Context Navigation

← Previous Period
Next Period →

Timeline

May 23, 2018:

10:42 PM Changeset in webkit [232148] by commit-queue@webkit.org

2 edits in trunk/Source/WebCore

Use ASCIILiteral with applicationBundleIsEqualTo in RuntimeApplicationChecksCocoa
https://bugs.webkit.org/show_bug.cgi?id=185935

Patch by Joseph Pecoraro <Joseph Pecoraro> on 2018-05-23
Reviewed by Yusuke Suzuki.

platform/cocoa/RuntimeApplicationChecksCocoa.mm:

(WebCore::MacApplication::isSafari):
(WebCore::MacApplication::isAppleMail):
(WebCore::MacApplication::isIBooks):
(WebCore::MacApplication::isITunes):
(WebCore::MacApplication::isMicrosoftMessenger):
(WebCore::MacApplication::isAdobeInstaller):
(WebCore::MacApplication::isAOLInstantMessenger):
(WebCore::MacApplication::isMicrosoftMyDay):
(WebCore::MacApplication::isMicrosoftOutlook):
(WebCore::MacApplication::isQuickenEssentials):
(WebCore::MacApplication::isAperture):
(WebCore::MacApplication::isVersions):
(WebCore::MacApplication::isHRBlock):
(WebCore::MacApplication::isIAdProducer):
(WebCore::MacApplication::isSolidStateNetworksDownloader):
(WebCore::IOSApplication::isMobileMail):
(WebCore::IOSApplication::isMobileSafari):
(WebCore::IOSApplication::isWebBookmarksD):
(WebCore::IOSApplication::isDumpRenderTree):
(WebCore::IOSApplication::isMobileStore):
(WebCore::IOSApplication::isSpringBoard):
(WebCore::IOSApplication::isWebApp):
(WebCore::IOSApplication::isIBooks):
(WebCore::IOSApplication::isIBooksStorytime):
(WebCore::IOSApplication::isTheSecretSocietyHiddenMystery):
(WebCore::IOSApplication::isCardiogram):
(WebCore::IOSApplication::isNike):

10:23 PM Changeset in webkit [232147] by Brent Fulgham

11 edits in trunk/Source/WebCore

Avoid keeping FormState alive longer than necessary
https://bugs.webkit.org/show_bug.cgi?id=185877
<rdar://problem/39329219>

Reviewed by Ryosuke Niwa.

A number of crash fixes were done to prevent FormState objects from being
accessed after their relevant Frames had been destroyed. Unfortunately, this
could cause the FormState to persist after the owning Frame had been
destroyed, resulting in nullptr dereferences.

This patch does the following:

Uses WeakPtr's for FormState objects passed to completion handlers, rather than RefPtr, since those completion handlers might fire as part of the clean-up process during Frame destruction. This allows us to use the FormState if they are still valid, but gracefully handle cases where a form submission is cancelled in-flight.
Moves FormState object as they pass through the loader.
Removes some extraneous WTFMove() calls being made on bare FormState pointers.
Changes FormSubmission to hold a RefPtr so we can move the FormState to the loader in the code path that uses it (the FormSubmission is always destroyed shortly afterwards).
Changes the trap from Bug 183704 so that it only fires if the FormState object is being retained more than once.

loader/DocumentLoader.cpp:

(WebCore::DocumentLoader::willSendRequest): Update for new CompletionHandler
signature.

loader/FormState.cpp:

(WebCore::FormState::willDetachPage): Revise trap to check for retain counts
above one.

loader/FormState.h:

(WebCore::FormState::weakPtrFactory const): Added.

loader/FormSubmission.h:

(WebCore::FormSubmission::state const): Revised for change to RefPtr.
(WebCore::FormSubmission::takeState): Added.

loader/FrameLoader.cpp:

(WebCore::FrameLoader::urlSelected): Update for new CompletionHandler signature.
(WebCore::FrameLoader::loadURLIntoChildFrame): Ditto.
(WebCore::FrameLoader::loadFrameRequest): Ditto.
(WebCore::FrameLoader::loadURL): Ditto.
(WebCore::FrameLoader::load): Ditto.
(WebCore::FrameLoader::loadWithNavigationAction): Ditto.
(WebCore::FrameLoader::loadWithDocumentLoader): Ditto.
(WebCore::FrameLoader::reloadWithOverrideEncoding): Ditto.
(WebCore::FrameLoader::reload): Ditto.
(WebCore::FrameLoader::loadPostRequest): Ditto.
(WebCore::FrameLoader::loadDifferentDocumentItem): Ditto.

loader/FrameLoader.h:
loader/NavigationScheduler.cpp:
loader/PolicyChecker.cpp:

(WebCore::PolicyChecker::checkNavigationPolicy):Revise to use WeakPtr for
FormState passed to the completion handler. Remove some extraneous WTFMove()
calls on bare pointers.
(WebCore::PolicyChecker::checkNewWindowPolicy): Ditto.

loader/PolicyChecker.h:
page/ContextMenuController.cpp:

(WebCore::openNewWindow): Revise for new signatures.
(WebCore::ContextMenuController::contextMenuItemSelected): Ditto.

9:45 PM Changeset in webkit [232146] by bshafiei@apple.com

7 edits in branches/safari-606.1.17.40-branch/Source

Versioning.

9:32 PM Changeset in webkit [232145] by bshafiei@apple.com

1 copy in tags/Safari-606.1.17.40.3

Tag Safari-606.1.17.40.3.

9:08 PM Changeset in webkit [232144] by Chris Dumez

6 edits in trunk/LayoutTests

Regression(r229831): fast/loader/javascript-url-iframe-remove-on-navigate-async-delegate.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=183885
<rdar://problem/38733968>

Reviewed by Youenn Fablet.

Stop loading an external URL as this is not what the test is testing and since the error message for
the external load causes flakiness.

TestExpectations:
fast/loader/javascript-url-iframe-remove-on-navigate-async-delegate-expected.txt:
fast/loader/javascript-url-iframe-remove-on-navigate-async-delegate.html:
fast/loader/javascript-url-iframe-remove-on-navigate-expected.txt:
fast/loader/javascript-url-iframe-remove-on-navigate.html:

8:01 PM Changeset in webkit [232143] by keith_miller@apple.com

5 edits in trunk/Source

Expose $vm if window.internals is exposed
https://bugs.webkit.org/show_bug.cgi?id=185900

Reviewed by Mark Lam.

This is useful for testing vm internals when running LayoutTests.

Source/JavaScriptCore:

runtime/JSGlobalObject.cpp:

(JSC::JSGlobalObject::init):
(JSC::JSGlobalObject::visitChildren):
(JSC::JSGlobalObject::exposeDollarVM):

runtime/JSGlobalObject.h:

Source/WebCore:

testing/js/WebCoreTestSupport.cpp:

(WebCoreTestSupport::injectInternalsObject):

6:18 PM Changeset in webkit [232142] by Dewei Zhu

3 edits in trunk/Websites/perf.webkit.org

OSBuildFetcher should respect maxRevision while finding OS builds to report.
https://bugs.webkit.org/show_bug.cgi?id=185925

Reviewed by Ryosuke Niwa.

server-tests/tools-os-build-fetcher-tests.js: Fix a typo in the unit tests.

Added unit tests for this change. Aslo convert an existing test using async.

tools/js/os-build-fetcher.js:

(prototype._fetchAvailableBuilds): It should also use 'maxRevision' to filter builds to be reported.
It should use 'minRevisionOrder' when no commit has ever been submitted.
(prototype._commitsForAvailableBuilds): Takes 'maxOrder' as fifth argument.
'minOrder' and 'maxOrder' should be inclusive.

6:18 PM Changeset in webkit [232141] by Dewei Zhu

2 edits in trunk/Websites/perf.webkit.org

Update ChartPane per change r231087.
https://bugs.webkit.org/show_bug.cgi?id=185570

Reviewed by Ryosuke Niwa.

On r231087 we change the return type of AnalysisTask.create, ChartPane needs to adapt
this change.

public/v3/pages/chart-pane.js: Adapted the change that AnalysisTask.create now returns

an AnalysisTask object.
(ChartPane.prototype.async._analyzeRange):

5:53 PM Changeset in webkit [232140] by commit-queue@webkit.org

2 edits in trunk/Websites/webkit.org

Remove Source Sans Pro from WebKit website
https://bugs.webkit.org/show_bug.cgi?id=185871

Patch by Roy Reapor <rreapor@apple.com> on 2018-05-23
Reviewed by Aakash Jain.

demos/spring/index.html:

5:50 PM Changeset in webkit [232139] by ddkilzer@apple.com

5 edits in trunk/Source/WebCore

Don't create the SubimageCache just to clear an image from it
<https://webkit.org/b/185757>

Reviewed by Said Abou-Hallawa.

To fix this we make SubimageCacheWithTimer::clearImage() a
static class method that checks whether the cache exists before
removing it. We also make SubimageCacheWithTimer::getImage() a
static class method, and move more methods into the
SubimageCacheWithTimer class and make them private to reduce API
footprint.

platform/graphics/cg/GraphicsContextCG.cpp:

(WebCore::GraphicsContext::drawNativeImage): Switch to use new
SubimageCacheWithTimer::getSubimage() static class method.

platform/graphics/cg/NativeImageCG.cpp:

(WebCore::clearNativeImageSubimages): Switch to use new
SubimageCacheWithTimer::clearImage() static class method which
returns early if the subimage cache has not been created yet.
This fixes the bug.

platform/graphics/cg/SubimageCacheWithTimer.cpp:

(WebCore::SubimageCacheWithTimer::s_cache): Allocate space for
static class variable.
(WebCore::SubimageCacheWithTimer::getSubimage): Replace instance
method with new static class method that gets the subimage cache
singleton and calls the subimage() instance method.
(WebCore::SubimageCacheWithTimer::clearImage): Replace instance
methdod with new static class method that returns early if the
static cache singleton doesn't exist (fixes the bug), otherwise
calls the clearImageAndSubimages() instance method.
(WebCore::SubimageCacheWithTimer::subimage): Rename from
getSubimage(). Use auto after renaming SubimageCache typedef
to SubimageCacheHashSet.
(WebCore::SubimageCacheWithTimer::clearImageAndSubimages):
Rename from clearImage(). Modernize loops.
(WebCore::SubimageCacheWithTimer::subimageCache): Change
WebCore::subimageCache() to a static class method that creates
the subimage cache singleton if it doesn't exist yet, and
returns it.
(WebCore::SubimageCacheWithTimer::subimageCacheExists): Add.
Returns false if the subimage cache singleton has not been
created yet.

platform/graphics/cg/SubimageCacheWithTimer.h:
Rename typedef SubimageCache to SubimageCacheHashSet to avoid general confusion.

(WebCore::SubimageCacheWithTimer::getSubimage):
(WebCore::SubimageCacheWithTimer::clearImage):

Change to static class methods.

(WebCore::SubimageCacheWithTimer::SubimageCacheWithTimer):

Make private.

(WebCore::SubimageCacheWithTimer::subimage):

Rename from getSubimage() and make private.

(WebCore::SubimageCacheWithTimer::clearImageAndSubimages):

Rename from clearImage() and make private.

(WebCore::SubimageCacheWithTimer::subimageCache):

Rename from WebCore::subimageCache() and make a private static class method.

(WebCore::SubimageCacheWithTimer::subimageCacheExists):

Add private static class method.

(WebCore::SubimageCacheWithTimer::s_cache):

Declare private static variable to hold singleton.

5:47 PM Changeset in webkit [232138] by keith_miller@apple.com

3 edits
1 add in trunk

Define length on CoW array should properly convert to writable
https://bugs.webkit.org/show_bug.cgi?id=185927

Reviewed by Yusuke Suzuki.

JSTests:

stress/cow-define-length-as-value.js: Added.

(test):

Source/JavaScriptCore:

runtime/JSArray.cpp:

(JSC::JSArray::setLength):

5:05 PM Changeset in webkit [232137] by commit-queue@webkit.org

4 edits in trunk/Tools

Use Helvetica on build.webkit.org webpages
https://bugs.webkit.org/show_bug.cgi?id=185894

Patch by Roy Reapor <rreapor@apple.com> on 2018-05-23
Reviewed by Aakash Jain.

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html:

5:00 PM Changeset in webkit [232136] by eric.carlson@apple.com

23 edits in trunk/Source

Avoid loading AVFoundation to check supported MIME types if possible
https://bugs.webkit.org/show_bug.cgi?id=185839
<rdar://problem/40182010>

Reviewed by Jer Noble.
Source/WebCore:

Avoid loading AVFoundation to call +[AVURLAssetClass audiovisualMIMETypes] as long as possible,
and when they are loaded send the list to the UI process so it can pass it to all extant
and all new web processes so they can won't have call it at all.

WebCore.xcodeproj/project.pbxproj:
platform/graphics/ImageDecoder.cpp:

(WebCore::ImageDecoder::create): Don't call ImageDecoderAVFObjC::canDecodeType if
ImageDecoderCG can decode the type so we don't have to load AVFoundation.
(WebCore::ImageDecoder::supportsMediaType): Return as soon as a decoder class says
it supports a media type to avoid calling more than one. Call ImageDecoderAVFObjC last.

platform/graphics/avfoundation/objc/AVFoundationMIMETypeCache.h:

(WebCore::AVFoundationMIMETypeCache::setCacheMIMETypesCallback):

platform/graphics/avfoundation/objc/AVFoundationMIMETypeCache.mm:

(WebCore::AVFoundationMIMETypeCache::singleton): Simplify.
(WebCore::AVFoundationMIMETypeCache::setSupportedTypes): Cache the supplied list of types
so we won't have to load AVFoundation when asked for types later.
(WebCore::AVFoundationMIMETypeCache::types):
(WebCore::AVFoundationMIMETypeCache::supportsContentType): New convenience routine.
(WebCore::AVFoundationMIMETypeCache::canDecodeType): Ditto.
(WebCore::AVFoundationMIMETypeCache::isAvailable const): New, check to see if AVFoundation.framework
is available without actually loading it.
(WebCore::AVFoundationMIMETypeCache::loadMIMETypes): Load types if possible.
(WebCore::AVFoundationMIMETypeCache::AVFoundationMIMETypeCache): Deleted.
(WebCore::AVFoundationMIMETypeCache::loadTypes): Deleted.

platform/graphics/avfoundation/objc/ImageDecoderAVFObjC.mm:

(WebCore::ImageDecoderAVFObjC::create): Use AVFoundationMIMETypeCache::isAvailable instead
of loading the frameworks.
(WebCore::ImageDecoderAVFObjC::supportsMediaType): Ditto.
(WebCore::ImageDecoderAVFObjC::supportsContentType): Use AVFoundationMIMETypeCache::supportsContentType.
(WebCore::ImageDecoderAVFObjC::canDecodeType): Use AVFoundationMIMETypeCache::canDecodeType.

platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:

(WebCore::MediaPlayerPrivateAVFoundationObjC::registerMediaEngine): ASSERT if the
AVFoundationMIMETypeCache is empty, it shouldn't be possible to get here in that state.
(WebCore::MediaPlayerPrivateAVFoundationObjC::supportsType): Use AVFoundationMIMETypeCache::supportsContentType.
(WebCore::MediaPlayerPrivateAVFoundationObjC::supportsKeySystem): Use AVFoundationMIMETypeCache::canDecodeType.

platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:

(WebCore::MediaPlayerPrivateMediaSourceAVFObjC::registerMediaEngine): ASSERT if the
AVFoundationMIMETypeCache is empty, it shouldn't be possible to get here in that state.
(WebCore::MediaPlayerPrivateMediaSourceAVFObjC::supportsType): Use AVFoundationMIMETypeCache::canDecodeType.

platform/graphics/cg/ImageDecoderCG.cpp:

(WebCore::ImageDecoderCG::canDecodeType): New.

platform/graphics/cg/ImageDecoderCG.h:

Source/WebKit:

Shared/WebProcessCreationParameters.cpp:

(WebKit::WebProcessCreationParameters::encode const): Encode mediaMIMETypes.
(WebKit::WebProcessCreationParameters::decode): Decode mediaMIMETypes.

Shared/WebProcessCreationParameters.h:

UIProcess/Cocoa/WebProcessProxyCocoa.mm:

(WebKit::mediaTypeCache): Static Vector of media MIME types.
(WebKit::WebProcessProxy::cacheMediaMIMETypes): Cache the type list and pass it to every other
process proxy.
(WebKit::WebProcessProxy::cacheMediaMIMETypesInternal): Cache the type list and pass it to the
web process.
(WebKit::WebProcessProxy::mediaMIMETypes): Return the cached type list.

UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::initializeNewWebProcess): Set parameters.mediaMIMETypes.

UIProcess/WebProcessProxy.h:
UIProcess/WebProcessProxy.messages.in: Add CacheMediaMIMETypes.

WebProcess/WebProcess.h:
WebProcess/WebProcess.messages.in: Add SetMediaMIMETypes.

WebProcess/cocoa/WebProcessCocoa.mm:

(WebKit::WebProcess::platformInitializeWebProcess): Cache the MIME types if the list isn't
empty, else register with AVFoundationMIMETypeCache to be notified when it loads types.
AVFoundationMIMETypeCache to
(WebKit::WebProcess::platformTerminate): Unregister with AVFoundationMIMETypeCache.
(WebKit::WebProcess::setMediaMIMETypes): Pass list of types to AVFoundationMIMETypeCache.

Source/WTF:

wtf/cocoa/SoftLinking.h: Add SOFT_LINK_FRAMEWORK_OPTIONAL_PREFLIGHT.

4:41 PM Changeset in webkit [232135] by aakash_jain@apple.com

1 edit
6 adds in trunk/Tools

Add Buildbot configuration for Opensource EWS
https://bugs.webkit.org/show_bug.cgi?id=185484

Rubber-stamped by Alexey Proskuryakov.

BuildSlaveSupport/ews-build/config.json: Added, config file.
BuildSlaveSupport/ews-build/factories.py: Added, basic template for factories.
BuildSlaveSupport/ews-build/loadConfig.py: Added.

(loadBuilderConfig): Loads the configuration.
(checkValidWorker): Checks if the worker is valid.
(checkValidBuilder): Checks if the builder is valid.
(checkWorkersAndBuildersForConsistency): Check if workers and builders are consistent.
(checkWorkersAndBuildersForConsistency._find_worker_with_name):
(getBlackListedTags): Returns a list of keywords which should not be listed as tags.
(getValidTags): Returns a list of valid tags.
(getTagsForBuilder): Return a list of tags for given builder.

BuildSlaveSupport/ews-build/loadConfig_unittest.py: unit-tests for loadConfig.py.
BuildSlaveSupport/ews-build/master.cfg: Configuration for buildbot.

4:04 PM Changeset in webkit [232134] by keith_miller@apple.com

2 edits in trunk/Source/JavaScriptCore

InPlaceAbstractState should filter variables at the tail from a GetLocal by their flush format
https://bugs.webkit.org/show_bug.cgi?id=185923

Reviewed by Saam Barati.

Previously, we could confuse AI by overly broadening a type. This happens when a block in a
loop has a local mutated following a GetLocal but never SetLocaled to the stack. For example,

Block 1:
@1: GetLocal(loc42, FlushedInt32);
@2: PutStructure(Check: Cell: @1);
@3: Jump(Block 1);

Would cause us to claim that loc42 could be either an int32 or a some cell. However,
the type of an local cannot change without writing to it.

This fixes a crash in destructuring-rest-element.js

dfg/DFGInPlaceAbstractState.cpp:

(JSC::DFG::InPlaceAbstractState::endBasicBlock):

3:55 PM Changeset in webkit [232133] by BJ Burg

2 edits in trunk/Source/WebKit

Web Automation: disable process swap on navigation when an automation session is active
https://bugs.webkit.org/show_bug.cgi?id=185552

Reviewed by Tim Horton.

UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::processForNavigationInternal):
This is not intended to work right now. Opt out to avoid crashing
later when a process is deallocated unexpectedly.

