WebKit-https.git
8 years agoUnreviewed; chromium test_expectations update.
senorblanco@chromium.org [Fri, 8 Jul 2011 20:28:21 +0000 (20:28 +0000)]
Unreviewed; chromium test_expectations update.

Add some MAC GPU failing tests from r90464.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake checkout_unittest more robust against files moving around
aroben@apple.com [Fri, 8 Jul 2011 20:02:23 +0000 (20:02 +0000)]
Make checkout_unittest more robust against files moving around

Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
Scripts directory

Reviewed by Adam Barth.

* Scripts/webkitpy/common/checkout/checkout_unittest.py:
(CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
object and use it to get the path to the Scripts directory, rather than hard-coding the
relative path from this file to Scripts.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90659 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUnreviewed; chromium test_expectations update.
senorblanco@chromium.org [Fri, 8 Jul 2011 20:01:37 +0000 (20:01 +0000)]
Unreviewed; chromium test_expectations update.

Add some WIN GPU failing tests from r90464 to test_expectations.txt.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90658 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUnreviewed; chromium test rebaseline.
senorblanco@chromium.org [Fri, 8 Jul 2011 19:54:21 +0000 (19:54 +0000)]
Unreviewed; chromium test rebaseline.

New baseline for a failing test on GPU linux introduced in r90646.

* platform/chromium-gpu-linux/platform/chromium/compositing/layout-width-change-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90657 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake TestFailures remember that run-webkit-tests timed out, even across reloads
aroben@apple.com [Fri, 8 Jul 2011 19:51:16 +0000 (19:51 +0000)]
Make TestFailures remember that run-webkit-tests timed out, even across reloads

Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload

Reviewed by Daniel Bates.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
When true, it indicates that there was an error retrieving the results for this build and
that the errorCallback should be called.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90656 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUnreviewed; chromium test expectations cleanup.
senorblanco@chromium.org [Fri, 8 Jul 2011 19:47:37 +0000 (19:47 +0000)]
Unreviewed; chromium test expectations cleanup.

Removed some duplicate expectations.

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90655 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoAdd framework for a new/dummy XMLDocumentParser
jpfau@apple.com [Fri, 8 Jul 2011 19:31:34 +0000 (19:31 +0000)]
Add framework for a new/dummy XMLDocumentParser
https://bugs.webkit.org/show_bug.cgi?id=63955

Reviewed by Adam Barth.

Added a dummy framework NewXMLDocumentParser

* WebCore.xcodeproj/project.pbxproj:
* dom/Document.cpp:
(WebCore::Document::createParser):
* xml/parser/NewXMLDocumentParser.cpp: Added.
(WebCore::NewXMLDocumentParser::NewXMLDocumentParser):
(WebCore::NewXMLDocumentParser::textPosition):
(WebCore::NewXMLDocumentParser::lineNumber):
(WebCore::NewXMLDocumentParser::insert):
(WebCore::NewXMLDocumentParser::append):
(WebCore::NewXMLDocumentParser::finish):
(WebCore::NewXMLDocumentParser::detach):
(WebCore::NewXMLDocumentParser::hasInsertionPoint):
(WebCore::NewXMLDocumentParser::finishWasCalled):
(WebCore::NewXMLDocumentParser::processingData):
(WebCore::NewXMLDocumentParser::prepareToStopParsing):
(WebCore::NewXMLDocumentParser::stopParsing):
(WebCore::NewXMLDocumentParser::isWaitingForScripts):
(WebCore::NewXMLDocumentParser::isExecutingScript):
(WebCore::NewXMLDocumentParser::executeScriptsWaitingForStylesheets):
* xml/parser/NewXMLDocumentParser.h: Added.
(WebCore::NewXMLDocumentParser::create):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoRemove unused function parameters.
ojan@chromium.org [Fri, 8 Jul 2011 19:27:44 +0000 (19:27 +0000)]
Remove unused function parameters.

Patch by David Reveman <reveman@chromium.org> on 2011-07-08
Reviewed by Ryosuke Niwa.

* platform/graphics/gpu/TilingData.cpp:
(WebCore::TilingData::textureOffset):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoTeach garden-o-matic how to display test results
abarth@webkit.org [Fri, 8 Jul 2011 19:26:32 +0000 (19:26 +0000)]
Teach garden-o-matic how to display test results
https://bugs.webkit.org/show_bug.cgi?id=64141

Reviewed by Ojan Vafai.

This patch includes basic infrastructure for probing build.chromium.org
for test results.  We only handle text and image tests, not anything
complicated like reftests.  Also, we're using the revision/build
independent results store on the server, so we're avoiding that
complication for now.

It's slightly hacky that we need to probe the server to see what kinds
of results exist.  A better solution would be to add CORS support to
the server or to use the local server to help.

* Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
* Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
* Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
* Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
* Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
* Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoREGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
dpranke@chromium.org [Fri, 8 Jul 2011 19:21:41 +0000 (19:21 +0000)]
REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
https://bugs.webkit.org/show_bug.cgi?id=64092

Reviewed by Tony Chang.

The code was incorrectly creating empty shards in the
shard_in_two case.

* Scripts/webkitpy/layout_tests/controllers/manager.py:
* Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90651 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoTeach buildbot to figure out how many webkitpy/webkitperl tests failed
aroben@apple.com [Fri, 8 Jul 2011 19:05:17 +0000 (19:05 +0000)]
Teach buildbot to figure out how many webkitpy/webkitperl tests failed

Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
tests failed when looking at build.webkit.org

Reviewed by Eric Seidel.

* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(TestWithFailureCount): New class that represents a test build step which has an associated
failure count. Eventually we should move more of our test classes to deriving from this.
(TestWithFailureCount.countFailures): Method for subclasses to override to say how many
failures occurred.

(TestWithFailureCount.commandComplete):
(TestWithFailureCount.evaluateCommand):
(TestWithFailureCount.getText):
(TestWithFailureCount.getText2):
These were all based on RunGtkAPITests.

(RunPythonTests): Changed to inherit from TestWithFailureCount.
(RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
failures.
(RunPerlTests): Changed to inherit from TestWithFailureCount.
(RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
failures.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoPatch by Kalev Lember <kalev@smartlink.ee> on 2011-07-08
commit-queue@webkit.org [Fri, 8 Jul 2011 18:59:27 +0000 (18:59 +0000)]
Patch by Kalev Lember <kalev@smartlink.ee> on 2011-07-08
Reviewed by Adam Roben.

Add missing _WIN32_WINNT and WINVER definitions
https://bugs.webkit.org/show_bug.cgi?id=59702

Moved _WIN32_WINNT and WINVER definitions to config.h so that they are
available for all source files.

In particular, wtf/FastMalloc.cpp uses CreateTimerQueueTimer and
DeleteTimerQueueTimer which are both guarded by
#if (_WIN32_WINNT >= 0x0500)
in MinGW headers.

* config.h:
* wtf/Assertions.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoEnsure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor...
aroben@apple.com [Fri, 8 Jul 2011 18:47:49 +0000 (18:47 +0000)]
Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration

Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
bots due to commit-log-editor errors

Reviewed by Adam Barth.

* Scripts/webkitpy/common/checkout/checkout_unittest.py:
(CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
$CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
commit-log-editor can find a value for the committer's email, and that the committer and
author email addresses match, which will prevent commit-log-editor from inserting a "Patch
by" line in the commit message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90648 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt][WK2] Views should expose QActions for basic navigation.
andreas.kling@nokia.com [Fri, 8 Jul 2011 18:37:38 +0000 (18:37 +0000)]
[Qt][WK2] Views should expose QActions for basic navigation.
https://bugs.webkit.org/show_bug.cgi?id=64174

Source/WebKit2:

Add navigationAction() methods to the views to provide default
QActions for Back, Forward, Stop and Reload.

Reviewed by Benjamin Poulain.

* UIProcess/API/qt/WKView.h:
* UIProcess/API/qt/qdesktopwebview.cpp:
(QDesktopWebView::navigationAction):
* UIProcess/API/qt/qdesktopwebview.h:
* UIProcess/API/qt/qtouchwebpage.cpp:
(QTouchWebPage::navigationAction):
* UIProcess/API/qt/qtouchwebpage.h:
* UIProcess/API/qt/qwebkittypes.h: Added.
* UIProcess/API/qt/tests/commonviewtests/tst_commonviewtests.cpp:
(tst_CommonViewTests::backAndForward):
(tst_CommonViewTests::reload):
(tst_CommonViewTests::stop):
* UIProcess/API/qt/tests/commonviewtests/webviewabstraction.cpp:
(WebViewAbstraction::triggerNavigationAction):
* UIProcess/API/qt/tests/commonviewtests/webviewabstraction.h:
* UIProcess/API/qt/tests/html/basic_page2.html: Added.
* UIProcess/qt/QtWebPageProxy.cpp:
(QtWebPageProxy::navigationAction):
* UIProcess/qt/QtWebPageProxy.h:
* WebKit2API.pri:

Tools:

Reviewed by Benjamin Poulain.

Bring the toolbar in Qt's MiniBrowser back to life.

* MiniBrowser/qt/BrowserView.cpp:
(BrowserView::navigationAction):
* MiniBrowser/qt/BrowserView.h:
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoPatch by David Reveman <reveman@chromium.org> on 2011-07-08
commit-queue@webkit.org [Fri, 8 Jul 2011 18:33:00 +0000 (18:33 +0000)]
Patch by David Reveman <reveman@chromium.org> on 2011-07-08
Reviewed by Stephen White.

[Chromium] Edge anti-aliasing for composited layers.
https://bugs.webkit.org/show_bug.cgi?id=61388

Source/WebCore:

Add transparent outer border to tiled layers and adjust vertex
coordinates so that use of a bilinear filter creates a smooth
layer edge.

Tests: platform/chromium/compositing/tiny-layer-rotated.html
       platform/chromium/compositing/huge-layer-rotated.html (existing)
       TilingDataTest in webkit_unit_tests

* platform/graphics/chromium/ContentLayerChromium.cpp:
Change maxUntiledSize to 510 to ensure that tiles are not greater
than 512 with outer borders.
(WebCore::ContentLayerChromium::updateLayerSize): We can't use the
layer size as tile size when we want to avoid tiling as this will
not be enough space to include possible outer borders. We instead use
an empty size, which allows the tiler to automatically adjust the
tile size to be large enough for the layer to fit in one tile.
(WebCore::ContentLayerChromium::createTilerIfNeeded):
(WebCore::ContentLayerChromium::setIsMask): Don't use border texels
for layer used as mask.

* platform/graphics/chromium/ContentLayerChromium.h:
(WebCore::ContentLayerChromium::m_borderTexels) Added.

