Timeline
Jun 17, 2016:
- 10:53 PM Changeset in webkit [202197] by
-
- 21 edits8 adds1 delete in trunk
:indeterminate pseudo-class should match radios whose group has no checked radio
https://bugs.webkit.org/show_bug.cgi?id=156270
Reviewed by Simon Fraser.
LayoutTests/imported/w3c:
- web-platform-tests/html/semantics/selectors/pseudo-classes/indeterminate-expected.txt:
One more pass on official tests :)
Source/WebCore:
The pseudo-class ":indeterminate" is supposed to match radio buttons
for which the entire group has no checked button.
Spec: https://html.spec.whatwg.org/#pseudo-classes:selector-indeterminate
The change is straightforward with one non-obvious choice:
I added matchesIndeterminatePseudoClass() in addition to shouldAppearIndeterminate().
The reason is shouldAppearIndeterminate() is used for styling and AX of elements
with an indeterminate states (check boxes and progress element). There is no such
UI for radio boxes.
I could have extended shouldAppearIndeterminate() to radio box
then filter out this case in RenderTheme. The problem is doing that would also requires
changes to the repaint logic to match :indeterminate. It seemed overkill to me to
change repaint() for a case that is never used in practice.
Tests: fast/css/pseudo-indeterminate-radio-buttons-basics.html
fast/css/pseudo-indeterminate-with-radio-buttons-style-invalidation.html
fast/selectors/detached-radio-button-checked-and-indeterminate-states.html
fast/selectors/pseudo-indeterminate-with-radio-buttons-style-update.html
- css/SelectorCheckerTestFunctions.h:
(WebCore::shouldAppearIndeterminate):
- dom/Element.cpp:
(WebCore::Element::matchesIndeterminatePseudoClass):
- dom/Element.h:
- dom/RadioButtonGroups.cpp:
(WebCore::RadioButtonGroup::setCheckedButton):
(WebCore::RadioButtonGroup::updateCheckedState):
(WebCore::RadioButtonGroup::remove):
(WebCore::RadioButtonGroup::setNeedsStyleRecalcForAllButtons):
(WebCore::RadioButtonGroups::hasCheckedButton):
- dom/RadioButtonGroups.h:
- html/CheckboxInputType.cpp:
(WebCore::CheckboxInputType::matchesIndeterminatePseudoClass):
(WebCore::CheckboxInputType::shouldAppearIndeterminate):
(WebCore::CheckboxInputType::supportsIndeterminateAppearance): Deleted.
- html/CheckboxInputType.h:
- html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::setChecked):
(WebCore::HTMLInputElement::matchesIndeterminatePseudoClass):
(WebCore::HTMLInputElement::shouldAppearIndeterminate):
(WebCore::HTMLInputElement::radioButtonGroups):
- html/HTMLInputElement.h:
- html/InputType.cpp:
(WebCore::InputType::matchesIndeterminatePseudoClass):
(WebCore::InputType::shouldAppearIndeterminate):
(WebCore::InputType::supportsIndeterminateAppearance): Deleted.
- html/InputType.h:
- html/RadioInputType.cpp:
(WebCore::RadioInputType::matchesIndeterminatePseudoClass):
(WebCore::RadioInputType::willDispatchClick): Deleted.
(WebCore::RadioInputType::didDispatchClick): Deleted.
(WebCore::RadioInputType::supportsIndeterminateAppearance): Deleted.
The iOS specific code is just plain wrong.
It was changing the indeterminate state of the input element.
The spec clearly says that state is only used by checkbox:
https://html.spec.whatwg.org/#dom-input-indeterminate
Moreover, the style update would not change the indeterminate state
of other buttons in the Button Group, which is just bizarre.
RenderThemeIOS does not make use of any of this with the current style.
- html/RadioInputType.h:
- style/StyleSharingResolver.cpp:
(WebCore::Style::SharingResolver::canShareStyleWithElement):
(WebCore::Style::canShareStyleWithControl): Deleted.
(WebCore::Style::SharingResolver::sharingCandidateHasIdenticalStyleAffectingAttributes): Deleted.
Style sharing is unified behind the selector matching which is neat.
LayoutTests:
There are two important aspect to cover for this change:
1) The style is updated correctly when a Button Group composition change.
2) When the checkness changes for a Button Group, all its elements
are invalidated to match :indeterminate.
- fast/forms/radio/indeterminate-radio.html:
This test was verifying that the property "indeterminate" of the input element
is not reflected to the style through :indeterminate.
I updated the test to still verify that except that we now match :indeterminate
before changing the property.
- fast/css/pseudo-indeterminate-radio-buttons-basics-expected.html: Added.
- fast/css/pseudo-indeterminate-radio-buttons-basics.html: Added.
- fast/css/pseudo-indeterminate-with-radio-buttons-style-invalidation-expected.txt: Added.
- fast/css/pseudo-indeterminate-with-radio-buttons-style-invalidation.html: Added.
Verify that we don't invalidate everything when the checked button changes.
We only need to invalidate everything if the checked state of the whole group changes.
- fast/selectors/detached-radio-button-checked-and-indeterminate-states-expected.txt: Added.
- fast/selectors/detached-radio-button-checked-and-indeterminate-states.html: Added.
- fast/selectors/pseudo-indeterminate-with-radio-buttons-style-update-expected.txt: Added.
- fast/selectors/pseudo-indeterminate-with-radio-buttons-style-update.html: Added.
- 8:33 PM Changeset in webkit [202196] by
-
- 21 edits2 deletes in trunk
Unreviewed, rolling out r202152.
https://bugs.webkit.org/show_bug.cgi?id=158897
The new test is very unstable, timing out frequently
(Requested by ap on #webkit).
Reverted changeset:
"Web Inspector: console.profile should use the new Sampling
Profiler"
https://bugs.webkit.org/show_bug.cgi?id=153499
http://trac.webkit.org/changeset/202152
- 8:29 PM Changeset in webkit [202195] by
-
- 6 edits2 deletes in trunk
Unreviewed, rolling out r202068, r202115, and r202128.
https://bugs.webkit.org/show_bug.cgi?id=158896
The new test is very unstable, timing out frequently
(Requested by ap on #webkit).
Reverted changesets:
"decompose4 return value is unchecked, leading to potentially
uninitialized data."
https://bugs.webkit.org/show_bug.cgi?id=158761
http://trac.webkit.org/changeset/202068
"[mac] LayoutTest transforms/undecomposable.html is a flaky
timeout"
https://bugs.webkit.org/show_bug.cgi?id=158816
http://trac.webkit.org/changeset/202115
"[mac] LayoutTest transforms/undecomposable.html is a flaky
timeout"
https://bugs.webkit.org/show_bug.cgi?id=158816
http://trac.webkit.org/changeset/202128
- 6:19 PM Changeset in webkit [202194] by
-
- 4 edits2 adds in trunk
AX: HTML indeterminate IDL attribute not mapped to checkbox value=2 for native checkboxes
https://bugs.webkit.org/show_bug.cgi?id=158876
<rdar://problem/26842619>
Reviewed by Joanmarie Diggs.
Source/WebCore:
The indeterminate state was not being reported for native checkboxes.
Also the isIndeterminate() method was relying on whether the appearance changed, which does not happen on Mac, so that
was not being reported correctly. Changed that to check the actual attribute.
Test: accessibility/checkbox-mixed-value.html
- accessibility/AccessibilityNodeObject.cpp:
(WebCore::AccessibilityNodeObject::isIndeterminate):
(WebCore::AccessibilityNodeObject::isPressed):
(WebCore::AccessibilityNodeObject::checkboxOrRadioValue):
- accessibility/AccessibilityObject.cpp:
(WebCore::AccessibilityObject::checkboxOrRadioValue):
LayoutTests:
- accessibility/checkbox-mixed-value-expected.txt: Added.
- accessibility/checkbox-mixed-value.html: Added.
- 6:09 PM Changeset in webkit [202193] by
-
- 3 edits2 copies in branches/safari-601-branch
Merged r202174. rdar://problem/26701322
- 6:06 PM Changeset in webkit [202192] by
-
- 4 edits2 adds in trunk
REGRESSION (r199819): CrashTracer: [GraphicsContext3D::getInternalFramebufferSize
https://bugs.webkit.org/show_bug.cgi?id=158895
<rdar://problem/26423617>
Reviewed by Zalan Bujtas.
Source/WebCore:
In r199819 we started resetting contexts if the page had too
many. Unfortunately there were entry points in the WebGL context
that didn't check for the validity of the object before trying
to access the lower level objects.
Test: webgl/many-contexts-access-after-loss.html
- html/canvas/WebGLRenderingContextBase.cpp:
(WebCore::WebGLRenderingContextBase::drawingBufferWidth): Return 0 if we're lost.
(WebCore::WebGLRenderingContextBase::drawingBufferHeight): Ditto.
LayoutTests:
New test to access the contexts after loss. Also cleaned
up the JavaScript code in the existing test (and made sure
it ran outside of the test system).
- webgl/many-contexts-access-after-loss-expected.txt: Added.
- webgl/many-contexts-access-after-loss.html: Added.
- webgl/many-contexts.html:
- 6:05 PM Changeset in webkit [202191] by
-
- 3 edits2 copies in branches/safari-601.1.46-branch
Merged r202174. rdar://problem/26701284
- 6:03 PM Changeset in webkit [202190] by
-
- 6 edits2 deletes in trunk
Unreviewed, rolling out r202186.
Broke the Apple Windows, Apple Yosemite, GTK, and WinCairo
builds.
Reverted changeset:
"File scheme should not allow access of a resource on a
different volume."
https://bugs.webkit.org/show_bug.cgi?id=158552
http://trac.webkit.org/changeset/202186
- 6:02 PM Changeset in webkit [202189] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed, rolling out r202187.
202186
Reverted changeset:
"Unreviewed clean-up after r202186."
http://trac.webkit.org/changeset/202187
- 5:25 PM Changeset in webkit [202188] by
-
- 4 edits in trunk/Source
Optimize parseCacheHeader() by using StringView
https://bugs.webkit.org/show_bug.cgi?id=158891
Reviewed by Darin Adler.
Source/WebCore:
Optimize parseCacheHeader() and avoid some temporary String allocations
by using StringView. We now strip the whitespaces in the input string
at the beginning of the function, at the same as as we strip the
control characters. We are then able to leverage StringView in the
rest of the function to get substrings without the need for extra
String allocations.
- platform/network/CacheValidation.cpp:
(WebCore::isControlCharacterOrSpace):
(WebCore::trimToNextSeparator):
(WebCore::parseCacheHeader):
Source/WTF:
Add a StringView::find() overload which takes a CharacterMatchFunction
to match the one on String.
- wtf/text/StringView.h:
(WTF::StringView::find):
- 5:09 PM Changeset in webkit [202187] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed clean-up after r202186.
- platform/FileSystem.cpp:
(WebCore::filesHaveSameVolume): Don't use C-style formatting.
- 5:03 PM Changeset in webkit [202186] by
-
- 6 edits2 adds in trunk
File scheme should not allow access of a resource on a different volume.
https://bugs.webkit.org/show_bug.cgi?id=158552
<rdar://problem/15307582>
Patch by Pranjal Jumde <pjumde@apple.com> on 2016-06-17
Reviewed by Brent Fulgham.
Source/WebCore:
Tests: Tools/TestWebKitAPI/Tests/mac/CrossPartitionFileSchemeAccess.mm
- page/SecurityOrigin.cpp:
(WebCore::SecurityOrigin::canDisplay):
- platform/FileSystem.cpp:
(WebCore::platformFileStat):
(WebCore::filesHaveSameVolume):
Returns true if the files are on the same volume
- platform/FileSystem.h:
Tools:
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
- TestWebKitAPI/Tests/mac/CrossPartitionFileSchemeAccess.html: Added.
- TestWebKitAPI/Tests/mac/CrossPartitionFileSchemeAccess.mm: Added.
(-[CrossPartitionFileSchemeAccessNavigationDelegate webView:didFinishNavigation:]):
When the main page load is complete fetch the contents of the document to check if iframe was loaded
(createPartition):
Create a disk image and load the contents of the file.
(cleanUp):
(TestWebKitAPI::TEST):
- 4:31 PM Changeset in webkit [202185] by
-
- 15 edits in trunk/Source/WebKit2
Support configurable autocapitalization.
https://bugs.webkit.org/show_bug.cgi?id=158860
rdar://problem/26231403
Reviewed by Tim Horton.
The behavior of spellchecker should be configurable to avoid
unwanted autocapitalization.
- UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::copy):
- UIProcess/API/APIPageConfiguration.h:
(API::PageConfiguration::initialCapitalizationEnabled):
(API::PageConfiguration::setInitialCapitalizationEnabled):
- UIProcess/API/C/WKPageConfigurationRef.cpp:
(WKPageConfigurationSetIntialCapitalizationEnabled):
- UIProcess/API/C/WKPageConfigurationRef.h:
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _initializeWithConfiguration:]):
- UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration _initialCapitalizationEnabled]):
(-[WKWebViewConfiguration _setInitialCapitalizationEnabled:]):
- UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
- UIProcess/TextChecker.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy):
(WebKit::WebPageProxy::checkTextOfParagraph):
(WebKit::WebPageProxy::getGuessesForWord):
- UIProcess/WebPageProxy.h:
- UIProcess/efl/TextCheckerEfl.cpp:
(WebKit::TextChecker::checkTextOfParagraph):
(WebKit::TextChecker::getGuessesForWord):
- UIProcess/gtk/TextCheckerGtk.cpp:
(WebKit::TextChecker::getGuessesForWord):
(WebKit::TextChecker::checkTextOfParagraph):
- UIProcess/ios/TextCheckerIOS.mm:
(WebKit::TextChecker::checkTextOfParagraph):
(WebKit::TextChecker::getGuessesForWord):
- UIProcess/mac/TextCheckerMac.mm:
(WebKit::TextChecker::checkTextOfParagraph):
(WebKit::TextChecker::getGuessesForWord):
- 4:27 PM Changeset in webkit [202184] by
-
- 14 edits in trunk/Source/WebKit2
Use WTF::NoncopyableFunction in NetworkDataTaskClient
https://bugs.webkit.org/show_bug.cgi?id=158887
Reviewed by Alex Christensen.
