UIProcess/API/efl/ewk_url_scheme_request.h:
* UIProcess/API/efl/ewk_url_scheme_request_private.h:
(EwkUrlSchemeRequest):
(EwkUrlSchemeRequest::create):
* UIProcess/API/efl/tests/test_ewk2_auth_request.cpp:
(onAuthenticationRequest):
(TEST_F):
* UIProcess/efl/DownloadManagerEfl.cpp:
(WebKit::DownloadManagerEfl::decideDestinationWithSuggestedFilename):
(WebKit::DownloadManagerEfl::didReceiveResponse):
(WebKit::DownloadManagerEfl::didCreateDestination):
(WebKit::DownloadManagerEfl::didReceiveData):
(WebKit::DownloadManagerEfl::didFail):
(WebKit::DownloadManagerEfl::didCancel):
(WebKit::DownloadManagerEfl::didFinish):
(WebKit::DownloadManagerEfl::registerDownload):
(WebKit::DownloadManagerEfl::downloadJob):
* UIProcess/efl/DownloadManagerEfl.h:
(DownloadManagerEfl):
* UIProcess/efl/PageLoadClientEfl.cpp:
(WebKit::PageLoadClientEfl::didReceiveAuthenticationChallengeInFrame):
* UIProcess/efl/RequestManagerClientEfl.cpp:
(WebKit::RequestManagerClientEfl::didReceiveURIRequest):

Tools:

* MiniBrowser/efl/main.c:
(auth_popup_close):
(on_authentication_request):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoJSStringCreateWithCFString() Should create an 8 bit String if possible
msaboff@apple.com [Tue, 6 Nov 2012 22:16:36 +0000 (22:16 +0000)]
JSStringCreateWithCFString() Should create an 8 bit String if possible
https://bugs.webkit.org/show_bug.cgi?id=101104

Reviewed by Darin Adler.

Try converting the CFString to an 8 bit string using CFStringGetBytes(...,
kCFStringEncodingISOLatin1, ...) and return the 8 bit string if successful.
If not proceed with 16 bit conversion.

* API/JSStringRefCF.cpp:
(JSStringCreateWithCFString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRemove branch from inside RenderObject::view now that renderer() is more expensive
commit-queue@webkit.org [Tue, 6 Nov 2012 22:12:14 +0000 (22:12 +0000)]
Remove branch from inside RenderObject::view now that renderer() is more expensive
https://bugs.webkit.org/show_bug.cgi?id=101277

Patch by Elliott Sprehn <esprehn@chromium.org> on 2012-11-06
Reviewed by Eric Seidel.

It was observed in Bug 100057 that calling renderer() repeatedly now that it has a branch
can be a performance regression. Now that we no longer keep a separate pointer for rare data
in Document, we can use that space for a pointer to the RenderView making RenderObject::view()
faster and removing the branch.

This is a 1% improvement on Parser/html5-full-render.html

This also cleans up the code because it turns out we don't need to have RenderObject::view() in
RenderView.h because we can just call Document::renderView() and not do toRenderView. This makes
it easier to find this method as it exists in the right header file now.

No new tests, this is just a refactor.

* WebCore.exp.in: Remove export of Document::renderView since it's inline now.
* dom/Document.cpp:
(WebCore::Document::Document):
(WebCore::Document::setRenderer):
(WebCore):
* dom/Document.h:
(WebCore::Document::renderView):
(Document):
* rendering/RenderObject.h:
(WebCore::RenderObject::view):
* rendering/RenderView.h:
(WebCore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago Delete Leopard and Snow Leopard versions of WKSI. These have not been updated...
ap@apple.com [Tue, 6 Nov 2012 22:06:50 +0000 (22:06 +0000)]
    Delete Leopard and Snow Leopard versions of WKSI. These have not been updated in ages.

        Rubber-stampted by Mark Rowe.

        * libWebKitSystemInterfaceLeopard.a: Removed.
        * libWebKitSystemInterfaceSnowLeopard.a: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoReduce direct m_symbolTable usage in CodeBlock
oliver@apple.com [Tue, 6 Nov 2012 22:05:49 +0000 (22:05 +0000)]
Reduce direct m_symbolTable usage in CodeBlock
https://bugs.webkit.org/show_bug.cgi?id=101391

Reviewed by Sam Weinig.

Simple refactoring.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::dump):
(JSC::CodeBlock::dumpStatistics):
(JSC::CodeBlock::nameForRegister):
* bytecode/CodeBlock.h:
(JSC::CodeBlock::isCaptured):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoLexer::scanRegExp, create 8 bit pattern and flag Identifiers from 16 bit source when...
msaboff@apple.com [Tue, 6 Nov 2012 22:01:51 +0000 (22:01 +0000)]
Lexer::scanRegExp, create 8 bit pattern and flag Identifiers from 16 bit source when possible
https://bugs.webkit.org/show_bug.cgi?id=101013

Reviewed by Darin Adler.

Changed scanRegExp so that it will create 8 bit identifiers from 8 bit sources and from 16 bit sources
whan all the characters are 8 bit.  Using two templated helpers, the "is all 8 bit" check is only performed
on 16 bit sources.  The first helper is orCharacter() that will accumulate the or value of all characters
only for 16 bit sources.  Replaced the helper Lexer::makeIdentifierSameType() with Lexer::makeRightSizedIdentifier().

* parser/Lexer.cpp:
(JSC::orCharacter<LChar>): Explicit template that serves as a placeholder.
(JSC::orCharacter<UChar>): Explicit template that actually or accumulates characters.
(JSC::Lexer::scanRegExp):
* parser/Lexer.h:
(Lexer):
(JSC::Lexer::makeRightSizedIdentifier<LChar>): New template that always creates an 8 bit Identifier.
(JSC::Lexer::makeRightSizedIdentifier<UChar>): New template that creates an 8 bit Identifier for 8 bit
data in a 16 bit source.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Pushing window.external.* events to webplatform
commit-queue@webkit.org [Tue, 6 Nov 2012 22:00:23 +0000 (22:00 +0000)]
[BlackBerry] Pushing window.external.* events to webplatform
https://bugs.webkit.org/show_bug.cgi?id=101386

Patch by Otto Derek Cheung <otcheung@rim.com> on 2012-11-06
Reviewed by Rob Buis.

Adding calls to trigger window.external events so we can handle them on the
browser side.

PR 170059

* Api/WebPageClient.h:
* WebCoreSupport/ChromeClientBlackBerry.cpp:
(WebCore::ChromeClientBlackBerry::addSearchProvider):
(WebCore::ChromeClientBlackBerry::isSearchProviderInstalled):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r133526.
jsbell@chromium.org [Tue, 6 Nov 2012 21:37:47 +0000 (21:37 +0000)]
Unreviewed, rolling out r133526.
http://trac.webkit.org/changeset/133526
https://bugs.webkit.org/show_bug.cgi?id=101388

May have caused 15% memory regression in Chromium (Requested
by jsbell|gardener on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-06

Source/WebCore:

* UseV8.cmake:
* WebCore.gypi:
* bindings/scripts/CodeGeneratorV8.pm:
(GenerateHeader):
(GenerateNormalAttrGetter):
(GenerateConstructorCallback):
(GenerateNamedConstructorCallback):
(GenerateToV8Converters):
(GetDomMapFunction):
(GetDomMapName):
* bindings/scripts/test/V8/V8Float64Array.h:
(WebCore::V8Float64Array::wrap):
* bindings/scripts/test/V8/V8TestActiveDOMObject.h:
(WebCore::V8TestActiveDOMObject::wrap):
* bindings/scripts/test/V8/V8TestCustomNamedGetter.h:
(WebCore::V8TestCustomNamedGetter::wrap):
* bindings/scripts/test/V8/V8TestEventConstructor.h:
(WebCore::V8TestEventConstructor::wrap):
* bindings/scripts/test/V8/V8TestEventTarget.h:
(WebCore::V8TestEventTarget::wrap):
* bindings/scripts/test/V8/V8TestException.h:
(WebCore::V8TestException::wrap):
* bindings/scripts/test/V8/V8TestInterface.h:
(WebCore::V8TestInterface::wrap):
* bindings/scripts/test/V8/V8TestMediaQueryListListener.h:
(WebCore::V8TestMediaQueryListListener::wrap):
* bindings/scripts/test/V8/V8TestNamedConstructor.h:
(WebCore::V8TestNamedConstructor::wrap):
* bindings/scripts/test/V8/V8TestNode.cpp:
(WebCore::V8TestNode::constructorCallback):
(WebCore::V8TestNode::wrapSlow):
* bindings/scripts/test/V8/V8TestObj.cpp:
(WebCore::TestObjV8Internal::readOnlyTestObjAttrAttrGetter):
* bindings/scripts/test/V8/V8TestObj.h:
(WebCore::V8TestObj::wrap):
* bindings/scripts/test/V8/V8TestSerializedScriptValueInterface.h:
(WebCore::V8TestSerializedScriptValueInterface::wrap):
* bindings/v8/DOMDataStore.cpp:
(WebCore::DOMDataStore::DOMDataStore):
(WebCore::DOMDataStore::~DOMDataStore):
(WebCore::DOMDataStore::current):
(WebCore::DOMDataStore::reportMemoryUsage):
* bindings/v8/DOMDataStore.h:
(WebCore::DOMDataStore::domNodeMap):
(WebCore::DOMDataStore::domObjectMap):
(DOMDataStore):
* bindings/v8/DOMWrapperMap.h:
(WebCore):
(DOMWrapperMap):
(WebCore::DOMWrapperMap::~DOMWrapperMap):
(WebCore::DOMWrapperHashMap::remove):
* bindings/v8/DOMWrapperWorld.h:
* bindings/v8/IntrusiveDOMWrapperMap.h: Copied from Source/WebKit/chromium/src/WebScriptController.cpp.
(WebCore):
(IntrusiveDOMWrapperMap):
(WebCore::IntrusiveDOMWrapperMap::weakCallback):
* bindings/v8/ScriptProfiler.cpp:
* bindings/v8/SerializedScriptValue.cpp:
(WebCore::neuterBinding):
* bindings/v8/V8DOMMap.cpp: Copied from Source/WebKit/chromium/src/WebScriptController.cpp.
(WebCore):
(WebCore::getDOMNodeMap):
(WebCore::getDOMObjectMap):
* bindings/v8/V8DOMMap.h: Copied from Source/WebKit/chromium/src/WebScriptController.cpp.
(WebCore):
* bindings/v8/V8DOMWindowShell.cpp:
* bindings/v8/V8DOMWrapper.cpp:
(WebCore::V8DOMWrapper::setJSWrapperForDOMNode):
(WebCore):
* bindings/v8/V8DOMWrapper.h:
(V8DOMWrapper):
(WebCore::V8DOMWrapper::getCachedWrapper):
(WebCore::V8DOMWrapper::setJSWrapperForDOMObject):
* bindings/v8/V8GCController.cpp:
* bindings/v8/V8NPObject.cpp:
* bindings/v8/WorkerContextExecutionProxy.cpp:
* bindings/v8/WorkerScriptController.cpp:
* bindings/v8/custom/V8HTMLImageElementConstructor.cpp:
(WebCore::v8HTMLImageElementConstructorCallback):

Source/WebKit/chromium:

* src/WebScriptController.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSome cookie code cleanup
ap@apple.com [Tue, 6 Nov 2012 21:24:37 +0000 (21:24 +0000)]
Some cookie code cleanup
https://bugs.webkit.org/show_bug.cgi?id=101375

Reviewed by Dan Bernstein.

* platform/CookieJar.h: Fixed style. Grouped functions by which storage they operate
on in Mac port, but didn't add explanatory comments yet, because this is different
in some ports.

* platform/mac/CookieJar.mm:
(WebCore): Removed special code for isHTTPOnly, it's present in Foundation in all
supported OS X versions.
(WebCore::filterCookies): Use -isHTTPOnly directly.
(WebCore::cookies): Get rid of a variable for URL, implicit conversion works just as well.
(WebCore::cookieRequestHeaderFieldValue): Ditto.
(WebCore::setCookies): Assert that no more than one cookie was created from one
Set-Cookie header, document.cookie can only be used to set one cookie at a time.
(WebCore::getRawCookies): Removed useless local variables.

* platform/network/CookieStorage.h: Removed an unneeded include, clarified a comment.

* platform/network/HTTPHeaderMap.h: Added a FIXME.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed chromium, chromium-mac-mountainlion gardening
dpranke@chromium.org [Tue, 6 Nov 2012 21:19:59 +0000 (21:19 +0000)]
Unreviewed chromium, chromium-mac-mountainlion gardening

Remove a bunch of passing tests, and clean up a couple of ML issues.

* platform/chromium-mac-lion/media/track/track-cue-rendering-horizontal-expected.png: Removed.
* platform/chromium-mac-lion/media/track/track-cue-rendering-horizontal-expected.txt: Removed.
* platform/chromium-mac/media/track/track-cue-rendering-horizontal-expected.png:
* platform/chromium-mac/media/track/track-cue-rendering-horizontal-expected.txt:
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIndentation of JSCell.h is wrong
fpizlo@apple.com [Tue, 6 Nov 2012 21:13:27 +0000 (21:13 +0000)]
Indentation of JSCell.h is wrong
https://bugs.webkit.org/show_bug.cgi?id=101379

Rubber stamped by Alexey Proskuryakov.

Just removed four spaces on a bunch of lines.

* runtime/JSCell.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoImplement MediaPlayerPrivate::didLoadingProgress for BlackBerry platform
commit-queue@webkit.org [Tue, 6 Nov 2012 21:08:35 +0000 (21:08 +0000)]
Implement MediaPlayerPrivate::didLoadingProgress for BlackBerry platform
https://bugs.webkit.org/show_bug.cgi?id=100378

Patch by John Griggs <jgriggs@rim.com> on 2012-11-06
Reviewed by Eric Carlson.

Implement this method for BlackBerry platform.

New test: LayoutTests/media/progress-events-generated-correctly.html

Source/WebCore:

* platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
(WebCore::MediaPlayerPrivate::MediaPlayerPrivate):
(WebCore::MediaPlayerPrivate::didLoadingProgress):
* platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.h:
(MediaPlayerPrivate):

LayoutTests:

* media/progress-events-generated-correctly-expected.txt: Added.
* media/progress-events-generated-correctly.html: Added.
* platform/chromium/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[WK2] Wrong scroll position when going back in history
commit-queue@webkit.org [Tue, 6 Nov 2012 20:44:40 +0000 (20:44 +0000)]
[WK2] Wrong scroll position when going back in history
https://bugs.webkit.org/show_bug.cgi?id=101358

Patch by Yael Aharon <yael.aharon@intel.com> on 2012-11-06
Reviewed by Kenneth Rohde Christiansen.

Reset the contentsSize after a new load was committed. We should not continue
to use the contentsSize of the previous page in any calculation.

* UIProcess/PageViewportController.cpp:
(WebKit::PageViewportController::didCommitLoad):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133659 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDOMImplementation should use ScriptWrappable
abarth@webkit.org [Tue, 6 Nov 2012 20:25:52 +0000 (20:25 +0000)]
DOMImplementation should use ScriptWrappable
https://bugs.webkit.org/show_bug.cgi?id=101279

Reviewed by Eric Seidel.

PerformanceTests:

This simple PerformanceTest measure how quickly we can access
document.implementation. The goal of this performance test is just to
validate that using ScriptWrappable actually makes the bindings faster.

* Bindings/document-implementation.html: Added.

Source/WebCore:

This patch makes DOMImplementation ScriptWrappable. Both the V8 and JSC
bindings automatically detect the presence of this base class and use
it to optimize getting and setting JavaScript wrappers.

This patch is a 23% progression on Bindings/document-implementation for
the V8 bindings and a 69% progression for the JSC bindings.

* dom/DOMImplementation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133657 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Rolled DEPS.
commit-queue@webkit.org [Tue, 6 Nov 2012 20:10:57 +0000 (20:10 +0000)]
Unreviewed.  Rolled DEPS.