3:34 PM Changeset in webkit [232132] by fpizlo@apple.com

14 edits
2 adds in trunk/Source

Speed up JetStream/base64
https://bugs.webkit.org/show_bug.cgi?id=185914

Reviewed by Michael Saboff.

Source/JavaScriptCore:

Make allocation fast paths ALWAYS_INLINE.

This is a 1% speed-up on SunSpider, mostly because of base64. It also speeds up pdfjs by
~6%.

CMakeLists.txt:
JavaScriptCore.xcodeproj/project.pbxproj:
heap/AllocatorInlines.h:

(JSC::Allocator::allocate const):

heap/CompleteSubspace.cpp:

(JSC::CompleteSubspace::allocateNonVirtual): Deleted.

heap/CompleteSubspace.h:
heap/CompleteSubspaceInlines.h: Added.

(JSC::CompleteSubspace::allocateNonVirtual):

heap/FreeListInlines.h:

(JSC::FreeList::allocate):

heap/IsoSubspace.cpp:

(JSC::IsoSubspace::allocateNonVirtual): Deleted.

heap/IsoSubspace.h:

(JSC::IsoSubspace::allocatorForNonVirtual):

heap/IsoSubspaceInlines.h: Added.

(JSC::IsoSubspace::allocateNonVirtual):

runtime/JSCellInlines.h:
runtime/VM.h:

Source/WTF:

Make Vector<>::append ALWAYS_INLINE.

wtf/Vector.h:

(WTF::Vector::append):
(WTF::minCapacity>::expandCapacity):
(WTF::minCapacity>::append):
(WTF::minCapacity>::tryAppend):

3:23 PM Changeset in webkit [232131] by bshafiei@apple.com

7 edits in branches/safari-606.1.17.20-branch/Source

Versioning.

3:21 PM Changeset in webkit [232130] by bshafiei@apple.com

7 edits in branches/safari-606.1.17.30-branch/Source

Versioning.

3:20 PM Changeset in webkit [232129] by bshafiei@apple.com

7 edits in branches/safari-606.1.17.40-branch/Source

Versioning.

3:19 PM Changeset in webkit [232128] by bshafiei@apple.com

1 copy in tags/Safari-606.1.17.40.2

Tag Safari-606.1.17.40.2.

3:18 PM Changeset in webkit [232127] by bshafiei@apple.com

1 copy in tags/Safari-606.1.17.30.2

Tag Safari-606.1.17.30.2.

3:17 PM Changeset in webkit [232126] by bshafiei@apple.com

1 copy in tags/Safari-606.1.17.20.1

Tag Safari-606.1.17.20.1.

3:03 PM Changeset in webkit [232125] by bshafiei@apple.com

7 edits in trunk/Source

Versioning.

3:01 PM Changeset in webkit [232124] by bshafiei@apple.com

1 copy in tags/Safari-606.1.18

Tag Safari-606.1.18.

1:22 PM Changeset in webkit [232123] by Chris Dumez

15 edits
3 adds in trunk

RenderLayer::scrollRectToVisible() should not propagate a subframe's scroll to its cross-origin parent
https://bugs.webkit.org/show_bug.cgi?id=185664
<rdar://problem/36185260>

Reviewed by Simon Fraser.

Source/WebCore:

RenderLayer::scrollRectToVisible() should not propagate a subframe's scroll to its
cross-origin parent. There was logic in FrameLoader::scrollToFragmentWithParentBoundary()
to temporarily set the 'safeToPropagateScrollToParent' flag to false on the cross-origin
ancestor frame during the call to FrameView::scrollToFragment(). This would correctly
prevent RenderLayer::scrollRectToVisible() to propagate the scroll to the cross-origin
ancestor frame when scrollRectToVisible() is called synchronously. However,
scrollRectToVisible() can get called asynchronously in case of a dirty layout, as part
of the post layout tasks.

To address the issue, we get rid of the safeToPropagateScrollToParent flag on FrameView
and instead update FrameView::safeToPropagateScrollToParent() to do the cross-origin
check. FrameView::safeToPropagateScrollToParent() is called by RenderLayer::scrollRectToVisible()
and this is a lot more robust than relying on a flag which gets temporarily set.

Test: http/tests/navigation/fragment-navigation-cross-origin-subframe-no-scrolling-parent.html

dom/Document.cpp:
dom/Document.h:
loader/FrameLoader.cpp:

(WebCore::FrameLoader::scrollToFragmentWithParentBoundary):

page/FrameView.cpp:

(WebCore::FrameView::FrameView):
(WebCore::FrameView::reset):
(WebCore::FrameView::safeToPropagateScrollToParent const):

page/FrameView.h:

LayoutTests:

Add layout test coverage.

http/tests/navigation/fragment-navigation-cross-origin-subframe-no-scrolling-parent-expected.txt: Added.
http/tests/navigation/fragment-navigation-cross-origin-subframe-no-scrolling-parent.html: Added.
http/tests/navigation/resources/clear-fragment.html: Added.

1:02 PM Changeset in webkit [232122] by msaboff@apple.com

3 edits
1 add in trunk

Date.parse() doesn't properly handle input outside of ES Spec limits
https://bugs.webkit.org/show_bug.cgi?id=185868

Reviewed by Mark Lam.

JSTests:

New test.

stress/date-parse-ranges.js: Added.

(shouldBe):
(throw.new.Error):
(shouldBeNaN):

Source/WTF:

Clamped date creation to +/-100,000,000 days relative to midnight at the beginning
of 01 January, 1970 UTC as per ecma262/#sec-time-values-and-time-range and
ecma262/#sec-date-time-string-format.

wtf/DateMath.cpp:

(WTF::ymdhmsToSeconds):
(WTF::parseES5DateFromNullTerminatedCharacters):

12:09 PM Changeset in webkit [232121] by youenn@apple.com

15 edits
1 delete in trunk

NetworkLoadChecker should check cached redirections
https://bugs.webkit.org/show_bug.cgi?id=185849

Reviewed by Chris Dumez.

LayoutTests/imported/w3c:

web-platform-tests/service-workers/service-worker/redirected-response.https-expected.txt:
web-platform-tests/service-workers/service-worker/registration-security-error.https-expected.txt:

Source/WebCore:

Covered by rebased tests.

loader/SubresourceLoader.cpp:

(WebCore::SubresourceLoader::willSendRequestInternal):

Log the case of a redirection with fetch error mode.

Source/WebKit:

NetworkProcess/NetworkLoadChecker.cpp:

(WebKit::NetworkLoadChecker::checkRedirection):
Set the resource error url as done by WebCore SubresourceLoader.

NetworkProcess/NetworkResourceLoader.cpp:

(WebKit::NetworkResourceLoader::retrieveCacheEntry):
Pass the resource request to dispatchWillSendRedirectedRequest now needs it.
(WebKit::NetworkResourceLoader::willSendRedirectedRequest):
Make sure that m_networkLoad is not null before cancelling it since we might be checking a cached redirection.
(WebKit::NetworkResourceLoader::continueWillSendRedirectedRequest):
Ensure the redirect response is coming from the Network before adding it to the cache.
(WebKit::NetworkResourceLoader::dispatchWillSendRequestForCacheEntry):
Call willSendRedirectedRequest to make sure the cached redirect is validated.

NetworkProcess/NetworkResourceLoader.h:

LayoutTests:

TestExpectations:
http/tests/fetch/redirectmode-and-preload-expected.txt:
http/tests/fetch/redirectmode-and-preload.html:

Removed tests that mix manual/error redirect mode with no-cors since this is no longer a valid possibility.

http/tests/xmlhttprequest/access-control-and-redirects-async-expected.txt:
http/tests/xmlhttprequest/access-control-and-redirects-expected.txt:
platform/mac-wk1/http/tests/xmlhttprequest/access-control-and-redirects-async-expected.txt: Removed.

11:38 AM Changeset in webkit [232120] by n_wang@apple.com

5 edits in trunk

AX: setValue on contenteditable should preserve whitespace
https://bugs.webkit.org/show_bug.cgi?id=185897

Reviewed by Chris Fleizach.

Source/WebCore:

RenderText is using its parent renderer's style to determine if
whitespace collapsing is necessary. So when setting the innerText
of the element in setValue, let's also set its style in order to
preserve whitespaces.

Modified an existing test to cover this change.

accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::setValue):

LayoutTests:

accessibility/mac/set-value-editable-types-expected.txt:
accessibility/mac/set-value-editable-types.html:

10:54 AM Changeset in webkit [232119] by Michael Catanzaro

2 edits in trunk/Source/WTF

Prohibit shrinking the FastBitVector
https://bugs.webkit.org/show_bug.cgi?id=181020

Reviewed by Oliver Hunt.

Prohibit shrinking the FastBitVector. It's not prepared for this and the current usage does
not require it.

wtf/FastBitVector.cpp:

(WTF::FastBitVectorWordOwner::resizeSlow):

10:50 AM Changeset in webkit [232118] by Michael Catanzaro

6 edits in trunk

[GTK] Silence GCC 8 warnings
https://bugs.webkit.org/show_bug.cgi?id=185556

Reviewed by Žan Doberšek.

Source/WebCore:

Silence remaining -Wcast-function-type warnings by casting through void (*)(void), aka
GCallback.

accessibility/atk/WebKitAccessibleHyperlink.cpp:

(webkitAccessibleHyperlinkSetProperty):
(webkitAccessibleHyperlinkGetType):

accessibility/atk/WebKitAccessibleWrapperAtk.cpp:

(webkitAccessibleGetType):

platform/network/soup/NetworkStorageSessionSoup.cpp:

(WebCore::NetworkStorageSession::getCredentialFromPersistentStorage):

Tools:

GCC is complaining about this strncpy. It's guaranteed to be safe, because the length of
source and destination are always identical. But switching to snprintf avoids the warning.

DumpRenderTree/TestNetscapePlugIn/Tests/FormValue.cpp:

(FormValue::NPP_GetValue):

10:39 AM Changeset in webkit [232117] by Antti Koivisto

5 edits in trunk

Increase the simulated memory size on PLATFORM(IOS_SIMULATOR) from 512MB to 1024MB
https://bugs.webkit.org/show_bug.cgi?id=185908

Reviewed by Geoffrey Garen.

Source/bmalloc:

We don't support 512MB devices anymore. This will make the simulator behave more
like a real device.

bmalloc/AvailableMemory.cpp:

(bmalloc::memorySizeAccordingToKernel):

Factor to a function.
Don't use availableMemoryGuess for the simulator value as it is not a guess.

(bmalloc::computeAvailableMemory):

Apply the same adjustments to the simulated value too.

LayoutTests:

platform/ios-simulator/fast/canvas/canvas-crash-expected.txt:
platform/ios-simulator/fast/canvas/canvas-skia-excessive-size-expected.txt:

10:31 AM Changeset in webkit [232116] by aboya@igalia.com

3 edits
1 add in trunk/LayoutTests

[GTK] Unreviewed test gardening
https://bugs.webkit.org/show_bug.cgi?id=185916

platform/gtk/TestExpectations:
platform/gtk/accessibility/table-roles-hierarchy-expected.txt:

9:59 AM Changeset in webkit [232115] by Ryan Haddad

2 edits
1 delete in trunk/Tools

Unreviewed, rolling out r232112.

The tests added with this change are failing on the bots.

Reverted changeset:

"test262/Runner.pm: add unit tests"
https://bugs.webkit.org/show_bug.cgi?id=185783
https://trac.webkit.org/changeset/232112

9:56 AM Changeset in webkit [232114] by Alan Bujtas

8 edits
1 copy
1 add in trunk/Source/WebCore

[LFC] Move sizing/positioning logic to helper classes
https://bugs.webkit.org/show_bug.cgi?id=185898

Reviewed by Antti Koivisto.

The idea here is to move all the sizing and positioning logic to helper classes so that
the formatting context code stays lean.
This is similar to the dedicated BlockMarginCollapse class for the collapsing logic.
The helper classes have only static functions. These static functions do not mutate the associated DisplayBoxes,
but instead they simply retun the computed values.

Sources.txt:
WebCore.xcodeproj/project.pbxproj:
layout/FormattingContext.cpp:

(WebCore::Layout::FormattingContext::computeOutOfFlowPosition const):
(WebCore::Layout::FormattingContext::computeOutOfFlowWidth const):
(WebCore::Layout::FormattingContext::computeFloatingWidth const):
(WebCore::Layout::FormattingContext::computeOutOfFlowHeight const):
(WebCore::Layout::FormattingContext::computeFloatingHeight const):
(WebCore::Layout::FormattingContext::computeOutOfFlowNonReplacedHeight const): Deleted.
(WebCore::Layout::FormattingContext::computeFloatingNonReplacedHeight const): Deleted.
(WebCore::Layout::FormattingContext::computeReplacedHeight const): Deleted.
(WebCore::Layout::FormattingContext::computeReplacedWidth const): Deleted.
(WebCore::Layout::FormattingContext::contentHeightForFormattingContextRoot const): Deleted.
(WebCore::Layout::FormattingContext::computeFloatingNonReplacedWidth const): Deleted.
(WebCore::Layout::FormattingContext::computeOutOfFlowNonReplacedWidth const): Deleted.
(WebCore::Layout::FormattingContext::computeOutOfFlowReplacedHeight const): Deleted.
(WebCore::Layout::FormattingContext::computeOutOfFlowReplacedWidth const): Deleted.
(WebCore::Layout::FormattingContext::computeOutOfFlowNonReplacedPosition const): Deleted.
(WebCore::Layout::FormattingContext::computeOutOfFlowReplacedPosition const): Deleted.
(WebCore::Layout::FormattingContext::shrinkToFitWidth const): Deleted.

layout/FormattingContext.h:
layout/FormattingContextGeometry.cpp: Copied from Source/WebCore/layout/FormattingContext.cpp.

(WebCore::Layout::contentHeightForFormattingContextRoot):
(WebCore::Layout::shrinkToFitWidth):
(WebCore::Layout::FormattingContextGeometry::outOfFlowNonReplacedHeight):
(WebCore::Layout::FormattingContextGeometry::outOfFlowNonReplacedWidth):
(WebCore::Layout::FormattingContextGeometry::outOfFlowReplacedHeight):
(WebCore::Layout::FormattingContextGeometry::outOfFlowReplacedWidth):
(WebCore::Layout::FormattingContextGeometry::floatingNonReplacedHeight):
(WebCore::Layout::FormattingContextGeometry::floatingNonReplacedWidth):
(WebCore::Layout::FormattingContextGeometry::floatingReplacedHeight):
(WebCore::Layout::FormattingContextGeometry::floatingReplacedWidth):
(WebCore::Layout::FormattingContextGeometry::outOfFlowNonReplacedPosition):
(WebCore::Layout::FormattingContextGeometry::outOfFlowReplacedPosition):
(WebCore::Layout::FormattingContextGeometry::replacedHeight):
(WebCore::Layout::FormattingContextGeometry::replacedWidth):

layout/FormattingContextGeometry.h: Copied from Source/WebCore/layout/blockformatting/BlockFormattingContext.h.
layout/blockformatting/BlockFormattingContext.cpp:

(WebCore::Layout::BlockFormattingContext::computeStaticPosition const):
(WebCore::Layout::BlockFormattingContext::computeInFlowHeight const):
(WebCore::Layout::BlockFormattingContext::computeInFlowWidth const):
(WebCore::Layout::BlockFormattingContext::computeInFlowNonReplacedWidth const): Deleted.
(WebCore::Layout::BlockFormattingContext::computeInFlowNonReplacedHeight const): Deleted.

layout/blockformatting/BlockFormattingContext.h:
layout/blockformatting/BlockFormattingContextGeometry.cpp: Added.

(WebCore::Layout::BlockFormattingContextGeometry::inFlowNonReplacedHeight):
(WebCore::Layout::BlockFormattingContextGeometry::inFlowNonReplacedWidth):
(WebCore::Layout::BlockFormattingContextGeometry::inFlowReplacedHeight):
(WebCore::Layout::BlockFormattingContextGeometry::inFlowReplacedWidth):
(WebCore::Layout::BlockFormattingContextGeometry::staticPosition):

layout/blockformatting/BlockFormattingContextGeometry.h: Copied from Source/WebCore/layout/blockformatting/BlockFormattingContext.h.
layout/displaytree/DisplayBox.h:

9:37 AM WebKitGTK/Gardening/Calendar edited by Michael Catanzaro

cgarcia trades days with csaavedra (diff)

9:09 AM Changeset in webkit [232113] by Antti Koivisto

4 edits
1 add in trunk

Page keeps reloading when viewing photos in google drive (due to too high canvas memory limits)
https://bugs.webkit.org/show_bug.cgi?id=185903
<rdar://problem/38420562>

Reviewed by Simon Fraser.

Source/WebCore:

The canvas memory usage limits don't work on iOS since the current 2GB minimum limit is
larger than the maximum process size.

html/HTMLCanvasElement.cpp:

(WebCore::maxActivePixelMemory):

Always base this on the reported ramSize() on iOS. Make it still fairly large to not risk breaking
any currently working content. In practice the limit computes to 448MB on device at the moment.

LayoutTests:

platform/ios-simulator/fast/canvas/canvas-crash-expected.txt:
platform/ios-simulator/fast/canvas/canvas-skia-excessive-size-expected.txt: Added.

8:33 AM Changeset in webkit [232112] by commit-queue@webkit.org

2 edits
13 adds in trunk/Tools

test262/Runner.pm: add unit tests
https://bugs.webkit.org/show_bug.cgi?id=185783

Patch by Valerie R Young <valerie@bocoup.com> on 2018-05-23
Reviewed by Aakash Jain.

Add unit tests to be run by test-webkitperl

Scripts/test262/Runner.pm:

(main):

If harness files cannot be found in a custom, supplies test262 path, then sure the webkit test262 harness files.
Lazily create the test262-results directory.
If an expectation file is supplied via command line, save to the supplied location instead of the detault webkit location.

(compileTest):
(processResult):
(getHarness):

Scripts/webkitperl/test262_unittest/README: Added.
Scripts/webkitperl/test262_unittest/fixtures/expectations-compare.yaml: Added.
Scripts/webkitperl/test262_unittest/fixtures/expectations.yaml: Added.
Scripts/webkitperl/test262_unittest/fixtures/test/expected-to-fail-now-failing-with-new-error.js: Added.
Scripts/webkitperl/test262_unittest/fixtures/test/expected-to-fail-now-failing.js: Added.
Scripts/webkitperl/test262_unittest/fixtures/test/expected-to-fail-now-passing.js: Added.

(f):

Scripts/webkitperl/test262_unittest/fixtures/test/expected-to-pass-now-failing.js: Added.
Scripts/webkitperl/test262_unittest/fixtures/test/fail.js: Added.
Scripts/webkitperl/test262_unittest/fixtures/test/pass.js: Added.

(f):

Scripts/webkitperl/test262_unittest/test262-runner-tests.pl: Added.

8:24 AM Changeset in webkit [232111] by Carlos Garcia Campos

6 edits in trunk/Source

[GTK] WebDriver: implement AutomationSessionClient::didDisconnectFromRemote
https://bugs.webkit.org/show_bug.cgi?id=185866

Reviewed by Brian Burg.

Source/WebDriver:

Close the dbus connection when receiving an empty target list.

glib/SessionHostGlib.cpp:

(WebDriver::SessionHost::setTargetList):

Source/WebKit:

To handle the case of the session being closed by the browser, for example in case of a network process
crash. This is currently causing WebDriver tests to timeout in the bot.

UIProcess/API/glib/WebKitAutomationSession.cpp: Add an implementation of didDisconnectFromRemote() to notify

the WebContext that the session will be closed.

