Context Navigation

← Previous Period
Next Period →

Timeline

Apr 12, 2020:

11:39 PM Changeset in webkit [259995] by Antti Koivisto

8 edits in trunk/LayoutTests/imported/w3c

Update imported/w3c/web-platform-tests/css/selectors/invalidation/is.html
https://bugs.webkit.org/show_bug.cgi?id=210418

Unreviewed.

Also some other minor updates from syncing css/selectors from WPT repo.

web-platform-tests/css/selectors/invalidation/defined.html:
web-platform-tests/css/selectors/invalidation/is-expected.txt:

Specificify subtest is correctly FAIL.

web-platform-tests/css/selectors/invalidation/is.html:
web-platform-tests/css/selectors/invalidation/selectorText-dynamic-001.html:
web-platform-tests/css/selectors/invalidation/sheet-going-away-001.html:
web-platform-tests/css/selectors/invalidation/sheet-going-away-002.html:
web-platform-tests/css/selectors/x-pseudo-element.html:

8:48 PM Changeset in webkit [259994] by Alan Coon

11 edits in branches/safari-609-branch/Source

Cherry-pick r259316. rdar://problem/61269751

IndexedDB: destroy WebIDBServer when session is removed in network process
https://bugs.webkit.org/show_bug.cgi?id=209606
<rdar://problem/59310081>

Reviewed by Geoffrey Garen.

Source/WebCore:

Rename immediateCloseForUserDelete to immediateClose as we now use it in destructor of IDBServer to make sure
everything in database finishes correctly.

Modules/indexeddb/server/IDBServer.cpp: (WebCore::IDBServer::IDBServer::~IDBServer): (WebCore::IDBServer::IDBServer::closeAndDeleteDatabasesModifiedSince): (WebCore::IDBServer::IDBServer::closeAndDeleteDatabasesForOrigins):
Modules/indexeddb/server/UniqueIDBDatabase.cpp: (WebCore::IDBServer::UniqueIDBDatabase::immediateClose): (WebCore::IDBServer::UniqueIDBDatabase::immediateCloseForUserDelete): Deleted.
Modules/indexeddb/server/UniqueIDBDatabase.h:

Source/WebKit:

Tested manually to verify WebIDBServer is removed and its thread ends when session is removed.

NetworkProcess/IndexedDB/WebIDBServer.cpp: (WebKit::WebIDBServer::~WebIDBServer): (WebKit::WebIDBServer::addConnection): (WebKit::WebIDBServer::removeConnection): (WebKit::WebIDBServer::close):
NetworkProcess/IndexedDB/WebIDBServer.h:
NetworkProcess/NetworkProcess.cpp: (WebKit::NetworkProcess::destroySession): (WebKit::NetworkProcess::connectionToWebProcessClosed):

Source/WTF:

Add function to kill CrossThreadTaskHandler and make thread finish. Also add a callback to be called before
thread finishes.

wtf/CrossThreadTaskHandler.cpp: (WTF::CrossThreadTaskHandler::CrossThreadTaskHandler): (WTF::CrossThreadTaskHandler::setCompletionCallback): (WTF::CrossThreadTaskHandler::kill):
wtf/CrossThreadTaskHandler.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@259316 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8:48 PM Changeset in webkit [259993] by Alan Coon

3 edits in branches/safari-609-branch/Source/WebCore

Cherry-pick r257901. rdar://problem/61269751

Remove unused variable m_allUniqueIDBDatabases in IDBServer
https://bugs.webkit.org/show_bug.cgi?id=208613

Reviewed by Chris Dumez.

Modules/indexeddb/server/IDBServer.h: (WebCore::IDBServer::IDBServer::addDatabase): Deleted. (WebCore::IDBServer::IDBServer::removeDatabase): Deleted.
Modules/indexeddb/server/UniqueIDBDatabase.cpp: (WebCore::IDBServer::UniqueIDBDatabase::UniqueIDBDatabase): (WebCore::IDBServer::UniqueIDBDatabase::~UniqueIDBDatabase):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@257901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8:37 PM Changeset in webkit [259992] by Lauro Moura

2 edits in trunk/LayoutTests

[WPE] Layout test gardening.

Some EME previous crashes are now failing.

Unreviewed test gardening.

platform/wpe/TestExpectations:

4:09 PM Changeset in webkit [259991] by Alan Bujtas

1 edit
2 adds in trunk/LayoutTests

[LFC][TFC] Add columnSpan test case (fixed width)
https://bugs.webkit.org/show_bug.cgi?id=210411

Unreviewed.

fast/layoutformattingcontext/table-fixed-width-colspans.html: Added.

12:51 PM Changeset in webkit [259990] by Darin Adler

10 edits in trunk/Source

Fix a few mispellings of descendant and propagation
https://bugs.webkit.org/show_bug.cgi?id=210409

Reviewed by Mark Lam.

Source/JavaScriptCore:

ftl/FTLAbstractHeap.h: "descendants"
offlineasm/ast.rb: "descendants"

Source/WebCore:

dom/Element.cpp:

(WebCore::Element::dispatchWheelEvent): "propagation"

dom/TreeScopeOrderedMap.cpp:

(WebCore::TreeScopeOrderedMap::getAllElementsById const):
"descendants". Also refactored this function a bit.

html/MediaElementSession.cpp:

(WebCore::MediaElementSession::canShowControlsManager const):
"descendants"

rendering/RenderFrameSet.cpp:

(WebCore::resetFrameRendererAndDescendants): "descendants"
(WebCore::RenderFrameSet::positionFrames): "descendants"
(WebCore::RenderFrameSet::positionFramesWithFlattening): "descendants"

Source/WebInspectorUI:

UserInterface/Views/TreeOutline.js:

(WI.TreeOutline.prototype._rememberTreeElement):
(WI.TreeOutline.prototype._forgetTreeElement):
(WI.TreeOutline.prototype.selectionControllerNumberOfItems):

12:34 PM Changeset in webkit [259989] by Lauro Moura

4 edits in trunk/LayoutTests

[GTK][WPE] Gardening some crashes and flakiness

Unreviewed gardening.

platform/gtk-wayland/TestExpectations:
platform/gtk/TestExpectations:
platform/wpe/TestExpectations:

12:33 PM Changeset in webkit [259988] by Darin Adler

5 edits in trunk/Source/WebCore

Refactor and tighten up the CSSVariableReferenceValue class
https://bugs.webkit.org/show_bug.cgi?id=210406

Reviewed by Anders Carlsson.

css/CSSCustomPropertyValue.h: Remove uneeded forward declaration of

CSSVariableReferenceValue, since it's not used here. Added inclde of
CSSVariableData.h since the use of Variant in this class does require
that header, which we were getting indirectly before from
CSSVariableReferenceValue.h in some translation units.

css/CSSVariableReferenceValue.cpp:

(WebCore::CSSVariableReferenceValue::CSSVariableReferenceValue): Moved here
from the header.
(WebCore::CSSVariableReferenceValue::create): Ditto.
(WebCore::CSSVariableReferenceValue::equals const): Ditto.
(WebCore::CSSVariableReferenceValue::customCSSText const): Use non-null to
indicate this is not serialized.

css/CSSVariableReferenceValue.h: Reduced includes, inlining, marked

constructor explicit, removed unneeded m_serialized boolean.

rendering/style/StyleCustomPropertyData.h: Remove unneeded include

of CSSVariableReferenceValue.h, not used here.

12:31 PM Changeset in webkit [259987] by Darin Adler

6 edits in trunk/Source

Fix some strange uses of start/endOfDocument
https://bugs.webkit.org/show_bug.cgi?id=210408

Reviewed by Wenson Hsieh.

Source/WebCore:

editing/VisibleSelection.cpp:

(WebCore::VisibleSelection::setStartAndEndFromBaseAndExtentRespectingGranularity):
Call startOfDocument and endOfDocument without unnecessarily turning a Position
into a VisiblePostion, since those functions just require any node from the document.

Source/WebKit:

WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::replaceDictatedText): Pass the document to startOfDocument,
rather than getting the documentelement, because the function works on any
node, so there's no reason to write code to get a particular one.
(WebKit::WebPage::applyAutocorrectionInternal): Ditto.

Source/WebKitLegacy/ios:

WebCoreSupport/WebFrameIOS.mm:

(-[WebFrame moveSelectionToStart]): Call startOfDocument on the document;
there is no need to pass a particular VisiblePosition, since all it gets
from the VisiblePosition is the document.
(-[WebFrame moveSelectionToEnd]): Ditto, for endOfDocument.
(-[WebFrame startPosition]): Call startOfDocument on the document rather
than using code to find the document element, since the function works
on any node.
(-[WebFrame endPosition]): Ditto.

12:28 PM Changeset in webkit [259986] by Darin Adler

2 edits in trunk/Source/WTF

[Cocoa] Minor tweaks to code to get locale strings to remove one-time-initialization booleans
https://bugs.webkit.org/show_bug.cgi?id=210410

Reviewed by Sam Weinig.

wtf/text/cocoa/TextBreakIteratorInternalICUCocoa.cpp:

(WTF::textBreakLocalePreference): Use auto, and no need to explicitly mark this inline.
(WTF::topLanguagePreference): Use auto.
(WTF::localeIDInBuffer): Renamed from getLocale. Changed to return a std::array and to
handle failure cases afterward rather than by initializing beforehand.
(WTF::getSearchLocale): Deleted.
(WTF::currentSearchLocaleID): Using "static const auto", wrote this in a straightforward
and compact way, using localeIDInBuffer and topLanguagePreference.
(WTF::textBreakLocale): Renamed from getTextBreakLocale and changed to return a
RetainPtr<CFStringRef>.
(WTF::currentTextBreakLocaleID): Using "static const auto", wrote this in a straightforward
and compact way, using localeIDInBuffer and textBreakLocale.

