WebKit-https.git
21 months agoAX: Treat AXChildrenInNavigationOrder as AXChildren before adding support for aria...
commit-queue@webkit.org [Sat, 23 Feb 2019 04:31:33 +0000 (04:31 +0000)]
AX: Treat AXChildrenInNavigationOrder as AXChildren before adding support for aria-flowto
https://bugs.webkit.org/show_bug.cgi?id=194923

Patch by Eric Liang <ericliang@apple.com> on 2019-02-22
Reviewed by Chris Fleizach.

Source/WebCore:

Added AXChildrenInNavigationOrder attribute that returns the same array as from AXChildren. It prevents AppKit from reordering elements from the fallback AXChildren attribute.

Test: accessibility/mac/children-in-navigation-order-returns-children.html

* accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
(-[WebAccessibilityObjectWrapper accessibilityAttributeValue:]):

LayoutTests:

Added test to check AXChildrenInNavigationOrder returns non-empty array from AXChildren.
* accessibility/mac/children-in-navigation-order-returns-children-expected.txt: Added.
* accessibility/mac/children-in-navigation-order-returns-children.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: Timelines: add UI for preventing auto-stop
drousso@apple.com [Sat, 23 Feb 2019 04:17:58 +0000 (04:17 +0000)]
Web Inspector: Timelines: add UI for preventing auto-stop
https://bugs.webkit.org/show_bug.cgi?id=194956

Reviewed by Joseph Pecoraro.

Add a checkbox to the navigation area of the Timelines tab that controls whether recordings
automatically stop (e.g. after "load" or a period of inactivity).

* UserInterface/Views/TimelineRecordingContentView.js:
(WI.TimelineRecordingContentView):
(WI.TimelineRecordingContentView.prototype.get navigationItems):
(WI.TimelineRecordingContentView.prototype._handleAutoStopCheckboxCheckedDidChange): Added.
(WI.TimelineRecordingContentView.prototype._handleTimelinesAutoStopSettingChanged): Added.

* UserInterface/Controllers/TimelineManager.js:
(WI.TimelineManager):
(WI.TimelineManager.prototype.capturingStopped):
(WI.TimelineManager.prototype._stopAutoRecordingSoon):
(WI.TimelineManager.prototype._resetAutoRecordingMaxTimeTimeout):
(WI.TimelineManager.prototype._resetAutoRecordingDeadTimeTimeout):
(WI.TimelineManager.prototype._handleTimelinesAutoStopSettingChanged):

* UserInterface/Base/Setting.js:
* Localizations/en.lproj/localizedStrings.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoREGRESSION(r?): Web Inspector: Popovers have inset shadows
drousso@apple.com [Sat, 23 Feb 2019 04:10:00 +0000 (04:10 +0000)]
REGRESSION(r?): Web Inspector: Popovers have inset shadows
https://bugs.webkit.org/show_bug.cgi?id=192963
<rdar://problem/46888679>

Reviewed by Brian Burg.

* UserInterface/Views/Popover.js:
(WI.Popover.prototype._drawBackground):
(WI.Popover.prototype._drawBackground.isolate): Added.
* UserInterface/Views/Popover.css:
(.popover):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: Styles Redesign: clicking CSS property or selector should always selec...
drousso@apple.com [Sat, 23 Feb 2019 04:07:58 +0000 (04:07 +0000)]
Web Inspector: Styles Redesign: clicking CSS property or selector should always select its text
https://bugs.webkit.org/show_bug.cgi?id=180791
<rdar://problem/36038366>

Reviewed by Brian Burg.

* UserInterface/Views/SpreadsheetSelectorField.js:
(WI.SpreadsheetSelectorField):
(WI.SpreadsheetSelectorField.prototype.startEditing):
(WI.SpreadsheetSelectorField.prototype.stopEditing):
(WI.SpreadsheetSelectorField.prototype._handleMouseDown): Added.
(WI.SpreadsheetSelectorField.prototype._handleMouseUp): Added.
(WI.SpreadsheetSelectorField.prototype._handleFocus): Deleted.

* UserInterface/Views/SpreadsheetCSSStyleDeclarationSection.js:
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.initialLayout):
* UserInterface/Views/SpreadsheetCSSStyleDeclarationSection.css:
(.spreadsheet-css-declaration .selector.editing:focus, .spreadsheet-css-declaration .selector > .matched): Added.
(.spreadsheet-css-declaration .selector.spreadsheet-selector-field):
(.spreadsheet-css-declaration .selector:focus, .spreadsheet-css-declaration .selector > .matched): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[WinCairo] Enable wk1/wk2 suffix for platform search path.
basuke.suzuki@sony.com [Sat, 23 Feb 2019 04:05:55 +0000 (04:05 +0000)]
[WinCairo] Enable wk1/wk2 suffix for platform search path.
https://bugs.webkit.org/show_bug.cgi?id=194846

Reviewed by Don Olmstead.

Tools:

Added _search_paths() and _port_specific_expectations_files() for
WinCairoPort.

* Scripts/webkitpy/port/win.py:
(WinCairoPort.default_baseline_search_path):
(WinCairoPort):
(WinCairoPort._port_specific_expectations_files):
(WinCairoPort._search_paths):

LayoutTests:

Added WebKitLegacy specific TestExpectations.

* platform/wincairo-wk1/TestExpectations: Added.
* platform/wincairo/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoProcessSwap.PageOverlayLayerPersistence fails on iOS and in debug builds
timothy_horton@apple.com [Sat, 23 Feb 2019 02:38:02 +0000 (02:38 +0000)]
ProcessSwap.PageOverlayLayerPersistence fails on iOS and in debug builds
https://bugs.webkit.org/show_bug.cgi?id=194963

Reviewed by Dean Jackson.

Source/WebCore:

Tested by existing failing API test.

* page/Page.cpp:
(WebCore::Page::installedPageOverlaysChanged): Deleted.
* page/Page.h:
(WebCore::Page::pageOverlayController):
* page/PageOverlayController.cpp:
(WebCore::PageOverlayController::installedPageOverlaysChanged):
(WebCore::PageOverlayController::detachViewOverlayLayers):
(WebCore::PageOverlayController::installPageOverlay):
(WebCore::PageOverlayController::uninstallPageOverlay):
(WebCore::PageOverlayController::willDetachRootLayer): Deleted.
* page/PageOverlayController.h:
As intended by r240940, move installedPageOverlaysChanged to PageOverlayController.
Also, make it ignore isInWindow state; otherwise, if you install a overlay
and then come into window, nothing installs the root layer. There is no
need for this code to follow in-window state manually anymore since
the DrawingArea and RenderLayerCompositor just hook the layers up when needed.

Make some methods private, and make detachViewOverlayLayers only touch
*view* overlays, so that we don't detach the document-relative root
layer when you drop to having no view overlays. This maintains
existing behavior because nothing was calling PageOverlayController::detachViewOverlayLayers.

Now there are no callers of willDetachRootLayer, so remove it.

Tools:

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
Do a `contains` check instead of `equals`, because in debug builds we
put the GraphicsLayer pointer in a prefix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRotation animations sometimes use the wrong origin (affects apple.com)
dino@apple.com [Sat, 23 Feb 2019 02:10:58 +0000 (02:10 +0000)]
Rotation animations sometimes use the wrong origin (affects apple.com)
https://bugs.webkit.org/show_bug.cgi?id=194878
<rdar://problem/43908047>

Reviewed by Simon Fraser.

Forgot another expected results file.

* legacy-animation-engine/animations/additive-transform-animations-expected.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: Dark Mode: DOM Content Loaded blue marker is too dark
nvasilyev@apple.com [Sat, 23 Feb 2019 01:51:15 +0000 (01:51 +0000)]
Web Inspector: Dark Mode: DOM Content Loaded blue marker is too dark
https://bugs.webkit.org/show_bug.cgi?id=194970
<rdar://problem/48330126>

Reviewed by Joseph Pecoraro.

* UserInterface/Views/TimelineRuler.css:
(@media (prefers-color-scheme: dark)):
(.timeline-ruler > .markers > .marker.dom-content-event):
Make the marker lighter for the dark mode.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoSkip the test added in r241932 on iOS for now.
rniwa@webkit.org [Sat, 23 Feb 2019 01:03:55 +0000 (01:03 +0000)]
Skip the test added in r241932 on iOS for now.

* platform/ios/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[iOS] Callout menu overlaps in-page controls when editing a comment in github.com...
wenson_hsieh@apple.com [Sat, 23 Feb 2019 00:48:16 +0000 (00:48 +0000)]
[iOS] Callout menu overlaps in-page controls when editing a comment in github.com's issue tracker
https://bugs.webkit.org/show_bug.cgi?id=194873
<rdar://problem/46701974>

Reviewed by Tim Horton.

Source/WebKit:

On the topic of supporting web-based rich text editors on iOS, one problematic area has always been handling
conflicts between platform UI (i.e., the system callout menu) and in-page text editing controls. This issue
comes up in websites that don't use the "hidden contenteditable" approach to rich text editing, but also show
additional controls in a toolbar or contextual menu above the selection. In these cases, what often happens is
that system controls overlap controls in the page.

Luckily, the iOS callout menu (i.e. the private UICalloutBar) is capable of presenting with a list of "evasion
rects" to avoid; if the callout bar would normally intersect with one of these rects, then a different
orientation that does not intersect with one of these rects is chosen instead. Currently, the only rect added
here by UIKit when presenting the callout menu is the bounding rect of the on-screen keyboard, but after
<rdar://problem/48128337>, we now have a generalized mechanism for offering additional evasion rects before
UIKit presents the callout menu.

This patch adopts the mechanism introduced in <rdar://problem/48128337>, and introduces a heuristic for
determining the approximate location of controls in the page which might overlap the callout menu. This
heuristic works by hit-testing for clickable (but non-editable) nodes above the bounds of the selection, which
are additionally not hit-tested by advancing outwards from any of the other edges of the selection bounds.
Additionally, any hit-tested nodes whose bounding rects are very large (relative to the content view size) are
ignored (this deals with scenarios where the body or a large container element has a click handler). We then add
the bounding rects of each of the nodes that fit this criteria to the list of rects for UIKit to avoid when
presenting the system callout menu.

The result is that WebKit will, by default, avoid overlapping anything that looks like controls in the page when
showing a callout menu in editable content. In practice, this fixes overlapping controls on most websites that
roll their own context menu or toolbar in their rich text editor.

Test: editing/selection/ios/avoid-showing-callout-menu-over-controls.html

* Platform/spi/ios/UIKitSPI.h:
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView requestAutocorrectionRectsForString:withCompletionHandler:]):
(-[WKContentView requestRectsToEvadeForSelectionCommandsWithCompletionHandler:]):
(-[WKContentView requestAutocorrectionContextWithCompletionHandler:]):

Drive-by: handle null completion handler arguments more gracefully, by raising an NSException and bailing before
attempting to invoke a nil block.

* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::requestEvasionRectsAboveSelection):

See above for more detail.

* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::requestEvasionRectsAboveSelection):

Tools:

Add a couple of UIScriptController methods to make callout menu testing on iOS easier (see below).

* DumpRenderTree/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::menuRect const):
(WTR::UIScriptController::isShowingMenu const):
* TestRunnerShared/UIScriptContext/Bindings/UIScriptController.idl:
* TestRunnerShared/UIScriptContext/UIScriptController.cpp:
(WTR::UIScriptController::menuRect const):

Add a function to query the bounds of the callout menu in content coordinates.

(WTR::UIScriptController::isShowingMenu const):

Add a function to query whether the callout menu is shown (i.e., has finished its appearance animation).

* TestRunnerShared/UIScriptContext/UIScriptController.h:
* WebKitTestRunner/cocoa/TestRunnerWKWebView.h:
* WebKitTestRunner/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::rectForMenuAction const):
(WTR::UIScriptController::menuRect const):
(WTR::UIScriptController::isShowingMenu const):
(WTR::findViewInHierarchyOfType): Deleted.

LayoutTests:

Add a test to ensure that the we dodge clickable elements when showing the callout bar.

* editing/selection/ios/avoid-showing-callout-menu-over-controls-expected.txt: Added.
* editing/selection/ios/avoid-showing-callout-menu-over-controls.html: Added.
* resources/ui-helper.js:
(window.UIHelper.waitForMenuToShow.return.new.Promise):
(window.UIHelper.waitForMenuToShow):
(window.UIHelper.menuRect):
(window.UIHelper):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[iOS] Break a reference cycle between PreviewLoader and ResourceLoader
aestes@apple.com [Sat, 23 Feb 2019 00:41:58 +0000 (00:41 +0000)]
[iOS] Break a reference cycle between PreviewLoader and ResourceLoader
https://bugs.webkit.org/show_bug.cgi?id=194964
<rdar://problem/48279441>

Reviewed by Alex Christensen.

When a document's QuickLook preview is loaded, a reference cycle is created between
PreviewLoader and ResourceLoader. Break the cycle by clearing m_previewLoader in
ResourceLoader::releaseResources().

Fixes leaks detected by `run-webkit-tests --leaks LayoutTests/quicklook`.

* loader/ResourceLoader.cpp:
(WebCore::ResourceLoader::releaseResources):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed build fix: need a WebCore:: namespace on ScrollingEventResult.
simon.fraser@apple.com [Sat, 23 Feb 2019 00:13:49 +0000 (00:13 +0000)]
Unreviewed build fix: need a WebCore:: namespace on ScrollingEventResult.

* UIProcess/RemoteLayerTree/ios/ScrollingTreeOverflowScrollingNodeIOS.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoDFGBytecodeParser should not declare that a node won't clobberExit if DFGFixupPhase...
rmorisset@apple.com [Sat, 23 Feb 2019 00:05:11 +0000 (00:05 +0000)]
DFGBytecodeParser should not declare that a node won't clobberExit if DFGFixupPhase can later declare it does clobberExit
https://bugs.webkit.org/show_bug.cgi?id=194953
<rdar://problem/47595253>

Reviewed by Saam Barati.

JSTests:

I could not make this work without the infinite loop, so I am using a watchdog to be able to use it as a regression test.

* stress/has-indexed-property-with-worsening-array-mode.js: Added.

Source/JavaScriptCore:

For each node that
(a) may or may not clobberExit depending on their arrayMode
(b) and get their arrayMode from profiling information in DFGBytecodeParser
(c) and can have their arrayMode refined by DFGFixupPhase,
We must make sure to be conservative in the DFGBytecodeParser and treat it as if it unconditionnally clobbered the exit.
Otherwise we will hit a validation failure after fixup if the next node was marked ExitValid and exits to the same semantic origin.

The list of nodes that fit (a) is:
- StringCharAt
- HasIndexProperty
- GetByVal
- PutByValDirect
- PutByVal
- PutByValAlias
- GetIndexedPropertyStorage

Out of these, the following also fit (b) and (c):
- HasIndexedProperty
- GetByVal
- PutByValDirect
- PutByVal

GetByVal already had "m_exitOK = false; // GetByVal must be treated as if it clobbers exit state, since FixupPhase may make it generic."
So we just have to fix the other three the same way.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):
(JSC::DFG::ByteCodeParser::handlePutByVal):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoCrash under IDBServer::IDBConnectionToClient::identifier() const
sihui_liu@apple.com [Fri, 22 Feb 2019 23:41:16 +0000 (23:41 +0000)]
Crash under IDBServer::IDBConnectionToClient::identifier() const
https://bugs.webkit.org/show_bug.cgi?id=194843
<rdar://problem/48203102>

Reviewed by Geoffrey Garen.

UniqueIDBDatabase should ignore requests from connections that are already closed.

Tests are hard to create without some tricks on UniqueIDBDatabase so this fix is verified manually.
One test is created by adding delay to UniqueIDBDatabase::openBackingStore on the background thread to make sure
disconnection of web process happens before UniqueIDBDatabase::didOpenBackingStore, because didOpenBackingStore
may start a version change transaction and ask for identifier from the connection that is already gone.