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-06

* DEPS:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSpeed up TransformationMatrix::multiply() on modern ARM
benjamin@webkit.org [Tue, 6 Nov 2012 20:10:46 +0000 (20:10 +0000)]
Speed up TransformationMatrix::multiply() on modern ARM
https://bugs.webkit.org/show_bug.cgi?id=101084

Reviewed by Gavin Barraclough.

Source/WebCore:

The main improvements are:
-Store the full source matrix in the registers d16 to d31. This way we avoid going
 back and forth to memory to use the operands.
 Since the full matrix is in register, we can also directly modify the memory of m_matrix,
 we no longer need the temporary matrix memcopied at the end.
-Use of LoadMultiple and StoreMultiple to load-store the matrix parameters.
-Use Multiply-Accumulate instead of VMUL followed by VADD. This half the number of instruction.
-On regular ARMv7, using a loop to reuse the same code for each row also improve the performance.

Depending on the hardware, the new code takes 20% to 42% less time than the basic implementation.

* platform/graphics/transforms/TransformationMatrix.cpp:
(WebCore::TransformationMatrix::multiply):
* platform/graphics/transforms/TransformationMatrix.h:
(TransformationMatrix):

Source/WTF:

* wtf/Platform.h:
Add CPU(ARM_VFP) for detecting VFP availability.
Add CPU(APPLE_ARMV7S) for the Apple ARMv7S architecture.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Unreviewed gardening.
jsbell@chromium.org [Tue, 6 Nov 2012 19:24:34 +0000 (19:24 +0000)]
[Chromium] Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIndentation of JSObject.h is wrong
fpizlo@apple.com [Tue, 6 Nov 2012 19:19:09 +0000 (19:19 +0000)]
Indentation of JSObject.h is wrong
https://bugs.webkit.org/show_bug.cgi?id=101313

Rubber stamped by Alexey Proskuryakov.

Just unindented code, since namespace bodies shouldn't be indented.

* runtime/JSObject.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[CSS Regions] Fix failing tests on Mac after sub-pixel layout was enabled
commit-queue@webkit.org [Tue, 6 Nov 2012 19:18:36 +0000 (19:18 +0000)]
[CSS Regions] Fix failing tests on Mac after sub-pixel layout was enabled
https://bugs.webkit.org/show_bug.cgi?id=101364

Patch by Andrei Bucur <abucur@adobe.com> on 2012-11-06
Reviewed by Simon Fraser.

There are no changes in the code. Only the tests had to be fixed.

* fast/regions/float-pushed-width-change-expected.html: Remove the unnecessary 0.5px margin.
* fast/regions/floats-basic-in-variable-width-regions-expected.html: Because of how regions are implemented, the left offset for lines is 50px (the float width) - X.5px (margin: auto gives a different value in the two regions). This translated into 37.5px which triggered a sub-pixel layout issue. The new width values changes the X.5px into a rounded value.
* fast/regions/floats-basic-in-variable-width-regions.html: See above
* fast/regions/render-region-renderer-expected.html: Remove unnecessary 0.5px height.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133648 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] platform/chromium-linux/fast/text/international/complex-joining-using...
jsbell@chromium.org [Tue, 6 Nov 2012 19:15:32 +0000 (19:15 +0000)]
[Chromium] platform/chromium-linux/fast/text/international/complex-joining-using-gpos.html flaky on linux debug
https://bugs.webkit.org/show_bug.cgi?id=101377

Unreviewed gardening. Add new expectation, remove expectation for reverted change.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoFix ChangeLog comment: r133645 was rubber stamped by, not reviwed by, Alexey.
fpizlo@apple.com [Tue, 6 Nov 2012 19:13:52 +0000 (19:13 +0000)]
Fix ChangeLog comment: r133645 was rubber stamped by, not reviwed by, Alexey.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIndentation of JSArray.h is wrong
fpizlo@apple.com [Tue, 6 Nov 2012 19:10:37 +0000 (19:10 +0000)]
Indentation of JSArray.h is wrong
https://bugs.webkit.org/show_bug.cgi?id=101314

Reviewed by Alexey Proskuryakov.

Just removing the indentation inside the namespace body.

* runtime/JSArray.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r133529 and r133562.
commit-queue@webkit.org [Tue, 6 Nov 2012 19:10:09 +0000 (19:10 +0000)]
Unreviewed, rolling out r133529 and r133562.
http://trac.webkit.org/changeset/133529
http://trac.webkit.org/changeset/133562
https://bugs.webkit.org/show_bug.cgi?id=101371

Caused 15% memory regression on Chromium page cyclers
(Requested by jsbell|gardener on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-06

Source/WebCore:

* platform/text/LineBreakIteratorPoolICU.h:
(WebCore::LineBreakIteratorPool::take):
(WebCore::LineBreakIteratorPool::put):
(LineBreakIteratorPool):
* platform/text/TextBreakIterator.cpp:
(WebCore::acquireLineBreakIterator):
* platform/text/TextBreakIterator.h:
(WebCore):
(WebCore::LazyLineBreakIterator::LazyLineBreakIterator):
(WebCore::LazyLineBreakIterator::get):
(WebCore::LazyLineBreakIterator::reset):
(LazyLineBreakIterator):
* platform/text/TextBreakIteratorICU.cpp:
(WebCore::acquireLineBreakIterator):
(WebCore::releaseLineBreakIterator):
* platform/text/gtk/TextBreakIteratorGtk.cpp:
(WebCore::setUpIterator):
(WebCore::acquireLineBreakIterator):
* platform/text/wince/TextBreakIteratorWinCE.cpp:
(WebCore::acquireLineBreakIterator):
* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::LineBreaker::nextLineBreak):
* rendering/RenderText.cpp:
(WebCore::RenderText::computePreferredLogicalWidths):
* rendering/RenderText.h:
(WebCore):
* rendering/break_lines.cpp:
(WebCore):
(WebCore::isBreakableSpace):
(WebCore::needsLineBreakIterator):
(WebCore::nextBreakablePosition):
(WebCore::nextBreakablePositionIgnoringNBSP):
* rendering/break_lines.h:
(WebCore):
(WebCore::isBreakable):

LayoutTests:

* css3/line-break/line-break-auto-centered-expected.html: Removed.
* css3/line-break/line-break-auto-centered.html: Removed.
* css3/line-break/line-break-auto-half-kana-expected.html: Removed.
* css3/line-break/line-break-auto-half-kana.html: Removed.
* css3/line-break/line-break-auto-hyphens-expected.html: Removed.
* css3/line-break/line-break-auto-hyphens.html: Removed.
* css3/line-break/line-break-auto-inseparables-expected.html: Removed.
* css3/line-break/line-break-auto-inseparables.html: Removed.
* css3/line-break/line-break-auto-iteration-marks-expected.html: Removed.
* css3/line-break/line-break-auto-iteration-marks.html: Removed.
* css3/line-break/line-break-auto-postfixes-expected.html: Removed.
* css3/line-break/line-break-auto-postfixes.html: Removed.
* css3/line-break/line-break-auto-prefixes-expected.html: Removed.
* css3/line-break/line-break-auto-prefixes.html: Removed.
* css3/line-break/line-break-auto-sound-marks-expected.html: Removed.
* css3/line-break/line-break-auto-sound-marks.html: Removed.
* css3/line-break/line-break-loose-centered-expected.html: Removed.
* css3/line-break/line-break-loose-centered.html: Removed.
* css3/line-break/line-break-loose-half-kana-expected.html: Removed.
* css3/line-break/line-break-loose-half-kana.html: Removed.
* css3/line-break/line-break-loose-hyphens-expected.html: Removed.
* css3/line-break/line-break-loose-hyphens.html: Removed.
* css3/line-break/line-break-loose-inseparables-expected.html: Removed.
* css3/line-break/line-break-loose-inseparables.html: Removed.
* css3/line-break/line-break-loose-iteration-marks-expected.html: Removed.
* css3/line-break/line-break-loose-iteration-marks.html: Removed.
* css3/line-break/line-break-loose-postfixes-expected.html: Removed.
* css3/line-break/line-break-loose-postfixes.html: Removed.
* css3/line-break/line-break-loose-prefixes-expected.html: Removed.
* css3/line-break/line-break-loose-prefixes.html: Removed.
* css3/line-break/line-break-loose-sound-marks-expected.html: Removed.
* css3/line-break/line-break-loose-sound-marks.html: Removed.
* css3/line-break/line-break-normal-centered-expected.html: Removed.
* css3/line-break/line-break-normal-centered.html: Removed.
* css3/line-break/line-break-normal-half-kana-expected.html: Removed.
* css3/line-break/line-break-normal-half-kana.html: Removed.
* css3/line-break/line-break-normal-hyphens-expected.html: Removed.
* css3/line-break/line-break-normal-hyphens.html: Removed.
* css3/line-break/line-break-normal-inseparables-expected.html: Removed.
* css3/line-break/line-break-normal-inseparables.html: Removed.
* css3/line-break/line-break-normal-iteration-marks-expected.html: Removed.
* css3/line-break/line-break-normal-iteration-marks.html: Removed.
* css3/line-break/line-break-normal-postfixes-expected.html: Removed.
* css3/line-break/line-break-normal-postfixes.html: Removed.
* css3/line-break/line-break-normal-prefixes-expected.html: Removed.
* css3/line-break/line-break-normal-prefixes.html: Removed.
* css3/line-break/line-break-normal-sound-marks-expected.html: Removed.
* css3/line-break/line-break-normal-sound-marks.html: Removed.
* css3/line-break/line-break-strict-centered-expected.html: Removed.
* css3/line-break/line-break-strict-centered.html: Removed.
* css3/line-break/line-break-strict-half-kana-expected.html: Removed.
* css3/line-break/line-break-strict-half-kana.html: Removed.
* css3/line-break/line-break-strict-hyphens-expected.html: Removed.
* css3/line-break/line-break-strict-hyphens.html: Removed.
* css3/line-break/line-break-strict-inseparables-expected.html: Removed.
* css3/line-break/line-break-strict-inseparables.html: Removed.
* css3/line-break/line-break-strict-iteration-marks-expected.html: Removed.
* css3/line-break/line-break-strict-iteration-marks.html: Removed.
* css3/line-break/line-break-strict-postfixes-expected.html: Removed.
* css3/line-break/line-break-strict-postfixes.html: Removed.
* css3/line-break/line-break-strict-prefixes-expected.html: Removed.
* css3/line-break/line-break-strict-prefixes.html: Removed.
* css3/line-break/line-break-strict-sound-marks-expected.html: Removed.
* css3/line-break/line-break-strict-sound-marks.html: Removed.
* platform/chromium-android/css3/line-break/line-break-auto-half-kana-expected.html: Removed.
* platform/chromium-android/css3/line-break/line-break-auto-sound-marks-expected.html: Removed.
* platform/chromium/css3/line-break/line-break-auto-half-kana-expected.html: Removed.
* platform/chromium/css3/line-break/line-break-auto-sound-marks-expected.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133644 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL][WK2] Declare all smart callbacks in EwkViewCallbacks.h
commit-queue@webkit.org [Tue, 6 Nov 2012 19:08:35 +0000 (19:08 +0000)]
[EFL][WK2] Declare all smart callbacks in EwkViewCallbacks.h
https://bugs.webkit.org/show_bug.cgi?id=101360

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-11-06
Reviewed by Kenneth Rohde Christiansen.

Declare all EwkView smart callbacks in EwkViewCallbacks.h
and remove remaining calls to evas_object_smart_callback_call().
It is a good idea to have all the callbacks declared in one
place for convenience and consistency. We also have better
argument type checking if we use the new callback mechanism.

* UIProcess/API/efl/EwkViewCallbacks.h:
(EwkViewCallbacks::CallBack::call):
(CallBack):
(EwkViewCallbacks):
* UIProcess/API/efl/EwkViewImpl.cpp:
(EwkViewImpl::informIconChange):
(EwkViewImpl::informURLChange):
(EwkViewImpl::createNewPage):
(EwkViewImpl::closePage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Enable the JavaScript i18n API for Android
peter@chromium.org [Tue, 6 Nov 2012 19:04:26 +0000 (19:04 +0000)]
[Chromium] Enable the JavaScript i18n API for Android
https://bugs.webkit.org/show_bug.cgi?id=101357

Reviewed by Adam Barth.

Source/WebCore:

Remove the conditional dependency on the v8-i18n API. The library is now
always available in a Chromium for Android checkout.

* WebCore.gyp/WebCore.gyp:

Source/WebKit/chromium:

Enable the JavaScript i18n API for Chromium on Android. This adds a
grand total of 18.7 KiB of data to the content shell apk.

* features.gypi:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, update chromium-mac-mountainlion expectations to remove passing tests.
dpranke@chromium.org [Tue, 6 Nov 2012 19:03:31 +0000 (19:03 +0000)]
Unreviewed, update chromium-mac-mountainlion expectations to remove passing tests.

* platform/chromium-mac-mountainlion/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDon't dump pixels for filter-empty-element-crash.
senorblanco@chromium.org [Tue, 6 Nov 2012 18:58:01 +0000 (18:58 +0000)]
Don't dump pixels for filter-empty-element-crash.
https://bugs.webkit.org/show_bug.cgi?id=101359

Reviewed by Adrienne Walker.

* css3/filters/filter-empty-element-crash.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago(r133351): Number of test failures from enabling SUBPIXEL_LAYOUT on Mac
simon.fraser@apple.com [Tue, 6 Nov 2012 18:51:00 +0000 (18:51 +0000)]
(r133351): Number of test failures from enabling SUBPIXEL_LAYOUT on Mac
https://bugs.webkit.org/show_bug.cgi?id=101220

New baselines for two more tests whose output was affected by enabling subpixel layout.

* platform/mac/fast/dynamic/subtree-parent-static-y-expected.txt:
* platform/mac/fast/images/repaint-subrect-grid-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd new test cases to layout test effect-reference and effect-reference-hw.
senorblanco@chromium.org [Tue, 6 Nov 2012 18:43:04 +0000 (18:43 +0000)]
Add new test cases to layout test effect-reference and effect-reference-hw.
https://bugs.webkit.org/show_bug.cgi?id=101366

Reviewed by Dean Jackson.

* css3/filters/effect-reference-hw.html:
* css3/filters/effect-reference.html:
Add some test cases for feBlend and feSpecularLighting.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoHook up the remote layer tree drawing area to an environment variable
andersca@apple.com [Tue, 6 Nov 2012 18:42:50 +0000 (18:42 +0000)]
Hook up the remote layer tree drawing area to an environment variable
https://bugs.webkit.org/show_bug.cgi?id=101267

Reviewed by Andreas Kling.

* UIProcess/API/mac/WKView.mm:
* UIProcess/mac/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::sizeDidChange):
(WebKit::RemoteLayerTreeDrawingAreaProxy::deviceScaleFactorDidChange):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] Two touchadjustment tests are failing on mac
commit-queue@webkit.org [Tue, 6 Nov 2012 18:40:26 +0000 (18:40 +0000)]
[chromium] Two touchadjustment tests are failing on mac
https://bugs.webkit.org/show_bug.cgi?id=100833

Patch by Rick Byers <rbyers@chromium.org> on 2012-11-06
Reviewed by Antonio Gomes.

Re-enable these two tests with appropriate tweaks.  Also avoid the platform dependant
behavior by switching to the Ahem font.

* platform/chromium/TestExpectations: Remove the failure expectations.
* touchadjustment/html-label.html: Tweak one very borderline case to be closer to pass
reliably.  Previously this passed only due to a bug in the test infrastructure fixed in r132929.
* touchadjustment/nested-touch.html: Tweak borderline case to compensate for inclusion of
the border in the measurement of the size of the div.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Unreviewed gardening.
jsbell@chromium.org [Tue, 6 Nov 2012 18:27:51 +0000 (18:27 +0000)]
[Chromium] Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Adding window.external to our port
commit-queue@webkit.org [Tue, 6 Nov 2012 18:25:02 +0000 (18:25 +0000)]
[BlackBerry] Adding window.external to our port
https://bugs.webkit.org/show_bug.cgi?id=100868

Patch by Otto Derek Cheung <otcheung@rim.com> on 2012-11-06
Reviewed by Rob Buis.