12:04 PM Changeset in webkit [259985] by Chris Dumez

2 edits in trunk/LayoutTests

fast/dom/timer-throttling-hidden-page.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=210355
<rdar://problem/61462972>

Reviewed by Sam Weinig.

The test was changing the page's visiblity to 'hidden' using testRunner.setPageVisibility("hidden")
in the first timer call and then expecting that by the 5th timer iteration, the page would actually
be hidden for timer throtting to kick in. While this happen to work most of the time, this would
cause flakiness because testRunner.setPageVisibility("hidden") involves an async IPC to the
UIProcess and then an IPC back to the WebProcess to update the page's activity state.

To address the issue, we now listen for the 'visibilitychange' event and only start the test / timer
once document.hidden actually returns true.

fast/dom/timer-throttling-hidden-page.html:

11:32 AM Changeset in webkit [259984] by Alan Bujtas

5 edits in trunk/Source/WebCore

[LFC][TFC] Add support for column spanners
https://bugs.webkit.org/show_bug.cgi?id=210403

Reviewed by Antti Koivisto.

Table width constraint computation with spanner support is as follows:

Collect each cells' width constraints.
Collect fixed column widths set by <colgroup>'s and <col>s.
Find the min/max width for each columns using the cell constraints and the <col> fixed widths but ignore column spans.
Distribute column spanning cells min/max widths.
Add them all up and return the computed min/max widths.

layout/FormattingContext.h:

(WebCore::Layout::FormattingContext::IntrinsicWidthConstraints::operator-=):

layout/tableformatting/TableFormattingContext.cpp:

(WebCore::Layout::TableFormattingContext::computedPreferredWidthForColumns):
(WebCore::Layout::TableFormattingContext::computeAndDistributeExtraHorizontalSpace):

layout/tableformatting/TableFormattingContextGeometry.cpp:

(WebCore::Layout::TableFormattingContext::Geometry::intrinsicWidthConstraintsForCell):

layout/tableformatting/TableGrid.cpp:

(WebCore::Layout::TableGrid::Columns::hasFixedColumnsOnly const):

layout/tableformatting/TableGrid.h:

10:53 AM Changeset in webkit [259983] by Alan Bujtas

4 edits in trunk/Source/WebCore

[LFC][TFC] Column, Row and Cell boxes are always ContainerBoxes
https://bugs.webkit.org/show_bug.cgi?id=210402

Reviewed by Antti Koivisto.

These boxes are always ContainerBox types.

layout/tableformatting/TableFormattingContext.cpp:

(WebCore::Layout::TableFormattingContext::layoutCell):
(WebCore::Layout::TableFormattingContext::ensureTableGrid):
(WebCore::Layout::TableFormattingContext::computedPreferredWidthForColumns):

layout/tableformatting/TableGrid.cpp:

(WebCore::Layout::TableGrid::Column::Column):
(WebCore::Layout::TableGrid::Columns::addColumn):
(WebCore::Layout::TableGrid::Rows::addRow):
(WebCore::Layout::TableGrid::Row::Row):
(WebCore::Layout::TableGrid::Cell::Cell):
(WebCore::Layout::TableGrid::appendCell):
(WebCore::Layout::TableGrid::insertCell):
(WebCore::Layout::TableGrid::removeCell):

layout/tableformatting/TableGrid.h:

(WebCore::Layout::TableGrid::Column::box const):
(WebCore::Layout::TableGrid::Row::box const):
(WebCore::Layout::TableGrid::Cell::box const):

10:21 AM Changeset in webkit [259982] by Alan Bujtas

7 edits in trunk/Source/WebCore

[LFC][TFC] Add support for fixed width columns
https://bugs.webkit.org/show_bug.cgi?id=210401

Reviewed by Antti Koivisto.

This is in preparation for adding support for spanner cells.
Fixed width columns (<col> and <td>) don't participate in the spanner width distribution.

layout/FormattingContext.h:

(WebCore::Layout::FormattingContext::IntrinsicWidthConstraints::operator-=):

layout/Verification.cpp:

(WebCore::Layout::areEssentiallyEqual):
(WebCore::Layout::LayoutContext::verifyAndOutputMismatchingLayoutTree):

layout/tableformatting/TableFormattingContext.cpp:

(WebCore::Layout::TableFormattingContext::computedIntrinsicWidthConstraints):
(WebCore::Layout::TableFormattingContext::computedPreferredWidthForColumns):
(WebCore::Layout::TableFormattingContext::computeAndDistributeExtraHorizontalSpace):
(WebCore::Layout::TableFormattingContext::computePreferredWidthForColumns): Deleted.
(WebCore::Layout::TableFormattingContext::useAsContentLogicalWidth): Deleted.

layout/tableformatting/TableFormattingContext.h:
layout/tableformatting/TableFormattingContextGeometry.cpp:

(WebCore::Layout::TableFormattingContext::Geometry::intrinsicWidthConstraintsForCell):

layout/tableformatting/TableGrid.cpp:

(WebCore::Layout::TableGrid::Column::isFixedWidth const):
(WebCore::Layout::TableGrid::Cell::isFixedWidth const):
(WebCore::Layout::TableGrid::Slot::Slot):
(WebCore::Layout::TableGrid::appendCell):
(WebCore::Layout::TableGrid::Column::setWidthConstraints): Deleted.
(WebCore::Layout::TableGrid::Column::widthConstraints const): Deleted.
(WebCore::Layout::TableGrid::Column::hasFixedWidth const): Deleted.
(WebCore::Layout::TableGrid::widthConstraints): Deleted.

layout/tableformatting/TableGrid.h:

(WebCore::Layout::TableGrid::setWidthConstraints):
(WebCore::Layout::TableGrid::widthConstraints):
(WebCore::Layout::TableGrid::Column::setHasFixedWidthCell):
(WebCore::Layout::TableGrid::Column::hasFixedWidthCell const):
(WebCore::Layout::TableGrid::Slot::cell const):
(WebCore::Layout::TableGrid::Slot::cell):
(WebCore::Layout::TableGrid::Slot::widthConstraints const):
(WebCore::Layout::TableGrid::Slot::setWidthConstraints):
(WebCore::Layout::TableGrid::Slot::hasColumnSpan const):
(WebCore::Layout::TableGrid::Slot::hasRowSpan const):
(WebCore::Layout::TableGrid::Slot::isColumnSpanned const):
(WebCore::Layout::TableGrid::Slot::isRowSpanned const):
(WebCore::Layout::TableGrid::hasComputedWidthConstraints const): Deleted.

9:34 AM Changeset in webkit [259981] by Alan Bujtas

7 edits in trunk/Source/WebCore

[LFC][TFC] Introduce dedicated SlotPosition/CellSpan structs
https://bugs.webkit.org/show_bug.cgi?id=210399

Reviewed by Antti Koivisto.

SlotPosition.column/row and CellSpan.column/row read better.

layout/LayoutUnits.h:

(WebCore::Layout::SlotPosition::SlotPosition):
(WebCore::Layout::operator==):
(WTF::SlotPositionHash::hash):
(WTF::SlotPositionHash::equal):
(WTF::HashTraits<WebCore::Layout::SlotPosition>::emptyValue):
(WTF::HashTraits<WebCore::Layout::SlotPosition>::constructDeletedValue):
(WTF::HashTraits<WebCore::Layout::SlotPosition>::isDeletedValue):

layout/layouttree/LayoutBox.cpp:

(WebCore::Layout::Box::setRowSpan):
(WebCore::Layout::Box::setColumnSpan):
(WebCore::Layout::Box::rowSpan const):
(WebCore::Layout::Box::columnSpan const):

layout/layouttree/LayoutBox.h:
layout/tableformatting/TableFormattingContext.cpp:

(WebCore::Layout::TableFormattingContext::computePreferredWidthForColumns):

layout/tableformatting/TableGrid.cpp:

(WebCore::Layout::TableGrid::Cell::Cell):
(WebCore::Layout::TableGrid::appendCell):

layout/tableformatting/TableGrid.h:

(WebCore::Layout::TableGrid::Cell::startColumn const):
(WebCore::Layout::TableGrid::Cell::endColumn const):
(WebCore::Layout::TableGrid::Cell::startRow const):
(WebCore::Layout::TableGrid::Cell::endRow const):
(WebCore::Layout::TableGrid::Cell::columnSpan const):
(WebCore::Layout::TableGrid::Cell::rowSpan const):
(WebCore::Layout::TableGrid::Cell::span const):
(WebCore::Layout::TableGrid::Cell::size const): Deleted.

9:05 AM Changeset in webkit [259980] by ddkilzer@apple.com

2 edits in trunk/Source/WTF

Follow-up: WTF::Persistence::Coder and WTF::Persistence::Decoder should use WARN_UNUSED_RETURN
<https://webkit.org/b/210238>
<rdar://problem/61491575>

Changes based on feedback from Alex Christensen and Darin Adler:

Remove WARN_UNUSED_RETURN from methods returning Optional<>.
Place WARN_UNUSED_RETURN consistently before the return type.

wtf/persistence/PersistentDecoder.h:

9:04 AM Changeset in webkit [259979] by Alan Bujtas

2 edits in trunk/Source/WebCore

[LFC][TFC] Add table support to BlockFormattingContext::Geometry::inFlowWidthAndMargin
https://bugs.webkit.org/show_bug.cgi?id=210400

Reviewed by Antti Koivisto.

