Context Navigation

← Previous Period
Next Period →

Timeline

Feb 9, 2020:

9:49 PM Changeset in webkit [256110] by basuke.suzuki@sony.com

2 edits in trunk/Source/WebCore

Replace redundant functions with variadic template function.
https://bugs.webkit.org/show_bug.cgi?id=207413

Reviewed by Darin Adler.

No new tests. Covered by existing tests.

loader/TextResourceDecoder.cpp:

(WebCore::bytesEqual):

8:47 PM Changeset in webkit [256109] by commit-queue@webkit.org

9 edits in trunk

[GTK][WPE] Expose allowTopNavigationToDataURL
https://bugs.webkit.org/show_bug.cgi?id=207384

Patch by Lauro Moura <Lauro Moura> on 2020-02-09
Reviewed by Adrian Perez de Castro.

Source/WebKit:

This patch exposes the new property "allow-top-navigation-to-data-urls"
from r255961 to the glib-based APIs.

UIProcess/API/glib/WebKitSettings.cpp:

(webKitSettingsSetProperty): Set new property.
(webKitSettingsGetProperty): Get new property.
(webkit_settings_class_init): Register new property.
(webkit_settings_get_allow_top_navigation_to_data_urls): Added.
(webkit_settings_set_allow_top_navigation_to_data_urls): Added.

UIProcess/API/gtk/WebKitSettings.h:
UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
UIProcess/API/wpe/WebKitSettings.h:
UIProcess/API/wpe/docs/wpe-1.0-sections.txt:

Tools:

Updates with new "allow-top-navigation-to-data-urls" property.

TestWebKitAPI/Tests/WebKitGLib/TestUIClient.cpp:

(testWebViewAllowModalDialogs): This test requires this property to be
set.

TestWebKitAPI/Tests/WebKitGLib/TestWebKitSettings.cpp:

(testWebKitSettings): Added tests for getter/setter.

6:04 PM Changeset in webkit [256108] by Keith Rollin

20 edits in trunk/Source

Re-enable LTO for ARM builds
https://bugs.webkit.org/show_bug.cgi?id=207402
<rdar://problem/49190767>

Reviewed by Sam Weinig.

Bug 190758 re-enabled LTO for Production builds for x86-family CPUs.
Enabling it for ARM was left out due to a compiler issue. That issue
has been fixed, and so now we can re-enable LTO for ARM.

Source/bmalloc:

Configurations/Base.xcconfig:

Source/JavaScriptCore:

Configurations/Base.xcconfig:

Source/ThirdParty/ANGLE:

Configurations/Base.xcconfig:

Source/ThirdParty/libwebrtc:

Configurations/Base.xcconfig:

Source/WebCore:

No new tests -- no new or changed functionality.

Configurations/Base.xcconfig:

Source/WebCore/PAL:

Configurations/Base.xcconfig:

Source/WebInspectorUI:

Configurations/Base.xcconfig:

Source/WebKit:

Configurations/Base.xcconfig:

Source/WebKitLegacy/mac:

Configurations/Base.xcconfig:

Source/WTF:

Configurations/Base.xcconfig:

5:51 PM Changeset in webkit [256107] by Antti Koivisto

7 edits in trunk/Source/WebCore

[LFC][Integration] Support intruding floats
https://bugs.webkit.org/show_bug.cgi?id=207099

Reviewed by Zalan Bujtas.

Add support for for floats placed by the legacy render tree layout intruding to IFC flow.

This was the last frequently used feature supported by the simple line layout but not
the integrated LFC (pagination is the final missing piece).

layout/floats/FloatingContext.cpp:

(WebCore::Layout::FloatingContext::remove): Deleted.

layout/floats/FloatingContext.h:
layout/floats/FloatingState.cpp:

(WebCore::Layout::FloatingState::FloatItem::FloatItem):
(WebCore::Layout::FloatingState::FloatingState):
(WebCore::Layout::FloatingState::remove): Deleted.

Delete some unused functions.

layout/floats/FloatingState.h:

(WebCore::Layout::FloatingState::FloatItem::isLeftPositioned const):
(WebCore::Layout::FloatingState::clear):
(WebCore::Layout::FloatingState::FloatItem::operator== const): Deleted.

layout/integration/LayoutIntegrationLineLayout.cpp:

(WebCore::LayoutIntegration::LineLayout::canUseFor):

Allow floats.

