Context Navigation

← Previous Period
Next Period →

Timeline

Aug 1, 2011:

8:47 PM Changeset in webkit [92166] by abarth@webkit.org

6 edits in trunk/Tools

garden-o-matic should call optimize-baselines when rebaselining tests
https://bugs.webkit.org/show_bug.cgi?id=65499

Reviewed by Dimitri Glazkov.

I took the opportunity to modernize this code to use some of our more
powerful primitives from base.

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
Scripts/webkitpy/tool/servers/gardeningserver.py:
Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:

8:09 PM Changeset in webkit [92165] by andersca@apple.com

5 edits in trunk/LayoutTests

Update Lion specific test results.

platform/mac/fast/forms/input-placeholder-visibility-1-expected.txt:
platform/mac/fast/forms/input-placeholder-visibility-3-expected.txt:
platform/mac/fast/forms/textarea-placeholder-visibility-1-expected.txt:
platform/mac/fast/forms/textarea-placeholder-visibility-2-expected.txt:

6:42 PM Changeset in webkit [92164] by macpherson@chromium.org

5 edits in trunk/Source/WebCore

Don't set m_fontDirty when setting zoom unless zoom has actually changed
https://bugs.webkit.org/show_bug.cgi?id=65092

Reviewed by Darin Adler.

No new tests as no functionality changed - this is an optimization that
should be logically equivalent to the current code.

The intent here is to avoid setting m_fontDirty unless the fornt information is actually dirty.

css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::styleForDocument):
(WebCore::CSSStyleSelector::applyProperty):
Use setZoom and setEffectiveZoom wrapper functions.

css/CSSStyleSelector.h:

(WebCore::CSSStyleSelector::setZoom):
Wrapper for m_style->setZoom() that automatically updates m_fontDirty.
(WebCore::CSSStyleSelector::setEffectiveZoom):
Wrapper for m_style->setEffectiveZoom that automatically updates m_fontDirty.

page/animation/AnimationBase.cpp:

(WebCore::AnimationBase::ensurePropertyMap):
Use

rendering/style/RenderStyle.h:

(WebCore::InheritedFlags::setZoom):
Return true only if underlying values change.
(WebCore::InheritedFlags::setZoomWithoutReturnValue):
Return void to match function pointer type where required.
(WebCore::InheritedFlags::setEffectiveZoom):
Return true only if underlying values change.

6:37 PM Changeset in webkit [92163] by timothy_horton@apple.com

1 edit
2 adds in trunk/LayoutTests

Unreviewed, rebaseline test results for recently added test.

platform/chromium-linux/svg/custom/repaint-shadow-expected.png: Added.
platform/qt/svg/custom/repaint-shadow-expected.txt: Added.

6:23 PM Changeset in webkit [92162] by zmo@google.com

37 edits
2 deletes in trunk/LayoutTests

Unreviewed, rebaseline, webkit gardener stuff.

platform/chromium-win-vista/fast/text/international/hindi-spacing-expected.png: Removed.
platform/chromium-win-vista/fast/text/international/khmer-selection-expected.png: Removed.
platform/chromium-win-xp/fast/text/backslash-to-yen-sign-euc-expected.png:
platform/chromium-win-xp/fast/text/backslash-to-yen-sign-expected.png:
platform/chromium-win-xp/fast/text/international/hindi-spacing-expected.png:
platform/chromium-win-xp/fast/text/international/khmer-selection-expected.png:
platform/chromium-win-xp/svg/batik/text/verticalText-expected.png:
platform/chromium-win/fast/text/backslash-to-yen-sign-euc-expected.png:
platform/chromium-win/fast/text/backslash-to-yen-sign-expected.png:
platform/chromium-win/fast/text/international/hindi-spacing-expected.png:
platform/chromium-win/fast/text/international/khmer-selection-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirLTR-anchorEnd-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirLTR-anchorMiddle-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirLTR-anchorStart-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirNone-anchorEnd-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirNone-anchorMiddle-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirNone-anchorStart-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirRTL-anchorEnd-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirRTL-anchorMiddle-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-dirRTL-anchorStart-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-inherited-dirLTR-anchorEnd-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-inherited-dirLTR-anchorMiddle-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-inherited-dirLTR-anchorStart-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-inherited-dirRTL-anchorEnd-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-inherited-dirRTL-anchorMiddle-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-inherited-dirRTL-anchorStart-expected.png:
platform/chromium-win/svg/W3C-I18N/text-anchor-no-markup-expected.png:
platform/chromium-win/svg/batik/filters/feTile-expected.png:
platform/chromium-win/svg/batik/filters/filterRegions-expected.png:
platform/chromium-win/svg/batik/text/longTextOnPath-expected.png:
platform/chromium-win/svg/batik/text/textAnchor-expected.png:
platform/chromium-win/svg/batik/text/textFeatures-expected.png:
platform/chromium-win/svg/batik/text/textLayout-expected.png:
platform/chromium-win/svg/batik/text/textProperties-expected.png:
platform/chromium-win/svg/batik/text/textProperties2-expected.png:
platform/chromium-win/svg/batik/text/textStyles-expected.png:
platform/chromium-win/svg/batik/text/verticalText-expected.png:
platform/chromium-win/svg/css/text-shadow-multiple-expected.png:

6:20 PM Changeset in webkit [92161] by commit-queue@webkit.org

6 edits in trunk/Source

Removing old source files in gyp files that slow build
https://bugs.webkit.org/show_bug.cgi?id=65503

Patch by Jean-luc Brouillet <jeanluc@chromium.org> on 2011-08-01
Reviewed by Adam Barth.

A number of stale files are listed in the gyp files. These slow the
build on Visual Studio 2010. Removing them.

Source/JavaScriptCore:

JavaScriptCore.gypi:

Source/WebCore:

WebCore.gypi:

Source/WebKit/chromium:

WebKit.gyp:

6:16 PM Changeset in webkit [92160] by Stephanie Lewis

4 edits in trunk/Tools

Finish reconfiguration started by Lucas Forschler
Reconfigured slaves:
Removed Leopard Intel Release (Build)
Removed Leopard Intel Release (Tests)
Remove apple-pixel-1
Added Lion slaves from disabled Leopard slaves

Added builders:
Lion Intel Release (Build)
Lion Intel Debug (Build)
Lion Intel Release (Tests)
Lion Intel Release WK2 (Tests)
Lion Intel Debug (Tests)
Lion Intel Debug WK2 (Tests)

Added triggers:
lion-intel-debug-tests
lion-intel-debug-test-wk2
lion-intel-release-tests
lion-intel-release-tests-wk2

Removed trigger:
leopard-intel-release-tests

Added PlatformSpecificScheduler:
mac-lion

Added "mac-lion" to build.py and build_unitttest.py

Reviewed by Mark Rowe.

BuildSlaveSupport/build.webkit.org-config/config.json:
Scripts/webkitpy/common/config/build.py:
Scripts/webkitpy/common/config/build_unittest.py:

5:59 PM Changeset in webkit [92159] by commit-queue@webkit.org

2 edits in trunk/Source/WebCore

Shard webcore_remaining for msvs official builds

https://bugs.webkit.org/show_bug.cgi?id=65475

Patch by Bradley Nelson <bradnelson@chromium.org> on 2011-08-01
Reviewed by Darin Fisher.

webcore_remaining has consistently exceeded Visual Studio 2008's
capacity to link when link time code generation is turned on (full
release builds). This has been dealt with the past by carving off
larger and larger pieces of webcore_remaining (ie webcore_html,
webcore_renderer etc).
A new option has been added to gyp to automate this sharding:
'msvs_shard': SHARD_COUNT,
Targets marked in this way will be built in several pieces
(target_0, target_1...).
Since the resulting project is less comprehensible in the IDE (and
since the linking limitation only afflicts official builds), this
option is only needed for those builds.
This patch shards webcore_remaining into 10 parts for official builds.

WebCore.gyp/WebCore.gyp:

5:51 PM Changeset in webkit [92158] by commit-queue@webkit.org

6 edits in trunk

WAI-ARIA textbox role should be AXTextArea for aria-multiline=true, AXTextField for aria-multiline=false
https://bugs.webkit.org/show_bug.cgi?id=65451

Patch by Alice Boxhall <aboxhall@chromium.org> on 2011-08-01
Reviewed by Chris Fleizach.

Source/WebCore:

accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::determineAriaRoleAttribute):

LayoutTests:

accessibility/aria-roles.html:
platform/mac/accessibility/aria-multiline.html:
platform/mac/accessibility/aria-roles-expected.txt:

5:49 PM Changeset in webkit [92157] by commit-queue@webkit.org

2 edits in trunk/Source/WebCore

Use threadsafeCopy() API when using Database Directory path variable
https://bugs.webkit.org/show_bug.cgi?id=65436

Patch by srikumar bonda <srikumar.b@gmail.com> on 2011-08-01
Reviewed by David Levin.

Use threadsafeCopy() API while using m_databaseDirectoryPath string variable
when the databaseGuard lock hasn't been taken in DatabaseTracker class.
This is just a safety fix.