Patching the method names to match the ones in the spec.

* WebCoreSupport/ExternalExtension.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoScriptWrappable should work for more than just Node
abarth@webkit.org [Tue, 6 Nov 2012 18:19:48 +0000 (18:19 +0000)]
ScriptWrappable should work for more than just Node
https://bugs.webkit.org/show_bug.cgi?id=101319

Reviewed by Eric Seidel.

This patch generalizes the inline cached wrapper code path to work with
all subclasses of ScriptWrappable, not just Node.

* bindings/js/JSDOMBinding.h:
(WebCore::setInlineCachedWrapper):
(WebCore::getInlineCachedWrapper):
(WebCore):
(WebCore::clearInlineCachedWrapper):
(WebCore::cacheWrapper):
* bindings/js/JSNodeCustom.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Update BB10 form theme.
commit-queue@webkit.org [Tue, 6 Nov 2012 18:17:36 +0000 (18:17 +0000)]
[BlackBerry] Update BB10 form theme.
https://bugs.webkit.org/show_bug.cgi?id=100760

Patch by Tiancheng Jiang <tijiang@rim.com> on 2012-11-06
Reviewed by Rob Buis.

Revert webkit/ce306dcc698199a6f7ce679daf0a30c25d3a3d43 slider theme
change which break media control and vertical slider.

RIM PR 236993
Internal Reviewed by Jeff Rogers.

* platform/blackberry/RenderThemeBlackBerry.cpp:
(WebCore::RenderThemeBlackBerry::paintSliderTrackRect):
(WebCore::RenderThemeBlackBerry::paintSliderThumb):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agocanonicalizedTitle() shouldn't convert 8 bit title strings to 16 bit
msaboff@apple.com [Tue, 6 Nov 2012 18:15:44 +0000 (18:15 +0000)]
canonicalizedTitle() shouldn't convert 8 bit title strings to 16 bit
https://bugs.webkit.org/show_bug.cgi?id=101105

Reviewed by Darin Adler.

Turned canonicalizedTitle() into a templated function based on character type.
Changed call in updateTitle() to check the bitness of the title string to call
the right template flavor of canonicalizedTitle().
Made supporting changes by adding displayBufferModifiedByEncoding(LChar*, ...)
and made TextEncoding::displayBuffer() a templated function as well.

No new tests needed, as functionality is unchanged.

* dom/Document.cpp:
(WebCore::canonicalizedTitle):
(WebCore::Document::updateTitle):
(WebCore::Document::displayBufferModifiedByEncodingInternal):
* dom/Document.h:
(WebCore::Document::displayBufferModifiedByEncoding):
* platform/text/TextEncoding.h:
(TextEncoding):
(WebCore::TextEncoding::displayBuffer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] css3/line-break/line-break-strict-sound-marks.html image flaky on Mac...
jsbell@chromium.org [Tue, 6 Nov 2012 18:15:04 +0000 (18:15 +0000)]
[Chromium] css3/line-break/line-break-strict-sound-marks.html image flaky on Mac 10.7
https://bugs.webkit.org/show_bug.cgi?id=101368

Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Page can't back to whole screen after zoom out and orientation change
zhajiang@rim.com [Tue, 6 Nov 2012 18:01:15 +0000 (18:01 +0000)]
[BlackBerry] Page can't back to whole screen after zoom out and orientation change
https://bugs.webkit.org/show_bug.cgi?id=101350

Reviewed by Rob Buis.
Patch by Jacky Jiang  <zhajiang@rim.com>.
Internally reviewed by Arvid Nilsson.

PR: 229321
When rotating device during touch interaction, bitmap zooming was
unexpected ended by resetBitmapZoomScale() so that it couldn't bounce
back. The call resetBitmapZoomScale() is ancient, which just resets the
scale of the bitmap before. Right now it only ends the bitmap zooming
which caused the issue. As UI viewport will be updated by zoomChanged()
right after that, we can remove this trouble maker.
Update the interaction views if the rotation is being changed so that
ZoomHandler can get the updated views.

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::zoomAboutPoint):
* Api/WebPageClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoFix RenderGeometryMap assertion when layers are scrolled during layout
simon.fraser@apple.com [Tue, 6 Nov 2012 17:57:08 +0000 (17:57 +0000)]
Fix RenderGeometryMap assertion when layers are scrolled during layout
https://bugs.webkit.org/show_bug.cgi?id=101292

Reviewed by Beth Dakin.

Source/WebCore:

When we set RenderLayer scroll positions as part of layout, we don't want
to update compositing layers right away. Updating compositing layers
requires that the entire layer tree has been updated from renderers,
so that the geometry of all RenderLayers can be trusted. When this state
was violated, RenderGeometryMap asserts.

Fix by bailing from updateCompositingLayersAfterScroll() if FrameView
tells us that we're doing layout. A full update of the compositing layers
will happen later anyway.

Test: compositing/geometry/geometry-map-scroll-during-layout-assertion.html

* rendering/RenderLayer.cpp:
(WebCore::frameViewFromLayer):
(WebCore::RenderLayer::updateCompositingLayersAfterScroll):

LayoutTests:

Test that hit the assertion before the fix.

* compositing/geometry/geometry-map-scroll-during-layout-assertion-expected.txt: Added.
* compositing/geometry/geometry-map-scroll-during-layout-assertion.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Unreviewed gardening - more rebaselines following bug 101115.
jsbell@chromium.org [Tue, 6 Nov 2012 17:54:42 +0000 (17:54 +0000)]
[Chromium] Unreviewed gardening - more rebaselines following bug 101115.

* platform/chromium-mac-lion/fast/block/lineboxcontain/block-glyphs-expected.png: Added.
* platform/chromium-mac-lion/fast/block/lineboxcontain/block-glyphs-replaced-expected.png: Added.
* platform/chromium-mac-lion/fast/block/lineboxcontain/glyphs-expected.png: Added.
* platform/chromium-mac-lion/fast/repaint/stacked-diacritics-expected.png: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] MediaStream API: Remove the valgrind issue with MockWebRTCPeerConnectionHa...
tommyw@google.com [Tue, 6 Nov 2012 17:50:03 +0000 (17:50 +0000)]
[chromium] MediaStream API: Remove the valgrind issue with MockWebRTCPeerConnectionHandler
https://bugs.webkit.org/show_bug.cgi?id=101336

Reviewed by Andreas Kling.

This patch changes the delete to delete [] to match the new[].

* DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoquoteCSSString() always creates a 16 bit string
msaboff@apple.com [Tue, 6 Nov 2012 17:42:03 +0000 (17:42 +0000)]
quoteCSSString() always creates a 16 bit string
https://bugs.webkit.org/show_bug.cgi?id=101004

Reviewed by Darin Adler.

Added a new templated helper based on character type from the logic of quoteCSSString() to process
the argument string based on its native bitness.

Functionality covered by existing tests.

* css/CSSParser.cpp:
(WebCore::quoteCSSStringInternal):
(WebCore::quoteCSSString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt][WK2] Fit-to-width broken on pages with viewport meta tag
abecsi@webkit.org [Tue, 6 Nov 2012 17:34:48 +0000 (17:34 +0000)]
[Qt][WK2] Fit-to-width broken on pages with viewport meta tag
https://bugs.webkit.org/show_bug.cgi?id=99715

Reviewed by Jocelyn Turcotte.

Source/WebCore:

Add a bool that indicates if the content had an explicit
initial-scale in the viewport meta tag.

* dom/ViewportArguments.cpp:
(WebCore::computeViewportAttributes):
* dom/ViewportArguments.h:
(ViewportAttributes):

Source/WebKit2:

The initial scale from the viewport attributes should only
be applied if the scale was explicitly specified in the
viewport meta tag.
If the initial scale is auto it should be calculated using
the final contents size, which might be larger than the
layout size, so that the content fits horizontally into
the view.

Also add QML unit tests that cover this functionality.

