⚠ Archived content — this site is no longer maintained.   Current WebKit documentation is at docs.webkit.org.

Timeline



Dec 20, 2007:

11:40 PM Changeset in webkit [28930] by ap@webkit.org
  • 2 edits in trunk/JavaScriptCore

Windows build fix

  • JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj: Copy npruntime_internal.h to WebKitBuild.
11:21 PM Changeset in webkit [28929] by eric@webkit.org
  • 5 edits in trunk/JavaScriptCore

2007-12-17 Eric Seidel <​eric@webkit.org>

Reviewed by mjs.

Split IfNode into IfNode and IfElseNode for speedup.
​http://bugs.webkit.org/show_bug.cgi?id=16470

SunSpider claims this is 1.003x as fast as before.
(This required running with --runs 15 to get consistent enough results to tell!)

  • kjs/grammar.y:
  • kjs/nodes.cpp: (KJS::IfNode::optimizeVariableAccess): (KJS::IfNode::execute): (KJS::IfNode::getDeclarations): (KJS::IfElseNode::optimizeVariableAccess): (KJS::IfElseNode::execute): (KJS::IfElseNode::getDeclarations):
  • kjs/nodes.h: (KJS::IfNode::): (KJS::IfElseNode::):
  • kjs/nodes2string.cpp: (KJS::IfNode::streamTo): (KJS::IfElseNode::streamTo):
10:57 PM Changeset in webkit [28928] by jhoneycutt@apple.com
  • 2 edits in trunk/WebCore

2007-12-20 Jon Honeycutt <​jhoneycutt@apple.com>

Reviewed by Kevin Decker.

<rdar://problem/5655797> REGRESSION(304-306A10) Safari does not find
shockwave plugin on freshly installed OS

Look for Director plugin in its native directory.

  • plugins/win/PluginDatabaseWin.cpp: (WebCore::PluginDatabaseWin::getPluginsInPaths): The Director plugin doesn't start with "np," so special case its filename. (WebCore::addMacromediaPluginPaths): Add Flash and "Shockwave 10" directories to the plugin paths. (WebCore::PluginDatabaseWin::defaultPluginPaths):
10:07 PM Changeset in webkit [28927] by Darin Adler
  • 2 edits in trunk/WebKit/mac

Reviewed by Oliver.

  • fix <rdar://problem/5658787> Selector -[WebView insertLineBreak:] is not implemented
  • WebView/WebView.mm: Added all selectors implemented by WebHTMLView to the list of selectors to forward here. The new ones are: changeBaseWritingDirection:, changeSpelling:, deleteToMark:, insertLineBreak:, moveParagraphBackwardAndModifySelection:, moveParagraphForwardAndModifySelection:, pageDownAndModifySelection:, pageUpAndModifySelection:, selectToMark:, setMark:, swapWithMark:, takeFindStringFromSelection:, toggleBaseWritingDirection:, and transpose:.
10:06 PM Changeset in webkit [28926] by Darin Adler
  • 2 edits in trunk/JavaScriptCore

Reviewed by Sam.

  • wtf/OwnPtr.h: (WTF::operator==): Added. (WTF::operator!=): Added.
7:56 PM Changeset in webkit [28925] by alp@webkit.org
  • 4 edits in trunk/WebCore

2007-12-20 Peter Kasting <​zerodpx@gmail.com>

Reviewed by Alp Toker.

​http://bugs.webkit.org/show_bug.cgi?id=16508
Fix regression in GIFImageDecoder.cpp: "Haeberli hack" led to wrongly
decoded transparent areas.

  • platform/image-decoders/gif/GIFImageDecoder.cpp: (WebCore::GIFImageDecoder::haveDecodedRow):
  • platform/image-decoders/gif/GIFImageDecoder.h:
  • platform/image-decoders/gif/GIFImageReader.cpp: (GIFImageReader::output_row):
6:55 PM Changeset in webkit [28924] by justin.garcia@apple.com
  • 10 edits
    6 adds in trunk

WebCore:

Reviewed by Oliver Hunt.

<rdar://problem/5543472> GoogleDocs: Safari hangs when creating a list from a particular selection


Still need to fix similar issues with the other operations that iterate
over selected paragraphs, like FormatBlock, Indent and Outdent (<rdar://problem/5658933>).

  • editing/IndentOutdentCommand.cpp: (WebCore::IndentOutdentCommand::indentRegion): Added a FIXME.
  • editing/IndentOutdentCommand.h: Removed an unused function.
  • editing/InsertListCommand.cpp: (WebCore::InsertListCommand::modifyRange): Renamed visibleStart to startOfSelection and visibleEnd to endOfSelection. Call the new selectionForParagraphIteration, which a) prevents operations like this one from being performed on a table that isn't fully selected (where the selection starts just before the table and ends inside it), and b) helps prevent paragraph iteration from going past the end of the selection. Call the new startOfNextParagraph, instead of using endOfParagraph(v).next(), since when v is in the last paragraph of the last cell of a table, that expression will return the position after the table, not the start of the next paragraph.
  • editing/htmlediting.cpp: (WebCore::enclosingListChild): Don't go above a table cell, so that list operations take effect inside the table cell where they are performed. (WebCore::selectionForParagraphIteration): Added, see above. (WebCore::indexForVisiblePosition): Moved from IndentOutdentCommand.cpp.
  • editing/htmlediting.h:
  • editing/visible_units.cpp: (WebCore::startOfNextParagraph): Added, see above.
  • editing/visible_units.h:
  • editing/TextIterator.h: (WebCore::TextIterator::exitNode): Added a FIXME.

LayoutTests:

Reviewed by Oliver Hunt.


<rdar://problem/5543472> GoogleDocs: Safari hangs when creating a list from a particular selection

  • editing/execCommand/5543472-1-expected.txt: Added.
  • editing/execCommand/5543472-1.html: Added.
  • editing/execCommand/5543472-2-expected.txt: Added.
  • editing/execCommand/5543472-2.html: Added.
  • editing/execCommand/5543472-3-expected.txt: Added.
  • editing/execCommand/5543472-3.html: Added.
6:18 PM Changeset in webkit [28923] by ggaren@apple.com
  • 6 edits in trunk/JavaScriptCore

Reviewed by Oliver Hunt.


AST optimization: Avoid NULL-checking ForNode's child nodes.


0.6% speedup on SunSpider.


This is a proof of concept patch that demonstrates how to optimize
grammar productions with optional components, like


for (optional; optional; optional) {

...

}


The parser emits NULL for an optional component that is not present.

Instead of checking for a NULL child at execution time, a node that
expects an optional component to be present more often than not checks
for a NULL child at construction time, and substitutes a viable
alternative node in its place.

(We'd like the parser to start emitting NULL a lot more once we teach
it to emit NULL for certain no-op productions like EmptyStatement and
VariableStatement, so, as a foundation, it's important for nodes with
NULL optional components to be fast.)

  • kjs/Parser.cpp: (KJS::Parser::didFinishParsing): Check for NULL SourceElements. Also, moved didFinishParsing into the .cpp file because adding a branch while it was in the header file caused a substantial and inexplicable performance regression. (Did I mention that GCC is crazy?)
  • kjs/grammar.y:
  • kjs/nodes.cpp: (KJS::BlockNode::BlockNode): Check for NULL SourceElements. (KJS::ForNode::optimizeVariableAccess): No need to check for NULL here. (KJS::ForNode::execute): No need to check for NULL here.
  • kjs/nodes.h: (KJS::ForNode::): Check for NULL SourceElements. Substitute a TrueNode because it's semantically harmless, and it evaluates to boolean in an efficient manner.
5:31 PM Changeset in webkit [28922] by mitz@apple.com
  • 2 edits in trunk/LayoutTests

Reviewed by John Sullivan.

  • reorganize skipped test list
  • platform/win/Skipped:
4:33 PM Changeset in webkit [28921] by alp@webkit.org
  • 8 edits in trunk

2007-12-20 Alp Toker <​alp@atoker.com>

Rubber-stamped by Maciej.

​http://bugs.webkit.org/show_bug.cgi?id=16542
[GTK] Text is missing with old Pango version

Back out commits r28880, r28876, r28865, r28864 which added Pango font
selection support. These changes caused a regression where no text was
displayed with older Pango versions.

4:24 PM Changeset in webkit [28920] by eric@webkit.org
  • 4 edits in trunk

2007-12-20 Eric Seidel <​eric@webkit.org>

Reviewed by mjs.

  • sunspider: increase default number of runs for more accurate numbers
3:59 PM Changeset in webkit [28919] by oliver@apple.com
  • 1 edit in trunk/JavaScriptCore/ChangeLog

ChangeLog entry corrections

3:46 PM Changeset in webkit [28918] by timothy@apple.com
  • 5 edits in trunk/WebCore

Reviewed by Sam Weinig.

<rdar://problem/5604897> Double-clicking on JavaScript console error
won't show the line where it occurred

Make the URLs clickable in the console, since double clicking is used for
word text selection. Makes the word "line" localizable. Messages that don't
have a URL will no longer print "undefined". Messages that have a line number
less than or equal to 0 will no longer be printed.

  • English.lproj/InspectorLocalizedStrings.js: Add "%s (line %d)".
  • page/inspector/ConsolePanel.js: Tweak how the DOM nodes are created and change the click handler to look for a link with the "console-message-url" class name.
  • page/inspector/SourcePanel.js: Call setupSourceFrameIfNeeded() if a line row is needed.
  • page/inspector/inspector.css: Update the look of the URL in the console.
3:43 PM Changeset in webkit [28917] by eric@webkit.org
  • 2 edits in trunk/JavaScriptCore

2007-12-20 Eric Seidel <​eric@webkit.org>

Reviewed by mjs.

Fix major Array regression introduced by 28899.

SunSpider claims this is at least 1.37x as fast as pre-regression. :)

  • kjs/array_instance.cpp: make Array's fast again!
3:39 PM Changeset in webkit [28916] by oliver@apple.com
  • 2 edits in trunk/JavaScriptCore

Slight logic reordering in JSImmediate::from(double)

Reviewed by Geoff.

This gives a 0.1% improvement in SunSpider.

3:30 PM Changeset in webkit [28915] by sfalken@apple.com
  • 1 copy in tags/Safari-523.15b

New tag.

3:30 PM Changeset in webkit [28914] by sfalken@apple.com
  • 2 edits in branches/Safari-3-branch/WebKit/win

Bump versions for submit

2:46 PM Changeset in webkit [28913] by eric@webkit.org
  • 2 edits in trunk/JavaScriptCore

2007-12-20 Eric Seidel <​eric@webkit.org>

Reviewed by Geoff, then re-rubber-stamped by Geoff after final search/replace and testing.

Small reworking of Date code for 4% speedup on Date tests (0.2% overall)
​http://bugs.webkit.org/show_bug.cgi?id=16537

Make msToYear human-readable
Make msToDayInMonth slightly more readable and avoid recalculating msToYear
Remove use of isInLeapYear to avoid calling msToYear
Remove dayInYear call by changing msToDayInMonth to dayInMonthFromDayInYear
Remove more duplicate calls to dayInYear and getUTCOffset for further speedup

  • kjs/DateMath.cpp: (KJS::daysFrom1970ToYear): (KJS::msToYear): (KJS::monthFromDayInYear): (KJS::checkMonth): (KJS::dayInMonthFromDayInYear): (KJS::dateToDayInYear): (KJS::getDSTOffsetSimple): (KJS::getDSTOffset): (KJS::gregorianDateTimeToMS): (KJS::msToGregorianDateTime):
2:39 PM Changeset in webkit [28912] by weinig@apple.com
  • 16 edits
    4 adds in trunk

WebCore:

Reviewed and landed by Sam Weinig.

​http://bugs.webkit.org/show_bug.cgi?id=15313
<rdar://problem/5514516>

The same-origin check was incorrect in two cases (both fixed in this
patch):

A) If both the source and the target have set their document.domain

to the same value, the protocol must also match in order for
access to be allowed. Without this requirement, the browser is
vulnerable to the following attack:

1) Suppose there is an HTTPS site (www.example.com) that sets

document.domain = "example.com".