UIProcess/API/glib/WebKitWebContext.cpp: Remove the automation session when closed.
UIProcess/API/glib/WebKitWebContextPrivate.h:

8:18 AM Changeset in webkit [232110] by commit-queue@webkit.org

4 edits
1 add in trunk

Conversion misspelled "Convertion" in error message string
https://bugs.webkit.org/show_bug.cgi?id=185436

Patch by Rick Waldron <waldron.rick@gmail.com> on 2018-05-23
JSTests:

Reviewed by Saam Barati, Michael Saboff.

bigIntTests.yaml:

Source/JavaScriptCore:

Reviewed by Saam Barati, Michael Saboff

runtime/JSBigInt.cpp:

(JSC::JSBigInt::toNumber const):

7:50 AM WebKitGTK/Gardening/Calendar edited by aboya@igalia.com

(diff)

7:46 AM Changeset in webkit [232109] by Alan Bujtas

2 edits in trunk/Source/WebCore

[LFC] Implement positioning for replaced out-of-flow elements
https://bugs.webkit.org/show_bug.cgi?id=185902

Reviewed by Antti Koivisto.

layout/FormattingContext.cpp:

(WebCore::Layout::FormattingContext::computeOutOfFlowReplacedPosition const):

5:37 AM Changeset in webkit [232108] by Yusuke Suzuki

2 edits in trunk/JSTests

Unreviewed, skip test if memoryLimited is specified
https://bugs.webkit.org/show_bug.cgi?id=185888

stress/regress-185888.js:

May 22, 2018:

11:49 PM Changeset in webkit [232107] by Manuel Rego Casasnovas

5 edits in trunk/LayoutTests

[css-text] W3C test suite gardening
https://bugs.webkit.org/show_bug.cgi?id=185879

Some more gardening on the css-text test suite.

Unreviewed test gardening.

TestExpectations:
platform/gtk/TestExpectations:
platform/ios/TestExpectations:
platform/mac/TestExpectations:

10:05 PM Changeset in webkit [232106] by Yusuke Suzuki

6 edits in trunk/Source/JavaScriptCore

[JSC] Clean up stringGetByValStubGenerator
https://bugs.webkit.org/show_bug.cgi?id=185864

Reviewed by Saam Barati.

We clean up stringGetByValStubGenerator.

Unify 32bit and 64bit implementations.
Rename stringGetByValStubGenerator to stringGetByValGenerator, move it to ThunkGenerators.cpp.
Remove string type check since this code is invoked only when we know regT0 is JSString*.
Do not tag Cell in stringGetByValGenerator side. 32bit code stores Cell with tag in JITPropertyAccess32_64 side.
Fix invalid use of loadPtr for StringImpl::flags. Should use load32.

jit/JIT.h:
jit/JITPropertyAccess.cpp:

(JSC::JIT::emitSlow_op_get_by_val):
(JSC::JIT::stringGetByValStubGenerator): Deleted.

jit/JITPropertyAccess32_64.cpp:

(JSC::JIT::emit_op_get_by_val):
(JSC::JIT::emitSlow_op_get_by_val):
(JSC::JIT::stringGetByValStubGenerator): Deleted.

jit/ThunkGenerators.cpp:

(JSC::stringGetByValGenerator):

jit/ThunkGenerators.h:

9:44 PM Changeset in webkit [232105] by Yusuke Suzuki

10 edits in trunk/Source/JavaScriptCore

[JSC] Use branchIfString/branchIfNotString instead of structure checkings
https://bugs.webkit.org/show_bug.cgi?id=185810

Reviewed by Saam Barati.

Let's use branchIfString/branchIfNotString helper functions instead of
checking structure with jsString's structure. It's easy to read. And
it emits less code since we do not need to embed string structure's
raw pointer in 32bit environment.

jit/JIT.h:
jit/JITInlines.h:

(JSC::JIT::emitLoadCharacterString):
(JSC::JIT::checkStructure): Deleted.

jit/JITOpcodes32_64.cpp:

(JSC::JIT::emitSlow_op_eq):
(JSC::JIT::compileOpEqJumpSlow):
(JSC::JIT::emitSlow_op_neq):

jit/JITPropertyAccess.cpp:

(JSC::JIT::stringGetByValStubGenerator):
(JSC::JIT::emitSlow_op_get_by_val):
(JSC::JIT::emitByValIdentifierCheck):

jit/JITPropertyAccess32_64.cpp:

(JSC::JIT::stringGetByValStubGenerator):
(JSC::JIT::emitSlow_op_get_by_val):

jit/JSInterfaceJIT.h:

(JSC::ThunkHelpers::jsStringLengthOffset): Deleted.
(JSC::ThunkHelpers::jsStringValueOffset): Deleted.

jit/SpecializedThunkJIT.h:

(JSC::SpecializedThunkJIT::loadJSStringArgument):

jit/ThunkGenerators.cpp:

(JSC::stringCharLoad):
(JSC::charCodeAtThunkGenerator):
(JSC::charAtThunkGenerator):

runtime/JSString.h:

9:04 PM Changeset in webkit [232104] by bshafiei@apple.com

17 edits in branches/safari-606.1.17.20-branch/Source

Cherry-pick r232090. rdar://problem/40306056

Optimized path zoom animation needs a valid UIImage and CGRect
https://bugs.webkit.org/show_bug.cgi?id=185883
<rdar://problem/40306056>

Reviewed by Jon Lee.

Source/WebCore:

Pass the bounding box of the element that was clicked onto
the UI process, so it can perform an animation from that spot.

This involved adding an IntRect to the ResourceRequest, and passing
that info into it from the HTMLAnchorElement, using a new struct
called SystemPreviewInfo.

html/HTMLAnchorElement.cpp: (WebCore::HTMLAnchorElement::handleClick):
loader/FrameLoadRequest.cpp: (WebCore::FrameLoadRequest::FrameLoadRequest):
loader/FrameLoadRequest.h: (WebCore::FrameLoadRequest::FrameLoadRequest): (WebCore::FrameLoadRequest::isSystemPreview const): (WebCore::FrameLoadRequest::systemPreviewRect const):
loader/FrameLoader.cpp: (WebCore::FrameLoader::urlSelected): (WebCore::FrameLoader::loadURL):
loader/FrameLoader.h: (WebCore::FrameLoader::urlSelected):
loader/FrameLoaderTypes.h:
platform/network/ResourceRequestBase.cpp: (WebCore::ResourceRequestBase::systemPreviewRect const): (WebCore::ResourceRequestBase::setSystemPreviewRect):
platform/network/ResourceRequestBase.h:

Source/WebKit:

Take the rectangle that was passed into the ResourceRequest and
use it for the origin of an animation into QuickLook.

Shared/WebCoreArgumentCoders.cpp: (IPC::ArgumentCoder<ResourceRequest>::encode): (IPC::ArgumentCoder<ResourceRequest>::decode):
UIProcess/Cocoa/DownloadClient.mm: (WebKit::DownloadClient::didStart):
UIProcess/Cocoa/SystemPreviewControllerCocoa.mm: (-[_WKPreviewControllerDelegate initWithSystemPreviewController:fromRect:]): (-[_WKPreviewControllerDelegate presentingViewController]): (-[_WKPreviewControllerDelegate previewController:frameForPreviewItem:inSourceView:]): (-[_WKPreviewControllerDelegate previewController:transitionImageForPreviewItem:contentRect:]): (WebKit::SystemPreviewController::start): (-[_WKPreviewControllerDelegate initWithSystemPreviewController:]): Deleted.
UIProcess/Downloads/DownloadProxy.h: (WebKit::DownloadProxy::systemPreviewDownloadRect const):
UIProcess/SystemPreviewController.h:
UIProcess/WebPageProxy.cpp: (WebKit::WebPageProxy::syncRootViewToScreen):
UIProcess/WebPageProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@232090 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8:47 PM Changeset in webkit [232103] by mmaxfield@apple.com

3 edits
2 adds in trunk

Text can wrap between hyphens and commas
https://bugs.webkit.org/show_bug.cgi?id=185899
<rdar://problem/40118983>

Reviewed by Zalan Bujtas.

Source/WebCore:

Chrome, Firefox, and ICU all agree that text shouldn't wrap there.

Test: fast/text/line-break-hyphen-comma.html

rendering/BreakLines.cpp:

LayoutTests:

fast/text/line-break-hyphen-comma-expected.html: Added.
fast/text/line-break-hyphen-comma.html: Added.

8:19 PM Changeset in webkit [232102] by bshafiei@apple.com

7 edits in branches/safari-606.1.17.40-branch/Source

Versioning.

8:17 PM Changeset in webkit [232101] by bshafiei@apple.com

7 edits in branches/safari-606.1.17.30-branch/Source

Versioning.

8:11 PM Changeset in webkit [232100] by mark.lam@apple.com

3 edits
1 add in trunk

BytecodeGeneratorification shouldn't add a ValueProfile if the JIT is disabled.
https://bugs.webkit.org/show_bug.cgi?id=185896
<rdar://problem/40471403>

Reviewed by Saam Barati.

JSTests:

stress/regress-185896.js: Added.

Source/JavaScriptCore:

bytecode/BytecodeGeneratorification.cpp:

(JSC::BytecodeGeneratorification::run):

8:06 PM Changeset in webkit [232099] by bshafiei@apple.com

7 edits in branches/safari-605-branch/Source

Versioning.

7:36 PM Changeset in webkit [232098] by Simon Fraser

10 edits in trunk/Source/WebCore

Add notifyutil callbacks to dump the memory cache, and the list of live Pages and Documents
https://bugs.webkit.org/show_bug.cgi?id=185853

Reviewed by Keith Miller.
Source/WebCore:

Add a notifyutil callback to dump the PageCache, which dumps the stats, and the list of live pages.
This can be invoked on iOS and macOS via "notifyutil -p com.apple.WebKit.showPageCache".

Add a notifyutil callback that dumps the list of all Pages, and the list of all Documents, with
Document pointer address and URL.
This can be invoked on iOS and macOS via "notifyutil -p com.apple.WebKit.showAllDocuments".

history/PageCache.cpp:

(WebCore::PageCache::PageCache):
(WebCore::PageCache::dump const):

history/PageCache.h:
loader/cache/CachedResource.h:

(WebCore::CachedResource::numberOfClients const):
(WebCore::CachedResource::count const): Deleted.

loader/cache/MemoryCache.cpp:

(WebCore::MemoryCache::MemoryCache):
(WebCore::MemoryCache::pruneLiveResourcesToSize):
(WebCore::MemoryCache::dumpStats):
(WebCore::MemoryCache::dumpLRULists const):

loader/cache/MemoryCache.h:
page/mac/PageMac.mm:

(WebCore::Page::platformInitialize):

Source/WebCore/PAL:

Compile registerNotifyCallback in release builds.

pal/Logging.cpp:

(PAL::registerNotifyCallback):

pal/Logging.h:

7:23 PM Changeset in webkit [232097] by Brent Fulgham

4 edits in trunk/Source/WebKit

Close access to "lsopen" for non-UI process
https://bugs.webkit.org/show_bug.cgi?id=185890
<rdar://problem/39686511>

Reviewed by Alexey Proskuryakov.