Use WTF::NoncopyableFunction in NetworkDataTaskClient instead of std::function
and consistently move it around. This avoids some unnecessary copying.
- NetworkProcess/Downloads/DownloadManager.cpp:
(WebKit::DownloadManager::willDecidePendingDownloadDestination):
(WebKit::DownloadManager::continueDecidePendingDownloadDestination):
- NetworkProcess/Downloads/DownloadManager.h:
- NetworkProcess/NetworkDataTask.h:
- NetworkProcess/NetworkLoad.cpp:
(WebKit::NetworkLoad::willPerformHTTPRedirection):
(WebKit::NetworkLoad::didReceiveChallenge):
(WebKit::NetworkLoad::didReceiveResponseNetworkSession):
(WebKit::NetworkLoad::continueCanAuthenticateAgainstProtectionSpace):
- NetworkProcess/NetworkLoad.h:
- NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::findPendingDownloadLocation):
- NetworkProcess/NetworkProcess.h:
- NetworkProcess/PingLoad.h:
- NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTask::didReceiveChallenge):
(WebKit::NetworkDataTask::didReceiveResponse):
(WebKit::NetworkDataTask::willPerformHTTPRedirection):
(WebKit::NetworkDataTask::tryPasswordBasedAuthentication):
- NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:didReceiveChallenge:completionHandler:]):
- Shared/Authentication/AuthenticationManager.cpp:
(WebKit::AuthenticationManager::addChallengeToChallengeMap):
(WebKit::AuthenticationManager::coalesceChallengesMatching):
(WebKit::AuthenticationManager::didReceiveAuthenticationChallenge):
(WebKit::AuthenticationManager::tryUseCertificateInfoForChallenge):
(WebKit::AuthenticationManager::useCredentialForSingleChallenge):
- Shared/Authentication/AuthenticationManager.h:
- Shared/Authentication/mac/AuthenticationManager.mac.mm:
(WebKit::AuthenticationManager::tryUseCertificateInfoForChallenge):
- 3:24 PM Changeset in webkit [202183] by
-
- 34 edits1 move2 adds in trunk
Web video playback controls should have RTL volume slider
https://bugs.webkit.org/show_bug.cgi?id=158856
<rdar://problem/25971769>
Patch by Antoine Quint <Antoine Quint> on 2016-06-17
Reviewed by Tim Horton.
Source/WebCore:
We reproduce the system used to propagate the page scale factor from the WebPage to the media controls to
propagate the user interface layout direction.
The Page exposes a new setUserInterfaceLayoutDirection() method which is set by the WebPage. The Page
then notifies the Document of a change, which propagates down to registered media elements, and finally sets
the usesLTRUserInterfaceLayoutDirection property on the media controller object in the injected JavaScript.
Based on the value of that property we toggle a new .uses-ltr-user-interface-layout-direction CSS class on the
.volume-box which applies a translate to the right and flips the volume controls on the x axis.
Since we're setting a new JS property from HTMLMediaController, we refactor much of the code out of the existing
pageScaleFactorChanged() and setPageScaleFactorProperty() into the new setControllerJSProperty() method so that
can easily set a named JS property with a given JSValue.
For testing purposes, we expose the WebCore::Page::setUserInterfaceLayoutDirection() method through Internals.
Test: fullscreen/video-controls-rtl.html
- Modules/mediacontrols/mediaControlsApple.css:
(video:-webkit-full-screen::-webkit-media-controls-panel .volume-box:not(.uses-ltr-user-interface-layout-direction)):
- Modules/mediacontrols/mediaControlsApple.js:
(Controller.prototype.set usesLTRUserInterfaceLayoutDirection):
- WebCore.xcodeproj/project.pbxproj:
- dom/Document.cpp:
(WebCore::Document::registerForUserInterfaceLayoutDirectionChangedCallbacks):
(WebCore::Document::unregisterForUserInterfaceLayoutDirectionChangedCallbacks):
(WebCore::Document::userInterfaceLayoutDirectionChanged):
- dom/Document.h:
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::registerWithDocument):
(WebCore::HTMLMediaElement::unregisterWithDocument):
(WebCore::HTMLMediaElement::updatePageScaleFactorJSProperty):
(WebCore::HTMLMediaElement::updateUsesLTRUserInterfaceLayoutDirectionJSProperty):
(WebCore::HTMLMediaElement::setControllerJSProperty):
(WebCore::HTMLMediaElement::didAddUserAgentShadowRoot):
(WebCore::HTMLMediaElement::pageScaleFactorChanged):
(WebCore::HTMLMediaElement::userInterfaceLayoutDirectionChanged):
(WebCore::setPageScaleFactorProperty): Deleted.
- html/HTMLMediaElement.h:
- page/Page.cpp:
(WebCore::Page::setUserInterfaceLayoutDirection):
- page/Page.h:
(WebCore::Page::userInterfaceLayoutDirection):
- platform/UserInterfaceLayoutDirection.h: Renamed from Source/WebKit2/UIProcess/UserInterfaceLayoutDirection.h.
- testing/Internals.cpp:
(WebCore::Internals::setUserInterfaceLayoutDirection):
- testing/Internals.h:
- testing/Internals.idl:
Source/WebKit2:
We now call setUserInterfaceLayoutDirection() on WebCore::Page to propagate
the user interface layout direction down to the Web content. To facilitate this,
the UserInterfaceLayoutDirection enum is now defined in WebCore.
- Shared/WebPageCreationParameters.h:
- UIProcess/API/Cocoa/WKWebView.mm:
(convertSystemLayoutDirection):
(toUserInterfaceLayoutDirection):
- UIProcess/API/gtk/PageClientImpl.h:
- UIProcess/Cocoa/WebViewImpl.h:
- UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::toUserInterfaceLayoutDirection):
(WebKit::WebViewImpl::userInterfaceLayoutDirection):
- UIProcess/PageClient.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::userInterfaceLayoutDirection):
(WebKit::WebPageProxy::setUserInterfaceLayoutDirection):
- UIProcess/WebPageProxy.h:
- UIProcess/efl/WebView.h:
- UIProcess/ios/PageClientImplIOS.h:
- UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::userInterfaceLayoutDirection):
- UIProcess/mac/PageClientImpl.h:
- UIProcess/mac/PageClientImpl.mm:
(WebKit::PageClientImpl::userInterfaceLayoutDirection):
- UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::PendingSwipeTracker::scrollEventCanBecomeSwipe):
(WebKit::ViewGestureController::isPhysicallySwipingLeft):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::m_userInterfaceLayoutDirection):
(WebKit::WebPage::setUserInterfaceLayoutDirection):
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::performNonEditingBehaviorForSelector):
LayoutTests:
New test that checks that the volume box has the correct CSS class name and the expected
CSS transform value applied to it when the user interface layout direction is right-to-left
and that changing the user interface layout direction dynamically update the media element
shadow tree as expected.
- fullscreen/video-controls-rtl-expected.txt: Added.
- fullscreen/video-controls-rtl.html: Added.
- 2:46 PM Changeset in webkit [202182] by
-
- 2 edits in trunk/LayoutTests
Update LayoutTest fast/canvas/canvas-composite-text-alpha.html
https://bugs.webkit.org/show_bug.cgi?id=158710
Reviewed by Alexey Proskuryakov.
Though this test passes, opening the test file directly reports multiple failures. Updating the
readback locations to get the correct results.
- fast/canvas/canvas-composite-text-alpha.html:
- 2:46 PM Changeset in webkit [202181] by
-
- 2 edits in trunk/LayoutTests
Marking media/video-controls-show-on-kb-or-ax-event.html as flaky on Mac
https://bugs.webkit.org/show_bug.cgi?id=158889
Unreviewed test gardening.
- platform/mac/TestExpectations:
- 2:30 PM Changeset in webkit [202180] by
-
- 2 edits in trunk/LayoutTests
Skip security/contentSecurityPolicy/video-with-file-url-allowed-by-media-src-star.html on Mac
https://bugs.webkit.org/show_bug.cgi?id=155196
Unreviewed test gardening.
- platform/mac/TestExpectations:
- 2:21 PM Changeset in webkit [202179] by
-
- 3 edits2 adds in trunk
REGRESSION: Double tap to zoom does not work on yahoo finance
https://bugs.webkit.org/show_bug.cgi?id=158886
<rdar://problem/24917760>
Reviewed by Simon Fraser.
Source/WebKit2:
We shouldn't enable fast click on pages that have
had to shrink to fit. In other words, we should enable
double-tap-to-zoom if the page has already zoomed.
Test: fast/events/ios/viewport-shrink-to-fit-allows-double-tap.html
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _allowsDoubleTapGestures]):
LayoutTests:
- fast/events/ios/viewport-shrink-to-fit-allows-double-tap-expected.txt: Added.
- fast/events/ios/viewport-shrink-to-fit-allows-double-tap.html: Added.
- 1:51 PM Changeset in webkit [202178] by
-
- 7 edits2 adds in trunk
TouchEvent should have a constructor
https://bugs.webkit.org/show_bug.cgi?id=158883
<rdar://problem/26063585>
Reviewed by Benjamin Poulain.
Source/WebCore:
TouchEvent should have a constructor:
Chrome already ships this:
Test: fast/events/touch/touch-event-constructor.html
- bindings/js/JSDictionary.cpp:
(WebCore::JSDictionary::convertValue):
- bindings/js/JSDictionary.h:
- dom/TouchEvent.cpp:
(WebCore::TouchEvent::TouchEvent):
- dom/TouchEvent.h:
- dom/TouchEvent.idl:
LayoutTests:
Add layout test coverage for the TouchEvent constructor.
- fast/events/touch/touch-event-constructor-expected.txt: Added.
- fast/events/touch/touch-event-constructor.html: Added.
- 1:21 PM Changeset in webkit [202177] by
-
- 3 edits in trunk/Source/WebCore
Potential null dereferencing on a detached positioned renderer.
https://bugs.webkit.org/show_bug.cgi?id=158879
Reviewed by Simon Fraser.
This patch fixes the case when the while loop to search for the absolute positioned ancestor
returns null (it happens when positioned renderer has been detached from the render tree).
Speculative fix.
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::markFixedPositionObjectForLayoutIfNeeded):
- rendering/RenderBlock.h:
- 1:18 PM Changeset in webkit [202176] by
-
- 9 edits in trunk
URL hash setter does not remove fragment identifier if argument is an empty string
https://bugs.webkit.org/show_bug.cgi?id=158869
<rdar://problem/26863430>
Reviewed by Darin Adler.
Source/WebCore:
URL hash setter and URLUtils hash setter should remove the fragment identifier
if set to "#" or "":
- https://url.spec.whatwg.org/#dom-url-hash
- https://html.spec.whatwg.org/multipage/semantics.html#dom-hyperlink-hash
This patch aligns our behavior with the specification and with other browsers
(tested Firefox and Chrome).
This patch also updates HTMLAnchorElement to inherit URLUtils to avoid code
duplication. HTMLAnchorElement already implements URLUtils in the IDL, as per
the specification:
No new tests, rebaselined existing tests.
- html/HTMLAnchorElement.cpp:
(WebCore::HTMLAnchorElement::origin): Deleted.
(WebCore::HTMLAnchorElement::text): Deleted.
(WebCore::HTMLAnchorElement::setText): Deleted.
(WebCore::HTMLAnchorElement::toString): Deleted.
(WebCore::HTMLAnchorElement::isLiveLink): Deleted.
(WebCore::HTMLAnchorElement::sendPings): Deleted.
(WebCore::HTMLAnchorElement::handleClick): Deleted.
(WebCore::HTMLAnchorElement::eventType): Deleted.
(WebCore::HTMLAnchorElement::treatLinkAsLiveForEventType): Deleted.
(WebCore::isEnterKeyKeydownEvent): Deleted.
(WebCore::shouldProhibitLinks): Deleted.
(WebCore::HTMLAnchorElement::willRespondToMouseClickEvents): Deleted.
(WebCore::rootEditableElementMap): Deleted.
(WebCore::HTMLAnchorElement::rootEditableElementForSelectionOnMouseDown): Deleted.
(WebCore::HTMLAnchorElement::clearRootEditableElementForSelectionOnMouseDown): Deleted.
(WebCore::HTMLAnchorElement::setRootEditableElementForSelectionOnMouseDown): Deleted.
- html/HTMLAnchorElement.h:
(WebCore::HTMLAnchorElement::invalidateCachedVisitedLinkHash): Deleted.
- html/URLUtils.h:
(WebCore::URLUtils<T>::setHash):
LayoutTests:
Update existing layout tests now that setting hash to "" or "#" clears
the fragment identifier.
- fast/dom/DOMURL/set-href-attribute-hash-expected.txt:
- fast/dom/DOMURL/set-href-attribute-hash.html:
- fast/dom/HTMLAnchorElement/script-tests/set-href-attribute-hash.js:
- fast/dom/HTMLAnchorElement/set-href-attribute-hash-expected.txt:
- 12:51 PM Changeset in webkit [202175] by
-
- 5 edits4 copies in branches/safari-601-branch
Merged r202151. rdar://problem/26701322
- 12:50 PM Changeset in webkit [202174] by
-
- 3 edits2 adds in trunk
Ignore case in the check for security origin inheritance
https://bugs.webkit.org/show_bug.cgi?id=158878
Reviewed by Alex Christensen.