(WebCore::LayoutIntegration::LineLayout::layout):
(WebCore::LayoutIntegration::LineLayout::prepareLayoutState):
(WebCore::LayoutIntegration::LineLayout::prepareFloatingState):

Translate floats from render tree to IFC float structures.

layout/integration/LayoutIntegrationLineLayout.h:

5:34 PM Changeset in webkit [256106] by Alan Bujtas

4 edits in trunk/Source/WebCore

[LFC][BFC] Move updatePositiveNegativeMarginValues out of MarginCollapse class
https://bugs.webkit.org/show_bug.cgi?id=207445
<rdar://problem/59297879>

Reviewed by Antti Koivisto.

MarginCollapse::updatePositiveNegativeMarginValues is just a static helper.
Let's split this function so that it can be just a regular member function (move the state updating to the BlockFormattingContext caller).

layout/blockformatting/BlockFormattingContext.cpp:

(WebCore::Layout::BlockFormattingContext::computeHeightAndMargin):

layout/blockformatting/BlockFormattingContext.h:
layout/blockformatting/BlockMarginCollapse.cpp:

(WebCore::Layout::BlockFormattingContext::MarginCollapse::resolvedPositiveNegativeMarginValues):
(WebCore::Layout::BlockFormattingContext::MarginCollapse::updatePositiveNegativeMarginValues): Deleted.

2:42 PM Changeset in webkit [256105] by Wenson Hsieh

2 edits in trunk/Source/WebKit

Fix the watchOS build after <rdar://problem/56134081>
https://bugs.webkit.org/show_bug.cgi?id=207443
<rdar://problem/59295173>

Reviewed by Maciej Stachowiak.

Use the version of the SPI that's guaranteed to be available across all platforms.

UIProcess/ios/WKContentViewInteraction.mm:

(gestureRecognizerModifierFlags):

1:36 PM Changeset in webkit [256104] by Kate Cheney

73 edits in trunk

[ Mac wk2 ] http/tests/resourceLoadStatistics/log-cross-site-load-with-link-decoration.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=207271
<rdar://problem/59190346>

Reviewed by Maciej Stachowiak.

Source/WebKit:

This patch adds a check to make sure a test is expecting a statistics
processing call before proceeding. Otherwise, this can cause
flakiness in ITP tests if data is removed/changed during the test.

NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:

(WebKit::ResourceLoadStatisticsStore::processStatisticsAndDataRecords):

LayoutTests:

This patch updates all ITP tests to follow the pattern
setEnableFeature(true) at test setup and setEnableFeature(false) when
finished. This enables/disables both ITP and the isRunningTest
parameter so tests don't have to make two calls. It also ensures that
each test which relies on statistics processing notifies the testRunner
so it knows to wait for a callback, and deletes unecessary calls
to setStatisticsNotifyPagesWhenDataRecordsWereScanned which is called
in installStatisticsDidScanDataRecordsCallback.

This will help limit flakiness by processing statistics only for ITP
tests which require it for their purpose.