* UIProcess/API/qt/tests/qmltests/WebView/tst_fitToView.qml:
* UIProcess/API/qt/tests/qmltests/common/test5.html: Added.
* UIProcess/PageViewportController.cpp:
(WebKit::PageViewportController::PageViewportController):
(WebKit::PageViewportController::pageTransitionViewportReady):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRefactor setting TEXTURE_MAPPER_GL
commit-queue@webkit.org [Tue, 6 Nov 2012 17:31:30 +0000 (17:31 +0000)]
Refactor setting TEXTURE_MAPPER_GL
https://bugs.webkit.org/show_bug.cgi?id=99758

Patch by Laszlo Gombos <l.gombos@samsung.com> on 2012-11-06
Reviewed by Noam Rosenthal.

Make the rule that sets the default for USE(3D_GRAPHICS) port independent.

If not set, set USE(TEXTURE_MAPPER_GL) when USE(TEXTURE_MAPPER) and
USE(3D_GRAPHICS) is enabled.

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Gardening, add pixel expectation for test.
rakuco@webkit.org [Tue, 6 Nov 2012 17:28:03 +0000 (17:28 +0000)]
[EFL] Gardening, add pixel expectation for test.

* platform/efl/svg/dynamic-updates/SVGUseElement-dom-requiredFeatures-expected.png: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] fast/dom/Window/customized-property-survives-gc.html slow on debug
jsbell@chromium.org [Tue, 6 Nov 2012 17:26:36 +0000 (17:26 +0000)]
[Chromium] fast/dom/Window/customized-property-survives-gc.html slow on debug
https://bugs.webkit.org/show_bug.cgi?id=101362

Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCSP 1.1: Tweak the script interface to match the spec.
mkwst@chromium.org [Tue, 6 Nov 2012 17:15:29 +0000 (17:15 +0000)]
CSP 1.1: Tweak the script interface to match the spec.
https://bugs.webkit.org/show_bug.cgi?id=101321

Reviewed by Adam Barth.

Source/WebCore:

       This patch brings WebKit in line with the changes made to the spec in
       https://dvcs.w3.org/hg/content-security-policy/rev/5a29424a37d4.
       Specifically, the following:

       - Renamed 'document.SecurityPolicy' to 'document.securityPolicy'
       - Converted the following to read-only boolean attributes:
 - 'allowsEval'
 - 'allowsInlineScript'
 - 'allowsInlineStyle'
 - 'isActive'

These changes only have effect for ports with CSP_NEXT enabled.

Tests: http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowinlinescript.html
       http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowinlinestyle.html

* dom/Document.idl:
* page/DOMSecurityPolicy.idl:
    Tweaking the IDL files to match the new definition of the
    functionlity in the spec.

LayoutTests:

* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowconnectionto.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-alloweval.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowfontfrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowformaction.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowframefrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowimagefrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowmediafrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowobjectfrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowplugintype.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowscriptfrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowstylefrom.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-isactive-expected.txt:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-isactive.html:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-reporturi-expected.txt:
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-reporturi.html:
* http/tests/security/contentSecurityPolicy/resources/securitypolicy-tests-base.js:
    s/document.SecurityPolicy/document.securityPolicy/g
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowinlinescript-expected.txt: Added.
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowinlinescript.html: Added.
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowinlinestyle-expected.txt: Added.
* http/tests/security/contentSecurityPolicy/1.1/securitypolicy-allowinlinestyle.html: Added.
    For whatever reason, we never wrote tests for these methods. Now we
    have, which is good.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed gardening after bug 101115 fixed.
dbarton@mathscribe.com [Tue, 6 Nov 2012 16:44:46 +0000 (16:44 +0000)]
Unreviewed gardening after bug 101115 fixed.

* platform/chromium-linux-x86/fast/block/lineboxcontain: Added.
* platform/chromium-linux-x86/fast/block/lineboxcontain/block-glyphs-expected.txt: Added.
* platform/chromium-linux-x86/fast/block/lineboxcontain/block-glyphs-replaced-expected.png: Added.
* platform/chromium-linux-x86/fast/block/lineboxcontain/block-glyphs-replaced-expected.txt: Added.
* platform/chromium-linux-x86/fast/block/lineboxcontain/glyphs-expected.png: Added.
* platform/chromium-linux-x86/fast/block/lineboxcontain/glyphs-expected.txt: Added.
* platform/chromium-linux/fast/block/lineboxcontain/block-glyphs-expected.png:
* platform/chromium-linux/fast/block/lineboxcontain/block-glyphs-expected.txt: Added.
* platform/chromium-linux/fast/block/lineboxcontain/block-glyphs-replaced-expected.png: Added.
* platform/chromium-linux/fast/block/lineboxcontain/block-glyphs-replaced-expected.txt: Added.
* platform/chromium-linux/fast/block/lineboxcontain/glyphs-expected.png: Added.
* platform/chromium-linux/fast/block/lineboxcontain/glyphs-expected.txt: Added.
* platform/chromium-linux/fast/repaint/stacked-diacritics-expected.png:
* platform/chromium-mac/fast/block/lineboxcontain/block-glyphs-expected.png:
* platform/chromium-mac/fast/block/lineboxcontain/block-glyphs-replaced-expected.png:
* platform/chromium-mac/fast/block/lineboxcontain/glyphs-expected.png:
* platform/chromium-mac/fast/repaint/stacked-diacritics-expected.png:
* platform/chromium-win-xp/fast/block/lineboxcontain: Added.
* platform/chromium-win-xp/fast/block/lineboxcontain/block-glyphs-replaced-expected.png: Added.
* platform/chromium-win-xp/fast/block/lineboxcontain/block-glyphs-replaced-expected.txt: Added.
* platform/chromium-win-xp/fast/block/lineboxcontain/glyphs-expected.png: Added.
* platform/chromium-win-xp/fast/block/lineboxcontain/glyphs-expected.txt: Added.
* platform/chromium-win/fast/block/lineboxcontain/block-glyphs-replaced-expected.png: Added.
* platform/chromium-win/fast/block/lineboxcontain/block-glyphs-replaced-expected.txt: Added.
* platform/chromium-win/fast/block/lineboxcontain/glyphs-expected.png: Added.
* platform/chromium-win/fast/block/lineboxcontain/glyphs-expected.txt: Added.
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd Adobe's bug tracker to committers.py
commit-queue@webkit.org [Tue, 6 Nov 2012 16:37:18 +0000 (16:37 +0000)]
Add Adobe's bug tracker to committers.py
https://bugs.webkit.org/show_bug.cgi?id=101338

Patch by Mihai Balan <mibalan@adobe.com> on 2012-11-06
Reviewed by Kentaro Hara.

Adding Adobe's bug tracker mailing list to committers.py
to enable its inclusion in the autocomplete list.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Make webkit_lint happy. Unreviewed.
senorblanco@chromium.org [Tue, 6 Nov 2012 16:22:43 +0000 (16:22 +0000)]
[Chromium] Make webkit_lint happy.  Unreviewed.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Unreviewed gardening.
rakuco@webkit.org [Tue, 6 Nov 2012 16:18:41 +0000 (16:18 +0000)]
[EFL] Unreviewed gardening.

Mark fast/dom/shadow/shadowdom-for-output-only-shadow.html as flaky.

* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Gardening. Unskip test that's been passing for a while.
rakuco@webkit.org [Tue, 6 Nov 2012 16:15:46 +0000 (16:15 +0000)]
[EFL] Gardening. Unskip test that's been passing for a while.

* platform/efl-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdded console.clear() method
commit-queue@webkit.org [Tue, 6 Nov 2012 16:14:27 +0000 (16:14 +0000)]
Added console.clear() method

Web Inspector: add console.clear()
https://bugs.webkit.org/show_bug.cgi?id=101021

Patch by Andrey Lushnikov <lushnikov@google.com> on 2012-11-06
Reviewed by Vsevolod Vlasov.