Source/WebCore:
Darin Adler commented in https://bugs.webkit.org/show_bug.cgi?id=158855:
"Are these comparisons intentionally case sensitive? Shouldn’t they ignore ASCII
case? We could use equalIgnoringASCIICase and equalLettersIgnoringASCIICase for
those two lines instead of using ==. URL::parse normalizes letters in the scheme
and host by using toASCIILower, but does not normalize letters elsewhere in the
URL, such as in the "blank" or "srcdoc" in the above URLs."
Test: http/tests/dom/window-open-about-uppercase-blank-and-access-document.html
- platform/URL.cpp:
(WebCore::URL::shouldInheritSecurityOriginFromOwner):
LayoutTests:
- http/tests/dom/window-open-about-uppercase-blank-and-access-document-expected.txt: Added.
- http/tests/dom/window-open-about-uppercase-blank-and-access-document.html: Added.
- 12:22 PM Changeset in webkit [202173] by
-
- 3 edits in trunk/Source/WTF
OOM Assertion failure in JSON.stringify.
https://bugs.webkit.org/show_bug.cgi?id=158794
<rdar://problem/26826254>
Reviewed by Saam Barati.
The bug was actually in StringBuilder::appendQuotedJSONString() where it failed
to detect an imminent unsigned int overflow. The fix is to use Checked<unsigned>
for the needed math, and RELEASE_ASSERT afterwards that we did not overflow.
I also added more assertions to detect sooner if any there are any problems with
StringBuilder's m_buffer or m_length being incorrectly sized. These assertions
have been run on the JSC and layout tests without any issue.
- wtf/text/StringBuilder.cpp:
(WTF::StringBuilder::resize):
(WTF::StringBuilder::allocateBuffer):
(WTF::StringBuilder::allocateBufferUpConvert):
(WTF::StringBuilder::reallocateBuffer<LChar>):
(WTF::StringBuilder::reallocateBuffer<UChar>):
(WTF::StringBuilder::reserveCapacity):
(WTF::StringBuilder::appendUninitializedSlow):
(WTF::StringBuilder::append):
(WTF::StringBuilder::appendQuotedJSONString):
- wtf/text/StringBuilder.h:
(WTF::StringBuilder::swap):
- 12:10 PM Changeset in webkit [202172] by
-
- 2 edits in trunk/Source/WebCore
Fix compilation errors when we enable DUMP_NODE_STATISTICS in Node.h
https://bugs.webkit.org/show_bug.cgi?id=158868
Reviewed by Alex Christensen.
Fix compilation errors in Node.cpp when we enable DUMP_NODE_STATISTICS
- dom/Node.cpp:
(WebCore::Node::dumpStatistics):
- 11:47 AM Changeset in webkit [202171] by
-
- 5 edits4 copies in branches/safari-601.1.46-branch
Merged r202151. rdar://problem/26701284
- 10:24 AM Changeset in webkit [202170] by
-
- 2 edits in trunk/LayoutTests
Skipping two new LayoutTests that rely on mouse events on ios-simulator.
https://bugs.webkit.org/show_bug.cgi?id=158855
Unreviewed test gardening.
- platform/ios-simulator/TestExpectations:
- 10:20 AM Changeset in webkit [202169] by
-
- 2 edits in trunk/Source/WebCore
[Win] Scrolling in popup menu scrolls past last entry.
https://bugs.webkit.org/show_bug.cgi?id=158870
Reviewed by Brent Fulgham.
When the popup has a scrollbar, the content size is not equal to the popup window size.
- platform/win/PopupMenuWin.cpp:
(WebCore::PopupMenuWin::contentsSize):
- 9:54 AM MathML/Early_2016_Refactoring edited by
- (diff)
- 9:54 AM MathML/Early_2016_Refactoring edited by
- (diff)
- 9:53 AM MathML/Early_2016_Refactoring edited by
- (diff)
- 9:30 AM Changeset in webkit [202168] by
-
- 27 edits1 add4 deletes in trunk
Refactor RenderMathMLRoot layout function to avoid using flexbox
https://bugs.webkit.org/show_bug.cgi?id=153987
Patch by Frederic Wang <fwang@igalia.com> on 2016-06-17
Reviewed by Brent Fulgham.
Source/WebCore:
No new tests, already covered by existing tests.
A case for RTL root has been added to roots.xhtml.
We reimplement RenderMathMLRoot without any flexbox or anonymous.
The anonymous RenderMathMLRadicalOperator used to draw the radical sign is replaced with
the MathOperator class introduced in bug 152244.
msqrt (row of children under a square root) is now implemented directly in RenderMathMLRoot,
so RenderMathMLSquareRoot is removed and RenderMathMLRoot now inherits from RenderMathMLRow.
- CMakeLists.txt: Remove files for RenderMathMLRadicalOperator and RenderMathMLSquareRoot.
- WebCore.xcodeproj/project.pbxproj: ditto.
- accessibility/AccessibilityRenderObject.cpp: Update code now that we do not use any
radical wrappers.
(WebCore::AccessibilityRenderObject::isMathRow): Now that RenderMathMLRoot inherits from
RenderMathMLRow, we must exclude MathRoot or otherwise some accessibility code may treat
roots as rows.
(WebCore::AccessibilityRenderObject::mathRadicandObject): Return the first child for
Root/SquareRoot or nullptr.
(WebCore::AccessibilityRenderObject::mathRootIndexObject): Return the second child for
Root and nullptr for SquareRoot.
- mathml/MathMLInlineContainerElement.cpp:
(WebCore::MathMLInlineContainerElement::childrenChanged): We no longer need a special case
for msqrt, it is treated as a normal RenderMathMLRow.
(WebCore::MathMLInlineContainerElement::createElementRenderer): Make msqrt create a
RenderMathMLRoot object.
- rendering/RenderObject.h:
(WebCore::RenderObject::isRenderMathMLRadicalOperator): Deleted.
- rendering/mathml/RenderMathMLBlock.cpp:
(WebCore::RenderMathMLBlock::mirrorIfNeeded): New function to mirror a child horizontal
offset according to the parent width.
(WebCore::RenderMathMLBlock::renderName):
- rendering/mathml/RenderMathMLBlock.h:
(WebCore::RenderMathMLBlock::mirrorIfNeeded): Moved from RenderMathMLScripts, just forward
call to the other mirrorIfNeeded function.
- rendering/mathml/RenderMathMLOperator.cpp: We no longer need this trailingSpaceError hack.
(WebCore::RenderMathMLOperator::trailingSpaceError): Deleted.
- rendering/mathml/RenderMathMLOperator.h: ditto.
- rendering/mathml/RenderMathMLRadicalOperator.cpp: Removed. The radical sign is now drawn
with a MathOperator.
- rendering/mathml/RenderMathMLRadicalOperator.h: Removed.
- rendering/mathml/RenderMathMLRoot.cpp: Complete refactoring to avoid using flexbox and
anonymous wrappers.
(WebCore::RenderMathMLRoot::RenderMathMLRoot): Set m_kind parameters to distinguish between
square root and general root and set the MathOperator member to draw the radical sign.
(WebCore::RenderMathMLRoot::isValid): Helper function to verify whether the child list is valid.
(WebCore::RenderMathMLRoot::getBase): Get the base of an mroot.
(WebCore::RenderMathMLRoot::getIndex): Get the index of an mroot.
(WebCore::RenderMathMLRoot::styleDidChange): Be sure to keep the style of the
MathOperator in sync with ours ; no need to skip empty roots.
(WebCore::RenderMathMLRoot::updateFromElement): Call the function from the new parent class ;
no need to skip empty roots.
(WebCore::RenderMathMLRoot::updateStyle): Remove the isEmpty ASSERT as it is valid to have
empty square root. Set the m_kernBeforeDegree, m_kernBeforeDegree members.
No need to set style for anonymous.
(WebCore::RenderMathMLRoot::computePreferredLogicalWidths): Implement this function.
(WebCore::RenderMathMLRoot::layoutBlock): Implement this function.
(WebCore::RenderMathMLRoot::paintChildren): Implement this function.
(WebCore::RenderMathMLRoot::paint): Remove the trailingSpaceError hack ;
paint the radical sign via MathOperator::paint
(WebCore::RenderMathMLRoot::baseWrapper): Deleted.
(WebCore::RenderMathMLRoot::radicalWrapper): Deleted.
(WebCore::RenderMathMLRoot::indexWrapper): Deleted.
(WebCore::RenderMathMLRoot::radicalOperator): Deleted.
(WebCore::RenderMathMLRoot::restructureWrappers): Deleted.
(WebCore::RenderMathMLRoot::addChild): Deleted.
(WebCore::RenderMathMLRoot::firstLineBaseline): Deleted.
(WebCore::RenderMathMLRoot::layout): Deleted.
(WebCore::RenderMathMLRootWrapper::createAnonymousWrapper): Deleted.
(WebCore::RenderMathMLRootWrapper::removeChildWithoutRestructuring): Deleted.
(WebCore::RenderMathMLRootWrapper::removeChild): Deleted.
- rendering/mathml/RenderMathMLRoot.h: Make RenderMathMLRoot inherit from RenderMathMLRow.
Make RenderMathMLRoot support <msqrt>.
Remove all the anonymous wrapper stuff and instead use a MathOperator for the radical symbol.
Update function declaration to implement layout without flexbox and add some helper functions.
- rendering/mathml/RenderMathMLRow.cpp: Allow to get the exact metrics of the chid row,
for use in RenderMathMLRoot.
(WebCore::RenderMathMLRow::computeLineVerticalStretch): rename parameters.
(WebCore::RenderMathMLRow::layoutRowItems): Set parameters to the final ascent, descent and
logical width of the chid row. Set the temporary logical width for RenderMathRoot before
laying the children out.
(WebCore::RenderMathMLRow::layoutBlock): Rename parameters ; add a dummy logicalWidth
parameter.
- rendering/mathml/RenderMathMLRow.h: Make some functions accessible or overridable by
RenderMathMLRoot. Make layoutRowItems return the final ascent, descent and logical width
after the chid row is laid out.
- rendering/mathml/RenderMathMLScripts.cpp: Move mirrorIfNeeded to RenderMathMLBlock.
(WebCore::RenderMathMLScripts::mirrorIfNeeded): Deleted.
- rendering/mathml/RenderMathMLScripts.h: Move mirrorIfNeeded to RenderMathMLBlock.
- rendering/mathml/RenderMathMLSquareRoot.cpp: Removed.
- rendering/mathml/RenderMathMLSquareRoot.h: Removed.
- rendering/mathml/MathOperator.cpp:
(WebCore::MathOperator::paint): Apply a mirroring scale transform to radical symbol
in RTL direction.
LayoutTests:
Update expectations for some MathML tests containing radical operators.
- TestExpectations: mathml/roots-removeChild.html works as expected now
that we do not manage anonymous wrappers anymore.
- mathml/presentation/roots.xhtml: Add one test for RTL roots.
- platform/gtk/mathml/presentation/mo-stretch-expected.txt: Update expectation to
take into change in the render tree.
- platform/gtk/mathml/presentation/roots-expected.txt: Ditto.
- platform/gtk/mathml/presentation/roots-expected.png: Update expectation to take
into account small rendering changes.
- platform/ios-simulator/mathml/presentation/roots-expected.txt: Ditto.
- platform/gtk/mathml/radical-fallback-expected.txt: Ditto.
- platform/gtk/mathml/radical-fallback-expected.png: Ditto.
- platform/mac/mathml/radical-fallback-expected.txt: Ditto.
- platform/mac/mathml/radical-fallback-expected.png: Ditto.
- platform/ios-simulator/mathml/radical-fallback-expected.txt: Ditto.
- platform/ios-simulator/mathml/radical-fallback-expected.png: Added.
- 8:43 AM Changeset in webkit [202167] by
-
- 33 edits1 add in trunk/Source/WebCore
Drop some unnecessary header includes
https://bugs.webkit.org/show_bug.cgi?id=158864
Reviewed by Alexey Proskuryakov.
Drop some unnecessary header includes to try and reduce build times.
- WebCore.xcodeproj/project.pbxproj:
- accessibility/AccessibilityList.cpp:
- css/CSSComputedStyleDeclaration.cpp:
- css/MediaQueryMatcher.cpp:
- css/StyleMedia.cpp:
- css/TransformFunctions.cpp:
- dom/NodeRenderStyle.h:
- dom/PseudoElement.h:
(isType): Deleted.
- html/HTMLTitleElement.cpp:
- html/shadow/MediaControlElementTypes.h:
- html/shadow/MediaControls.cpp:
- inspector/InspectorDOMAgent.h:
- inspector/InspectorLayerTreeAgent.h:
- inspector/InspectorPageAgent.cpp:
- page/scrolling/AsyncScrollingCoordinator.cpp:
- page/scrolling/ScrollingCoordinator.h:
- rendering/BidiRun.h:
- rendering/BorderEdge.h:
- rendering/RenderElement.h:
- rendering/RenderObject.h:
(WebCore::AnnotatedRegionValue::operator==): Deleted.
(WebCore::AnnotatedRegionValue::operator!=): Deleted.
- rendering/RenderObjectEnums.h: Added.
- rendering/RenderTheme.h:
- rendering/SimpleLineLayoutFlowContents.h:
- rendering/SimpleLineLayoutTextFragmentIterator.h:
- rendering/TextPainter.h:
- rendering/style/RenderStyle.h:
(WebCore::pseudoElementRendererIsNeeded):
- rendering/style/ShapeValue.cpp:
- rendering/style/ShapeValue.h:
- style/ClassChangeInvalidation.cpp:
- style/ClassChangeInvalidation.h:
- style/InlineTextBoxStyle.h:
- style/StyleUpdate.cpp:
- 6:35 AM Changeset in webkit [202166] by
-
- 4 edits in trunk/Source/WebCore
[iOS] Throw away linked code when navigating to a new page.