* Modules/indexeddb/server/IDBConnectionToClient.cpp:
(WebCore::IDBServer::IDBConnectionToClient::connectionToClientClosed):
* Modules/indexeddb/server/IDBConnectionToClient.h:
(WebCore::IDBServer::IDBConnectionToClient::isClosed):
* Modules/indexeddb/server/UniqueIDBDatabase.cpp:
(WebCore::IDBServer::UniqueIDBDatabase::clearStalePendingOpenDBRequests):
(WebCore::IDBServer::UniqueIDBDatabase::handleDatabaseOperations):
(WebCore::IDBServer::UniqueIDBDatabase::operationAndTransactionTimerFired):
* Modules/indexeddb/server/UniqueIDBDatabase.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoB3ReduceStrength: missing peephole optimizations for binary operations
rmorisset@apple.com [Fri, 22 Feb 2019 22:54:23 +0000 (22:54 +0000)]
B3ReduceStrength: missing peephole optimizations for binary operations
https://bugs.webkit.org/show_bug.cgi?id=194252

Reviewed by Saam Barati.

Adds several sets of optimizations for BitAnd, BitOr and BitXor.
Using BitAnd distributivity over BitOr and BitXor:
  Turn any of these (for Op == BitOr || Op == BitXor):
        Op(BitAnd(x1, x2), BitAnd(x1, x3))
        Op(BitAnd(x2, x1), BitAnd(x1, x3))
        Op(BitAnd(x1, x2), BitAnd(x3, x1))
        Op(BitAnd(x2, x1), BitAnd(x3, x1))
   Into this: BitAnd(Op(x2, x3), x1)
   And any of these:
        Op(BitAnd(x1, x2), x1)
        Op(BitAnd(x2, x1), x1)
        Op(x1, BitAnd(x1, x2))
        Op(x1, BitAnd(x2, x1))
   Into this: BitAnd(Op(x2, x1), x1)
   This second set is equivalent to doing x1 => BitAnd(x1, x1), and then applying the first set.
Using de Morgan laws (we represent not as BitXor with allOnes):
  BitAnd(BitXor(x1, allOnes), BitXor(x2, allOnes)) => BitXor(BitOr(x1, x2), allOnes)
  BitOr(BitXor(x1, allOnes), BitXor(x2, allOnes) => BitXor(BitAnd(x1, x2), allOnes)
  BitOr(BitXor(x, allOnes), c) => BitXor(BitAnd(x, ~c), allOnes)
  BitAnd(BitXor(x, allOnes), c) => BitXor(BitOr(x, ~c), allOnes)
The latter two are equivalent to doing c => BitXor(~c, allOnes), and then applying the former two.

All of these transformations either reduce the number of operations (which we always do when possible), or bring the expression closer to having:
  - BitXor with all ones at the outermost
  - then BitAnd
  - then other BitXor
  - then BitOr at the innermost.
These transformations that don't directly reduce the number of operations are still useful for normalization (helping things like CSE), and also can enable
more optimizations (for example BitXor with all ones can easily cancel each other once they are all at the outermost level).

* b3/B3ReduceStrength.cpp:
* b3/testb3.cpp:
(JSC::B3::testBitAndNotNot):
(JSC::B3::testBitAndNotImm):
(JSC::B3::testBitOrAndAndArgs):
(JSC::B3::testBitOrAndSameArgs):
(JSC::B3::testBitOrNotNot):
(JSC::B3::testBitOrNotImm):
(JSC::B3::testBitXorAndAndArgs):
(JSC::B3::testBitXorAndSameArgs):
(JSC::B3::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRegression(PSON) Unable to preview password-protected documents on iCloud.com
cdumez@apple.com [Fri, 22 Feb 2019 21:17:37 +0000 (21:17 +0000)]
Regression(PSON) Unable to preview password-protected documents on iCloud.com
https://bugs.webkit.org/show_bug.cgi?id=194954
<rdar://problem/48127957>

Reviewed by Alex Christensen.

Source/WebKit:

Make sure the ProvisionalPageProxy forwards to the WebPageProxy the QuickLook-related
IPC on iOS.

* UIProcess/ProvisionalPageProxy.cpp:
(WebKit::ProvisionalPageProxy::didRequestPasswordForQuickLookDocumentInMainFrame):
(WebKit::ProvisionalPageProxy::didReceiveMessage):
* UIProcess/ProvisionalPageProxy.h:
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didRequestPasswordForQuickLookDocumentInMainFrame):
(WebKit::WebPageProxy::didRequestPasswordForQuickLookDocumentInMainFrameShared):

Tools:

Add API test coverage.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
(-[PSONNavigationDelegate _webViewDidRequestPasswordForQuickLookDocument:]):
(-[PSONNavigationDelegate _webView:didStartLoadForQuickLookDocumentInMainFrameWithFileName:uti:]):
(-[PSONNavigationDelegate _webView:didFinishLoadForQuickLookDocumentInMainFrame:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[iOS] Sandbox should allow mach lookups related to media capturing
pvollan@apple.com [Fri, 22 Feb 2019 20:44:28 +0000 (20:44 +0000)]
[iOS] Sandbox should allow mach lookups related to media capturing
https://bugs.webkit.org/show_bug.cgi?id=194951
<rdar://problem/48013875>

Reviewed by Brent Fulgham.

* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[JSC] putNonEnumerable in JSWrapperMap is too costly
ysuzuki@apple.com [Fri, 22 Feb 2019 20:20:49 +0000 (20:20 +0000)]
[JSC] putNonEnumerable in JSWrapperMap is too costly
https://bugs.webkit.org/show_bug.cgi?id=194935

Reviewed by Mark Lam.

When we convert Objective-C blocks to JS objects, we need to set up a corresponding function object correctly.
During this allocation, we call [JSValue defineProperty:descriptor] to connect a "prototype" object and "constructor" object.
The problem is that this API has a particularly costly implementation:

    [[_context globalObject][@"Object"] invokeMethod:@"defineProperty" withArguments:@[ self, key, descriptor ]];

This wraps each JS objects appear in this code with Objective-C wrapper. And we convert a NSDictionary to JSObject, which
has "writable", "enumerable", "configurable", "value" fields, and call the "defineProperty" JS function through Objective-C wrapper.
This allocates many Objective-C wrappers and JS objects for descriptors. Since JSC has a direct C++ API "defineOwnProperty", we should
bypass these Objective-C APIs and call JSC's code directly.

This patch changes `putNonEnumerable` implementation, from calling [JSValue defineProperty:descriptor] to calling JSC C++ code directly.
We do not change [JSValue defineProperty:descriptor] implementation for now because of two reasons. (1) This is not used in our benchmarks
except for this (converting an Objective-C block to a JS object) one path. And (2) even if we were to re-write [JSValue defineProperty:descriptor]
to be more optimized, we would still want to call the JSC C++ version of defineProperty directly here to avoid NSDictionary allocation for a descriptor.

* API/APIUtils.h:
(setException):
* API/JSWrapperMap.mm:
(putNonEnumerable):
(copyMethodsToObject):
(-[JSObjCClassInfo allocateConstructorAndPrototypeInContext:]):
(-[JSObjCClassInfo wrapperForObject:inContext:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed, build fix after r241954
ysuzuki@apple.com [Fri, 22 Feb 2019 19:15:42 +0000 (19:15 +0000)]
Unreviewed, build fix after r241954
https://bugs.webkit.org/show_bug.cgi?id=194939

Renaming setCanAccessHeap was incomplete.

* runtime/SmallStrings.cpp:
(JSC::SmallStrings::initializeCommonStrings):
* runtime/VM.cpp:
(JSC::VM::~VM):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[JSC] SmallStringsStorage is unnecessary
ysuzuki@apple.com [Fri, 22 Feb 2019 19:04:40 +0000 (19:04 +0000)]
[JSC] SmallStringsStorage is unnecessary
https://bugs.webkit.org/show_bug.cgi?id=194939

Reviewed by Mark Lam.

SmallStrings hold common small JSStrings. Their underlying StringImpl is also held by SmallStringsStorage.
But it is duplicate since we can get StringImpl from small JSStrings. This patch removes SmallStringsStorage,
and get StringImpls from JSStrings if necessary.

We also add m_canAccessHeap flag to SmallStrings. At the time of VM destruction, JSStrings are destroyed when
VM's Heap is finalized. We must not touch JSStrings before VM's heap (and JSStrings in SmallStrings) is initialized,
and after VM's Heap is destroyed. We add this m_canAccessHeap flag to allow users to get StringImpl during the
this sensitive period. If m_canAccessHeap is false, we get StringImpl from AtomicStringImpl::add.

* runtime/SmallStrings.cpp:
(JSC::SmallStrings::initializeCommonStrings):
(JSC::SmallStrings::singleCharacterStringRep):
(JSC::SmallStringsStorage::rep): Deleted.
(JSC::SmallStringsStorage::SmallStringsStorage): Deleted.
(JSC::SmallStrings::createSingleCharacterString): Deleted.
* runtime/SmallStrings.h:
(JSC::SmallStrings::setCanAccessHeap):
* runtime/VM.cpp:
(JSC::VM::VM):
(JSC::VM::~VM):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: Remove unused Timeline record selection code path
commit-queue@webkit.org [Fri, 22 Feb 2019 18:46:29 +0000 (18:46 +0000)]
Web Inspector: Remove unused Timeline record selection code path
https://bugs.webkit.org/show_bug.cgi?id=194931

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2019-02-22
Reviewed by Devin Rousso.

* UserInterface/Views/TimelineOverview.js:
(WI.TimelineOverview):
(WI.TimelineOverview.prototype.userSelectedRecord): Deleted.
* UserInterface/Views/TimelineRecordingContentView.js:
(WI.TimelineRecordingContentView):
(WI.TimelineRecordingContentView.prototype.timelineOverviewUserSelectedRecord): Deleted.
* UserInterface/Views/TimelineView.js:
(WI.TimelineView.prototype.userSelectedRecordFromOverview): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[GTK] Navigation gesture improvements
commit-queue@webkit.org [Fri, 22 Feb 2019 18:46:25 +0000 (18:46 +0000)]
[GTK] Navigation gesture improvements
https://bugs.webkit.org/show_bug.cgi?id=194943

Patch by Alexander Mikhaylenko <exalm7659@gmail.com> on 2019-02-22
Reviewed by Michael Catanzaro.

Cancel the gesture if progress is less than 0.5 and velocity is not high enough.

Allow to continue the gesture during animation. Introduce finished state to be used
when showing snapshot after the animation ends.

Fix duration calculation, also slow it down so that the initial velocity matches
what it was during the gesture.

* UIProcess/ViewGestureController.h: Add shouldCancel() and State::Finishing to SwipeProgressTracker.
* UIProcess/gtk/ViewGestureControllerGtk.cpp:
(WebKit::ViewGestureController::SwipeProgressTracker::handleEvent):
Fix velocity calculation, allow scrolling during State::Animating.
(WebKit::ViewGestureController::SwipeProgressTracker::shouldCancel): Added.
(WebKit::ViewGestureController::SwipeProgressTracker::startAnimation): Use shouldCancel() and fix duration calculation.
(WebKit::ViewGestureController::SwipeProgressTracker::endAnimation): Set state to State::Finishing when the animation ends.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoREGRESSION(PSON) Scroll position is sometimes not restored on history navigation
cdumez@apple.com [Fri, 22 Feb 2019 18:24:57 +0000 (18:24 +0000)]
REGRESSION(PSON) Scroll position is sometimes not restored on history navigation
https://bugs.webkit.org/show_bug.cgi?id=194924
<rdar://problem/48216125>

Reviewed by Geoffrey Garen.

Source/WebKit:

When process-swapping, we would create a new WebPage in the new process, which would
call restoreSessionInternal() to restore the HistoryItems based on the UIProcess's
backforward list. The issue is that this session restoring would send HistoryItem
updates back to the UIProcess. Without PSON, this would be unnecessary but harmless.
With PSON though, this may end up overwriting values set by the previous process,
such as the scroll position.

Address the issue by temporarily disabling the HistoryItem update notifications to
the UIProcess while restoring a session.

* UIProcess/API/Cocoa/WKBackForwardListItem.mm:
(-[WKBackForwardListItem _scrollPosition]):
* UIProcess/API/Cocoa/WKBackForwardListItemPrivate.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::restoreSessionInternal):

Tools:

Add API test coverage.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoInput type "formatSetInlineTextDirection" is dispatched when changing paragraph-level...
wenson_hsieh@apple.com [Fri, 22 Feb 2019 17:09:27 +0000 (17:09 +0000)]
Input type "formatSetInlineTextDirection" is dispatched when changing paragraph-level text direction
https://bugs.webkit.org/show_bug.cgi?id=194703
<rdar://problem/48111775>

Reviewed by Ryosuke Niwa.

Source/WebCore:

Currently, when changing text direction, WebKit always sends input events of type formatSetInlineTextDirection,
even when changing paragraph text direction. Instead, we should be emitting formatSetBlockTextDirection in this
scenario. This is problematic when using the context menus on macOS to change writing direction, since changing
"Selection Direction" is currently indistinguishable from changing "Paragraph Direction".

To fix this, we split EditAction::SetWritingDirection into EditAction::SetInlineWritingDirection and
EditAction::SetBlockWritingDirection, which emit inline and block text direction input events, respectively.

Tests: fast/events/before-input-events-prevent-block-text-direction.html
       fast/events/before-input-events-prevent-inline-text-direction.html

* editing/CompositeEditCommand.cpp:
(WebCore::CompositeEditCommand::apply):
* editing/EditAction.cpp:
(WebCore::undoRedoLabel):
* editing/EditAction.h:
* editing/EditCommand.cpp:
(WebCore::inputTypeNameForEditingAction):
* editing/Editor.cpp:
(WebCore::inputEventDataForEditingStyleAndAction):
(WebCore::Editor::setBaseWritingDirection):
* editing/EditorCommand.cpp:
(WebCore::executeMakeTextWritingDirectionLeftToRight):
(WebCore::executeMakeTextWritingDirectionNatural):
(WebCore::executeMakeTextWritingDirectionRightToLeft):

Source/WebKitLegacy/win:

* WebCoreSupport/WebEditorClient.cpp:
(undoNameForEditAction):

LayoutTests:

Rebaseline some existing tests to expect input events of type "formatSetBlockTextDirection" instead of
"formatSetInlineTextDirection" when changing paragraph text direction; additionally, add a new layout test that
changes the inline text direction in some Bidi text, and verify that "formatSetInlineTextDirection" is emitted
in this scenario, and that calling `preventDefault()` in the beforeinput event handler causes no change to be
made.

* editing/input/ios/rtl-keyboard-input-on-focus-expected.txt:
* fast/events/before-input-events-prevent-block-text-direction-expected.txt: Added.
* fast/events/before-input-events-prevent-block-text-direction.html: Renamed from LayoutTests/fast/events/before-input-events-prevent-text-direction.html.
* fast/events/before-input-events-prevent-inline-text-direction-expected.txt: Added.
* fast/events/before-input-events-prevent-inline-text-direction.html: Added.
* fast/events/before-input-events-prevent-text-direction-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed, disable API test added in r241928 on iOS.
cdumez@apple.com [Fri, 22 Feb 2019 16:48:38 +0000 (16:48 +0000)]
Unreviewed, disable API test added in r241928 on iOS.

The cache is not enabled on devices with less than 3GB of RAM.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241948 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRemove stripLeadingAndTrailingWhitespace from MathMLElement.cpp
commit-queue@webkit.org [Fri, 22 Feb 2019 16:31:17 +0000 (16:31 +0000)]
Remove stripLeadingAndTrailingWhitespace from MathMLElement.cpp
https://bugs.webkit.org/show_bug.cgi?id=160172

Patch by Rob Buis <rbuis@igalia.com> on 2019-02-22
Reviewed by Frédéric Wang.

Remove stripLeadingAndTrailingWhitespace and use stripLeadingAndTrailingHTTPSpaces
from HTTPParsers instead.

No new tests, already covered by MathML tests.

* mathml/MathMLElement.cpp:
(WebCore::MathMLElement::stripLeadingAndTrailingWhitespace): Deleted.
* mathml/MathMLElement.h:
* mathml/MathMLPresentationElement.cpp:
(WebCore::MathMLPresentationElement::parseMathMLLength):
* mathml/MathMLTokenElement.cpp:
(WebCore::MathMLTokenElement::convertToSingleCodePoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUpdate some media logging
eric.carlson@apple.com [Fri, 22 Feb 2019 13:52:18 +0000 (13:52 +0000)]
Update some media logging
https://bugs.webkit.org/show_bug.cgi?id=194915

Reviewed by Jer Noble.

No new tests, no functional change.

* Modules/mediasource/SourceBuffer.cpp:
(WebCore::SourceBuffer::evictCodedFrames):
(WebCore::SourceBuffer::provideMediaData):
(WebCore::SourceBuffer::trySignalAllSamplesInTrackEnqueued):

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::checkPlaybackTargetCompatablity):
(WebCore::HTMLMediaElement::loadResource):
(WebCore::HTMLMediaElement::updateActiveTextTrackCues):
(WebCore::HTMLMediaElement::canTransitionFromAutoplayToPlay const):
(WebCore::HTMLMediaElement::seekTask):
(WebCore::HTMLMediaElement::playInternal):
(WebCore::HTMLMediaElement::pauseInternal):
(WebCore::HTMLMediaElement::setLoop):
(WebCore::HTMLMediaElement::setControls):
(WebCore::HTMLMediaElement::sourceWasRemoved):

* html/MediaElementSession.cpp:
(WebCore::convertEnumerationToString):

* html/MediaElementSession.h:
(WTF::LogArgument<WebCore::MediaPlaybackDenialReason>::toString):

* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
(WebCore::SourceBufferPrivateAVFObjC::enqueueSample):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFix unitless usage of mathsize
commit-queue@webkit.org [Fri, 22 Feb 2019 10:42:43 +0000 (10:42 +0000)]
Fix unitless usage of mathsize
https://bugs.webkit.org/show_bug.cgi?id=194940

Patch by Rob Buis <rbuis@igalia.com> on 2019-02-22
Reviewed by Frédéric Wang.

Source/WebCore:

Convert unitless lengths to percentage values to correct the computed
font size.

* mathml/MathMLElement.cpp:
(WebCore::convertToPercentageIfNeeded):
(WebCore::MathMLElement::collectStyleForPresentationAttribute):

LayoutTests:

Tests lengths-1.html and length-3.html now pass.

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed. Bump version numbers
aperez@igalia.com [Fri, 22 Feb 2019 09:18:24 +0000 (09:18 +0000)]
Unreviewed. Bump version numbers

* Source/cmake/OptionsWPE.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoCache CompactVariableMap::Handle instead of VariableEnvironment for UnlinkedFunctionE...
tzagallo@apple.com [Fri, 22 Feb 2019 09:08:35 +0000 (09:08 +0000)]
Cache CompactVariableMap::Handle instead of VariableEnvironment for UnlinkedFunctionExecutable
https://bugs.webkit.org/show_bug.cgi?id=194706

Reviewed by Saam Barati.

In https://bugs.webkit.org/show_bug.cgi?id=194583 we started using a
CompactVariableMap::Handle instead of VariableEnvironment for
UnlinkedFunctionExecutables, but we were creating the full environment
to encode the executable in the bytecode cache. This patch changes it so
that we cache the handle instead of the environment. This avoids duplicating
the VariableEnvironment whenever we have to cache two handles that point
to the environment.

* bytecode/UnlinkedFunctionExecutable.h:
* parser/VariableEnvironment.cpp:
(JSC::CompactVariableMap::get):
* parser/VariableEnvironment.h:
* runtime/CachedTypes.cpp:
(JSC::CachedCompactVariableEnvironment::encode):
(JSC::CachedCompactVariableEnvironment::decode const):
(JSC::CachedCompactVariableMapHandle::encode):
(JSC::CachedCompactVariableMapHandle::decode const):
(JSC::CachedFunctionExecutable::encode):
(JSC::CachedFunctionExecutable::decode const):
(JSC::UnlinkedFunctionExecutable::UnlinkedFunctionExecutable):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed. Bump version numbers
carlosgc@webkit.org [Fri, 22 Feb 2019 07:32:29 +0000 (07:32 +0000)]
Unreviewed. Bump version numbers

* Source/cmake/OptionsGTK.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[WPE][GTK] No API documentation generated for WebKitUserContentFilterStore
aperez@igalia.com [Fri, 22 Feb 2019 07:19:38 +0000 (07:19 +0000)]
[WPE][GTK] No API documentation generated for WebKitUserContentFilterStore
https://bugs.webkit.org/show_bug.cgi?id=194908

Reviewed by Carlos Garcia Campos.

* UIProcess/API/gtk/docs/webkit2gtk-4.0.types: Add entry for webkit_user_content_filter_store_get_type().
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml: Add entry for WebKitUserContentFilterStore.
* UIProcess/API/wpe/docs/wpe-docs.sgml: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agowpewebkit-2.23.90 Failing to build on Fedora
carlosgc@webkit.org [Fri, 22 Feb 2019 07:06:03 +0000 (07:06 +0000)]
wpewebkit-2.23.90 Failing to build on Fedora
https://bugs.webkit.org/show_bug.cgi?id=194922

Unreviewed. Fix path to WebKitUserContentFilterStore.h in WPE_API_INSTALLED_HEADERS.

* PlatformWPE.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoHardcode Visual Viewports on everywhere except iOS WK1
simon.fraser@apple.com [Fri, 22 Feb 2019 07:04:03 +0000 (07:04 +0000)]
Hardcode Visual Viewports on everywhere except iOS WK1
https://bugs.webkit.org/show_bug.cgi?id=194928

Reviewed by Zalan Bujtas.
Source/WebCore:

Remove the WK1 and WK2 preferences and MiniBrowser menu item for "visual viewports",
change the default value of the Setting to 'true', and hardcode WebView on iOS to
set it to false. The setting has shipped for several years and there's no need to turn
it off now.

Similarly, disable the "Visual Viewport API" on iOS WK1, since it makes no sense if
Visual Viewports are not enabled.

Remove the "visualViewportEnabled" flag and unused code paths from scrolling tree code
that only runs in WK2

* page/Settings.yaml:
* page/scrolling/AsyncScrollingCoordinator.cpp:
(WebCore::AsyncScrollingCoordinator::frameViewLayoutUpdated):
(WebCore::AsyncScrollingCoordinator::reconcileScrollingState):
(WebCore::AsyncScrollingCoordinator::visualViewportEnabled const): Deleted.
* page/scrolling/AsyncScrollingCoordinator.h:
* page/scrolling/ScrollingStateFrameScrollingNode.cpp:
(WebCore::ScrollingStateFrameScrollingNode::ScrollingStateFrameScrollingNode):
(WebCore::ScrollingStateFrameScrollingNode::setAllPropertiesChanged):
(WebCore::ScrollingStateFrameScrollingNode::dumpProperties const):
(WebCore::ScrollingStateFrameScrollingNode::setVisualViewportEnabled): Deleted.
* page/scrolling/ScrollingStateFrameScrollingNode.h:
* page/scrolling/ScrollingTree.cpp:
(WebCore::ScrollingTree::commitTreeState):
* page/scrolling/ScrollingTree.h:
(WebCore::ScrollingTree::visualViewportEnabled const): Deleted.
(WebCore::ScrollingTree::setVisualViewportEnabled): Deleted.
* page/scrolling/ScrollingTreeFrameScrollingNode.cpp:
(WebCore::ScrollingTreeFrameScrollingNode::layoutViewportForScrollPosition const):
* page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.mm:
(WebCore::ScrollingTreeFrameScrollingNodeMac::setScrollPositionWithoutContentEdgeConstraints):
(WebCore::ScrollingTreeFrameScrollingNodeMac::setScrollLayerPosition):

Source/WebKit:

Remove the WK1 and WK2 preferences and MiniBrowser menu item for "visual viewports",
change the default value of the Setting to 'true', and hardcode WebView on iOS to
set it to false. The setting has shipped for several years and there's no need to turn
it off now.

Similarly, disable the "Visual Viewport API" on iOS WK1, since it makes no sense if
Visual Viewports are not enabled.

Remove the "visualViewportEnabled" flag and unused code paths from scrolling tree code
that only runs in WK2

* Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateFrameScrollingNode>::encode):
(ArgumentCoder<ScrollingStateFrameScrollingNode>::decode):
* Shared/WebPreferences.yaml:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _visualViewportEnabled]): Deleted.
(-[WKPreferences _setVisualViewportEnabled:]): Deleted.
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::commitLayerTree):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.h:
(WebKit::RemoteScrollingCoordinatorProxy::visualViewportEnabled const): Deleted.
* UIProcess/RemoteLayerTree/ios/RemoteScrollingCoordinatorProxyIOS.mm:
(WebKit::RemoteScrollingCoordinatorProxy::customFixedPositionRect const):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentView.mm:
(-[WKContentView didUpdateVisibleRect:unobscuredRect:contentInsets:unobscuredRectInScrollViewCoordinates:obscuredInsets:unobscuredSafeAreaInsets:inputViewBounds:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:enclosedInScrollableAncestorView:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::computeCustomFixedPositionRect const):
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::WebAutomationSessionProxy::computeElementLayout):
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::overrideBoolPreferenceForTestRunner):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getFocusedElementInformation):
(WebKit::WebPage::dynamicViewportSizeUpdate):
(WebKit::WebPage::updateVisibleContentRects):

Source/WebKitLegacy/mac:

Remove the WK1 and WK2 preferences and MiniBrowser menu item for "visual viewports",
change the default value of the Setting to 'true', and hardcode WebView on iOS to
set it to false. The setting has shipped for several years and there's no need to turn
it off now.

Similarly, disable the "Visual Viewport API" on iOS WK1, since it makes no sense if
Visual Viewports are not enabled.

Remove the "visualViewportEnabled" flag and unused code paths from scrolling tree code
that only runs in WK2

* WebView/WebPreferenceKeysPrivate.h:
* WebView/WebPreferences.mm:
(+[WebPreferences initialize]):
(-[WebPreferences visualViewportEnabled]): Deleted.
(-[WebPreferences setVisualViewportEnabled:]): Deleted.
* WebView/WebPreferencesPrivate.h:
* WebView/WebView.mm:
(-[WebView _preferencesChanged:]):

Tools:

Remove the WK1 and WK2 preferences and MiniBrowser menu item for "visual viewports",
change the default value of the Setting to 'true', and hardcode WebView on iOS to
set it to false. The setting has shipped for several years and there's no need to turn
it off now.

Similarly, disable the "Visual Viewport API" on iOS WK1, since it makes no sense if
Visual Viewports are not enabled.

Remove the "visualViewportEnabled" flag and unused code paths from scrolling tree code
that only runs in WK2

* MiniBrowser/mac/SettingsController.h:
* MiniBrowser/mac/SettingsController.m:
(-[SettingsController init]):
(-[SettingsController _populateMenu]):
(-[SettingsController validateMenuItem:]):
(-[SettingsController visualViewportEnabled]): Deleted.
(-[SettingsController toggleVisualViewportEnabled:]): Deleted.
* MiniBrowser/mac/WK1BrowserWindowController.m:
(-[WK1BrowserWindowController didChangeSettings]):
* MiniBrowser/mac/WK2BrowserWindowController.m:
(-[WK2BrowserWindowController didChangeSettings]):

LayoutTests:

Remove "(visual viewport enabled 1)" from scrolling tree dumps. Remove calls
to "internals.settings.setVisualViewportEnabled(true)"

* fast/dom/elementFromPoint-scaled-scrolled-layout-viewport.html:
* fast/events/ios/rotation/resources/rotation-utils.js:
* fast/visual-viewport/client-coordinates-relative-to-layout-viewport.html:
* fast/visual-viewport/client-rects-relative-to-layout-viewport-zoomed.html:
* fast/visual-viewport/client-rects-relative-to-layout-viewport.html:
* fast/visual-viewport/nonzoomed-rects.html:
* fast/visual-viewport/resize-event-fired-window-resized.html:
* fast/visual-viewport/resize-event-fired.html:
* fast/visual-viewport/rtl-nonzoomed-rects.html:
* fast/visual-viewport/rtl-zoomed-rects.html:
* fast/visual-viewport/rubberbanding-viewport-rects-extended-background.html:
* fast/visual-viewport/rubberbanding-viewport-rects-header-footer.html:
* fast/visual-viewport/rubberbanding-viewport-rects.html:
* fast/visual-viewport/scroll-event-fired.html:
* fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolled-down-expected.txt:
* fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolled-down-then-up-expected.txt:
* fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolled-down-then-up.html:
* fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolled-down.html:
* fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolling-layers-state-expected.txt:
* fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolling-layers-state.html:
* fast/visual-viewport/viewport-dimensions-exclude-custom-scrollbars.html:
* fast/visual-viewport/viewport-dimensions-exclude-scrollbars.html:
* fast/visual-viewport/viewport-dimensions-iframe.html:
* fast/visual-viewport/viewport-dimensions-under-page-zoom.html:
* fast/visual-viewport/viewport-dimensions.html:
* fast/visual-viewport/visual-viewport-same-object.html:
* fast/visual-viewport/zoomed-fixed-header-and-footer.html:
* fast/visual-viewport/zoomed-fixed-scroll-down-then-up.html:
* fast/visual-viewport/zoomed-fixed.html:
* fast/visual-viewport/zoomed-rects.html:
* fast/visual-viewport/zoomed-scroll-into-view-fixed.html:
* fast/visual-viewport/zoomed-scroll-to-anchor-in-position-fixed.html:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/coordinated-frame-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/coordinated-frame-gain-scrolling-ancestor-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/coordinated-frame-in-fixed-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/coordinated-frame-lose-scrolling-ancestor-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/fixed-inside-frame-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/gain-scrolling-node-parent-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/lose-scrolling-node-parent-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/overflow-in-fixed-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/remove-coordinated-frame-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/remove-scrolling-role-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/reparent-across-compositing-layers-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/reparent-with-layer-removal-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/scrolling-tree-includes-frame-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/scrolling-tree-is-z-order-expected.txt:
* platform/ios-wk2/scrollingcoordinator/scrolling-tree/toggle-coordinated-frame-scrolling-expected.txt:
* scrollingcoordinator/scrolling-tree/coordinated-frame-expected.txt:
* scrollingcoordinator/scrolling-tree/coordinated-frame-gain-scrolling-ancestor-expected.txt:
* scrollingcoordinator/scrolling-tree/coordinated-frame-in-fixed-expected.txt:
* scrollingcoordinator/scrolling-tree/coordinated-frame-lose-scrolling-ancestor-expected.txt:
* scrollingcoordinator/scrolling-tree/fixed-inside-frame-expected.txt:
* scrollingcoordinator/scrolling-tree/gain-scrolling-node-parent-expected.txt:
* scrollingcoordinator/scrolling-tree/lose-scrolling-node-parent-expected.txt:
* scrollingcoordinator/scrolling-tree/overflow-in-fixed-expected.txt:
* scrollingcoordinator/scrolling-tree/remove-coordinated-frame-expected.txt:
* scrollingcoordinator/scrolling-tree/remove-scrolling-role-expected.txt:
* scrollingcoordinator/scrolling-tree/reparent-across-compositing-layers-expected.txt:
* scrollingcoordinator/scrolling-tree/reparent-with-layer-removal-expected.txt:
* scrollingcoordinator/scrolling-tree/scrolling-tree-includes-frame-expected.txt:
* scrollingcoordinator/scrolling-tree/scrolling-tree-is-z-order-expected.txt:
* scrollingcoordinator/scrolling-tree/toggle-coordinated-frame-scrolling-expected.txt:
* scrollingcoordinator/scrolling-tree/toggle-coordinated-frame-scrolling.html: Don't insert the first scrolling tree
dump until the end, to avoid it affecting main frame height before we get the second layer tree.
* tiled-drawing/scrolling/clamp-out-of-bounds-scrolls-expected.txt:
* tiled-drawing/scrolling/fixed/absolute-inside-fixed-expected.txt:
* tiled-drawing/scrolling/fixed/absolute-inside-out-of-view-fixed-expected.txt:
* tiled-drawing/scrolling/fixed/fixed-in-overflow-expected.txt:
* tiled-drawing/scrolling/fixed/fixed-position-out-of-view-expected.txt:
* tiled-drawing/scrolling/fixed/fixed-position-out-of-view-negative-zindex-expected.txt:
* tiled-drawing/scrolling/fixed/four-bars-expected.txt:
* tiled-drawing/scrolling/fixed/four-bars-with-header-and-footer-expected.txt:
* tiled-drawing/scrolling/fixed/four-bars-zoomed-expected.txt:
* tiled-drawing/scrolling/fixed/negative-scroll-offset-expected.txt:
* tiled-drawing/scrolling/fixed/negative-scroll-offset-in-view-expected.txt:
* tiled-drawing/scrolling/fixed/nested-fixed-expected.txt:
* tiled-drawing/scrolling/fixed/percentage-inside-fixed-expected.txt:
* tiled-drawing/scrolling/frames/scroll-region-after-frame-layout-expected.txt:
* tiled-drawing/scrolling/scrolling-tree-after-scroll-expected.txt:
* tiled-drawing/scrolling/scrolling-tree-slow-scrolling-expected.txt:
* tiled-drawing/scrolling/sticky/negative-scroll-offset-expected.txt:
* tiled-drawing/scrolling/sticky/sticky-horizontal-expected.txt:
* tiled-drawing/scrolling/sticky/sticky-vertical-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[GTK] Epiphany searching for plugins even if plugins are disabled
carlosgc@webkit.org [Fri, 22 Feb 2019 07:02:05 +0000 (07:02 +0000)]
[GTK] Epiphany searching for plugins even if plugins are disabled
https://bugs.webkit.org/show_bug.cgi?id=194352

Reviewed by Michael Catanzaro.

Check pluginsEnabled setting before trying to get plugins from UI process.

* WebProcess/Plugins/WebPluginInfoProvider.cpp:
(WebKit::WebPluginInfoProvider::populatePluginCache):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoSome refinements for Node and Document
darin@apple.com [Fri, 22 Feb 2019 06:46:37 +0000 (06:46 +0000)]
Some refinements for Node and Document
https://bugs.webkit.org/show_bug.cgi?id=194764

Reviewed by Ryosuke Niwa.

Source/WebCore:

* accessibility/AccessibilityObject.cpp:
(WebCore::AccessibilityObject::press): Use shadowHost instead of
deprecatedShadowAncestorNode.
(WebCore::AccessibilityObject::axObjectCache const): Tweak coding style.
(WebCore::AccessibilityObject::focusedUIElement const): Use existing page
function to streamline.

* accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::accessKey const): Use
attributeWithoutSynchronization for efficiency and consistency with other
code working with the accesskey attribute.

* dom/ContainerNode.cpp:
(WebCore::ContainerNode::childrenChanged): Added call to
invalidateAccessKeyCache, replacing old scheme tied to style recalculation.

* dom/DecodedDataDocumentParser.cpp:
(WebCore::DecodedDataDocumentParser::appendBytes): Update for name and return
type change of createDecoderIfNeeded to decoder.
(WebCore::DecodedDataDocumentParser::flush): Ditto.

* dom/Document.cpp:
(WebCore::Document::elementForAccessKey): Renamed from getElementForAccessKey
to match WebKit coding style. Changed to use unique_ptr instead of separate
boolean to keep track of map validity status.
(WebCore::Document::buildAccessKeyCache): Renamed from buildAccessKeyMap to
clarify that this is a cache. Changed to use composedTreeDescendants rather
than explicit calls to shadowRoot. Use add instead of set so that first element
in document order wins, instead of last element in document order. Updated
to make a new map in a new unique_ptr instead of populating a map.
(WebCore::Document::invalidateAccessKeyCacheSlowCase): Renamed from
invalidateAccessKeyMap, and left an inline part in the header so the fast case
of quickly checking for a null pointer can be inlined.
(WebCore::Document::doctype const): Use downcast instead of static_cast.
(WebCore::Document::scheduleStyleRecalc): Moved call to invalidateAccessKeyMap
from here to childrenChanged and accesskey attribute change handling.
(WebCore::Document::processFormatDetection): Set m_isTelephoneNumberParsingAllowed
directly since this is the only place that does it and we don't need to factor
that one line of code into a function.
(WebCore::Document::getOverrideStyle): Moved to header since it's just a stub
that always returns nullptr and can be inlined.
(WebCore::Document::setIsTelephoneNumberParsingAllowed): Deleted.
(WebCore::Document::ensureTemplateDocument): Removed nullptr frame argument to
the create function, since Document::create now always involves no frame.
(WebCore::Document::didAssociateFormControl): Changed argument type to a reference
and simplified the logic with a local variable.
(WebCore::Document::didAssociateFormControlsTimerFired): Simplified the null
checks and rearranged things so that m_associatedFormControls will always
get cleared even if the document is no longer associated with a page.

* dom/Document.h: Removed unnnecessary explicit values for enumerations (first
value is always zero, etc.) and formatted simple enumerations in a single line
for easier reading. Moved Document::create fucntion bodies out of line, removed
the frame argument from the simple "create with URL" overload and made the frame
argument for createNonRenderedPlaceholder be a reference rather than a pointer.
Renamed getElementByAccessKey to elementForAccessKey, invalidateAccessKeyMap to
invalidateAccessKeyCache, buildAccessKeyMap to buildAccessKeCache,
m_elementsByAccessKey to m_accessKeyCache and changed its type.
Removed bogus "DOM methods" comment, unused setParserFeature friend declaration,
setIsTelephoneNumberParsingAllowed function, and m_accessKeyMapValid flag.

* dom/Document.idl: Added comment highlighting that getOverrideStyle is just a
placeholder returning null.

* dom/Element.cpp:
(WebCore::Element::attributeChanged): Call invalidateAccessKeyCache when the
value of the accesskey attribute is changed. Also moved the class attribute code
so the attributes here are in alphabetical order (only class and id were out of
alphabetical order).

* dom/Node.cpp:
(WebCore::Node::isDescendantOrShadowDescendantOf const): Rewrote to no longer
use deprecatedShadowAncestorNode and used boolean operators to make it a
single line and easier to understand. Also added a FIXME since the
containsIncludingShadowDOM function is so similar, yet differently written.
(WebCore::Node::contains const): Rewrote as a single line to make this easier
to read and to *slightly* improve the speed in the "this == node" case.
(WebCore::Node::containsIncludingHostElements const): Use downcast.
(WebCore::Node::deprecatedShadowAncestorNode const): Deleted.

* dom/Node.h: Deleted now-unused deprecatedShadowAncestorNode.

* editing/ReplaceSelectionCommand.cpp:
(WebCore::ReplacementFragment::ReplacementFragment): Rewrote to use shadowHost
instead of deprecatedShadowAncestorNode.

* html/FormAssociatedElement.cpp:
(WebCore::FormAssociatedElement::resetFormOwner): Pass reference to
didAssociateFormControl.
(WebCore::FormAssociatedElement::formAttributeChanged): Ditto.

* html/HTMLAreaElement.cpp:
(WebCore::HTMLAreaElement::parseAttribute): Removed special case for accesskey
attribute, because we want to call the base class parseAttribute in that case.

* html/HTMLFormElement.cpp:
(WebCore::HTMLFormElement::insertedIntoAncestor): Pass reference to
didAssociateFormControl.

* html/HTMLSelectElement.cpp:
(WebCore::HTMLSelectElement::parseAttribute): Removed special case for accesskey
attribute with mysterious FIXME, because we want to call the base class
parseAttribute in that case. The old code had no effect before; the access key
logic would still find the attribute; if the goal is to ignore the attribute
for these elements we will need a different solution.
* html/HTMLTextAreaElement.cpp:
(WebCore::HTMLTextAreaElement::parseAttribute): Ditto.

* loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::DocumentLoader): Removed code that was always passing
nullptr as the frame for the newly created DocumentWriter. It used m_frame, but
m_frame was always null.
(WebCore::DocumentLoader::attachToFrame): Pass reference to DocumentWriter::setFrame.

* loader/DocumentWriter.cpp:
(WebCore::DocumentWriter::DocumentWriter): Deleted. The old version set m_frame to
the passed in frame, which was always nullptr, and initialized some booleans and an
enumeration, which are now initialized in the class definition.
(WebCore::DocumentWriter::createDocument): Pass a reference for the frame.
(WebCore::DocumentWriter::begin): Updated now that State is an enum class.
(WebCore::DocumentWriter::decoder): Renamed from createDecoderIfNeeded and changed
to return a reference.
(WebCore::DocumentWriter::addData): Use RELEASE_ASSERT instead of if/CRASH and
updated now that State is an enum class.
(WebCore::DocumentWriter::insertDataSynchronously): Updated now that State is an
enum class.
(WebCore::DocumentWriter::end): Ditto.

* loader/DocumentWriter.h: Removed unneeded include and forward declaration.
Removed the frame pointer argument to the constructor, caller was always passing a
nullptr. Changed argument to setFrame to be a reference. Renamed createDecoderIfNeeded
to decoder and changed it to return a reference. Initialized m_frame,
m_hasReceivedSomeData, m_encodingWasChosenByUser, and m_state here so we don't need
to initialize them in a constructor. Renamed the enum from WriterState to State since
it's a member of DocumentWriter already, and made it an enum class rather than ending
each enumeration value with WritingState.

* page/DragController.cpp:
(WebCore::isEnabledColorInput): Removed boolean argument setToShadowAncestor. The
one caller that formerly passed true now calls the new hasEnabledColorInputAsShadowHost
function instead.
(WebCore::hasEnabledColorInputAsShadowHost): Added.
(WebCore::elementUnderMouse): Use shadowHost instead of deprecatedShadowAncestorNode.
Also added FIXME since it seems this should instead be using parentElementInComposedTree.
(WebCore::DragController::concludeEditDrag): Removed "false" argument to isEnabledColorInput.
(WebCore::DragController::canProcessDrag): Removed "true" argument to isEnabledColorInput
and added call to hasEnabledColorInputAsShadowHost. Also put the value of the node to drag
into a local variable to simplify code.
(WebCore::DragController::draggableElement const): Removed "false" argument to isEnabledColorInput.

* page/EventHandler.cpp:
(WebCore::EventHandler::handleAccessKey): Update name of elementForAccessKey.

* page/FocusController.cpp:
(WebCore::clearSelectionIfNeeded): Use shadowHost instead of deprecatedShadowAncestorNode.

* workers/service/context/ServiceWorkerThreadProxy.cpp:
(WebCore::createPageForServiceWorker): Pass reference instead of pointer for frame to
Document::createNonRenderedPlaceholder.

Source/WebInspectorUI:

* UserInterface/Models/NativeFunctionParameters.js: Removed getOverrideStyle
because this deprecated non-working function is no longer something web authors
should be thinking about or typing.

Source/WebKit:

* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMDocumentGtk.cpp:
(webkit_dom_document_get_override_style): Changed to return nullptr directly instead of
calling a bunch of WebCore code first. No real value in checking arguments or doing any work.

* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::shouldAlwaysUsePluginDocument const): Deleted.
No need to override the function inherited from the base class, nor to keep track of the
unnecessary task of implementing this some day.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h: Updated for the above.

Source/WebKitLegacy/mac:

* DOM/DOMDocument.mm:
(-[DOMDocument getOverrideStyle:pseudoElement:]): Return nil without calling
through to WebCore or even looking at the arguments.
(-[DOMDocument getOverrideStyle::]): Ditto.

LayoutTests:

fast/forms/access-key-shadow-and-ordering-expected.txt: Added.
fast/forms/access-key-shadow-and-ordering.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFix the test failure following r241918
dbates@webkit.org [Fri, 22 Feb 2019 06:46:06 +0000 (06:46 +0000)]
Fix the test failure following r241918
(https://bugs.webkit.org/show_bug.cgi?id=194906)

For some reason strict cookies are sent via HTTP in WK1 and not in WK2. Will investigate in <https://bugs.webkit.org/show_bug.cgi?id=194933>.
This sub-test failure occurs with and without the patch. As the primary purpose of this test
was to ensure correctness for Same Site lax cookies, I've amended the test and test result
for now.

* http/tests/cookies/same-site/user-load-cross-site-redirect-expected.txt:
* http/tests/cookies/same-site/user-load-cross-site-redirect.php:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRemoved a duplicate ChangeLog entry.
aestes@apple.com [Fri, 22 Feb 2019 04:29:53 +0000 (04:29 +0000)]
Removed a duplicate ChangeLog entry.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUpdate JSScript SPI based on feedback
sbarati@apple.com [Fri, 22 Feb 2019 04:21:54 +0000 (04:21 +0000)]
Update JSScript SPI based on feedback
https://bugs.webkit.org/show_bug.cgi?id=194517

Reviewed by Keith Miller.

This patch updates the JSScript SPI in the following ways:
- JSScript can now represent both modules and programs. This is a property
of the script determined during creation.
- JSScript now takes a sourceURL during construction. For modules, this acts
as the module identifier.
- JSScript now has SPI for writing the cache out to disk. We don't do this
automatically.
- JSScript will load the bytecode cache on creation if it exists.
- We retrofit these new requirements on the prior JSScript SPI that
we're going to remove as soon as we can: https://bugs.webkit.org/show_bug.cgi?id=194909.
Previous SPI assumes all JSScripts are modules. Previous SPI also assigns
a sourceURL to the JSScript based on what the module loader decided the
identifier should be. We'll remove this once we remove the old SPI.

This patch also adds SPI to JSContext to evaluate a JSScript. For modules,
this is like returning the result of doing dynamic import. For programs,
this does normal program evaluation.

This patch also fixes a bug in generateBytecode/generateModuleBytecode where
we would try to cache the bytecode even if recursivelyGenerateUnlinkedCodeBlock
returned null. E.g, if the script had a syntax error.

When writing tests, I also discovered that someone previously broke
testapi. This patch also fixes those failures. They were broken when
we switched to using a testapiScripts directory to hold our test .js
scripts.

* API/JSAPIGlobalObject.h:
* API/JSAPIGlobalObject.mm:
(JSC::JSAPIGlobalObject::moduleLoaderResolve):
(JSC::JSAPIGlobalObject::moduleLoaderFetch):
(JSC::JSAPIGlobalObject::loadAndEvaluateJSScriptModule):
* API/JSBase.cpp:
(JSEvaluateScriptInternal):
(JSEvaluateScript):
* API/JSBaseInternal.h: Added.
* API/JSContext.mm:
(-[JSContext evaluateScript:withSourceURL:]):
(-[JSContext evaluateJSScript:]):
* API/JSContextPrivate.h:
* API/JSScript.h:
* API/JSScript.mm:
(+[JSScript scriptWithSource:inVirtualMachine:]):
(+[JSScript scriptFromASCIIFile:inVirtualMachine:withCodeSigning:andBytecodeCache:]):
(createError):
(+[JSScript scriptOfType:inVirtualMachine:withSourceURL:andSource:andBytecodeCache:error:]):
(+[JSScript scriptOfType:inVirtualMachine:memoryMappedFromASCIIFile:withSourceURL:andBytecodeCache:error:]):
(-[JSScript cacheBytecodeWithError:]):
(-[JSScript sourceURL]):
(-[JSScript type]):
(-[JSScript jsSourceCode]):
(-[JSScript writeCache:]):
(-[JSScript setSourceURL:]):
(-[JSScript forceRecreateJSSourceCode]):
(-[JSScript writeCache]): Deleted.
(-[JSScript jsSourceCode:]): Deleted.
* API/JSScriptInternal.h:
* API/tests/FunctionOverridesTest.cpp:
(testFunctionOverrides):
* API/tests/testapi.c:
(main):
* API/tests/testapi.mm:
(tempFile):
(testModuleBytecodeCache):
(testProgramBytecodeCache):
(testBytecodeCacheWithSyntaxError):
(testProgramJSScriptException):
(testLoadBasicFileLegacySPI):
(+[JSContextMemoryMappedLoaderDelegate newContext]):
(-[JSContextMemoryMappedLoaderDelegate context:fetchModuleForIdentifier:withResolveHandler:andRejectHandler:]):
(testLoadBasicFile):
(+[JSContextAugmentedLoaderDelegate newContext]):
(-[JSContextAugmentedLoaderDelegate context:fetchModuleForIdentifier:withResolveHandler:andRejectHandler:]):
(testJSScriptURL):
(testObjectiveCAPI):
(testBytecodeCache): Deleted.
* API/tests/testapiScripts/foo.js: Added.
* JavaScriptCore.xcodeproj/project.pbxproj:
* runtime/Completion.cpp:
(JSC::generateBytecode):
(JSC::generateModuleBytecode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoDiscard cached processes when clearing website data store
rniwa@webkit.org [Fri, 22 Feb 2019 02:58:40 +0000 (02:58 +0000)]
Discard cached processes when clearing website data store
https://bugs.webkit.org/show_bug.cgi?id=194894

Reviewed by Chris Dumez.