Use a slightly modified shrink-to-fit logic to compute the table width.

layout/blockformatting/BlockFormattingContextGeometry.cpp:

(WebCore::Layout::BlockFormattingContext::Geometry::inFlowWidthAndMargin):

8:46 AM Changeset in webkit [259978] by Alan Bujtas

5 edits in trunk/Source/WebCore

[LFC][TFC] Cleanup class/struct/variable names in TableGrid/TableFormattingContext
https://bugs.webkit.org/show_bug.cgi?id=210397

Reviewed by Antti Koivisto.

This is in preparation for the column spanner work.

layout/tableformatting/TableFormattingContext.cpp:

(WebCore::Layout::TableFormattingContext::layoutInFlowContent):
(WebCore::Layout::TableFormattingContext::layoutCell):
(WebCore::Layout::TableFormattingContext::positionTableCells):
(WebCore::Layout::TableFormattingContext::setComputedGeometryForRows):
(WebCore::Layout::TableFormattingContext::setComputedGeometryForSections):
(WebCore::Layout::TableFormattingContext::ensureTableGrid):
(WebCore::Layout::TableFormattingContext::computePreferredWidthForColumns):
(WebCore::Layout::TableFormattingContext::computeAndDistributeExtraHorizontalSpace):
(WebCore::Layout::TableFormattingContext::useAsContentLogicalWidth):
(WebCore::Layout::TableFormattingContext::layoutTableCellBox): Deleted.

layout/tableformatting/TableFormattingContext.h:
layout/tableformatting/TableGrid.cpp:

(WebCore::Layout::TableGrid::Column::Column):
(WebCore::Layout::TableGrid::Column::hasFixedWidth const):
(WebCore::Layout::TableGrid::Columns::addColumn):
(WebCore::Layout::TableGrid::Columns::addAnonymousColumn):
(WebCore::Layout::TableGrid::Rows::addRow):
(WebCore::Layout::TableGrid::Row::Row):
(WebCore::Layout::TableGrid::Cell::Cell):
(WebCore::Layout::TableGrid::Slot::Slot):
(WebCore::Layout::TableGrid::slot):
(WebCore::Layout::TableGrid::appendCell):
(WebCore::Layout::TableGrid::insertCell):
(WebCore::Layout::TableGrid::removeCell):
(WebCore::Layout::TableGrid::widthConstraints):
(WebCore::Layout::TableGrid::ColumnsContext::addColumn): Deleted.
(WebCore::Layout::TableGrid::CellInfo::CellInfo): Deleted.
(WebCore::Layout::TableGrid::SlotInfo::SlotInfo): Deleted.

layout/tableformatting/TableGrid.h:

(WebCore::Layout::TableGrid::totalHorizontalSpacing const):
(WebCore::Layout::TableGrid::hasComputedWidthConstraints const):
(WebCore::Layout::TableGrid::Column::box const):
(WebCore::Layout::TableGrid::Columns::list):
(WebCore::Layout::TableGrid::Columns::list const):
(WebCore::Layout::TableGrid::Columns::size const):
(WebCore::Layout::TableGrid::Columns::logicalWidth const):
(WebCore::Layout::TableGrid::Row::logicalBottom const):
(WebCore::Layout::TableGrid::Row::box const):
(WebCore::Layout::TableGrid::Rows::list):
(WebCore::Layout::TableGrid::Rows::rowList const):
(WebCore::Layout::TableGrid::Rows::size const):
(WebCore::Layout::TableGrid::Cell::startColumn const):
(WebCore::Layout::TableGrid::Cell::endColumn const):
(WebCore::Layout::TableGrid::Cell::startRow const):
(WebCore::Layout::TableGrid::Cell::endRow const):
(WebCore::Layout::TableGrid::Cell::columnSpan const):
(WebCore::Layout::TableGrid::Cell::rowSpan const):
(WebCore::Layout::TableGrid::Cell::position const):
(WebCore::Layout::TableGrid::Cell::size const):
(WebCore::Layout::TableGrid::Cell::box const):
(WebCore::Layout::TableGrid::columns const):
(WebCore::Layout::TableGrid::columns):
(WebCore::Layout::TableGrid::rows const):
(WebCore::Layout::TableGrid::rows):
(WebCore::Layout::TableGrid::cells):
(WebCore::Layout::TableGrid::CellInfo::startColumn const): Deleted.
(WebCore::Layout::TableGrid::CellInfo::endColumn const): Deleted.
(WebCore::Layout::TableGrid::CellInfo::startRow const): Deleted.
(WebCore::Layout::TableGrid::CellInfo::endRow const): Deleted.
(WebCore::Layout::TableGrid::CellInfo::columnSpan const): Deleted.
(WebCore::Layout::TableGrid::CellInfo::rowSpan const): Deleted.
(WebCore::Layout::TableGrid::Column::columnBox const): Deleted.
(WebCore::Layout::TableGrid::ColumnsContext::columns): Deleted.
(WebCore::Layout::TableGrid::ColumnsContext::columns const): Deleted.
(WebCore::Layout::TableGrid::ColumnsContext::logicalWidth const): Deleted.
(WebCore::Layout::TableGrid::columnsContext const): Deleted.
(WebCore::Layout::TableGrid::columnsContext): Deleted.

6:27 AM Changeset in webkit [259977] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.28

Merge r258978 - [GTK] Add user agent quirk for auth.mayohr.com
https://bugs.webkit.org/show_bug.cgi?id=209378

Patch by Ting-Wei Lan <Ting-Wei Lan> on 2020-03-25
Reviewed by Michael Catanzaro.

Source/WebCore:

This site blocks the login page completely and asks the user to download
Google Chrome when using our standard user agent. It only blocks the
login page. After logging in, it works fine without user agent quirks.

Note that it is required to hide Version/X from the user agent string
for the site to recognize it as a valid Chrome user agent. Since Chrome
itself does not use Version/X, it should be safe to modify the existing
quirk instead of adding a new one to handle it.

platform/UserAgentQuirks.cpp:

(WebCore::urlRequiresChromeBrowser):
(WebCore::urlRequiresNoSafariVersion):
(WebCore::UserAgentQuirks::quirksForURL):
(WebCore::UserAgentQuirks::stringForQuirk):

platform/UserAgentQuirks.h:
platform/glib/UserAgentGLib.cpp:

(WebCore::buildUserAgentString):

Tools:

TestWebKitAPI/Tests/WebCore/UserAgentQuirks.cpp:

(TestWebKitAPI::assertUserAgentForURLHasNoSafariVersionQuirk):
(TestWebKitAPI::TEST):

6:27 AM Changeset in webkit [259976] by Carlos Garcia Campos

16 edits in releases/WebKitGTK/webkit-2.28/Source

Merge r259284 - [GTK][WPE] Jumpy rendering of fixed-element layers while scrolling
https://bugs.webkit.org/show_bug.cgi?id=209466

Reviewed by Carlos Garcia Campos.

Source/WebCore:

Avoid intermittent state application that can occur when asynchronous
scrolling is done on the dedicated thread while the general scene update
is being done in parallel on the composition thread, leading to partial
scrolling updates that visually present themselves as e.g. fixed
elements "jumping" around the view.

Instead of the staging state of a given Nicosia::CompositionLayer, the
scrolling nodes now update the base state with the given scrolling
change. At the end of the update, inside the UpdateScope descructor,
the updated states inside the scene are flushed into the staging phase
before they are adopted by the composition thread.

page/scrolling/nicosia/ScrollingTreeFixedNode.cpp:

(WebCore::ScrollingTreeFixedNode::applyLayerPositions):

page/scrolling/nicosia/ScrollingTreeFrameScrollingNodeNicosia.cpp:

(WebCore::ScrollingTreeFrameScrollingNodeNicosia::repositionScrollingLayers):
(WebCore::ScrollingTreeFrameScrollingNodeNicosia::repositionRelatedLayers):

page/scrolling/nicosia/ScrollingTreeOverflowScrollProxyNode.cpp:

(WebCore::ScrollingTreeOverflowScrollProxyNode::applyLayerPositions):

page/scrolling/nicosia/ScrollingTreeOverflowScrollingNodeNicosia.cpp:

(WebCore::ScrollingTreeOverflowScrollingNodeNicosia::repositionScrollingLayers):

page/scrolling/nicosia/ScrollingTreePositionedNode.cpp:

(WebCore::ScrollingTreePositionedNode::applyLayerPositions):

page/scrolling/nicosia/ScrollingTreeStickyNode.cpp:

(WebCore::ScrollingTreeStickyNode::applyLayerPositions):

platform/graphics/nicosia/NicosiaPlatformLayer.h:

(Nicosia::CompositionLayer::accessStaging): Deleted.

platform/graphics/nicosia/NicosiaSceneIntegration.cpp:

(Nicosia::SceneIntegration::SceneIntegration):
(Nicosia::SceneIntegration::invalidate):
(Nicosia::SceneIntegration::UpdateScope::~UpdateScope):

platform/graphics/nicosia/NicosiaSceneIntegration.h:

(Nicosia::SceneIntegration::create):

platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.cpp:

(WebCore::CoordinatedGraphicsLayer::syncPosition):
Don't signal the related attribute delta flag since this change is now
strictly applied by the scrolling thread.
(WebCore::CoordinatedGraphicsLayer::syncBoundsOrigin): Ditto.