No changes to functionality so no new tests.

storage/DatabaseTracker.cpp:

(WebCore::DatabaseTracker::trackerDatabasePath):
(WebCore::DatabaseTracker::deleteOrigin):

5:44 PM Changeset in webkit [92156] by zmo@google.com

1 edit
2 adds in trunk/LayoutTests

Unreviewed, rebaseline, webkit gardener stuff.

platform/chromium-mac/svg/custom/repaint-shadow-expected.png: Added.
platform/chromium-win/svg/custom/repaint-shadow-expected.png: Added.

5:21 PM Changeset in webkit [92155] by Michael Nordman

11 edits in trunk/Source/WebCore

[Chromium] WebSQLDatabase version handling is broken in multi-process browsers.
https://bugs.webkit.org/show_bug.cgi?id=65486

The WebCore::AbstractDatabase class maintains a global in-memory map of
the version numbers associated with open database files, but that map is
not reliable in a multi-process system like Chrome. So instead of relying
on the cached values in that map, we read the value from the database (and
update the cached value) where possible. There are two edge cases where that's
not possible because the scriptable interface requires synchronous access
to the version: the .version attribute getter and the .openDatabase() method.
In those cases, we have no choice but to use the potentially stale cached value.

Reviewed by Darin Fisher.

No new tests. Existing layout tests cover the version handling functionality.

storage/AbstractDatabase.cpp:

(WebCore::AbstractDatabase::version):
(WebCore::AbstractDatabase::performOpenAndVerify):
(WebCore::AbstractDatabase::getVersionFromDatabase):
(WebCore::AbstractDatabase::setVersionInDatabase):
(WebCore::AbstractDatabase::setExpectedVersion):
(WebCore::AbstractDatabase::getCachedVersion):
(WebCore::AbstractDatabase::setCachedVersion):
(WebCore::AbstractDatabase::getActualVersionForTransaction):

storage/AbstractDatabase.h:

(WebCore::AbstractDatabase::expectedVersion):

storage/ChangeVersionWrapper.cpp:

(WebCore::ChangeVersionWrapper::handleCommitFailedAfterPostflight):

storage/ChangeVersionWrapper.h:
storage/Database.cpp:

(WebCore::Database::openDatabase):

storage/DatabaseSync.cpp:

(WebCore::DatabaseSync::openDatabaseSync):
(WebCore::DatabaseSync::changeVersion):

storage/SQLTransaction.cpp:

(WebCore::SQLTransaction::SQLTransaction):
(WebCore::SQLTransaction::executeSQL):
(WebCore::SQLTransaction::openTransactionAndPreflight):
(WebCore::SQLTransaction::runCurrentStatement):
(WebCore::SQLTransaction::postflightAndCommit):

storage/SQLTransaction.h:
storage/SQLTransactionSync.cpp:

(WebCore::SQLTransactionSync::SQLTransactionSync):
(WebCore::SQLTransactionSync::executeSQL):
(WebCore::SQLTransactionSync::begin):

storage/SQLTransactionSync.h:

5:10 PM Changeset in webkit [92154] by levin@chromium.org

4 edits in trunk/Source/JavaScriptCore

currentThread is too slow!
https://bugs.webkit.org/show_bug.cgi?id=64577

Reviewed by Darin Adler and Dmitry Titov.

The problem is that currentThread results in a pthread_once call which always takes a lock.
With this change, currentThread is 10% faster than isMainThread in release mode and only
5% slower than isMainThread in debug.

wtf/ThreadIdentifierDataPthreads.cpp:

(WTF::ThreadIdentifierData::initializeOnce): Remove the pthread once stuff
which is no longer needed because this is called from initializeThreading().
(WTF::ThreadIdentifierData::identifier): Remove the initializeKeyOnce call because
intialization of the pthread key should already be done.
(WTF::ThreadIdentifierData::initialize): Ditto.

wtf/ThreadIdentifierDataPthreads.h:
wtf/ThreadingPthreads.cpp:

(WTF::initializeThreading): Acquire the pthread key here.

5:07 PM Changeset in webkit [92153] by abarth@webkit.org

5 edits
2 adds in trunk/Tools

webkit-patch needs to be able to "optimize" the storage of baselines on disk
https://bugs.webkit.org/show_bug.cgi?id=65418

Reviewed by Dimitri Glazkov.

If we're not careful when rebaselining tests, we can end up with lots
of duplicate expected results files in the tree. This patch adds a
webkit-patch command that optimizes the storage of expected results on
disk.

This command is similar to deduplicate-tests, except that it can move
test results around rather than just remove duplicate results.

Unfortunately, this problem is very tricky because the baseline search
structure is a hypergraph. This patch include a huerstic optimizer
that appears to work on a bunch of examples I've tried. We'll likely
need to refine it as gain experience.

Scripts/webkitpy/common/system/filesystem.py:
Scripts/webkitpy/tool/commands/rebaseline.py:

4:49 PM Changeset in webkit [92152] by timothy_horton@apple.com

5 edits
3 adds in trunk

Severe shadow repaint issues with SVGText elements
https://bugs.webkit.org/show_bug.cgi?id=63648
<rdar://problem/7632269>

Reviewed by Darin Adler.

Make SVG elements which inherit from RenderSVGBlock take text-shadow
overflow into account when determining whether or not they should draw

Test: svg/custom/repaint-shadow.svg

rendering/RenderBox.h:

(WebCore::RenderBox::visualOverflowRect):

rendering/svg/RenderSVGBlock.cpp:

(WebCore::RenderSVGBlock::visualOverflowRect):

rendering/svg/RenderSVGBlock.h:

4:25 PM Changeset in webkit [92151] by crogers@google.com

2 edits in trunk/LayoutTests

One more test_expectations.txt fix for windows web audio
https://bugs.webkit.org/show_bug.cgi?id=65496

Unreviewed.

platform/chromium/test_expectations.txt:

4:12 PM Changeset in webkit [92150] by zmo@google.com

9 edits in trunk

Unreviewed, rolling out r92022.
http://trac.webkit.org/changeset/92022
https://bugs.webkit.org/show_bug.cgi?id=65203

try to see if this is causing a bunch of browser tests failues
on win bots - will recommit if not

Source/WebCore:

platform/graphics/chromium/FontChromiumWin.cpp:

(WebCore::TransparencyAwareFontPainter::TransparencyAwareFontPainter::TransparencyAwareFontPainter):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareFontPainter::init):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareFontPainter::initializeForGDI):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareFontPainter::~TransparencyAwareFontPainter):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareGlyphPainter::TransparencyAwareGlyphPainter):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareGlyphPainter::~TransparencyAwareGlyphPainter):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareGlyphPainter::drawGlyphs):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareUniscribePainter::hdc):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareUniscribePainter::TransparencyAwareUniscribePainter):
(WebCore::TransparencyAwareFontPainter::TransparencyAwareUniscribePainter::~TransparencyAwareUniscribePainter):
(WebCore::drawGlyphsWin):
(WebCore::Font::drawComplexText):

platform/graphics/chromium/UniscribeHelper.cpp:

(WebCore::UniscribeHelper::draw):

platform/graphics/skia/PlatformContextSkia.cpp:

(WebCore::PlatformContextSkia::isNativeFontRenderingAllowed):

platform/graphics/skia/PlatformContextSkia.h:
platform/graphics/skia/SkiaFontWin.cpp:

(WebCore::windowsCanHandleDrawTextShadow):
(WebCore::windowsCanHandleTextDrawing):
(WebCore::windowsCanHandleTextDrawingWithoutShadow):

platform/graphics/skia/SkiaFontWin.h:

LayoutTests:

platform/chromium/test_expectations.txt:

3:45 PM Changeset in webkit [92149] by rniwa@webkit.org

2 edits in trunk/LayoutTests

Remove time out expectation since media/track has been added to the skipped list on Qt.

platform/qt/test_expectations.txt:

3:32 PM Changeset in webkit [92148] by fpizlo@apple.com

4 edits in trunk/Source/JavaScriptCore

DFG JIT sometimes creates speculation check data structures that have
invalid information about the format of a register
https://bugs.webkit.org/show_bug.cgi?id=65490

Reviewed by Gavin Barraclough.

The code now makes sure to (1) always have correct and up-to-date
information about register format at the time that a speculation
check is emitted, (2) assert that speculation data is correct
inside the speculation check implementation, and (3) avoid creating
speculation data altogether if compilation has already failed, since
at that point the format data is almost guaranteed to be bogus.

dfg/DFGNonSpeculativeJIT.cpp:

(JSC::DFG::EntryLocation::EntryLocation):

dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculationCheck::SpeculationCheck):
(JSC::DFG::SpeculativeJIT::fillSpeculateCell):
(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGSpeculativeJIT.h:

(JSC::DFG::SpeculativeJIT::speculationCheck):

3:22 PM Changeset in webkit [92147] by Dimitri Glazkov

5 edits in trunk/Tools

Teach TestExpectationSerializer about parsed expectations.
https://bugs.webkit.org/show_bug.cgi?id=65444

This is a bit largish in size, but the essentials are:

TestExpectationSerializer, given a parsed TestExpectationLine, can now spit out a pretty string.
We now store original content of the expectation line on TestExpectationLine.
Lots of tests.

Reviewed by Adam Barth.