http/tests/resourceLoadStatistics/classify-as-non-prevalent-based-on-mixed-statistics-database.html:
http/tests/resourceLoadStatistics/classify-as-non-prevalent-based-on-mixed-statistics.html:
http/tests/resourceLoadStatistics/classify-as-non-prevalent-based-on-subresource-under-top-frame-origins-database.html:
http/tests/resourceLoadStatistics/classify-as-non-prevalent-based-on-subresource-under-top-frame-origins.html:
http/tests/resourceLoadStatistics/classify-as-non-prevalent-based-on-subresource-unique-redirects-to-database.html:
http/tests/resourceLoadStatistics/classify-as-non-prevalent-based-on-subresource-unique-redirects-to.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-mixed-statistics-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-mixed-statistics.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-sub-frame-under-top-frame-origins-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-sub-frame-under-top-frame-origins.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-subresource-redirect-to-prevalent-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-subresource-redirect-to-prevalent.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-subresource-under-top-frame-origins-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-subresource-under-top-frame-origins.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-subresource-unique-redirects-to-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-subresource-unique-redirects-to.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-top-frame-redirect-to-prevalent-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-top-frame-redirect-to-prevalent.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-top-frame-unique-redirects-to-database.html:
http/tests/resourceLoadStatistics/classify-as-prevalent-based-on-top-frame-unique-redirects-to.html:
http/tests/resourceLoadStatistics/classify-as-very-prevalent-based-on-mixed-statistics-database.html:
http/tests/resourceLoadStatistics/classify-as-very-prevalent-based-on-mixed-statistics.html:
http/tests/resourceLoadStatistics/clear-in-memory-and-persistent-store-database.html:
http/tests/resourceLoadStatistics/clear-in-memory-and-persistent-store-one-hour-database.html:
http/tests/resourceLoadStatistics/clear-in-memory-and-persistent-store-one-hour.html:
http/tests/resourceLoadStatistics/clear-in-memory-and-persistent-store.html:
http/tests/resourceLoadStatistics/non-prevalent-resource-with-user-interaction-database.html:
http/tests/resourceLoadStatistics/non-prevalent-resource-with-user-interaction.html:
http/tests/resourceLoadStatistics/non-prevalent-resource-without-user-interaction-database.html:
http/tests/resourceLoadStatistics/non-prevalent-resource-without-user-interaction.html:
http/tests/resourceLoadStatistics/non-sandboxed-iframe-redirect-localhost-to-ip-to-localhost-database.html:
http/tests/resourceLoadStatistics/non-sandboxed-iframe-redirect-localhost-to-ip-to-localhost.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-ip-to-localhost-to-ip-database.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-ip-to-localhost-to-ip.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-localhost-to-ip-to-localhost-database.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-localhost-to-ip-to-localhost.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-ip-to-localhost-to-ip-database.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-ip-to-localhost-to-ip.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-localhost-to-ip-to-localhost-database.html:
http/tests/resourceLoadStatistics/non-sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-localhost-to-ip-to-localhost.html:
http/tests/resourceLoadStatistics/prevalent-resource-handled-keydown-database.html:
http/tests/resourceLoadStatistics/prevalent-resource-handled-keydown.html:
http/tests/resourceLoadStatistics/prevalent-resource-unhandled-keydown-database.html:
http/tests/resourceLoadStatistics/prevalent-resource-unhandled-keydown.html:
http/tests/resourceLoadStatistics/prevalent-resource-with-user-interaction-database.html:
http/tests/resourceLoadStatistics/prevalent-resource-with-user-interaction-timeout-database.html:
http/tests/resourceLoadStatistics/prevalent-resource-with-user-interaction-timeout.html:
http/tests/resourceLoadStatistics/prevalent-resource-with-user-interaction.html:
http/tests/resourceLoadStatistics/prevalent-resource-without-user-interaction-database.html:
http/tests/resourceLoadStatistics/prevalent-resource-without-user-interaction.html:
http/tests/resourceLoadStatistics/prune-statistics-database.html:
http/tests/resourceLoadStatistics/prune-statistics.html:
http/tests/resourceLoadStatistics/resources/util.js:

(setEnableFeature):

http/tests/resourceLoadStatistics/sandboxed-iframe-redirect-ip-to-localhost-to-ip-database.html:
http/tests/resourceLoadStatistics/sandboxed-iframe-redirect-ip-to-localhost-to-ip.html:
http/tests/resourceLoadStatistics/sandboxed-iframe-redirect-localhost-to-ip-to-localhost-database.html:
http/tests/resourceLoadStatistics/sandboxed-iframe-redirect-localhost-to-ip-to-localhost.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-ip-to-localhost-to-ip-database.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-ip-to-localhost-to-ip.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-localhost-to-ip-to-localhost-database.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-non-sandboxed-iframe-redirect-localhost-to-ip-to-localhost.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-ip-to-localhost-to-ip-database.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-ip-to-localhost-to-ip.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-localhost-to-ip-to-localhost-database.html:
http/tests/resourceLoadStatistics/sandboxed-nesting-iframe-with-sandboxed-iframe-redirect-localhost-to-ip-to-localhost.html:
http/tests/resourceLoadStatistics/telemetry-generation-advanced-functionality-database.html:
http/tests/resourceLoadStatistics/telemetry-generation-basic-functionality-database.html:
http/tests/resourceLoadStatistics/telemetry-generation.html:
http/tests/resourceLoadStatistics/user-interaction-in-cross-origin-sub-frame-database.html:
http/tests/resourceLoadStatistics/user-interaction-in-cross-origin-sub-frame.html:

12:47 PM Changeset in webkit [256103] by weinig@apple.com

15 edits in trunk

Move JavaScriptCore related feature defines from FeatureDefines.xcconfig to PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=207436

Reviewed by Tim Horton.

Source/JavaScriptCore:

Configurations/FeatureDefines.xcconfig:

Remove ENABLE_FAST_JIT_PERMISSIONS and ENABLE_FTL_JIT.

Source/WebCore:

Configurations/FeatureDefines.xcconfig:

Remove ENABLE_FAST_JIT_PERMISSIONS and ENABLE_FTL_JIT.

Source/WebCore/PAL:

Configurations/FeatureDefines.xcconfig:

Remove ENABLE_FAST_JIT_PERMISSIONS and ENABLE_FTL_JIT.

Source/WebKit:

Configurations/FeatureDefines.xcconfig:

Remove ENABLE_FAST_JIT_PERMISSIONS and ENABLE_FTL_JIT.

Source/WebKitLegacy/mac:

Configurations/FeatureDefines.xcconfig:

Remove ENABLE_FAST_JIT_PERMISSIONS and ENABLE_FTL_JIT.

Source/WTF:

wtf/PlatformEnable.h:

Add default values for ENABLE_FAST_JIT_PERMISSIONS and ENABLE_SEPARATED_WX_HEAP.

wtf/PlatformEnableCocoa.h:

Added ENABLE_FAST_JIT_PERMISSIONS (from FeatureDefines.xcconfig), ENABLE_SEPARATED_WX_HEAP
(from PlatformEnable.h) and ENABLE_FTL_JIT (from both FeatureDefines.xcconfig and PlatformEnable.h).

Tools:

TestWebKitAPI/Configurations/FeatureDefines.xcconfig:

Remove ENABLE_FAST_JIT_PERMISSIONS and ENABLE_FTL_JIT.

12:24 PM WebKitGTK/2.28.x edited by Adrian Perez de Castro

(diff)

12:22 PM WebKitGTK/2.28.x edited by Adrian Perez de Castro

(diff)

9:57 AM Changeset in webkit [256102] by Simon Fraser

4 edits in trunk/Source/WebCore

Optimize Style::determineChange()
https://bugs.webkit.org/show_bug.cgi?id=207438

Reviewed by Antti Koivisto.

Style::determineChange() called RenderStyle::operator!=() before the testing some other
sets of properties for inequality.

It's faster to call inheritedEqual() etc before the full style compare. Also optimize
comparing alignItems() and justifyItems() by adding a helper function that first tests
for equality of the m_rareNonInheritedData pointer. These (added in r213480) return Inherit
because align-items and justify-items affect child layout.

This is a ~4% progression on some MotionMark subtests. Time under TreeResolver::createAnimatedElementUpdate()
drops from 2.4% to 0.9%.

rendering/style/RenderStyle.cpp:

(WebCore::RenderStyle::descendantAffectingNonInheritedPropertiesEqual const):

rendering/style/RenderStyle.h:
style/StyleChange.cpp:

(WebCore::Style::determineChange):

9:46 AM Changeset in webkit [256101] by Alan Bujtas

5 edits in trunk/Source/WebCore

[LFC] FloatingContext::constraints should take a vertical range instead of just vertical position.
https://bugs.webkit.org/show_bug.cgi?id=207440
<rdar://problem/59295022>

Reviewed by Antti Koivisto.

In an IFC, this covers the entire line when searching for intrusive float to shrink the available horizontal space.
(This fix is limited to content with fixed line height)

layout/blockformatting/BlockFormattingContext.cpp:

(WebCore::Layout::BlockFormattingContext::usedAvailableWidthForFloatAvoider const):

layout/floats/FloatingContext.cpp:

(WebCore::Layout::FloatingContext::constraints const):

layout/floats/FloatingContext.h:
layout/inlineformatting/InlineFormattingContext.cpp:

(WebCore::Layout::InlineFormattingContext::lineLayout):
(WebCore::Layout::InlineFormattingContext::constraintsForLine):

Feb 8, 2020:

8:12 PM Changeset in webkit [256100] by Alan Bujtas

7 edits in trunk/Source/WebCore

[LFC][BFC] Replace "estimated" term with "precomputed"
https://bugs.webkit.org/show_bug.cgi?id=207433
<rdar://problem/59288794>

Reviewed by Antti Koivisto.

The term "pre-computed" describes better what this value is.

layout/MarginTypes.h:

(WebCore::Layout::EstimatedMarginBefore::usedValue const): Deleted.

layout/blockformatting/BlockFormattingContext.cpp:

(WebCore::Layout::BlockFormattingContext::usedAvailableWidthForFloatAvoider const):
(WebCore::Layout::BlockFormattingContext::precomputeVerticalPosition):
(WebCore::Layout::BlockFormattingContext::precomputeVerticalPositionForAncestors):
(WebCore::Layout::BlockFormattingContext::precomputeVerticalPositionForFormattingRoot):
(WebCore::Layout::BlockFormattingContext::computePositionToAvoidFloats):
(WebCore::Layout::BlockFormattingContext::computeHeightAndMargin):
(WebCore::Layout::BlockFormattingContext::setPrecomputedMarginBefore):
(WebCore::Layout::BlockFormattingContext::hasPrecomputedMarginBefore const):
(WebCore::Layout::BlockFormattingContext::computeEstimatedVerticalPosition): Deleted.
(WebCore::Layout::BlockFormattingContext::computeEstimatedVerticalPositionForAncestors): Deleted.
(WebCore::Layout::BlockFormattingContext::setEstimatedMarginBefore): Deleted.
(WebCore::Layout::BlockFormattingContext::hasEstimatedMarginBefore const): Deleted.

layout/blockformatting/BlockFormattingContext.h:
layout/blockformatting/BlockMarginCollapse.cpp:

(WebCore::Layout::BlockFormattingContext::MarginCollapse::hasClearance const):
(WebCore::Layout::BlockFormattingContext::MarginCollapse::updateMarginAfterForPreviousSibling):
(WebCore::Layout::BlockFormattingContext::MarginCollapse::positiveNegativeValues const):
(WebCore::Layout::BlockFormattingContext::MarginCollapse::precomputedMarginBefore):
(WebCore::Layout::BlockFormattingContext::MarginCollapse::estimatedMarginBefore): Deleted.

layout/displaytree/DisplayBox.cpp:

(WebCore::Display::Box::Box):

layout/displaytree/DisplayBox.h:

(WebCore::Display::Box::setHasPrecomputedMarginBefore):
(WebCore::Display::Box::invalidatePrecomputedMarginBefore):
(WebCore::Display::Box::top const):
(WebCore::Display::Box::topLeft const):
(WebCore::Display::Box::setVerticalMargin):
(WebCore::Display::Box::setHasEstimatedMarginBefore): Deleted.
(WebCore::Display::Box::invalidateEstimatedMarginBefore): Deleted.

7:45 PM Changeset in webkit [256099] by Alan Bujtas

3 edits in trunk/Source/WebCore

[LFC][BFC] Pre-compute vertical position only when we really need it.
https://bugs.webkit.org/show_bug.cgi?id=207432
<rdar://problem/59288539>

Reviewed by Antti Koivisto.

Instead of pre-computing the vertical position at computeStaticVerticalPosition, let's compute it when

we are about to layout a new non-float-avoider/float clear FC (e.g. so that intrusive floats can shrink the lines inside IFCs)
we are about to compute the non-static vertical position of a float avoider block level box.

layout/blockformatting/BlockFormattingContext.cpp:

(WebCore::Layout::BlockFormattingContext::layoutInFlowContent):
(WebCore::Layout::BlockFormattingContext::computeStaticVerticalPosition):
(WebCore::Layout::BlockFormattingContext::computeEstimatedVerticalPositionForAncestors):
(WebCore::Layout::BlockFormattingContext::precomputeVerticalPositionForFormattingRoot):
(WebCore::Layout::BlockFormattingContext::computePositionToAvoidFloats):
(WebCore::Layout::BlockFormattingContext::computeVerticalPositionForFloatClear):
(WebCore::Layout::BlockFormattingContext::computeStaticPosition): Deleted.
(WebCore::Layout::BlockFormattingContext::computeEstimatedVerticalPositionForFormattingRoot): Deleted.
(WebCore::Layout::BlockFormattingContext::computeEstimatedVerticalPositionForFloatClear): Deleted.
(WebCore::Layout::BlockFormattingContext::hasPrecomputedMarginBefore const): Deleted.
(WebCore::Layout::BlockFormattingContext::computeFloatingPosition): Deleted.

layout/blockformatting/BlockFormattingContext.h:

6:51 PM Changeset in webkit [256098] by Alan Bujtas

5 edits in trunk/Source/WebCore

[LFC][BFC] Remove establishesBlockFormattingContextOnly and establishesInlineFormattingContextOnly
https://bugs.webkit.org/show_bug.cgi?id=207431
<rdar://problem/59288366>

Reviewed by Antti Koivisto.

The current caller of establishesInlineFormattingContextOnly is actually interested in whether the box is float avoider
(and the debug-only caller of establishesBlockFormattingContextOnly should instead check if the box also establishes an IFC).

layout/LayoutContext.cpp:

(WebCore::Layout::LayoutContext::createFormattingContext):

layout/blockformatting/BlockFormattingContext.cpp:

(WebCore::Layout::BlockFormattingContext::computeEstimatedVerticalPositionForFormattingRoot):

layout/layouttree/LayoutBox.cpp:

(WebCore::Layout::Box::establishesInlineFormattingContextOnly const): Deleted.
(WebCore::Layout::Box::establishesBlockFormattingContextOnly const): Deleted.

layout/layouttree/LayoutBox.h:

6:01 PM Changeset in webkit [256097] by Alan Bujtas

2 edits in trunk/Source/WebCore

[LFC][BFC] Adjust Box::isFloatAvoider() for Inline and Independent FCs.
https://bugs.webkit.org/show_bug.cgi?id=207430
<rdar://problem/59288236>

Reviewed by Antti Koivisto.

A block container box that establishes both BFC and IFC is not a float avoider (add support for independent formatting context too).

layout/layouttree/LayoutBox.cpp:

(WebCore::Layout::Box::isFloatAvoider const):

5:29 PM Changeset in webkit [256096] by weinig@apple.com

2 edits in trunk/Source/WTF

Migrate definition of USE_PLATFORM_REGISTERS_WITH_PROFILE to PlatformUse.h
https://bugs.webkit.org/show_bug.cgi?id=207428

Reviewed by Tim Horton.

wtf/PlatformUse.h:

Move definition of USE_PLATFORM_REGISTERS_WITH_PROFILE from AdditionalFeatureDefines.h, where
it did not belong.

5:11 PM Changeset in webkit [256095] by Simon Fraser

8 edits
8 adds in trunk

Extent of a composited animation should not include the untransformed position
https://bugs.webkit.org/show_bug.cgi?id=207434

Reviewed by Sam Weinig.
Source/WebCore:

To determine whether to create ("attach") backing store for layers with transform animations,
we compute the union of all the states of the animation as an "extent", and ask whether it intersects
the coverage rect in GraphicsLayerCA.

The non-Web Animations transition code did this correctly, just unioning the bounds transformed by
the start and end state.

The non-Web Animations keyframe code, and the Web Animations KeyframeEffect code (used for both CSS transitions
and animations) also unioned with the unanimated bounds, which could create overly large extents in some
cases, contributing to jetsams on iOS (rdar://problem/59020443).

Fix KeyframeAnimation and KeyframeEffect to not union with the untransformed bounds.

Tests: compositing/backing/backing-store-attachment-animating-outside-viewport.html

legacy-animation-engine/compositing/backing/backing-store-attachment-animating-outside-viewport.html

animation/KeyframeEffect.cpp:

(WebCore::KeyframeEffect::computeExtentOfTransformAnimation const):

page/animation/KeyframeAnimation.cpp:

(WebCore::KeyframeAnimation::computeExtentOfTransformAnimation const):

LayoutTests:

New test that checks backing store attachment and overlap for an element which is positioned in-view,
but is move offscreen by the animation.

New baselines for overlap tests, since overlap no longer considers the unanimated position.

compositing/backing/backing-store-attachment-animating-outside-viewport-expected.txt: Added.
compositing/backing/backing-store-attachment-animating-outside-viewport.html: Added.
compositing/layer-creation/translate-animation-overlap-expected.txt:
compositing/layer-creation/translate-scale-animation-overlap-expected.txt:
legacy-animation-engine/compositing/backing/backing-store-attachment-animating-outside-viewport-expected.txt: Added.
legacy-animation-engine/compositing/backing/backing-store-attachment-animating-outside-viewport.html: Added.
legacy-animation-engine/compositing/layer-creation/translate-animation-overlap-expected.txt:
legacy-animation-engine/compositing/layer-creation/translate-scale-animation-overlap-expected.txt:
platform/ios-wk2/compositing/backing/backing-store-attachment-animating-outside-viewport-expected.txt: Added.
platform/ios-wk2/legacy-animation-engine/compositing/backing/backing-store-attachment-animating-outside-viewport-expected.txt: Added.

1:54 PM Changeset in webkit [256094] by Alan Bujtas

2 edits in trunk/Source/WebCore

[LFC][BFC] Box::isFloatAvoider should not check for isFloatingPositioned()
https://bugs.webkit.org/show_bug.cgi?id=207429
<rdar://problem/59288204>

Reviewed by Antti Koivisto.