* platform/graphics/chromium/LayerTilerChromium.cpp:
(WebCore::LayerTilerChromium::tileTexRect): Added.
(WebCore::LayerTilerChromium::tileLayerRect): m_tileSize is no
longer the correct layer size. Size of bounds with border should
be the correct layer size.
(WebCore::LayerTilerChromium::growLayerToContain): Adjust texture
size to include outer borders and handle empty m_tileSize.
(WebCore::LayerTilerChromium::invalidateRect): Use size of rectangle
returned by tileTexRect instead of m_tileSize for texture size.
(WebCore::LayerTilerChromium::prepareToUpdate): Ditto.
(WebCore::LayerTilerChromium::draw): Compute and intersect tile edges
instead of using tile coordinates directly. Edges are adjusted to
include outer borders and make sure all partially covered pixels are
processed.
(WebCore::LayerTilerChromium::drawTexturedQuad): Pass quad to
shader using point uniform.
(WebCore::LayerTilerChromium::invalidateRect): Invalidate old layer
area to clear any contents left from previous layer size.

* platform/graphics/chromium/LayerTilerChromium.h:
(WebCore::LayerTilerChromium::tileTexRect) Added.
(WebCore::LayerTilerChromium::drawTexturedQuad): Add quad parameter.

* platform/graphics/chromium/ShaderChromium.cpp:
(WebCore::VertexShaderPosTexTransform::getShaderString) Get X/Y vertex
components from point uniform.
(WebCore::VertexShaderPosTexTransform::VertexShaderPosTexTransform)
Added point uniform.
(WebCore::VertexShaderPosTexTransform::init) Ditto.
(WebCore::VertexShaderPosTexTransform::pointLocation) Added.

* platform/graphics/chromium/ShaderChromium.h:
(WebCore::VertexShaderPosTexTransform::pointLocation) Added.

* platform/graphics/gpu/TilingData.cpp:
(WebCore::TilingData::tileBoundsWithOuterBorder): Added.
(WebCore::TilingData::computeNumTiles): Adjust for outer border.
(WebCore::TilingData::tileXIndexFromSrcCoord): Ditto.
(WebCore::TilingData::tileYIndexFromSrcCoord): Ditto.
(WebCore::TilingData::tileSizeX): Ditto.
(WebCore::TilingData::tileSizeY): Ditto.
(WebCore::TilingData::intersectDrawQuad): Ditto.
(WebCore::TilingData::textureOffset): Ditto.

* platform/graphics/gpu/TilingData.h:
(WebCore::TilingData::tileBoundsWithOuterBorder): Added.

Source/WebKit:

Track changes to TilingData class which now uses an outer border.

* chromium/tests/TilingDataTest.cpp:

LayoutTests:

* platform/chromium/compositing/tiny-layer-rotated-expected.png: Added.
* platform/chromium/compositing/tiny-layer-rotated-expected.txt: Added.
* platform/chromium/compositing/tiny-layer-rotated.html: Added.
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[skia] don't rely on lockPixels failure to detect gpu-backed device (in prep for...
commit-queue@webkit.org [Fri, 8 Jul 2011 18:13:47 +0000 (18:13 +0000)]
[skia] don't rely on lockPixels failure to detect gpu-backed device (in prep for skia roll)
https://bugs.webkit.org/show_bug.cgi?id=64162

Patch by Mike Reed <reed@google.com> on 2011-07-08
Reviewed by Stephen White.

No new tests. preparing for skia roll, where lockPixels always succeeds (but slowly for gpu-backed)

* platform/graphics/skia/ImageBufferSkia.cpp:
(WebCore::putImageData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt][WK2] Get rid of the check for TILED_BACKING_STORE in Qt
benjamin.poulain@nokia.com [Fri, 8 Jul 2011 18:06:20 +0000 (18:06 +0000)]
[Qt][WK2] Get rid of the check for TILED_BACKING_STORE in Qt
https://bugs.webkit.org/show_bug.cgi?id=64175

Reviewed by Kenneth Rohde Christiansen.

Qt no longer supports building WebKit2 without TILED_BACKING_STORE.

* UIProcess/API/qt/qtouchwebpage.cpp:
(QTouchWebPagePrivate::onScaleChanged):
* UIProcess/qt/QtWebPageProxy.cpp:
(QtWebPageProxy::setResizesToContentsUsingLayoutSize):
* UIProcess/qt/TiledDrawingAreaProxyQt.cpp:
* UIProcess/qt/TiledDrawingAreaTileQt.cpp:
* WebProcess/WebPage/qt/TiledDrawingAreaQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90644 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSource/JavaScriptCore: Rename "makeSecure" to "fill" and remove the support for displ...
commit-queue@webkit.org [Fri, 8 Jul 2011 17:54:05 +0000 (17:54 +0000)]
Source/JavaScriptCore: Rename "makeSecure" to "fill" and remove the support for displaying last character
to avoid layering violatation.
https://bugs.webkit.org/show_bug.cgi?id=59114

Patch by Chang Shu <cshu@webkit.org> on 2011-07-08
Reviewed by Alexey Proskuryakov.

* JavaScriptCore.exp:
* JavaScriptCore.order:
* wtf/text/StringImpl.cpp:
(WTF::StringImpl::fill):
* wtf/text/StringImpl.h:
* wtf/text/WTFString.h:
(WTF::String::fill):

Source/WebCore: Update calling sites after function renamed.
https://bugs.webkit.org/show_bug.cgi?id=59114

Patch by Chang Shu <cshu@webkit.org> on 2011-07-08
Reviewed by Alexey Proskuryakov.

No new tests, just refactoring.

* editing/visible_units.cpp:
(WebCore::previousBoundary):
(WebCore::nextBoundary):
* rendering/RenderText.cpp:
(WebCore::RenderText::setTextInternal):

Source/WebKit/win: Update calling sites after function renamed.
https://bugs.webkit.org/show_bug.cgi?id=59114

Patch by Chang Shu <cshu@webkit.org> on 2011-07-08
Reviewed by Alexey Proskuryakov.

* WebKitGraphics.cpp:
(WebDrawText):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[CSSRegions]Parse content: -webkit-from-flow
commit-queue@webkit.org [Fri, 8 Jul 2011 17:29:10 +0000 (17:29 +0000)]
[CSSRegions]Parse content: -webkit-from-flow
https://bugs.webkit.org/show_bug.cgi?id=63133

Patch by Mihnea Ovidenie <mihnea@adobe.com> on 2011-07-08
Reviewed by David Hyatt.

Source/WebCore:

Test: fast/regions/content-webkit-from-flow-parsing.html

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::contentToCSSValue):
* css/CSSParser.cpp:
(WebCore::CSSParser::parseContent):
(WebCore::CSSParser::parseFromFlowContent):
* css/CSSParser.h:
* css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::cleanup):
(WebCore::CSSPrimitiveValue::getStringValue):
(WebCore::CSSPrimitiveValue::cssText):
* css/CSSPrimitiveValue.h:
* css/CSSStyleSelector.cpp:
(WebCore::CSSStyleSelector::applyProperty):
* rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::diff):
* rendering/style/RenderStyle.h:
(WebCore::InheritedFlags::regionThread):
(WebCore::InheritedFlags::setRegionThread):
(WebCore::InheritedFlags::initialRegionThread):
* rendering/style/StyleRareNonInheritedData.cpp:
(WebCore::StyleRareNonInheritedData::StyleRareNonInheritedData):
(WebCore::StyleRareNonInheritedData::operator==):
* rendering/style/StyleRareNonInheritedData.h:

LayoutTests:

* fast/regions/content-webkit-from-flow-parsing-expected.txt: Added.
* fast/regions/content-webkit-from-flow-parsing.html: Added.
* fast/regions/script-tests/content-webkit-from-flow-parsing.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoCheck activeDocumentLoader() in
japhet@chromium.org [Fri, 8 Jul 2011 17:24:51 +0000 (17:24 +0000)]
Check activeDocumentLoader() in
WebFrameImpl::currentHistoryItem() and return
early if null, since that should mean we're
shutting down.
https://bugs.webkit.org/show_bug.cgi?id=52923

Reviewed by Darin Fisher.

No known repro, so no new test.

* src/WebFrameImpl.cpp:
(WebKit::WebFrameImpl::currentHistoryItem):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoREGRESSION (r90557): http/tests/inspector/network/network-embed.html fails
vsevik@chromium.org [Fri, 8 Jul 2011 16:13:42 +0000 (16:13 +0000)]
REGRESSION (r90557): http/tests/inspector/network/network-embed.html fails
https://bugs.webkit.org/show_bug.cgi?id=64103

Unreviewed test fix (typo).

* http/tests/inspector/network/network-embed.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoTeach TestFailures how to find test names in commit-log-editor-style commit messages
aroben@apple.com [Fri, 8 Jul 2011 16:09:24 +0000 (16:09 +0000)]
Teach TestFailures how to find test names in commit-log-editor-style commit messages

TestFailures was relying on Trac turning the list of modified files in our commit messages
into an HTML list. But Trac only does this when the list of modified files is indented.
commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
list, which was confusing TestFailures.

TestFailures now does much simpler parsing of the commit message (i.e., just a substring
search) without relying at all on its structure.

Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
fast/dom/HTMLProgressElement/progress-element-markup.html on Windows

Reviewed by David Kilzer.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
(Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
message, just return its text.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForRegressionRange): Instead of searching for test names in
each commit's list of modified files, just search for test names anywhere in the commit's
message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: styles-disable-then-delete was flaky.
pfeldman@chromium.org [Fri, 8 Jul 2011 16:04:28 +0000 (16:04 +0000)]
Web Inspector: styles-disable-then-delete was flaky.
https://bugs.webkit.org/show_bug.cgi?id=64170

Reviewed by Yury Semikhatsky.

* inspector/styles/styles-add-invalid-property.html:
* inspector/styles/styles-disable-then-delete.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: inspector/styles/styles-url-linkify.html is flaky.
vsevik@chromium.org [Fri, 8 Jul 2011 16:01:07 +0000 (16:01 +0000)]
Web Inspector: inspector/styles/styles-url-linkify.html is flaky.
https://bugs.webkit.org/show_bug.cgi?id=64171

Added image files because the behavior differs when they are not available.

Unreviewed test fix: added image files.

* inspector/styles/resources/fromcss.png: Added.
* inspector/styles/resources/iframed.png: Added.
* inspector/styles/styles-url-linkify-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agosheriffbot is too spammy in IRC
abarth@webkit.org [Fri, 8 Jul 2011 15:58:29 +0000 (15:58 +0000)]
sheriffbot is too spammy in IRC
https://bugs.webkit.org/show_bug.cgi?id=64153

Reviewed by Eric Seidel.

Reporting failures in IRC worked well when the tree was greener than it
is today.  Nowadays, this feature mostly just results in spam about
false positives.  If we reach a greener state, we can bring this code
back.

(Another possibility is to restrict this feature to builder bots,
rather than including testers, as we did before this patch.)