Source/WebKit:

Clear the process cache when clearing the website data store so that there is no way to infer
which site the user had visited by observing for which sites WebContent processes had been cached.

There is one sublty in WebsiteDataStore::removeData that we have to delay the clearing of
the web process cache until the next run loop because SuspendedPageProxy::~SuspendedPageProxy
invokes WebProcessProxy::maybeShutDown in the next run loop. We also have to disable the process
cache during this time as it would otherwise trigger the responsiveness check of WebContent process
can take arbitrarily long time.

* UIProcess/API/Cocoa/WKProcessPool.mm:
(-[WKProcessPool _processCacheCapacity]): Added for testing.
* UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::addProcess): Avoid adding web processes to the cache while the suspended
pages are being cleared.
* UIProcess/WebProcessCache.h:
(WebKit::WebProcessCache::disabled const): Added.
(WebKit::WebProcessCache::setDisabled): Added.
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::handleMemoryPressureWarning):
(WebKit::WebProcessPool::clearSuspendedPages): Added.
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeData):

Tools:

Added a test case.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
(TestWebKitAPI.ProcessSwap.NumberOfCachedProcesses): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAdd more doesGC() assertions.
mark.lam@apple.com [Fri, 22 Feb 2019 02:02:32 +0000 (02:02 +0000)]
Add more doesGC() assertions.
https://bugs.webkit.org/show_bug.cgi?id=194911
<rdar://problem/48285723>

Reviewed by Saam Barati and Yusuke Suzuki.

* dfg/DFGOSRExit.cpp:
(JSC::DFG::OSRExit::compileOSRExit):
- Set expectDoesGC here because we no longer have to worry about missing store
  barriers in optimized code after this point.  This will prevent false positive
  assertion failures arising from functions called beneath compileOSRExit().

(JSC::DFG::OSRExit::compileExit):
- Add a comment to explain why the generated ramp needs to set expectDoesGC even
  though compileOSRExit() also sets it.  Reason: compileOSRExit() is only called
  for the first OSR from this code origin, the generated ramp is called for many
  subsequents OSR exits from this code origin.

* ftl/FTLOSRExitCompiler.cpp:
(JSC::FTL::compileStub):
- Added a comment for the equivalent reason to the one above.

(JSC::FTL::compileFTLOSRExit):
- Set expectDoesGC here because we no longer have to worry about missing store
  barriers in optimized code after this point.  This will prevent false positive
  assertion failures arising from functions called beneath compileFTLOSRExit().

* heap/CompleteSubspace.cpp:
(JSC::CompleteSubspace::tryAllocateSlow):
* heap/CompleteSubspaceInlines.h:
(JSC::CompleteSubspace::allocateNonVirtual):
- assert expectDoesGC.

* heap/DeferGC.h:
(JSC::DeferGC::~DeferGC):
- assert expectDoesGC.
- Also added WTF_FORBID_HEAP_ALLOCATION to DeferGC, DeferGCForAWhile, and DisallowGC
  because all 3 should be stack allocated RAII objects.

* heap/GCDeferralContext.h:
* heap/GCDeferralContextInlines.h:
(JSC::GCDeferralContext::~GCDeferralContext):
- Added WTF_FORBID_HEAP_ALLOCATION.
- assert expectDoesGC.

* heap/Heap.cpp:
(JSC::Heap::collectNow):
(JSC::Heap::collectAsync):
(JSC::Heap::collectSync):
(JSC::Heap::stopIfNecessarySlow):
(JSC::Heap::collectIfNecessaryOrDefer):
* heap/HeapInlines.h:
(JSC::Heap::acquireAccess):
(JSC::Heap::stopIfNecessary):
* heap/LargeAllocation.cpp:
(JSC::LargeAllocation::tryCreate):
* heap/LocalAllocatorInlines.h:
(JSC::LocalAllocator::allocate):
- conservatively assert expectDoesGC on these functions that may trigger a GC
  though they don't always do.

* runtime/DisallowScope.h:
- DisallowScope should be stack allocated because it's an RAII object.

* runtime/JSCellInlines.h:
(JSC::tryAllocateCellHelper):
- Remove the expectDoesGC assertion because it is now covered by assertions in
  CompleteSubspace, LargeAllocation, and LocalAllocator.

* runtime/RegExpMatchesArray.h:
(JSC::createRegExpMatchesArray):
- assert expectDoesGC.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoClicking "Go Back" on a safe browsing warning before a WKWebView has loaded any page...
commit-queue@webkit.org [Fri, 22 Feb 2019 01:58:34 +0000 (01:58 +0000)]
Clicking "Go Back" on a safe browsing warning before a WKWebView has loaded any page should request to close the WKWebView
https://bugs.webkit.org/show_bug.cgi?id=194914
<rdar://problem/47586889>

Patch by Alex Christensen <achristensen@webkit.org> on 2019-02-21
Reviewed by Geoffrey Garen.

Source/WebKit:

* UIProcess/API/APIUIClient.h:
(API::UIClient::didShowSafeBrowsingWarning):
(API::UIClient::didClickGoBackFromSafeBrowsingWarning): Deleted.
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::didClickGoBackFromSafeBrowsingWarning): Deleted.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNavigationAction):

Tools:

* TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm:
(-[SafeBrowsingNavigationDelegate webViewDidClose:]):
(TEST):
(-[SafeBrowsingNavigationDelegate _webViewDidClickGoBackFromSafeBrowsingWarning:]): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoTurn a high-value UI-side-compositing assertion into a release assert
timothy_horton@apple.com [Fri, 22 Feb 2019 01:51:32 +0000 (01:51 +0000)]
Turn a high-value UI-side-compositing assertion into a release assert
https://bugs.webkit.org/show_bug.cgi?id=194887

Reviewed by Simon Fraser.

* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.cpp:
(WebKit::PlatformCALayerRemote::recursiveBuildTransaction):
There is absolutely no situation in which this assert should fire
if WebKit is behaving correctly, and it will basically always result
in a UI process crash (or at least confusion), so make it a release assert.
This would have saved us a lot of stress in bug 194845.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[MSVC] Fix compilation errors with lambdas in ResourceLoadStatistics
don.olmstead@sony.com [Fri, 22 Feb 2019 01:05:08 +0000 (01:05 +0000)]
[MSVC] Fix compilation errors with lambdas in ResourceLoadStatistics
https://bugs.webkit.org/show_bug.cgi?id=194913

Reviewed by Brent Fulgham.

MSVC has problems with the scoping of `this` within a nested lambda. To work around the
behavior the enclosing lambda's wrapping of `this` is used within the nested lambda.

* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::processStatisticsAndDataRecords):
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccessGranted):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[JSC] Use Fast Malloc as much as possible
ysuzuki@apple.com [Fri, 22 Feb 2019 01:04:01 +0000 (01:04 +0000)]
[JSC] Use Fast Malloc as much as possible
https://bugs.webkit.org/show_bug.cgi?id=194316

Reviewed by Mark Lam.

We should use Fast Malloc as much as possible to offer the whole memory view to bmalloc.

* inspector/scripts/codegen/cpp_generator_templates.py:
* inspector/scripts/tests/all/expected/definitions-with-mac-platform.json-result:
* inspector/scripts/tests/generic/expected/enum-values.json-result:
* inspector/scripts/tests/generic/expected/events-with-optional-parameters.json-result:
* inspector/scripts/tests/generic/expected/generate-domains-with-feature-guards.json-result:
* inspector/scripts/tests/mac/expected/definitions-with-mac-platform.json-result:
* jit/ExecutableAllocator.h:
* jsc.cpp:
* runtime/JSRunLoopTimer.h:
* tools/VMInspector.h:
* wasm/WasmThunks.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoLeak of CFErrorRef objects (1.92 Kbytes) in com.apple.WebKit.WebContent.Development...
ddkilzer@apple.com [Fri, 22 Feb 2019 00:03:25 +0000 (00:03 +0000)]
Leak of CFErrorRef objects (1.92 Kbytes) in com.apple.WebKit.WebContent.Development running WebKit layout tests on iOS Simulator
<https://webkit.org/b/194761>

Reviewed by Myles Maxfield.

* WebKitTestRunner/InjectedBundle/cocoa/ActivateFontsCocoa.mm:
(WTR::activateSystemCoreWebFonts): Log `error` to console if
CTFontManagerRegisterFontsForURL() fails (macOS only).  Update
to use ARC-compliant cast.
(WTR::activateFonts): Change `0` to `nullptr`.  Update to use
ARC-compliant cast.
(WTR::installFakeHelvetica): Log `error` to console if
CTFontManagerRegisterFontsForURL() fails, but only on supported
platforms (macOS Mojave/iOS 12 and newer, not iOS Simulator).
Call CFRelease() on `error` to fix the leak.  Update to use
ARC-compliant cast.
(WTR::uninstallFakeHelvetica): Update to use ARC-compliant
cast.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFix API test crashes after r241855.
cdumez@apple.com [Thu, 21 Feb 2019 23:59:29 +0000 (23:59 +0000)]
Fix API test crashes after r241855.

Reviewed by Ryosuke Niwa.

* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::addProcessIfPossible):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoSame Site Lax cookies are not sent with cross-site redirect from client-initiated...
dbates@webkit.org [Thu, 21 Feb 2019 23:54:40 +0000 (23:54 +0000)]
Same Site Lax cookies are not sent with cross-site redirect from client-initiated load
https://bugs.webkit.org/show_bug.cgi?id=194906
<rdar://problem/44305947>

Reviewed by Brent Fulgham.

Source/WebCore:

Ensure that a request for a top-level navigation is annotated as such regardless of whether
the request has a computed Same Site policy.

"New loads" initiated by a the client (Safari) either by API or a human either explicitly
typing a URL in the address bar or Command + clicking a hyperlink to open it in a new window/tab
are always considered Same Site. This is by definition from the spec. [1] as we aren't navigating
from an existing page. (Command + click should be thought of as a convenience to the user from
having to copy the hyperlink's URL, create a new window, and paste the URL into the address bar).
Currently the frame loader marks a request as a top-level navigation if and only if the request
does not have a pre-computed Same Site policy. However, "New loads" have a pre-computed Same Site
policy. So, these loads would never be marked as a top-level navigation by the frame loading code.
Therefore, if the "new load" turned out to be a cross-site redirect then WebKit would incorrectly
tell the networking stack that the load was a cross-site, non-top-level navigation, and per the
Same Site spec [2], the networking stack would not send Same Site Lax cookies. Instead,
WebKit should unconditionally ensure that requests are marked as a top-level navigation, if applicable.

[1] See Note for (1) in  <https://tools.ietf.org/html/draft-ietf-httpbis-rfc6265bis-02#section-5.2>
[2] <https://tools.ietf.org/html/draft-ietf-httpbis-rfc6265bis-02#section-5.3.7.1>

Test: http/tests/cookies/same-site/user-load-cross-site-redirect.php

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::addExtraFieldsToRequest): Unconditionally update the request's top-
level navigation bit.
* platform/network/ResourceRequestBase.cpp:
(WebCore::ResourceRequestBase::setAsIsolatedCopy): Unconditionally copy a request's top-
level navigation bit.

LayoutTests:

Add a test that is representative of a user loading a cross-site page that redirects
to a page that expects Same Site Lax cookies.

* http/tests/cookies/same-site/user-load-cross-site-redirect-expected.txt: Added.
* http/tests/cookies/same-site/user-load-cross-site-redirect.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFlaky API Test: TestWebKitAPI.ProcessSwap.SuspendedPageLimit
cdumez@apple.com [Thu, 21 Feb 2019 23:53:11 +0000 (23:53 +0000)]
Flaky API Test: TestWebKitAPI.ProcessSwap.SuspendedPageLimit
https://bugs.webkit.org/show_bug.cgi?id=194481

Reviewed by Alex Christensen.

Give the processes a chance to shutdown and wait until we reached the expected number
of processes.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFix the macOS build
timothy_horton@apple.com [Thu, 21 Feb 2019 23:16:20 +0000 (23:16 +0000)]
Fix the macOS build

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoLayout Test fast/text/international/khmer-selection.html is crashing
pvollan@apple.com [Thu, 21 Feb 2019 23:12:36 +0000 (23:12 +0000)]
Layout Test fast/text/international/khmer-selection.html is crashing
https://bugs.webkit.org/show_bug.cgi?id=191368

Reviewed by Brent Fulgham.

Source/WebCore:

GlyphBuffer's offset array wasn't getting filled by UniscribeController.
Our underlining code requires this array.

Uniscribe gives us a character -> glyph mapping, so we just have to compute
the inverse and give it to the GlyphBuffer.

This patch is written by Myles C. Maxfield.

Test: fast/text/international/khmer-selection.html.

* platform/graphics/GlyphBuffer.h:
(WebCore::GlyphBuffer::add):
* platform/graphics/displaylists/DisplayListItems.cpp:
(WebCore::DisplayList::DrawGlyphs::generateGlyphBuffer const):
* platform/graphics/win/UniscribeController.cpp:
(WebCore::UniscribeController::advance):
(WebCore::UniscribeController::itemizeShapeAndPlace):
(WebCore::UniscribeController::shapeAndPlaceItem):
* platform/graphics/win/UniscribeController.h:

LayoutTests:

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agocontentfiltering tests leak documents
aestes@apple.com [Thu, 21 Feb 2019 23:11:58 +0000 (23:11 +0000)]
contentfiltering tests leak documents
https://bugs.webkit.org/show_bug.cgi?id=189434
<rdar://44239943>

Reviewed by Simon Fraser.

Changed ContentFilter to capture the blocked Frame as a WeakPtr to break a reference cycle.

This fixes world leaks in several tests in LayoutTests/contentfiltering/.

* bindings/js/ScriptController.h:
* loader/ContentFilter.cpp:
(WebCore::ContentFilter::didDecide):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoIndexedDB: leak UniqueIDBDatabase in layout tests
sihui_liu@apple.com [Thu, 21 Feb 2019 23:06:23 +0000 (23:06 +0000)]
IndexedDB: leak UniqueIDBDatabase in layout tests
https://bugs.webkit.org/show_bug.cgi?id=194870
<rdar://problem/48163812>

Reviewed by Geoffrey Garen.

UniqueIDBDatabase owns a pointer to itself after it is hard closed. It should release the pointer when it
receives confirmation from clients and all pending tasks are done. UniqueIDBDatabase already checks whether the
pointer should be released when a database task finishes, but it didn't perform a check when a confirm message
is received.

No new test as the order of task completion and confirmation arrival is uncertain.

* Modules/indexeddb/server/UniqueIDBDatabase.cpp:
(WebCore::IDBServer::UniqueIDBDatabase::confirmDidCloseFromServer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFix ProcessSwap.PageOverlayLayerPersistence on macOS
timothy_horton@apple.com [Thu, 21 Feb 2019 22:55:05 +0000 (22:55 +0000)]
Fix ProcessSwap.PageOverlayLayerPersistence on macOS

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
I am a backwards.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agocontentfiltering tests leak documents
aestes@apple.com [Thu, 21 Feb 2019 22:51:13 +0000 (22:51 +0000)]
contentfiltering tests leak documents
https://bugs.webkit.org/show_bug.cgi?id=189434
<rdar://44239943>

Reviewed by Simon Fraser.

Changed ContentFilter to capture the blocked Frame as a WeakPtr to break a reference cycle.

This fixes world leaks in several tests in LayoutTests/contentfiltering/.

* bindings/js/ScriptController.h:
* loader/ContentFilter.cpp:
(WebCore::ContentFilter::didDecide):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoFix ProcessSwap.PageOverlayLayerPersistence on macOS
timothy_horton@apple.com [Thu, 21 Feb 2019 22:50:14 +0000 (22:50 +0000)]
Fix ProcessSwap.PageOverlayLayerPersistence on macOS

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:
We can't find the UI-side layer on macOS because we don't use UI-side compositing.
We could force on UI-side compositing, but I feel like the test might be more
valuable exercising the normal path on each platform (though the original
bug was also UI-side specific). Though we can't validate that the overlay
layer was installed successfully, there are layout tests that do that.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRotation animations sometimes use the wrong origin (affects apple.com)
dino@apple.com [Thu, 21 Feb 2019 22:36:32 +0000 (22:36 +0000)]
Rotation animations sometimes use the wrong origin (affects apple.com)
https://bugs.webkit.org/show_bug.cgi?id=194878
<rdar://problem/43908047>

Reviewed by Simon Fraser.

Forgot the expected results file.

* animations/additive-transform-animations-expected.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAPI::HTTPCookieStore should expose setCookies()
achristensen@apple.com [Thu, 21 Feb 2019 22:34:17 +0000 (22:34 +0000)]
API::HTTPCookieStore should expose setCookies()
https://bugs.webkit.org/show_bug.cgi?id=194861

Reviewed by Brent Fulgham.

Pass a Vector<Cookie> instead of just one cookie at a time.
This will add the ability to set multiple cookies without an IPC roundtrip per cookie.

* NetworkProcess/Cookies/WebCookieManager.cpp:
(WebKit::WebCookieManager::setCookie):
* NetworkProcess/Cookies/WebCookieManager.h:
* NetworkProcess/Cookies/WebCookieManager.messages.in:
* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::setCookies):
(API::HTTPCookieStore::setCookie): Deleted.
* UIProcess/API/APIHTTPCookieStore.h:
* UIProcess/API/Cocoa/WKHTTPCookieStore.mm:
(-[WKHTTPCookieStore setCookie:completionHandler:]):
* UIProcess/WebCookieManagerProxy.cpp:
(WebKit::WebCookieManagerProxy::setCookies):
(WebKit::WebCookieManagerProxy::setCookie): Deleted.
* UIProcess/WebCookieManagerProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoMove UIWebTouchEventsGestureRecognizer.activeTouchesByIdentifier to SPI
commit-queue@webkit.org [Thu, 21 Feb 2019 21:43:43 +0000 (21:43 +0000)]
Move UIWebTouchEventsGestureRecognizer.activeTouchesByIdentifier to SPI
https://bugs.webkit.org/show_bug.cgi?id=194531
<rdar://problem/47714562>

Patch by Antoine Quint <graouts@apple.com> on 2019-02-21
Reviewed by Tim Horton.

Follow-up commit to ensure this SPI is only called on newer versions of iOS.

Source/WebKit:

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cancelPointersForGestureRecognizer:]):

Source/WTF:

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoCrash under RemoteLayerTreePropertyApplier::applyProperties when reattaching to old...
timothy_horton@apple.com [Thu, 21 Feb 2019 21:34:26 +0000 (21:34 +0000)]
Crash under RemoteLayerTreePropertyApplier::applyProperties when reattaching to old process
https://bugs.webkit.org/show_bug.cgi?id=194845
<rdar://problem/47944579>

Reviewed by Antti Koivisto.

Source/WebKit:

New test: ProcessSwap.PageOverlayLayerPersistence

* Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
(WebKit::RemoteLayerTreeTransaction::LayerProperties::notePropertiesChanged):
Keep track of all properties that have ever changed on a layer, so we
can re-send them if a layer moves between contexts.

* WebProcess/WebPage/DrawingArea.cpp:
(WebKit::DrawingArea::~DrawingArea):
(WebKit::DrawingArea::removeMessageReceiverIfNeeded):
* WebProcess/WebPage/DrawingArea.h:
(WebKit::DrawingArea::layerHostDidFlushLayers):
Make it possible to tear down DrawingArea's MessageReceiver before it is destroyed,
so that we can keep two DrawingAreas alive in a single process for a short time.

(WebKit::DrawingArea::adoptLayersFromDrawingArea):
Add adoptLayersFromDrawingArea; see below for its only useful implementation.

* WebProcess/WebPage/RemoteLayerTree/GraphicsLayerCARemote.cpp:
(WebKit::GraphicsLayerCARemote::GraphicsLayerCARemote):
(WebKit::GraphicsLayerCARemote::~GraphicsLayerCARemote):
(WebKit::GraphicsLayerCARemote::createPlatformCALayer):
(WebKit::GraphicsLayerCARemote::createPlatformCALayerForEmbeddedView):
(WebKit::GraphicsLayerCARemote::moveToContext):
* WebProcess/WebPage/RemoteLayerTree/GraphicsLayerCARemote.h:
Make it possible to move a GraphicsLayerCARemote between RemoteLayerTreeContexts.

* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.cpp:
(WebKit::PlatformCALayerRemote::create):
(WebKit::PlatformCALayerRemote::createForEmbeddedView):
(WebKit::PlatformCALayerRemote::~PlatformCALayerRemote):
(WebKit::PlatformCALayerRemote::moveToContext):
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemote.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCALayerRemoteCustom.mm:
(WebKit::PlatformCALayerRemoteCustom::create):
(WebKit::PlatformCALayerRemoteCustom::clone const):
Make it possible to move a PlatformCALayerRemote between RemoteLayerTreeContexts.

* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeContext.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeContext.mm:
(WebKit::RemoteLayerTreeContext::~RemoteLayerTreeContext):
(WebKit::RemoteLayerTreeContext::adoptLayersFromContext):
(WebKit::RemoteLayerTreeContext::layerDidEnterContext):
(WebKit::RemoteLayerTreeContext::layerWillLeaveContext):
(WebKit::RemoteLayerTreeContext::graphicsLayerDidEnterContext):
(WebKit::RemoteLayerTreeContext::graphicsLayerWillLeaveContext):
(WebKit::RemoteLayerTreeContext::layerWasCreated): Deleted.
(WebKit::RemoteLayerTreeContext::layerWillBeDestroyed): Deleted.
Keep track of all GraphicsLayerCARemote instances in the context, like we
do for PlatformCALayerRemote, so that we can update their context backpointers if needed.

Also make it possible to move all outstanding layers to a new context.

* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.h:
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::adoptLayersFromDrawingArea):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::reinitializeWebPage):
When doing a DrawingArea swap, transition all layers from the old context
to the new one. In order to do this, we temporarily keep both DrawingAreas
alive, but make use of the new mechanism to remove the old one's MessageReceiver
before installing the new one, so that destroying the old one later doesn't
remove it (avoiding re-introducing bug 189481).

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[CMake][Win] Fix !USE(CF) build of WebCore
don.olmstead@sony.com [Thu, 21 Feb 2019 21:06:37 +0000 (21:06 +0000)]
[CMake][Win] Fix !USE(CF) build of WebCore
https://bugs.webkit.org/show_bug.cgi?id=194879

Reviewed by Konstantin Tokarev.

* PlatformAppleWin.cmake:
* PlatformWin.cmake:
* PlatformWinCairo.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241876 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: Canvas: recordings with a single frame sometimes missing TreeElement
drousso@apple.com [Thu, 21 Feb 2019 19:59:01 +0000 (19:59 +0000)]
Web Inspector: Canvas: recordings with a single frame sometimes missing TreeElement
https://bugs.webkit.org/show_bug.cgi?id=194687

Reviewed by Joseph Pecoraro.

* UserInterface/Views/CanvasSidebarPanel.js:
(WI.CanvasSidebarPanel):
Wait until the entire `Recording` has been processed before setting the selected action.

* UserInterface/Views/CanvasSidebarPanel.css:
(.sidebar > .panel.navigation.canvas.has-recordings > .content > .recording-content):
Drive-by: ensure that the recording `TreeOutline` container element has the full height to
take advantage of for virtualization.

* UserInterface/Views/RecordingStateDetailsSidebarPanel.js:
(WI.RecordingStateDetailsSidebarPanel.prototype.set recording):
Drive-by: clear the `DataGrid` list when changing recordings since we removed all subviews.
git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241875 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: DOM Debugger: node-specific event breakpoints aren't cleared/recalcula...
drousso@apple.com [Thu, 21 Feb 2019 18:56:36 +0000 (18:56 +0000)]
Web Inspector: DOM Debugger: node-specific event breakpoints aren't cleared/recalculated on navigation
https://bugs.webkit.org/show_bug.cgi?id=194874
<rdar://problem/48253415>

Reviewed by Joseph Pecoraro.

* UserInterface/Controllers/DOMManager.js:
(WI.DOMManager.prototype._setDocument):
(WI.DOMManager.prototype.removeBreakpointForEventListener):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241874 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRotation animations sometimes use the wrong origin (affects apple.com)
dino@apple.com [Thu, 21 Feb 2019 18:14:06 +0000 (18:14 +0000)]
Rotation animations sometimes use the wrong origin (affects apple.com)
https://bugs.webkit.org/show_bug.cgi?id=194878
<rdar://problem/43908047>

Follow-up commit to ensure this change only affects newer versions
of iOS.

* wtf/Platform.h: Add a version check.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241873 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: macOS user agent option should be available in device settings menu
drousso@apple.com [Thu, 21 Feb 2019 18:00:50 +0000 (18:00 +0000)]
Web Inspector: macOS user agent option should be available in device settings menu
https://bugs.webkit.org/show_bug.cgi?id=194885
<rdar://problem/48259253>

Reviewed by Joseph Pecoraro.

* UserInterface/Base/Main.js:
(WI._handleDeviceSettingsToolbarButtonClicked):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241870 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[LFC][Floats] Add support for placing formatting roots in-between floats.
zalan@apple.com [Thu, 21 Feb 2019 17:58:15 +0000 (17:58 +0000)]
[LFC][Floats] Add support for placing formatting roots in-between floats.
https://bugs.webkit.org/show_bug.cgi?id=194902

Reviewed by Antti Koivisto.

Source/WebCore:

This patch add support for placing a formatting root box in-between existing floats.
The initial vertical position of a formatting root is its static position which can make the box
placed above exsiting floats (whereas we can never place a regular float above existing floats.)

Test: fast/block/block-only/floats-and-block-formatting-roots.html

* layout/blockformatting/BlockFormattingContext.cpp:
(WebCore::Layout::BlockFormattingContext::computePositionToAvoidFloats const):
* layout/floats/FloatingContext.cpp:
(WebCore::Layout::FloatPair::LeftRightIndex::isEmpty const):
(WebCore::Layout::FloatPair::isEmpty const):
(WebCore::Layout::FloatPair::operator* const):
(WebCore::Layout::Iterator::operator* const):
(WebCore::Layout::begin):
(WebCore::Layout::end):
(WebCore::Layout::FloatingContext::positionForFloat const):
(WebCore::Layout::FloatingContext::positionForFormattingContextRoot const):
(WebCore::Layout::findAvailablePosition):
(WebCore::Layout::FloatingContext::findPositionForFloatBox const):
(WebCore::Layout::FloatingContext::findPositionForFormattingContextRoot const):
(WebCore::Layout::FloatPair::FloatPair):
(WebCore::Layout::FloatPair::left const):
(WebCore::Layout::FloatPair::right const):
(WebCore::Layout::FloatPair::intersects const):
(WebCore::Layout::FloatPair::operator == const):
(WebCore::Layout::FloatPair::horizontalConstraints const):
(WebCore::Layout::FloatPair::bottom const):
(WebCore::Layout::Iterator::operator++):
(WebCore::Layout::Iterator::set):
(WebCore::Layout::FloatingPair::isEmpty const): Deleted.
(WebCore::Layout::FloatingPair::verticalConstraint const): Deleted.
(WebCore::Layout::FloatingContext::positionForFloatAvoiding const): Deleted.
(WebCore::Layout::FloatingContext::floatingPosition const): Deleted.
(WebCore::Layout::FloatingPair::FloatingPair): Deleted.
(WebCore::Layout::FloatingPair::left const): Deleted.
(WebCore::Layout::FloatingPair::right const): Deleted.
(WebCore::Layout::FloatingPair::intersects const): Deleted.
(WebCore::Layout::FloatingPair::operator == const): Deleted.
(WebCore::Layout::FloatingPair::horizontalConstraints const): Deleted.
(WebCore::Layout::FloatingPair::bottom const): Deleted.
* layout/floats/FloatingContext.h:

Tools:

* LayoutReloaded/misc/LFC-passing-tests.txt:

LayoutTests:

* fast/block/block-only/floats-and-block-formatting-roots-expected.html: Added.
* fast/block/block-only/floats-and-block-formatting-roots.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241869 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed API test fix after r241855.
cdumez@apple.com [Thu, 21 Feb 2019 17:24:51 +0000 (17:24 +0000)]
Unreviewed API test fix after r241855.

* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::~WebProcessProxy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241868 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[WPE] Do not hardcode WPEBackend-fdo library name for linking tests
aperez@igalia.com [Thu, 21 Feb 2019 17:08:07 +0000 (17:08 +0000)]
[WPE] Do not hardcode WPEBackend-fdo library name for linking tests
https://bugs.webkit.org/show_bug.cgi?id=194901

Unreviewed build fix.

.:

* Source/cmake/FindWPEBackend-fdo.cmake: Use WPEBACKEND_FDO as prefix
for the output variables instead of WPEBackend-fdo, for consistency
with other usages of of find_package_handle_standard_args().

Tools:

* TestWebKitAPI/PlatformWPE.cmake: Set TestWebKitAPIBase to be linked
against the detected ${WPEBACKEND_FDO_LIBRARIES} instead of hardcoding
WPEBackend-fdo-0.1 as library name.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241866 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoMove UIWebTouchEventsGestureRecognizer.activeTouchesByIdentifier to SPI
graouts@webkit.org [Thu, 21 Feb 2019 13:55:51 +0000 (13:55 +0000)]
Move UIWebTouchEventsGestureRecognizer.activeTouchesByIdentifier to SPI
https://bugs.webkit.org/show_bug.cgi?id=194531
<rdar://problem/47714562>

Reviewed by Dean Jackson.

* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cancelPointersForGestureRecognizer:]): We update the referenced Radar since we need to keep the call
to respondsToSelector until the SPI has shipped for OpenSource builds to work.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241865 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[GTK] Gardening for unexpected passes
dpino@igalia.com [Thu, 21 Feb 2019 12:29:17 +0000 (12:29 +0000)]
[GTK] Gardening for unexpected passes
https://bugs.webkit.org/show_bug.cgi?id=194854

Unreviewed test gardening.

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241864 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUpdate MIME type parser
commit-queue@webkit.org [Thu, 21 Feb 2019 08:39:00 +0000 (08:39 +0000)]
Update MIME type parser
https://bugs.webkit.org/show_bug.cgi?id=180526

Patch by Rob Buis <rbuis@igalia.com> on 2019-02-21
Reviewed by Darin Adler.

LayoutTests/imported/w3c:

Update improved test expectations.

* web-platform-tests/xhr/overridemimetype-blob-expected.txt:

Source/WebCore:

Further testing showed the MIME parser needs these fixes:
- stripWhitespace is wrong for removing HTTP whitespace, use
  stripLeadingAndTrailingHTTPSpaces instead.
- HTTP Token code points checking for Rfc2045 and Mimesniff were
  mixed up, use the dedicated isValidHTTPToken for Mimesniff mode.
- Quoted Strings were not unescaped/escaped, this seems ok for
  serializing but is wrong when gettings individual parameter values.
  Implement [1] and [2] Step 2.4 to properly unescape and escape.