Box::isFloatAvoider() already checks for establishesBlockFormattingContext() (and floats do establish BFCs).

layout/layouttree/LayoutBox.cpp:

(WebCore::Layout::Box::isFloatAvoider const):

1:50 PM Changeset in webkit [256093] by ysuzuki@apple.com

4 edits in trunk

[WTF] Try using 75% load factor for HashTable
https://bugs.webkit.org/show_bug.cgi?id=207183

Reviewed by Mark Lam.

Source/WTF:

We know that hash-table is one of the most memory consuming part in WebKit.
By analyzing many production hash-table implementations[1], I found that many
of them are using 75% load-factor while our one is 50%.

This patch changes the load-factor from 50% to 75%. But we pick 75% only for
small tables which capacity is <= 1024 based on collected data by micro-benchmarking.
The collected data is telling that longer probe-length affects on performance if table
size gets larger.

[1]: LLVM DenseMap, Abseil's, rust's, and so on.

wtf/HashTable.h:

(WTF::HashTableCapacityForSize::shouldExpand):
(WTF::HashTableCapacityForSize::capacityForSize):
(WTF::HashTable::shouldExpand const):
(WTF::KeyTraits>::computeBestTableSize):

Tools:

TestWebKitAPI/Tests/WTF/HashSet.cpp:

(TestWebKitAPI::testInitialCapacity):

12:36 PM Changeset in webkit [256092] by Wenson Hsieh

2 edits in trunk/Source/WebKit

REGRESSION (r256072): ScrollViewScrollabilityTests.ScrollableWithOverflowHiddenAndInputView fails
https://bugs.webkit.org/show_bug.cgi?id=207425
Followup to <rdar://problem/56960774>

Reviewed by Tim Horton.

The failing test ScrollViewScrollabilityTests.ScrollableWithOverflowHiddenAndInputView relies on -[WKWebView
_zoomToFocusRect:...:] to modify the zoom scale of WKScrollView in order for the next layer tree transaction to
have a zoom scale not equal to the initial scale, which in turn ensures that we make the UIScrollView scrollable
after the next remote layer tree update.

The change in r256072 caused us to bail early in -_zoomToFocusRect: and -_zoomToPoint: in the case where the
scroll view is not scrollable (i.e. the page has overflow: hidden); this means that if all the following
conditions are met, we now avoid zooming into focused elements, whereas we would've zoomed before the change:

The user is focusing an element on an iPhone.
The document body has overflow: hidden;, or the WebKit client has called -setScrollEnabled:NO.
Either an SPI client has vended an empty input view and input accessory view, or the focused element has

inputmode="none".

To fix the test and restore zooming behavior, we partially revert the change in r256072, such that we no longer
bail from WKWebView's zooming codepaths if the scrollEnabled property is NO.

UIProcess/API/ios/WKWebViewIOS.mm:

(-[WKWebView _zoomToPoint:atScale:animated:]):
(-[WKWebView _zoomToFocusRect:selectionRect:insideFixed:fontSize:minimumScale:maximumScale:allowScaling:forceScroll:]):

10:29 AM Changeset in webkit [256091] by weinig@apple.com

26 edits in trunk

Move trivial definitions from FeatureDefines.xcconfig to PlatformEnableCocoa.h
https://bugs.webkit.org/show_bug.cgi?id=207155

Reviewed by Tim Horton.
Source/JavaScriptCore:

Move all trivial definitions (just ENABLE_FOO = ENABLE_FOO; or ENABLE_BAR = ;)
from the FeatureDefines.xcconfigs to PlatformEnableCocoa.h, ensuring each one
also has a default value in PlatformEnable.h

To support the move, DerivedSources.make has been updated to generate the list
of ENABLE_* features by directly from preprocessing Platform.h, rather than
just getting the partial list from the xcconfig file.

Configurations/FeatureDefines.xcconfig:
DerivedSources.make:

Source/WebCore:

To support the move, DerivedSources.make and generate-unified-sources.sh have been
updated to generate the list of ENABLE_* features by directly from preprocessing
Platform.h, rather than just getting the partial list from the xcconfig file.

Configurations/FeatureDefines.xcconfig:
Configurations/WebCore.xcconfig:
DerivedSources.make:
Scripts/generate-unified-sources.sh:

Source/WebCore/PAL:

Configurations/FeatureDefines.xcconfig:
Configurations/PAL.xcconfig:

Remove quotes as they should not be necessary (other uses of BUILT_PRODUCTS_DIR
are not quoted) and complicate scripts that want to use HEADER_SEARCH_PATHS.

