Timeline

Jun 4, 2019:

11:34 PM Changeset in webkit [246100] by bshafiei@apple.com

• 2 edits in branches/safari-608.1.27-branch/Source/WebCore

Cherry-pick r246099. rdar://problem/48006291

iPadOS: Google calendars new event picker moves and flickers in landscape with software keyboard
​https://bugs.webkit.org/show_bug.cgi?id=198556

Reviewed by Wenson Hsieh.

Fix the issuse that it's really hard to create a new event using Google calendar on iPadOS with
a software keyboard enabled because tapping on date brings up software keyboard and erratically
shifts up/down the new event dialog/pane.

Add a site specific quirk to pretend these input elements have inputmode=none for now.

No new tests since this is a site specific quirk.

• page/Quirks.cpp: (WebCore::Quirks::needsInputModeNoneImplicitly const):

git-svn-id: ​https://svn.webkit.org/repository/webkit/trunk@246099 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11:09 PM Changeset in webkit [246099] by rniwa@webkit.org

• 2 edits in trunk/Source/WebCore

iPadOS: Google calendars new event picker moves and flickers in landscape with software keyboard
​https://bugs.webkit.org/show_bug.cgi?id=198556

Reviewed by Wenson Hsieh.

Fix the issuse that it's really hard to create a new event using Google calendar on iPadOS with
a software keyboard enabled because tapping on date brings up software keyboard and erratically
shifts up/down the new event dialog/pane.

Add a site specific quirk to pretend these input elements have inputmode=none for now.

No new tests since this is a site specific quirk.

• page/Quirks.cpp:

(WebCore::Quirks::needsInputModeNoneImplicitly const):

10:41 PM Changeset in webkit [246098] by bshafiei@apple.com

• 2 edits in branches/safari-608.1.27-branch/Source/WebKit

Cherry-pick r246077. rdar://problem/51164741

Uploading third-party applications from /Applications needs additional syscall access
​https://bugs.webkit.org/show_bug.cgi?id=198537
<rdar://problem/51164741>

Reviewed by Brent Fulgham.

• WebProcess/com.apple.WebProcess.sb.in:

git-svn-id: ​https://svn.webkit.org/repository/webkit/trunk@246077 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10:00 PM Changeset in webkit [246097] by Chris Dumez

• 4 edits in trunk

Cookies set via [WKHTTPCookieStore setCookie:] on store right after constructing WKWebView get lost
​https://bugs.webkit.org/show_bug.cgi?id=198553
<rdar://problem/51317144>

Reviewed by Geoffrey Garen.

Source/WebKit:

If you call [WKHTTPCookieStore setCookie:] right after you construct the WKWebView and before the
WebContent process has finished launching, then WebsiteDataStore::processPoolForCookieStorageOperations()
would return null, even though there is already a view/page/WebProcessProxy/WebProcessPool for this data
store. As a result, the cookie would get added to the WebsiteDataStore's m_pendingCookies but it will
not be used since we've already previously launched a network process when we constructed the web view.

The reason processPoolForCookieStorageOperations() would return null is because WebsiteDataStore::processPools()
relies on WebProcessLifetimeObserver::processes() but processes only register themselves with the
WebProcessLifetimeObservers when they have pages *and* after they are finished launching.

This patch updates processPoolForCookieStorageOperations() to fallback to iterating over all process pools
and find a process pool with a process using the current data store and which has pages. This way, even if
the process is still launching, we'll properly find its associated process pool.

• UIProcess/WebsiteData/WebsiteDataStore.cpp:

(WebKit::WebsiteDataStore::processPoolForCookieStorageOperations):

Tools:

Add API test coverage.

• TestWebKitAPI/Tests/WebKitCocoa/WKHTTPCookieStore.mm:

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

9:15 PM Changeset in webkit [246096] by Alan Bujtas

• 4 edits
  2 adds in trunk

[ContentChangeObserver] Gmail text editing controls require two taps
​https://bugs.webkit.org/show_bug.cgi?id=198541
<rdar://problem/51375055>

Reviewed by Simon Fraser.

Source/WebCore:

When the animation completes we should also check if the newly visible content is also clickable and report it accordingly.
When the animated content is not clickable, we need to proceed with click instead of stopping at hover.

Test: fast/events/touch/ios/content-observation/100ms-delay-10ms-transition-on-mousemove-no-clickable.html

• page/ios/ContentChangeObserver.cpp:

(WebCore::isConsideredClickable):
(WebCore::ContentChangeObserver::didFinishTransition):
(WebCore::ContentChangeObserver::adjustObservedState):
(WebCore::ContentChangeObserver::StyleChangeScope::~StyleChangeScope):
(WebCore::ContentChangeObserver::StyleChangeScope::isConsideredClickable const): Deleted. -> Turn it into a static function so that didFinishTransition could call it as well.

• page/ios/ContentChangeObserver.h:

LayoutTests:

• fast/events/touch/ios/content-observation/100ms-delay-10ms-transition-on-mousemove-no-clickable.html: Added.

8:14 PM Changeset in webkit [246095] by Michael Catanzaro

• 10 edits in trunk

Fix miscellaneous build warnings
​https://bugs.webkit.org/show_bug.cgi?id=198544

Reviewed by Don Olmstead.

Source/JavaScriptCore:

Silence -Wclass-memaccess warning in this dangerous code.

• wasm/WasmInstance.cpp:

(JSC::Wasm::Instance::Instance):

Source/WebCore:

Carefully silence -Wsign-compare warnings.

• contentextensions/DFABytecodeCompiler.cpp:

(WebCore::ContentExtensions::DFABytecodeCompiler::compile):

• inspector/InspectorCanvas.cpp:

(WebCore::InspectorCanvas::indexForData):

• xml/XSLStyleSheetLibxslt.cpp:

(WebCore::XSLStyleSheet::parseString):

Source/WebKit:

Carefully silence -Wsign-compare warnings.

• NetworkProcess/cache/NetworkCacheData.cpp:

(WebKit::NetworkCache::readOrMakeSalt):

Tools:

When converting to PRIVATE include directories, we accidentally dropped SYSTEM here. The
naming convention used here is a bit confusing: the *_SYSTEM_INCLUDE_DIRECTORIES actually
uses both SYSTEM and PRIVATE. We should probably clarify this.

• TestWebKitAPI/PlatformGTK.cmake:

7:21 PM Changeset in webkit [246094] by Jon Davis

• 2 edits in trunk/Websites/webkit.org

Improve performance of the lazy image loader
​https://bugs.webkit.org/show_bug.cgi?id=198530

Reviewed by Devin Rousso.

• wp-content/themes/webkit/scripts/global.js:

(enableScrollableTables):
(lazyLoadImages):
(findParentMenu): Deleted.
(i.m.menus.m.targetMenu.targetMenuClass.indexOf): Deleted.
(inView): Deleted.
(): Deleted.
(loadImage.else.img.onload): Deleted.
(loadImage): Deleted.
(onMovement): Deleted.
(updateImages): Deleted.

6:03 PM Changeset in webkit [246093] by youenn@apple.com

• 8 edits in trunk

getUserMedia requests should be processed sequentially in UIProcess
​https://bugs.webkit.org/show_bug.cgi?id=198430
<rdar://problem/51311420>

Reviewed by Eric Carlson.

Source/WebKit:

Before the patch, we process all incoming gum/gdm requests in parallel.
We now queueu them and process them one at a time.
This allows to take into consideration state changes triggered by one request for the next one.
In particular, if a user grants a request, this might grant the next one as well.

To implement that, we keep a reference of the current request to process.
We queue other requests happening whenever another request comes.
When the request is processed, we look at the next one in the queue.
To ensure we do not stop processing the queue for no good reason, some refactoring is done:

• queue processing happens when sending back IPC response to WebProcess.
• denyRequest/grantRequest are consistently called in the manager proxy.

• UIProcess/UserMediaPermissionRequestManagerProxy.cpp:

(WebKit::UserMediaPermissionRequestManagerProxy::invalidatePendingRequests):
Invalidate pregranted requests as well.
(WebKit::UserMediaPermissionRequestManagerProxy::denyRequest):
Renamed from userMediaAccessWasDenied to denyRequest.
This method is now consistently used whenever the manager proxy wants to deny the request.
It does the IPC to the WebProcess and triggers processing of the next request.
(WebKit::UserMediaPermissionRequestManagerProxy::grantRequest):
Renamed from userMediaAccessWasGranted to grantRequest.
(WebKit::UserMediaPermissionRequestManagerProxy::finishGrantingRequest):
This method is now consistently used whenever the manager proxy wants to deny the request.
It does the IPC to the WebProcess and triggers processing of the next request.
(WebKit::UserMediaPermissionRequestManagerProxy::rejectionTimerFired):
We now keep a queue of request instead of request IDs to make the deny code path more consistent.
(WebKit::UserMediaPermissionRequestManagerProxy::requestUserMediaPermissionForFrame):
(WebKit::UserMediaPermissionRequestManagerProxy::processNextUserMediaRequestIfNeeded):
(WebKit::UserMediaPermissionRequestManagerProxy::startProcessingUserMediaPermissionRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionRequest):
To make sure we do not process a different request, we keep a pointer to the request and compare it with the current media request.
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionInvalidRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::processUserMediaPermissionValidRequest):
(WebKit::UserMediaPermissionRequestManagerProxy::viewIsBecomingVisible):

• UIProcess/UserMediaPermissionRequestManagerProxy.h:

(WebKit::UserMediaPermissionRequestManagerProxy::denyRequest):

• UIProcess/UserMediaPermissionRequestProxy.cpp:

(WebKit::setDeviceAsFirst):
(WebKit::UserMediaPermissionRequestProxy::allow):
(WebKit::UserMediaPermissionRequestProxy::deny):

• UIProcess/UserMediaPermissionRequestProxy.h:

Tools:

• TestWebKitAPI/Tests/WebKit/GetUserMediaReprompt.mm:

(-[GetUserMediaRepromptUIDelegate _webView:requestMediaCaptureAuthorization:decisionHandler:]):
(TestWebKitAPI::TEST):

• TestWebKitAPI/Tests/WebKit/getUserMedia.html:

5:40 PM Changeset in webkit [246092] by Keith Rollin

• 2 edits in trunk/Source/WebCore

Fix 64-bit vs 32-bit mismatch in ISOFairPlayStreamingPsshBox.cpp
​https://bugs.webkit.org/show_bug.cgi?id=198539
<rdar://problem/51410358>

Reviewed by Alex Christensen.

Both ISOFairPlayStreamingKeyAssetIdBox and
ISOFairPlayStreamingKeyContextBox have Vector<> data members. The
parse() members of these classes call Vector<>::resize() on these
members. In both cases, the type of the parameter passed is a
uint64_t. However, resize() takes a size_t. On some platforms, size_t
is a 32-bit value, leading to a compile-time type mismatch error. Fix
this by changing the type of the value passed to parse() into a
size_t.

No new tests -- no new functionality.

• platform/graphics/avfoundation/ISOFairPlayStreamingPsshBox.cpp:

(WebCore::ISOFairPlayStreamingKeyAssetIdBox::parse):
(WebCore::ISOFairPlayStreamingKeyContextBox::parse):

5:26 PM Changeset in webkit [246091] by Keith Rollin

• 2 edits in trunk/Source/WebCore

Fix 64-bit vs 32-bit mismatch in TileController.cpp
​https://bugs.webkit.org/show_bug.cgi?id=198540
<rdar://problem/51410851>

Reviewed by Alex Christensen.

TileController::blankPixelCountForTiles calculates its result as a
uint64_t, but returns it as an unsigned. The former is a 64-bit value,
while the latter can be a 32-bit value on some platforms. This
mismatch can lead to a compile-time error. Fix this by explicitly
casting the 64-bit value to an "unsigned".

No new tests -- no new functionality.

• platform/graphics/ca/TileController.cpp:

(WebCore::TileController::blankPixelCountForTiles):

5:23 PM Changeset in webkit [246090] by commit-queue@webkit.org

• 7 edits in trunk/Source/WebKit

Unreviewed, rolling out r246086.
​https://bugs.webkit.org/show_bug.cgi?id=198549

Causing Internal build failures (Requested by ShawnRoberts on
#webkit).

