Source/WebCore/ChangeLog

   1 2011-09-30  Dan Bernstein  <mitz@apple.com>
   2
   3         <rdar://problem/10208291> first-letter in generated before content with display: table is not updated
   4
   5         Reviewed by Sam Weinig.
   6
   7         Test: fast/css-generated-content/first-letter-in-nested-before.html
   8
   9         * rendering/RenderObjectChildList.cpp:
  10         (WebCore::RenderObjectChildList::updateBeforeAfterContent): Preserve the style type of anonymous
  11         containers around the generated content when updating their style. This preserves the first-letter-ness
  12         of the inline wrapping the first letter, which allows updateFirstLetter() to identify it later.
  13
  14 2011-09-30  Cary Clark  <caryclark@google.com>
  15
  16         Inset focus ring (Skia on Mac)
  17         https://bugs.webkit.org/show_bug.cgi?id=69166
  18         http://code.google.com/p/chromium/issues/detail?id=97956
  19
  20         This focus ring was drawn too large and was left tracks
  21         in subsequent invalidations. Make it more closely match
  22         the CG version.
  23
  24         Reviewed by Stephen White.
  25
  26         No new tests. This platform is not enabled.
  27
  28         * platform/graphics/skia/GraphicsContextSkia.cpp:
  29         (WebCore::getFocusRingOutset):
  30         (WebCore::GraphicsContext::drawFocusRing):
  31
  32 2011-09-30  Sheriff Bot  <webkit.review.bot@gmail.com>
  33
  34         Unreviewed, rolling out r96422.
  35         http://trac.webkit.org/changeset/96422
  36         https://bugs.webkit.org/show_bug.cgi?id=69170
  37
  38         Broke Windows build and kling wanted to roll it out as well
  39         (Requested by rniwa on #webkit).
  40
  41         * rendering/InlineBox.cpp:
  42         (WebCore::InlineBox::prevOnLineExists):
  43         * rendering/InlineBox.h:
  44         (WebCore::InlineBox::InlineBox):
  45
  46 2011-09-30  David Hyatt  <hyatt@apple.com>
  47
  48         https://bugs.webkit.org/show_bug.cgi?id=69167
  49
  50         [CSS3 Regions] Fix some bugs in the computation of the logical left and width offsets
  51         for blocks in variable width regions. Specifically, this patch addresses a minor bug
  52         with overflow:hidden/scroll/auto block placement and with RTL directionality.
  53
  54         Reviewed by Sam Weinig.
  55
  56         Added five new tests in fast/regions.
  57
  58         * rendering/RenderBlock.cpp:
  59         (WebCore::RenderBlock::logicalRightOffsetForContent):
  60         Avoid a double fetch of the border box in a given region by not calling
  61         logicalLeftOffsetForContent. No correctness change here. It's just cleaner.
  62
  63         * rendering/RenderBox.cpp:
  64         (WebCore::RenderBox::borderBoxRectInRegion):
  65         Change the shifted optimization to be cached for "self" rather than for "containing block".
  66         Rework the adjustment loop to handle changes in directionality and to fix correctness
  67         issues with RTL.
  68
  69         (WebCore::RenderBox::containingBlockLogicalWidthForContentInRegion):
  70         Don't double adjust if an object uses line width as the containing block width. In that
  71         case we already computed the line width in the region, so we don't need to apply any
  72         additional fixup.
  73
  74         (WebCore::RenderBox::renderBoxRegionInfo):
  75         * rendering/RenderBoxRegionInfo.h:
  76         (WebCore::RenderBoxRegionInfo::RenderBoxRegionInfo):
  77         (WebCore::RenderBoxRegionInfo::isShifted):
  78         Changed the shifted code to be on self rather than on containing block.
  79
  80         * rendering/RenderFlowThread.cpp:
  81         (WebCore::RenderFlowThread::layout):
  82         Refactor the layout code for RenderFlowThread to properly set up region rects assuming
  83         right alignment in RTL code. This change forces us to compute our maximum logical width
  84         first before looping through the regions again to set their content rects.
  85
  86 2011-09-30  Andreas Kling  <kling@webkit.org>
  87
  88         REGRESSION(r82611) InlineBox has 33 bits of bitset, causing alignment issues and extra memory use.
  89         https://bugs.webkit.org/show_bug.cgi?id=64914
  90
  91         Reviewed by Antti Koivisto.
  92
  93         Remove InlineBox::prevOnLineExists() and its two accompanying bitfields
  94         since nobody is using them anymore. nextOnLineExists() is still used by
  95         GTK+ accessibility code.
  96
  97         Also added a compile-time assertion to guard against future bloating of
  98         the InlineBox class.
  99
 100         * rendering/InlineBox.cpp:
 101         (WebCore::SameSizeAsInlineBox::~SameSizeAsInlineBox):
 102         * rendering/InlineBox.h:
 103         (WebCore::InlineBox::InlineBox):
 104
 105 2011-09-30  Pierre Rossi  <pierre.rossi@gmail.com>
 106
 107         [Qt] Build fix: Qt::escape is deprecated in Qt5
 108         https://bugs.webkit.org/show_bug.cgi?id=69162
 109
 110         Use QString::toHtmlEscaped in the Qt5 case.
 111
 112         Reviewed by Andreas Kling.
 113
 114         No new tests needed.
 115
 116         * WebCore.pro: adjust the include path accordingly
 117         in the v8 case.
 118
 119 2011-09-30  Ryosuke Niwa  <rniwa@webkit.org>
 120
 121         Remove unused ReplaceSelectionCommand::copyStyleToChildren
 122         https://bugs.webkit.org/show_bug.cgi?id=69153
 123
 124         Reviewed by Antonio Gomes.
 125
 126         Removed the function because it's no called anywhere.
 127
 128         * editing/ReplaceSelectionCommand.cpp:
 129         * editing/ReplaceSelectionCommand.h:
 130
 131 2011-09-30  Aaron Colwell  <acolwell@chromium.org>
 132
 133         Fix EnabledAtRuntime support for constants.
 134         https://bugs.webkit.org/show_bug.cgi?id=67311
 135
 136         Reviewed by Adam Barth.
 137
 138         * bindings/scripts/CodeGeneratorV8.pm:
 139         (GenerateImplementation):
 140
 141 2011-09-29  David Hyatt  <hyatt@apple.com>
 142
 143         https://bugs.webkit.org/show_bug.cgi?id=69043
 144
 145         [CSS3 Regions] Blocks split across regions with variable width need to size
 146         differently in each region.
 147
 148         This patch adds the capability for blocks to have variable width and positioning
 149         when split across regions. It is very rudimentary and so far only operates on basic
 150         normal flow RenderBlocks. Future patches will expand the support to cover other layout
 151         constructs like flexible boxes and tables.
 152
 153         Reviewed by Sam Weinig.
 154
 155         Added new tests in fast/regions and updated existing results.
 156
 157         * WebCore.xcodeproj/project.pbxproj:
 158         * rendering/HitTestResult.cpp:
 159         (WebCore::HitTestResult::HitTestResult):
 160         (WebCore::HitTestResult::operator=):
 161         * rendering/HitTestResult.h:
 162         (WebCore::HitTestResult::region):
 163         (WebCore::HitTestResult::setRegion):
 164         The hit test result now holds the active RenderRegion so that hit testing can adjust
 165         block widths and positions when hit testing their border boxes.
 166
 167         * rendering/PaintInfo.h:
 168         (WebCore::PaintInfo::PaintInfo):
 169         The paint info struct now holds the active RenderRegion so that painting can adjust
 170         block widths and positions when painting borders and backgrounds and shadows.
 171
 172         * rendering/RenderBlock.cpp:
 173         (WebCore::RenderBlock::layoutBlock):
 174         If our width ever changes, we invalidate all of our cached RenderBoxRegionInfo in all
 175         regions that we span.
 176
 177         (WebCore::RenderBlock::clearRenderBoxRegionInfo):
 178         Called to clear out our cached region-specific information in all regions that we span.
 179
 180         (WebCore::RenderBlock::borderBoxRectInRegionAtPosition):
 181         Returns the borderBoxRect for the region at the specified vertical offset. This rect
 182         can be both shifted horizontally and have a different width from our original border rect.
 183
 184         (WebCore::RenderBlock::logicalLeftOffsetForContent):
 185         (WebCore::RenderBlock::logicalRightOffsetForContent):
 186         Modified to call borderBoxRectInRegionAtPosition so that lines will fit inside
 187         the content rect of the border box rect for the specific region.
 188
 189         * rendering/RenderBlock.h:
 190         (WebCore::RenderBlock::logicalRightOffsetForContent):
 191         Modified to call logicalLeftOffsetForContent since it's cleaner.
 192
 193         * rendering/RenderBox.cpp:
 194         (WebCore::RenderBox::borderBoxRectInRegion):
 195         Called to compute the border box rect in a specific region. The result is cached in a HashMap
 196         in the region itself so that subsequent lookups are fast.
 197
 198         (WebCore::RenderBox::nodeAtPoint):
 199         Modified to use the border box rect in the active region when hit testing the backgrounds of boxes.
 200
 201         (WebCore::RenderBox::paintBoxDecorations):
 202         Modified to use the border box rect in the active region when painting the backgrounds of boxes.
 203
 204         (WebCore::RenderBox::containingBlockLogicalWidthForContent):
 205         (WebCore::RenderBox::containingBlockLogicalWidthForContentInRegion):
 206         (WebCore::RenderBox::computeLogicalWidth):
 207         (WebCore::RenderBox::computeLogicalWidthInRegion):
 208         Helpers for computing logical widths and margins in a specific region. The result is then cached
 209         in a HashMap in the region.
 210
 211         (WebCore::RenderBox::renderBoxRegionInfo):
 212         The function for obtaining the region-specific information for a given box.
 213
 214         * rendering/RenderBox.h:
 215         (WebCore::RenderBox::borderBoxRectInRegion):
 216         New function for returning the border box rect in a given region.
 217
 218         * rendering/RenderBoxModelObject.cpp:
 219         (WebCore::RenderBoxModelObject::paintFillLayerExtended):
 220         Make sure mask painting honors the active region.
 221
 222         * rendering/RenderBoxRegionInfo.h: Added.
 223         (WebCore::RenderBoxRegionInfo::RenderBoxRegionInfo):
 224         (WebCore::RenderBoxRegionInfo::logicalLeft):
 225         (WebCore::RenderBoxRegionInfo::logicalWidth):
 226         (WebCore::RenderBoxRegionInfo::containingBlockChainIsShifted):
 227         New class held by RenderRegions that caches box-specific info for a given region. This will eventually
 228         expand to include custom styles as well, but for now it is limited to a new logical left and a new
 229         logical width, along with a bit for optimizing accumulated shifting when painting/hit testing to avoid
 230         too much groveling up the containing block chain.
 231
 232         * rendering/RenderFlowThread.cpp:
 233         (WebCore::RenderFlowThread::RenderFlowThread):
 234         Removed the region fitting optimization, since eventually everyone is going to care, and there's no reason
 235         to limit it with the results now being cached.
 236
 237         (WebCore::RenderFlowThread::layout):
 238         All box-specific region information is always cleared whenever the regions are invalidated.
 239
 240         (WebCore::RenderFlowThread::computeLogicalWidth):
 241         Modified to set up RenderBoxRegionInfo for the flow thread in all regions.
 242
 243         (WebCore::RenderFlowThread::paintIntoRegion):
 244         (WebCore::RenderFlowThread::hitTestRegion):
 245         Modified to take the region instead of the region's rectangle so that painting and hit testing of the flow
 246         thread layer tree can properly receive the active region.
 247
 248         (WebCore::RenderFlowThread::removeRenderBoxRegionInfo):
 249         Helper for removing a box's information from all regions. For now it grovels through every region, so eventually
 250         we may want to have a cache of the start/end regions for a given box somewhere.
 251
 252         * rendering/RenderFlowThread.h:
 253         Removal of the region fitting stuff.
 254
 255         * rendering/RenderLayer.cpp:
 256         (WebCore::RenderLayer::paint):
 257         (WebCore::RenderLayer::paintOverlayScrollbars):
 258         (WebCore::RenderLayer::paintLayer):
 259         (WebCore::RenderLayer::paintList):
 260         (WebCore::RenderLayer::paintPaginatedChildLayer):
 261         (WebCore::RenderLayer::paintChildLayerIntoColumns):
 262         * rendering/RenderLayer.h:
 263         * rendering/RenderLayerBacking.cpp:
 264         (WebCore::RenderLayerBacking::paintIntoLayer):
 265         Modified to pass the current region down through painting functions so that it is known at paint time.
 266
 267         * rendering/RenderObject.cpp:
 268         (WebCore::RenderObject::RenderObject):
 269         Added a new bit to RenderObjects, inRenderFlowThread(), so that it is quick to determine whether or not an
 270         object needs flow thread special casing.
 271
 272         (WebCore::RenderObject::enclosingRenderFlowThread):
 273         Modified enclosingRenderFlowThread to be able to quickly return 0 if the object is not in a flow thread.
 274
 275         (WebCore::RenderObject::containerForRepaint):
 276         Same.
 277
 278         * rendering/RenderObject.h:
 279         (WebCore::RenderObject::setParent):
 280         setParent now updates inRenderFlowThread() state.
 281
 282         (WebCore::RenderObject::inRenderFlowThread):
 283         (WebCore::RenderObject::setInRenderFlowThread):
 284         Adding the new bit.
 285
 286         * rendering/RenderObjectChildList.cpp:
 287         (WebCore::RenderObjectChildList::removeChildNode):
 288         Make sure when an object is removed from a RenderFlowThread that it deletes its box-specific information
 289         in all regions.
 290
 291         * rendering/RenderRegion.cpp:
 292         (WebCore::RenderRegion::~RenderRegion):
 293         (WebCore::RenderRegion::paintReplaced):
 294         (WebCore::RenderRegion::nodeAtPoint):
 295         (WebCore::RenderRegion::renderBoxRegionInfo):
 296         (WebCore::RenderRegion::setRenderBoxRegionInfo):
 297         (WebCore::RenderRegion::removeRenderBoxRegionInfo):
 298         (WebCore::RenderRegion::deleteAllRenderBoxRegionInfo):
 299         (WebCore::RenderRegion::matchesRenderFlowThreadLogicalWidth):
 300         * rendering/RenderRegion.h:
 301         Added the new HashMap for holding box-specific region information. Also added an additional optimization to
 302         check if a specific region matches the overall width of the RenderFlowThread. If it does, we don't need to
 303         cache box-specific information for that region.
 304
 305         * rendering/RenderReplica.cpp:
 306         (WebCore::RenderReplica::paint):
 307         * rendering/RenderScrollbarPart.cpp:
 308         (WebCore::RenderScrollbarPart::paintIntoRect):
 309         * rendering/svg/SVGImageBufferTools.cpp:
 310         (WebCore::SVGImageBufferTools::renderSubtreeToImageBuffer):
 311         Modified the PaintInfo construction to include the region argument. It's not optional so that callers will
 312         have to consider it (since masks illustrated this is necessary).
 313
 314 2011-09-30  Tim Horton  <timothy_horton@apple.com>
 315
 316         feBlend uses a table of function pointers which reduces inlineability inside the main loop
 317         https://bugs.webkit.org/show_bug.cgi?id=69154
 318         <rdar://problem/10215221>
 319
 320         Reviewed by Darin Adler.
 321
 322         Don't use a table of function pointers inside the feBlend inner loop, instead
 323         use switch and inline functions, bringing a 20% performance gain across the
 324         board to feBlend.
 325
 326         No new tests, minor performance improvement.
 327
 328         * platform/graphics/filters/FEBlend.cpp:
 329         (WebCore::normal):
 330         (WebCore::multiply):
 331         (WebCore::screen):
 332         (WebCore::darken):
 333         (WebCore::lighten):
 334         (WebCore::FEBlend::apply):
 335
 336 2011-09-30  Mark Hahnenberg  <mhahnenberg@apple.com>
 337
 338         Add getCallData to MethodTable in ClassInfo
 339         https://bugs.webkit.org/show_bug.cgi?id=69024
 340
 341         Reviewed by Sam Weinig.
 342
 343         No new tests.
 344
 345         * WebCore.exp.in:
 346
 347         Changed getCallData from private to protected to allow subclasses who
 348         don't override getCallData themselves to reference it in their own
 349         method tables when calling the CREATE_METHOD_TABLE macro.
 350         * bridge/runtime_method.h:
 351
 352 2011-09-30  Ned Holbrook  <nholbrook@apple.com>
 353
 354         Complex spaces with synthetic bold are too wide
 355         https://bugs.webkit.org/show_bug.cgi?id=69033
 356
 357         Reviewed by Dan Bernstein.
 358
 359         Test: fast/text/complex-synthetic-bold-space-width.html
 360
 361         * platform/graphics/mac/ComplexTextController.cpp:
 362         (WebCore::ComplexTextController::adjustGlyphsAndAdvances): Subtract synthetic bold offset from spaceWidth.
 363         * platform/graphics/win/UniscribeController.cpp:
 364         (WebCore::UniscribeController::shapeAndPlaceItem): Ditto.
 365
 366 2011-09-30  Pavel Feldman  <pfeldman@google.com>
 367
 368         Web Inspector: [chromium] expose inspector protocol version to the embedder.
 369         https://bugs.webkit.org/show_bug.cgi?id=69092
 370
 371         This change also introduces major/minor inspector version components.
 372
 373         Reviewed by Yury Semikhatsky.
 374
 375         * CMakeLists.txt:
 376         * CodeGenerators.pri:
 377         * DerivedSources.make:
 378         * GNUmakefile.am:
 379         * GNUmakefile.list.am:
 380         * WebCore.gyp/WebCore.gyp:
 381         * WebCore.gypi:
 382         * inspector/Inspector-0.1.json: Renamed from Source/WebCore/inspector/Inspector.draft-01.json.
 383         * inspector/Inspector.json:
 384         * inspector/generate-inspector-protocol-version: Renamed from Source/WebCore/inspector/validate-protocol-compatibility.
 385
 386 2011-09-29  Antti Koivisto  <antti@apple.com>
 387
 388         https://bugs.webkit.org/show_bug.cgi?id=69106
 389         Universal attribute selectors disable style sharing
 390
 391         Reviewed by Dave Hyatt.
 392
 393         Selectors of type [foo="bar"] ended up marking every element style with the affectedByAttributeSelectors bit
 394         rendering style sharing inoperative. This happens on http://www.whatwg.org/specs/web-apps/current-work/ for example.
 395
 396         Instead we now mark style with affectedByUncommonAttributeSelectors bit only if an attribute selector actually
 397         matches the element. Before sharing, we also check the current element against collected attribute rules.
 398         We can share the style if neither element was affected.
 399
 400         This speeds up style matching and applying ~15% on full HTML5 spec (=~0.7s). Sharing percentage goes from 0% to ~30%.
 401         Increased sharing should also save a substantial amount of memory.
 402
 403         * css/CSSSelector.h:
 404         (WebCore::CSSSelector::isAttributeSelector):
 405         * css/CSSStyleSelector.cpp:
 406         (WebCore::RuleData::containsUncommonAttributeSelector):
 407         (WebCore::collectSpecialRulesInDefaultStyle):
 408         (WebCore::assertNoSiblingRulesInDefaultStyle):
 409         (WebCore::CSSStyleSelector::CSSStyleSelector):
 410         (WebCore::CSSStyleSelector::matchRules):
 411         (WebCore::CSSStyleSelector::matchesRuleSet):
 412         (WebCore::CSSStyleSelector::canShareStyleWithElement):
 413         (WebCore::CSSStyleSelector::locateSharedStyle):
 414         (WebCore::CSSStyleSelector::styleForElement):
 415         (WebCore::selectorListContainsUncommonAttributeSelector):
 416         (WebCore::isCommonAttributeSelectorAttribute):
 417         (WebCore::containsUncommonAttributeSelector):
 418         (WebCore::RuleData::RuleData):
 419         (WebCore::collectFeaturesFromSelector):
 420         (WebCore::collectFeaturesFromList):
 421         * css/CSSStyleSelector.h:
 422         * css/SelectorChecker.cpp:
 423         (WebCore::SelectorChecker::checkOneSelector):
 424         * rendering/style/RenderStyle.cpp:
 425         (WebCore::RenderStyle::RenderStyle):
 426         * rendering/style/RenderStyle.h:
 427         (WebCore::InheritedFlags::affectedByUncommonAttributeSelectors):
 428         (WebCore::InheritedFlags::setAffectedByUncommonAttributeSelectors):
 429
 430 2011-09-30  James Robinson  <jamesr@chromium.org>
 431
 432         [chromium] Add WebKit API for sending input events to the compositor thread
 433         https://bugs.webkit.org/show_bug.cgi?id=69117
 434
 435         Reviewed by Darin Fisher.
 436
 437         Convert to using an externally-provided thread instead of one managed by CCThreadProxy internally. Most changes
 438         in these files are mechanically changing references from ccThread to s_ccThread to reflect the class-static
 439         nature of this pointer, the rest are removing the old thread creation codepath.
 440
 441         These changes are covered by the CC* unit tests and by the gpu layout tests when run with the
 442         use_threaded_compositing=1 gyp define.
 443
 444         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
 445         * platform/graphics/chromium/cc/CCLayerTreeHost.h:
 446         * platform/graphics/chromium/cc/CCThreadProxy.cpp:
 447         (WebCore::CCThreadProxy::setThread):
 448         (WebCore::CCThreadProxy::CCThreadProxy):
 449         (WebCore::CCThreadProxy::~CCThreadProxy):
 450         (WebCore::CCThreadProxy::compositeAndReadback):
 451         (WebCore::CCThreadProxy::finishAllRendering):
 452         (WebCore::CCThreadProxy::initializeLayerRenderer):
 453         (WebCore::CCThreadProxy::setNeedsCommit):
 454         (WebCore::CCThreadProxy::setNeedsCommitAndRedraw):
 455         (WebCore::CCThreadProxy::setNeedsRedraw):
 456         (WebCore::CCThreadProxy::start):
 457         (WebCore::CCThreadProxy::stop):
 458         (WebCore::CCThreadProxy::beginFrameAndCommit):
 459         (WebCore::CCThreadProxy::scheduleDrawTaskOnCCThread):
 460         * platform/graphics/chromium/cc/CCThreadProxy.h:
 461
 462 2011-09-29  Fady Samuel  <fsamuel@chromium.org>
 463
 464         Fixed Hit testing on Framesets when Frameset is transformed
 465         https://bugs.webkit.org/show_bug.cgi?id=67740
 466
 467         Reviewed by Simon Fraser.
 468
 469         Tests: fast/frames/frame-set-rotation-hit.html
 470                fast/frames/frame-set-scaling-hit.html
 471
 472         * rendering/RenderFrameSet.cpp:
 473         (WebCore::RenderFrameSet::userResize):
 474
 475 2011-09-29  Dirk Pranke  <dpranke@chromium.org>
 476
 477         fix gyp warnings in chromium for renamed files
 478         https://bugs.webkit.org/show_bug.cgi?id=69116
 479
 480         Reviewed by Adam Barth.
 481
 482         * WebCore.gypi:
 483
 484 2011-09-29  Martin Robinson  <mrobinson@igalia.com>
 485
 486         [Freetype] Some text in Planet GNOME renders in the wrong place
 487         https://bugs.webkit.org/show_bug.cgi?id=69099
 488
 489         Reviewed by Dirk Schulze.
 490
 491         Test: platform/gtk/fonts/synthetic-oblique-positioning.html
 492
 493         Fold the oblique transform into the font matrix itself rather than transforming
 494         the CTM of the context at render time. Not only does this fix the issue, it
 495         prevents unnecessary work on every paint.
 496
 497         * platform/graphics/cairo/FontCairo.cpp:
 498         No longer set the synthetic oblique transformation matrix on the context
 499         when rendering the text. Instead it is now folded into the TM of the font
 500         itself.
 501         * platform/graphics/freetype/FontPlatformDataFreeType.cpp:
 502         (WebCore::FontPlatformData::initializeWithFontFace): Fold the oblique transform into the scaled font.
 503         * platform/graphics/win/FontPlatformDataCairoWin.cpp:
 504         (WebCore::FontPlatformData::FontPlatformData): Ditto.
 505
 506 2011-09-29  Dan Bernstein  <mitz@apple.com>
 507
 508         Follow-up fix for <rdar://problem/10191243> Glyph variants (line final swashes) appear where they should not
 509
 510         Reviewed by Simon Fraser.
 511
 512         * platform/graphics/cocoa/FontPlatformDataCocoa.mm:
 513         (WebCore::FontPlatformData::ctFont): Check for Hoefler Text Black Italic as well.
 514
 515 2011-09-29  Cary Clark  <caryclark@google.com>
 516
 517         Enable LCD text in Skia on Mac
 518         https://bugs.webkit.org/show_bug.cgi?id=68734
 519
 520         Reviewed by Stephen White.
 521
 522         No new tests. Existing layout tests are generated
 523         with LCD text disabled for pixel comparisons.
 524
 525         Duplicate the logic in FontMac.mm to pass settings
 526         for antialiasing and smoothing. Also disable smoothing
 527         for DumpRenderTree.
 528
 529         * platform/graphics/skia/FontSkia.cpp:
 530         (WebCore::setupPaint):
 531         (WebCore::Font::drawGlyphs):
 532
 533 2011-09-29  Varun Jain  <varunjain@google.com>
 534
 535          Implement flick gesture in Chromium Gesture Recognizer
 536          https://bugs.webkit.org/show_bug.cgi?id=67930
 537
 538          Reviewed by Dimitri Glazkov.
 539
 540          Test: Source/WebKit/chromium/tests/InnerGestureRecognizerTest.cpp
 541
 542          * platform/chromium/GestureRecognizerChromium.cpp:
 543          (WebCore::GestureRecognizerChromium::isOverMinFlickSpeed):
 544          (WebCore::GestureRecognizerChromium::appendScrollGestureEnd):
 545          (WebCore::GestureRecognizerChromium::updateValues):
 546          (WebCore::GestureRecognizerChromium::scrollEnd):
 547          * platform/chromium/GestureRecognizerChromium.h:
 548
 549 2011-09-29  Noel Gordon  <noel.gordon@gmail.com>
 550
 551         [chromium] canvas.toDataURL("image/jpeg"): use libjpeg-turbo data swizzle
 552         https://bugs.webkit.org/show_bug.cgi?id=67402
 553
 554         Reviewed by Kenneth Russell.
 555
 556         libjpeg-turbo can read directly from the input data for the premultiplied BRGX
 557         (SkBitmap) encoding case.
 558
 559         No change in behavior, covered by existing canvas 2d/3d tests
 560         canvas/philip/tests/toDataURL.jpeg.alpha.html
 561         fast/canvas/webgl/premultiplyalpha-test.html
 562
 563         * platform/image-encoders/skia/JPEGImageEncoder.cpp:
 564         (WebCore::encodePixels): If JCS_EXTENSIONS is defined (libjpeg-turbo), swizzle
 565         the input BRGX pixels directly within libjpeg-turbo.
 566
 567 2011-09-29  Adam Barth  <abarth@webkit.org>
 568
 569         Fixenate the build.
 570
 571         * bindings/v8/V8DOMWindowShell.cpp:
 572         (WebCore::V8DOMWindowShell::namedItemAdded):
 573
 574 2011-09-29  Ryosuke Niwa  <rniwa@webkit.org>
 575
 576         Remove direct reads to m_firstNodeInserted and m_lastLeafInserted in ReplaceSelectionCommand
 577         https://bugs.webkit.org/show_bug.cgi?id=69023
 578
 579         Reviewed by Enrica Casucci.
 580
 581         Converted to removeUnrenderedTextNodesAtEnds to use InsertedNodes instead of m_firstNodeInserted
 582         and m_lastLeafInserted. Extracted the logic to update nodes as InsertedNodes::willRemoveNode
 583         (old one was renamed to InsertedNodes::willRemoveNodePreservingChildren).
 584
 585         Also extracted shouldPerformSmartReplace and addSpacesForSmartReplace out of doApply,
 586         and rewrote the logic to obtain endNode and startNode using startOfInsertedContent and
 587         endOfInsertedContent instead of m_firstNodeInserted and m_lastLeafInserted.
 588
 589         Finally, replaced the nullity checks of m_firstNodeInserted and m_lastLeafInserted in
 590         completeHTMLReplacement by nullity checks of start and end positions.
 591
 592         * dom/Node.cpp:
 593         (WebCore::Node::traversePreviousSibling): Added.
 594         * dom/Node.h:
 595         * editing/ReplaceSelectionCommand.cpp:
 596         (WebCore::ReplaceSelectionCommand::InsertedNodes::willRemoveNodePreservingChildren): Renamed from
 597         willRemoveNode.
 598         (WebCore::ReplaceSelectionCommand::InsertedNodes::willRemoveNode): Extracted from
 599         removeUnrenderedTextNodesAtEnds.
 600         (WebCore::ReplaceSelectionCommand::removeRedundantStylesAndKeepStyleSpanInline):
 601         (WebCore::ReplaceSelectionCommand::removeUnrenderedTextNodesAtEnds):
 602         (WebCore::ReplaceSelectionCommand::positionAtEndOfInsertedContent): Made const.
 603         (WebCore::ReplaceSelectionCommand::positionAtStartOfInsertedContent): Made const.
 604         (WebCore::ReplaceSelectionCommand::handleStyleSpans): Takes firstNodeInserted instead of directly
 605         accessing m_firstNodeInserted.
 606         (WebCore::ReplaceSelectionCommand::doApply):
 607         (WebCore::ReplaceSelectionCommand::shouldPerformSmartReplace):
 608         (WebCore::ReplaceSelectionCommand::addSpacesForSmartReplace):
 609         (WebCore::ReplaceSelectionCommand::completeHTMLReplacement):
 610         * editing/ReplaceSelectionCommand.h:
 611
 612 2011-09-29  Andreas Kling  <kling@webkit.org>
 613
 614         Shrink HTMLAnchorElement on 32-bit.
 615         https://bugs.webkit.org/show_bug.cgi?id=69094
 616
 617         Reviewed by Antti Koivisto.
 618
 619         * html/HTMLAnchorElement.h: Pack members into a bitfield.
 620
 621 2011-09-29  Dan Bernstein  <mitz@apple.com>
 622
 623         <rdar://problem/10191243> Glyph variants (line final swashes) appear where they should not
 624
 625         Reviewed by Simon Fraser.
 626
 627         Test: fast/text/line-initial-and-final-swashes.html
 628
 629         Hoefler Text Italic enables line-initial and -final swashes by default, so disable them. This
 630         change targets only this known-bad font rather than all fonts, because at least one font (Khmer MN)
 631         incorrectly claims to have the line-initial feature enabled, but disabling it actually does
 632         something different and bad (breaking some combining marks).
 633
 634         * platform/graphics/cocoa/FontPlatformDataCocoa.mm:
 635         (WebCore::createFeatureSettingDictionary): Added this helper function.
 636         (WebCore::cascadeToLastResortFontDescriptor): Deployed WTF_ARRAY_LENGTH().
 637         (WebCore::cascadeToLastResortAndDisableSwashesFontDescriptor): Added. Returns a font descriptor
 638         that, in addition to having a cascade list consisting of the last resort font, also has feature
 639         settings to disable line-initial and line-final swashes.
 640         (WebCore::FontPlatformData::ctFont): For Hoefler Text Italic, use cascadeToLastResortAndDisableSwashesFontDescriptor().
 641
 642 2011-09-29  Mark Hahnenberg  <mhahnenberg@apple.com>
 643
 644         Unreviewed: resetting baseline for code generator bindings
 645
 646         Unreviewed build fix
 647
 648         No new tests.
 649
 650         * bindings/scripts/test/JS/JSTestObj.cpp:
 651         * bindings/scripts/test/JS/JSTestObj.h:
 652
 653 2011-09-29  Nate Chapin  <japhet@chromium.org>
 654
 655         [V8, chromium] More logging to determine cause of a null
 656         v8::Context in V8DOMWindowShell::namedItemAdded().
 657
 658         https://bugs.webkit.org/show_bug.cgi?id=68099
 659
 660         Reviewed by Adam Barth.
 661
 662         * bindings/v8/V8DOMWindowShell.cpp:
 663         (WebCore::V8DOMWindowShell::namedItemAdded):
 664
 665 2011-09-29  Alexey Proskuryakov  <ap@apple.com>
 666
 667         https://bugs.webkit.org/show_bug.cgi?id=69040
 668         ScrollbarThemeComposite requires a ScrollView to draw scroll corner
 669
 670         Reviewed by Simon Fraser.
 671
 672         No new tests. This will be needed later.
 673
 674         * platform/ScrollbarThemeComposite.cpp:
 675         (WebCore::pageForScrollView):
 676         (WebCore::ScrollbarThemeComposite::paintScrollCorner):
 677         Use pageForScrollView() function which already existed in this file, used in another similar
 678         location.
 679
 680         * platform/chromium/FramelessScrollView.cpp:
 681         * platform/chromium/FramelessScrollView.h:
 682         * platform/gtk/ScrollbarThemeGtk.cpp:
 683         * platform/gtk/ScrollbarThemeGtk.h:
 684         * platform/wx/ScrollbarThemeWx.cpp:
 685         * platform/wx/ScrollbarThemeWx.h:
 686         All these overrides are no longer needed, ScrollbarThemeComposite will do the right thing.
 687
 688         * platform/qt/ScrollbarThemeQt.cpp: (WebCore::ScrollbarThemeQt::paintScrollCorner):
 689         Removed a special case for updatingControlTints phase. The same case is present in cross-platform
 690         code now, and Qt doesn't have any custom subclasses of ScrollableArea or ScrollView to need
 691         special handling.
 692         This was added in r37377 without a bug or much ChangeLog explanation.
 693
 694 2011-09-29  Mark Hahnenberg  <mhahnenberg@apple.com>
 695
 696         De-virtualize JSCell::visitChildrenVirtual and remove all other visitChildrenVirtual methods
 697         https://bugs.webkit.org/show_bug.cgi?id=68839
 698
 699         Reviewed by Geoffrey Garen.
 700
 701         No new tests.
 702
 703         Removed the remaining visitChildrenVirtual methods.  This patch completes the process of
 704         de-virtualizing visitChildren.
 705
 706         * WebCore.exp.in:
 707         * bindings/js/JSAttrCustom.cpp:
 708         * bindings/js/JSAudioContextCustom.cpp:
 709         * bindings/js/JSCSSRuleCustom.cpp:
 710         * bindings/js/JSCSSStyleDeclarationCustom.cpp:
 711         * bindings/js/JSCanvasRenderingContextCustom.cpp:
 712         * bindings/js/JSDOMGlobalObject.cpp:
 713         (WebCore::JSDOMGlobalObject::~JSDOMGlobalObject):
 714         (WebCore::JSDOMGlobalObject::finishCreation):
 715         * bindings/js/JSDOMGlobalObject.h:
 716         * bindings/js/JSDOMWindowCustom.cpp:
 717         * bindings/js/JSDOMWindowShell.cpp:
 718         * bindings/js/JSDOMWindowShell.h:
 719         * bindings/js/JSJavaScriptAudioNodeCustom.cpp:
 720         * bindings/js/JSMessageChannelCustom.cpp:
 721         * bindings/js/JSMessagePortCustom.cpp:
 722         * bindings/js/JSNamedNodeMapCustom.cpp:
 723         * bindings/js/JSNodeCustom.cpp:
 724         * bindings/js/JSNodeFilterCustom.cpp:
 725         * bindings/js/JSNodeIteratorCustom.cpp:
 726         * bindings/js/JSSVGElementInstanceCustom.cpp:
 727         * bindings/js/JSSharedWorkerCustom.cpp:
 728         * bindings/js/JSStyleSheetCustom.cpp:
 729         * bindings/js/JSTreeWalkerCustom.cpp:
 730         * bindings/js/JSWebGLRenderingContextCustom.cpp:
 731         * bindings/js/JSWorkerContextCustom.cpp:
 732         * bindings/js/JSXMLHttpRequestCustom.cpp:
 733         * bindings/js/JSXPathResultCustom.cpp:
 734         * bindings/scripts/CodeGeneratorJS.pm:
 735         (GenerateHeader):
 736         (GenerateImplementation):
 737         * bridge/qt/qt_instance.cpp:
 738         * bridge/qt/qt_runtime.cpp:
 739         * bridge/qt/qt_runtime.h:
 740
 741 2011-09-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 742
 743         [Qt] Fix build against Qt5 after refactor of widgets out of QtGUi
 744
 745         QWidget and friends now live in the QtWidgets library. We update
 746         includes in implementation files and private headers to us the
 747         non-module-prefixed path, and leave the lookup for the include
 748         path. For public headers we have to ifdef the includes as the
 749         user might now have the modules we need in his QT config.
 750
 751         Finally, QSGCanvas is no longer a QWidget but a QWindow, so we
 752         have to update our code and use windowHandle() for setting the
 753         parent relationships.
 754
 755         https://bugs.webkit.org/show_bug.cgi?id=68687
 756
 757         Reviewed by Andreas Kling.
 758
 759         * WebCore.pri:
 760         * platform/graphics/qt/GraphicsLayerQt.cpp:
 761
 762 2011-09-29  Sheriff Bot  <webkit.review.bot@gmail.com>
 763
 764         Unreviewed, rolling out r96340.
 765         http://trac.webkit.org/changeset/96340
 766         https://bugs.webkit.org/show_bug.cgi?id=69098
 767
 768         Caused 10 tests to crash in Debug (Requested by
 769         abarth|gardener on #webkit).
 770
 771         * accessibility/AXObjectCache.cpp:
 772         (WebCore::AXObjectCache::AXObjectCache):
 773         (WebCore::AXObjectCache::~AXObjectCache):
 774         (WebCore::AXObjectCache::remove):
 775         (WebCore::AXObjectCache::childrenChanged):
 776         * accessibility/AXObjectCache.h:
 777         * accessibility/AccessibilityMenuList.cpp:
 778         (WebCore::AccessibilityMenuList::childrenChanged):
 779         * accessibility/AccessibilityMenuList.h:
 780         * accessibility/AccessibilityMenuListPopup.cpp:
 781         (WebCore::AccessibilityMenuListPopup::childrenChanged):
 782         * accessibility/AccessibilityMenuListPopup.h:
 783         * accessibility/AccessibilityObject.h:
 784         (WebCore::AccessibilityObject::childrenChanged):
 785         * accessibility/AccessibilityRenderObject.cpp:
 786         (WebCore::startOfContinuations):
 787         (WebCore::AccessibilityRenderObject::updateAccessibilityRole):
 788         (WebCore::AccessibilityRenderObject::childrenChanged):
 789         * accessibility/AccessibilityRenderObject.h:
 790         * rendering/RenderObject.cpp:
 791         (WebCore::RenderObject::willBeDestroyed):
 792
 793 2011-09-29  Andreas Kling  <kling@webkit.org>
 794
 795         Shrink FontFallbackList.
 796         https://bugs.webkit.org/show_bug.cgi?id=69093
 797
 798         Reviewed by Antti Koivisto.
 799
 800         Reduce the size of FontFallbackList by one CPU word, decreasing memory
 801         consumption by 300 kB (on 64-bit) when loading the full HTML5 spec.
 802
 803         * platform/graphics/FontCache.h:
 804         * platform/graphics/FontCache.cpp:
 805         (WebCore::FontCache::generation):
 806
 807             Store the FontCache generation as an ushort rather than uint.
 808
 809         * platform/graphics/FontFallbackList.cpp:
 810         (WebCore::FontFallbackList::FontFallbackList):
 811         * platform/graphics/FontFallbackList.h:
 812
 813             Pack enum and bool members in a bitfield.
 814
 815 2011-09-29  Adam Barth  <abarth@webkit.org>
 816
 817         We should ignore the return value of GetRealNamedProperty
 818         https://bugs.webkit.org/show_bug.cgi?id=68840
 819
 820         Reviewed by Nate Chapin.
 821
 822         Instead of skipping the lookup process by using the result of real
 823         named property, we should just indiciate that it wasn't handled by the
 824         interceptor.
 825
 826         Test: http/tests/security/window-named-valueOf.html
 827
 828         * bindings/v8/custom/V8DOMWindowCustom.cpp:
 829         (WebCore::V8DOMWindow::namedPropertyGetter):
 830
 831 2011-09-29  Chris Fleizach  <cfleizach@apple.com>
 832
 833         ARIA live regions don't trigger notifications for elements that aren't in the AX tree
 834         https://bugs.webkit.org/show_bug.cgi?id=62289
 835
 836         If an ARIA Live region udpates an element that is not in the AX object cache, then the Live region
 837         notification is not sent. To fix this, the childrenChanged() method needs to actually create
 838         the appropriate objects, but since that method gets called during a render tree update, we've learned
 839         that it's generally not safe to create objects.
 840
 841         Instead a one shot timer can be fired that will update and create the necessary objects so that the
 842         correct notification can be sent.
 843
 844         Reviewed by Darin Adler.
 845
 846         Test: platform/mac/accessibility/aria-liveregion-without-element-access.html
 847
 848         * accessibility/AXObjectCache.cpp:
 849         (WebCore::AXObjectCache::AXObjectCache):
 850         (WebCore::AXObjectCache::~AXObjectCache):
 851         (WebCore::AXObjectCache::remove):
 852         (WebCore::AXObjectCache::childrenUpdateTimerFired):
 853         (WebCore::AXObjectCache::childrenChanged):
 854         * accessibility/AXObjectCache.h:
 855         * accessibility/AccessibilityMenuList.cpp:
 856         (WebCore::AccessibilityMenuList::childrenChanged):
 857         * accessibility/AccessibilityMenuList.h:
 858         * accessibility/AccessibilityMenuListPopup.cpp:
 859         (WebCore::AccessibilityMenuListPopup::childrenChanged):
 860         * accessibility/AccessibilityMenuListPopup.h:
 861         * accessibility/AccessibilityObject.h:
 862         (WebCore::AccessibilityObject::childrenChanged):
 863         * accessibility/AccessibilityRenderObject.cpp:
 864         (WebCore::startOfContinuations):
 865         (WebCore::AccessibilityRenderObject::updateAccessibilityRole):
 866         (WebCore::AccessibilityRenderObject::childrenChanged):
 867         * accessibility/AccessibilityRenderObject.h:
 868         * rendering/RenderObject.cpp:
 869         (WebCore::RenderObject::willBeDestroyed):
 870
 871 2011-09-29  Martin Robinson  <mrobinson@igalia.com>
 872
 873         [GTK] Dragging a selection does not produce a drag image
 874         https://bugs.webkit.org/show_bug.cgi?id=69064
 875
 876         Reviewed by Andreas Kling.
 877
 878         Moved Chromium's generic implementation of Frame::nodeImage and
 879         Frame::dragImageForSeletion to Frame.cpp and simply excluded via
 880         #ifdefs platforms that do not have their own implementation. Removed
 881         all empty implementations of these functions and Qt's which was
 882         functionally identical.
 883
 884         No new tests. Drag-and-drop drag image functionality changes
 885         are incredibly difficult to test in a non-manual way.
 886
 887         * CMakeListsEfl.txt: Delist FrameEfl.cpp.
 888         * GNUmakefile.list.am: Delist FrameGtk.cpp.
 889         * WebCore.gypi: Delist removed Frame*.cpp.
 890         * WebCore.pro: Delist FrameQt.cpp.
 891         * page/Frame.cpp: Moved FrameChromium implementation here.
 892         (WebCore::ScopedFramePaintingState::ScopedFramePaintingState):
 893         (WebCore::ScopedFramePaintingState::~ScopedFramePaintingState):
 894         (WebCore::Frame::nodeImage):
 895         (WebCore::Frame::dragImageForSelection):
 896         * page/chromium/FrameChromium.cpp: Removed.
 897         * page/efl/FrameEfl.cpp: Removed.
 898         * page/gtk/FrameGtk.cpp: Removed.
 899         * page/qt/FrameQt.cpp: Removed.
 900         * page/wx/FrameWx.cpp: Removed.
 901         * platform/gtk/ClipboardGtk.cpp:
 902         (WebCore::ClipboardGtk::createDragImage): Call nodeImage to get
 903         the drag image for a node.
 904
 905 2011-09-29  Noel Gordon  <noel.gordon@gmail.com>
 906
 907         [chromium skia] PNGImageEncoder: hoist constants out of the encoding loop
 908         https://bugs.webkit.org/show_bug.cgi?id=68988
 909
 910         Reviewed by Kenneth Russell.
 911
 912         No new tests.  Covered by existing canvas 2d and 3d tests.
 913         canvas/philip/tests/toDataURL.png.*.html
 914         fast/canvas/toDataURL-alpha.html
 915         fast/canvas/webgl/premultiplyalpha-test.html
 916
 917         * platform/image-encoders/skia/PNGImageEncoder.cpp:
 918         (WebCore::encodePixels):  Move constant out of the encoding loop.
 919         (WebCore::PNGImageEncoder::encode):  Consistency & style: call encodePixels()
 920         just like we do in the JPEG encoder.
 921
 922 2011-09-29  Iain Merrick  <husky@google.com>
 923
 924         Add unit test for CCLayerSorter
 925         https://bugs.webkit.org/show_bug.cgi?id=68622
 926
 927         Minor refactoring for testability:
 928         - Made pointInTriangle public.
 929         - Added LayerShape to decouple LayerIntersector and GraphNode.
 930         - Added a public wrapper function for LayerIntersector.
 931
 932         Reviewed by James Robinson.
 933
 934         * platform/graphics/chromium/cc/CCLayerSorter.cpp:
 935         (WebCore::CCLayerSorter::pointInTriangle):
 936         (WebCore::CCLayerSorter::calculateZDiff):
 937         (WebCore::CCLayerSorter::LayerIntersector::LayerIntersector):
 938         (WebCore::CCLayerSorter::LayerIntersector::go):
 939         (WebCore::CCLayerSorter::LayerIntersector::checkZDiff):
 940         (WebCore::CCLayerSorter::LayerIntersector::layerZFromProjectedPoint):
 941         (WebCore::CCLayerSorter::checkOverlap):
 942         (WebCore::CCLayerSorter::LayerShape::LayerShape):
 943         (WebCore::CCLayerSorter::createGraphNodes):
 944         * platform/graphics/chromium/cc/CCLayerSorter.h:
 945         (WebCore::CCLayerSorter::LayerShape::LayerShape):
 946         (WebCore::CCLayerSorter::GraphNode::GraphNode):
 947
 948 2011-09-29  Ilya Tikhonovsky  <loislo@chromium.org>
 949
 950         Web Inspector: UI performance test for network panel is incorrect. The time is including a timeout between scheduled refresh and actual refresh.
 951         https://bugs.webkit.org/show_bug.cgi?id=69087
 952
 953         The fix has the changes for the test and minor perf framework improvements.
 954
 955         Reviewed by Yury Semikhatsky.
 956
 957         * inspector/front-end/NetworkPanel.js:
 958         (WebInspector.NetworkLogView.prototype._defaultRefreshDelay.500._scheduleRefresh):
 959
 960 2011-09-29  Vsevolod Vlasov  <vsevik@chromium.org>
 961
 962         Web Inspector: Make search-in-resource test simpler.
 963         https://bugs.webkit.org/show_bug.cgi?id=69025
 964
 965         Changed search-in-resource test.
 966         Removed error parameter from content provider's searchInContent callback.
 967
 968         Reviewed by Pavel Feldman.
 969
 970         * inspector/InspectorPageAgent.cpp:
 971         (WebCore::InspectorPageAgent::searchInResource):
 972         (WebCore::InspectorPageAgent::searchInResources):
 973         * inspector/front-end/Resource.js:
 974         (WebInspector.Resource.prototype.searchInContent):
 975
 976 2011-09-29  Hans Wennborg  <hans@chromium.org>
 977
 978         IndexedDB: Use LevelDB also for in-memory databases
 979         https://bugs.webkit.org/show_bug.cgi?id=68903
 980
 981         Reviewed by Steve Block.
 982
 983         Add LevelDBDatabase::openInMemory() which uses leveldb::NewMemEnv()
 984         to create in-memory LevelDB databases.
 985
 986         Use this in IDBLeveLDBBackingStore::open() when the caller passes in
 987         an empty file path.
 988         This happens in Chromium's incognito mode, and when running layout
 989         tests.
 990
 991         Fix IDBSQLiteBackingStore::backingStoreExists() so it doesn't create
 992         files when passed in an empty file path, but uses the in-memory mode
 993         instead.
 994
 995         Existing layout tests will all be run in-memory.
 996
 997         * platform/leveldb/LevelDBDatabase.cpp:
 998         (WebCore::LevelDBDatabase::~LevelDBDatabase):
 999         (WebCore::openDB):
1000         (WebCore::LevelDBDatabase::open):
1001         (WebCore::LevelDBDatabase::openInMemory):
1002         * platform/leveldb/LevelDBDatabase.h:
1003         * storage/IDBLevelDBBackingStore.cpp:
1004         (WebCore::IDBLevelDBBackingStore::open):
1005         * storage/IDBSQLiteBackingStore.cpp:
1006         (WebCore::IDBSQLiteBackingStore::backingStoreExists):
1007
1008 2011-09-29  Ilya Tikhonovsky  <loislo@chromium.org>
1009
1010         Web Inspector: speed-up Network panel. Change _staleResources type from array to object.
1011         https://bugs.webkit.org/show_bug.cgi?id=69081
1012
1013         There is a test with 30 requests.
1014         For the each stage of loading a resource we have an entry in _staleResources array. There are at least 4 stages per request.
1015         NetworkLogView._refresh function is creating/updating the resource row for the each such entry.
1016         This array can be replaced with a hash map just because the resource associated with the entry is the same for all the entries with the same request id.
1017
1018         Reviewed by Pavel Feldman.
1019
1020         Test: inspector/performance/resources/network-append-30-requests.html
1021
1022         * inspector/front-end/NetworkPanel.js:
1023         (WebInspector.NetworkLogView):
1024         (WebInspector.NetworkLogView.prototype._invalidateAllItems):
1025         (WebInspector.NetworkLogView.prototype.refresh):
1026         (WebInspector.NetworkLogView.prototype._reset):
1027         (WebInspector.NetworkLogView.prototype._refreshResource):
1028
1029 2011-09-28  Pavel Feldman  <pfeldman@google.com>
1030
1031         Web Inspector: make inspector protocol validation a part of the build process.
1032         https://bugs.webkit.org/show_bug.cgi?id=68999
1033
1034         Since we'd like to commit to inspector protocol backwards compatibility, violating it
1035         should be a build failure.
1036
1037         Reviewed by Yury Semikhatsky.
1038
1039         * CMakeLists.txt:
1040         * CodeGenerators.pri:
1041         * DerivedSources.make:
1042         * GNUmakefile.am:
1043         * WebCore.gyp/WebCore.gyp:
1044         * inspector/Inspector.draft-01.json:
1045         * inspector/Inspector.json:
1046         * inspector/generate-inspector-idl:
1047         * inspector/generate-protocol-externs:
1048         * inspector/validate-protocol-compatibility:
1049
1050 2011-09-27  Pavel Feldman  <pfeldman@google.com>
1051
1052         Web Inspector: Scripts panel without folders causes errors when creating content scripts section.
1053         https://bugs.webkit.org/show_bug.cgi?id=68827
1054
1055         Reviewed by Yury Semikhatsky.
1056
1057         * inspector/front-end/ScriptsPanel.js:
1058         (WebInspector.ScriptsPanel.prototype._addOptionToFilesSelect.insertOrdered.optionCompare):
1059         (WebInspector.ScriptsPanel.prototype._addOptionToFilesSelect.insertOrdered):
1060         (WebInspector.ScriptsPanel.prototype._addOptionToFilesSelect):
1061
1062 2011-09-28  Andrey Kosyakov  <caseq@chromium.org>
1063
1064         Web Inspector: network log view refresh optimizations
1065         https://bugs.webkit.org/show_bug.cgi?id=69010
1066
1067         Reviewed by Pavel Feldman.
1068
1069         * inspector/front-end/NetworkPanel.js:
1070         (WebInspector.NetworkLogView.prototype.refresh):
1071
1072 2011-09-29  Pavel Podivilov  <podivilov@chromium.org>
1073
1074         Web Inspector: split SourceFile.js into RawSourceCode.js and UISourceCodeContentProviders.js.
1075         https://bugs.webkit.org/show_bug.cgi?id=69068
1076
1077         Reviewed by Yury Semikhatsky.
1078
1079         * WebCore.gypi:
1080         * WebCore.vcproj/WebCore.vcproj:
1081         * inspector/front-end/RawSourceCode.js: Renamed from Source/WebCore/inspector/front-end/SourceFile.js.
1082         (WebInspector.RawSourceCode): Moved from SourceFile.js.
1083         (WebInspector.UILocation): Moved from SourceFile.js.
1084         * inspector/front-end/UISourceCodeContentProviders.js: Added. Moved ContentProvider implementations from SourceFile.js
1085         (WebInspector.ScriptContentProvider):
1086         (WebInspector.ScriptContentProvider.prototype.requestContent):
1087         (WebInspector.ConcatenatedScriptsContentProvider):
1088         (WebInspector.ConcatenatedScriptsContentProvider.prototype.requestContent):
1089         (WebInspector.ConcatenatedScriptsContentProvider.prototype._concatenateScriptsContent.appendChunk):
1090         (WebInspector.ConcatenatedScriptsContentProvider.prototype._concatenateScriptsContent):
1091         (WebInspector.ResourceContentProvider):
1092         (WebInspector.ResourceContentProvider.prototype.requestContent):
1093         (WebInspector.StaticContentProvider):
1094         (WebInspector.StaticContentProvider.prototype.requestContent):
1095         * inspector/front-end/WebKit.qrc:
1096         * inspector/front-end/inspector.html:
1097
1098 2011-09-29  Philippe Normand  <pnormand@igalia.com>
1099
1100         [GStreamer] 2 flaky media tests
1101         https://bugs.webkit.org/show_bug.cgi?id=67407
1102
1103         Reviewed by Martin Robinson.
1104
1105         Workaround for GStreamer bug#639941. In GStreamer
1106         0.10.35 basesink reports wrong duration in case of EOS and
1107         negative playback rate. This workaround fixes two media tests,
1108         media/video-timeupdate-reverse-play.html and
1109         media/video-reverse-play-duration.html.
1110
1111         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
1112         (WebCore::MediaPlayerPrivateGStreamer::prepareToPlay):
1113         (WebCore::MediaPlayerPrivateGStreamer::currentTime):
1114         (WebCore::MediaPlayerPrivateGStreamer::didEnd):
1115
1116 2011-09-29  Philippe Normand  <pnormand@igalia.com>
1117
1118         [GStreamer] fullscreen video pause/play fails
1119         https://bugs.webkit.org/show_bug.cgi?id=66936
1120
1121         Reviewed by Martin Robinson.
1122
1123         Don't use the identity element to avoid painting of the in-window
1124         video. Instead simply make the sink aware of the fullscreen state
1125         and ignore buffers if fullscreen and autovideosink are
1126         active. Also fixed two deadlocks happening when a paused pipeline
1127         is switched to fullscreen and when fullscreen is disabled for a
1128         paused pipeline.
1129
1130         * platform/graphics/gstreamer/GStreamerGWorld.cpp:
1131         (WebCore::GStreamerGWorld::enterFullscreen):
1132         (WebCore::GStreamerGWorld::exitFullscreen):
1133         * platform/graphics/gstreamer/GStreamerGWorld.h:
1134         (WebCore::GStreamerGWorld::isFullscreen):
1135         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
1136         (WebCore::MediaPlayerPrivateGStreamer::createGSTPlayBin):
1137         * platform/graphics/gstreamer/VideoSinkGStreamer.cpp:
1138         (webkit_video_sink_render):
1139         (webkit_video_sink_new):
1140         * platform/graphics/gstreamer/VideoSinkGStreamer.h:
1141
1142 2011-09-29  Tim Horton  <timothy_horton@apple.com>
1143
1144         REGRESSION(87010): elements in ECMA-cloud neither filled nor blurred
1145         https://bugs.webkit.org/show_bug.cgi?id=68679
1146         <rdar://problem/10204649>
1147
1148         Reviewed by Nikolas Zimmermann.
1149
1150         Strip prefixes from SVG attributes before testing if they're supported.
1151         Namespaced attributes will be matched using their namespace instead of
1152         the (user-choosable) prefix.
1153
1154         Test: svg/custom/xlink-custom-namespace.svg
1155
1156         * svg/SVGAElement.cpp:
1157         (WebCore::SVGAElement::isSupportedAttribute):
1158         * svg/SVGAnimateMotionElement.cpp:
1159         (WebCore::SVGAnimateMotionElement::isSupportedAttribute):
1160         * svg/SVGAnimateTransformElement.cpp:
1161         (WebCore::SVGAnimateTransformElement::isSupportedAttribute):
1162         * svg/SVGAnimationElement.cpp:
1163         (WebCore::SVGAnimationElement::isSupportedAttribute):
1164         * svg/SVGCircleElement.cpp:
1165         (WebCore::SVGCircleElement::isSupportedAttribute):
1166         * svg/SVGClipPathElement.cpp:
1167         (WebCore::SVGClipPathElement::isSupportedAttribute):
1168         * svg/SVGComponentTransferFunctionElement.cpp:
1169         (WebCore::SVGComponentTransferFunctionElement::isSupportedAttribute):
1170         * svg/SVGCursorElement.cpp:
1171         (WebCore::SVGCursorElement::isSupportedAttribute):
1172         * svg/SVGElement.h:
1173         (WebCore::SVGAttributeHashTranslator::hash):
1174         (WebCore::SVGAttributeHashTranslator::equal):
1175         * svg/SVGEllipseElement.cpp:
1176         (WebCore::SVGEllipseElement::isSupportedAttribute):
1177         * svg/SVGExternalResourcesRequired.h:
1178         * svg/SVGFEBlendElement.cpp:
1179         (WebCore::SVGFEBlendElement::isSupportedAttribute):
1180         * svg/SVGFEColorMatrixElement.cpp:
1181         (WebCore::SVGFEColorMatrixElement::isSupportedAttribute):
1182         * svg/SVGFEComponentTransferElement.cpp:
1183         (WebCore::SVGFEComponentTransferElement::isSupportedAttribute):
1184         * svg/SVGFECompositeElement.cpp:
1185         (WebCore::SVGFECompositeElement::isSupportedAttribute):
1186         * svg/SVGFEConvolveMatrixElement.cpp:
1187         (WebCore::SVGFEConvolveMatrixElement::isSupportedAttribute):
1188         * svg/SVGFEDiffuseLightingElement.cpp:
1189         (WebCore::SVGFEDiffuseLightingElement::isSupportedAttribute):
1190         * svg/SVGFEDisplacementMapElement.cpp:
1191         (WebCore::SVGFEDisplacementMapElement::isSupportedAttribute):
1192         * svg/SVGFEDropShadowElement.cpp:
1193         (WebCore::SVGFEDropShadowElement::isSupportedAttribute):
1194         * svg/SVGFEGaussianBlurElement.cpp:
1195         (WebCore::SVGFEGaussianBlurElement::isSupportedAttribute):
1196         * svg/SVGFEImageElement.cpp:
1197         (WebCore::SVGFEImageElement::isSupportedAttribute):
1198         * svg/SVGFELightElement.cpp:
1199         (WebCore::SVGFELightElement::isSupportedAttribute):
1200         * svg/SVGFEMergeNodeElement.cpp:
1201         (WebCore::SVGFEMergeNodeElement::isSupportedAttribute):
1202         * svg/SVGFEMorphologyElement.cpp:
1203         (WebCore::SVGFEMorphologyElement::isSupportedAttribute):
1204         * svg/SVGFEOffsetElement.cpp:
1205         (WebCore::SVGFEOffsetElement::isSupportedAttribute):
1206         * svg/SVGFESpecularLightingElement.cpp:
1207         (WebCore::SVGFESpecularLightingElement::isSupportedAttribute):
1208         * svg/SVGFETileElement.cpp:
1209         (WebCore::SVGFETileElement::isSupportedAttribute):
1210         * svg/SVGFETurbulenceElement.cpp:
1211         (WebCore::SVGFETurbulenceElement::isSupportedAttribute):
1212         * svg/SVGFilterElement.cpp:
1213         (WebCore::SVGFilterElement::isSupportedAttribute):
1214         * svg/SVGFilterPrimitiveStandardAttributes.cpp:
1215         (WebCore::SVGFilterPrimitiveStandardAttributes::isSupportedAttribute):
1216         * svg/SVGForeignObjectElement.cpp:
1217         (WebCore::SVGForeignObjectElement::isSupportedAttribute):
1218         * svg/SVGGElement.cpp:
1219         (WebCore::SVGGElement::isSupportedAttribute):
1220         * svg/SVGGradientElement.cpp:
1221         (WebCore::SVGGradientElement::isSupportedAttribute):
1222         * svg/SVGImageElement.cpp:
1223         (WebCore::SVGImageElement::isSupportedAttribute):
1224         * svg/SVGLineElement.cpp:
1225         (WebCore::SVGLineElement::isSupportedAttribute):
1226         * svg/SVGLinearGradientElement.cpp:
1227         (WebCore::SVGLinearGradientElement::isSupportedAttribute):
1228         * svg/SVGMPathElement.cpp:
1229         (WebCore::SVGMPathElement::isSupportedAttribute):
1230         * svg/SVGMarkerElement.cpp:
1231         (WebCore::SVGMarkerElement::isSupportedAttribute):
1232         * svg/SVGMaskElement.cpp:
1233         (WebCore::SVGMaskElement::isSupportedAttribute):
1234         * svg/SVGPathElement.cpp:
1235         (WebCore::SVGPathElement::isSupportedAttribute):
1236         * svg/SVGPatternElement.cpp:
1237         (WebCore::SVGPatternElement::isSupportedAttribute):
1238         * svg/SVGPolyElement.cpp:
1239         (WebCore::SVGPolyElement::isSupportedAttribute):
1240         * svg/SVGRadialGradientElement.cpp:
1241         (WebCore::SVGRadialGradientElement::isSupportedAttribute):
1242         * svg/SVGRectElement.cpp:
1243         (WebCore::SVGRectElement::isSupportedAttribute):
1244         * svg/SVGScriptElement.cpp:
1245         (WebCore::SVGScriptElement::isSupportedAttribute):
1246         * svg/SVGStopElement.cpp:
1247         (WebCore::SVGStopElement::isSupportedAttribute):
1248         * svg/SVGStyleElement.cpp:
1249         (WebCore::SVGStyleElement::isSupportedAttribute):
1250         * svg/SVGStyledTransformableElement.cpp:
1251         (WebCore::SVGStyledTransformableElement::isSupportedAttribute):
1252         * svg/SVGSymbolElement.cpp:
1253         (WebCore::SVGSymbolElement::isSupportedAttribute):
1254         * svg/SVGTRefElement.cpp:
1255         (WebCore::SVGTRefElement::isSupportedAttribute):
1256         * svg/SVGTextContentElement.cpp:
1257         (WebCore::SVGTextContentElement::isSupportedAttribute):
1258         * svg/SVGTextElement.cpp:
1259         (WebCore::SVGTextElement::isSupportedAttribute):
1260         * svg/SVGTextPathElement.cpp:
1261         (WebCore::SVGTextPathElement::isSupportedAttribute):
1262         * svg/SVGTextPositioningElement.cpp:
1263         (WebCore::SVGTextPositioningElement::isSupportedAttribute):
1264         * svg/SVGURIReference.cpp:
1265         (WebCore::SVGURIReference::addSupportedAttributes):
1266         * svg/SVGUseElement.cpp:
1267         (WebCore::SVGUseElement::isSupportedAttribute):
1268         * svg/SVGViewElement.cpp:
1269         (WebCore::SVGViewElement::isSupportedAttribute):
1270
1271 2011-09-28  Kentaro Hara  <haraken@chromium.org>
1272
1273         "ScriptValue" arguments of PopStateEvent methods should be "const ScriptValue&"
1274         https://bugs.webkit.org/show_bug.cgi?id=68984
1275
1276         Reviewed by Darin Adler.
1277
1278         Just a follow-up fix for the bug 68345. For performance,
1279         "ScriptValue" arguments of PopStateEvent methods should be "const ScriptValue&".
1280
1281         Tests: fast/events/constructors/pop-state-event-constructor.html
1282                fast/events/fire-popstate-event.html
1283
1284         * dom/PopStateEvent.cpp:
1285         (WebCore::PopStateEvent::PopStateEvent):
1286         (WebCore::PopStateEvent::create):
1287         (WebCore::PopStateEvent::initPopStateEvent):
1288         * dom/PopStateEvent.h:
1289
1290 2011-09-28  Martin Robinson  <mrobinson@igalia.com>
1291
1292         [GTK] Move drag-and-drop code which can be shared with WebKit2 to WebCore
1293         https://bugs.webkit.org/show_bug.cgi?id=66890
1294
1295         Added a GtkDragAndDropHelper class to hold all the logic that is common
1296         between WebKit1 and WebKit2. This will allow greater code sharing between
1297         the two ports.
1298
1299         Reviewed by Philippe Normand.
1300
1301         No new tests. This patch should not change behavior.
1302
1303         * GNUmakefile.list.am: Added the GtkDragAndDropHelper to the sources list.
1304         * platform/gtk/GtkDragAndDropHelper.cpp: Added.
1305         * platform/gtk/GtkDragAndDropHelper.h: Added.
1306
1307 2011-09-28  Dan Bernstein  <mitz@apple.com>
1308
1309         <rdar://problem/10196497> first-letter after list marker not updated correctly
1310
1311         Reviewed by Simon Fraser.
1312
1313         Test: fast/dynamic/first-letter-after-list-marker.html
1314
1315         * rendering/RenderBlock.cpp:
1316         (WebCore::RenderBlock::updateFirstLetter): Improved the logic for continuing past list markers
1317         when trying to locate a first letter to update.
1318
1319 2011-09-28  Kent Tamura  <tkent@chromium.org>
1320
1321         REGRESSION(r93858): Can't type anything into input elements when maxlength is greater than 2^31
1322         https://bugs.webkit.org/show_bug.cgi?id=68981
1323
1324         Reviewed by Darin Adler.
1325
1326         * html/parser/HTMLParserIdioms.cpp:
1327         (WebCore::parseHTMLInteger):
1328         Check the failure of charactersToIntStrict().
1329         (WebCore::parseHTMLNonNegativeInteger):
1330         Check the failure of charactersToUIntStrict().
1331
1332 2011-09-28  Antoine Labour  <piman@chromium.org>
1333
1334         Remove unused createRootLayerPainter from CCLayerTreeHostClient
1335         https://bugs.webkit.org/show_bug.cgi?id=69042
1336
1337         Reviewed by James Robinson.
1338
1339         No new test needed: no semantic change.
1340
1341         * platform/graphics/chromium/cc/CCLayerTreeHost.h:
1342
1343 2011-09-28  Luke Macpherson   <macpherson@chromium.org>
1344
1345         Only resolve direction aware properties when properties are direction aware.
1346         https://bugs.webkit.org/show_bug.cgi?id=68968
1347
1348         Reviewed by Eric Seidel.
1349
1350         No new tests / no functionality changed.
1351
1352         Profiling showed a ~8% improvement in applyProperty by not calling resolveDirectionAwareProperty for every property.
1353
1354         * css/CSSStyleSelector.cpp:
1355         (WebCore::CSSStyleSelector::applyProperty):
1356
1357 2011-09-28  Simon Fraser  <simon.fraser@apple.com>
1358
1359         fast/dom/scroll-element-to-rect.html fails on WK1 Mac port
1360         https://bugs.webkit.org/show_bug.cgi?id=68815
1361
1362         Reviewed by Dan Bernstein.
1363
1364         FrameView::scrollElementToRect() was incorrectly using Element::boundsInWindowSpace(),
1365         which is window-relative (not web view-relative), and has flipped coordinates
1366         in WebKit1.
1367
1368         Change to use Node::getRect() which is what the author intended.
1369
1370         * dom/Element.h:
1371         * page/FrameView.cpp:
1372         (WebCore::FrameView::scrollElementToRect):
1373
1374 2011-09-28  Peter Beverloo  <peter@chromium.org>
1375
1376         Don't clamp cubic-bezier timing functions between 0 and 1
1377         https://bugs.webkit.org/show_bug.cgi?id=45761
1378
1379         Remove the limitation of clamping the cubic-bezier timing function
1380         values between 0 and 1, following the specification change made on
1381         April 5 of this year.
1382
1383         Reviewed by Dean Jackson.
1384
1385         Tests: transitions/cubic-bezier-overflow-color.html
1386                transitions/cubic-bezier-overflow-length.html
1387                transitions/cubic-bezier-overflow-shadow.html
1388                transitions/cubic-bezier-overflow-svg-length.html
1389                transitions/cubic-bezier-overflow-transform.html
1390
1391         * css/CSSParser.cpp:
1392         (WebCore::CSSParser::parseCubicBezierTimingFunctionValue):
1393
1394 2011-09-28  Chris Rogers  <crogers@google.com>
1395
1396         DelayNode must set the context on delayTime AudioParam to support automation
1397         https://bugs.webkit.org/show_bug.cgi?id=68828
1398
1399         Reviewed by Kenneth Russell.
1400
1401         Test: webaudio/delaynode-scheduling.html
1402
1403         * webaudio/DelayNode.cpp:
1404         (WebCore::DelayNode::DelayNode):
1405
1406 2011-09-28  Sergey Glazunov  <serg.glazunov@gmail.com>
1407
1408         JavaScript generated documents don't inherit the cookie URL
1409         https://bugs.webkit.org/show_bug.cgi?id=69003
1410
1411         Reviewed by Adam Barth.
1412
1413         Test: http/tests/security/cookies/cookie-theft-with-javascript-doc.html
1414
1415         * dom/Document.h:
1416         (WebCore::Document::setCookieURL):
1417         * loader/DocumentWriter.cpp:
1418         (WebCore::DocumentWriter::replaceDocument):
1419         (WebCore::DocumentWriter::begin):
1420         * loader/DocumentWriter.h:
1421
1422 2011-09-27  Dimitri Glazkov  <dglazkov@chromium.org>
1423
1424         REGRESSION(r95573): Crash when loading SVG documents in a flattened frame or any SVG document in Chromium/Mac.
1425         https://bugs.webkit.org/show_bug.cgi?id=68938
1426
1427         Reviewed by David Hyatt.
1428
1429         Test: all svg tests in LayoutTests.
1430
1431         * rendering/RenderBlock.cpp:
1432         (WebCore::RenderBlock::computeBlockPreferredLogicalWidths): Added a null-check for containingBlock.
1433
1434 2011-09-28  Una Sabovic  <una.sabovic@palm.com>
1435
1436         Setting innerText to an empty string on editable div loses focus
1437         https://bugs.webkit.org/show_bug.cgi?id=62092
1438
1439         Reviewed by Ryosuke Niwa.
1440
1441         When selection start or end node is being deleted do not clear the selection.
1442         Instead update the start/end position to an equivalent parent-anchored positions.
1443         When text is inserted/deleted update selection per range modification spec:
1444         http://www.w3.org/TR/DOM-Level-2-Traversal-Range/ranges.html#Level-2-Range-Mutation
1445
1446         Tests: editing/selection/document-mutation.html
1447                editing/selection/editable-div-clear-on-keydown.html
1448
1449         * editing/FrameSelection.cpp:
1450         (WebCore::FrameSelection::respondToNodeModification):
1451         (WebCore::updatePositionAfterAdoptingTextReplacement):
1452         (WebCore::FrameSelection::textWillBeReplaced):
1453         * editing/htmlediting.cpp:
1454         (WebCore::updatePositionForNodeRemoval):
1455
1456 2011-09-28  Joseph Pecoraro  <joepeck@webkit.org>
1457
1458         FTPDirectoryDocument Shows Garbled String for Last Modified Date
1459         https://bugs.webkit.org/show_bug.cgi?id=69011
1460
1461         Reviewed by Dan Bernstein.
1462
1463         Force WTF::String concatenation instead of an accidental char* pointer addition.
1464
1465         * html/FTPDirectoryDocument.cpp:
1466         (WebCore::processFileDateString):
1467
1468 2011-09-28  Nate Chapin  <japhet@chromium.org>
1469
1470         Fix a regression in r96060. ResourceLoader shouldn't get
1471         initialized with a null DocumentLoader.
1472         https://bugs.webkit.org/show_bug.cgi?id=69004
1473
1474         Reviewed by Csaba Osztrogonác.
1475
1476         This should fix a debug crash in some platforms in
1477         plugins/js-from-destroy.html
1478
1479         * loader/ResourceLoader.cpp:
1480         (WebCore::ResourceLoader::init): Move the activeDocumentLoader() null checks back
1481             to SubresouceLoader::create().
1482         * loader/SubresourceLoader.cpp:
1483         (WebCore::SubresourceLoader::create):
1484
1485 2011-09-28  Adam Barth  <abarth@webkit.org>
1486
1487         CSP doesn't work for a wide variety of cases
1488         https://bugs.webkit.org/show_bug.cgi?id=68921
1489
1490         Reviewed by Darin Adler.
1491
1492         Patch suggested by Sam Weinig.  It's unclear to me how to test this
1493         change because all our tests run on non-default ports, which is why we
1494         have this bug in the first place.  Mozilla uses a proxy while testing
1495         so they can test with URLs like http://example.com, but we don't have
1496         such a facility.
1497
1498         * page/ContentSecurityPolicy.cpp:
1499         (WebCore::CSPSource::portMatches):
1500
1501 2011-09-28  Fady Samuel  <fsamuel@chromium.org>
1502
1503         [Chromium] Seperate GTK specific Gyp rules from X11 Gyp rules
1504         https://bugs.webkit.org/show_bug.cgi?id=68936
1505
1506         Reviewed by Tony Chang.
1507
1508         Chromium need to be built without Gtk for future version of ChromiumOS.
1509
1510         * WebCore.gyp/WebCore.gyp:
1511
1512 2011-09-28  Tom Sepez  <tsepez@chromium.org>
1513
1514         Revert change which broke displaying end script tags in view-source, instead
1515         deal with any trailing </script> tag included by mistake in the XSSAuditor
1516         itself.  Correct tests to detect the missing close tags.
1517         https://bugs.webkit.org/show_bug.cgi?id=68898
1518
1519         Reviewed by Adam Barth.
1520
1521         * html/parser/HTMLSourceTracker.cpp:
1522         (WebCore::HTMLSourceTracker::end):
1523         * html/parser/HTMLTokenizer.cpp:
1524         (WebCore::HTMLTokenizer::nextToken):
1525         * html/parser/XSSAuditor.cpp:
1526         (WebCore::startsHTMLEndTagAt):
1527         (WebCore::XSSAuditor::snippetForJavaScript):
1528
1529 2011-09-28  Adam Barth  <abarth@webkit.org>
1530
1531         Remove empty directories.
1532
1533         * editing/brew: Removed.
1534         * page/brew: Removed.
1535         * platform/brew: Removed.
1536         * platform/graphics/brew: Removed.
1537         * platform/network/brew: Removed.
1538         * platform/text/brew: Removed.
1539
1540 2011-09-28  Ilya Tikhonovsky  <loislo@chromium.org>
1541
1542         Web Inspector: optional arguments support for the frontend needs to be extended.
1543         https://bugs.webkit.org/show_bug.cgi?id=69005
1544
1545         Generator supports optional arguments but they are transferred by value.
1546         It is not suitable if the used type doesn't have operator bool.
1547         I'll transfer such arguments by pointer.
1548
1549         Reviewed by Pavel Feldman.
1550
1551         Build is the test.
1552
1553         * inspector/CodeGeneratorInspector.pm:
1554         (generateFrontendFunction):
1555         (paramTypeTraits):
1556         * inspector/InspectorDebuggerAgent.cpp:
1557         (WebCore::InspectorDebuggerAgent::didParseSource):
1558         * inspector/InspectorResourceAgent.cpp:
1559         (WebCore::InspectorResourceAgent::didFailLoading):
1560
1561 2011-09-28  Antaryami Pandia  <antaryami.pandia@motorola.com>
1562
1563         Reflected attribute input.size wraps on negative values (Chrome), or
1564         returns them (Safari).
1565         https://bugs.webkit.org/show_bug.cgi?id=44886
1566
1567         Reviewed by Darin Adler.
1568
1569         Test: fast/dom/HTMLInputElement/input-size-attribute.html
1570
1571         * html/HTMLInputElement.cpp:
1572         (WebCore::HTMLInputElement::parseMappedAttribute):
1573
1574 2011-09-28  Vsevolod Vlasov  <vsevik@chromium.org>
1575
1576         Web Inspector: Add support for search in single resource to page agent.
1577         https://bugs.webkit.org/show_bug.cgi?id=68998
1578
1579         Reviewed by Pavel Feldman.
1580
1581         Test: http/tests/inspector/search/search-in-resource.html
1582
1583         * inspector/ContentSearchUtils.cpp:
1584         (WebCore::ContentSearchUtils::getRegularExpressionMatchesByLines):
1585         * inspector/ContentSearchUtils.h:
1586         * inspector/Inspector.json:
1587         * inspector/InspectorPageAgent.cpp:
1588         (WebCore::buildObjectForSearchMatch):
1589         (WebCore::InspectorPageAgent::searchInResource):
1590         * inspector/InspectorPageAgent.h:
1591         * inspector/front-end/Resource.js:
1592         (WebInspector.Resource.prototype.searchInContent):
1593         * inspector/front-end/SourceFile.js:
1594         (WebInspector.ResourceContentProvider.prototype.requestContent):
1595         (WebInspector.ResourceContentProvider.prototype.searchInContent):
1596         * inspector/front-end/UISourceCode.js:
1597         (WebInspector.ContentProvider.prototype.requestContent):
1598         (WebInspector.ContentProvider.prototype.searchInContent):
1599
1600 2011-09-28  Ilya Tikhonovsky  <loislo@chromium.org>
1601
1602         Web Inspector: It'd be useful to have performance stats for the back-end to front-end communication channel.
1603         https://bugs.webkit.org/show_bug.cgi?id=69002
1604
1605         Reviewed by Yury Semikhatsky.
1606
1607         * inspector/CodeGeneratorInspector.pm:
1608
1609 2011-09-27  Pavel Podivilov  <podivilov@chromium.org>
1610
1611         Web Inspector: live edit cursor always appears at the beginning of the file.
1612         https://bugs.webkit.org/show_bug.cgi?id=68900
1613
1614         Reviewed by Pavel Feldman.
1615
1616         * inspector/front-end/TextViewer.js:
1617         (WebInspector.TextEditorMainPanel.prototype._updateSelectionOnStartEditing):
1618
1619 2011-09-28  Vsevolod Vlasov  <vsevik@chromium.org>
1620
1621         Web Inspector: Refactor searching in inspector, add test (fixed non regex search).
1622         https://bugs.webkit.org/show_bug.cgi?id=68930
1623
1624         Reviewed by Yury Semikhatsky.
1625
1626         Search moved out from page agent.
1627         Fixed search for non regex case which is not yet used.
1628
1629         Test: http/tests/inspector/search/search-in-resources.html
1630
1631         * CMakeLists.txt:
1632         * GNUmakefile.list.am:
1633         * WebCore.gypi:
1634         * WebCore.pro:
1635         * WebCore.vcproj/WebCore.vcproj:
1636         * WebCore.xcodeproj/project.pbxproj:
1637         * inspector/ContentSearchUtils.cpp: Added.
1638         (WebCore::ContentSearchUtils::createSearchRegexSource):
1639         (WebCore::ContentSearchUtils::createSearchRegex):
1640         (WebCore::ContentSearchUtils::countRegularExpressionMatches):
1641         * inspector/ContentSearchUtils.h: Added.
1642         * inspector/InspectorPageAgent.cpp:
1643         (WebCore::hasTextContent):
1644         (WebCore::InspectorPageAgent::cachedResourceContent):
1645         (WebCore::textContentForCachedResource):
1646         (WebCore::buildObjectForSearchResult):
1647         (WebCore::InspectorPageAgent::searchInResources):
1648         * inspector/front-end/utilities.js:
1649         ():
1650
1651 2011-09-28  Kentaro Hara  <haraken@chromium.org>
1652
1653         Implement a PopStateEvent constructor for V8
1654         https://bugs.webkit.org/show_bug.cgi?id=68896
1655
1656         Reviewed by Adam Barth.
1657
1658         Test: fast/events/constructors/pop-state-event-constructor.html
1659
1660         * bindings/v8/custom/V8EventConstructors.cpp: Added a PopStateEvent constructor.
1661         * dom/PopStateEvent.idl: Added a 'V8CustomConstructor' attribute.
1662
1663 2011-09-28  No'am Rosenthal  <noam.rosenthal@nokia.com>
1664
1665         [Texmap] Allow TextureMapperGL to work without a GraphicsContext
1666         https://bugs.webkit.org/show_bug.cgi?id=68980
1667
1668         Reviewed by Andreas Kling.
1669
1670         Perform a null-check for m_context and initialize that variable.
1671         If a graphics-context doesn't exist, we don't need to reset it.
1672
1673         No new functionality so no new tests.
1674
1675         * platform/graphics/opengl/TextureMapperGL.cpp:
1676         (WebCore::TextureMapperGL::TextureMapperGL):
1677         (WebCore::TextureMapperGL::beginPainting):
1678         (WebCore::TextureMapperGL::endPainting):
1679
1680 2011-09-28  Kentaro Hara  <haraken@chromium.org>
1681
1682         Implement an ErrorEvent constructor for V8
1683         https://bugs.webkit.org/show_bug.cgi?id=68336
1684
1685         Reviewed by Adam Barth.
1686
1687         Test: fast/events/constructors/error-event-constructor.html
1688
1689         * bindings/v8/OptionsObject.cpp:
1690         (WebCore::OptionsObject::getKeyValue): Returns an unsigned value corresponding to a given key.
1691         * bindings/v8/OptionsObject.h:
1692         * bindings/v8/custom/V8EventConstructors.cpp: Added the ErrorEvent constructor.
1693         * dom/ErrorEvent.idl: Added a 'V8CustomConstructor' attribute.
1694
1695 2011-09-27  Andy Estes  <aestes@apple.com>
1696
1697         WebKitLinkedOnOrAfter() check is ineffective for Solar Walk app-specific hack.
1698         https://bugs.webkit.org/show_bug.cgi?id=68863
1699
1700         Reviewed by Darin Adler.
1701
1702         The SolarWalk binary does not link against WebKit.framework directly,
1703         so calling NSVersionOfLinkTimeLibrary() will not return the version of
1704         WebKit against which SolarWalk was linked. Since Solar Walk has
1705         released an update that addresses the issue for which this app-specific
1706         hack was originally added, we should just remove the hack.
1707
1708         * WebCore.exp.in: Remove applicationIsSolarWalkMac().
1709         * platform/RuntimeApplicationChecks.cpp: Ditto.
1710         * platform/RuntimeApplicationChecks.h: Ditto.
1711
1712 2011-09-28  Tim Horton  <timothy_horton@apple.com>
1713
1714         Fix potential SVG performance regression (over-invalidation of caches) from 96052
1715         https://bugs.webkit.org/show_bug.cgi?id=68941
1716         <rdar://problem/10196224>
1717
1718         Reviewed by Zoltan Herczeg.
1719
1720         Only invalidate the cache of a filter when target style changes, as of the different types of cached
1721         resources, only filters make use of the style of the target element and thus need to be invalidated
1722         when the target style changes.
1723
1724         No new tests, performance improvement.
1725
1726         * rendering/svg/SVGResourcesCache.cpp:
1727         (WebCore::SVGResourcesCache::clientUpdatedFromElement):
1728
1729 2011-09-28  Adenilson Cavalcanti  <adenilson.silva@openbossa.org>
1730
1731         Missing initialization of member in ImageFrameQt class
1732         https://bugs.webkit.org/show_bug.cgi?id=68943
1733
1734         Reviewed by Kenneth Rohde Christiansen.
1735
1736         No new tests, this fixes internal implementation detail.
1737
1738         * platform/image-decoders/qt/ImageFrameQt.cpp:
1739         (WebCore::ImageFrame::ImageFrame):
1740         (WebCore::ImageFrame::operator=):
1741
1742 2011-09-27  Luke Macpherson   <macpherson@chromium.org>
1743
1744         Defer call to CSSValue::isPrimitiveValue(), saves ~4% in CSSStyleSelector::applyProperty().
1745         https://bugs.webkit.org/show_bug.cgi?id=68964
1746
1747         Reviewed by Eric Seidel.
1748
1749         No new tests / no functionality changed.
1750
1751         Doing value->isPrimitiveValue() is relatviely expensive, so moving it after the early returns but
1752         before the result is used saves a significant number of cycles. (Tested with Shark profiler in Safari).
1753
1754         * css/CSSStyleSelector.cpp:
1755         (WebCore::CSSStyleSelector::applyProperty):
1756
1757 2011-09-27  Ryosuke Niwa  <rniwa@webkit.org>
1758
1759         Simplify ReplaceSelectionCommand::positionAtStartOfInsertedContent
1760         https://bugs.webkit.org/show_bug.cgi?id=68939
1761
1762         Reviewed by Darin Adler.
1763
1764         Simplified ReplaceSelectionCommand::positionAtStartOfInsertedContent.
1765
1766         This change revealed a bug in removeUnrenderedTextNodesAtEnds that text nodes without any visible
1767         text at ends are not removed when it has a render object. Fixed the bug by checking the length of
1768         the rendered text. (Tested by editing/pasteboard/pasting-word-in-div-extra-line.html)
1769
1770         This further revealed that caretMaxRenderedOffset doesn't return an offset and caretMaxRenderedOffset
1771         on InlineBox, InlineTextBox, RenderObject, RenderBR, RenderPlaced are never called. To address this
1772         issue, renamed caretMaxRenderedOffset to renderedTextLength for RenderText and removed the rest.
1773
1774         * dom/Position.cpp:
1775         (WebCore::Position::rendersInDifferentPosition):
1776         * editing/ReplaceSelectionCommand.cpp:
1777         (WebCore::nodeHasVisibleRenderText): Added.
1778         (WebCore::ReplaceSelectionCommand::removeUnrenderedTextNodesAtEnds): Calls nodeHasVisibleRenderText.
1779         (WebCore::ReplaceSelectionCommand::positionAtStartOfInsertedContent): Simplified.
1780         * editing/visible_units.cpp:
1781         (WebCore::startOfParagraph): Calls renderedTextLength.
1782         (WebCore::endOfParagraph): Ditto.
1783         * rendering/InlineBox.cpp: Removed caretMaxRenderedOffset.
1784         * rendering/InlineBox.h: Ditto.
1785         * rendering/InlineTextBox.cpp: Ditto.
1786         * rendering/InlineTextBox.h: Ditto.
1787         * rendering/RenderBR.cpp: Ditto.
1788         * rendering/RenderBR.h: Ditto.
1789         * rendering/RenderObject.cpp: Ditto.
1790         * rendering/RenderObject.h: Ditto.
1791         * rendering/RenderReplaced.cpp: Ditto.
1792         * rendering/RenderReplaced.h: Ditto.
1793         * rendering/RenderText.cpp:
1794         (WebCore::RenderText::renderedTextLength): Renamed from caretMaxRenderedOffset.
1795         * rendering/RenderText.h:
1796
1797 2011-09-27  James Robinson  <jamesr@chromium.org>
1798
1799         [chromium] LayerRenderChromium asserts about leaking textures.
1800         https://bugs.webkit.org/show_bug.cgi?id=68190
1801
1802         Reviewed by Kenneth Russell.
1803
1804         This introduces an interface TextureAllocator that the TextureManager uses to allocate/deallocate textures.
1805         This means that TextureManager does not need to depend directly on GraphicsContext3D, other than for the format
1806         enum type, making it more testable.  It also allows us to track the allocations by subclassing TextureAllocator
1807         to make sure we aren't leaking any textures.
1808
1809         * WebCore.gypi:
1810         * platform/graphics/chromium/Canvas2DLayerChromium.cpp:
1811         (WebCore::Canvas2DLayerChromium::updateCompositorResources):
1812         * platform/graphics/chromium/Canvas2DLayerChromium.h:
1813         * platform/graphics/chromium/ImageLayerChromium.cpp:
1814         (WebCore::ImageLayerTextureUpdater::updateTextureRect):
1815         * platform/graphics/chromium/LayerChromium.h:
1816         (WebCore::LayerChromium::updateCompositorResources):
1817         * platform/graphics/chromium/LayerRendererChromium.cpp:
1818         (WebCore::LayerRendererChromium::LayerRendererChromium):
1819         (WebCore::LayerRendererChromium::releaseRenderSurfaceTextures):
1820         (WebCore::LayerRendererChromium::drawLayers):
1821         (WebCore::LayerRendererChromium::useRenderSurface):
1822         (WebCore::LayerRendererChromium::initializeSharedObjects):
1823         (WebCore::LayerRendererChromium::cleanupSharedObjects):
1824         * platform/graphics/chromium/LayerRendererChromium.h:
1825         (WebCore::LayerRendererChromium::renderSurfaceTextureAllocator):
1826         (WebCore::LayerRendererChromium::contentsTextureAllocator):
1827         * platform/graphics/chromium/LayerTextureUpdater.h:
1828         * platform/graphics/chromium/LayerTextureUpdaterCanvas.cpp:
1829         (WebCore::LayerTextureUpdaterBitmap::updateTextureRect):
1830         (WebCore::LayerTextureUpdaterSkPicture::updateTextureRect):
1831         * platform/graphics/chromium/LayerTextureUpdaterCanvas.h:
1832         * platform/graphics/chromium/ManagedTexture.cpp:
1833         (WebCore::ManagedTexture::bindTexture):
1834         (WebCore::ManagedTexture::framebufferTexture2D):
1835         * platform/graphics/chromium/ManagedTexture.h:
1836         * platform/graphics/chromium/TextureManager.cpp:
1837         (WebCore::TextureManager::memoryUseBytes):
1838         (WebCore::TextureManager::deleteEvictedTextures):
1839         (WebCore::TextureManager::evictAndDeleteAllTextures):
1840         (WebCore::TextureManager::removeTexture):
1841         (WebCore::TextureManager::allocateTexture):
1842         (WebCore::TextureManager::requestTexture):
1843         * platform/graphics/chromium/TextureManager.h:
1844         (WebCore::TextureAllocator::~TextureAllocator):
1845         * platform/graphics/chromium/TiledLayerChromium.cpp:
1846         (WebCore::TiledLayerChromium::updateCompositorResources):
1847         * platform/graphics/chromium/TiledLayerChromium.h:
1848         * platform/graphics/chromium/TrackingTextureAllocator.cpp: Added.