Source/WebKit:

Configurations/FeatureDefines.xcconfig:
DerivedSources.make:
Scripts/generate-unified-sources.sh:

Source/WebKitLegacy:

To support the move, generate-unified-sources.sh has been updated to generate
the list of ENABLE_* features by directly from preprocessing Platform.h, rather
than just getting the partial list from the xcconfig file.

scripts/generate-unified-sources.sh:

Source/WebKitLegacy/mac:

Configurations/FeatureDefines.xcconfig:
MigrateHeaders.make:

Source/WTF:

ENABLE_INSPECTOR_TELEMETRY, ENABLE_TEXT_AUTOSIZING, ENABLE_WEBGL, ENABLE_METER_ELEMENT,
ENABLE_VIDEO_TRACK, ENABLE_WEB_AUDIO and ENABLE_CURSOR_VISIBILITY are all now always
set to 1 for Cocoa (if not defined already), as this was already the case via the xcconfig.

wtf/PlatformEnable.h:
wtf/PlatformEnableCocoa.h:

Tools:

TestWebKitAPI/Configurations/FeatureDefines.xcconfig:

7:48 AM Changeset in webkit [256090] by wilander@apple.com

3 edits in trunk/Source/WebKit

Resource Load Statistics: Hold off non-cookie website data deletion until an hour after user interaction
https://bugs.webkit.org/show_bug.cgi?id=207418
<rdar://problem/58550164>

Reviewed by Chris Dumez.

This change makes sure there is at least a one-hour or older timestamp
for user interaction for some website before activating deletion of
non-cookie website data for sites the user has not interacted with.
This ensures that a fresh start such as after a reset or on a new
device doesn't interpret the lack of data as lack of user interaction.

No new tests. Existing tests make sure we don't regress the functionality.

NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:

(WebKit::ResourceLoadStatisticsDatabaseStore::registrableDomainsToRemoveWebsiteDataFor):

NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:

(WebKit::ResourceLoadStatisticsMemoryStore::registrableDomainsToRemoveWebsiteDataFor):

6:10 AM Changeset in webkit [256089] by commit-queue@webkit.org

4 edits
2 adds in trunk

Crash in RenderTreeBuilder::Table::findOrCreateParentForChild with multicol spanner
https://bugs.webkit.org/show_bug.cgi?id=206917

Patch by Doug Kelly <Doug Kelly> on 2020-02-08
Reviewed by Zalan Bujtas.

Source/WebCore:

During render tree construction, multi-column spanners are moved from their original tree position to
next to the enclosing anonymous fragmented flow and we mark their previous location with a spanner placeholder.
If we try to add a new renderer right before the spanner (beforeChild is the spanner renderer), we need to use
the spanner's original position as the insertion point.
This patch addresses the mismatching position issue by adjusting the spanner beforeChild right before
we start searching for the final insertion point.

Test: fast/multicol/spanner-crash-when-finding-table-parent.html

rendering/updating/RenderTreeBuilder.cpp:

(WebCore::RenderTreeBuilder::attach):

rendering/updating/RenderTreeBuilderTable.cpp:

(WebCore::RenderTreeBuilder::Table::findOrCreateParentForChild):

LayoutTests:

fast/multicol/spanner-crash-when-finding-table-parent-expected.txt: Added.
fast/multicol/spanner-crash-when-finding-table-parent.html: Added.

2:35 AM Changeset in webkit [256088] by basuke.suzuki@sony.com

5 edits
2 deletes in trunk/Source/bmalloc

[bmalloc] VMHeap can be merge into Heap
https://bugs.webkit.org/show_bug.cgi?id=207410

Reviewed by Yusuke Suzuki.

VMHeap has only one member function in it and Heap is the only client of that.
No member variable is defined. It does nothing special with its context as a class.
It is safe to merge the function into Heap.

CMakeLists.txt:
bmalloc.xcodeproj/project.pbxproj:
bmalloc/Heap.cpp:

(bmalloc::Heap::allocateLarge):
(bmalloc::Heap::tryAllocateLargeChunk): Moved from VMHeap.

bmalloc/Heap.h:
bmalloc/VMHeap.cpp: Removed.
bmalloc/VMHeap.h: Removed.

Note: See TracTimeline for information about the timeline view.

Download in other formats:

RSS Feed