Source/WebKit:

Move the Nicosia::SceneIntegration ownership into the
CompositingCoordinator class, along with the SceneIntegration::Client
inheritance. LayerTreeHost in turn now implements the updateScene()
method that triggers a scene update when invoked.

WebProcess/WebPage/CoordinatedGraphics/CompositingCoordinator.cpp:

(WebKit::CompositingCoordinator::CompositingCoordinator):
(WebKit::CompositingCoordinator::invalidate):
(WebKit::CompositingCoordinator::attachLayer):
(WebKit::CompositingCoordinator::requestUpdate):

WebProcess/WebPage/CoordinatedGraphics/CompositingCoordinator.h:
WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:

(WebKit::LayerTreeHost::LayerTreeHost):
(WebKit::LayerTreeHost::~LayerTreeHost):
(WebKit::LayerTreeHost::updateScene):
(WebKit::LayerTreeHost::sceneIntegration): Deleted.
(WebKit::LayerTreeHost::requestUpdate): Deleted.

WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.h:

6:27 AM Changeset in webkit [259975] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r258759 - [GStreamer] White-list vp09 in the codec registry
https://bugs.webkit.org/show_bug.cgi?id=209288

Reviewed by Xabier Rodriguez-Calvar.

platform/graphics/gstreamer/GStreamerRegistryScanner.cpp:

(WebCore::GStreamerRegistryScanner::initialize):

6:17 AM Changeset in webkit [259974] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Tools

Merge r259339 - [GTK] MiniBrowser: valgrind claims 'Invalid write of size 8' on close
https://bugs.webkit.org/show_bug.cgi?id=209471

Reviewed by Adrian Perez de Castro.

This is because we need to remove the weak pointer added in child windows to their parent.

MiniBrowser/gtk/BrowserWindow.c:

(browserWindowDispose): Remove the parentWindow weak ref.
(browser_window_class_init): Add dispose.

6:17 AM Changeset in webkit [259973] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259712 - Ensure base cache path exists before calculating disk cache capacity
https://bugs.webkit.org/show_bug.cgi?id=209338

Reviewed by Adrian Perez de Castro.

The base cache path might not exist yet in Cache::open() since it's Storage::open() the one calling
FileSystem::makeAllDirectories() with the final cache directory. In such case, computeCapacity() fails because
FileSystem::getVolumeFreeSpace() needs an existing file.

NetworkProcess/cache/NetworkCache.cpp:

(WebKit::NetworkCache::Cache::open):

6:17 AM Changeset in webkit [259972] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259944 - [GTK] Use gdk_window_move_to_rect() to position popup menus
https://bugs.webkit.org/show_bug.cgi?id=210069

Reviewed by Michael Catanzaro.

UIProcess/gtk/WebPopupMenuProxyGtk.cpp:

(WebKit::WebPopupMenuProxyGtk::showPopupMenu):

6:17 AM Changeset in webkit [259971] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259943 - [GTK] Page is not drawn with PSON, hardware acceleration and address typed in URL bar
https://bugs.webkit.org/show_bug.cgi?id=208091

Reviewed by Michael Catanzaro.

This happens only when accelerated compositing mode policy is set to always. In this case, the
bindSurfaceToWebPage wayland message is received before the WebPageProxy has updated the web page identifier. We
should always check if there's a provisional page and get its web page identifier.

UIProcess/gtk/WaylandCompositor.cpp:

(WebKit::WaylandCompositor::bindSurfaceToWebPage):

6:16 AM Changeset in webkit [259970] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259942 - [GTK][PSON] Crash in WaylandCompositor::Surface destructor with USE_WPE_RENDERER disabled
https://bugs.webkit.org/show_bug.cgi?id=209345

Reviewed by Michael Catanzaro.

In case of PSON the web page identifier can change fo the same web page proxy. When swapping processes the new
web page calls bindSurfaceToWebPage(), but we still have the web page proxy bound to the previous surface. We
just need to unbind the current surface before binding the new one.

UIProcess/gtk/WaylandCompositor.cpp:

(WebKit::WaylandCompositor::bindSurfaceToWebPage):

6:16 AM Changeset in webkit [259969] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.28/Source

Merge r258831 - [WPE] AsyncScrolling: horizontal scrolling is inverted
https://bugs.webkit.org/show_bug.cgi?id=208638

Source/WebCore:

Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2020-03-23
Reviewed by Adrian Perez de Castro.

Scrolling down should scroll to the right and up to the left..

page/scrolling/nicosia/ScrollingTreeFrameScrollingNodeNicosia.cpp:

(WebCore::ScrollingTreeFrameScrollingNodeNicosia::handleWheelEvent): Invert also the x axis.

page/scrolling/nicosia/ScrollingTreeOverflowScrollingNodeNicosia.cpp:

(WebCore::ScrollingTreeOverflowScrollingNodeNicosia::handleWheelEvent): Ditto.

Source/WebKit:

Patch by Chris Lord <Chris Lord> on 2020-03-23
Reviewed by Adrian Perez de Castro.

UIProcess/API/wpe/ScrollGestureController.cpp:

(WebKit::ScrollGestureController::handleEvent):

6:16 AM Changeset in webkit [259968] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r259073 - [GStreamer] Fix missing NULL-check in setSyncOnClock
https://bugs.webkit.org/show_bug.cgi?id=209609

Unreviewed, simple fix.

platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:

(WebCore::setSyncOnClock): Some systems are configured so that
audio sinks are not available. Make sure not to crash when asking
to sync with a NULL sink.

6:04 AM Changeset in webkit [259967] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259459 - Unreviewed fix after r259382

Rubber-stamped by Carlos Garcia Campos.

No new tests needed.

UIProcess/API/glib/WebKitWebContext.cpp:

(webkit_web_context_register_uri_scheme): Change g_return_if_fail()
calls to actually check that the passed scheme is not one of the
special ones.

6:04 AM Changeset in webkit [259966] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259382 - [WPE][GTK] Public API should not allow trying to register a special URI scheme
https://bugs.webkit.org/show_bug.cgi?id=209900

Reviewed by Carlos Garcia Campos.

No new tests needed.

UIProcess/API/glib/WebKitWebContext.cpp:

(webkit_web_context_register_uri_scheme): Use g_return_if_fail() to
check at the public API level whether the passed URI scheme is
special and bail out early.

6:04 AM Changeset in webkit [259965] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259380 - [GTK] [2.28.0] The Yelp build crashes if DISPLAY is not set
https://bugs.webkit.org/show_bug.cgi?id=209431

Reviewed by Carlos Alberto Lopez Perez.

Remove the assert when display is not X11 nor Wayland, and simply return false to ensure acceleration is not used.

UIProcess/gtk/AcceleratedBackingStore.cpp:

(WebKit::AcceleratedBackingStore::checkRequirements):

6:04 AM Changeset in webkit [259964] by Carlos Garcia Campos

3 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r259113 - [WPE] Unnecessary gl synchronization when using an OpenMAX video decoder and GLES2
https://bugs.webkit.org/show_bug.cgi?id=209647

Reviewed by Adrian Perez de Castro.

Don't perform the call to gst_gl_sync_meta_wait_cpu() when using an OpenMAX decoder,
as we don't need synchronization in that case and the internal call to glFinish()
casues an important fps drop.

platform/graphics/gstreamer/GStreamerCommon.h:
platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:

(WebCore::GstVideoFrameHolder::waitForCPUSync):
(WebCore::MediaPlayerPrivateGStreamer::createGSTPlayBin):

6:04 AM Changeset in webkit [259963] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r259110 - Update Chrome and Firefox versions in user agent quirks
https://bugs.webkit.org/show_bug.cgi?id=209631

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-27
Reviewed by Carlos Garcia Campos.

platform/UserAgentQuirks.cpp:

(WebCore::UserAgentQuirks::stringForQuirk):

6:04 AM Changeset in webkit [259962] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r259044 - [GTK] Crash in WebKit::LayerTreeHost::LayerTreeHost with bubblewrap sandbox enabled
https://bugs.webkit.org/show_bug.cgi?id=209106

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-26
Reviewed by Carlos Garcia Campos.

Don't bind the WaylandCompositor socket unless we're running under Wayland and it's actually
started successfully.

UIProcess/Launcher/glib/BubblewrapLauncher.cpp:

(WebKit::bindWayland):

6:03 AM Changeset in webkit [259961] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebKit

Merge r258923 - [GTK] Crash in WebKit::LayerTreeHost::LayerTreeHost with bubblewrap sandbox enabled
https://bugs.webkit.org/show_bug.cgi?id=209106

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-24
Reviewed by Adrian Perez de Castro.

The bubblewrap sandbox is blocking access to WebKit's nested Wayland compositor. Apparently
nobody has ever tested the nested Wayland compositor with the sandbox enabled until now.

UIProcess/Launcher/glib/BubblewrapLauncher.cpp:

(WebKit::bindWayland):

6:03 AM Changeset in webkit [259960] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WTF

Merge r259134 - REGRESSION(r258857): Broke aarch64 JSCOnly CI
https://bugs.webkit.org/show_bug.cgi?id=209670

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-27
Reviewed by Carlos Alberto Lopez Perez.

Change aarch64 to use 4 KB rather than 64 KB as the ceiling on page size.