* Scripts/webkitpy/tool/commands/sheriffbot.py:
* Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: NetworkPanel search failed if the matched sting is in the query part...
loislo@chromium.org [Fri, 8 Jul 2011 15:53:56 +0000 (15:53 +0000)]
Web Inspector: NetworkPanel search failed if the matched sting is in the query part of url
https://bugs.webkit.org/show_bug.cgi?id=64167

Reviewed by Yury Semikhatsky.

* inspector/front-end/NetworkPanel.js:
(WebInspector.NetworkPanel.prototype._matchResource):
(WebInspector.NetworkPanel.prototype._highlightNthMatchedResource):
(WebInspector.NetworkDataGridNode.prototype._refreshNameCell):
* inspector/front-end/Resource.js:
(WebInspector.Resource.prototype.get folder):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt] Enable HTTP Pipelining by default
benjamin.poulain@nokia.com [Fri, 8 Jul 2011 15:50:25 +0000 (15:50 +0000)]
[Qt] Enable HTTP Pipelining by default
https://bugs.webkit.org/show_bug.cgi?id=64169

Reviewed by Andreas Kling.

QNetworkAccessManager disables HTTP pipelining by default. We enable it by
setting an attribute on the request.

* platform/network/qt/ResourceRequestQt.cpp:
(WebCore::ResourceRequest::toNetworkRequest):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt] http/tests/plugins/get-url.html is crashing on the bot
abecsi@webkit.org [Fri, 8 Jul 2011 15:46:29 +0000 (15:46 +0000)]
[Qt] http/tests/plugins/get-url.html is crashing on the bot
https://bugs.webkit.org/show_bug.cgi?id=64168

Unreviewed gardening.

The cause of the crash revealed itself as a side effect thus
it is better to expect the test to CRASH rather than skipping it.

* platform/qt/Skipped: Unskip test.
* platform/qt/test_expectations.txt: Mark as CRASH

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoEnsure commit-log-editor adds a "Patch by" line when the author and committer are...
aroben@apple.com [Fri, 8 Jul 2011 15:22:34 +0000 (15:22 +0000)]
Ensure commit-log-editor adds a "Patch by" line when the author and committer are different

Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
by" line. But some patches (like rollout patches) don't contain that line. Now we always add
"Patch by" to the commit log regardless of the ChangeLog's contents.

Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
should have been abarth@webkit.org

Reviewed by Anders Carlsson.

* Scripts/commit-log-editor:
(createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
before. If there is no "Reviewed by" line, try to put it just above the first modified file.
If all else fails, put it at the end of the commit message.
(patchAuthorshipString): Added. Code came from createCommitMessage.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt] http/tests/plugins/get-url.html is crashing on the bot
abecsi@webkit.org [Fri, 8 Jul 2011 14:58:00 +0000 (14:58 +0000)]
[Qt] http/tests/plugins/get-url.html is crashing on the bot
https://bugs.webkit.org/show_bug.cgi?id=64168

Unreviewed gardening.

* platform/qt/Skipped: Skip the test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: CSS inspector gets confused about specificity of !important properties
apavlov@chromium.org [Fri, 8 Jul 2011 14:51:00 +0000 (14:51 +0000)]
Web Inspector: CSS inspector gets confused about specificity of !important properties
https://bugs.webkit.org/show_bug.cgi?id=64074

Reviewed by Yury Semikhatsky.

Source/WebCore:

* inspector/front-end/StylesSidebarPane.js:
(WebInspector.StylesSidebarPane.prototype._markUsedProperties):

LayoutTests:

* inspector/elements/elements-panel-styles-expected.txt:
* inspector/elements/resources/elements-panel-styles.css:
(#container .foo):
(.foo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[WK2] Do not forward touch events to the web process when it does not need them
benjamin.poulain@nokia.com [Fri, 8 Jul 2011 14:42:27 +0000 (14:42 +0000)]
[WK2] Do not forward touch events to the web process when it does not need them
https://bugs.webkit.org/show_bug.cgi?id=64164

Reviewed by Kenneth Rohde Christiansen.

Source/JavaScriptCore:

Add a convenience function to obtain a reference to the last element of a Deque.

* wtf/Deque.h:
(WTF::Deque::last):

Source/WebKit2:

The call to ChromeClient::needTouchEvent() is now forwarded to the WebPageProxy
to change the way events are delivered.

When the WebPage does not need touch events, and there is no queued touch events,
the incoming events just bounce back through PageClient::doneWithTouchEvent().

In the case when new events come to WebPageProxy and there are still touch events
incoming from the WebProcess, the new events are deferred with the corresponding
pending touch events.
Deferring the new events iafter the corresponding forwarded event ensure
the delivery is always done in order when PageClient::doneWithTouchEvent()
is called.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy):
(WebKit::WebPageProxy::handleTouchEvent):
(WebKit::WebPageProxy::needTouchEvents):
(WebKit::WebPageProxy::didReceiveEvent):
(WebKit::WebPageProxy::processDidCrash):
* UIProcess/WebPageProxy.h:
(WebKit::QueuedTouchEvents:::forwardedEvent):
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::needTouchEvents):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Chromium] Unreviewed, clang build fix.
apavlov@chromium.org [Fri, 8 Jul 2011 14:34:55 +0000 (14:34 +0000)]
[Chromium] Unreviewed, clang build fix.

* inspector/InspectorStyleTextEditor.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[GTK] testwebview API test fails after http://trac.webkit.org/changeset/90471
kov@webkit.org [Fri, 8 Jul 2011 13:42:27 +0000 (13:42 +0000)]
[GTK] testwebview API test fails after trac.webkit.org/changeset/90471
https://bugs.webkit.org/show_bug.cgi?id=64159

Rework the icon-uri change test so that the condition for quitting
the mainloop is the icon-uri change itself, along with a timeout
to avoid taking too long in case of failure. Since the conditions
for considering a page loaded were changed we can't rely on that
for this test anymore.

Patch by Gustavo Noronha Silva <gustavo.noronha@collabora.com> on 2011-07-08
Reviewed by Xan Lopez.

* tests/testwebview.c:
(timeout_cb): error out if it takes too long for the icon-uri
change to happen.
(icon_uri_changed_cb): quit the loop here instead of waiting on
the page being loaded.
(test_webkit_web_view_icon_uri):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: add support for drag'n'drop of non-elements (comments, text, etc.)
pfeldman@chromium.org [Fri, 8 Jul 2011 13:25:21 +0000 (13:25 +0000)]
Web Inspector: add support for drag'n'drop of non-elements (comments, text, etc.)
https://bugs.webkit.org/show_bug.cgi?id=64163

Reviewed by Yury Semikhatsky.

* inspector/Inspector.json:
* inspector/InspectorDOMAgent.cpp:
(WebCore::InspectorDOMAgent::moveTo):
* inspector/front-end/ElementsTreeOutline.js:
(WebInspector.ElementsTreeOutline.prototype._isValidDragSourceOrTarget):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-08 Andrey Kosyakov <caseq@chromium.org>
caseq@chromium.org [Fri, 8 Jul 2011 13:21:08 +0000 (13:21 +0000)]
2011-07-08  Andrey Kosyakov  <caseq@chromium.org>

        Web Inspector: secure access to extensions API
        https://bugs.webkit.org/show_bug.cgi?id=64080

        Reviewed by Pavel Feldman.

        * inspector/front-end/ExtensionAPI.js:
        (WebInspector.injectedExtensionAPI.Panels.prototype.create):
        (WebInspector.injectedExtensionAPI.ExtensionSidebarPaneImpl.prototype.setPage):
        * inspector/front-end/ExtensionPanel.js:
        (WebInspector.ExtensionPanel):
        * inspector/front-end/ExtensionServer.js:
        (WebInspector.ExtensionServer):
        (WebInspector.ExtensionServer.prototype._onCreatePanel):
        (WebInspector.ExtensionServer.prototype._onSetSidebarPage):
        (WebInspector.ExtensionServer.prototype._addExtensions):
        (WebInspector.ExtensionServer.prototype._onWindowMessage):
        (WebInspector.ExtensionServer.prototype._registerSubscriptionHandler):
        (WebInspector.ExtensionServer.prototype._expandResourcePath):
        (WebInspector.ExtensionServer.prototype._normalizePath):

2011-07-08  Andrey Kosyakov  <caseq@chromium.org>

        Web Inspector: secure access to extensions API
        https://bugs.webkit.org/show_bug.cgi?id=64080

        Reviewed by Pavel Feldman.

        * inspector/extensions/extensions.html: add explicit base paths to extension resource being loaded.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoReviewed by Andreas Kling.
zherczeg@webkit.org [Fri, 8 Jul 2011 13:19:25 +0000 (13:19 +0000)]
Reviewed by Andreas Kling.

Adding myself as a reviewer.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoAdd BUG modifier which was accidentally removed in the previous commit
abecsi@webkit.org [Fri, 8 Jul 2011 12:55:27 +0000 (12:55 +0000)]
Add BUG modifier which was accidentally removed in the previous commit

Unreviewed.

* platform/qt/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt] Some tests are flaky with NRWT
abecsi@webkit.org [Fri, 8 Jul 2011 12:29:39 +0000 (12:29 +0000)]
[Qt] Some tests are flaky with NRWT
https://bugs.webkit.org/show_bug.cgi?id=64002

Unreviewed gardening.

* platform/qt/test_expectations.txt: Skip fast/forms/textfield-overflow-by-value-update.html
because it sometimes does not produce any output.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUnreviewed, rolling out r90615.
commit-queue@webkit.org [Fri, 8 Jul 2011 11:56:56 +0000 (11:56 +0000)]
Unreviewed, rolling out r90615.
http://trac.webkit.org/changeset/90615
https://bugs.webkit.org/show_bug.cgi?id=64158

broke inspector/extensions/extensions.html (Requested by caseq
on #webkit).

* inspector/front-end/ExtensionAPI.js:
(WebInspector.injectedExtensionAPI.Panels.prototype.create):
(WebInspector.injectedExtensionAPI.ExtensionSidebarPaneImpl.prototype.setPage):
(WebInspector.injectedExtensionAPI.expandURL):
* inspector/front-end/ExtensionPanel.js:
(WebInspector.ExtensionPanel):
* inspector/front-end/ExtensionServer.js:
(WebInspector.ExtensionServer):
(WebInspector.ExtensionServer.prototype._onCreatePanel):
(WebInspector.ExtensionServer.prototype._onSetSidebarPage):
(WebInspector.ExtensionServer.prototype._addExtensions):
(WebInspector.ExtensionServer.prototype._onWindowMessage):
(WebInspector.ExtensionServer.prototype._registerSubscriptionHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoRebaseline expected file after r90567
abecsi@webkit.org [Fri, 8 Jul 2011 11:53:55 +0000 (11:53 +0000)]
Rebaseline expected file after r90567

Unreviewed.

* inspector/styles/styles-url-linkify-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: Adding CSS properties results in messy style rules
apavlov@chromium.org [Fri, 8 Jul 2011 11:32:40 +0000 (11:32 +0000)]
Web Inspector: Adding CSS properties results in messy style rules
https://bugs.webkit.org/show_bug.cgi?id=63622

Reviewed by Pavel Feldman.

Source/WebCore:

Test: inspector/styles/styles-formatting.html

* CMakeLists.txt:
* GNUmakefile.list.am:
* WebCore.gypi:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* css/CSSPropertySourceData.cpp:
(WebCore::SourceRange::length):
* css/CSSPropertySourceData.h:
* inspector/InspectorStyleSheet.cpp:
(WebCore::InspectorStyle::InspectorStyle):
(WebCore::InspectorStyle::setPropertyText):
(WebCore::InspectorStyle::toggleProperty):
(WebCore::InspectorStyle::applyStyleText):
(WebCore::InspectorStyle::newLineAndWhitespaceDelimiters):
* inspector/InspectorStyleSheet.h:
* inspector/InspectorStyleTextEditor.cpp: Added.
(WebCore::InspectorStyleTextEditor::InspectorStyleTextEditor):
(WebCore::InspectorStyleTextEditor::insertProperty):
(WebCore::InspectorStyleTextEditor::replaceProperty):
(WebCore::InspectorStyleTextEditor::removeProperty):
(WebCore::InspectorStyleTextEditor::enableProperty):
(WebCore::InspectorStyleTextEditor::disableProperty):
(WebCore::InspectorStyleTextEditor::disabledIndexByOrdinal):
(WebCore::InspectorStyleTextEditor::shiftDisabledProperties):
(WebCore::InspectorStyleTextEditor::internalReplaceProperty):
* inspector/InspectorStyleTextEditor.h: Added.
(WebCore::InspectorStyleTextEditor::styleText):

LayoutTests:

* inspector/styles/styles-formatting-expected.txt: Added.
* inspector/styles/styles-formatting.html: Added.
* inspector/styles/styles-new-API-expected.txt:
* inspector/styles/styles-new-API.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt][WK2] Add basic support for panning gestures to the QTouchWebView
commit-queue@webkit.org [Fri, 8 Jul 2011 11:25:08 +0000 (11:25 +0000)]
[Qt][WK2] Add basic support for panning gestures to the QTouchWebView
https://bugs.webkit.org/show_bug.cgi?id=64105

Patch by Benjamin Poulain <benjamin@webkit.org> on 2011-07-08
Reviewed by Kenneth Rohde Christiansen.

This patch adds basic support for the panning gesture on the UI process side.

The events coming back from the WebProcess are processed through
the QtPanGestureRecognizer to recognize the pan gesture. When the
gesture is recognized, the actions are performed on the view through
the TouchViewInterface.

Currently, the viewport just move the page around without limit.
This will be improved when a physics engine is integrated.

* UIProcess/API/qt/qtouchwebview.cpp:
(QTouchWebViewPrivate::scroll):
* UIProcess/API/qt/qtouchwebview.h:
* UIProcess/API/qt/qtouchwebview_p.h:
* UIProcess/qt/QtPanGestureRecognizer.cpp: Added.
(WebKit::QtPanGestureRecognizer::QtPanGestureRecognizer):
(WebKit::QtPanGestureRecognizer::recognize):
(WebKit::QtPanGestureRecognizer::reset):
* UIProcess/qt/QtPanGestureRecognizer.h: Added.
* UIProcess/qt/TouchViewInterface.cpp:
(WebKit::TouchViewInterface::panGestureStarted):
(WebKit::TouchViewInterface::panGestureRequestScroll):
(WebKit::TouchViewInterface::panGestureEnded):
(WebKit::TouchViewInterface::panGestureCancelled):
* UIProcess/qt/TouchViewInterface.h:
* UIProcess/qt/qtouchwebpageproxy.cpp:
(QTouchWebPageProxy::QTouchWebPageProxy):
(QTouchWebPageProxy::processDidCrash):
(QTouchWebPageProxy::doneWithTouchEvent):
* UIProcess/qt/qtouchwebpageproxy.h:
* WebKit2.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: Quotes are rendered as &quot; in the DOM tree
apavlov@chromium.org [Fri, 8 Jul 2011 11:16:24 +0000 (11:16 +0000)]
Web Inspector: Quotes are rendered as &quot; in the DOM tree
https://bugs.webkit.org/show_bug.cgi?id=64154

Reviewed by Yury Semikhatsky.

Source/WebCore:

Since we currently rely on setting textContent rather than innerHTML for the DOM tree elements,
escapeHTML() calls unnecessarily HTML-escape certain characters in the DOM tree contents.

* inspector/front-end/ElementsTreeOutline.js:
(WebInspector.ElementsTreeElement.prototype._buildAttributeDOM):
():

LayoutTests:

* inspector/elements/elements-panel-structure-expected.txt:
* inspector/elements/elements-panel-structure.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: Remove unused code from InspectorAgent.h.
vsevik@chromium.org [Fri, 8 Jul 2011 11:03:08 +0000 (11:03 +0000)]
Web Inspector: Remove unused code from InspectorAgent.h.
https://bugs.webkit.org/show_bug.cgi?id=64120

Reviewed by Joseph Pecoraro.

* inspector/InspectorAgent.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-08 Andrey Kosyakov <caseq@chromium.org>
caseq@chromium.org [Fri, 8 Jul 2011 10:39:20 +0000 (10:39 +0000)]
2011-07-08  Andrey Kosyakov  <caseq@chromium.org>

        Web Inspector: secure access to extensions API
        https://bugs.webkit.org/show_bug.cgi?id=64080

        Reviewed by Pavel Feldman.

        * inspector/front-end/ExtensionAPI.js:
        (WebInspector.injectedExtensionAPI.Panels.prototype.create):
        (WebInspector.injectedExtensionAPI.ExtensionSidebarPaneImpl.prototype.setPage):
        * inspector/front-end/ExtensionPanel.js:
        (WebInspector.ExtensionPanel):
        * inspector/front-end/ExtensionServer.js:
        (WebInspector.ExtensionServer):
        (WebInspector.ExtensionServer.prototype._onCreatePanel):
        (WebInspector.ExtensionServer.prototype._onSetSidebarPage):
        (WebInspector.ExtensionServer.prototype._addExtensions):
        (WebInspector.ExtensionServer.prototype._onWindowMessage):
        (WebInspector.ExtensionServer.prototype._registerSubscriptionHandler):
        (WebInspector.ExtensionServer.prototype._expandResourcePath):
        (WebInspector.ExtensionServer.prototype._normalizePath):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: typing undefined = 1 in console crashes browser
yurys@chromium.org [Fri, 8 Jul 2011 10:12:56 +0000 (10:12 +0000)]
Web Inspector: typing undefined = 1 in console crashes browser
https://bugs.webkit.org/show_bug.cgi?id=64155

Source/WebCore:

Do not access undefined value directly when producing JSON objects as undefined
may be overriden by the inspected page.

Reviewed by Pavel Feldman.

Test: inspector/console/console-eval-undefined-override.html

* inspector/InjectedScriptSource.js:
(.):
():

LayoutTests:

Reviewed by Pavel Feldman.

* inspector/console/console-eval-undefined-override-expected.txt: Added.
* inspector/console/console-eval-undefined-override.html: Added.
* platform/chromium/inspector/console/console-eval-undefined-override-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-08 Andrey Kosyakov <caseq@chromium.org>
caseq@chromium.org [Fri, 8 Jul 2011 08:37:12 +0000 (08:37 +0000)]
2011-07-08  Andrey Kosyakov  <caseq@chromium.org>

        Unreviewed. Moving test expectation from platform-specific to generic.

        * fast/dom/HTMLLinkElement/link-and-subresource-test-nonexistent-expected.txt: Renamed from LayoutTests/platform/chromium-linux/fast/dom/HTMLLinkElement/link-and-subresource-test-nonexistent-expected.txt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWebSocket: Clear cookies after test in httponly-cookie.pl
yutak@chromium.org [Fri, 8 Jul 2011 06:41:26 +0000 (06:41 +0000)]
WebSocket: Clear cookies after test in httponly-cookie.pl
https://bugs.webkit.org/show_bug.cgi?id=64145

Reviewed by Kent Tamura.

This test leaks cookies, which would cause a test failure if we put
a copy of this test under http/tests/websocket/tests.

To avoid this issue, cookies should be removed before the test ends.

* http/tests/websocket/tests/hixie76/httponly-cookie.pl:
After the test, send an XHR request to itself with a query parameter,
which erases cookies used in this test by setting "Max-Age=0".

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[wx] Unreviewed build fix, remove old files from the tree and update paths to new...
kevino@webkit.org [Fri, 8 Jul 2011 04:59:40 +0000 (04:59 +0000)]
[wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[wx] Unreviewed build fix, more fixes for install name issues.
kevino@webkit.org [Fri, 8 Jul 2011 04:53:29 +0000 (04:53 +0000)]
[wx] Unreviewed build fix, more fixes for install name issues.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Chromium] Update baseline files for input-appearance-range.html.
tkent@chromium.org [Fri, 8 Jul 2011 03:53:39 +0000 (03:53 +0000)]
[Chromium] Update baseline files for input-appearance-range.html.

* platform/chromium-linux/fast/forms/input-appearance-range-expected.png:
* platform/chromium-mac-leopard/fast/forms/input-appearance-range-expected.png:
* platform/chromium-mac/fast/forms/input-appearance-range-expected.png: Removed.
* platform/chromium-win/fast/forms/input-appearance-range-expected.png:
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agodump-as-markup.js should support shadow tree
morrita@google.com [Fri, 8 Jul 2011 02:46:38 +0000 (02:46 +0000)]
dump-as-markup.js should support shadow tree
https://bugs.webkit.org/show_bug.cgi?id=62447

Reviewed by Ryosuke Niwa.

- Added shadow support to Markup._get()
- Updated expectations caused by this change.
- Added explanatory test cases for shadows.

* editing/inserting/5607069-2-expected.txt:
* editing/inserting/5607069-3-expected.txt:
* editing/pasteboard/copy-null-characters-expected.txt:
* editing/selection/dump-as-markup-expected.txt:
* editing/selection/dump-as-markup-form-text-expected.txt:
* fast/parser/object-with-textarea-fallback-expected.txt:
* fast/dom/HTMLMeterElement/meter-element-markup-expected.txt: Added.
* fast/dom/HTMLMeterElement/meter-element-markup.html: Added.
* fast/dom/HTMLProgressElement/progress-element-markup-expected.txt: Added.
* fast/dom/HTMLProgressElement/progress-element-markup.html: Added.
* fast/dom/HTMLMeterElement/meter-element-markup.html:
* fast/dom/HTMLProgressElement/progress-element-markup.html:
* resources/dump-as-markup.js:
(Markup._get):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agotest-webkitpy fails on chromium win
dpranke@chromium.org [Fri, 8 Jul 2011 02:22:39 +0000 (02:22 +0000)]
test-webkitpy fails on chromium win
https://bugs.webkit.org/show_bug.cgi?id=64137

Reviewed by Eric Seidel.

test-webkitpy doesn't play nicely with the multiprocessing
module on win32. An earlier change actually reenabled the tests
on win32 by mistake.