This change also tries to avoid hard to read uses of find.

Test: ParsedContentType.Serialize

[1] https://fetch.spec.whatwg.org/#collect-an-http-quoted-string
[2] https://mimesniff.spec.whatwg.org/#serializing-a-mime-type

* platform/network/ParsedContentType.cpp:
(WebCore::skipSpaces):
(WebCore::parseToken):
(WebCore::isNotQuoteOrBackslash):
(WebCore::collectHTTPQuotedString):
(WebCore::containsNonTokenCharacters):
(WebCore::parseQuotedString):
(WebCore::ParsedContentType::parseContentType):
(WebCore::ParsedContentType::create):
(WebCore::ParsedContentType::setContentType):
(WebCore::containsNonQuoteStringTokenCharacters):
(WebCore::ParsedContentType::setContentTypeParameter):
(WebCore::ParsedContentType::serialize const):
(WebCore::substringForRange): Deleted.
(WebCore::isNonTokenCharacter): Deleted.
(WebCore::isNonQuotedStringTokenCharacter): Deleted.
* platform/network/ParsedContentType.h:

Tools:

Add tests involving leading and trailing whitespace, non-token
characters and quoted strings.

* TestWebKitAPI/Tests/WebCore/ParsedContentType.cpp:
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241863 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[JSC] Remove WatchpointSet creation for SymbolTable entries if VM::canUseJIT() return...
ysuzuki@apple.com [Thu, 21 Feb 2019 05:59:53 +0000 (05:59 +0000)]
[JSC] Remove WatchpointSet creation for SymbolTable entries if VM::canUseJIT() returns false
https://bugs.webkit.org/show_bug.cgi?id=194891

Reviewed by Geoffrey Garen.

WatchpointSet in SymbolTable is used to fold the value into a constant in JIT tiers. And it is
not useful under the non-JIT mode. This patch avoids creation of WatchpointSet in SymbolTable
if VM::canUseJIT() returns false.

* llint/LowLevelInterpreter32_64.asm:
* llint/LowLevelInterpreter64.asm:
* runtime/SymbolTable.cpp:
(JSC::SymbolTableEntry::addWatchpoint): Deleted.
* runtime/SymbolTable.h:
(JSC::SymbolTableEntry::isWatchable const):
(JSC::SymbolTableEntry::watchpointSet):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241862 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoREGRESSION (240698): Fixed position banners flicker and move when scrolling on iOS
simon.fraser@apple.com [Thu, 21 Feb 2019 03:32:50 +0000 (03:32 +0000)]
REGRESSION (240698): Fixed position banners flicker and move when scrolling on iOS
https://bugs.webkit.org/show_bug.cgi?id=194889
rdar://problem/47755552

Reviewed by Tim Horton.

After r240698 we could commit scrolling changes for a fixed node where the "viewportRectAtLastLayout" and the layer
position were mismatched; this happened when AsyncScrollingCoordinator::reconcileScrollingState() came back from the UI process
with an unstable update and set a new layoutViewport, then some other layout triggered a compositing tree update. During the tree
update, we'd update the fixed scrolling node with the new viewport, and an old layer position.

Fix by ensuring that we only update the geometry info for a scrolling tree node when we update layer geometry for the corresponding
layer.

Not currently testable.

* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::updateBackingAndHierarchy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241860 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoMake programmatic frame scrolling work on iOS
antti@apple.com [Thu, 21 Feb 2019 03:28:11 +0000 (03:28 +0000)]
Make programmatic frame scrolling work on iOS
https://bugs.webkit.org/show_bug.cgi?id=194886

Reviewed by Simon Fraser.

Source/WebKit:

* UIProcess/RemoteLayerTree/ios/ScrollingTreeScrollingNodeDelegateIOS.mm:
(WebKit::ScrollingTreeScrollingNodeDelegateIOS::commitStateAfterChildren):

Don't move based on the layout scroll position. This just overrides the user scroll position.
Remove ScrolledContentsLayer checks, we only need to deal with the ScrollContainerLayer here.

(WebKit::ScrollingTreeScrollingNodeDelegateIOS::setScrollLayerPosition):

Scroll the UIScrollView correctly. This is called as a result of setting the RequestedScrollPosition property.
Remove scroll origin code, it doesn't look correct (and is untested).

LayoutTests:

Test by Frederic Wang.

* fast/scrolling/ios/programmatic-scroll-iframe-expected.html: Added.
* fast/scrolling/ios/programmatic-scroll-iframe.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241859 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[CMake][Win] Only build DumpRenderTree when WebKit Legacy is enabled
don.olmstead@sony.com [Thu, 21 Feb 2019 02:55:50 +0000 (02:55 +0000)]
[CMake][Win] Only build DumpRenderTree when WebKit Legacy is enabled
https://bugs.webkit.org/show_bug.cgi?id=194884

Reviewed by Michael Catanzaro.

* PlatformWin.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241858 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoWeb Inspector: RTL: Scripts Timeline loading indicator has incorrect margins
commit-queue@webkit.org [Thu, 21 Feb 2019 02:45:24 +0000 (02:45 +0000)]
Web Inspector: RTL: Scripts Timeline loading indicator has incorrect margins
https://bugs.webkit.org/show_bug.cgi?id=194882

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2019-02-20
Reviewed by Matt Baker.

* UserInterface/Views/TimelineRecordingContentView.css:
(.content-view.timeline-recording > .content-browser .recording-progress > .status > .indeterminate-progress-spinner):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241857 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoURL percent-encode operations should use checked arithmetic for buffer allocation...
commit-queue@webkit.org [Thu, 21 Feb 2019 01:36:04 +0000 (01:36 +0000)]
URL percent-encode operations should use checked arithmetic for buffer allocation length
https://bugs.webkit.org/show_bug.cgi?id=194877
<rdar://problem/48212062>

Patch by Alex Christensen <achristensen@webkit.org> on 2019-02-20
Reviewed by Tim Horton.

* wtf/URLHelpers.cpp:
(WTF::URLHelpers::userVisibleURL):
* wtf/cocoa/NSURLExtras.mm:
(WTF::dataWithUserTypedString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241856 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[PSON] Make sure hung processes are not kept alive by suspended pages or process...
cdumez@apple.com [Thu, 21 Feb 2019 01:25:46 +0000 (01:25 +0000)]
[PSON] Make sure hung processes are not kept alive by suspended pages or process caching
https://bugs.webkit.org/show_bug.cgi?id=194881
<rdar://problem/48249014>

Reviewed by Geoffrey Garen.

After we construct a SuspendedPageProxy and before we send the IPC to the WebProcess to
ask it to suspend, start a 10 seconds timer. If the process does not answer the request
to suspend before the timer fires, we destroy the SuspendedPageProxy so that we do not
keep a hung process around.

For the WebProcessCache, we now call WebProcessProxy::isResponsive() on the process
before adding it to the cache. Internally, this relies on an IPC handshake with the
WebProcess. If the process is not responsive, we do not add it to the cache and we
shut it down. If it is responsive then we proceed normally with adding it to the
cache.

* UIProcess/SuspendedPageProxy.cpp:
(WebKit::SuspendedPageProxy::SuspendedPageProxy):
(WebKit::SuspendedPageProxy::didProcessRequestToSuspend):
(WebKit::SuspendedPageProxy::suspensionTimedOut):
* UIProcess/SuspendedPageProxy.h:
* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::addProcessIfPossible):
(WebKit::WebProcessCache::addProcess):
* UIProcess/WebProcessCache.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):
(WebKit::WebProcessProxy::maybeShutDown):
(WebKit::WebProcessProxy::isResponsive):
* UIProcess/WebProcessProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241855 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAdd API test for <rdar://problem/47471222>
cdumez@apple.com [Thu, 21 Feb 2019 01:09:39 +0000 (01:09 +0000)]
Add API test for <rdar://problem/47471222>
https://bugs.webkit.org/show_bug.cgi?id=194847

Reviewed by Alex Christensen.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241854 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed build fix after r241823.
cdumez@apple.com [Thu, 21 Feb 2019 00:51:57 +0000 (00:51 +0000)]
Unreviewed build fix after r241823.

* UIProcess/SuspendedPageProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241853 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRotation animations sometimes use the wrong origin (affects apple.com)
dino@apple.com [Thu, 21 Feb 2019 00:48:37 +0000 (00:48 +0000)]
Rotation animations sometimes use the wrong origin (affects apple.com)
https://bugs.webkit.org/show_bug.cgi?id=194878
<rdar://problem/43908047>

Reviewed by Simon Fraser.

Source/WebCore:

Some versions of CoreAnimation apply additive animations in reverse
order. Detect this and reverse the list of animations we provide.

Update the existing animations/additive-transform-animations.html test to
be a ref-test that would identify this failure. Previously it relied on
a pixel test.

* platform/graphics/ca/GraphicsLayerCA.cpp: Use
    HAVE_CA_WHERE_ADDITIVE_TRANSFORMS_ARE_REVERSED to decide whether or
    not to flip the list of animations (and mark the correct ones as
    additive).
(WebCore::GraphicsLayerCA::appendToUncommittedAnimations):
(WebCore::GraphicsLayerCA::createTransformAnimationsFromKeyframes):

Source/WTF:

* wtf/Platform.h: Add HAVE(CA_WHERE_ADDITIVE_TRANSFORMS_ARE_REVERSED).

LayoutTests:

Update an old pixel test to be a new ref test, and remove all the old platform-specific
-expected versions.

* animations/additive-transform-animations-expected.png: Removed.
* animations/additive-transform-animations.html:
* legacy-animation-engine/animations/additive-transform-animations-expected.png: Removed.
* legacy-animation-engine/animations/additive-transform-animations.html:
* platform/gtk/animations/additive-transform-animations-expected.png: Removed.
* platform/gtk/animations/additive-transform-animations-expected.txt: Removed.
* platform/gtk/legacy-animation-engine/animations/additive-transform-animations-expected.png: Removed.
* platform/gtk/legacy-animation-engine/animations/additive-transform-animations-expected.txt: Removed.
* platform/ios/animations/additive-transform-animations-expected.txt: Removed.
* platform/ios/legacy-animation-engine/animations/additive-transform-animations-expected.txt: Removed.
* platform/mac/animations/additive-transform-animations-expected.txt: Removed.
* platform/mac/legacy-animation-engine/animations/additive-transform-animations-expected.txt: Removed.
* platform/win/animations/additive-transform-animations-expected.txt: Removed.
* platform/win/legacy-animation-engine/animations/additive-transform-animations-expected.txt: Removed.
* platform/wincairo/animations/additive-transform-animations-expected.txt: Removed.
* platform/wincairo/legacy-animation-engine/animations/additive-transform-animations-expected.txt: Removed.
* platform/wpe/animations/additive-transform-animations-expected.txt: Removed.
* platform/wpe/legacy-animation-engine/animations/additive-transform-animations-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241852 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[Win] Guard CF usage in RenderThemeWin
don.olmstead@sony.com [Thu, 21 Feb 2019 00:19:16 +0000 (00:19 +0000)]
[Win] Guard CF usage in RenderThemeWin
https://bugs.webkit.org/show_bug.cgi?id=194875

Reviewed by Alex Christensen.

No new tests. No change in behavior.

Add #if USE(CF) checks to RenderThemeWin so it can compile without CF support.

* rendering/RenderThemeWin.cpp:
(WebCore::RenderThemeWin::stringWithContentsOfFile):
(WebCore::RenderThemeWin::mediaControlsStyleSheet):
(WebCore::RenderThemeWin::mediaControlsScript):
* rendering/RenderThemeWin.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241851 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed, rolling out r241838.
achristensen@apple.com [Thu, 21 Feb 2019 00:12:44 +0000 (00:12 +0000)]
Unreviewed, rolling out r241838.

Broke API tests

Reverted changeset:

"Resolve WebsiteDataStore/HTTPCookieStore ref cycle"
https://bugs.webkit.org/show_bug.cgi?id=194868
https://trac.webkit.org/changeset/241838

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241850 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAdd code to validate expected GC activity modelled by doesGC() against what the runti...
mark.lam@apple.com [Thu, 21 Feb 2019 00:10:43 +0000 (00:10 +0000)]
Add code to validate expected GC activity modelled by doesGC() against what the runtime encounters.
https://bugs.webkit.org/show_bug.cgi?id=193938
<rdar://problem/47616277>

Reviewed by Michael Saboff, Saam Barati, and Robin Morisset.

In DFG::SpeculativeJIT::compile() and FTL::LowerDFGToB3::compileNode(), before
emitting code / B3IR for each DFG node, we emit a write to set Heap::m_expectDoesGC
to the value returned by doesGC() for that node.  In the runtime (i.e. in allocateCell()
and functions that can resolve a rope), we assert that Heap::m_expectDoesGC is
true.

This validation code is currently only enabled for debug builds.  It is disabled
for release builds by default, but it can easily be made to run on release builds
as well by forcing ENABLE_DFG_DOES_GC_VALIDATION to 1 in Heap.h.

To allow this validation code to run on release builds as well, the validation uses
RELEASE_ASSERT instead of ASSERT.

To ensure that Heap.h is #include'd for all files that needs to do this validation
(so that the validation code is accidentally disabled), we guard the validation
code with an if conditional on constexpr bool validateDFGDoesGC (instead of using
a #if ENABLE(DFG_DOES_GC_VALIDATION)).  This way, if Heap.h isn't #include'd, the
validation code will fail to build (no silent failures).

Currently, all JSC tests and Layout tests should pass with this validation enabled
in debug builds.  We'll only see new failures if there's a regression or if new
tests reveal a previously untested code path that has an undetected issue.

* dfg/DFGOSRExit.cpp:
(JSC::DFG::OSRExit::executeOSRExit):
(JSC::DFG::OSRExit::compileExit):
* dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileNode):
* ftl/FTLOSRExitCompiler.cpp:
(JSC::FTL::compileStub):
* heap/Heap.h:
(JSC::Heap::expectDoesGC const):
(JSC::Heap::setExpectDoesGC):
(JSC::Heap::addressOfExpectDoesGC):
* jit/JITArithmetic.cpp:
(JSC::JIT::emit_compareAndJump):
* runtime/JSCellInlines.h:
(JSC::tryAllocateCellHelper):
* runtime/JSString.h:
(JSC::jsSingleCharacterString):
(JSC::JSString::toAtomicString const):
(JSC::JSString::toExistingAtomicString const):
(JSC::JSString::value const):
(JSC::JSString::tryGetValue const):
(JSC::JSRopeString::unsafeView const):
(JSC::JSRopeString::viewWithUnderlyingString const):
(JSC::JSString::unsafeView const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241849 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoCrash in DOMWindowExtension::suspendForPageCache
rniwa@webkit.org [Thu, 21 Feb 2019 00:06:27 +0000 (00:06 +0000)]
Crash in DOMWindowExtension::suspendForPageCache
https://bugs.webkit.org/show_bug.cgi?id=194871

Reviewed by Chris Dumez.

This is a speculative fix for a crash in DOMWindowExtension::suspendForPageCache.

We think it's possible for DOMWindowExtension::suspendForPageCache notifying the clients via
dispatchWillDisconnectDOMWindowExtensionFromGlobalObject to remove other DOMWindowExtension's.
Check that each DOMWindowProperty is still in m_properties before invoking suspendForPageCache
to avoid the crash.

* page/DOMWindow.cpp:
(WebCore::DOMWindow::willDestroyCachedFrame):
(WebCore::DOMWindow::willDestroyDocumentInFrame):
(WebCore::DOMWindow::willDetachDocumentFromFrame):
(WebCore::DOMWindow::suspendForPageCache):
(WebCore::DOMWindow::resumeFromPageCache):
* page/DOMWindowExtension.cpp:
(WebCore::DOMWindowExtension::suspendForPageCache):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241848 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[bmalloc] bmalloc::Heap is allocated even though we use system malloc mode
ysuzuki@apple.com [Thu, 21 Feb 2019 00:03:17 +0000 (00:03 +0000)]
[bmalloc] bmalloc::Heap is allocated even though we use system malloc mode
https://bugs.webkit.org/show_bug.cgi?id=194836

Reviewed by Mark Lam.

Previously, bmalloc::Heap holds DebugHeap, and delegates allocation and deallocation to debug heap.
However, bmalloc::Heap is large. We would like to avoid initialization of bmalloc::Heap under the
system malloc mode.

This patch extracts out DebugHeap from bmalloc::Heap, and logically puts this in a boundary of
bmalloc::api. bmalloc::api delegates allocation and deallocation to DebugHeap if DebugHeap is enabled.
Otherwise, using bmalloc's usual mechanism. The challenge is that we would like to keep bmalloc fast
path fast.

1. For IsoHeaps, we use the similar techniques done in Cache. If the debug mode is enabled, we always go
   to the slow path of the IsoHeap allocation, and keep IsoTLS::get() returning nullptr. In the slow path,
   we just fallback to the usual bmalloc::api::tryMalloc implementation. This is efficient because bmalloc
   continues using the fast path.

2. For the other APIs, like freeLargeVirtual, we just put DebugHeap check because this API itself takes fair
   amount of time. Then debug heap check does not matter.

* bmalloc/Allocator.cpp:
(bmalloc::Allocator::reallocateImpl):
* bmalloc/Cache.cpp:
(bmalloc::Cache::tryAllocateSlowCaseNullCache):
(bmalloc::Cache::allocateSlowCaseNullCache):
(bmalloc::Cache::deallocateSlowCaseNullCache):
(bmalloc::Cache::tryReallocateSlowCaseNullCache):
(bmalloc::Cache::reallocateSlowCaseNullCache):
(): Deleted.
(bmalloc::debugHeap): Deleted.
* bmalloc/DebugHeap.cpp:
* bmalloc/DebugHeap.h:
(bmalloc::DebugHeap::tryGet):
* bmalloc/Heap.cpp:
(bmalloc::Heap::Heap):
(bmalloc::Heap::footprint):
(bmalloc::Heap::tryAllocateLarge):
(bmalloc::Heap::deallocateLarge):
* bmalloc/Heap.h:
(bmalloc::Heap::debugHeap): Deleted.
* bmalloc/IsoTLS.cpp:
(bmalloc::IsoTLS::IsoTLS):
(bmalloc::IsoTLS::isUsingDebugHeap): Deleted.
(bmalloc::IsoTLS::debugMalloc): Deleted.
(bmalloc::IsoTLS::debugFree): Deleted.
* bmalloc/IsoTLS.h:
* bmalloc/IsoTLSInlines.h:
(bmalloc::IsoTLS::allocateSlow):
(bmalloc::IsoTLS::deallocateSlow):
* bmalloc/ObjectType.cpp:
(bmalloc::objectType):
* bmalloc/ObjectType.h:
* bmalloc/Scavenger.cpp:
(bmalloc::Scavenger::Scavenger):
* bmalloc/bmalloc.cpp:
(bmalloc::api::tryLargeZeroedMemalignVirtual):
(bmalloc::api::freeLargeVirtual):
(bmalloc::api::scavenge):
(bmalloc::api::isEnabled):
(bmalloc::api::setScavengerThreadQOSClass):
(bmalloc::api::commitAlignedPhysical):
(bmalloc::api::decommitAlignedPhysical):
(bmalloc::api::enableMiniMode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241847 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoREGRESSION: [ iOS ] Layout Test editing/input/ios/rtl-keyboard-input-on-focus.html...
wenson_hsieh@apple.com [Wed, 20 Feb 2019 23:34:50 +0000 (23:34 +0000)]
REGRESSION: [ iOS ] Layout Test editing/input/ios/rtl-keyboard-input-on-focus.html is a Timeout
https://bugs.webkit.org/show_bug.cgi?id=194601
<rdar://problem/48080316>

Reviewed by Tim Horton.

Following r241311, if a web view becomes first responder and is then moved offscreen (or obscured, hidden, or in
the case of WebKitTestRunner, its UIWindow loses its status as keyWindow), we end up holding on to the input
view update deferral token indefinitely, waiting for the current focused element to be blurred or refocused.

This also manifests other user-facing bugs, the most common of which is the keyboard occasionally remaining
onscreen after typing a URL in the unified field in MobileSafari and hitting Return, in the case where there is
no autofocused element on the page.

To fix this, when becoming the first responder, additionally install a callback to detect when the page is
finished handling the activity state change, and invalidate the input deferral token then. This retains the
behavior where calling -becomeFirstResponder on the web view while a different view is focused will keep the
keyboard stable, since the focused element message from the web process should be dispatched when handling the
activity state change within the web process.

Of course, the web process may not be responsive at all while the web view is still in the view hierarchy, in
which case we may still end up deferring input view updates indefinitely. In this case, we maintain a separate
watchdog timer with a short delay, after which we unconditionally invalidate the token.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::installActivityStateChangeCompletionHandler):

Move the implementation of installActivityStateChangeCompletionHandler into cross-platform code.

* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentView.mm:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanupInteraction]):
(-[WKContentView _cancelPreviousResetInputViewDeferralRequest]):
(-[WKContentView _scheduleResetInputViewDeferralAfterBecomingFirstResponder]):
(-[WKContentView _resetInputViewDeferral]):
(-[WKContentView becomeFirstResponderForWebView]):
(-[WKContentView resignFirstResponderForWebView]):
(-[WKContentView _commitPotentialTapFailed]):
(-[WKContentView _didNotHandleTapAsClick:]):
(-[WKContentView _didCompleteSyntheticClick]):

Funnel all existing calls that reset _inputViewDeferralToken to nullptr, such that they go through a helper
method instead that also cancels any scheduled requests to clear the token.

* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::activityStateDidChange):

Respond to all pending callbacks after handling the activity state change.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241846 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoRegression(PSON) "Reload without content extensions" does not work when the main...
cdumez@apple.com [Wed, 20 Feb 2019 23:30:03 +0000 (23:30 +0000)]
Regression(PSON) "Reload without content extensions" does not work when the main resource is blocked
https://bugs.webkit.org/show_bug.cgi?id=194872
<rdar://problem/47924500>

Reviewed by Alex Christensen.

Source/WebKit:

[WKWebView _reloadWithoutContentBlockers] relies on a ReloadOption flag that is passed to WebCore
instead of using WebsitePolicies.contentBlockersEnabled flag. If the reload causes a process swap
due to PSON, then the new process does not know about this ReloadOption and fails to honor it.

Since the modern way to do this is WebsitePolicies, and since WebsitePolicies are properly
propagated cross-process in case of process swap, this patch updates _reloadWithoutContentBlockers
to set a flag on the Navigation which we use to later set the WebsitePolicies.contentBlockersEnabled
flag in WebPageProxy::receivedNavigationPolicyDecision().

* UIProcess/API/APINavigation.h:
(API::Navigation::setUserContentExtensionsEnabled):
(API::Navigation::userContentExtensionsEnabled const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::reload):
(WebKit::WebPageProxy::receivedNavigationPolicyDecision):

Tools:

Add API test coverage.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241845 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoUnreviewed, rolling out r241817.
tsavell@apple.com [Wed, 20 Feb 2019 23:26:54 +0000 (23:26 +0000)]
Unreviewed, rolling out r241817.

Caused 4 API failures

Reverted changeset:

"[GTK] Epiphany searching for plugins even if plugins are
disabled"
https://bugs.webkit.org/show_bug.cgi?id=194352
https://trac.webkit.org/changeset/241817

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241844 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAdd a git utility method that allows the caller to determine if a specific commitish...
jer.noble@apple.com [Wed, 20 Feb 2019 23:25:23 +0000 (23:25 +0000)]
Add a git utility method that allows the caller to determine if a specific commitish is within a specified range of commits.
https://bugs.webkit.org/show_bug.cgi?id=191332

Reviewed by Dean Jackson.

* Scripts/git-is-in-range: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241843 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAlways call CompletionHandlers after r240909
commit-queue@webkit.org [Wed, 20 Feb 2019 22:52:18 +0000 (22:52 +0000)]
Always call CompletionHandlers after r240909
https://bugs.webkit.org/show_bug.cgi?id=194823

Patch by Alex Christensen <achristensen@webkit.org> on 2019-02-20
Reviewed by Ryosuke Niwa.

* loader/PolicyChecker.cpp:
(WebCore::PolicyChecker::checkNavigationPolicy):
(WebCore::PolicyChecker::checkNewWindowPolicy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241842 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[Xcode] Add SDKVariant.xcconfig to various Xcode projects
aestes@apple.com [Wed, 20 Feb 2019 22:37:28 +0000 (22:37 +0000)]
[Xcode] Add SDKVariant.xcconfig to various Xcode projects
https://bugs.webkit.org/show_bug.cgi?id=194869

Rubber-stamped by Jer Noble.

Source/bmalloc:

* bmalloc.xcodeproj/project.pbxproj:

Source/JavaScriptCore:

* JavaScriptCore.xcodeproj/project.pbxproj:

Source/ThirdParty/ANGLE:

* ANGLE.xcodeproj/project.pbxproj:

Source/ThirdParty/libwebrtc:

* libwebrtc.xcodeproj/project.pbxproj:

Source/WebCore:

* WebCore.xcodeproj/project.pbxproj:

Source/WebInspectorUI:

* WebInspectorUI.xcodeproj/project.pbxproj:

Source/WebKit:

* WebKit.xcodeproj/project.pbxproj:

Source/WebKitLegacy:

* WebKitLegacy.xcodeproj/project.pbxproj:

Source/WTF:

* WTF.xcodeproj/project.pbxproj:

Tools:

* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* ImageDiff/ImageDiff.xcodeproj/project.pbxproj:
* MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
* MobileMiniBrowser/MobileMiniBrowser.xcodeproj/project.pbxproj:
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241841 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agodrawImage() clears the canvas if it's the source of the image and globalCompositeOper...
commit-queue@webkit.org [Wed, 20 Feb 2019 22:28:13 +0000 (22:28 +0000)]
drawImage() clears the canvas if it's the source of the image and globalCompositeOperation is "copy"
https://bugs.webkit.org/show_bug.cgi?id=194746

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2019-02-20
Reviewed by Dean Jackson.

Source/WebCore:

Test: fast/canvas/canvas-drawImage-composite-copy.html

If the source canvas of drawImage() is the same as the destination and
globalCompositeOperation is set to "copy", copy the srcRect from the
canvas to a temporary buffer before calling clearCanvas() then drawImage
from this temporary buffer.

* html/canvas/CanvasRenderingContext2DBase.cpp:
(WebCore::CanvasRenderingContext2DBase::drawImage):
* platform/graphics/ImageBuffer.cpp:
(WebCore::ImageBuffer::copyRectToBuffer):
* platform/graphics/ImageBuffer.h:

LayoutTests:

* fast/canvas/canvas-drawImage-composite-copy-expected.html: Added.
* fast/canvas/canvas-drawImage-composite-copy.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241840 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAdded macOS Mojave builds to the WebKit Build Archives.
jond@apple.com [Wed, 20 Feb 2019 22:25:23 +0000 (22:25 +0000)]
Added macOS Mojave builds to the WebKit Build Archives.
https://bugs.webkit.org/show_bug.cgi?id=190908

Reviewed by Lucas Forschler.

* wp-content/themes/webkit/build-archives.php:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241839 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoResolve WebsiteDataStore/HTTPCookieStore ref cycle
commit-queue@webkit.org [Wed, 20 Feb 2019 22:22:17 +0000 (22:22 +0000)]
Resolve WebsiteDataStore/HTTPCookieStore ref cycle
https://bugs.webkit.org/show_bug.cgi?id=194868

Patch by Alex Christensen <achristensen@webkit.org> on 2019-02-20
Reviewed by Geoffrey Garen.

* UIProcess/API/APIHTTPCookieStore.cpp:
(API::HTTPCookieStore::HTTPCookieStore):
(API::HTTPCookieStore::cookies):
(API::HTTPCookieStore::setCookie):
(API::HTTPCookieStore::deleteCookie):
(API::HTTPCookieStore::registerObserver):
(API::HTTPCookieStore::unregisterObserver):
(API::HTTPCookieStore::cookieManagerDestroyed):
(API::HTTPCookieStore::registerForNewProcessPoolNotifications):
(API::HTTPCookieStore::ref const):
(API::HTTPCookieStore::deref const):
* UIProcess/API/APIHTTPCookieStore.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241838 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[bmalloc] DebugHeap::malloc does not have "try" version.
ysuzuki@apple.com [Wed, 20 Feb 2019 22:22:07 +0000 (22:22 +0000)]
[bmalloc] DebugHeap::malloc does not have "try" version.
https://bugs.webkit.org/show_bug.cgi?id=194837

Reviewed by Mark Lam.

Since DebugHeap::malloc does not have "try" version, our tryAllocate implementation does not work well with DebugHeap.
This patch adds crashOnFailure flag to DebugHeap::malloc.

* bmalloc/Cache.cpp:
(bmalloc::Cache::tryAllocateSlowCaseNullCache):
(bmalloc::Cache::allocateSlowCaseNullCache):
* bmalloc/DebugHeap.cpp:
(bmalloc::DebugHeap::malloc):
* bmalloc/DebugHeap.h:
* bmalloc/IsoTLS.cpp:
(bmalloc::IsoTLS::debugMalloc):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241837 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[ Mac WK2 ] REGRESSION (r231450) Layout Test http/tests/security/contentSecurityPolic...
tsavell@apple.com [Wed, 20 Feb 2019 22:22:07 +0000 (22:22 +0000)]
[ Mac WK2 ] REGRESSION (r231450) Layout Test http/tests/security/contentSecurityPolicy/block-all-mixed-content/insecure-image-in-iframe-with-enforced-and-report-policies.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=194350

Unreviewed test gardening.

Patch by Shawn Roberts <sroberts@apple.com> on 2019-02-20

* platform/mac-wk2/TestExpectations: Marking as flaky until a fix lands

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241836 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[ Mac WK1 ] REGRESSION (r230006) Layout Test media/modern-media-controls/compact...
tsavell@apple.com [Wed, 20 Feb 2019 22:04:04 +0000 (22:04 +0000)]
[ Mac WK1 ] REGRESSION (r230006) Layout Test media/modern-media-controls/compact-media-controls/compact-media-controls-layout.html is a flaky text diff failure
https://bugs.webkit.org/show_bug.cgi?id=194309

Unreviewed test gardening.

Patch by Shawn Roberts <sroberts@apple.com> on 2019-02-20

* platform/mac-wk1/TestExpectations: Marking as flaky until a fix lands

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241835 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months ago[ Mac WK2 iOS Debug ] REGRESSION (r240713) Layout Test scrollingcoordinator/scrolling...
tsavell@apple.com [Wed, 20 Feb 2019 21:49:53 +0000 (21:49 +0000)]
[ Mac WK2 iOS Debug ] REGRESSION (r240713) Layout Test scrollingcoordinator/scrolling-tree/fixed-inside-frame.html is flaky failure
https://bugs.webkit.org/show_bug.cgi?id=194253

Unreviewed test gardening.

Patch by Shawn Roberts <sroberts@apple.com> on 2019-02-20

* platform/ios-wk2/TestExpectations: Marking as flaky until a fix lands
* platform/mac-wk2/TestExpectations: ditto

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241834 268f45cc-cd09-0410-ab3c-d52691b4dbfc

21 months agoAttempt to fix the Apple Internal build following r241828
dbates@webkit.org [Wed, 20 Feb 2019 21:38:40 +0000 (21:38 +0000)]
Attempt to fix the Apple Internal build following r241828
(https://trac.webkit.org/changeset/241828/webkit)

Conditionally include a private header.

* pal/spi/ios/UIKitSPI.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@241833 268f45cc-cd09-0410-ab3c-d52691b4dbfc