Source/WebCore:

Adds native implementation of console.clear

Test: inspector/console/console-clear-function.html

* English.lproj/localizedStrings.js: Adds 'Console was cleared' string
* inspector/ConsoleMessage.cpp:
(WebCore::messageTypeValue): Adds ClearMessageType
* inspector/Inspector.json:
* inspector/InspectorConsoleAgent.cpp:
(WebCore::InspectorConsoleAgent::addMessageToConsole):
* inspector/front-end/ConsoleMessage.js:
(WebInspector.ConsoleMessageImpl.prototype._formatMessage): Display strategy for "clear" message
* inspector/front-end/ConsoleModel.js:
* inspector/front-end/inspector.css:
(.console-info): Adds style for the 'console was cleared' message
* page/Console.cpp: backend binging for console.clear()
(WebCore::Console::clear):
(WebCore):
* page/Console.h:
(Console):
* page/Console.idl: Adds 'clear' message to protocol
* page/ConsoleTypes.h:

Source/WebKit/mac:

* WebCoreSupport/WebChromeClient.mm:
(stringForMessageType): Added ClearMessageType case to switch

LayoutTests:

* inspector/console/console-clear-function-expected.txt: Added.
* inspector/console/console-clear-function.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] fast/images/exif-orientation-image-document.html failing on WK2 bots
rakuco@webkit.org [Tue, 6 Nov 2012 16:11:03 +0000 (16:11 +0000)]
[EFL] fast/images/exif-orientation-image-document.html failing on WK2 bots
https://bugs.webkit.org/show_bug.cgi?id=101344

Unreviewed gardening.

Patch by Thiago Marcos P. Santos <thiago.santos@intel.com> on 2012-11-06

* platform/efl-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Rolled DEPS.
senorblanco@chromium.org [Tue, 6 Nov 2012 16:06:24 +0000 (16:06 +0000)]
Unreviewed.  Rolled DEPS.

* DEPS:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Unreviewed gardening.
senorblanco@chromium.org [Tue, 6 Nov 2012 16:00:35 +0000 (16:00 +0000)]
[Chromium] Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoHeap-buffer-overflow in WebCore::TextTrackCueList::add
commit-queue@webkit.org [Tue, 6 Nov 2012 15:54:05 +0000 (15:54 +0000)]
Heap-buffer-overflow in WebCore::TextTrackCueList::add
https://bugs.webkit.org/show_bug.cgi?id=101018

Patch by Aaron Colwell <acolwell@chromium.org> on 2012-11-06
Reviewed by Eric Carlson.

Source/WebCore:

Added an extra check to avoid using a negative array index when a cue
is added to the beginning of the list.

Test case added to LayoutTests/media/track/track-add-remove-cue.html.

* html/track/TextTrackCueList.cpp:
(WebCore::TextTrackCueList::add):

LayoutTests:

Added a test case to verify that adding a cue to the beginning of a non-empty list doesn't crash.

* media/track/track-add-remove-cue-expected.txt:
* media/track/track-add-remove-cue.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRegression(r132681): Heap-use-after-free in WebCore::RenderTextTrackCue::layout
commit-queue@webkit.org [Tue, 6 Nov 2012 15:42:48 +0000 (15:42 +0000)]
Regression(r132681): Heap-use-after-free in WebCore::RenderTextTrackCue::layout
https://bugs.webkit.org/show_bug.cgi?id=100981

Patch by Aaron Colwell <acolwell@chromium.org> on 2012-11-06
Reviewed by Eric Carlson.

Source/WebCore:

Fixing a TextTrackCue use after free bug. textTrackRemoveCues() needs to be called when
an HTMLTrackElement is removed from an HTMLMediaElement so that references to
TextTrackCues are removed from m_cueTree.

Test: media/track/track-remove-by-setting-innerHTML.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::willRemoveTrack):

LayoutTests:

Added a test that removes track elements by setting innerHTML on the track's parent.

* media/track/track-remove-by-setting-innerHTML-expected.txt: Added.
* media/track/track-remove-by-setting-innerHTML.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[skia] Implement reference (url) filters on composited layers.
senorblanco@chromium.org [Tue, 6 Nov 2012 15:34:13 +0000 (15:34 +0000)]
[skia] Implement reference (url) filters on composited layers.
https://bugs.webkit.org/show_bug.cgi?id=100142

Reviewed by James Robinson.

Source/Platform:

* chromium/public/WebLayer.h:
Expose Layer::setFilter() to the WebKit API.

Source/WebCore:

Covered by css3/filters/effect-reference-hw.html

* WebCore.gyp/WebCore.gyp:
* WebCore.gypi:
Add new files.
* WebCore.xcodeproj/project.pbxproj:
Mark FilterEffect.h as "private".
* platform/graphics/chromium/GraphicsLayerChromium.cpp:
(WebCore::GraphicsLayerChromium::setFilters):
Use the new SkImageFilter builder and setLayer() infrastructure, but
only if there's a reference filter in the chain.
* platform/graphics/filters/FEBlend.h:
* platform/graphics/filters/FEColorMatrix.h:
* platform/graphics/filters/FEGaussianBlur.h:
* platform/graphics/filters/FELighting.h:
* platform/graphics/filters/FEMorphology.h:
* platform/graphics/filters/FilterEffect.h:
(WebCore::FilterEffect::createImageFilter):
Add createImageFilter() factory function for building SkImageFilters.
* platform/graphics/filters/FilterOperation.h:
(WebCore::FilterOperation::ReferenceFilterOperation::filterEffect):
(WebCore::FilterOperation::ReferenceFilterOperation::setFilterEffect):
Add a FilterEffect member to ReferenceFilterOperation, and accessors
for it.
* platform/graphics/filters/skia/FEBlendSkia.cpp:
(WebCore::FEBlend::createImageFilter):
Implement SkImageFilter building for FEBlend filter.
* platform/graphics/filters/skia/FEColorMatrixSkia.cpp:
(WebCore::createColorFilter):
(WebCore::FEColorMatrix::platformApplySkia):
Refactor creation of SkColorFilter from an FEColorMatrix filter.
(WebCore::FEColorMatrix::createImageFilter):
Implement SkImageFilter building for FEColorMatrix filter.
* platform/graphics/filters/skia/FEGaussianBlurSkia.cpp:
(WebCore::FEGaussianBlur::createImageFilter):
Implement SkImageFilter building for FEGaussianBlur filter.
* platform/graphics/filters/skia/FELightingSkia.cpp:
(WebCore::FELighting::createImageFilter):
Implement SkImageFilter building for FELighting filters.
* platform/graphics/filters/skia/FEMorphologySkia.cpp:
(WebCore::FEMorphology::createImageFilter):
Implement SkImageFilter building for FEMorphology filter.
* platform/graphics/filters/skia/SkiaImageFilterBuilder.cpp: Added.
(WebCore::SkiaImageFilterBuilder::SkiaImageFilterBuilder):
(WebCore::SkiaImageFilterBuilder::build):
* platform/graphics/filters/skia/SkiaImageFilterBuilder.h:
Implement new helper class for building an SkImageFilter DAG from
a FilterEffect DAG.  Can also build from FilterOperations list.
* rendering/FilterEffectRenderer.cpp:
(WebCore::FilterEffectRenderer::build):
When building FilterEffects from a reference filter, also store the
generated DAG on the ReferenceFilterOperation itself.
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::updateOrRemoveFilterEffect):
Even when in composited mode, if there's a reference filter in the
chain, build the FilterEffect DAG from it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] Enable fixed position compositing on Android DRT
commit-queue@webkit.org [Tue, 6 Nov 2012 15:32:19 +0000 (15:32 +0000)]
[chromium] Enable fixed position compositing on Android DRT
https://bugs.webkit.org/show_bug.cgi?id=100263

Patch by Sami Kyostila <skyostil@chromium.org> on 2012-11-06
Reviewed by James Robinson.

Chrome on Android forces composited mode for every page and creates
stacking contexts and graphics layers for fixed positioned elements.
Change the DumpRenderTree settings to match this.

* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::resetWebSettings):
* DumpRenderTree/chromium/TestShell.h:
(TestShell::setAcceleratedCompositingForFixedPositionEnabled):
(TestShell):
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
(WebPreferences::applyTo):
* DumpRenderTree/chromium/WebPreferences.h:
(WebPreferences):
* Scripts/webkitpy/layout_tests/port/chromium_android.py:
(ChromiumAndroidPort.additional_drt_flag):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Automatically go fullscreen on video play
commit-queue@webkit.org [Tue, 6 Nov 2012 15:27:54 +0000 (15:27 +0000)]
[BlackBerry] Automatically go fullscreen on video play
https://bugs.webkit.org/show_bug.cgi?id=101100

Patch by Max Feil <mfeil@rim.com> on 2012-11-06
Reviewed by Eric Carlson.

Source/WebCore:

There is a requirement to have HTML5 video automatically enter
fullscreen when a video starts playing (PR131774). This change
implements this feature, with restrictions. The main restriction
is adherence to WebKit's philosophy of only entering fullscreen
due to a user gesture. This is important in order to avoid
pop-up advertisements and other unwanted fullscreen content.
One consequence of this is that video elements with the autoplay
attribute will not automatically enter fullscreen.

Other caveats:
- This feature applies only to "small screen" devices where
automatically going fullscreen makes more sense.
- Fullscreen will only be entered automatically when the
video is played from the beginning (current time is zero).
It is assumed that if the user is resuming play from a paused
state and is not in fullscreen mode, then they exited fullscreen
mode intentionally.

Test: platform/blackberry/media/video-automatic-fullscreen.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::mediaPlayerEnterFullscreen):
(WebCore):
(WebCore::HTMLMediaElement::mediaPlayerIsFullscreen):
(WebCore::HTMLMediaElement::mediaPlayerIsFullscreenPermitted):
* html/HTMLMediaElement.h:
(HTMLMediaElement):
* platform/graphics/MediaPlayer.h:
(WebCore::MediaPlayerClient::mediaPlayerEnterFullscreen):
(WebCore::MediaPlayerClient::mediaPlayerIsFullscreen):
(WebCore::MediaPlayerClient::mediaPlayerIsFullscreenPermitted):
* platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
(WebCore::MediaPlayerPrivate::play):
(WebCore::MediaPlayerPrivate::waitMetadataTimerFired):
(WebCore::MediaPlayerPrivate::conditionallyGoFullscreenAfterPlay):
(WebCore):
* platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.h:
(MediaPlayerPrivate):

LayoutTests:

Test that fullscreen is entered automatically when play is
pressed (PR131774). This test applies to handheld (small screen)
devices only, not tablets. The html code for this test was based
on video-controls-fullscreen-volume.html and modified.

* platform/blackberry/media/video-automatic-fullscreen-expected.txt: Added.
* platform/blackberry/media/video-automatic-fullscreen.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCoordinated Graphics: Remove a parent member in WebLayerInfo.
commit-queue@webkit.org [Tue, 6 Nov 2012 15:20:42 +0000 (15:20 +0000)]
Coordinated Graphics: Remove a parent member in WebLayerInfo.
https://bugs.webkit.org/show_bug.cgi?id=101142

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-06
Reviewed by Noam Rosenthal.

The parent member is not used in UI Process, and GraphicsLayer::setParent is
internally used only, so we don't need to send the parent layer id to UI Process.

* Shared/WebLayerTreeInfo.h:
(WebKit::WebLayerInfo::WebLayerInfo):
* WebProcess/WebPage/CoordinatedGraphics/CoordinatedGraphicsLayer.cpp:
(WebCore::CoordinatedGraphicsLayer::syncLayerState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCoordinated Graphics: Remove two methods' virtual modifiers in CoordinatedGraphicsLayer.
commit-queue@webkit.org [Tue, 6 Nov 2012 15:16:12 +0000 (15:16 +0000)]
Coordinated Graphics: Remove two methods' virtual modifiers in CoordinatedGraphicsLayer.
https://bugs.webkit.org/show_bug.cgi?id=101141

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-06
Reviewed by Noam Rosenthal.

The methods exist for only CoordinatedGraphicsLayer and we have no plan to
extend CoordinatedGraphicsLayer.

* WebProcess/WebPage/CoordinatedGraphics/CoordinatedGraphicsLayer.h:
(CoordinatedGraphicsLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[WK2][EFL][GTK] early return of checkSpellingOfString treats correct words as misspelled
g.czajkowski@samsung.com [Tue, 6 Nov 2012 15:14:50 +0000 (15:14 +0000)]
[WK2][EFL][GTK] early return of checkSpellingOfString treats correct words as misspelled
https://bugs.webkit.org/show_bug.cgi?id=101215

Reviewed by Gustavo Noronha Silva.

Save the misspelled location and length of the string to the default values (-1, 0)
at the beginning of the checkSpellingOfString method.
It assumes that the words are spelled correctly and early return of method doesn't
cause assertions in findFirstMisspelling (WebCore/editing/TextCheckingHelper.cpp).

* platform/text/enchant/TextCheckerEnchant.cpp:
(TextCheckerEnchant::checkSpellingOfString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[TexMap] Remove unused textures in the texture pool.
commit-queue@webkit.org [Tue, 6 Nov 2012 15:12:19 +0000 (15:12 +0000)]
[TexMap] Remove unused textures in the texture pool.
https://bugs.webkit.org/show_bug.cgi?id=100706

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-06
Reviewed by Noam Rosenthal.

Currently, we do not remove textures in the texture pool of
TextureMapper. The texture pool is destroyed when TextureMapper is
destroyed. It means the texture pool consumes texture memory until its
destruction. This patch removes textures if the textures have been not
used in 3 seconds as LayerTreeCoordinator manages UpdateAtlases.

Changing cache policy is not testable in layout tests.

* platform/graphics/texmap/TextureMapper.cpp:
(WebCore::BitmapTexturePoolEntry::BitmapTexturePoolEntry):
(WebCore::BitmapTexturePoolEntry::markUsed):
(WebCore::BitmapTexturePoolEntry::compareTimeLastUsed):
(BitmapTexturePoolEntry):
(WebCore):
(BitmapTexturePool):
(WebCore::BitmapTexturePool::BitmapTexturePool):
(WebCore::BitmapTexturePool::scheduleReleaseUnusedTextures):
(WebCore::BitmapTexturePool::releaseUnusedTexturesTimerFired):
(WebCore::BitmapTexturePool::acquireTexture):
(WebCore::TextureMapper::acquireTextureFromPool):
(WebCore::TextureMapper::TextureMapper):
(WebCore::TextureMapper::~TextureMapper):
* platform/graphics/texmap/TextureMapper.h:
(WebCore):
(TextureMapper):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: NMI instrument CachedMetadata. It uses ~300k on plus.google.com
loislo@chromium.org [Tue, 6 Nov 2012 14:45:05 +0000 (14:45 +0000)]
Web Inspector: NMI instrument CachedMetadata. It uses ~300k on plus.google.com
https://bugs.webkit.org/show_bug.cgi?id=101067

Reviewed by Yury Semikhatsky.

Plain vanilla instrumentation for CachedMetadata class.

* CMakeLists.txt:
* GNUmakefile.list.am:
* Target.pri:
* WebCore.gypi:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* loader/CachedMetadata.cpp: Added.
(WebCore):
(WebCore::CachedMetadata::reportMemoryUsage):
* loader/CachedMetadata.h:
(CachedMetadata):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[CoordinatedGraphics] Access to LayerTreeRenderer::m_renderQueue should be thread...
kbalazs@webkit.org [Tue, 6 Nov 2012 14:40:11 +0000 (14:40 +0000)]
[CoordinatedGraphics] Access to LayerTreeRenderer::m_renderQueue should be thread safe
https://bugs.webkit.org/show_bug.cgi?id=101341

Reviewed by Noam Rosenthal.

The queue can be accessed from public API so we should make
sure it is not updated concurrently with a threaded render loop.

* UIProcess/CoordinatedGraphics/LayerTreeRenderer.cpp:
(WebKit::LayerTreeRenderer::syncRemoteContent):
(WebKit::LayerTreeRenderer::appendUpdate):
* UIProcess/CoordinatedGraphics/LayerTreeRenderer.h:
(LayerTreeRenderer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@133599 268f45cc-cd09-0410-ab3c-d52691b4dbfc