This change is definitely incorrect, because it will break our internal aarch64 CI that uses
64 KB pages. But maybe it will fix the public aarch64 CI bot that is using 4 KB pages?
Further investigation is required, because 64 KB should have been a safe value for all
platforms, but first step is to commit this and see what happens.

wtf/PageBlock.h:

6:03 AM Changeset in webkit [259959] by Carlos Garcia Campos

7 edits in releases/WebKitGTK/webkit-2.28/Source

Merge r258857 - REGRESSION(r249808): [GTK] Crash in JSC Config::permanentlyFreeze() on architecture ppc64el
https://bugs.webkit.org/show_bug.cgi?id=209236

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-23
Reviewed by Mark Lam.

Source/JavaScriptCore:

heap/MarkedBlock.h: Use new CeilingOnPageSize.
runtime/JSCConfig.cpp:

(JSC::Config::permanentlyFreeze): Use pageSize instead of vmPageSize.

runtime/JSCConfig.h: Use new CeilingOnPageSize.

Source/WTF:

Add new CeilingOnPageSize constants, for use in JSC, in order to centralize our compile-time
page size guessing into one place. Improve the implementation of pageSize() to
RELEASE_ASSERT() when CeilingOnPageSize is wrong, so we can detect and fix it if so. (It
will be even easier to detect if we change RELEASE_ASSERT_WITH_MESSAGE() to actually print
its message in release builds.) Change pageSize() to use sysconf(_SC_PAGESIZE), which is
specified by POSIX, instead of getpagesize(), which is nonstandard.

wtf/PageBlock.cpp:

(WTF::systemPageSize):
(WTF::pageSize):

wtf/PageBlock.h:

6:03 AM Changeset in webkit [259958] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/JavaScriptCore

Merge r258463 - Unreviewed. Fix windows build by making configSizeToProtect stay 4KB.

runtime/JSCConfig.h:

6:03 AM Changeset in webkit [259957] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.28/Source

Merge r258460 - configSizeToProtect should be 16KB
https://bugs.webkit.org/show_bug.cgi?id=209068

Reviewed by Keith Miller.

Source/bmalloc:

bmalloc/Gigacage.h:

Source/JavaScriptCore:

runtime/JSCConfig.h:

6:03 AM Changeset in webkit [259956] by Carlos Garcia Campos

4 edits in releases/WebKitGTK/webkit-2.28/Source

Merge r257662 - Clean up code with how we choose Gigacage sizes and whether or not to use Wasm fast memory
https://bugs.webkit.org/show_bug.cgi?id=208392

Reviewed by Yusuke Suzuki.

Source/bmalloc:

bmalloc/Gigacage.h:

Source/JavaScriptCore:

runtime/OptionsList.h:

6:03 AM Changeset in webkit [259955] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r258755 - Fix build with gstreamer 1.12
https://bugs.webkit.org/show_bug.cgi?id=209296

Patch by Mike Gorse <mgorse@suse.com> on 2020-03-20
Reviewed by Philippe Normand.

No new tests (build fix only).

platform/graphics/gstreamer/GLVideoSinkGStreamer.cpp:

(webKitGLVideoSinkChangeState): Add GST_VERSION_CHECK around check for
GST_STATE_CHANGE_READY_TO_READY.

6:03 AM Changeset in webkit [259954] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/JavaScriptCore

Merge r258719 - Unreviewed, build fix after r258717
https://bugs.webkit.org/show_bug.cgi?id=199295

llint/LowLevelInterpreter.asm:

6:03 AM Changeset in webkit [259953] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/JavaScriptCore

Merge r258717 - sanitizeStackForVMImpl writes below stack pointer, triggers huge warning spam from valgrind
https://bugs.webkit.org/show_bug.cgi?id=199295

Reviewed by Mark Lam.

During sanitizeStackForVMImpl, we should not access to the region beyond the stack-pointer.
This patch changes stack-pointer while sanitizeStackForVMImpl is zero-filling the old stack region.

llint/LowLevelInterpreter.asm:

6:03 AM Changeset in webkit [259952] by Carlos Garcia Campos

6 edits in releases/WebKitGTK/webkit-2.28/JSTests

Merge r258710 - [JSC][BigEndians] Several JSC stress tests failing
https://bugs.webkit.org/show_bug.cgi?id=194007

Reviewed by Michael Catanzaro.

Adapt the DataView tests to be able to pass on little endian as well
as the big endian hardware. Do so by introducing method that determine
the endianess and based on its value it saves the expected results in
the right format.

stress/dataview-get-cse.js:

(getIsLittleEndian):
(adjustForEndianess):
(test3.foo):
(test3):

stress/dataview-jit-get.js:

(getIsLittleEndian):
(test1.adjustForEndianess):
(test1):
(test2.adjustForEndianess):
(test2):
(test3.adjustForEndianess):
(test3):
(adjustForEndianessUint32):
(test4):
(test5):
(test6):
(test7):
(test8):

stress/dataview-jit-neuter.js:

(getIsLittleEndian):
(adjustForEndianess):
(test):
(test2):

stress/dataview-jit-set-nan.js:

(getIsLittleEndian):
(adjustForEndianessFloat32):
(test):

stress/dataview-jit-set.js:

(getIsLittleEndian):
(readHex):
(adjustForEndianessUint16):
(test):
(test2):
(adjustForEndianessUint32):
(test3):
(test4):
(adjustForEndianessFloat32):
(test5):
(adjustForEndianessFloat64):
(test6):
(isLittleEndian): Deleted.

6:03 AM Changeset in webkit [259951] by Carlos Garcia Campos

3 edits
2 adds in releases/WebKitGTK/webkit-2.28

Merge r258662 - Fix ReplaceSelectionCommand::InsertedNodes::willRemoveNodePreservingChildren crash
https://bugs.webkit.org/show_bug.cgi?id=208312

Patch by Eugene But <eugenebut@chromium.org> on 2020-03-18
Reviewed by Ryosuke Niwa

ReplaceSelectionCommand::InsertedNodes::willRemoveNodePreservingChildren
was crashing on dereferencing m_firstNodeInserted pointer. Before the crash
ReplaceSelectionCommand::InsertedNodes object received the following calls:

respondToNodeInsertion() with node A, which set m_firstNodeInserted and m_lastNodeInserted to A
willRemoveNode() with node B, which left m_firstNodeInserted and m_lastNodeInserted unchanged (A)
(node A was destroyed setting m_firstNodeInserted and m_lastNodeInserted to null)
respondToNodeInsertion() with node C, which set m_firstNodeInserted and m_lastNodeInserted to C
willRemoveNodePreservingChildren() with node C, which set m_firstNodeInserted to null and crashed

This patch checks m_firstNodeInserted before dereferencing and sets m_lastNodeInserted to null if
m_firstNodeInserted became null. It seems like having non-null value for m_lastNodeInserted would
be an invalid state.

Test: editing/pasteboard/insert-apple-style-span-after-timeout.html

editing/ReplaceSelectionCommand.cpp:

(WebCore::ReplaceSelectionCommand::InsertedNodes::willRemoveNodePreservingChildren):

LayoutTests:
Test for ReplaceSelectionCommand::InsertedNodes::willRemoveNodePreservingChildren crash fix
https://bugs.webkit.org/show_bug.cgi?id=208312

Patch by Eugene But <eugenebut@chromium.org> on 2020-03-18
Reviewed by Ryosuke Niwa

This test insers empty Apple-style-span after timeout to a non-empty document.

editing/pasteboard/insert-apple-style-span-after-timeout.html:

6:03 AM Changeset in webkit [259950] by Carlos Garcia Campos

3 edits
2 adds in releases/WebKitGTK/webkit-2.28

Merge r258532 - A change event gets dispatched when textarea gets changed without focus
https://bugs.webkit.org/show_bug.cgi?id=202144

Patch by ChangSeok Oh <ChangSeok Oh> on 2020-03-16
Reviewed by Ryosuke Niwa.

Source/WebCore:

A crash happens in WebCore::ValidationMessage::buildBubbleTree. An immediate reason
is that DOM tree is modified in buildBubbleTree triggered by a timer.
The function calls document.updateLayout() that causes a change event
for textarea to fire when something changed in the textarea.
This bug is not reproduced on Mac because buildBubbleTree is not called.
See ValidationMessage::setMessage.
On the other hand, the root cause of this issue is triggering the change event
for textarea even if it is not focused when a change is made. This behavior
is different to what Gecko and Chromium do. When loading the test, they do not
trigger the change event although the textarea is filled by the script
since the textarea is not focused. Only when we manually make a change (meaning
the textarea is focused by user input), the event gets dispatched. To fix it,
setChangedSinceLastFormControlChangeEvent(true) is moved below the focus check
in HTMLTextAreaElement::subtreeHasChanged();

Test: fast/forms/textfield-onchange-without-focus.html

html/HTMLTextAreaElement.cpp:

(WebCore::HTMLTextAreaElement::subtreeHasChanged):

LayoutTests:

The test should be identical to the extected result without crash.

fast/forms/textfield-onchange-without-focus-expected.html: Added.
fast/forms/textfield-onchange-without-focus.html: Added.

6:03 AM Changeset in webkit [259949] by Carlos Garcia Campos

4 edits in releases/WebKitGTK/webkit-2.28

Merge r258808 - [GTK] Use ${PYTHON_EXECUTABLE} to run generate-gtkdoc
https://bugs.webkit.org/show_bug.cgi?id=208970

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-21
Reviewed by Konstantin Tokarev.

There's no need to rely on shebangs when executing python scripts from CMake. This should
ensure the script always works even if the shebang does not.

