WebKit-https.git
11 years ago2008-11-12 Tor Arne Vestbø <tavestbo@trolltech.com>
vestbo@webkit.org [Thu, 13 Nov 2008 15:20:43 +0000 (15:20 +0000)]
2008-11-12  Tor Arne Vestbø  <tavestbo@trolltech.com>

        Reviewed by Simon Hausmann.

        Inital implementation of PluginPackage and PluginView for Mac NPAPI support in WebCore

        Currently only used by QtWebKit. See also:

        https://bugs.webkit.org/show_bug.cgi?id=21427

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38371 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Ariya Hidayat <ariya.hidayat@trolltech.com>
vestbo@webkit.org [Thu, 13 Nov 2008 15:20:19 +0000 (15:20 +0000)]
2008-11-12  Ariya Hidayat  <ariya.hidayat@trolltech.com>

        Rubber-stamped by Tor Arne Vestbø.

        Mention the color role change (r38331) in the API doc.

        * Api/qwebpage.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38370 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-13 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Thu, 13 Nov 2008 12:53:35 +0000 (12:53 +0000)]
2008-11-13  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Cameron Zwarich.

        - remove immediate checks from native codegen for known non-immediate cases like "this"

        ~.5% speedup on v8 benchmarks

        In the future we can extend this model to remove all sorts of
        typechecks based on local type info or type inference.

        I also added an assertion to verify that all slow cases linked as
        many slow case jumps as the corresponding fast case generated, and
        fixed the pre-existing cases where this was not true.

        * VM/CTI.cpp:
        (JSC::CTI::emitJumpSlowCaseIfNotJSCell):
        (JSC::CTI::linkSlowCaseIfNotJSCell):
        (JSC::CTI::compileBinaryArithOp):
        (JSC::CTI::compileBinaryArithOpSlowCase):
        (JSC::CTI::privateCompileMainPass):
        (JSC::CTI::privateCompileSlowCases):
        * VM/CTI.h:
        * VM/CodeBlock.h:
        (JSC::CodeBlock::isKnownNotImmediate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38369 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-13 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Thu, 13 Nov 2008 11:26:38 +0000 (11:26 +0000)]
2008-11-13  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Maciej Stachowiak.

        Bug 21943: Avoid needless reads of temporary values in CTI code
        <https://bugs.webkit.org/show_bug.cgi?id=21943>

        If an opcode needs to load a virtual register and a previous opcode left
        the contents of that virtual register in a machine register, use the
        value in the machine register rather than getting it from memory.

        In order to perform this optimization, it is necessary to know the
        jump tagets in the CodeBlock. For temporaries, the only problematic
        jump targets are binary logical operators and the ternary conditional
        operator. However, if this optimization were to be extended to local
        variable registers as well, other jump targets would need to be
        included, like switch statement cases and the beginnings of catch
        blocks.

        This optimization also requires that the fast case and the slow case
        of an opcode use emitPutResult() on the same register, which was chosen
        to be eax, as that is the register into which we read the first operand
        of opcodes. In order to make this the case, we needed to add some mov
        instructions to the slow cases of some instructions.

        This optimizaton is not applied whenever compileBinaryArithOp() is used
        to compile an opcode, because different machine registers may be used to
        store the final result. It seems possible to rewrite the code generation
        in compileBinaryArithOp() to allow for this optimization.

        This optimization is also not applied when generating slow cases,
        because some fast cases overwrite the value of eax before jumping to the
        slow case. In the future, it may be possible to apply this optimization
        to slow cases as well, but it did not seem to be a speedup when testing
        an early version of this patch.

        This is a 1.0% speedup on SunSpider and a 6.3% speedup on the V8
        benchmark suite.

        * VM/CTI.cpp:
        (JSC::CTI::killLastResultRegister):
        (JSC::CTI::emitGetArg):
        (JSC::CTI::emitGetPutArg):
        (JSC::CTI::emitGetCTIParam):
        (JSC::CTI::emitGetFromCallFrameHeader):
        (JSC::CTI::emitPutResult):
        (JSC::CTI::emitCTICall):
        (JSC::CTI::CTI):
        (JSC::CTI::compileOpCall):
        (JSC::CTI::compileOpStrictEq):
        (JSC::CTI::emitSlowScriptCheck):
        (JSC::CTI::compileBinaryArithOp):
        (JSC::CTI::privateCompileMainPass):
        (JSC::CTI::privateCompileSlowCases):
        (JSC::CTI::privateCompileGetByIdProto):
        (JSC::CTI::privateCompilePatchGetArrayLength):
        * VM/CTI.h:
        * VM/CodeBlock.h:
        (JSC::CodeBlock::isTemporaryRegisterIndex):
        * bytecompiler/CodeGenerator.cpp:
        (JSC::CodeGenerator::emitLabel):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38368 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Dmitry Titov <dimich@chromium.org>
ap@webkit.org [Thu, 13 Nov 2008 07:35:50 +0000 (07:35 +0000)]
2008-11-12  Dmitry Titov  <dimich@chromium.org>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=22216

        * Scripts/build-webkit:
        Addes ENABLE_WORKERS to build-webkit so it's possible to build with workers enabled.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38367 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Darin Adler.
ap@webkit.org [Thu, 13 Nov 2008 07:32:43 +0000 (07:32 +0000)]
    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=22200
        Crash when close event is dispatched to MessagePort due to document destruction

        Test: fast/events/message-port-onclose.html

        * dom/MessagePort.cpp: (WebCore::MessagePort::contextDestroyed):
        Don't dispatch close event, even though HTML5 currently says that we should. It is not
        right for GC to have observable effects.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38366 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Darin Adler.
ap@webkit.org [Thu, 13 Nov 2008 07:27:12 +0000 (07:27 +0000)]
    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=22203
        Implement Worker messaging

        No test cases included, because this functionality is disabled by default.

        The implementation is known to still have many race condition, but works quite well for
        testing.

        * WebCore.xcodeproj/project.pbxproj: Added WorkerTask.{h,cpp}.

        * bindings/js/JSDOMBinding.cpp:
        (WebCore::markActiveObjectsForContext): Re-worded comments a little.
        (WebCore::markCrossHeapDependentObjectsForContext): Existing cross-heap GC protocol was
        incorrect, changed it to a much simpler (but still incorrect) version.

        * dom/WorkerTask.cpp: Added.
        (WebCore::WorkerTask::~WorkerTask):
        * dom/WorkerTask.h: Added.
        Tasks posted to workers implement this new interface.

        * bindings/js/JSDedicatedWorkerCustom.cpp:
        (WebCore::JSDedicatedWorker::mark):
        (WebCore::JSDedicatedWorker::connect):
        * dom/DedicatedWorker.h:
        * dom/DedicatedWorker.idl:
        Auto-generate event listener attributes. Renamed startConversation() to connect(), tracking
        WHATWG discussions.

        * dom/DedicatedWorker.cpp:
        (WebCore::WorkerConnectTask::WorkerConnectTask): A task that performs worker-side connect()
        operations.
        (WebCore::DedicatedWorker::DedicatedWorker): Initialize WorkerThread pointer.
        (WebCore::DedicatedWorker::connect): Connect() creates a pair of entangled ports, and posts
        one to worker. Since message port registration in ScriptExecutionContext is not thread safe,
        this port starts with a null context pointer.
        (WebCore::DedicatedWorker::notifyFinished): Since Worker methods should work immediately
        after creation, we have to queue tasks until after a WorkerThread object is created. Then we
        forward all queued tasks to its queue.

        * dom/EventTarget.cpp: (WebCore::EventTarget::toWorkerContext):
        * dom/EventTarget.h:
        * bindings/js/JSEventTarget.cpp: (WebCore::toJS):
        Added cases for WorkerContext, which is now an EventTarget, too.

        * bindings/js/JSWorkerContext.h: Added JSWorkerContext::put() to make onconnect settable.
        * bindings/js/JSWorkerContext.cpp:
        (WebCore::JSWorkerContext::mark): Mark event listeners.
        (WebCore::JSWorkerContext::createPrototype): Fixed a typo, use the right StructureID.
        (WebCore::JSWorkerContext::put): Implemented.
        (WebCore::jsWorkerContextPrototypeFunctionAddEventListener): Added an EventTarget implementation.
        (WebCore::jsWorkerContextPrototypeFunctionRemoveEventListener): Ditto.
        (WebCore::jsWorkerContextPrototypeFunctionDispatchEvent): Ditto.
        (WebCore::jsWorkerContextOnconnect): Added.
        (WebCore::setJSWorkerContextOnconnect): Added.

        * bindings/js/WorkerScriptController.cpp: (WebCore::WorkerScriptController::evaluate):
        Made it actually work by adding necessary setup.

        * bindings/js/WorkerScriptController.h: (WebCore::WorkerScriptController::initScriptIfNeeded):
        Check the right variable - it is the wrapper that may not be initialized yet.

        * dom/ActiveDOMObject.cpp:
        (WebCore::ActiveDOMObject::ActiveDOMObject):
        (WebCore::ActiveDOMObject::~ActiveDOMObject):
        Assert being called from the correct thread, as active DOM object tracking is not thread safe.

        * dom/ScriptExecutionContext.h: Added a Task interface and a postTask() method, to be used
        for asynchronously executing tasks in context's thread.

        * dom/ScriptExecutionContext.cpp:
        (WebCore::ProcessMessagesSoonTask): Changed from a Timer to a Task
        (WebCore::ScriptExecutionContext::ScriptExecutionContext): Removed m_firedMessagePortTimer.
        It was an optimization that couldn't be easily preserved without introducing race conditions
        in multithreading case.
        (WebCore::ScriptExecutionContext::processMessagePortMessagesSoon): Use postTask().
        (WebCore::ScriptExecutionContext::dispatchMessagePortEvents): Added a comment explaining
        why it's OK to not ref() ports in a frozen copy.
        (WebCore::ScriptExecutionContext::createdMessagePort): Assert that we're not being called
        from a wrong thread.
        (WebCore::ScriptExecutionContext::destroyedMessagePort): Ditto.
        (WebCore::ScriptExecutionContextTaskTimer): Part of
        ScriptExecutionContext::Task implementation - use Timer if posting from main thread to main
        thread.
        (WebCore::ScriptExecutionContextTaskWorkerTask): Another part - use WorkerTask if posting
        to a worker.
        (WebCore::PerformTaskContext::PerformTaskContext): Finally, use callOnMainThread() if posting
        to main thread from a secondary one.
        (WebCore::performTask): A helper function for callOnMainThread().
        (WebCore::ScriptExecutionContext::postTask): Use one of the above implementations.

        * dom/MessagePort.h: Fixed message queue to keep EventData pointers - otherwise, we would
        ref/deref EventData::message from different threads, which is not allowed.