Close down access to 'lsopen' in the iOS sandboxes. These operations are
performed by the UIProcess on behalf of these helper processes.

Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
Resources/SandboxProfiles/ios/com.apple.WebKit.Storage.sb:
Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:

7:15 PM Changeset in webkit [232096] by bshafiei@apple.com

1 copy in tags/Safari-605.3.4

Tag Safari-605.3.4.

7:00 PM Changeset in webkit [232095] by bshafiei@apple.com

7 edits in branches/safari-605-branch/Source

Versioning.

6:45 PM Changeset in webkit [232094] by dino@apple.com

2 edits in trunk/Source/WebCore

Optimized path zoom animation needs a valid UIImage and CGRect
https://bugs.webkit.org/show_bug.cgi?id=185883
<rdar://problem/40306056>

Build fix.

loader/FrameLoadRequest.h:

5:17 PM Changeset in webkit [232093] by commit-queue@webkit.org

7 edits in trunk/Source/WebCore

Unreviewed, rolling out r232081.
https://bugs.webkit.org/show_bug.cgi?id=185895

Broke two API tests (Requested by bfulgham_ on #webkit).

Reverted changeset:

"Avoid keeping FormState alive longer than necessary"
https://bugs.webkit.org/show_bug.cgi?id=185877
https://trac.webkit.org/changeset/232081

5:16 PM Changeset in webkit [232092] by Yusuke Suzuki

4 edits in trunk

[JSC] Fix CachedCall's argument count if RegExp has named captures
https://bugs.webkit.org/show_bug.cgi?id=185587

Reviewed by Mark Lam.

JSTests:

test262/expectations.yaml:

Source/JavaScriptCore:

If the given RegExp has named captures, the argument count of CachedCall in String#replace
should be increased by one. This causes crash with assertion in test262. This patch corrects
the argument count.

This patch also unifies source.is8Bit()/!source.is8Bit() code since they are now completely
the same.

runtime/StringPrototype.cpp:

(JSC::replaceUsingRegExpSearch):

5:12 PM Changeset in webkit [232091] by commit-queue@webkit.org

2 edits in trunk/LayoutTests

Update test expectations for newly imported css-text layout tests
https://bugs.webkit.org/show_bug.cgi?id=185872

Unreviewed test gardening.

Patch by David Fenton <david_fenton@apple.com> on 2018-05-22

TestExpectations:

4:56 PM Changeset in webkit [232090] by dino@apple.com

17 edits in trunk/Source

Optimized path zoom animation needs a valid UIImage and CGRect
https://bugs.webkit.org/show_bug.cgi?id=185883
<rdar://problem/40306056>

Reviewed by Jon Lee.

Source/WebCore:

Pass the bounding box of the element that was clicked onto
the UI process, so it can perform an animation from that spot.

This involved adding an IntRect to the ResourceRequest, and passing
that info into it from the HTMLAnchorElement, using a new struct
called SystemPreviewInfo.

html/HTMLAnchorElement.cpp:

(WebCore::HTMLAnchorElement::handleClick):

loader/FrameLoadRequest.cpp:

(WebCore::FrameLoadRequest::FrameLoadRequest):

loader/FrameLoadRequest.h:

(WebCore::FrameLoadRequest::FrameLoadRequest):
(WebCore::FrameLoadRequest::isSystemPreview const):
(WebCore::FrameLoadRequest::systemPreviewRect const):

loader/FrameLoader.cpp:

(WebCore::FrameLoader::urlSelected):
(WebCore::FrameLoader::loadURL):

loader/FrameLoader.h:

(WebCore::FrameLoader::urlSelected):

loader/FrameLoaderTypes.h:
platform/network/ResourceRequestBase.cpp:

(WebCore::ResourceRequestBase::systemPreviewRect const):
(WebCore::ResourceRequestBase::setSystemPreviewRect):

platform/network/ResourceRequestBase.h:

Source/WebKit:

Take the rectangle that was passed into the ResourceRequest and
use it for the origin of an animation into QuickLook.

Shared/WebCoreArgumentCoders.cpp:

(IPC::ArgumentCoder<ResourceRequest>::encode):
(IPC::ArgumentCoder<ResourceRequest>::decode):

UIProcess/Cocoa/DownloadClient.mm:

(WebKit::DownloadClient::didStart):

UIProcess/Cocoa/SystemPreviewControllerCocoa.mm:

(-[_WKPreviewControllerDelegate initWithSystemPreviewController:fromRect:]):
(-[_WKPreviewControllerDelegate presentingViewController]):
(-[_WKPreviewControllerDelegate previewController:frameForPreviewItem:inSourceView:]):
(-[_WKPreviewControllerDelegate previewController:transitionImageForPreviewItem:contentRect:]):
(WebKit::SystemPreviewController::start):
(-[_WKPreviewControllerDelegate initWithSystemPreviewController:]): Deleted.

UIProcess/Downloads/DownloadProxy.h:

(WebKit::DownloadProxy::systemPreviewDownloadRect const):

UIProcess/SystemPreviewController.h:
UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::syncRootViewToScreen):

UIProcess/WebPageProxy.h:

4:54 PM Changeset in webkit [232089] by mark.lam@apple.com

16 edits
2 adds in trunk

StringImpl utf8 conversion should not fail silently.
https://bugs.webkit.org/show_bug.cgi?id=185888
<rdar://problem/40464506>

Reviewed by Filip Pizlo.

JSTests:

stress/regress-185888.js: Added.

Source/JavaScriptCore:

dfg/DFGLazyJSValue.cpp:

(JSC::DFG::LazyJSValue::dumpInContext const):

runtime/DateConstructor.cpp:

(JSC::constructDate):
(JSC::dateParse):

runtime/JSDateMath.cpp:

(JSC::parseDate):

runtime/JSDateMath.h:

Source/WTF:

WTF.xcodeproj/project.pbxproj:
wtf/CMakeLists.txt:
wtf/PrintStream.cpp:

(WTF::printExpectedCStringHelper):
(WTF::printInternal):

wtf/text/StringImpl.cpp:

(WTF::StringImpl::utf8Impl):
(WTF::StringImpl::utf8ForCharacters):
(WTF::StringImpl::tryUtf8ForRange const):
(WTF::StringImpl::tryUtf8 const):
(WTF::StringImpl::utf8 const):
(WTF::StringImpl::utf8ForRange const): Deleted.

wtf/text/StringImpl.h:
wtf/text/StringView.cpp:

(WTF::StringView::tryUtf8 const):
(WTF::StringView::utf8 const):

wtf/text/StringView.h:
wtf/text/UTF8ConversionError.h: Added.
wtf/text/WTFString.cpp:

(WTF::String::tryUtf8 const):
(WTF::String::utf8 const):

wtf/text/WTFString.h:

4:54 PM Changeset in webkit [232088] by Chris Dumez

2 edits in trunk/Source/WebCore

[POSIX] Use access() instead of stat() in FileSystem::fileExists()
https://bugs.webkit.org/show_bug.cgi?id=185882

Reviewed by Geoffrey Garen.

Use access() instead of stat() in FileSystem::fileExists(). stat() returns a lot of information we
do not leverage and local benchmarking on macOS shows access() being > 80% faster than stat():
stat: 0.31567 (lower is better)
access: 0.16074 (lower is better)
stat: 0.303665 (lower is better)
access: 0.165468 (lower is better)

platform/posix/FileSystemPOSIX.cpp:

(WebCore::FileSystem::fileExists):

4:45 PM Changeset in webkit [232087] by bshafiei@apple.com

1 copy in tags/Safari-606.1.17.40.1

Tag Safari-606.1.17.40.1.

4:45 PM Changeset in webkit [232086] by bshafiei@apple.com

1 copy in tags/Safari-606.1.17.30.1

Tag Safari-606.1.17.30.1.

4:42 PM Changeset in webkit [232085] by Ryan Haddad

9 edits
4 adds in trunk

Test infrastructure websites should use jQuery and jQuery UI hosted on webkit.org
https://bugs.webkit.org/show_bug.cgi?id=185788

Patch by Roy Reapor <rreapor@apple.com> on 2018-05-22
Reviewed by Aakash Jain.

Tools:

BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/rebaseline.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html:
TestResultServer/static-dashboards/run-embedded-unittests.html:
TestResultServer/static-dashboards/run-unittests.html:

Websites/webkit.org:

ajax/libs/jquery/jquery-1.6.2.min.js: Added.
ajax/libs/jqueryui: Added.
ajax/libs/jqueryui/jquery-ui-1.8.15.css: Added.

4:37 PM Changeset in webkit [232084] by Wenson Hsieh

1 edit
17 adds in trunk/Websites/webkit.org

Add some additional demo content to webkit.org
https://bugs.webkit.org/show_bug.cgi?id=185892

Reviewed by Tim Horton.

Adds three demo pages, and some placeholder assets.

demos/extrazoom/article/constitution.jpg: Added.
demos/extrazoom/article/index.html: Added.
demos/extrazoom/login/index.html: Added.
demos/extrazoom/login/webkit.svg: Added.
demos/extrazoom/recipes/butterchicken.jpg: Added.
demos/extrazoom/recipes/chowder.jpg: Added.
demos/extrazoom/recipes/hummus.jpg: Added.
demos/extrazoom/recipes/index.html: Added.
demos/extrazoom/recipes/padthai.jpg: Added.
demos/extrazoom/recipes/pizza.jpg: Added.
demos/extrazoom/recipes/salmon.jpg: Added.
demos/extrazoom/recipes/steak.jpg: Added.
demos/extrazoom/recipes/tacos.jpg: Added.