Scripts/webkitpy/layout_tests/models/test_configuration.py: Fixed an issue where we didn't

populate macros correctly, also fixed an issue where we incorrectly returned an empty list
for all configurations.

Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added tests for those.
Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed vestiges of code long gone,

refactored TestExpectationsSerializer to recognize and correctly serialize parsed TestExpectationLine instances,
changed to store and use TestExpectationLine.original_string.

Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.

3:09 PM Changeset in webkit [92146] by fpizlo@apple.com

2 edits in trunk/Source/JavaScriptCore

REGRESSION(r92092): Build fails on 64 bit
https://bugs.webkit.org/show_bug.cgi?id=65458

Reviewed by Oliver Hunt.

The build was broken because some compilers were smart enough to see
an array index out of bounds due to the decision fuction for when to
go from precise size classes to imprecise size classes being broken:
it would assume that sizes in the range 97..128 belonged to a precise
size class when in fact they belonged to an imprecise one.

In fact, the code would have run correctly, by way of a fluke, because
though the 4th precise size class (for 97..128) didn't exist, the next
array over from m_preciseSizeClasses was m_impreciseSizeClasses, and
its first entry would have been a size class that is appropriate for
allocations in the range 97..128. However, this relies on specific
ordering of fields in NewSpace, so it's still a bug.

This fixes the bug by ensuring that allocations larger than 96 use
the imprecise size classes.

heap/NewSpace.h:

(JSC::NewSpace::sizeClassFor):

2:46 PM Changeset in webkit [92145] by nduca@chromium.org

2 edits in trunk/Source/WebKit/chromium

[chromium] Add null check to setSwapbuffersCompletCallbackChromium
https://bugs.webkit.org/show_bug.cgi?id=65485

Reviewed by James Robinson.

src/GraphicsContext3DChromium.cpp:

(WebCore::SwapBuffersCompleteCallbackAdapter::create):

2:41 PM Changeset in webkit [92144] by crogers@google.com

1 edit
5 adds in trunk/LayoutTests

Add audionode.html and gain.html web audio layout tests
https://bugs.webkit.org/show_bug.cgi?id=65276

Reviewed by Kenneth Russell.

webaudio/audionode-expected.txt: Added.
webaudio/audionode.html: Added.
webaudio/gain-expected.wav: Added.
webaudio/gain.html: Added.
webaudio/resources/buffer-loader.js: Added.

(BufferLoader):
(BufferLoader.prototype.loadBuffer.request.onload):
(BufferLoader.prototype.loadBuffer.request.onerror):
(BufferLoader.prototype.loadBuffer):
(BufferLoader.prototype.load):

2:22 PM Changeset in webkit [92143] by commit-queue@webkit.org

18 edits
2 adds in trunk

.: REGRESSION (r39725?): Resources removed from document can not be freed until the document is deleted
https://bugs.webkit.org/show_bug.cgi?id=61006

Patch by Scott Graham <scottmg@chromium.org> on 2011-08-01
Reviewed by Antti Koivisto.

Update exports for test harness.

Source/autotools/symbols.filter:

Source/WebCore: REGRESSION (r39725?): Resources removed from document can not be freed
until the document is deleted
https://bugs.webkit.org/show_bug.cgi?id=61006

Patch by Scott Graham <scottmg@chromium.org> on 2011-08-01
Reviewed by Antti Koivisto.

Upon completing a load start a Timer to iterate through
CachedResourceLoader's m_documentResources map to check for any items
that have only one reference (thus being the reference in the map
itself). The map should really be weak, but because the
CachedResourceHandle achieves bookkeeping work in addition to
reference counting, this is a simpler and more localized way to free
the used memory while maintaining the other behaviour (when
CachedResource is used as proxy).

With this patch the testcase at
https://bugs.webkit.org/attachment.cgi?id=93850 should no longer
consume 400MB of ram on load. Test added for crash discovered in
previous revision, but no tests for memory usage.

Test: http/tests/inspector/network/disabled-cache-crash.html

WebCore.exp.in:
loader/cache/CachedResource.h:

(WebCore::CachedResource::hasOneHandle):

loader/cache/CachedResourceLoader.cpp:

(WebCore::CachedResourceLoader::CachedResourceLoader):
(WebCore::CachedResourceLoader::loadDone):
(WebCore::CachedResourceLoader::garbageCollectDocumentResourcesTimerFired):

loader/cache/CachedResourceLoader.h:
testing/Internals.cpp:

(WebCore::Internals::disableMemoryCache):

testing/Internals.h:
testing/Internals.idl:

Source/WebKit2: REGRESSION (r39725?): Resources removed from document can not be freed until the document is deleted
https://bugs.webkit.org/show_bug.cgi?id=61006

Patch by Scott Graham <scottmg@chromium.org> on 2011-08-01
Reviewed by Antti Koivisto.

Update exports for test harness.

win/WebKit2.def:
win/WebKit2CFLite.def:

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=61006

Test for CachedResourceLoader. Not caused by cache-disabling but very
difficult to reproduce when cache is active, so use cache disable in
inspector to exercise code.

Patch by Scott Graham <scottmg@chromium.org> on 2011-08-01
Reviewed by Antti Koivisto.

http/tests/inspector/network/disabled-cache-crash-expected.txt: Added.
http/tests/inspector/network/disabled-cache-crash.html: Added.
platform/gtk/Skipped:
platform/mac/Skipped:
platform/qt/Skipped:
platform/win/Skipped:

2:18 PM Changeset in webkit [92142] by jochen@chromium.org

5 edits
7 adds in trunk

Source/WebCore: Never override the policy URL on form submissions.
https://bugs.webkit.org/show_bug.cgi?id=61809

Reviewed by Adam Barth.

Tests: http/tests/security/cookies/third-party-cookie-blocking-main-frame.html

http/tests/security/cookies/third-party-cookie-blocking-user-action.html
http/tests/security/cookies/third-party-cookie-blocking.html

loader/FrameLoader.cpp:

(WebCore::FrameLoader::loadURL):
(WebCore::FrameLoader::addExtraFieldsToSubresourceRequest):
(WebCore::FrameLoader::addExtraFieldsToMainResourceRequest):
(WebCore::FrameLoader::addExtraFieldsToRequest):
(WebCore::FrameLoader::loadPostRequest):
(WebCore::FrameLoader::loadDifferentDocumentItem):

loader/FrameLoader.h:

LayoutTests: Require explicit user action to override the policy URL on form submissions.
https://bugs.webkit.org/show_bug.cgi?id=61809

Reviewed by Adam Barth.

http/tests/loading/redirect-methods-expected.txt:
http/tests/security/cookies/resources/set-a-cookie.php: Added.
http/tests/security/cookies/third-party-cookie-blocking-expected.txt: Added.
http/tests/security/cookies/third-party-cookie-blocking-main-frame-expected.txt: Added.
http/tests/security/cookies/third-party-cookie-blocking-main-frame.html: Added.
http/tests/security/cookies/third-party-cookie-blocking-user-action-expected.txt: Added.
http/tests/security/cookies/third-party-cookie-blocking-user-action.html: Added.
http/tests/security/cookies/third-party-cookie-blocking.html: Added.

2:16 PM Changeset in webkit [92141] by rniwa@webkit.org

5 edits in trunk/LayoutTests

media/track tests should be skipped on all platforms until feature is fully implemented.
https://bugs.webkit.org/show_bug.cgi?id=65482

Patch by Anna Cavender <annacc@chromium.org> on 2011-08-01
Reviewed by Ryosuke Niwa.

platform/gtk/Skipped:
platform/mac/Skipped:
platform/qt/Skipped:
platform/win/Skipped:

2:10 PM Changeset in webkit [92140] by rniwa@webkit.org

2 edits in trunk/Source/WebCore

Update comment added in r92139 per Darin's suggestion.

css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::styleForElement):

2:08 PM Changeset in webkit [92139] by rniwa@webkit.org

4 edits
2 adds in trunk

Search field in designMode causes a crash
https://bugs.webkit.org/show_bug.cgi?id=65362

Reviewed by Dimitri Glazkov.

Source/WebCore:

The crashed was caused by editing code inadvertently deleting search and cancel buttons in the design mode.
Fixed the bug by avoid inheriting user-modify property from the shadow host.

Test: editing/input/search-field-crash-in-designmode.html

css/CSSStyleSelector.cpp:

(WebCore::isAtShadowBoundary): Moved.
(WebCore::CSSStyleSelector::styleForElement): Overrides -webkit-user-modify by the initial value after m_style
inherited the values from m_parentStyle but before applying matched rules.

dom/Node.cpp:

(WebCore::Node::rendererIsEditable): Ignore page's editability inside a shadow DOM.

LayoutTests:

Add a test to ensure WebKit doesn't crash when a user tries to delete inside
an empty search field in the design mode.

editing/input/search-field-crash-in-designmode-expected.txt: Added.
editing/input/search-field-crash-in-designmode.html: Added.

1:40 PM Changeset in webkit [92138] by commit-queue@webkit.org

2 edits in trunk/Source/WebCore

[Soup] Set default max-conns to a more sane default value and fix comment
https://bugs.webkit.org/show_bug.cgi?id=65335

Patch by Marco Peereboom <marco@peereboom.us> on 2011-08-01
Reviewed by Martin Robinson.

Rigged the xxxterm web browser and validated that the default values were set properly.

platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::ResourceHandle::defaultSession):

1:39 PM Changeset in webkit [92137] by abarth@webkit.org

3 edits in trunk/Tools

Remove unused functionality from gardeningserver
https://bugs.webkit.org/show_bug.cgi?id=65479

Reviewed by Dimitri Glazkov.

We now handle these operations directly from the front-end. That means
they work even without the gardening server running!

Also, add some missing files to the list of static files.

Scripts/webkitpy/tool/servers/gardeningserver.py:
Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:

12:55 PM Changeset in webkit [92136] by Dimitri Glazkov

3 edits in trunk/Tools

Introduce SpecifierSorter, a thing that knows how specifiers should be ordered.
https://bugs.webkit.org/show_bug.cgi?id=65413

Reviewed by Adam Barth.

Scripts/webkitpy/layout_tests/models/test_configuration.py: Added SpecifierSorter and converted

existing code to use it.

Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests and cleaned up a bit.

12:25 PM Changeset in webkit [92135] by abarth@webkit.org

14 edits in trunk/Tools

Refactor bugzilla.js for use by garden-o-matic
https://bugs.webkit.org/show_bug.cgi?id=65450

Reviewed by Dimitri Glazkov.

This patch refactors bugzilla.js to use the AsynchronousCache and
updates the style to use a module instead of an object. This patch
then fixes all the existing code that uses this class to use the new
API style.

This main benefit of this patch is we remove the tricky manual caching
and this code is now available to use in garden-o-matic (since the
dependency on Utilities.js is now gone).

I ran all the unit tests and poked around in TestFailures a bit to see
that everything seemed to be working properly.

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:

(FailingTestsBugForm):
(FailingTestsBugForm.prototype._createBugTitle):

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js:

(FlakyTestBugForm):

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js:

(NewBugForm):
(NewBugForm.prototype.domElement):

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js:

(TestRelatedBugForm):

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:

(ViewController.prototype._displayBuilder.start):
(ViewController.prototype._displayBuilder):
(ViewController.prototype._domForAuxiliaryUIElements):
(ViewController.prototype._domForNewAndExistingBugs.bugzilla.quickSearch):
(ViewController.prototype._domForPossiblyFlakyTests.flakyList.appendChildren):
(ViewController.prototype._domForPossiblyFlakyTests):

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:

12:23 PM Changeset in webkit [92134] by abarth@webkit.org

4 edits
1 add in trunk/Tools

garden-o-matic shouldn't require local server to determine if compile failed
https://bugs.webkit.org/show_bug.cgi?id=65446

Reviewed by Dimitri Glazkov.

This patch adds some basic bindings for the buildbot JSON API. I
wanted to use Buildbot.js an Builder.js from the original TestFailures,
but they make a number of WebKit-specific assumptions. We'll
eventually need to reconcile these pieces of code.

I should really add tests for this code, but that will require me to
build a better mock network. I'll be happy to do that in a follow-up
patch.

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js: Added.
- It's very unclear to me whether didFail is correct. Supposedly the first entry here is the exit code of the step, which is 0 on success. I suspect we'll need to improve this function over time.
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:

11:45 AM Changeset in webkit [92133] by abarth@webkit.org

2 edits in trunk/Tools

garden-o-matic should show all failures, even ones that it can't find a regression range for
https://bugs.webkit.org/show_bug.cgi?id=65447

Reviewed by Dimitri Glazkov.

Previously, we just wouldn't display test failures that we couldn't
identify a regression range for (or whose regression range had fallen
off the page). This patch just displays them at the end of the list,
which is slightly (but not much) better.

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:

11:03 AM Changeset in webkit [92132] by inferno@chromium.org

3 edits
2 adds in trunk

Regression(82144): Crash in TrailingObjects::updateMidpointsForTrailingBoxes
https://bugs.webkit.org/show_bug.cgi?id=65137

Source/WebCore:

Fix the looping condition to prevent trailingSpaceMidpoint from becoming negative.

Reviewed by Dave Hyatt.

Test: fast/block/update-midpoints-for-trailing-boxes-crash.html

rendering/RenderBlockLineLayout.cpp:

(WebCore::TrailingObjects::updateMidpointsForTrailingBoxes):

LayoutTests:

ASSERTION FAILED: trailingSpaceMidpoint >= 0

Reviewed by Dave Hyatt.

fast/block/update-midpoints-for-trailing-boxes-crash-expected.txt: Added.
fast/block/update-midpoints-for-trailing-boxes-crash.html: Added.

9:51 AM Changeset in webkit [92131] by Csaba Osztrogonác

3 edits
4 adds in trunk/LayoutTests

[Qt] Unreviewed gardening after r92127.

platform/qt-4.8/css2.1/t0602-c13-inh-underlin-00-e-expected.png: Added.
platform/qt-4.8/css2.1/t0602-c13-inh-underlin-00-e-expected.txt: Added.
platform/qt-4.8/editing/selection/4397952-expected.png: Added.
platform/qt-4.8/editing/selection/4397952-expected.txt: Added.
platform/qt-mac/Skipped: Skip failing tests.
platform/qt-wk2/Skipped: Skip failing tests.

9:10 AM Changeset in webkit [92130] by commit-queue@webkit.org

3 edits in trunk/Source/WebCore

[Qt] OpenGLShims does not build on ARM
https://bugs.webkit.org/show_bug.cgi?id=65465

Patch by Benjamin Poulain <benjamin@webkit.org> on 2011-08-01
Reviewed by Noam Rosenthal.

Fix the build with OpenGL ES 2:
-lookupOpenGLFunctionAddress() was defined but not used for OpenGL ES.
-glBlitFramebuffer() and glRenderbufferStorageMultisample() are not part of the specification.
When those are available as platform extension, the extension has been added.
-GLchar is not defined on some platform. The patch adds the same typedef as the official definition
to avoid conflicts.

platform/graphics/cairo/OpenGLShims.cpp:

(lookupOpenGLFunctionAddress):
(WebCore::initializeOpenGLShims):

platform/graphics/cairo/OpenGLShims.h:

9:09 AM Changeset in webkit [92129] by andersca@apple.com

2 edits in trunk/LayoutTests

Add fast/forms/selection-direction.html to the Skipped list.

platform/wk2/Skipped:

9:02 AM Changeset in webkit [92128] by jeffm@apple.com

3 edits in trunk/Tools

Windows build fix after r92096.

TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Use #include, not #import, and don't include Foundation.h.
TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Add InjectedBundleControllerWin.cpp.

9:01 AM Changeset in webkit [92127] by Csaba Osztrogonác

105 edits
112 adds in trunk/LayoutTests

[Qt] Update expected results and unskip now passing tests.

Patch by Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu> on 2011-08-01
Reviewed by Csaba Osztrogonác.

platform/qt/Skipped:
platform/qt/css2.1/ [...]: Updated.
platform/qt/editing/ [...]: Updated.

6:47 AM Changeset in webkit [92126] by pfeldman@chromium.org

4 edits in trunk/LayoutTests

2011-08-01 Pavel Feldman <pfeldman@google.com>

Not reviewed: unflake inspector test that was just added.

inspector/debugger/scripts-sorting-expected.txt:
inspector/debugger/scripts-sorting.html:
platform/chromium/test_expectations.txt:

5:45 AM Changeset in webkit [92125] by yurys@chromium.org

2 edits in trunk/Source/WebCore

2011-08-01 Yury Semikhatsky <yurys@chromium.org>

Unreviewed. Trivial fix for inspector/debugger/debugger-completions-on-call-frame.html after r92122.

inspector/front-end/ConsoleView.js:

5:25 AM Changeset in webkit [92124] by hayato@chromium.org

11 edits
4 adds in trunk

Add support for getting an element in shadow root by its id into a window.internals object.
https://bugs.webkit.org/show_bug.cgi?id=64587

Reviewed by Hajime Morita.

Source/autotools/symbols.filter:

Source/WebCore:

Test: fast/dom/shadow/get-element-by-id-in-shadow-root.html

WebCore.exp.in:
testing/Internals.cpp:

(WebCore::Internals::getElementByIdInShadowRoot):

testing/Internals.h:
testing/Internals.idl:

Source/WebKit2:

win/WebKit2.def:
win/WebKit2CFLite.def:

LayoutTests:

fast/dom/shadow/get-element-by-id-in-shadow-root-expected.txt: Added.
fast/dom/shadow/get-element-by-id-in-shadow-root.html: Added.
fast/dom/shadow/resources/create-dom.js: Added.

(createShadow):
(createDom):

4:53 AM Changeset in webkit [92123] by commit-queue@webkit.org

7 edits in trunk

build: Fix finding the headers for GStreamer

The cmake files to find the various GStreamer packages were all
checking for the header gst/gst.h. However if gst-plugins-base is
installed into a separate prefix from gstreamer then all of these
tests would only pick up the gstreamer include path so the build
would fail. This patch changes it to try and find a file
appropriate to each package.

https://bugs.webkit.org/show_bug.cgi?id=64933

Patch by Neil Roberts <neil@linux.intel.com> on 2011-08-01
Reviewed by Martin Robinson.

Source/cmake/FindGStreamer-App.cmake:
Source/cmake/FindGStreamer-Base.cmake:
Source/cmake/FindGStreamer-Interfaces.cmake:
Source/cmake/FindGStreamer-Pbutils.cmake:
Source/cmake/FindGStreamer-Plugins-Base.cmake:
Source/cmake/FindGStreamer-Video.cmake:

4:42 AM Changeset in webkit [92122] by yurys@chromium.org

11 edits
3 adds in trunk

Web Inspector: typing an expression in an iframe leads to multiple "Unsafe JavaScript attempt to access frame..." errors
https://bugs.webkit.org/show_bug.cgi?id=65457

Source/WebCore:

Console completions are now done using evaluation which returns a JSON object with all property names rather than a remote
object. Also Runtime.evaluate and Runtime.callFunctionOn commands were extended with an optional parameter that allows to
get result as JSON value.

Reviewed by Pavel Feldman.

Test: http/tests/inspector/console-cd-completions.html

inspector/InjectedScript.cpp:

(WebCore::InjectedScript::evaluate):
(WebCore::InjectedScript::callFunctionOn):

inspector/InjectedScript.h:
inspector/InjectedScriptSource.js:

(.):
():

inspector/Inspector.json:
inspector/InspectorRuntimeAgent.cpp:

(WebCore::asBool):
(WebCore::InspectorRuntimeAgent::evaluate):
(WebCore::InspectorRuntimeAgent::callFunctionOn):

inspector/InspectorRuntimeAgent.h:
inspector/front-end/ConsoleView.js:

(WebInspector.ConsoleView.prototype._completions.evaluated.getCompletions):
(WebInspector.ConsoleView.prototype._completions.evaluated):
(WebInspector.ConsoleView.prototype._completions.receivedPropertySet):
(WebInspector.ConsoleView.prototype._completions):
(WebInspector.ConsoleView.prototype.evalInInspectedWindow):

inspector/front-end/RemoteObject.js:

(WebInspector.RemoteObject):
(WebInspector.RemoteObject.prototype.callFunction):
(WebInspector.RemoteObject.prototype.callFunctionJSON):

inspector/front-end/WatchExpressionsSidebarPane.js:

(WebInspector.WatchExpressionsSection.prototype.update):

LayoutTests:

Reviewed by Pavel Feldman.

http/tests/inspector/console-cd-completions-expected.txt: Added.
http/tests/inspector/console-cd-completions.html: Added.
http/tests/inspector/resources/console-cd-completions-iframe.html: Added.

4:38 AM Changeset in webkit [92121] by tonyg@chromium.org

2 edits in trunk/LayoutTests

[chromium] http/tests/inspector/resource-tree/resource-tree-document-url.html occasionally times out
https://bugs.webkit.org/show_bug.cgi?id=65463

Unreviewed expectations update.

platform/chromium/test_expectations.txt:

4:30 AM Changeset in webkit [92120] by tonyg@chromium.org

2 edits in trunk/LayoutTests

[chromium] http/tests/cache/history-only-cached-subresource-loads-max-age-https.html frequently times out on Vista
https://bugs.webkit.org/show_bug.cgi?id=65462

Unreviewed expectations update.

platform/chromium/test_expectations.txt:

4:22 AM Changeset in webkit [92119] by tonyg@chromium.org

2 edits in trunk/LayoutTests

[chromium] Temporarily expect scripts-sorting.html to fail
https://bugs.webkit.org/show_bug.cgi?id=65461

Unreviewed gardening.

platform/chromium/test_expectations.txt:

4:18 AM Changeset in webkit [92118] by commit-queue@webkit.org

8 edits
2 adds in trunk

BORDER attribute with the object tag, using percentage values not working.
https://bugs.webkit.org/show_bug.cgi?id=65176

Patch by Mihnea Ovidenie <mihnea@adobe.com> on 2011-08-01
Reviewed by Hajime Morita.

Source/WebCore:

When border presentational attribute for object element has % in it, it should be parsed using HTML parser rules.

Test: fast/borders/border-width-percent.html

html/HTMLElement.cpp:

(WebCore::HTMLElement::parseBorderWidthAttribute):

html/HTMLElement.h:
html/HTMLImageElement.cpp:
html/HTMLObjectElement.cpp:

(WebCore::HTMLObjectElement::parseMappedAttribute):

LayoutTests:

Added the tests with border width with percentages in a new file. Moved them
from images border test file.

fast/borders/border-width-percent-expected.txt: Added.
fast/borders/border-width-percent.html: Added.
fast/images/border-expected.txt:
fast/images/script-tests/border.js:

4:16 AM Changeset in webkit [92117] by tonyg@chromium.org

2 edits
1 delete in trunk/LayoutTests

[chromium] Update expectations for fast/text/international/bidi-mirror-he-ar.html
https://bugs.webkit.org/show_bug.cgi?id=65460

Unreviewed expectations update.

platform/chromium-win/fast/text/international/bidi-mirror-he-ar-expected.png:

4:16 AM Changeset in webkit [92116] by yutak@chromium.org

8 edits in trunk/Source/WebCore

WebSocket: Receive URL and subprotocol in WebSocketChannel::connect()
https://bugs.webkit.org/show_bug.cgi?id=65367

Reviewed by Kent Tamura.

To implement multiple subprotocols support (bug 65247), WebSocket::connect() will need to validate
the value of subprotocols after constructing WebSocketChannel, because the result depends on which
WebSocket protocol is used, which is obtained from WebSocketChannel::useHixie76Protocol(). This
means the subprotocol value will not be available at the time of WebSocketChannel construction.

This change moves URL and subprotocol arguments in WebSocketChannel constructor to
WebSocketChannel::connect(), which allows WebSocket::connect() function to check the subprotocol
value before the actual connection is established.

Relocating URL argument is technically not necessary, but seemed legitimate in terms of functional
correspondence between WebSocket and WebSocketChannel (constructor versus connect()).

No change in behavior, thus no new tests.

websockets/ThreadableWebSocketChannel.cpp:

Remove "url" and "protocol" arguments from constructor and add them in connect().
(WebCore::ThreadableWebSocketChannel::create):

websockets/ThreadableWebSocketChannel.h:
websockets/WebSocket.cpp:

(WebCore::WebSocket::connect):

websockets/WebSocketChannel.cpp: Same as ThreadableWebSocketChannel.

(WebCore::WebSocketChannel::WebSocketChannel):
(WebCore::WebSocketChannel::connect):
InspectorInstrumentation::didCreateWebSocket() call was moved to connect() because it needs URL.
This does not change behavior, because connect() is guaranteed to be called immediately after
WebSocketChannel is constructed.

websockets/WebSocketChannel.h:

(WebCore::WebSocketChannel::create):

websockets/WorkerThreadableWebSocketChannel.cpp: Same as ThreadableWebSocketChannel.

(WebCore::WorkerThreadableWebSocketChannel::WorkerThreadableWebSocketChannel):
(WebCore::WorkerThreadableWebSocketChannel::connect):
(WebCore::WorkerThreadableWebSocketChannel::Peer::Peer):
(WebCore::WorkerThreadableWebSocketChannel::Peer::connect):
(WebCore::WorkerThreadableWebSocketChannel::Bridge::mainThreadCreateWebSocketChannel):
(WebCore::WorkerThreadableWebSocketChannel::Bridge::Bridge):
(WebCore::WorkerThreadableWebSocketChannel::mainThreadConnect):
(WebCore::WorkerThreadableWebSocketChannel::Bridge::connect):

websockets/WorkerThreadableWebSocketChannel.h:

(WebCore::WorkerThreadableWebSocketChannel::create):
(WebCore::WorkerThreadableWebSocketChannel::Peer::create):
(WebCore::WorkerThreadableWebSocketChannel::Bridge::create):

3:19 AM Changeset in webkit [92115] by Csaba Osztrogonác

2 edits in trunk/Source/WebKit/qt

[Qt] Fix tst_QWebPage::loadHtml5Video() API test
https://bugs.webkit.org/show_bug.cgi?id=65452

Patch by Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu> on 2011-08-01
Reviewed by Csaba Osztrogonác.

tests/qwebpage/tst_qwebpage.cpp: Mark failing test case as expected fail

(tst_QWebPage::loadHtml5Video):

3:18 AM Changeset in webkit [92114] by vsevik@chromium.org

1 edit
2 adds in trunk/LayoutTests

Web Inspector: Add network panel sidebar width test.
https://bugs.webkit.org/show_bug.cgi?id=65314

Reviewed by Yury Semikhatsky.

http/tests/inspector/network/network-sidebar-width-expected.txt: Added.
http/tests/inspector/network/network-sidebar-width.html: Added.

2:53 AM Changeset in webkit [92113] by commit-queue@webkit.org

2 edits in trunk/Source/WebKit2

Unreviewed, rolling out r92108.
http://trac.webkit.org/changeset/92108
https://bugs.webkit.org/show_bug.cgi?id=65459

breaks SL compile (Requested by tonyg-cr on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2011-08-01

Shared/WebCoreArgumentCoders.cpp:

(CoreIPC::::encode):
(CoreIPC::::decode):

2:36 AM Changeset in webkit [92112] by tonyg@chromium.org

2 edits in trunk/LayoutTests

http/tests/inspector/resource-tree/resource-tree-reload.html is flaky on chromium win
https://bugs.webkit.org/show_bug.cgi?id=60110

Marking it flaky on Linux as well.

platform/chromium/test_expectations.txt:

2:05 AM Changeset in webkit [92111] by tonyg@chromium.org

2 edits
1 copy in trunk/LayoutTests

Fix windows text rendering expectations for fast/text/justify-ideograph-leading-expansion.html
https://bugs.webkit.org/show_bug.cgi?id=65455

Unreviewed expectations update.

platform/chromium-win-xp/fast/text/justify-ideograph-leading-expansion-expected.png: Copied from LayoutTests/platform/chromium-win/fast/text/justify-ideograph-leading-expansion-expected.png.
platform/chromium-win/fast/text/justify-ideograph-leading-expansion-expected.png:

2:04 AM Changeset in webkit [92110] by Csaba Osztrogonác

5 edits
4 copies in trunk/LayoutTests

REGRESSION (r92048): svg/custom/external-paintserver-reference.svg, svg/custom/linking-base-external-reference.xhtml failing on SnowLeopard Intel Release (Tests)
https://bugs.webkit.org/show_bug.cgi?id=65432

Revert accidentally committed changes, and update Qt specific expected results.

platform/qt/svg/custom/external-paintserver-reference-expected.png: Copied from LayoutTests/svg/custom/external-paintserver-reference-expected.png.
platform/qt/svg/custom/external-paintserver-reference-expected.txt: Copied from LayoutTests/svg/custom/external-paintserver-reference-expected.txt.
platform/qt/svg/custom/linking-base-external-reference-expected.png: Copied from LayoutTests/svg/custom/linking-base-external-reference-expected.png.
platform/qt/svg/custom/linking-base-external-reference-expected.txt: Copied from LayoutTests/svg/custom/linking-base-external-reference-expected.txt.
svg/custom/external-paintserver-reference-expected.png: Revert r92048.
svg/custom/external-paintserver-reference-expected.txt: Revert r92048.
svg/custom/linking-base-external-reference-expected.png: Revert r92048.
svg/custom/linking-base-external-reference-expected.txt: Revert r92048.

1:57 AM Changeset in webkit [92109] by pfeldman@chromium.org

3 edits in trunk/LayoutTests

2011-08-01 Pavel Feldman <pfeldman@google.com>

Not reviewed: unflake inspector test that was just added.

inspector/debugger/scripts-sorting-expected.txt:
inspector/debugger/scripts-sorting.html:

1:49 AM Changeset in webkit [92108] by commit-queue@webkit.org

2 edits in trunk/Source/WebKit2

Custom cursors cause the WebProcess to crash
https://bugs.webkit.org/show_bug.cgi?id=64802

Patch by Amruth Raj <amruthraj@motorola.com> on 2011-08-01
Reviewed by Darin Adler.

Shared/WebCoreArgumentCoders.cpp:

(CoreIPC::::encode): In case of a NULL cursor image, encode a bool indicating that.
(CoreIPC::::decode): Decode the image only if required.

1:48 AM Changeset in webkit [92107] by tonyg@chromium.org

2 edits in trunk/LayoutTests

[chromium] fast/css/outline-auto-empty-rects.html broken by r92059
https://bugs.webkit.org/show_bug.cgi?id=65453

This patch just expects the test to fail for now.

Unreviewed expectations update.

platform/chromium/test_expectations.txt:

1:31 AM Changeset in webkit [92106] by macpherson@chromium.org

5 edits in trunk/Source/WebCore

Add iterator to CSSValueList
https://bugs.webkit.org/show_bug.cgi?id=65297

Reviewed by Darin Adler.

No new tests / refactoring only.

css/CSSPrimitiveValue.h:

(WebCore::CSSPrimitiveValue::isLength):
Add shorthand to determine if this primitive value is a length.

css/CSSStyleSelector.cpp:

Use CSSValueListIterator throughout.
(WebCore::CSSStyleSelector::applyProperty):
(WebCore::CSSStyleSelector::applyPageSizeProperty):
(WebCore::CSSStyleSelector::createTransformOperations):

css/CSSValueList.cpp:

(WebCore::CSSValueList::copy):
Use itemWithoutBoundsCheck() instead of item().

css/CSSValueList.h:

Add CSSValueListIterator and CSSValueListInspector class definitions.
(WebCore::CSSValueList::item)
Provide inline definition of item.
(WebCore::CSSValueListIterator::CSSValueListIterator):
(WebCore::CSSValueListIterator::hasMore):
Return true if there are more values to consume, including the current value.
(WebCore::CSSValueListIterator::value):
Return the value at the current position.
(WebCore::CSSValueListIterator::next):
Move the iterator forward to the next item.
(WebCore::CSSValueListIterator::index):
Return the current position in the list.
(WebCore::CSSValueListInspector::item):
Return the item at a given index.
(WebCore::CSSValueListInspector::first):
Return the first item in the list.
(WebCore::CSSValueListInspector::second):
Return the second item in the list.
(WebCore::CSSValueListInspector::length):
Return the size of the underlying list.

1:17 AM Changeset in webkit [92105] by pfeldman@chromium.org

4 edits
2 adds in trunk

Web Inspector: group scripts by folder in the scripts selector.
https://bugs.webkit.org/show_bug.cgi?id=65321

Reviewed by Yury Semikhatsky.

Source/WebCore:

Test: inspector/debugger/scripts-sorting.html

inspector/front-end/ScriptsPanel.js:

(WebInspector.ScriptsPanel.prototype._sourceFileAdded.get if):
(WebInspector.ScriptsPanel.prototype._addOptionToFilesSelect.insertOrdered.optionCompare):
(WebInspector.ScriptsPanel.prototype._addOptionToFilesSelect.insertOrdered):
(WebInspector.ScriptsPanel.prototype._addOptionToFilesSelect):
(WebInspector.ScriptsPanel.prototype._folderAndDisplayNameForScriptURL):
(WebInspector.ScriptsPanel.prototype.reset):
(WebInspector.SourceFrameDelegateForScriptsPanel.prototype.suggestedFileName):

LayoutTests:

http/tests/inspector/debugger-test.js:

(initialize_DebuggerTest):

inspector/debugger/scripts-sorting-expected.txt: Added.
inspector/debugger/scripts-sorting.html: Added.

1:13 AM Changeset in webkit [92104] by pfeldman@chromium.org

5 edits in trunk

Web Inspector: [protocol] methods where all parameters are optional should not require empty params object.
https://bugs.webkit.org/show_bug.cgi?id=65331

Reviewed by Yury Semikhatsky.

Source/WebCore:

inspector/CodeGeneratorInspector.pm:

(generateBackendFunction):
(generateArgumentGetters):

LayoutTests:

inspector/report-protocol-errors-expected.txt:

1:04 AM Changeset in webkit [92103] by abarth@webkit.org

2 edits in trunk/Source/WebCore

application/json MIME type shouldn't cause warnings in <script> tags
https://bugs.webkit.org/show_bug.cgi?id=65448

Reviewed by Pavel Feldman.

This patch stops the web inspector from logging a ton of spam warnings
when a web page uses a JSONP service (which points script tags at
application/json resources).

inspector/front-end/inspector.js:

(WebInspector.reset):

12:26 AM Changeset in webkit [92102] by yutak@chromium.org

10 edits in trunk/Source

WebSocket: Rename SocketStreamHandleClient member functions
https://bugs.webkit.org/show_bug.cgi?id=65374

Reviewed by Kent Tamura.

Source/WebCore:

Rename SocketStreamHandleClient member functions to avoid conflict of the names of
FileReaderLoaderClient functions.

No change in behavior, thus no new tests are added.

platform/network/SocketStreamHandleClient.h:

(WebCore::SocketStreamHandleClient::didOpenSocketStream):
(WebCore::SocketStreamHandleClient::didCloseSocketStream):
(WebCore::SocketStreamHandleClient::didReceiveSocketStreamData):
(WebCore::SocketStreamHandleClient::didFailSocketStream):

platform/network/brew/SocketStreamHandleBrew.cpp:

(WebCore::SocketStreamHandlePrivate::socketConnected):
(WebCore::SocketStreamHandlePrivate::socketReadyRead):
(WebCore::SocketStreamHandlePrivate::socketClosed):
(WebCore::SocketStreamHandlePrivate::socketError):

platform/network/cf/SocketStreamHandleCFNet.cpp:

(WebCore::SocketStreamHandle::addCONNECTCredentials):
(WebCore::SocketStreamHandle::readStreamCallback):
(WebCore::SocketStreamHandle::writeStreamCallback):
(WebCore::SocketStreamHandle::reportErrorToClient):
(WebCore::SocketStreamHandle::platformClose):

platform/network/qt/SocketStreamHandleQt.cpp:

(WebCore::SocketStreamHandlePrivate::socketConnected):
(WebCore::SocketStreamHandlePrivate::socketReadyRead):
(WebCore::SocketStreamHandlePrivate::socketClosedCallback):
(WebCore::SocketStreamHandlePrivate::socketErrorCallback):

platform/network/soup/SocketStreamHandleSoup.cpp:

(WebCore::SocketStreamHandle::connected):
(WebCore::SocketStreamHandle::readBytes):
(WebCore::SocketStreamHandle::platformSend):
(WebCore::SocketStreamHandle::platformClose):

websockets/WebSocketChannel.cpp:

(WebCore::WebSocketChannel::didOpenSocketStream):
(WebCore::WebSocketChannel::didCloseSocketStream):
(WebCore::WebSocketChannel::didReceiveSocketStreamData):
(WebCore::WebSocketChannel::didFailSocketStream):
(WebCore::WebSocketChannel::resumeTimerFired):

websockets/WebSocketChannel.h:

Source/WebKit/chromium:

src/SocketStreamHandle.cpp:

(WebCore::SocketStreamHandleInternal::didOpenStream):
(WebCore::SocketStreamHandleInternal::didReceiveData):
(WebCore::SocketStreamHandleInternal::didClose):
(WebCore::SocketStreamHandleInternal::didFail):

Jul 31, 2011:

11:05 PM Changeset in webkit [92101] by abarth@webkit.org

4 edits
1 delete in trunk/LayoutTests

These results appear to differ only in text rendering.

platform/chromium-win-xp/fast/text/international/bidi-mirror-he-ar-expected.png:
platform/chromium-win-xp/fast/text/justify-ideograph-leading-expansion-expected.png: Removed.
platform/chromium-win/fast/text/international/bidi-mirror-he-ar-expected.png:
platform/chromium-win/fast/text/justify-ideograph-leading-expansion-expected.png:

10:36 PM Changeset in webkit [92100] by abarth@webkit.org

1 edit
10 adds in trunk/LayoutTests

Adding expected results. These results appear to differ only in the scrollbar.

platform/chromium-linux/svg/custom/external-paintserver-reference-expected.png: Added.
platform/chromium-linux/svg/custom/linking-base-external-reference-expected.png: Added.
platform/chromium-mac/svg/custom/external-paintserver-reference-expected.png: Added.
platform/chromium-mac/svg/custom/external-paintserver-reference-expected.txt: Added.
platform/chromium-mac/svg/custom/linking-base-external-reference-expected.png: Added.
platform/chromium-mac/svg/custom/linking-base-external-reference-expected.txt: Added.
platform/chromium-win/svg/custom/external-paintserver-reference-expected.png: Added.
platform/chromium-win/svg/custom/external-paintserver-reference-expected.txt: Added.
platform/chromium-win/svg/custom/linking-base-external-reference-expected.png: Added.
platform/chromium-win/svg/custom/linking-base-external-reference-expected.txt: Added.

10:33 PM Changeset in webkit [92099] by abarth@webkit.org

2 edits in trunk/LayoutTests

Update baseline. This patch might cause Linux to start failing, but in
that case, we can add a more specific Linux result. This approach is
somewhat of an exploration of this automated rebaselining algorithm.

platform/chromium-win/svg/batik/text/textLength-expected.txt:

9:25 PM Changeset in webkit [92098] by commit-queue@webkit.org

3 edits
2 deletes in trunk

Unreviewed, rolling out r92094.
http://trac.webkit.org/changeset/92094
https://bugs.webkit.org/show_bug.cgi?id=65443

Triggered many asserts (Requested by abarth on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2011-07-31

Source/WebCore:

dom/EventDispatcher.cpp:

(WebCore::EventDispatcher::dispatchEvent):

LayoutTests:

fast/events/event-target-reset-expected.txt: Removed.
fast/events/event-target-reset.html: Removed.

9:17 PM Changeset in webkit [92097] by andersca@apple.com

2 edits in trunk/Tools

DownloadDecideDestinationCrash WebKit2 API tests crashes
https://bugs.webkit.org/show_bug.cgi?id=65442

Reviewed by Sam Weinig.

Copy the 18-characters.html file to the destination directory so we won't
crash trying to make an URL for it.

TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:

9:10 PM Changeset in webkit [92096] by andersca@apple.com

2 adds in trunk/Tools/TestWebKitAPI

Add forgotten files.

8:32 PM Changeset in webkit [92095] by andersca@apple.com

8 edits in trunk/Tools

Fix SpacebarScrolling WebKit2 API test
https://bugs.webkit.org/show_bug.cgi?id=65431

Reviewed by Sam Weinig.

Add a way to do platform-specific bundle initialization, and use it to disable smooth
scrolling. Also make it possible for a test to load the injected bundle without actually running
a test from it.

TestWebKitAPI/InjectedBundleController.cpp:

(TestWebKitAPI::InjectedBundleController::initialize):

TestWebKitAPI/InjectedBundleController.h:
TestWebKitAPI/PlatformUtilities.cpp:

(TestWebKitAPI::Util::createContextWithInjectedBundle):

TestWebKitAPI/PlatformUtilities.h:
TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:

(TestWebKitAPI::TEST):

TestWebKitAPI/mac/InjectedBundleControllerMac.mm: Added.

(TestWebKitAPI::InjectedBundleController::platformInitialize):

TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Added.

(TestWebKitAPI::InjectedBundleController::platformInitialize):

TestWebKitAPI/win/TestWebKitAPI.vcproj:

7:53 PM Changeset in webkit [92094] by dominicc@chromium.org

3 edits
2 adds in trunk

Event target set to null post event dispatch
https://bugs.webkit.org/show_bug.cgi?id=63934

Reviewed by Dimitri Glazkov.

Source/WebCore:

Test: fast/events/event-target-reset.html

dom/EventDispatcher.cpp:

(WebCore::EventDispatcher::dispatchEvent): Reset target to outermost
shadow host.

LayoutTests:

fast/events/event-target-reset-expected.txt: Added.
fast/events/event-target-reset.html: Added.

6:50 PM Changeset in webkit [92093] by barraclough@apple.com

6 edits in trunk

https://bugs.webkit.org/show_bug.cgi?id=64679
Fix bugs in Array.prototype this handling.

Unreviewed - rolling out r91290.

Source/JavaScriptCore:

Looks like the wild wild web isn't ready for this yet.

This change broke http://slides.html5rocks.com/#landing-slide.
Interestingly, this might only be due to our lack of bind support -
it looks like this site is calling Array.prototype.slice as a part
of its bind implementation.

runtime/ArrayPrototype.cpp:

(JSC::arrayProtoFuncJoin):
(JSC::arrayProtoFuncConcat):
(JSC::arrayProtoFuncPop):
(JSC::arrayProtoFuncPush):
(JSC::arrayProtoFuncReverse):
(JSC::arrayProtoFuncShift):
(JSC::arrayProtoFuncSlice):
(JSC::arrayProtoFuncSort):
(JSC::arrayProtoFuncSplice):
(JSC::arrayProtoFuncUnShift):
(JSC::arrayProtoFuncFilter):
(JSC::arrayProtoFuncMap):
(JSC::arrayProtoFuncEvery):
(JSC::arrayProtoFuncForEach):
(JSC::arrayProtoFuncSome):
(JSC::arrayProtoFuncReduce):
(JSC::arrayProtoFuncReduceRight):
(JSC::arrayProtoFuncIndexOf):
(JSC::arrayProtoFuncLastIndexOf):

LayoutTests:

Checking in failing results - we can't fix this bug just yet
for web compatibility reasons.

fast/js/array-prototype-properties-expected.txt:
ietestcenter/Javascript/15.4.4.14-5-28-expected.txt:
ietestcenter/Javascript/15.4.4.15-5-28-expected.txt:

4:48 PM Changeset in webkit [92092] by fpizlo@apple.com

2 edits in trunk/Source/JavaScriptCore

JSC GC lays out size classes under wrong assumptions about expected
object size.
https://bugs.webkit.org/show_bug.cgi?id=65437

Reviewed by Oliver Hunt.

Changed the atom size - which is both the smallest allocation size and
the smallest possible stepping unit for size class spacing - from
8 bytes to 4 pointer-size words. This is a 1% win on SunSpider.

heap/MarkedBlock.h:

4:17 PM Changeset in webkit [92091] by rniwa@webkit.org

2 edits in trunk/LayoutTests

Fix fast/dom/Range/getBoundingClientRect.html after r92089.

fast/dom/Range/getBoundingClientRect.html:

2:14 PM Changeset in webkit [92090] by rniwa@webkit.org

2 edits in trunk/LayoutTests

[Chromium] editing/execCommand/4786404-1.html,editing/execCommand/break-out-of-empty-list-item.html are slow/timeout
https://bugs.webkit.org/show_bug.cgi?id=65019

This bug was invalid; neither test is slow nor timing out.

platform/chromium/test_expectations.txt:

2:05 PM Changeset in webkit [92089] by weinig@apple.com

7 edits in trunk

REGRESSION: getBoundingClientRect() method of Range incorrectly returns null for collapsed Range
https://bugs.webkit.org/show_bug.cgi?id=65324

Reviewed by Ryosuke Niwa.

Source/WebCore:

Update to more closely match the spec.

dom/Range.cpp:

(WebCore::Range::getClientRects):
Return an empty ClientRectList instead of null when the range is empty.

(WebCore::Range::getBoundingClientRect):
Always return a ClientRect, even for an empty bounding rect.

LayoutTests:

fast/dom/Range/getBoundingClientRect-expected.txt:
fast/dom/Range/getBoundingClientRect.html:
fast/dom/Range/getClientRects-expected.txt:
fast/dom/Range/getClientRects.html:

Add test cases for empty and collapsed ranges.

1:44 PM Changeset in webkit [92088] by rniwa@webkit.org

10 edits
2 copies
4 adds in trunk

[HTML5] Implement the selectionDirection property on input and textarea
https://bugs.webkit.org/show_bug.cgi?id=60403

Reviewed by Darin Adler.

Source/WebCore:

Implements selectionDirection property as specified in HTML5 section 4.10.20:
http://www.whatwg.org/specs/web-apps/current-work/#textFieldSelection

selectionDirection is a property on input and textarea elements, which, on getting returns either "none",
"forward", or "backward" corresponding to the current selection direction, and on setting, modifies
the direction of the current selection. When selection doesn't exist or direction is unknown, "none"
is returned on getting. Setting the property value to neither of above values will result in setting "none".

This patch also adds the optional third argument, direction, to setSelectionRange.

Test: fast/forms/selection-direction.html

fast/forms/selection-wrongtype.html

bindings/js/JSHTMLInputElementCustom.cpp:

(WebCore::JSHTMLInputElement::selectionDirection): Added.
(WebCore::JSHTMLInputElement::setSelectionDirection): Added.
(WebCore::JSHTMLInputElement::setSelectionRange): Processes the third argument.

bindings/v8/custom/V8HTMLInputElementCustom.cpp:

(WebCore::V8HTMLInputElement::selectionDirectionAccessorGetter): Added.
(WebCore::V8HTMLInputElement::selectionDirectionAccessorSetter): Added.
(WebCore::V8HTMLInputElement::setSelectionRangeCallback): Processes the third argument.

html/HTMLTextAreaElement.cpp:

(WebCore::HTMLTextAreaElement::updateFocusAppearance): Calls hasCachedSelection.

html/HTMLInputElement.cpp:

(WebCore::HTMLInputElement::updateFocusAppearance): Ditto.
(WebCore::HTMLInputElement::setValue): Calls cacheSelection with SelectionHasNoDirection.

html/HTMLTextAreaElement.idl: Added selectionDirection and the third argument to setSelectionRange.
html/HTMLInputElement.idl: Ditto.
html/HTMLTextFormControlElement.cpp:

(WebCore::HTMLTextFormControlElement::HTMLTextFormControlElement): Initializes
m_cachedSelectionDirection.
(WebCore::HTMLTextFormControlElement::setSelectionStart): Calls setSelectionRange.
(WebCore::HTMLTextFormControlElement::setSelectionEnd): Ditto.
(WebCore::HTMLTextFormControlElement::setSelectionDirection): Added.
(WebCore::HTMLTextFormControlElement::select): Calls setSelectionRange with SelectionHasNoDirection.
(WebCore::HTMLTextFormControlElement::setSelectionRange): Added the version that takes direction
string as the third argument, which calls the version that takes two ints and TextFieldSelectionDirection.
(WebCore::HTMLTextFormControlElement::selectionStart): Calls hasCachedSelection.
(WebCore::HTMLTextFormControlElement::selectionEnd): Ditto.
(WebCore::directionString): Added.
(WebCore::HTMLTextFormControlElement::selectionDirection): Added.
(WebCore::HTMLTextFormControlElement::computeSelectionDirection): Added.
(WebCore::HTMLTextFormControlElement::selection): Calls hasCachedSelection.
(WebCore::HTMLTextFormControlElement::restoreCachedSelection): Calls setSelectionRange.
(WebCore::HTMLTextFormControlElement::selectionChanged): Calls cacheSelection.

html/HTMLTextFormControlElement.h:

(WebCore::HTMLTextFormControlElement::cacheSelection): Takes TextFieldSelectionDirection.
(WebCore::HTMLTextFormControlElement::hasCachedSelection): Extracted from hasCachedSelectionStart;
cacheSelection is the only function that modifies m_cacheSelectionStart and m_cacheSelectionEnd
and none of its callers ever set one of them be -1 so they are always both -1 or both non-negative.
Thus, having two separate functions for m_cacheSelectionStart and m_cacheSelectionEnd was redundant.

LayoutTests:

Added a test to ensure WebKit sets and gets selectionDirection property.

Also added a test to ensure WebKit throws exceptions when retrieving or assigning values to selectionStart,
selectionEnd, and selectionDirection property of the input element when the element isn't a text field.

fast/forms/selection-direction-expected.txt: Added.
fast/forms/selection-direction.html: Added.
fast/forms/selection-wrongtype-expected.txt: Added.
fast/forms/selection-wrongtype.html: Added.
platform/mac/editing/deleting/delete-all-text-in-text-field-assertion-expected.txt: Copied from

editing/deleting/delete-all-text-in-text-field-assertion-expected.txt; editing delegate change.

platform/win/editing/deleting/delete-all-text-in-text-field-assertion-expected.txt: Copied from

editing/deleting/delete-all-text-in-text-field-assertion-expected.txt since Windows port fallbacks to Mac.

1:40 PM Changeset in webkit [92087] by Dimitri Glazkov

3 edits in trunk/Tools

Use set comparison to find duplicate or overlapping specifiers in test expectations.
https://bugs.webkit.org/show_bug.cgi?id=65430

Reviewed by Adam Barth.

Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed to use set comparisons, because length

of a set is wrong.

Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Added some unit tests to actually catch

regressions.

1:32 PM Changeset in webkit [92086] by mitz@apple.com

1 edit
2 adds in trunk/LayoutTests

Add expected results for Snow Leopard and earlier for fast/css/font-family-pictograph.html

Fixes REGRESSION (r91777): fast/css/font-family-pictograph.html failing on SnowLeopard Intel Release (Tests)
https://bugs.webkit.org/show_bug.cgi?id=65433

platform/mac-snowleopard/fast/css: Added.
platform/mac-snowleopard/fast/css/font-family-pictograph-expected.txt: Added.

1:13 PM Changeset in webkit [92085] by fpizlo@apple.com

2 edits in trunk/Source/JavaScriptCore

DFG non-speculative JIT does not optimize PutByVal
https://bugs.webkit.org/show_bug.cgi?id=65424

Reviewed by Gavin Barraclough.

Added code to emit PutByVal inline fast path.

dfg/DFGNonSpeculativeJIT.cpp:

(JSC::DFG::NonSpeculativeJIT::compile):

12:03 PM Changeset in webkit [92084] by fpizlo@apple.com

6 edits in trunk/Source/JavaScriptCore

The JSC garbage collector returns memory to the operating system too
eagerly.
https://bugs.webkit.org/show_bug.cgi?id=65382

Reviewed by Oliver Hunt.

This introduces a memory reuse model similar to the one in FastMalloc.
A periodic scavenger thread runs in the background and returns half the
free memory to the OS on each timer fire. New block allocations first
attempt to get the memory from the collector's internal pool, reverting
to OS allocation only when this pool is empty.

heap/Heap.cpp:

(JSC::Heap::Heap):
(JSC::Heap::~Heap):
(JSC::Heap::destroy):
(JSC::Heap::waitForRelativeTimeWhileHoldingLock):
(JSC::Heap::waitForRelativeTime):
(JSC::Heap::blockFreeingThreadStartFunc):
(JSC::Heap::blockFreeingThreadMain):
(JSC::Heap::allocateBlock):
(JSC::Heap::freeBlocks):
(JSC::Heap::releaseFreeBlocks):

heap/Heap.h:
heap/MarkedBlock.cpp:

(JSC::MarkedBlock::destroy):
(JSC::MarkedBlock::MarkedBlock):
(JSC::MarkedBlock::initForCellSize):
(JSC::MarkedBlock::reset):

heap/MarkedBlock.h:
wtf/Platform.h:

9:30 AM Changeset in webkit [92083] by commit-queue@webkit.org

6 edits in trunk/Source/WebKit/chromium

[Chromium] Remove the parameterless WebFrame::createAssociatedURLLoader
overload, and change the overload that takes WebURLLoaderOptions to
use a default value. This is equivalent to the original parameterless
overload.
https://bugs.webkit.org/show_bug.cgi?id=65280

Patch by Bill Budge <bbudge@chromium.org> on 2011-07-31
Reviewed by Darin Fisher.

public/WebFrame.h:
src/AssociatedURLLoader.cpp:
src/AssociatedURLLoader.h:
src/WebFrameImpl.cpp:
src/WebFrameImpl.h:

1:48 AM Changeset in webkit [92082] by dbates@webkit.org

2 edits in trunk/Source/WebCore

[GTK] Remove unused variable coreChild in WebCore::getChildForTable()
https://bugs.webkit.org/show_bug.cgi?id=65427

This variable was unused when it was initially added in <http://trac.webkit.org/changeset/89660> (bug #62718)
and it is unused today. We should remove it. This will also resolve a compiler warning.

Reviewed by Xan Lopez.