This patch also fixes a few cases where path names will trip
things up on win32 (e.g., by testing unix paths on windows).
We do not lose any real coverage here with those fixes.

* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
* Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[ImageDiff] Calculate/print difference right after reading baseline image.
commit-queue@webkit.org [Fri, 8 Jul 2011 02:13:05 +0000 (02:13 +0000)]
[ImageDiff] Calculate/print difference right after reading baseline image.
https://bugs.webkit.org/show_bug.cgi?id=64117

Patch by Leandro Pereira <leandro@profusion.mobi> on 2011-07-07
Reviewed by Kent Tamura.

Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
it.  Whichever happens first.

* DumpRenderTree/efl/ImageDiff.cpp:
(main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
commit-queue@webkit.org [Fri, 8 Jul 2011 01:16:07 +0000 (01:16 +0000)]
[EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
https://bugs.webkit.org/show_bug.cgi?id=64107

Patch by Raphael Kubo da Costa <kubo@profusion.mobi> on 2011-07-07
Reviewed by Kent Tamura.

The pointer is later managed by BitmapContext, but it was being
de-refed and deleted earlier by our RefPtr, causing crashes in all
pixel tests.

We now call leakRef() to make sure it is not removed when our
createBitmapContextFromWebView goes out of context.

* DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
(createBitmapContextFromWebView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWebSocket: Use jsTestIsAsync and finishJSTest() in httponly-cookie.pl
yutak@chromium.org [Fri, 8 Jul 2011 00:49:40 +0000 (00:49 +0000)]
WebSocket: Use jsTestIsAsync and finishJSTest() in httponly-cookie.pl
https://bugs.webkit.org/show_bug.cgi?id=64087

Reviewed by Kent Tamura.

* http/tests/websocket/tests/hixie76/httponly-cookie.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSwitch HitTestResult to to new layout types
eae@chromium.org [Fri, 8 Jul 2011 00:17:24 +0000 (00:17 +0000)]
Switch HitTestResult to to new layout types
https://bugs.webkit.org/show_bug.cgi?id=64131

Reviewed by Eric Seidel.

No new tests, no functionality changes.

* page/MouseEventWithHitTestResults.cpp:
(WebCore::MouseEventWithHitTestResults::localPoint):
* page/MouseEventWithHitTestResults.h:
* rendering/HitTestResult.cpp:
(WebCore::HitTestResult::HitTestResult):
(WebCore::HitTestResult::addNodeToRectBasedTestResult):
(WebCore::HitTestResult::rectForPoint):
* rendering/HitTestResult.h:
(WebCore::HitTestResult::point):
(WebCore::HitTestResult::localPoint):
(WebCore::HitTestResult::setPoint):
(WebCore::HitTestResult::setLocalPoint):
(WebCore::HitTestResult::rectForPoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoDFG JIT does not implement op_construct.
commit-queue@webkit.org [Thu, 7 Jul 2011 23:54:57 +0000 (23:54 +0000)]
DFG JIT does not implement op_construct.
https://bugs.webkit.org/show_bug.cgi?id=64066

Patch by Filip Pizlo <fpizlo@apple.com> on 2011-07-07
Reviewed by Gavin Barraclough.

* dfg/DFGAliasTracker.h:
(JSC::DFG::AliasTracker::recordConstruct):
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::addCall):
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGJITCodeGenerator.cpp:
(JSC::DFG::JITCodeGenerator::emitCall):
* dfg/DFGNode.h:
* dfg/DFGNonSpeculativeJIT.cpp:
(JSC::DFG::NonSpeculativeJIT::compile):
* dfg/DFGOperations.cpp:
* dfg/DFGOperations.h:
* dfg/DFGRepatch.cpp:
(JSC::DFG::dfgLinkFor):
* dfg/DFGRepatch.h:
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compile):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoDFG JIT does not implement get_by_id prototype caching.
commit-queue@webkit.org [Thu, 7 Jul 2011 23:48:39 +0000 (23:48 +0000)]
DFG JIT does not implement get_by_id prototype caching.
https://bugs.webkit.org/show_bug.cgi?id=64077

Patch by Filip Pizlo <fpizlo@apple.com> on 2011-07-07
Reviewed by Gavin Barraclough.

* dfg/DFGRepatch.cpp:
(JSC::DFG::emitRestoreScratch):
(JSC::DFG::linkRestoreScratch):
(JSC::DFG::tryCacheGetByID):
* runtime/JSObject.h:
(JSC::JSObject::addressOfPropertyAtOffset):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSwitch remaining paint functions to new layout types
leviw@chromium.org [Thu, 7 Jul 2011 23:17:55 +0000 (23:17 +0000)]
Switch remaining paint functions to new layout types
https://bugs.webkit.org/show_bug.cgi?id=64116

Reviewed by Eric Seidel.

Switching all functions with IntPoint paintOffsets to the new Layout types.

No new tests, no functionality changes.

* rendering/EllipsisBox.cpp:
(WebCore::EllipsisBox::paint):
(WebCore::EllipsisBox::paintSelection):
* rendering/EllipsisBox.h:
* rendering/InlineBox.cpp:
(WebCore::InlineBox::paint):
* rendering/InlineBox.h:
* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::paint):
* rendering/InlineFlowBox.h:
* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paint):
(WebCore::InlineTextBox::paintCustomHighlight):
* rendering/InlineTextBox.h:
* rendering/LayoutTypes.h:
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::paintColumnRules):
(WebCore::RenderBlock::paintColumnContents):
(WebCore::RenderBlock::paintContents):
(WebCore::RenderBlock::paintChildren):
(WebCore::RenderBlock::paintCaret):
(WebCore::RenderBlock::paintObject):
(WebCore::RenderBlock::paintFloats):
(WebCore::RenderBlock::paintEllipsisBoxes):
(WebCore::RenderBlock::paintContinuationOutlines):
(WebCore::RenderBlock::paintSelection):
* rendering/RenderBlock.h:
* rendering/RenderBox.cpp:
(WebCore::RenderBox::paintCustomHighlight):
* rendering/RenderBox.h:
* rendering/RenderInline.cpp:
(WebCore::RenderInline::paintOutline):
(WebCore::RenderInline::paintOutlineForLine):
* rendering/RenderInline.h:
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::paintOverflowControls):
(WebCore::RenderLayer::paintScrollCorner):
(WebCore::RenderLayer::paintResizer):
* rendering/RenderLayer.h:
* rendering/RenderLineBoxList.cpp:
(WebCore::RenderLineBoxList::paint):
* rendering/RenderListBox.cpp:
(WebCore::RenderListBox::paintScrollbar):
(WebCore::RenderListBox::paintItemForeground):
(WebCore::RenderListBox::paintItemBackground):
* rendering/RenderListBox.h:
* rendering/RenderObject.cpp:
(WebCore::RenderObject::paintFocusRing):
(WebCore::RenderObject::paintOutline):
* rendering/RenderObject.h:
* rendering/RenderReplaced.cpp:
(WebCore::RenderReplaced::shouldPaint):
* rendering/RenderReplaced.h:
* rendering/RenderScrollbarPart.cpp:
(WebCore::RenderScrollbarPart::paintIntoRect):
* rendering/RenderScrollbarPart.h:
* rendering/RenderTableCell.cpp:
(WebCore::RenderTableCell::paintBackgroundsBehindCell):
* rendering/RenderTableSection.cpp:
(WebCore::RenderTableSection::paintCell):
* rendering/RenderTableSection.h:
* rendering/RenderTextControl.cpp:
(WebCore::RenderTextControl::paintPlaceholder):
* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::paintEllipsisBox):
(WebCore::RootInlineBox::paintCustomHighlight):
(WebCore::RootInlineBox::paint):
* rendering/RootInlineBox.h:
* rendering/mathml/RenderMathMLBlock.cpp:
(WebCore::RenderMathMLBlock::paint):
* rendering/mathml/RenderMathMLFraction.cpp:
(WebCore::RenderMathMLFraction::paint):
* rendering/mathml/RenderMathMLRoot.cpp:
(WebCore::RenderMathMLRoot::paint):
* rendering/mathml/RenderMathMLSquareRoot.cpp:
(WebCore::RenderMathMLSquareRoot::paint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWork around Skia PDF's lack of inverted path support.
commit-queue@webkit.org [Thu, 7 Jul 2011 23:16:19 +0000 (23:16 +0000)]
Work around Skia PDF's lack of inverted path support.
https://bugs.webkit.org/show_bug.cgi?id=64032

Patch by Steve VanDeBogart <vandebo@chromium.org> on 2011-07-07
Reviewed by James Robinson.

The trick used in http://neugierig.org/software/chromium/notes/2010/07/clipping.html
to support antialiased clips doesn't work when printing to Skia's PDF backend because
the backend does not support inverted paths. This manifests as rounded buttons not being
drawn when printing, tracked as Chrome bug 79519.

However, when the output is a vector device, like PDF, we don't need antialiased clips.
It's up to the PDF rendering engine to do that.  So we can simply disable the antialiased
clip code if the output is a vector device.

I think the fix isn't testable because it requires examining the printed output.

* platform/graphics/skia/PlatformContextSkia.cpp:
(WebCore::PlatformContextSkia::clipPathAntiAliased):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoDFG JIT method_check implementation does not link to optimized get_by_id
commit-queue@webkit.org [Thu, 7 Jul 2011 23:07:41 +0000 (23:07 +0000)]
DFG JIT method_check implementation does not link to optimized get_by_id
slow path.
https://bugs.webkit.org/show_bug.cgi?id=64073

Patch by Filip Pizlo <fpizlo@apple.com> on 2011-07-07
Reviewed by Gavin Barraclough.

* dfg/DFGRepatch.cpp:
(JSC::DFG::dfgRepatchGetMethodFast):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[chromium] Add text baseline for spanOverlapsCanvas
jamesr@google.com [Thu, 7 Jul 2011 22:45:29 +0000 (22:45 +0000)]
[chromium] Add text baseline for spanOverlapsCanvas

* platform/chromium-gpu-win/compositing/layer-creation/spanOverlapsCanvas-expected.txt: Added.
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSwitch rendering tree selection code to to new layout types
eae@chromium.org [Thu, 7 Jul 2011 22:37:55 +0000 (22:37 +0000)]
Switch rendering tree selection code to to new layout types
https://bugs.webkit.org/show_bug.cgi?id=63762

Reviewed by Eric Seidel.

Switch selection getters and selection gap calculation methods over to
new layout abstraction.