Source/cmake/GtkDoc.cmake:

Tools:

We are allowed to use /usr/bin/env in this script because it's not installed (so not
packaged). Therefore we can make the shebang compatible with FreeBSD in this particularl
case.

gtkdoc/generate-gtkdoc:

6:03 AM Changeset in webkit [259948] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Tools

Merge r258328 - [GTK] Use #!/usr/bin/python3 shebang to run generate-gtkdoc
https://bugs.webkit.org/show_bug.cgi?id=208970

Patch by Michael Catanzaro <Michael Catanzaro> on 2020-03-12
Reviewed by Daniel Bates.

Fedora and Ubuntu are both patching generate-gtkdoc to use #!/usr/bin/python3 rather than

probably has similar rules.

Note this is the only script that attempts to use /usr/bin/python in tarball builds.
Everything else that does so is probably only used by webkit-build.

https://docs.fedoraproject.org/en-US/packaging-guidelines/Python/#_multiple_python_runtimes

gtkdoc/generate-gtkdoc:

6:02 AM Changeset in webkit [259947] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r258542 - [GStreamer][MSE] Playback rate update support
https://bugs.webkit.org/show_bug.cgi?id=208454

Reviewed by Xabier Rodriguez-Calvar.

Implement playback rate update support for the MSE player. Also
includes drive-by logging cleanups.

platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:

(WebCore::MediaPlayerPrivateGStreamer::updatePlaybackRate):
(WebCore::MediaPlayerPrivateGStreamer::setRate):
(WebCore::MediaPlayerPrivateGStreamer::setPreservesPitch):

platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:

(WebCore::MediaPlayerPrivateGStreamer::pipeline const):

platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.cpp:

(WebCore::MediaPlayerPrivateGStreamerMSE::seek):
(WebCore::MediaPlayerPrivateGStreamerMSE::doSeek):
(WebCore::MediaPlayerPrivateGStreamerMSE::maybeFinishSeek):
(WebCore::MediaPlayerPrivateGStreamerMSE::seekCompleted):
(WebCore::MediaPlayerPrivateGStreamerMSE::updatePlaybackRate): Deleted.
(WebCore::MediaPlayerPrivateGStreamerMSE::setRate): Deleted.

platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.h:

6:02 AM Changeset in webkit [259946] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r258199 - Unreviewed, !USE(GSTREAMER_GL) build fix after r258197.

platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:

(WebCore::GstVideoFrameHolder::platformLayerBuffer):

6:02 AM WebKitGTK/2.28.x edited by Carlos Garcia Campos

(diff)

6:02 AM Changeset in webkit [259945] by Carlos Garcia Campos

10 edits in releases/WebKitGTK/webkit-2.28/Source/WebCore

Merge r258197 - [GStreamer][GL] External OES textures rendering support
https://bugs.webkit.org/show_bug.cgi?id=208572

Patch by Philippe Normand <philn@igalia.com> on 2020-03-10
Reviewed by Žan Doberšek.

For hardware that contains native YUV samplers, some drivers may
only support external-oes import of YUV textures, so by supporting
this texture target in the MediaPlayer we can avoid some costly
operations in the pipeline.

platform/graphics/gstreamer/GLVideoSinkGStreamer.cpp:

(webKitGLVideoSinkSetMediaPlayerPrivate):

platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:

(WebCore::GstVideoFrameHolder::GstVideoFrameHolder):
(WebCore::GstVideoFrameHolder::platformLayerBuffer):
(WebCore::MediaPlayerPrivateGStreamer::paint):

platform/graphics/gstreamer/VideoTextureCopierGStreamer.cpp:

(WebCore::VideoTextureCopierGStreamer::copyVideoTextureToPlatformTexture):

platform/graphics/texmap/TextureMapperGL.cpp:

(WebCore::TextureMapperGL::drawTexturedQuadWithProgram):
(WebCore::TextureMapperGL::drawTextureExternalOES):

platform/graphics/texmap/TextureMapperGL.h:
platform/graphics/texmap/TextureMapperPlatformLayerBuffer.cpp:

(WebCore::TextureMapperPlatformLayerBuffer::clone):
(WebCore::TextureMapperPlatformLayerBuffer::paintToTextureMapper):

platform/graphics/texmap/TextureMapperPlatformLayerBuffer.h:
platform/graphics/texmap/TextureMapperShaderProgram.cpp:

(WebCore::TextureMapperShaderProgram::create):

platform/graphics/texmap/TextureMapperShaderProgram.h:

2:28 AM Changeset in webkit [259944] by Carlos Garcia Campos

2 edits in trunk/Source/WebKit

[GTK] Use gdk_window_move_to_rect() to position popup menus
https://bugs.webkit.org/show_bug.cgi?id=210069

Reviewed by Michael Catanzaro.

UIProcess/gtk/WebPopupMenuProxyGtk.cpp:

(WebKit::WebPopupMenuProxyGtk::showPopupMenu):

2:27 AM Changeset in webkit [259943] by Carlos Garcia Campos

2 edits in trunk/Source/WebKit

[GTK] Page is not drawn with PSON, hardware acceleration and address typed in URL bar
https://bugs.webkit.org/show_bug.cgi?id=208091

Reviewed by Michael Catanzaro.

UIProcess/gtk/WaylandCompositor.cpp:

(WebKit::WaylandCompositor::bindSurfaceToWebPage):

2:26 AM Changeset in webkit [259942] by Carlos Garcia Campos

2 edits in trunk/Source/WebKit

[GTK][PSON] Crash in WaylandCompositor::Surface destructor with USE_WPE_RENDERER disabled
https://bugs.webkit.org/show_bug.cgi?id=209345

Reviewed by Michael Catanzaro.

UIProcess/gtk/WaylandCompositor.cpp:

(WebKit::WaylandCompositor::bindSurfaceToWebPage):

12:42 AM Changeset in webkit [259941] by Ross Kirsling

8 edits in trunk

[ECMA-402] WebKit Intl does not allow calendar and numberingSystem options
https://bugs.webkit.org/show_bug.cgi?id=209784

Reviewed by Myles C. Maxfield.

JSTests:

stress/intl-datetimeformat.js:

Fix two test cases for old behavior.

test262/expectations.yaml:

Mark 16 test cases as passing.

Source/JavaScriptCore:

As an alternative to using ca and nu extensions in the locale string:

the Intl.DateTimeFormat constructor needs to be able to take calendar and numberingSystem options https://tc39.es/ecma402/#sec-initializedatetimeformat
the Intl.NumberFormat needs to be able to take a numberingSystem option https://tc39.es/ecma402/#sec-initializenumberformat

Since we already support ca and nu, this is a very simple addition.
The only interesting part is that we must verify that values for these options are 3-8 alphanumeric characters.

runtime/IntlDateTimeFormat.cpp:

(JSC::IntlDateTimeFormat::initializeDateTimeFormat):

runtime/IntlNumberFormat.cpp:

(JSC::IntlNumberFormat::initializeNumberFormat):
(JSC::IntlNumberFormat::resolvedOptions):

runtime/IntlObject.cpp:

(JSC::isUnicodeLocaleIdentifierType):

runtime/IntlObject.h:

Apr 11, 2020:

9:32 PM Changeset in webkit [259940] by ddkilzer@apple.com

11 edits in trunk/Source/WebKit

Follow-up: Add WARN_UNUSED_RETURN to decode methods in Source/WebKit/Platform/IPC
<https://webkit.org/b/210316>
<rdar://problem/61559305>

Changes based on feedback from Alex Christensen and Darin Adler:

Remove WARN_UNUSED_RETURN from methods returning Optional<>.
Place WARN_UNUSED_RETURN consistently before the return type.

Platform/IPC/ArgumentCoder.h:
Platform/IPC/ArgumentCoders.cpp:
Platform/IPC/ArgumentCoders.h:

(IPC::ArgumentCoder<OptionSet<T>>::decode):
(IPC::ArgumentCoder<Optional<T>>::decode):
(IPC::ArgumentCoder<Box<T>>::decode):
(IPC::TupleDecoderImpl::decode):
(IPC::TupleDecoderImpl<Type>::decode):
(IPC::TupleDecoder::decode):
(IPC::TupleDecoder<0>::decode):
(IPC::VariantCoder::decode):

Platform/IPC/Attachment.h:
Platform/IPC/DataReference.h:
Platform/IPC/Decoder.h:
Platform/IPC/FormDataReference.h:

(IPC::FormDataReference::decode):

Platform/IPC/ImageDataReference.h:

(IPC::ImageDataReference::decode):

Platform/IPC/SharedBufferDataReference.h:

(IPC::SharedBufferDataReference::decode):

Platform/IPC/StringReference.h:

8:13 PM Changeset in webkit [259939] by Jack Lee

3 edits
2 adds in trunk

Infinite loop in InsertListCommand::doApply()
https://bugs.webkit.org/show_bug.cgi?id=210354
<rdar://problem/61427778>

Reviewed by Darin Adler.

Source/WebCore:

Function startOfNextParagraph may return an empty position. Added null check to exit the while loop
and stop looking for next paragraph.

Test: editing/inserting/insert-list-end-of-table.html

editing/InsertListCommand.cpp:

(WebCore::InsertListCommand::doApply):

LayoutTests:

Added a regression test for the crash.

editing/inserting/insert-list-end-of-table-expected.txt: Added.
editing/inserting/insert-list-end-of-table.html: Added.

5:43 PM Changeset in webkit [259938] by Wenson Hsieh

9 edits in trunk

[macOS] [WK1] Touch Bar flashes when typing in Vietnamese in Mail
https://bugs.webkit.org/show_bug.cgi?id=210394
<rdar://problem/60099560>

Reviewed by Tim Horton.