        * dom/MessagePort.cpp:
        (WebCore::MessagePortCloseEventTask): Use a task instead of a timer to work across threads.
        (WebCore::MessagePort::EventData::create): Updated for EventData being refcountable now.
        (WebCore::MessagePort::EventData::EventData): Ditto.
        (WebCore::MessagePort::MessagePort): ScriptExecutionContext is now allowed to be null at
        first, because we need to create ports for posting to other threads, and it is not possible
        to register in a context from another thread.
        (WebCore::MessagePort::clone): Always create ports with null contexts - it is now message
        receiver's job to set the context.
        (WebCore::MessagePort::postMessage): Enable posting to ports that are not attached to any
        context yet.
        (WebCore::MessagePort::startConversation): Ditto. Data port is always posted unattached.
        (WebCore::MessagePort::contextDestroyed): Assert that we had a context.
        (WebCore::MessagePort::attachToContext): Called when receiving a data port to register in
        context.
        (WebCore::MessagePort::scriptExecutionContext): Moved from header, as the function is virtual.
        (WebCore::MessagePort::dispatchMessages): Attach data port to receiving context. Use postTask().
        (WebCore::MessagePort::queueCloseEvent): Use postTask().
        (WebCore::MessagePort::hasPendingActivity): Reworded comment a little. As mentioned above,
        MessagePort cross-heap GC is still quite wrong.

        * dom/WorkerContext.h: Made WorkerContext an event target, added onconnect attribute.
        * dom/WorkerContext.cpp: Keep a pointer to WorkerThread. It is only used for debug assertions
        now, but there is no harm in tracking it in release builds, too.

        * dom/WorkerThread.cpp:
        (WebCore::WorkerThread::create): WorkerThread is refcountable, construct with create().
        (WebCore::WorkerThread::workerThread): Implemented a message loop.
        * dom/WorkerThread.h:
        (WebCore::WorkerThread::threadID): Also only used for assertions.
        (WebCore::WorkerThread::messageQueue): Return a reference to queue, so clients can post to it.

        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::postMessage): MessagePort::clone() no longer takes a context, as it
        always sets it to null.
        (WebCore::DOMWindow::postMessageTimerFired): Attach data port to receiving context.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38365 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Dirk Schulze <vbs85@gmx.de>
krit@webkit.org [Thu, 13 Nov 2008 07:17:53 +0000 (07:17 +0000)]
2008-11-12  Dirk Schulze  <vbs85@gmx.de>

WebCore:

        Reviewed Darin Adler.

        use strokeColor.alpha() instead of fillColor.alpha()
        on GraphicsContext::strokePath()

        [Canvas] Transparent fillStyle stops line rendering
        https://bugs.webkit.org/show_bug.cgi?id=21888

        Test: fast/canvas/canvas-stroke-empty-fill.html

        * ChangeLog:
        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::GraphicsContext::strokePath):

LayoutTests:

        Reviewed by Darin Adler.

        Check behavior of canvas.stroke() on transparent fillStyle()

        * fast/canvas/canvas-stroke-empty-fill-expected.txt: Added.
        * fast/canvas/canvas-stroke-empty-fill.html: Added.
        * fast/canvas/canvas-stroke-empty-fill.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38364 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Stephanie Lewis <slewis@apple.com>
slewis@apple.com [Thu, 13 Nov 2008 07:17:17 +0000 (07:17 +0000)]
2008-11-12  Stephanie Lewis  <slewis@apple.com>

        Fix Mac build.

        * Panels/WebAuthenticationPanel.m:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38363 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Stephanie <slewis@apple.com>
slewis@apple.com [Thu, 13 Nov 2008 06:08:46 +0000 (06:08 +0000)]
2008-11-12  Stephanie  <slewis@apple.com>

        Reviewed by Mark Rowe.

        Fix context leak in XSLStyleSheets.
        See bug https://bugs.webkit.org/show_bug.cgi?id=15715

        * xml/XSLStyleSheet.cpp:
        (WebCore::XSLStyleSheet::parseString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38362 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebKit:
sullivan@apple.com [Thu, 13 Nov 2008 04:47:56 +0000 (04:47 +0000)]
WebKit:

2008-11-12  John Sullivan  <sullivan@apple.com>

        fixed <rdar://problem/3839110> Authentication dialogs talk about passwords being sent "in the clear"

        Reviewed by Tim Hatcher

        * English.lproj/Localizable.strings:
        * StringsNotToBeLocalized.txt:
        updated for this change

WebKit/mac:

2008-11-12  John Sullivan  <sullivan@apple.com>

        fixed <rdar://problem/3839110> Authentication dialogs talk about passwords being sent "in the clear"

        Reviewed by Tim Hatcher

        * Panels/WebAuthenticationPanel.m:
        (-[WebAuthenticationPanel setUpForChallenge:]):
        use "unencrypted" instead of "in the clear". Also, use the "Your log-in information
        will be sent securely" version when the receiving server is https, regardless of
        whether it uses basic or digest authentication.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38361 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Alp Toker <alp@nuanti.com>
alp@webkit.org [Thu, 13 Nov 2008 04:26:39 +0000 (04:26 +0000)]
2008-11-12  Alp Toker  <alp@nuanti.com>

        autotools: List web worker-related headers needed for a successful
        build. distcheck is passing again with this change.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38360 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Alp Toker <alp@nuanti.com>
alp@webkit.org [Thu, 13 Nov 2008 02:30:09 +0000 (02:30 +0000)]
2008-11-12  Alp Toker  <alp@nuanti.com>

        autotools build system fix-up only. Add FloatQuad.h to the source
        lists and sort them.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38356 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Darin Adler <darin@apple.com>
darin@apple.com [Thu, 13 Nov 2008 02:04:58 +0000 (02:04 +0000)]
2008-11-12  Darin Adler  <darin@apple.com>

        Reviewed by Mark Rowe.

        - fix https://bugs.webkit.org/show_bug.cgi?id=22223
          <rdar://problem/6366864> REGRESSION(r38245): "View Source" crashes the browser

        * WebView/WebFrameView.mm: (-[WebFrameView viewDidMoveToWindow]):
        Add missing null check.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38355 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Pierre-Olivier Latour <pol@apple.com>
mitz@apple.com [Thu, 13 Nov 2008 01:41:19 +0000 (01:41 +0000)]
2008-11-12  Pierre-Olivier Latour  <pol@apple.com>

        Reviewed by Dan Bernstein.

        Updated reference images so that all pixel tests now "pass" on 10.4.11 PowerPC + r38347.
        https://bugs.webkit.org/show_bug.cgi?id=21821

        Checked-in specific reference images for pixel tests that are valid on Tiger but fail
        because difference with reference image due to text rendering is greater than 1.0%

        * platform/mac-tiger/svg/carto.net: Added.
        * platform/mac-tiger/svg/carto.net/selectionlist-expected.checksum: Added.
        * platform/mac-tiger/svg/carto.net/selectionlist-expected.png: Added.
        * platform/mac-tiger/svg/custom/feComponentTransfer-Gamma-expected.checksum: Added.
        * platform/mac-tiger/svg/custom/feComponentTransfer-Gamma-expected.png: Added.
        * platform/mac-tiger/svg/text/text-align-01-b-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-align-01-b-expected.png: Added.
        * platform/mac-tiger/svg/text/text-align-02-b-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-align-02-b-expected.png: Added.
        * platform/mac-tiger/svg/text/text-align-04-b-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-align-04-b-expected.png: Added.
        * platform/mac-tiger/svg/text/text-fonts-01-t-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-fonts-01-t-expected.png: Added.
        * platform/mac-tiger/svg/text/text-text-03-b-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-text-03-b-expected.png: Added.
        * platform/mac-tiger/svg/text/text-tselect-01-b-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-tselect-01-b-expected.png: Added.
        * platform/mac-tiger/svg/text/text-tspan-01-b-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-tspan-01-b-expected.png: Added.
        * platform/mac-tiger/svg/text/text-ws-01-t-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-ws-01-t-expected.png: Added.
        * platform/mac-tiger/svg/text/text-ws-02-t-expected.checksum: Added.
        * platform/mac-tiger/svg/text/text-ws-02-t-expected.png: Added.
        * platform/mac-tiger/svg/W3C-SVG-1.1/filters-color-01-b-expected.checksum: Added.
        * platform/mac-tiger/svg/W3C-SVG-1.1/filters-color-01-b-expected.png: Added.
        * platform/mac-tiger/svg/W3C-SVG-1.1/filters-comptran-01-b-expected.checksum: Added.
        * platform/mac-tiger/svg/W3C-SVG-1.1/filters-comptran-01-b-expected.png: Added.

        svg/batik/paints/gradientLimit.svg is rendering different on Tiger and Leopard
        https://bugs.webkit.org/show_bug.cgi?id=22219

        * platform/mac-tiger/svg/batik: Added.
        * platform/mac-tiger/svg/batik/paints: Added.
        * platform/mac-tiger/svg/batik/paints/gradientLimit-expected.checksum: Added.
        * platform/mac-tiger/svg/batik/paints/gradientLimit-expected.png: Added.

        svg/custom/mask-excessive-malloc.svg is failing on Tiger
        https://bugs.webkit.org/show_bug.cgi?id=22218

        * platform/mac-tiger/svg/custom/mask-excessive-malloc-expected.checksum: Added.
        * platform/mac-tiger/svg/custom/mask-excessive-malloc-expected.png: Added.

        svg/W3C-SVG-1.1/pservers-grad-17-b.svg is failing on Leopard
        https://bugs.webkit.org/show_bug.cgi?id=22220

        * platform/mac-leopard/svg: Added.
        * platform/mac-leopard/svg/W3C-SVG-1.1: Added.
        * platform/mac-leopard/svg/W3C-SVG-1.1/pservers-grad-17-b-expected.checksum: Added.
        * platform/mac-leopard/svg/W3C-SVG-1.1/pservers-grad-17-b-expected.png: Added.
        * platform/mac/svg/W3C-SVG-1.1/pservers-grad-17-b-expected.checksum:
        * platform/mac/svg/W3C-SVG-1.1/pservers-grad-17-b-expected.png:

        svg/custom/foreign-object-skew.svg is failing on Leopard
        https://bugs.webkit.org/show_bug.cgi?id=22221

        * platform/mac-leopard/svg/custom: Added.
        * platform/mac-leopard/svg/custom/foreign-object-skew-expected.checksum: Added.
        * platform/mac-leopard/svg/custom/foreign-object-skew-expected.png: Added.
        * platform/mac/svg/custom/foreign-object-skew-expected.checksum:
        * platform/mac/svg/custom/foreign-object-skew-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38354 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Thu, 13 Nov 2008 01:32:11 +0000 (01:32 +0000)]
2008-11-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein

        https://bugs.webkit.org/show_bug.cgi?id=22212

        Rename absoluteOutlineBox() to absoluteOutlineBounds(), to indicate that
        it may be the bounding outline box of a transformed element, and fix it
        to respect transforms by mapping the outline box to an absolute quad and
        taking the enclosing rect.

        RenderBox::localToAbsoluteQuad() can no longer assert that there is no
        LayoutState, but LayoutState cannot be used during quad mapping (it knows
        nothing about transforms).

        Finally, fix a bug in RenderBox::localToAbsoluteQuad() which was getting
        borderTopExtra() from the object, rather than its container (as localToAbsolute()
        does).

        Test: fast/repaint/transform-absolute-child.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38353 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agodyslexics untie\!
ggaren@apple.com [Thu, 13 Nov 2008 01:22:43 +0000 (01:22 +0000)]
dyslexics untie\!

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38352 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Geoffrey Garen <ggaren@apple.com>
ggaren@apple.com [Thu, 13 Nov 2008 01:20:50 +0000 (01:20 +0000)]
2008-11-12  Geoffrey Garen  <ggaren@apple.com>

        Roll in r38322, onw that test failures have been fixed.

        * fast/js/global-recursion-on-full-stack-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38351 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Geoffrey Garen <ggaren@apple.com>
ggaren@apple.com [Thu, 13 Nov 2008 00:48:23 +0000 (00:48 +0000)]
2008-11-12  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Fixed https://bugs.webkit.org/show_bug.cgi?id=22192
        +37 failures in fast/profiler

        along with Darin's review comments in
        https://bugs.webkit.org/show_bug.cgi?id=22174
        Simplified op_call by nixing its responsibility for moving the value of
        "this" into the first argument slot

        * VM/Machine.cpp:
        (JSC::returnToThrowTrampoline):
        (JSC::throwStackOverflowError):
        (JSC::Machine::cti_register_file_check):
        (JSC::Machine::cti_op_call_arityCheck):
        (JSC::Machine::cti_vm_throw): Moved the throw logic into a function, since
        functions are better than macros.

        * bytecompiler/CodeGenerator.cpp:
        (JSC::CodeGenerator::emitCall):
        (JSC::CodeGenerator::emitConstruct): Ensure that the function register
        is preserved if profiling is enabled, since the profiler uses that
        register.

        * runtime/JSGlobalData.h: Renamed throwReturnAddress to exceptionLocation,
        because I had a hard time understanding what "throwReturnAddress" meant.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38349 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Alp Toker <alp@nuanti.com>
alp@webkit.org [Thu, 13 Nov 2008 00:44:05 +0000 (00:44 +0000)]
2008-11-12  Alp Toker  <alp@nuanti.com>

        Build fixes for r38346. FloatQuad.cpp is missing:

        1) #include "config.h"
        2) #include <algorithm> for various mathematical definitions it uses.

        * platform/graphics/FloatQuad.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38348 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Pierre-Olivier Latour <pol@apple.com>
mitz@apple.com [Thu, 13 Nov 2008 00:24:40 +0000 (00:24 +0000)]
2008-11-12  Pierre-Olivier Latour  <pol@apple.com>

        Reviewed by Dan Bernstein.

        Skip fast/repaint/body-background-image.html test as it is not reliable.

        https://bugs.webkit.org/show_bug.cgi?id=21913

        * platform/mac/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38347 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Wed, 12 Nov 2008 23:43:52 +0000 (23:43 +0000)]
2008-11-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dave Hyatt

        https://bugs.webkit.org/show_bug.cgi?id=21942
        https://bugs.webkit.org/show_bug.cgi?id=18557

        Add methods which can be used to map renderer-local rectangles
        to quads in absolute coordinates, taking transforms into account:
            localToAbsoluteQuad() converts a local rect into an absolute quad.
            collectAbsoluteLineBoxQuads() is an analogue of addLineBoxRects()
                that works with quads.
            absoluteQuads() is an analogue of absoluteRects(), for quads.

        Use the quad methods to fix the inspector highlight for transformed
        elements.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38346 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Wed, 12 Nov 2008 23:04:31 +0000 (23:04 +0000)]
2008-11-12  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Fix stupid bug.

        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView createPlugin]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38345 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Dan Bernstein <mitz@apple.com>
mitz@apple.com [Wed, 12 Nov 2008 22:53:56 +0000 (22:53 +0000)]
2008-11-12  Dan Bernstein  <mitz@apple.com>

        - remove an empty directory

        * platform/mac-leopard/http: Removed.
        * platform/mac-leopard/http/tests: Removed.
        * platform/mac-leopard/http/tests/security: Removed.
        * platform/mac-leopard/http/tests/security/dataURL: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38344 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Fix Bug 22178: Build errors in WebKit Cairo port
aroben@apple.com [Wed, 12 Nov 2008 21:54:52 +0000 (21:54 +0000)]
    Fix Bug 22178: Build errors in WebKit Cairo port

        <https://bugs.webkit.org/show_bug.cgi?id=22178>

        Reviewed by Mark Rowe.

        * platform/graphics/cairo/PatternCairo.cpp: Correct build error under
        Visual Studio.
        (WebCore::Pattern::createPlatformPattern):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38343 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Fix a typo in the include path for the Release_Cairo configuration
aroben@apple.com [Wed, 12 Nov 2008 21:49:51 +0000 (21:49 +0000)]
    Fix a typo in the include path for the Release_Cairo configuration

        * WebCore.vcproj/WebCore.vcproj: Fixed "caro" -> "cairo" typo.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38342 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoTouch precompiled header wrapper to fix build.
sfalken@apple.com [Wed, 12 Nov 2008 21:02:54 +0000 (21:02 +0000)]
Touch precompiled header wrapper to fix build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38341 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoSet svn:mime-type to image/png on .png files that did not have it
mitz@apple.com [Wed, 12 Nov 2008 20:44:15 +0000 (20:44 +0000)]
Set svn:mime-type to image/png on .png files that did not have it

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38340 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Wed, 12 Nov 2008 20:28:44 +0000 (20:28 +0000)]
2008-11-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dave Hyatt

        https://bugs.webkit.org/show_bug.cgi?id=17840
        Patch 2

        Code cleanup in RenderBox::computeAbsoluteRepaintRect().

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::computeAbsoluteRepaintRect):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38339 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Wed, 12 Nov 2008 20:22:28 +0000 (20:22 +0000)]
2008-11-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dave Hyatt

        https://bugs.webkit.org/show_bug.cgi?id=17840

        Fix repainting when the transform on a positioned element changes.
        In computeAbsoluteRepaintRect(), add the offsets due to relative positioning,
        and being in a positioned container after applying the transform to the local
        rect.

        Tests: fast/repaint/transform-absolute-in-positioned-container.html
               fast/repaint/transform-relative-position.html

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::computeAbsoluteRepaintRect):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38338 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
justin.garcia@apple.com [Wed, 12 Nov 2008 20:08:06 +0000 (20:08 +0000)]
WebCore:

2008-11-12  Justin Garcia  <justin.garcia@apple.com>

        Reviewed by Beth Dakin.

        <rdar://problem/5495723> Selecting and deleting quoted text quotes other text
        <rdar://problem/4775313> Deleting lines from the bottom of a messages leaves the last blank line quoted

        We don't want to merge into a block if it will mean changing the quote level of content after deleting
        selections that contain a whole number paragraphs plus a line break, since it is unclear to most users
        that such a selection actually ends at the start of the next paragraph.  Instead we want to completely
        remove the selected paragraph(s) and all evidence of the first one's quote level.  This matches TextEdit behavior
        for indented paragraphs.

        * editing/DeleteSelectionCommand.cpp:
        (WebCore::DeleteSelectionCommand::initializePositionData): For the selections described above, do not
        try to merge after the deletion.  Instead try and prune the start block or blocks if they've been emptied
        so that we remove evidence of the deleted paragraphs' quote level.
        (WebCore::DeleteSelectionCommand::mergeParagraphs): Try and prune the start block(s) if necessary.  Also
        make sure that the caret is placed correctly so that it ends up on the same line that the deleted selection
        started on, instead of one higher.
        * editing/DeleteSelectionCommand.h:
        * editing/htmlediting.cpp:
        (WebCore::numEnclosingMailBlockquotes): Added.
        * editing/htmlediting.h:

LayoutTests:

2008-11-12  Justin Garcia  <justin.garcia@apple.com>

        Reviewed by Beth Dakin.

        <rdar://problem/5495723> Selecting and deleting quoted text quotes other text
        <rdar://problem/4775313> Deleting lines from the bottom of a messages leaves the last blank line quoted

        * editing/deleting/5495723-expected.txt: Added.
        * editing/deleting/5495723.html: Added.
        * platform/mac/editing/deleting/delete-4038408-fix-expected.txt: Change demonstrates that we
        now behave correctly when pressing delete when the caret is just after empty quoted paragraphs.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38337 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoTouch precompiled header wrapper to try and fix build.
sfalken@apple.com [Wed, 12 Nov 2008 19:40:12 +0000 (19:40 +0000)]
Touch precompiled header wrapper to try and fix build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38336 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Pierre-Olivier Latour <pol@apple.com>
mitz@apple.com [Wed, 12 Nov 2008 18:35:26 +0000 (18:35 +0000)]
2008-11-12  Pierre-Olivier Latour  <pol@apple.com>

        Reviewed by Dan Bernstein.

        Updated reference images so that all pixel tests now "pass" on 10.5.5 Intel + r38318.
        https://bugs.webkit.org/show_bug.cgi?id=21821

        Removed empty directory.

        * platform/mac-leopard/http: Removed.
        * platform/mac-leopard/http/tests: Removed.
        * platform/mac-leopard/http/tests/security: Removed.
        * platform/mac-leopard/http/tests/security/dataURL: Removed.

        Test doesn't repaint entire view so result is corrupted by previous test
        https://bugs.webkit.org/show_bug.cgi?id=21916

        * platform/mac/Skipped:

        Test is passing but original reference image was invalid.

        * platform/mac/fast/backgrounds/size/backgroundSize05-expected.checksum:
        * platform/mac/fast/backgrounds/size/backgroundSize05-expected.png:

        Invalid image tiling
        https://bugs.webkit.org/show_bug.cgi?id=21915

        * platform/mac/fast/backgrounds/size/backgroundSize19-expected.checksum:
        * platform/mac/fast/backgrounds/size/backgroundSize19-expected.png:

        SVG not rendering
        https://bugs.webkit.org/show_bug.cgi?id=21910

        * platform/mac/fast/backgrounds/svg-as-background-1-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-background-1-expected.png:
        * platform/mac/fast/backgrounds/svg-as-background-2-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-background-2-expected.png:
        * platform/mac/fast/backgrounds/svg-as-background-3-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-background-3-expected.png:
        * platform/mac/fast/backgrounds/svg-as-background-4-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-background-4-expected.png:
        * platform/mac/fast/backgrounds/svg-as-background-5-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-background-5-expected.png:
        * platform/mac/fast/backgrounds/svg-as-background-6-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-background-6-expected.png:
        * platform/mac/fast/backgrounds/svg-as-mask-expected.checksum:
        * platform/mac/fast/backgrounds/svg-as-mask-expected.png:
        * platform/mac/fast/borders/svg-as-border-image-2-expected.checksum:
        * platform/mac/fast/borders/svg-as-border-image-2-expected.png:
        * platform/mac/fast/borders/svg-as-border-image-expected.checksum:
        * platform/mac/fast/borders/svg-as-border-image-expected.png:
        * platform/mac/fast/images/svg-as-background-expected.checksum:
        * platform/mac/fast/images/svg-as-background-expected.png:
        * platform/mac/fast/images/svg-as-image-expected.checksum:
        * platform/mac/fast/images/svg-as-image-expected.png:
        * platform/mac/fast/images/svg-as-relative-image-expected.checksum:
        * platform/mac/fast/images/svg-as-relative-image-expected.png:
        * platform/mac/fast/images/svg-as-tiled-background-expected.checksum:
        * platform/mac/fast/images/svg-as-tiled-background-expected.png:

        Possible Flash failure
        https://bugs.webkit.org/show_bug.cgi?id=21912

        * platform/mac/fast/dynamic/flash-replacement-test-expected.checksum:
        * platform/mac/fast/dynamic/flash-replacement-test-expected.png:

        Repaint failure
        https://bugs.webkit.org/show_bug.cgi?id=21913

        * platform/mac/fast/repaint/body-background-image-expected.checksum:
        * platform/mac/fast/repaint/body-background-image-expected.png:

        Excessive repainting
        https://bugs.webkit.org/show_bug.cgi?id=21914

        * platform/mac/fast/repaint/float-overflow-expected.checksum:
        * platform/mac/fast/repaint/float-overflow-expected.png:
        * platform/mac/fast/repaint/float-overflow-right-expected.checksum:
        * platform/mac/fast/repaint/float-overflow-right-expected.png:

        No video images
        https://bugs.webkit.org/show_bug.cgi?id=21911

        * platform/mac/media/video-aspect-ratio-expected.checksum:
        * platform/mac/media/video-aspect-ratio-expected.png:
        * platform/mac/media/video-controls-rendering-expected.checksum:
        * platform/mac/media/video-controls-rendering-expected.png:
        * platform/mac/media/video-display-toggle-expected.checksum:
        * platform/mac/media/video-display-toggle-expected.png:
        * platform/mac/media/video-layer-crash-expected.checksum:
        * platform/mac/media/video-layer-crash-expected.png:
        * platform/mac/media/video-transformed-expected.checksum:
        * platform/mac/media/video-transformed-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38335 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoJavaScriptCore:
barraclough@apple.com [Wed, 12 Nov 2008 11:31:05 +0000 (11:31 +0000)]
JavaScriptCore:

2008-11-12  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Cameron Zwarich.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=22201
        Integer conversion in array.length was safe signed values,
        but the length is unsigned.

        * VM/CTI.cpp:
        (JSC::CTI::privateCompilePatchGetArrayLength):

LayoutTests:

2008-11-12  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Cameron Zwarich.

        Test for https://bugs.webkit.org/show_bug.cgi?id=22201

        * fast/js/pic/cached-array-length-access-expected.txt: Added.
        * fast/js/pic/cached-array-length-access.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38334 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Tor Arne Vestbø <tavestbo@trolltech.com>
vestbo@webkit.org [Wed, 12 Nov 2008 10:51:27 +0000 (10:51 +0000)]
2008-11-10  Tor Arne Vestbø  <tavestbo@trolltech.com>

        Reviewed by Simon Hausmann.

        Move _web_encodingForResource from WebKit into WebCore and change return type

        This change is needed to implement NSAPI in WebCore for Mac, see:

        https://bugs.webkit.org/show_bug.cgi?id=21427

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38333 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Tor Arne Vestbø <tavestbo@trolltech.com>
vestbo@webkit.org [Wed, 12 Nov 2008 10:44:55 +0000 (10:44 +0000)]
2008-11-10  Tor Arne Vestbø  <tavestbo@trolltech.com>

        Reviewed by Simon Hausmann.

        Moved the implementation of _webkit_isCaseInsensitiveEqualToString to
        WebCore's WebCoreNSStringExtras as _stringIsCaseInsensitiveEqualToString.

        This change is needed to implement NSAPI in WebCore for Mac, see:

        https://bugs.webkit.org/show_bug.cgi?id=21427

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38332 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Ariya Hidayat <ariya.hidayat@trolltech.com>
vestbo@webkit.org [Wed, 12 Nov 2008 09:51:21 +0000 (09:51 +0000)]
2008-11-11  Ariya Hidayat  <ariya.hidayat@trolltech.com>

        Reviewed by Tor Arne Vestbø and Simon Hausmann.

        Use QPalette::Base (instead of QPalette::Background) for the page
        background.

        * Api/qwebframe.cpp:
        (QWebFramePrivate::updateBackground):
        * Api/qwebview.cpp:
        (QWebView::QWebView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38331 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-12 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Wed, 12 Nov 2008 09:34:22 +0000 (09:34 +0000)]
2008-11-12  Cameron Zwarich  <zwarich@apple.com>

        Rubber-stamped by Mark Rowe.

        Roll out r38322 due to test failures on the bots.

        JavaScriptCore:

        * VM/CTI.cpp:
        (JSC::CTI::compileOpCallSetupArgs):
        (JSC::CTI::compileOpCall):
        (JSC::CTI::privateCompileMainPass):
        (JSC::CTI::privateCompileSlowCases):
        * VM/CTI.h:
        * VM/CodeBlock.cpp:
        (JSC::CodeBlock::dump):
        * VM/Machine.cpp:
        (JSC::Machine::callEval):
        (JSC::Machine::dumpCallFrame):
        (JSC::Machine::dumpRegisters):
        (JSC::Machine::execute):
        (JSC::Machine::privateExecute):
        (JSC::Machine::throwStackOverflowPreviousFrame):
        (JSC::Machine::cti_register_file_check):
        (JSC::Machine::cti_op_call_arityCheck):
        (JSC::Machine::cti_op_call_NotJSFunction):
        (JSC::Machine::cti_op_construct_JSConstruct):
        (JSC::Machine::cti_op_construct_NotJSConstruct):
        (JSC::Machine::cti_op_call_eval):
        (JSC::Machine::cti_vm_throw):
        * VM/Machine.h:
        * bytecompiler/CodeGenerator.cpp:
        (JSC::CodeGenerator::emitCall):
        (JSC::CodeGenerator::emitCallEval):
        (JSC::CodeGenerator::emitConstruct):
        * bytecompiler/CodeGenerator.h:
        * parser/Nodes.cpp:
        (JSC::EvalFunctionCallNode::emitCode):
        (JSC::FunctionCallValueNode::emitCode):
        (JSC::FunctionCallResolveNode::emitCode):
        (JSC::FunctionCallBracketNode::emitCode):
        (JSC::FunctionCallDotNode::emitCode):
        * parser/Nodes.h:
        (JSC::ScopeNode::neededConstants):

        LayoutTests:

        * fast/js/global-recursion-on-full-stack-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38330 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Stephanie <slewis@apple.com>
slewis@apple.com [Wed, 12 Nov 2008 03:30:12 +0000 (03:30 +0000)]
2008-11-11  Stephanie  <slewis@apple.com>

        Reviewed by Dan Bernstein.

        Null check image data before setting it as image source.

        Test: fast/images/image-empty-data.html

        * platform/graphics/BitmapImage.cpp:
        (WebCore::BitmapImage::destroyDecodedData):

        Test loading an image with no data.

        * fast/images/image-empty-data-expected.txt: Added.
        * fast/images/image-empty-data.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38329 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Adele Peterson <adele@apple.com>
adele@apple.com [Wed, 12 Nov 2008 03:09:38 +0000 (03:09 +0000)]
2008-11-11  Adele Peterson  <adele@apple.com>

        Reviewed by Tim Hatcher.

        Remove ifdef so future OS versions can use the new delegate method.

        * platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::didSendBodyDataDelegateExists):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38328 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Antti Koivisto and Sam Weinig.
eric@webkit.org [Wed, 12 Nov 2008 01:53:37 +0000 (01:53 +0000)]
    Reviewed by Antti Koivisto and Sam Weinig.

        Clean up HTMLTokenizer a litle
        https://bugs.webkit.org/show_bug.cgi?id=22188

        Rename pendingScripts to m_pendingScripts
        Rename scriptNode to m_scriptNode make the type specific
        Rename pendingSrc to m_pendingSrc
        Rename currentPrependingSrc to m_currentPrependingSrc
        Rename noMoreData to m_noMoreData and cBuffer to m_cBuffer
        Remove long-since-dead kdDebug and qDebug calls
        Rename brokenComments to m_brokenComments
        Remove HTMLTokenizer includes and document the rest
        Rename src to m_src
        Rename parser to m_parser and make it an OwnPtr
        Rename inWrite to m_inWrite and jsProxy to scriptController
        Rename brokenServer to m_brokenServer
        Rename buffer to m_buffer and dest to m_dest
        Rename size to m_bufferSize
        Rename attrName to m_attrName
        Rename searchStopper to m_searchStopper and searchStopperLen to m_searchStopperLen
        Rename scriptCode* to m_scriptCode* and change scriptCodeMax to m_scriptCodeCapacity
        Rename scriptStartLineno to m_currentScriptTagStartLineNumber and tagStartLineno to m_currentTagStartLineNumber
        Rename scriptSrc to m_scriptTagSrcAttrValue and scriptSrcCharset to m_scriptTagCharsetAttrValue -- a bit unwieldy, but more precise
        Rename flat to selfClosingTag
        Rename currToken to m_currentToken

        * css/CSSParser.cpp:
        * dom/XMLTokenizer.cpp:
        * dom/XMLTokenizerLibxml2.cpp:
        * html/HTMLDocument.cpp:
        * html/HTMLElement.cpp:
        * html/HTMLFormControlElement.cpp:
        * html/HTMLParser.h:
        * html/HTMLTokenizer.cpp:
        (WebCore::Token::addAttribute):
        (WebCore::HTMLTokenizer::HTMLTokenizer):
        (WebCore::HTMLTokenizer::reset):
        (WebCore::HTMLTokenizer::begin):
        (WebCore::HTMLTokenizer::processListing):
        (WebCore::HTMLTokenizer::parseSpecial):
        (WebCore::HTMLTokenizer::scriptHandler):
        (WebCore::HTMLTokenizer::scriptExecution):
        (WebCore::HTMLTokenizer::parseComment):
        (WebCore::HTMLTokenizer::parseServer):
        (WebCore::HTMLTokenizer::parseProcessingInstruction):
        (WebCore::HTMLTokenizer::parseText):
        (WebCore::HTMLTokenizer::parseEntity):
        (WebCore::HTMLTokenizer::parseTag):
        (WebCore::HTMLTokenizer::write):
        (WebCore::HTMLTokenizer::processingData):
        (WebCore::HTMLTokenizer::end):
        (WebCore::HTMLTokenizer::finish):
        (WebCore::HTMLTokenizer::processToken):
        (WebCore::HTMLTokenizer::processDoctypeToken):
        (WebCore::HTMLTokenizer::~HTMLTokenizer):
        (WebCore::HTMLTokenizer::enlargeBuffer):
        (WebCore::HTMLTokenizer::enlargeScriptBuffer):
        (WebCore::HTMLTokenizer::notifyFinished):
        (WebCore::HTMLTokenizer::setSrc):
        * html/HTMLTokenizer.h:
        (WebCore::Token::Token):
        (WebCore::Token::reset):
        (WebCore::HTMLTokenizer::processingContentWrittenByScript):
        (WebCore::HTMLTokenizer::htmlParser):
        (WebCore::HTMLTokenizer::checkBuffer):
        (WebCore::HTMLTokenizer::checkScriptBuffer):
        * html/HTMLViewSourceDocument.h:
        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::leftRelOffset):
        (WebCore::RenderBlock::rightRelOffset):
        (WebCore::RenderBlock::lineWidth):
        * xml/XSLTProcessor.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38327 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Wed, 12 Nov 2008 01:37:26 +0000 (01:37 +0000)]
2008-11-11  Simon Fraser  <simon.fraser@apple.com>

        Update leaves animation sample.

        * blog-files/leaves/index.html:
        * blog-files/leaves/leaves.css:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38326 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Dean Jackson <dino@apple.com>
dino@apple.com [Wed, 12 Nov 2008 01:35:08 +0000 (01:35 +0000)]
2008-11-11  Dean Jackson  <dino@apple.com>

        Removed quotes in animation name value.

        * blog-files/pulse.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38325 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Dean Jackson <dino@apple.com>
dino@apple.com [Wed, 12 Nov 2008 00:49:58 +0000 (00:49 +0000)]
2008-11-11  Dean Jackson  <dino@apple.com>

        Added a smaller version of a screenshot.

        * blog-files/leaves-screenshot-small.jpg: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38324 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Dean Jackson <dino@apple.com>
dino@apple.com [Wed, 12 Nov 2008 00:32:54 +0000 (00:32 +0000)]
2008-11-11  Dean Jackson  <dino@apple.com>

        Add a bunch of examples for CSS Animation blog draft.

        * blog-files/bounce.html: Added.
        * blog-files/bounce.png: Added.
        * blog-files/leaves-screenshot.jpg: Added.
        * blog-files/leaves/images/apple-touch-icon.png: Added.
        * blog-files/leaves/images/backgroundLeaves.jpg: Added.
        * blog-files/leaves/images/realLeaf1.png: Added.
        * blog-files/leaves/images/realLeaf2.png: Added.
        * blog-files/leaves/images/realLeaf3.png: Added.
        * blog-files/leaves/images/realLeaf4.png: Added.
        * blog-files/leaves/images/textBackground.png: Added.
        * blog-files/leaves/index.html: Added.
        * blog-files/leaves/leaves.css: Added.
        * blog-files/leaves/leaves.js: Added.
        * blog-files/pulse.html: Added.
        * blog-files/pulse.png: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38323 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Geoffrey Garen <ggaren@apple.com>
ggaren@apple.com [Wed, 12 Nov 2008 00:32:38 +0000 (00:32 +0000)]
2008-11-11  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Darin Adler.

        Fixed https://bugs.webkit.org/show_bug.cgi?id=22174
        Simplified op_call by nixing its responsibility for moving the value of
        "this" into the first argument slot.

        Instead, the caller emits an explicit load or mov instruction, or relies
        on implicit knowledge that "this" is already in the first argument slot.
        As a result, two operands to op_call are gone: firstArg and thisVal.

        SunSpider and v8 tests show no change in bytecode or CTI.

        * VM/CTI.cpp:
        (JSC::CTI::compileOpCallSetupArgs):
        (JSC::CTI::compileOpCallEvalSetupArgs):
        (JSC::CTI::compileOpConstructSetupArgs): Split apart these three versions
        of setting up arguments to op_call, because they're more different than
        they are the same -- even more so with this patch.

        (JSC::CTI::compileOpCall): Updated for the fact that op_construct doesn't
        match op_call anymore.

        (JSC::CTI::privateCompileMainPass):
        (JSC::CTI::privateCompileSlowCases): Merged a few call cases. Updated
        for changes mentioned above.

        * VM/CTI.h:

        * VM/CodeBlock.cpp:
        (JSC::CodeBlock::dump): Updated for new bytecode format of call / construct.

        * VM/Machine.cpp:
        (JSC::Machine::callEval): Updated for new bytecode format of call / construct.

        (JSC::Machine::dumpCallFrame):
        (JSC::Machine::dumpRegisters): Simplified these debugging functions,
        taking advantage of the new call frame layout.

        (JSC::Machine::execute): Fixed up the eval version of execute to be
        friendlier to calls in the new format.

        (JSC::Machine::privateExecute): Implemented the new call format in
        bytecode.

        (JSC::Machine::cti_op_call_NotJSFunction):
        (JSC::Machine::cti_op_construct_JSConstruct):
        (JSC::Machine::cti_op_construct_NotJSConstruct):
        (JSC::Machine::cti_op_call_eval): Updated CTI helpers to match the new
        call format.

        Fixed a latent bug in stack overflow checking that is now hit because
        the register layout has changed a bit -- namely: when throwing a stack
        overflow exception inside an op_call helper, we need to account for the
        fact that the current call frame is only half-constructed, and use the
        parent call frame instead.

        * VM/Machine.h:

        * bytecompiler/CodeGenerator.cpp:
        (JSC::CodeGenerator::emitCall):
        (JSC::CodeGenerator::emitCallEval):
        (JSC::CodeGenerator::emitConstruct):
        * bytecompiler/CodeGenerator.h: Updated codegen to match the new call
        format.

        * parser/Nodes.cpp:
        (JSC::EvalFunctionCallNode::emitCode):
        (JSC::FunctionCallValueNode::emitCode):
        (JSC::FunctionCallResolveNode::emitCode):
        (JSC::FunctionCallBracketNode::emitCode):
        (JSC::FunctionCallDotNode::emitCode):
        * parser/Nodes.h:
        (JSC::ScopeNode::neededConstants): ditto

2008-11-10  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Darin Adler.

        Updated a test after fixing https://bugs.webkit.org/show_bug.cgi?id=22174
        Simplified op_call by nixing its responsibility for moving the value of
        "this" into the first argument slot.

        * fast/js/global-recursion-on-full-stack-expected.txt: This test passes
        a little differently now, because the register layout has changed.
        Specifically, the stack overflow now happens in the call to f() instead
        of the initiation of the <script> tag, so it is caught, and it does not
        log an exception to the console.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38322 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Sam Weinig <sam@webkit.org>
weinig@apple.com [Wed, 12 Nov 2008 00:28:57 +0000 (00:28 +0000)]
2008-11-11  Sam Weinig  <sam@webkit.org>

        Reviewed by Dan Bernstein.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=22189
        Track CachedScript decoded data

        We now track the decoded size of CachedScripts in the Cache allowing
        them to accounted for when deciding when to evict data.  The decoded
        data itself can be evicted when Cache pressure gets high enough through
        CachedScript::destroyDecodedData.

        * loader/CachedResource.h:
        (WebCore::CachedResource::allClientsRemoved):
        (WebCore::CachedResource::destroyDecodedData):
        * loader/CachedScript.cpp:
        (WebCore::CachedScript::script):
        (WebCore::CachedScript::data):
        (WebCore::CachedScript::destroyDecodedData):
        * loader/CachedScript.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38321 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 23:39:15 +0000 (23:39 +0000)]
2008-11-11  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Geoff Garen.

        Remove pointless dependencies on the now-deleted kjs directory and
        mentions of it in comments that are no longer relevant.

        JavaScriptCore:

        * ForwardingHeaders/JavaScriptCore/JSLock.h: Removed.

        WebCore:

        * WebCore.pro:
        * webcore-base.bkl:

        WebKit/qt:

        * WebKit_pch.h:

        WebKit/wx:

        * presets/wxwebkit.bkl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38320 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Fix: https://bugs.webkit.org/show_bug.cgi?id=22187
adachan@apple.com [Tue, 11 Nov 2008 23:26:21 +0000 (23:26 +0000)]
    Fix: https://bugs.webkit.org/show_bug.cgi?id=22187
        Bug 22187: CLEARTYPE_QUALITY flag is not supported on Win2000

        Update window versions to correspond to Windows XP.

        Reviewed by Steve Falkenburg.

        * win/tools/vsprops/common.vsprops:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38319 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
mitz@apple.com [Tue, 11 Nov 2008 22:45:16 +0000 (22:45 +0000)]
WebCore:

        Reviewed by Adam Roben.

        WebCore part of adding a master volume control for media elements in a WebView

        * WebCore.base.exp: Exported WebCore::Page::setMediaVolume().
        * dom/Document.cpp:
        (WebCore::Document::mediaVolumeDidChange): Added. Called by the Page
        when the media volume is changed.
        (WebCore::Document::registerForMediaVolumeCallbacks): Added. Allows
        elements to register for a callback when the media volume is changed.
        (WebCore::Document::unregisterForMediaVolumeCallbacks): Added. Allows
        elements to unregister for the callback.
        * dom/Document.h:
        * dom/Element.h:
        (WebCore::Element::mediaVolumeDidChange): Added an empty implementation.
        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::HTMLMediaElement): Added a call to
        registerForMediaVolumeCallbacks().
        (WebCore::HTMLMediaElement::~HTMLMediaElement): Added a call to
        unregisterForMediaVolumeCallbacks().
        (WebCore::HTMLMediaElement::updateVolume): Changed to multiply the
        element's intrinsic volume by the master volume from the page.
        (WebCore::HTMLMediaElement::mediaVolumeDidChange): Added. Calls
        updateVolume().
        * html/HTMLMediaElement.h:
        * page/Page.cpp:
        (WebCore::Page::Page): Initialize m_mediaVolume to 1.
        (WebCore::Page::setMediaVolume): Added. Calls mediaVolumeDidChange() on
        every document in the page when the master volume changes.
        * page/Page.h:
        (WebCore::Page::mediaVolume): Added this getter.

WebKit/mac:

        Reviewed by Adam Roben.

        WebKit/mac part of adding a master volume control for media elements in a WebView

        * WebView/WebView.mm:
        (-[WebView setMediaVolume:]): Added.
        (-[WebView mediaVolume]): Added.
        * WebView/WebViewPrivate.h:

WebKit/win:

        Reviewed by Adam Roben.

        WebKit/win part of adding a master volume control for media elements in a WebView
        * Interfaces/IWebViewPrivate.idl: Added setMediaVolume() and
        mediaVolume().
        * WebView.cpp:
        (WebView::setMediaVolume): Added.
        (WebView::mediaVolume): Added.
        * WebView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38318 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Pierre-Olivier Latour <pol@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 22:31:45 +0000 (22:31 +0000)]
2008-11-11  Pierre-Olivier Latour  <pol@apple.com>

        Reviewed by Mark Rowe.

        Removed invalid bases incorrectly checked-in in r38305.

        https://bugs.webkit.org/show_bug.cgi?id=22185

        * platform/mac-tiger/fast/text/international/complex-character-based-fallback-expected.checksum: Removed.
        * platform/mac-tiger/fast/text/international/complex-character-based-fallback-expected.png: Removed.
        * platform/mac-tiger/fast/text/international/complex-character-based-fallback-expected.txt: Removed.
        * platform/mac-tiger/fast/text/international/hindi-spacing-expected.checksum: Removed.
        * platform/mac-tiger/fast/text/international/hindi-spacing-expected.png: Removed.
        * platform/mac-tiger/fast/text/international/hindi-spacing-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38317 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFix broken dependencies building JavaScriptCore on a freezing cold cat, caused
mrowe@apple.com [Tue, 11 Nov 2008 22:25:14 +0000 (22:25 +0000)]
Fix broken dependencies building JavaScriptCore on a freezing cold cat, caused
by failure to update all instances of "kjs" to their new locations.

* JavaScriptCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38316 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Aaron Golden <agolden@apple.com>
ddkilzer@apple.com [Tue, 11 Nov 2008 22:13:59 +0000 (22:13 +0000)]
2008-11-11  Aaron Golden  <agolden@apple.com>

        Bug 22134: -[WebHistoryItem dictionaryRepresentation] accesses past the end of a vector

        <https://bugs.webkit.org/show_bug.cgi?id=22134>

        Reviewed by Geoff Garen.

        * History/WebHistoryItem.mm:
        (-[WebHistoryItem initFromDictionaryRepresentation:]):
        (-[WebHistoryItem dictionaryRepresentation]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38315 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Rubber-stamped by Adam Roben.
ap@webkit.org [Tue, 11 Nov 2008 21:49:34 +0000 (21:49 +0000)]
    Rubber-stamped by Adam Roben.

        * wtf/AVLTree.h: (WTF::AVLTree::Iterator::start_iter):
        Fix indentation a little more.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38314 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 21:11:04 +0000 (21:11 +0000)]
2008-11-11  Cameron Zwarich  <zwarich@apple.com>

        Rubber-stamped by Sam Weinig.

        Clean up EvalCodeCache to match our coding style a bit more.

        * VM/EvalCodeCache.h:
        (JSC::EvalCodeCache::get):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38313 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 David Hyatt <hyatt@apple.com>
hyatt@apple.com [Tue, 11 Nov 2008 20:43:57 +0000 (20:43 +0000)]
2008-11-11  David Hyatt  <hyatt@apple.com>

        https://bugs.webkit.org/show_bug.cgi?id=22182

        Fix problems with scrollbars when the OS setting for clicks in the track jumping the thumb to that location
        is turned on.  Make sure to only jump the thumb if the click happens in the track.  Also fix some math
        errors when jumping the thumb that cause dragging after the jump to incorrectly jump the thumb again.

        Reviewed by Adele

        * platform/Scrollbar.cpp:
        (WebCore::Scrollbar::mouseDown):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38312 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 20:06:11 +0000 (20:06 +0000)]
2008-11-11  Cameron Zwarich  <zwarich@apple.com>

        Rubber-stamped by Sam Weinig.

        Bug 22179: Move EvalCodeCache from CodeBlock.h into its own file
        <https://bugs.webkit.org/show_bug.cgi?id=22179>

        * GNUmakefile.am:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * VM/CodeBlock.h:
        * VM/EvalCodeCache.h: Copied from VM/CodeBlock.h.
        * VM/Machine.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38311 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
bdakin@apple.com [Tue, 11 Nov 2008 19:58:31 +0000 (19:58 +0000)]
WebCore:

2008-11-11  Beth Dakin  <bdakin@apple.com>

        Reviewed by Darin Adler.

        Fix for <rdar://problem/5089327> Too much indentation when pasting
        quoted paragraphs

        This patch re-addresses pasting blockquotes into blockquotes. It
        backs out most of revision 38273. 38273 fixed the bug by inserting
        the pasted content as a sibling blockquote node to the pre-existing
        blockquote node. The problem with that is that by default,
        blockquotes have a giant margin, so visually, this can be weird.
        This patch instead inserts the pasted content as siblings of the
        the content already inside the outer blockquote, and then removes
        the blockquote node from the pasted content itself, so that it
        doesn't nest itself into the outer blockquote.
        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::doApply):

        Move contains() from Element to Node since there is nothing
        Element-specific about it.
        * dom/Element.cpp:
        * dom/Element.h:
        * dom/Node.cpp:
        (WebCore::Node::contains):
        * dom/Node.h:

LayoutTests:

2008-11-11  Beth Dakin  <bdakin@apple.com>

        Reviewed by Darin Adler.

        Tests for <rdar://problem/5089327> Too much indentation when
        pasting quoted paragraphs

        This is a new test:
        * editing/pasteboard/paste-blockquote-into-blockquote-3.html:
        Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-3-expected.checksum: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-3-expected.png: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-3-expected.txt: Added.

        The results for these tests previously reflected the behavior
        implemented in revision 38273. We have since decided that that
        behavior is wrong and we changed it with this patch.
        * platform/mac/editing/pasteboard/merge-end-blockquote-
        expected.checksum:
        * platform/mac/editing/pasteboard/merge-end-blockquote-
        expected.png:
        * platform/mac/editing/pasteboard/merge-end-blockquote-
        expected.txt:
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-2-expected.txt:
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-
        expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38310 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 David Hyatt <hyatt@apple.com>
hyatt@apple.com [Tue, 11 Nov 2008 19:50:18 +0000 (19:50 +0000)]
2008-11-11  David Hyatt  <hyatt@apple.com>

        https://bugs.webkit.org/show_bug.cgi?id=17997

        Stop using the new Leopard API tiling call for scaled tiles when drawing background images.  The method
        is buggy.  Fall back to using the slower (and more correct) pattern tiling that Tiger still uses.  This
        is a Leopard-only workaround, since the API has been fixed on Snow Leopard.

        Reviewed by John Sullivan

        * platform/graphics/cg/ImageCG.cpp:
        (WebCore::Image::drawPattern):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38309 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Tue, 11 Nov 2008 19:41:10 +0000 (19:41 +0000)]
2008-11-11  Dirk Schulze  <krit@webkit.org>

        Reviewed by Dave Hyatt.

        Pattern transform check was too strict for Cg. Scaling can cause a wrong
        color on the borders of the scaled object.

        Correct expected result of the tests.

        * fast/canvas/canvas-pattern-transform-expected.txt:
        * fast/canvas/canvas-pattern-transform.html:
        * fast/canvas/canvas-pattern-transform.js:
        * fast/canvas/canvas-radial-gradient-spreadMethod-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38308 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 David Hyatt <hyatt@apple.com>
hyatt@apple.com [Tue, 11 Nov 2008 19:18:26 +0000 (19:18 +0000)]
2008-11-10  David Hyatt  <hyatt@apple.com>

        Disable the Aqua look for menu lists when full page zoom is turned on.  Because the control is buggy
        when scaling and because it's impossible to paint into an offscreen bitmap without completely rewriting
        how we draw the control on Mac, our only real option for now is to just switch to our styled look when zoomed.

        Reviewed by Adam Roben

        * rendering/RenderThemeMac.mm:
        (WebCore::RenderThemeMac::isControlStyled):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38307 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 18:31:51 +0000 (18:31 +0000)]
2008-11-11  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Sam Weinig.

        Remove the 'm_' prefix from the fields of the SwitchRecord struct.

        * VM/CTI.cpp:
        (JSC::CTI::privateCompile):
        * VM/CTI.h:
        (JSC::SwitchRecord):
        (JSC::SwitchRecord::SwitchRecord):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38306 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Pierre-Olivier Latour <pol@apple.com>
mitz@apple.com [Tue, 11 Nov 2008 18:26:32 +0000 (18:26 +0000)]
2008-11-11  Pierre-Olivier Latour  <pol@apple.com>

        Rubber-stamped by Dan Bernstein.

        https://bugs.webkit.org/show_bug.cgi?id=21821

        Removed pixel results in platform/mac-tiger that were only slightly different
        (within 1% tolerance) from the platform/mac results.
        Removed results in platform/mac-tiger where the results in platform/mac were
        also correct for Tiger.
        Also updated results in platform/mac-tiger that were out of date.

        Made sure every test that has a specific mac-tiger render tree output also has matching images.

        (file listing elided)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38305 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Darin Adler <darin@apple.com>
darin@apple.com [Tue, 11 Nov 2008 17:59:01 +0000 (17:59 +0000)]
2008-11-11  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        - https://bugs.webkit.org/show_bug.cgi?id=22160
        speculative fix for <rdar://problem/5995450> Nil deref because nextLinePosition
        (previousLinePosition, too) uses a nil node

        * editing/visible_units.cpp:
        (WebCore::nextLinePosition): Use RenderObject::node instead of RenderObject::element
        to check editingIgnoresContent. It makes sense to use the associated DOM node for
        this, even in cases of anonymous content, and it avoids the case where element() can
        be 0.
        (WebCore::previousLinePosition): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38304 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 17:54:00 +0000 (17:54 +0000)]
2008-11-11  Cameron Zwarich  <zwarich@apple.com>

        Rubber-stamped by Sam Weinig.

        Make asInteger() a static function so that it has internal linkage.

        * VM/CTI.cpp:
        (JSC::asInteger):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38303 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Dirk Schulze <vbs85@gmx.de>
krit@webkit.org [Tue, 11 Nov 2008 17:40:54 +0000 (17:40 +0000)]
2008-11-11  Dirk Schulze  <vbs85@gmx.de>

WebCore:

        Reviewed by Darin Adler.

        Removed default: in applySpreadMethod. This causes the problems
        in radial gradients. We don't need CAIRO_EXTEND_NONE in SVG, Canvas
        or CSS.

        * platform/graphics/cairo/GraphicsContextCairo.cpp:
        (WebCore::applySpreadMethod):

LayoutTests:

        Reviewed by Darin Adler.

        Check behavior of default spreadMethod on radial gradients.

        * fast/canvas/canvas-radial-gradient-spreadMethod-expected.txt: Added.
        * fast/canvas/canvas-radial-gradient-spreadMethod.html: Added.
        * fast/canvas/canvas-radial-gradient-spreadMethod.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38302 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Ariya Hidayat <ariya.hidayat@trolltech.com>
vestbo@webkit.org [Tue, 11 Nov 2008 13:32:01 +0000 (13:32 +0000)]
2008-11-11  Ariya Hidayat  <ariya.hidayat@trolltech.com>

        Reviewed by Tor Arne Vestbø.

        Fix crash when rendering the file chooser.
        Also proper implementation for multi-selected files.

        * platform/qt/FileChooserQt.cpp:
        (WebCore::FileChooser::basenameForWidth):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38301 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Dirk Schulze <vbs85@gmx.de>
krit@webkit.org [Tue, 11 Nov 2008 11:35:29 +0000 (11:35 +0000)]
2008-11-11  Dirk Schulze  <vbs85@gmx.de>

        Reviewed by Oliver Hunter.

        Added test to check behavior on transformations of patterns in canvas.

        * fast/canvas/canvas-pattern-transform-expected.txt: Added.
        * fast/canvas/canvas-pattern-transform.html: Added.
        * fast/canvas/canvas-pattern-transform.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38300 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agogit-svn-id: https://svn.webkit.org/repository/webkit/trunk@38299 268f45cc-cd09-0410...
krit@webkit.org [Tue, 11 Nov 2008 11:05:49 +0000 (11:05 +0000)]
git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38299 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 11 Nov 2008 09:51:31 +0000 (09:51 +0000)]
2008-11-11  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Mark Rowe.

        - shrink CodeBlock and AST related Vectors to exact fit (5-10M savings on membuster test)

        No perf regression combined with the last patch (each seems like a small regression individually)

        * bytecompiler/CodeGenerator.cpp:
        (JSC::CodeGenerator::generate):
        * parser/Nodes.h:
        (JSC::SourceElements::releaseContentsIntoVector):
        * wtf/Vector.h:
        (WTF::Vector::shrinkToFit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38298 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Pierre-Olivier Latour <pol@apple.com>
mitz@apple.com [Tue, 11 Nov 2008 09:11:30 +0000 (09:11 +0000)]
2008-11-11  Pierre-Olivier Latour  <pol@apple.com>

        Rubber-stamped by Dan Bernstein

        https://bugs.webkit.org/show_bug.cgi?id=21821

        Remaining set of updated expected pixel results. This includes http/,
        media/, plugins/, scrollbars/, tables/, transforms/, transitions/ and
        webarchive/.

        (file listing elided)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38297 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 11 Nov 2008 09:10:58 +0000 (09:10 +0000)]
2008-11-11  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Mark Rowe.

        - remove inline capacity from declaration stacks (15M savings on membuster test)

        No perf regression on SunSpider or V8 test combined with other upcoming memory improvement patch.

        * JavaScriptCore.exp:
        * parser/Nodes.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38296 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-11 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 08:06:47 +0000 (08:06 +0000)]
2008-11-11  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Oliver Hunt.

        While r38286 removed the need for the m_callFrame member variable of
        CTI, it should be also be removed.

        * VM/CTI.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38295 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 07:57:57 +0000 (07:57 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Oliver Hunt.

        Make CTI::asInteger() a non-member function, since it needs no access to
        any of CTI's member variables.

        * VM/CTI.cpp:
        (JSC::asInteger):
        * VM/CTI.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38294 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Stephanie Lewis <slewis@apple.com>
slewis@apple.com [Tue, 11 Nov 2008 07:06:41 +0000 (07:06 +0000)]
2008-11-10  Stephanie Lewis  <slewis@apple.com>

        Reviewed by Dan Bernstein.

        Clear the image source cache even if we haven't decoded any of the image data.
        We use the image source cache when calculating layout.

        * loader/Cache.cpp:
        (WebCore::Cache::pruneDeadResources):
        * platform/graphics/BitmapImage.cpp:
        (WebCore::BitmapImage::destroyDecodedData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38293 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 06:04:37 +0000 (06:04 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Maciej Stachowiak.

        Use 'value' instead of 'js' in CTI as a name for JSValue* to match our
        usual convention elsewhere.

        * VM/CTI.cpp:
        (JSC::CTI::emitGetArg):
        (JSC::CTI::emitGetPutArg):
        (JSC::CTI::getConstantImmediateNumericArg):
        (JSC::CTI::printOpcodeOperandTypes):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38292 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 05:55:50 +0000 (05:55 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Maciej Stachowiak.

        Make CTI::getConstant() a member function of CodeBlock instead.

        * VM/CTI.cpp:
        (JSC::CTI::emitGetArg):
        (JSC::CTI::emitGetPutArg):
        (JSC::CTI::getConstantImmediateNumericArg):
        (JSC::CTI::printOpcodeOperandTypes):
        (JSC::CTI::privateCompileMainPass):
        * VM/CTI.h:
        * VM/CodeBlock.h:
        (JSC::CodeBlock::getConstant):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38291 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 05:45:14 +0000 (05:45 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Sam Weinig.

        Rename CodeBlock::isConstant() to isConstantRegisterIndex().

        * VM/CTI.cpp:
        (JSC::CTI::emitGetArg):
        (JSC::CTI::emitGetPutArg):
        (JSC::CTI::getConstantImmediateNumericArg):
        (JSC::CTI::printOpcodeOperandTypes):
        (JSC::CTI::privateCompileMainPass):
        * VM/CodeBlock.h:
        (JSC::CodeBlock::isConstantRegisterIndex):
        * bytecompiler/CodeGenerator.cpp:
        (JSC::CodeGenerator::emitEqualityOp):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38290 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agonon-CTI build fix take 2
barraclough@apple.com [Tue, 11 Nov 2008 05:40:13 +0000 (05:40 +0000)]
non-CTI build fix take 2

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38289 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agonon-CTI build fix
barraclough@apple.com [Tue, 11 Nov 2008 05:31:49 +0000 (05:31 +0000)]
non-CTI build fix

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38288 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 05:22:55 +0000 (05:22 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Sam Weinig.

        Remove the unused labels member variable of CodeBlock.

        * VM/CodeBlock.h:
        * VM/LabelID.h:
        (JSC::LabelID::setLocation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38287 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Tue, 11 Nov 2008 05:09:46 +0000 (05:09 +0000)]
2008-11-10  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Camron Zwarich.

        Batch compile the set of static trampolines at the point Machine is constructed, using a single allocation.
        Refactor out m_callFrame from CTI, since this is only needed to access the global data (instead store a
        pointer to the global data directly, since this is available at the point the Machine is constructed).
        Add a method to align the code buffer, to allow JIT generation for multiple trampolines in one block.

        * VM/CTI.cpp:
        (JSC::CTI::getConstant):
        (JSC::CTI::emitGetArg):
        (JSC::CTI::emitGetPutArg):
        (JSC::CTI::getConstantImmediateNumericArg):
        (JSC::CTI::printOpcodeOperandTypes):
        (JSC::CTI::CTI):
        (JSC::CTI::compileBinaryArithOp):
        (JSC::CTI::privateCompileMainPass):
        (JSC::CTI::privateCompileGetByIdProto):
        (JSC::CTI::privateCompileGetByIdChain):
        (JSC::CTI::privateCompileCTIMachineTrampolines):
        (JSC::CTI::freeCTIMachineTrampolines):
        * VM/CTI.h:
        (JSC::CTI::compile):
        (JSC::CTI::compileGetByIdSelf):
        (JSC::CTI::compileGetByIdProto):
        (JSC::CTI::compileGetByIdChain):
        (JSC::CTI::compilePutByIdReplace):
        (JSC::CTI::compilePutByIdTransition):
        (JSC::CTI::compileCTIMachineTrampolines):
        (JSC::CTI::compilePatchGetArrayLength):
        * VM/Machine.cpp:
        (JSC::Machine::initialize):
        (JSC::Machine::~Machine):
        (JSC::Machine::execute):
        (JSC::Machine::tryCTICachePutByID):
        (JSC::Machine::tryCTICacheGetByID):
        (JSC::Machine::cti_op_call_JSFunction):
        (JSC::Machine::cti_vm_lazyLinkCall):
        * VM/Machine.h:
        * masm/X86Assembler.h:
        (JSC::JITCodeBuffer::isAligned):
        (JSC::X86Assembler::):
        (JSC::X86Assembler::align):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::JSGlobalData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38286 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 11 Nov 2008 03:51:24 +0000 (03:51 +0000)]
2008-11-10  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Antti Koivisto.

        - Make Vector::clear() release the Vector's memory (1MB savings on membuster)
        https://bugs.webkit.org/show_bug.cgi?id=22170

        * wtf/Vector.h:
        (WTF::VectorBufferBase::deallocateBuffer): Set capacity to 0 as
        well as size, otherwise shrinking capacity to 0 can fail to reset
        the capacity and thus cause a future crash.
        (WTF::Vector::~Vector): Shrink size not capacity; we only need
        to call destructors, the buffer will be freed anyway.
        (WTF::Vector::clear): Change this to shrinkCapacity(0), not just shrink(0).
        (WTF::::shrinkCapacity): Use shrink() instead of resize() for case where
        the size is greater than the new capacity, to work with types that have no
        default constructor.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38285 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Tue, 11 Nov 2008 03:38:18 +0000 (03:38 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Maciej Stachowiak.

        Split multiple definitions into separate lines.

        * VM/CTI.cpp:
        (JSC::CTI::compileBinaryArithOp):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38284 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 01:26:35 +0000 (01:26 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Move renewGState to the base class.

        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView renewGState]):
        * Plugins/WebNetscapePluginView.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38283 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 01:23:36 +0000 (01:23 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Move start, stop and all the related methods down to WebBaseNetscapePluginView.

        * Plugins/WebBaseNetscapePluginView.h:
        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView updateAndSetWindow]):
        (-[WebBaseNetscapePluginView addWindowObservers]):
        (-[WebBaseNetscapePluginView removeWindowObservers]):
        (-[WebBaseNetscapePluginView start]):
        (-[WebBaseNetscapePluginView stop]):
        (-[WebBaseNetscapePluginView viewWillMoveToWindow:]):
        (-[WebBaseNetscapePluginView viewWillMoveToSuperview:]):
        (-[WebBaseNetscapePluginView viewDidMoveToWindow]):
        (-[WebBaseNetscapePluginView viewWillMoveToHostWindow:]):
        (-[WebBaseNetscapePluginView viewDidMoveToHostWindow]):
        (-[WebBaseNetscapePluginView windowWillClose:]):
        (-[WebBaseNetscapePluginView windowBecameKey:]):
        (-[WebBaseNetscapePluginView windowResignedKey:]):
        (-[WebBaseNetscapePluginView windowDidMiniaturize:]):
        (-[WebBaseNetscapePluginView windowDidDeminiaturize:]):
        (-[WebBaseNetscapePluginView loginWindowDidSwitchFromUser:]):
        (-[WebBaseNetscapePluginView loginWindowDidSwitchToUser:]):
        (-[WebBaseNetscapePluginView preferencesHaveChanged:]):
        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38282 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 01:13:29 +0000 (01:13 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Move rightMouseDown, rightMouseUp and sendActivateEvent to the base plugin view.
        Add stubs for createPlugin, loadStream, shouldStop and destroyPlugin.

        * Plugins/WebBaseNetscapePluginView.h:
        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView windowFocusChanged:]):
        (-[WebBaseNetscapePluginView createPlugin]):
        (-[WebBaseNetscapePluginView loadStream]):
        (-[WebBaseNetscapePluginView destroyPlugin]):
        (-[WebBaseNetscapePluginView sendActivateEvent:]):
        (-[WebBaseNetscapePluginView rightMouseDown:]):
        (-[WebBaseNetscapePluginView rightMouseUp:]):
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView windowFocusChanged:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38281 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 00:51:26 +0000 (00:51 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Factor plug-in type specific code out to three new methods, createPlugin, destroyPlugin and loadStream.

        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView createPlugin]):
        (-[WebNetscapePluginView loadStream]):
        (-[WebNetscapePluginView start]):
        (-[WebNetscapePluginView shouldStop]):
        (-[WebNetscapePluginView destroyPlugin]):
        (-[WebNetscapePluginView stop]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38280 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 00:38:21 +0000 (00:38 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Fix Tiger build.

        * Plugins/WebNetscapePluginPackage.m:
        (-[WebNetscapePluginPackage load]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38279 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Tue, 11 Nov 2008 00:29:54 +0000 (00:29 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        Move even more code down to WebBaseNetscapePluginView, get rid of some unnecessary methods.

        * Plugins/WebBaseNetscapePluginView.h:
        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView focusChanged]):
        (-[WebBaseNetscapePluginView visibleRect]):
        (-[WebBaseNetscapePluginView acceptsFirstResponder]):
        (-[WebBaseNetscapePluginView setHasFocus:]):
        (-[WebBaseNetscapePluginView becomeFirstResponder]):
        (-[WebBaseNetscapePluginView resignFirstResponder]):
        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView focusChanged]):
        (-[WebNetscapePluginView tellQuickTimeToChill]):
        (-[WebNetscapePluginView updateAndSetWindow]):
        (-[WebNetscapePluginView start]):
        (-[WebNetscapePluginView stop]):
        (-[WebNetscapePluginView viewWillMoveToWindow:]):
        (-[WebNetscapePluginView createPluginScriptableObject]):
        (-[WebNetscapePluginView pluginView:receivedData:]):
        (-[WebNetscapePluginView pluginView:receivedError:]):
        (-[WebNetscapePluginView pluginViewFinishedLoading:]):
        (-[WebNetscapePluginView inputContext]):
        (-[WebNetscapePluginView hasMarkedText]):
        (-[WebNetscapePluginView insertText:]):
        (-[WebNetscapePluginView markedRange]):
        (-[WebNetscapePluginView selectedRange]):
        (-[WebNetscapePluginView setMarkedText:selectedRange:]):
        (-[WebNetscapePluginView unmarkText]):
        (-[WebNetscapePluginView validAttributesForMarkedText]):
        (-[WebNetscapePluginView attributedSubstringFromRange:]):
        (-[WebNetscapePluginView characterIndexForPoint:]):
        (-[WebNetscapePluginView doCommandBySelector:]):
        (-[WebNetscapePluginView firstRectForCharacterRange:]):
        (-[WebNetscapePluginView _viewHasMoved]):
        (-[WebNetscapePluginView _redeliverStream]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38278 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Mon, 10 Nov 2008 23:43:18 +0000 (23:43 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Move timer handling code down to WebBaseNetscapePluginView.

        * Plugins/WebBaseNetscapePluginView.h:
        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView stopTimers]):
        (-[WebBaseNetscapePluginView startTimers]):
        (-[WebBaseNetscapePluginView restartTimers]):
        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView stopTimers]):
        (-[WebNetscapePluginView startTimers]):
        (-[WebNetscapePluginView scheduleTimerWithInterval:repeat:timerFunc:]):
        * WebView/WebHTMLView.mm:
        (-[WebHTMLView _resumeNullEventsForAllNetscapePlugins]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38277 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebKit:
andersca@apple.com [Mon, 10 Nov 2008 22:59:13 +0000 (22:59 +0000)]
WebKit:

2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Add WebHostedNetscapePluginView to the project.

        * WebKit.xcodeproj/project.pbxproj:

WebKit/mac:

2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Move a bunch of methods and ivars up to WebBaseNetscapePluginView.

        * Plugins/WebBaseNetscapePluginView.h:
        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView dealloc]):
        (-[WebBaseNetscapePluginView finalize]):
        (-[WebBaseNetscapePluginView removeTrackingRect]):
        (-[WebBaseNetscapePluginView resetTrackingRect]):
        (-[WebBaseNetscapePluginView dataSource]):
        (-[WebBaseNetscapePluginView webFrame]):
        (-[WebBaseNetscapePluginView webView]):
        (-[WebBaseNetscapePluginView currentWindow]):
        * Plugins/WebNetscapePluginEventHandlerCarbon.mm:
        (WebNetscapePluginEventHandlerCarbon::windowFocusChanged):
        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView sendEvent:isDrawRect:]):
        (-[WebNetscapePluginView sendActivateEvent:]):
        (-[WebNetscapePluginView restartTimers]):
        (-[WebNetscapePluginView setHasFocus:]):
        (-[WebNetscapePluginView mouseDown:]):
        (-[WebNetscapePluginView mouseUp:]):
        (-[WebNetscapePluginView mouseEntered:]):
        (-[WebNetscapePluginView mouseExited:]):
        (-[WebNetscapePluginView handleMouseMoved:]):
        (-[WebNetscapePluginView mouseDragged:]):
        (-[WebNetscapePluginView scrollWheel:]):
        (-[WebNetscapePluginView keyUp:]):
        (-[WebNetscapePluginView keyDown:]):
        (-[WebNetscapePluginView flagsChanged:]):
        (-[WebNetscapePluginView updateAndSetWindow]):
        (-[WebNetscapePluginView setWindowIfNecessary]):
        (-[WebNetscapePluginView start]):
        (-[WebNetscapePluginView stop]):
        (-[WebNetscapePluginView isStarted]):
        (-[WebNetscapePluginView dealloc]):
        (-[WebNetscapePluginView finalize]):
        (-[WebNetscapePluginView drawRect:]):
        (-[WebNetscapePluginView windowBecameKey:]):
        (-[WebNetscapePluginView preferencesHaveChanged:]):
        (-[WebNetscapePluginView evaluateJavaScriptPluginRequest:]):
        (-[WebNetscapePluginView webFrame:didFinishLoadWithReason:]):
        (-[WebNetscapePluginView scheduleTimerWithInterval:repeat:timerFunc:]):
        (-[WebNetscapePluginView _viewHasMoved]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38276 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Mon, 10 Nov 2008 22:27:34 +0000 (22:27 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker and Mark Rowe.

        Add a WebHostedNetscapePluginView class.

        * Plugins/WebHostedNetscapePluginView.h: Added.
        * Plugins/WebHostedNetscapePluginView.mm: Added.

        * Plugins/WebNetscapePluginPackage.h:
        Use the right define.

        * Plugins/WebPluginDatabase.mm:
        (-[WebPluginDatabase removePluginInstanceViewsFor:]):
        Check for WebBaseNetscapePluginView.

        (-[WebPluginDatabase destroyAllPluginInstanceViews]):
        Check for WebBaseNetscapePluginView.

        * WebCoreSupport/WebFrameLoaderClient.mm:
        (NetscapePluginWidget::NetscapePluginWidget):
        (NetscapePluginWidget::handleEvent):
        Use WebBaseNetscapePluginView.

        (netscapePluginViewClass):
        New function that returns the right netscape plugin view type to use.

        (WebFrameLoaderClient::createPlugin):
        Get the right class.

        * WebKitPrefix.h:
        Prefix the #define with WTF_.

        * WebView/WebHTMLView.mm:
        (-[NSArray _web_makePluginViewsPerformSelector:withObject:]):
        Check for WebBaseNetscapePluginView.

        * WebView/WebView.mm:
        Remove an unnecessary include.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38275 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFix Bug 22161: Assertion failure in RenderThemeWin::systemColor when loading microsof...
aroben@apple.com [Mon, 10 Nov 2008 22:12:57 +0000 (22:12 +0000)]
Fix Bug 22161: Assertion failure in RenderThemeWin::systemColor when loading microsoftpdc.com

WebCore:

        Fix Bug 22161: Assertion failure in RenderThemeWin::systemColor when
        loading microsoftpdc.com

        <https://bugs.webkit.org/show_bug.cgi?id=22161>

        Reviewed by Dan Bernstein.

        Test: fast/css/outline-invert-assertion.html

        * rendering/RenderThemeWin.cpp:
        (WebCore::RenderThemeWin::systemColor): If we can't determine a good
        Win32 system color to use, defer to RenderTheme instead of asserting.
        This matches what RenderThemeMac does.

LayoutTests:

        Test for Bug 22161: Assertion failure in RenderThemeWin::systemColor
        when loading microsoftpdc.com

        <https://bugs.webkit.org/show_bug.cgi?id=22161>

        Note that this test would only fail if DumpRenderTree used
        RenderThemeWin, which it doesn't. I did test that locally, though.

        Reviewed by Dan Bernstein.

        * fast/css/outline-invert-assertion-expected.txt: Added.
        * fast/css/outline-invert-assertion.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38274 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
bdakin@apple.com [Mon, 10 Nov 2008 21:51:22 +0000 (21:51 +0000)]
WebCore:

2008-11-10  Justin Garcia  <justin.garcia@apple.com>

        Reviewed by Beth Dakin.

        <rdar://problem/4037481> REGRESSION (Mail): pasting quoted text
        into quoted text yields double-quoting

        Don't nest inserted content in Mail blockquotes.  Perform a
        BreakBlockquoteCommand if we're in
        one and insert the incoming fragment between the split blockquotes.

        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::shouldMerge): Renamed from ->
        source and to -> destination.
        (WebCore::ReplaceSelectionCommand::doApply):

LayoutTests:

2008-11-10  Beth Dakin  <bdakin@apple.com>

        Rubber stamped by Justin Garcia.

        New tests for <rdar://problem/4037481> REGRESSION (Mail): pasting
        quoted text into quoted text yields double-quoting
        * editing/pasteboard/paste-blockquote-into-blockquote-2.html:
        Added.
        * editing/pasteboard/paste-blockquote-into-blockquote.html: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-2-expected.checksum: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-2-expected.png: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-
        blockquote-2-expected.txt: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-
        expected.checksum: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-
        expected.png: Added.
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-
        expected.txt: Added.

        Test with new results because of this patch:
        * platform/mac/editing/pasteboard/merge-end-blockquote-
        expected.checksum:
        * platform/mac/editing/pasteboard/merge-end-blockquote-
        expected.png:
        * platform/mac/editing/pasteboard/merge-end-blockquote-
        expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38273 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoCommit the current configuration of build.webkit.org.
mrowe@apple.com [Mon, 10 Nov 2008 21:47:26 +0000 (21:47 +0000)]
Commit the current configuration of build.webkit.org.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38272 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoCommit the current configuration of build.webkit.org.
mrowe@apple.com [Mon, 10 Nov 2008 21:45:32 +0000 (21:45 +0000)]
Commit the current configuration of build.webkit.org.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38271 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Mon, 10 Nov 2008 21:33:22 +0000 (21:33 +0000)]
2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        Don't use individual ivars for each plug-in vtable function. Instead, get them from the plugin package.

        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView initWithFrame:pluginPackage:URL:baseURL:MIMEType:attributeKeys:attributeValues:loadManually:DOMElement:]):
        * Plugins/WebNetscapePluginPackage.h:
        * Plugins/WebNetscapePluginPackage.m:
        (-[WebNetscapePluginPackage load]):
        (-[WebNetscapePluginPackage _unloadWithShutdown:]):
        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView sendEvent:isDrawRect:]):
        (-[WebNetscapePluginView setWindowIfNecessary]):
        (-[WebNetscapePluginView start]):
        (-[WebNetscapePluginView initWithFrame:pluginPackage:URL:baseURL:MIMEType:attributeKeys:attributeValues:loadManually:DOMElement:]):
        (-[WebNetscapePluginView createPluginScriptableObject]):
        (-[WebNetscapePluginView evaluateJavaScriptPluginRequest:]):
        (-[WebNetscapePluginView webFrame:didFinishLoadWithReason:]):
        (-[WebNetscapePluginView loadPluginRequest:]):
        (-[WebNetscapePluginView _createPlugin]):
        (-[WebNetscapePluginView _destroyPlugin]):
        (-[WebNetscapePluginView _printedPluginBitmap]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38270 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebKit:
andersca@apple.com [Mon, 10 Nov 2008 21:00:28 +0000 (21:00 +0000)]
WebKit:

2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Add a new WebBaseNetscapePluginView class.

        * WebKit.xcodeproj/project.pbxproj:

WebKit/mac:

2008-11-10  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Add a new WebBaseNetscapePluginView class.

        * Plugins/WebBaseNetscapePluginView.h: Added.
        * Plugins/WebBaseNetscapePluginView.mm: Added.
        (-[WebBaseNetscapePluginView initWithFrame:pluginPackage:URL:baseURL:MIMEType:attributeKeys:attributeValues:loadManually:DOMElement:]):
        (-[WebBaseNetscapePluginView setAttributeKeys:andValues:]):
        (-[WebBaseNetscapePluginView handleMouseMoved:]):
        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView cut:]):
        (-[WebNetscapePluginView copy:]):
        (-[WebNetscapePluginView paste:]):
        (-[WebNetscapePluginView selectAll:]):
        (-[WebNetscapePluginView start]):
        (-[WebNetscapePluginView initWithFrame:pluginPackage:URL:baseURL:MIMEType:attributeKeys:attributeValues:loadManually:DOMElement:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38269 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2008-11-10 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Mon, 10 Nov 2008 20:25:35 +0000 (20:25 +0000)]
2008-11-10  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Geoff Garen.

        Bug 22162: Remove cachedValueGetter from the JavaScriptCore API implementation
        <https://bugs.webkit.org/show_bug.cgi?id=22162>

        There is no more need for the cachedValueGetter hack now that we have
        PropertySlot::setValue(), so we should remove it.

        * API/JSCallbackObject.h:
        * API/JSCallbackObjectFunctions.h:
        (JSC::::getOwnPropertySlot):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@38268 268f45cc-cd09-0410-ab3c-d52691b4dbfc