2) A network attacker redirects the browser to ​http://www.example.com/

a) injects script to set document.domain = "example.com", and
b) opens a window to ​https://www.example.com/

3) Now the network attacker can inject script into the HTTPS page,

stealing cookies and issuing banking transactions.

B) If only one of the source and target has set document.domain, then

access should be denied. With this behavior, the browser is
vulnerable to the following attack:

1) Suppose ​http://foo.example.com/ opens an iframe to

​http://foo.example.com/frame.html that
a) sets document.domain = "example.com", and
b) opens an iframe to ​http://bar.example.com/
This is a common usage of document.domain for cross-domain
communication, see for example:

​http://www.collinjackson.com/research/papers/fp801-jackson.pdf

2) The inner-most iframe, which is from bar.example.com, sets

document.domain = "example.com".

3) Now the inner-most iframe can inject script into the middle

iframe (say via document.write). This bar.example.com script
now has access to the outer-most frame (from foo.example.com).

Both these changes cause WebKit to match the behavior of Firefox 2 and
IE6 in these cases. This patch includes regression tests for both
issues.

Internet Explorer 7 and Opera 9 are more strict in that they require
the port numbers to match when both pages have document.domain set.
Opera 9 allows access when only one page has set document.domain, but
this is a security vulnerability.

Tests: http/tests/security/cross-frame-access-child-explicit-domain.html

http/tests/security/cross-frame-access-parent-explicit-domain.html

  • bindings/js/kjs_window.cpp: (KJS::createWindow): (KJS::Window::allowsAccessFrom):
  • dom/Document.cpp: (WebCore::Document::domain): (WebCore::Document::setDomain): (WebCore::Document::initSecurityOrigin):
  • dom/Document.h: (WebCore::Document::securityOrigin):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::begin): (WebCore::FrameLoader::checkCallImplicitClose): (WebCore::FrameLoader::shouldAllowNavigation):
  • platform/SecurityOrigin.cpp: (WebCore::SecurityOrigin::setForURL): (WebCore::SecurityOrigin::createForFrame): (WebCore::SecurityOrigin::canAccess):
  • platform/SecurityOrigin.h: (WebCore::SecurityOrigin::domain):
  • storage/Database.cpp: (WebCore::Database::openDatabase): (WebCore::Database::Database): (WebCore::Database::securityOriginData):
  • storage/Database.h: (WebCore::Database::databaseDebugName):
  • storage/DatabaseTracker.cpp: (WebCore::DatabaseTracker::canEstablishDatabase):
  • storage/SQLTransaction.cpp: (WebCore::SQLTransaction::postflightAndCommit): (WebCore::SQLTransaction::cleanupAfterTransactionErrorCallback):

LayoutTests:

Reviewed and landed by Sam Weinig.

Update LayoutTests for ​http://bugs.webkit.org/show_bug.cgi?id=15313

  • http/tests/security/cross-frame-access-child-explicit-domain-expected.txt: Added.
  • http/tests/security/cross-frame-access-child-explicit-domain.html: Added.
  • http/tests/security/cross-frame-access-custom-expected.txt:
  • http/tests/security/cross-frame-access-parent-explicit-domain-expected.txt: Added.
  • http/tests/security/cross-frame-access-parent-explicit-domain.html: Added.
  • http/tests/security/cross-frame-access-port-explicit-domain-expected.txt:
  • http/tests/security/cross-frame-access-protocol-explicit-domain-expected.txt:
  • http/tests/security/cross-frame-access-protocol-explicit-domain.html:
2:17 PM Changeset in webkit [28911] by mrowe@apple.com
  • 14 edits
    2 adds in trunk

2007-12-20 Rodney Dawes <​dobey@wayofthemonkey.com>

Reviewed by Darin Adler.

Proxy includes of npruntime.h or npapi.h through npruntime_internal.h
Include stdio.h in npapi.h for the use of FILE with XP_UNIX defined
This is for building with X11, as some type and enum names conflict
with #define names in X11 headers.
​http://bugs.webkit.org/show_bug.cgi?id=15669

  • JavaScriptCore.xcodeproj/project.pbxproj:
  • bindings/NP_jsobject.h:
  • bindings/npapi.h:
  • bindings/npruntime.cpp:
  • bindings/npruntime_impl.h:
  • bindings/npruntime_priv.h:
  • bindings/npruntime_internal.h:
  • bindings/testbindings.cpp:
  • bindings/c/c_class.h:
  • bindings/c/c_runtime.h:
  • bindings/c/c_utility.h:

2007-12-20 Rodney Dawes <​dobey@wayofthemonkey.com>

Reviewed by Darin Adler.

Define XP_UNIX when building with plugins on X11.
Use new npruntime_internal.h instead of npruntime.h.
​http://bugs.webkit.org/show_bug.cgi?id=15669

  • ForwardingHeaders/bindings/npruntime_internal.h
  • WebCore.pro:
  • html/HTMLPlugInElement.h:
2:16 PM Changeset in webkit [28910] by kmccullough@apple.com
  • 10 edits
    1 add in trunk

WebKit/win:

Reviewed by Steve.

  • <rdar://5656485> Drosera: Win: Nightly does not reliably connect to WebKit.
  • Drosera and Safari need to use the same ProgIDs in order for CoCreateInstance to work properly. The most robust way to do this is for WebKit to dynamically publish those ProgIDs.
  • ForEachCoClass.cpp: Added. (setUseOpenSourceWebKit): (progIDForClass):
  • ForEachCoClass.h: Added functionality to publish ProgIDs.
  • WebKit.vcproj/WebKit.def: Exposed added functions for applications.
  • WebKit.vcproj/WebKit.vcproj: Added ForEachCoClass.cpp
  • WebKit.vcproj/WebKit_debug.def: Exposed added functions fori applications.

WebKitTools:

Reviewed by Steve.

  • <rdar://5656485> Drosera: Win: Nightly does not reliably connect to WebKit.
  • Drosera and Safari need to use the same ProgIDs in order for CoCreateInstance to work properly. The most robust way to do this is for WebKit to dynamically publish those ProgIDs.
  • Drosera/win/DebuggerClient.cpp: Use the published ProgIDs (DebuggerClient::createWebViewWithRequest):
  • Drosera/win/Drosera.cpp: Use the published ProgIDs and rename some statics. (Drosera::initUI): (Drosera::attach): (Drosera::attemptToCreateServerConnection):
  • Drosera/win/Drosera.vcproj/Drosera.vcproj: Fix a runtime failure.
  • Drosera/win/ServerConnection.cpp: Reformat the connection function and make use of the published ProgIDs. (ServerConnection::attemptToCreateServerConnection):
2:10 PM Changeset in webkit [28909] by Stephanie Lewis
  • 2 edits in trunk/WebKitTools/Scripts

revert script check-ins that I didn't mean to commit

2:05 PM Changeset in webkit [28908] by Stephanie Lewis
  • 6 edits
    2 adds in trunk

Any attribute name start with a unicode which like #xx00(x could be any hex number[0-9a-f]) will cause HTMLTokenizer parse error
Actually any unicode characters which great than 255 in attribute name will cause Webkit parse the attribute name wrong. So after comparing
the same scenario in IE 6/7, FireFox 2/3, Opera, we should treat those characters as part of attribute name.

  • html/HTMLTokenizer.cpp: (WebCore::HTMLTokenizer::parseEntity): Handle Unicode Entity Name by using ASCII version of findEntity. (WebCore::HTMLTokenizer::parseTag): Let type of ptr match type of cBuffer.
  • html/HTMLTokenizer.h: Change type of cBuffer from char to UChar.
1:57 PM Changeset in webkit [28907] by Darin Adler
  • 13 edits in trunk/JavaScriptCore

Same patch as last time with the test failures problem fixed.

  • kjs/function.cpp: (KJS::GlobalFuncImp::callAsFunction): Make sure to check the completion type from newExec to see if the execute raised an exception.
1:22 PM Changeset in webkit [28906] by sfalken@apple.com
  • 1 copy in tags/Safari-523.14b

New tag.

1:22 PM Changeset in webkit [28905] by sfalken@apple.com
  • 2 edits in branches/Safari-3-branch/WebKit/win

Bump versions for submit

1:21 PM Changeset in webkit [28904] by mrowe@apple.com
  • 2 edits in tags/Safari-5523.15.1

Versioning.

1:21 PM Changeset in webkit [28903] by mrowe@apple.com
  • 2 edits in tags/Safari-5523.15.1/WebCore

Merge r28057.

1:06 PM Changeset in webkit [28902] by kdecker@apple.com
  • 2 edits in trunk/WebKit/mac

Reviewed by Anders.


Fixed: <rdar://problem/5638288> REGRESSION: Flash movies show up in other tabs above the page (16373)

  • Plugins/WebBaseNetscapePluginView.mm: (-[WebBaseNetscapePluginView updateAndSetWindow]): QuickDraw plug-ins must manually be told when to stop writing to the window backing store. The problem was that change-set 28400 introduced an early return which prevented this necessary operation. The fix is to limit the scope of the early return to CG and GL plug-ins and to tweak the needsFocus check to prevent an exception from occurring in QuickDraw-based plug-ins.
1:01 PM Changeset in webkit [28901] by kmccullough@apple.com
  • 2 edits in trunk/WebKitTools
  • <rdar://problem/5658317> REGRESSION: 20+ leaks seen on buildbots.
  • Build bot and leak fix.
  • DumpRenderTree/LayoutTestController.cpp: (pathToLocalResourceCallback):
12:40 PM Changeset in webkit [28900] by mrowe@apple.com
  • 1 copy in tags/Safari-5523.15.1

New tag.

12:33 PM Changeset in webkit [28899] by Darin Adler
  • 14 edits in trunk/JavaScriptCore
  • roll out that last change -- it was causing test failures; I'll check it back in after fixing them
12:10 PM Changeset in webkit [28898] by mrowe@apple.com
  • 2 edits in tags/Safari-4523.12.2

Versioning.

12:10 PM Changeset in webkit [28897] by mrowe@apple.com
  • 2 edits in tags/Safari-4523.12.2/WebCore

Merge r28057.

12:08 PM Changeset in webkit [28896] by mrowe@apple.com
  • 2 edits in tags/Safari-5523.10.6

Versioning.

12:08 PM Changeset in webkit [28895] by mrowe@apple.com
  • 2 edits in tags/Safari-5523.10.6/WebCore

Merge r28057.

12:05 PM Changeset in webkit [28894] by mrowe@apple.com
  • 2 edits in branches/Safari-3-branch/WebCore

Merge r28057.

11:40 AM Changeset in webkit [28893] by mrowe@apple.com
  • 1 copy in tags/Safari-4523.12.2

New tag.

11:40 AM Changeset in webkit [28892] by mrowe@apple.com
  • 1 copy in tags/Safari-5523.10.6

New tag.

11:21 AM Changeset in webkit [28891] by eric@webkit.org
  • 10 edits in trunk

2007-12-20 Eric Seidel <​eric@webkit.org>

Reviewed by Nikolas Zimmermann.

WebKit claims to support SVG feature strings it shouldn't
​http://bugs.webkit.org/show_bug.cgi?id=15480

  • dom/DOMImplementation.cpp: (WebCore::isSVG10Feature): (WebCore::isSVG11Feature):
10:49 AM Changeset in webkit [28890] by sullivan@apple.com
  • 2 edits in trunk/WebCore

Reviewed by Oliver and Geoff


  • fix <rdar://problem/5536858> Yellow highlight for find results is sometimes shorter than white "hole" behind it
  • rendering/RenderText.cpp: (WebCore::RenderText::addLineBoxRects): respect useSelectionHeight in all cases; we were only respecting it in one of the two cases
10:28 AM Changeset in webkit [28889] by mitz@apple.com
  • 2 edits in trunk/WebCore

Reviewed by Darin Adler.

  • fix <rdar://problem/5656368> REGRESSION(3.0.4-ToT): Acid2 test fails to render anything
  • platform/win/ScrollViewWin.cpp: (WebCore::ScrollView::maximumScroll): Changed back to return the maximum scroll offsets even if scrolling is not allowed, because navigation and scrollTo should still work. (WebCore::ScrollView::wheelEvent): Added early return if scrolling is not allowed.
10:09 AM Changeset in webkit [28888] by alp@webkit.org
  • 4 edits
    2 adds in trunk

2007-12-17 Tony Chang <​idealisms@gmail.com>

Reviewed by Darin.

Test: fast/text/reset-drag-on-mouse-down.html

9:42 AM Changeset in webkit [28887] by Darin Adler
  • 13 edits in trunk/JavaScriptCore

Reviewed by Eric.

SuSpider shows 2.4% speedup.

Stop using completions in the execution engine.
Instead, the completion type and label target are both
stored in the ExecState.

  • API/JSContextRef.cpp: Removed unneeded include of "completion.h".
  • bindings/runtime_method.cpp: Removed unused execute function.
  • bindings/runtime_method.h: Ditto.
  • kjs/ExecState.h: Added completionType, breakOrContinueTarget, setCompletionType, setNormalCompletion, setBreakCompletion, setContinueCompletion, setReturnValueCompletion, setThrowCompletion, setInterruptedCompletion, m_completionType, and m_breakOrContinueTarget.
  • kjs/completion.h: Removed constructor and getter for target for break and continue from Completion. This class is now only used for the public API to Interpreter and such.
  • kjs/date_object.h: Removed unused execute function.
  • kjs/function.cpp: (KJS::FunctionImp::callAsFunction): Removed some unneeded exception processing. Updated to call the new execute function and to get the completion type from the ExecState. Merged in the execute function, which repeated some of the same logic and was called only from here. (KJS::GlobalFuncImp::callAsFunction): More of the same for eval.
  • kjs/function.h: Removed execute.
  • kjs/interpreter.cpp: (KJS::Interpreter::evaluate): Added code to convert the result of execut into a Completion.
  • kjs/nodes.cpp: (KJS::Node::setErrorCompletion): Renamed from createErrorCompletion. Now sets the completion type in the ExecState. (KJS::Node::rethrowException): Now sets the completion type in the ExecState. (KJS::StatementNode::hitStatement): Now sets the completion type in the ExecState. (KJS::VarStatementNode::execute): Updated to put completion type in the ExecState instead of a Completion object. (KJS::statementListExecute): Ditto. Also changed the for loop to use indices instead of iterators. (KJS::BlockNode::execute): Updated return type. (KJS::EmptyStatementNode::execute): Updated to put completion type in the ExecState instead of a Completion object. (KJS::ExprStatementNode::execute): Ditto. (KJS::IfNode::execute): Ditto. (KJS::DoWhileNode::execute): Ditto. Also streamlined the logic a little to make the normal case a little faster and moved the end outside the loop so that "break" can do a break. (KJS::WhileNode::execute): Ditto. (KJS::ForNode::execute): Ditto. (KJS::ForInNode::execute): Ditto. (KJS::ContinueNode::execute): Updated to put completion type in the ExecState instead of a Completion object. (KJS::BreakNode::execute): Ditto. (KJS::ReturnNode::execute): Ditto. (KJS::WithNode::execute): Ditto. (KJS::CaseClauseNode::executeStatements): Ditto. Also renamed to have execute in its name to reflect the fact that it's a member of the same family of functions. (KJS::CaseBlockNode::executeBlock): Ditto. (KJS::SwitchNode::execute): Ditto. (KJS::LabelNode::execute): Ditto. (KJS::ThrowNode::execute): Ditto. (KJS::TryNode::execute): Ditto. (KJS::ProgramNode::execute): Ditto. (KJS::EvalNode::execute): Ditto. (KJS::FunctionBodyNode::execute): Ditto. (KJS::FuncDeclNode::execute): Ditto.
  • kjs/nodes.h: Renamed setErrorCompletion to createErrorCompletion, made hitStatement protected, changed return value of execute to a JSValue, renamed evalStatements to executeStatements, and evalBlock to executeBlock.
  • kjs/number_object.h: Removed unused execute function.
5:17 AM QtWebKitTodo edited by zecke2@selfish.org
(diff)
2:31 AM Changeset in webkit [28886] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

Added Radar number.

  • kjs/nodes.cpp: (KJS::ProgramNode::processDeclarations):
1:41 AM Changeset in webkit [28885] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

Linux build fix: config.h has to come first.

  • kjs/error_object.cpp:
1:32 AM Changeset in webkit [28884] by ggaren@apple.com
  • 35 edits
    11 adds in trunk

JavaScriptCore:

Reviewed by Oliver Hunt.


Optimized global access to global variables, using a symbol table.


SunSpider reports a 1.5% overall speedup, a 6.2% speedup on 3d-morph,
and a whopping 33.1% speedup on bitops-bitwise-and.

  • API/JSCallbackObjectFunctions.h: Replaced calls to JSObject:: with calls to Base::, since JSObject is not always our base class. This was always a bug, but the bug is even more apparent after some of my changes.

(KJS::::staticFunctionGetter): Replaced use of getDirect with call to
getOwnPropertySlot. Global declarations are no longer stored in the
property map, so a call to getDirect is insufficient for finding
override properties.

  • API/testapi.c:
  • API/testapi.js: Added test for the getDirect change mentioned above.
  • kjs/ExecState.cpp:
  • kjs/ExecState.h: Dialed back the optimization to store a direct pointer to the localStorage buffer. One ExecState can grow the global object's localStorage without another ExecState's knowledge, so ExecState can't store a direct pointer to the localStorage buffer unless/until we invent a way to update all the relevant ExecStates.
  • kjs/JSGlobalObject.cpp: Inserted the symbol table into get and put operations. (KJS::JSGlobalObject::reset): Reset the symbol table and local storage, too. Also, clear the property map here, removing the need for a separate call.
  • kjs/JSVariableObject.cpp:
  • kjs/JSVariableObject.h: Added support for saving localStorage and the symbol table to the back/forward cache, and restoring them.
  • kjs/function.cpp: (KJS::GlobalFuncImp::callAsFunction): Renamed progNode to evalNode because it's an EvalNode, not a ProgramNode.
  • kjs/lookup.h: (KJS::cacheGlobalObject): Replaced put with faster putDirect, since that's how the rest of lookup.h works. putDirect is safe here because cacheGlobalObject is only used for objects whose names are not valid identifiers.
  • kjs/nodes.cpp: The good stuff!

(KJS::EvalNode::processDeclarations): Replaced hasProperty with
the new hasOwnProperty, which is slightly faster.

  • kjs/object.h: Nixed clearProperties because clear() does this job now.
  • kjs/property_map.cpp:
  • kjs/property_map.h: More back/forward cache support.


  • wtf/Vector.h: (WTF::::grow): Added fast non-branching grow function. I used it in an earlier version of this patch, even though it's not used anymore.

JavaScriptGlue:

Build fix.

  • ForwardingHeaders/wtf/VectorTraits.h: Added.

WebCore:

Reviewed by Oliver Hunt.

Build support:

  • ForwardingHeaders/kjs/SymbolTable.h: Added.
  • ForwardingHeaders/wtf/VectorTraits.h: Added.
  • bindings/js/JSDOMWindowCustom.cpp: (WebCore::JSDOMWindow::customGetOwnPropertySlot): Replaced use of getDirectLocation with getOwnPropertySlot. getDirectLocation is no longer valid, since global declarations are not stored in the property map.

(WebCore::JSDOMWindow::customPut): Replaced use of JSObject::put with
JSGlobalObject::put. JSObject::put is no longer valid, since global
declarations are not stored in the property map.

  • bindings/js/kjs_window.cpp: Replaced JSObject:: calls with Base:: calls, since JSObject is not our base class. This was always a bug, but the bug is even more apparent after some of my changes.

(KJS::Window::clear): Removed call to clearProperties because
JSGlobalObject::reset takes care of that now.

  • history/CachedPage.cpp:
  • history/CachedPage.h: Added support for saving a symbol table and localStorage to the page cache, and restoring it.

WebKit/mac:

Reviewed by Oliver Hunt.

Build fix.

  • ForwardingHeaders/kjs/SymbolTable.h: Added.
  • ForwardingHeaders/wtf/VectorTraits.h: Added.

LayoutTests:

Reviewed by Oliver Hunt.


Added some tests to verify some of the changes I made while optimizing
global access to global variables.

  • fast/dom/Window/resources/window-property-clearing-iframe0.html: Added.
  • fast/dom/Window/resources/window-property-clearing-iframe1.html: Added.
  • fast/dom/Window/window-property-clearing-expected.txt: Added.
  • fast/dom/Window/window-property-clearing.html: Added.
  • fast/dom/getter-on-window-object2-expected.txt: Added.
  • fast/dom/getter-on-window-object2.html: Added.

Checked in failing results for these const tests. The symbol table
optimization broke const. (We didn't know this before because our only
tests used global variables.)

  • fast/js/const-expected.txt:
  • fast/js/kde/const-expected.txt:
  • fast/js/resources/for-in-avoid-duplicates.js: Fixed a typo I noticed. Not related to this patch.
  • fast/dom/Window/window-property-shadowing.html: Changed this test to use "this" instead of "window". The fact that "window" worked before, despite an overriding / shadowing var declaration, was a bug.

Dec 19, 2007:

10:40 PM Changeset in webkit [28883] by mitz@apple.com
  • 2 edits in trunk/LayoutTests
  • remove two more tests that pass now due to changed font fallback
  • platform/win/Skipped:
10:29 PM BuildingGtk edited by chad.kittel@gmail.com
added Lunar-Linux's package information. (diff)
10:21 PM Changeset in webkit [28882] by mitz@apple.com
  • 7 edits in trunk

WebCore:

Reviewed by Darin Adler and Dave Hyatt.

  • Substitute the user's preferred standard font for an unknown primary font before falling back on the platform's last resort font

Cannot be tested in DumpRenderTree because it sets the standard font to
Times, which is also the hard-coded last resort font on Mac.

  • css/CSSFontSelector.cpp: (WebCore::CSSFontSelector::getFontData): Changed the early bail out condition to allow getting generic font families from settings even if there are not @font-face rules. Fixed a typo that mapped fantasy to cursive. Added a mapping from -webkit-standard to the standard family.
  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::CSSStyleSelector): Changed to always create a font selector. (WebCore::CSSRuleSet::addRulesFromSheet): (WebCore::CSSStyleSelector::applyProperty): Changed to always use the font selector.
  • css/CSSStyleSelector.h: (WebCore::CSSStyleSelector::fontSelector):
  • platform/graphics/FontCache.cpp: (WebCore::FontCache::getFontData): Try the user's preferred standard font before the platform's last resort font.

LayoutTests:

  • remove passing tests following the change to font fallback
  • platform/win/Skipped: Removed tests that failed only because they specified as primary font family a font family that was not available on Windows and therefore were rendered using the platform last resort font, Times New Roman. Now they are rendered using the 'standard' font family, which is set to Times on both Mac and Windows when running the tests.
8:24 PM Changeset in webkit [28881] by ddkilzer@apple.com
  • 3 edits in trunk/WebKitTools

Make svn-apply/svn-unapply work with patches from git-format-patch.

Reviewed by Darin.

  • Scripts/svn-apply: (patch): If 'Index:' can't be found in the text passed in, print it out with a warning and return early.
  • Scripts/svn-unapply: (patch): Ditto.
7:03 PM Changeset in webkit [28880] by alp@webkit.org
  • 2 edits in trunk/WebCore

2007-12-19 Sven Herzberg <​sven@imendio.com>

Reviewed by Alp Toker.

Scaled font destruction takes place in FontData::platformDestroy(). No
need to do it in FontPlatformData::~FontPlatformData().

Destroying platform data in platformDestroy() is a convention we
borrow from the Mac and Win ports.

  • platform/graphics/gtk/FontPlatformDataGtk.cpp:
5:12 PM Changeset in webkit [28879] by alice.liu@apple.com
  • 3 edits in trunk/WebCore

build fix (Windows)

Changed uint to unsigned int. uint caused Windows build breakage

  • page/Page.cpp: (WebCore::Page::markAllMatchesForText):
  • page/Page.h:
4:23 PM Changeset in webkit [28878] by alp@webkit.org
  • 6 edits in trunk

2007-12-19 Christian Dywan <​christian@twotoasts.de>

Reviewed by Alp Toker.

​http://bugs.webkit.org/show_bug.cgi?id=16222
[GTK] Implement inline search and highlighting of matching strings.

4:02 PM Changeset in webkit [28877] by ggaren@apple.com
  • 4 edits in trunk/WebCore

Reviewed by Sam Weinig, Dan Bernstein.

Tiger build fix: restored some graphics code still needed on Tiger.

  • platform/graphics/GraphicsTypes.h:
  • platform/graphics/cg/GraphicsContextCG.cpp:
  • platform/graphics/mac/GraphicsContextMac.mm: (WebCore::GraphicsContext::setCompositeOperation):
3:22 PM Changeset in webkit [28876] by alp@webkit.org
  • 2 edits in trunk/WebCore

2007-12-19 Alp Toker <​alp@atoker.com>

Build fix for Pango < 1.18.0 breakage introduced in r28864. Use Fc and
the Pango backend API in these cases.

  • platform/graphics/gtk/FontPlatformDataGtk.cpp: (WebCore::FontPlatformData::FontPlatformData): (WebCore::FontPlatformData::~FontPlatformData):
2:51 PM Changeset in webkit [28875] by alice.liu@apple.com
  • 7 edits
    2 adds in trunk

WebCore:

Reviewed by Darin.

Fixed <rdar://problem/5592485> Safari crashed trying to get a motorcycle insurance quote
on Geico.com WebCore::Document::inPageCache()

Calling Node::willRemove on the focusedNode would immediately tell the document to remove
the focused node, and trigger JS events. This means that the document is mutated while
the engine is trying to tell all child nodes that it's about to removed. To avoid
crashing, we need to hold off on mutating the document until node traversal is finished.

  • dom/ContainerNode.cpp: (WebCore::ContainerNode::removeChild): (WebCore::ContainerNode::removeChildren):
  • dom/Node.cpp:
  • dom/Node.h: (WebCore::Node::willRemove):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::clear):

LayoutTests:

Reviewed by Darin.

Fixed <rdar://problem/5592485> Safari crashed trying to get a motorcycle insurance quote
on Geico.com WebCore::Document::inPageCache()

  • fast/events/nested-event-remove-node-crash-expected.txt: Added.
  • fast/events/nested-event-remove-node-crash.html: Added.
1:57 PM Changeset in webkit [28874] by aboule@apple.com
  • 6 edits
    4 adds in trunk

LayoutTests:

Reviewed by Dan Bernstein.

Adding tests for canvas compositing modes to make sure the results don't change after:
<rdar://problem/5640059> GraphicsContext::setCompositeOperation should use CGContextSetBlendMode

The following compositing modes don't pass however that is covered by another bug:
source-in
source-out
destination-in
destination-atop
copy
<rdar://problem/5651783> Some canvas tag compositing modes don't render correctly

  • fast/canvas/canvas-composite.html: Added.
  • platform/mac/fast/canvas/canvas-composite-expected.checksum: Added.
  • platform/mac/fast/canvas/canvas-composite-expected.png: Added.
  • platform/mac/fast/canvas/canvas-composite-expected.txt: Added.

WebCore:

Reviewed by Dan Bernstein.

Test: fast/canvas/canvas-composite.html

Fix for:
<rdar://problem/5640059> GraphicsContext::setCompositeOperation should use CGContextSetBlendMode

This fix makes setCompositeOperation consistent across all platforms that use CG.

The following compositing modes don't pass however that is covered by another bug:
source-in
source-out
destination-in
destination-atop
copy
<rdar://problem/5651783> Some canvas tag compositing modes don't render correctly

  • platform/graphics/GraphicsTypes.h:
  • platform/graphics/cg/GraphicsContextCG.cpp: (WebCore::GraphicsContext::setCompositeOperation):
  • platform/graphics/mac/GraphicsContextMac.mm:
  • platform/win/GraphicsContextWin.cpp:
1:45 PM Changeset in webkit [28873] by mitz@apple.com
  • 3 edits
    4 adds in trunk

WebCore:

Reviewed by Dave Hyatt.

  • fix <rdar://problem/5650045> REGRESSION: major layout problems in svn r28754 WebKit on Vox.com page

Test: fast/dynamic/subtree-boundary-percent-height.html

  • rendering/RenderObject.cpp: (WebCore::objectIsRelayoutBoundary): Do not allow overflows with percent heights because sometimes they compute to 'auto'.

LayoutTests:

Reviewed by Dave Hyatt.

  • test for <rdar://problem/5650045> REGRESSION: major layout problems in svn r28754 WebKit on Vox.com page
  • fast/dynamic/subtree-boundary-percent-height.html: Added.
  • platform/mac/fast/dynamic/subtree-boundary-percent-height-expected.checksum: Added.
  • platform/mac/fast/dynamic/subtree-boundary-percent-height-expected.png: Added.
  • platform/mac/fast/dynamic/subtree-boundary-percent-height-expected.txt: Added.
1:21 PM Changeset in webkit [28872] by hyatt@apple.com
  • 1 edit in trunk/WebCore/page/Settings.cpp

Fix unsigned/enum problem on Win32. r=darin.

12:26 PM Changeset in webkit [28871] by Beth Dakin
  • 2 edits in trunk/WebCore

Oops! Typo!

12:15 PM Changeset in webkit [28870] by hyatt@apple.com
  • 1 edit in trunk/WebCore/page/Settings.cpp

Fix unsigned/enum problem on Win32. r=darin.

12:14 PM Changeset in webkit [28869] by hyatt@apple.com
  • 3 edits in trunk/WebCore/page

Fix unsigned/enum problem on Win32. r=darin.

12:11 PM Changeset in webkit [28868] by weinig@apple.com
  • 14 edits in trunk/WebCore

Reviewed by Maciej.

​http://bugs.webkit.org/show_bug.cgi?id=16511
Speed up ClassNodeList and NamedNodeList by using the caching mechanism employed by ChildNodeList.

  • dom/ChildNodeList.cpp: Use the caching NodeList constructor to turn on caching. (WebCore::ChildNodeList::ChildNodeList):
  • dom/ClassNodeList.cpp: (WebCore::ClassNodeList::ClassNodeList):
  • dom/ClassNodeList.h:

Move getElementsByName and getElementsByClassName to Node so they
can use easily employ the caching already used by ChildNodeLists. In the case of
getElementsByClassName, this reduces code duplication in Element as well

  • dom/Document.cpp:
  • dom/Document.h:

Move getElementsByClassName to Node.

  • dom/Element.cpp:
  • dom/Element.h:
  • dom/NameNodeList.cpp: Use the caching NodeList constructor to turn on caching. (WebCore::NameNodeList::NameNodeList): (WebCore::NameNodeList::item):
  • dom/NameNodeList.h:

Add maps of caches for ClassNodeLists and NameNodeList to NodeListsNodeData.

  • dom/Node.cpp: (WebCore::TagNodeList::TagNodeList): (WebCore::Node::Node): (WebCore::Node::~Node): (WebCore::Node::childNodes): (WebCore::Node::registerNodeList): (WebCore::Node::getElementsByName): (WebCore::Node::getElementsByClassName):
  • dom/Node.h: Make m_nodeLists an OwnPtr. Moved getElementsByName and getElementsByClassName here

Allow subclasses to choose whether they want to receive the notifications using a new bit.

  • dom/NodeList.cpp: (WebCore::NodeList::NodeList):
  • dom/NodeList.h: (WebCore::NodeList::needsNotifications):
11:24 AM Changeset in webkit [28867] by hyatt@apple.com
  • 19 edits in trunk/WebCore

Add support for GDI text on Windows.

11:13 AM Changeset in webkit [28866] by hyatt@apple.com
  • 5 edits in trunk/WebKit/win

Add fourth preference value for GDI text font smoothing to WebKit.

10:41 AM Changeset in webkit [28865] by alp@webkit.org
  • 2 edits in trunk

2007-12-19 Alp Toker <​alp@atoker.com>

Remove the cairo-ft pkg-config check. The GTK+ port doesn't use
cairo-ft directly following changes in r28864.

  • WebKit.pri:
10:00 AM Changeset in webkit [28864] by alp@webkit.org
  • 6 edits in trunk/WebCore

2007-12-19 Sven Herzberg <​sven@imendio.com>

Reviewed by Alp Toker.

Replace the fontconfig/freetype based font management with a pango
based one. Fixes:
​http://bugs.webkit.org/show_bug.cgi?id=15229

  • platform/gtk/FontDataGtk.cpp (FontData::platformDestroy()): updated the platform specific destroy code (FontData::containsCharacters()): implemented font coverage with pango
  • platform/gtk/FontPlatformData.h: replaced fontconfig specific members with pango-specific ones
  • platform/gtk/FontPlatformDataGtk.cpp: added static members for the FontPlatformData class (FontPlatformData::FontPlatformData()): implemented the font-matching with a PangoFontDescription instead of an FcPattern; initialize the scaled font by using the API for PangoCairoFont (FontPlatformData::init()): initialize the PangoFontMap and set up a hash table to translate the font family name into a font family (FontPlatformData::isFixedPitch()): implemented by querying the PangoFontFamily (FontPlatformData::operator==): compare the FontPlatformData by comparing the font pointers or the described fonts
  • platform/gtk/GlyphPageTreeNodeGtk.cpp (pango_font_get_glyph()): added a function to query a glyph from a PangoFont (GlyphPage::fill()): implemented the fill function with Pango instead of fontconfig/freetype
9:36 AM Changeset in webkit [28863] by kevino@webkit.org
  • 2 edits in trunk/WebKitTools

wx build process improvements, including added checks and outdated changes after the move to trunk.

Reviewed by Alp Toker

9:00 AM Changeset in webkit [28862] by alp@webkit.org
  • 4 edits in trunk/WebKit/gtk

2007-12-19 Alp Toker <​alp@atoker.com>

Reviewed by Holger Freyther.

Delete when Destroy functions are called to avoid leaks

This matches the Mac port.

  • WebCoreSupport/ChromeClientGtk.cpp:
  • WebCoreSupport/ContextMenuClientGtk.cpp:
  • WebCoreSupport/InspectorClientGtk.cpp:
8:55 AM Changeset in webkit [28861] by alp@webkit.org
  • 2 edits in trunk/WebCore

2007-12-19 Alp Toker <​alp@atoker.com>

Reviewed by Holger Freyther.

Improve graphics operator approximations

These changes match Cairo's own CG approximations.

  • platform/graphics/cairo/GraphicsContextCairo.cpp: (WebCore::toCairoOperator):
8:55 AM Changeset in webkit [28860] by alp@webkit.org
  • 2 edits in trunk/WebCore

2007-12-19 Alp Toker <​alp@atoker.com>

Reviewed by Holger Freyther.

Check the bounding box before doing a full hit test

  • platform/graphics/cairo/PathCairo.cpp: (WebCore::Path::contains):
1:10 AM Changeset in webkit [28859] by mrowe@apple.com
  • 3 edits
    1 add in trunk/JavaScriptCore

Build fix for non-Mac platforms. Move NodeInfo into its own header so that the YYTYPE declaration in grammar.h is able to declare members of that type.

12:41 AM Changeset in webkit [28858] by weinig@apple.com
  • 4 edits in trunk/WebCore

Reviewed by Geoff.

Fix for <rdar://problem/5646478>
REGRESSION: fast/events/event-view-toString fails on Leopard

This fixes an issue where we were incorrectly setting the lastInPrototypeChain
for the JSDOMWindow (the global object) before a call to setPrototype overwrote
it. This fixes it by passing the prototype up the constructor chain so that it
is set before any calls can be made.

  • bindings/js/kjs_window.cpp: (KJS::Window::Window):
  • bindings/js/kjs_window.h:
  • bindings/scripts/CodeGeneratorJS.pm:
12:13 AM Changeset in webkit [28857] by oliver@apple.com
  • 2 edits in trunk/JavaScriptCore

Make appendToVarDeclarationList static

RS=Weinig

Note: See TracTimeline for information about the timeline view.