Context Navigation

← Previous Period
Next Period →

Timeline

Aug 28, 2019:

11:49 PM Changeset in webkit [249247] by mark.lam@apple.com

5 edits
1 add in trunk

DFG/FTL: We should prefetch structures and do a loadLoadFence before doing PrototypeChainIsSane checks.
https://bugs.webkit.org/show_bug.cgi?id=201281
<rdar://problem/54028228>

Reviewed by Yusuke Suzuki and Saam Barati.

JSTests:

stress/structure-storedPrototype-should-only-assert-on-the-mutator-thread.js: Added.

Source/JavaScriptCore:

This (see title above) is already the preferred idiom used in most places in our
compiler, except for 2: DFG's SpeculativeJIT::compileGetByValOnString() and FTL's
compileStringCharAt(). Consider the following:

bool prototypeChainIsSane = false;
if (globalObject->stringPrototypeChainIsSane()) {

...
m_graph.registerAndWatchStructureTransition(globalObject->stringPrototype()->structure(vm()));
m_graph.registerAndWatchStructureTransition(globalObject->objectPrototype()->structure(vm()));

prototypeChainIsSane = globalObject->stringPrototypeChainIsSane();

}

What's essential for correctness here is that the stringPrototype and objectPrototype
structures be loaded before the loads in the second stringPrototypeChainIsSane()
check. Without a loadLoadFence before the second stringPrototypeChainIsSane()
check, we can't guarantee that. Elsewhere in the compiler, the preferred idiom
for doing this right is to pre-load the structures first, do a loadLoadFence, and
then do the IsSane check just once after e.g.

Structure* arrayPrototypeStructure = globalObject->arrayPrototype()->structure(m_vm);
Structure* objectPrototypeStructure = globalObject->objectPrototype()->structure(m_vm);

if (arrayPrototypeStructure->transitionWatchpointSetIsStillValid() has loadLoadFences.

&& objectPrototypeStructure->transitionWatchpointSetIsStillValid() has loadLoadFences.
&& globalObject->arrayPrototypeChainIsSane()) {

m_graph.registerAndWatchStructureTransition(arrayPrototypeStructure);
m_graph.registerAndWatchStructureTransition(objectPrototypeStructure);
...

}

This patch changes DFG's SpeculativeJIT::compileGetByValOnString() and FTL's
compileStringCharAt() to follow the same idiom.

We also fix a bad assertion in Structure::storedPrototype() and
Structure::storedPrototypeObject(). The assertion is only correct when those
methods are called from the mutator thread. The assertion has been updated to
only check its test condition if the current thread is the mutator thread.

dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::compileGetByValOnString):

ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileStringCharAt):

runtime/StructureInlines.h:

(JSC::Structure::storedPrototype const):
(JSC::Structure::storedPrototypeObject const):

11:46 PM Changeset in webkit [249246] by Devin Rousso

3 edits in trunk/Source/WebInspectorUI

Web Inspector: REGRESSION(r249078): JavaScript autocomplete doesn't work when evaluating properties of values
https://bugs.webkit.org/show_bug.cgi?id=201226

Reviewed by Joseph Pecoraro.

r249078 modified WI.JavaScriptRuntimeCompletionProvider to use arrays of property names
instead of objects for completion, but a few code paths were missed.

UserInterface/Controllers/JavaScriptRuntimeCompletionProvider.js:

(WI.JavaScriptRuntimeCompletionProvider.prototype.completionControllerCompletionsNeeded.evaluated):
(WI.JavaScriptRuntimeCompletionProvider.prototype.completionControllerCompletionsNeeded.receivedPropertyNamesFromEvaluate):
(WI.JavaScriptRuntimeCompletionProvider.prototype.completionControllerCompletionsNeeded.receivedObjectPropertyNames): Added.
(WI.JavaScriptRuntimeCompletionProvider.prototype.completionControllerCompletionsNeeded.receivedArrayPropertyNames):

UserInterface/Models/CallFrame.js:

(WI.CallFrame.prototype.collectScopeChainVariableNames):
(WI.CallFrame.prototype.collectScopeChainVariableNames.propertiesCollected):

11:26 PM Changeset in webkit [249245] by Fujii Hironori

2 edits in trunk/Tools

[Win] MiniBrowser crashes in WKURLCopyString if WKPageCopyActiveURL returns null
https://bugs.webkit.org/show_bug.cgi?id=201215

Reviewed by Don Olmstead.

MiniBrowser crashed if it was going to go to a unreachable page
because WKPageCopyActiveURL returned a nullptr.

MiniBrowser/win/WebKitBrowserWindow.cpp:

(createString): Added null checking of the argument.

11:13 PM Changeset in webkit [249244] by zandobersek@gmail.com

2 edits in trunk/Source/WebCore

[Nicosia] Implement layer representation retain, release mechanics
https://bugs.webkit.org/show_bug.cgi?id=201133

Reviewed by Carlos Garcia Campos.

page/scrolling/nicosia/ScrollingStateNodeNicosia.cpp:

(WebCore::LayerRepresentation::retainPlatformLayer):
Type-cast the layer object to the reference-counted
Nicosia::PlatformLayer type and reference that object.
(WebCore::LayerRepresentation::releasePlatformLayer):
Ditto, but dereference the object.

11:13 PM Changeset in webkit [249243] by zandobersek@gmail.com

2 edits in trunk/Source/WebCore

[Nicosia] Add Nicosia::PlatformLayer::accessStaging() helper
https://bugs.webkit.org/show_bug.cgi?id=201132

Reviewed by Carlos Garcia Campos.

platform/graphics/nicosia/NicosiaPlatformLayer.h:

(Nicosia::CompositionLayer::accessStaging):
Add an accessor into the staging state of the Nicosia::PlatformLayer
object. This will be needed for the application of scrolling changes
in the asynchronous scrolling system.

11:12 PM Changeset in webkit [249242] by zandobersek@gmail.com

3 edits in trunk/Source/WebCore

[CoordGraphics] Expose Nicosia layer as the underlying CoordinatedGraphicsLayer platform layer
https://bugs.webkit.org/show_bug.cgi?id=201131

Reviewed by Carlos Garcia Campos.

Add the CoordinatedGraphicsLayer::platformLayer() override, returning
the Nicosia::CompositionLayer object as the underlying platform layer.
This will come in handy for asynchronous scrolling.

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

(WebCore::CoordinatedGraphicsLayer::platformLayer const):

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

11:11 PM Changeset in webkit [249241] by zandobersek@gmail.com

2 edits in trunk/Source/WebCore

[Nicosia] Polish ScrollingCoordinator implementation
https://bugs.webkit.org/show_bug.cgi?id=201130

Reviewed by Carlos Garcia Campos.

These changes primarily mirror the Mac implementation of this class.

page/scrolling/nicosia/ScrollingCoordinatorNicosia.cpp:

(WebCore::ScrollingCoordinatorNicosia::pageDestroyed):
Upon releasing the ThreadedScrollingTree, it should also be invalidated
on the scrolling thread.
(WebCore::ScrollingCoordinatorNicosia::commitTreeState):
Invoke willCommitTree() at the beginning of this method. Additionally
the pending-commit-count is incremented for the ThreadedScrollingTree
object.

10:57 PM Changeset in webkit [249240] by Kocsen Chung

4 edits in branches/safari-608-branch

Cherry-pick r249224. rdar://problem/54819205

Null check webFrame when creating a print preview to prevent a crash.
https://bugs.webkit.org/show_bug.cgi?id=201237
<rdar://problem/51618863>

Reviewed by Tim Horton.

Source/WebKit:

Move and expend a null check to keep from crashing when making a print preview.

UIProcess/mac/WKPrintingView.mm: (-[WKPrintingView _drawPreview:]): (-[WKPrintingView drawRect:]):

Tools:

Test to verify that if we don't have the WebPageProxy, we will not crash when making a print preview.

TestWebKitAPI/Tests/WebKitCocoa/UIDelegate.mm: (TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@249224 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10:57 PM Changeset in webkit [249239] by Kocsen Chung

3 edits
3 adds in branches/safari-608-branch

Cherry-pick r249156. rdar://problem/54775048

Crash under WebCore::jsNotificationConstructorPermission
https://bugs.webkit.org/show_bug.cgi?id=201186
<rdar://problem/53962833>

Reviewed by Youenn Fablet.

Source/WebCore:

Update the Notification API implementation to null-check the page before using. The page becomes null
when using the API in a frame that gets detached from its parent while in the middle of running
script.

Test: http/tests/notifications/request-in-detached-frame.html

Modules/notifications/Notification.cpp: (WebCore::Notification::permission): (WebCore::Notification::requestPermission):

LayoutTests:

Add layout test coverage.

http/tests/notifications/request-in-detached-frame-expected.txt: Added.
http/tests/notifications/request-in-detached-frame.html: Added.
http/tests/notifications/resources/request-in-detached-frame-subframe.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@249156 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10:57 PM Changeset in webkit [249238] by Kocsen Chung

3 edits in branches/safari-608-branch/Source/WebCore

Cherry-pick r249154. rdar://problem/54775050

Disabled devices should not be taken into account when searching for a capture device
https://bugs.webkit.org/show_bug.cgi?id=201183
<rdar://problem/54353440>

Reviewed by Jer Noble.

Manually tested.

platform/mediastream/mac/CoreAudioCaptureDeviceManager.cpp: (WebCore::CoreAudioCaptureDeviceManager::coreAudioDeviceWithUID): We currently keep a list of devices, some of which might be disabled. We should not take into account disabled devices, only enabled devices when doing this search.
platform/mediastream/mac/CoreAudioCaptureSource.cpp: (WebCore::CoreAudioSharedUnit::setupAudioUnit): Improve logging.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@249154 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8:14 PM Changeset in webkit [249237] by rniwa@webkit.org

19 edits in trunk

Make tabIndex IDL attribute reflect its content attribute
https://bugs.webkit.org/show_bug.cgi?id=199606
<rdar://problem/52811448>

Reviewed by Chris Dumez.

LayoutTests/imported/w3c:

web-platform-tests/html/dom/reflection-misc-expected.txt: Rebaselined now that test cases for summary are passing.

Source/WebCore:

This patch makes tabIndex IDL attribute no longer return 0 when the element is focusable
to match the latest HTML5 specification. Instead, the IDL attribute simply reflect the tabindex
content attribute with some elements having 0 as the default tab index (see r248784):
https://html.spec.whatwg.org/multipage/interaction.html#the-tabindex-attribute

The practical implication is that tabIndex IDL attribute on a root editable element (a.k.a.
editing host in HTML5 spec term), body element in design mode, and media elements with
media controls would start returning -1 instead of 0.

Mechanically, this is accomplished by removing the special case in Element::tabIndexForBindings
when supportsFocus returned true. The effect, if any, this patch has on each element which
overrides Element::supportsFocus is summarized as follows (indentation simplies inheritance):

HTMLAnchorElement -> No effect since defaultTabIndex returns 0.

HTMLAreaElement -> Ditto.

HTMLBodyElement -> Changes to return -1 in design mode.
HTMLElement -> Changes to return -1 on a root editable element.
HTMLFormControlElement

HTMLButtonElement -> No effect since defaultTabIndex returns 0.
HTMLFieldSetElement -> No effect since this is an override to use HTMLElement's supportsFocus.
HTMLFormControlElementWithState

HTMLKeygenElement -> No effect since defaultTabIndex returns 0.
HTMLSelectElement -> Ditto.
HTMLTextFormControlElement -> Ditto.

HTMLInputElement -> Ditto.
HTMLTextAreaElement -> Ditto.

HTMLOutputElement -> No effect since this is an override to use HTMLElement's supportsFocus.

HTMLFrameElementBase - No change. Added defaultTabIndex on HTMLIFrameElement and HTMLFrameElement

to returns 0.

HTMLImageElement - No impact since it only affects when an image is set to be editable via SPI.
HTMLMediaElement - Changes to return -1 when media controls is present.
HTMLPlugInElement - applet and embed elements change to return -1 when the plugin is available.
HTMLSummaryElement - No change. Added defaultTabIndex to return 0 when it's active to match

supportsFocus as well as the HTML5 specification.

MathMLElement - No effect since tabIndex IDL attribute does not exist in MathML.
SVGAElement - No effect since defaultTabIndex returns 0.
SVGClipPathElement - No effect since it always returns false.
SVGDefsElement - No effect since it always returns false.

Tests: fast/dom/tabindex-defaults.html

plugins/focus.html

dom/Element.cpp:

(WebCore::Element::tabIndexForBindings const): Made the change.

html/HTMLFrameElement.cpp:

(WebCore::HTMLFrameElement::defaultTabIndex const): Added to preserve the existing behavior.

html/HTMLFrameElement.h:
html/HTMLIFrameElement.cpp:

(WebCore::HTMLIFrameElement::defaultTabIndex const): Ditto.

html/HTMLIFrameElement.h:
html/HTMLObjectElement.cpp:

(WebCore::HTMLObjectElement::defaultTabIndex const): Added. Always return 0 to match the spec.

html/HTMLObjectElement.h:
html/HTMLSummaryElement.cpp:

(WebCore::HTMLSummaryElement::defaultTabIndex const): Added. Return 0 when the this summary
is the active summary element of the details element.

html/HTMLSummaryElement.h:

LayoutTests:

Added test cases and assertions.

fast/dom/tabindex-defaults-expected.txt:
fast/dom/tabindex-defaults.html: Added test cases for iframe, frame, object, video, summary, and SVG elements.

Also blur the active element to avoid any race conditions.

plugins/focus-expected.txt:
plugins/focus.html:
svg/custom/tabindex-order-expected.txt:
svg/custom/tabindex-order.html: Made the sequential navigation code not rely on tabIndex IDL attribute.

8:01 PM Changeset in webkit [249236] by Simon Fraser

3 edits in trunk/Source/WebCore

Make FillLayer::hasImage() inline
https://bugs.webkit.org/show_bug.cgi?id=201265

Reviewed by Zalan Bujtas.

FillLayer::hasImage() shows up on profiles because it's called from hot functions like
isTransparent() and hasMask(), so make a basic inline version that doens't have
to walk the list.

rendering/style/FillLayer.cpp:

(WebCore::FillLayer::hasImageInAnyLayer const):
(WebCore::FillLayer::hasImage const): Deleted.

rendering/style/FillLayer.h:

(WebCore::FillLayer::hasImage const):

7:57 PM Changeset in webkit [249235] by Alan Coon

7 edits in branches/safari-608.2.11.0-branch/Source

Versioning.

7:57 PM Changeset in webkit [249234] by Alan Coon

7 edits in branches/safari-608.2.11.1-branch/Source

Versioning.

7:40 PM Changeset in webkit [249233] by commit-queue@webkit.org

3 edits in trunk/Source/WebCore

REGRESSION: String check: “realtime” Suggesting “real time”
https://bugs.webkit.org/show_bug.cgi?id=201107
<rdar://problem/46372620>

Patch by Peng Liu <Peng Liu> on 2019-08-28
Reviewed by Jer Noble.

Update Localizable.strings.

No new test.

en.lproj/Localizable.strings:
platform/LocalizedStrings.cpp:

(WebCore::localizedMediaControlElementString):

7:24 PM Changeset in webkit [249232] by Simon Fraser

2 edits in trunk/Source/WebCore

PaintFrequencyTracker triggers too many calls to MonotonicTime::now() on layer painting
https://bugs.webkit.org/show_bug.cgi?id=201261

Reviewed by Zalan Bujtas.

MonotonicTime::now() shows up as expensive when painting layer-heavy content, because PaintFrequencyTracker
makes two calls per layer paint.

Halve the number of calls by storing m_lastPaintTime at the start of the paint; doing so doesn't substantially
change the behavior of the tracker.

rendering/PaintFrequencyTracker.h:

(WebCore::PaintFrequencyTracker::begin):
(WebCore::PaintFrequencyTracker::end):

7:10 PM Changeset in webkit [249231] by Kocsen Chung

1 copy in branches/safari-608.2.11.1-branch

New branch.

7:07 PM Changeset in webkit [249230] by timothy_horton@apple.com

4 edits
2 adds in trunk

Reloading a web view with a fixed-width viewport and variable content width restores the previous page scale, shouldn't
https://bugs.webkit.org/show_bug.cgi?id=201256
<rdar://problem/54809509>

Reviewed by Simon Fraser.

Source/WebKit:

WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::restorePageState):
When restoring page state from a history item, if the saved scale was equal to the
initial scale at the time it was saved, ignore the saved scale and use the current
initial scale instead.

Normally this doesn't matter because a given page's initial scale doesn't usually change
between loads, but it totally can! See the test for one example of a way an API client
might cause this; you could also imagine something similar happening if the actual
page content changed.

Tools:

TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
TestWebKitAPI/Tests/WebKit/long-email-viewport.html: Added.
TestWebKitAPI/Tests/WebKitCocoa/ReloadWithDifferingInitialScale.mm: Added.

(TestWebKitAPI::TEST):
Add a test. I left many comments because I had a great deal of trouble
writing this test and wanted to document my findings.

7:06 PM Changeset in webkit [249229] by Kocsen Chung

1 copy in branches/safari-608.2.11.0-branch

New branch.

6:37 PM Changeset in webkit [249228] by Alan Coon

1 copy in tags/Safari-609.1.3

Tag Safari-609.1.3.

5:44 PM Changeset in webkit [249227] by Jonathan Bedard

2 edits in trunk/Tools

results.webkit.org: Do not display branch selector if only one branches available
https://bugs.webkit.org/show_bug.cgi?id=201244

Rubber-stamped by Aakash Jain.

resultsdbpy/resultsdbpy/view/static/js/drawer.js: Hide the branch selector's

container if there aren't multiple branches to choose from.

5:01 PM Changeset in webkit [249226] by Wenson Hsieh

2 edits in trunk/Tools

WKAttachmentTests.DropFolderAsAttachmentAndMoveByDragging fails in some iOS simulator configurations
https://bugs.webkit.org/show_bug.cgi?id=201241
<rdar://problem/54317588>

Reviewed by Tim Horton.

For reasons that are still unknown, it's possible for iOS 13 simulators to get into a state where the IPC
communication delay between the web and UI processes can become extraordinarily long. Under these circumstances,
the drag and drop harness fails to simulate a drop, since it ends up firing all scheduled calls to
-dropInteraction:sessionDidUpdate: before the first response from the web process arrives in the UI process, so
it believes that the web view has rejected the drop from being handled.

Instead, make the drag and drop simulator robust by ensuring a presentation update between drop session updates,
to make sure that the web process has had time to send a response after each update.

TestWebKitAPI/ios/DragAndDropSimulatorIOS.mm:

(-[DragAndDropSimulator _advanceProgress]):

4:11 PM Changeset in webkit [249225] by mark.lam@apple.com

3 edits
1 add in trunk

Placate exception check validation in DFG's operationHasGenericProperty().
https://bugs.webkit.org/show_bug.cgi?id=201245
<rdar://problem/54777512>

Reviewed by Robin Morisset.

JSTests:

stress/missing-exception-check-in-operationHasGenericProperty.js: Added.

Source/JavaScriptCore:

dfg/DFGOperations.cpp:

4:07 PM Changeset in webkit [249224] by Megan Gardner

4 edits in trunk

Null check webFrame when creating a print preview to prevent a crash.
https://bugs.webkit.org/show_bug.cgi?id=201237
<rdar://problem/51618863>

Reviewed by Tim Horton.

Source/WebKit:

Move and expend a null check to keep from crashing when making a print preview.

UIProcess/mac/WKPrintingView.mm:

(-[WKPrintingView _drawPreview:]):
(-[WKPrintingView drawRect:]):

Tools:

Test to verify that if we don't have the WebPageProxy, we will not crash when making a print preview.

TestWebKitAPI/Tests/WebKitCocoa/UIDelegate.mm:

(TEST):

3:52 PM Changeset in webkit [249223] by Ross Kirsling

5 edits in trunk/Source

Unreviewed. Restabilize non-unified build.

Source/JavaScriptCore:

runtime/PropertySlot.h:

Source/WebCore:

inspector/agents/WebConsoleAgent.h:
loader/ResourceLoadObserver.h:

3:32 PM Changeset in webkit [249222] by Simon Fraser

5 edits in trunk/Source/WebCore

Devirtualize RenderBox::visualOverflowRect()
https://bugs.webkit.org/show_bug.cgi?id=201231

Reviewed by Zalan Bujtas.

The only override of RenderBox::visualOverflowRect() was in RenderView, for "paintsEntireContents" views, and as
far as I can tell this is not necessary. visualOverflowRect() is hot when called from RenderLayer::localBoundingBox() --
this shows in profiles when scrolling large patch reviews, so making it non-virtual is a performance enhancement.

RenderLayer::localBoundingBox() can also just call visualOverflowRect(), since that returns borderBoxRect()
when there is no overflow.

rendering/RenderBox.h:

(WebCore::RenderBox::visualOverflowRect const):

rendering/RenderLayer.cpp:

(WebCore::performOverlapTests): Minor optimization to avoid a call to boundingBox().
(WebCore::RenderLayer::calculateClipRects const):

rendering/RenderView.cpp:

(WebCore::RenderView::visualOverflowRect const): Deleted.

rendering/RenderView.h:

2:24 PM Changeset in webkit [249221] by mark.lam@apple.com

4 edits in trunk/Source/JavaScriptCore

Wasm's AirIRGenerator::addLocal() and B3IRGenerator::addLocal() are doing unnecessary overflow checks.
https://bugs.webkit.org/show_bug.cgi?id=201006
<rdar://problem/52053991>

Reviewed by Yusuke Suzuki.

We already ensured that it is not possible to overflow in Wasm::FunctionParser's
parse(). It is unnecessary and misleading to do those overflow checks in
AirIRGenerator and B3IRGenerator. The only check that is necessary is that
m_locals.tryReserveCapacity() is successful, otherwise, we have an out of memory
situation.

This patch changes these unnecessary checks to assertions instead.

wasm/WasmAirIRGenerator.cpp:

(JSC::Wasm::AirIRGenerator::addLocal):

wasm/WasmB3IRGenerator.cpp:

(JSC::Wasm::B3IRGenerator::addLocal):

wasm/WasmValidate.cpp:

(JSC::Wasm::Validate::addLocal):

2:23 PM Changeset in webkit [249220] by Devin Rousso

2 edits in trunk/LayoutTests

Unreviewed, fix test failure after r249173

inspector/timeline/line-column-expected.txt:

1:46 PM Changeset in webkit [249219] by Alan Coon

1 copy in tags/Safari-608.2.11

Tag Safari-608.2.11.

1:34 PM Changeset in webkit [249218] by commit-queue@webkit.org

11 edits in trunk/Source/WebCore

Create ANGLE EGL Context with all extensions disabled by default
https://bugs.webkit.org/show_bug.cgi?id=200900

Patch by Austin Eng <enga@chromium.org> on 2019-08-28
Reviewed by Alex Christensen.

In WebGL, extensions must be explicitly requested before they are enabled.
Fixes the following WebGL conformance tests with the ANGLE backend
LayoutTests/webgl/*/conformance/extensions/ext-blend-minmax.html
LayoutTests/webgl/*/conformance/extensions/ext-frag-depth.html
LayoutTests/webgl/*/conformance/extensions/ext-shader-texture-lod.html
LayoutTests/webgl/*/conformance/extensions/ext-sRGB.html
LayoutTests/webgl/*/conformance/extensions/oes-standard-derivatives.html
LayoutTests/webgl/*/conformance/extensions/oes-texture-float.html
LayoutTests/webgl/*/conformance/extensions/webgl-compressed-texture-s3tc.html
LayoutTests/webgl/*/conformance/glsl/misc/shader-with-dfdx.frag.html
LayoutTests/webgl/*/conformance/glsl/variables/glsl-built-ins.html
LayoutTests/webgl/*/conformance/textures/misc/texture-npot-video.html
LayoutTests/webgl/*/conformance/textures/misc/texture-npot.html

html/canvas/ANGLEInstancedArrays.cpp:

(WebCore::ANGLEInstancedArrays::ANGLEInstancedArrays):
(WebCore::ANGLEInstancedArrays::supported):

html/canvas/WebGLCompressedTextureASTC.cpp:

(WebCore::WebGLCompressedTextureASTC::WebGLCompressedTextureASTC):

html/canvas/WebGLCompressedTextureATC.cpp:

(WebCore::WebGLCompressedTextureATC::WebGLCompressedTextureATC):

html/canvas/WebGLCompressedTexturePVRTC.cpp:

(WebCore::WebGLCompressedTexturePVRTC::WebGLCompressedTexturePVRTC):

html/canvas/WebGLCompressedTextureS3TC.cpp:

(WebCore::WebGLCompressedTextureS3TC::WebGLCompressedTextureS3TC):
(WebCore::WebGLCompressedTextureS3TC::supported):

html/canvas/WebGLDebugShaders.cpp:

(WebCore::WebGLDebugShaders::WebGLDebugShaders):

html/canvas/WebGLDepthTexture.cpp:

(WebCore::WebGLDepthTexture::WebGLDepthTexture):

html/canvas/WebGLDrawBuffers.cpp:

(WebCore::WebGLDrawBuffers::WebGLDrawBuffers):
(WebCore::WebGLDrawBuffers::supported):

platform/graphics/angle/GraphicsContext3DANGLE.cpp:

(WebCore::GraphicsContext3D::reshapeFBOs):
(WebCore::GraphicsContext3D::validateDepthStencil):

platform/graphics/cocoa/GraphicsContext3DCocoa.mm:

(WebCore::GraphicsContext3D::GraphicsContext3D):

1:23 PM Changeset in webkit [249217] by commit-queue@webkit.org

63 edits
1 add in trunk/Source

All image drawing functions should take an argument of type ImagePaintingOptions
https://bugs.webkit.org/show_bug.cgi?id=201059

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2019-08-28
Reviewed by Simon Fraser.

Source/WebCore:

In this patch:
-- All the image drawing function will take an argument of type

ImagePaintingOptions instead of passing individual options.

-- ImagePaintingOptions is made like a set of options. It can be

initialized with any number of options and the order does not matter

-- Image::originalSize() is deleted because it returns size() and none

of the Image concrete classes overrides this implementation.

Headers.cmake:
WebCore.xcodeproj/project.pbxproj:
html/ImageBitmap.cpp:

(WebCore::ImageBitmap::createPromise):
(WebCore::ImageBitmap::createFromBuffer):

html/canvas/CanvasRenderingContext2DBase.cpp:

(WebCore::CanvasRenderingContext2DBase::drawImage):
(WebCore::drawImageToContext):
(WebCore::CanvasRenderingContext2DBase::fullCanvasCompositedDrawImage):

platform/graphics/BitmapImage.cpp:

(WebCore::BitmapImage::draw):
(WebCore::BitmapImage::drawPattern):

platform/graphics/BitmapImage.h:
platform/graphics/CrossfadeGeneratedImage.cpp:

(WebCore::drawCrossfadeSubimage):
(WebCore::CrossfadeGeneratedImage::draw):
(WebCore::CrossfadeGeneratedImage::drawPattern):

platform/graphics/CrossfadeGeneratedImage.h:
platform/graphics/CustomPaintImage.cpp:

(WebCore::CustomPaintImage::draw):
(WebCore::CustomPaintImage::drawPattern):

platform/graphics/CustomPaintImage.h:
platform/graphics/GeneratedImage.h:

(WebCore::GeneratedImage::draw):
(WebCore::GeneratedImage::drawPattern):

platform/graphics/GradientImage.cpp:

(WebCore::GradientImage::draw):
(WebCore::GradientImage::drawPattern):

platform/graphics/GradientImage.h:
platform/graphics/GraphicsContext.cpp:

(WebCore::GraphicsContext::drawImage):
(WebCore::GraphicsContext::drawTiledImage):
(WebCore::GraphicsContext::drawImageBuffer):
(WebCore::GraphicsContext::drawConsumingImageBuffer):

platform/graphics/GraphicsContext.h:

(WebCore::GraphicsContext::drawNativeImage):
(WebCore::GraphicsContext::drawImage):
(WebCore::GraphicsContext::drawTiledImage):
(WebCore::GraphicsContext::drawImageBuffer):
(WebCore::GraphicsContext::drawPattern):
(WebCore::GraphicsContext::drawConsumingImageBuffer):
(WebCore::ImagePaintingOptions::ImagePaintingOptions): Deleted.
(WebCore::ImagePaintingOptions::usesDefaultInterpolation const): Deleted.

platform/graphics/GraphicsContextImpl.cpp:

(WebCore::GraphicsContextImpl::drawImageImpl):
(WebCore::GraphicsContextImpl::drawTiledImageImpl):

platform/graphics/GraphicsContextImpl.h:
platform/graphics/Image.cpp:

(WebCore::Image::drawPattern):
(WebCore::Image::drawTiled):
(WebCore::Image::computeIntrinsicDimensions):

platform/graphics/Image.h:

(WebCore::Image::drawPattern):
(WebCore::Image::draw):
(WebCore::Image::drawTiled):
(WebCore::Image::originalSize const): Deleted.

platform/graphics/ImageBuffer.h:

(WebCore::ImageBuffer::draw):
(WebCore::ImageBuffer::drawPattern):
(WebCore::ImageBuffer::drawConsuming):

platform/graphics/ImagePaintingOptions.h: Added.

(WebCore::ImagePaintingOptions::ImagePaintingOptions):
(WebCore::ImagePaintingOptions::compositeOperator const):
(WebCore::ImagePaintingOptions::blendMode const):
(WebCore::ImagePaintingOptions::decodingMode const):
(WebCore::ImagePaintingOptions::orientation const):
(WebCore::ImagePaintingOptions::interpolationQuality const):
(WebCore::ImagePaintingOptions::setOption):

platform/graphics/NamedImageGeneratedImage.cpp:

(WebCore::NamedImageGeneratedImage::draw):
(WebCore::NamedImageGeneratedImage::drawPattern):

platform/graphics/NamedImageGeneratedImage.h:
platform/graphics/NativeImage.h:
platform/graphics/cairo/CairoOperations.cpp:

(WebCore::Cairo::drawShadowLayerBuffer):
(WebCore::Cairo::drawShadowImage):
(WebCore::Cairo::drawNativeImage):
(WebCore::Cairo::drawPattern):

platform/graphics/cairo/CairoOperations.h:
platform/graphics/cairo/GraphicsContextCairo.cpp:

(WebCore::GraphicsContext::drawNativeImage):
(WebCore::GraphicsContext::drawPattern):

platform/graphics/cairo/GraphicsContextImplCairo.cpp:

(WebCore::GraphicsContextImplCairo::drawNativeImage):
(WebCore::GraphicsContextImplCairo::drawPattern):

platform/graphics/cairo/GraphicsContextImplCairo.h:
platform/graphics/cairo/ImageBufferCairo.cpp:

(WebCore::ImageBuffer::drawConsuming):
(WebCore::ImageBuffer::draw):
(WebCore::ImageBuffer::drawPattern):

platform/graphics/cairo/NativeImageCairo.cpp:

(WebCore::drawNativeImage):

platform/graphics/cg/GraphicsContext3DCG.cpp:

(WebCore::GraphicsContext3D::paintToCanvas):

platform/graphics/cg/GraphicsContextCG.cpp:

(WebCore::GraphicsContext::drawNativeImage):
(WebCore::GraphicsContext::drawPattern):

platform/graphics/cg/ImageBufferCG.cpp:

(WebCore::ImageBuffer::drawConsuming):
(WebCore::ImageBuffer::draw):
(WebCore::ImageBuffer::drawPattern):

platform/graphics/cg/NativeImageCG.cpp:

(WebCore::drawNativeImage):

platform/graphics/cg/PDFDocumentImage.cpp:

(WebCore::PDFDocumentImage::draw):

platform/graphics/cg/PDFDocumentImage.h:
platform/graphics/displaylists/DisplayListItems.cpp:

(WebCore::DisplayList::DrawNativeImage::DrawNativeImage):
(WebCore::DisplayList::DrawNativeImage::apply const):
(WebCore::DisplayList::DrawPattern::DrawPattern):
(WebCore::DisplayList::DrawPattern::apply const):

platform/graphics/displaylists/DisplayListItems.h:

(WebCore::DisplayList::DrawNativeImage::create):
(WebCore::DisplayList::DrawPattern::create):
(WebCore::DisplayList::DrawPattern::DrawPattern):

platform/graphics/displaylists/DisplayListRecorder.cpp:

(WebCore::DisplayList::Recorder::drawNativeImage):
(WebCore::DisplayList::Recorder::drawPattern):

platform/graphics/displaylists/DisplayListRecorder.h:
platform/graphics/filters/FEBlend.cpp:

(WebCore::FEBlend::platformApplySoftware):

platform/graphics/filters/FEComposite.cpp:

(WebCore::FEComposite::platformApplySoftware):

platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:

(WebCore::MediaPlayerPrivateGStreamerBase::paint):

platform/graphics/nicosia/cairo/NicosiaCairoOperationRecorder.cpp:

(Nicosia::CairoOperationRecorder::drawNativeImage):
(Nicosia::CairoOperationRecorder::drawPattern):

platform/graphics/nicosia/cairo/NicosiaCairoOperationRecorder.h:
platform/graphics/win/Direct2DOperations.cpp:

(WebCore::Direct2D::drawNativeImage):
(WebCore::Direct2D::drawPattern):

platform/graphics/win/Direct2DOperations.h:
platform/graphics/win/GraphicsContextImplDirect2D.h:
platform/graphics/win/ImageCGWin.cpp:

(WebCore::BitmapImage::getHBITMAPOfSize):
(WebCore::BitmapImage::drawFrameMatchingSourceSize):

platform/graphics/win/ImageCairoWin.cpp:

(WebCore::BitmapImage::getHBITMAPOfSize):
(WebCore::BitmapImage::drawFrameMatchingSourceSize):

platform/graphics/win/NativeImageDirect2D.cpp:

(WebCore::drawNativeImage):

platform/ios/DragImageIOS.mm:

(WebCore::createDragImageFromImage):

rendering/RenderBoxModelObject.cpp:

(WebCore::RenderBoxModelObject::paintFillLayerExtended):

rendering/RenderEmbeddedObject.cpp:

(WebCore::RenderEmbeddedObject::paintSnapshotImage):

rendering/RenderImage.cpp:

(WebCore::RenderImage::paintReplaced):
(WebCore::RenderImage::paintIntoRect):

rendering/RenderSnapshottedPlugIn.cpp:

(WebCore::RenderSnapshottedPlugIn::paintSnapshot):

rendering/RenderThemeMac.mm:

(WebCore::RenderThemeMac::paintSnapshottedPluginOverlay):

svg/graphics/SVGImage.cpp:

(WebCore::SVGImage::drawForContainer):
(WebCore::SVGImage::nativeImageForCurrentFrame):
(WebCore::SVGImage::nativeImage):
(WebCore::SVGImage::drawPatternForContainer):
(WebCore::SVGImage::draw):

svg/graphics/SVGImage.h:
svg/graphics/SVGImageForContainer.cpp:

(WebCore::SVGImageForContainer::draw):
(WebCore::SVGImageForContainer::drawPattern):

svg/graphics/SVGImageForContainer.h:

Source/WebKit:

GraphicsContext::drawImage() now takes an ImagePaintingOptions.

WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::imagePositionInformation):

12:57 PM Changeset in webkit [249216] by commit-queue@webkit.org

3 edits
2 adds in trunk

XLinkNames namespace is required before the 'href' attribute of SVG animate elements
https://bugs.webkit.org/show_bug.cgi?id=201227

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2019-08-28
Reviewed by Ryosuke Niwa.

Source/WebCore:

To get the animation target element, get the value of the "href" attribute
or the "xlink:href" attribute.

Tests: svg/custom/href-svg-namespace-animate-target.svg

svg/animation/SVGSMILElement.cpp:

(WebCore::SVGSMILElement::buildPendingResource):

LayoutTests:

svg/custom/href-svg-namespace-animate-target-expected.svg: Added.
svg/custom/href-svg-namespace-animate-target.svg: Added.

12:51 PM Changeset in webkit [249215] by Truitt Savell

4 edits in trunk/Source/WebCore

Unreviewed, rolling out r249209.

Broke 19 webgpu/ tests

Reverted changeset:

"[WHLSL] Inline typedef'd types during Metal code generation
to simplify generated code while also making it easier to
read"
https://bugs.webkit.org/show_bug.cgi?id=201185
https://trac.webkit.org/changeset/249209

12:46 PM Changeset in webkit [249214] by mmaxfield@apple.com

3 edits
2 adds in trunk

[WHLSL] Matrices need to have correct alignment
https://bugs.webkit.org/show_bug.cgi?id=201212

Reviewed by Robin Morisset.

Source/WebCore:

Matrices have particular alignment requirements and size requirements.

Type | Alignment | Size

float2x2 | 8 | 16
float2x3 | 16 | 32
float2x4 | 16 | 32
float3x2 | 8 | 24
float3x3 | 16 | 48
float3x4 | 16 | 48
float4x2 | 8 | 32
float4x3 | 16 | 64
float4x4 | 16 | 64

These are important because they may be a member of a struct, and we don't want to misplace
every successive item in the struct.

Test: webgpu/whlsl/matrix-alignment.html

Modules/webgpu/WHLSL/Metal/WHLSLNativeTypeWriter.cpp:

(WebCore::WHLSL::Metal::writeNativeType):

LayoutTests:

Test the alignment and size of float matrices.

Intentionally don't test bool matrices, because they can't be placed in buffers,
meaning their size and alignment is unobservable.

webgpu/whlsl/matrix-alignment-expected.txt: Added.
webgpu/whlsl/matrix-alignment.html: Added.

12:36 PM Changeset in webkit [249213] by Jonathan Bedard

10 edits
1 delete in trunk/Tools

results.webkit.org: Move drawer to the right, open by default
https://bugs.webkit.org/show_bug.cgi?id=200977

Rubber-stamped by Aakash Jain.

The drawer was both unpopular and not discoverable. After feedback from bot watchers,
moving the drawer to the right side of the screen, opening it by default and have it displace
instead of hide the main content.

resultsdbpy/resultsdbpy/view/commit_view_unittest.py:

(CommitViewUnittest.test_drawer): Support new drawer style.
(CommitViewUnittest.test_range_slider): Ditto.
(CommitViewUnittest.test_one_line_switch): Ditto.
(CommitViewUnittest.test_branch_selection): Ditto.

resultsdbpy/resultsdbpy/view/static/css/drawer.css: Removed.
resultsdbpy/resultsdbpy/view/static/js/drawer.js:

(setEnableRecursive): Disable all elements underneath this one
(Drawer): Make drawer a sidebar.

resultsdbpy/resultsdbpy/view/static/library/css/webkit.css:

(.header>.title, .topbar>.title): Overflow of titles should be hidden.
(:root): Move boldInverseColor into webkit.css.
(@media (prefers-color-scheme: dark)):
(.sidebar): Sidebars on mobile should behave like drawers.
(.sidebar.hidden): Add concept of hidden sidebar.
(.mobile-sidebar-control): Add a control to collapse sidebar that only exists on mobile.
(.mobile-sidebar-control.display):
(.main.under-topbar-with-actions):
(@media screen and (min-width: 600px) and (orientation: landscape)):
(.sidebar.left.hidden):
(.sidebar.right.hidden):
(.main.right.hidden):
(.main.left.hidden):
(@media screen and (min-width: 768px) and (orientation: landscape)):
(.sidebar.right):
(a.disabled): Add ability to disable a link.
(.desktop-control): Add a control which only exists on desktop.

resultsdbpy/resultsdbpy/view/templates/base.html: Add hamburger drawer-button for mobile.
resultsdbpy/resultsdbpy/view/templates/commits.html: Use main since Drawer is now a sidebar.
resultsdbpy/resultsdbpy/view/templates/documentation.html: Ditto.
resultsdbpy/resultsdbpy/view/templates/search.html: Ditto.
resultsdbpy/resultsdbpy/view/templates/suite_results.html: Ditto.
resultsdbpy/resultsdbpy/view/view_routes_unittest.py:

(WebSiteTestCase.toggle_drawer): Support new drawer style.
(WebSiteTestCase.find_input_with_name): Ditto.

12:34 PM Changeset in webkit [249212] by commit-queue@webkit.org

14 edits
1 add in trunk

Implement HTMLOrForeignElement
https://bugs.webkit.org/show_bug.cgi?id=201219

Patch by Rob Buis <rbuis@igalia.com> on 2019-08-28
Reviewed by Ryosuke Niwa.

Source/WebCore:

Add the HTMLOrForeignElement [1] interface to share properties
and methods between HTML, SVG and MathML.

[1] https://github.com/mathml-refresh/mathml/issues/83

CMakeLists.txt:
DerivedSources-input.xcfilelist:
DerivedSources-output.xcfilelist:
DerivedSources.make:
Sources.txt:
WebCore.xcodeproj/project.pbxproj:
dom/StaticRange.h:
html/HTMLElement.idl:
html/HTMLOrForeignElement.idl: Added.
svg/SVGElement.idl:

LayoutTests:

Adjust test expectations to new HTMLOrForeignElement interface.

js/dom/dom-static-property-for-in-iteration-expected.txt:
platform/mac-wk2/js/dom/dom-static-property-for-in-iteration-expected.txt:
platform/mac/inspector/model/remote-object-dom-expected.txt:

12:19 PM Changeset in webkit [249211] by dino@apple.com

2 edits in trunk/Source/WebKitLegacy/mac

GenerateTAPI WebKitLegacy.tbd fails on internal iOS Simulator builds
https://bugs.webkit.org/show_bug.cgi?id=201200

Reviewed by Simon Fraser.

We want to include Foundation/NSURLDownload.h if we're on
a newer iOS.

Misc/WebDownload.h:

11:37 AM Changeset in webkit [249210] by Alan Coon

7 edits in branches/safari-608-branch/Source

Versioning.

11:11 AM Changeset in webkit [249209] by weinig@apple.com

4 edits in trunk/Source/WebCore

[WHLSL] Inline typedef'd types during Metal code generation to simplify generated code while also making it easier to read
https://bugs.webkit.org/show_bug.cgi?id=201185

Reviewed by Saam Barati.

The current Metal code generation generates many unnecessary typedefs during TypeNamer::emitMetalTypeDefinitions
such as 'typedef float2 type19;' that can be removed by just using resolved type whereever the typedef would have
been used. The only types that actually need to be predefined are the types that declare a new struct or enum, which
means we can stop emitting for AST::TypeReference, AST::Pointer, AST::Array, and AST::TypeDefinition. Instead, the
relevent mangledNameForType constructs the computed name on the fly.

This is a 25% speedup in the metal code generation phase.

Modules/webgpu/WHLSL/Metal/WHLSLFunctionWriter.cpp:

(WebCore::WHLSL::Metal::FunctionDefinitionWriter::visit):
Switch to using auto rather than Mangled*Name when it is obvious and the actual return type
may need to become more complicated (e.g. MangledOrNativeTypeName).

Modules/webgpu/WHLSL/Metal/WHLSLTypeNamer.h:
Modules/webgpu/WHLSL/Metal/WHLSLTypeNamer.cpp:

(WebCore::WHLSL::Metal::TypeNamer::visit):
There is no need to generate typenames for AST::TypeDefinitions if they are never used.

(WebCore::WHLSL::Metal::TypeNamer::generateUniquedTypeName):
The only UnnamedType that needs a generated typename now is AST::ArrayReferenceType, which emits a struct interface.

(WebCore::WHLSL::Metal::TypeNamer::emitUnnamedTypeDefinition):
Stop generating type definitions for AST::TypeReference, AST::Pointer, AST::Array.

(WebCore::WHLSL::Metal::TypeNamer::emitNamedTypeDefinition):
Stop generating type definitions for AST::TypeDefinition.

(WebCore::WHLSL::Metal::TypeNamer::emitMetalTypeDefinitions):
Update for rename from m_unnamedTypeMapping to m_arrayReferenceTypeMapping.

(WebCore::WHLSL::Metal::TypeNamer::mangledNameForType):
Look into the types and generate a concrete type name when necessary. Lazily cache the constructed
names for AST::ArrayType and AST::Pointer.

10:40 AM Changeset in webkit [249208] by Keith Rollin

13 edits in trunk

Remove support for macOS < 10.13 (part 2)
https://bugs.webkit.org/show_bug.cgi?id=201197
<rdar://problem/54759985>

Source/JavaScriptCore:

Update conditionals that reference WK_MACOS_1013 and suffixes like
_MACOS_SINCE_1013, assuming that we're always building on 10.13 or
later and that these conditionals are always True or False.