Source/WebCore:

See WebKitLegacy/mac/ChangeLog for more details.

Currently, many users of TemporarySelectionChange use it to temporarily avoid propagating selection changes to
the client layer during temporary selection changes. This involves creating a TemporarySelectionChange without
a new selection, but with the IgnoreSelectionChanges option specified, which makes us call `Editor::
setIgnoreSelectionChanges` to suppress selection change notifications.

Do a bit of cleanup in this area by introducing IgnoreSelectionChangeForScope, which wraps a
TemporarySelectionChange and makes it easier for a handful of call sites that currently use
TemporarySelectionChange to hide selection changes from the client layer to get their desired behavior.

Test: CandidateTests.DoNotHideCandidatesDuringTextReplacement

editing/Editor.cpp:

(WebCore::Editor::respondToChangedSelection):

editing/Editor.h:

(WebCore::TemporarySelectionChange::TemporarySelectionChange):
(WebCore::IgnoreSelectionChangeForScope::IgnoreSelectionChangeForScope):

page/DragController.cpp:

(WebCore::DragController::performDragOperation):
(WebCore::DragController::insertDroppedImagePlaceholdersAtCaret):

Replace these:

TemporarySelectionChange ignoreSelectionChanges { frame, WTF::nullopt, TemporarySelectionOption::IgnoreSelectionChanges };

...with these instead:

IgnoreSelectionChangeForScope ignoreSelectionChanges { *frame };

Source/WebKitLegacy/mac:

In recent versions of macOS, changes to the animation of the candidate list touch bar item when it is expanded
or collapsed using -updateWithInsertionPointVisibility: mean that back-to-back calls to hide and show the
candidate list during the same runloop now result in a visible flicker, whereas it would remain still in prior
releases.

Combined with how -[WebHTMLView insertText:] in WebKitLegacy causes multiple selection change updates if the
argument is an attributed string with the text replacement attribute, this means that there is now a visible
flicker in the touch bar when using input methods that rely on the NSTextInputReplacementRangeAttributeName
attribute. One such input method is Simple Telex (Vietnamese), when typing causes characters in other parts of
the word to gain diacritics.

To fix this, we make two small adjustments in WebKitLegacy (see below).

WebView/WebHTMLView.mm:

(-[WebHTMLView insertText:]):

While inserting text (which may be an NSAttibutedString or NSString), we may come across an attribute indicating
which part of the document to replace. In this case, we will select this range in the document before inserting
the given text.

This selection change is propagated to the client layer, which makes us briefly think that we have a ranged
selection, causing the flicker described above. Address this by suppressing selection change notifications while
we're selecting the range to replace, using the new IgnoreSelectionChangeForScope RAII object (see WebCore
changes for more detail).

WebView/WebView.mm:

(-[WebView updateTextTouchBar]):

When updating the text touch bar (which contains the candidate list item) after a selection change, we currently
handle the case where selection changes are ignored by hiding the candidate list; this doesn't seem right, since
selection changes that are ignored should simply be ignored by the client. To fix this, push the
ignoreSelectionChanges() check into the if statement.

Tools:

Add an API test that inserts an attributed string with the text replacement attribute, and verifies that we
show the candidate list when inserting the string, and also don't temporarily hide the candidate list in the
process of doing so.

To achieve this, we swizzle out calls to -[NSCandidateListTouchBarItem updateWithInsertionPointVisibility:] to
detect when the candidate list is either hidden or made visible.

TestWebKitAPI/Tests/TestWebKitAPI/mac/AppKitSPI.h:
TestWebKitAPI/Tests/mac/CandidateTests.mm:

Fix a leak in an adjacent test too, while I'm touching this file.

(TestWebKitAPI::updateCandidateListWithVisibility):

4:56 PM Changeset in webkit [259937] by ddkilzer@apple.com

3 edits in trunk/Source/WebKit

Follow-up: Add WARN_UNUSED_RETURN to decode methods in WebCoreArgumentCoders
<https://webkit.org/b/210322>
<rdar://problem/61565148>

Changes based on feedback from Alex Christensen and Darin Adler:

Remove WARN_UNUSED_RETURN from methods returning Optional<>.
Place WARN_UNUSED_RETURN consistently before the return type.

Shared/WebCoreArgumentCoders.cpp:
Shared/WebCoreArgumentCoders.h:

4:49 PM Changeset in webkit [259936] by Simon Fraser

4 edits
4 adds in trunk

[Async overflow] Can't scroll overflow:scroll in sideways-scrollable RTL document
https://bugs.webkit.org/show_bug.cgi?id=210389

Reviewed by Tim Horton.

Source/WebCore:

ScrollingTree::handleWheelEvent() converts the event coordinates from view to "content"
coordinates, but we then jump into hit-testing on CALayers. In a sideways-scrollable
RTL document, the root content layer has a negative X offset which corresponds to
scrollOrigin; we need to map the point into the coordinate space of this layer
before entering layer-based hit-testing.

Tests: fast/scrolling/mac/async-scroll-overflow-rtl-zoomed.html

fast/scrolling/mac/async-scroll-overflow-rtl.html

page/scrolling/ScrollingTree.cpp:

(WebCore::ScrollingTree::handleWheelEvent):

page/scrolling/mac/ScrollingTreeMac.mm:

(ScrollingTreeMac::scrollingNodeForPoint):

LayoutTests:

fast/scrolling/mac/async-scroll-overflow-rtl-expected.txt: Added.
fast/scrolling/mac/async-scroll-overflow-rtl-zoomed-expected.txt: Added.
fast/scrolling/mac/async-scroll-overflow-rtl-zoomed.html: Added.
fast/scrolling/mac/async-scroll-overflow-rtl.html: Added.

4:21 PM Changeset in webkit [259935] by dbates@webkit.org

3 edits in trunk/Source/WebKit

Add text interaction delegate functions
https://bugs.webkit.org/show_bug.cgi?id=210395
<rdar://problem/61650549>

Reviewed by Wenson Hsieh.

Implement some delegate callbacks that are used to toggle selection reveal,
placeholder text, and selection assistance.

UIProcess/ios/WKContentViewInteraction.h:
UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView _willBeginTextInteractionInTextInputContext:]): Added.
(-[WKContentView _didFinishTextInteractionInTextInputContext:]): Added.

1:17 PM Changeset in webkit [259934] by Diego Pino Garcia

2 edits in trunk/LayoutTests

[GTK] Gardening of encrypted-media tests
https://bugs.webkit.org/show_bug.cgi?id=210391

Unreviewed gardening.

platform/gtk/TestExpectations:

12:36 PM Changeset in webkit [259933] by Darin Adler

22 edits in trunk/Source

Move more from live range to SimpleRange: callers of absoluteTextRects
https://bugs.webkit.org/show_bug.cgi?id=210369

Reviewed by Anders Carlsson.

Source/WebCore:

dom/Node.cpp:

(WebCore::Node::textRects const): Deleted.

dom/Node.h: Updated for the above.

dom/Range.cpp:

(WebCore::Range::absoluteBoundingBox const): Updated since absoluteTextRects
no longer has a RangeInFixedPosition* argument.
(WebCore::Range::absoluteTextRects const): Removed the unused RangeInFixedPosition*
argument.

dom/Range.h: Got rid of unused RangeInFixedPosition type and also removed

RangeInFixedPosition* argument from the absoluteTextRects function. Later will
remove absoluteTextRects entirely.

editing/Editor.cpp:

(WebCore::Editor::firstRectForRange const): Use RenderObject::absoluteTextQuads
and unitedBoundingBoxes rather than using RenderObject::absoluteBoundingBoxRectForRange.

editing/cocoa/DataDetection.mm:

(WebCore::DataDetection::detectContentInRange): Use SimpleRange rather than
live ranges.

html/HTMLTextFormControlElement.cpp:

(WebCore::setContainerAndOffsetForRange): Moved from int to unsigned.
(WebCore::HTMLTextFormControlElement::selection const): Return Optional<SimpleRange>
rather than a live range.

html/HTMLTextFormControlElement.h: Updated for the change above.

page/TextIndicator.cpp:

(WebCore::initializeIndicator): Updated since absoluteTextRects no longer takes
a RangeInFixedPosition* argument.

rendering/HighlightData.h: Removed stray obsolete declaration.

rendering/RenderObject.cpp:

(WebCore::RenderObject::absoluteBoundingBoxRectForRange): Deleted. Callers can
use absoluteTextQuads directly. We need to cut down on the number of separate
functions that are not really separate concepts, and this was used in only one place.
(WebCore::RenderObject::absoluteTextRects): Added. Replaces Range::absoluteTextRects
for all callers outside the live range class and will eventually replace it entirely.

rendering/RenderObject.h: Updated for the above.

Source/WebKit:

WebProcess/InjectedBundle/API/mac/WKDOMNode.mm:

(-[WKDOMNode textRects]): Moved implementation here. There's no reason for Node
to have a textRects function, since it's a layout/rendering operation and a
single node is also an unusual special case, but for now at least we will keep
this legacy method.

WebProcess/InjectedBundle/API/mac/WKDOMRange.mm:

(-[WKDOMRange textRects]): Changed to call RenderObject::absoluteTextRects
instead of Range::absoluteTextRects.

WebProcess/WebPage/FindController.cpp:

(WebKit::FindController::updateFindUIAfterPageScroll): Use
RenderObject::absoluteTextRects instead of Range::absoluteRectRects.
(WebKit::FindController::findStringMatches): Ditto.