Reverted changeset:

"Remove some unused selection code"
​https://bugs.webkit.org/show_bug.cgi?id=198451
​https://trac.webkit.org/changeset/246086

5:11 PM Changeset in webkit [246089] by youenn@apple.com

• 3 edits in trunk/LayoutTests

Layout test landed flaky in 245873 [ Release ] http/wpt/service-workers/service-worker-networkprocess-crash.html is a flaky failure
​https://bugs.webkit.org/show_bug.cgi?id=198435
<rdar://problem/51314295>

Reviewed by Alex Christensen.

Make the test more robust by retrying the null test more than once.

• http/wpt/service-workers/service-worker-networkprocess-crash-expected.txt:
• http/wpt/service-workers/service-worker-networkprocess-crash.html:

4:44 PM Changeset in webkit [246088] by timothy_horton@apple.com

• 2 edits in trunk/Source/WebKit

Turn on background color estimation for all link indicators
​https://bugs.webkit.org/show_bug.cgi?id=198547
<rdar://problem/51140196>

Reviewed by Dean Jackson.

• WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::computeAndSendEditDragSnapshot):
(WebKit::linkIndicatorPositionInformation):

4:31 PM Changeset in webkit [246087] by Chris Dumez

• 3 edits
  3 adds in trunk

Crash when calling XMLHttpRequest.setRequestHeader() in a worker
​https://bugs.webkit.org/show_bug.cgi?id=198534
<rdar://problem/51393912>

Reviewed by Alex Christensen.

Source/WebCore:

Make sure the script execution context is a Document because calling document()
to get the settings.

Test: fast/workers/worker-xhr-setRequestHeader.html

• xml/XMLHttpRequest.cpp:

(WebCore::XMLHttpRequest::setRequestHeader):

LayoutTests:

Add layout test coverage.

• fast/workers/resources/worker-xhr-setRequestHeader.js: Added.
• fast/workers/worker-xhr-setRequestHeader-expected.txt: Added.
• fast/workers/worker-xhr-setRequestHeader.html: Added.

4:27 PM Changeset in webkit [246086] by timothy_horton@apple.com

• 7 edits in trunk/Source/WebKit

Remove some unused selection code
​https://bugs.webkit.org/show_bug.cgi?id=198451

Reviewed by Wenson Hsieh.

• Platform/spi/ios/UIKitSPI.h:
• Shared/ios/GestureTypes.h:
• UIProcess/ios/WKContentViewInteraction.mm:

(toGestureType):
(toUIWKGestureType):

• UIProcess/ios/WKPDFView.mm:

(-[WKPDFView actionSheetAssistant:shareElementWithURL:rect:]):

• WebProcess/WebPage/WebPage.h:
• WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::selectWithGesture):
(WebKit::WebPage::selectTextWithGranularityAtPoint):
(WebKit::canShrinkToTextSelection): Deleted.
(WebKit::hasCustomLineHeight): Deleted.
(WebKit::WebPage::rangeForWebSelectionAtPosition): Deleted.

3:56 PM Changeset in webkit [246085] by ddkilzer@apple.com

• 2 edits in trunk/Tools

REGRESSION (r244557): Leak of WKNSString in WTR::runOpenPanel() while running WebKit layout tests
<​https://webkit.org/b/198538>
<rdar://problem/51407719>

Reviewed by Darin Adler.

• WebKitTestRunner/TestController.cpp:

(WTR::runOpenPanel): Use adoptWK() to fix the leak.

3:08 PM Changeset in webkit [246084] by ysuzuki@apple.com

• 2 edits in trunk/Source/JavaScriptCore

Unreviewed, update exception scope for putByIndexBeyondVectorLength
​https://bugs.webkit.org/show_bug.cgi?id=198477

• runtime/JSObject.cpp:

(JSC::JSObject::putByIndexBeyondVectorLength):

2:53 PM Changeset in webkit [246083] by Antti Koivisto

• 9 edits
  12 adds in trunk

Sticky positioning is jumpy in many overflow cases
​https://bugs.webkit.org/show_bug.cgi?id=198532
<rdar://problem/51400532>

Reviewed by Simon Fraser.

Source/WebCore:

Tests: scrollingcoordinator/ios/sticky-overflow-no-stacking-context-no-stick-1.html

scrollingcoordinator/ios/sticky-overflow-no-stacking-context-no-stick-2.html
scrollingcoordinator/ios/sticky-overflow-no-stacking-context-stick-1.html
scrollingcoordinator/ios/sticky-overflow-no-stacking-context-stick-2.html
scrollingcoordinator/ios/sticky-overflow-stacking-context-no-stick.html
scrollingcoordinator/ios/sticky-overflow-stacking-context-stick.html

• page/scrolling/ScrollingTree.cpp:

(WebCore::ScrollingTree::notifyRelatedNodesAfterScrollPositionChange):
(WebCore::ScrollingTree::notifyRelatedNodesRecursive):

Simplify for relatedNodeScrollPositionDidChange removal.

• page/scrolling/ScrollingTree.h:
• page/scrolling/ScrollingTreeNode.cpp:

(WebCore::ScrollingTreeNode::relatedNodeScrollPositionDidChange): Deleted.

• page/scrolling/ScrollingTreeNode.h:
• page/scrolling/cocoa/ScrollingTreeFixedNode.mm:

(WebCore::ScrollingTreeFixedNode::applyLayerPositions):

• page/scrolling/cocoa/ScrollingTreePositionedNode.h:
• page/scrolling/cocoa/ScrollingTreePositionedNode.mm:

(WebCore::ScrollingTreePositionedNode::scrollOffsetSinceLastCommit const):

Factor into a function.

(WebCore::ScrollingTreePositionedNode::applyLayerPositions):
(WebCore::ScrollingTreePositionedNode::relatedNodeScrollPositionDidChange): Deleted.

We can't bail out based on changed node as that makes us compute different positions based on what the change root is.
Since all relatedNodeScrollPositionDidChange functions now always simply call applyLayerPositions we can remove the whole thing.

• page/scrolling/cocoa/ScrollingTreeStickyNode.mm:

(WebCore::ScrollingTreeStickyNode::applyLayerPositions):

Implement taking into account that the containing scroller may not be our ancestor.

LayoutTests:

• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-no-stick-1-expected.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-no-stick-1.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-no-stick-2-expected.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-no-stick-2.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-stick-1-expected.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-stick-1.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-stick-2-expected.html: Added.
• scrollingcoordinator/ios/sticky-overflow-no-stacking-context-stick-2.html: Added.
• scrollingcoordinator/ios/sticky-overflow-stacking-context-no-stick-expected.html: Added.
• scrollingcoordinator/ios/sticky-overflow-stacking-context-no-stick.html: Added.
• scrollingcoordinator/ios/sticky-overflow-stacking-context-stick-expected.html: Added.
• scrollingcoordinator/ios/sticky-overflow-stacking-context-stick.html: Added.

2:40 PM Changeset in webkit [246082] by aakash_jain@apple.com

• 2 edits in trunk/Tools

[ews-app] Add authentication while fetching bugs
​https://bugs.webkit.org/show_bug.cgi?id=198415
<rdar://problem/51298710>

Reviewed by Jonathan Bedard.

• BuildSlaveSupport/ews-app/ews/common/bugzilla.py:

(Bugzilla._fetch_attachment_json): Use api_key if configured in environment variable.
(BugzillaBeautifulSoup.authenticate): Method to authenticate, logic copied from webkitpy/common/net/bugzilla/bugzilla.py
(BugzillaBeautifulSoup._load_query):

2:27 PM Changeset in webkit [246081] by Kocsen Chung

• 7 edits in branches/safari-608.1.27-branch/Source

Versioning.

2:27 PM Changeset in webkit [246080] by aakash_jain@apple.com

• 2 edits in trunk/Tools

[ews-build] Do not display unnecessary steps in the Buildbot build page UI
​https://bugs.webkit.org/show_bug.cgi?id=198218
<rdar://problem/51104544>

Reviewed by Jonathan Bedard.

• BuildSlaveSupport/ews-build/steps.py:

1:54 PM Changeset in webkit [246079] by sihui_liu@apple.com

• 9 edits in trunk

WKWebsiteDataStore API fails to fetch web storage data for non-persistent data store
​https://bugs.webkit.org/show_bug.cgi?id=198317
Source/WebKit:

<rdar://problem/51244662>

Reviewed by Alex Christensen.

Use LocalStorageNameSpace instead of SessionStorageNameSpace for localStorage in ephemeral session or
websiteDataStore.

Reland r245943 as test has been fixed in r246012.

• NetworkProcess/WebStorage/StorageManager.cpp:

(WebKit::StorageManager::StorageArea::isEphemeral const):
(WebKit::StorageManager::StorageArea::removeListener):
(WebKit::StorageManager::StorageArea::setItems):
(WebKit::StorageManager::StorageArea::openDatabaseAndImportItemsIfNeeded const):
(WebKit::StorageManager::LocalStorageNamespace::~LocalStorageNamespace):
(WebKit::StorageManager::LocalStorageNamespace::getOrCreateStorageArea):
(WebKit::StorageManager::LocalStorageNamespace::clearAllStorageAreas):
(WebKit::StorageManager::LocalStorageNamespace::ephemeralOrigins const):
(WebKit::StorageManager::LocalStorageNamespace::cloneTo):
(WebKit::StorageManager::StorageManager):
(WebKit::StorageManager::cloneSessionStorageNamespace):
(WebKit::StorageManager::getLocalStorageOrigins):
(WebKit::StorageManager::getLocalStorageOriginDetails):
(WebKit::StorageManager::deleteLocalStorageEntriesForOrigin):
(WebKit::StorageManager::deleteLocalStorageOriginsModifiedSince):
(WebKit::StorageManager::deleteLocalStorageEntriesForOrigins):
(WebKit::StorageManager::createLocalStorageMap):
(WebKit::StorageManager::createTransientLocalStorageMap):
(WebKit::StorageManager::createSessionStorageMap):
(WebKit::StorageManager::destroyStorageMap):
(WebKit::StorageManager::getValues):
(WebKit::StorageManager::setItem):
(WebKit::StorageManager::removeItem):
(WebKit::StorageManager::clear):
(WebKit::StorageManager::suspend):
(WebKit::StorageManager::resume):
(WebKit::StorageManager::StorageArea::isSessionStorage const): Deleted.

• NetworkProcess/WebStorage/StorageManager.h:

(): Deleted.

• WebProcess/WebStorage/StorageAreaMap.cpp:

(WebKit::StorageAreaMap::dispatchStorageEvent):
(WebKit::StorageAreaMap::dispatchSessionStorageEvent):
(WebKit::StorageAreaMap::connect):

• WebProcess/WebStorage/StorageNamespaceImpl.cpp:

(WebKit::StorageNamespaceImpl::createEphemeralLocalStorageNamespace):
(WebKit::StorageNamespaceImpl::createLocalStorageNamespace):

• WebProcess/WebStorage/StorageNamespaceImpl.h:
• WebProcess/WebStorage/WebStorageNamespaceProvider.cpp:

(WebKit::WebStorageNamespaceProvider::createLocalStorageNamespace):

Tools:

Reviewed by Alex Christensen.

• TestWebKitAPI/Tests/WebKitCocoa/WKWebsiteDatastore.mm:

(TestWebKitAPI::TEST):

1:42 PM Changeset in webkit [246078] by Alan Coon

• 1 copy in tags/Safari-608.1.27.2

Tag Safari-608.1.27.2.

12:37 PM Changeset in webkit [246077] by achristensen@apple.com

• 2 edits in trunk/Source/WebKit

Uploading third-party applications from /Applications needs additional syscall access
​https://bugs.webkit.org/show_bug.cgi?id=198537
<rdar://problem/51164741>

Reviewed by Brent Fulgham.

• WebProcess/com.apple.WebProcess.sb.in:

12:15 PM Changeset in webkit [246076] by commit-queue@webkit.org

• 9 edits
  1 copy
  1 add in trunk

[WinCairo] Implement cpu and memory measuring functions.
​https://bugs.webkit.org/show_bug.cgi?id=198466

Patch by Takashi Komori <​Takashi.Komori@sony.com> on 2019-06-04
Reviewed by Don Olmstead.

.:

• Source/cmake/OptionsWin.cmake:

Source/WebCore:

Tests: inspector/memory/tracking.html

inspector/cpu-profiler/tracking.html

• PlatformWinCairo.cmake:
• page/ResourceUsageThread.h:
• page/win/ResourceUsageOverlayWin.cpp: Copied from Tools/WebKitTestRunner/InjectedBundle/win/TestRunnerWin.cpp.

(WebCore::ResourceUsageOverlay::platformInitialize):
(WebCore::ResourceUsageOverlay::platformDestroy):

• page/win/ResourceUsageThreadWin.cpp: Added.

(WebCore::ResourceUsageThread::platformSaveStateBeforeStarting):
(WebCore::fileTimeToUint64):
(WebCore::getCurrentCpuTime):
(WebCore::cpuUsage):
(WebCore::memoryUsage):
(WebCore::ResourceUsageThread::platformCollectCPUData):
(WebCore::ResourceUsageThread::platformCollectMemoryData):

Tools:

• WebKitTestRunner/InjectedBundle/win/TestRunnerWin.cpp:

(WTR::TestRunner::inspectorTestStubURL):

LayoutTests:

• platform/wincairo/TestExpectations:

12:06 PM Changeset in webkit [246075] by Tadeu Zagallo

• 3 edits
  1 add in trunk

Argument elimination should check transitive dependents for interference
​https://bugs.webkit.org/show_bug.cgi?id=198520
<rdar://problem/50863343>

Reviewed by Filip Pizlo.

JSTests:

• stress/argument-elimination-inline-rest-past-kill.js: Added.

(f2):
(f3):

Source/JavaScriptCore:

Consider the following program:

a: CreateRest
-->

b: CreateRest

<--
c: Spread(@a)
d: Spread(@b)
e: NewArrayWithSpread(@a, @b)
f: KillStack(locX)
g: LoadVarargs(@e)

Suppose @b reads locX, then we cannot transform @e to PhantomNewArraySpread, since that would
move the stack access from @b into @g, and that stack location is no longer valid at that point.

We fix that by computing a set of all inline call frames that any argument elimination candidate
depends on and checking each of them for interference in eliminateCandidatesThatInterfere.

• dfg/DFGArgumentsEliminationPhase.cpp:

12:00 PM Changeset in webkit [246074] by commit-queue@webkit.org

• 4 edits in trunk

[Pointer Events] Only allow pointer capture if the pointer is in the active buttons state
​https://bugs.webkit.org/show_bug.cgi?id=198479

Patch by Antoine Quint <Antoine Quint> on 2019-06-04
Reviewed by Dean Jackson.

LayoutTests/imported/w3c:

Mark WPT progression.

• web-platform-tests/pointerevents/pointerevent_setpointercapture_inactive_button_mouse-expected.txt:

Source/WebCore:

The Pointer Events specification says that pointer capture can only be engaged provided the pointer is
in the active buttons state, which means that it has dispatched a "pointerdown" event more recently than
it has a "pointerup" event.

This is tested by web-platform-tests/pointerevents/pointerevent_setpointercapture_inactive_button_mouse.html.

That test showed a few issues that this patch addresses. First, we would update the pointerIsPressed state to
"true" only after a "pointerdown" event had been dispatched. This is incorrect since setPointerCapture() can,
and is likely to, be called during handling of a "pointerdown" event. So we now call pointerEventWillBeDispatched()
prior to dispatching a PointerEvent with a mouse type, which we only did previously for a PointerEvent with a
touch or pen type. If the event is "pointerdown", we set "pointerIsPressed" to true on the CapturingData object
matching the given pointer, and to false if the event is "pointerup".

Finally, we must also ensure that "pointerIsPressed" is set to true when creating CapturingData for a PointerEvent
with a touch or pen type since these types of pointer events implictly set capture.

• page/PointerCaptureController.cpp:

(WebCore::PointerCaptureController::setPointerCapture):
(WebCore::PointerCaptureController::dispatchEvent):
(WebCore::PointerCaptureController::pointerEventWillBeDispatched):
(WebCore::PointerCaptureController::pointerEventWasDispatched):

11:27 AM Changeset in webkit [246073] by ysuzuki@apple.com

• 25 edits
  2 copies
  1 delete in trunk/Source/JavaScriptCore

[JSC] InferredValue should not be a JSCell
​https://bugs.webkit.org/show_bug.cgi?id=198407

Reviewed by Filip Pizlo.

Allocating InferredValue as a JSCell is too costly in terms of memory. Gmail has 90000 FunctionExecutables. And each gets
InferredValue, which takes 32 bytes. So it takes 2.7 MB memory footprint.

In this patch, we introduce a new container InferredValue<>. Which is similar to WriteBarrier<> container, but it replaces
the existing InferredValue cells with one pointer size field. The implementation of InferredValue<> is similar to
InlineWatchpointSet. But we encode JSCell* too to the pointer data of InlineWatchpointSet. So sizeof(InferredValue<>) is one
pointer size while it keeps Watchpoint feature and JSCell holder feature.

InferredValue<> needs validation in GC finalize phase. So this patch also makes SymbolTable Iso-allocated.

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

(JSC::ObjectAllocationProfileBase<Derived>::initializeProfile):

• bytecode/Watchpoint.h:
• dfg/DFGAbstractInterpreterInlines.h:

(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

• dfg/DFGByteCodeParser.cpp:

(JSC::DFG::ByteCodeParser::get):
(JSC::DFG::ByteCodeParser::parseBlock):

• dfg/DFGClobberize.h:

(JSC::DFG::clobberize):

• dfg/DFGClobbersExitState.cpp:

(JSC::DFG::clobbersExitState):

• dfg/DFGDesiredWatchpoints.cpp:

(JSC::DFG::SymbolTableAdaptor::add):
(JSC::DFG::FunctionExecutableAdaptor::add):
(JSC::DFG::DesiredWatchpoints::addLazily):
(JSC::DFG::DesiredWatchpoints::reallyAdd):
(JSC::DFG::DesiredWatchpoints::areStillValid const):
(JSC::DFG::DesiredWatchpoints::dumpInContext const):
(JSC::DFG::InferredValueAdaptor::add): Deleted.

• dfg/DFGDesiredWatchpoints.h:

(JSC::DFG::SymbolTableAdaptor::hasBeenInvalidated):
(JSC::DFG::SymbolTableAdaptor::dumpInContext):
(JSC::DFG::FunctionExecutableAdaptor::hasBeenInvalidated):
(JSC::DFG::FunctionExecutableAdaptor::dumpInContext):
(JSC::DFG::DesiredWatchpoints::isWatched):
(JSC::DFG::InferredValueAdaptor::hasBeenInvalidated): Deleted.
(JSC::DFG::InferredValueAdaptor::dumpInContext): Deleted.

• dfg/DFGObjectAllocationSinkingPhase.cpp:
• dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::compileNewFunction):
(JSC::DFG::SpeculativeJIT::compileCreateActivation):

• ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileCreateActivation):
(JSC::FTL::DFG::LowerDFGToB3::compileNewFunction):

• heap/Heap.cpp:

(JSC::Heap::finalizeUnconditionalFinalizers):

• runtime/FunctionExecutable.cpp:

(JSC::FunctionExecutable::FunctionExecutable):
(JSC::FunctionExecutable::finishCreation):
(JSC::FunctionExecutable::visitChildren):

• runtime/FunctionExecutable.h:
• runtime/FunctionExecutableInlines.h: Copied from Source/JavaScriptCore/runtime/InferredValueInlines.h.

(JSC::FunctionExecutable::finalizeUnconditionally):

• runtime/InferredValue.cpp: Removed.
• runtime/InferredValue.h:

(JSC::InferredValue::inferredValue):
(JSC::InferredValue::InferredValue):
(JSC::InferredValue::~InferredValue):
(JSC::InferredValue::stateOnJSThread const):
(JSC::InferredValue::state const):
(JSC::InferredValue::hasBeenInvalidated const):
(JSC::InferredValue::isStillValid const):
(JSC::InferredValue::invalidate):
(JSC::InferredValue::isBeingWatched const):
(JSC::InferredValue::notifyWrite):
(JSC::InferredValue::isThin):
(JSC::InferredValue::isFat):
(JSC::InferredValue::decodeState):
(JSC::InferredValue::encodeState):
(JSC::InferredValue::isThin const):
(JSC::InferredValue::isFat const):
(JSC::InferredValue::fat):
(JSC::InferredValue::fat const):
(JSC::InferredValue::inflate):
(JSC::InferredValue<JSCellType>::InferredValueWatchpointSet::notifyWriteSlow):
(JSC::InferredValue<JSCellType>::notifyWriteSlow):
(JSC::InferredValue<JSCellType>::add):
(JSC::InferredValue<JSCellType>::inflateSlow):
(JSC::InferredValue<JSCellType>::freeFat):

• runtime/InferredValueInlines.h:

(JSC::InferredValue<JSCellType>::finalizeUnconditionally):
(JSC::InferredValue::finalizeUnconditionally): Deleted.

• runtime/JSFunctionInlines.h:

(JSC::JSFunction::createWithInvalidatedReallocationWatchpoint):

• runtime/JSSymbolTableObject.h:

(JSC::JSSymbolTableObject::setSymbolTable):

• runtime/SymbolTable.cpp:

(JSC::SymbolTable::finishCreation):
(JSC::SymbolTable::visitChildren):

• runtime/SymbolTable.h:
• runtime/SymbolTableInlines.h: Copied from Source/JavaScriptCore/runtime/InferredValueInlines.h.

(JSC::SymbolTable::finalizeUnconditionally):

• runtime/VM.cpp:

(JSC::VM::VM):

• runtime/VM.h:

10:59 AM Changeset in webkit [246072] by Keith Rollin

• 2 edits in trunk/Source/WebCore

Fix 32-bit/64-bit mismatch in PointerCaptureController::elementWasRemoved
​https://bugs.webkit.org/show_bug.cgi?id=198501
<rdar://problem/51370464>

Reviewed by Chris Dumez.

keyAndValue.key is assigned to pointerId. KeyAndValue.key is a
int64_t, whereas pointerId is a PointerID, aka int32_t. This mismatch
is normally just a warning, but breaks builds where warnings are
treated as errors.

This issue is not encountered in most builds because the warning is
disabled in the majority of build configurations. But there are some
where the warning is not disabled, and so those builds break.

Address this conversion error/warning by explicitly casting
keyAndValue.key to a PointerID (and adding a debug check to make sure
the cast is OK).

No new tests -- no new functionality.

• page/PointerCaptureController.cpp:

(WebCore::PointerCaptureController::elementWasRemoved):

10:56 AM Changeset in webkit [246071] by Tadeu Zagallo

• 4 edits
  1 add in trunk

Argument elimination should check for negative indices in GetByVal
​https://bugs.webkit.org/show_bug.cgi?id=198302
<rdar://problem/51188095>

Reviewed by Filip Pizlo.

JSTests:

• stress/eliminate-arguments-negative-rest-access.js: Added.

(inlinee):
(opt):

Source/JavaScriptCore:

In DFG::ArgumentEliminationPhase, the index is treated as unsigned, but there's no check
for overflow in the addition. In compileGetMyArgumentByVal, there's a check for overflow,
but the index is treated as signed, resulting in an index lower than numberOfArgumentsToSkip.

• dfg/DFGArgumentsEliminationPhase.cpp:
• ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileGetMyArgumentByVal):

10:45 AM Changeset in webkit [246070] by graouts@webkit.org

• 14 edits
  4 adds in trunk

[Pointer Events] Expose navigator.maxTouchPoints
​https://bugs.webkit.org/show_bug.cgi?id=198468
<rdar://problem/51273029>

Reviewed by Chris Dumez.

Source/WebCore:

Expose the navigator.maxTouchPoints property when Pointer Events are enabled both at compile-time and run-time.
We return a canned value for this on iOS touch-enabled devices that matches the number of simultaneous touches
supported by the system, which is 5. In fact, iPad support more simultaneous touches, but it doesn't seem worthy
to expose this level of granularity due to fingerprinting practices. In practice, what really matters is returning
0, 1 or more than 1 for this value to identify multi-touch support.

Test: js/dom/navigator-maxtouchpoints.html

• CMakeLists.txt:
• DerivedSources-input.xcfilelist:
• DerivedSources-output.xcfilelist:
• DerivedSources.make:
• Sources.txt:
• WebCore.xcodeproj/project.pbxproj:
• dom/NavigatorMaxTouchPoints.idl: Added.
• page/Navigator.idl:
• page/Navigator.h:

(WebCore::Navigator::maxTouchPoints const):

LayoutTests:

• js/dom/navigator-maxtouchpoints-expected.txt: Added.
• js/dom/navigator-maxtouchpoints.html: Added.
• platform/mac-highsierra-wk1/fast/dom/navigator-detached-no-crash-expected.txt:
• platform/mac-highsierra/fast/dom/navigator-detached-no-crash-expected.txt:
• platform/mac-wk1/fast/dom/navigator-detached-no-crash-expected.txt:
• platform/mac-wk2/fast/dom/navigator-detached-no-crash-expected.txt:
• platform/win/js/dom/navigator-maxtouchpoints-expected.txt: Added.

10:09 AM Changeset in webkit [246069] by Chris Dumez

• 3 edits in trunk/Source/WebKit

Lazily construct the NetworkHTTPSUpgradeChecker
​https://bugs.webkit.org/show_bug.cgi?id=198500

Reviewed by Alex Christensen.

Lazily construct the NetworkHTTPSUpgradeChecker since the feature is off by default and this
avoids unnecessarily opening its database in the common case.

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

(WebKit::NetworkProcess::networkHTTPSUpgradeChecker):

9:11 AM Changeset in webkit [246068] by youenn@apple.com

• 2 edits in trunk/Source/WebKit

REGRESSION (r245913) [ Debug ] ASSERTION FAILED: m_swConnectionsByIdentifier.contains(connection.serverConnectionIdentifier()) Layout Test http/wpt/service-workers/update-service-worker.https.html is a flaky crash
​https://bugs.webkit.org/show_bug.cgi?id=198434
<rdar://problem/51313917>

Reviewed by Alex Christensen.

Iterate through the connectionID-based connection map instead of the sessionID-based map to notify them their connection is lost.

• WebProcess/Network/NetworkProcessConnection.cpp:

(WebKit::NetworkProcessConnection::didClose):

8:52 AM Changeset in webkit [246067] by Alan Bujtas

• 4 edits in trunk/Source/WebCore

[LFC][IFC] Decouple float placement and line shrinking
​https://bugs.webkit.org/show_bug.cgi?id=198528
<rdar://problem/51397638>

Reviewed by Antti Koivisto.

In LineLayout::placeInlineItems() float handling should be only about shrinking the current line, the actual
float placement should happen later when we construct the the display boxes/runs. It enables the preferred width
computation to call placeInlineItems() to gather line widths without accidentally mutating the layout context.

• layout/inlineformatting/InlineFormattingContext.h:
• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::InlineFormattingContext::LineLayout::placeInlineItems const):
(WebCore::Layout::InlineFormattingContext::LineLayout::layout const):
(WebCore::Layout::InlineFormattingContext::LineLayout::createDisplayRuns const):
(WebCore::Layout::InlineFormattingContext::LineLayout::handleFloat const): Deleted.

• layout/inlineformatting/InlineItem.h:

8:49 AM Changeset in webkit [246066] by Alan Bujtas

• 4 edits in trunk/Source/WebCore

[LFC][IFC] Add hard line break handling to LineBreaker
​https://bugs.webkit.org/show_bug.cgi?id=198503
<rdar://problem/51373482>

Reviewed by Antti Koivisto.

LineBreaker should simply return BreakingContext::Keep with the breaking opportunity of yes.

• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::InlineFormattingContext::LineLayout::placeInlineItems const):

• layout/inlineformatting/InlineLineBreaker.cpp:

(WebCore::Layout::LineBreaker::breakingContext):
(WebCore::Layout::LineBreaker::wordBreakingBehavior const):
(WebCore::Layout::LineBreaker::isAtBreakingOpportunity):

• layout/inlineformatting/InlineLineBreaker.h:

8:48 AM Changeset in webkit [246065] by Alan Bujtas

• 4 edits in trunk/Source/WebCore

[LFC][IFC] Remove InlineItem::width
​https://bugs.webkit.org/show_bug.cgi?id=198502
<rdar://problem/51371744>

Reviewed by Antti Koivisto.

InlineItems are supposd to work across subsequent layouts (and in preferred width computation as well) so they should
not hold on to layout information (run width). This would not work with split runs either.

• layout/inlineformatting/InlineFormattingContext.h:
• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::UncommittedContent::runs):
(WebCore::Layout::UncommittedContent::isEmpty const):
(WebCore::Layout::UncommittedContent::size const):
(WebCore::Layout::UncommittedContent::add):
(WebCore::Layout::UncommittedContent::reset):
(WebCore::Layout::InlineFormattingContext::LineLayout::placeInlineItems const):
(WebCore::Layout::InlineFormattingContext::LineLayout::computedIntrinsicWidth const):
(WebCore::Layout::InlineFormattingContext::LineLayout::createDisplayRuns const):
(): Deleted.
(WebCore::Layout::InlineFormattingContext::LineLayout::commitInlineItemToLine const): Deleted.

• layout/inlineformatting/InlineItem.h:

(WebCore::Layout::InlineItem::style const):
(): Deleted.
(WebCore::Layout::InlineItem::setWidth): Deleted.
(WebCore::Layout::InlineItem::width const): Deleted.

8:46 AM Changeset in webkit [246064] by Chris Dumez

• 2 edits in trunk/Source/WebKit

[iOS] UIProcess' background task expiration handler may get called after the app is foreground again
​https://bugs.webkit.org/show_bug.cgi?id=198380
<rdar://problem/49762471>

Reviewed by Geoff Garen.

UIProcess' background task expiration handler may get called after the app is foreground again. When
this happens, we already have a foreground assertion on behalf of the WebContent process, and the view
is visible. We would send the WillSuspendImminently IPC to the WebContent process, which would freeze
its layers, even though it is visible on screen.

To address the issue, we now check if the app is visible in the expiration handler. If it is visible,
we do not notify clients of imminent suspension. Instead, we end the background task right away and
call _updateBackgroundTask asynchronously to start a new background task if necessary.

• UIProcess/ios/ProcessAssertionIOS.mm:

(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):

8:26 AM Changeset in webkit [246063] by Alan Bujtas

• 4 edits in trunk/Source/WebCore

[LFC][IFC] Move run width measuring out of LineBreaker
​https://bugs.webkit.org/show_bug.cgi?id=198491
<rdar://problem/51363554>

Reviewed by Antti Koivisto.

LineBreaker should not need to deal with measuring runs.
This is also in preparation for removing InlineItem::width().

• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::inlineItemWidth):
(WebCore::Layout::InlineFormattingContext::LineLayout::placeInlineItems const):
(WebCore::Layout::InlineFormattingContext::LineLayout::computedIntrinsicWidth const):

• layout/inlineformatting/InlineLineBreaker.cpp:

(WebCore::Layout::LineBreaker::breakingContext):
(WebCore::Layout::LineBreaker::LineBreaker): Deleted.
(WebCore::Layout::LineBreaker::runWidth const): Deleted.
(WebCore::Layout::LineBreaker::textWidth const): Deleted.

• layout/inlineformatting/InlineLineBreaker.h:

7:54 AM Changeset in webkit [246062] by Alan Bujtas

• 4 edits in trunk/Source/WebCore

[LFC][IFC] Remove redundant InlineItem::width() calls.
​https://bugs.webkit.org/show_bug.cgi?id=198489
<rdar://problem/51360390>

Reviewed by Antti Koivisto.

This is in preparation for removing InlineItem::width().

• layout/inlineformatting/InlineFormattingContextLineLayout.cpp:

(WebCore::Layout::InlineFormattingContext::LineLayout::handleFloat const):
(WebCore::Layout::InlineFormattingContext::LineLayout::commitInlineItemToLine const):

• layout/inlineformatting/InlineLine.cpp:

(WebCore::Layout::Line::appendNonBreakableSpace):
(WebCore::Layout::Line::appendInlineContainerStart):
(WebCore::Layout::Line::appendInlineContainerEnd):

• layout/inlineformatting/InlineLine.h:

7:34 AM Changeset in webkit [246061] by commit-queue@webkit.org

• 15 edits
  4 adds in trunk

The "mouseenter" and "pointerenter" events are fired from the bottom up
​https://bugs.webkit.org/show_bug.cgi?id=198036
<rdar://problem/50940350>

Patch by Antoine Quint <Antoine Quint> on 2019-06-04
Reviewed by Darin Adler.

Source/WebCore:

Ensure "mouseenter" and "pointerenter" events are dispatched from the bottom up to match the UI Events spec
at ​https://w3c.github.io/uievents/#events-mouseevent-event-order. We also fix the issue where "pointerevent"
and "pointerleave" events were dispatched as bubbling events on iOS which is not correct and was caught by the
new iOS test.

Tests: pointerevents/ios/enter-leave-order.html

pointerevents/mouse/enter-leave-order.html

• dom/ios/PointerEventIOS.cpp:

(WebCore::typeCanBubble):
(WebCore::PointerEvent::PointerEvent):

• page/EventHandler.cpp:

(WebCore::EventHandler::updateMouseEventTargetNode):

• page/PointerCaptureController.cpp:

(WebCore::PointerCaptureController::dispatchEventForTouchAtIndex):

LayoutTests:

• fast/events/mouseenter-mouseleave-capture-expected.txt:
• fast/events/mouseenter-mouseleave-expected.txt:
• fast/events/mouseenterleave-on-subframe-expected.txt:
• fast/events/shadow-event-path-expected.txt:
• fast/shadow-dom/mouseenter-mouseleave-across-shadow-boundary-expected.txt:
• fast/shadow-dom/mouseenter-mouseleave-inside-shadow-tree-expected.txt:
• fast/shadow-dom/mouseenter-mouseleave-on-slot-parent-expected.txt:
• platform/mac-wk1/TestExpectations:
• platform/mac-wk2/fast/events/shadow-event-path-expected.txt:
• platform/mac/fast/events/shadow-event-path-2-expected.txt:
• pointerevents/ios/enter-leave-order-expected.txt: Added.
• pointerevents/ios/enter-leave-order.html: Added.
• pointerevents/mouse/enter-leave-order-expected.txt: Added.
• pointerevents/mouse/enter-leave-order.html: Added.

4:14 AM Changeset in webkit [246060] by Tadeu Zagallo

• 15 edits
  2 adds in trunk/Source/JavaScriptCore

JSScript should not keep bytecode cache in memory
​https://bugs.webkit.org/show_bug.cgi?id=198482

Reviewed by Saam Barati.

When JSScript writes to the cache, we keep the in-memory serialized bytecode alive.
Instead, we should only ever hold the memory mapped bytecode cache to avoid using
too much memory.

• API/JSScript.mm:

(-[JSScript writeCache:]):

• API/tests/testapi.mm:

(testBytecodeCacheWithSyntaxError):

• CMakeLists.txt:
• JavaScriptCore.xcodeproj/project.pbxproj:
• Sources.txt:
• jsc.cpp:
• parser/SourceProvider.h:
• runtime/BytecodeCacheError.cpp: Added.

(JSC::BytecodeCacheError::StandardError::isValid const):
(JSC::BytecodeCacheError::StandardError::message const):
(JSC::BytecodeCacheError::WriteError::isValid const):
(JSC::BytecodeCacheError::WriteError::message const):
(JSC::BytecodeCacheError::operator=):
(JSC::BytecodeCacheError::isValid const):
(JSC::BytecodeCacheError::message const):

• runtime/BytecodeCacheError.h: Added.

(JSC::BytecodeCacheError::StandardError::StandardError):
(JSC::BytecodeCacheError::WriteError::WriteError):

• runtime/CachedBytecode.h:

(JSC::CachedBytecode::create):

• runtime/CachedTypes.cpp:

(JSC::Encoder::Encoder):
(JSC::Encoder::release):
(JSC::Encoder::releaseMapped):
(JSC::encodeCodeBlock):
(JSC::encodeFunctionCodeBlock):

• runtime/CachedTypes.h:
• runtime/CodeCache.cpp:

(JSC::serializeBytecode):

• runtime/CodeCache.h:
• runtime/Completion.cpp:

(JSC::generateProgramBytecode):
(JSC::generateModuleBytecode):

• runtime/Completion.h:

3:37 AM Changeset in webkit [246059] by zandobersek@gmail.com

• 4 edits in trunk/Tools

[misc] Remove JSCOnly ARMv7 Traditional bot
​https://bugs.webkit.org/show_bug.cgi?id=198524

Reviewed by Carlos Garcia Campos.

This is not a supported configuration anymore, so the buildbot slave can
be removed from configuration.

• BuildSlaveSupport/build.webkit.org-config/config.json:
• BuildSlaveSupport/build.webkit.org-config/public_html/dashboard/Scripts/WebKitBuildbot.js:

(WebKitBuildbot):

• BuildSlaveSupport/build.webkit.org-config/steps_unittest.py:

3:04 AM Changeset in webkit [246058] by Carlos Garcia Campos

• 2 edits in trunk/Source/WebKit

[GTK] Crash when re-entering AC mode after r245957
​https://bugs.webkit.org/show_bug.cgi?id=198522

Reviewed by Žan Doberšek.

When AcceleratedBackingStoreWayland::displayBuffer() is called and we have already left AC mode, we early
dispatch the image and release it. The problem is that images are reused, so the given image can be the
committed one. In that case we end up freeing it again in the next AcceleratedBackingStoreWayland::paint()
call.

• UIProcess/gtk/AcceleratedBackingStoreWayland.cpp:

(WebKit::AcceleratedBackingStoreWayland::displayBuffer): Do not release the given image if it's the committed one.

12:38 AM Changeset in webkit [246057] by commit-queue@webkit.org

• 8 edits
  1 copy
  7 adds in trunk

JS wrapper of target in ResizeObserverEntry/ResizeObserver shouldn't get collected ahead
​https://bugs.webkit.org/show_bug.cgi?id=197457

Patch by Cathie Chen <cathiechen> on 2019-06-04
Reviewed by Ryosuke Niwa.

Source/WebCore:

Add JSCustomMarkFunction to make sure JS wrappers wouldn't be collected when JSResizeObserverEntry live.

For ResizeObserver, if targets are removed, it will get fired for the last time. We also need to keep these JS
wrappers live. So add these targets to a GCReachableRef list once they're observed.

Add element-leak.html to test the targets with entry.target.myEntry = entry could be released properly.

Tests: resize-observer/element-leak.html

resize-observer/resize-observer-entry-keeps-js-wrapper-of-target-alive.html
resize-observer/resize-observer-keeps-js-wrapper-of-target-alive.html

• Sources.txt:
• WebCore.xcodeproj/project.pbxproj:
• bindings/js/JSResizeObserverEntryCustom.cpp: Added.

(WebCore::JSResizeObserverEntry::visitAdditionalChildren):

• page/ResizeObserver.cpp:

(WebCore::ResizeObserver::observe):
(WebCore::ResizeObserver::removeAllTargets):
(WebCore::ResizeObserver::removeObservation):
(WebCore::ResizeObserver::stop):

• page/ResizeObserver.h:
• page/ResizeObserverEntry.idl:

LayoutTests:

• platform/win/TestExpectations:
• resize-observer/element-leak-expected.txt: Added.
• resize-observer/element-leak.html: Added.
• resize-observer/resize-observer-entry-keeps-js-wrapper-of-target-alive-expected.txt: Added.
• resize-observer/resize-observer-entry-keeps-js-wrapper-of-target-alive.html: Added.
• resize-observer/resize-observer-keeps-js-wrapper-of-target-alive-expected.txt: Added.
• resize-observer/resize-observer-keeps-js-wrapper-of-target-alive.html: Added.
• resize-observer/resources/element-leak-frame.html: Added.

Jun 3, 2019:

11:36 PM Changeset in webkit [246056] by aestes@apple.com

• 16 edits
  5 adds in trunk

[Apple Pay] Disable script injection when canMakePayment APIs are called and return true
​https://bugs.webkit.org/show_bug.cgi?id=198448
<rdar://problem/51323694>

Reviewed by Alex Christensen.

Source/WebCore:

Previously, only an active Apple Pay session would disable script injection in restricted
WKWebViews. However, this can result in websites rendering non-functional Apple Pay buttons
due to the race between the hosting app calling -evaluateJavaScript:completionHandler: and
the website calling canMakePayment APIs to determine whether to draw a button.

This patch makes it so that, if a website calls ApplePaySession's canMakePayments or
canMakePaymentsWithActiveCard, or PaymentRequest's canMakePayment, in a web view that has no
injected scripts, and those calls return true, future script injections from the hosting app
will be blocked.

Also, this patch removes the restrictions on the openPaymentSetup, supportsVersion, and
validatedPaymentNetwork APIs, since those APIs do not reveal transaction information and are
not used to determine whether to draw buttons.

Added new API tests.

• Modules/applepay/PaymentCoordinator.cpp:

(WebCore::PaymentCoordinator::supportsVersion const):
(WebCore::PaymentCoordinator::canMakePayments):
(WebCore::PaymentCoordinator::canMakePaymentsWithActiveCard):
(WebCore::PaymentCoordinator::openPaymentSetup):
(WebCore::PaymentCoordinator::beginPaymentSession):
(WebCore::PaymentCoordinator::validatedPaymentNetwork const):
(WebCore::PaymentCoordinator::setApplePayIsActiveIfAllowed const):
(WebCore::PaymentCoordinator::shouldAllowUserAgentScripts const):
(WebCore::PaymentCoordinator::shouldAllowApplePay const): Deleted.

• Modules/applepay/PaymentCoordinator.h:
• dom/Document.cpp:

(WebCore::Document::isApplePayActive const):
(WebCore::Document::setApplePayIsActive):
(WebCore::Document::hasStartedApplePaySession const): Deleted.
(WebCore::Document::setHasStartedApplePaySession): Deleted.

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

(WebCore::Internals::setApplePayIsActive):
(WebCore::Internals::setHasStartedApplePaySession): Deleted.

• testing/Internals.h:
• testing/Internals.idl:

Tools:

• TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
• TestWebKitAPI/Tests/WebKitCocoa/ApplePay.mm:

(-[TestApplePayAvailableScriptMessageHandler userContentController:didReceiveScriptMessage:]):
(-[TestApplePayActiveSessionScriptMessageHandler userContentController:didReceiveScriptMessage:]):
(TestWebKitAPI::TEST):
(TestWebKitAPI::runActiveSessionTest):
(-[TestApplePayScriptMessageHandler initWithAPIsAvailableExpectation:canMakePaymentsExpectation:]): Deleted.
(-[TestApplePayScriptMessageHandler userContentController:didReceiveScriptMessage:]): Deleted.

• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-active-session.html:
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-availability-existing-object.html: Added.
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-availability-in-iframe.html:
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-availability.html:
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-can-make-payment.html: Added.
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-can-make-payments-with-active-card.html: Added.
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay-can-make-payments.html: Added.
• TestWebKitAPI/Tests/WebKitCocoa/apple-pay.js: Added.

(applePayRequestBase):
(applePayPaymentRequest):
(applePayMethod):

• TestWebKitAPI/cocoa/TestProtocol.mm:

(-[TestProtocol startLoading]):

10:05 PM Changeset in webkit [246055] by bshafiei@apple.com

• 5 edits in branches/safari-608.1.27-branch

Cherry-pick r246042. rdar://problem/50667125

Tweak the text and underline color for data detected text.
​https://bugs.webkit.org/show_bug.cgi?id=198487
rdar://problem/50667125

Reviewed by Devin Rousso.

Source/WebCore:

Tests: Color.RGBToHSL API tests

• editing/cocoa/DataDetection.mm: (WebCore::DataDetection::detectContentInRange): Use currentcolor so semantic text colors work. Force the lightness of the underline color to the middle, and multiply the alpha by 38%, so the color will appear on light and dark backgrounds, since only one color can be specified.
• platform/graphics/Color.cpp: (WebCore::Color::getHSL const): Return hue in [0...6) range to easily round-trip with makeRGBAFromHSLA().

Tools:

• TestWebKitAPI/Tests/WebCore/Color.cpp: (TestWebKitAPI::TEST): Added Color.RGBToHSL tests.

git-svn-id: ​https://svn.webkit.org/repository/webkit/trunk@246042 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10:01 PM Changeset in webkit [246054] by bshafiei@apple.com

• 7 edits in branches/safari-608.1.27-branch/Source

Versioning.

6:28 PM Changeset in webkit [246053] by Chris Dumez

• 3 edits in trunk/Source/WebKit

[iOS] Do not prevent app suspension for more than 20 seconds after getting backgrounded
​https://bugs.webkit.org/show_bug.cgi?id=198488
<rdar://problem/50837208>

Reviewed by Geoff Garen.

Do not prevent app suspension for more than 20 seconds after getting backgrounded on iOS. We
do this by implementing our own expiration handler which notifies our child processes of
their imminent suspension before ending the background task that was preventing suspension.

• UIProcess/ios/ProcessAssertionIOS.mm:

(isBackgroundState):
(-[WKProcessAssertionBackgroundTaskManager init]):
(-[WKProcessAssertionBackgroundTaskManager _scheduleTimeoutTask]):
(-[WKProcessAssertionBackgroundTaskManager _cancelTimeoutTask]):
(-[WKProcessAssertionBackgroundTaskManager _backgroundTaskExpired]):
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
(-[WKProcessAssertionBackgroundTaskManager _releaseBackgroundTask]):

5:25 PM Changeset in webkit [246052] by rmorisset@apple.com

• 4 edits in trunk/Source/WebCore

[WHLSL] Parsing and lexing the standard library is slow
​https://bugs.webkit.org/show_bug.cgi?id=192890
<rdar://problem/50746335>

Reviewed by Myles Maxfield.

The main idea is to avoid backtracking by instead peeking at the next token (and occasionally at the one after that).
This implies a few things:

• We can replace the stack of tokens by a trivial ring buffer of size 2 (holding the next token and the one after, or WTF::nullopt if we are at the end of the file).
• We now have "completeFooExpression" functions, to avoid having to reparse the prefix of some expression, if we find half-way through what it is.

I also fixed the following parser bug:

• ​https://bugs.webkit.org/show_bug.cgi?id=198305 [WHLSL] Multiple variables with initializers in a declaration statement crashes the compiler

which was due to a mistake I made in the grammar

Finally I added two new macros: CONSUME_TYPE and PARSE to eliminate about 500 lines of error propagation boilerplate.

There are still lots of ways of improving the parser and lexer, such as:

• finishing the conversion of tokens in the lexer, not bothering with allocating string views
• make two special tokens Invalid and EOF, to remove the overhead of Optional
• make peekTypes and consumeTypes use templates to avoid constructing a Vector and calling find on it.
• Turn the entire lexer into a proper automata, not going through the same characters again and again (this is certainly the largest win by far)
• Remove the last few pieces of backtracking from the parser.

The current patch is already enough to make parsing the full standard library (something like 85k lines) approximately 260ms.
This is still longer than I would like, but nowhere near the bottleneck any longer because of some other parts of the compiler.

• Modules/webgpu/WHLSL/WHLSLLexer.h:

(WebCore::WHLSL::Lexer::Lexer):
(WebCore::WHLSL::Lexer::consumeToken):
(WebCore::WHLSL::Lexer::peek):
(WebCore::WHLSL::Lexer::peekFurther):
(WebCore::WHLSL::Lexer::state const):
(WebCore::WHLSL::Lexer::setState):
(WebCore::WHLSL::Lexer::unconsumeToken): Deleted.

• Modules/webgpu/WHLSL/WHLSLParser.cpp:

(WebCore::WHLSL::Parser::parse):
(WebCore::WHLSL::Parser::peek):
(WebCore::WHLSL::Parser::peekTypes):
(WebCore::WHLSL::Parser::tryType):
(WebCore::WHLSL::Parser::tryTypes):
(WebCore::WHLSL::Parser::consumeTypes):
(WebCore::WHLSL::Parser::parseConstantExpression):
(WebCore::WHLSL::Parser::parseTypeArgument):
(WebCore::WHLSL::Parser::parseTypeArguments):
(WebCore::WHLSL::Parser::parseTypeSuffixAbbreviated):
(WebCore::WHLSL::Parser::parseTypeSuffixNonAbbreviated):
(WebCore::WHLSL::Parser::parseType):
(WebCore::WHLSL::Parser::parseTypeDefinition):
(WebCore::WHLSL::Parser::parseResourceSemantic):
(WebCore::WHLSL::Parser::parseSpecializationConstantSemantic):
(WebCore::WHLSL::Parser::parseStageInOutSemantic):
(WebCore::WHLSL::Parser::parseSemantic):
(WebCore::WHLSL::Parser::parseQualifiers):
(WebCore::WHLSL::Parser::parseStructureElement):
(WebCore::WHLSL::Parser::parseStructureDefinition):
(WebCore::WHLSL::Parser::parseEnumerationDefinition):
(WebCore::WHLSL::Parser::parseEnumerationMember):
(WebCore::WHLSL::Parser::parseNativeTypeDeclaration):
(WebCore::WHLSL::Parser::parseNumThreadsFunctionAttribute):
(WebCore::WHLSL::Parser::parseAttributeBlock):
(WebCore::WHLSL::Parser::parseParameter):
(WebCore::WHLSL::Parser::parseParameters):
(WebCore::WHLSL::Parser::parseFunctionDefinition):
(WebCore::WHLSL::Parser::parseComputeFunctionDeclaration):
(WebCore::WHLSL::Parser::parseVertexFragmentFunctionDeclaration):
(WebCore::WHLSL::Parser::parseRegularFunctionDeclaration):
(WebCore::WHLSL::Parser::parseOperatorFunctionDeclaration):
(WebCore::WHLSL::Parser::parseFunctionDeclaration):
(WebCore::WHLSL::Parser::parseNativeFunctionDeclaration):
(WebCore::WHLSL::Parser::parseBlock):
(WebCore::WHLSL::Parser::parseBlockBody):
(WebCore::WHLSL::Parser::parseIfStatement):
(WebCore::WHLSL::Parser::parseSwitchStatement):
(WebCore::WHLSL::Parser::parseSwitchCase):
(WebCore::WHLSL::Parser::parseForLoop):
(WebCore::WHLSL::Parser::parseWhileLoop):
(WebCore::WHLSL::Parser::parseDoWhileLoop):
(WebCore::WHLSL::Parser::parseVariableDeclaration):
(WebCore::WHLSL::Parser::parseVariableDeclarations):
(WebCore::WHLSL::Parser::parseStatement):
(WebCore::WHLSL::Parser::parseEffectfulExpression):
(WebCore::WHLSL::Parser::parseEffectfulAssignment):
(WebCore::WHLSL::Parser::parseExpression):
(WebCore::WHLSL::Parser::parseTernaryConditional):
(WebCore::WHLSL::Parser::completeTernaryConditional):
(WebCore::WHLSL::Parser::parseAssignment):
(WebCore::WHLSL::Parser::completeAssignment):
(WebCore::WHLSL::Parser::parsePossibleTernaryConditional):
(WebCore::WHLSL::Parser::parsePossibleLogicalBinaryOperation):
(WebCore::WHLSL::Parser::completePossibleLogicalBinaryOperation):
(WebCore::WHLSL::Parser::parsePossibleRelationalBinaryOperation):
(WebCore::WHLSL::Parser::completePossibleRelationalBinaryOperation):
(WebCore::WHLSL::Parser::parsePossibleShift):
(WebCore::WHLSL::Parser::completePossibleShift):
(WebCore::WHLSL::Parser::parsePossibleAdd):
(WebCore::WHLSL::Parser::completePossibleAdd):
(WebCore::WHLSL::Parser::parsePossibleMultiply):
(WebCore::WHLSL::Parser::completePossibleMultiply):
(WebCore::WHLSL::Parser::parsePossiblePrefix):
(WebCore::WHLSL::Parser::parsePossibleSuffix):
(WebCore::WHLSL::Parser::parseCallExpression):
(WebCore::WHLSL::Parser::parseTerm):
(WebCore::WHLSL::Parser::parseAddressSpaceType): Deleted.
(WebCore::WHLSL::Parser::parseNonAddressSpaceType): Deleted.
(WebCore::WHLSL::Parser::parseEntryPointFunctionDeclaration): Deleted.
(WebCore::WHLSL::Parser::parseEffectfulPrefix): Deleted.
(WebCore::WHLSL::Parser::parseEffectfulSuffix): Deleted.

• Modules/webgpu/WHLSL/WHLSLParser.h:

(WebCore::WHLSL::Parser::Error::dump const):

2:59 PM Changeset in webkit [246051] by Alan Coon

• 1 copy in tags/Safari-608.1.27.1

Tag Safari-608.1.27.1.

2:10 PM Changeset in webkit [246050] by sihui_liu@apple.com

• 2 edits in trunk/Tools

[ Mac WK2 ] TestWebKitAPI.WKWebView.LocalStorageProcessCrashes is a flaky timeout when run locally.
​https://bugs.webkit.org/show_bug.cgi?id=198423
<rdar://problem/51305247>

Reviewed by Youenn Fablet.

IDBDatabase receives error event on network process crash, so we don't need to check it periodically.

• TestWebKitAPI/Tests/WebKitCocoa/local-storage-process-crashes.html:

2:07 PM Changeset in webkit [246049] by youenn@apple.com

• 7 edits
  2 adds in trunk

Allow resizing of camera video feeds to very small resolutions
​https://bugs.webkit.org/show_bug.cgi?id=198421

Reviewed by Alex Christensen.

Source/WebCore:

Before the patch, the minimum capture resolution was set to 120.
Since this is a bit arbitrary and there are some use cases for lower resolution,
We now allow down to 1x1 resolution.

Test: fast/mediastream/getUserMedia-video-rescaling.html

• platform/mediastream/RealtimeVideoSource.cpp:

(WebCore::RealtimeVideoSource::updateCapabilities):
Allow 1x1 when computing min/max resolutions.
(WebCore::RealtimeVideoSource::bestSupportedSizeAndFrameRate):
In case width and height are provided, we pick the closest preset that matches and resize accordingly.
If width or height is provided, we pick the closest preset tat matches and applies its aspect ratio.

LayoutTests:

Update tests to expect 1x1 minimum capture resolution.

• fast/mediastream/MediaStreamTrack-getCapabilities-expected.txt:
• fast/mediastream/apply-constraints-video-expected.txt:
• fast/mediastream/apply-constraints-video.html:
• fast/mediastream/get-user-media-constraints-expected.txt:
• fast/mediastream/getUserMedia-video-rescaling-expected.txt: Added.
• fast/mediastream/getUserMedia-video-rescaling.html: Added.

2:07 PM Changeset in webkit [246048] by Wenson Hsieh

• 12 edits
  1 copy in trunk/Source

Implement an internal switch to turn idempotent text autosizing and viewport rules off
​https://bugs.webkit.org/show_bug.cgi?id=198460
<rdar://problem/51324526>

Reviewed by Tim Horton.

Source/WebCore:

Add a new WebCore setting for viewport shrink-to-fit-content heuristics; additionally, tweak the existing
idempotent text autosizing setting to default to false (this is overridden by preferences at the WebKit layer).

• page/Settings.yaml:

Source/WebKit:

• Shared/WebPreferences.yaml:

Add new preferences to control viewport shrink-to-fit-content and idempotent text autosizing.

• Shared/WebPreferencesDefaultValues.cpp:

Add the non-iOS implementation of defaultTextAutosizingUsesIdempotentMode, and also wrap these functions in the
WebKit namespace.

(WebKit::defaultTextAutosizingUsesIdempotentMode):
(defaultPassiveTouchListenersAsDefaultOnDocument): Deleted.
(defaultCustomPasteboardDataEnabled): Deleted.

• Shared/WebPreferencesDefaultValues.h:

Move some existing default preference value helper functions into the WebKit namespace.

• Shared/ios/WebPreferencesDefaultValuesIOS.mm: Copied from Source/WebKit/Shared/WebPreferencesDefaultValues.cpp.

Add an iOS-specific file for WebPreferencesDefaultValues, so that we can check for the user interface idiom.

(WebKit::defaultTextAutosizingUsesIdempotentMode):

• SourcesCocoa.txt:
• WebKit.xcodeproj/project.pbxproj:
• WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::immediatelyShrinkToFitContent):

Add an early return for the case where shrink-to-fit-content is explicitly disabled via the new preference.

2:04 PM Changeset in webkit [246047] by Alan Coon

• 3 edits in branches/safari-607-branch/Source/WebCore

Revert r246001. rdar://problem/50871594

1:58 PM Changeset in webkit [246046] by Devin Rousso

• 6 edits in trunk

Web Inspector: remove RemoteObject.prototype.getPropertyDescriptorsAsObject
​https://bugs.webkit.org/show_bug.cgi?id=198395

Reviewed by Matt Baker.

Source/WebInspectorUI:

When calling RemoteObject.prototype.getPropertyDescriptorsAsObject, if one of the returned
property descriptors is named __proto__, it will replace the __proto__ of the plain
object created for that function with the WI.PropertyDescriptor itself, altering the
prototype chain.

• UserInterface/Protocol/RemoteObject.js:

(WI.RemoteObject.prototype.getPropertyDescriptorsAsObject): Deleted.

• UserInterface/Models/AuditTestCase.js:

(WI.AuditTestCase.prototype.async run.async parseResponse.async resultArrayForEach):

• UserInterface/Views/ErrorObjectView.js:

(WI.ErrorObjectView.prototype.update):

LayoutTests:

• inspector/runtime/awaitPromise.html:

1:49 PM Changeset in webkit [246045] by commit-queue@webkit.org

• 25 edits
  32 adds in trunk

Implement imagesrcset and imagesizes attributes on link rel=preload
​https://bugs.webkit.org/show_bug.cgi?id=192950

Patch by Rob Buis <​rbuis@igalia.com> on 2019-06-03
Reviewed by Youenn Fablet.

LayoutTests/imported/w3c:

Import relevant tests for this feature.

• web-platform-tests/preload/dynamic-adding-preload-imagesrcset-expected.txt: Added.
• web-platform-tests/preload/dynamic-adding-preload-imagesrcset.html: Added.
• web-platform-tests/preload/link-header-preload-expected.txt: Added.
• web-platform-tests/preload/link-header-preload-imagesrcset-expected.txt: Added.
• web-platform-tests/preload/link-header-preload-imagesrcset.html: Added.
• web-platform-tests/preload/link-header-preload-imagesrcset.html.headers: Added.
• web-platform-tests/preload/link-header-preload-nonce-expected.txt: Added.
• web-platform-tests/preload/link-header-preload-nonce.html: Added.
• web-platform-tests/preload/link-header-preload-nonce.html.headers: Added.
• web-platform-tests/preload/link-header-preload.html: Added.
• web-platform-tests/preload/link-header-preload.html.headers: Added.
• web-platform-tests/preload/onload-event-expected.txt: Added.
• web-platform-tests/preload/onload-event.html: Added.
• web-platform-tests/preload/preload-csp.sub-expected.txt:
• web-platform-tests/preload/preload-csp.sub.html:
• web-platform-tests/preload/preload-default-csp.sub-expected.txt:
• web-platform-tests/preload/preload-default-csp.sub.html:
• web-platform-tests/preload/preload-with-type-expected.txt: Added.
• web-platform-tests/preload/preload-with-type.html: Added.
• web-platform-tests/preload/resources/A4.ogv: Added.
• web-platform-tests/preload/resources/A4.ogv.sub.headers: Added.
• web-platform-tests/preload/resources/cross-origin-module.py: Added.

(main):

• web-platform-tests/preload/resources/dummy-preloads-subresource.css: Added.
• web-platform-tests/preload/resources/dummy-preloads-subresource.css.sub.headers: Added.
• web-platform-tests/preload/resources/empty.html: Added.
• web-platform-tests/preload/resources/empty.html.sub.headers: Added.
• web-platform-tests/preload/resources/foo.vtt.sub.headers: Added.
• web-platform-tests/preload/resources/module1.js: Added.
• web-platform-tests/preload/resources/module2.js: Added.
• web-platform-tests/preload/resources/preload_helper.js:

(verifyNumberOfResourceTimingEntries):
(numberOfResourceTimingEntries):

Source/WebCore:

Implement imagesrcset and imagesizes attributes for both Link header
and link element.

Tests: imported/w3c/web-platform-tests/preload/dynamic-adding-preload-imagesrcset.html

imported/w3c/web-platform-tests/preload/link-header-preload-imagesrcset.html
imported/w3c/web-platform-tests/preload/link-header-preload-nonce.html
imported/w3c/web-platform-tests/preload/link-header-preload.html
imported/w3c/web-platform-tests/preload/onload-event.html
imported/w3c/web-platform-tests/preload/preload-with-type.html

• html/HTMLAttributeNames.in:
• html/HTMLLinkElement.cpp:

(WebCore::HTMLLinkElement::process):

• html/HTMLLinkElement.idl:
• loader/LinkHeader.cpp:

(WebCore::paramterNameFromString):
(WebCore::LinkHeader::setValue):
(WebCore::LinkHeader::LinkHeader):

• loader/LinkHeader.h:

(WebCore::LinkHeader::imageSrcSet const):
(WebCore::LinkHeader::imageSizes const):
(WebCore::LinkHeader::isViewportDependent const):

• loader/LinkLoader.cpp:

(WebCore::LinkLoader::loadLinksFromHeader):
(WebCore::LinkLoader::preloadIfNeeded):
(WebCore::LinkLoader::loadLink):

• loader/LinkLoader.h:
• page/RuntimeEnabledFeatures.h:

(WebCore::RuntimeEnabledFeatures::linkPreloadResponsiveImagesEnabled const):
(WebCore::RuntimeEnabledFeatures::setLinkPreloadResponsiveImagesEnabled):

Source/WebKit:

Add a runtime flag for link preload responsive images.

• Shared/WebPreferences.yaml:

Source/WebKitLegacy/mac:

Add a runtime flag for link preload responsive images.

• WebView/WebPreferenceKeysPrivate.h:
• WebView/WebPreferences.mm:

(+[WebPreferences initialize]):
(-[WebPreferences linkPreloadResponsiveImagesEnabled]):
(-[WebPreferences setLinkPreloadResponsiveImagesEnabled:]):

• WebView/WebPreferencesPrivate.h:
• WebView/WebView.mm:

(-[WebView _preferencesChanged:]):

Tools:

Add a runtime flag for link preload responsive images.

• DumpRenderTree/mac/DumpRenderTree.mm:

(enableExperimentalFeatures):

LayoutTests:

Write special expectation for link-header-preload-imagesrcset.html because test runner
uses different dimensions on iOS.

• platform/ios-simulator-12-wk2/imported/w3c/web-platform-tests/preload/dynamic-adding-preload-imagesrcset-expected.txt: Added.
• platform/ios-simulator-12-wk2/imported/w3c/web-platform-tests/preload/link-header-preload-imagesrcset-expected.txt: Added.

1:12 PM Changeset in webkit [246044] by don.olmstead@sony.com

• 6 edits
  1 add in trunk/Tools

[CMake] Use WEBKIT_EXECUTABLE in ImageDiff
​https://bugs.webkit.org/show_bug.cgi?id=198490

Reviewed by Konstantin Tokarev.

Use WEBKIT_EXECUTABLE in ImageDiff. Change CMake variables to use the proper
convention. Add a shared CoreGraphics.cmake instead of just including PlatformMac
for AppleWin builds.

• ImageDiff/CMakeLists.txt:
• ImageDiff/Cairo.cmake:
• ImageDiff/CoreGraphics.cmake: Added.
• ImageDiff/PlatformGTK.cmake:
• ImageDiff/PlatformMac.cmake:
• ImageDiff/PlatformWin.cmake:

12:39 PM Changeset in webkit [246043] by aestes@apple.com

• 6 edits in trunk

Memory-cached main resources continue to load after the client decides a content policy of PolicyAction::Download
​https://bugs.webkit.org/show_bug.cgi?id=198469
<rdar://problem/50512713>

Reviewed by Youenn Fablet.

Source/WebCore:

When a document is loaded from the memory cache it does not have a main resource loader, but
DocumentLoader::continueAfterContentPolicy relies on being able to call
ResourceLoader::didFail on the main resource loader to cancel the provisional navigation
when the client decides a content policy of PolicyAction::Download.

This means that memory-cached main resources continue to load even after WebKit has started
to download the main resource. The expected behavior is for the provisional navigation to
fail once the download starts, like what happens when there is a main resource loader.

This patch teaches DocumentLoader::continueAfterContentPolicy to call
stopLoadingForPolicyChange() in the case of a null main resource loader. This will dispatch
didFailProvisionalNavigation and remove the DocumentLoader as a client of its
CachedRawResource to prevent it from delivering any cached data.

Added a new API test.

• loader/DocumentLoader.cpp:

(WebCore::DocumentLoader::continueAfterContentPolicy):

Tools:

• TestWebKitAPI/Tests/WebKitCocoa/Download.mm:

(-[TestDownloadNavigationResponseFromMemoryCacheDelegate webView:didStartProvisionalNavigation:]):
(-[TestDownloadNavigationResponseFromMemoryCacheDelegate webView:didFailProvisionalNavigation:withError:]):
(-[TestDownloadNavigationResponseFromMemoryCacheDelegate webView:didFinishNavigation:]):
(-[TestDownloadNavigationResponseFromMemoryCacheDelegate _downloadDidStart:]):
(-[TestDownloadNavigationResponseFromMemoryCacheDelegate webView:decidePolicyForNavigationResponse:decisionHandler:]):
(TEST):

• TestWebKitAPI/cocoa/TestProtocol.h:
• TestWebKitAPI/cocoa/TestProtocol.mm:

(+[TestProtocol additionalResponseHeaders]):
(+[TestProtocol setAdditionalResponseHeaders:]):
(-[TestProtocol startLoading]):

12:14 PM Changeset in webkit [246042] by timothy@apple.com

• 5 edits in trunk

Tweak the text and underline color for data detected text.
​https://bugs.webkit.org/show_bug.cgi?id=198487
rdar://problem/50667125

Reviewed by Devin Rousso.

Source/WebCore:

Tests: Color.RGBToHSL API tests

• editing/cocoa/DataDetection.mm:

(WebCore::DataDetection::detectContentInRange): Use currentcolor so semantic text colors work.
Force the lightness of the underline color to the middle, and multiply the alpha by 38%,
so the color will appear on light and dark backgrounds, since only one color can be specified.

• platform/graphics/Color.cpp:

(WebCore::Color::getHSL const): Return hue in [0...6) range to easily round-trip with makeRGBAFromHSLA().

Tools:

• TestWebKitAPI/Tests/WebCore/Color.cpp:

(TestWebKitAPI::TEST): Added Color.RGBToHSL tests.

11:42 AM Changeset in webkit [246041] by Caio Lima

• 22 edits
  9 adds in trunk

[ESNext][BigInt] Implement support for ""
​https://bugs.webkit.org/show_bug.cgi?id=190799

Reviewed by Saam Barati.

JSTests:

• stress/big-int-exp-basic.js: Added.
• stress/big-int-exp-jit-osr.js: Added.
• stress/big-int-exp-jit-untyped.js: Added.
• stress/big-int-exp-jit.js: Added.
• stress/big-int-exp-negative-exponent.js: Added.
• stress/big-int-exp-to-primitive.js: Added.
• stress/big-int-exp-type-error.js: Added.
• stress/big-int-exp-wrapped-value.js: Added.
• stress/value-pow-ai-rule.js: Added.

Source/JavaScriptCore:

We are introducing support for BigInt into "" operator. This Patch
also includes changes into DFG, introducing a new node "ValuePow" that
is responsible to handle UntypedUse and BigIntUse.

• dfg/DFGAbstractInterpreterInlines.h:

(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

ValuePow(Untyped, Untyped) still can propagate constant if AI proves
it. We are doing so if AI proves rhs and lhs as numbers.

• dfg/DFGByteCodeParser.cpp:

(JSC::DFG::ByteCodeParser::parseBlock):

When compiling op_pow, we first verify if rhs and lhs can be any Int
or number. If this happen, we emit ArithPow, otherwise we fallback to
ValuePow and rely on fixup to convert it to ArithPow if possible.

• dfg/DFGClobberize.h:

(JSC::DFG::clobberize):

We only clobberize world if ValuePow is UntypedUse. Otherwise, we can
properly support CSE.

• dfg/DFGDoesGC.cpp:

(JSC::DFG::doesGC):

JSBigInt::exponentiate allocates JSBigInts to perform calculation and
it can trigger GC. ValuePow(UntypedUse) can trigger GC because it can
execute user code.

• dfg/DFGFixupPhase.cpp:

(JSC::DFG::FixupPhase::fixupArithPow):
(JSC::DFG::FixupPhase::fixupNode):

• dfg/DFGNodeType.h:
• dfg/DFGOperations.cpp:
• dfg/DFGOperations.h:
• dfg/DFGPredictionPropagationPhase.cpp:
• dfg/DFGSafeToExecute.h:

(JSC::DFG::safeToExecute):

• dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::compileValuePow):

• dfg/DFGSpeculativeJIT.h:
• dfg/DFGSpeculativeJIT32_64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

• dfg/DFGSpeculativeJIT64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

• dfg/DFGValidate.cpp:
• ftl/FTLCapabilities.cpp:

(JSC::FTL::canCompile):

• ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileValuePow):

• runtime/CommonSlowPaths.cpp:

(JSC::SLOW_PATH_DECL):

We are adding proper support to BigInt on op_pow. The specification
defines that we can only apply pow when both operands have the same
type after calling ToNumeric().

• runtime/JSBigInt.cpp:

(JSC::JSBigInt::exponentiate):

• runtime/JSBigInt.h:

11:27 AM Changeset in webkit [246040] by ysuzuki@apple.com

• 4 edits
  3 adds in trunk

[JSC] JSObject::attemptToInterceptPutByIndexOnHole should use getPrototype instead of getPrototypeDirect
​https://bugs.webkit.org/show_bug.cgi?id=198477
<rdar://problem/51299504>

Reviewed by Saam Barati.

Source/JavaScriptCore:

JSObject::attemptToInterceptPutByIndexOnHole uses getPrototypeDirect, but it should use getPrototype to
handle getPrototype methods in derived JSObject classes correctly.

• runtime/JSArrayInlines.h:

(JSC::JSArray::pushInline):

• runtime/JSObject.cpp:

(JSC::JSObject::putByIndex):
(JSC::JSObject::attemptToInterceptPutByIndexOnHoleForPrototype):
(JSC::JSObject::attemptToInterceptPutByIndexOnHole):
(JSC::JSObject::putByIndexBeyondVectorLength):

LayoutTests:

Ensure that JSWindow::getPrototype is used.

• http/tests/security/cross-frame-access-object-getPrototypeOf-in-put-expected.txt: Added.
• http/tests/security/cross-frame-access-object-getPrototypeOf-in-put.html: Added.
• http/tests/security/resources/cross-frame-iframe-for-object-getPrototypeOf-in-put-test.html: Added.

10:04 AM Changeset in webkit [246039] by don.olmstead@sony.com

• 17 edits in trunk

[CMake] Add WebKit::JavaScriptCore target
​https://bugs.webkit.org/show_bug.cgi?id=198403

Reviewed by Konstantin Tokarev.

Source/JavaScriptCore:

Create the WebKit::JavaScriptCore target and use that to propagate headers. Use
WEBKIT_COPY_FILES instead of WEBKIT_MAKE_FORWARDING_HEADERS.

• CMakeLists.txt:
• shell/CMakeLists.txt:

Source/WebCore:

Use the WebKit::JavaScriptCore target.

• CMakeLists.txt:

Source/WebCore/PAL:

Use the WebKit::JavaScriptCore target.

• pal/CMakeLists.txt:

Source/WebKitLegacy:

Use the WebKit::JavaScriptCore target.

• CMakeLists.txt:

Tools:

Use the WebKit::JavaScriptCore target.

• DumpRenderTree/CMakeLists.txt:
• MiniBrowser/win/CMakeLists.txt:
• TestWebKitAPI/CMakeLists.txt:
• TestWebKitAPI/PlatformGTK.cmake:
• TestWebKitAPI/PlatformWPE.cmake:
• TestWebKitAPI/glib/CMakeLists.txt:
• WebKitTestRunner/CMakeLists.txt:

9:48 AM Changeset in webkit [246038] by commit-queue@webkit.org

• 9 edits in trunk/Source

Unreviewed, rolling out r246022.
​https://bugs.webkit.org/show_bug.cgi?id=198486

Causing Internal build failures and JSC test failures
(Requested by ShawnRoberts on #webkit).

Reverted changeset:

"Reenable Gigacage on ARM64."
​https://bugs.webkit.org/show_bug.cgi?id=198453
​https://trac.webkit.org/changeset/246022

9:37 AM Changeset in webkit [246037] by Devin Rousso

• 2 edits in trunk/LayoutTests

Flaky Test: inspector/canvas/recording.html
​https://bugs.webkit.org/show_bug.cgi?id=198470

Reviewed by Timothy Hatcher.

• inspector/canvas/recording.html:

The ordering of <canvas> in WI.canvasManager.canvases isn't always the same, so we should
perform the action on every <canvas> since we know we'll be recording at least one of them.

8:34 AM Changeset in webkit [246036] by zandobersek@gmail.com

• 2 edits in trunk/Source/WebCore

[Nicosia] Disable async scrolling until implemented
​https://bugs.webkit.org/show_bug.cgi?id=198476
<rdar://problem/51351478>

Unreviewed follow-up to r246033.

• page/scrolling/ScrollingCoordinator.cpp:

Expand the PLATFORM(IOS) guard to IOS_FAMILY.

7:28 AM Changeset in webkit [246035] by berto@igalia.com

• 3 edits in trunk/Source/WebDriver

REGRESSION(r245567): [GTK][WPE] [2.25.1] WebKitWebDriver gets installed as WebDriver
​https://bugs.webkit.org/show_bug.cgi?id=198480

Patch by Carlos Garcia Campos <​cgarcia@igalia.com> on 2019-06-03
Reviewed by Michael Catanzaro.

We are now using the WEBKIT_EXECUTABLE macro that uses target_OUTPUT_NAME variable, but we define
taget_Process_OUTPUT_NAME.

• PlatformGTK.cmake:
• PlatformWPE.cmake:

4:34 AM Changeset in webkit [246034] by Darin Adler

• 30 edits in trunk

Finish cleanup of String::number for floating point
​https://bugs.webkit.org/show_bug.cgi?id=198471

Reviewed by Yusuke Suzuki.

Source/JavaScriptCore:

• dfg/DFGStrengthReductionPhase.cpp:

(JSC::DFG::StrengthReductionPhase::handleNode): Use String::number instead of
String::numberToStringECMAScript, since that's now the default.

• parser/ParserArena.h:

(JSC::IdentifierArena::makeNumericIdentifier): Ditto.

• runtime/JSONObject.cpp:

(JSC::Stringifier::appendStringifiedValue): Use appendNumber instead of
builder.appendECMAScriptNumber, since that's now the default.

• runtime/NumberPrototype.cpp:

(JSC::toStringWithRadix): Use String::number instead of
String::numberToStringECMAScript, since that's now the default.
(JSC::numberProtoFuncToExponential): Ditto.
(JSC::numberProtoFuncToFixed): Ditto.
(JSC::numberProtoFuncToPrecision): Ditto.
(JSC::numberToStringInternal): Ditto.

• runtime/NumericStrings.h:

(JSC::NumericStrings::add): Ditto.

• wasm/WasmBBQPlan.cpp:

(JSC::Wasm::BBQPlan::prepare): Ditto.

Source/WebCore:

• css/CSSCursorImageValue.cpp:

(WebCore::CSSCursorImageValue::customCSSText const): Use makeString instead
of StringBuilder since it is more terse and possibly more efficient as well.

• css/DOMMatrixReadOnly.cpp:

(WebCore::DOMMatrixReadOnly::toString const): Ditto.

• css/WebKitCSSMatrix.cpp:

(WebCore::WebKitCSSMatrix::toString const): Ditto.

• html/parser/HTMLParserIdioms.cpp:

(WebCore::serializeForNumberType): Use String::number instead of
String::numberToStringECMAScript since that is now the default.

• html/shadow/MediaControlElementTypes.cpp:

(WebCore::MediaControlVolumeSliderElement::setVolume): Ditto.

• html/shadow/MediaControlElements.cpp:

(WebCore::MediaControlTimelineElement::setPosition): Ditto.

• platform/Decimal.cpp:

(WebCore::Decimal::fromDouble): Ditto.

• rendering/RenderListMarker.cpp:

(WebCore::toRoman): Return a String instead of taking a StringBuilder argument.
Use LetterCase enum instead of bool for the uppercase vs. lowercase argument, to
be less mysterious at call sites. Use unsigned for string lengths since that's
what WTF::String uses.
(WebCore::toAlphabeticOrNumeric): Ditto. Also updated since SequenceType is now
an enum class.
(WebCore::toSymbolic): Use String::createUninitialized instead of StringBuilder.
Return a String instead of taking a StringBuilder argument. Straighten out the
algorithm so it's easier to understand.
(WebCore::toAlphabetic): Updated since SequenceType is now an enum class.
(WebCore::toNumeric): Ditto.
(WebCore::toHebrew): Return a String instead of taking a StringBuilder argument.
Use unsigned for string lengths since that's what WTF::String uses.
(WebCore::toArmenianUnder10000): Use unsigned and the LetterCase enum class.
(WebCore::toArmenian): Return a String instead of taking a StringBuilder argument.
Also use unsigned for string lengths since that's what WTF::String uses.
(WebCore::toGeorgian): Ditto.
(WebCore::toCJKIdeographic): Ditto.
(WebCore::listMarkerSuffix): Tweaked a comment.
(WebCore::listMarkerText): Use String return values instead of StringBuilder.
(WebCore::RenderListMarker::paint): Use String::createUninitialized instead of
StringBuilder since it's a bit more efficient. Use makeString instead of turning
single characters into strings since that's more efficient.

• svg/SVGPathUtilities.cpp:

(WebCore::buildStringFromPath): Use appendNumber instead of appendShortestFormNumber
since that is now the default.

Source/WebKit:

• NetworkProcess/Classifier/WebResourceLoadStatisticsTelemetry.cpp:

(WebKit::submitTopList): Use makeString instead of StringBuilder since it
is more terse and possibly more efficient as well.

• UIProcess/API/APIUserScript.cpp:

(API::UserScript::generateUniqueURL): Ditto.

• UIProcess/API/APIUserStyleSheet.cpp:

(API::UserStyleSheet::generateUniqueURL): Ditto.

Source/WTF:

• wtf/JSONValues.cpp:

(WTF::JSONImpl::Value::writeJSON const): Use appendNumber instead of
appendECMAScriptNumber, since that's now the default.

• wtf/text/StringBuilder.cpp:

(WTF::StringBuilder::appendNumber): Renamed from appendShortestFormNumber.

• wtf/text/StringBuilder.h: Removed appendShortestFormNumber and

appendECMAScriptNumber, replacing both with overload of appendNumber.

• wtf/text/WTFString.cpp:

(WTF::String::number): Renamed from numberToStringShortest.

• wtf/text/WTFString.h: Removed numberToStringShortest and

numberToStringECMAScript, replacing both with overload of number.

Tools:

• TestWebKitAPI/Tests/WTF/WTFString.cpp:

(TestWebKitAPI::testStringNumber): Go back to calling String::number instead of
String::numberToStringECMAScript.

• WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:

(WTR::dumpFrameScrollPosition): Use appendNumber instead of appendECMAScriptNumber
since that is now the default.

2:45 AM Changeset in webkit [246033] by zandobersek@gmail.com

• 24 edits in trunk

[Nicosia] Disable async scrolling until implemented
​https://bugs.webkit.org/show_bug.cgi?id=198476

Reviewed by Antti Koivisto.

.:

Disable async scrolling on Nicosia-using ports until the underlying
implementation is complete, allowing easier refactorings of the general
async scrolling system.

• Source/cmake/OptionsGTK.cmake:
• Source/cmake/OptionsPlayStation.cmake:
• Source/cmake/OptionsWPE.cmake:

Source/WebCore:

Build ScrollingCoordinator::create() method on iOS or whenever
ASYNC_SCROLLING is not enabled.

Additionally, the USE(COORDINATED_GRAPHICS) build guards are removed
throughout the async scrolling classes as they aren't required now
and shouldn't be in the future.

• page/scrolling/ScrollingCoordinator.cpp:
• page/scrolling/ScrollingStateFixedNode.cpp:
• page/scrolling/ScrollingStateFixedNode.h:
• page/scrolling/ScrollingStateFrameHostingNode.cpp:
• page/scrolling/ScrollingStateFrameHostingNode.h:
• page/scrolling/ScrollingStateFrameScrollingNode.cpp:
• page/scrolling/ScrollingStateFrameScrollingNode.h:
• page/scrolling/ScrollingStateNode.cpp:
• page/scrolling/ScrollingStateNode.h:
• page/scrolling/ScrollingStateOverflowScrollingNode.cpp:
• page/scrolling/ScrollingStateOverflowScrollingNode.h:
• page/scrolling/ScrollingStatePositionedNode.cpp:
• page/scrolling/ScrollingStatePositionedNode.h:
• page/scrolling/ScrollingStateScrollingNode.cpp:
• page/scrolling/ScrollingStateScrollingNode.h:
• page/scrolling/ScrollingStateStickyNode.cpp:
• page/scrolling/ScrollingStateStickyNode.h:
• page/scrolling/ScrollingStateTree.cpp:
• page/scrolling/ScrollingStateTree.h:

2:43 AM Changeset in webkit [246032] by Fujii Hironori

• 2 edits in trunk/Source/WebKit

REGRESSION(r240862) [WinCairo][WK2] position:fixed elements don't show
​https://bugs.webkit.org/show_bug.cgi?id=198478

Reviewed by Carlos Garcia Campos.

All position:fixed elements didn't show in WinCairo WebKit2 since
r240862 which turns acceleratedCompositingEnabled on for WebKit2.
WinCairo WebKit2 doesn't have accelerated compositing support yet.
Let's do the same with GTK port.

• WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::updatePreferences): Conditioned out the code
enabling acceleratedCompositingEnabled for PLATFORM(WIN).

2:29 AM Changeset in webkit [246031] by graouts@webkit.org

• 2 edits in trunk/Source/WebCore

[Pointer Events] Check that capturing data managed by the PointerCaptureController gets cleared upon navigation
​https://bugs.webkit.org/show_bug.cgi?id=198191

Unreviewed post-commit review feedback suggested by Darin Adler.

• page/PointerCaptureController.cpp:

(WebCore::PointerCaptureController::reset):

1:07 AM Changeset in webkit [246030] by Carlos Garcia Campos

• 3 edits in trunk/Source/WebKit

REGRESSION(r244182): [CoordinatedGraphics] Related view doesn't receive graphic updates
​https://bugs.webkit.org/show_bug.cgi?id=198418

Reviewed by Žan Doberšek.

This is because we need a DisplayRefreshMonitor per drawing area, but the one from the first ThreradedCompositor
is always reused because they use the same display ID.

• WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:

(WebKit::DrawingAreaCoordinatedGraphics::enterAcceleratedCompositingMode): Use a unique display ID derived from
WebPage's ID.
(WebKit::DrawingAreaCoordinatedGraphics::exitAcceleratedCompositingMode): Use the primary display ID.

• WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:

(WebKit::LayerTreeHost::LayerTreeHost): Get the current web page display ID instead of inventing one.
(WebKit::LayerTreeHost::setIsDiscardable): Do not change display ID.