No new tests, no functionality changes.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::selectionTop):
(WebCore::InlineTextBox::selectionBottom):
(WebCore::InlineTextBox::selectionHeight):
(WebCore::InlineTextBox::isSelected):
* rendering/InlineTextBox.h:
* rendering/RenderBR.h:
(WebCore::RenderBR::selectionRectForRepaint):
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::selectionGapRectsForRepaint):
(WebCore::RenderBlock::selectionGaps):
(WebCore::RenderBlock::inlineSelectionGaps):
(WebCore::RenderBlock::blockSelectionGaps):
(WebCore::RenderBlock::blockSelectionGap):
(WebCore::RenderBlock::logicalLeftSelectionGap):
(WebCore::RenderBlock::logicalRightSelectionGap):
(WebCore::RenderBlock::logicalLeftSelectionOffset):
(WebCore::RenderBlock::logicalRightSelectionOffset):
(WebCore::RenderBlock::positionForPointWithInlineChildren):
(WebCore::RenderBlock::desiredColumnWidth):
* rendering/RenderBlock.h:
(WebCore::RenderBlock::selectionRectForRepaint):
* rendering/RenderListMarker.cpp:
(WebCore::RenderListMarker::selectionRectForRepaint):
* rendering/RenderListMarker.h:
* rendering/RenderObject.h:
(WebCore::RenderObject::selectionRect):
(WebCore::RenderObject::selectionRectForRepaint):
* rendering/RenderReplaced.cpp:
(WebCore::RenderReplaced::selectionRectForRepaint):
* rendering/RenderReplaced.h:
* rendering/RenderSelectionInfo.h:
(WebCore::RenderSelectionInfo::RenderSelectionInfo):
* rendering/RenderText.cpp:
(WebCore::RenderText::selectionRectForRepaint):
* rendering/RenderText.h:
* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
* rendering/RootInlineBox.h:
(WebCore::RootInlineBox::selectionHeight):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoReviewed by Alexey Proskuryakov.
gavinp@chromium.org [Thu, 7 Jul 2011 22:29:57 +0000 (22:29 +0000)]
Reviewed by Alexey Proskuryakov.

fast/dom/HTMLLinkElement/link-and-subresource-test.html is flaky on chromium debug bots
https://bugs.webkit.org/show_bug.cgi?id=60097

The culprit was that CachedResource:stopLoading() was using *this
after a call to checkNotify(), which isn't kosher.  This patch
uses a CachedResourceHandle to keep the CachedResource alive.

Source/WebCore:

The test is a very close copy of the eponymous
link-and-subresource-test.html, only substituting invalid
resources for the valid ones in that test.  The reproduction is
timing related, and happens much more consistantly with an invalid
resource for whatever reason.
Test: fast/dom/HTMLLinkElement/link-and-subresource-test-nonexistent.html

* loader/cache/CachedResource.cpp:
(WebCore::CachedResource::stopLoading):

LayoutTests:

The test is a very close copy of the eponymous
link-and-subresource-test.html, only substituting invalid
resources for the valid ones in that test.  The reproduction is
timing related, and happens much more consistantly with an invalid
resource for whatever reason.

* fast/dom/HTMLLinkElement/link-and-subresource-test-nonexistent-expected.txt: Added.
* fast/dom/HTMLLinkElement/link-and-subresource-test-nonexistent.html: Added.
* platform/gtk/Skipped:
* platform/mac/Skipped:
* platform/qt/Skipped:
* platform/win/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUpdate TestFailures's title and header immediately upon navigation
aroben@apple.com [Thu, 7 Jul 2011 22:20:21 +0000 (22:20 +0000)]
Update TestFailures's title and header immediately upon navigation

Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks

Reviewed by Daniel Bates.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
title.
(ViewController.prototype._displayBuilder): Set the page's title and clear out any old
content immediately instead of waiting for the first set of results for the history
analyzer. As we receive new results from the analyzer we'll just clear out the main content
area and replace it with the new info.
(ViewController.prototype._displayTesters): Set the page's title and clear out any old
content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
in a little use of our removeAllChildren helper, too.
(ViewController.prototype._setTitle): Added. Set's the page's title and the header text.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90594 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoPatch by Jeff Timanus <twiz@chromium.org> on 2011-07-07
ojan@chromium.org [Thu, 7 Jul 2011 22:09:50 +0000 (22:09 +0000)]
Patch by Jeff Timanus <twiz@chromium.org> on 2011-07-07
Reviewed by Stephen White.

Mark layout tests as failing in Chromium after skia roll r1799:r1814
https://bugs.webkit.org/show_bug.cgi?id=64121

* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90593 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoPatch by James Robinson <jamesr@chromium.org> on 2011-07-07
commit-queue@webkit.org [Thu, 7 Jul 2011 22:09:36 +0000 (22:09 +0000)]
Patch by James Robinson <jamesr@chromium.org> on 2011-07-07
Reviewed by Kenneth Russell.

Use v8::AdjustAmountOfExternalAllocatedMemory for ArrayBuffers
https://bugs.webkit.org/show_bug.cgi?id=42912

This calls v8's AdjustAmountOfExternalAllocatedMemory when ArrayBuffers are allocated/deallocated so that V8's
garbage collection heuristics can account for the memory held by these objects.  On the new test page, this
reduces the peak memory use from 5BG+ (or a crash in 32-bit systems) to <300MB.

Test: WebCore/manual-tests/array-buffer-memory.html

* html/canvas/ArrayBuffer.cpp:
(WebCore::ArrayBuffer::~ArrayBuffer):
(WebCore::ArrayBuffer::tryAllocate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90592 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMove all code related to cachedSelection to HTMLTextFormControlElement
rniwa@webkit.org [Thu, 7 Jul 2011 21:37:26 +0000 (21:37 +0000)]
Move all code related to cachedSelection to HTMLTextFormControlElement
https://bugs.webkit.org/show_bug.cgi?id=64118

Reviewed by Alexey Proskuryakov.

Moved m_cachedSelectionStart and m_cachedSelectionEnd from HTMLInputElement
and HTMLTextAreaElement to HTMLTextFormControlElement

Also removed cached selection related functions from RenderTextControl,
RenderTextControlSingleLine, and RenderTextControlMultiLine because they were
merely providing wrapper functions to enable polymorphism between
input and textarea elements and their WML equivalents.

* editing/FrameSelection.cpp:
(WebCore::FrameSelection::notifyRendererOfSelectionChange): Calls HTMLTextFormControlElement's
selectionChanged instead of RenderTextControl's.
* html/HTMLFormControlElement.cpp:
(WebCore::HTMLTextFormControlElement::HTMLTextFormControlElement): Initialize
m_cachedSelectionStart and m_cachedSelectionEnd.
(WebCore::HTMLTextFormControlElement::selectionStart):
(WebCore::HTMLTextFormControlElement::selectionEnd):
(WebCore::HTMLTextFormControlElement::selection):
(WebCore::HTMLTextFormControlElement::restoreCachedSelection): Added.
(WebCore::HTMLTextFormControlElement::selectionChanged): Extracted from
RenderTextControl::selectionChanged.
* html/HTMLFormControlElement.h:
(WebCore::HTMLTextFormControlElement::cacheSelection): Extracted from HTMLInputElement
and HTMLTextAreaElement.
(WebCore::HTMLTextFormControlElement::hasCachedSelectionStart): Added.
(WebCore::HTMLTextFormControlElement::hasCachedSelectionEnd): Added.
* html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::HTMLInputElement):
(WebCore::HTMLInputElement::updateFocusAppearance):
* html/HTMLInputElement.h:
* html/HTMLTextAreaElement.cpp:
(WebCore::HTMLTextAreaElement::HTMLTextAreaElement):
(WebCore::HTMLTextAreaElement::updateFocusAppearance):
* html/HTMLTextAreaElement.h:
(WebCore::HTMLTextAreaElement::isEmptyValue):
* rendering/RenderTextControl.cpp:
(WebCore::setSelectionRange):
* rendering/RenderTextControl.h:
* rendering/RenderTextControlMultiLine.cpp:
* rendering/RenderTextControlMultiLine.h:
* rendering/RenderTextControlSingleLine.cpp:
* rendering/RenderTextControlSingleLine.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90591 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSwitch addFocusRingRects to to new layout types
eae@chromium.org [Thu, 7 Jul 2011 21:16:15 +0000 (21:16 +0000)]
Switch addFocusRingRects to to new layout types
https://bugs.webkit.org/show_bug.cgi?id=64114

Reviewed by Eric Seidel.

No new tests, no functionality changes.

* rendering/LayoutTypes.h:
(WebCore::flooredLayoutPoint):
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::addFocusRingRects):
* rendering/RenderBlock.h:
* rendering/RenderBox.cpp:
(WebCore::RenderBox::addFocusRingRects):
* rendering/RenderBox.h:
* rendering/RenderInline.cpp:
(WebCore::RenderInline::addFocusRingRects):
* rendering/RenderInline.h:
* rendering/RenderListBox.cpp:
(WebCore::RenderListBox::addFocusRingRects):
* rendering/RenderListBox.h:
* rendering/RenderObject.h:
(WebCore::RenderObject::addFocusRingRects):
* rendering/RenderTextControl.cpp:
(WebCore::RenderTextControl::addFocusRingRects):
* rendering/RenderTextControl.h:
* rendering/svg/RenderSVGContainer.cpp:
(WebCore::RenderSVGContainer::addFocusRingRects):
* rendering/svg/RenderSVGContainer.h:
* rendering/svg/RenderSVGImage.cpp:
(WebCore::RenderSVGImage::addFocusRingRects):
* rendering/svg/RenderSVGImage.h:
* rendering/svg/RenderSVGPath.cpp:
(WebCore::RenderSVGPath::addFocusRingRects):
* rendering/svg/RenderSVGPath.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90590 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[chromium] Reduce compositor texture memory by skipping layers and clipping surfaces
enne@google.com [Thu, 7 Jul 2011 21:06:58 +0000 (21:06 +0000)]
[chromium] Reduce compositor texture memory by skipping layers and clipping surfaces
https://bugs.webkit.org/show_bug.cgi?id=64052

Reviewed by James Robinson.

Layers and surfaces that are entirely transparent are now skipped.
Parent scissor rects are now applied to the content rect of surfaces
so that offscreen surfaces can be skipped.

Landing this for vangelis@chromium.org.

Covered by existing tests.

* platform/graphics/chromium/LayerRendererChromium.cpp:
(WebCore::calculateVisibleRect):
(WebCore::calculateVisibleLayerRect):
(WebCore::LayerRendererChromium::paintLayerContents):
(WebCore::LayerRendererChromium::drawLayers):
(WebCore::LayerRendererChromium::updatePropertiesAndRenderSurfaces):
(WebCore::LayerRendererChromium::updateCompositorResources):
(WebCore::LayerRendererChromium::drawLayer):
* platform/graphics/chromium/RenderSurfaceChromium.cpp:
(WebCore::RenderSurfaceChromium::RenderSurfaceChromium):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90589 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUnreviewed, rolling out r90581.
commit-queue@webkit.org [Thu, 7 Jul 2011 20:45:10 +0000 (20:45 +0000)]
Unreviewed, rolling out r90581.
http://trac.webkit.org/changeset/90581
https://bugs.webkit.org/show_bug.cgi?id=64124

Broke a bunch of inspector tests (Requested by ojan on
#webkit).

* inspector/front-end/ExtensionAPI.js:
(WebInspector.injectedExtensionAPI.Panels.prototype.create):
(WebInspector.injectedExtensionAPI.ExtensionSidebarPaneImpl.prototype.setPage):
(WebInspector.injectedExtensionAPI.expandURL):
* inspector/front-end/ExtensionPanel.js:
(WebInspector.ExtensionPanel):
* inspector/front-end/ExtensionServer.js:
(WebInspector.ExtensionServer):
(WebInspector.ExtensionServer.prototype._onCreatePanel):
(WebInspector.ExtensionServer.prototype._onSetSidebarPage):
(WebInspector.ExtensionServer.prototype._addExtensions):
(WebInspector.ExtensionServer.prototype._onWindowMessage):
(WebInspector.ExtensionServer.prototype._registerSubscriptionHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90588 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[chromium] Fix crash when compositing gets disabled mid-paint
enne@google.com [Thu, 7 Jul 2011 20:37:44 +0000 (20:37 +0000)]
[chromium] Fix crash when compositing gets disabled mid-paint
https://bugs.webkit.org/show_bug.cgi?id=64119

Reviewed by James Robinson.

WebKit paint can disable compositing and call setRootLayer(0). Adding
a check for this in drawLayers prevents a null pointer deref.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90587 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoEncode jump and link sizes into the appropriate enums
oliver@apple.com [Thu, 7 Jul 2011 20:26:19 +0000 (20:26 +0000)]
Encode jump and link sizes into the appropriate enums
https://bugs.webkit.org/show_bug.cgi?id=64123

Reviewed by Sam Weinig.

Finally kill off the out of line jump and link size arrays,
so we can avoid icky loads and constant fold the linking arithmetic.

* assembler/ARMv7Assembler.cpp:
* assembler/ARMv7Assembler.h:
(JSC::ARMv7Assembler::jumpSizeDelta):
(JSC::ARMv7Assembler::computeJumpType):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90586 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-07 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Thu, 7 Jul 2011 20:17:00 +0000 (20:17 +0000)]
2011-07-07  Dirk Schulze  <krit@webkit.org>

        SVGAnimateTransform does not support calcMode=discrete
        https://bugs.webkit.org/show_bug.cgi?id=63914

        Reviewed by Rob Buis.

        CalcMode discrete specifies that the animation function will jump from one value to the next without any interpolation.
        Implemented calcMode discrete by checking current progress of animation. If we are in the first half of the animation,
        we use the start value of the aniamtion, end value for the second half of the animation. The key time at 50% is used on
        all other animations as well as on other SVG viewers.

        Added a manual test. DRT crashes on an automated test with the SVG animation API. Opened a new bug report: 64104.

        * manual-tests/svg-animateTransform-calcMode-discrete.svg: Added.
        * svg/SVGAnimateTransformElement.cpp:
        (WebCore::SVGAnimateTransformElement::calculateAnimatedValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90585 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake Checkout use SCM's Executive instead of conjuring up its own
aroben@apple.com [Thu, 7 Jul 2011 19:13:46 +0000 (19:13 +0000)]
Make Checkout use SCM's Executive instead of conjuring up its own

This will improve integration with the rest of webkitpy, particularly when invoked via
webkit-patch.

Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
Executive inappropriately

Reviewed by Adam Barth.

* Scripts/webkitpy/common/checkout/checkout.py:
(Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
our own use. SCM might have some super-special Executive that it uses under the covers, and
we want to use it, too!

* Scripts/webkitpy/common/checkout/checkout_unittest.py:
(CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
method to call through to Executive.run_command.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90584 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake Term/ReadKey.pm an optional dependency of commit-log-editor
aroben@apple.com [Thu, 7 Jul 2011 18:47:22 +0000 (18:47 +0000)]
Make Term/ReadKey.pm an optional dependency of commit-log-editor

When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.

Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
Term/ReadKey.pm module

Reviewed by Xan Lopez.

* Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
that existing commit messages will be preserved. (Users can manually remove existing commit
messages using git-reset.)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90583 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoReviewed by Rob Buis.
krit@webkit.org [Thu, 7 Jul 2011 18:43:44 +0000 (18:43 +0000)]
Reviewed by Rob Buis.

Source/WebCore:

SVGAnimatedPath needs fallback for 'by' animation
https://bugs.webkit.org/show_bug.cgi?id=63865

Added fallback to 'to' animation for 'by' animated SVGAnimatedPathAnimator.

Clenup of SVGAnimateElement. Removed all switches since we support all AnimatedTypes
with the exception of AnimatedEnumeration. AnimatedEnumeration gets animated as AnimatedString at the moment.

Test: svg/animations/svgpath-animation-1.html

* svg/SVGAnimateElement.cpp:
(WebCore::getPropertyValue):
(WebCore::inheritsFromProperty):
(WebCore::attributeValueIsCurrentColor):
(WebCore::SVGAnimateElement::adjustForCurrentColor):
(WebCore::SVGAnimateElement::determineAnimatedAttributeType):
(WebCore::SVGAnimateElement::calculateAnimatedValue):
(WebCore::SVGAnimateElement::calculateFromAndToValues):
(WebCore::SVGAnimateElement::calculateFromAndByValues):
(WebCore::SVGAnimateElement::resetToBaseValue):
(WebCore::SVGAnimateElement::applyResultsToTarget):
(WebCore::SVGAnimateElement::calculateDistance):
* svg/SVGAnimatedPath.cpp:
(WebCore::SVGAnimatedPathAnimator::calculateFromAndByValues):

LayoutTests:

SVGAnimatedPath needs fallback for 'by' animation
https://bugs.webkit.org/show_bug.cgi?id=63865

Test fallback to 'to' animation for 'by' animated SVGPath.

* svg/animations/script-tests/svgpath-animation-1.js: Added.
(sample1):
(sample2):
(sample3):
(executeTest):
* svg/animations/svgpath-animation-1-expected.txt: Added.
* svg/animations/svgpath-animation-1.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90582 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-07 Andrey Kosyakov <caseq@chromium.org>
caseq@chromium.org [Thu, 7 Jul 2011 18:38:18 +0000 (18:38 +0000)]
2011-07-07  Andrey Kosyakov  <caseq@chromium.org>

        Web Inspector: secure access to extensions API
        https://bugs.webkit.org/show_bug.cgi?id=64080

        Reviewed by Pavel Feldman.

        * inspector/front-end/ExtensionAPI.js:
        (WebInspector.injectedExtensionAPI.Panels.prototype.create):
        (WebInspector.injectedExtensionAPI.ExtensionSidebarPaneImpl.prototype.setPage):
        * inspector/front-end/ExtensionPanel.js:
        (WebInspector.ExtensionPanel):
        * inspector/front-end/ExtensionServer.js:
        (WebInspector.ExtensionServer):
        (WebInspector.ExtensionServer.prototype._onCreatePanel):
        (WebInspector.ExtensionServer.prototype._onSetSidebarPage):
        (WebInspector.ExtensionServer.prototype._addExtensions):
        (WebInspector.ExtensionServer.prototype._onWindowMessage):
        (WebInspector.ExtensionServer.prototype._registerSubscriptionHandler):
        (WebInspector.ExtensionServer.prototype._expandResourcePath):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90581 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-07 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 7 Jul 2011 18:15:06 +0000 (18:15 +0000)]
2011-07-07  Gustavo Noronha Silva  <gns@gnome.org>

        Unreviewed.

        Re-skip the inspector tests, they are still timing out on the
        release bot for some reason.

        * platform/gtk/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90580 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoKeep testing that commit messages containing Unicode are handled properly by webkitpy
aroben@apple.com [Thu, 7 Jul 2011 18:02:03 +0000 (18:02 +0000)]
Keep testing that commit messages containing Unicode are handled properly by webkitpy

Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
correctly handle Unicode in commit messages

Reviewed by Eric Seidel.

* Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
places in the ChangeLog that will end up being included in the commit message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90579 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt][WK2] Remove Symbian code.
andreas.kling@nokia.com [Thu, 7 Jul 2011 17:45:32 +0000 (17:45 +0000)]
[Qt][WK2] Remove Symbian code.
https://bugs.webkit.org/show_bug.cgi?id=64101

Reviewed by Benjamin Poulain.

Source/WebKit2:

* DerivedSources.pro:
* Platform/CoreIPC/Attachment.h:
* Platform/CoreIPC/Connection.h:
* Platform/SharedMemory.h:
* Platform/qt/SharedMemorySymbian.cpp: Removed.
* UIProcess/API/qt/tests/tests.pri:
* UIProcess/Launcher/qt/ProcessLauncherQt.cpp:
* WebKit2.pri:
* WebKit2.pro:
* WebProcess.pro:
* config.h:

Tools:

* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
(BrowserWindow::screenshot):
* MiniBrowser/qt/MiniBrowser.pro:
* MiniBrowser/qt/MiniBrowserApplication.h:
(WindowOptions::WindowOptions):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90578 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
andreas.kling@nokia.com [Thu, 7 Jul 2011 17:43:30 +0000 (17:43 +0000)]
[Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
https://bugs.webkit.org/show_bug.cgi?id=64099

Reviewed by Benjamin Poulain.

Source/WebKit2:

* Shared/qt/WebEventFactoryQt.cpp:
(WebKit::WebEventFactory::createWebWheelEvent):
* UIProcess/API/qt/qdesktopwebview_p.h:
* UIProcess/qt/QtWebPageProxy.cpp:
(QtWebPageProxy::QtWebPageProxy):
(QtWebPageProxy::setCursor):
(QtWebPageProxy::registerEditCommand):
(QtWebPageProxy::clearAllEditCommands):
(QtWebPageProxy::canUndoRedo):
(QtWebPageProxy::executeUndoRedo):
(QtWebPageProxy::updateAction):
(QtWebPageProxy::webActionTriggered):
(QtWebPageProxy::triggerAction):
(QtWebPageProxy::action):
* UIProcess/qt/QtWebPageProxy.h:
* UIProcess/qt/TouchViewInterface.cpp:
(WebKit::TouchViewInterface::didChangeCursor):
* UIProcess/qt/TouchViewInterface.h:
* UIProcess/qt/ViewInterface.h:
* UIProcess/qt/WebUndoCommandQt.cpp:
(WebUndoCommandQt::WebUndoCommandQt):
* UIProcess/qt/WebUndoCommandQt.h:
* WebProcess/WebCoreSupport/qt/WebDragClientQt.cpp:
(WebKit::WebDragClient::startDrag):

Tools:

* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::openFile):
(BrowserWindow::screenshot):
(BrowserWindow::loadURLListFromFile):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90577 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSkip another test that uses beginDragWithFiles on Windows
aroben@apple.com [Thu, 7 Jul 2011 17:39:56 +0000 (17:39 +0000)]
Skip another test that uses beginDragWithFiles on Windows

* platform/win/Skipped: Added fast/forms/file-input-change-event.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90576 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUpdated change log
mitz@apple.com [Thu, 7 Jul 2011 17:33:06 +0000 (17:33 +0000)]
Updated change log

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90575 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoThe return value of SharedBuffer::createWithContentsOfFile must have valid m_size.
commit-queue@webkit.org [Thu, 7 Jul 2011 17:29:20 +0000 (17:29 +0000)]
The return value of SharedBuffer::createWithContentsOfFile must have valid m_size.
https://bugs.webkit.org/show_bug.cgi?id=63313

Patch by Kyusun Kim <maniagoon@company100.net> on 2011-07-07
Reviewed by Eric Seidel.

* platform/posix/SharedBufferPOSIX.cpp:
(WebCore::SharedBuffer::createWithContentsOfFile):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90574 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago<rdar://problem/9737435> Re-enable -webkit-column-break-inside: avoid
mitz@apple.com [Thu, 7 Jul 2011 17:27:46 +0000 (17:27 +0000)]
<rdar://problem/9737435> Re-enable -webkit-column-break-inside: avoid
https://bugs.webkit.org/show_bug.cgi?id=64102

Reviewed by Simon Fraser.

Source/WebCore:

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::adjustForUnsplittableChild): Treat the child as unsplittable if
'-webkit-column-break-inside: avoid' was specified.

LayoutTests:

* fast/multicol/break-properties-expected.txt:
* fast/multicol/break-properties.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90573 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt][WK2] Views should know about WebProcess crash/relaunch.
andreas.kling@nokia.com [Thu, 7 Jul 2011 17:27:33 +0000 (17:27 +0000)]
[Qt][WK2] Views should know about WebProcess crash/relaunch.
https://bugs.webkit.org/show_bug.cgi?id=64093

Reviewed by Benjamin Poulain.

Add processDidCrash() and didRelaunchProcess() to ViewInterface.

QDesktopWebView now displays a simple sad smiley ":(" when the
web process crashes.

* UIProcess/API/qt/qdesktopwebview.cpp:
(QDesktopWebViewPrivate::QDesktopWebViewPrivate):
(paintCrashedPage):
(QDesktopWebView::paint):
(QDesktopWebViewPrivate::processDidCrash):
(QDesktopWebViewPrivate::didRelaunchProcess):
* UIProcess/API/qt/qdesktopwebview_p.h:
* UIProcess/qt/QtWebPageProxy.cpp:
(QtWebPageProxy::QtWebPageProxy):
(QtWebPageProxy::didRelaunchProcess):
(QtWebPageProxy::processDidCrash):
* UIProcess/qt/QtWebPageProxy.h:
* UIProcess/qt/TouchViewInterface.cpp:
(WebKit::TouchViewInterface::showContextMenu):
(WebKit::TouchViewInterface::hideContextMenu):
(WebKit::TouchViewInterface::processDidCrash):
(WebKit::TouchViewInterface::didRelaunchProcess):
* UIProcess/qt/TouchViewInterface.h:
* UIProcess/qt/ViewInterface.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90572 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoCompletely ignore too-many-failures builds in TestFailures in most circumstances
aroben@apple.com [Thu, 7 Jul 2011 17:18:09 +0000 (17:18 +0000)]
Completely ignore too-many-failures builds in TestFailures in most circumstances

Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
really use them to perform regression analysis. The only time we want to pay attention to
too-many-failures builds is when we're trying to explain when the current bout of
too-many-failures started.

Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
in a build that didn't even run it (because it exited early due to too many failues)

Reviewed by David Kilzer.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
too-many-failures handling that would cause us to blame builds that didn't even run a given
test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
build was itself a too-many-failures build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90571 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt]REGRESSION(r90471): It made 4 fast/notifications tests fail on Qt
abecsi@webkit.org [Thu, 7 Jul 2011 17:17:18 +0000 (17:17 +0000)]
[Qt]REGRESSION(r90471): It made 4 fast/notifications tests fail on Qt
https://bugs.webkit.org/show_bug.cgi?id=63255

Unreviewed gardening.

* platform/qt/Skipped: Skip tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90570 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWeb Inspector: visual regressions in network panel
caseq@chromium.org [Thu, 7 Jul 2011 17:15:20 +0000 (17:15 +0000)]
Web Inspector: visual regressions in network panel
https://bugs.webkit.org/show_bug.cgi?id=64089

Reviewed by Pavel Feldman.

* inspector/front-end/networkPanel.css:
(.data-grid.network-log-grid tr.filler td):
(.network.panel .sidebar):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90569 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoPatch by Julien Chaffraix <jchaffraix@webkit.org> on 2011-07-07
jchaffraix@webkit.org [Thu, 7 Jul 2011 17:05:25 +0000 (17:05 +0000)]
Patch by Julien Chaffraix <jchaffraix@webkit.org> on 2011-07-07
Reviewed by David Hyatt.

Partial layout when a flex-box has visibility: collapse
https://bugs.webkit.org/show_bug.cgi?id=63776

Source/WebCore:

Tests: fast/flexbox/crash-button-input-autofocus.html
       fast/flexbox/crash-button-keygen.html
       fast/flexbox/crash-button-relayout.html

The issue is that FlexBoxIterator would skip any child if it has visibility: collapsed.
However if one of the child is anonymous, it may wrap some other child that would be skipped.
Now FlexBoxIterator is called during the layout phase and thus some nodes would not relayouted
as expected.

* rendering/RenderDeprecatedFlexibleBox.cpp:
(WebCore::FlexBoxIterator::next): When iterating, don't skip anonymous content as there may
be real content hiding below.

LayoutTests:

Those tests checks some variation of the same underlying issue.

* fast/flexbox/crash-button-input-autofocus-expected.txt: Added.
* fast/flexbox/crash-button-input-autofocus.html: Added.
* fast/flexbox/crash-button-keygen-expected.txt: Added.
* fast/flexbox/crash-button-keygen.html: Added.
* fast/flexbox/crash-button-relayout-expected.txt: Added.
* fast/flexbox/crash-button-relayout.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90568 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-07 Vsevolod Vlasov <vsevik@chromium.org>
vsevik@chromium.org [Thu, 7 Jul 2011 16:53:51 +0000 (16:53 +0000)]
2011-07-07  Vsevolod Vlasov  <vsevik@chromium.org>

        Web Inspector: URL links in styles open new tab instead of showing resources panel.
        https://bugs.webkit.org/show_bug.cgi?id=64090

        Reviewed by Pavel Feldman.

        * inspector/styles/styles-url-linkify-expected.txt:
        * inspector/styles/styles-url-linkify.html:
2011-07-07  Vsevolod Vlasov  <vsevik@chromium.org>

        Web Inspector: URL links in styles open new tab instead of showing resources panel.
        https://bugs.webkit.org/show_bug.cgi?id=64090

        Reviewed by Pavel Feldman.

        * inspector/front-end/StylesSidebarPane.js:
        (WebInspector.StylePropertyTreeElement.prototype.updateTitle.linkifyURL):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90567 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoREGRESSION(90557) http/tests/inspector/network/network-embed.html fails
abecsi@webkit.org [Thu, 7 Jul 2011 16:52:27 +0000 (16:52 +0000)]
REGRESSION(90557) http/tests/inspector/network/network-embed.html fails
https://bugs.webkit.org/show_bug.cgi?id=64103

Unreviewed gardening.

* platform/qt/Skipped: Skip test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90566 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-07 Andrey Kosyakov <caseq@chromium.org>
caseq@chromium.org [Thu, 7 Jul 2011 16:48:27 +0000 (16:48 +0000)]
2011-07-07  Andrey Kosyakov  <caseq@chromium.org>

        Web Inspector: active panel does not receive resize event when drawer is resized
        https://bugs.webkit.org/show_bug.cgi?id=64094

        Reviewed by Pavel Feldman.

        * inspector/front-end/Drawer.js:
        (WebInspector.Drawer.prototype._statusBarDragging):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90565 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoTeach webkitpy's Checkout class to use commit-log-editor to create commit messages
aroben@apple.com [Thu, 7 Jul 2011 16:36:32 +0000 (16:36 +0000)]
Teach webkitpy's Checkout class to use commit-log-editor to create commit messages

Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
commit-log-editor's

Reviewed by David Kilzer.

* Scripts/webkitpy/common/checkout/checkout.py:
(Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
the modified ChangeLogs, to generate the commit message, rather than trying to generate one
ourselves.

* Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
to match commit-log-editor's format.
(CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
directories so we can see how the various entries get labeled in the commit message.
(CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
instance that knows how to find commit-log-editor and pass it to our Checkout instance.
Don't bother capturing output, since there shouldn't be any.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90564 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoAdd a --print-log option to commit-log-editor
aroben@apple.com [Thu, 7 Jul 2011 16:36:23 +0000 (16:36 +0000)]
Add a --print-log option to commit-log-editor

When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
generates a commit message from those ChangeLogs, prints it to stdout, and exits.

Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
readable than commit-log-editor's

Reviewed by David Kilzer.

* Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
option, which calls through to createCommitMessage, prints the result, and exits.
(printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90563 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoExtract commit-log-editor's commit-message-generation code into a separate function
aroben@apple.com [Thu, 7 Jul 2011 16:35:33 +0000 (16:35 +0000)]
Extract commit-log-editor's commit-message-generation code into a separate function

As a bonus, we now skip a bunch of work in the case where we are generating a commit message
for staged git changes.

Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
readable than commit-log-editor's

Reviewed by David Kilzer.

* Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
file so that it will be defined even if we aren't operating on an existing log message.
Moved a bunch of top-level code to generate the commit message from here...
(createCommitMessage): ...to here.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90562 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-07-07 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Thu, 7 Jul 2011 16:20:12 +0000 (16:20 +0000)]
2011-07-07  Pavel Feldman  <pfeldman@chromium.org>

        Web Inspector: prevent default action during drag'n'drop in the Elements panel.
        https://bugs.webkit.org/show_bug.cgi?id=64081

        Reviewed by Yury Semikhatsky.

        * inspector/front-end/ElementsTreeOutline.js:
        (WebInspector.ElementsTreeOutline):
        (WebInspector.ElementsTreeOutline.prototype._ondragstart):
        (WebInspector.ElementsTreeOutline.prototype._ondragover):
        (WebInspector.ElementsTreeOutline.prototype._ondragend):
        * inspector/front-end/treeoutline.js:
        (TreeOutline.prototype.treeElementFromPoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90561 268f45cc-cd09-0410-ab3c-d52691b4dbfc