WebKit-https.git
5 years agoWeb Inspector: Reduce synchronous view layouts
mattbaker@apple.com [Thu, 19 Nov 2015 05:44:32 +0000 (05:44 +0000)]
Web Inspector: Reduce synchronous view layouts
https://bugs.webkit.org/show_bug.cgi?id=151058

Reviewed by Timothy Hatcher.

Replace calls to View.updateLayout with needsLayout, when a synchronous layout
isn't absolutely necessary.

* UserInterface/Models/BackForwardEntry.js:
(WebInspector.BackForwardEntry.prototype.prepareToShow):

* UserInterface/Views/ApplicationCacheFrameContentView.js:
(WebInspector.ApplicationCacheFrameContentView.prototype.updateLayout): Deleted.
View subclasses shouldn't override updateLayout.

* UserInterface/Views/ButtonNavigationItem.js:
(WebInspector.ButtonNavigationItem.prototype.set label):

* UserInterface/Views/ContentBrowser.js:
(WebInspector.ContentBrowser.prototype._contentViewSelectionPathComponentDidChange):
(WebInspector.ContentBrowser.prototype._currentContentViewDidChange):
(WebInspector.ContentBrowser.prototype._contentViewNavigationItemsDidChange):

* UserInterface/Views/DatabaseTableContentView.js:
(WebInspector.DatabaseTableContentView.prototype._queryFinished):
Just update DataGrid's layout, not the whole view. The grid is the only subview
so the result is identical, but the intent is cleared this way.
(WebInspector.DatabaseTableContentView.prototype.updateLayout): Deleted.
View subclasses shouldn't override updateLayout.

* UserInterface/Views/LogContentView.js:
(WebInspector.LogContentView.prototype.didAppendConsoleMessageView):
(WebInspector.LogContentView.prototype.promptDidChangeHeight): Deleted.
Unused code.

* UserInterface/Views/Sidebar.js:
(WebInspector.Sidebar.prototype.set width):
(WebInspector.Sidebar.prototype.set collapsed):

* UserInterface/Views/TextResourceContentView.js:
(WebInspector.TextResourceContentView.prototype._contentWillPopulate):
Add TextEditor as a subview.
(WebInspector.TextResourceContentView.prototype.updateLayout): Deleted.
View subclasses shouldn't override updateLayout.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoInternal perf dashboard takes forever to load
rniwa@webkit.org [Thu, 19 Nov 2015 04:48:42 +0000 (04:48 +0000)]
Internal perf dashboard takes forever to load
https://bugs.webkit.org/show_bug.cgi?id=151430

Rubber-stamped by Antti Koivisto.

Fix a few performance problems with the perf dashboard v2 UI.

* public/v2/app.js:
(App.DashboardRow._createPane): Set "inDashboard" to true.
(App.Pane._fetch): Immediately show the cached chart instead of waiting for the refetched data which invokes
a PHP JSON API. Also don't fetch the analysis tasks when the chart is shown in the dashboard since we don't
show annotate charts in the dashboard.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoModern IDB:Make in-memory Index cursors work.
beidson@apple.com [Thu, 19 Nov 2015 04:36:00 +0000 (04:36 +0000)]
Modern IDB:Make in-memory Index cursors work.
https://bugs.webkit.org/show_bug.cgi?id=151278

Reviewed by Alex Christensen.

Source/WebCore:

Tests: storage/indexeddb/modern/index-cursor-1.html
       storage/indexeddb/modern/index-cursor-2.html
       storage/indexeddb/modern/index-cursor-3.html

* CMakeLists.txt:
* WebCore.xcodeproj/project.pbxproj:

* Modules/indexeddb/client/IDBIndexImpl.cpp:
(WebCore::IDBClient::IDBIndex::openCursor):
(WebCore::IDBClient::IDBIndex::openKeyCursor):

* Modules/indexeddb/server/IndexValueEntry.cpp:
(WebCore::IDBServer::IndexValueEntry::removeKey):
(WebCore::IDBServer::IndexValueEntry::Iterator::Iterator):
(WebCore::IDBServer::IndexValueEntry::Iterator::key):
(WebCore::IDBServer::IndexValueEntry::Iterator::isValid):
(WebCore::IDBServer::IndexValueEntry::Iterator::invalidate):
(WebCore::IDBServer::IndexValueEntry::Iterator::operator++):
(WebCore::IDBServer::IndexValueEntry::begin):
(WebCore::IDBServer::IndexValueEntry::reverseBegin):
(WebCore::IDBServer::IndexValueEntry::find):
(WebCore::IDBServer::IndexValueEntry::reverseFind):
* Modules/indexeddb/server/IndexValueEntry.h:
(WebCore::IDBServer::IndexValueEntry::Iterator::Iterator):
(WebCore::IDBServer::IndexValueEntry::unique):

* Modules/indexeddb/server/IndexValueStore.cpp:
(WebCore::IDBServer::IndexValueStore::removeEntriesWithValueKey):
(WebCore::IDBServer::IndexValueStore::lowestKeyWithRecordInRange):
(WebCore::IDBServer::IndexValueStore::lowestIteratorInRange):
(WebCore::IDBServer::IndexValueStore::highestReverseIteratorInRange):
(WebCore::IDBServer::IndexValueStore::find):
(WebCore::IDBServer::IndexValueStore::reverseFind):
(WebCore::IDBServer::IndexValueStore::Iterator::Iterator):
(WebCore::IDBServer::IndexValueStore::Iterator::nextIndexEntry):
(WebCore::IDBServer::IndexValueStore::Iterator::operator++):
(WebCore::IDBServer::IndexValueStore::Iterator::invalidate):
(WebCore::IDBServer::IndexValueStore::Iterator::isValid):
(WebCore::IDBServer::IndexValueStore::Iterator::key):
(WebCore::IDBServer::IndexValueStore::Iterator::primaryKey):
* Modules/indexeddb/server/IndexValueStore.h:
(WebCore::IDBServer::IndexValueStore::Iterator::Iterator):

* Modules/indexeddb/server/MemoryIDBBackingStore.cpp:
(WebCore::IDBServer::MemoryIDBBackingStore::openCursor):

* Modules/indexeddb/server/MemoryIndex.cpp:
(WebCore::IDBServer::MemoryIndex::cursorDidBecomeClean):
(WebCore::IDBServer::MemoryIndex::cursorDidBecomeDirty):
(WebCore::IDBServer::MemoryIndex::objectStoreCleared):
(WebCore::IDBServer::MemoryIndex::notifyCursorsOfValueChange):
(WebCore::IDBServer::MemoryIndex::notifyCursorsOfAllRecordsChanged):
(WebCore::IDBServer::MemoryIndex::putIndexKey):
(WebCore::IDBServer::MemoryIndex::removeRecord):
(WebCore::IDBServer::MemoryIndex::removeEntriesWithValueKey):
(WebCore::IDBServer::MemoryIndex::maybeOpenCursor):
* Modules/indexeddb/server/MemoryIndex.h:
(WebCore::IDBServer::MemoryIndex::valueStore):
(WebCore::IDBServer::MemoryIndex::objectStore):

* Modules/indexeddb/server/MemoryIndexCursor.cpp: Added.
(WebCore::IDBServer::MemoryIndexCursor::MemoryIndexCursor):
(WebCore::IDBServer::MemoryIndexCursor::~MemoryIndexCursor):
(WebCore::IDBServer::MemoryIndexCursor::currentData):
(WebCore::IDBServer::MemoryIndexCursor::iterate):
(WebCore::IDBServer::MemoryIndexCursor::indexRecordsAllChanged):
(WebCore::IDBServer::MemoryIndexCursor::indexValueChanged):
* Modules/indexeddb/server/MemoryIndexCursor.h:

* Modules/indexeddb/server/MemoryObjectStore.cpp:
(WebCore::IDBServer::MemoryObjectStore::indexForIdentifier):
(WebCore::IDBServer::MemoryObjectStore::valueForKey):
* Modules/indexeddb/server/MemoryObjectStore.h:

* Modules/indexeddb/shared/IDBCursorInfo.cpp:
(WebCore::IDBCursorInfo::objectStoreCursor):
(WebCore::IDBCursorInfo::indexCursor):
(WebCore::IDBCursorInfo::IDBCursorInfo):
(WebCore::IDBCursorInfo::isDirectionNoDuplicate):
(WebCore::IDBCursorInfo::isolatedCopy):
* Modules/indexeddb/shared/IDBCursorInfo.h:
(WebCore::IDBCursorInfo::objectStoreIdentifier):

LayoutTests:

* storage/indexeddb/modern/cursor-1-expected.txt:
* storage/indexeddb/modern/index-cursor-1-expected.txt: Added.
* storage/indexeddb/modern/index-cursor-1.html: Added.
* storage/indexeddb/modern/index-cursor-2-expected.txt: Added.
* storage/indexeddb/modern/index-cursor-2.html: Added.
* storage/indexeddb/modern/index-cursor-3-expected.txt: Added.
* storage/indexeddb/modern/index-cursor-3.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Add some missing load/store to FTLB3Output
benjamin@webkit.org [Thu, 19 Nov 2015 04:15:04 +0000 (04:15 +0000)]
[JSC] Add some missing load/store to FTLB3Output
https://bugs.webkit.org/show_bug.cgi?id=151427

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-18
Reviewed by Filip Pizlo.

* ftl/FTLB3Output.cpp:
(JSC::FTL::Output::load8SignExt32):
(JSC::FTL::Output::load8ZeroExt32):
(JSC::FTL::Output::load16SignExt32):
(JSC::FTL::Output::load16ZeroExt32):
(JSC::FTL::Output::loadFloatToDouble):
* ftl/FTLB3Output.h:
(JSC::FTL::Output::load32):
(JSC::FTL::Output::loadPtr):
(JSC::FTL::Output::loadDouble):
(JSC::FTL::Output::load8SignExt32): Deleted.
(JSC::FTL::Output::load8ZeroExt32): Deleted.
(JSC::FTL::Output::load16SignExt32): Deleted.
(JSC::FTL::Output::load16ZeroExt32): Deleted.
(JSC::FTL::Output::loadFloatToDouble): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAssertion failure in RenderTreePosition::computeNextSibling
antti@apple.com [Thu, 19 Nov 2015 04:14:24 +0000 (04:14 +0000)]
Assertion failure in RenderTreePosition::computeNextSibling
https://bugs.webkit.org/show_bug.cgi?id=151337
rdar://problem/23250075

Reviewed by Zalan Bujtas.

Source/WebCore:

Test: fast/html/details-mathml-crash.html

* html/ads: Added.
* style/StyleResolveTree.cpp:
(WebCore::Style::resolveChildAtShadowBoundary):

    Factor common code for resolving child here from resolveShadowTree.

(WebCore::Style::resolveShadowTree):

    We don't need StyleResolverParentPusher because shadow tree uses different style resolver anyway.

(WebCore::Style::resolveSlotAssignees):

    This needs to call renderTreePosition.invalidateNextSibling() if there is a renderer already.
    Achieve this by calling the new common function resolveChildAtShadowBoundary.

LayoutTests:

Test case by Pranjal Jumde.

* fast/html/details-mathml-crash-expected.txt: Added.
* fast/html/details-mathml-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWebGL slow video to texture
jer.noble@apple.com [Thu, 19 Nov 2015 03:20:11 +0000 (03:20 +0000)]
WebGL slow video to texture
https://bugs.webkit.org/show_bug.cgi?id=129626

Reviewed by Dean Jackson.

Support a direct GPU-to-GPU copy of video textures. Add a new AVPlayerItemVideoOutput which,
when lazily-created, will emit CVPixelBuffers which are guaranteed to be compatible with
OpenGL framebuffers. Then, use a CVOpenGLTextureCache object to convert those CVPixelBuffers
to OpenGL textures. Once the video frame is in an OpenGL texture, use an OpenGL framebuffer
to copy the underlying video texture memory to the destination texture.

The copy step uses glTexImage2D, which requires format and type parameters, so change the
signature of copyVideoTextureToPlatformTexture() to pass those parameters in.

* html/HTMLVideoElement.cpp:
(WebCore::HTMLVideoElement::copyVideoTextureToPlatformTexture): Changed signature.
* html/HTMLVideoElement.h:
* html/canvas/WebGLRenderingContextBase.cpp:
(WebCore::WebGLRenderingContextBase::texImage2D): Changed signature.
* platform/graphics/MediaPlayer.cpp:
(WebCore::MediaPlayer::copyVideoTextureToPlatformTexture): Changed signature.
* platform/graphics/MediaPlayer.h:
* platform/graphics/MediaPlayerPrivate.h:
(WebCore::MediaPlayerPrivateInterface::copyVideoTextureToPlatformTexture): Changed signature.
* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::createOpenGLVideoOutput): Create an OpenGL-compatible
    AVPlayerItemVideoOutput.
(WebCore::MediaPlayerPrivateAVFoundationObjC::destroyOpenGLVideoOutput): Destroy same.
(WebCore::MediaPlayerPrivateAVFoundationObjC::updateLastOpenGLImage): Cache the current frame, if available.
(WebCore::MediaPlayerPrivateAVFoundationObjC::copyVideoTextureToPlatformTexture): Convert the
    current frame to a texture, and use that texture to render into the destination texture.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix typos in r192605
benjamin@webkit.org [Thu, 19 Nov 2015 03:09:51 +0000 (03:09 +0000)]
Fix typos in r192605

* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::DFG::LowerDFGToLLVM::compileUpsilon):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFTL should be able to compile a small function with B3
benjamin@webkit.org [Thu, 19 Nov 2015 02:48:48 +0000 (02:48 +0000)]
FTL should be able to compile a small function with B3
https://bugs.webkit.org/show_bug.cgi?id=151423

Patch by Filip Pizlo  <fpizlo@apple.com> and Benjamin Poulain <bpoulain@apple.com> on 2015-11-18
Reviewed by Filip Pizlo.

* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::lower):
* dfg/DFGPlan.cpp:
(JSC::DFG::Plan::compileInThreadImpl):
* ftl/FTLAbbreviatedTypes.h:
* ftl/FTLB3Output.cpp:
(JSC::FTL::Output::Output):
(JSC::FTL::Output::~Output):
(JSC::FTL::Output::initialize):
(JSC::FTL::Output::appendTo):
(JSC::FTL::Output::lockedStackSlot):
(JSC::FTL::Output::load):
(JSC::FTL::Output::store):
(JSC::FTL::Output::branch):
* ftl/FTLB3Output.h:
(JSC::FTL::Output::newBlock):
(JSC::FTL::Output::setOrigin):
(JSC::FTL::Output::origin):
(JSC::FTL::Output::framePointer):
(JSC::FTL::Output::constBool):
(JSC::FTL::Output::constInt32):
(JSC::FTL::Output::constIntPtr):
(JSC::FTL::Output::constInt64):
(JSC::FTL::Output::constDouble):
(JSC::FTL::Output::upsilon):
(JSC::FTL::Output::phi):
(JSC::FTL::Output::add):
(JSC::FTL::Output::sub):
(JSC::FTL::Output::mul):
(JSC::FTL::Output::neg):
(JSC::FTL::Output::doubleAdd):
(JSC::FTL::Output::doubleSub):
(JSC::FTL::Output::doubleMul):
(JSC::FTL::Output::doubleDiv):
(JSC::FTL::Output::doubleNeg):
(JSC::FTL::Output::bitAnd):
(JSC::FTL::Output::bitOr):
(JSC::FTL::Output::bitXor):
(JSC::FTL::Output::shl):
(JSC::FTL::Output::aShr):
(JSC::FTL::Output::lShr):
(JSC::FTL::Output::load64):
(JSC::FTL::Output::store32):
(JSC::FTL::Output::store64):
(JSC::FTL::Output::storePtr):
(JSC::FTL::Output::storeDouble):
(JSC::FTL::Output::addPtr):
(JSC::FTL::Output::address):
(JSC::FTL::Output::below):
(JSC::FTL::Output::isZero32):
(JSC::FTL::Output::notZero32):
(JSC::FTL::Output::isZero64):
(JSC::FTL::Output::notZero64):
(JSC::FTL::Output::isNull):
(JSC::FTL::Output::notNull):
(JSC::FTL::Output::testIsZero32):
(JSC::FTL::Output::testNonZero32):
(JSC::FTL::Output::testIsZero64):
(JSC::FTL::Output::testNonZero64):
(JSC::FTL::Output::testIsZeroPtr):
(JSC::FTL::Output::testNonZeroPtr):
(JSC::FTL::Output::call):
(JSC::FTL::Output::operation):
(JSC::FTL::Output::jump):
(JSC::FTL::Output::branch):
(JSC::FTL::Output::ret):
(JSC::FTL::Output::unreachable):
(JSC::FTL::Output::check):
(JSC::FTL::Output::anchor):
(JSC::FTL::Output::addIncomingToPhi):
(JSC::FTL::Output::~Output): Deleted.
(JSC::FTL::Output::appendTo): Deleted.
(JSC::FTL::Output::param): Deleted.
(JSC::FTL::Output::load): Deleted.
(JSC::FTL::Output::store): Deleted.
(JSC::FTL::Output::store16): Deleted.
* ftl/FTLCommonValues.cpp:
(JSC::FTL::CommonValues::CommonValues):
(JSC::FTL::CommonValues::initializeConstants):
* ftl/FTLCommonValues.h:
* ftl/FTLJITFinalizer.cpp:
(JSC::FTL::JITFinalizer::finalizeFunction):
* ftl/FTLLink.cpp:
(JSC::FTL::link):
* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::DFG::LowerDFGToLLVM::LowerDFGToLLVM):
(JSC::FTL::DFG::LowerDFGToLLVM::lower):
(JSC::FTL::DFG::LowerDFGToLLVM::createPhiVariables):
(JSC::FTL::DFG::LowerDFGToLLVM::compileBlock):
(JSC::FTL::DFG::LowerDFGToLLVM::safelyInvalidateAfterTermination):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNode):
(JSC::FTL::DFG::LowerDFGToLLVM::compileUpsilon):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePhi):
(JSC::FTL::DFG::LowerDFGToLLVM::compileToThis):
(JSC::FTL::DFG::LowerDFGToLLVM::compileValueAdd):
(JSC::FTL::DFG::LowerDFGToLLVM::compileStrCat):
(JSC::FTL::DFG::LowerDFGToLLVM::compileArithAddOrSub):
(JSC::FTL::DFG::LowerDFGToLLVM::compileArrayifyToStructure):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetById):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutById):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetIndexedPropertyStorage):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetByVal):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutByVal):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutAccessorById):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutGetterSetterById):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutAccessorByVal):
(JSC::FTL::DFG::LowerDFGToLLVM::compileArrayPush):
(JSC::FTL::DFG::LowerDFGToLLVM::compileArrayPop):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCreateActivation):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNewFunction):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCreateScopedArguments):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCreateClonedArguments):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNewArray):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNewArrayBuffer):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNewArrayWithSize):
(JSC::FTL::DFG::LowerDFGToLLVM::compileToStringOrCallStringConstructor):
(JSC::FTL::DFG::LowerDFGToLLVM::compileToPrimitive):
(JSC::FTL::DFG::LowerDFGToLLVM::compileStringCharAt):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCallOrConstruct):
(JSC::FTL::DFG::LowerDFGToLLVM::compileTailCall):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCallOrConstructVarargs):
(JSC::FTL::DFG::LowerDFGToLLVM::compileLoadVarargs):
(JSC::FTL::DFG::LowerDFGToLLVM::compileSwitch):
(JSC::FTL::DFG::LowerDFGToLLVM::compileInvalidationPoint):
(JSC::FTL::DFG::LowerDFGToLLVM::compileIn):
(JSC::FTL::DFG::LowerDFGToLLVM::compileHasIndexedProperty):
(JSC::FTL::DFG::LowerDFGToLLVM::compileHasGenericProperty):
(JSC::FTL::DFG::LowerDFGToLLVM::compileHasStructureProperty):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetDirectPname):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetPropertyEnumerator):
(JSC::FTL::DFG::LowerDFGToLLVM::compileToIndexString):
(JSC::FTL::DFG::LowerDFGToLLVM::didOverflowStack):
(JSC::FTL::DFG::LowerDFGToLLVM::allocatePropertyStorage):
(JSC::FTL::DFG::LowerDFGToLLVM::reallocatePropertyStorage):
(JSC::FTL::DFG::LowerDFGToLLVM::getById):
(JSC::FTL::DFG::LowerDFGToLLVM::nonSpeculativeCompare):
(JSC::FTL::DFG::LowerDFGToLLVM::contiguousPutByValOutOfBounds):
(JSC::FTL::DFG::LowerDFGToLLVM::switchStringSlow):
(JSC::FTL::DFG::LowerDFGToLLVM::doubleToInt32):
(JSC::FTL::DFG::LowerDFGToLLVM::sensibleDoubleToInt32):
(JSC::FTL::DFG::LowerDFGToLLVM::lazySlowPath):
(JSC::FTL::DFG::LowerDFGToLLVM::jsValueToStrictInt52):
(JSC::FTL::DFG::LowerDFGToLLVM::doubleToStrictInt52):
(JSC::FTL::DFG::LowerDFGToLLVM::vmCall):
(JSC::FTL::DFG::LowerDFGToLLVM::callCheck):
(JSC::FTL::DFG::LowerDFGToLLVM::lowBlock):
(JSC::FTL::DFG::LowerDFGToLLVM::appendOSRExit):
(JSC::FTL::DFG::LowerDFGToLLVM::callStackmap):
(JSC::FTL::DFG::LowerDFGToLLVM::probe):
(JSC::FTL::DFG::LowerDFGToLLVM::crash):
(JSC::FTL::lowerDFGToLLVM):
(JSC::FTL::DFG::LowerDFGToLLVM::vmCallNoExceptions): Deleted.
* ftl/FTLOutput.cpp:
(JSC::FTL::Output::sensibleDoubleToInt):
* ftl/FTLOutput.h:
(JSC::FTL::Output::ceil64):
(JSC::FTL::Output::ctlz32):
(JSC::FTL::Output::addWithOverflow32):
(JSC::FTL::Output::subWithOverflow32):
(JSC::FTL::Output::mulWithOverflow32):
(JSC::FTL::Output::addWithOverflow64):
(JSC::FTL::Output::subWithOverflow64):
(JSC::FTL::Output::mulWithOverflow64):
(JSC::FTL::Output::doubleAbs):
(JSC::FTL::Output::doubleSin):
(JSC::FTL::Output::doubleCos):
(JSC::FTL::Output::doublePow):
(JSC::FTL::Output::doublePowi):
(JSC::FTL::Output::doubleSqrt):
(JSC::FTL::Output::doubleLog):
(JSC::FTL::Output::call):
(JSC::FTL::Output::trap):
* ftl/FTLState.cpp:
(JSC::FTL::State::State): Deleted.
* ftl/FTLState.h:
* ftl/FTLWeight.h:
(JSC::FTL::Weight::frequencyClass):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WK1] Crash loading Blink layout test fast/dom/Window/property-access-on-cached-windo...
jiewen_tan@apple.com [Thu, 19 Nov 2015 01:42:12 +0000 (01:42 +0000)]
[WK1] Crash loading Blink layout test fast/dom/Window/property-access-on-cached-window-after-frame-removed.html
https://bugs.webkit.org/show_bug.cgi?id=150198
<rdar://problem/23136026>

Reviewed by Brent Fulgham.

Source/WebCore:

Test: fast/dom/Window/property-access-on-cached-window-after-frame-removed.html

Properties of a contentWindow could be accessed even if the frame who owns the window is
detached. Therefore, check whether the document loader is still alive before using it.

* page/PerformanceTiming.cpp:
(WebCore::PerformanceTiming::monotonicTimeToIntegerMilliseconds):

Tools:

* WebKitTestRunner/InjectedBundle/mac/TestRunnerMac.mm:
(WTR::TestRunner::inspectorTestStubURL):
Since WebInspectorUI.framework is not available for iOS, the framework
and corresponding functions are disabled in iOS.

LayoutTests:

* fast/dom/Window/666869-expected.txt: Added.
* fast/dom/Window/666869.html: Added.
Test case is from Mozilla.
* fast/dom/Window/property-access-on-cached-window-after-frame-removed-expected.txt: Added.
* fast/dom/Window/property-access-on-cached-window-after-frame-removed.html: Added.
* fast/dom/Window/resources/window-property-collector.js: Added.
(collectProperties):
(emitExpectedResult):
(collectPropertiesHelper):
Test case is from Blink r168256:
https://codereview.chromium.org/131113003
* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoThere is a bug when default parameter values are mixed with destructuring parameter...
sbarati@apple.com [Thu, 19 Nov 2015 01:26:36 +0000 (01:26 +0000)]
There is a bug when default parameter values are mixed with destructuring parameter values
https://bugs.webkit.org/show_bug.cgi?id=151369

Reviewed by Geoffrey Garen and Mark Lam.

Relanding this after a rollout.

This patch changes our parser to no longer declare destructuring
parameters as "var"s. This is a weird bug that just happened
to work in a world without default parameter values. In a world with
default parameter values this is just completely wrong. It would
incorrectly transform this program:
```function foo(a = function() { b = 40; }, {b}) { a(); return b; }; foo(undefined, {b: 42}); // Should return 40```
into
```function foo(a = function() { b = 40; }, {b}) { var b; a(); return b; }; foo(undefined, {b:42}); // Returns 42, not 40.```
Which is wrong because we end up with two distinct bindings of "b" when
there should only be one.

* parser/Parser.cpp:
(JSC::Parser<LexerType>::parseVariableDeclarationList):
(JSC::Parser<LexerType>::createBindingPattern):
(JSC::Parser<LexerType>::parseDestructuringPattern):
* parser/Parser.h:
(JSC::Scope::declareParameter):
(JSC::Scope::getUsedVariables):
(JSC::Parser::strictMode):
(JSC::Parser::isValidStrictMode):
(JSC::Parser::declareParameter):
(JSC::Parser::breakIsValid):
(JSC::Scope::declareBoundParameter): Deleted.
(JSC::Parser::declareBoundParameter): Deleted.
* tests/stress/es6-default-parameters.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMediaStream: Implement MediaDevices.getSupportedConstraints
eric.carlson@apple.com [Thu, 19 Nov 2015 01:24:02 +0000 (01:24 +0000)]
MediaStream: Implement MediaDevices.getSupportedConstraints
https://bugs.webkit.org/show_bug.cgi?id=151394

Reviewed by Brent Fulgham.

Source/WebCore:

Test: fast/mediastream/MediaDevices-getSupportedConstraints.html

* CMakeLists.txt: Add MediaTrackSupportedConstraints and JSMediaTrackSupportedConstraintsCustom.

* DerivedSources.make: Ditto.

* Modules/mediastream/MediaDevices.cpp:
(WebCore::MediaDevices::getSupportedConstraints): New.
* Modules/mediastream/MediaDevices.h:
* Modules/mediastream/MediaDevices.idl:

* Modules/mediastream/MediaTrackSupportedConstraints.h: Added. Wrapper around a
  RealtimeMediaSourceSupportedConstraints.
(WebCore::MediaTrackSupportedConstraints::create):
(WebCore::MediaTrackSupportedConstraints::supportsWidth):
(WebCore::MediaTrackSupportedConstraints::supportsHeight):
(WebCore::MediaTrackSupportedConstraints::supportsAspectRatio):
(WebCore::MediaTrackSupportedConstraints::supportsFrameRate):
(WebCore::MediaTrackSupportedConstraints::supportsFacingMode):
(WebCore::MediaTrackSupportedConstraints::supportsVolume):
(WebCore::MediaTrackSupportedConstraints::supportsSampleRate):
(WebCore::MediaTrackSupportedConstraints::supportsSampleSize):
(WebCore::MediaTrackSupportedConstraints::supportsEchoCancellation):
(WebCore::MediaTrackSupportedConstraints::supportsDeviceId):
(WebCore::MediaTrackSupportedConstraints::supportsGroupId):
(WebCore::MediaTrackSupportedConstraints::MediaTrackSupportedConstraints):
* Modules/mediastream/MediaTrackSupportedConstraints.idl: Added.

* WebCore.xcodeproj/project.pbxproj: Add JSMediaTrackSupportedConstraintsCustom.cpp,
  RealtimeMediaSourceSupportedConstraints.h, and MediaTrackSupportedConstraints.*.

A MediaTrackSupportedConstraints only contains the properties supported by the currently
available capture devices, so implement getOwnPropertySlot and getOwnPropertyNames so
we don't have to declare any attributes in the idl file.
* bindings/js/JSMediaTrackSupportedConstraintsCustom.cpp: Added.
(WebCore::JSMediaTrackSupportedConstraints::getOwnPropertySlotDelegate):
(WebCore::JSMediaTrackSupportedConstraints::getOwnPropertyNames):

* platform/mediastream/RealtimeMediaSourceCenter.h:
* platform/mediastream/RealtimeMediaSourceSupportedConstraints.h: Added.
(WebCore::RealtimeMediaSourceSupportedConstraints::RealtimeMediaSourceSupportedConstraints):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsWidth):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsWidth):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsHeight):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsHeight):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsAspectRatio):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsAspectRatio):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsFrameRate):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsFrameRate):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsFacingMode):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsFacingMode):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsVolume):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsVolume):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsSampleRate):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsSampleRate):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsSampleSize):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsSampleSize):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsEchoCancellation):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsEchoCancellation):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsDeviceId):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsDeviceId):
(WebCore::RealtimeMediaSourceSupportedConstraints::supportsGroupId):
(WebCore::RealtimeMediaSourceSupportedConstraints::setSupportsGroupId):

* platform/mediastream/mac/RealtimeMediaSourceCenterMac.cpp:
(WebCore::RealtimeMediaSourceCenterMac::RealtimeMediaSourceCenterMac): Initialize supported constraints.
* platform/mediastream/mac/RealtimeMediaSourceCenterMac.h:

* platform/mock/MockRealtimeMediaSource.cpp: Delete some dead code.

* platform/mock/MockRealtimeMediaSourceCenter.cpp:
(WebCore::MockRealtimeMediaSourceCenter::MockRealtimeMediaSourceCenter): Initialize supported constraints.
* platform/mock/MockRealtimeMediaSourceCenter.h:

LayoutTests:

* fast/mediastream/MediaDevices-getSupportedConstraints-expected.txt: Added.
* fast/mediastream/MediaDevices-getSupportedConstraints.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoProgress towards implementing Downloads with NETWORK_SESSION
commit-queue@webkit.org [Thu, 19 Nov 2015 01:17:14 +0000 (01:17 +0000)]
Progress towards implementing Downloads with NETWORK_SESSION
https://bugs.webkit.org/show_bug.cgi?id=151414

Reviewed begrudgingly by Brady Eidson.

Source/WebCore:

There is no change in behavior except that SessionIDs are sent across IPC and passed as parameters,
and they are not used yet.

Patch by Alex Christensen <achristensen@webkit.org> on 2015-11-18

* loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::continueAfterContentPolicy):
Removed an unused default. Compiler warnings will let us know if we add an unhandled case to this switch.
* loader/EmptyClients.h:
* loader/FrameLoaderClient.h:
Pass SessionIDs around.

Source/WebKit/mac:

Patch by Alex Christensen <achristensen@webkit.org> on 2015-11-18

* WebCoreSupport/WebFrameLoaderClient.h:
* WebCoreSupport/WebFrameLoaderClient.mm:
(WebFrameLoaderClient::detachedFromParent3):
(WebFrameLoaderClient::convertMainResourceLoadToDownload):

Source/WebKit2:

Patch by Alex Christensen <achristensen@webkit.org> on 2015-11-18

* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::storageSession):
(WebKit::NetworkConnectionToWebProcess::startDownload):
(WebKit::NetworkConnectionToWebProcess::convertMainResourceLoadToDownload):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkConnectionToWebProcess.messages.in:
* NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::NetworkLoad):
(WebKit::NetworkLoad::didReceiveChallenge):
(WebKit::NetworkLoad::didReceiveResponse):
(WebKit::NetworkLoad::didReceiveData):
* NetworkProcess/NetworkLoad.h:
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::deleteWebsiteDataForOrigins):
(WebKit::NetworkProcess::downloadRequest):
(WebKit::NetworkProcess::resumeDownload):
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::cleanup):
(WebKit::NetworkResourceLoader::didConvertHandleToDownload):
(WebKit::NetworkResourceLoader::abort):
* NetworkProcess/NetworkResourceLoader.h:
* NetworkProcess/NetworkSession.h:
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(toNSURLSessionResponseDisposition):
(-[NetworkSessionDelegate URLSession:dataTask:didReceiveResponse:completionHandler:]):
(-[NetworkSessionDelegate URLSession:dataTask:didReceiveData:]):
(-[NetworkSessionDelegate URLSession:downloadTask:didFinishDownloadingToURL:]):
(-[NetworkSessionDelegate URLSession:downloadTask:didWriteData:totalBytesWritten:totalBytesExpectedToWrite:]):
(-[NetworkSessionDelegate URLSession:downloadTask:didResumeAtOffset:expectedTotalBytes:]):
(-[NetworkSessionDelegate URLSession:dataTask:didBecomeDownloadTask:]):
* Shared/Authentication/AuthenticationManager.cpp:
(WebKit::AuthenticationManager::useCredentialForSingleChallenge):
(WebKit::AuthenticationManager::continueWithoutCredentialForSingleChallenge):
(WebKit::AuthenticationManager::cancelSingleChallenge):
(WebKit::AuthenticationManager::performDefaultHandlingForSingleChallenge):
(WebKit::AuthenticationManager::rejectProtectionSpaceAndContinueForSingleChallenge):
* Shared/Downloads/Download.h:
* Shared/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::DownloadManager):
(WebKit::DownloadManager::startDownload):
(WebKit::DownloadManager::convertHandleToDownload):
(WebKit::DownloadManager::resumeDownload):
(WebKit::DownloadManager::cancelDownload):
(WebKit::DownloadManager::downloadFinished):
(WebKit::DownloadManager::didCreateDownload):
* Shared/Downloads/DownloadManager.h:
(WebKit::DownloadManager::isDownloading):
(WebKit::DownloadManager::activeDownloadCount):
* Shared/Downloads/mac/DownloadMac.mm:
(WebKit::Download::start):
(WebKit::Download::startWithHandle):
(WebKit::Download::resume):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::download):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::canCachePage):
(WebKit::WebFrameLoaderClient::convertMainResourceLoadToDownload):
(WebKit::WebFrameLoaderClient::createFrame):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
* WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::startDownload):
(WebKit::WebFrame::convertMainResourceLoadToDownload):
(WebKit::WebFrame::source):
* WebProcess/WebPage/WebFrame.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::downloadRequest):
(WebKit::WebProcess::resumeDownload):
(WebKit::WebProcess::releasePageCache):
(WebKit::WebProcess::fetchWebsiteData):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSnippefy op_mul for the baseline JIT.
mark.lam@apple.com [Thu, 19 Nov 2015 00:54:37 +0000 (00:54 +0000)]
Snippefy op_mul for the baseline JIT.
https://bugs.webkit.org/show_bug.cgi?id=151393

Reviewed by Geoffrey Garen.

Benchmarks shows that perf is neutral on x86 and x86_64 with the DFG enabled.

With the DFG disabled (relying on the baseline JIT for perf), LongSpider
3d-morph shows a 7.6% regression.  However, there are other benchmarks that shows
a progression e.g. on Kraken, audio-beat-detection and audio-fft.

Upon inspection of the generated code for 3d-morph, the only differences is the
added use of a scratch register for the result as well as a jump around the
code that handles double types.  It does not look like we're generating bad code.
I'll consider the perf acceptable in aggregate.

* CMakeLists.txt:
* JavaScriptCore.vcxproj/JavaScriptCore.vcxproj:
* JavaScriptCore.vcxproj/JavaScriptCore.vcxproj.filters:
* JavaScriptCore.xcodeproj/project.pbxproj:

* bytecode/PolymorphicAccess.cpp:
(JSC::AccessCase::generate):
* jit/AssemblyHelpers.h:
(JSC::AssemblyHelpers::boxInt32):
* jit/IntrinsicEmitter.cpp:
(JSC::AccessCase::emitIntrinsicGetter):
- Changed AssemblyHelpers::boxInt32() to take a TagRegistersMode.
  The pre-existing boxInt32() always assume that the tag registers are not
  available.  Since we should assume we have tag registers by default, I also
  changed all the other clients to explicitly specify a more of
  DoNotHaveTagRegisters.  That is except for the snippet generators that do have
  the tag registers.

* jit/JIT.h:
* jit/JITArithmetic.cpp:
(JSC::JIT::compileBinaryArithOpSlowCase):
(JSC::JIT::emit_op_div):
(JSC::JIT::emitSlow_op_add):
(JSC::JIT::emit_op_mul):
(JSC::JIT::emitSlow_op_mul):
(JSC::JIT::emit_op_sub):
(JSC::JIT::compileBinaryArithOp): Deleted.
* jit/JITArithmetic32_64.cpp:
(JSC::JIT::emitBinaryDoubleOp):
(JSC::JIT::emit_op_div):
(JSC::JIT::emit_op_mul): Deleted.
(JSC::JIT::emitSlow_op_mul): Deleted.
* jit/JITMulGenerator.cpp: Added.
(JSC::JITMulGenerator::generateFastPath):
* jit/JITMulGenerator.h: Added.
(JSC::JITMulGenerator::JITMulGenerator):
(JSC::JITMulGenerator::didEmitFastPath):
(JSC::JITMulGenerator::endJumpList):
(JSC::JITMulGenerator::slowPathJumpList):

* tests/stress/op_mul.js: Added.
(o1.valueOf):
(generateScenarios):
(printScenarios):
(testCases.func):
(func):
(initializeTestCases):
(stringifyIfNeeded):
(isIdentical):
(runTest):
- Tests that JIT op_mul results are equivalent to the expected values as
  defined by the LLINT.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemove some unnecessary jumps in snippet code.
mark.lam@apple.com [Thu, 19 Nov 2015 00:30:37 +0000 (00:30 +0000)]
Remove some unnecessary jumps in snippet code.
https://bugs.webkit.org/show_bug.cgi?id=151415

Reviewed by Geoffrey Garen.

Previously, the snippet generators always emit a jump at the end of the fast
path.  In the baseline JIT and FTL, this results in a jump to the very next
instruction.  I've change it to assume that the fast path will just fall thru,
and let the client decide instead if it wants/needs a jump or not after the fast
path.

I also changed the generators to provide a didEmitFastPath() query explicitly
declare if they actually generated the fast path, instead of having the client
infer it from an empty endJumpList.

Benchmarks show that perf is neutral with this change (tested on x86_64).

* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileValueAdd):
(JSC::DFG::SpeculativeJIT::compileArithSub):
* ftl/FTLCompile.cpp:
(JSC::FTL::mmAllocateDataSection):
* jit/JITAddGenerator.cpp:
(JSC::JITAddGenerator::generateFastPath):
* jit/JITAddGenerator.h:
(JSC::JITAddGenerator::didEmitFastPath):
(JSC::JITAddGenerator::endJumpList):
(JSC::JITAddGenerator::slowPathJumpList):
* jit/JITArithmetic.cpp:
(JSC::JIT::emit_op_add):
(JSC::JIT::emit_op_sub):
* jit/JITSubGenerator.cpp:
(JSC::JITSubGenerator::generateFastPath):
* jit/JITSubGenerator.h:
(JSC::JITSubGenerator::didEmitFastPath):
(JSC::JITSubGenerator::endJumpList):
(JSC::JITSubGenerator::slowPathJumpList):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[Mac] Swipe tests depend on user preferences
ap@apple.com [Thu, 19 Nov 2015 00:20:41 +0000 (00:20 +0000)]
[Mac] Swipe tests depend on user preferences
https://bugs.webkit.org/show_bug.cgi?id=151403

Reviewed by Tim Horton.

* WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: (WTR::InjectedBundle::platformInitialize):
* WebKitTestRunner/mac/main.mm: (setDefaultsToConsistentValuesForTesting):
Override these preferences to match the default for a new OS X installation.

* DumpRenderTree/mac/DumpRenderTree.mm: (setDefaultsToConsistentValuesForTesting):
Swipe tests are not enabled in WK1, but let's override the default anyway.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, rolling out r192436 and r192586.
commit-queue@webkit.org [Thu, 19 Nov 2015 00:03:26 +0000 (00:03 +0000)]
Unreviewed, rolling out r192436 and r192586.
https://bugs.webkit.org/show_bug.cgi?id=151417

Caused a hang in the inspector and a crash @ google.com.
(Requested by saamyjoon on #webkit).

Reverted changesets:

"Allow any LeftHandSideExpression as a valid
AssignmentElement"
https://bugs.webkit.org/show_bug.cgi?id=151026
http://trac.webkit.org/changeset/192436

"There is a bug when default parameter values are mixed with
destructuring parameter values"
https://bugs.webkit.org/show_bug.cgi?id=151369
http://trac.webkit.org/changeset/192586

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] TailCalls should not trigger assertion failures when running with the jsc profi...
keith_miller@apple.com [Thu, 19 Nov 2015 00:01:32 +0000 (00:01 +0000)]
[JSC] TailCalls should not trigger assertion failures when running with the jsc profiler on.
https://bugs.webkit.org/show_bug.cgi?id=151359

Reviewed by Geoffrey Garen.

If we have the profiler on when parsing tail calls in the dfg we would emit a CountExecutions node
following a TailCall node as our bytecode for tail calls always emits a op_ret following an op_tail_call.
This trips assertions that no nodes follow a terminal in a basic block. This patch fixes this issue by
not emiting a CountExecutions when we have a tail call that is terminal i.e. the tail caller is not
inlined into another function or the tail caller is inlined but is in tail position itself.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::handleCall):
(JSC::DFG::ByteCodeParser::handleVarargsCall):
(JSC::DFG::ByteCodeParser::parseBlock):
* tests/stress/tail-call-profiler.js: Added.
(tail):
(inlineTail):
(inlineTailVarArgs):
(inlineTailTernary):
(body):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoResourceUsageOverlay should have better accounting for reclaimable memory.
akling@apple.com [Wed, 18 Nov 2015 23:48:48 +0000 (23:48 +0000)]
ResourceUsageOverlay should have better accounting for reclaimable memory.
<https://webkit.org/b/151407>

Reviewed by Anders Carlsson.

Add code to inspect the purgeable state of VM regions when traversing the
web process VM map, and track reclaimable regions of memory separately.

Memory categories that have some amount of reclaimable memory are now
displayed with the reclaimable amount in brackets, e.g "123.45 MB [56.78MB]"

* page/ResourceUsageOverlay.h:
* page/cocoa/ResourceUsageOverlayCocoa.mm:
(WebCore::ResourceUsageOverlay::platformDraw):
(WebCore::TagInfo::TagInfo):
(WebCore::pagesPerVMTag):
(WebCore::runSamplerThread):
(WebCore::dirtyPagesPerVMTag): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Probes sidebar panel sections should use NavigationBar
mattbaker@apple.com [Wed, 18 Nov 2015 23:47:52 +0000 (23:47 +0000)]
Web Inspector: Probes sidebar panel sections should use NavigationBar
https://bugs.webkit.org/show_bug.cgi?id=151373

Reviewed by Brian Burg.

Replaces image elements with a NavigationBar instance. The "Clear samples"
button is disabled when no samples exist in the probe set.

* Localizations/en.lproj/localizedStrings.js:
New navigation item labels.

* UserInterface/Models/ProbeSet.js:
(WebInspector.ProbeSet.prototype._sampleCollected):
Dispatch new event WebInspector.ProbeSet.Event.SampleAdded.

* UserInterface/Views/DetailsSection.css:
(.details-section > .header > .options > .navigation-bar):
(.details-section > .header > .options > .navigation-bar > .item):
Define consistent style for showing a NavigationBar inside the section header.

* UserInterface/Views/DetailsSection.js:
(WebInspector.DetailsSection):
Add "options" class name in the base class.

* UserInterface/Views/ProbeDetailsSidebarPanel.css:
(.details-section.probe-set .options > :matches(.probe-remove, .probe-clear-samples, .probe-add)): Deleted.
(.details-section.probe-set .options > .probe-clear-samples): Deleted.
(.details-section.probe-set .options > .probe-remove): Deleted.
(.details-section.probe-set .options > .probe-add): Deleted.
Remove <img> button styles.

* UserInterface/Views/ProbeSetDetailsSection.js:
(WebInspector.ProbeSetDetailsSection):
(WebInspector.ProbeSetDetailsSection.prototype._addProbeButtonClicked.createProbeFromEnteredExpression):
(WebInspector.ProbeSetDetailsSection.prototype._addProbeButtonClicked):
(WebInspector.ProbeSetDetailsSection.prototype._clearSamplesButtonClicked):
(WebInspector.ProbeSetDetailsSection.prototype._probeSetSamplesChanged):
Update "Clear samples" enabled state.
(WebInspector.ProbeSetDetailsSection.prototype._probeSetHasSamples):
Helper function to check if the probe set is empty.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192594 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoClean up / optimize NetworkCacheSubresourcesEntry
cdumez@apple.com [Wed, 18 Nov 2015 23:21:38 +0000 (23:21 +0000)]
Clean up / optimize NetworkCacheSubresourcesEntry
https://bugs.webkit.org/show_bug.cgi?id=151405

Reviewed by Antti Koivisto.

Clean up / optimize NetworkCacheSubresourcesEntry:
- Remove m_sourceStorageRecord data member
- Avoid copying the Key when constructing it

* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::SpeculativeLoadManager::PendingFrameLoad::encodeAsSubresourcesRecord):
* NetworkProcess/cache/NetworkCacheSubresourcesEntry.cpp:
(WebKit::NetworkCache::SubresourcesEntry::SubresourcesEntry):
* NetworkProcess/cache/NetworkCacheSubresourcesEntry.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192593 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Client Blocked Resource Requests causes Crash under InspectorPageAgent...
commit-queue@webkit.org [Wed, 18 Nov 2015 23:17:21 +0000 (23:17 +0000)]
Web Inspector: Client Blocked Resource Requests causes Crash under InspectorPageAgent::cachedResource
https://bugs.webkit.org/show_bug.cgi?id=151398

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2015-11-18
Reviewed by Brian Burg.

Source/WebCore:

Test: inspector/network/client-blocked-load.html

* inspector/InspectorPageAgent.cpp:
(WebCore::InspectorPageAgent::cachedResource):
Gracefully handle null request.

* loader/cache/CachedResourceLoader.cpp:
(WebCore::CachedResourceLoader::cachedResource):
ASSERT if someone tried to pass a null URL.

LayoutTests:

* inspector/network/client-blocked-load-expected.txt: Added.
* inspector/network/client-blocked-load.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192592 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Make LBasicBlock into an alias for B3::BasicBlock. #ifdef anything that does...
benjamin@webkit.org [Wed, 18 Nov 2015 23:14:54 +0000 (23:14 +0000)]
[JSC] Make LBasicBlock into an alias for B3::BasicBlock. #ifdef anything that does not compile
https://bugs.webkit.org/show_bug.cgi?id=151381

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-18
Reviewed by Filip Pizlo.

* JavaScriptCore.xcodeproj/project.pbxproj:
* ftl/FTLAbbreviatedTypes.h:
* ftl/FTLAbbreviations.h:
* ftl/FTLAbstractHeap.cpp:
(JSC::FTL::AbstractHeap::decorateInstruction):
(JSC::FTL::IndexedAbstractHeap::IndexedAbstractHeap):
* ftl/FTLAbstractHeap.h:
(JSC::FTL::AbstractHeap::AbstractHeap):
* ftl/FTLAbstractHeapRepository.cpp:
(JSC::FTL::AbstractHeapRepository::AbstractHeapRepository):
* ftl/FTLAbstractHeapRepository.h:
* ftl/FTLB3Output.h:
(JSC::FTL::Output::newBlock):
(JSC::FTL::Output::insertNewBlocksBefore):
(JSC::FTL::Output::addIncomingToPhi):
(JSC::FTL::Output::Output): Deleted.
(JSC::FTL::Output::initialize): Deleted.
* ftl/FTLCommonValues.cpp:
(JSC::FTL::CommonValues::CommonValues):
* ftl/FTLCommonValues.h:
* ftl/FTLIntrinsicRepository.cpp:
* ftl/FTLIntrinsicRepository.h:
* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::DFG::LowerDFGToLLVM::LowerDFGToLLVM):
(JSC::FTL::DFG::LowerDFGToLLVM::lower):
(JSC::FTL::DFG::LowerDFGToLLVM::compileArithAddOrSub):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutById):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCreateDirectArguments):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNewArrayWithSize):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCallOrConstruct):
(JSC::FTL::DFG::LowerDFGToLLVM::compileTailCall):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCallOrConstructVarargs):
(JSC::FTL::DFG::LowerDFGToLLVM::compileForwardVarargs):
(JSC::FTL::DFG::LowerDFGToLLVM::compileIn):
(JSC::FTL::DFG::LowerDFGToLLVM::compileInstanceOf):
(JSC::FTL::DFG::LowerDFGToLLVM::getById):
(JSC::FTL::DFG::LowerDFGToLLVM::allocateCell):
(JSC::FTL::DFG::LowerDFGToLLVM::buildSwitch):
(JSC::FTL::DFG::LowerDFGToLLVM::lazySlowPath):
(JSC::FTL::DFG::LowerDFGToLLVM::crash):
* ftl/FTLOutput.cpp:
(JSC::FTL::Output::Output):
* ftl/FTLOutput.h:
(JSC::FTL::Output::addIncomingToPhi):
* ftl/FTLState.cpp:
(JSC::FTL::State::dumpState):
* ftl/FTLValueRange.cpp:
(JSC::FTL::ValueRange::decorateInstruction):
* ftl/FTLValueRange.h:
(JSC::FTL::ValueRange::ValueRange):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192591 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoImplement the B3 equivalent of FTLCompile
fpizlo@apple.com [Wed, 18 Nov 2015 23:05:34 +0000 (23:05 +0000)]
Implement the B3 equivalent of FTLCompile
https://bugs.webkit.org/show_bug.cgi?id=151370

Reviewed by Benjamin Poulain.

This adds a B3 version of FTLCompile and gets the data structures related to installing an FTL
compilation (i.e. the finalizer and JITCode) to be aware of B3. That requires stubbing a lot of stuff
out and also simplifying a lot of code around having everything inside a single contiguous chunk of
JIT code rather than some opaque JIT code handles plus miscellaneous side-codes. This compiles but
crashes because lowering isn't done yet.

* JavaScriptCore.xcodeproj/project.pbxproj:
* b3/B3Procedure.cpp:
(JSC::B3::Procedure::addDataSection):
(JSC::B3::Procedure::calleeSaveRegisters):
(JSC::B3::Procedure::addValueIndex):
* b3/B3Procedure.h:
(JSC::B3::Procedure::code):
* ftl/FTLB3Compile.cpp: Added.
(JSC::FTL::compile):
* ftl/FTLCompile.cpp:
(JSC::FTL::mmAllocateDataSection):
* ftl/FTLExceptionHandlerManager.cpp:
(JSC::FTL::ExceptionHandlerManager::callOperationExceptionTarget):
(JSC::FTL::ExceptionHandlerManager::lazySlowPathExceptionTarget):
(JSC::FTL::ExceptionHandlerManager::getByIdOSRExit):
* ftl/FTLJITCode.cpp:
(JSC::FTL::JITCode::~JITCode):
(JSC::FTL::JITCode::initializeB3Code):
(JSC::FTL::JITCode::initializeExitThunks):
(JSC::FTL::JITCode::addDataSection):
(JSC::FTL::JITCode::initializeAddressForCall):
(JSC::FTL::JITCode::initializeArityCheckEntrypoint):
(JSC::FTL::JITCode::addressForCall):
(JSC::FTL::JITCode::contains):
(JSC::FTL::JITCode::exitThunks):
(JSC::FTL::JITCode::ftl):
* ftl/FTLJITCode.h:
(JSC::FTL::JITCode::b3Code):
(JSC::FTL::JITCode::handles):
(JSC::FTL::JITCode::dataSections):
* ftl/FTLJITFinalizer.cpp:
(JSC::FTL::JITFinalizer::codeSize):
(JSC::FTL::JITFinalizer::finalizeFunction):
* ftl/FTLJITFinalizer.h:
* ftl/FTLLink.cpp:
(JSC::FTL::link):
* ftl/FTLOSRExit.cpp:
(JSC::FTL::OSRExit::codeLocationForRepatch):
(JSC::FTL::OSRExit::gatherRegistersToSpillForCallIfException):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192590 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoExpose WebCore's AudioPlaybackRequiresUserGesture setting to WebKit & WebKit2
jer.noble@apple.com [Wed, 18 Nov 2015 23:04:36 +0000 (23:04 +0000)]
Expose WebCore's AudioPlaybackRequiresUserGesture setting to WebKit & WebKit2
https://bugs.webkit.org/show_bug.cgi?id=151274

Reviewed by Anders Carlsson.

Source/WebKit/mac:

Add a WebKit preference key, WebKitAudioPlaybackRequiresUserGesture, and a matching WebPreference
property, and transfer those preferences to WebCore's settings when they change.

* WebView/WebPreferenceKeysPrivate.h:
* WebView/WebPreferences.mm:
(+[WebPreferences initialize]):
(-[WebPreferences audioPlaybackRequiresUserGesture]):
(-[WebPreferences setAudioPlaybackRequiresUserGesture:]):
* WebView/WebPreferencesPrivate.h:
* WebView/WebView.mm:
(-[WebView _preferencesChanged:]):

Source/WebKit2:

Add a new, private, WKWebViewConfiguration property, _requiresUserActionForAudioPlayback,
and use that configuration property to set WebCore's AudioPlaybackRequiresUserGesture setting
during page creation.

* Shared/WebPreferencesDefinitions.h:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetAudioPlaybackRequiresUserGesture): Add legacy C API for setting this new WK preference.
(WKPreferencesGetAudioPlaybackRequiresUserGesture): Ditto.
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]): Set the page configuration value from preferences.
* UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]): Default the value of the new configuration property to YES for iOS.
(-[WKWebViewConfiguration copyWithZone:]): Copy the value across.
(-[WKWebViewConfiguration _requiresUserActionForAudioPlayback]): Simple getter.
(-[WKWebViewConfiguration _setRequiresUserActionForAudioPlayback:]): Simple setter.
* UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences): Set WebCore's audioPlaybackRequiresUserGesture Setting.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192589 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdding a maintenance file for the old WordPress blog.
jond@apple.com [Wed, 18 Nov 2015 22:30:45 +0000 (22:30 +0000)]
Adding a maintenance file for the old WordPress blog.

Reviewed by Timothy Hatcher.

* blog/wp-content/maintenance.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192588 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, skip additional mediastream tests that crash when painting a mock video...
wenson_hsieh@apple.com [Wed, 18 Nov 2015 22:21:23 +0000 (22:21 +0000)]
Unreviewed, skip additional mediastream tests that crash when painting a mock video source.

These tests were not skipped previously because the crashes are timing-dependent and do not
always reproduce. See https://bugs.webkit.org/show_bug.cgi?id=151353 for more details.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192587 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoThere is a bug when default parameter values are mixed with destructuring parameter...
sbarati@apple.com [Wed, 18 Nov 2015 22:01:19 +0000 (22:01 +0000)]
There is a bug when default parameter values are mixed with destructuring parameter values
https://bugs.webkit.org/show_bug.cgi?id=151369

Reviewed by Geoffrey Garen and Mark Lam.

This patch changes our parser to no longer declare destructuring
parameters as "var"s. This is a weird bug that just happened
to work in a world without default parameter values. In a world with
default parameter values this is just completely wrong. It would
incorrectly transform this program:
```function foo(a = function() { b = 40; }, {b}) { a(); return b; }; foo(undefined, {b: 42}); // Should return 40```
into
```function foo(a = function() { b = 40; }, {b}) { var b; a(); return b; }; foo(undefined, {b:42}); // Returns 42, not 40.```
Which is wrong because we end up with two distinct bindings of "b" when
there should only be one.

* parser/Parser.cpp:
(JSC::Parser<LexerType>::parseVariableDeclarationList):
(JSC::Parser<LexerType>::createBindingPattern):
(JSC::Parser<LexerType>::parseDestructuringPattern):
* parser/Parser.h:
(JSC::Scope::declareParameter):
(JSC::Scope::getUsedVariables):
(JSC::Parser::strictMode):
(JSC::Parser::isValidStrictMode):
(JSC::Parser::declareParameter):
(JSC::Parser::breakIsValid):
(JSC::Scope::declareBoundParameter): Deleted.
(JSC::Parser::declareBoundParameter): Deleted.
* tests/stress/es6-default-parameters.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192586 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Timeline Recording across page navigations behaves poorly
commit-queue@webkit.org [Wed, 18 Nov 2015 21:46:33 +0000 (21:46 +0000)]
Web Inspector: Timeline Recording across page navigations behaves poorly
https://bugs.webkit.org/show_bug.cgi?id=151112

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2015-11-18
Reviewed by Timothy Hatcher.

* inspector/InspectorPageAgent.cpp:
(WebCore::InspectorPageAgent::frameStartedLoading): Deleted.
Don't reset the execution stopwatch on page navigation.
If a timeline is actively being recorded on the frontend
then all new timestamps suddenly downshifted towards zero
introduces bad data.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192585 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Watch Expressions details section should use NavigationBar
mattbaker@apple.com [Wed, 18 Nov 2015 21:36:40 +0000 (21:36 +0000)]
Web Inspector: Watch Expressions details section should use NavigationBar
https://bugs.webkit.org/show_bug.cgi?id=151377

Reviewed by Timothy Hatcher.

Replaces image elements with a NavigationBar instance. The refresh and
clear buttons are disabled if no watch expressions exist.

* Localizations/en.lproj/localizedStrings.js:
New navigation item labels.

* UserInterface/Views/ScopeChainDetailsSidebarPanel.css:
(.details-section.watch-expressions .options > *): Deleted.
(.details-section.watch-expressions .options > *:active): Deleted.
(.details-section.watch-expressions .options > .watch-expression-add): Deleted.
(.details-section.watch-expressions .options > .watch-expression-clear): Deleted.
(.details-section.watch-expressions .options > .watch-expression-refresh): Deleted.
Remove <img> button styles.

* UserInterface/Views/ScopeChainDetailsSidebarPanel.js:
(WebInspector.ScopeChainDetailsSidebarPanel):
(WebInspector.ScopeChainDetailsSidebarPanel.prototype._addWatchExpressionButtonClicked.presentPopoverOverTargetElement):
Use the navigation item element.
(WebInspector.ScopeChainDetailsSidebarPanel.prototype._updateWatchExpressionsNavigationBar):
Enable/disable refresh and clear buttons.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192584 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] ASSERTION FAILED: temporaryFilePath.last() == '/' in WebCore::openTemporaryFile()
dbates@webkit.org [Wed, 18 Nov 2015 19:47:08 +0000 (19:47 +0000)]
[iOS] ASSERTION FAILED: temporaryFilePath.last() == '/' in WebCore::openTemporaryFile()
https://bugs.webkit.org/show_bug.cgi?id=151392
<rdar://problem/23595341>

Reviewed by Alexey Proskuryakov.

Workaround <rdar://problem/23579077> where confstr(_CS_DARWIN_USER_TEMP_DIR, ..., ...) retrieves
the path to the user temporary directory without a trailing slash in the iOS simulator.

Source/WebCore:

* platform/mac/FileSystemMac.mm:
(WebCore::openTemporaryFile): Add a path separator (/) between the directory path and filename.

Source/WebKit2:

* Shared/mac/SandboxExtensionMac.mm:
(WebKit::SandboxExtension::createHandleForTemporaryFile): Add a path separator (/) between the directory path and filename.

Source/WTF:

* wtf/DataLog.cpp:
(WTF::initializeLogFileOnce):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192583 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoNull dereference in Performance::Performance(WebCore::Frame*)
cdumez@apple.com [Wed, 18 Nov 2015 19:07:54 +0000 (19:07 +0000)]
Null dereference in Performance::Performance(WebCore::Frame*)
https://bugs.webkit.org/show_bug.cgi?id=151390

Reviewed by Brady Eidson.

Based on the stack trace, it appears the DocumentLoader can be null
when constructing the Performance object. This patch thus adds a null
check before trying to dereference it.

No new tests, was not able to reproduce.

* page/DOMWindow.cpp:
(WebCore::DOMWindow::navigator):
(WebCore::DOMWindow::performance):
* page/Performance.cpp:
(WebCore::Performance::Performance):
(WebCore::Performance::scriptExecutionContext):
* page/Performance.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192582 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoVersioning.
bshafiei@apple.com [Wed, 18 Nov 2015 19:02:00 +0000 (19:02 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192581 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, skip remaining mediastream tests that crash when painting a mock video...
wenson_hsieh@apple.com [Wed, 18 Nov 2015 18:29:39 +0000 (18:29 +0000)]
Unreviewed, skip remaining mediastream tests that crash when painting a mock video source.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192579 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WinCairo][MediaFoundation] The main thread can sometimes be blocked forever when...
peavo@outlook.com [Wed, 18 Nov 2015 18:24:57 +0000 (18:24 +0000)]
[WinCairo][MediaFoundation] The main thread can sometimes be blocked forever when ending a media session.
https://bugs.webkit.org/show_bug.cgi?id=151387

Reviewed by Alex Christensen.

This happens because the main thread is waiting for the sample scheduler thread to finish,
but sometimes it never does. This can be resolved by emptying the scheduler's sample queue
before requesting the scheduler thread to finish. This will make sure the scheduler thread
returns to its main loop quickly, and can handle the termination request.

* platform/graphics/win/MediaPlayerPrivateMediaFoundation.cpp:
(WebCore::MediaPlayerPrivateMediaFoundation::VideoScheduler::stopScheduler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192578 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMarking two new mediastream/RTCPeerConnection tests as crashing on debug during inves...
ryanhaddad@apple.com [Wed, 18 Nov 2015 17:40:25 +0000 (17:40 +0000)]
Marking two new mediastream/RTCPeerConnection tests as crashing on debug during investigation.
https://bugs.webkit.org/show_bug.cgi?id=151304

Unreviewed test gardening.

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192577 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMarking storage/indexeddb/modern/aborted-put.html as flaky on mac-wk1
ryanhaddad@apple.com [Wed, 18 Nov 2015 17:25:32 +0000 (17:25 +0000)]
Marking storage/indexeddb/modern/aborted-put.html as flaky on mac-wk1
https://bugs.webkit.org/show_bug.cgi?id=151389

Unreviewed test gardening.

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192576 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWebRTC: Initial testing of updated RTCPeerConnection API
adam.bergkvist@ericsson.com [Wed, 18 Nov 2015 14:52:29 +0000 (14:52 +0000)]
WebRTC: Initial testing of updated RTCPeerConnection API
https://bugs.webkit.org/show_bug.cgi?id=151304

Reviewed by Eric Carlson.

Source/WebCore:

Add an empty implementation of MediaEndpointPeerConnection which
realizes the PeerConnectionBackend interface. This makes it possible
to construct an RTCPeerConnection and do initial testing.

Tests: fast/mediastream/RTCPeerConnection-overloaded-operations-params.html
       fast/mediastream/RTCPeerConnection-overloaded-operations.html

Also unskip three existing RTCPeerConnection tests on GTK.

* CMakeLists.txt:
* Modules/mediastream/MediaEndpointPeerConnection.cpp: Added.
(WebCore::createMediaEndpointPeerConnection):
(WebCore::MediaEndpointPeerConnection::MediaEndpointPeerConnection):
(WebCore::MediaEndpointPeerConnection::createOffer):
(WebCore::MediaEndpointPeerConnection::createAnswer):
(WebCore::MediaEndpointPeerConnection::setLocalDescription):
(WebCore::MediaEndpointPeerConnection::localDescription):
(WebCore::MediaEndpointPeerConnection::currentLocalDescription):
(WebCore::MediaEndpointPeerConnection::pendingLocalDescription):
(WebCore::MediaEndpointPeerConnection::setRemoteDescription):
(WebCore::MediaEndpointPeerConnection::remoteDescription):
(WebCore::MediaEndpointPeerConnection::currentRemoteDescription):
(WebCore::MediaEndpointPeerConnection::pendingRemoteDescription):
(WebCore::MediaEndpointPeerConnection::setConfiguration):
(WebCore::MediaEndpointPeerConnection::addIceCandidate):
(WebCore::MediaEndpointPeerConnection::getStats):
(WebCore::MediaEndpointPeerConnection::stop):
(WebCore::MediaEndpointPeerConnection::markAsNeedingNegotiation):
* Modules/mediastream/MediaEndpointPeerConnection.h: Added.
* WebCore.xcodeproj/project.pbxproj:

LayoutTests:

Add two new tests and unskip three old ones (details below).

* fast/mediastream/RTCPeerConnection-overloaded-operations-expected.txt: Added.
* fast/mediastream/RTCPeerConnection-overloaded-operations-params-expected.txt: Added.
* fast/mediastream/RTCPeerConnection-overloaded-operations-params.html: Added.
Feeds some good but mostly bad input to the overloaded operations to verify JS built-ins
argument checking.
* fast/mediastream/RTCPeerConnection-overloaded-operations.html: Added.
Verifies that the correct overloaded operation is invoked. For examaple, createOffer()
without arguments should invoke the promise based function, but the same operation with
two function arguments should invoke the legacy callback version.
* platform/gtk/TestExpectations:
Unskip fast/mediastream/RTCIceCandidate.html, fast/mediastream/RTCPeerConnection.html
and fast/mediastream/RTCSessionDescription.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192575 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[Cairo] SolidStroke broken in drawLine after r191658
commit-queue@webkit.org [Wed, 18 Nov 2015 13:25:22 +0000 (13:25 +0000)]
[Cairo] SolidStroke broken in drawLine after r191658
https://bugs.webkit.org/show_bug.cgi?id=151385

Patch by Alejandro G. Castro <alex@igalia.com> on 2015-11-18
Reviewed by Carlos Garcia Campos.

Fix the drawLine function after r191658, we have to make sure the
color is set when line is SolidStroke.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192574 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Grid Layout] inline margins not honored when not using stretch in row-axis alignment
jfernandez@igalia.com [Wed, 18 Nov 2015 12:53:26 +0000 (12:53 +0000)]
[CSS Grid Layout] inline margins not honored when not using stretch in row-axis alignment
https://bugs.webkit.org/show_bug.cgi?id=151323

Reviewed by Sergio Villar Senin.

Source/WebCore:

There are some situations where we avoid to compute the inline-axis
margins when computing the logical width of a box. One of those situations
is when we have set an override width; this only affects for now to flex
and grid items. We also follow this approach when setting the logical
width based on the restrictions of 'auto' value in the 'min-width'
property.

This behavior is not correct, since there is no reason to avoid
computing this margins, in the general case. I think this logic was
designed as an optimization for flexbox, which was already computing
the margins by its own, but it's not applicable in the general case, so
grid needs these margins to be computed properly.

For the shrink-to-fit behavior we can add some grid related logic to the
already defined RenderBox function "sizesLogicalWidthToFitContent",
so we avoid the override width.

Tests: fast/css-grid-layout/grid-item-auto-sized-align-justify-margin-border-padding.html
       fast/css-grid-layout/min-width-height-auto-and-margins.html

* rendering/RenderBox.cpp:
(WebCore::RenderBox::computeLogicalWidthInRegion):
(WebCore::RenderBox::sizesLogicalWidthToFitContent):
* rendering/RenderGrid.cpp:
(WebCore::defaultAlignmentChangedFromStretchInRowAxis):
(WebCore::selfAlignmentChangedFromStretchInRowAxis):
(WebCore::RenderGrid::styleDidChange):
(WebCore::RenderGrid::applyStretchAlignmentToChildIfNeeded):
(WebCore::selfAlignmentChangedFromStretchInColumnAxis): Deleted.
(WebCore::RenderGrid::computeMarginLogicalHeightForChild): Deleted.
(WebCore::RenderGrid::availableAlignmentSpaceForChildBeforeStretching): Deleted.

LayoutTests:

Tests to verify that margins are computed properly on grid items
using an alignment value different than stretch or when min-width
is auto.

* fast/css-grid-layout/grid-item-auto-sized-align-justify-margin-border-padding-expected.txt: Added.
* fast/css-grid-layout/grid-item-auto-sized-align-justify-margin-border-padding.html: Added.
* fast/css-grid-layout/min-width-height-auto-and-margins-expected.txt: Added.
* fast/css-grid-layout/min-width-height-auto-and-margins.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192573 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK] Update xserver version in jhbuild
carlosgc@webkit.org [Wed, 18 Nov 2015 12:43:39 +0000 (12:43 +0000)]
[GTK] Update xserver version in jhbuild
https://bugs.webkit.org/show_bug.cgi?id=151383

Reviewed by Alejandro G. Castro.

In r192568 I assumed the bots would have a recent enough Xvfb
to use the -displayfd option, but I forgot we build our own
xserver just for Xvfb. The version we build is too old, so
we could just use an updated version (the same bots have
installed, for example) to use the -displayfd option.

* gtk/jhbuild.modules: Update xserver to 1.16.4.
* gtk/patches/xserver-fix-glx-init.patch: Deleted.
* gtk/patches/xserver-remove-bogus-dependencies.patch: Updated to
apply on 1.16.4.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192572 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWebRTC: [GTK] Skip failing MediaStream tests
adam.bergkvist@ericsson.com [Wed, 18 Nov 2015 11:30:48 +0000 (11:30 +0000)]
WebRTC: [GTK] Skip failing MediaStream tests
https://bugs.webkit.org/show_bug.cgi?id=151360

Reviewed by Eric Carlson.

* platform/gtk/TestExpectations:
Two tests skipped. Awaits resolution of https://webkit.org/b/151344.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192571 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAX: Shadow DOM video player controls menus need aria-owns on the trigger buttons
commit-queue@webkit.org [Wed, 18 Nov 2015 11:12:42 +0000 (11:12 +0000)]
AX: Shadow DOM video player controls menus need aria-owns on the trigger buttons
https://bugs.webkit.org/show_bug.cgi?id=127065

Patch by Aaron Chu <arona.chu@gmail.com> on 2015-11-18
Reviewed by Darin Adler.

Source/WebCore:

Test: media/accessibility-closed-captions-has-aria-owns.html

* Modules/mediacontrols/mediaControlsApple.js:
(Controller.prototype.createControls):
(Controller.prototype.buildCaptionMenu):
* Modules/mediacontrols/mediaControlsBase.js:
(Controller.prototype.createControls):
(Controller.prototype.buildCaptionMenu):

LayoutTests:

* media/accessibility-closed-captions-has-aria-owns-expected.txt: Added.
* media/accessibility-closed-captions-has-aria-owns.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192570 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed. Fix webkitpy tests after r192568.
carlosgc@webkit.org [Wed, 18 Nov 2015 11:02:43 +0000 (11:02 +0000)]
Unreviewed. Fix webkitpy tests after r192568.

Instead of ignoring exception when closing mock fds, use a helper
function that the unit test can override to do nothing.

* Scripts/webkitpy/port/xvfbdriver.py:
(XvfbDriver._xvfb_close_pipe):
(XvfbDriver._xvfb_run):
* Scripts/webkitpy/port/xvfbdriver_unittest.py:
(XvfbDriverTest.make_driver):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192569 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[XvfbDriver] Fail to run all layout tests when X server started with -displayfd option
carlosgc@webkit.org [Wed, 18 Nov 2015 09:05:28 +0000 (09:05 +0000)]
[XvfbDriver] Fail to run all layout tests when X server started with -displayfd option
https://bugs.webkit.org/show_bug.cgi?id=151135

Reviewed by Darin Adler.

The XvfbDriver uses the x server command line to check the
displays that are currently in use. This doesn't work when X
server was started with -displayfd option. This option is used to let
the server find the display id available that is written to the
given file descriptor. With this option xorg doesn't need to
create the lock files in tmp either. The -displayfd option is also
available in Xvfb, so we could use it too. That would simplify the
code, fixing also race conditions between the check for available
displays and Xvfb opening the connection, we wouldn't need to wait
for 4 seconds after launching Xvfb, and all lock files we are
using won't be needed either.

* Scripts/webkitpy/port/xvfbdriver.py:
(XvfbDriver._xvfb_pipe): Helper function to create the pipe, only
needed to be overriden by unit tests.
(XvfbDriver._xvfb_read_display_id): Helper function to read from
the pipe, only needed to be overriden by unit tests.
(XvfbDriver._xvfb_run): Run Xvfb with -displayfd option, using a
pipe to read the display id.
(XvfbDriver._start): Call _xvfb_run() and remove the code to run
Xvfb for a given display.
(XvfbDriver.stop): Remove code to release and delete file locks.
* Scripts/webkitpy/port/xvfbdriver_unittest.py:
(XvfbDriverTest.make_driver):
(XvfbDriverTest.test_start):
(XvfbDriverTest.test_start_arbitrary_worker_number):
(XvfbDriverTest.test_stop):
(XvfbDriverTest.assertDriverStartSuccessful): Deleted.
(XvfbDriverTest): Deleted.
(XvfbDriverTest.test_stop.FakeXvfbProcess): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192568 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoNull dereference loading Blink layout test editing/execCommand/indent-button-crash...
carlosgc@webkit.org [Wed, 18 Nov 2015 07:52:47 +0000 (07:52 +0000)]
Null dereference loading Blink layout test editing/execCommand/indent-button-crash.html
https://bugs.webkit.org/show_bug.cgi?id=151187

Reviewed by Darin Adler.

Source/WebCore:

This is a merge of Blink r174671:
https://codereview.chromium.org/291143002

Fixes imported/blink/editing/execCommand/indent-button-crash.html.

* editing/ApplyBlockElementCommand.cpp:
(WebCore::ApplyBlockElementCommand::doApply):

LayoutTests:

Unskip the test.

* platform/gtk/TestExpectations:
* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192567 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION(r192459): [GTK] User agent string is broken after r192459
carlosgc@webkit.org [Wed, 18 Nov 2015 07:48:06 +0000 (07:48 +0000)]
REGRESSION(r192459): [GTK] User agent string is broken after r192459
https://bugs.webkit.org/show_bug.cgi?id=151347

Reviewed by Žan Doberšek.

.:

Pass UA version numbers as strings to the build.

* Source/cmake/OptionsGTK.cmake:

Source/WebCore:

Remove the incorrect macro and simply use the given values
USER_AGENT_GTK_MAJOR_VERSION and USER_AGENT_GTK_MINOR_VERSION that
are now strings.

* platform/gtk/UserAgentGtk.cpp:
(WebCore::versionForUAString):
(WebCore::buildUserAgentString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192566 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agorun-webkit-tests should not truncate persistent lines
ap@apple.com [Wed, 18 Nov 2015 06:27:09 +0000 (06:27 +0000)]
run-webkit-tests should not truncate persistent lines
https://bugs.webkit.org/show_bug.cgi?id=151376
rdar://problem/23162775

Reviewed by Antti Koivisto.

It is OK to truncate progress lines, but it's not OK to truncate error (or otherwise
unexpected result) messages that remain on the screen to be read. In practice, this
almost always truncates the test path, which is important to see.

* Scripts/webkitpy/layout_tests/views/printing.py:
(Printer._test_status_line):
(Printer.print_finished_test):
* Scripts/webkitpy/layout_tests/views/printing_unittest.py:
(Testprinter.test_test_status_line):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192565 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSimple line layout: Add letter-spacing support.
zalan@apple.com [Wed, 18 Nov 2015 03:00:43 +0000 (03:00 +0000)]
Simple line layout: Add letter-spacing support.
https://bugs.webkit.org/show_bug.cgi?id=151362

Reviewed by Antti Koivisto.

This enables us to use simple line layout on letter-spacing content.
(fixme: webkit.org/b/151368 -> Repaint rect is not computed correctly when negative letter-spacing applied)

Source/WebCore:

Test: fast/text/simple-line-letterspacing.html

* rendering/SimpleLineLayout.cpp:
(WebCore::SimpleLineLayout::canUseForStyle):
* rendering/SimpleLineLayoutFunctions.cpp:
(WebCore::SimpleLineLayout::paintFlow): RenderLineBoxList tests vertical intersection only.
* rendering/SimpleLineLayoutTextFragmentIterator.cpp:
(WebCore::SimpleLineLayout::TextFragmentIterator::runWidth):

LayoutTests:

* fast/text/simple-line-letterspacing-expected.html: Added.
* fast/text/simple-line-letterspacing.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192564 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Untangle the dependencies on FTLAbbreviations a bit
benjamin@webkit.org [Wed, 18 Nov 2015 01:53:44 +0000 (01:53 +0000)]
[JSC] Untangle the dependencies on FTLAbbreviations a bit
https://bugs.webkit.org/show_bug.cgi?id=151375

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-17
Reviewed by Geoffrey Garen.

FTLAbbreviations was included in many places but in most cases
we just need the Abbreviated types. Update the #include accordingly.

Also remove the IntrinsicRepository dependency from the B3Output.

* ftl/FTLAbstractHeap.cpp:
* ftl/FTLAbstractHeap.h:
* ftl/FTLAbstractHeapRepository.cpp:
* ftl/FTLB3Output.h:
(JSC::FTL::Output::Output):
(JSC::FTL::Output::stackmapIntrinsic):
(JSC::FTL::Output::frameAddressIntrinsic):
(JSC::FTL::Output::patchpointInt64Intrinsic):
(JSC::FTL::Output::patchpointVoidIntrinsic):
* ftl/FTLCommonValues.cpp:
* ftl/FTLCommonValues.h:
* ftl/FTLFormattedValue.h:
* ftl/FTLIntrinsicRepository.cpp:
* ftl/FTLOSRExit.h:
* ftl/FTLOSRExitCompilationInfo.h:
* ftl/FTLStackmapArgumentList.h:
* ftl/FTLState.cpp:
* ftl/FTLState.h:
* ftl/FTLTypedPointer.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192561 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Move Timeline creation into a recording
commit-queue@webkit.org [Wed, 18 Nov 2015 01:34:49 +0000 (01:34 +0000)]
Web Inspector: Move Timeline creation into a recording
https://bugs.webkit.org/show_bug.cgi?id=151367

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2015-11-17
Reviewed by Timothy Hatcher.

* UserInterface/Controllers/TimelineManager.js:
(WebInspector.TimelineManager.prototype._loadNewRecording): Deleted.
* UserInterface/Models/Timeline.js:
(WebInspector.Timeline.create):
(WebInspector.Timeline.prototype.get type):
(WebInspector.Timeline): Deleted.
(WebInspector.Timeline.prototype.get recording): Deleted.
* UserInterface/Models/TimelineRecording.js:
(WebInspector.TimelineRecording):
(WebInspector.TimelineRecording.prototype.get readonly):
(WebInspector.TimelineRecording.prototype.unloaded):
(WebInspector.TimelineRecording.prototype.reset):
(WebInspector.TimelineRecording.prototype.isWritable): Deleted.
* UserInterface/Views/TimelineOverview.js:
(WebInspector.TimelineOverview):
* UserInterface/Views/TimelineRecordingContentView.js:
(WebInspector.TimelineRecordingContentView.prototype.shown):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192560 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoReorganize the graphics benchmark string table
commit-queue@webkit.org [Wed, 18 Nov 2015 01:29:15 +0000 (01:29 +0000)]
Reorganize the graphics benchmark string table
https://bugs.webkit.org/show_bug.cgi?id=151334

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-11-17
Reviewed by Simon Fraser.

Make the graphics benchmark string table be an object of sub-objects. Each
sub-object represents an associative array, the key is the string name and
the value is the string data.

* Animometer/resources/extensions.js:
(ResultsDashboard.prototype.toJSON):
(ResultsTable.prototype._showGraph):
(ResultsTable.prototype._showJSON):
(ResultsTable.prototype._isNoisyMeasurement):
(ResultsTable.prototype._isNoisyTest):
(ResultsTable.prototype._showTest):
(ResultsTable.prototype._showSuite):
(ResultsTable.prototype._showIteration):
* Animometer/resources/sampler.js:
(Sampler.prototype.toJSON):
* Animometer/resources/strings.js:
* Animometer/runner/resources/animometer.js:
(window.benchmarkRunnerClient.didFinishLastIteration):
(window.suitesManager.updateLocalStorageFromJSON):
(window.benchmarkController.showResults):
(window.benchmarkController.showJson):
(window.benchmarkController.showTestGraph):
(window.benchmarkController.showTestJSON):
* Animometer/runner/resources/tests.js:
* Animometer/tests/resources/main.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192559 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoB3::generate should separate out the final Air codegen, so that it can be done outsid...
fpizlo@apple.com [Wed, 18 Nov 2015 01:27:28 +0000 (01:27 +0000)]
B3::generate should separate out the final Air codegen, so that it can be done outside the Graph safepoint
https://bugs.webkit.org/show_bug.cgi?id=151371

Reviewed by Benjamin Poulain.

One of the FTL optimizations is that while the expensive backend is running, we are at a "graph
safepoint" that allows the VM to do GCs and other dangerous and time-sensitive things without
waiting for the compilation thread. While in the safepoint, we cannot do anything that touches
anything other than the backend's state. That means, for example, that we wouldn't be able to run
any of the stackmap generation callbacks, since those need to mess with DFG state.

That means that we need to separate the B3 pipeline into "preparation" and "generation".
Preparation is all of the expensive stuff: all B3 phases, lowering to Air, all Air phases.
Generation is just the part where we turn fully lowered Air into machine code. Generation also
happens to be the part where we call stackmap generation callbacks. In other words, preparation
is exactly the stuff that should go into the graph safepoint, while generation is the stuff that
we want to do after we emerge from the safepoint.

Because the tests were using the higher-level Compilation API, I didn't have to change any test
code. The FTL will not use that high-level API.

* b3/B3Compilation.cpp:
(JSC::B3::Compilation::Compilation):
* b3/B3Generate.cpp:
(JSC::B3::prepareForGeneration):
(JSC::B3::generate):
(JSC::B3::generateToAir):
* b3/B3Generate.h:
* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::LowerToAir):
(JSC::B3::lowerToAir):
* b3/B3LowerToAir.h:
* b3/B3Procedure.cpp:
(JSC::B3::Procedure::Procedure):
* b3/B3Procedure.h:
(JSC::B3::Procedure::takeByproducts):
(JSC::B3::Procedure::code):
* b3/air/AirCode.h:
(JSC::B3::Air::Code::proc):
(JSC::B3::Air::Code::lastPhaseName):
* b3/air/AirGenerate.cpp:
(JSC::B3::Air::prepareForGeneration):
(JSC::B3::Air::generate):
* b3/air/AirGenerate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192558 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Breakpoint condition field should use CodeMirror
mattbaker@apple.com [Wed, 18 Nov 2015 00:23:56 +0000 (00:23 +0000)]
Web Inspector: Breakpoint condition field should use CodeMirror
https://bugs.webkit.org/show_bug.cgi?id=151333

Reviewed by Joseph Pecoraro.

Updated breakpoint popover's condition field to use CodeMirror.

* UserInterface/Controllers/BreakpointPopoverController.js:
Removed keyboard shortcuts. Now handled by CodeMirror.
(WebInspector.BreakpointPopoverController.prototype.completionControllerShouldAllowEscapeCompletion):
Prevent handling of escape to allow tabbing to next element.
(WebInspector.BreakpointPopoverController.prototype._createPopoverContent):
Create CodeMirror instance.
(WebInspector.BreakpointPopoverController.prototype._conditionCodeMirrorChanged):
(WebInspector.BreakpointPopoverController.prototype._conditionCodeMirrorBeforeChange):
Enforce single-line editor.
(WebInspector.BreakpointPopoverController.prototype._conditionCodeMirrorEscapeOrEnterKey):
(WebInspector.BreakpointPopoverController): Deleted.
(WebInspector.BreakpointPopoverController.prototype.appendContextMenuItems.editBreakpoint):
Removed condition element select. CodeMirror sets focus when popover content is created.
(WebInspector.BreakpointPopoverController.prototype._popoverConditionInputChanged): Deleted.
Replaced by BreakpointPopoverController.prototype._conditionCodeMirrorChanged.
(WebInspector.BreakpointPopoverController.prototype._popoverConditionInputKeyDown): Deleted.
Replaced by BreakpointPopoverController.prototype._conditionCodeMirrorEscapeOrEnterKey.

* UserInterface/Views/BreakpointPopoverController.css:
(.edit-breakpoint-popover-condition):
(.edit-breakpoint-popover-condition > .CodeMirror):
(#edit-breakpoint-popover-condition): Deleted.
Replace selector id with class, styled CodeMirror element to match original input element.

* UserInterface/Views/CSSStyleDeclarationTextEditor.css:
(.css-style-text-editor > .CodeMirror .CodeMirror-placeholder): Deleted.
Moved common CodeMirror placeholder text style to CodeMirrorOverrides.css.

* UserInterface/Views/CodeMirrorOverrides.css:
(.CodeMirror .CodeMirror-placeholder):
Add placeholder style.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192551 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDisable flattening the stage iframe of the graphics benchmark when running on iOS
said@apple.com [Wed, 18 Nov 2015 00:16:56 +0000 (00:16 +0000)]
Disable flattening the stage iframe of the graphics benchmark when running on iOS
https://bugs.webkit.org/show_bug.cgi?id=151361

Reviewed by Simon Fraser.

Use fixed size for stage iframe of the graphics benchmark to disable
flattening the iframe while animating the particles. Also ensure the
bouncing particles do not go outside the iframe's boundaries.

* Animometer/runner/resources/animometer.css:
(section#running > #running-test > iframe):
(@media screen and (min-device-width: 1800px)):
* Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
(BouncingParticle.prototype.animate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192550 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUse the media queries to dynamically set the stage for the graphics benchmark
commit-queue@webkit.org [Wed, 18 Nov 2015 00:08:38 +0000 (00:08 +0000)]
Use the media queries to dynamically set the stage for the graphics benchmark
https://bugs.webkit.org/show_bug.cgi?id=151327

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-11-17
Reviewed by Simon Fraser.

Clean setting the benchmark css rules by using the media queries. Accessing
document.stylesheets.cssRules seems to be unreliable. Fix the test harness
load event listener invocation. And also remove the options for normalizing
the stage resolution for retina display.

* Animometer/resources/extensions.js:
(window.DocumentExtension.insertCssRuleAfter): Deleted.
* Animometer/runner/animometer.html:
* Animometer/runner/resources/animometer.css:
(@media screen and (min-device-width: 1800px)):
* Animometer/runner/resources/animometer.js:
(window.sectionsManager.setupRunningSectionStyle):
(window.benchmarkController.initialize):
(window.sectionsManager.setupSectionStyle): Deleted.
* Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
(BouncingParticlesStage):
* Animometer/tests/text/layering-text.html:
* Animometer/tests/text/resources/layering-text.js:
(LayeringTextStage):
(LayeringTextStage.prototype._setFontSize): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192549 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, fix a ChangeLog entry to have the right filename.
fpizlo@apple.com [Wed, 18 Nov 2015 00:07:34 +0000 (00:07 +0000)]
Unreviewed, fix a ChangeLog entry to have the right filename.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192548 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFTL::State should be able to refer to B3::Procedure
fpizlo@apple.com [Wed, 18 Nov 2015 00:05:47 +0000 (00:05 +0000)]
FTL::State should be able to refer to B3::Procedure

Rubber stamped by Benjamin Poulain.

* ftl/FTLState.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192547 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Add an empty FTLOutput for B3 that crashes on any operation
benjamin@webkit.org [Tue, 17 Nov 2015 23:44:01 +0000 (23:44 +0000)]
[JSC] Add an empty FTLOutput for B3 that crashes on any operation
https://bugs.webkit.org/show_bug.cgi?id=151366

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-17
Reviewed by Filip Pizlo.

* JavaScriptCore.xcodeproj/project.pbxproj:
* ftl/B3Output.h: Added.
* ftl/FTLAbbreviatedTypes.h:
* ftl/FTLAbstractHeap.cpp:
* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::DFG::LowerDFGToLLVM::createPhiVariables):
* ftl/FTLOutput.cpp:
* ftl/FTLOutput.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192546 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMarking storage/indexeddb/modern/get-index-failures.html as flaky on mac-wk1
ryanhaddad@apple.com [Tue, 17 Nov 2015 23:15:31 +0000 (23:15 +0000)]
Marking storage/indexeddb/modern/get-index-failures.html as flaky on mac-wk1
https://bugs.webkit.org/show_bug.cgi?id=151364

Unreviewed test gardening.

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192545 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Do not copy the adjacency list when we just need to manipulate them
commit-queue@webkit.org [Tue, 17 Nov 2015 23:10:14 +0000 (23:10 +0000)]
[JSC] Do not copy the adjacency list when we just need to manipulate them
https://bugs.webkit.org/show_bug.cgi?id=151343

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-17
Reviewed by Geoffrey Garen.

* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::precoloredCoalescingHeuristic):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::conservativeHeuristic):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192544 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Remove FTLOutput operations that act directly on floats
commit-queue@webkit.org [Tue, 17 Nov 2015 22:48:19 +0000 (22:48 +0000)]
[JSC] Remove FTLOutput operations that act directly on floats
https://bugs.webkit.org/show_bug.cgi?id=151342

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-17
Reviewed by Geoffrey Garen.

* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::DFG::LowerDFGToLLVM::compileArithFRound):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetByVal):
(JSC::FTL::DFG::LowerDFGToLLVM::compilePutByVal):
* ftl/FTLOutput.cpp:
(JSC::FTL::Output::store):
* ftl/FTLOutput.h:
(JSC::FTL::Output::loadFloatToDouble):
(JSC::FTL::Output::fround):
(JSC::FTL::Output::loadFloat): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192543 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd a FTL_USES_B3 compile-time setting and set it to 0.
fpizlo@apple.com [Tue, 17 Nov 2015 22:46:53 +0000 (22:46 +0000)]
Add a FTL_USES_B3 compile-time setting and set it to 0.

Rubber stamped by Benjamin Poulain.

* dfg/DFGCommon.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192542 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WinCairo][MediaFoundation] Current playback time is not shown.
peavo@outlook.com [Tue, 17 Nov 2015 22:41:24 +0000 (22:41 +0000)]
[WinCairo][MediaFoundation] Current playback time is not shown.
https://bugs.webkit.org/show_bug.cgi?id=151357

Reviewed by Alex Christensen.

We need to implement the currentTime() method.

* platform/graphics/win/MediaPlayerPrivateMediaFoundation.cpp:
(WebCore::MediaPlayerPrivateMediaFoundation::durationDouble):
(WebCore::MediaPlayerPrivateMediaFoundation::currentTime):
(WebCore::MediaPlayerPrivateMediaFoundation::paused):
(WebCore::MediaPlayerPrivateMediaFoundation::CustomVideoPresenter::paintCurrentFrame):
(WebCore::MediaPlayerPrivateMediaFoundation::CustomVideoPresenter::currentTime):
(WebCore::MediaPlayerPrivateMediaFoundation::CustomVideoPresenter::isActive):
* platform/graphics/win/MediaPlayerPrivateMediaFoundation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192541 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCheckAdd/Mul should have commutativity optimizations in B3->Air lowering
fpizlo@apple.com [Tue, 17 Nov 2015 22:31:40 +0000 (22:31 +0000)]
CheckAdd/Mul should have commutativity optimizations in B3->Air lowering
https://bugs.webkit.org/show_bug.cgi?id=151214

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:

This is an overhaul of CheckAdd/CheckSub/CheckMul that fixes bugs, improves codegen, and
simplifies the contract between B3 and its client.

Previously, the idea was that the operands to the Air BranchAdd/Sub/Mul matched the children of
the B3 CheckAdd/Sub/Mul, or at least, that's what the B3::CheckSpecial would make you believe.
This meant that B3/Air had to avoid optimizations that would break this protocol. This prevented
commutativity optimizations on CheckAdd/Mul and it also prevented strength reduction from
CheckMul(x, 2) to CheckAdd(x, x), for example. Those optimizations would break things because the
client's Stackmap generation callback was allowed to assume that the first entry in params.reps
was the first child. Also, there was a contract between B3 and its client that for CheckAdd/Sub,
the client would undo the operation by doing the opposite operation with the params.reps[0] as the
source and params.reps[1] as the destination.

This not only prevented commutativity optimizations, it also led to bugs. Here are two bugs that
we had:

- Add(x, x) would not work. The client would be told to undo the add using %x as both the source
  and destination. The client would use a sub() instruction. The result would not be x - it would
  be zero.

- Mul where the result got coalesced with one of the sources. You can't undo a multiplication, so
  you need to keep the inputs alive until after the result is computed - i.e. the inputs are late
  uses. I initially thought I worked around this by using a three-operand form of Mul, but of
  course that doesn't actually fix the issue.

This patch fixes these issues comprehensively by introducing the following changes:

The params.reps corresponding to the first two children of CheckAdd/Sub/Mul and the first child of
Check are always garbage: if you want to know the values of those children in the slow path, pass
them as additional stackmap children. This makes it clear to the compiler whose values you
actually care about, and frees the compiler to reorder and commute the non-stackmap children.

The "undo" of an Add or Sub is handled internally by B3: the client doesn't have to know anything
about undoing. B3 does it. The contract is simply that if a B3::Value* is a stackmap child of a
CheckXYZ, then the corresponding entry in params.reps inside the stackmap generator callback will
contain the value of that Value*. For Add and Sub, B3 undoes the operation. For Add(x, x), the
undo uses the carry flag and some shift magic. For Mul, B3 uses LateUse:

A new kind of Arg::Role called Arg::LateUse is introduced: This kind of use means that the use
happens at the start of the execution of the next instruction. None of the built-in Air opcodes
use LateUse, but it is used by B3::CheckSpecial. We use it to implement CheckMul.

Finally, this change fixes testComplex to actually create many live variables. This revealed a
really dumb pathology in allocateStack(), and this patch fixes it. Basically, it's a bad idea to
build interference graphs by repeatedly recreating the same cliques.

* assembler/MacroAssemblerX86Common.h:
(JSC::MacroAssemblerX86Common::test32):
(JSC::MacroAssemblerX86Common::setCarry):
(JSC::MacroAssemblerX86Common::invert):
* b3/B3CheckSpecial.cpp:
(JSC::B3::Air::numB3Args):
(JSC::B3::CheckSpecial::Key::Key):
(JSC::B3::CheckSpecial::Key::dump):
(JSC::B3::CheckSpecial::CheckSpecial):
(JSC::B3::CheckSpecial::forEachArg):
(JSC::B3::CheckSpecial::isValid):
(JSC::B3::CheckSpecial::generate):
(JSC::B3::CheckSpecial::dumpImpl):
(JSC::B3::CheckSpecial::deepDumpImpl):
* b3/B3CheckSpecial.h:
(JSC::B3::CheckSpecial::Key::Key):
(JSC::B3::CheckSpecial::Key::operator==):
(JSC::B3::CheckSpecial::Key::operator!=):
(JSC::B3::CheckSpecial::Key::opcode):
(JSC::B3::CheckSpecial::Key::numArgs):
(JSC::B3::CheckSpecial::Key::stackmapRole):
(JSC::B3::CheckSpecial::Key::hash):
* b3/B3CheckValue.cpp:
(JSC::B3::CheckValue::~CheckValue):
(JSC::B3::CheckValue::convertToAdd):
(JSC::B3::CheckValue::CheckValue):
* b3/B3CheckValue.h:
* b3/B3Const32Value.cpp:
(JSC::B3::Const32Value::checkMulConstant):
(JSC::B3::Const32Value::checkNegConstant):
(JSC::B3::Const32Value::divConstant):
* b3/B3Const32Value.h:
* b3/B3Const64Value.cpp:
(JSC::B3::Const64Value::checkMulConstant):
(JSC::B3::Const64Value::checkNegConstant):
(JSC::B3::Const64Value::divConstant):
* b3/B3Const64Value.h:
* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::appendBinOp):
(JSC::B3::Air::LowerToAir::lower):
* b3/B3Opcode.h:
* b3/B3PatchpointSpecial.cpp:
(JSC::B3::PatchpointSpecial::~PatchpointSpecial):
(JSC::B3::PatchpointSpecial::forEachArg):
(JSC::B3::PatchpointSpecial::isValid):
* b3/B3ReduceStrength.cpp:
* b3/B3StackmapSpecial.cpp:
(JSC::B3::StackmapSpecial::forEachArgImpl):
* b3/B3StackmapSpecial.h:
* b3/B3StackmapValue.cpp:
(JSC::B3::StackmapValue::append):
(JSC::B3::StackmapValue::appendSomeRegister):
(JSC::B3::StackmapValue::setConstrainedChild):
* b3/B3StackmapValue.h:
* b3/B3Validate.cpp:
* b3/B3Value.cpp:
(JSC::B3::Value::checkMulConstant):
(JSC::B3::Value::checkNegConstant):
(JSC::B3::Value::divConstant):
* b3/B3Value.h:
* b3/air/AirAllocateStack.cpp:
(JSC::B3::Air::allocateStack):
* b3/air/AirArg.cpp:
(WTF::printInternal):
* b3/air/AirArg.h:
(JSC::B3::Air::Arg::isAnyUse):
(JSC::B3::Air::Arg::isEarlyUse):
(JSC::B3::Air::Arg::isLateUse):
(JSC::B3::Air::Arg::isDef):
(JSC::B3::Air::Arg::forEachTmp):
(JSC::B3::Air::Arg::isUse): Deleted.
* b3/air/AirGenerate.cpp:
(JSC::B3::Air::generate):
* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::build):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::allocate):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::InterferenceEdge::hash):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::InterferenceEdge::dump):
(JSC::B3::Air::addSpillAndFillToProgram):
(JSC::B3::Air::iteratedRegisterCoalescingOnType):
(JSC::B3::Air::iteratedRegisterCoalescing):
* b3/air/AirLiveness.h:
(JSC::B3::Air::Liveness::Liveness):
(JSC::B3::Air::Liveness::LocalCalc::LocalCalc):
(JSC::B3::Air::Liveness::LocalCalc::live):
(JSC::B3::Air::Liveness::LocalCalc::takeLive):
(JSC::B3::Air::Liveness::LocalCalc::execute):
* b3/air/AirOpcode.opcodes:
* b3/air/AirReportUsedRegisters.cpp:
(JSC::B3::Air::reportUsedRegisters):
* b3/air/AirSpillEverything.cpp:
(JSC::B3::Air::spillEverything):
* b3/testb3.cpp:
(JSC::B3::testMulArg):
(JSC::B3::testMulArgStore):
(JSC::B3::testMulAddArg):
(JSC::B3::testMulArgs):
(JSC::B3::testComplex):
(JSC::B3::testSimpleCheck):
(JSC::B3::testCheckLessThan):
(JSC::B3::testCheckMegaCombo):
(JSC::B3::testCheckAddImm):
(JSC::B3::testCheckAddImmCommute):
(JSC::B3::testCheckAddImmSomeRegister):
(JSC::B3::testCheckAdd):
(JSC::B3::testCheckAdd64):
(JSC::B3::testCheckSubImm):
(JSC::B3::testCheckSubBadImm):
(JSC::B3::testCheckSub):
(JSC::B3::testCheckSub64):
(JSC::B3::testCheckNeg):
(JSC::B3::testCheckNeg64):
(JSC::B3::testCheckMul):
(JSC::B3::testCheckMulMemory):
(JSC::B3::testCheckMul2):
(JSC::B3::testCheckMul64):
(JSC::B3::run):

Source/WTF:

Disable my failed attempts at perfect forwarding, since they were incorrect, and caused compile
errors if you tried to pass an argument that bound to lvalue. This shouldn't affect performance of
anything we care about, and performance tests seem to confirm that it's all good.

* wtf/ScopedLambda.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192540 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAir should lay out code optimally
fpizlo@apple.com [Tue, 17 Nov 2015 22:29:54 +0000 (22:29 +0000)]
Air should lay out code optimally
https://bugs.webkit.org/show_bug.cgi?id=150478

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:

This adds a phase that optimizes code layout using something that's worked well for me in the past.
Basically, it just forces pre-ordering on the CFG, except that:

- Blocks that are only reachable by Rare control flow are scheduled separately, all the way at the
  end.

- Successors of the same frequency class are pushed in ascending order of frequency, so that the most
  frequent successor is scheduled immediately after.

This also adds the requisite branch flipping, so that a branch's taken successor is not the
fall-through block. We want the fall-through to be the not-taken successor if at all possible.

* JavaScriptCore.xcodeproj/project.pbxproj:
* b3/B3BlockWorklist.h:
* b3/B3GenericFrequentedBlock.h:
(JSC::B3::GenericFrequentedBlock::frequency):
(JSC::B3::GenericFrequentedBlock::isRare):
(JSC::B3::GenericFrequentedBlock::dump):
* b3/B3Procedure.h:
* b3/air/AirArg.h:
(JSC::B3::Air::Arg::isDoubleCond):
(JSC::B3::Air::Arg::isCondition):
(JSC::B3::Air::Arg::isSpecial):
(JSC::B3::Air::Arg::asDoubleCondition):
(JSC::B3::Air::Arg::isInvertible):
(JSC::B3::Air::Arg::inverted):
* b3/air/AirBasicBlock.h:
(JSC::B3::Air::BasicBlock::index):
(JSC::B3::Air::BasicBlock::setIndex):
(JSC::B3::Air::BasicBlock::size):
(JSC::B3::Air::BasicBlock::begin):
(JSC::B3::Air::BasicBlock::containsPredecessor):
(JSC::B3::Air::BasicBlock::frequency):
* b3/air/AirBlockWorklist.h: Added.
* b3/air/AirCode.cpp:
(JSC::B3::Air::Code::findNextBlock):
* b3/air/AirCode.h:
(JSC::B3::Air::Code::proc):
(JSC::B3::Air::Code::at):
(JSC::B3::Air::Code::operator[]):
(JSC::B3::Air::Code::blockList):
* b3/air/AirGenerate.cpp:
(JSC::B3::Air::generate):
* b3/air/AirOpcode.opcodes:
* b3/air/AirOptimizeBlockOrder.cpp: Added.
(JSC::B3::Air::optimizeBlockOrder):
* b3/air/AirOptimizeBlockOrder.h: Added.

Source/WTF:

* wtf/GraphNodeWorklist.h:
(WTF::GraphNodeWorklist::push):
(WTF::GraphNodeWorklist::isEmpty):
(WTF::GraphNodeWorklist::notEmpty):
(WTF::GraphNodeWorklist::pop):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192539 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] JSPropertyNameEnumerator could be destructorless.
akling@apple.com [Tue, 17 Nov 2015 22:13:41 +0000 (22:13 +0000)]
[JSC] JSPropertyNameEnumerator could be destructorless.
<https://webkit.org/b/151242>

Reviewed by Mark Lam.

Make JSPropertyNameEnumerator destructorless and have it store the property names
cache in CopiedSpace. This was the most popular occupant of 64-byte destructor cells
in MarkedSpace, so making it destructorless gets rid of some ill-filled MarkedBlocks.

This patch had two issues on 32-bit platforms when first landed:

- Copied space allocations are required to be 8-byte divisible in size.

- WriteBarrier<Unknown> and WriteBarrier<JSString> are not the same size on 32-bit;
  the former is a 64-bit EncodedJSValue internally, and the latter is a 32-bit JSCell*.
  My patch was reinterpret_cast'ing a WriteBarrier<JSString> to a WriteBarrier<Unknown>
  when passing to SlotVisitor::appendValues(), which led to invalid addresses getting
  marked and strings getting GC'd prematurely.

* heap/CopyToken.h:
* runtime/JSPropertyNameEnumerator.cpp:
(JSC::JSPropertyNameEnumerator::finishCreation):
(JSC::JSPropertyNameEnumerator::visitChildren):
(JSC::JSPropertyNameEnumerator::copyBackingStore):
(JSC::JSPropertyNameEnumerator::destroy): Deleted.
* runtime/JSPropertyNameEnumerator.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192536 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRefactoring: move branchMul32's imm arg to the 3rd argument to be consistent.
mark.lam@apple.com [Tue, 17 Nov 2015 22:09:07 +0000 (22:09 +0000)]
Refactoring: move branchMul32's imm arg to the 3rd argument to be consistent.
https://bugs.webkit.org/show_bug.cgi?id=151358

Reviewed by Saam Barati.

branch32, branchAdd32, and branchSub32 all have it as the 3rd argument.
branchMul32 is the odd man out.  This patch makes branchMul32 consistent with
the others.

* assembler/MacroAssembler.h:
(JSC::MacroAssembler::branchAdd32):
(JSC::MacroAssembler::branchMul32):
* assembler/MacroAssemblerARM.h:
(JSC::MacroAssemblerARM::branchMul32):
* assembler/MacroAssemblerARM64.h:
(JSC::MacroAssemblerARM64::branchMul32):
* assembler/MacroAssemblerARMv7.h:
(JSC::MacroAssemblerARMv7::branchMul32):
* assembler/MacroAssemblerMIPS.h:
(JSC::MacroAssemblerMIPS::branchMul32):
* assembler/MacroAssemblerSH4.h:
(JSC::MacroAssemblerSH4::branchNeg32):
(JSC::MacroAssemblerSH4::branchMul32):
* assembler/MacroAssemblerX86Common.h:
(JSC::MacroAssemblerX86Common::branchMul32):
* jit/JITArithmetic.cpp:
(JSC::JIT::emit_op_mul):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192535 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUse the JITAddGenerator snippet in the DFG.
mark.lam@apple.com [Tue, 17 Nov 2015 21:55:33 +0000 (21:55 +0000)]
Use the JITAddGenerator snippet in the DFG.
https://bugs.webkit.org/show_bug.cgi?id=151266

Reviewed by Geoffrey Garen.

No tests added because the op_add.js stress test already tests for correctness
(using the LLINT as a reference).

Performance-wise, the difference from the pre-existing DFG implementation is
insignificant (at least as measured on x86 and x86_64).  We're moving forward
with adopting this implementation because it unifies the 32-bit and 64-bit
implementations, as well as lays ground work for a repatching inline cache
implementation of op_add later.

* dfg/DFGAbstractValue.cpp:
(JSC::DFG::AbstractValue::resultType):
- Made an assertion less restrictive.  For ValueAdd operands, the DFG thinks that
  the operand can also be empty (though we should never see this in practice).

* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):
- Add fallback to unused type operands.

* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileValueAdd):
- Introduce a common function to compile the ValueAdd node.

* dfg/DFGSpeculativeJIT.h:
(JSC::DFG::JSValueOperand::JSValueOperand):
- Add the forwarding constructor so that we can use Optional<JSValueOperand>.

* dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
- Changed to use the common compileValueAdd().

* jit/AssemblyHelpers.h:
(JSC::AssemblyHelpers::moveValue):
- Similar to moveTrustedValue() but used for untrusted constants.

* jit/JITAddGenerator.cpp:
(JSC::JITAddGenerator::generateFastPath):
* jit/JITAddGenerator.h:
(JSC::JITAddGenerator::JITAddGenerator):
- Updated to take the left or right operand as a constant.  This is necessary
  because the client should not be making assumptions about whether the snippet
  will determine the operation to be commutative or not.  Instead, the client
  should just pass in the operands and let the snippet do any operand order
  swapping if necessary.

* jit/JITArithmetic.cpp:
(JSC::JIT::emit_op_add):
- Updated to use the new JITAddGenerator interface.

* tests/stress/op_add.js:
(stringifyIfNeeded):
(runTest):
- Made test output more clear about when string results are expected.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192531 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoIt's best for the DFG to always have some guess of basic block frequency
fpizlo@apple.com [Tue, 17 Nov 2015 21:41:43 +0000 (21:41 +0000)]
It's best for the DFG to always have some guess of basic block frequency
https://bugs.webkit.org/show_bug.cgi?id=151350

Reviewed by Geoffrey Garen.

It'll simplify things for B3 if we always have some estimate of block frequency, even if it's not
a great estimate. Using NaN as an estimate is probably worse than using any non-NaN number.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::inlineCall):
(JSC::DFG::ByteCodeParser::handleInlining):
(JSC::DFG::ByteCodeParser::parseCodeBlock):
* dfg/DFGCriticalEdgeBreakingPhase.cpp:
(JSC::DFG::CriticalEdgeBreakingPhase::breakCriticalEdge):
* dfg/DFGLoopPreHeaderCreationPhase.cpp:
(JSC::DFG::createPreHeader):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192529 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoA window with a hung tab waits 5s before closing
ggaren@apple.com [Tue, 17 Nov 2015 21:31:57 +0000 (21:31 +0000)]
A window with a hung tab waits 5s before closing
https://bugs.webkit.org/show_bug.cgi?id=151319

Reviewed by Anders Carlsson.

I manually tested that I did not break unload events upon window close
by running run-webkit-httpd and invoking a 1x1 image load of
http://127.0.0.1:8000/navigation/resources/save-Ping.php from an unload
handler.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::tryClose): Use the sudden termination bit to
close without waiting for a response.

* UIProcess/WebProcessProxy.h:
(WebKit::WebProcessProxy::isSuddenTerminationEnabled): Expose the bit.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192528 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoReserved VM pool established in r187125 is likely too conservative
msaboff@apple.com [Tue, 17 Nov 2015 21:30:01 +0000 (21:30 +0000)]
Reserved VM pool established in r187125 is likely too conservative
https://bugs.webkit.org/show_bug.cgi?id=151351

Reviewed by Filip Pizlo.

Reduce the VM allocation reserved pool from 25% to 15% for ARM32.

* jit/ExecutableAllocator.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192527 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSplit SimpleLineLayout::canUseFor into canUseForStyle and canUseForFontAndText.
zalan@apple.com [Tue, 17 Nov 2015 21:16:07 +0000 (21:16 +0000)]
Split SimpleLineLayout::canUseFor into canUseForStyle and canUseForFontAndText.
https://bugs.webkit.org/show_bug.cgi?id=151338

Reviewed by Myles C. Maxfield.

No change in functionality.

* rendering/SimpleLineLayout.cpp:
(WebCore::SimpleLineLayout::canUseForFontAndText):
(WebCore::SimpleLineLayout::canUseForStyle):
(WebCore::SimpleLineLayout::canUseFor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192526 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFTLLazySlowPaths should be able to handle being passed the zero register as a location
sbarati@apple.com [Tue, 17 Nov 2015 20:52:26 +0000 (20:52 +0000)]
FTLLazySlowPaths should be able to handle being passed the zero register as a location
https://bugs.webkit.org/show_bug.cgi?id=151193

Reviewed by Geoffrey Garen.

On arm64, SP and ZR are the same register number. The meaning
of the register number being SP or ZR is dependent on context of
the instruction and the register within the instruction. LLVM may
prove that a value is zero, or sometimes, we will lower a
value as a constant zero (for example, we might compile
CreateDirectArguments of an inlined call frame and we might know
that the arguments have a length of zero). LazySlowPaths should
be able to gracefully handle being passed a stackmap location
with a gpr value of SP by moving zero into another register
and replacing the location's register with the new register.
This way, no lazy slow path will ever have access to a location with a GPR
value of SP.  This patch makes this work by using a scratch register
allocator when we need to do this maneuver of moving zero into a scratch
register.  Inside FTLCompile, we track if we need to move zero into a register,
and if so, into which register. We actually emit the necessary
instructions to move zero into this register, and to spill reused
registers if necessary, while generating the code for the lazy slow
path itself.

* ftl/FTLCompile.cpp:
(JSC::FTL::mmAllocateDataSection):
* ftl/FTLDWARFRegister.cpp:
(JSC::FTL::DWARFRegister::reg):
* ftl/FTLLazySlowPath.cpp:
(JSC::FTL::LazySlowPath::LazySlowPath):
(JSC::FTL::LazySlowPath::generate):
* ftl/FTLLazySlowPath.h:
(JSC::FTL::LazySlowPath::createGenerator):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192525 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Support Doubles with B3's Mul.
mark.lam@apple.com [Tue, 17 Nov 2015 20:24:12 +0000 (20:24 +0000)]
[JSC] Support Doubles with B3's Mul.
https://bugs.webkit.org/show_bug.cgi?id=151355

Reviewed by Filip Pizlo.

* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::lower):
* b3/air/AirOpcode.opcodes:
* b3/testb3.cpp:
(JSC::B3::testMulLoadTwice):
(JSC::B3::testMulArgDouble):
(JSC::B3::testMulArgsDouble):
(JSC::B3::testMulArgImmDouble):
(JSC::B3::testMulImmArgDouble):
(JSC::B3::testMulImmsDouble):
(JSC::B3::testDivArgDouble):
(JSC::B3::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192524 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemoved useless variable scriptResult from testExecutionTimeLimit
commit-queue@webkit.org [Tue, 17 Nov 2015 19:56:01 +0000 (19:56 +0000)]
Removed useless variable scriptResult from testExecutionTimeLimit
https://bugs.webkit.org/show_bug.cgi?id=151331

Patch by Matthew Daiter <mdaiter8121@gmail.com> on 2015-11-17
Reviewed by Alex Christensen.

* API/tests/ExecutionTimeLimitTest.cpp: Removed scriptResult
(testExecutionTimeLimit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192520 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Support Doubles with B3's Sub.
mark.lam@apple.com [Tue, 17 Nov 2015 19:41:27 +0000 (19:41 +0000)]
[JSC] Support Doubles with B3's Sub.
https://bugs.webkit.org/show_bug.cgi?id=151322

Reviewed by Filip Pizlo.

* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::lower):
* b3/air/AirOpcode.opcodes:

* b3/testb3.cpp:
(JSC::B3::testSubImmArg32):
(JSC::B3::testSubArgDouble):
(JSC::B3::testSubArgsDouble):
(JSC::B3::testSubArgImmDouble):
(JSC::B3::testSubImmArgDouble):
(JSC::B3::testSubImmsDouble):
(JSC::B3::testBitAndArgs):
(JSC::B3::negativeZero):
(JSC::B3::posInfinity):
(JSC::B3::negInfinity):
(JSC::B3::doubleOperands):
(JSC::B3::run):
- Added RUN_UNARY and RUN_BINARY macros to auto generate permutations
  of operands for the tests.
- Added SubDouble tests using the new macros.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192519 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoModern IDB: Support IDBObjectStore.indexNames.
beidson@apple.com [Tue, 17 Nov 2015 19:41:26 +0000 (19:41 +0000)]
Modern IDB: Support IDBObjectStore.indexNames.
https://bugs.webkit.org/show_bug.cgi?id=151341

Reviewed by Alex Christensen.

Source/WebCore:

No new tests (Covered by existing storage/indexeddb/modern/objectstore-attributes.html).

* Modules/indexeddb/client/IDBObjectStoreImpl.cpp:
(WebCore::IDBClient::IDBObjectStore::indexNames):
* Modules/indexeddb/shared/IDBObjectStoreInfo.cpp:
(WebCore::IDBObjectStoreInfo::indexNames):
* Modules/indexeddb/shared/IDBObjectStoreInfo.h:

LayoutTests:

* storage/indexeddb/modern/objectstore-attributes-expected.txt:
* storage/indexeddb/modern/objectstore-attributes.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192518 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCommit working changes from build.webkit.org
lforschler@apple.com [Tue, 17 Nov 2015 19:30:14 +0000 (19:30 +0000)]
Commit working changes from build.webkit.org

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192517 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] IteratedRegisterCoalescingAllocator's freeze can add zombie Tmps to our colorin...
commit-queue@webkit.org [Tue, 17 Nov 2015 19:11:22 +0000 (19:11 +0000)]
[JSC] IteratedRegisterCoalescingAllocator's freeze can add zombie Tmps to our coloring algorithm
https://bugs.webkit.org/show_bug.cgi?id=151345

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-17
Reviewed by Filip Pizlo.

The extended test of https://bugs.webkit.org/show_bug.cgi?id=151214 revealed a problem with
the Move freezing step of the iterated register allocator.

When freezing a move related Tmp, we go over all the move instructions and update the other
Tmp of the move as needed. If that Tmp is no longer Move related and is of low degree it can
be simplified too.

The problem arise when one of those Tmp was already coalesced. For example, say we have

  1: move %tmp42, %tmp43
  2: move %tmp42, %tmp44

The first move cannot be coalesced for some reason and is moved to the Active worklist.
The second move is coalesced and %tmp42 is now %tmp44.

When nothing can be simplified or coalesced, we look at freezing a Move-related Tmp.
Let's say we pick %tmp43. The code of freeze() was going to all the Move instructions
to update them. Here you would find @1 above, get %tmp42 as the other Tmp. Since %tmp42
is no longer move related, it is added to the simplify work list.
-> We just added a coalesced Tmp to our worklist!

To fix that, I get the alias of every other Tmp before updating its status.

There is still a problem: multiple Moves instructions may have been coalesced to the same
Tmp. We must not add a Tmp twice to a worklist.
To avoid duplicates, I rely on the fact that the freeze worklist should not contains
aliased Tmps. Before adding something to the simplify worklist, I check that it was
in the freeze worklist.

I cannot find how this problem would have been avoided in the original paper. I suspect
that may have been a bug in the pseudo-code.

* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::freeze):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::freezeMoves):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192516 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, skip crashing mediastream tests
wenson_hsieh@apple.com [Tue, 17 Nov 2015 19:05:29 +0000 (19:05 +0000)]
Unreviewed, skip crashing mediastream tests

Tests in fast/mediastream/ that attempt to render mock video source frames crash due to
use-after-free. https://bugs.webkit.org/show_bug.cgi?id=151353 tracks these crashes.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192515 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WK1] Mark test failure to get bots green after r192499
bfulgham@apple.com [Tue, 17 Nov 2015 17:58:42 +0000 (17:58 +0000)]
[WK1] Mark test failure to get bots green after r192499
https://bugs.webkit.org/show_bug.cgi?id=151055

* webgl/1.0.3/151055_asan.html: Mark webgl/1.0.3/151055_asan.html as failing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192514 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoASSERTION FAILED: contentSize >= 0 in WebCore::RenderFlexibleBox::adjustChildSizeForM...
svillar@igalia.com [Tue, 17 Nov 2015 10:13:19 +0000 (10:13 +0000)]
ASSERTION FAILED: contentSize >= 0 in WebCore::RenderFlexibleBox::adjustChildSizeForMinAndMax
https://bugs.webkit.org/show_bug.cgi?id=151025

Reviewed by Darin Adler.

Source/WebCore:

Negative margins could make RenderFlexibleBox compute negative
intrinsic sizes. Clamp intrinsic sizes to 0.

Test: css3/flexbox/negative-margins-assert.html

* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::computeIntrinsicLogicalWidths):

LayoutTests:

* css3/flexbox/negative-margins-assert-expected.txt: Added.
* css3/flexbox/negative-margins-assert.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192513 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[css-grid] Fix alignment with gutters and negative free spaces
svillar@igalia.com [Tue, 17 Nov 2015 10:11:43 +0000 (10:11 +0000)]
[css-grid] Fix alignment with gutters and negative free spaces
https://bugs.webkit.org/show_bug.cgi?id=151307

Reviewed by Zalan Bujtas.

Source/WebCore:

This is a regression caused by r192154. The problem was that
the free space was not properly computed when the grid
container was narrower than the track sizes + gutters size for
a given axis. The culprit was an unneeded clamp to 0 which was
polluting the freeSpace computation as the track sizing
algorithm works perfectly OK with negative freeSpace values.

* rendering/RenderGrid.cpp:
(WebCore::RenderGrid::computeTrackSizesForDirection):

LayoutTests:

* fast/css-grid-layout/grid-gutters-and-alignment-expected.txt:
* fast/css-grid-layout/grid-gutters-and-alignment.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192512 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GStreamer] Use RunLoop instead of GMainLoop in AudioFileReaderGStreamer
carlosgc@webkit.org [Tue, 17 Nov 2015 09:57:42 +0000 (09:57 +0000)]
[GStreamer] Use RunLoop instead of GMainLoop in AudioFileReaderGStreamer
https://bugs.webkit.org/show_bug.cgi?id=151256

Reviewed by Žan Doberšek.

Use RunLoop instead of the platform specific code. The AudioBus
can be created from any thread, so we create a helper thread to
ensure we don't use the main RunLoop.

This patch also includes some code cleanups:
  - Uses smart pointers when possible.
  - Fixes uninitialized members in constructors.
  - Makes private members private.
  - Uses lambdas instead of static non-members functions.
  - nullptr instead of 0 in some places.

* platform/audio/gstreamer/AudioFileReaderGStreamer.cpp:
(WebCore::AudioFileReader::createWeakPtr):
(WebCore::AudioFileReader::deinterleavePadAddedCallback):
(WebCore::AudioFileReader::deinterleaveReadyCallback):
(WebCore::AudioFileReader::decodebinPadAddedCallback):
(WebCore::AudioFileReader::AudioFileReader):
(WebCore::AudioFileReader::~AudioFileReader):
(WebCore::AudioFileReader::handleSample):
(WebCore::AudioFileReader::handleMessage):
(WebCore::AudioFileReader::handleNewDeinterleavePad):
(WebCore::AudioFileReader::deinterleavePadsConfigured):
(WebCore::AudioFileReader::plugDeinterleave):
(WebCore::AudioFileReader::decodeAudioForBusCreation):
(WebCore::AudioFileReader::createBus):
(WebCore::createBusFromAudioFile):
(WebCore::createBusFromInMemoryAudioFile):
(WebCore::onAppsinkPullRequiredCallback): Deleted.
(WebCore::messageCallback): Deleted.
(WebCore::onGStreamerDeinterleavePadAddedCallback): Deleted.
(WebCore::onGStreamerDeinterleaveReadyCallback): Deleted.
(WebCore::onGStreamerDecodebinPadAddedCallback): Deleted.
* platform/graphics/gstreamer/GRefPtrGStreamer.cpp:
(WTF::adoptGRef):
(WTF::refGPtr<GstBufferList>):
(WTF::derefGPtr<GstBufferList>):
* platform/graphics/gstreamer/GRefPtrGStreamer.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192511 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK] Web Process crashes on reparenting a WebView with AC mode on
carlosgc@webkit.org [Tue, 17 Nov 2015 08:37:21 +0000 (08:37 +0000)]
[GTK] Web Process crashes on reparenting a WebView with AC mode on
https://bugs.webkit.org/show_bug.cgi?id=151139

Reviewed by Mario Sanchez Prada.

When the web view is reparented, the widget is first unrealized,
and then realized again when added to the new parent. In the
second realize, the old redirected XComposite window is destroyed
and a new one is created, but the web process is still using the
old redirected window ID. As soon as the redirected window is
destroyed and the web process tries to use the window ID, it
crashes due to a BadDrawable X error. We have to notify the web
process as soon as the web view is unrealized to stop using the
current window ID and exit accelerated compositing mode until a
new window ID is given. This notification needs to be synchronous,
because the window can be destroyed in the UI process before the
message is received in the web process.

* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseRealize): Add an assert to ensure we never have
a redirected window when the view is realized. Also check drawing
area is not nullptr, since it can be destroyed at any time if the
web process crashes.
(webkitWebViewBaseUnrealize): Call
DrawingAreaProxyImpl::destroyNativeSurfaceHandleForCompositing()
and destroy the redirected XComposite window.
* UIProcess/DrawingAreaProxyImpl.cpp:
(WebKit::DrawingAreaProxyImpl::destroyNativeSurfaceHandleForCompositing):
Send DestroyNativeSurfaceHandleForCompositing synchronous messsage
to the web process.
* UIProcess/DrawingAreaProxyImpl.h:
* WebProcess/WebPage/DrawingArea.h:
* WebProcess/WebPage/DrawingArea.messages.in: Add
DestroyNativeSurfaceHandleForCompositing message.
* WebProcess/WebPage/DrawingAreaImpl.cpp:
(WebKit::DrawingAreaImpl::destroyNativeSurfaceHandleForCompositing):
Set the native surface handler for compositing to 0 to reset it.
* WebProcess/WebPage/DrawingAreaImpl.h:
* WebProcess/WebPage/gtk/LayerTreeHostGtk.cpp:
(WebKit::LayerTreeHostGtk::makeContextCurrent): Return false
early always when layer tree context ID is 0, even if we already
have a context.
(WebKit::LayerTreeHostGtk::setNativeSurfaceHandleForCompositing):
Cancel any pending layer flush when setting a new handler.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192510 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed EFL Gardening on 16th Nov.
hs85.jeong@samsung.com [Tue, 17 Nov 2015 08:11:37 +0000 (08:11 +0000)]
Unreviewed EFL Gardening on 16th Nov.
https://bugs.webkit.org/show_bug.cgi?id=151339

* platform/efl/fast/text/decorations-with-text-combine-expected.txt:
* platform/efl/fast/text/international/text-combine-image-test-expected.txt:
* platform/efl/fast/text/mark-matches-overflow-clip-expected.txt: Added.
* platform/efl/fast/text/text-combine-shrink-on-color-change-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192509 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Make FTLOutput's load8() and load16() compatible with B3
benjamin@webkit.org [Tue, 17 Nov 2015 06:20:21 +0000 (06:20 +0000)]
[JSC] Make FTLOutput's load8() and load16() compatible with B3
https://bugs.webkit.org/show_bug.cgi?id=151336

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-16
Reviewed by Filip Pizlo.

B3 does not have 8bit and 16bit types. Make FTLOutput abstract enough
to handle LLVM IR and B3.

* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetArrayLength):
(JSC::FTL::DFG::LowerDFGToLLVM::compileGetByVal):
(JSC::FTL::DFG::LowerDFGToLLVM::compileStringCharAt):
(JSC::FTL::DFG::LowerDFGToLLVM::compileStringCharCodeAt):
(JSC::FTL::DFG::LowerDFGToLLVM::compileNotifyWrite):
(JSC::FTL::DFG::LowerDFGToLLVM::compileSwitch):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCheckHasInstance):
(JSC::FTL::DFG::LowerDFGToLLVM::compileCheckWatchdogTimer):
(JSC::FTL::DFG::LowerDFGToLLVM::speculateTruthyObject):
(JSC::FTL::DFG::LowerDFGToLLVM::boolify):
(JSC::FTL::DFG::LowerDFGToLLVM::equalNullOrUndefined):
(JSC::FTL::DFG::LowerDFGToLLVM::switchStringRecurse):
(JSC::FTL::DFG::LowerDFGToLLVM::isObject):
(JSC::FTL::DFG::LowerDFGToLLVM::isNotObject):
(JSC::FTL::DFG::LowerDFGToLLVM::isArrayType):
(JSC::FTL::DFG::LowerDFGToLLVM::isExoticForTypeof):
(JSC::FTL::DFG::LowerDFGToLLVM::isType):
(JSC::FTL::DFG::LowerDFGToLLVM::speculateNonNullObject):
(JSC::FTL::DFG::LowerDFGToLLVM::loadCellState):
(JSC::FTL::DFG::LowerDFGToLLVM::emitStoreBarrier):
* ftl/FTLOutput.h:
(JSC::FTL::Output::load8SignExt32):
(JSC::FTL::Output::load8ZeroExt32):
(JSC::FTL::Output::load16SignExt32):
(JSC::FTL::Output::load16ZeroExt32):
(JSC::FTL::Output::constInt8): Deleted.
(JSC::FTL::Output::load8): Deleted.
(JSC::FTL::Output::load16): Deleted.
(JSC::FTL::Output::store8): Deleted.
(JSC::FTL::Output::storeFloat): Deleted.
(JSC::FTL::Output::isZero8): Deleted.
(JSC::FTL::Output::notZero8): Deleted.
(JSC::FTL::Output::testIsZero8): Deleted.
(JSC::FTL::Output::testNonZero8): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192508 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemove the option for animating using setInterval from the graphics benchmark
commit-queue@webkit.org [Tue, 17 Nov 2015 03:29:43 +0000 (03:29 +0000)]
Remove the option for animating using setInterval from the graphics benchmark
https://bugs.webkit.org/show_bug.cgi?id=151283

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-11-16
Reviewed by Darin Adler.

Get rid of the option to run animation using setInterval().

* Animometer/tests/resources/main.js:
(Animator.prototype.animate):
(Benchmark):
(Benchmark.prototype.start):
(Animator.prototype.start): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192505 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[MediaStream] VideoTrack should respond to MediaStreamTrack state changes
eric.carlson@apple.com [Tue, 17 Nov 2015 02:52:16 +0000 (02:52 +0000)]
[MediaStream] VideoTrack should respond to MediaStreamTrack state changes
https://bugs.webkit.org/show_bug.cgi?id=151299

Reviewed by Jer Noble.

Source/WebCore:

Test: fast/mediastream/MediaStream-video-element-track-stop.html

* Modules/mediastream/MediaStreamTrack.cpp:
(WebCore::MediaStreamTrack::ended): Check m_ended.
(WebCore::MediaStreamTrack::stopProducingData): Set m_ended before telling private track to
  stop so we won't fire an 'ended' event.
(WebCore::MediaStreamTrack::trackEnded): Set m_ended, add comments from spec.
* Modules/mediastream/MediaStreamTrack.h:

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::mediaPlayerTimeChanged): Live streams have infinite duration,
  not indefinite.

* platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.h:
* platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::platformLayer): Return NULL when displayMode
  is None.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::currentDisplayMode): Calculate current
  displayMode based on MediaStream state.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::updateDisplayMode): Renamed from
  setPausedImageVisible, use m_displayMode. Don't start/stop clock, it makes more sense
  to do that in play and pause.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::play): Start clock.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::pause): Stop clock.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::currentReadyState): Call updateDisplayMode.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::activeStatusChanged): Ditto.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::updateIntrinsicSize):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::createPreviewLayers): Always try to create
  both layers if necessary.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::renderingModeChanged): New, update displayMode
  and call back to the media element.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::characteristicsChanged):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::updateTracks): Get rid of some unused locals.
  React to a change in the enabled video track.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::paintCurrentFrameInContext): Paint black when
  the active video track is disabled.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::setPausedImageVisible): Deleted.

* platform/mediastream/MediaStreamPrivate.cpp:
(WebCore::MediaStreamPrivate::updateActiveState): Split the logic of updating the active video
  track out into updateActiveVideoTrack.
(WebCore::MediaStreamPrivate::hasVideo): Check for track.ended.
(WebCore::MediaStreamPrivate::hasAudio): Ditto.
(WebCore::MediaStreamPrivate::paintCurrentFrameInContext): Don't call the source directly.
(WebCore::MediaStreamPrivate::updateActiveVideoTrack): New.
(WebCore::MediaStreamPrivate::trackEnabledChanged): Call updateActiveVideoTrack.
(WebCore::MediaStreamPrivate::trackEnded): new.
* platform/mediastream/MediaStreamPrivate.h:

* platform/mediastream/MediaStreamTrackPrivate.cpp:
(WebCore::MediaStreamTrackPrivate::setEnabled): Call observers if enabled state changes.
(WebCore::MediaStreamTrackPrivate::endTrack): Call observers.
(WebCore::MediaStreamTrackPrivate::paintCurrentFrameInContext): New.
(WebCore::MediaStreamTrackPrivate::preventSourceFromStopping): Add comments.
* platform/mediastream/MediaStreamTrackPrivate.h:

* platform/mock/MockRealtimeVideoSource.cpp:
(WebCore::MockRealtimeVideoSource::paintCurrentFrameInContext): Remove a commented-out line.

LayoutTests:

* fast/mediastream/MediaStream-video-element-expected.txt:
* fast/mediastream/MediaStream-video-element-track-stop-expected.txt: Added.
* fast/mediastream/MediaStream-video-element-track-stop.html: Added.
* fast/mediastream/MediaStream-video-element.html:
* platform/gtk/TestExpectations: Skip new test.
* platform/ios-simulator/TestExpectations: Ditto.
* platform/mac-wk2/TestExpectations: Ditto.
* platform/win/TestExpectations: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192503 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix CMake build and make PluginProcess executable
commit-queue@webkit.org [Tue, 17 Nov 2015 02:10:35 +0000 (02:10 +0000)]
Fix CMake build and make PluginProcess executable
https://bugs.webkit.org/show_bug.cgi?id=151332

Patch by Alex Christensen <achristensen@webkit.org> on 2015-11-16
Reviewed by Tim Horton.

.:

* Source/cmake/OptionsMac.cmake:

Source/WebKit2:

* PlatformMac.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192502 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMulti-WKWebView find completes after the first view
timothy_horton@apple.com [Tue, 17 Nov 2015 01:21:06 +0000 (01:21 +0000)]
Multi-WKWebView find completes after the first view
https://bugs.webkit.org/show_bug.cgi?id=151325

Reviewed by Anders Carlsson.

* UIProcess/mac/WKTextFinderClient.mm:
(-[WKTextFinderClient findMatchesForString:relativeToMatch:findOptions:maxResults:resultCollector:]):
Update progress before calling the completion handler; NSTextFinder assumes
that if the progress doesn't change, the search is complete.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192500 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSource/WebCore:
bfulgham@apple.com [Tue, 17 Nov 2015 01:19:10 +0000 (01:19 +0000)]
Source/WebCore:
Fixes the buffer-overflow when reading characters from textRun
https://bugs.webkit.org/attachment.cgi?bugid=151055
<rdar://problem/23251789>

Patch by Pranjal Jumde <pjumde@apple.com> on 2015-11-16
Reviewed by Brent Fulgham.

* platform/graphics/FontCascade.cpp

LayoutTests:
Checks for buffer-overflows when reading characters from textRun
https://bugs.webkit.org/attachment.cgi?bugid=151055
<rdar://problem/23251789>

Patch by Pranjal Jumde <pjumde@apple.com> on 2015-11-16
Reviewed by Brent Fulgham.

* webgl/1.0.3/151055_asan-expected.txt
* webgl/1.0.3/151055_asan.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192499 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix a typo in AirIteratedRegisterCoalescing
benjamin@webkit.org [Tue, 17 Nov 2015 00:53:05 +0000 (00:53 +0000)]
Fix a typo in AirIteratedRegisterCoalescing

I forgot to fix that review comment from Geoff.

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-16

* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::build):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192498 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Add support for the extra registers that can be clobbered by Specials
benjamin@webkit.org [Tue, 17 Nov 2015 00:44:28 +0000 (00:44 +0000)]
[JSC] Add support for the extra registers that can be clobbered by Specials
https://bugs.webkit.org/show_bug.cgi?id=151246

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-16
Reviewed by Geoffrey Garen.

Specials can clobber arbitrary registers. This was not handled correctly by Air
and nothing was preventing us from re-allocating those registers.

This patch adds support for the extra clobbered registers in the two register allocators.

I also fixed the re-spilling FIXME of the iterated allocator because the test might
not always converge without it. Since we are at maximum register pressure at the patch point,
we could be always spilling the return value, which would loop forever.

To fix the re-spilling, I just kept a Set of every value spilled or filled so far. When selecting
a spill candidate, we never pick a Tmp from that set.

* b3/air/AirGenerate.cpp:
(JSC::B3::Air::generate):
* b3/air/AirHandleCalleeSaves.cpp:
(JSC::B3::Air::handleCalleeSaves):
* b3/air/AirInst.h:
* b3/air/AirInstInlines.h:
(JSC::B3::Air::Inst::forEachDefAndExtraClobberedTmp):
* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::IteratedRegisterCoalescingAllocator):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::build):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::addEdges):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::selectSpill):
(JSC::B3::Air::addSpillAndFillToProgram):
(JSC::B3::Air::iteratedRegisterCoalescingOnType):
(JSC::B3::Air::iteratedRegisterCoalescing):
* b3/air/AirSpillEverything.cpp:
(JSC::B3::Air::spillEverything):
* b3/testb3.cpp:
(JSC::B3::testSimplePatchpointWithoutOuputClobbersGPArgs):
(JSC::B3::testSimplePatchpointWithOuputClobbersGPArgs):
(JSC::B3::testSimplePatchpointWithoutOuputClobbersFPArgs):
(JSC::B3::testSimplePatchpointWithOuputClobbersFPArgs):
(JSC::B3::run):
* jit/RegisterSet.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192497 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWebFrameLoaderClient::dispatchDecidePolicyForNavigationAction calls an std::function...
ryuan.choi@navercorp.com [Tue, 17 Nov 2015 00:26:57 +0000 (00:26 +0000)]
WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction calls an std::function after it's been moved from
https://bugs.webkit.org/show_bug.cgi?id=151248

Reviewed by Darin Adler.

Like r188287, calling an empty std::function results in a std::bad_function_call
exception being thrown when sendSync is failed.

* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192496 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoBuild fix after r192492
benjamin@webkit.org [Tue, 17 Nov 2015 00:14:54 +0000 (00:14 +0000)]
Build fix after r192492

Last second change broke the build.

* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::contains):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::takeMove):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::takeLastMove):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192495 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoHighlight the alarming test results in the graphics benchmark results page
commit-queue@webkit.org [Mon, 16 Nov 2015 23:52:46 +0000 (23:52 +0000)]
Highlight the alarming test results in the graphics benchmark results page
https://bugs.webkit.org/show_bug.cgi?id=151286

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-11-16
Reviewed by Simon Fraser.

When showing the results of a test in the graphics benchmark the following
criteria is going to be applied:

1. If the standard deviation of the test complexity or the frame rate is
   equal to or more than 10%, the standard deviation and the test name
   will be displayed in red.
2. If the average frame rate is not in the range = [(desired_frame_rate - 2)
   .. (desired_frame_rate + 2)], the average frame rate and the test name will
   be displayed in red.

* Animometer/resources/extensions.js:
(ResultsTable.prototype._showHeaderRow):
(ResultsTable.prototype._showHeader):
(ResultsTable.prototype._showEmptyCell):
(ResultsTable.prototype._showText):
(ResultsTable.prototype._showFixedNumber):
(ResultsTable.prototype.):
(ResultsTable.prototype._showGraph):
(ResultsTable.prototype._showJSON):
(ResultsTable.prototype._isAlarmingMeasurement):
(ResultsTable.prototype._isAlarmingTestResults):
(ResultsTable.prototype._showEmptyCells):
(ResultsTable.prototype._showEmptyRow):
(ResultsTable.prototype._showTest):
(ResultsTable.prototype._showSuite):
(ResultsTable.prototype._showIteration):
(ResultsTable.prototype.showRecord):
(ResultsTable.prototype.showIterations):
(ResultsTable.prototype._showEmpty): Deleted.
* Animometer/runner/resources/animometer.js:
(window.benchmarkRunnerClient.didFinishLastIteration):
* Animometer/tests/resources/stage.js:
(StageBenchmark.prototype.showResults):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192494 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Add trivial lowering for B3's Div with doubles
benjamin@webkit.org [Mon, 16 Nov 2015 23:49:22 +0000 (23:49 +0000)]
[JSC] Add trivial lowering for B3's Div with doubles
https://bugs.webkit.org/show_bug.cgi?id=151292

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-16
Reviewed by Geoffrey Garen.

Filip had already made the constant propagation for it.
The Air Opcode was all we had left.

* b3/B3LowerToAir.cpp:
(JSC::B3::Air::LowerToAir::lower):
* b3/air/AirOpcode.opcodes:
* b3/testb3.cpp:
(JSC::B3::testDivArgDouble):
(JSC::B3::testDivArgsDouble):
(JSC::B3::testDivArgImmDouble):
(JSC::B3::testDivImmArgDouble):
(JSC::B3::testDivImmsDouble):
(JSC::B3::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192493 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[JSC] Speed up the coalescing-related operation of the iterated register allocator
benjamin@webkit.org [Mon, 16 Nov 2015 23:47:10 +0000 (23:47 +0000)]
[JSC] Speed up the coalescing-related operation of the iterated register allocator
https://bugs.webkit.org/show_bug.cgi?id=151290

Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-11-16
Reviewed by Geoffrey Garen.

One step closer to removing the Hash structures:

For the coalescing operation, we need to keep track of Move instructions. We do not strictly
need those to be the Air Move, just any abstract operation that copy a Tmp into another Tmp.

In this patch, I exploit that to remove the Hash structure related to the Inst. Instead of
using the Move Inst, we just keep track of the Use() and Def() of the instructions.
Those are added in the global list m_coalescingCandidates and the index in that list represent
the move for the remaining of the algorithm.

With Moves transformed into dense indices, we can start using arrays to make fast sets.

The m_activeMoves Set is easy since we only need simple add/remove/contains. It is transformed
into a BitVector.
The bit vector is always fully allocated to allow for quick uniform access. The assumtion is that
activeMoves will contains a few values for non trivial cases.

The worklist m_worklistMoves is more complicated. I want it to be ordered to coalesce moves starting
at the top of blocks. Having a fast remove() operation is also useful for mass coalescing.
It also needs Set operations, especially a fast contains().

For m_worklistMoves, I created a new structure: OrderedMoveSet.
It contains a list of ordered values, and a map of each value to its position in the list.

This resembles Briggs' Sparse Set but it is not exactly the same. When removing a value,
I set a special marker in the map (UINT_MAX). The reason is that I want contains() to be fast
instead of remove(). The marker in the map allows contains() with a single memory operation instead of two.

* b3/air/AirIteratedRegisterCoalescing.cpp:
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::build):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::allocate):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::forEachNodeMoves):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::isMoveRelated):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::enableMovesOnValue):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::coalesce):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::combine):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::freezeMoves):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::dumpWorkLists):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::addMove):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::isEmpty):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::contains):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::takeMove):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::takeLastMove):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::returnMove):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::OrderedMoveSet::clear):
(JSC::B3::Air::IteratedRegisterCoalescingAllocator::assignColors): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192492 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoClean referencing the options object in the graphics benchmark
commit-queue@webkit.org [Mon, 16 Nov 2015 23:22:16 +0000 (23:22 +0000)]
Clean referencing the options object in the graphics benchmark
https://bugs.webkit.org/show_bug.cgi?id=151284

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-11-16
Reviewed by Simon Fraser.

Get rid of the Benchmark.options member and rely only on the private member
Benchmark._options. The animator need to have its own options member instead
of accessing it from its reference to Benchmark object.

* Animometer/tests/bouncing-particles/resources/bouncing-canvas-particles.js:
(BouncingCanvasParticlesAnimator):
(BouncingCanvasParticlesBenchmark.prototype.createAnimator):
* Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
(BouncingParticle.prototype.animate):
(BouncingParticlesAnimator):
(BouncingParticlesBenchmark.prototype.createAnimator):
* Animometer/tests/examples/resources/canvas-electrons.js:
(CanvasElectronsStage.prototype.animate):
(CanvasElectronsAnimator):
(CanvasElectronsBenchmark.prototype.createAnimator):
(window.benchmarkClient.create):
* Animometer/tests/examples/resources/canvas-stars.js:
(CanvasStarsStage.prototype.animate):
(CanvasStarsAnimator):
(CanvasStarsBenchmark.prototype.createAnimator):
(window.benchmarkClient.create):
* Animometer/tests/resources/main.js:
(BenchmarkState.prototype.currentProgress):
(Animator):
(Animator.prototype.animate):
(Benchmark):
(Benchmark.prototype.update):
* Animometer/tests/resources/stage.js:
(Stage.prototype.clear):
(StageAnimator):
(StageBenchmark.prototype.createAnimator):
(StageBenchmark.prototype.tune):
(StageBenchmark.prototype.showResults):
* Animometer/tests/simple/resources/simple-canvas.js:
(SimpleCanvasStage.prototype.animate):
(SimpleCanvasAnimator):
(SimpleCanvasAnimator.prototype.animate):
(SimpleCanvasBenchmark.prototype.createAnimator):
* Animometer/tests/template/resources/template-canvas.js:
(TemplateCanvasStage.prototype.animate):
(TemplateCanvasBenchmark.prototype.createAnimator):
(window.benchmarkClient.create):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@192491 268f45cc-cd09-0410-ab3c-d52691b4dbfc