4:20 PM Changeset in webkit [232083] by sihui_liu@apple.com

10 edits in trunk

Source/WebKit:
[iOS] TestWebKitAPI.WebKit.WKHTTPCookieStoreWithoutProcessPool fails because cookies use different files with/without processpool
https://bugs.webkit.org/show_bug.cgi?id=185831

Reviewed by Chris Dumez.

Started to use uiProcessCookieStorageIdentifier for iOS: make sure cookies handled without
processpool would use the same storage file as when processpool exists.

NetworkProcess/NetworkProcess.h:
NetworkProcess/NetworkProcessCreationParameters.cpp:

(WebKit::NetworkProcessCreationParameters::encode const):
(WebKit::NetworkProcessCreationParameters::decode):

NetworkProcess/NetworkProcessCreationParameters.h:
NetworkProcess/cocoa/NetworkProcessCocoa.mm:

(WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):

Shared/WebProcessCreationParameters.cpp:

(WebKit::WebProcessCreationParameters::encode const):
(WebKit::WebProcessCreationParameters::decode):

Shared/WebProcessCreationParameters.h:
UIProcess/Cocoa/WebProcessPoolCocoa.mm:

(WebKit::WebProcessPool::platformInitializeWebProcess):
(WebKit::WebProcessPool::platformInitializeNetworkProcess):

Tools:
[iOS] TestWebKitAPI.WebKit.WKHTTPCookieStoreWithoutProcessPool is crashing with reason: 'Completion handler passed to -[CookieUIDelegate webView:runJavaScriptAlertPanelWithMessage:initiatedByFrame:completionHandler:] was not called'
https://bugs.webkit.org/show_bug.cgi?id=185831

Reviewed by Chris Dumez.

Updated the test to prevent crash after assertion fails.

TestWebKitAPI/Tests/WebKitCocoa/WKHTTPCookieStore.mm:

(-[CookieUIDelegate webView:runJavaScriptAlertPanelWithMessage:initiatedByFrame:completionHandler:]):

3:41 PM Changeset in webkit [232082] by Chris Dumez

46 edits
1 move
1 add in trunk

Regression(AsyncPolicyDelegates): Box.app login Window is blank
https://bugs.webkit.org/show_bug.cgi?id=185832
<rdar://problem/40307871>

Reviewed by Geoffrey Garen.

Source/WebKit:

Moved WeakObjCPtr.h header from WebKit/ to wtf/ so that it can be used in
WebKitLegacy code.

UIProcess/API/Cocoa/WKBrowsingContextController.mm:
UIProcess/API/Cocoa/WKBrowsingContextControllerInternal.h:
UIProcess/API/Cocoa/WKConnection.mm:
UIProcess/API/Cocoa/WKHTTPCookieStore.mm:
UIProcess/API/Cocoa/WKProcessGroup.mm:
UIProcess/API/Cocoa/WKProcessPool.mm:
UIProcess/API/Cocoa/WKScriptMessage.mm:
UIProcess/API/Cocoa/WKWebView.mm:
UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
UIProcess/API/Cocoa/_WKAutomationSession.mm:
UIProcess/API/Cocoa/_WKDownload.mm:
UIProcess/API/Cocoa/_WKElementAction.mm:
UIProcess/ApplicationStateTracker.h:
UIProcess/Cocoa/AutomationClient.h:
UIProcess/Cocoa/AutomationSessionClient.h:
UIProcess/Cocoa/DiagnosticLoggingClient.h:
UIProcess/Cocoa/DownloadClient.h:
UIProcess/Cocoa/FindClient.h:
UIProcess/Cocoa/FullscreenClient.h:
UIProcess/Cocoa/IconLoadingDelegate.h:
UIProcess/Cocoa/NavigationState.h:
UIProcess/Cocoa/UIDelegate.h:
UIProcess/Cocoa/VideoFullscreenManagerProxy.mm:
UIProcess/Cocoa/WKReloadFrameErrorRecoveryAttempter.mm:
UIProcess/Cocoa/WebViewImpl.h:
UIProcess/ios/ViewGestureControllerIOS.mm:
UIProcess/ios/WKActionSheetAssistant.mm:
UIProcess/ios/WKContentViewInteraction.mm:
UIProcess/ios/WKPDFView.mm:

(-[WKPDFView web_setContentProviderData:suggestedFilename:]):

UIProcess/ios/WKScrollView.mm:
UIProcess/mac/WKInspectorViewController.mm:
UIProcess/mac/WKInspectorWKWebView.mm:
WebKit.xcodeproj/project.pbxproj:
WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:

Source/WebKitLegacy/mac:

We used to have a bug where where we would fail to wait for the policy decision for
the navigation response from the client and the load would keep going, racing with
the client's policy decision. If the client did not respond in time, the behavior
would be the same as "Use" policy action.

Box.app fails to make any policy decision in its decidePolicyForMIMEType delegate
but the load happened to proceed anyway due to our bug. Now that we've fixed the
WebKit bug, however, the load would hang because the completion handler for the
decidePolicyForNavigationResponse would never get called.

To work around the issue, I made the policy listener weak on the WebFrameLoaderClient
instead of retaining it. If the policy listener object gets destroyed because getting
resolved, we now use "Use" policy action in its dealloc function to maintain previous
behavior.

WebCoreSupport/WebFrameLoaderClient.h:
WebCoreSupport/WebFrameLoaderClient.mm:

(WebFrameLoaderClient::cancelPolicyCheck):
(WebFrameLoaderClient::setUpPolicyListener):
(-[WebFramePolicyListener dealloc]):

Source/WTF:

Moved WeakObjCPtr.h from WebKit2 to WTF with RetainPtr.h, so that it can be used
in WebKitLegacy code.

WTF.xcodeproj/project.pbxproj:
wtf/WeakObjCPtr.h: Renamed from Source/WebKit/Shared/mac/WeakObjCPtr.h.

Tools:

Add API test coverage.

TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
TestWebKitAPI/Tests/WebKit/cocoa/WeakObjCPtr.mm:
TestWebKitAPI/Tests/mac/NoPolicyDelegateResponse.mm: Added.

(-[NoPolicyDelegateDecisionDelegate webView:decidePolicyForNavigationAction:request:frame:decisionListener:]):
(-[NoPolicyDelegateDecisionDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]):
(-[NoPolicyDelegateDecisionDelegate webView:didFinishLoadForFrame:]):
(TestWebKitAPI::TEST):

TestWebKitAPI/cocoa/TestWKWebView.mm:

(-[WKWebView _isBackground]):

3:01 PM Changeset in webkit [232081] by Brent Fulgham

7 edits in trunk/Source/WebCore

Avoid keeping FormState alive longer than necessary
https://bugs.webkit.org/show_bug.cgi?id=185877
<rdar://problem/39329219>

Reviewed by Ryosuke Niwa.

This patch does the following:

Changes to use WeakPtr's for FormState objects passed to completion handlers, rather than RefPtr, since those completion handlers might fire as part of the clean-up process during Frame destruction. This allows us to use the FormState if they are still valid, but gracefully handle cases where a form submission is cancelled in-flight.
Removes some extraneous WTFMove() calls being made on bare FormState pointers.
Changes the trap from Bug 183704 so that it only fires if the FormState object is being retained more than once.

loader/DocumentLoader.cpp:

(WebCore::DocumentLoader::willSendRequest): Update for new CompletionHandler
signature.

loader/FormState.cpp:

(WebCore::FormState::willDetachPage): Revise trap to check for retain counts
above one.

loader/FormState.h:

(WebCore::FormState::weakPtrFactory const): Added.

loader/FrameLoader.cpp:

(WebCore::FrameLoader::loadFrameRequest): Revise to use WeakPtr for FormState
passed to the completion handler.
(WebCore::FrameLoader::loadURL): Update for new CompletionHandler signature.
(WebCore::FrameLoader::load): Ditto.
(WebCore::FrameLoader::loadWithDocumentLoader): Ditto.
(WebCore::FrameLoader::loadPostRequest): Ditto.

loader/PolicyChecker.cpp:

(WebCore::PolicyChecker::checkNavigationPolicy): Revise to use WeakPtr for
FormState passed to the completion handler. Remove some extraneous WTFMove()
calls on bare pointers.
(WebCore::PolicyChecker::checkNewWindowPolicy): Ditto.

loader/PolicyChecker.h:

2:33 PM Changeset in webkit [232080] by keith_miller@apple.com

11 edits
1 delete in trunk/Source/JavaScriptCore

Remove the UnconditionalFinalizer class
https://bugs.webkit.org/show_bug.cgi?id=185881

Reviewed by Filip Pizlo.

The only remaining user of this API is
JSWebAssemblyCodeBlock. This patch changes, JSWebAssemblyCodeBlock
to use the newer template based API and removes the old class.

JavaScriptCore.xcodeproj/project.pbxproj:
bytecode/CodeBlock.h:
heap/Heap.cpp:

(JSC::Heap::finalizeUnconditionalFinalizers):

heap/Heap.h:
heap/SlotVisitor.cpp:

(JSC::SlotVisitor::addUnconditionalFinalizer): Deleted.

heap/SlotVisitor.h:
heap/UnconditionalFinalizer.h: Removed.
wasm/js/JSWebAssemblyCodeBlock.cpp:

(JSC::JSWebAssemblyCodeBlock::JSWebAssemblyCodeBlock):
(JSC::JSWebAssemblyCodeBlock::visitChildren):
(JSC::JSWebAssemblyCodeBlock::finalizeUnconditionally):
(JSC::JSWebAssemblyCodeBlock::UnconditionalFinalizer::finalizeUnconditionally): Deleted.

wasm/js/JSWebAssemblyCodeBlock.h:
wasm/js/JSWebAssemblyModule.h:

CMakeLists.txt:
JavaScriptCore.xcodeproj/project.pbxproj:
bytecode/CodeBlock.h:
heap/Heap.cpp:

(JSC::Heap::finalizeUnconditionalFinalizers):

heap/Heap.h:
heap/SlotVisitor.cpp:

(JSC::SlotVisitor::addUnconditionalFinalizer): Deleted.

heap/SlotVisitor.h:
heap/UnconditionalFinalizer.h: Removed.
wasm/js/JSWebAssemblyCodeBlock.cpp:

wasm/js/JSWebAssemblyCodeBlock.h:
wasm/js/JSWebAssemblyModule.h:

2:13 PM Changeset in webkit [232079] by sihui_liu@apple.com

7 edits
1 add in trunk

Conversion between SecurityOriginData and DatabaseIdentifier is asymmetric when port is null
https://bugs.webkit.org/show_bug.cgi?id=185715

Reviewed by Geoffrey Garen.

Source/WebCore:

Fixed the issue of null port when converting between SecurityOriginData and DatabaseIdentifier.

Test: WKWebView.LocalStorageFetchDataRecords.

page/SecurityOriginData.cpp:

(WebCore::SecurityOriginData::fromDatabaseIdentifier):

Source/WebKit:

Add getter for origins in WKWebsiteDataRecord for testing.

UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:

(-[WKWebsiteDataRecord _originsString]):

UIProcess/API/Cocoa/WKWebsiteDataRecordPrivate.h:

Tools:

Add API test coverage.

TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
TestWebKitAPI/Tests/WebKitCocoa/LocalStorageDatabaseTracker.mm: Added.

(-[LocalStorageUIDelegate webView:runJavaScriptAlertPanelWithMessage:initiatedByFrame:completionHandler:]):
(TEST):

1:52 PM Changeset in webkit [232078] by beidson@apple.com

2 edits in trunk/Source/WebKit

Rename the "Web content is visible" process assertion.
https://bugs.webkit.org/show_bug.cgi?id=185878

Reviewed by Chris Dumez.

UIProcess/ios/ProcessAssertionIOS.mm:

(WebKit::ProcessAssertion::ProcessAssertion):

1:48 PM Changeset in webkit [232077] by aestes@apple.com

6 edits in trunk/Source/WebKit

[Wi-Fi Assertions] Drop assertions on process suspension
https://bugs.webkit.org/show_bug.cgi?id=185844
<rdar://problem/40352319>

Reviewed by Daniel Bates.

NetworkProcess/NetworkProcess.cpp:

(WebKit::NetworkProcess::actualPrepareToSuspend):
(WebKit::NetworkProcess::processDidResume):

NetworkProcess/NetworkProcess.h:
NetworkProcess/cocoa/NetworkProcessCocoa.mm:

(WebKit::NetworkProcess::platformPrepareToSuspend):
(WebKit::NetworkProcess::platformProcessDidResume):

12:55 PM Changeset in webkit [232076] by keith_miller@apple.com

2 edits in trunk/Source/JavaScriptCore

Unreviewed, fix internal build.

runtime/JSImmutableButterfly.cpp:

12:47 PM Changeset in webkit [232075] by sbarati@apple.com

3 edits in trunk/Source/JavaScriptCore

DFG::LICMPhase should attempt to hoist edge type checks if hoisting the whole node fails
https://bugs.webkit.org/show_bug.cgi?id=144525

Reviewed by Filip Pizlo.

This patch teaches LICM to fall back to hoisting a node's type checks when
hoisting the entire node fails.

This patch follow the same principles we use when deciding to hoist nodes in general:

If the pre header is control equivalent to where the current check is, we

go ahead and hoist the check.

Otherwise, if hoisting hasn't failed before, we go ahead and gamble and

hoist the check. If hoisting failed in the past, we will not hoist the check.

dfg/DFGLICMPhase.cpp:

(JSC::DFG::LICMPhase::attemptHoist):

dfg/DFGUseKind.h:

(JSC::DFG::checkMayCrashIfInputIsEmpty):

12:20 PM Changeset in webkit [232074] by fpizlo@apple.com

30 edits
5 deletes in trunk/Source

Get rid of TLCs
https://bugs.webkit.org/show_bug.cgi?id=185846

Rubber stamped by Geoffrey Garen.
Source/JavaScriptCore:

This removes support for thread-local caches from the GC in order to speed up allocation a
bit.

We added TLCs as part of Spectre mitigations, which we have since removed.

We will want some kind of TLCs eventually, since they allow us to:

have a global GC, which may be a perf optimization at some point.
allocate objects from JIT threads, which we've been wanting to do for a while.

This change keeps the most interesting aspect of TLCs, which is the
LocalAllocator/BlockDirectory separation. This means that it ought to be easy to implement
TLCs again in the future if we wanted this feature.

This change removes the part of TLCs that causes a perf regression, namely that Allocator is
an offset that requires a bounds check and lookup that makes the rest of the allocation fast
path dependent on the load of the TLC. Now, Allocator is really just a LocalAllocator*, so
you can directly use it to allocate. This removes two loads and a check from the allocation
fast path. In hindsight, I probably could have made that whole thing more efficient, had I
allowed us to have a statically known set of LocalAllocators. This would have removed the
bounds check (one load and one branch) and it would have made it possible to CSE the load of
the TLC data structure, since that would no longer resize. But that's a harder change that
this patch, and we don't need it right now.

While reviewing the allocation hot paths, I found that CreateThis had an unnecessary branch
to check if the allocator is null. I removed that check. AssemblyHelpers::emitAllocate() does
that check already. Previously, the TLC bounds check doubled as this check.

This is a 1% speed-up on Octane and a 2.3% speed-up on TailBench. However, the Octane
speed-up on my machine includes an 8% regexp speed-up. I've found that sometimes regexp
speeds up or slows down by 8% depending on which path I build JSC from. Without that 8%, this
is still an Octane speed-up due to 2-4% speed-ups in earley, boyer, raytrace, and splay.

JavaScriptCore.xcodeproj/project.pbxproj:
Sources.txt:
bytecode/ObjectAllocationProfileInlines.h:

(JSC::ObjectAllocationProfile::initializeProfile):

dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::compileCreateThis):

ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileMakeRope):
(JSC::FTL::DFG::LowerDFGToB3::compileMaterializeNewObject):
(JSC::FTL::DFG::LowerDFGToB3::allocatePropertyStorageWithSizeImpl):
(JSC::FTL::DFG::LowerDFGToB3::allocateHeapCell):
(JSC::FTL::DFG::LowerDFGToB3::allocateObject):
(JSC::FTL::DFG::LowerDFGToB3::allocatorForSize):

heap/Allocator.cpp:

(JSC::Allocator::cellSize const):

heap/Allocator.h:

(JSC::Allocator::Allocator):
(JSC::Allocator::localAllocator const):
(JSC::Allocator::operator== const):
(JSC::Allocator::offset const): Deleted.

heap/AllocatorInlines.h:

(JSC::Allocator::allocate const):
(JSC::Allocator::tryAllocate const): Deleted.

heap/BlockDirectory.cpp:

(JSC::BlockDirectory::BlockDirectory):
(JSC::BlockDirectory::~BlockDirectory):

heap/BlockDirectory.h:

(JSC::BlockDirectory::allocator const): Deleted.

heap/CompleteSubspace.cpp:

(JSC::CompleteSubspace::allocateNonVirtual):
(JSC::CompleteSubspace::allocatorForSlow):
(JSC::CompleteSubspace::tryAllocateSlow):

heap/CompleteSubspace.h:
heap/Heap.cpp:

(JSC::Heap::Heap):

heap/Heap.h:

(JSC::Heap::threadLocalCacheLayout): Deleted.

heap/IsoSubspace.cpp:

(JSC::IsoSubspace::IsoSubspace):
(JSC::IsoSubspace::allocateNonVirtual):

heap/IsoSubspace.h:

(JSC::IsoSubspace::allocatorForNonVirtual):

heap/LocalAllocator.cpp:

(JSC::LocalAllocator::LocalAllocator):
(JSC::LocalAllocator::~LocalAllocator):

heap/LocalAllocator.h:

(JSC::LocalAllocator::cellSize const):
(JSC::LocalAllocator::tlc const): Deleted.

heap/ThreadLocalCache.cpp: Removed.
heap/ThreadLocalCache.h: Removed.
heap/ThreadLocalCacheInlines.h: Removed.
heap/ThreadLocalCacheLayout.cpp: Removed.
heap/ThreadLocalCacheLayout.h: Removed.
jit/AssemblyHelpers.cpp:

(JSC::AssemblyHelpers::emitAllocateWithNonNullAllocator):
(JSC::AssemblyHelpers::emitAllocate):
(JSC::AssemblyHelpers::emitAllocateVariableSized):

jit/JITOpcodes.cpp:

(JSC::JIT::emit_op_create_this):

runtime/JSLock.cpp:

(JSC::JSLock::didAcquireLock):

runtime/VM.cpp:

(JSC::VM::VM):
(JSC::VM::~VM):

runtime/VM.h:
runtime/VMEntryScope.cpp:

(JSC::VMEntryScope::~VMEntryScope):

runtime/VMEntryScope.h:

Source/WTF:

wtf/Platform.h:

12:16 PM Changeset in webkit [232073] by commit-queue@webkit.org

2 edits in trunk/LayoutTests

Update test expectations for newly imported css-text layout tests
https://bugs.webkit.org/show_bug.cgi?id=185872

Unreviewed test gardening.

Patch by David Fenton <david_fenton@apple.com> on 2018-05-22

TestExpectations: Flag tests that are not passing right now.

12:08 PM Changeset in webkit [232072] by Brent Fulgham

3 edits in trunk/Source/WebKit

REGRESSION(r229093): Re-enable Network Extension support in the WebContent process (Take 2)
https://bugs.webkit.org/show_bug.cgi?id=185874
<rdar://problem/40454404>

Reviewed by Eric Carlson.

Add back a necessary XPC connection after locking down the network features in r229093.

Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:
WebProcess/com.apple.WebProcess.sb.in:

11:31 AM Changeset in webkit [232071] by sihui_liu@apple.com

2 edits in trunk/Tools

Unreviewed, added myself as a WebKit committer.

Scripts/webkitpy/common/config/contributors.json:

11:04 AM Changeset in webkit [232070] by keith_miller@apple.com

75 edits
9 adds in trunk

We should have a CoW storage for NewArrayBuffer arrays.
https://bugs.webkit.org/show_bug.cgi?id=185003

Reviewed by Filip Pizlo.

JSTests:

stress/cow-convert-contiguous-to-array-storage.js: Added.