1849         (WebCore::TrackingTextureAllocator::TrackingTextureAllocator):
1850         (WebCore::TrackingTextureAllocator::~TrackingTextureAllocator):
1851         (WebCore::TrackingTextureAllocator::createTexture):
1852         (WebCore::TrackingTextureAllocator::deleteTexture):
1853         * platform/graphics/chromium/TrackingTextureAllocator.h: Copied from Source/WebCore/platform/graphics/chromium/ManagedTexture.h.
1854         (WebCore::TrackingTextureAllocator::create):
1855         (WebCore::TrackingTextureAllocator::currentMemoryUseBytes):
1856         * platform/graphics/chromium/VideoLayerChromium.cpp:
1857         (WebCore::VideoLayerChromium::updateCompositorResources):
1858         (WebCore::VideoLayerChromium::updateTexture):
1859         * platform/graphics/chromium/VideoLayerChromium.h:
1860         * platform/graphics/chromium/WebGLLayerChromium.cpp:
1861         (WebCore::WebGLLayerChromium::updateCompositorResources):
1862         * platform/graphics/chromium/WebGLLayerChromium.h:
1863         * platform/graphics/chromium/cc/CCHeadsUpDisplay.cpp:
1864         (WebCore::CCHeadsUpDisplay::draw):
1865         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
1866         (WebCore::CCLayerTreeHost::~CCLayerTreeHost):
1867         (WebCore::CCLayerTreeHost::deleteContentsTexturesOnCCThread):
1868         (WebCore::CCLayerTreeHost::commitToOnCCThread):
1869         (WebCore::CCLayerTreeHost::didRecreateGraphicsContext):
1870         (WebCore::CCLayerTreeHost::updateCompositorResources):
1871         * platform/graphics/chromium/cc/CCLayerTreeHost.h:
1872         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
1873         (WebCore::CCLayerTreeHostImpl::contentsTextureAllocator):
1874         (WebCore::CCLayerTreeHostImpl::initializeLayerRenderer):
1875         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.h:
1876         * platform/graphics/chromium/cc/CCRenderSurface.cpp:
1877         (WebCore::CCRenderSurface::drawSurface):
1878         * platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
1879         (WebCore::CCSingleThreadProxy::setNeedsCommit):
1880         (WebCore::CCSingleThreadProxy::stop):
1881         (WebCore::CCSingleThreadProxy::recreateContextIfNeeded):
1882         (WebCore::CCSingleThreadProxy::commitIfNeeded):
1883         * platform/graphics/chromium/cc/CCThreadProxy.cpp:
1884         (WebCore::CCThreadProxy::commitOnCCThread):
1885         (WebCore::CCThreadProxy::layerTreeHostClosedOnCCThread):
1886 2011-09-27  James Robinson  <jamesr@chromium.org>
1887
1888         [chromium] Only initiate the beginFrameAndCommit sequence if a commit has been requested
1889         https://bugs.webkit.org/show_bug.cgi?id=68967
1890
1891         Reviewed by Kenneth Russell.
1892
1893         When updating the scheduler state, we should only initiate a new commit flow if a commit has been requested (as
1894         opposed to only a redraw).
1895
1896         Covered by the unit test CCLayerTreeHostTestSetNeedsRedraw with USE(THREADED_COMPOSITING) set to true.
1897
1898         * platform/graphics/chromium/cc/CCThreadProxy.cpp:
1899         (WebCore::CCThreadProxy::updateSchedulerStateOnCCThread):
1900
1901 2011-09-27  Kentaro Hara  <haraken@chromium.com>
1902
1903         Implement a PageTransitionEvent constructor for V8
1904         https://bugs.webkit.org/show_bug.cgi?id=68067
1905
1906         Reviewed by Adam Barth.
1907
1908         Test: fast/events/constructors/page-transition-event-constructor.html
1909
1910         * bindings/v8/custom/V8EventConstructors.cpp: Added the PageTransitionEvent constructor.
1911         * dom/PageTransitionEvent.idl: Added a 'V8CustomConstructor' attribute.
1912
1913 2011-09-27  Kentaro Hara  <haraken@chromium.org>
1914
1915         Implement a MessageEvent constructor for JSC
1916         https://bugs.webkit.org/show_bug.cgi?id=68883
1917
1918         Reviewed by Adam Barth.
1919
1920         The spec for MessageEvent is here:
1921         http://www.whatwg.org/specs/web-apps/current-work/#messageevent
1922
1923         Currently, some tests in fast/events/constructors/message-event-constructor.html
1924         are failing or crashing in DRT, as we commented in the test file.
1925         This is because MessageEvent.data is implemented as SerializedScriptValue,
1926         and thus it cannot keep ScriptValue passed by JavaScript.
1927         This is the same issue as the bug (https://bugs.webkit.org/show_bug.cgi?id=68345).
1928         We will soon make a follow-up patch to fix these failures, after this
1929         patch is landed.
1930
1931         Test: fast/events/constructors/message-event-constructor.html
1932
1933         * bindings/generic/EventConstructors.h: Added a definition for the MessageEvent constructor.
1934         * bindings/js/JSDictionary.cpp:
1935         (WebCore::JSDictionary::convertValue): Generates MessagePortArray from the list of message ports in the format of JSValues.
1936         * bindings/js/JSDictionary.h:
1937         * bindings/js/JSEventConstructors.cpp: Added #includes for MessageEvent.
1938         * dom/MessageEvent.cpp:
1939         (WebCore::MessageEventInit::MessageEventInit):
1940         (WebCore::MessageEvent::MessageEvent):
1941         (WebCore::MessageEvent::initMessageEvent):
1942         * dom/MessageEvent.h: Added a definition for MessageEvent. Removed an extra leading spaces.
1943         (WebCore::MessageEvent::create):
1944         (WebCore::MessageEvent::origin):
1945         (WebCore::MessageEvent::lastEventId):
1946         (WebCore::MessageEvent::source):
1947         (WebCore::MessageEvent::ports):
1948         (WebCore::MessageEvent::dataType):
1949         (WebCore::MessageEvent::dataAsSerializedScriptValue):
1950         (WebCore::MessageEvent::dataAsString):
1951         (WebCore::MessageEvent::dataAsBlob):
1952         (WebCore::MessageEvent::dataAsArrayBuffer):
1953         * dom/MessageEvent.idl: Makes MessageEvent constructible.
1954
1955 2011-09-27  Eric Uhrhane  <ericu@chromium.org>
1956
1957         [Chromium/FileWriter] race condition in FileWriter completion can lead to assert
1958         https://bugs.webkit.org/show_bug.cgi?id=67684
1959
1960         Reviewed by David Levin.
1961
1962         Tests: fast/filesystem/file-writer-abort-continue.html
1963                fast/filesystem/file-writer-abort.html
1964
1965         Track the state of the backend and be prepared for reentrant user
1966         requests.  Limit recursion depth to an arbitrary small constant.
1967         * fileapi/FileWriter.cpp: Lots of event-handling changes.
1968         * fileapi/FileWriter.h:
1969
1970 2011-09-27  Mihai Parparita  <mihaip@chromium.org>
1971
1972         Unreviewed, rolling out r96141.
1973         http://trac.webkit.org/changeset/96141
1974         https://bugs.webkit.org/show_bug.cgi?id=68190
1975
1976         Breaks PrerenderBrowserTest.PrerenderHTML5Video in
1977         browser_tests
1978
1979         * WebCore.gypi:
1980         * platform/graphics/chromium/Canvas2DLayerChromium.cpp:
1981         (WebCore::Canvas2DLayerChromium::updateCompositorResources):
1982         * platform/graphics/chromium/Canvas2DLayerChromium.h:
1983         * platform/graphics/chromium/ImageLayerChromium.cpp:
1984         (WebCore::ImageLayerTextureUpdater::updateTextureRect):
1985         * platform/graphics/chromium/LayerChromium.h:
1986         (WebCore::LayerChromium::updateCompositorResources):
1987         * platform/graphics/chromium/LayerRendererChromium.cpp:
1988         (WebCore::LayerRendererChromium::LayerRendererChromium):
1989         (WebCore::LayerRendererChromium::releaseRenderSurfaceTextures):
1990         (WebCore::LayerRendererChromium::drawLayers):
1991         (WebCore::LayerRendererChromium::createLayerTexture):
1992         (WebCore::LayerRendererChromium::deleteLayerTexture):
1993         (WebCore::LayerRendererChromium::useRenderSurface):
1994         (WebCore::LayerRendererChromium::initializeSharedObjects):
1995         (WebCore::LayerRendererChromium::cleanupSharedObjects):
1996         * platform/graphics/chromium/LayerRendererChromium.h:
1997         (WebCore::LayerRendererChromium::setContentsTextureMemoryUseBytes):
1998         * platform/graphics/chromium/LayerTextureUpdater.h:
1999         * platform/graphics/chromium/LayerTextureUpdaterCanvas.cpp:
2000         (WebCore::LayerTextureUpdaterBitmap::updateTextureRect):
2001         (WebCore::LayerTextureUpdaterSkPicture::updateTextureRect):
2002         * platform/graphics/chromium/LayerTextureUpdaterCanvas.h:
2003         * platform/graphics/chromium/ManagedTexture.cpp:
2004         (WebCore::ManagedTexture::bindTexture):
2005         (WebCore::ManagedTexture::framebufferTexture2D):
2006         * platform/graphics/chromium/ManagedTexture.h:
2007         * platform/graphics/chromium/TextureManager.cpp:
2008         (WebCore::memoryUseBytes):
2009         (WebCore::TextureManager::deleteEvictedTextures):
2010         (WebCore::TextureManager::evictAndDeleteAllTextures):
2011         (WebCore::TextureManager::removeTexture):
2012         (WebCore::TextureManager::allocateTexture):
2013         (WebCore::TextureManager::requestTexture):
2014         * platform/graphics/chromium/TextureManager.h:
2015         * platform/graphics/chromium/TiledLayerChromium.cpp:
2016         (WebCore::TiledLayerChromium::updateCompositorResources):
2017         * platform/graphics/chromium/TiledLayerChromium.h:
2018         * platform/graphics/chromium/TrackingTextureAllocator.cpp: Removed.
2019         * platform/graphics/chromium/TrackingTextureAllocator.h: Removed.
2020         * platform/graphics/chromium/VideoLayerChromium.cpp:
2021         (WebCore::VideoLayerChromium::updateCompositorResources):
2022         (WebCore::VideoLayerChromium::updateTexture):
2023         * platform/graphics/chromium/VideoLayerChromium.h:
2024         * platform/graphics/chromium/WebGLLayerChromium.cpp:
2025         (WebCore::WebGLLayerChromium::updateCompositorResources):
2026         * platform/graphics/chromium/WebGLLayerChromium.h:
2027         * platform/graphics/chromium/cc/CCHeadsUpDisplay.cpp:
2028         (WebCore::CCHeadsUpDisplay::draw):
2029         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
2030         (WebCore::CCLayerTreeHost::~CCLayerTreeHost):
2031         (WebCore::CCLayerTreeHost::deleteContentsTextures):
2032         (WebCore::CCLayerTreeHost::commitTo):
2033         (WebCore::CCLayerTreeHost::didRecreateGraphicsContext):
2034         (WebCore::CCLayerTreeHost::updateCompositorResources):
2035         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp.rej: Removed.
2036         * platform/graphics/chromium/cc/CCLayerTreeHost.h:
2037         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
2038         (WebCore::CCLayerTreeHostImpl::initializeLayerRenderer):
2039         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.h:
2040         * platform/graphics/chromium/cc/CCRenderSurface.cpp:
2041         (WebCore::CCRenderSurface::drawSurface):
2042         * platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
2043         (WebCore::CCSingleThreadProxy::setNeedsCommit):
2044         (WebCore::CCSingleThreadProxy::stop):
2045         (WebCore::CCSingleThreadProxy::recreateContextIfNeeded):
2046         (WebCore::CCSingleThreadProxy::commitIfNeeded):
2047         * platform/graphics/chromium/cc/CCThreadProxy.cpp:
2048         (WebCore::CCThreadProxy::commitOnCCThread):
2049         (WebCore::CCThreadProxy::layerTreeHostClosedOnCCThread):
2050
2051 2011-09-27  James Robinson  <jamesr@chromium.org>
2052
2053         [chromium] Guard accelerated skia drawing logic with its own define to remove dependency on USE(THREADED_COMPOSITING)
2054         https://bugs.webkit.org/show_bug.cgi?id=68953
2055
2056         Reviewed by Kenneth Russell.
2057
2058         The accelerated drawing code is incompatible with threaded compositing. Rather than tying the two behaviors
2059         together via USE(THREADED_COMPOSITING), this patch guards the accelerated drawing code with its own USE() guard
2060         so the features can advance independently. Today, it's not safe to set both of these flags.
2061
2062         * platform/graphics/chromium/ContentLayerChromium.cpp:
2063         (WebCore::ContentLayerChromium::createTextureUpdater):
2064         * platform/graphics/chromium/LayerTextureUpdaterCanvas.cpp:
2065         * platform/graphics/chromium/LayerTextureUpdaterCanvas.h:
2066
2067 2011-09-27  Luke Macpherson   <macpherson@chromium.org>
2068
2069         Slightly improve performance of CSSStyleApplyProperty handler lookup.
2070         https://bugs.webkit.org/show_bug.cgi?id=68868
2071
2072         Reviewed by Eric Seidel.
2073
2074         No new tests as no functionality changed.
2075
2076         * css/CSSStyleApplyProperty.h:
2077         (WebCore::CSSStyleApplyProperty::propertyHandler):
2078         Make propertyHandler() public and remove redirecting functions.
2079         * css/CSSStyleSelector.cpp:
2080         (WebCore::CSSStyleSelector::applyProperty):
2081         Perform property handler lookup once and reuse the result.
2082
2083 2011-09-27  Kent Tamura  <tkent@chromium.org>
2084
2085         [V8] element.dataset.nonExistingKey should return undefined.
2086         https://bugs.webkit.org/show_bug.cgi?id=68877
2087
2088         Reviewed by Nate Chapin.
2089
2090         * bindings/v8/custom/V8DOMStringMapCustom.cpp:
2091         (WebCore::V8DOMStringMap::namedPropertyGetter): Uses v8StringOrUndefined() instead of v8String().
2092
2093 2011-09-23  Mark Hahnenberg  <mhahnenberg@apple.com>
2094
2095         Add static version of JSCell::visitChildren
2096         https://bugs.webkit.org/show_bug.cgi?id=68404
2097
2098         Reviewed by Darin Adler.
2099
2100         No new tests.
2101
2102         In this patch we just extract the bodies of the virtual visitChildren methods
2103         throughout the JSCell inheritance hierarchy out into static methods, which are
2104         now called from the virtual methods.  This is an intermediate step in trying to
2105         move the virtual-ness of visitChildren into our own custom vtable stored in
2106         ClassInfo.  We need to convert the methods to static methods in order to be
2107         able to more easily store and refer to them in our custom vtable since normal
2108         member methods store some implicit information in their types, making it
2109         impossible to store them generically in ClassInfo.
2110
2111         * WebCore.exp.in:
2112         * bindings/js/JSAttrCustom.cpp:
2113         (WebCore::JSAttr::visitChildrenVirtual):
2114         (WebCore::JSAttr::visitChildren):
2115         * bindings/js/JSAudioContextCustom.cpp:
2116         (WebCore::JSAudioContext::visitChildrenVirtual):
2117         (WebCore::JSAudioContext::visitChildren):
2118         * bindings/js/JSCSSRuleCustom.cpp:
2119         (WebCore::JSCSSRule::visitChildrenVirtual):
2120         (WebCore::JSCSSRule::visitChildren):
2121         * bindings/js/JSCSSStyleDeclarationCustom.cpp:
2122         (WebCore::JSCSSStyleDeclaration::visitChildrenVirtual):
2123         (WebCore::JSCSSStyleDeclaration::visitChildren):
2124         * bindings/js/JSCanvasRenderingContextCustom.cpp:
2125         (WebCore::JSCanvasRenderingContext::visitChildrenVirtual):
2126         (WebCore::JSCanvasRenderingContext::visitChildren):
2127         * bindings/js/JSDOMGlobalObject.cpp:
2128         (WebCore::JSDOMGlobalObject::visitChildrenVirtual):
2129         (WebCore::JSDOMGlobalObject::visitChildren):
2130         * bindings/js/JSDOMGlobalObject.h:
2131         * bindings/js/JSDOMWindowCustom.cpp:
2132         (WebCore::JSDOMWindow::visitChildrenVirtual):
2133         (WebCore::JSDOMWindow::visitChildren):
2134         * bindings/js/JSDOMWindowShell.cpp:
2135         (WebCore::JSDOMWindowShell::visitChildrenVirtual):
2136         (WebCore::JSDOMWindowShell::visitChildren):
2137         * bindings/js/JSDOMWindowShell.h:
2138         * bindings/js/JSJavaScriptAudioNodeCustom.cpp:
2139         (WebCore::JSJavaScriptAudioNode::visitChildrenVirtual):
2140         (WebCore::JSJavaScriptAudioNode::visitChildren):
2141         * bindings/js/JSMessageChannelCustom.cpp:
2142         (WebCore::JSMessageChannel::visitChildrenVirtual):
2143         (WebCore::JSMessageChannel::visitChildren):
2144         * bindings/js/JSMessagePortCustom.cpp:
2145         (WebCore::JSMessagePort::visitChildrenVirtual):
2146         (WebCore::JSMessagePort::visitChildren):
2147         * bindings/js/JSNamedNodeMapCustom.cpp:
2148         (WebCore::JSNamedNodeMap::visitChildrenVirtual):
2149         (WebCore::JSNamedNodeMap::visitChildren):
2150         * bindings/js/JSNodeCustom.cpp:
2151         (WebCore::JSNode::visitChildrenVirtual):
2152         (WebCore::JSNode::visitChildren):
2153         * bindings/js/JSNodeFilterCustom.cpp:
2154         (WebCore::JSNodeFilter::visitChildrenVirtual):
2155         (WebCore::JSNodeFilter::visitChildren):
2156         * bindings/js/JSNodeIteratorCustom.cpp:
2157         (WebCore::JSNodeIterator::visitChildrenVirtual):
2158         (WebCore::JSNodeIterator::visitChildren):
2159         * bindings/js/JSSVGElementInstanceCustom.cpp:
2160         (WebCore::JSSVGElementInstance::visitChildrenVirtual):
2161         (WebCore::JSSVGElementInstance::visitChildren):
2162         * bindings/js/JSSharedWorkerCustom.cpp:
2163         (WebCore::JSSharedWorker::visitChildrenVirtual):
2164         (WebCore::JSSharedWorker::visitChildren):
2165         * bindings/js/JSStyleSheetCustom.cpp:
2166         (WebCore::JSStyleSheet::visitChildrenVirtual):
2167         (WebCore::JSStyleSheet::visitChildren):
2168         * bindings/js/JSTreeWalkerCustom.cpp:
2169         (WebCore::JSTreeWalker::visitChildrenVirtual):
2170         (WebCore::JSTreeWalker::visitChildren):
2171         * bindings/js/JSWebGLRenderingContextCustom.cpp:
2172         (WebCore::JSWebGLRenderingContext::visitChildrenVirtual):
2173         (WebCore::JSWebGLRenderingContext::visitChildren):
2174         * bindings/js/JSWorkerContextCustom.cpp:
2175         (WebCore::JSWorkerContext::visitChildrenVirtual):
2176         (WebCore::JSWorkerContext::visitChildren):
2177         * bindings/js/JSXMLHttpRequestCustom.cpp:
2178         (WebCore::JSXMLHttpRequest::visitChildrenVirtual):
2179         (WebCore::JSXMLHttpRequest::visitChildren):
2180         * bindings/js/JSXPathResultCustom.cpp:
2181         (WebCore::JSXPathResult::visitChildrenVirtual):
2182         (WebCore::JSXPathResult::visitChildren):
2183         * bindings/scripts/CodeGeneratorJS.pm:
2184         (GenerateHeader):
2185         (GenerateImplementation):
2186         * bindings/scripts/test/JS/JSTestObj.cpp:
2187         (WebCore::JSTestObj::visitChildrenVirtual):
2188         (WebCore::JSTestObj::visitChildren):
2189         * bindings/scripts/test/JS/JSTestObj.h:
2190         * bridge/qt/qt_instance.cpp:
2191         (JSC::Bindings::QtRuntimeObject::visitChildrenVirtual):
2192         (JSC::Bindings::QtRuntimeObject::visitChildren):
2193         * bridge/qt/qt_runtime.cpp:
2194         (JSC::Bindings::QtRuntimeMetaMethod::visitChildrenVirtual):
2195         (JSC::Bindings::QtRuntimeMetaMethod::visitChildren):
2196         * bridge/qt/qt_runtime.h:
2197         * workers/WorkerContext.h:
2198
2199 <<<<<<< .mine
2200 2011-09-27  Mark Hahnenberg  <mhahnenberg@apple.com>
2201
2202         Add static version of JSCell::getCallData
2203         https://bugs.webkit.org/show_bug.cgi?id=68741
2204
2205         Reviewed by Darin Adler.
2206
2207         No new tests.
2208
2209         In this patch we just extract the bodies of the virtual getCallData methods
2210         throughout the JSCell inheritance hierarchy out into static methods, which are
2211         now called from the virtual methods.  This is an intermediate step in trying to
2212         move the virtual-ness of getCallData into our own method table stored in
2213         ClassInfo.  We need to convert the methods to static methods because static methods
2214         can be represented as function pointers rather than pointers to member functions, and
2215         function pointers are smaller and faster to call than pointers to member functions.
2216
2217         * WebCore.exp.in:
2218         * bindings/js/JSCallbackData.cpp:
2219         (WebCore::JSCallbackData::invokeCallback):
2220         * bindings/js/JSCustomXPathNSResolver.cpp:
2221         (WebCore::JSCustomXPathNSResolver::lookupNamespaceURI):
2222         * bindings/js/JSErrorHandler.cpp:
2223         (WebCore::JSErrorHandler::handleEvent):
2224         * bindings/js/JSEventListener.cpp:
2225         (WebCore::JSEventListener::handleEvent):
2226         * bindings/js/JSHTMLAllCollectionCustom.cpp:
2227         (WebCore::JSHTMLAllCollection::getCallDataVirtual):
2228         (WebCore::JSHTMLAllCollection::getCallData):
2229         * bindings/js/JSHTMLAppletElementCustom.cpp:
2230         (WebCore::JSHTMLAppletElement::getCallDataVirtual):
2231         (WebCore::JSHTMLAppletElement::getCallData):
2232         * bindings/js/JSHTMLEmbedElementCustom.cpp:
2233         (WebCore::JSHTMLEmbedElement::getCallDataVirtual):
2234         (WebCore::JSHTMLEmbedElement::getCallData):
2235         * bindings/js/JSHTMLObjectElementCustom.cpp:
2236         (WebCore::JSHTMLObjectElement::getCallDataVirtual):
2237         (WebCore::JSHTMLObjectElement::getCallData):
2238         * bindings/js/JSInjectedScriptHostCustom.cpp:
2239         (WebCore::JSInjectedScriptHost::evaluate):
2240         * bindings/js/JSPluginElementFunctions.cpp:
2241         (WebCore::runtimeObjectGetCallData):
2242         * bindings/scripts/CodeGeneratorJS.pm:
2243         (GenerateHeader):
2244         * bridge/objc/objc_runtime.h:
2245         * bridge/objc/objc_runtime.mm:
2246         (JSC::Bindings::ObjcFallbackObjectImp::getCallDataVirtual):
2247         (JSC::Bindings::ObjcFallbackObjectImp::getCallData):
2248         * bridge/qt/qt_runtime.cpp:
2249         (JSC::Bindings::QtRuntimeMetaMethod::getCallDataVirtual):
2250         (JSC::Bindings::QtRuntimeMetaMethod::getCallData):
2251         (JSC::Bindings::QtRuntimeConnectionMethod::call):
2252         (JSC::Bindings::QtRuntimeConnectionMethod::getCallDataVirtual):
2253         (JSC::Bindings::QtRuntimeConnectionMethod::getCallData):
2254         (JSC::Bindings::isJavaScriptFunction):
2255         * bridge/qt/qt_runtime.h:
2256         * bridge/runtime_method.cpp:
2257         (JSC::RuntimeMethod::getCallDataVirtual):
2258         (JSC::RuntimeMethod::getCallData):
2259         * bridge/runtime_method.h:
2260         * bridge/runtime_object.cpp:
2261         (JSC::Bindings::RuntimeObject::getCallDataVirtual):
2262         (JSC::Bindings::RuntimeObject::getCallData):
2263         * bridge/runtime_object.h:
2264
2265 =======
2266 2011-09-27  Dean Jackson  <dino@apple.com>
2267
2268         Unreviewed. Remove duplicate FilterOperation{s}.h that were
2269         accidentally added to the top level of the Xcode project.
2270
2271         * WebCore.xcodeproj/project.pbxproj:
2272
2273 >>>>>>> .r96163
2274 2011-09-27  David Hyatt  <hyatt@apple.com>
2275
2276         https://bugs.webkit.org/show_bug.cgi?id=68940
2277
2278         Narrow the float/lines pagination heuristic to only kick in if
2279         the previous line broke cleanly and if the floats are occurring
2280         at the start of the line.
2281
2282         Reviewed by Dan Bernstein.
2283
2284         * rendering/RenderBlockLineLayout.cpp:
2285         (WebCore::RenderBlock::positionNewFloatOnLine):
2286
2287 2011-09-27  James Robinson  <jamesr@chromium.org>
2288
2289         Add a mechanism to test for the compositing tree mutated during painting
2290         https://bugs.webkit.org/show_bug.cgi?id=68738
2291
2292         Reviewed by Adam Barth.
2293
2294         Sets a static bool during GraphicsLayer::paintGraphicsLayerContents and ASSERT()s that we never create or
2295         destroy a GraphicsLayer inside this function. Painting should never mutate the GraphicsLayer tree.
2296
2297         Test: compositing/video/video-with-invalid-source.html
2298
2299         * platform/graphics/GraphicsLayer.cpp:
2300         (WebCore::GraphicsLayer::GraphicsLayer):
2301         (WebCore::GraphicsLayer::~GraphicsLayer):
2302         (WebCore::GraphicsLayer::paintGraphicsLayerContents):
2303
2304 2011-09-27  Ojan Vafai  <ojan@chromium.org>
2305
2306         take padding/border on flexbox into account with direction:rtl
2307         https://bugs.webkit.org/show_bug.cgi?id=68846
2308
2309         Reviewed by David Hyatt.
2310
2311         Use the logicalWidth of the flexbox instead of it's availableLogicalWidth
2312         so that we include the flexbox's border and padding.
2313
2314         Also, change all uses of availableLogicalWidth to contentLogicalWidth.
2315         The former takes multi-column into account. Multi-column flexboxes don't
2316         make any sense, so there's no point in calling availableLogicalWidth.
2317
2318         * rendering/RenderFlexibleBox.cpp:
2319         (WebCore::RenderFlexibleBox::flowAwareLogicalWidth):
2320         (WebCore::RenderFlexibleBox::flowAwareContentLogicalWidth):
2321         (WebCore::RenderFlexibleBox::computePreferredLogicalWidth):
2322         (WebCore::RenderFlexibleBox::runFreeSpaceAllocationAlgorithmInlineDirection):
2323         (WebCore::RenderFlexibleBox::layoutAndPlaceChildrenInlineDirection):
2324         * rendering/RenderFlexibleBox.h:
2325
2326 2011-09-27  Sheriff Bot  <webkit.review.bot@gmail.com>
2327
2328         Unreviewed, rolling out r96139.
2329         http://trac.webkit.org/changeset/96139
2330         https://bugs.webkit.org/show_bug.cgi?id=68933
2331
2332         Broke table-percent-height.html on Mac bots (Requested by
2333         mwenge2 on #webkit).
2334
2335         * rendering/RenderBox.cpp:
2336         (WebCore::RenderBox::computeReplacedLogicalWidthUsing):
2337         (WebCore::RenderBox::computeReplacedLogicalHeightUsing):
2338         * rendering/RenderBox.h:
2339
2340 2011-09-27  Tim Horton  <timothy_horton@apple.com>
2341
2342         REGRESSION(65665): Pattern size being clamped to SVG size can prevent transformed elements from being fully covered by userSpaceOnUse patterns
2343         https://bugs.webkit.org/show_bug.cgi?id=67700
2344         <rdar://problem/10125102>
2345
2346         Reviewed by Darin Adler.
2347
2348         Clamp all resources to the same size, 4096x4096 (arbitrarily chosen), instead of to the size
2349         of the <svg> element. This fixes the case where a transformed element displays part of a resource
2350         outside of the size of the <svg> element.
2351
2352         When drawing an oversized pattern into its tile image, scale the content down to fit. When drawing
2353         the tile image to the screen, scale it back up to fit the expected area. This will cause pixelation
2354         when patterns are over the 4k limit.
2355
2356         Tests: svg/custom/transformed-pattern-clamp-svg-root.svg, svg/custom/oversized-pattern-scale.svg
2357
2358         * rendering/svg/RenderSVGResourceClipper.cpp:
2359         (WebCore::RenderSVGResourceClipper::applyClippingToContext):
2360         * rendering/svg/RenderSVGResourceGradient.cpp:
2361         (WebCore::createMaskAndSwapContextForTextGradient):
2362         (WebCore::clipToTextMask):
2363         * rendering/svg/RenderSVGResourceMasker.cpp:
2364         (WebCore::RenderSVGResourceMasker::applyResource):
2365         * rendering/svg/RenderSVGResourcePattern.cpp:
2366         (WebCore::RenderSVGResourcePattern::applyResource):
2367         (WebCore::RenderSVGResourcePattern::createTileImage):
2368         * rendering/svg/RenderSVGResourcePattern.h:
2369         * rendering/svg/SVGImageBufferTools.cpp:
2370         (WebCore::SVGImageBufferTools::clampedAbsoluteTargetRect):
2371         * rendering/svg/SVGImageBufferTools.h:
2372
2373 2011-09-27  Dan Bernstein  <mitz@apple.com>
2374
2375         <rdar://problem/10098679> Assertion failure in RenderLayer::paintPaginatedChildLayer()
2376
2377         Reviewed by Simon Fraser.
2378
2379         Test: fast/dynamic/layer-no-longer-paginated.html
2380
2381         FrameView::layout() calls adjustViewSize() before calling RenderLayer::updateLayerPositions().
2382         The former may trigger painting with a layer tree that is not entirely up-to-date. Specifically,
2383         the isPaginated() state of a layer may be incorrect, leading to the assertion in this bug. Instead
2384         of asserting, return early and count on the upcoming updateLayerPositions() to repaint as needed.
2385
2386         * rendering/RenderLayer.cpp:
2387         (WebCore::RenderLayer::paintPaginatedChildLayer): Replaced the assertion with an early return.
2388
2389 2011-09-27  Ojan Vafai  <ojan@chromium.org>
2390
2391         offsetTop/offsetLeft return the wrong values for horizontal-bt/vertical-rl writing modes
2392         https://bugs.webkit.org/show_bug.cgi?id=68304
2393
2394         Reviewed by David Hyatt.
2395
2396         When grabbing the x/y values of the RenderBox, we need to take writing mode
2397         flipping into account.
2398
2399         Test: fast/dom/offset-position-writing-modes.html
2400
2401         * rendering/RenderBox.cpp:
2402         (WebCore::RenderBox::locationIncludingFlipping):
2403         * rendering/RenderBox.h:
2404         (WebCore::RenderBox::yFlippedForWritingMode):
2405         (WebCore::RenderBox::xFlippedForWritingMode):
2406         * rendering/RenderBoxModelObject.cpp:
2407         (WebCore::RenderBoxModelObject::offsetLeft):
2408         (WebCore::RenderBoxModelObject::offsetTop):
2409
2410 2011-09-27  Tim Horton  <timothy_horton@apple.com>
2411
2412         Rapidly refreshing a feMorphology[erode] with r=0 can sometimes cause display corruption
2413         https://bugs.webkit.org/show_bug.cgi?id=68816
2414         <rdar://problem/10186468>
2415
2416         Reviewed by Simon Fraser.
2417
2418         If a filter returns without writing into its result buffer, make sure to return an cleared buffer.
2419
2420         Test: svg/filters/feMorphology-zero-radius.svg
2421
2422         * platform/graphics/filters/FEMorphology.cpp:
2423         (WebCore::FEMorphology::apply):
2424         * platform/graphics/filters/FETurbulence.cpp:
2425         (WebCore::FETurbulence::apply):
2426
2427 2011-09-27  Ryosuke Niwa  <rniwa@webkit.org>
2428
2429         CompositeEditCommand::prune should remove subtree at once
2430         https://bugs.webkit.org/show_bug.cgi?id=68866
2431
2432         Reviewed by Darin Adler.
2433
2434         Extracted the logic to find the highest ancestor to remove as highestNodeToRemoveInPruning from prune.
2435         This reduces the number of node removals from O(n) to O(1) where n is the depth of the tree.
2436
2437         * editing/CompositeEditCommand.cpp:
2438         (WebCore::hasARenderedDescendant): Takes excludedNode in addition to node. excludedNode is used to ignore
2439         the child node from which we climbed up the tree in highestNodeToRemoveInPruning.
2440         (WebCore::highestNodeToRemoveInPruning): Extracted from prune.
2441         (WebCore::CompositeEditCommand::prune):
2442         (WebCore::CompositeEditCommand::breakOutOfEmptyMailBlockquotedParagraph):
2443
2444 2011-09-27  David Hyatt  <hyatt@apple.com>
2445
2446         https://bugs.webkit.org/show_bug.cgi?id=68922
2447
2448         Paginated floats should not grow the height of a block when they don't end
2449         up being tied to a line. Floats encountered in skipLeadingWhitespace that
2450         paginate can't immediately grow the height of the block, since there may not
2451         end up being any actual line content. We only want to push the height of the
2452         block down if we have actual line content that we want to keep with the
2453         float.
2454
2455         Make positionNewFloatOnLine check if the line is empty or not, and if it is,
2456         it just puts the desired pagination strut into LineInfo. If and when actual
2457         line content is encountered and setEmpty becomes false, we'll grow the block
2458         height then.
2459
2460         Reviewed by Dan Bernstein.
2461
2462         Added fast/multicol/float-paginate-empty-lines.html and
2463         fast/regions/webkit-flow-double-pagination-float-push.html.
2464
2465         * rendering/RenderBlock.h:
2466         * rendering/RenderBlockLineLayout.cpp:
2467         (WebCore::LineWidth::LineWidth):
2468         (WebCore::LineWidth::fitsOnLine):
2469         (WebCore::LineWidth::currentWidth):
2470         (WebCore::LineWidth::uncommittedWidth):
2471         (WebCore::LineWidth::committedWidth):
2472         (WebCore::LineWidth::availableWidth):
2473         (WebCore::LineWidth::addUncommittedWidth):
2474         (WebCore::LineWidth::commit):
2475         (WebCore::LineWidth::computeAvailableWidthFromLeftAndRight):
2476         (WebCore::LineWidth::updateAvailableWidth):
2477         (WebCore::LineWidth::shrinkAvailableWidthForNewFloatIfNeeded):
2478         (WebCore::LineWidth::applyOverhang):
2479         (WebCore::LineWidth::fitBelowFloats):
2480         (WebCore::LineInfo::LineInfo):
2481         (WebCore::LineInfo::floatPaginationStrut):
2482         (WebCore::LineInfo::setEmpty):
2483         (WebCore::LineInfo::setFloatPaginationStrut):
2484         (WebCore::RenderBlock::LineBreaker::skipLeadingWhitespace):
2485         (WebCore::RenderBlock::LineBreaker::nextLineBreak):
2486         (WebCore::RenderBlock::positionNewFloatOnLine):
2487
2488 2011-09-27  Dean Jackson  <dino@apple.com>
2489
2490         Add 'filter' value to RenderStyle
2491         https://bugs.webkit.org/show_bug.cgi?id=68471
2492
2493         Reviewed by Simon Fraser.
2494
2495         Add a filter property to RenderStyle when ENABLE(CSS_FILTERS)
2496         is turned on. Similar to -webkit-transform, -webkit-filter is a
2497         list of FilterOperations, each identifying what type of operation
2498         it is. This change simply adds the basic objects, it does not
2499         parse the property to generate the list.
2500
2501         * CMakeLists.txt:
2502         * GNUmakefile.list.am:
2503         * WebCore.gypi:
2504         * WebCore.pro:
2505         * WebCore.vcproj/WebCore.vcproj:
2506         * WebCore.xcodeproj/project.pbxproj:
2507         * platform/graphics/filters/FilterOperation.h: Added.
2508         (WebCore::FilterOperation::~FilterOperation):
2509         (WebCore::FilterOperation::operator!=):
2510         (WebCore::FilterOperation::getOperationType):
2511         (WebCore::FilterOperation::isSameType):
2512         (WebCore::FilterOperation::FilterOperation):
2513         (WebCore::ReferenceFilterOperation::create):
2514         (WebCore::ReferenceFilterOperation::reference):
2515         (WebCore::ReferenceFilterOperation::operator==):
2516         (WebCore::ReferenceFilterOperation::ReferenceFilterOperation):
2517         (WebCore::BasicColorMatrixFilterOperation::create):
2518         (WebCore::BasicColorMatrixFilterOperation::amount):
2519         (WebCore::BasicColorMatrixFilterOperation::operator==):
2520         (WebCore::BasicColorMatrixFilterOperation::BasicColorMatrixFilterOperation):
2521         (WebCore::BasicComponentTransferFilterOperation::create):
2522         (WebCore::BasicComponentTransferFilterOperation::amount):
2523         (WebCore::BasicComponentTransferFilterOperation::operator==):
2524         (WebCore::BasicComponentTransferFilterOperation::BasicComponentTransferFilterOperation):
2525         (WebCore::GammaFilterOperation::create):
2526         (WebCore::GammaFilterOperation::amplitude):
2527         (WebCore::GammaFilterOperation::exponent):
2528         (WebCore::GammaFilterOperation::offset):
2529         (WebCore::GammaFilterOperation::operator==):
2530         (WebCore::GammaFilterOperation::GammaFilterOperation):
2531         (WebCore::BlurFilterOperation::create):
2532         (WebCore::BlurFilterOperation::stdDeviationX):
2533         (WebCore::BlurFilterOperation::stdDeviationY):
2534         (WebCore::BlurFilterOperation::operator==):
2535         (WebCore::BlurFilterOperation::BlurFilterOperation):
2536         (WebCore::SharpenFilterOperation::create):
2537         (WebCore::SharpenFilterOperation::radius):
2538         (WebCore::SharpenFilterOperation::threshold):
2539         (WebCore::SharpenFilterOperation::amount):
2540         (WebCore::SharpenFilterOperation::operator==):
2541         (WebCore::SharpenFilterOperation::SharpenFilterOperation):
2542         (WebCore::DropShadowFilterOperation::create):
2543         (WebCore::DropShadowFilterOperation::shadow):
2544         (WebCore::DropShadowFilterOperation::operator==):
2545         (WebCore::DropShadowFilterOperation::DropShadowFilterOperation):
2546         * platform/graphics/filters/FilterOperations.cpp: Added.
2547         (WebCore::FilterOperations::FilterOperations):
2548         (WebCore::FilterOperations::operator==):
2549         * platform/graphics/filters/FilterOperations.h: Added.
2550         (WebCore::FilterOperations::operator!=):
2551         (WebCore::FilterOperations::clear):
2552         (WebCore::FilterOperations::operations):
2553         (WebCore::FilterOperations::size):
2554         (WebCore::FilterOperations::at):
2555         * rendering/style/RenderStyle.cpp:
2556         (WebCore::RenderStyle::RenderStyle):
2557         * rendering/style/RenderStyle.h:
2558         (WebCore::InheritedFlags::filter):
2559         (WebCore::InheritedFlags::hasFilter):
2560         (WebCore::InheritedFlags::setFilter):
2561         (WebCore::InheritedFlags::initialFilter):
2562         * rendering/style/StyleAllInOne.cpp: Added include of new StyleFilterData.
2563         * rendering/style/StyleFilterData.cpp: Added.
2564         (WebCore::StyleFilterData::StyleFilterData):
2565         (WebCore::StyleFilterData::operator==):
2566         * rendering/style/StyleFilterData.h: Added.
2567         (WebCore::StyleFilterData::create):
2568         (WebCore::StyleFilterData::copy):
2569         (WebCore::StyleFilterData::operator!=):
2570         * rendering/style/StyleRareNonInheritedData.cpp:
2571         (WebCore::StyleRareNonInheritedData::StyleRareNonInheritedData):
2572         (WebCore::StyleRareNonInheritedData::operator==):
2573         * rendering/style/StyleRareNonInheritedData.h: New filter DataRef.
2574
2575 2011-09-27  James Robinson  <jamesr@chromium.org>
2576
2577         [chromium] LayerRenderChromium asserts about leaking textures.
2578         https://bugs.webkit.org/show_bug.cgi?id=68190
2579
2580         Reviewed by Kenneth Russell.
2581
2582         This introduces an interface TextureAllocator that the TextureManager uses to allocate/deallocate textures.
2583         This means that TextureManager does not need to depend directly on GraphicsContext3D, other than for the format
2584         enum type, making it more testable.  It also allows us to track the allocations by subclassing TextureAllocator
2585         to make sure we aren't leaking any textures.
2586
2587         * WebCore.gypi:
2588         * platform/graphics/chromium/Canvas2DLayerChromium.cpp:
2589         (WebCore::Canvas2DLayerChromium::updateCompositorResources):
2590         * platform/graphics/chromium/Canvas2DLayerChromium.h:
2591         * platform/graphics/chromium/ImageLayerChromium.cpp:
2592         (WebCore::ImageLayerTextureUpdater::updateTextureRect):
2593         * platform/graphics/chromium/LayerChromium.h:
2594         (WebCore::LayerChromium::updateCompositorResources):
2595         * platform/graphics/chromium/LayerRendererChromium.cpp:
2596         (WebCore::LayerRendererChromium::LayerRendererChromium):
2597         (WebCore::LayerRendererChromium::releaseRenderSurfaceTextures):
2598         (WebCore::LayerRendererChromium::drawLayers):
2599         (WebCore::LayerRendererChromium::useRenderSurface):
2600         (WebCore::LayerRendererChromium::initializeSharedObjects):
2601         (WebCore::LayerRendererChromium::cleanupSharedObjects):
2602         * platform/graphics/chromium/LayerRendererChromium.h:
2603         (WebCore::LayerRendererChromium::renderSurfaceTextureAllocator):
2604         (WebCore::LayerRendererChromium::contentsTextureAllocator):
2605         * platform/graphics/chromium/LayerTextureUpdater.h:
2606         * platform/graphics/chromium/LayerTextureUpdaterCanvas.cpp:
2607         (WebCore::LayerTextureUpdaterBitmap::updateTextureRect):
2608         (WebCore::LayerTextureUpdaterSkPicture::updateTextureRect):
2609         * platform/graphics/chromium/LayerTextureUpdaterCanvas.h:
2610         * platform/graphics/chromium/ManagedTexture.cpp:
2611         (WebCore::ManagedTexture::bindTexture):
2612         (WebCore::ManagedTexture::framebufferTexture2D):
2613         * platform/graphics/chromium/ManagedTexture.h:
2614         * platform/graphics/chromium/TextureManager.cpp:
2615         (WebCore::TextureManager::memoryUseBytes):
2616         (WebCore::TextureManager::deleteEvictedTextures):
2617         (WebCore::TextureManager::evictAndDeleteAllTextures):
2618         (WebCore::TextureManager::removeTexture):
2619         (WebCore::TextureManager::allocateTexture):
2620         (WebCore::TextureManager::requestTexture):
2621         * platform/graphics/chromium/TextureManager.h:
2622         (WebCore::TextureAllocator::~TextureAllocator):
2623         * platform/graphics/chromium/TiledLayerChromium.cpp:
2624         (WebCore::TiledLayerChromium::updateCompositorResources):
2625         * platform/graphics/chromium/TiledLayerChromium.h:
2626         * platform/graphics/chromium/TrackingTextureAllocator.cpp: Added.
2627         (WebCore::TrackingTextureAllocator::TrackingTextureAllocator):
2628         (WebCore::TrackingTextureAllocator::~TrackingTextureAllocator):
2629         (WebCore::TrackingTextureAllocator::createTexture):
2630         (WebCore::TrackingTextureAllocator::deleteTexture):
2631         * platform/graphics/chromium/TrackingTextureAllocator.h: Copied from Source/WebCore/platform/graphics/chromium/ManagedTexture.h.
2632         (WebCore::TrackingTextureAllocator::create):
2633         (WebCore::TrackingTextureAllocator::currentMemoryUseBytes):
2634         * platform/graphics/chromium/VideoLayerChromium.cpp:
2635         (WebCore::VideoLayerChromium::updateCompositorResources):
2636         (WebCore::VideoLayerChromium::updateTexture):
2637         * platform/graphics/chromium/VideoLayerChromium.h:
2638         * platform/graphics/chromium/WebGLLayerChromium.cpp:
2639         (WebCore::WebGLLayerChromium::updateCompositorResources):
2640         * platform/graphics/chromium/WebGLLayerChromium.h:
2641         * platform/graphics/chromium/cc/CCHeadsUpDisplay.cpp:
2642         (WebCore::CCHeadsUpDisplay::draw):
2643         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
2644         (WebCore::CCLayerTreeHost::~CCLayerTreeHost):
2645         (WebCore::CCLayerTreeHost::deleteContentsTexturesOnCCThread):
2646         (WebCore::CCLayerTreeHost::commitToOnCCThread):
2647         (WebCore::CCLayerTreeHost::didRecreateGraphicsContext):
2648         (WebCore::CCLayerTreeHost::updateCompositorResources):
2649         * platform/graphics/chromium/cc/CCLayerTreeHost.h:
2650         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
2651         (WebCore::CCLayerTreeHostImpl::contentsTextureAllocator):
2652         (WebCore::CCLayerTreeHostImpl::initializeLayerRenderer):
2653         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.h:
2654         * platform/graphics/chromium/cc/CCRenderSurface.cpp:
2655         (WebCore::CCRenderSurface::drawSurface):
2656         * platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
2657         (WebCore::CCSingleThreadProxy::setNeedsCommit):
2658         (WebCore::CCSingleThreadProxy::stop):
2659         (WebCore::CCSingleThreadProxy::recreateContextIfNeeded):
2660         (WebCore::CCSingleThreadProxy::commitIfNeeded):
2661         * platform/graphics/chromium/cc/CCThreadProxy.cpp:
2662         (WebCore::CCThreadProxy::commitOnCCThread):
2663         (WebCore::CCThreadProxy::layerTreeHostClosedOnCCThread):
2664
2665 2011-06-28  Robert Hogan  <robert@webkit.org>
2666
2667         Reviewed by David Hyatt.
2668
2669         Replaced elements squeezed when width is specified as percentage inside a table with Auto layout
2670         https://bugs.webkit.org/show_bug.cgi?id=29447
2671
2672         If inserting a 'replaced' element (e.g. image, plugin) in a table cell that is not descendant from
2673         a block with fixed layout then do not squeeze the element, let it use its intrinsic width and height.
2674
2675         Test: fast/replaced/table-percent-width.html
2676
2677         * rendering/RenderBox.cpp:
2678         (WebCore::avoidSqueezingWidth):
2679         (WebCore::avoidSqueezingHeight):
2680         (WebCore::RenderBox::containingBlockReplacedLogicalWidthForContent):
2681         (WebCore::RenderBox::computeReplacedLogicalWidthUsing):
2682         (WebCore::RenderBox::computeReplacedLogicalHeightUsing):
2683         * rendering/RenderBox.h:
2684
2685 2011-09-27  Simon Fraser  <simon.fraser@apple.com>
2686
2687         https://bugs.webkit.org/show_bug.cgi?id=67858
2688
2689         Roll r96070 back in now that the crash has been fixed in r96130.
2690
2691         Reviewed by Darin Adler.
2692
2693         When non-overlay scrollbars are hidden on a composited iframe, nothing invalidated
2694         the scrollbar areas or the scroll corner, so the scrollbars appear to remain.
2695
2696         Fix by invalidating the scrollbars and scroll corner when they are removed. Invalidation
2697         on scrollbar creation appears to happen via updating the scrollbar style.
2698
2699         Tested by compositing/iframes/repaint-after-losing-scrollbars.html which no longer shows
2700         stale scrollbars when run manually, even though the green squares are missing from the
2701         pixel result (bug 67878).
2702
2703         * page/FrameView.cpp:
2704         (WebCore::FrameView::updateScrollCorner): Pass the corner rect into invalidateScrollCorner().
2705         * platform/ScrollView.cpp:
2706         (WebCore::ScrollView::setHasHorizontalScrollbar): Invalidate the scrollbar area if hiding it.
2707         (WebCore::ScrollView::setHasVerticalScrollbar): Ditto.
2708         (WebCore::ScrollView::updateScrollbars): In the case where both scrollbars are going away,
2709         compute the scroll corner rect while we still have scrollbars, and then invalidate it
2710         explicitly. (updateScrollCorner() doesn't, because it doesn't have access to the old corner
2711         rect.)
2712         * platform/ScrollableArea.cpp:
2713         (WebCore::ScrollableArea::invalidateScrollCorner): Pass the rect in, because we can't
2714         compute it in the case where the scrollbars are going away.
2715         * platform/ScrollableArea.h: Pass in a rect to invalidateScrollCorner(), which matches
2716         invalidateScrollbar().
2717         * rendering/RenderLayerCompositor.cpp:
2718         (WebCore::RenderLayerCompositor::destroyRootLayer): Pass the corner rect into invalidateScrollCorner().
2719         * rendering/RenderScrollbarPart.cpp: Ditto.
2720         (WebCore::RenderScrollbarPart::imageChanged): Ditto.
2721
2722 2011-09-27  Mihai Parparita  <mihaip@chromium.org>
2723
2724         Fix Chromium Mac build after r96130.
2725
2726         * page/FrameView.cpp:
2727         (WebCore::FrameView::layerForOverhangAreas):
2728
2729 2011-09-27  Kaustubh Atrawalkar  <kaustubh@motorola.com>
2730
2731         Autofocus on readonly inputs does not focus the element.
2732         https://bugs.webkit.org/show_bug.cgi?id=24092
2733
2734         Reviewed by Ryosuke Niwa.
2735
2736         Readonly input elements should be autofocusable. Removed the check.
2737
2738         Tests: fast/forms/autofocus-readonly-attribute.html
2739
2740         * html/HTMLFormControlElement.cpp:
2741         (WebCore::shouldAutofocus):
2742
2743 2011-09-27  Simon Fraser  <simon.fraser@apple.com>
2744
2745         Clean up how FrameView accesses the RenderView
2746         https://bugs.webkit.org/show_bug.cgi?id=68914
2747
2748         Reviewed by Sam Weinig.
2749
2750         Clean up how FrameView accesses the content renderer of its
2751         frame. Previously, this was done in several different ways,
2752         only some of which did null-checking.
2753
2754         Use an inline method to avoid having to expose Frame
2755         in the header.
2756
2757         Standardize the terminology to use 'root' for this RenderView.
2758
2759         * page/FrameView.cpp:
2760         (WebCore::rootRenderer):
2761         (WebCore::FrameView::setFrameRect):
2762         (WebCore::FrameView::adjustViewSize):
2763         (WebCore::FrameView::updateCompositingLayers):
2764         (WebCore::FrameView::clearBackingStores):
2765         (WebCore::FrameView::restoreBackingStores):
2766         (WebCore::FrameView::layerForHorizontalScrollbar):
2767         (WebCore::FrameView::layerForVerticalScrollbar):
2768         (WebCore::FrameView::layerForScrollCorner):
2769         (WebCore::FrameView::layerForOverhangAreas):
2770         (WebCore::FrameView::syncCompositingStateForThisFrame):
2771         (WebCore::FrameView::hasCompositedContent):
2772         (WebCore::FrameView::enterCompositingMode):
2773         (WebCore::FrameView::isSoftwareRenderable):
2774         (WebCore::FrameView::didMoveOnscreen):
2775         (WebCore::FrameView::willMoveOffscreen):
2776         (WebCore::FrameView::layout):
2777         (WebCore::FrameView::embeddedContentBox):
2778         (WebCore::FrameView::contentsInCompositedLayer):
2779         (WebCore::FrameView::scrollContentsFastPath):
2780         (WebCore::FrameView::scrollContentsSlowPath):
2781         (WebCore::FrameView::maintainScrollPositionAtAnchor):
2782         (WebCore::FrameView::scrollPositionChanged):
2783         (WebCore::FrameView::repaintFixedElementsAfterScrolling):
2784         (WebCore::FrameView::visibleContentsResized):
2785         (WebCore::FrameView::scheduleRelayoutOfSubtree):
2786         (WebCore::FrameView::needsLayout):
2787         (WebCore::FrameView::setNeedsLayout):
2788         (WebCore::FrameView::performPostLayoutTasks):
2789         (WebCore::FrameView::updateControlTints):
2790         (WebCore::FrameView::paintContents):
2791         (WebCore::FrameView::forceLayoutForPagination):
2792         (WebCore::FrameView::adjustPageHeightDeprecated):
2793         (WebCore::FrameView::isVerticalDocument):
2794         (WebCore::FrameView::isFlippedDocument):
2795
2796 2011-09-27  Dominic Mazzoni  <dmazzoni@google.com>
2797
2798         AXObjectCache cleared unnecessarily when non-top Document is detached.
2799         https://bugs.webkit.org/show_bug.cgi?id=68636
2800
2801         Reviewed by Chris Fleizach.
2802
2803         Test: accessibility/deleting-iframe-destroys-axcache.html
2804
2805         * dom/Document.cpp:
2806         (WebCore::Document::attach):
2807         (WebCore::Document::detach):
2808
2809 2011-09-27  Sheriff Bot  <webkit.review.bot@gmail.com>
2810
2811         Unreviewed, rolling out r96108, r96111, r96113, and r96116.
2812         http://trac.webkit.org/changeset/96108
2813         http://trac.webkit.org/changeset/96111
2814         http://trac.webkit.org/changeset/96113
2815         http://trac.webkit.org/changeset/96116
2816         https://bugs.webkit.org/show_bug.cgi?id=68913
2817
2818         Wait for working Qt5 (Requested by ossy on #webkit).
2819
2820         * WebCore.pri:
2821         * platform/graphics/qt/GraphicsLayerQt.cpp:
2822
2823 2011-09-27  Csaba Osztrogonác  <ossy@webkit.org>
2824
2825         Fix ENABLE(SQL_DATABASE)=0 build after r95919
2826         https://bugs.webkit.org/show_bug.cgi?id=68902
2827
2828         r95919 enabled OFFLINE_WEB_APPLICATIONS by default and
2829         it needs SQLite stuff even if ENABLE_SQL_DATABASE=0.
2830
2831         Reviewed by Adam Barth.
2832
2833         * platform/sql/SQLiteAuthorizer.cpp:
2834         * platform/sql/SQLiteDatabase.cpp:
2835         * platform/sql/SQLiteFileSystem.cpp:
2836         * platform/sql/SQLiteStatement.cpp:
2837         * platform/sql/SQLiteTransaction.cpp:
2838         * storage/DatabaseAuthorizer.cpp:
2839
2840 2011-09-27  Julien Chaffraix  <jchaffraix@webkit.org>
2841
2842         Crash because CSSPrimitiveValue::computeLengthDouble assumes fontMetrics are available
2843         https://bugs.webkit.org/show_bug.cgi?id=66291
2844
2845         Reviewed by Darin Adler.
2846
2847         Test: fast/canvas/crash-set-font.html
2848
2849         This is Yet Another Missing updateFont (similar to bug 57756 and likely others). Here the issue is that
2850         applying one of the font properties could mutate the parent style's font if m_parentStyle == m_style.
2851         We would then query the newly created font when applying CSSPropertyFontSize, which has no font fallback
2852         list as Font::update was never called.
2853
2854         The right fix would be to refactor of how we handle fonts to avoid such manual updates (see bug 62390).
2855         Until this happens, it is better not to crash.
2856
2857         * css/CSSStyleSelector.cpp:
2858         (WebCore::CSSStyleSelector::applyProperty): Added updateFont() here as the fonts could have been
2859         mutated by the previous property change. Also added a comment explaining why it is safe to do it
2860         this way.
2861
2862 2011-09-27  No'am Rosenthal  <noam.rosenthal@nokia.com>
2863
2864         [Texmap] Code cleanup: remove unused boundingRect/visibleRect calculations
2865         https://bugs.webkit.org/show_bug.cgi?id=68897
2866
2867         Reviewed by Andreas Kling.
2868
2869         No new functionality so no new tests.
2870
2871         * platform/graphics/texmap/TextureMapperNode.cpp:
2872         (WebCore::TextureMapperNode::computeAllTransforms):
2873         (WebCore::TextureMapperNode::computeTiles):
2874         (WebCore::TextureMapperNode::syncCompositingState):
2875         * platform/graphics/texmap/TextureMapperNode.h:
2876
2877 2011-09-27  No'am Rosenthal  <noam.rosenthal@nokia.com>
2878
2879         [Texmap][Qt] Refactor texture-upload to allow direct chunk update
2880         https://bugs.webkit.org/show_bug.cgi?id=68808
2881
2882         Add a function to BitmapTexture for direct pixel updates.
2883         Modify BitmapTextureGL::endPaint to use that function. Since the BGRA
2884         to RGBA swizzling is done inside that function, there's no need for the
2885         RGBA32PremultipliedBuffer class to contain such function. Also,
2886         RGBA32PremultipliedBuffer was renamed to BGRA32PremultipliedBuffer, correcting
2887         an old mistake.
2888
2889         Reviewed by Andreas Kling.
2890
2891         No new tests. Existing tests in LayoutTests/compositing test this.
2892
2893         * platform/graphics/opengl/TextureMapperGL.cpp:
2894         (WebCore::BitmapTextureGL::beginPaint):
2895         (WebCore::BitmapTextureGL::endPaint):
2896         (WebCore::swizzleBGRAToRGBA):
2897         (WebCore::BitmapTextureGL::updateContents):
2898         * platform/graphics/opengl/TextureMapperGL.h:
2899         (WebCore::BGRA32PremultimpliedBuffer::~BGRA32PremultimpliedBuffer):
2900         * platform/graphics/qt/TextureMapperQt.cpp:
2901         (WebCore::BitmapTextureQt::updateContents):
2902         (WebCore::BGRA32PremultimpliedBufferQt::data):
2903         (WebCore::BGRA32PremultimpliedBuffer::create):
2904         * platform/graphics/qt/TextureMapperQt.h:
2905         * platform/graphics/texmap/TextureMapper.h:
2906
2907 2011-09-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2908
2909         [Qt] Fix build against Qt5 after refactor of widgets out of QtGUi
2910
2911         QWidget and friends now live in the QtWidgets library. We update
2912         includes in implementation files and private headers to us the
2913         non-module-prefixed path, and leave the lookup for the include
2914         path. For public headers we have to ifdef the includes as the
2915         user might now have the modules we need in his QT config.
2916
2917         Finally, QSGCanvas is no longer a QWidget but a QWindow, so we
2918         have to update our code and use windowHandle() for setting the
2919         parent relationships.
2920
2921         https://bugs.webkit.org/show_bug.cgi?id=68687
2922
2923         Reviewed by Andreas Kling.
2924
2925         * WebCore.pri:
2926         * platform/graphics/qt/GraphicsLayerQt.cpp:
2927
2928 2011-09-27  Pavel Feldman  <pfeldman@google.com>
2929
2930         Web Inspector: split DOM.attributesUpdated into attributeModified and attributeRemoved.
2931         Send attribute name and value within the event.
2932         https://bugs.webkit.org/show_bug.cgi?id=68613
2933
2934         Reviewed by Yury Semikhatsky.
2935
2936         * dom/Element.cpp:
2937         (WebCore::Element::setAttribute):
2938         (WebCore::Element::removeAttribute):
2939         * inspector/Inspector.draft-01.json:
2940         * inspector/Inspector.json:
2941         * inspector/InspectorDOMAgent.cpp:
2942         (WebCore::InspectorDOMAgent::didModifyDOMAttr):
2943         (WebCore::InspectorDOMAgent::didRemoveDOMAttr):
2944         * inspector/InspectorDOMAgent.h:
2945         * inspector/InspectorInstrumentation.cpp:
2946         (WebCore::InspectorInstrumentation::didModifyDOMAttrImpl):
2947         (WebCore::InspectorInstrumentation::didRemoveDOMAttrImpl):
2948         * inspector/InspectorInstrumentation.h:
2949         (WebCore::InspectorInstrumentation::didModifyDOMAttr):
2950         (WebCore::InspectorInstrumentation::didRemoveDOMAttr):
2951         * inspector/front-end/DOMAgent.js:
2952         (WebInspector.DOMNode.prototype._addAttribute):
2953         (WebInspector.DOMNode.prototype._setAttribute):
2954         (WebInspector.DOMNode.prototype._removeAttribute):
2955         (WebInspector.DOMAgent.prototype._attributeModified):
2956         (WebInspector.DOMAgent.prototype._attributeRemoved):
2957         (WebInspector.DOMAgent.prototype._inlineStyleInvalidated):
2958         (WebInspector.DOMAgent.prototype._loadNodeAttributes):
2959         (WebInspector.DOMDispatcher.prototype.attributeModified):
2960         (WebInspector.DOMDispatcher.prototype.attributeRemoved):
2961         * inspector/front-end/ElementsPanel.js:
2962         (WebInspector.ElementsPanel):
2963         (WebInspector.ElementsPanel.prototype._attributesUpdated):
2964         * inspector/front-end/MetricsSidebarPane.js:
2965         (WebInspector.MetricsSidebarPane):
2966         (WebInspector.MetricsSidebarPane.prototype._attributesUpdated):
2967         * inspector/front-end/StylesSidebarPane.js:
2968         (WebInspector.StylesSidebarPane):
2969         (WebInspector.StylesSidebarPane.prototype._attributesModified):
2970         (WebInspector.StylesSidebarPane.prototype._attributesRemoved):
2971         (WebInspector.StylesSidebarPane.prototype._styleInvalidated):
2972         (WebInspector.StylePropertyTreeElement.prototype.event):
2973         (WebInspector.StylePropertyTreeElement.prototype):
2974         * inspector/validate-protocol-compatibility:
2975
2976 2011-09-27  Sheriff Bot  <webkit.review.bot@gmail.com>
2977
2978         Unreviewed, rolling out r96070 and r96075.
2979         http://trac.webkit.org/changeset/96070
2980         http://trac.webkit.org/changeset/96075
2981         https://bugs.webkit.org/show_bug.cgi?id=68893
2982
2983         WK2 tests started crashing after r96070 for SL and Qt
2984         (Requested by torarne on #webkit).
2985
2986         * page/FrameView.cpp:
2987         (WebCore::FrameView::updateScrollCorner):
2988         * platform/ScrollView.cpp:
2989         (WebCore::ScrollView::setHasHorizontalScrollbar):
2990         (WebCore::ScrollView::setHasVerticalScrollbar):
2991         (WebCore::ScrollView::updateScrollbars):
2992         * platform/ScrollableArea.cpp:
2993         (WebCore::ScrollableArea::invalidateScrollCorner):
2994         * platform/ScrollableArea.h:
2995         * rendering/RenderLayerCompositor.cpp:
2996         (WebCore::RenderLayerCompositor::destroyRootLayer):
2997         * rendering/RenderScrollbarPart.cpp:
2998         (WebCore::RenderScrollbarPart::imageChanged):
2999
3000 2011-09-27  Andrey Kosyakov  <caseq@chromium.org>
3001
3002         Web Inspector: JS exception upon clicking on "Word Wrap" checkbox in Settings screen
3003         https://bugs.webkit.org/show_bug.cgi?id=68888
3004
3005         Reviewed by Pavel Feldman.
3006
3007         * inspector/front-end/ElementsPanel.js:
3008         (WebInspector.ElementsPanel.prototype._domWordWrapSettingChanged):
3009
3010 2011-09-21  Pavel Podivilov  <podivilov@chromium.org>
3011
3012         Web Inspector: migrate RawSourceCode clients to SourceMapping class.
3013         https://bugs.webkit.org/show_bug.cgi?id=68524
3014
3015         Clients should use uiSourceCode(), rawLocationToUILocation(), uiLocationToRawLocation() methods of SourceMapping class.
3016         Initially, RawSourceCode may not have associated SourceMapping, so it is natural to extract this methods and associated state to a separate class.
3017
3018         Reviewed by Yury Semikhatsky.
3019
3020         * inspector/front-end/BreakpointManager.js:
3021         (WebInspector.BreakpointManager.prototype._materializeBreakpoint):
3022         (WebInspector.BreakpointManager.prototype._breakpointDebuggerLocationChanged):
3023         * inspector/front-end/DebuggerPresentationModel.js:
3024         (WebInspector.DebuggerPresentationModel.prototype.linkifyLocation.updateAnchor):
3025         (WebInspector.DebuggerPresentationModel.prototype.linkifyLocation):
3026         (WebInspector.DebuggerPresentationModel.prototype._addScript):
3027         (WebInspector.DebuggerPresentationModel.prototype._updateSourceMapping):
3028         (WebInspector.DebuggerPresentationModel.prototype._restoreBreakpoints):
3029         (WebInspector.DebuggerPresentationModel.prototype._restoreConsoleMessages):
3030         (WebInspector.DebuggerPresentationModel.prototype._consoleMessageAdded):
3031         (WebInspector.DebuggerPresentationModel.prototype._createPresentationMessage):
3032         (WebInspector.DebuggerPresentationModel.prototype.continueToLine):
3033         (WebInspector.PresentationCallFrame.prototype.get url):
3034         (WebInspector.PresentationCallFrame.prototype.sourceLine.sourceMappingUpdated):
3035         (WebInspector.PresentationCallFrame.prototype.sourceLine):
3036         (WebInspector.DebuggerPresentationModelResourceBinding.prototype.canSetContent):
3037         (WebInspector.DebuggerPresentationModelResourceBinding.prototype.setContent):
3038         * inspector/front-end/SourceFile.js:
3039         (WebInspector.RawSourceCode.prototype.get sourceMapping):
3040
3041 2011-09-27  Vineet Chaudhary  <vineet.chaudhary@motorola.com>
3042
3043         wrap attribute of textarea element cannot be accessed by JavaScript.
3044         https://bugs.webkit.org/show_bug.cgi?id=68592
3045
3046         Reviewed by Kent Tamura.
3047
3048         Added JS interface for wrap attribute to HTMLTextAreaElement.idl.
3049
3050         Test: fast/forms/textarea-wrap-attribute.html
3051
3052         * html/HTMLTextAreaElement.idl:
3053
3054 2011-09-27  Xan Lopez  <xlopez@igalia.com>
3055
3056         [GTK] Add compatibility methods for DOM bindings
3057         https://bugs.webkit.org/show_bug.cgi?id=68884
3058
3059         Reviewed by Philippe Normand.
3060
3061         Add compatibility methods for our DOM bindings.
3062
3063         * bindings/gobject/WebKitDOMCustom.cpp:
3064         (webkit_dom_blob_slice): alias to the new method name.
3065         (webkit_dom_html_form_element_dispatch_form_change): this was
3066         removed from WebCore, so just print a warning about it.
3067         (webkit_dom_html_form_element_dispatch_form_input): ditto.
3068         * bindings/gobject/WebKitDOMCustom.h:
3069
3070 2011-09-27  Ryosuke Niwa  <rniwa@webkit.org>
3071
3072         Encapsulate m_firstNodeInserted and m_lastLeafInserted in node insertion logic
3073         https://bugs.webkit.org/show_bug.cgi?id=68875
3074
3075         Reviewed by Kent Tamura.
3076
3077         Added InsertedNode class inside ReplaceSelection to encapsulate m_firstNodeInserted and m_lastLeafInserted.
3078         Deployed it in removeRedundantStylesAndKeepStyleSpanInline, doApply, and insertAsListItems of ReplaceSelectionCommand.
3079
3080         * editing/DeleteSelectionCommand.cpp:
3081         (WebCore::DeleteSelectionCommand::removeNode):
3082         (WebCore::DeleteSelectionCommand::handleGeneralDelete):
3083         * editing/ReplaceSelectionCommand.cpp:
3084         (WebCore::ReplaceSelectionCommand::removeRedundantStylesAndKeepStyleSpanInline):
3085         (WebCore::ReplaceSelectionCommand::doApply):
3086         (WebCore::ReplaceSelectionCommand::insertAsListItems):
3087         * editing/ReplaceSelectionCommand.h:
3088         (WebCore::ReplaceSelectionCommand::InsertedNodes::respondToNodeInsertion):
3089         (WebCore::ReplaceSelectionCommand::InsertedNodes::willRemoveNode):
3090         (WebCore::ReplaceSelectionCommand::InsertedNodes::firstNodeInserted):
3091         (WebCore::ReplaceSelectionCommand::InsertedNodes::lastLeafInserted):
3092         (WebCore::ReplaceSelectionCommand::InsertedNodes::pastLastLeaf):
3093         * editing/htmlediting.cpp:
3094         (WebCore::updatePositionForNodeRemoval): Moved from DeleteSelectionCommand.cpp.
3095         * editing/htmlediting.h:
3096
3097 2011-09-27  Arun Patole  <bmf834@motorola.com>
3098
3099         Audio element doesn't emit the 'playing' event every time it starts playing, after it has finished playing.
3100         https://bugs.webkit.org/show_bug.cgi?id=60972
3101
3102         Reviewed by Philippe Normand.
3103
3104         The paused attribute should be set to true and the media element should emit a 'paused' at the end of playback.
3105         WHATWG revision r6562: http://html5.org/tools/web-apps-tracker?from=6561&to=6562
3106
3107         Test: media/media-element-play-after-eos.html
3108
3109         * html/HTMLMediaElement.cpp:
3110         (WebCore::HTMLMediaElement::mediaPlayerTimeChanged): set m_paused to true and schedule 'pause' event when playback ended.
3111
3112 2011-09-27  Rakesh KN  <rakesh.kn@motorola.com>
3113
3114         <input> with autofocus doesn't lose focus when it has a certain onblur listener
3115         https://bugs.webkit.org/show_bug.cgi?id=68513
3116
3117         Reviewed by Kent Tamura.
3118
3119         Test: fast/forms/autofocus-focus-only-once.html
3120
3121         These changes make sure that an element is focused only once when autofocus attribute is used.
3122
3123         * html/HTMLFormControlElement.cpp:
3124         (WebCore::HTMLFormControlElement::HTMLFormControlElement):
3125         (WebCore::shouldAutofocus):
3126         (WebCore::HTMLFormControlElement::attach):
3127         * html/HTMLFormControlElement.h:
3128         (WebCore::HTMLFormControlElement::hasAutofocused):
3129         (WebCore::HTMLFormControlElement::setAutofocused):
3130
3131 2011-09-26  Kentaro Hara  <haraken@chromium.org>
3132
3133         Implement PopStateEvent.state with SerializedScriptValue and ScriptValue
3134         https://bugs.webkit.org/show_bug.cgi?id=68345
3135
3136         Reviewed by Adam Barth.
3137
3138         Previously, the following test cases fail or crash:
3139
3140         - shouldBe("new PopStateEvent('eventType', { state: object1 }).state", "object1") -> FAIL
3141         - new PopStateEvent('eventType', { state: document }).state -> CRASH in DRT
3142
3143         This is because PopStateEvent.state is implemented not as ScriptValue but as SerializedScriptValue.
3144         However, we cannot simply change the type of PopStateEvent.state to ScriptValue,
3145         since PopStateEvent can be constructed in the context that does not know ScriptValue.
3146         For example, Document.cpp calls PopStateEvent::create() with SerializedScriptValue
3147         popped from HistoryItem, but we cannot deserialize the SerializedScriptValue into
3148         the corresponding ScriptValue here because the deserialization requires ExecState.
3149         In other words, although we want to store PopStateEvent.state by ScriptValue internally,
3150         PopStateEvent still needs to provide an API to construct it with SerializedScriptValue.
3151         With these observations, this patch makes the following changes:
3152
3153         - If PopStateEvent is constructed with ScriptValue, it is stored as ScriptValue internally.
3154         When PopStateEvent.state is called, the ScriptValue is returned.
3155         - If PopStateEvent is constructed with SerializedScriptValue, it is stored as
3156         SerializedScriptValue internally (since we cannot deserialize it into ScriptValue
3157         at this point). When PopStateEvent.state is called, the SerializedScriptValue is
3158         deserialized into the corresponding ScriptValue, and the ScriptValue is returned.
3159
3160         Tests: fast/events/constructors/pop-state-event-constructor.html
3161                fast/events/fire-popstate-event.html
3162
3163         * GNUmakefile.list.am: Added JSPopStateEventCustom.cpp.
3164         * UseJSC.cmake: Ditto.
3165         * WebCore.gypi: Ditto.
3166         * WebCore.pro: Ditto.
3167         * WebCore.xcodeproj/project.pbxproj: Ditto.
3168         * bindings/js/JSBindingsAllInOne.cpp: Ditto.
3169         * bindings/js/JSPopStateEventCustom.cpp:
3170         (WebCore::JSPopStateEvent::state): Custom getter for PopStateEvent.state.
3171         * bindings/v8/custom/V8PopStateEventCustom.cpp:
3172         (WebCore::V8PopStateEvent::stateAccessorGetter): Custom getter for PopStateEvent.state.
3173         * dom/PopStateEvent.cpp:
3174         (WebCore::PopStateEventInit::PopStateEventInit): Added initialization code for PopStateEvent.m_state.
3175         (WebCore::PopStateEvent::PopStateEvent): Ditto.
3176         (WebCore::PopStateEvent::create): Ditto.
3177         (WebCore::PopStateEvent::initPopStateEvent): Ditto.
3178         * dom/PopStateEvent.h:
3179         (WebCore::PopStateEvent::serializedState): Getter.
3180         (WebCore::PopStateEvent::state): Getter.
3181         * dom/PopStateEvent.idl: Change the type of 'stateArg' and 'state' to DOMObject. Added [CustomGetter] to 'state'.
3182
3183 2011-09-09  Simon Fraser  <simon.fraser@apple.com>
3184
3185         Pixel result shows that compositing/iframes/repaint-after-losing-scrollbars.html is failing
3186         https://bugs.webkit.org/show_bug.cgi?id=67858
3187
3188         Reviewed by Darin Adler.
3189
3190         When non-overlay scrollbars are hidden on a composited iframe, nothing invalidated
3191         the scrollbar areas or the scroll corner, so the scrollbars appear to remain.
3192
3193         Fix by invalidating the scrollbars and scroll corner when they are removed. Invalidation
3194         on scrollbar creation appears to happen via updating the scrollbar style.
3195
3196         Tested by compositing/iframes/repaint-after-losing-scrollbars.html which no longer shows
3197         stale scrollbars when run manually, even though the green squares are missing from the
3198         pixel result (bug 67878).
3199
3200         * page/FrameView.cpp:
3201         (WebCore::FrameView::updateScrollCorner): Pass the corner rect into invalidateScrollCorner().
3202         * platform/ScrollView.cpp:
3203         (WebCore::ScrollView::setHasHorizontalScrollbar): Invalidate the scrollbar area if hiding it.
3204         (WebCore::ScrollView::setHasVerticalScrollbar): Ditto.
3205         (WebCore::ScrollView::updateScrollbars): In the case where both scrollbars are going away,
3206         compute the scroll corner rect while we still have scrollbars, and then invalidate it
3207         explicitly. (updateScrollCorner() doesn't, because it doesn't have access to the old corner
3208         rect.)
3209         * platform/ScrollableArea.cpp:
3210         (WebCore::ScrollableArea::invalidateScrollCorner): Pass the rect in, because we can't
3211         compute it in the case where the scrollbars are going away.
3212         * platform/ScrollableArea.h: Pass in a rect to invalidateScrollCorner(), which matches
3213         invalidateScrollbar().
3214         * rendering/RenderLayerCompositor.cpp:
3215         (WebCore::RenderLayerCompositor::destroyRootLayer): Pass the corner rect into invalidateScrollCorner().
3216         * rendering/RenderScrollbarPart.cpp: Ditto.
3217         (WebCore::RenderScrollbarPart::imageChanged): Ditto.
3218
3219 2011-09-09  Simon Fraser  <simon.fraser@apple.com>
3220
3221         Translucent scrollbars on composited layers render incorrectly
3222         https://bugs.webkit.org/show_bug.cgi?id=58515
3223
3224         Reviewed by Sam Weinig.
3225
3226         Scrollbars in composited elements were getting drawn twice,
3227         because r41203 moved the call to paintOverflowControls() out of
3228         RenderLayer::paintLayer(), but forgot to change RenderLayerBacking::paintIntoLayer().
3229
3230         Test: compositing/scrollbar-painting.html
3231
3232         * rendering/RenderLayerBacking.cpp:
3233         (WebCore::RenderLayerBacking::paintIntoLayer):
3234
3235 2011-09-26  Nat Duca  <nduca@chromium.org>
3236
3237         [chromium] Make CCThreadProxy draw
3238         https://bugs.webkit.org/show_bug.cgi?id=67417
3239
3240         Update the CCThreadProxy to correctly implement the CCProxy
3241         interface, do all the right committing and updating steps, and
3242         draw a picture on the screen.
3243
3244         Reviewed by James Robinson.
3245
3246         * platform/graphics/IntRect.h:
3247         * platform/graphics/chromium/LayerRendererChromium.cpp:
3248         (WebCore::LayerRendererChromium::~LayerRendererChromium):
3249         * platform/graphics/chromium/cc/CCHeadsUpDisplay.cpp:
3250         (WebCore::CCHeadsUpDisplay::enabled):
3251         * platform/graphics/chromium/cc/CCLayerImpl.cpp:
3252         (WebCore::CCLayerImpl::CCLayerImpl):
3253         (WebCore::CCLayerImpl::~CCLayerImpl):
3254         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
3255         (WebCore::CCLayerTreeHost::CCLayerTreeHost):
3256         (WebCore::CCLayerTreeHost::commitTo):
3257         (WebCore::CCLayerTreeHost::commitComplete):
3258         (WebCore::CCLayerTreeHost::setNeedsRedraw):
3259         * platform/graphics/chromium/cc/CCLayerTreeHost.h:
3260         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
3261         (WebCore::CCLayerTreeHostImpl::CCLayerTreeHostImpl):
3262         (WebCore::CCLayerTreeHostImpl::~CCLayerTreeHostImpl):
3263         * platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
3264         (WebCore::CCSingleThreadProxy::finishAllRendering):
3265         (WebCore::CCSingleThreadProxy::setNeedsCommit):
3266         (WebCore::CCSingleThreadProxy::commitIfNeeded):
3267         * platform/graphics/chromium/cc/CCThreadProxy.cpp:
3268         (WebCore::CCThreadProxy::CCThreadProxy):
3269         (WebCore::CCThreadProxy::~CCThreadProxy):
3270         (WebCore::CCThreadProxy::compositeAndReadback):
3271         (WebCore::CCThreadProxy::drawLayersAndReadbackOnCCThread):
3272         (WebCore::CCThreadProxy::finishAllRendering):
3273         (WebCore::CCThreadProxy::isStarted):
3274         (WebCore::CCThreadProxy::setNeedsCommit):
3275         (WebCore::CCThreadProxy::setNeedsCommitAndRedraw):
3276         (WebCore::CCThreadProxy::setNeedsRedraw):
3277         (WebCore::CCThreadProxy::start):
3278         (WebCore::CCThreadProxy::stop):
3279         (WebCore::CCThreadProxy::finishAllRenderingOnCCThread):
3280         (WebCore::CCThreadProxy::createBeginFrameAndCommitTaskOnCCThread):
3281         (WebCore::CCThreadProxy::beginFrameAndCommit):
3282         (WebCore::CCThreadProxy::commitOnCCThread):
3283         (WebCore::CCThreadProxy::scheduleDrawTaskOnCCThread):
3284         (WebCore::CCThreadProxy::drawLayersAndPresentOnCCThread):
3285         (WebCore::CCThreadProxy::drawLayersOnCCThread):
3286         (WebCore::CCThreadProxy::updateSchedulerStateOnCCThread):
3287         * platform/graphics/chromium/cc/CCThreadProxy.h:
3288
3289 2011-09-26  Adam Klein  <adamk@chromium.org>
3290
3291         [MutationObservers] implement MutationRecord
3292         https://bugs.webkit.org/show_bug.cgi?id=68824
3293
3294         Reviewed by Darin Adler.
3295
3296         Implements MutationRecord as specified in the thread at
3297         http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/1622.html,
3298         including some minor naming changes from the original proposal.
3299
3300         This is a small part of the MutationObserver API, see
3301         https://bugs.webkit.org/show_bug.cgi?id=68729 for the metabug covering
3302         this feature.
3303
3304         No new tests as this isn't yet exposed to the platform.
3305
3306         * CMakeLists.txt:
3307         * DerivedSources.cpp:
3308         * DerivedSources.make:
3309         * GNUmakefile.list.am:
3310         * WebCore.gypi:
3311         * WebCore.pro:
3312         * WebCore.vcproj/WebCore.vcproj:
3313         * WebCore.xcodeproj/project.pbxproj:
3314         * dom/MutationRecord.cpp: Added.
3315         (WebCore::MutationRecord::createChildList):
3316         (WebCore::MutationRecord::createAttributes):
3317         (WebCore::MutationRecord::createCharacterData):
3318         (WebCore::MutationRecord::MutationRecord):
3319         (WebCore::MutationRecord::~MutationRecord):
3320         * dom/MutationRecord.h: Added.
3321         (WebCore::MutationRecord::target):
3322         (WebCore::MutationRecord::addedNodes):
3323         (WebCore::MutationRecord::removedNodes):
3324         (WebCore::MutationRecord::previousSibling):
3325         (WebCore::MutationRecord::nextSibling):
3326         (WebCore::MutationRecord::attributeName):
3327         (WebCore::MutationRecord::attributeNamespace):
3328         (WebCore::MutationRecord::oldValue):
3329         (WebCore::MutationRecord::setOldValue):
3330         * dom/MutationRecord.idl: Added.
3331
3332 2011-09-26  Kentaro Hara  <haraken@chromium.org>
3333
3334         Implement a CloseEvent constructor for V8
3335         https://bugs.webkit.org/show_bug.cgi?id=68793
3336
3337         Reviewed by Adam Barth.
3338
3339         Test: fast/events/constructors/close-event-constructor.html
3340
3341         * bindings/v8/OptionsObject.cpp:
3342         (WebCore::OptionsObject::getKey): Just removed an extra space.
3343         (WebCore::OptionsObject::getKeyValue): Returns an unsigned short value corresponding to the given key.
3344         * bindings/v8/OptionsObject.h:
3345         * bindings/v8/custom/V8EventConstructors.cpp: Added the CloseEvent constructor.
3346         * websockets/CloseEvent.idl: Added a 'V8CustomConstructor' attribute.
3347
3348 2011-09-26  Nate Chapin  <japhet@chromium.org>
3349
3350         Clean up CachedResource::load(). Collapse its
3351         logic into a single callsite, taking just a
3352         CachedResourceLoader and a ResourceLoaderOptions.
3353
3354         1. Its 'incremental' parameter is redundant.
3355         2. With 'incremental' removed, the SecurityCheckPolicy is the only
3356            parameter difference between the CachedResource::load() variants.
3357            Making it a part of ResourceLoaderOptions removes yet another random
3358            enum that is passed around individually.
3359         3. We currently have to call setResourceLoaderOptions() before load() on a
3360            new CachedResource, so we should just take the ResourceLoaderOptions in load().
3361
3362         https://bugs.webkit.org/show_bug.cgi?id=67443
3363
3364         Reviewed by Antti Koivisto.
3365
3366         No new tests, refactor only.
3367
3368         * loader/DocumentThreadableLoader.cpp:
3369         * loader/FrameLoaderTypes.h: Move SecurityCheckPolicy to ResourceLoaderOptions.h.
3370         * loader/MainResourceLoader.cpp:
3371         * loader/NetscapePlugInStreamLoader.cpp:
3372         * loader/ResourceLoadScheduler.cpp:
3373         * loader/ResourceLoadScheduler.h:
3374         * loader/ResourceLoader.cpp: Enforce SecurityCheckPolicy here instead of SubresourceLoader.
3375         * loader/ResourceLoaderOptions.h: Store SecurityCheckPolicy on ResourceLoaderOptions.
3376         * loader/SubresourceLoader.cpp:
3377         * loader/SubresourceLoader.h:
3378         * loader/cache/CachedFont.cpp:
3379         * loader/cache/CachedFont.h:
3380         * loader/cache/CachedImage.cpp:
3381         * loader/cache/CachedImage.h:
3382         * loader/cache/CachedResource.cpp:
3383         * loader/cache/CachedResource.h:
3384         * loader/cache/CachedResourceLoader.cpp:
3385         (WebCore::defaultCachedResourceOptions): Define the default ResourceLoaderOptions for
3386             CachedResource loads here instead of as a default parameter on requestResource().
3387         * loader/cache/CachedResourceLoader.h:
3388         * loader/cache/CachedResourceRequest.cpp:
3389         (WebCore::CachedResourceRequest::didReceiveData): Remove m_incremental, since its purpose
3390             is already enforced in the data() implementation of every CachedResource that doesn't want
3391             incremental loads.
3392         * loader/cache/CachedResourceRequest.h:
3393
3394 2011-09-26  Joshua Bell  <jsbell@chromium.org>
3395
3396         IndexedDB: Second half of IDBFactory.getDatabaseNames implementation
3397         https://bugs.webkit.org/show_bug.cgi?id=68818
3398
3399         Reviewed by Tony Chang.
3400
3401         Test: storage/indexeddb/factory-basics.html
3402
3403         Adds ability to enumerate IndexedDB databases within an origin.
3404
3405         * bindings/v8/custom/V8IDBAnyCustom.cpp:
3406         (WebCore::toV8):
3407         * storage/IDBAny.cpp:
3408         (WebCore::IDBAny::domStringList):
3409         (WebCore::IDBAny::set):
3410         * storage/IDBAny.h:
3411         * storage/IDBBackingStore.h:
3412         * storage/IDBCallbacks.h:
3413         * storage/IDBFactory.cpp:
3414         (WebCore::IDBFactory::getDatabaseNames):
3415         * storage/IDBFactory.h:
3416         * storage/IDBFactory.idl:
3417         * storage/IDBFactoryBackendImpl.cpp:
3418         (WebCore::IDBFactoryBackendImpl::getDatabaseNames):
3419         (WebCore::IDBFactoryBackendImpl::open):
3420         (WebCore::IDBFactoryBackendImpl::openBackingStore):
3421         * storage/IDBFactoryBackendImpl.h:
3422         * storage/IDBFactoryBackendInterface.h:
3423         * storage/IDBLevelDBBackingStore.cpp:
3424         (WebCore::IDBLevelDBBackingStore::getDatabaseNames):
3425         (WebCore::IDBLevelDBBackingStore::getObjectStores):
3426         (WebCore::IDBLevelDBBackingStore::getIndexes):
3427         * storage/IDBLevelDBBackingStore.h:
3428         * storage/IDBLevelDBCoding.cpp:
3429         (WebCore::IDBLevelDBCoding::DatabaseNameKey::encodeMinKeyForOrigin):
3430         (WebCore::IDBLevelDBCoding::DatabaseNameKey::encodeStopKeyForOrigin):
3431         * storage/IDBLevelDBCoding.h:
3432         * storage/IDBRequest.cpp:
3433         (WebCore::IDBRequest::onSuccess):
3434         * storage/IDBRequest.h:
3435         * storage/IDBSQLiteBackingStore.cpp:
3436         (WebCore::IDBSQLiteBackingStore::getDatabaseNames):
3437         * storage/IDBSQLiteBackingStore.h:
3438
3439 2011-09-26  Tim Horton  <timothy_horton@apple.com>
3440
3441         <animateColor> applied to filtered ellipse does not update
3442         https://bugs.webkit.org/show_bug.cgi?id=68457
3443         <rdar://problem/10154777>
3444
3445         Reviewed by Darin Adler.
3446
3447         Invalidate the filter's cache when style changes take place so that style changes are respected.
3448
3449         Test: svg/filters/animate-fill.svg
3450
3451         * rendering/svg/SVGResourcesCache.cpp:
3452         (WebCore::SVGResourcesCache::clientStyleChanged):
3453         (WebCore::SVGResourcesCache::clientUpdatedFromElement):
3454
3455 2011-09-26  Max Perepelitsyn  <pph34r@gmail.com>
3456
3457         Set but unused variables cleanup in v8 bindings (gcc 4.6)
3458        &n