Source/WebKitLegacy/ios:

WebCoreSupport/WebFrameIOS.mm:

(-[WebFrame closestCaretRectInMarkedTextRangeForPoint:]): Use
RenderObject::absoluteTextRects instead of Range::absoluteTextRects.
Also added some missing null checks and reorganized the function a
bit to make it more direct and slightly cleaerer.

Source/WebKitLegacy/mac:

DOM/DOM.mm:

(-[DOMNode textRects]): Moved implementation here. There's no reason for Node
to have a textRects function, since it's a layout/rendering operation and a
single node is also an unusual special case, but for now we will keep this
legacy method for compatibility with old clients.
(-[DOMRange textRects]): Changed to call RenderObject::absoluteTextRects
instead of Range::absoluteTextRects.

WebView/WebFrame.mm:

(-[WebFrame _rectsForRange:]): Changed to call -[DOMRange textRects] instead
of Range::absoluteTextRects. Don't really need two identical methods but
probably need to keep this around for compatiblitiy.

11:46 AM Changeset in webkit [259932] by Diego Pino Garcia

2 edits in trunk/LayoutTests

[GTK] Gardening of mediastream tests
https://bugs.webkit.org/show_bug.cgi?id=210386

Unreviewed gardening.

platform/gtk/TestExpectations:

11:36 AM Changeset in webkit [259931] by Devin Rousso

5 edits
2 adds in trunk/Source/WebCore

REGRESSION (Safari 13.1?): Web Inspector: Debugger hang at breakpoint when using Keyboard Maestro
https://bugs.webkit.org/show_bug.cgi?id=210177
<rdar://problem/61485723>

Reviewed by Joseph Pecoraro.

Partial revert of r251036 <https://webkit.org/b/202716> to go back to using AppKit APIs
instead of CFRunLoopRunInMode. Only seems to affect WebKitLegacy.

inspector/PageScriptDebugServer.h:
inspector/PageScriptDebugServer.cpp:

(WebCore::PageScriptDebugServer::runEventLoopWhilePausedInternal):
(WebCore::PageScriptDebugServer::platformShouldContinueRunningEventLoopWhilePaused):

inspector/mac/PageScriptDebugServerMac.mm: Added.

(WebCore::PageScriptDebugServer::platformShouldContinueRunningEventLoopWhilePaused):

SourcesCocoa.txt:
WebCore.xcodeproj/project.pbxproj:

8:35 AM Changeset in webkit [259930] by Darin Adler

19 edits in trunk/Source/WebCore

Use Node::length to replace Node::maxCharacterOffset and lastOffsetInNode; switch more offsets from int to unsigned
https://bugs.webkit.org/show_bug.cgi?id=210246

Reviewed by Antti Koivisto.

The recently-added Node::length, which matches the DOM specification terminology for node offsets as used in ranges, is the same as the existing maxCharacterOffset and lastOffsetInNode functions. Deleted all uses of those and replaced them with calls to Node::length. One of the benefits of this is that Node::length is implemented more efficiently and is not a virtual function. Another is consistently matching the DOM specification terminology.
Many offsets, including the ones in live ranges, are currently implemented as signed in WebKit, but are specified as unsigned in the DOM and HTML specifications. This has very little observable effect from JavaScript that can affect website compatibility, but it's still helpful to be consistent both with the specification and internally. Accordingly, changed some of these to unsigned; more to come later.

accessibility/AXObjectCache.cpp:

(WebCore::AXObjectCache::previousBoundary): Use length instead of
maxCharacterOffset.

dom/CharacterData.cpp:

(WebCore::CharacterData::maxCharacterOffset const): Deleted.

dom/CharacterData.h: Deleted maxCharacterOffset override.

dom/DocumentMarkerController.cpp:

(WebCore::DocumentMarkerController::shiftMarkers): Use length instead of
maxCharacterOffset.

dom/Node.cpp:

(WebCore::Node::maxCharacterOffset const): Deleted.

dom/Node.h: Deleted maxCharacterOffset.

dom/Position.cpp:

(WebCore::Position::computeOffsetInContainerNode const): Use length instead
of lastOffsetInNode.

dom/Position.h:

(WebCore::lastOffsetInNode): Deleted.
(WebCore::lastPositionInNode): Use length instead of lastOffsetInNode.
(WebCore::minOffsetForNode): Use length instead of maxCharacterOffset.
(WebCore::offsetIsBeforeLastNodeOffset): Ditto.

dom/RangeBoundaryPoint.h:

(WebCore::RangeBoundaryPoint::setToEndOfNode): Use length instead of
maxCharacterOffset.

editing/ApplyBlockElementCommand.cpp:

(WebCore::isNewLineAtPosition): Use length instead of maxCharacterOffset.
(WebCore::ApplyBlockElementCommand::rangeForParagraphSplittingTextNodesIfNeeded):
Ditto.

editing/ApplyStyleCommand.cpp:

(WebCore::ApplyStyleCommand::removeInlineStyle): Use length instead of
maxCharacterOffset.

editing/Editing.cpp:

(WebCore::lastOffsetForEditing): Use length instead of mmaxCharacterOffset
and countChildNodes. Also improved the comment here.

editing/EditingStyle.cpp:

(WebCore::EditingStyle::styleAtSelectionStart): Use length instead of
maxCharacterOffset.

editing/InsertListCommand.cpp:

(WebCore::InsertListCommand::doApplyForSingleParagraph): Use length instead
of lastOffsetInNode.

editing/TextIterator.cpp:

(WebCore::SimplifiedBackwardsTextIterator::SimplifiedBackwardsTextIterator):
USe length instead of lastOffsetInNode.

editing/VisibleUnits.cpp:

(WebCore::previousBoundary): Use length instead of maxCharacterOffset.

8:14 AM Changeset in webkit [259929] by Devin Rousso

6 edits in trunk/Source/WebInspectorUI

Web Inspector: CSS: variables should have a go-to arrow to quickly jump to the definition
https://bugs.webkit.org/show_bug.cgi?id=195775
<rdar://problem/48905785>

Reviewed by Timothy Hatcher.

Often when debugging/editing styles that use CSS variables, the desired workflow is to
modify the variable itself, not that specific usage. This is especially helpful when editing
colors saved in CSS variables.

Add an always visible go-to arrow immediately after the variable (e.g. --XYZ (->)) that
will jump to the declaration of the variable elsewhere in the style cascade.

UserInterface/Views/SpreadsheetStyleProperty.js:

(WI.SpreadsheetStyleProperty.prototype._addVariableTokens):

UserInterface/Views/SpreadsheetCSSStyleDeclarationEditor.css:

(.spreadsheet-style-declaration-editor .property .select-variable-property): Added.

UserInterface/Base/Setting.js:
UserInterface/Views/SettingsTabContentView.js:

(WI.SettingsTabContentView.prototype._createExperimentalSettingsView):

Localizations/en.lproj/localizedStrings.js:

7:07 AM Changeset in webkit [259928] by Diego Pino Garcia

2 edits in trunk/LayoutTests

[GTK] Gardening, mark some WebInspector tests as Slow
https://bugs.webkit.org/show_bug.cgi?id=210380

Unreviewed gardening.

platform/gtk/TestExpectations:

5:36 AM Changeset in webkit [259927] by aakash_jain@apple.com

2 edits in trunk/Tools

Buildbot: Force crash log submission after each test run (follow-up fix)
https://bugs.webkit.org/show_bug.cgi?id=210202
<rdar://problem/60508929>

Unreviewed follow-up fix.

BuildSlaveSupport/ews-build/steps.py:

(WaitForCrashCollection): Remove spaces from step name.

5:07 AM Changeset in webkit [259926] by aakash_jain@apple.com

2 edits in trunk/Tools

Buildbot: Force crash log submission after each test run (follow-up fix)
https://bugs.webkit.org/show_bug.cgi?id=210202
<rdar://problem/60508929>

Unreviewed follow-up fix.

BuildSlaveSupport/wait-for-crash-collection: Improved error message.

4:28 AM Changeset in webkit [259925] by ddkilzer@apple.com

3 edits in trunk/Tools

check-webkit-style should warn about 'decode' functions with missing WARN_UNUSED_RETURN attribute
<https://webkit.org/b/210336>
<rdar://problem/61585835>

Reviewed by Jonathan Bedard.

Scripts/webkitpy/style/checkers/cpp.py:

(_FunctionState.post_modifiers): Add. Returns string of text
that appears after the function definition's list of parameters.
This is used to find the WARN_UNUSED_RETURN attribute in
function declarations in header files.
(_FunctionState.has_attribute): Add. Takes a string (or regex)
and searchs before and after the function definition for the
attribute.
(_FunctionState.has_return_type): Add. Takes a string (or
regex) and checks the return type of the fuction.
(_FunctionState.is_static): Add. Returns True for functions
with the 'static' modifier.
(check_function_definition): Add check for decode() functions
with missing WARN_UNUSED_RETURN attribute.
(CppChecker): Add 'security/missing_warn_unused_return' to the
list of enabled style checkers.

Scripts/webkitpy/style/checkers/cpp_unittest.py:

(CppStyleTestBase.perform_function_definition_check): Add. This
runs the check_function_definition() function in cpp.py.
(CppStyleTest.test_decode_functions_missing_warn_unused_return): Add.
Tests for the new warning.

2:41 AM Changeset in webkit [259924] by Diego Pino Garcia

2 edits in trunk/LayoutTests

[GTK] Gardening of flaky failures
https://bugs.webkit.org/show_bug.cgi?id=210376

Context Navigation

Timeline

Apr 12, 2020:

Apr 11, 2020:

Download in other formats: