Timeline


and

02/26/09:

20:00 Changeset [41278] by adele@apple.com
  • 5 edits in trunk/LayoutTests

2009-02-26 Adele Peterson <adele@apple.com>

Reviewed by Alice Liu.

Make the success/failure of this test more obvious.

  • editing/selection/designmode-no-caret.html:
  • platform/mac/editing/selection/designmode-no-caret-expected.checksum:
  • platform/mac/editing/selection/designmode-no-caret-expected.png:
  • platform/mac/editing/selection/designmode-no-caret-expected.txt:
19:32 Changeset [41277] by barraclough@apple.com
  • 6 edits in trunk

JavaScriptCore:

2009-02-26 Gavin Barraclough <barraclough@apple.com>

Reviewed by Geoff Garen.

Fix bug #23614. Switches on double precision values were incorrectly
truncating the scrutinee value. E.g.:

switch (1.1) { case 1: print("FAIL"); }

Was resulting in FAIL.

  • interpreter/Interpreter.cpp: (JSC::Interpreter::privateExecute):
  • jit/JITStubs.cpp: (JSC::JITStubs::cti_op_switch_imm):

LayoutTests:

2009-02-26 Gavin Barraclough <barraclough@apple.com>

Rubber stamped by Geoff Garen.

Layout test for bug #23614.

  • fast/js/resources/switch-behaviour.js:
  • fast/js/switch-behaviour-expected.txt:
17:15 Changeset [41276] by adele@apple.com
  • 8 edits in trunk

WebKit/mac:

2009-02-26 Adele Peterson <adele@apple.com>

Reviewed by Geoff Garen.

Fix for <rdar://problem/6618166>
https://bugs.webkit.org/show_bug.cgi?id=24216
(REGRESSION r36919) Safari 4 Beta causes MSN Messenger's text entry field to lose focus after entering a message

During a series of firstResponder changes, at some point while the WebHTMLView was losing first responder status,
we incorrectly marked the page as active, and then when the WebHTMLView became first responder again, setActive did nothing.
This change restores behavior from before r36919 to check if the WebHTMLView is in the middle of losing first responder when calling setActive.

In addition to updating editing/selection/designmode-no-caret.html results, I also made sure the test cases that were
fixed in r36919 and r38570 are still fixed.

  • WebView/WebHTMLView.mm: (-[WebHTMLView resignFirstResponder]): Keep track if we're in the process of resigning first responder. (-[WebHTMLView _isResigningFirstResponder]): Added.
  • WebView/WebHTMLViewInternal.h:
  • WebView/WebView.mm: (-[WebView _updateFocusedAndActiveStateForFrame:]): Don't set the page to be active if the document view is currently resigning first responder.

LayoutTests:

2009-02-26 Adele Peterson <adele@apple.com>

Reviewed by Geoff Garen.

Updating results for <rdar://problem/6618166>
https://bugs.webkit.org/show_bug.cgi?id=24216
(REGRESSION r36919) Safari 4 Beta causes MSN Messenger's text entry field to lose focus after entering a message

We changed these results shortly after r36919, but the original result is actually correct.
The test's text indicates that we expect a caret to be placed in this case.

  • platform/mac/editing/selection/designmode-no-caret-expected.checksum:
  • platform/mac/editing/selection/designmode-no-caret-expected.png:
  • platform/mac/editing/selection/designmode-no-caret-expected.txt:
17:09 Changeset [41275] by barraclough@apple.com
  • 4 edits in trunk/JavaScriptCore

2009-02-26 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Integer Immediate representation need not be canonical in x86 JIT code.
On x86-64 we already have loosened the requirement that the int immediate
representation in canonical, we should bring x86 into line.

This patch is a minor (~0.5%) improvement on sunspider & v8-tests, and
should reduce memory footoprint (reduces JIT code size).

  • jit/JIT.cpp: (JSC::JIT::compileOpStrictEq): (JSC::JIT::privateCompileSlowCases):
  • jit/JIT.h: (JSC::JIT::emitJumpIfImmediateNumber): (JSC::JIT::emitJumpIfNotImmediateNumber):
  • jit/JITArithmetic.cpp: (JSC::JIT::putDoubleResultToJSNumberCellOrJSImmediate): (JSC::JIT::compileBinaryArithOp):
16:20 Changeset [41274] by darin@chromium.org
  • 7 edits in trunk/WebCore

2009-02-26 Stephen White <senorblanco@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=23957

Fix for SVG gradient and pattern text for Chromium/skia.
Added accessors for the fill and stroke gradients to
GraphicsContext. Changed the paintSkiaText function to take a
GraphicsContext, so we can check for gradients/patterns.
Changed the skiaDrawText function to set the SkPaint shader
correctly, and to scale up the gradient shader matrix to
encompass the entire text string. Also offset each glyph
separately, rather than transforming the canvas, so that the
gradient/pattern stays fixed relative to the text origin.

  • platform/graphics/GraphicsContext.cpp: (WebCore::GraphicsContext::getFillGradient): (WebCore::GraphicsContext::getStrokeGradient): (WebCore::GraphicsContext::getFillPattern): (WebCore::GraphicsContext::getStrokePattern):
  • platform/graphics/GraphicsContext.h:
  • platform/graphics/chromium/FontChromiumWin.cpp: (WebCore::Font::drawGlyphs):
  • platform/graphics/chromium/UniscribeHelper.cpp: (WebCore::UniscribeHelper::draw):
  • platform/graphics/skia/SkiaFontWin.cpp: (WebCore::windowsCanHandleTextDrawing): (WebCore::skiaDrawText): (WebCore::paintSkiaText):
  • platform/graphics/skia/SkiaFontWin.h:
15:29 Changeset [41273] by dglazkov@chromium.org
  • 1 edit in trunk/WebCore/ChangeLog

Fix ChangeLog entry

15:27 Changeset [41272] by dglazkov@chromium.org
  • 1 edit
    4 adds in trunk/WebCore

2009-02-26 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

http://codereview.chromium.org/27217
Add custom V8 bindings for Navigator, Clipboard, Document, and Node.

  • bindings/v8/custom/V8ClipboardCustom.cpp: Added. (WebCore::ACCESSOR_GETTER): (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8DocumentCustom.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8NavigatorCustom.cpp: Added. (WebCore::ACCESSOR_GETTER):
  • bindings/v8/custom/V8NodeCustom.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
15:00 Changeset [41271] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-02-26 Gustavo Noronha Silva <gns@gnome.org>

Unreviewed build fix for building with GNOME Keyring enabled.

  • platform/network/soup/webkit-soup-auth-dialog.c: (show_auth_dialog):
14:45 Changeset [41270] by kov@webkit.org
  • 7 edits
    2 adds in trunk

2009-02-26 Xan Lopez <xan@gnome.org>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=16947
[GTK] Missing HTTP Auth challenge

Add HTTP authentication dialog with optional GNOME Keyring
storage.

  • GNUmakefile.am:
  • platform/network/ResourceHandleInternal.h: (WebCore::ResourceHandleInternal::ResourceHandleInternal):
  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::currentToplevelCallback): (WebCore::ResourceHandle::startHttp): (WebCore::ResourceHandle::start):
  • platform/network/soup/webkit-soup-auth-dialog.c: Added. (webkit_soup_auth_dialog_class_init): (webkit_soup_auth_dialog_init): (webkit_soup_auth_dialog_session_feature_init): (free_authData): (set_password_callback): (response_callback): (show_auth_dialog): (find_password_callback): (session_authenticate): (attach):
  • platform/network/soup/webkit-soup-auth-dialog.h: Added.
13:11 Changeset [41269] by eric@webkit.org
  • 3 edits in trunk/WebCore

2009-02-25 Ojan Vafai <ojan@chromium.org> and Eric Seidel <eric@webkit.org>

Reviewed by Dave Hyatt.

After Ojan's positionForCoordinates fix http://trac.webkit.org/changeset/41191
svg/custom/pointer-events-image.svg and svg/custom/pointer-events-text.svg
started failing because Ojan's new code was now *correctly* calling through to
SVG asking for the closest text offset in the last line box, instead of
just returning the offset at the end of the last line box when clicking below a box.

But! The SVG code was wrong, in that it returned the character offset of the last
character when you asked for a character offset after the end of the box, instead
it should return the offset *after* the last character. This patch fixes
that behavior by reordering the last two clauses in svgCharacterHitsPosition.

The SVG positionForCoordinates function is still wrong, and I've added some FIXMEs
to document what's wrong. I've also cleaned up the code a bit so it's clearer
what it is doing (which also makes more obvious what's wrong with it).

  • rendering/RenderSVGInlineText.cpp: (WebCore::RenderSVGInlineText::positionForCoordinates):
  • rendering/SVGInlineTextBox.cpp: (WebCore::SVGInlineTextBoxClosestCharacterToPositionWalker::SVGInlineTextBoxClosestCharacterToPositionWalker): (WebCore::SVGInlineTextBoxClosestCharacterToPositionWalker::chunkPortionCallback): (WebCore::SVGInlineTextBoxClosestCharacterToPositionWalker::offsetOfHitCharacter): (WebCore::SVGInlineTextBox::closestCharacterToPosition): (WebCore::SVGInlineTextBox::svgCharacterHitsPosition):
12:40 Changeset [41268] by darin@chromium.org
  • 3 edits in trunk/WebCore

2009-02-26 Darin Fisher <darin@chromium.org>

Reviewed by Eric Seidel.

Minor cleanup of ChromiumBridge:
Removing unused matchesMIMEType method and s/Javascript/JavaScript/

  • platform/chromium/ChromiumBridge.h:
  • platform/chromium/MimeTypeRegistryChromium.cpp: (WebCore::MIMETypeRegistry::isSupportedImageMIMEType): (WebCore::MIMETypeRegistry::isSupportedJavaScriptMIMEType): (WebCore::MIMETypeRegistry::isSupportedNonImageMIMEType):
12:36 Changeset [41267] by hyatt@apple.com
  • 3 edits
    4 adds in trunk

WebCore:

2009-02-26 David Hyatt <hyatt@apple.com>

Reviewed by Sam Weinig

Fix text-bottom vertical alignment. It was incorrectly aligning the bottom of the descent instead of including
the line-height below the descent.

Added fast/inline/vertical-align-text-bottom.html

  • rendering/RenderBoxModelObject.cpp: (WebCore::RenderBoxModelObject::verticalPosition):

LayoutTests:

2009-02-26 David Hyatt <hyatt@apple.com>

Reviewed by Sam Weinig

Fix text-bottom vertical alignment. It was incorrectly aligning the bottom of the descent instead of including
the line-height below the descent.

  • fast/inline/vertical-align-text-bottom.html: Added.
  • platform/mac/fast/inline/vertical-align-text-bottom-expected.checksum: Added.
  • platform/mac/fast/inline/vertical-align-text-bottom-expected.png: Added.
  • platform/mac/fast/inline/vertical-align-text-bottom-expected.txt: Added.
12:11 Changeset [41266] by krit@webkit.org
  • 12 edits
    5 deletes in trunk

2009-02-26 Dirk Schulze <krit@webkit.org>

Reviewed by Eric Seidel.

Make SVG Masking platform independet with the use of ImageBuffer::getImageData(),
ImageBuffer::putImageData() and GraphicsContext::clipToImageBuffer(). Every platform has
just to implement this three methods to support SVG Masking now.


Make SVG Masking platform aware
https://bugs.webkit.org/show_bug.cgi?id=19243

  • GNUmakefile.am:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • svg/graphics/SVGResourceMasker.cpp: (WebCore::SVGResourceMasker::applyMask):
  • svg/graphics/cairo/SVGResourceMaskerCairo.cpp: Removed.
  • svg/graphics/cg/SVGResourceMaskerCg.cpp: Removed.
  • svg/graphics/cg/SVGResourceMaskerCg.mm: Removed.
  • svg/graphics/qt/SVGResourceMaskerQt.cpp: Removed.
  • svg/graphics/skia/SVGResourceMaskerSkia.cpp: Removed.

LayoutTests:
Corrected some tests after the new SVG Masking implementation for Mac.

Make SVG Masking platform aware
https://bugs.webkit.org/show_bug.cgi?id=19243

  • platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.checksum:
  • platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.png:
  • platform/mac/svg/custom/grayscale-gradient-mask-expected.checksum:
  • platform/mac/svg/custom/grayscale-gradient-mask-expected.png:
  • svg/custom/mask-excessive-malloc.svg:
12:09 Changeset [41265] by treat@webkit.org
  • 5 edits
    3 adds in trunk

2009-02-26 Yong Li <yong.li@torchmobile.com>

Reviewed by Darin Adler, Antti Koivisto and Alexey Proskuryakov.

Test: http/tests/misc/slow-preload-cancel.html

https://bugs.webkit.org/show_bug.cgi?id=24133
Clear all pending preloads in the DocLoader object when we decide to
cancel its all requests.

  • loader/DocLoader.cpp: (WebCore::DocLoader::clearPendingPreloads):
  • loader/DocLoader.h:
  • loader/loader.cpp: (WebCore::Loader::cancelRequests):
12:07 Changeset [41264] by andersca@apple.com
  • 5 edits in trunk/WebKit/mac

2009-02-25 Anders Carlsson <andersca@apple.com>

Reviewed by Kevin Decker.

Fix <rdar://problem/6623697>.


  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::print): Ask the plug-in host to print, create a CGImage of the returned bytes and draw the image into the passed in context.


  • Plugins/Hosted/WebHostedNetscapePluginView.mm: (-[WebHostedNetscapePluginView drawRect:]): When printing, call NetscapePluginInstanceProxy::print.


  • Plugins/Hosted/WebKitPluginHost.defs:
11:43 Changeset [41263] by dglazkov@chromium.org
  • 1 edit
    3 adds in trunk/WebCore

2009-02-26 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24182
Add NodeFilter, NodeIterator, and TreeWalker custom V8 bindings.

  • bindings/v8/custom/V8NodeFilterCustom.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8NodeIteratorCustom.cpp: Added. (WebCore::toV8): (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8TreeWalkerCustom.cpp: Added. (WebCore::toV8): (WebCore::CALLBACK_FUNC_DECL):
11:40 Changeset [41262] by ap@webkit.org
  • 4 edits in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23500
KURL::parse() incorrectly compares its result to original string

  • platform/KURL.cpp: (WebCore::KURL::parse): Take string length into account.
11:35 Changeset [41261] by dglazkov@chromium.org
  • 1 edit
    1 add in trunk/WebCore

2009-02-26 Ojan Vafai <ojan@chromium.org>

Reviewed by Kevin McCullough.

Manual test for inspector node highlighting.

  • manual-tests/inspector/highlight-nodes.html: Added.
11:32 Changeset [41260] by darin@chromium.org
  • 2 edits in trunk/WebCore

2009-02-16 Anantanarayanan Iyengar <ananta@chromium.org>

Reviewed by Darin Fisher.

https://bugs.webkit.org/show_bug.cgi?id=23973
ScrollView::scrollContents can be invoked during view shutdown. In
this scenario the FrameView::hostWindow method can return NULL, which
indicates that the frame/page is being destroyed. This causes a crash
when we try to dereference a NULL hostWindow pointer. Fix is to add a
NULL check for this.

  • platform/ScrollView.cpp: (WebCore::ScrollView::scrollContents):
11:11 Changeset [41259] by dglazkov@chromium.org
  • 3 edits
    2 adds in trunk

WebCore:

2009-02-26 Rahul Kuchhal <kuchhal@chromium.org>

Reviewed by Dave Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=24003
Fix a crash caused by unsafe type conversion.

Test: fast/block/positioning/absolute-in-inline-rtl-4.html

  • rendering/RenderBox.cpp: (WebCore::RenderBox::calcAbsoluteHorizontalValues):

LayoutTests:

2009-02-26 Rahul Kuchhal <kuchhal@chromium.org>

Reviewed by Dave Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=24003
Fix a crash caused by unsafe type conversion.

  • fast/block/positioning/absolute-in-inline-rtl-4-expected.txt: Added.
  • fast/block/positioning/absolute-in-inline-rtl-4.html: Added.
10:59 Changeset [41258] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

2009-02-26 Ojan Vafai <ojan@chromium.org>

Reviewed by Adam Roben.

https://bugs.webkit.org/show_bug.cgi?id=24202
Have drawNodeHighlight clip instead of clearing. This makes it work for
both Safari and Chromium since Chromium draws the inspector highlighting
in the same buffer as the page.

  • inspector/InspectorController.cpp: (WebCore::quadToPath): (WebCore::drawOutlinedQuad): (WebCore::drawOutlinedQuadWithClip): (WebCore::drawHighlightForBox):
10:45 Changeset [41257] by hyatt@apple.com
  • 3 edits
    4 adds in trunk

WebCore:

2009-02-26 David Hyatt <hyatt@apple.com>

Reviewed by Dan Bernstein & Darin Adler

Make sure the border/padding are properly omitted at the start of an inline that is a continuation.


Added fast/inline/inline-continuation-borders.html

  • rendering/InlineFlowBox.cpp: (WebCore::InlineFlowBox::determineSpacingForFlowBoxes):

LayoutTests:

2009-02-26 David Hyatt <hyatt@apple.com>

Reviewed by Dan Bernstein & Darin Adler

Make sure the border/padding are properly omitted at the start of an inline that is a continuation.

  • fast/inline/inline-continuation-borders.html: Added.
  • platform/mac/fast/inline/inline-continuation-borders-expected.checksum: Added.
  • platform/mac/fast/inline/inline-continuation-borders-expected.png: Added.
  • platform/mac/fast/inline/inline-continuation-borders-expected.txt: Added.
10:30 Changeset [41256] by ap@webkit.org
  • 4 edits in trunk/WebKit

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24024
REGRESSION (r39845): Assertion failure in -[WebHistoryItem dictionaryRepresentation] when
archiving a submission to about:blank

I don't know how to make an automated test for this bug.

  • History/WebHistory.mm: (-[WebHistory _visitedURL:withTitle:method:wasFailure:]):
  • WebHistory.cpp: (WebHistory::visitedURL): Account for the fact that HTTP method may be non-empty for non-HTTP requests.
10:21 Changeset [41255] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2009-02-26 Simon Fraser <simon.fraser@apple.com>

Build fix, no review.

Try to fix the wx build after r41218.

  • WebCoreSources.bkl:
10:19 Changeset [41254] by pam@chromium.org
  • 1 edit
    2 adds in trunk/LayoutTests

2009-02-25 Pamela Greene <pam@chromium.org>

Reviewed by Darin Fisher.

Add test that custom DOM properties survive garbage collection.
https://bugs.webkit.org/show_bug.cgi?id=24142

  • fast/dom/Window/customized-property-survives-gc-expected.txt: Added.
  • fast/dom/Window/customized-property-survives-gc.html: Added.
10:17 Changeset [41253] by ap@webkit.org
  • 4 edits
    2 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=19527
ASSERTION FAILED: containerA && containerB

Test: fast/dom/Range/bug-19527.html

  • dom/Range.cpp: (WebCore::Range::isPointInRange): (WebCore::Range::comparePoint): Bring the behavior in line with current Firefox, making it impossible for these methods to pass a null container to compareBoundaryPoints().
09:08 Changeset [41252] by zecke@webkit.org
  • 2 edits in trunk/JavaScriptCore

2009-02-26 Carol Szabo <carol.szabo@nokia.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24099
ARM Compiler Warnings in pcre_exec.cpp

  • pcre/pcre_exec.cpp: (match):
08:48 Changeset [41251] by zecke@webkit.org
  • 3 edits
    1 add in trunk

2009-02-26 Xan Lopez <xan@gnome.org>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=16947
[GTK] Missing HTTP Auth challenge

Take marshallers to be built from a manually maintained list
instead of grepping the sources.

It's much faster, especially so now that we want to add
marshallers from WebCore too. A system to only take into account
the modified files when generating the marshallers from sources
could be hacked, but I think it's overkill considering how rarely
a new marshaller is added.

08:40 Changeset [41250] by zecke@webkit.org
  • 1 edit in trunk/WebKit/gtk/ChangeLog

Fix the ChangeLog s/frame/content.

I should have read more about the bug history and should have
done the substitution before landing.

08:33 Changeset [41249] by zecke@webkit.org
  • 2 edits in trunk/WebKit/gtk

2009-02-26 Xan Lopez <xan@gnome.org>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=24193
[GTK] Checkbuttons not activated with space

Do not swallow key events with GtkIMContext for non-editable
frames (FIXME: I don't think frame is the right name here).

  • WebCoreSupport/EditorClientGtk.cpp: (WebKit::EditorClient::handleInputMethodKeydown):
08:27 Changeset [41248] by zecke@webkit.org
  • 2 edits in trunk/WebKit/gtk

2009-02-26 Xan Lopez <xan@gnome.org>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=24103
[GTK] Use correct return value for WebView button-release handler

We are returning whatever the core code tells us it did, but this
does not play well with the GTK+ model. GTK+ in general expects it
will see a button-release if it saw a button-press and no
motion/leave/etc events in between. EventHandler.cpp will, in some
cases, not handle press but handle release, confusing the parent
container of the WebView.

As a workaround return always FALSE for button-release (this is
the same than the Windows port does).

  • webkit/webkitwebview.cpp:
08:21 Changeset [41247] by zecke@webkit.org
  • 4 edits in trunk/WebKit/gtk

2009-02-26 Xan Lopez <xan@gnome.org>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=17585
[gtk] get|set encoding api

Add functions to get and set a custom encoding an a view.

  • webkit/webkitprivate.h:
  • webkit/webkitwebview.cpp: (webkit_web_view_set_custom_encoding): (webkit_web_view_get_custom_encoding):
  • webkit/webkitwebview.h:
08:14 Changeset [41246] by zecke@webkit.org
  • 2 edits in trunk/WebCore

Make font-variant:small-caps work properly (bug #20358)

Set the computed size of the small caps font data rather than the specified size
so that the small-caps text is rendered smaller

06:13 Changeset [41245] by treat@webkit.org
  • 2 edits in trunk/WebCore

2009-02-26 Benjamin Meyer <benjamin.meyer@torchmobile.com>

Reviewed by George Staikos.

https://bugs.webkit.org/show_bug.cgi?id=24062
QNetworkCookieJar expects the url and not the policy url. Sending the
policy url will cause QNetworkCookieJar to behave incorrectly. One
example would be a cookie that does not have a path or domain.
QNetworkCookieJar will use the url it is given to fill in default values.
This allows setting cookies on the url of the main frame from an iFrame
when the cookie should be set on the url of the iFrame.

Originally noticed on http://writer.zoho.com/jsp/home.jsp?serviceurl=/index.do

  • platform/qt/CookieJarQt.cpp: (WebCore::setCookies):
03:09 Changeset [41244] by staikos@webkit.org
  • 3 edits in trunk/WebCore

2009-02-26 Charles Wei <charles.wei@torchmobile.com.cn>

Reviewed by George Staikos.

make WebKit/Qt compile with SVG disabled

  • WebCore.pro:
  • platform/graphics/qt/FontPlatformDataQt.cpp:

02/25/09:

23:04 Changeset [41243] by cwzwarich@webkit.org
  • 3 edits
    3 adds in trunk

2009-02-25 Cameron Zwarich <cwzwarich@uwaterloo.ca>

Reviewed by Gavin Barraclough.

Bug 24086: Regression (r40993): WebKit crashes after logging in to lists.zenbe
<https://bugs.webkit.org/show_bug.cgi?id=24086>
<rdar://problem/6625111>

The numeric sort optimization in r40993 generated bytecode for a function
without generating JIT code. This breaks an assumption in some parts of
the JIT's function calling logic that the presence of a CodeBlock implies
the existence of JIT code.

In order to fix this, we simply generate JIT code whenever we check whether
a function is a numeric sort function. This only incurs an additional cost
in the case when the function is a numeric sort function, in which case it
is not expensive to generate JIT code for it.

JavaScriptCore:

  • runtime/ArrayPrototype.cpp: (JSC::isNumericCompareFunction):

LayoutTests:

  • fast/js/resources/sort-no-jit-code-crash.js: Added.
  • fast/js/sort-no-jit-code-crash-expected.txt: Added.
  • fast/js/sort-no-jit-code-crash.html: Added.
20:31 Changeset [41242] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-02-25 Gustavo Noronha Silva <gns@gnome.org>

Unreviewed build fix. Add WebCore/workers to the list of paths
searched by the bindings generator perl script.

  • GNUmakefile.am:
19:45 Changeset [41241] by weinig@apple.com
  • 2 edits in trunk/WebCore

2009-02-25 Sam Weinig <sam@webkit.org>

Reviewed by Geoffrey Garen.

Add canvas to the list of RenderObjects that can mark a page as
visually not empty.

  • rendering/RenderHTMLCanvas.cpp: (WebCore::RenderHTMLCanvas::RenderHTMLCanvas):
18:19 Changeset [41240] by eric.carlson@apple.com
  • 2 edits in trunk/WebCore

2009-02-25 Eric Carlson <eric.carlson@apple.com>

Reviewed by Dave Hyatt.

  • rendering/RenderBox.cpp: (WebCore::RenderBox::nodeAtPoint): Remove assertion fired when child has layer as RenderMedia with controls always has a layer.
18:12 Changeset [41239] by simon.fraser@apple.com
  • 3 edits in trunk/WebKitTools

2009-02-25 Simon Fraser <simon.fraser@apple.com>

Reviewed by Eric Seidel

Fix spew about a WebView being deallocated while key value observers are
still registered with it by making sure that we always stop observing
_isUsingAcceleratedCompositing in -[DumpRenderTreeWindow close].

  • DumpRenderTree/mac/DumpRenderTree.mm: (dumpRenderTree):
  • DumpRenderTree/mac/DumpRenderTreeWindow.mm: (-[DumpRenderTreeWindow close]):
17:24 Changeset [41238] by mitz@apple.com
  • 3 edits
    4 adds in trunk

WebCore:

Reviewed by Simon Fraser.

Test: fast/gradients/background-clipped.html

  • rendering/RenderBoxModelObject.cpp: (WebCore::RenderBoxModelObject::paintFillLayerExtended): Intersect the destination rectangle passed to drawTiledImage() with the dirty rectangle. This makes it more likely for the destination rect to be contained in a single tile rect, which results in a faster code path being taken down the road (just drawing a single tile instead of tiling).

LayoutTests:

Reviewed by Simon Fraser.

  • fast/gradients/background-clipped.html: Added.
  • platform/mac/fast/gradients/background-clipped-expected.checksum: Added.
  • platform/mac/fast/gradients/background-clipped-expected.png: Added.
  • platform/mac/fast/gradients/background-clipped-expected.txt: Added.
16:35 Changeset [41237] by eric@webkit.org
  • 8 edits in trunk

2009-02-25 Ojan Vafai <ojan@chromium.org>

Reviewed by Eric Seidel.

Fix test regressions from positionForCoordinates patch.
https://bugs.webkit.org/show_bug.cgi?id=24148

  • rendering/RenderBlock.cpp: (WebCore::positionForPointWithInlineChildren): Remove ASSERT that placeholder text codepath is hitting.
16:27 Changeset [41236] by cfleizach@apple.com
  • 21 edits
    2 adds in trunk

Bug 24143: Crash occurs at WebCore::AccessibilityTable::isTableExposableThroughAccessibility() when applying a link in GMail
https://bugs.webkit.org/show_bug.cgi?id=24143

When an AX object is marked dirty, do not create AX elements while going up the parent chain.
Do not allow AXRenderObjects to remove their own IDs from the cache, all the cache to do that work
Make sure the AXObjectWrapper's have an object before calling them
In AXObjectCache, change get -> getOrCreate. Use get() to only retrieve an element if it exists

16:16 Changeset [41235] by bdakin@apple.com
  • 2 edits in trunk/WebCore

2009-02-25 Beth Dakin <bdakin@apple.com>

Reviewed by Geoff Garen.

Re-working of <rdar://problem/6487249> repro crash in
WebCore::CSSParser::parseFillImage copying entire contents of this
page
-and corresponding: https://bugs.webkit.org/show_bug.cgi?id=24172

Fixes a bunch of layout test failures I caused!

  • css/CSSParser.cpp: (WebCore::CSSParser::parseValue): (WebCore::CSSParser::parseContent): (WebCore::CSSParser::parseFillImage): (WebCore::CSSParser::parseFontFaceSrc): (WebCore::CSSParser::parseBorderImage):
15:59 Changeset [41234] by dglazkov@chromium.org
  • 1 edit
    8 adds in trunk/WebCore

2009-02-25 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24174
Add more V8 custom bindings.

  • bindings/v8/custom/V8CSSStyleDeclarationCustom.cpp: Added. (WebCore::hasCSSPropertyNamePrefix): (WebCore::cssPropertyName): (WebCore::NAMED_PROPERTY_GETTER): (WebCore::NAMED_PROPERTY_SETTER):
  • bindings/v8/custom/V8DOMStringListCustom.cpp: Added. (WebCore::INDEXED_PROPERTY_GETTER): (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8EventCustom.cpp: Added. (WebCore::ACCESSOR_SETTER): (WebCore::ACCESSOR_GETTER):
  • bindings/v8/custom/V8HTMLDocumentCustom.cpp: Added. (WebCore::NAMED_PROPERTY_DELETER): (WebCore::NAMED_PROPERTY_SETTER): (WebCore::NAMED_PROPERTY_GETTER):
  • bindings/v8/custom/V8HTMLFrameSetElementCustom.cpp: Added. (WebCore::NAMED_PROPERTY_GETTER):
  • bindings/v8/custom/V8NamedNodeMapCustom.cpp: Added. (WebCore::INDEXED_PROPERTY_GETTER): (WebCore::NAMED_PROPERTY_GETTER):
  • bindings/v8/custom/V8NodeListCustom.cpp: Added. (WebCore::NAMED_PROPERTY_GETTER):
  • bindings/v8/custom/V8StyleSheetListCustom.cpp: Added. (WebCore::NAMED_PROPERTY_GETTER):
15:52 Changeset [41233] by dglazkov@chromium.org
  • 6 edits in trunk/WebCore

2009-02-25 Scott Violet <sky@google.com>

Reviewed by Dave Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=24171
Provides a RenderTheme method for getting the scroll bar size and
changes RenderListBox to use it. RenderTheme returns a size of regular,
and Mac's override to return small. Changes ScrollbarThemeChromium to
use the scrollbarsize of the scrollbar when getting the size
instead of passing in no-args.

  • platform/chromium/ScrollbarThemeChromium.cpp: (WebCore::ScrollbarThemeChromium::trackRect): (WebCore::ScrollbarThemeChromium::buttonSize):
  • rendering/RenderListBox.cpp: (WebCore::RenderListBox::createScrollbar):
  • rendering/RenderTheme.h: (WebCore::RenderTheme::scrollbarControlSizeForPart):
  • rendering/RenderThemeChromiumMac.h: (WebCore::RenderThemeChromiumMac::scrollbarControlSizeForPart):
  • rendering/RenderThemeMac.h: (WebCore::RenderThemeMac::scrollbarControlSizeForPart):
15:44 Changeset [41232] by ggaren@apple.com
  • 12 edits
    3 adds in trunk

JavaScriptCore:

2009-02-25 Geoffrey Garen <ggaren@apple.com>

Reviewed by Maciej Stachowiak.


Fixed <rdar://problem/6611174> REGRESSION (r36701): Unable to select
messages on hotmail (24052)


The bug was that for-in enumeration used a cached prototype chain without
validating that it was up-to-date.


This led me to refactor prototype chain caching so it was easier to work
with and harder to get wrong.


After a bit of inlining, this patch is performance-neutral on SunSpider
and the v8 benchmarks.

  • interpreter/Interpreter.cpp: (JSC::Interpreter::tryCachePutByID): (JSC::Interpreter::tryCacheGetByID):
  • jit/JITStubs.cpp: (JSC::JITStubs::tryCachePutByID): (JSC::JITStubs::tryCacheGetByID): (JSC::JITStubs::cti_op_get_by_id_proto_list): Use the new refactored goodness. See lines beginning with "-" and smile.
  • runtime/JSGlobalObject.h: (JSC::Structure::prototypeForLookup): A shout out to const.
  • runtime/JSPropertyNameIterator.h: (JSC::JSPropertyNameIterator::next): We can use a pointer comparison to see if our cached structure chain is equal to the object's structure chain, since in the case of a cache hit, we share references to the same structure chain.
  • runtime/Operations.h: (JSC::countPrototypeChainEntriesAndCheckForProxies): Use the new refactored goodness.
  • runtime/PropertyNameArray.h: (JSC::PropertyNameArray::PropertyNameArray): (JSC::PropertyNameArray::setShouldCache): (JSC::PropertyNameArray::shouldCache): Renamed "cacheable" to "shouldCache" to communicate that the client is specifying a recommendation, not a capability.


  • runtime/Structure.cpp: (JSC::Structure::Structure): No need to initialize a RefPtr. (JSC::Structure::getEnumerablePropertyNames): Moved some code into helper functions.

(JSC::Structure::prototypeChain): New centralized accessor for a prototype
chain. Revalidates on every access, since the objects in the prototype
chain may have mutated.

(JSC::Structure::isValid): Helper function for revalidating a cached
prototype chain.

(JSC::Structure::getEnumerableNamesFromPropertyTable):
(JSC::Structure::getEnumerableNamesFromClassInfoTable): Factored out of
getEnumerablePropertyNames.

  • runtime/Structure.h:
  • runtime/StructureChain.cpp: (JSC::StructureChain::StructureChain):
  • runtime/StructureChain.h: (JSC::StructureChain::create): No need for structureChainsAreEqual, since we use pointer equality now. Refactored StructureChain to make a little more sense and eliminate special cases for null prototypes.

LayoutTests:

2009-02-24 Geoffrey Garen <ggaren@apple.com>

Reviewed by Maciej Stachowiak.


Added a test for <rdar://problem/6611174> REGRESSION (r36701): Unable to
select messages on hotmail (24052)

  • fast/js/for-in-cached-expected.txt: Added.
  • fast/js/for-in-cached.html: Added.
  • fast/js/resources/for-in-cached.js: Added. (forIn):
14:58 Changeset [41231] by bdakin@apple.com
  • 3 edits
    2 adds in trunk

WebCore:

2009-02-25 Beth Dakin <bdakin@apple.com>

Reviewed by Darin Adler.

Fix for <rdar://problem/6487249> repro crash in
WebCore::CSSParser::parseFillImage copying entire contents of this
page
-and corresponding: https://bugs.webkit.org/show_bug.cgi?id=24172

  • css/CSSParser.cpp: (WebCore::CSSParser::parseValue): Null-check m_styleSheet

LayoutTests:

2009-02-25 Beth Dakin <bdakin@apple.com>

Reviewed by Darin Adler.

Test for <rdar://problem/6487249> repro crash in
WebCore::CSSParser::parseFillImage copying entire contents of this
page
-and corresponding: https://bugs.webkit.org/show_bug.cgi?id=24172

  • editing/pasteboard/copy-crash-with-extraneous-attribute-expected.txt: Added.
  • editing/pasteboard/copy-crash-with-extraneous-attribute.html: Added.
13:37 Changeset [41230] by treat@webkit.org
  • 2 edits in trunk/WebCore

2009-02-25 Adam Treat <adam.treat@torchmobile.com>

Fix the Qt build.

  • WebCore.pro:
12:59 Changeset [41229] by simon.fraser@apple.com
  • 2 edits in trunk/WebKit/mac

2009-02-25 Simon Fraser <simon.fraser@apple.com>

Build fix with ACCELERATED_COMPOSITING turned on.

I missed a spot in my last commit in renaming to
_stoppedAcceleratedCompositingForFrame:

  • WebView/WebHTMLView.mm: (-[WebHTMLView close]):
12:45 Changeset [41228] by treat@webkit.org
  • 2 edits in trunk/WebKitTools

2009-02-25 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Alexey Proskuryakov.

Do not queue the calls to 'DumpRenderTree::dump()' as this can result
in more than one call as a test that calls 'notifyDone()' can then be
subsequently fully loaded and initiate a second dump. Also make sure
to stop any existing page load that is happening before running the next
test. Combined this serves to produce 217 more passing tests for the
Qt port or roughly 5% at this point.

  • DumpRenderTree/qt/DumpRenderTree.cpp: (WebCore::DumpRenderTree::DumpRenderTree): (WebCore::DumpRenderTree::open):
12:39 Changeset [41227] by jmalonzo@webkit.org
  • 3 edits in trunk/WebCore

2009-02-25 Jan Michael Alonzo <jmalonzo@webkit.org>

Gtk build fix. Not reviewed.

Add files to the build per r41218.
Rename WebkitPoint.h to WebKitPoint.h in DOMWindow.cpp

  • GNUmakefile.am:
  • page/DOMWindow.cpp:
12:29 Changeset [41226] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-02-25 Zan Dobersek <zandobersek@gmail.com>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=24043
When faced with URLs with unsupported protocol on Gtk port,
report the error through an idle function and return true, so
a proper resource handle is created.

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::reportUnknownProtocolError): (WebCore::ResourceHandle::start):
12:22 Changeset [41225] by sfalken@apple.com
  • 16 edits in trunk/WebCore

Windows build fix.
Use struct to forward declare ResourceRequest.

  • history/HistoryItem.h:
  • inspector/InspectorController.h:
  • loader/DocumentThreadableLoader.h:
  • loader/FrameLoaderClient.h:
  • loader/MainResourceLoader.h:
  • loader/SubresourceLoader.h:
  • loader/SubresourceLoaderClient.h:
  • loader/ThreadableLoader.h:
  • loader/WorkerThreadableLoader.h:
  • platform/CrossThreadCopier.h:
  • platform/network/ResourceHandle.h:
  • platform/network/ResourceHandleClient.h:
  • platform/network/ResourceRequestBase.h:
  • platform/network/cf/ResourceRequestCFNet.h:
  • xml/XMLHttpRequest.h:
12:12 Changeset [41224] by simon.fraser@apple.com
  • 10 edits in trunk

2009-02-25 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=23854

Make an observable property, _isUsingAcceleratedCompositing, on
WebView that DumpRenderTree can use to specialize behavior.

This is implemented via a count of Frames that are using
accelerated compositing.

11:56 Changeset [41223] by alp@webkit.org
  • 4 edits in trunk/WebKit/win

2009-02-25 Andreia Gaita <shana@jitted.com>

Reviewed by Alp Toker.

Update Win def files with recent JavaScriptCore API additions.

  • WebKit.vcproj/WebKit.def:
  • WebKit.vcproj/WebKit_Cairo.def:
  • WebKit.vcproj/WebKit_debug.def:
11:50 Changeset [41222] by sfalken@apple.com
  • 3 edits in trunk/WebCore

Partial Windows build fix.


  • DerivedSources.cpp:
  • WebCore.vcproj/WebCore.vcproj:
11:43 Changeset [41221] by krit@webkit.org
  • 2 edits in trunk/WebCore

2009-02-25 Dirk Schulze <krit@webkit.org>

Reviewed by Oliver Hunt.

Ported arcTo to Qt. Qt has no native support for arcTo. This changes
calculate the behavior of arcTo and draws it with lineTo and arc.

[QT] implement Canvas arcTo
https://bugs.webkit.org/show_bug.cgi?id=23873

  • platform/graphics/qt/PathQt.cpp: (WebCore::Path::addArcTo):
11:29 Changeset [41220] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2009-02-25 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

Remove idl files from Resources, and sort.

  • WebCore.xcodeproj/project.pbxproj:
11:11 Changeset [41219] by sfalken@apple.com
  • 2 edits in trunk/JavaScriptCore

Use timeBeginPeriod to enable timing resolution greater than 16ms in command line jsc for Windows.
Allows more accurate reporting of benchmark times via command line jsc.exe. Doesn't affect WebKit's use of JavaScriptCore.


Reviewed by Adam Roben.

  • jsc.cpp: (main):
11:09 Changeset [41218] by cmarrin@apple.com
  • 15 edits
    6 adds in trunk

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=23943

Added webkitConvertPointFromNodeToPage and webkitConvertPointFromPageToNode on
the window object. Also added WebKitPoint object, which is passed in and out
of these functions.

10:06 Changeset [41217] by levin@chromium.org
  • 3 edits in trunk/WebCore

2009-02-25 Jian Li <jianli@chromium.org>

Reviewed by Alexey Proskuryakov.

Remove "#if ENABLE(WORKERS)" wrap from CrossThreadCopier files.
https://bugs.webkit.org/show_bug.cgi?id=24145

  • platform/CrossThreadCopier.cpp:
  • platform/CrossThreadCopier.h:
10:06 Changeset [41216] by levin@chromium.org
  • 8 edits in trunk/WebCore

2009-02-25 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 23688: ThreadableLoader needs a sync implementation for Workers.
<https://bugs.webkit.org/show_bug.cgi?id=23688>

No observable change in behavior, so no test.

  • loader/ThreadableLoader.cpp: (WebCore::ThreadableLoader::loadResourceSynchronously):
  • loader/ThreadableLoaderClientWrapper.h: (WebCore::ThreadableLoaderClientWrapper::clearClient): (WebCore::ThreadableLoaderClientWrapper::done): (WebCore::ThreadableLoaderClientWrapper::didFinishLoading): (WebCore::ThreadableLoaderClientWrapper::didFail): (WebCore::ThreadableLoaderClientWrapper::didFailRedirectCheck): (WebCore::ThreadableLoaderClientWrapper::ThreadableLoaderClientWrapper): Expose whether the loader is done (based on what callbacks were done).
  • loader/WorkerThreadableLoader.cpp: (WebCore::WorkerThreadableLoader::WorkerThreadableLoader): (WebCore::WorkerThreadableLoader::loadResourceSynchronously): Each loader is given its own mode so that only its callbacks get through the run loop.

The xhr spec says that the readystatechange events are synchronous, so in the case of a
nested sync xhr no readystatechange events should be fired for the outer xhr.

(WebCore::WorkerThreadableLoader::MainThreadBridge::MainThreadBridge):

  • loader/WorkerThreadableLoader.h: (WebCore::WorkerThreadableLoader::create): (WebCore::WorkerThreadableLoader::done):
  • workers/WorkerRunLoop.cpp: (WebCore::WorkerRunLoop::WorkerRunLoop):
  • workers/WorkerRunLoop.h: (WebCore::WorkerRunLoop::createUniqueId): Simple method to create a uniqueId on demand with respect to the run loop.
  • workers/WorkerThread.h:
10:06 Changeset [41215] by levin@chromium.org
  • 10 edits in trunk

WebCore:

2009-02-25 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 24089: ThreadableLoader::loadResourceSynchronously should do callbacks like the async code.
<https://bugs.webkit.org/show_bug.cgi?id=24089>

Make threadable loader callbacks to happen during the sync load call.

Changes the behavior of sync xhr for insecure redirects in two ways:
+ Sends an error event instead of an abort event (which is the same as async xhr's behavior).
+ Throws a network exception which is what other browsers do and what the spec
says to do (http://www.w3.org/TR/XMLHttpRequest/).

  • loader/DocumentThreadableLoader.cpp: (WebCore::DocumentThreadableLoader::loadResourceSynchronously):
  • loader/DocumentThreadableLoader.h:
  • loader/ThreadableLoader.cpp: (WebCore::ThreadableLoader::loadResourceSynchronously):
  • loader/ThreadableLoader.h:
  • xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::XMLHttpRequest): (WebCore::XMLHttpRequest::loadRequestSynchronously): (WebCore::XMLHttpRequest::loadRequestAsynchronously): (WebCore::XMLHttpRequest::didFail): (WebCore::XMLHttpRequest::didFailRedirectCheck):
  • xml/XMLHttpRequest.h:

LayoutTests:

2009-02-25 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 24089: ThreadableLoader::loadResourceSynchronously should do callbacks like the async code.
<https://bugs.webkit.org/show_bug.cgi?id=24089>

Modified the test to account for behavior change and output more information to better detect
future changes in behavior.

  • http/tests/xmlhttprequest/xmlhttprequest-unsafe-redirect-expected.txt:
  • http/tests/xmlhttprequest/xmlhttprequest-unsafe-redirect.html:
09:55 Changeset [41214] by cmarrin@apple.com
  • 5 edits
    8 adds in trunk

Reviewed by David Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=23368

Added logic to correctly hit test accelerated layers.

09:01 Changeset [41213] by ap@webkit.org
  • 3 edits
    2 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24067
REGRESSION: Crash in WebCore::Document::initSecurityContext

The crash started to happen when we removed a check for frame->document() being null.
However, the original document shouldn't be null here, because it is needed to alias
security origins. So, this patch fixes the crash by correcting security origin behavior.

Test: http/tests/security/aboutBlank/window-open-self-about-blank.html
This tests for not crashing, and for inheriting the domain from the document being replaced.
Preserving the aliasing cannot be tested for automatically, because we'd need a non-trivial
domain to remove a prefix from.

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::begin): Create a new document before clearing the frame, so that Document::initSecurityContext() could access the old one.
08:58 Changeset [41212] by darin@chromium.org
  • 2 edits in trunk/WebCore

2009-02-25 Jay Campan <jcampan@google.com>

Reviewed by Darin Fisher.

https://bugs.webkit.org/show_bug.cgi?id=24066

Items in drop-downs were not painted correctly. Makes sure the
PopupListBox invalidates in the coordinates of the window as this is
FramelessScrollView::invalidateRect paints to.

  • platform/chromium/PopupMenuChromium.cpp: (WebCore::PopupListBox::invalidateRow):
08:14 QtWebKitContrib edited by henrique@bastos.net
added Zack Rusin's irc nickname (diff)
06:12 Changeset [41211] by treat@webkit.org
  • 8 edits
    3 adds in trunk

2009-02-25 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Zack Rusin.

https://bugs.webkit.org/show_bug.cgi?id=24158
Implement the queue*() methods of the layoutTestController and begin
implementing the dump of the back/forward list. This results in 2% more
tests passing as well as 23 currently skipped tests now passing.

  • DumpRenderTree/qt/DumpRenderTree.cpp: (WebCore::DumpRenderTree::dumpBackForwardList): (WebCore::DumpRenderTree::dump):
  • DumpRenderTree/qt/DumpRenderTree.h:
  • DumpRenderTree/qt/DumpRenderTree.pro:
  • DumpRenderTree/qt/jsobjects.cpp: (findFrameNamed): (LoadItem::invoke): (ReloadItem::invoke): (ScriptItem::invoke): (BackForwardItem::invoke): (LayoutTestController::reset): (LayoutTestController::processWork): (LayoutTestController::maybeDump): (LayoutTestController::queueBackNavigation): (LayoutTestController::queueForwardNavigation): (LayoutTestController::queueLoad): (LayoutTestController::queueReload): (LayoutTestController::queueScript):
  • DumpRenderTree/qt/jsobjects.h: (LayoutTestController::shouldDumpBackForwardList): (LayoutTestController::dumpBackForwardList):
04:12 Changeset [41210] by hausmann@webkit.org
  • 2 edits in trunk/WebKit/qt

2009-02-25 Kavindra Palaraja <kavindra.palaraja@nokia.com>

Reviewed by Simon Hausmann.

Fix the documentation for the linkHovered signal.

  • Api/qwebpage.cpp:

02/24/09:

22:18 Changeset [41209] by simon.fraser@apple.com
  • 6 edits
    13 adds in trunk

2009-02-24 Simon Fraser <simon.fraser@apple.com>

Reviewed by Anders Carlsson

https://bugs.webkit.org/show_bug.cgi?id=15081

Make display:none work for applet, emebed and object elements
by calling rendererIsNeeded() on superclasses.

Tests: fast/replaced/applet-display-none.html

fast/replaced/embed-display-none.html
fast/replaced/object-display-none.html

  • html/HTMLAppletElement.cpp: (WebCore::HTMLAppletElement::rendererIsNeeded):
  • html/HTMLElement.cpp: (WebCore::HTMLElement::rendererIsNeeded):
  • html/HTMLEmbedElement.cpp: (WebCore::HTMLEmbedElement::rendererIsNeeded):
  • html/HTMLObjectElement.cpp: (WebCore::HTMLObjectElement::rendererIsNeeded):
18:50 Changeset [41208] by simon.fraser@apple.com
  • 24 edits
    2 adds in trunk/WebCore

2009-02-24 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dave Hyatt

https://bugs.webkit.org/show_bug.cgi?id=24137

Fix localToAbsolute() and absoluteToLocal() to map points through 3d transforms,
taking perspective and transform-style: preserve-3d into account.

In order to support transform-style: preserve-3d, which keeps elements in a
3d space, we have to carry along an accumulated matrix in TransformState.
We also need to apply the perspective from the parent, if any, with the
appropriate origin.

18:38 Changeset [41207] by weinig@apple.com
  • 3 edits
    2 adds in trunk

WebCore:

2009-02-24 Sam Weinig <sam@webkit.org>

Reviewed by David Hyatt.

Fix for https://bugs.webkit.org/show_bug.cgi?id=23990
Regression (r40837): JavaScript image popup doesn't work

Make the getClientRects and getBoundingClientRect methods return rects
relative to the viewport.

Test: fast/dom/getBoundingClientRect-getClientRects-relative-to-viewport.html

  • dom/Element.cpp: (WebCore::Element::getClientRects): (WebCore::Element::getBoundingClientRect):

LayoutTests:

2009-02-24 Sam Weinig <sam@webkit.org>

Reviewed by David Hyatt.

Test for https://bugs.webkit.org/show_bug.cgi?id=23990
Regression (r40837): JavaScript image popup doesn't work

  • fast/dom/getBoundingClientRect-getClientRects-relative-to-viewport-expected.txt: Added.
  • fast/dom/getBoundingClientRect-getClientRects-relative-to-viewport.html: Added.
17:46 Changeset [41206] by eric@webkit.org
  • 2 edits in trunk/WebCore

2009-02-24 Ojan Vafai <ojan@chromium.org>

Reviewed by Eric Seidel.

SVG pages don't have a body or an html element!
Don't return early if there is no body.

  • editing/VisiblePosition.cpp: (WebCore::VisiblePosition::canonicalPosition):
17:34 Changeset [41205] by hyatt@apple.com
  • 1 edit in trunk/LayoutTests/platform/mac/fast/block/float/033-expected.txt

Update some layout test results from my previous checkin.

17:33 Changeset [41204] by hyatt@apple.com
  • 3 edits in trunk/LayoutTests/platform/mac/fast/repaint

Update some layout test results from my previous checkin.

17:21 Changeset [41203] by hyatt@apple.com
  • 18 edits
    8 adds in trunk

WebCore:

2009-02-23 David Hyatt <hyatt@apple.com>

Reviewed by Eric Seidel

https://bugs.webkit.org/show_bug.cgi?id=23740, painting order wrong for normal flow elements with overflow: hidden


This patch reworks the painting of overflow. There is now the concept of a "self-painting layer." All
layers are considered to be self-painting except for overflow layers that are normal flow (and that don't have
reflections or masks).


If an overflow layer is not self-painting, then it ends up painted by its parent just like any other normal flow object.
The only difference is that the clip has to be pushed and popped when painting the object's children.


The lightweight clipping scheme used for controls has been extended to cover overflow now in this simplified case. With
the code consolidated into reusable push/pop functions, all of the renderers that use overflow have been patched to
use the new functions.


Hit testing has also been patched to check the overflow clip rect first before recurring into children.


Scrollbar paint has been moved into RenderBlock for now, since none of the table objects support scrollbars
yet, and scrollbar hit testing was already there anyway. Now the two code paths are more symmetrical.


Masks are now treated like normal flow layers (just like reflections).

A couple of test cases have been added to fast/overflow to test the stacking order.

  • rendering/InlineFlowBox.cpp: (WebCore::InlineFlowBox::nodeAtPoint): (WebCore::InlineFlowBox::paint):
  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::repaintOverhangingFloats): (WebCore::RenderBlock::paint): (WebCore::RenderBlock::paintChildren): (WebCore::RenderBlock::paintObject): (WebCore::RenderBlock::paintFloats): (WebCore::RenderBlock::insertFloatingObject): (WebCore::RenderBlock::floatRect): (WebCore::RenderBlock::lowestPosition): (WebCore::RenderBlock::rightmostPosition): (WebCore::RenderBlock::leftmostPosition): (WebCore::RenderBlock::addOverhangingFloats): (WebCore::RenderBlock::nodeAtPoint): (WebCore::RenderBlock::hitTestContents):
  • rendering/RenderBox.cpp: (WebCore::RenderBox::nodeAtPoint): (WebCore::RenderBox::pushContentsClip): (WebCore::RenderBox::popContentsClip):
  • rendering/RenderBox.h: (WebCore::RenderBox::paintObject):
  • rendering/RenderBoxModelObject.cpp: (WebCore::RenderBoxModelObject::hasSelfPaintingLayer):
  • rendering/RenderBoxModelObject.h:
  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::paintLayer): (WebCore::RenderLayer::hitTestLayer): (WebCore::RenderLayer::shouldBeNormalFlowOnly): (WebCore::RenderLayer::isSelfPaintingLayer):
  • rendering/RenderLayer.h:
  • rendering/RenderTable.cpp: (WebCore::RenderTable::paint): (WebCore::RenderTable::paintObject): (WebCore::RenderTable::nodeAtPoint):
  • rendering/RenderTable.h:
  • rendering/RenderTableCell.cpp: (WebCore::RenderTableCell::paint):
  • rendering/RenderTableRow.cpp: (WebCore::RenderTableRow::nodeAtPoint): (WebCore::RenderTableRow::paint):
  • rendering/RenderTableRow.h:
  • rendering/RenderTableSection.cpp: (WebCore::RenderTableSection::paint): (WebCore::RenderTableSection::paintObject): (WebCore::RenderTableSection::nodeAtPoint):
  • rendering/RenderTableSection.h:
  • rendering/RenderTreeAsText.cpp: (WebCore::write): (WebCore::writeLayers):

LayoutTests:

2009-02-24 David Hyatt <hyatt@apple.com>

Reviewed by Eric Seidel

https://bugs.webkit.org/show_bug.cgi?id=23740, painting order wrong for normal flow elements with overflow: hidden


This patch reworks the painting of overflow. There is now the concept of a "self-painting layer." All
layers are considered to be self-painting except for overflow layers that are normal flow (and that don't have
reflections or masks).


If an overflow layer is not self-painting, then it ends up painted by its parent just like any other normal flow object.
The only difference is that the clip has to be pushed and popped when painting the object's children.


The lightweight clipping scheme used for controls has been extended to cover overflow now in this simplified case. With
the code consolidated into reusable push/pop functions, all of the renderers that use overflow have been patched to
use the new functions.


Hit testing has also been patched to check the overflow clip rect first before recurring into children.


Scrollbar paint has been moved into RenderBlock for now, since none of the table objects support scrollbars
yet, and scrollbar hit testing was already there anyway. Now the two code paths are more symmetrical.


Masks are now treated like normal flow layers (just like reflections).


A couple of test cases have been added to fast/overflow to test the stacking order.

  • fast/overflow/overflow-float-stacking.html: Added.
  • fast/overflow/overflow-stacking.html: Added.
  • platform/mac/fast/overflow/overflow-float-stacking-expected.checksum: Added.
  • platform/mac/fast/overflow/overflow-float-stacking-expected.png: Added.
  • platform/mac/fast/overflow/overflow-float-stacking-expected.txt: Added.
  • platform/mac/fast/overflow/overflow-stacking-expected.checksum: Added.
  • platform/mac/fast/overflow/overflow-stacking-expected.png: Added.
  • platform/mac/fast/overflow/overflow-stacking-expected.txt: Added.
17:05 Changeset [41202] by levin@chromium.org
  • 2 edits in trunk/WebCore

2009-02-24 David Levin <levin@chromium.org>

Reviewed by NOBODY (build fix).

Attempted build fix for wx-mac.

  • WebCoreSources.bkl:
16:54 Changeset [41201] by simon.fraser@apple.com
  • 2 edits in trunk/LayoutTests

2009-02-24 Simon Fraser <simon.fraser@apple.com>

Reviewed by Eric Seidel

https://bugs.webkit.org/show_bug.cgi?id=24135

Update the expected value after testcase change.

  • fast/events/offsetX-offsetY-expected.txt:
16:46 Changeset [41200] by levin@chromium.org
  • 3 edits in trunk/WebCore

2009-02-24 David Levin <levin@chromium.org>

Reviewed by NOBODY (build fix).

Attempted build fixes for qt-linux and wx-mac.

  • WebCore.pro:
  • webcore-base.bkl:
16:32 Changeset [41199] by weinig@apple.com
  • 2 edits in trunk/WebCore

2009-02-24 Ojan Vafai <ojan@chromium.org>

Reviewed by Sam Weinig

Fix null pointer error. If the node is the Document, then ownerDocument()
returns null, document() does not.

  • editing/VisiblePosition.cpp: (WebCore::VisiblePosition::canonicalPosition):
16:20 Changeset [41198] by levin@chromium.org
  • 3 edits in trunk/WebCore

2009-02-24 Jian Li <jianli@chromium.org>

Reviewed by NOBODY (attempted build fixes).

Fix build break for Windows and Linux.

  • GNUmakefile.am:
  • WebCore.vcproj/WebCore.vcproj:
16:18 Changeset [41197] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

2009-02-24 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24131
Fix-up COM/RefCounted dichotomy in Chromium port.

  • page/chromium/AccessibilityObjectWrapper.h: (WebCore::AccessibilityObjectWrapper::AccessibilityObjectWrapper): Added explicit setting of recount.
16:13 Changeset [41196] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

2009-02-24 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24141
Add clarity to V8 interceptor helper function.

  • bindings/v8/custom/V8HTMLPlugInElementCustom.cpp: (WebCore::NAMED_PROPERTY_GETTER): Renamed to notHandledByInterceptor. (WebCore::NAMED_PROPERTY_SETTER): Ditto. (WebCore::INDEXED_PROPERTY_GETTER): Ditto. (WebCore::INDEXED_PROPERTY_SETTER): Ditto.
15:57 Changeset [41195] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2009-02-24 Simon Fraser <simon.fraser@apple.com>

Reviewed by Eric Seidel

Minor numeric cleanup: convert float literals to doubles.

  • platform/graphics/transforms/RotateTransformOperation.cpp: (WebCore::RotateTransformOperation::blend):
15:22 Changeset [41194] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

2009-02-24 Mark Mentovai <mark@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24139
Add missing include.

  • bindings/v8/ScriptInstance.h:
15:16 Changeset [41193] by levin@chromium.org
  • 8 edits
    20 moves
    1 add in trunk/WebCore

2009-02-24 Jian Li <jianli@chromium.org>

Reviewed by Alexey Proskuryakov.

Move worker related files from dom directory to worker directory under WebCore.
https://bugs.webkit.org/show_bug.cgi?id=24123

  • DerivedSources.make:
  • GNUmakefile.am:
  • WebCore.pro:
  • WebCore.scons:
  • WebCore.vcproj/MigrateIDLAndScripts:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • platform/CrossThreadCopier.cpp: Renamed from WebCore/dom/CrossThreadCopier.cpp.
  • platform/CrossThreadCopier.h: Renamed from WebCore/dom/CrossThreadCopier.h.
  • workers/GenericWorkerTask.h: Renamed from WebCore/dom/GenericWorkerTask.h.
  • workers/Worker.cpp: Renamed from WebCore/dom/Worker.cpp.
  • workers/Worker.h: Renamed from WebCore/dom/Worker.h.
  • workers/Worker.idl: Renamed from WebCore/dom/Worker.idl.
  • workers/WorkerContext.cpp: Renamed from WebCore/dom/WorkerContext.cpp.
  • workers/WorkerContext.h: Renamed from WebCore/dom/WorkerContext.h.
  • workers/WorkerContext.idl: Renamed from WebCore/dom/WorkerContext.idl.
  • workers/WorkerContextProxy.h: Renamed from WebCore/dom/WorkerContextProxy.h.
  • workers/WorkerLocation.cpp: Renamed from WebCore/dom/WorkerLocation.cpp.
  • workers/WorkerLocation.h: Renamed from WebCore/dom/WorkerLocation.h.
  • workers/WorkerLocation.idl: Renamed from WebCore/dom/WorkerLocation.idl.
  • workers/WorkerMessagingProxy.cpp: Renamed from WebCore/dom/WorkerMessagingProxy.cpp.
  • workers/WorkerMessagingProxy.h: Renamed from WebCore/dom/WorkerMessagingProxy.h.
  • workers/WorkerObjectProxy.h: Renamed from WebCore/dom/WorkerObjectProxy.h.
  • workers/WorkerRunLoop.cpp: Renamed from WebCore/dom/WorkerRunLoop.cpp.
  • workers/WorkerRunLoop.h: Renamed from WebCore/dom/WorkerRunLoop.h.
  • workers/WorkerThread.cpp: Renamed from WebCore/dom/WorkerThread.cpp.
  • workers/WorkerThread.h: Renamed from WebCore/dom/WorkerThread.h.
14:57 Changeset [41192] by simon.fraser@apple.com
  • 3 edits in trunk/WebCore

2009-02-24 Simon Fraser <simon.fraser@apple.com>

Fix build when ACCELERATED_COMPOSITING is turned on
(overflowList() -> normalFlowList()).

  • rendering/RenderLayerBacking.cpp: (WebCore::RenderLayerBacking::hasNonCompositingContent): (WebCore::RenderLayerBacking::paintIntoLayer):
  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::calculateCompositedBounds): (WebCore::RenderLayerCompositor::computeCompositingRequirements): (WebCore::RenderLayerCompositor::rebuildCompositingLayerTree): (WebCore::RenderLayerCompositor::recursiveRepaintLayerRect): (WebCore::RenderLayerCompositor::layerHas3DContent):
14:28 Changeset [41191] by eric@webkit.org
  • 8 edits
    12 adds
    5 deletes in trunk

2009-02-05 Ojan Vafai <ojan@chromium.org> and Eric Seidel <eric@webkit.org>

Reviewed by Dave Hyatt.

Make cursor positions match IE6/IE7/FF3 when clicking in margins/padding
around divs inside editable regions.
https://bugs.webkit.org/show_bug.cgi?id=23605

Fix clicks outside editable regions from focusing the editable region.
https://bugs.webkit.org/show_bug.cgi?id=23607

Removed editing/selection/contenteditable-click-outside.html as it's
not as useful as our new tests.

Clean up RenderBlock::positionForCoordinates to remove dead code,
duplicate code, and generally make it more readable.

Tests: editing/selection/click-in-margins-inside-editable-div.html

editing/selection/click-in-padding-with-multiple-line-boxes.html
editing/selection/click-outside-editable-div.html

  • editing/VisiblePosition.cpp: (WebCore::VisiblePosition::canonicalPosition):
  • rendering/RenderBlock.cpp: (WebCore::positionForPointRespectingEditingBoundaries): (WebCore::positionForPointWithInlineChildren): (WebCore::RenderBlock::positionForCoordinates): (WebCore::RenderBlock::updateFirstLetter):
14:22 Changeset [41190] by weinig@apple.com
  • 13 edits in trunk

WebCore:

2009-02-24 Sam Weinig <sam@webkit.org>

Reviewed by Geoffrey Garen.

Related to <rdar://problem/6590295>
Allow disabling javascript: urls.

  • WebCore.base.exp:
  • html/HTMLAnchorElement.cpp: (WebCore::HTMLAnchorElement::parseMappedAttribute):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::executeIfJavaScriptURL):
  • page/Page.cpp: (WebCore::Page::Page): (WebCore::Page::setJavaScriptURLsAreAllowed): (WebCore::Page::javaScriptURLsAreAllowed):
  • page/Page.h:

WebKit/mac:

2009-02-24 Sam Weinig <sam@webkit.org>

Reviewed by Geoffrey Garen.

Related to <rdar://problem/6590295>
Allow disabling javascript: urls.

  • WebView/WebView.mm: (-[WebView _setJavaScriptURLsAreAllowed:]):
  • WebView/WebViewPrivate.h:

WebKit/win:

2009-02-24 Sam Weinig <sam@webkit.org>

Reviewed by Geoffrey Garen.

Related to <rdar://problem/6590295>
Allow disabling javascript: urls.

  • Interfaces/IWebViewPrivate.idl:
  • WebView.cpp: (WebView::setJavaScriptURLsAreAllowed):
  • WebView.h:
14:19 Changeset [41189] by simon.fraser@apple.com
  • 4 edits in trunk

2009-02-24 Simon Fraser <simon.fraser@apple.com>

Reviewed by Darin Adler

https://bugs.webkit.org/show_bug.cgi?id=24135

Round the FloatPoint returned by absoluteToLocal(), rather than
truncating it.

  • dom/MouseRelatedEvent.cpp: (WebCore::MouseRelatedEvent::receivedTarget):
13:34 Changeset [41188] by bdakin@apple.com
  • 5 edits
    4 adds in trunk

WebCore:

2009-02-24 Beth Dakin <bdakin@apple.com>

Reviewed by Sam Weinig.

Fix for https://bugs.webkit.org/show_bug.cgi?id=24004 REGRESSION:
Ordered list item marker misaligned when line height is not 1.2
-and corresponding <rdar://problem/6602506>

This is a regression from http://trac.webkit.org/changeset/40880
where createInlineBox was taken off of RenderObject and moved to
RenderBox. The problem was that the RenderBox version still needs
to be virtual because there is still an existing implementation in
RenderListMarker.

  • rendering/RenderBox.h:
  • rendering/RenderListMarker.cpp: (WebCore::RenderListMarker::createInlineBox):
  • rendering/RenderListMarker.h:

LayoutTests:

2009-02-24 Beth Dakin <bdakin@apple.com>

Reviewed by Sam Weinig.

Test for https://bugs.webkit.org/show_bug.cgi?id=24004 REGRESSION:
Ordered list item marker misaligned when line height is not 1.2
-and corresponding <rdar://problem/6602506>

  • fast/lists/list-marker-with-line-height.html: Added.
  • platform/mac/fast/lists/list-marker-with-line-height-expected.checksum: Added.
  • platform/mac/fast/lists/list-marker-with-line-height-expected.png: Added.
  • platform/mac/fast/lists/list-marker-with-line-height-expected.txt: Added.
12:47 Changeset [41187] by aroben@apple.com
  • 3 edits in trunk/WebKit/win

Windows build fix

  • WebKit.vcproj/WebKit.def:
  • WebKit.vcproj/WebKit_debug.def: Export WTF::ThreadCondition::timedWait.
12:23 Changeset [41186] by dglazkov@chromium.org
  • 1 edit
    6 adds in trunk/WebCore

2009-02-24 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24128
Upstream more V8 custom bindings: constructors, XSLTProcessor.

  • bindings/v8/custom/V8DOMParserConstructor.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8MessageChannelConstructor.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8WebKitCSSMatrixConstructor.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8XMLSerializerConstructor.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8XPathEvaluatorConstructor.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
  • bindings/v8/custom/V8XSLTProcessorCustom.cpp: Added. (WebCore::CALLBACK_FUNC_DECL):
11:46 Changeset [41185] by treat@webkit.org
  • 2 edits in trunk/LayoutTests

2009-02-24 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Simon Fraser.

Adding these tests back in and removing from the Skipped list as
these tests are all now passing on my machine as of latest revision.

  • platform/qt/Skipped:
10:54 Changeset [41184] by levin@chromium.org
  • 6 edits in trunk/WebCore

2009-02-24 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 24090: WorkerThreadableLoader needs to be able to post tasks for a mode.
<https://bugs.webkit.org/show_bug.cgi?id=24090>

No observable change in behavior, so no test.

  • dom/WorkerMessagingProxy.cpp: (WebCore::WorkerMessagingProxy::postMessageToWorkerContext): (WebCore::WorkerMessagingProxy::postTaskToWorkerContext): Fixed the bug that m_unconfirmedMessageCount was getting incremented for non-message tasks.

(WebCore::WorkerMessagingProxy::postTaskForModeToWorkerContext):

  • dom/WorkerMessagingProxy.h:
  • loader/ThreadableLoader.cpp: (WebCore::ThreadableLoader::create):
  • loader/WorkerThreadableLoader.cpp: (WebCore::WorkerThreadableLoader::WorkerThreadableLoader): (WebCore::WorkerThreadableLoader::MainThreadBridge::MainThreadBridge): (WebCore::WorkerThreadableLoader::MainThreadBridge::didSendData): (WebCore::WorkerThreadableLoader::MainThreadBridge::didReceiveResponse): (WebCore::WorkerThreadableLoader::MainThreadBridge::didReceiveData): (WebCore::WorkerThreadableLoader::MainThreadBridge::didFinishLoading): (WebCore::WorkerThreadableLoader::MainThreadBridge::didFail): (WebCore::WorkerThreadableLoader::MainThreadBridge::didFailRedirectCheck): (WebCore::WorkerThreadableLoader::MainThreadBridge::didReceiveAuthenticationCancellation): Changed these methods to post task using a mode.
  • loader/WorkerThreadableLoader.h: (WebCore::WorkerThreadableLoader::create):
10:17 Changeset [41183] by treat@webkit.org
  • 2 edits in trunk/WebKitTools

2009-02-24 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Alexey Proskuryakov.

Don't print out that you are generating new results if you are not in fact
generating new results and disable generating new results by default for
the all ports other than the canonical Apple Mac port since this can litter
the source directory with hundreds of new results since other ports
are not as up to date.

  • Scripts/run-webkit-tests:
08:33 Changeset [41182] by mrowe@apple.com
  • 4 copies
    1 add in releases/Apple/Safari 4 Public Beta

Tag the Safari 4 Public Beta release.

This release consists of JavaScriptCore-x528.15, JavaScriptGlue-x528.15, WebCore-x528.16 and WebKit-x528.16.

08:23 Changeset [41181] by treat@webkit.org
  • 2 edits in trunk/WebKitTools

2009-02-24 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Cameron Zwarich.

The Qt port does not support these yet nor does the nm check work with
QMake based build.

  • Scripts/webkitdirs.pm:
08:12 Changeset [41180] by treat@webkit.org
  • 2 edits in trunk/LayoutTests

2009-02-24 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Cameron Zwarich.

Remove tests from the Qt Skipped list that no longer even exist.

  • platform/qt/Skipped:
07:23 Changeset [41179] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

2009-02-24 Geoffrey Garen <ggaren@apple.com>

Build fix?

  • GNUmakefile.am:
05:42 Changeset [41178] by ap@webkit.org
  • 3 edits
    4 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24091
<rdar://problem/6468660> Start of redirect chain ends up as master entry in Application Cache

Test: http/tests/appcache/access-via-redirect.php

  • loader/appcache/ApplicationCacheGroup.cpp: (WebCore::ApplicationCacheGroup::finishedLoadingMainResource): Use the URL we ended up with, not the original one.
05:26 Changeset [41177] by mrowe@apple.com
  • 2 edits in trunk/JavaScriptGlue

<rdar://problem/6407844> JavaScriptGlue should use -Os rather than -O2

Rubber-stamped by Dan Bernstein.

  • Configurations/Base.xcconfig:
04:35 Changeset [41176] by mrowe@apple.com
  • 11 edits in trunk

<rdar://problem/6259220> Rename AVAILABLE_AFTER_WEBKIT_VERSION_3_1 (etc.) to match the other macros

Reviewed by Oliver Hunt.

  • API/JSBasePrivate.h:
  • API/JSContextRef.h:
  • API/JSObjectRef.h:
  • API/WebKitAvailability.h:
  • Carbon/CarbonUtils.h:
  • Carbon/HIWebView.h:
  • Plugins/WebPlugin.h:
  • Plugins/WebPluginViewFactory.h:
  • WebView/WebUIDelegate.h:
02:21 Changeset [41175] by ap@webkit.org
  • 2 edits in trunk/LayoutTests

Reviewed by Mark Rowe ("awesome!").

Make test for <https://bugs.webkit.org/show_bug.cgi?id=15707> more reliable.

  • fast/dom/onload-open.html: Added waitUntilDone().
01:44 Changeset [41174] by levin@chromium.org
  • 3 edits in trunk/WebCore

2009-02-24 Jian Li <jianli@chromium.org>

Reviewed by Alexey Proskuryakov.

Change to include WorkerObjectProxy.h instead of WorkerMessagingProxy.h in WorkerContext.cpp and WorkerScriptController.cpp.
https://bugs.webkit.org/show_bug.cgi?id=24112

  • bindings/js/WorkerScriptController.cpp:
  • dom/WorkerContext.cpp:
01:39 Changeset [41173] by mrowe@apple.com
  • 2 edits in trunk/WebKit/mac

2009-02-24 Peter Ammon <pammon@apple.com>

Reviewed by Mark Rowe.

Fix <rdar://problem/6251410> Services can modify non-editable content in Safari

  • WebView/WebHTMLView.mm: (-[WebHTMLView validRequestorForSendType:returnType:]): Return self only if we can handle both the send and return type. We should also handle a nil send or return type by ignoring the argument and returning whether we can handle the other type passed in.
01:22 Changeset [41172] by jmalonzo@webkit.org
  • 4 edits in trunk

[Gtk] add options for 3D transforms and HTML5 channel messaging to the build
https://bugs.webkit.org/show_bug.cgi?id=24072

Reviewed by Alexey Proskuryakov.

Add options for toggling 3D transforms and HTML5 channel messaging
support on or off.

Also fix the web-workers option. It should be web-workers and not workers.

  • configure.ac:

Allow toggling of 3D transforms and HTML5 channel messaging
support for the Autotools (Gtk) build.

Also add '--gtk' in the build-webkit help doc and fix autotools
option for web-workers support.

  • Scripts/build-webkit:
00:57 Changeset [41171] by levin@chromium.org
  • 2 edits in trunk/LayoutTests

2009-02-24 David Levin <levin@chromium.org>

Reviewed by Dave Hyatt.

Update expected results due to better results introduced by r41090.

  • platform/mac/media/video-layer-crash-expected.txt:

02/23/09:

23:47 Changeset [41170] by antti@apple.com
  • 2 edits in trunk/WebCore

2009-02-23 Antti Koivisto <antti@apple.com>

Reviewed by Oliver Hunt.

<rdar://problem/6613796> Extended text codecs registered on webview creation

Comparing a text encoding with string "GBK" ended up constructing
TextEncoding("GBK") which in turn initialized all extended
text codecs.

  • platform/text/TextCodecICU.cpp: (WebCore::TextCodecICU::decode):
20:29 Changeset [41169] by weinig@apple.com
  • 2 edits in trunk/WebCore

2009-02-23 Sam Weinig <sam@webkit.org>

Fix Qt build.

  • bridge/qt/qt_runtime.cpp: (JSC::Bindings::valueRealType):
19:58 Changeset [41168] by ggaren@apple.com
  • 22 edits in trunk/JavaScriptCore

2009-02-23 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig.

Next step in splitting JIT functionality out of the Interpreter class:
Moved vptr storage from Interpreter to JSGlobalData, so it could be shared
between Interpreter and JITStubs, and moved the *Trampoline JIT stubs
into the JITStubs class. Also added a VPtrSet class to encapsulate vptr
hacks during JSGlobalData initialization.


SunSpider says 0.4% faster. Meh.

  • JavaScriptCore.exp:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • interpreter/Interpreter.cpp: (JSC::Interpreter::Interpreter): (JSC::Interpreter::tryCacheGetByID): (JSC::Interpreter::privateExecute):
  • interpreter/Interpreter.h:
  • jit/JIT.cpp: (JSC::JIT::privateCompileMainPass): (JSC::JIT::privateCompile): (JSC::JIT::privateCompileCTIMachineTrampolines):
  • jit/JIT.h: (JSC::JIT::compileCTIMachineTrampolines):
  • jit/JITCall.cpp: (JSC::JIT::compileOpCall): (JSC::JIT::compileOpCallSlowCase):
  • jit/JITPropertyAccess.cpp: (JSC::JIT::privateCompilePatchGetArrayLength):
  • jit/JITStubs.cpp: (JSC::JITStubs::JITStubs): (JSC::JITStubs::tryCacheGetByID): (JSC::JITStubs::cti_vm_dontLazyLinkCall): (JSC::JITStubs::cti_op_get_by_val): (JSC::JITStubs::cti_op_get_by_val_byte_array): (JSC::JITStubs::cti_op_put_by_val): (JSC::JITStubs::cti_op_put_by_val_array): (JSC::JITStubs::cti_op_put_by_val_byte_array): (JSC::JITStubs::cti_op_is_string):
  • jit/JITStubs.h: (JSC::JITStubs::ctiArrayLengthTrampoline): (JSC::JITStubs::ctiStringLengthTrampoline): (JSC::JITStubs::ctiVirtualCallPreLink): (JSC::JITStubs::ctiVirtualCallLink): (JSC::JITStubs::ctiVirtualCall):
  • runtime/ArrayPrototype.cpp: (JSC::arrayProtoFuncPop): (JSC::arrayProtoFuncPush):
  • runtime/FunctionPrototype.cpp: (JSC::functionProtoFuncApply):
  • runtime/JSArray.h: (JSC::isJSArray):
  • runtime/JSByteArray.h: (JSC::asByteArray): (JSC::isJSByteArray):
  • runtime/JSCell.h:
  • runtime/JSFunction.h:
  • runtime/JSGlobalData.cpp: (JSC::VPtrSet::VPtrSet): (JSC::JSGlobalData::JSGlobalData): (JSC::JSGlobalData::create): (JSC::JSGlobalData::sharedInstance):
  • runtime/JSGlobalData.h:
  • runtime/JSString.h: (JSC::isJSString):
  • runtime/Operations.h: (JSC::jsLess): (JSC::jsLessEq):
  • wrec/WREC.cpp: (JSC::WREC::Generator::compileRegExp):
18:57 Changeset [41167] by mrowe@apple.com
  • 1 edit in tags/Safari-6530.1/WebCore/page/AccessibilityTable.cpp

Work around <rdar://problem/6610801>.

17:47 Changeset [41166] by mrowe@apple.com
  • 5 edits
    1 add in tags/Safari-6530.1/WebKit/mac

Merge r41148.

17:46 Changeset [41165] by mrowe@apple.com
  • 1 add in tags/Safari-6530.1/WebKit/mac/Plugins/Hosted/WebTextInputWindowController.m

Merge r41124.

17:46 Changeset [41164] by mrowe@apple.com
  • 1 edit in tags/Safari-6530.1/WebKit/mac/ChangeLog

Merge r41124.

17:45 Changeset [41163] by mrowe@apple.com
  • 3 edits in tags/Safari-6530.1/WebKit

Merge r41123.

17:21 Changeset [41162] by jchaffraix@webkit.org
  • 5 edits
    2 adds in trunk

WebCore:

2009-02-23 Julien Chaffraix <jchaffraix@webkit.org>

Reviewed by Darin Adler.

Bug 23956: Safari crashes when cloneNode fails (cloning a XML element with an invalid nodeName)

The crash occurred because Document::cloneNode would call Document::createElementNS. Unfortunately
element created with createElement could have a wrong nodeName (createElement sets the string as the
localName without checking for a prefix).

The fix is to call Document::createElement(const QualifiedName&, bool) that will not do any checks on the QualifiedName
and will always succeed.
Also rolled-out the HTMLElement specialisation of clonedNode as it was equivalent to what is done now (added an ASSERT
to ensure that).

Test: fast/dom/cloneNode.html

  • dom/Element.cpp: (WebCore::Element::cloneNode): Call createElement(const QualifiedName&, bool) instead of createElementNS as it will always return an element.
  • html/HTMLElement.cpp:
  • html/HTMLElement.h: Removed HTMLElement::clonedNode as it is equivalent to what is now done.

LayoutTests:

2009-02-23 Julien Chaffraix <jchaffraix@webkit.org>

Reviewed by Darin Adler.

Test for bug 23956: Safari crashes when cloneNode fails (cloning a XML element with an invalid nodeName)

  • fast/dom/cloneNode-expected.txt: Added.
  • fast/dom/cloneNode.html: Added.
16:53 Changeset [41161] by mrowe@apple.com
  • 1 copy in tags/Safari-6530.1

New tag.

16:29 Changeset [41160] by oliver@apple.com
  • 2 edits in trunk/JavaScriptCore

Bug 23787: Allow JIT to generate SSE2 code if using GCC
<https://bugs.webkit.org/show_bug.cgi?id=23787>

Reviewed by Oliver Hunt
Contributed by Csaba Osztrogonac

GCC version of the cpuid check.

16:04 Changeset [41159] by dglazkov@chromium.org
  • 1 edit
    12 adds in trunk/WebCore

2009-02-23 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24109
Upstream V8 Script abstractions, all except ScriptController.

  • bindings/v8/ScriptCachedFrameData.h: Added. (WebCore::ScriptCachedFrameData::ScriptCachedFrameData): (WebCore::ScriptCachedFrameData::~ScriptCachedFrameData): (WebCore::ScriptCachedFrameData::restore): (WebCore::ScriptCachedFrameData::clear): (WebCore::ScriptCachedFrameData::domWindow):
  • bindings/v8/ScriptCallFrame.cpp: Added. (WebCore::ScriptCallFrame::ScriptCallFrame): (WebCore::ScriptCallFrame::~ScriptCallFrame): (WebCore::ScriptCallFrame::argumentAt):
  • bindings/v8/ScriptCallFrame.h: Added. (WebCore::ScriptCallFrame::functionName): (WebCore::ScriptCallFrame::sourceURL): (WebCore::ScriptCallFrame::lineNumber): (WebCore::ScriptCallFrame::argumentCount):
  • bindings/v8/ScriptCallStack.cpp: Added. (WebCore::ScriptCallStack::ScriptCallStack): (WebCore::ScriptCallStack::~ScriptCallStack): (WebCore::ScriptCallStack::at):
  • bindings/v8/ScriptCallStack.h: Added. (WebCore::ScriptCallStack::size):
  • bindings/v8/ScriptInstance.cpp: Added. (WebCore::V8ScriptInstance::V8ScriptInstance): (WebCore::V8ScriptInstance::~V8ScriptInstance): (WebCore::V8ScriptInstance::instance): (WebCore::V8ScriptInstance::clear): (WebCore::V8ScriptInstance::set):
  • bindings/v8/ScriptInstance.h: Added. (WebCore::V8ScriptInstance::create):
  • bindings/v8/ScriptSourceCode.h: Added. (WebCore::ScriptSourceCode::ScriptSourceCode): (WebCore::ScriptSourceCode::isEmpty): (WebCore::ScriptSourceCode::source): (WebCore::ScriptSourceCode::url): (WebCore::ScriptSourceCode::startLine):
  • bindings/v8/ScriptState.h: Added. (WebCore::ScriptState::hadException): (WebCore::ScriptState::setException): (WebCore::ScriptState::exception):
  • bindings/v8/ScriptString.h: Added. (WebCore::ScriptString::ScriptString): (WebCore::ScriptString::operator String): (WebCore::ScriptString::isNull): (WebCore::ScriptString::size): (WebCore::ScriptString::operator=): (WebCore::ScriptString::operator+=):
  • bindings/v8/ScriptValue.cpp: Added. (WebCore::ScriptValue::getString):
  • bindings/v8/ScriptValue.h: Added. (WebCore::ScriptValue::ScriptValue): (WebCore::ScriptValue::operator=): (WebCore::ScriptValue::operator==): (WebCore::ScriptValue::operator!=): (WebCore::ScriptValue::isNull): (WebCore::ScriptValue::isUndefined): (WebCore::ScriptValue::clear): (WebCore::ScriptValue::~ScriptValue): (WebCore::ScriptValue::v8Value):
15:20 Changeset [41158] by treat@webkit.org
  • 2 edits in trunk/WebCore

2009-02-23 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Timothy Hatcher.

https://bugs.webkit.org/show_bug.cgi?id=24106
The Qt port is crashing on exit because the tear down procedure involves
the WebCore::InspectorController trying to access the JS execution context
for a page that is being deleted. This patch amends the inspector so
that it does not try and access the execution context of the WebCore::Page
in the midst of deletion.

  • inspector/InspectorController.cpp: (WebCore::InspectorController::inspectedPageDestroyed): (WebCore::InspectorController::stopUserInitiatedProfiling):
14:37 Changeset [41157] by levin@chromium.org
  • 8 edits in trunk/WebCore

2009-02-23 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 24088: ThreadableLoaderClient::didFailWillSendRequestCheck isn't wired up completely for workers and could use a better name.
<https://bugs.webkit.org/show_bug.cgi?id=24088>

No observable change in behavior, so no test.

  • loader/DocumentThreadableLoader.cpp: (WebCore::DocumentThreadableLoader::create):
  • loader/ThreadableLoaderClient.h: (WebCore::ThreadableLoaderClient::didFailRedirectCheck):
  • loader/ThreadableLoaderClientWrapper.h: (WebCore::ThreadableLoaderClientWrapper::didFailRedirectCheck):
  • loader/WorkerThreadableLoader.cpp: (WebCore::workerContextDidFailRedirectCheck): (WebCore::WorkerThreadableLoader::MainThreadBridge::didFailRedirectCheck):
  • loader/WorkerThreadableLoader.h:
  • xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::didFinishLoading):
  • xml/XMLHttpRequest.h:
14:37 Changeset [41156] by levin@chromium.org
  • 7 edits in trunk

JavaScriptCore:

2009-02-23 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 24047: Need to simplify nested if's in WorkerRunLoop::runInMode
<https://bugs.webkit.org/show_bug.cgi?id=24047>

  • wtf/MessageQueue.h: (WTF::MessageQueue::infiniteTime): Allows for one to call waitForMessageFilteredWithTimeout and wait forever.

(WTF::MessageQueue::alwaysTruePredicate):
(WTF::MessageQueue::waitForMessage):
Made waitForMessage call waitForMessageFilteredWithTimeout, so that there is less
duplicate code.

(WTF::MessageQueue::waitForMessageFilteredWithTimeout):

  • wtf/ThreadingQt.cpp: (WTF::ThreadCondition::timedWait):
  • wtf/ThreadingWin.cpp: (WTF::ThreadCondition::timedWait): Made these two implementations consistent with the pthread and gtk implementations. Currently, the time calculations would overflow when passed large values.

WebCore:

2009-02-23 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

Bug 24047: Need to simplify nested if's in WorkerRunLoop::runInMode
<https://bugs.webkit.org/show_bug.cgi?id=24047>

Made a nested if inside of WorkerRunLoop::runInMode a lot simpler by
using only MessageQueue::waitForMessageFilteredWithTimeout instead
of three different MessageQueue methods.

No observable change in behavior, so no test.

  • dom/WorkerRunLoop.cpp: (WebCore::ModePredicate::operator()): Minor clean-up to able to pass a const ref point for ModePredicate into runInMode. (WebCore::WorkerRunLoop::runInMode):
  • dom/WorkerRunLoop.h:
14:08 Changeset [41155] by andersca@apple.com
  • 2 edits in trunk/WebKit/mac

2009-02-23 Anders Carlsson <andersca@apple.com>

Reviewed by Geoffrey Garen and Darin Adler.

WebKit side of <rdar://problem/6613151>.


Make sure to vm_deallocate all memory we get from MIG callbacks.


  • Plugins/Hosted/NetscapePluginHostProxy.mm: (DataDeallocator::DataDeallocator): (DataDeallocator::~DataDeallocator): Add a simple deallocator class.


(WKPCStatusText):
(WKPCLoadURL):
(WKPCBooleanAndDataReply):
(WKPCEvaluate):
(WKPCGetStringIdentifier):
(WKPCInvoke):
(WKPCInvokeDefault):
(WKPCConstruct):
(WKPCSetProperty):
Use the new deallocator class.

12:40 Changeset [41154] by hyatt@apple.com
  • 4 edits in trunk/WebCore

2009-02-23 David Hyatt <hyatt@apple.com>

In preparation for making layers for multicol objects (so that they can properly split child layers
into multiple columns), rename all of the "overflowOnly" and "overflowList" members and functions
of RenderLayer to use the term "normal flow" instead.

Reviewed by Cameron Zwarich

  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::RenderLayer): (WebCore::RenderLayer::~RenderLayer): (WebCore::RenderLayer::setHasVisibleContent): (WebCore::RenderLayer::enclosingCompositingLayer): (WebCore::RenderLayer::addChild): (WebCore::RenderLayer::removeChild): (WebCore::RenderLayer::paintLayer): (WebCore::RenderLayer::hitTestLayer): (WebCore::RenderLayer::dirtyNormalFlowList): (WebCore::RenderLayer::updateNormalFlowList): (WebCore::RenderLayer::collectLayers): (WebCore::RenderLayer::updateLayerListsIfNeeded): (WebCore::RenderLayer::shouldBeNormalFlowOnly): (WebCore::RenderLayer::styleChanged):
  • rendering/RenderLayer.h: (WebCore::RenderLayer::isNormalFlowOnly): (WebCore::RenderLayer::normalFlowList):
  • rendering/RenderTreeAsText.cpp: (WebCore::writeLayers):
12:16 Changeset [41153] by hyatt@apple.com
  • 4 edits
    4 adds in trunk

WebCore:

2009-02-23 David Hyatt <hyatt@apple.com>

Fix the stacking order for column rules in multi-column layout. Column rules should paint as part of the background of an element, just
after all other components of the background have been painted. This allows negative z-index children to still paint on top of the
column rules (rather than ending up above the background of the box but behind the column rules).

Reviewed by Eric Seidel

Added fast/multicol/column-rules-stacking.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::paintColumnRules): (WebCore::RenderBlock::paintColumnContents): (WebCore::RenderBlock::paintObject):
  • rendering/RenderBlock.h:

LayoutTests:

2009-02-23 David Hyatt <hyatt@apple.com>

Fix the stacking order for column rules in multi-column layout. Column rules should paint as part of the background of an element, just
after all other components of the background have been painted. This allows negative z-index children to still paint on top of the
column rules (rather than ending up above the background of the box but behind the column rules).

Reviewed by Eric Seidel

  • fast/multicol/column-rules-stacking.html: Added.
  • platform/mac/fast/multicol/column-rules-stacking-expected.checksum: Added.
  • platform/mac/fast/multicol/column-rules-stacking-expected.png: Added.
  • platform/mac/fast/multicol/column-rules-stacking-expected.txt: Added.
12:11 Changeset [41152] by treat@webkit.org
  • 3 edits in trunk/WebCore

2009-02-23 Adam Treat <adam.treat@torchmobile.com>

Reviewed by David Hyatt.

No testcases have been added or modified since this patch should not result in
a behavior change for ports that have layout tests enabled.

Currently, the implementation of GraphicsContext::drawLineForText amongst
the various ports differ in that some of them are honoring the context's
strokeStyle when drawing a text-decoration and some of them are not.
For instance, Apple's Mac port *does not* honor the context's strokeStyle(),
but the Cairo implementation does and has an explicit workaround that
sets the strokeStyle() temporarily.

This patch fixes so that all ports are consistent by explicitly making sure
to set the GraphicsContext strokeStyle to SolidStroke whenever
painting the text-decoration of an InlineFlowBox or InlineTextBox as these
should always use a solid stroke.

This patch addresses these bugs:
https://bugs.webkit.org/show_bug.cgi?id=19364
https://bugs.webkit.org/show_bug.cgi?id=15659

  • rendering/InlineFlowBox.cpp: (WebCore::InlineFlowBox::paintTextDecorations):
  • rendering/InlineTextBox.cpp: (WebCore::InlineTextBox::paintDecoration):
11:52 Changeset [41151] by treat@webkit.org
  • 4 edits in trunk/WebKit/qt

2009-02-23 Adam Treat <adam.treat@torchmobile.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24094
Make sure to empty the cache on exit to get rid of superfluous leak messages
for CachedResource's in order to keep the leak messages accurate. This
is analagous to what the Mac port is doing on application exit when all
WebView's have been closed.

  • Api/qwebpage.cpp: (QWebPagePrivate::_q_cleanupLeakMessages): (QWebPage::QWebPage):
  • Api/qwebpage.h:
  • Api/qwebpage_p.h:
11:35 Changeset [41150] by dglazkov@chromium.org
  • 3 edits
    3 adds in trunk/WebCore

2009-02-23 Scott Violet <sky@google.com>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24098
Bugs in ClipboardChromium

Fixes the following bugs in ClipboardChromium:

  • It's possible for the extension to be empty, resulting in a bad file name, for example, 'foo.' or just '.'.
  • We weren't restricting the size of the file to MAX_PATH.
  • We weren't removing characters that are invalid for file system names.
  • platform/chromium/ClipboardChromium.cpp: (WebCore::writeImageToDataObject):
  • platform/chromium/ClipboardChromium.h:
  • platform/chromium/ClipboardChromiumLinux.cpp: Added. (WebCore::ClipboardChromium::validateFileName):
  • platform/chromium/ClipboardChromiumMac.cpp: Added. (WebCore::ClipboardChromium::validateFileName):
  • platform/chromium/ClipboardChromiumWin.cpp: Added. (WebCore::isInvalidFileCharacter): (WebCore::ClipboardChromium::validateFileName):
09:48 Changeset [41149] by dglazkov@chromium.org
  • 2 edits in trunk/JavaScriptCore

2009-02-23 Jeremy Moskovich <jeremy@chromium.org>

Reviewed by Adam Roben.

https://bugs.webkit.org/show_bug.cgi?id=24096
PLATFORM(MAC)->PLATFORM(CF) since we want to use the CF functions in Chrome on OS X.

  • wtf/CurrentTime.cpp:
09:20 Changeset [41148] by andersca@apple.com
  • 5 edits in trunk/WebKit/mac

2009-02-23 Anders Carlsson <andersca@apple.com>

Reviewed by Darin Adler.

Fix <rdar://problem/6450656>.


  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::insertText): Add insert text which just calls the new WKPH function.


  • Plugins/Hosted/WebHostedNetscapePluginView.mm: (-[WebHostedNetscapePluginView inputContext]): Get the input context from the shared input panel.


(-[WebHostedNetscapePluginView keyDown:]):
Let the shared input panel have a go at the event first.


  • Plugins/Hosted/WebKitPluginHost.defs: Add new InsertText function.
08:19 Changeset [41147] by hausmann@webkit.org
  • 4 edits in trunk

2009-02-23 Thiago Macieira <thiago.macieira@nokia.com>

Reviewed by Simon Hausmann.

Fix the Copyright notices in a few files

07:08 Changeset [41146] by kov@webkit.org
  • 2 edits
    3 adds in trunk

2009-02-23 Gustavo Noronha Silva <gns@gnome.org>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=24011
KURL's setPort doesn't unset port if 0 is given

Adding regression tests for change done on KURL's setPort in
r41076. It is worth noting that Firefox doesn't render this test
correctly since it raises an exception when javascript tries to
change the iframe's contentWindow.location.port, though other
properties are modifiable.

  • fast/loader/location-port-expected.txt: Added.
  • fast/loader/location-port.html: Added.
  • fast/loader/resources/location-port-iframe.html: Added.

Also added bug URL to the WebCore entry for the r41076 change.

06:46 Changeset [41145] by ap@webkit.org
  • 708 edits in trunk/LayoutTests/dom/xhtml/level3/core

Rubber-stamped by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=15690
Bug in layout test framework (setUpPage vs. setUpPageStatus)

  • (Many JS files in dom/): (setUpPage): Replaced setUpPage with setUpPageStatus where appropriate, downstreaming a W3C fix.

Committing dom/xhtml/level3.

06:23 Changeset [41144] by ap@webkit.org
  • 723 edits in trunk/LayoutTests/dom/xhtml/level2

Rubber-stamped by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=15690
Bug in layout test framework (setUpPage vs. setUpPageStatus)

  • (Many JS files in dom/): (setUpPage): Replaced setUpPage with setUpPageStatus where appropriate, downstreaming a W3C fix.

Committing dom/xhtml/level2.

06:03 Changeset [41143] by ap@webkit.org
  • 238 edits in trunk/LayoutTests/dom/xhtml/level1/core

Rubber-stamped by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=15690
Bug in layout test framework (setUpPage vs. setUpPageStatus)

  • (Many JS files in dom/): (setUpPage): Replaced setUpPage with setUpPageStatus where appropriate, downstreaming a W3C fix.

Committing dom/xhtml/level1.

05:56 Changeset [41142] by ap@webkit.org
  • 721 edits in trunk/LayoutTests/dom/html/level2

Rubber-stamped by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=15690
Bug in layout test framework (setUpPage vs. setUpPageStatus)

  • (Many JS files in dom/): (setUpPage): Replaced setUpPage with setUpPageStatus where appropriate, downstreaming a W3C fix.

Committing dom/html/level2.

05:34 Changeset [41141] by ap@webkit.org
  • 238 edits in trunk/LayoutTests/dom/html/level1/core

Rubber-stamped by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=15690
Bug in layout test framework (setUpPage vs. setUpPageStatus)

  • (Many JS files in dom/): (setUpPage): Replaced setUpPage with setUpPageStatus where appropriate, downstreaming a W3C fix.

Committing dom/html/level1.

05:27 Changeset [41140] by ap@webkit.org
  • 66 edits in trunk/LayoutTests

Rubber-stamped by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=15690
Bug in layout test framework (setUpPage vs. setUpPageStatus)

  • (Many JS files in dom/): (setUpPage): Replaced setUpPage with setUpPageStatus where appropriate, downstreaming a W3C fix.

First part: dom/svg.

03:24 Changeset [41139] by christian@webkit.org
  • 8 edits in trunk

Allow retrieving the Soup session and modify the code to take
into account users changing features on it.

https://bugs.webkit.org/show_bug.cgi?id=22624
[SOUP][GTK] Need API to get SoupSession from WebKit.

03:15 Changeset [41138] by christian@webkit.org
  • 9 edits in trunk

Remove CURL support in the Gtk port, the HTTP backend is SOUP now.

https://bugs.webkit.org/show_bug.cgi?id=22624
[SOUP][GTK] Need API to get SoupSession from WebKit.

02:00 Changeset [41137] by jmalonzo@webkit.org
  • 3 edits in trunk/WebKit/gtk

2009-02-23 Jan Michael Alonzo <jmalonzo@webkit.org>

Reviewed by Mark Rowe.

[GTK] Back / Forward history menus are flipped
https://bugs.webkit.org/show_bug.cgi?id=22694

Don't call g_list_reverse when returning the back or forward list.

  • tests/main.c: (test_webkit_web_back_forward_list_order): (main):
  • webkit/webkitwebbackforwardlist.cpp: (_WebKitWebBackForwardListPrivate::webkit_web_back_forward_list_get_forward_list_with_limit): (_WebKitWebBackForwardListPrivate::webkit_web_back_forward_list_get_back_list_with_limit):
01:44 Changeset [41136] by ap@webkit.org
  • 6 edits
    2 adds in trunk

Reviewed by Mark Rowe.

Test: fast/dom/empty-hash-and-search.html

https://bugs.webkit.org/show_bug.cgi?id=21147
hash property returns incorrect value for links w/o hash

Make hash() and search() behavior for empty and missing parts match IE and Firefox.

  • platform/KURL.cpp: (WebCore::KURL::query): Changed to return query without '?', as it is already done for ref(). (WebCore::KURL::prettyURL): Append the query with the question mark.
  • page/Location.cpp: (WebCore::Location::search): Return an empty string if query is empty or missing. (WebCore::Location::hash): Return an empty string for empty hashes, not only missing ones.
  • dom/WorkerLocation.cpp: (WebCore::WorkerLocation::search): (WebCore::WorkerLocation::hash): Match document.location fixes above.
  • html/HTMLAnchorElement.cpp: (WebCore::HTMLAnchorElement::hash): (WebCore::HTMLAnchorElement::search): Return an empty string for empty and missing parts.
01:21 Changeset [41135] by ap@webkit.org
  • 3 edits
    2 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=20184
SELECT with no name generates invalid query string

Test: fast/forms/select-no-name.html

  • html/HTMLSelectElement.cpp: (WebCore::HTMLSelectElement::appendFormData): Added a check for empty name.
01:10 Changeset [41134] by ap@webkit.org
  • 2 edits in trunk/WebCore

Apply review comments for the previous check-in (forgot to save the file, oops).

  • dom/ContainerNode.cpp: (WebCore::ContainerNode::removeChildren):
01:03 Changeset [41133] by ap@webkit.org
  • 3 edits
    3 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=15707
Crash when manipulating document from within an iframe onload function

Test: fast/dom/onload-open.html

  • dom/ContainerNode.cpp: (WebCore::ContainerNode::removeChildren): Protect the container and the current node, because anything can happen when dispatching events.
00:58 Changeset [41132] by mrowe@apple.com
  • 2 edits in trunk/WebKit/mac

Fix the build after r41126.

  • Plugins/Hosted/NetscapePluginInstanceProxy.mm:

(WebKit::NetscapePluginInstanceProxy::invoke):
(WebKit::NetscapePluginInstanceProxy::invokeDefault):
(WebKit::NetscapePluginInstanceProxy::construct):

00:52 Changeset [41131] by ap@webkit.org
  • 3 edits
    2 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=18970
Numerically named input fields cause document.forms loop problems

Test: fast/forms/numeric-input-name.html

  • bindings/scripts/CodeGeneratorJS.pm: Try index getter before name getter, even if the latter overrides properties.
00:44 Changeset [41130] by ap@webkit.org
  • 11 edits in trunk/WebCore

Reviewed by Sam Weinig.

https://bugs.webkit.org/show_bug.cgi?id=24059
Tokenizer::write() return value is never used

  • dom/Tokenizer.h:
  • dom/XMLTokenizer.cpp: (WebCore::XMLTokenizer::write):
  • dom/XMLTokenizer.h:
  • html/HTMLTokenizer.cpp: (WebCore::HTMLTokenizer::write):
  • html/HTMLTokenizer.h:
  • loader/FTPDirectoryDocument.cpp: (WebCore::FTPDirectoryTokenizer::write):
  • loader/ImageDocument.cpp: (WebCore::ImageTokenizer::write):
  • loader/MediaDocument.cpp: (WebCore::MediaTokenizer::write):
  • loader/PluginDocument.cpp: (WebCore::PluginTokenizer::write):
  • loader/TextDocument.cpp: (WebCore::TextTokenizer::write): Made write() return void, not bool.

02/22/09:

22:22 Changeset [41129] by mitz@apple.com
  • 2 edits in trunk/WebKit/mac

Reviewed by Darin Adler.

  • fix <rdar://problem/5966123> REGRESSION (r30741): Generic Sun Applet loading logo appears half off screen
  • WebCoreSupport/WebFrameLoaderClient.mm: Correct a copy & paste error in r30741, and assign the height value, rather than the width, to the "height" parameter.
16:53 Changeset [41128] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

2009-02-22 Geoffrey Garen <ggaren@apple.com>

Build fix?

  • GNUmakefile.am:
16:34 Changeset [41127] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

2009-02-22 Geoffrey Garen <ggaren@apple.com>

Build fix.

  • GNUmakefile.am:
15:26 Changeset [41126] by ggaren@apple.com
  • 40 edits
    2 copies in trunk

JavaScriptCore:

2009-02-22 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig.

Next step in splitting JIT functionality out of the Interpreter class:
Created a JITStubs class and renamed Interpreter::cti_* to JITStubs::cti_*.


Also, moved timeout checking into its own class, located in JSGlobalData,
so both the Interpreter and the JIT could have access to it.

  • JavaScriptCore.exp:
  • JavaScriptCore.pri:
  • JavaScriptCore.scons:
  • JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • JavaScriptCoreSources.bkl:
  • interpreter/CallFrame.h:
  • interpreter/Interpreter.cpp: (JSC::Interpreter::Interpreter): (JSC::Interpreter::privateExecute):
  • interpreter/Interpreter.h:
  • interpreter/Register.h:
  • jit/JIT.cpp: (JSC::): (JSC::JIT::emitTimeoutCheck): (JSC::JIT::privateCompileMainPass): (JSC::JIT::privateCompileSlowCases): (JSC::JIT::privateCompile): (JSC::JIT::privateCompileCTIMachineTrampolines):
  • jit/JIT.h:
  • jit/JITArithmetic.cpp: (JSC::JIT::compileFastArithSlow_op_lshift): (JSC::JIT::compileFastArithSlow_op_rshift): (JSC::JIT::compileFastArithSlow_op_bitand): (JSC::JIT::compileFastArithSlow_op_mod): (JSC::JIT::compileFastArith_op_mod): (JSC::JIT::compileFastArithSlow_op_post_inc): (JSC::JIT::compileFastArithSlow_op_post_dec): (JSC::JIT::compileFastArithSlow_op_pre_inc): (JSC::JIT::compileFastArithSlow_op_pre_dec): (JSC::JIT::compileFastArith_op_add): (JSC::JIT::compileFastArith_op_mul): (JSC::JIT::compileFastArith_op_sub): (JSC::JIT::compileBinaryArithOpSlowCase): (JSC::JIT::compileFastArithSlow_op_add): (JSC::JIT::compileFastArithSlow_op_mul):
  • jit/JITCall.cpp: (JSC::JIT::compileOpCall): (JSC::JIT::compileOpCallSlowCase):
  • jit/JITPropertyAccess.cpp: (JSC::JIT::compileGetByIdHotPath): (JSC::JIT::compilePutByIdHotPath): (JSC::JIT::compileGetByIdSlowCase): (JSC::JIT::compilePutByIdSlowCase): (JSC::JIT::privateCompilePutByIdTransition): (JSC::JIT::patchGetByIdSelf): (JSC::JIT::patchPutByIdReplace): (JSC::JIT::privateCompilePatchGetArrayLength): (JSC::JIT::privateCompileGetByIdSelf): (JSC::JIT::privateCompileGetByIdProto): (JSC::JIT::privateCompileGetByIdChain): (JSC::JIT::privateCompilePutByIdReplace):
  • jit/JITStubs.cpp: (JSC::JITStubs::tryCachePutByID): (JSC::JITStubs::tryCacheGetByID): (JSC::JITStubs::cti_op_convert_this): (JSC::JITStubs::cti_op_end): (JSC::JITStubs::cti_op_add): (JSC::JITStubs::cti_op_pre_inc): (JSC::JITStubs::cti_timeout_check): (JSC::JITStubs::cti_register_file_check): (JSC::JITStubs::cti_op_loop_if_less): (JSC::JITStubs::cti_op_loop_if_lesseq): (JSC::JITStubs::cti_op_new_object): (JSC::JITStubs::cti_op_put_by_id_generic): (JSC::JITStubs::cti_op_get_by_id_generic): (JSC::JITStubs::cti_op_put_by_id): (JSC::JITStubs::cti_op_put_by_id_second): (JSC::JITStubs::cti_op_put_by_id_fail): (JSC::JITStubs::cti_op_get_by_id): (JSC::JITStubs::cti_op_get_by_id_second): (JSC::JITStubs::cti_op_get_by_id_self_fail): (JSC::JITStubs::cti_op_get_by_id_proto_list): (JSC::JITStubs::cti_op_get_by_id_proto_list_full): (JSC::JITStubs::cti_op_get_by_id_proto_fail): (JSC::JITStubs::cti_op_get_by_id_array_fail): (JSC::JITStubs::cti_op_get_by_id_string_fail): (JSC::JITStubs::cti_op_instanceof): (JSC::JITStubs::cti_op_del_by_id): (JSC::JITStubs::cti_op_mul): (JSC::JITStubs::cti_op_new_func): (JSC::JITStubs::cti_op_call_JSFunction): (JSC::JITStubs::cti_op_call_arityCheck): (JSC::JITStubs::cti_vm_dontLazyLinkCall): (JSC::JITStubs::cti_vm_lazyLinkCall): (JSC::JITStubs::cti_op_push_activation): (JSC::JITStubs::cti_op_call_NotJSFunction): (JSC::JITStubs::cti_op_create_arguments): (JSC::JITStubs::cti_op_create_arguments_no_params): (JSC::JITStubs::cti_op_tear_off_activation): (JSC::JITStubs::cti_op_tear_off_arguments): (JSC::JITStubs::cti_op_profile_will_call): (JSC::JITStubs::cti_op_profile_did_call): (JSC::JITStubs::cti_op_ret_scopeChain): (JSC::JITStubs::cti_op_new_array): (JSC::JITStubs::cti_op_resolve): (JSC::JITStubs::cti_op_construct_JSConstruct): (JSC::JITStubs::cti_op_construct_NotJSConstruct): (JSC::JITStubs::cti_op_get_by_val): (JSC::JITStubs::cti_op_get_by_val_byte_array): (JSC::JITStubs::cti_op_resolve_func): (JSC::JITStubs::cti_op_sub): (JSC::JITStubs::cti_op_put_by_val): (JSC::JITStubs::cti_op_put_by_val_array): (JSC::JITStubs::cti_op_put_by_val_byte_array): (JSC::JITStubs::cti_op_lesseq): (JSC::JITStubs::cti_op_loop_if_true): (JSC::JITStubs::cti_op_negate): (JSC::JITStubs::cti_op_resolve_base): (JSC::JITStubs::cti_op_resolve_skip): (JSC::JITStubs::cti_op_resolve_global): (JSC::JITStubs::cti_op_div): (JSC::JITStubs::cti_op_pre_dec): (JSC::JITStubs::cti_op_jless): (JSC::JITStubs::cti_op_not): (JSC::JITStubs::cti_op_jtrue): (JSC::JITStubs::cti_op_post_inc): (JSC::JITStubs::cti_op_eq): (JSC::JITStubs::cti_op_lshift): (JSC::JITStubs::cti_op_bitand): (JSC::JITStubs::cti_op_rshift): (JSC::JITStubs::cti_op_bitnot): (JSC::JITStubs::cti_op_resolve_with_base): (JSC::JITStubs::cti_op_new_func_exp): (JSC::JITStubs::cti_op_mod): (JSC::JITStubs::cti_op_less): (JSC::JITStubs::cti_op_neq): (JSC::JITStubs::cti_op_post_dec): (JSC::JITStubs::cti_op_urshift): (JSC::JITStubs::cti_op_bitxor): (JSC::JITStubs::cti_op_new_regexp): (JSC::JITStubs::cti_op_bitor): (JSC::JITStubs::cti_op_call_eval): (JSC::JITStubs::cti_op_throw): (JSC::JITStubs::cti_op_get_pnames): (JSC::JITStubs::cti_op_next_pname): (JSC::JITStubs::cti_op_push_scope): (JSC::JITStubs::cti_op_pop_scope): (JSC::JITStubs::cti_op_typeof): (JSC::JITStubs::cti_op_is_undefined): (JSC::JITStubs::cti_op_is_boolean): (JSC::JITStubs::cti_op_is_number): (JSC::JITStubs::cti_op_is_string): (JSC::JITStubs::cti_op_is_object): (JSC::JITStubs::cti_op_is_function): (JSC::JITStubs::cti_op_stricteq): (JSC::JITStubs::cti_op_nstricteq): (JSC::JITStubs::cti_op_to_jsnumber): (JSC::JITStubs::cti_op_in): (JSC::JITStubs::cti_op_push_new_scope): (JSC::JITStubs::cti_op_jmp_scopes): (JSC::JITStubs::cti_op_put_by_index): (JSC::JITStubs::cti_op_switch_imm): (JSC::JITStubs::cti_op_switch_char): (JSC::JITStubs::cti_op_switch_string): (JSC::JITStubs::cti_op_del_by_val): (JSC::JITStubs::cti_op_put_getter): (JSC::JITStubs::cti_op_put_setter): (JSC::JITStubs::cti_op_new_error): (JSC::JITStubs::cti_op_debug): (JSC::JITStubs::cti_vm_throw):
  • jit/JITStubs.h: (JSC::):
  • runtime/JSFunction.h:
  • runtime/JSGlobalData.cpp: (JSC::JSGlobalData::JSGlobalData):
  • runtime/JSGlobalData.h:
  • runtime/JSGlobalObject.cpp:
  • runtime/JSGlobalObject.h:
  • runtime/TimeoutChecker.cpp: Copied from interpreter/Interpreter.cpp. (JSC::TimeoutChecker::TimeoutChecker): (JSC::TimeoutChecker::reset): (JSC::TimeoutChecker::didTimeOut):
  • runtime/TimeoutChecker.h: Copied from interpreter/Interpreter.h. (JSC::TimeoutChecker::setTimeoutInterval): (JSC::TimeoutChecker::ticksUntilNextCheck): (JSC::TimeoutChecker::start): (JSC::TimeoutChecker::stop):

WebCore:

2009-02-20 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig.

Updated for JavaScriptCore changes to timeout checking.

  • bindings/js/JSCustomPositionCallback.cpp: (WebCore::JSCustomPositionCallback::handleEvent):
  • bindings/js/JSCustomPositionErrorCallback.cpp: (WebCore::JSCustomPositionErrorCallback::handleEvent):
  • bindings/js/JSCustomSQLStatementCallback.cpp: (WebCore::JSCustomSQLStatementCallback::handleEvent):
  • bindings/js/JSCustomSQLStatementErrorCallback.cpp: (WebCore::JSCustomSQLStatementErrorCallback::handleEvent):
  • bindings/js/JSCustomSQLTransactionCallback.cpp: (WebCore::JSCustomSQLTransactionCallback::handleEvent):
  • bindings/js/JSCustomSQLTransactionErrorCallback.cpp: (WebCore::JSCustomSQLTransactionErrorCallback::handleEvent):
  • bindings/js/JSCustomVoidCallback.cpp: (WebCore::JSCustomVoidCallback::handleEvent):
  • bindings/js/JSCustomXPathNSResolver.cpp: (WebCore::JSCustomXPathNSResolver::lookupNamespaceURI):
  • bindings/js/JSDOMWindowBase.cpp: (WebCore::JSDOMWindowBase::JSDOMWindowBase): (WebCore::JSDOMWindowBase::commonJSGlobalData):
  • bindings/js/JSEventListener.cpp: (WebCore::JSAbstractEventListener::handleEvent):
  • bindings/js/ScheduledAction.cpp: (WebCore::ScheduledAction::executeFunctionInContext):
  • bindings/js/ScriptController.cpp: (WebCore::ScriptController::evaluate):
  • bindings/js/WorkerScriptController.cpp: (WebCore::WorkerScriptController::evaluate): (WebCore::WorkerScriptController::forbidExecution):
  • bindings/objc/WebScriptObject.mm: (-[WebScriptObject callWebScriptMethod:withArguments:]): (-[WebScriptObject evaluateWebScript:]):
  • bridge/NP_jsobject.cpp: (_NPN_InvokeDefault): (_NPN_Invoke): (_NPN_Evaluate): (_NPN_Construct):
  • bridge/jni/jni_jsobject.mm: (JavaJSObject::call): (JavaJSObject::eval):
02:37 Changeset [41125] by mrowe@apple.com
  • 2 edits in trunk/WebKitTools

Fix assertion failures in editing/pasteboard/paste-RTFD.html and editing/pasteboard/paste-TIFF.html in 64-bit.

Reviewed by Oliver Hunt.

  • DumpRenderTree/mac/DumpRenderTree.mm:

(swizzleAllMethods): When adding a new method to a class, use the implementation and type of the new method rather
than of an arbitrary existing method on the class.

Note: See TracTimeline for information about the timeline view.