<https://webkit.org/b/153851>
Reviewed by Antti Koivisto.
When navigating to a new page, tell JSC to throw out any linked code it has lying around.
Linked code is tied to a specific global object, and as we're creating a new one for the
new page, none of it is useful to us here.
In the event that the user navigates back, the cost of relinking some code will be far
lower than the memory cost of keeping all of it around.
This was in-tree before but was rolled out due to regressing JSBench. It was a slowdown
due to the benchmark harness using top-level navigations to drive the tests.
This new version avoids that problem by only throwing out code if we haven't navigated
in the last 2 seconds. This also prevents excessive work in response to redirects.
I've also moved this into MemoryPressureHandler so we don't make a mess in FrameLoader.
- loader/FrameLoader.cpp:
(WebCore::FrameLoader::commitProvisionalLoad):
- platform/MemoryPressureHandler.cpp:
(WebCore::MemoryPressureHandler::jettisonExpensiveObjectsOnTopLevelNavigation):
- platform/MemoryPressureHandler.h:
- 4:50 AM Changeset in webkit [202165] by
-
- 2 edits in trunk/Tools
Unreviewed. Added myself to the list of committers.
- Scripts/webkitpy/common/config/contributors.json:
- 4:46 AM Changeset in webkit [202164] by
-
- 2 edits3 deletes in trunk/LayoutTests
Update test-expected file for failing fullscreen test
https://bugs.webkit.org/show_bug.cgi?id=158838
Patch by Romain Bellessort <romain.bellessort@crf.canon.fr> on 2016-06-17
Reviewed by Carlos Garcia Campos.
Used MAC version of the test as the default version since only MAC and
GTK do not skip this test and expect the same result. Removed
test-expected files specific to each port when there was one. Prior to
this patch, GTK test was failing unexpectedly.
- fullscreen/video-controls-override-expected.txt:
- platform/efl/fullscreen/video-controls-override-expected.txt: Removed.
- platform/gtk/fullscreen/video-controls-override-expected.txt: Removed.
- platform/mac/fullscreen/video-controls-override-expected.txt: Removed.
- 4:19 AM Changeset in webkit [202163] by
-
- 2 edits in trunk/Source/WebKit/win
[Win] Rendering is not scaled correctly if process is DPI aware.
https://bugs.webkit.org/show_bug.cgi?id=158781
Reviewed by Brent Fulgham.
Scale the root graphics layer with the device scale factor.
- WebView.cpp:
(WebView::setAcceleratedCompositing):
- 3:34 AM Changeset in webkit [202162] by
-
- 13 edits in trunk
CORS preflight with a non-200 response should be a preflight failure
https://bugs.webkit.org/show_bug.cgi?id=111008
Reviewed by Darin Adler.
LayoutTests/imported/w3c:
- web-platform-tests/XMLHttpRequest/data-uri-expected.txt:
Source/WebCore:
Covered by rebased tests.
- Modules/fetch/FetchResponse.h: Making use of ResourceResponse::isSuccessful.
- loader/CrossOriginPreflightChecker.cpp:
(WebCore::CrossOriginPreflightChecker::validatePreflightResponse): Checking that response status is code is
successful. If not, calling preflight failure callback.
(WebCore::CrossOriginPreflightChecker::startPreflight): Putting in manual redirection mode so that redirection
responses are processed as other responses.
- loader/ResourceLoaderOptions.h:
(WebCore::ResourceLoaderOptions::fetchOptions): Adding a non-const getter and fixing const getter to return a
const reference.
(WebCore::ResourceLoaderOptions::setFetchOptions): Passing options by reference.
- platform/network/ResourceResponseBase.cpp:
(WebCore::ResourceResponseBase::isSuccessful): Utility function.
- platform/network/ResourceResponseBase.h:
LayoutTests:
- http/tests/xmlhttprequest/access-control-and-redirects-async-expected.txt:
- http/tests/xmlhttprequest/access-control-and-redirects-async.html:
- http/tests/xmlhttprequest/access-control-basic-get-fail-non-simple-expected.txt:
- http/tests/xmlhttprequest/workers/access-control-basic-get-fail-non-simple-expected.txt:
- 2:27 AM MathML/Early_2016_Refactoring edited by
- (diff)
- 2:24 AM Changeset in webkit [202161] by
-
- 8 edits6 adds in trunk
MathOperator: Add fallback mechanisms for stretching and mirroring radical symbols
https://bugs.webkit.org/show_bug.cgi?id=156836
Source/WebCore:
Patch by Frederic Wang <fwang@igalia.com> on 2016-06-17
Reviewed by Sergio Villar Senin.
Some platforms do not have OpenType MATH fonts pre-installed and thus can not draw stretchy
operators using size variants or glyph assembly. This is especially problematic for the
radical symbol which is used to write roots. Currently, we have some fallback code to draw
that symbol using graphical primitives but it is a bit complex and makes the style of radical
inconsistent with the font used. We solve these issues by just scaling the base glyph via a
scale transform. Such scale transform is also used to mirror the radical symbol so that we
have some support for right-to-left roots until we can do glyph-level mirroring
via the OpenType rtlm feature.
Test: mathml/radical-fallback.html
- rendering/mathml/MathOperator.cpp: Add a constant for the code point U+221A of the radical.
(WebCore::MathOperator::reset): In general, we don't need any vertical scaling for radical
symbols so m_radicalVerticalScale is initialized to 1.
(WebCore::MathOperator::calculateStretchyData): If we don't have a font with a MATH table and we
try streching a radical, then we update the vertical metrics to match the target size and
set m_radicalVerticalScale to the value necessary to make the base glyph scaled to that size.
(WebCore::MathOperator::paint): For a radical operator, we may apply a scale transform of
parameters (radicalHorizontalScale, m_radicalVerticalScale) in order to support RTL
mirroring or vertical stretching.
- rendering/mathml/MathOperator.h: We add a m_radicalVerticalScale member to indicate the
scaling to apply to the base radical glyph when the stretchy fallback is necessary.
(WebCore::MathOperator::isStretched): The operator is also considered stretched when the
m_radicalVerticalScale is applied to the base size.
- rendering/mathml/RenderMathMLRadicalOperator.cpp: Remove code specific to the old fallback mechanism.
- rendering/mathml/RenderMathMLRadicalOperator.h: Ditto.
LayoutTests:
Reviewed by Sergio Villar Senin.
We add a pixel test to verify the result of stretching a radical operator.
We test RTL/LTR direction as well as default/sans-serif font in order to check the fallback code.
- mathml/radical-fallback.html: Added.
- platform/gtk/mathml/radical-fallback-expected.png: Added.
- platform/gtk/mathml/radical-fallback-expected.txt: Added.
- platform/ios-simulator/mathml/radical-fallback-expected.txt: Added.
- platform/mac/mathml/radical-fallback-expected.png: Added.
- platform/mac/mathml/radical-fallback-expected.txt: Added.
- platform/ios-simulator/mathml/presentation/roots.txt: Updated.
Jun 16, 2016:
- 11:23 PM Changeset in webkit [202160] by
-
- 4 edits4 deletes in trunk
Unreviewed, rolling out r202147.
https://bugs.webkit.org/show_bug.cgi?id=158867
Broke scrolling tests on iOS Simulator (Requested by ap on
#webkit).
Reverted changeset:
"Focus event dispatched in iframe causes parent document to
scroll incorrectly"
https://bugs.webkit.org/show_bug.cgi?id=158629
http://trac.webkit.org/changeset/202147
- 10:19 PM Changeset in webkit [202159] by
-
- 8 edits12 adds in trunk
:in-range & :out-of-range CSS pseudo-classes shouldn't match disabled or readonly inputs
https://bugs.webkit.org/show_bug.cgi?id=156530
Patch by Benjamin Poulain <bpoulain@apple.com> on 2016-06-16
Reviewed by Simon Fraser.
LayoutTests/imported/w3c:
- web-platform-tests/html/semantics/selectors/pseudo-classes/inrange-outofrange-expected.txt:
Source/WebCore:
Elements should only match :in-range and :out-of-range
when they are candidate for constraint validation.
Tests: fast/css/pseudo-in-range-on-disabled-input-basics.html
fast/css/pseudo-in-range-on-readonly-input-basics.html
fast/css/pseudo-in-range-out-of-range-on-disabled-input-trivial.html
fast/css/pseudo-out-of-range-on-disabled-input-basics.html
fast/css/pseudo-out-of-range-on-readonly-input-basics.html
fast/selectors/in-range-out-of-range-style-update.html
- html/BaseDateAndTimeInputType.cpp:
(WebCore::BaseDateAndTimeInputType::minOrMaxAttributeChanged):
- html/NumberInputType.cpp:
(WebCore::NumberInputType::minOrMaxAttributeChanged):
I forgot to handle style update in r202143.
This is covered by the new style invalidation test.
- html/BaseDateAndTimeInputType.h:
- html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::isInRange):
(WebCore::HTMLInputElement::isOutOfRange):
LayoutTests:
- fast/css/pseudo-in-range-on-disabled-input-basics-expected.html: Added.
- fast/css/pseudo-in-range-on-disabled-input-basics.html: Added.
- fast/css/pseudo-in-range-on-readonly-input-basics-expected.html: Added.
- fast/css/pseudo-in-range-on-readonly-input-basics.html: Added.
- fast/css/pseudo-in-range-out-of-range-on-disabled-input-trivial-expected.html: Added.
- fast/css/pseudo-in-range-out-of-range-on-disabled-input-trivial.html: Added.
- fast/css/pseudo-out-of-range-on-disabled-input-basics-expected.html: Added.
- fast/css/pseudo-out-of-range-on-disabled-input-basics.html: Added.
- fast/css/pseudo-out-of-range-on-readonly-input-basics-expected.html: Added.
- fast/css/pseudo-out-of-range-on-readonly-input-basics.html: Added.
- fast/selectors/in-range-out-of-range-style-update-expected.txt: Added.
- fast/selectors/in-range-out-of-range-style-update.html: Added.
- 9:56 PM Changeset in webkit [202158] by
-
- 5 edits1 add in trunk
[EFL][GTK] LayoutTest doesn't run on Ubuntu 16.04
https://bugs.webkit.org/show_bug.cgi?id=158141
Reviewed by Michael Catanzaro.
Tools:
- Scripts/webkitpy/port/base.py: Detect php version in system, then use it.
(Port._is_debian_php_version_7):
(Port._debian_php_version):
(Port._apache_config_file_name_for_platform):
- efl/install-dependencies:
- gtk/install-dependencies:
LayoutTests:
- http/conf/debian-httpd-2.4-php7.conf: Added.
- 9:48 PM Changeset in webkit [202157] by
-
- 25 edits5 adds in trunk
Baseline JIT should be concurrent
https://bugs.webkit.org/show_bug.cgi?id=158755
Reviewed by Geoffrey Garen.
Source/JavaScriptCore:
This makes the baseline JIT concurrent. We want it to be concurrent because it takes up
about 1% of PLT3 and 10% of JSBench (though the JSBench number might be down from recent
optimizations).
The idea is really simple: I separated the compile and link phases of JIT::privateCompile(),
and arranged to call the compile phase from another thread. This doesn't reuse the old
DFG::Worklist code, because that code does things we don't need (like compilation plan
cancellation to allow GC to interleave with compilations) and is structured in a way that
would have required more changes to the baseline JIT. Also, I think that code uses the wrong
API, and as a result, clients of that API have a bad time. For example, it's never clear who
has the responsibility of setting the JIT thresholds and the DFG::Worklist goes to great
lengths to try to help its client set those things correctly, but since it doesn't set them
directly, the client then has to have additional complex logic to combine what it learned
from the Worklist and what it knows to set the thresholds. This patch takes a simpler
approach: the JITWorklist takes complete control over scheduling compilations. It's like a
combination of DFG::Worklist and operationOptimize().
Because the baseline JIT runs quickly, we can take some shortcuts. The JITWorklist requires
that all of its plans complete before a GC begins. This ensures that we don't have to worry
about interactions between the concurrent baseline JIT and the GC.
I needed to do a bunch of minor changes to the JIT to handle the races that emerged. For
example, I needed to do things to opcodes that read profiling both in the main path code
generator and the slow path one. One trick I used was to create a copy of the instruction
stream and provide that for anyone interested in the original value of the profiles. Most
code still uses the CodeBlock's instruction stream because it may emit JIT code that points
at the stream.
This also fixes a LLInt bug in prototype caching. This bug was revealed by this change
because more of our LayoutTests now run in LLInt.
This looks like it might be a ~1% Octane speed-up (on command line) and a ~0.7% PLT3
speed-up. This also looks like a ~2% JSBench speed-up.
- CMakeLists.txt:
- JavaScriptCore.xcodeproj/project.pbxproj:
- debugger/Debugger.cpp:
(JSC::Debugger::setSteppingMode):
(JSC::Debugger::toggleBreakpoint):
(JSC::Debugger::clearBreakpoints):
(JSC::Debugger::clearDebuggerRequests):
- dfg/DFGOSRExitPreparation.cpp:
(JSC::DFG::prepareCodeOriginForOSRExit):
- heap/Heap.cpp:
(JSC::Heap::didFinishIterating):
(JSC::Heap::completeAllJITPlans):
(JSC::Heap::deleteAllCodeBlocks):
(JSC::Heap::collectImpl):
(JSC::Heap::completeAllDFGPlans): Deleted.
- heap/Heap.h:
- heap/HeapInlines.h:
(JSC::Heap::forEachCodeBlock):
- jit/JIT.cpp:
(JSC::JIT::emitNotifyWrite):
(JSC::JIT::privateCompileMainPass):
(JSC::JIT::privateCompileSlowCases):
(JSC::JIT::compileWithoutLinking):
(JSC::JIT::link):
(JSC::JIT::privateCompile):
(JSC::JIT::privateCompileExceptionHandlers):
- jit/JIT.h:
(JSC::JIT::compile):
(JSC::JIT::getSlowCase):
(JSC::JIT::linkSlowCase):
(JSC::JIT::linkDummySlowCase):
- jit/JITInlines.h:
(JSC::JIT::emitTagBool):
(JSC::JIT::originalInstruction):
- jit/JITPropertyAccess32_64.cpp:
(JSC::JIT::emitSlow_op_put_to_scope):
- jit/JITPropertyAccess.cpp:
(JSC::JIT::emitSlow_op_put_by_val):
(JSC::JIT::emit_op_resolve_scope):
(JSC::JIT::emitSlow_op_resolve_scope):
(JSC::JIT::emit_op_get_from_scope):
(JSC::JIT::emitSlow_op_get_from_scope):
(JSC::JIT::emit_op_put_to_scope):
(JSC::JIT::emitSlow_op_put_to_scope):
- jit/JITWorklist.cpp: Added.
(JSC::JITWorklist::Plan::Plan):
(JSC::JITWorklist::Plan::compileInThread):
(JSC::JITWorklist::Plan::finalize):
(JSC::JITWorklist::Plan::codeBlock):
(JSC::JITWorklist::Plan::vm):
(JSC::JITWorklist::Plan::isFinishedCompiling):
(JSC::JITWorklist::Plan::isFinalized):
(JSC::JITWorklist::JITWorklist):
(JSC::JITWorklist::~JITWorklist):
(JSC::JITWorklist::completeAllForVM):
(JSC::JITWorklist::poll):
(JSC::JITWorklist::compileLater):
(JSC::JITWorklist::compileNow):
(JSC::JITWorklist::runThread):
(JSC::JITWorklist::finalizePlans):
(JSC::JITWorklist::instance):
- jit/JITWorklist.h: Added.
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::jitCompileAndSetHeuristics):
- runtime/CommonSlowPaths.cpp:
(JSC::SLOW_PATH_DECL):
- runtime/CommonSlowPaths.h:
(JSC::CommonSlowPaths::tryCachePutToScopeGlobal):
(JSC::CommonSlowPaths::tryCacheGetFromScopeGlobal):
- runtime/VM.cpp:
(JSC::VM::~VM):
Source/WTF:
The concurrent baseline JIT needs to be able to clone bytecode to get a consistent snapshot.
So, this adds such a method.
- wtf/RefCountedArray.h:
(WTF::RefCountedArray::RefCountedArray):
(WTF::RefCountedArray::clone):
Tools:
Need to disable concurrent JIT when running profiler tests. We should have been doing this
all along.
- Scripts/run-jsc-stress-tests:
LayoutTests:
This change revealed a latent bug in the LLInt. The non-llint version of this new test
would usually fail because it was still in LLInt. This new test always fails.
- js/dom/prototype-chain-caching-with-impure-get-own-property-slot-traps-llint.html: Added.
- js/dom/prototype-chain-caching-with-impure-get-own-property-slot-traps-llint-expected.txt: Added.
- js/dom/script-tests/prototype-chain-caching-with-impure-get-own-property-slot-traps-llint.js: Added.
- 9:36 PM MathML/Early_2016_Refactoring edited by
- (diff)
- 9:34 PM Changeset in webkit [202156] by
-
- 22 edits in trunk
Add separate MathOperator for selection/measuring/drawing of stretchy operators
https://bugs.webkit.org/show_bug.cgi?id=152244
Patch by Frederic Wang <fwang@igalia.com> on 2016-06-16
Reviewed by Brent Fulgham.
Source/WebCore:
We complete the class to select, measure and draw stretchy operators that is independent
from RenderMathMLOperator. That way, we will be able use stretchy operator without having
to introduce & manage anonymous RenderMathMLOperator's
(e.g for <mroot>, <msqrt> and <mfenced>).
No new tests, already covered by existing tests.
- rendering/mathml/MathOperator.cpp:
(WebCore::ascentForGlyph): Add this helper function to get glyph ascent.
(WebCore::descentForGlyph): Add this helper function to get glyph descent.
(WebCore::MathOperator::reset): Initialize all the data and calculate ascent/descent of the
base glyph.
(WebCore::MathOperator::setSizeVariant): Set the width/ascent/descent.
(WebCore::MathOperator::setGlyphAssembly): Ditto.
(WebCore::MathOperator::calculateDisplayStyleLargeOperator): Remove the STIX Word hack and
change m_maxPreferredWidth to use the actual width instead.
(WebCore::MathOperator::stretchTo): New functions to execute the actual operator streching.
(WebCore::MathOperator::fillWithVerticalExtensionGlyph): Add a FIXME for bug 155434.
(WebCore::MathOperator::fillWithHorizontalExtensionGlyph): Align all the glyph baselines on
the same axis, given by m_ascent.
Add a FIXME for bug 155434.
(WebCore::MathOperator::paintHorizontalGlyphAssembly): Ditto.
(WebCore::MathOperator::paint): Public function to do the painting.
(WebCore::MathOperator::paintVerticalGlyphAssembly): Deleted.
- rendering/mathml/MathOperator.h: Update declarations and make most of the members private.
(WebCore::MathOperator::ascent): Function to expose m_ascent.
(WebCore::MathOperator::descent): Function to expose m_descent.
- rendering/mathml/RenderMathMLOperator.cpp:
(WebCore::RenderMathMLOperator::stretchTo): Forward the stretching call to MathOperator.
(WebCore::RenderMathMLOperator::computePreferredLogicalWidths): Unfold advanceForGlyph
since we delete RenderMathMLOperator::advanceForGlyph. Just rely on
MathOperator::maxPreferredWidth to determine the preferred width of stretchy operators.
For horizontal operators, we just use the width of the base glyph.
Finally, we remove the dirty flag on preferred logical width.
(WebCore::RenderMathMLOperator::rebuildTokenContent): Reinit the MathOperator instance.
(WebCore::RenderMathMLOperator::updateFromElement): Force more updates of
RenderMathMLOperator to avoid test breakage.
(WebCore::RenderMathMLOperator::styleDidChange): Call MathOperator::reset to take into
account style change.
(WebCore::RenderMathMLOperator::updateStyle): Remove unused code.
(WebCore::RenderMathMLOperator::firstLineBaseline): Use MathOperator::ascent() function.
(WebCore::RenderMathMLOperator::computeLogicalHeight): Use MathOperator::ascent() and
MathOperator::descent() functions to calculate the height.
(WebCore::RenderMathMLOperator::paint): Only stretched operators are treated specially.
We center horizontal operator and forward the paint() call to MathOperator.
(WebCore::RenderMathMLOperator::trailingSpaceError): The error is now just the difference
between the values returned by MathOperator::maxPreferredWidth() and
MathOperator::width().
(WebCore::boundsForGlyph): Deleted.
(WebCore::heightForGlyph): Deleted.
(WebCore::advanceWidthForGlyph): Deleted.
(WebCore::RenderMathMLOperator::updateStyle): Deleted.
LayoutTests:
Update expectations for some MathML tests containing operators/radicals
to take into account small rendering changes.
- platform/gtk/mathml/opentype/horizontal-expected.png:
- platform/gtk/mathml/opentype/horizontal-expected.txt:
- platform/gtk/mathml/opentype/horizontal-munderover-expected.txt:
- platform/gtk/mathml/opentype/large-operators-expected.png:
- platform/gtk/mathml/opentype/large-operators-expected.txt:
- platform/gtk/mathml/opentype/opentype-stretchy-expected.png:
- platform/gtk/mathml/opentype/opentype-stretchy-expected.txt:
- platform/gtk/mathml/opentype/opentype-stretchy-horizontal-expected.png:
- platform/gtk/mathml/opentype/opentype-stretchy-horizontal-expected.txt:
- platform/gtk/mathml/opentype/vertical-expected.png:
- platform/gtk/mathml/opentype/vertical-expected.txt:
- platform/gtk/mathml/presentation/mo-stretch-expected.png:
- platform/gtk/mathml/presentation/mo-stretch-expected.txt:
- platform/gtk/mathml/presentation/roots-expected.txt:
- platform/ios-simulator/mathml/opentype/horizontal-munderover-expected.txt:
- platform/mac/mathml/opentype/opentype-stretchy-horizontal-expected.png:
- platform/mac/mathml/opentype/opentype-stretchy-horizontal-expected.txt:
- 8:51 PM Changeset in webkit [202155] by
-
- 3 edits1 copy1 move7 adds5 deletes in trunk
CSP: Content Security Policy should allow '*' to match the originating page's scheme
https://bugs.webkit.org/show_bug.cgi?id=158811
<rdar://problem/26819568>
Reviewed by Daniel Bates.
Source/WebCore:
Tests: security/contentSecurityPolicy/image-with-file-url-allowed-by-img-src-star.html
security/contentSecurityPolicy/link-with-file-url-allowed-by-style-src-star.html
security/contentSecurityPolicy/script-with-file-url-allowed-by-script-src-star.html
security/contentSecurityPolicy/video-with-file-url-allowed-by-media-src-star.html
- page/csp/ContentSecurityPolicySourceList.cpp:
(WebCore::ContentSecurityPolicySourceList::isProtocolAllowedByStar):
LayoutTests:
- security/contentSecurityPolicy/image-with-file-url-allowed-by-img-src-star-expected.html: Added.
- security/contentSecurityPolicy/image-with-file-url-allowed-by-img-src-star.html: Added.
- security/contentSecurityPolicy/image-with-file-url-blocked-by-img-src-star-expected.html: Removed.
- security/contentSecurityPolicy/image-with-file-url-blocked-by-img-src-star.html: Removed.
- security/contentSecurityPolicy/link-with-file-url-allowed-by-style-src-star-expected.html: Added.
- security/contentSecurityPolicy/link-with-file-url-allowed-by-style-src-star.html: Added.
- security/contentSecurityPolicy/link-with-file-url-blocked-by-style-src-star-expected.html: Removed.
- security/contentSecurityPolicy/link-with-file-url-blocked-by-style-src-star.html: Removed.
- security/contentSecurityPolicy/resources/alert-pass.js: Added.
- security/contentSecurityPolicy/script-with-file-url-allowed-by-script-src-star-expected.txt: Added.
- security/contentSecurityPolicy/script-with-file-url-allowed-by-script-src-star.html: Added.
- security/contentSecurityPolicy/video-with-file-url-allowed-by-media-src-star-expected.html: Copied from LayoutTests/security/contentSecurityPolicy/video-with-file-url-blocked-by-media-src-star.html.
- security/contentSecurityPolicy/video-with-file-url-allowed-by-media-src-star.html: Renamed from LayoutTests/security/contentSecurityPolicy/video-with-file-url-blocked-by-media-src-star.html.
- security/contentSecurityPolicy/video-with-file-url-blocked-by-media-src-star-expected.html: Removed.
- 7:50 PM Changeset in webkit [202154] by
-
- 8 edits in trunk/Source
No need to ref connection in lambda inside NetworkResourceLoader::tryStoreAsCacheEntry()
https://bugs.webkit.org/show_bug.cgi?id=158862
Reviewed by Darin Adler.
Source/WebKit2:
No need to ref connection in lambda inside NetworkResourceLoader::tryStoreAsCacheEntry().
We already ref the NetworkResourceLoader which hold a ref to the connection. Also update
the lambda capture to use Ref<> for the NetworkResourceLoader instead of RefPtr<>.
Switch callback type from std::function to WTF::NoncopyableFunction so we can capture a
Ref<>.
- NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::tryStoreAsCacheEntry):
- NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::store):
- NetworkProcess/cache/NetworkCache.h:
- NetworkProcess/cache/NetworkCacheStorage.cpp:
(WebKit::NetworkCache::Storage::WriteOperation::WriteOperation):
- NetworkProcess/cache/NetworkCacheStorage.h:
Source/WTF:
Add NoncopyableFunction constructor that takes a nullptr_t in, in order
to match the std::function API and make porting from one to the other
easier.
- wtf/NoncopyableFunction.h:
- 7:30 PM Changeset in webkit [202153] by
-
- 2 edits in trunk/Source/WebCore
Add HTTPHeaderMap::set() overload taking a NSString*
https://bugs.webkit.org/show_bug.cgi?id=158857
Reviewed by Darin Adler.
Add HTTPHeaderMap::set() overloading taking a NSString* in addition to
the one taking a CFStringRef. It is useful for the Cocoa implementation
of ResourceRequest::doUpdateResourceRequest().
- platform/network/HTTPHeaderMap.h:
(WebCore::HTTPHeaderMap::set):
- 7:26 PM Changeset in webkit [202152] by
-
- 21 edits2 adds in trunk
Web Inspector: console.profile should use the new Sampling Profiler
https://bugs.webkit.org/show_bug.cgi?id=153499
<rdar://problem/24352431>
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2016-06-16
Reviewed by Timothy Hatcher.
Source/JavaScriptCore:
Currently console.profile/profileEnd behave slightly differently
between JSContext and Web inspection. Unifying will be part of:
<https://webkit.org/b/158753> Generalize the concept of Instruments on the backend
Both JSContext and Web inspection keep track of active
profiles started and stopped via console.profile/profileEnd.
JSContext inspection sends its programmatic start/stop
via the ScriptProfiler domain.
Web inspection sends its programmatic start/stop
via the Timeline domain, and also will start/stop backend
list of Instruments.
The functional differences between these is that for JSContext
inspection, console.profile only starts/stops the ScriptProfiler
domain, and does not auto-start other instruments. This isn't really
a problem right now given the instruments available for JSContext
inspection; but it will be nice to unify as we add more instruments.
Also, JSContext inspection won't have "Profile (name)" records in
its Events view, since those are currently generated only by the
Web's Timeline domain.
- inspector/protocol/ScriptProfiler.json:
- inspector/protocol/Timeline.json:
Events to inform the frontend of programmatic start/stop.
- debugger/Debugger.h:
- inspector/agents/InspectorDebuggerAgent.cpp:
(Inspector::InspectorDebuggerAgent::breakpointsActive):
(Inspector::InspectorDebuggerAgent::isPaused):
- inspector/agents/InspectorDebuggerAgent.h:
Expose breakpoints active state, since programmatic recording
will temporarily disabled breakpoints if needed.
- inspector/JSGlobalObjectConsoleClient.cpp:
(Inspector::JSGlobalObjectConsoleClient::JSGlobalObjectConsoleClient):
(Inspector::JSGlobalObjectConsoleClient::profile):
(Inspector::JSGlobalObjectConsoleClient::profileEnd):
(Inspector::JSGlobalObjectConsoleClient::startConsoleProfile):
(Inspector::JSGlobalObjectConsoleClient::stopConsoleProfile):
- inspector/JSGlobalObjectConsoleClient.h:
- inspector/JSGlobalObjectInspectorController.cpp:
(Inspector::JSGlobalObjectInspectorController::JSGlobalObjectInspectorController):
- inspector/agents/InspectorScriptProfilerAgent.cpp:
(Inspector::InspectorScriptProfilerAgent::programmaticCaptureStarted):
(Inspector::InspectorScriptProfilerAgent::programmaticCaptureStopped):
- inspector/agents/InspectorScriptProfilerAgent.h:
JSContext implementation of console.profile/profileEnd.
Source/WebCore:
Test: inspector/timeline/setInstruments-programmatic-capture.html
- inspector/InspectorTimelineAgent.cpp:
(WebCore::InspectorTimelineAgent::startFromConsole):
(WebCore::InspectorTimelineAgent::stopFromConsole):
(WebCore::InspectorTimelineAgent::mainFrameStartedLoading):
(WebCore::InspectorTimelineAgent::startProgrammaticCapture):
(WebCore::InspectorTimelineAgent::stopProgrammaticCapture):
(WebCore::InspectorTimelineAgent::toggleInstruments):
(WebCore::InspectorTimelineAgent::toggleScriptProfilerInstrument):
(WebCore::InspectorTimelineAgent::toggleHeapInstrument):
(WebCore::InspectorTimelineAgent::toggleMemoryInstrument):
(WebCore::InspectorTimelineAgent::toggleTimelineInstrument):
- inspector/InspectorTimelineAgent.h:
Web implementation of console.profile/profileEnd.
Make helpers for startings / stopping instruments.
Source/WebInspectorUI:
- UserInterface/Controllers/TimelineManager.js:
(WebInspector.TimelineManager.prototype.programmaticCaptureStarted):
(WebInspector.TimelineManager.prototype.programmaticCaptureStopped):
(WebInspector.TimelineManager.prototype.scriptProfilerProgrammaticCaptureStarted):
(WebInspector.TimelineManager.prototype.scriptProfilerProgrammaticCaptureStopped):
For programmatic captures, automatically show the Script Timeline, since
that is guarenteed to have been started by the backend. Start capturing
without creating a new recording.
- UserInterface/Models/TimelineRecording.js:
(WebInspector.TimelineRecording.prototype.stop):
(WebInspector.TimelineRecording.prototype.addScriptInstrumentForProgrammaticCapture):
When stopping for programmatic capture we don't need to disable
instruments, the backend would have already done this.
- UserInterface/Protocol/ScriptProfilerObserver.js:
(WebInspector.ScriptProfilerObserver.prototype.programmaticCaptureStarted):
(WebInspector.ScriptProfilerObserver.prototype.programmaticCaptureStopped):
(WebInspector.ScriptProfilerObserver):
- UserInterface/Protocol/TimelineObserver.js:
(WebInspector.TimelineObserver.prototype.programmaticCaptureStarted):
(WebInspector.TimelineObserver.prototype.programmaticCaptureStopped):
(WebInspector.TimelineObserver):
Pass through Web and JSContext programmatic capture events.
LayoutTests:
- inspector/timeline/setInstruments-programmatic-capture-expected.txt: Added.
- inspector/timeline/setInstruments-programmatic-capture.html: Added.
New test that the backend auto-starts instruments during programmatic capture.
- 6:10 PM Changeset in webkit [202151] by
-
- 5 edits4 adds in trunk
Restrict security origin inheritance to empty, about:blank, and about:srcdoc URLs
https://bugs.webkit.org/show_bug.cgi?id=158855
<rdar://problem/26142632>
Reviewed by Alex Christensen.
Source/WebCore:
Tests: http/tests/dom/window-open-about-blank-and-access-document.html
http/tests/dom/window-open-about-webkit-org-and-access-document.html
Document.cpp previously checked whether a document should inherit its owner's
security origin by checking if the URL is either empty or blank. URL.cpp in
turn only checks if the protocol is "about:" in the isBlankURL() function.
Thus all about:* URLs inherited security origin. This patch restricts
security origin inheritance to empty, about:blank, and about:srcdoc URLs.
Quotes and links from the WHATWG spec regarding about:srcdoc:
7.1 Browsing contexts
A browsing context can have a creator browsing context, the browsing context
that was responsible for its creation. If a browsing context has a parent
browsing context, then that is its creator browsing context. Otherwise, if the
browsing context has an opener browsing context, then that is its creator
browsing context. Otherwise, the browsing context has no creator browsing
context.
https://html.spec.whatwg.org/multipage/browsers.html#concept-document-bc
7.1.1 Nested browsing contexts
Certain elements (for example, iframe elements) can instantiate further
browsing contexts. These are called nested browsing contexts. If a browsing
context P has a Document D with an element E that nests another browsing
context C inside it, then C is said to be nested through D, and E is said to
be the browsing context container of C. If the browsing context container
element E is in the Document D, then P is said to be the parent browsing
context of C and C is said to be a child browsing context of P. Otherwise,
the nested browsing context C has no parent browsing context.
https://html.spec.whatwg.org/multipage/browsers.html#nested-browsing-context
4.8.5 The iframe element
The iframe element represents a nested browsing context.
...
If the srcdoc attribute is specified
Navigate the element's child browsing context to a new response whose
url list consists of about:srcdoc ...
https://html.spec.whatwg.org/multipage/embedded-content.html#attr-iframe-srcdoc
- dom/Document.cpp:
(WebCore::Document::initSecurityContext):
Now uses the URL::shouldInheritSecurityOriginFromOwner() function instead.
(WebCore::Document::initContentSecurityPolicy):
Now uses the URL::shouldInheritSecurityOriginFromOwner() function instead.
(WebCore::shouldInheritSecurityOriginFromOwner): Deleted.
Moved to URL::shouldInheritSecurityOriginFromOwner() and restricted the check.
- platform/URL.cpp:
(WebCore::URL::shouldInheritSecurityOriginFromOwner):
- platform/URL.h:
Moved the function from Document and restricted the check to only allow
security origin inheritance for empty, about:blank, and about:srcdoc URLs.
LayoutTests:
- http/tests/dom/window-open-about-blank-and-access-document-expected.txt: Added.
- http/tests/dom/window-open-about-blank-and-access-document.html: Added.
- http/tests/dom/window-open-about-webkit-org-and-access-document-expected.txt: Added.
- http/tests/dom/window-open-about-webkit-org-and-access-document.html: Added.
- 5:45 PM Changeset in webkit [202150] by
-
- 4 edits40 moves2 adds in trunk/LayoutTests
Move most of CSP tests into security/contentSecurityPolicy/
https://bugs.webkit.org/show_bug.cgi?id=158859
Reviewed by Daniel Bates.
- platform/mac-wk1/TestExpectations:
- platform/mac/TestExpectations:
- platform/wk2/TestExpectations:
- security/contentSecurityPolicy/blocks-video-expected.txt: Renamed from LayoutTests/media/csp-blocks-video-expected.txt.
- security/contentSecurityPolicy/blocks-video.html: Renamed from LayoutTests/media/csp-blocks-video.html.
- security/contentSecurityPolicy/font-loading-block-all-expected.txt: Renamed from LayoutTests/fast/text/font-loading-csp-block-all-expected.txt.
- security/contentSecurityPolicy/font-loading-block-all.html: Renamed from LayoutTests/fast/text/font-loading-csp-block-all.html.
- security/contentSecurityPolicy/image-with-blob-url-allowed-by-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-blob-url-allowed-by-csp-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html.
- security/contentSecurityPolicy/image-with-blob-url-allowed-by-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-blob-url-allowed-by-csp-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html.
- security/contentSecurityPolicy/image-with-blob-url-blocked-by-img-src-star-expected.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-blob-url-blocked-by-csp-img-src-star-expected.html.
- security/contentSecurityPolicy/image-with-blob-url-blocked-by-img-src-star.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-blob-url-blocked-by-csp-img-src-star.html.
- security/contentSecurityPolicy/image-with-data-url-allowed-by-img-src-star-expected.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-data-url-allowed-by-csp-img-src-star-expected.html.
- security/contentSecurityPolicy/image-with-data-url-allowed-by-img-src-star.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-data-url-allowed-by-csp-img-src-star.html.
- security/contentSecurityPolicy/image-with-file-url-allowed-by-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-file-url-allowed-by-csp-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html.
- security/contentSecurityPolicy/image-with-file-url-allowed-by-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-file-url-allowed-by-csp-img-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html.
- security/contentSecurityPolicy/image-with-file-url-blocked-by-img-src-star-expected.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-file-url-blocked-by-csp-img-src-star-expected.html.
- security/contentSecurityPolicy/image-with-file-url-blocked-by-img-src-star.html: Renamed from LayoutTests/fast/dom/HTMLImageElement/image-with-file-url-blocked-by-csp-img-src-star.html.
- security/contentSecurityPolicy/link-with-blob-url-allowed-by-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-blob-url-allowed-by-csp-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html.
- security/contentSecurityPolicy/link-with-blob-url-allowed-by-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-blob-url-allowed-by-csp-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html.
- security/contentSecurityPolicy/link-with-blob-url-blocked-by-style-src-star-expected.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-blob-url-blocked-by-csp-style-src-star-expected.html.
- security/contentSecurityPolicy/link-with-blob-url-blocked-by-style-src-star.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-blob-url-blocked-by-csp-style-src-star.html.
- security/contentSecurityPolicy/link-with-data-url-allowed-by-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-data-url-allowed-by-csp-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html.
- security/contentSecurityPolicy/link-with-data-url-allowed-by-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-data-url-allowed-by-csp-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html.
- security/contentSecurityPolicy/link-with-data-url-blocked-by-style-src-star-expected.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-data-url-blocked-by-csp-style-src-star-expected.html.
- security/contentSecurityPolicy/link-with-data-url-blocked-by-style-src-star.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-data-url-blocked-by-csp-style-src-star.html.
- security/contentSecurityPolicy/link-with-file-url-allowed-by-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-file-url-allowed-by-csp-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html.
- security/contentSecurityPolicy/link-with-file-url-allowed-by-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-file-url-allowed-by-csp-style-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html.
- security/contentSecurityPolicy/link-with-file-url-blocked-by-style-src-star-expected.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-file-url-blocked-by-csp-style-src-star-expected.html.
- security/contentSecurityPolicy/link-with-file-url-blocked-by-style-src-star.html: Renamed from LayoutTests/fast/dom/HTMLLinkElement/link-with-file-url-blocked-by-csp-style-src-star.html.
- security/contentSecurityPolicy/resources/worker-inherits-blocks-eval.js: Renamed from LayoutTests/fast/workers/resources/worker-inherits-csp-blocks-eval.js.
(catch):
- security/contentSecurityPolicy/resources/worker-inherits-blocks-xhr.js: Renamed from LayoutTests/fast/workers/resources/worker-inherits-csp-blocks-xhr.js.
(catch):
- security/contentSecurityPolicy/video-with-blob-url-allowed-by-media-src-star-expected.html: Renamed from LayoutTests/media/video-with-blob-url-allowed-by-csp-media-src-star-expected.html.
- security/contentSecurityPolicy/video-with-blob-url-allowed-by-media-src-star.html: Renamed from LayoutTests/media/video-with-blob-url-allowed-by-csp-media-src-star.html.
- security/contentSecurityPolicy/video-with-data-url-allowed-by-media-src-star-expected.html: Renamed from LayoutTests/media/video-with-data-url-allowed-by-csp-media-src-star-expected.html.
- security/contentSecurityPolicy/video-with-data-url-allowed-by-media-src-star.html: Renamed from LayoutTests/media/video-with-data-url-allowed-by-csp-media-src-star.html.
- security/contentSecurityPolicy/video-with-file-url-allowed-by-media-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html: Renamed from LayoutTests/media/video-with-file-url-allowed-by-csp-media-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled-expected.html.
- security/contentSecurityPolicy/video-with-file-url-allowed-by-media-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html: Renamed from LayoutTests/media/video-with-file-url-allowed-by-csp-media-src-star-with-AllowContentSecurityPolicySourceStarToMatchAnyProtocol-enabled.html.
- security/contentSecurityPolicy/video-with-file-url-blocked-by-media-src-star-expected.html: Renamed from LayoutTests/media/video-with-file-url-blocked-by-csp-media-src-star-expected.html.
- security/contentSecurityPolicy/video-with-file-url-blocked-by-media-src-star.html: Renamed from LayoutTests/media/video-with-file-url-blocked-by-csp-media-src-star.html.
- security/contentSecurityPolicy/worker-inherits-blocks-eval-expected.txt: Renamed from LayoutTests/fast/workers/worker-inherits-csp-blocks-eval-expected.txt.
- security/contentSecurityPolicy/worker-inherits-blocks-eval.html: Renamed from LayoutTests/fast/workers/worker-inherits-csp-blocks-eval.html.
- security/contentSecurityPolicy/worker-inherits-blocks-xhr-expected.txt: Renamed from LayoutTests/fast/workers/worker-inherits-csp-blocks-xhr-expected.txt.
- security/contentSecurityPolicy/worker-inherits-blocks-xhr.html: Renamed from LayoutTests/fast/workers/worker-inherits-csp-blocks-xhr.html.
- 5:00 PM Changeset in webkit [202149] by
-
- 2 edits in trunk/Tools
[Win] Accessibility implementation unable to recurse through document (sometimes) to find named elements
https://bugs.webkit.org/show_bug.cgi?id=140798
Reviewed by Brent Fulgham.
Make sure layout is up-to-date before trying to find named elements.
- DumpRenderTree/win/AccessibilityControllerWin.cpp:
(AccessibilityController::rootElement):
- 4:39 PM Changeset in webkit [202148] by
-
- 9 edits in trunk/Source/WebKit2
[WK2] Improve serialization of SubresourcesEntry to network disk cache
https://bugs.webkit.org/show_bug.cgi?id=158851
Reviewed by Antti Koivisto.
Improve serialization of SubresourcesEntry to network disk cache:
- Do not bother serializing SubresourceInfo's first party for cookies and HTTP headers data members if the resource is transient. This is because those are never used for transient resources. This patch also makes it so that we do not even bother initializing those data members if the transient flag is set.
- Update SubresourceInfo::encode() / decode() to encode and decode the HTTPHeaderMap data member with the right NetworkCache::Coder template specialization instead of using directly HTTPHeaderMap::decode() and HTTPHeaderMap::encode(). HTTPHeaderMap::decode() / encode() is unsafe in the disk cache case (see r200394).
Also update entry types as so:
- "resource" -> "Resource"
- "subresources" -> "SubResources"
These entry types are used a folder names in the disk cache and other
folder names were capitalized. Antti suggested this was a good time
to harmonize since we're bumping the cache version.
- NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::constructRevalidationRequest):
(WebKit::NetworkCache::SpeculativeLoadManager::startSpeculativeRevalidation):
- NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.h:
- NetworkProcess/cache/NetworkCacheStorage.h:
- NetworkProcess/cache/NetworkCacheSubresourcesEntry.cpp:
(WebKit::NetworkCache::SubresourceInfo::encode):
(WebKit::NetworkCache::SubresourceInfo::decode):
- NetworkProcess/cache/NetworkCacheSubresourcesEntry.h:
(WebKit::NetworkCache::SubresourceInfo::SubresourceInfo):
(WebKit::NetworkCache::SubresourceInfo::isTransient):
(WebKit::NetworkCache::SubresourceInfo::firstPartyForCookies):
(WebKit::NetworkCache::SubresourceInfo::requestHeaders):
- 4:34 PM Changeset in webkit [202147] by
-
- 4 edits4 adds in trunk
Source/WebCore:
[iOS] Focus event dispatched in iframe causes parent document to scroll incorrectly
https://bugs.webkit.org/show_bug.cgi?id=158629
rdar://problem/26521616
Reviewed by Enrica Casucci.
When focussing elements in iframes, the page could scroll to an incorrect location.
This happened because code in Element::focus() tried to disable scrolling on focus,
but did so only for the current frame, so ancestor frames got programmatically scrolled.
On iOS we handle the scrolling in the UI process, so never want the web process to
do programmatic scrolling.
Fix by changing the focus and cache restore code to use SelectionRevealMode::DoNotReveal,
rather than manually prohibiting frame scrolling.
Tests: fast/forms/ios/focus-input-in-iframe.html
fast/forms/ios/programmatic-focus-input-in-iframe.html
- dom/Element.cpp:
(WebCore::Element::focus):
- history/CachedPage.cpp:
(WebCore::CachedPage::restore):
LayoutTests:
Focus event dispatched in iframe causes parent document to scroll incorrectly
https://bugs.webkit.org/show_bug.cgi?id=158629
rdar://problem/26521616
Reviewed by Enrica Casucci.
Tests for user-initiated and programmatic focus in frames.
- fast/forms/ios/focus-input-in-iframe-expected.txt: Added.
- fast/forms/ios/focus-input-in-iframe.html: Added.
- fast/forms/ios/programmatic-focus-input-in-iframe.html: Added.
- fast/forms/ios/programmatic-focus-input-in-iframe-expected.txt: Added.
- 4:27 PM Changeset in webkit [202146] by
-
- 3 edits2 adds in trunk
[New Block-Inside-Inline Model] Do not attempt to re-run margin collapsing on the block sequence.
https://bugs.webkit.org/show_bug.cgi?id=158854
Reviewed by David Hyatt.
Source/WebCore:
Test: fast/block/inside-inlines/crash-on-first-line-change.html
- rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlockFlow::marginCollapseLinesFromStart):
LayoutTests:
- fast/block/inside-inlines/crash-on-first-line-change-expected.txt: Added.
- fast/block/inside-inlines/crash-on-first-line-change.html: Added.
- 4:03 PM Changeset in webkit [202145] by
-
- 2 edits in trunk/Source/WebCore
Include cstdlib before using std::atexit
https://bugs.webkit.org/show_bug.cgi?id=158681
Patch by Ting-Wei Lan <Ting-Wei Lan> on 2016-06-16
Reviewed by Brent Fulgham.
- platform/graphics/PlatformDisplay.cpp:
- 3:55 PM Changeset in webkit [202144] by
-
- 2 edits in trunk/Source/WebCore
Use StringView::toAtomicString() in HTMLImageElement::setBestFitURLAndDPRFromImageCandidate()
https://bugs.webkit.org/show_bug.cgi?id=158853
Reviewed by Brent Fulgham.
Use StringView::toAtomicString() in HTMLImageElement::setBestFitURLAndDPRFromImageCandidate()
as m_bestFitImageURL data member is an AtomicString. This avoids constructing a String and
then atomizing it.
- html/HTMLImageElement.cpp:
(WebCore::HTMLImageElement::setBestFitURLAndDPRFromImageCandidate):
- 3:47 PM Changeset in webkit [202143] by
-
- 15 edits6 adds in trunk
:in-range & :out-of-range CSS pseudo-classes shouldn't match inputs without range limitations
https://bugs.webkit.org/show_bug.cgi?id=156558
Patch by Benjamin Poulain <bpoulain@apple.com> on 2016-06-16
Reviewed by Simon Fraser.
LayoutTests/imported/w3c:
- web-platform-tests/html/semantics/selectors/pseudo-classes/inrange-outofrange-expected.txt:
One of the previous match was erroneous.
Our results are still very far from being correct. There are several
bugs affecting our range validation.
Source/WebCore:
The pseudo selectors :in-range and :out-of-range should only
apply if:
-minimum/maximum are defined for the input type
-the input value is/is-not suffering from underflow/overflow.
Only certain types have a valid minimum and maximum:
-number
-range
-date
-month
-week
-time
-datetime-local
Of those, only one has a default minimum and maximum: range.
For all the others, the minimum or maximum is only defined
if the min/max attribute is defined and valid.
This patch addresses these constraints for number and range.
The date types range validation is severely broken and is
left untouched. It really needs a clean rewrite.
Tests: fast/css/pseudo-in-range-basics.html
fast/css/pseudo-in-range-out-of-range-trivial.html
fast/css/pseudo-out-of-range-basics.html
- html/DateInputType.cpp:
(WebCore::DateInputType::createStepRange):
- html/DateTimeInputType.cpp:
(WebCore::DateTimeInputType::createStepRange):
- html/DateTimeLocalInputType.cpp:
(WebCore::DateTimeLocalInputType::createStepRange):
- html/InputType.cpp:
(WebCore::InputType::isInRange):
(WebCore::InputType::isOutOfRange):
Notice the isEmpty() shortcut.
A value can only overflow/underflow if it is not empty.
- html/MonthInputType.cpp:
(WebCore::MonthInputType::createStepRange):
- html/NumberInputType.cpp:
(WebCore::NumberInputType::createStepRange):
- html/RangeInputType.cpp:
(WebCore::RangeInputType::createStepRange):
- html/StepRange.cpp:
(WebCore::StepRange::StepRange):
- html/StepRange.h:
(WebCore::StepRange::hasRangeLimitations):
- html/WeekInputType.cpp:
(WebCore::WeekInputType::createStepRange):
LayoutTests:
- fast/css/pseudo-in-range-basics-expected.html: Added.
- fast/css/pseudo-in-range-basics.html: Added.
- fast/css/pseudo-in-range-out-of-range-trivial-expected.html: Added.
- fast/css/pseudo-in-range-out-of-range-trivial.html: Added.
- fast/css/pseudo-out-of-range-basics-expected.html: Added.
- fast/css/pseudo-out-of-range-basics.html: Added.
- 3:44 PM Changeset in webkit [202142] by
-
- 6 edits in trunk/Source
Fix macOS Sierra build
https://bugs.webkit.org/show_bug.cgi?id=158849
Reviewed by Tim Horton.
Source/WebCore:
Add WebCore:: qualifiers for IOSurface, to avoid conflicts with the IOSurface Objective-C class.
Also, add an asLayerContents() getter that will return an id that's suitable for setting
as the contents of a CALayer.
- platform/graphics/cocoa/IOSurface.h:
- platform/graphics/cocoa/IOSurface.mm:
Source/WebKit2:
Add WebCore:: qualifiers for IOSurface, to avoid conflicts with the IOSurface Objective-C class
and adopt IOSurface::asLayerContents().
- Shared/mac/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::decode):
(WebKit::RemoteLayerBackingStore::bytesPerPixel):
(WebKit::RemoteLayerBackingStore::swapToValidFrontBuffer):
(WebKit::RemoteLayerBackingStore::applyBackingStoreToLayer):
(WebKit::RemoteLayerBackingStore::setBufferVolatility):
(WebKit::RemoteLayerBackingStore::Buffer::discard):
- UIProcess/mac/ViewSnapshotStore.mm:
(WebKit::ViewSnapshot::create):
(WebKit::ViewSnapshot::ViewSnapshot):
(WebKit::ViewSnapshot::asLayerContents):
- 3:18 PM Changeset in webkit [202141] by
-
- 3 edits in trunk/Source/JavaScriptCore
Kraken/stanford-crypto-pbkdf2.js sometimes crashes with an OSR assertion in FTL
https://bugs.webkit.org/show_bug.cgi?id=158850
Reviewed by Keith Miller.
Bytecode liveness was incorrectly claiming that all tail-deleted locals are live! That's
crazy! We never noticed this because extending OSR liveness is usually not a showstopper and
until recently we didn't have a lot of tail-call test cases to play with. Well, we do now,
thanks to the increasing reliance on tail calls in our builtins.
- dfg/DFGGraph.cpp:
(JSC::DFG::Graph::localsLiveInBytecode): Fix the bug and add some optional tracing. Also restructure the code so that we don't break to return true, since that's counterintuitive.
- ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::buildExitArguments): Make this assertion print more useful information.
- 3:06 PM Changeset in webkit [202140] by
-
- 5 edits in trunk/Source/WebKit2
When a WebsiteDataStore finds no ProcessPools and creates one from scratch, it should represent the data store's config.
rdar://problem/26586991 and https://bugs.webkit.org/show_bug.cgi?id=158844
Reviewed by Anders Carlsson.
- UIProcess/API/APIProcessPoolConfiguration.cpp:
(API::ProcessPoolConfiguration::createWithWebsiteDataStoreConfiguration):
- UIProcess/API/APIProcessPoolConfiguration.h:
- UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::WebsiteDataStore):
(WebKit::WebsiteDataStore::fetchData):
(WebKit::WebsiteDataStore::removeData):
(WebKit::WebsiteDataStore::processPools):
- UIProcess/WebsiteData/WebsiteDataStore.h:
- 2:47 PM Changeset in webkit [202139] by
-
- 2 edits in trunk/Source/WebCore
REGRESSION(r196217): 3% JSBench regression on iPhone 5.
<https://webkit.org/b/158848>
<rdar://problem/26609622>
Unreviewed rollout.
Don't jettison linked code on every top-level navigation as that was hurting JSBench on iPhone 5.
- loader/FrameLoader.cpp:
(WebCore::FrameLoader::commitProvisionalLoad):
- 2:35 PM Changeset in webkit [202138] by
-
- 2 edits in trunk/Source/WebInspectorUI
REGRESSION (r201538): Web Inspector: Border over Allocations Views is too thick
https://bugs.webkit.org/show_bug.cgi?id=158846
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2016-06-16
Reviewed by Timothy Hatcher.
- UserInterface/Views/HeapAllocationsTimelineView.css:
(.timeline-view.heap-allocations > .content-view-container > .content-view): Deleted.
- 1:41 PM Changeset in webkit [202137] by
-
- 1 copy in tags/Safari-602.1.37
New tag.
- 1:39 PM Changeset in webkit [202136] by
-
- 12 edits in trunk
Unreviewed, rolling out r202002 and r202111.
Ryosuke says this was a JSBench regression on iOS
Reverted changesets:
"Make HashMap and HashSet work with Refs"
https://bugs.webkit.org/show_bug.cgi?id=158638
http://trac.webkit.org/changeset/202002
"Improve HashMap and HashSet support for Ref"
https://bugs.webkit.org/show_bug.cgi?id=158789
http://trac.webkit.org/changeset/202111
- 1:39 PM Changeset in webkit [202135] by
-
- 5 edits in trunk/Source
Versioning.
- 12:50 PM Changeset in webkit [202134] by
-
- 2 edits in trunk/Source/WebInspectorUI
Unreviewed build fix.
- UserInterface/Views/DataGrid.js:
(WebInspector.DataGrid.prototype.createSettings):
"showColumn" -> "setColumnVisible".
- 12:39 PM Changeset in webkit [202133] by
-
- 14 edits in trunk/Source/WebInspectorUI
Web Inspector: DataGrid should manage its own settings
https://bugs.webkit.org/show_bug.cgi?id=158675
<rdar://problem/26761338>
Reviewed by Timothy Hatcher.
Move grid sorting settings (sort order, column identifier) into DataGrid,
removing duplicated code from various views. A new grid method, "createSettings",
takes an identifier which is used as a prefix for naming the grid's settings.
So that the new behavior is easy to use and hard to misuse, default values
for each setting are set to the current value of the associated property.
The result is that:
grid.sortColumnIdentifier = "columnA";
grid.createSettings("myGrid");
will use "columnA" as the default value the first time the setting is created.
- UserInterface/Views/ApplicationCacheFrameContentView.js:
(WebInspector.ApplicationCacheFrameContentView.prototype._createDataGrid):
- UserInterface/Views/CookieStorageContentView.js:
(WebInspector.CookieStorageContentView.prototype._rebuildTable):
- UserInterface/Views/DOMStorageContentView.js:
(WebInspector.DOMStorageContentView):
- UserInterface/Views/DataGrid.js:
(WebInspector.DataGrid):
(WebInspector.DataGrid.prototype.createSettings):
(WebInspector.DataGrid.prototype.get identifier): Deleted.
(WebInspector.DataGrid.prototype.set identifier): Deleted.
(WebInspector.DataGrid.prototype.set sortOrderSetting): Deleted.
(WebInspector.DataGrid.prototype.set sortColumnIdentifierSetting): Deleted.
- UserInterface/Views/HeapAllocationsTimelineView.js:
(WebInspector.HeapAllocationsTimelineView):
- UserInterface/Views/HeapSnapshotContentView.js:
(WebInspector.HeapSnapshotContentView):
- UserInterface/Views/LayerTreeDetailsSidebarPanel.js:
(WebInspector.LayerTreeDetailsSidebarPanel.prototype._buildDataGridSection):
- UserInterface/Views/LayoutTimelineView.js:
(WebInspector.LayoutTimelineView):
- UserInterface/Views/NetworkGridContentView.js:
(WebInspector.NetworkGridContentView):
- UserInterface/Views/NetworkTimelineView.js:
(WebInspector.NetworkTimelineView):
- UserInterface/Views/ProfileView.js:
(WebInspector.ProfileView):
- UserInterface/Views/RenderingFrameTimelineView.js:
(WebInspector.RenderingFrameTimelineView):
- UserInterface/Views/ScriptDetailsTimelineView.js:
(WebInspector.ScriptDetailsTimelineView):
- 11:49 AM Changeset in webkit [202132] by
-
- 14 edits6 adds in trunk
[iOS WK2] On iPad, indirect focussing of a text field doesn't always scroll to the correct location
https://bugs.webkit.org/show_bug.cgi?id=158828
Reviewed by Enrica Casucci.
Source/WebKit2:
WebPage::getAssistedNodeInformation() unconditionally set information.selectionRect()
to a 1x1 rect at the last interaction location. This caused -[WKWebView _zoomToFocusRect:...]
to early return under the !forceScroll clause if the last interaction rect was visible, but
the elementRect was not. This would happen, for example, if a click in a <button> focussed
an input on some other portion of the page. This behavior is iPad-only, because forceScroll
is only NO there.
Fix by making getAssistedNodeInformation() only set the selectionRect to the last interaction
location if that is inside of the elementRect (this caters to the case where a tap is inside
an input which is larger than the screen). Otherwise, set it to the empty rect.
Make this testable by allowing a test to opt into the iPad-style scrolling/zooming behavior
in this code path (essentially, making forceScroll NO for testing), via testing SPI on WKWebView,
plumbed through to WKContentView.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView forceIPadStyleZoomOnInputFocus]):
(-[WKWebView setForceIPadStyleZoomOnInputFocus:]):
- UIProcess/API/Cocoa/WKWebViewPrivate.h:
- UIProcess/ios/WKContentViewInteraction.h:
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _displayFormNodeInputView]):
(-[WKContentView requiresAccessoryView:]):
(-[WKContentView inputAccessoryView]):
(-[WKContentView forceIPadStyleZoomOnInputFocus]):
(-[WKContentView setForceIPadStyleZoomOnInputFocus:]):
(-[WKContentView requiresAccessoryView]): Remove redundant returns, and allow testing to
override the "isIPad" condition.
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getAssistedNodeInformation):
Tools:
Expose "forceIPadStyleZoomOnInputFocus" on UIScriptController, which allows iPad-style
zooming behavior on <input> focus in the iPhone simulator, which is used for testing.
- WebKitTestRunner/UIScriptContext/Bindings/UIScriptController.idl:
- WebKitTestRunner/UIScriptContext/UIScriptController.cpp:
(WTR::UIScriptController::forceIPadStyleZoomOnInputFocus):
(WTR::UIScriptController::setForceIPadStyleZoomOnInputFocus):
- WebKitTestRunner/UIScriptContext/UIScriptController.h:
- WebKitTestRunner/ios/TestControllerIOS.mm:
(WTR::TestController::platformResetStateToConsistentValues):
- WebKitTestRunner/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::forceIPadStyleZoomOnInputFocus):
(WTR::UIScriptController::setForceIPadStyleZoomOnInputFocus):
LayoutTests:
Add various tests for focusing form fields. focus-input-via-button-ipad.html tests
the change in the current patch.
- fast/forms/ios/focus-input-in-fixed-expected.txt: Added.
- fast/forms/ios/focus-input-in-fixed.html: Added.
- fast/forms/ios/focus-input-via-button-ipad-expected.txt: Added.
- fast/forms/ios/focus-input-via-button-ipad.html: Added.
- fast/forms/ios/focus-long-textarea-expected.txt: Added.
- fast/forms/ios/focus-long-textarea.html: Added.
- fast/forms/ios/resources/zooming-test-utils.js:
(testZoomAfterTap):
- 11:25 AM Changeset in webkit [202131] by
-
- 8 edits in trunk/Source/JavaScriptCore
Add collecting of LLINT slow path stats.
https://bugs.webkit.org/show_bug.cgi?id=158829
Reviewed by Keith Miller.
- llint/LLIntData.cpp:
(JSC::LLInt::Data::dumpStats):
- llint/LLIntData.h:
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
- llint/LLIntSlowPaths.h:
- llint/LowLevelInterpreter.asm:
- llint/LowLevelInterpreter32_64.asm:
- llint/LowLevelInterpreter64.asm:
- 11:12 AM Changeset in webkit [202130] by
-
- 5 edits2 adds in trunk
WebRTC: Check type of this in RTCPeerConnection JS built-in functions
https://bugs.webkit.org/show_bug.cgi?id=151303
Reviewed by Youenn Fablet.
Source/WebCore:
Check type of 'this' in RTCPeerConnection JS built-in functions.
Test: fast/mediastream/RTCPeerConnection-js-built-ins-check-this.html
- Modules/mediastream/RTCPeerConnection.js:
(createOffer):
(createAnswer):
(setLocalDescription):
(setRemoteDescription):
(addIceCandidate):
(getStats):
Reject if 'this' isn't of type RTCPeerConnection.
- Modules/mediastream/RTCPeerConnectionInternals.js:
(isRTCPeerConnection):
Add helper function to perform type check. Needs further robustifying.
LayoutTests:
Verify that the RTCPeerConnection JS built-in methods checks 'this'. The test has expected
failures (bug: http://webkit.org/b/158831).
- fast/mediastream/RTCPeerConnection-js-built-ins-check-this-expected.txt: Added.
- fast/mediastream/RTCPeerConnection-js-built-ins-check-this.html: Added.
- platform/mac/TestExpectations:
The mac port does not build with WEB_RTC enabled yet.
- 10:54 AM Changeset in webkit [202129] by
-
- 15 edits1 add in trunk
Forward/Back keyboard shortcuts need to flip for RTL
https://bugs.webkit.org/show_bug.cgi?id=158823
<rdar://problem/25975359>
Reviewed by Darin Adler.
Source/WebKit2:
Pipe the UserInterfaceLayoutDirection down to the WebContentProcess on initialization
and when it changes, and use it to flip the behavior of command-left and command-right
when in RTL mode.
Tests:
WebKit2_CommandBackForwardTestWKView.LTR
WebKit2_CommandBackForwardTestWKView.RTL
WebKit2_CommandBackForwardTestWKWebView.LTR
WebKit2_CommandBackForwardTestWKWebView.RTL
- Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode):
(WebKit::WebPageCreationParameters::decode):
- Shared/WebPageCreationParameters.h:
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _webViewPrintFormatter]):
(-[WKWebView setSemanticContentAttribute:]):
(-[WKWebView _printOperationWithPrintInfo:forFrame:]):
(-[WKWebView setUserInterfaceLayoutDirection:]):
- UIProcess/API/mac/WKView.mm:
(-[WKView setUserInterfaceLayoutDirection:]):
- UIProcess/Cocoa/WebViewImpl.h:
- UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::windowIsFrontWindowUnderMouse):
(WebKit::toUserInterfaceLayoutDirection):
(WebKit::WebViewImpl::userInterfaceLayoutDirection):
(WebKit::WebViewImpl::setUserInterfaceLayoutDirection):
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
(WebKit::WebPageProxy::userInterfaceLayoutDirection):
(WebKit::WebPageProxy::setUserInterfaceLayoutDirection):
- UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::hasHadSelectionChangesFromUserInteraction):
- UIProcess/ios/WKContentView.mm:
(-[WKContentView _commonInitializationWithProcessPool:configuration:]):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::m_userInterfaceLayoutDirection):
(WebKit::WebPage::setResourceCachingDisabled):
(WebKit::WebPage::setUserInterfaceLayoutDirection):
(WebKit::m_shouldDispatchFakeMouseMoveEvents): Deleted.
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/WebPage.messages.in:
- WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::performNonEditingBehaviorForSelector):
Tools:
Add tests for flipping the behavior of command-left and command-right
under RTL mode.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
- TestWebKitAPI/Tests/WebKit2Cocoa/CommandBackForward.mm: Added.
(-[CommandBackForwardOffscreenWindow isKeyWindow]):
(-[CommandBackForwardOffscreenWindow isVisible]):
(simulateCommandArrow):
(WebKit2_CommandBackForwardTest::SetUp):
(WebKit2_CommandBackForwardTestWKView::didFinishLoadForFrame):
(WebKit2_CommandBackForwardTestWKView::SetUp):
(WebKit2_CommandBackForwardTestWKView::loadFiles):
(-[CommandBackForwardNavigationDelegate webView:didFinishNavigation:]):
(WebKit2_CommandBackForwardTestWKWebView::SetUp):
(WebKit2_CommandBackForwardTestWKWebView::loadFiles):
- 10:43 AM Changeset in webkit [202128] by
-
- 2 edits in trunk/LayoutTests
[mac] LayoutTest transforms/undecomposable.html is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=158816
Another attempt to deflake this test. Or is it unflake? I don't know.
- transforms/undecomposable.html:
- 10:10 AM Changeset in webkit [202127] by
-
- 3 edits2 adds in trunk
Sporadic crash in HashTableAddResult following CSSValuePool::createFontFamilyValue
https://bugs.webkit.org/show_bug.cgi?id=158297
Reviewed by Darin Adler.
Source/WebCore:
In an effort to reduce the flash of unstyled content, we force all elements
to have display: none during an external stylesheet load. We do this by
ignoring the CSS cascade and forcing all elements to have a placeholder style
which hardcodes display: none. (This is necessary to make elements created by
script during the stylesheet load not flash.)
This style is exposed to web content via getComputedStyle(), which means it
needs to maintain the invariant that font-families can never be null strings.
We enforce this by forcing the font-family to be the standard font name.
Test: fast/text/placeholder-renderstyle-null-font.html
- style/StyleTreeResolver.cpp:
(WebCore::Style::ensurePlaceholderStyle):
LayoutTests:
- fast/text/placeholder-renderstyle-null-font-expected.txt: Added.
- fast/text/placeholder-renderstyle-null-font.html: Added.
- 9:32 AM Changeset in webkit [202126] by
-
- 3 edits in trunk/Source/WebCore
Avoid some temporary String allocations for common HTTP headers in ResourceResponse::platformLazyInit()
https://bugs.webkit.org/show_bug.cgi?id=158827
Reviewed by Darin Adler.
Add a HTTPHeaderMap::set() overload taking in a CFStringRef. The
implementation has a fast path which gets the internal characters
of the CFStringRef when possible and constructs a StringView for
it in order to call findHTTPHeaderName(). As a result, we avoid
allocating a temporary String when findHTTPHeaderName() succeeds.
This new HTTPHeaderMap::set() overload is called from both the
CF and Cocoa implementations of ResourceResponse::platformLazyInit().
I have confirmed locally on both Mac and iOS that the fast path
is used ~93% of the time. CFStringGetCStringPtr() returns null in
rare cases, causing the regular code path to be used.
- platform/network/HTTPHeaderMap.cpp:
(WebCore::HTTPHeaderMap::set):
- platform/network/HTTPHeaderMap.h:
