Timeline
Feb 8, 2015:
- 11:09 PM Changeset in webkit [179820] by
-
- 10 edits2 adds in trunk
AX: VoiceOver appears unresponsive when JavaScript alerts are triggered via focus or blur events
https://bugs.webkit.org/show_bug.cgi?id=140485
Reviewed by Anders Carlsson.
Source/WebCore:
If setting an accessibility attribute results in a modal alert being displayed, it can cause VoiceOver
to hang. A simple solution is perform the actual work after a short delay, which will ensure the call
returns without hanging.
Test: platform/mac/accessibility/setting-attributes-is-asynchronous.html
- accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
(-[WebAccessibilityObjectWrapper accessibilitySetValue:forAttribute:]):
(-[WebAccessibilityObjectWrapper _accessibilitySetValue:forAttribute:]):
Tools:
Implement takeFocus() as a way to set focus through accessibility wrappers.
- DumpRenderTree/mac/AccessibilityUIElementMac.mm:
(AccessibilityUIElement::takeFocus):
- WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
(WTR::AccessibilityUIElement::takeFocus):
LayoutTests:
Modify tests that relied on setting behavior and immediately checking results. Those
tests now need to retrieve results after a short timeout.
- accessibility/textarea-selected-text-range-expected.txt:
- accessibility/textarea-selected-text-range.html:
- platform/mac/accessibility/select-element-selection-with-optgroups.html:
- platform/mac/accessibility/setting-attributes-is-asynchronous-expected.txt: Added.
- platform/mac/accessibility/setting-attributes-is-asynchronous.html: Added.
- 7:44 PM Changeset in webkit [179819] by
-
- 10 edits8 adds in trunk
Add parsing support for CSS Selector L4's case-insensitive attribute
https://bugs.webkit.org/show_bug.cgi?id=141373
Reviewed by Darin Adler.
Source/WebCore:
This patch adds parsing for the case-insensitive attribute value
matching of CSS Selectors Level 4: http://dev.w3.org/csswg/selectors-4/#attribute-case
Excuse of a grammar: http://dev.w3.org/csswg/selectors-4/#grammar
This patch also covers serialization for CSSOM. The serialization
is defined here: http://dev.w3.org/csswg/cssom/#serializing-selectors
Matching is completely ignored in this patch. All the simple selectors
are treated as regular attribute selectors.
Tests: fast/css/parsing-css-attribute-case-insensitive-value-1.html
fast/css/parsing-css-attribute-case-insensitive-value-2.html
fast/css/parsing-css-attribute-case-insensitive-value-3.html
fast/css/parsing-css-attribute-case-insensitive-value-4.html
- css/CSSGrammar.y.in:
- css/CSSParserValues.h:
(WebCore::CSSParserSelector::setAttributeValueMatchingIsCaseInsensitive):
- css/CSSSelector.cpp:
(WebCore::CSSSelector::CSSSelector):
(WebCore::CSSSelector::selectorText):
- css/CSSSelector.h:
(WebCore::CSSSelector::CSSSelector):
(WebCore::CSSSelector::setAttributeValueMatchingIsCaseInsensitive):
(WebCore::CSSSelector::attributeValueMatchingIsCaseInsensitive):
LayoutTests:
- fast/css/css-selector-text-expected.txt:
- fast/css/css-selector-text.html:
- fast/css/css-set-selector-text-expected.txt:
- fast/css/css-set-selector-text.html:
Basic round-trip serialization through CSSOM.
- fast/css/parsing-css-attribute-case-insensitive-value-1-expected.txt: Added.
- fast/css/parsing-css-attribute-case-insensitive-value-1.html: Added.
Simple cases by themself and used in complex selectors.
- fast/css/parsing-css-attribute-case-insensitive-value-2-expected.txt: Added.
- fast/css/parsing-css-attribute-case-insensitive-value-2.html: Added.
Less simple cases, all kinds of valid syntax for case-insensitive attributes.
- fast/css/parsing-css-attribute-case-insensitive-value-3-expected.txt: Added.
- fast/css/parsing-css-attribute-case-insensitive-value-3.html: Added.
Cases that must be treated as invalid selectors.
- fast/css/parsing-css-attribute-case-insensitive-value-4-expected.txt: Added.
- fast/css/parsing-css-attribute-case-insensitive-value-4.html: Added.
Verify that invalid rules do not affect surrounding valid rules.
- 6:56 PM Changeset in webkit [179818] by
-
- 3 edits in trunk/LayoutTests
[iOS] Skip js/dom/create-lots-of-workers.html on ios-simulator
Already skipped on mac; marked as flakey on efl and gtk.
Fix tracked by:
REGRESSION: js/dom/create-lots-of-workers.html frequently crashes (sometimes in js/dom/cross-frame-bad-time.html)
<http://webkit.org/b/129758>
<rdar://problem/19760988>
- platform/ios-simulator/TestExpectations: Skip test.
- platform/mac/TestExpectations: Add comment about skipping on
ios-simulator.
- 6:42 PM Changeset in webkit [179817] by
-
- 2 edits in trunk/Source/JavaScriptCore
Remove a few duplicate propagation steps from the DFG's PredictionPropagation phase
https://bugs.webkit.org/show_bug.cgi?id=141363
Reviewed by Darin Adler.
- dfg/DFGPredictionPropagationPhase.cpp:
(JSC::DFG::PredictionPropagationPhase::propagate):
Some blocks were duplicated, they probably evolved separately
to the same state.
- 6:40 PM Changeset in webkit [179816] by
-
- 3 edits in trunk/Source/JavaScriptCore
Remove useless declarations and a stale comment from DFGByteCodeParser.h
https://bugs.webkit.org/show_bug.cgi?id=141361
Reviewed by Darin Adler.
The comment refers to the original form of the ByteCodeParser:
parse(Graph&, JSGlobalData*, CodeBlock*, unsigned startIndex);
That form is long dead, the comment is more misleading than anything.
- dfg/DFGByteCodeParser.cpp:
- dfg/DFGByteCodeParser.h:
- 6:38 PM Changeset in webkit [179815] by
-
- 3 edits in trunk/Source/JavaScriptCore
Encapsulate DFG::Plan's beforeFTL timestamp
https://bugs.webkit.org/show_bug.cgi?id=141360
Reviewed by Darin Adler.
Make the attribute private, it is an internal state.
Rename beforeFTL->timeBeforeFTL for readability.
- dfg/DFGPlan.cpp:
(JSC::DFG::Plan::compileInThread):
(JSC::DFG::Plan::compileInThreadImpl):
- dfg/DFGPlan.h:
- 6:37 PM Changeset in webkit [179814] by
-
- 2 edits in trunk/Source/JavaScriptCore
Remove DFGNode::hasArithNodeFlags()
https://bugs.webkit.org/show_bug.cgi?id=141319
Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-02-08
Reviewed by Michael Saboff.
- dfg/DFGNode.h:
(JSC::DFG::Node::hasArithNodeFlags): Deleted.
Unused code is unused.
- 5:50 PM Changeset in webkit [179813] by
-
- 2 edits in trunk/Source/WebCore
Fix CMake-based build.
- CMakeLists.txt: Added a dependency on the CMakeLists.txt itself, analogous
to the one I added in DerivedSources.make.
- 5:40 PM Changeset in webkit [179812] by
-
- 2 edits in trunk/Source/WebCore
Fix build.
- bindings/js/JSEventListener.h: Removed a call to forwardEventListeners.
- 5:25 PM Changeset in webkit [179811] by
-
- 2 edits in trunk/LayoutTests
fullscreen/full-screen-plugin.html is very flaky on Yosemite WK2
https://bugs.webkit.org/show_bug.cgi?id=141364
Reviewed by Sam Weinig.
Make the test wait for the plug-in to become available.
- fullscreen/full-screen-plugin.html:
- 5:13 PM Changeset in webkit [179810] by
-
- 88 edits4 deletes in trunk
Remove the SVG instance tree
https://bugs.webkit.org/show_bug.cgi?id=140602
Reviewed by Dean Jackson.
Source/WebCore:
- CMakeLists.txt: Removed SVGElementInstance source files.
- DerivedSources.cpp: Ditto.
- DerivedSources.make: Ditto.
- WebCore.vcxproj/WebCore.vcxproj: Ditto.
- WebCore.vcxproj/WebCore.vcxproj.filters: Ditto.
- WebCore.xcodeproj/project.pbxproj: Ditto.
- bindings/js/JSBindingsAllInOne.cpp: Ditto.
- bindings/js/JSEventListener.cpp:
(WebCore::forwardsEventListeners): Deleted. Only returned true for JSSVGElementInstance.
(WebCore::correspondingElementWrapper): Deleted. Only used for JSSVGElementInstance.
(WebCore::createJSEventListenerForAttribute): Deleted. Argument type was JSSVGElementInstance.
(WebCore::createJSEventListenerForAdd): Removed most of the code; later we can delete this entirely.
- bindings/js/JSEventListener.h: Removed the overload of createJSEventListenerForAttribute
that takes a JSSVGElementInstance.
- bindings/js/JSSVGElementInstanceCustom.cpp: Removed.
- dom/ContainerNodeAlgorithms.h: Updated comment to reflect the fact that
this code is really now only used for ContainerNode and no longer needs to
exist in a generic form.
- dom/EventTarget.h: Removed forward declaration of SVGElementInstance.
- svg/SVGElement.h: Ditto.
- dom/EventTargetFactory.in: Removed SVGElementInstance.
- svg/SVGElementInstance.cpp: Removed.
- svg/SVGElementInstance.h: Removed.
- svg/SVGElementInstance.idl: Removed.
- svg/SVGUseElement.cpp:
(WebCore::SVGUseElement::insertedInto): Removed obsolete comment.
(WebCore::SVGUseElement::instanceTreeIsLoading): Deleted. Unused
function that I forgot to delete in my last patch. It also had a
glaring mistake, a missing "return" before the recursive call to
itself that would cause it to return false when it should return true.
- svg/SVGUseElement.h: Removed instanceTreeIsLoading.
- dom/EventDispatcher.cpp: Removed include of SVGElementInstance.h.
- page/EventHandler.cpp: Ditto.
- rendering/svg/RenderSVGViewportContainer.cpp: Ditto.
- svg/SVGAElement.cpp: Ditto.
- svg/SVGAllInOne.cpp: Ditto.
- svg/SVGAnimateMotionElement.cpp: Ditto.
- svg/SVGAnimatedTypeAnimator.h: Ditto.
- svg/SVGAnimationElement.cpp: Ditto.
- svg/SVGCircleElement.cpp: Ditto.
- svg/SVGClipPathElement.cpp: Ditto.
- svg/SVGComponentTransferFunctionElement.cpp: Ditto.
- svg/SVGCursorElement.cpp: Ditto.
- svg/SVGElement.cpp: Ditto.
- svg/SVGEllipseElement.cpp: Ditto.
- svg/SVGFEBlendElement.cpp: Ditto.
- svg/SVGFEColorMatrixElement.cpp: Ditto.
- svg/SVGFECompositeElement.cpp: Ditto.
- svg/SVGFEConvolveMatrixElement.cpp: Ditto.
- svg/SVGFEDiffuseLightingElement.cpp: Ditto.
- svg/SVGFEDisplacementMapElement.cpp: Ditto.
- svg/SVGFEDropShadowElement.cpp: Ditto.
- svg/SVGFEGaussianBlurElement.cpp: Ditto.
- svg/SVGFEImageElement.cpp: Ditto.
- svg/SVGFELightElement.cpp: Ditto.
- svg/SVGFEMergeNodeElement.cpp: Ditto.
- svg/SVGFEMorphologyElement.cpp: Ditto.
- svg/SVGFEOffsetElement.cpp: Ditto.
- svg/SVGFESpecularLightingElement.cpp: Ditto.
- svg/SVGFETileElement.cpp: Ditto.
- svg/SVGFETurbulenceElement.cpp: Ditto.
- svg/SVGFilterElement.cpp: Ditto.
- svg/SVGFilterPrimitiveStandardAttributes.cpp: Ditto.
- svg/SVGForeignObjectElement.cpp: Ditto.
- svg/SVGGElement.cpp: Ditto.
- svg/SVGGradientElement.cpp: Ditto.
- svg/SVGGraphicsElement.cpp: Ditto.
- svg/SVGImageElement.cpp: Ditto.
- svg/SVGLineElement.cpp: Ditto.
- svg/SVGLinearGradientElement.cpp: Ditto.
- svg/SVGMarkerElement.cpp: Ditto.
- svg/SVGMaskElement.cpp: Ditto.
- svg/SVGPathElement.cpp: Ditto.
- svg/SVGPatternElement.cpp: Ditto.
- svg/SVGPolyElement.cpp: Ditto.
- svg/SVGRadialGradientElement.cpp: Ditto.
- svg/SVGRectElement.cpp: Ditto.
- svg/SVGSVGElement.cpp: Ditto.
- svg/SVGScriptElement.cpp: Ditto.
- svg/SVGStopElement.cpp: Ditto.
- svg/SVGSymbolElement.cpp: Ditto.
- svg/SVGTRefElement.cpp: Ditto.
- svg/SVGTextContentElement.cpp: Ditto.
- svg/SVGTextElement.cpp: Ditto.
- svg/SVGTextPathElement.cpp: Ditto.
- svg/SVGTextPositioningElement.cpp: Ditto.
Tools:
- Scripts/check-for-global-initializers: Removed special case for
SVGElementInstance.o.
LayoutTests:
Last step: Remove SVGElementInstance class itself.
- js/dom/global-constructors-attributes-expected.txt: Removed SVGElementInstance.
- platform/efl/js/dom/global-constructors-attributes-expected.txt: Ditto.
- platform/gtk/js/dom/global-constructors-attributes-expected.txt: Ditto.
- platform/ios-sim-deprecated/fast/dom/Window/window-property-descriptors-expected.txt: Ditto.
- platform/ios-sim-deprecated/fast/js/global-constructors-expected.txt: Ditto.
- platform/ios-sim-deprecated/js/dom/global-constructors-attributes-expected.txt: Ditto.
- platform/mac-mavericks/js/dom/global-constructors-attributes-expected.txt: Ditto.
- platform/mac-mountainlion/js/dom/global-constructors-attributes-expected.txt: Ditto.
- platform/mac/js/dom/global-constructors-attributes-expected.txt: Ditto.
- platform/win/js/dom/global-constructors-attributes-expected.txt: Ditto.
- svg/custom/global-constructors-expected.txt: Ditto.
- svg/custom/script-tests/global-constructors.js: Ditto.
- svg/dom/svg2-inheritance-expected.txt: Ditto.
- svg/dom/svg2-inheritance.html: Ditto.
- 4:30 PM Changeset in webkit [179809] by
-
- 2 edits in trunk/Source/WebKit2
Null deref in _clearImmediateActionState when closing a view with a DataDetectors popover open
https://bugs.webkit.org/show_bug.cgi?id=141377
<rdar://problem/19711203>
Reviewed by Darin Adler.
- UIProcess/mac/WKImmediateActionController.mm:
(-[WKImmediateActionController _clearImmediateActionState]):
We can have already detached the page when DataDetectors calls us back
in interactionStoppedHandler. While we have kept a strong reference to the
page in the interactionStoppedHandler block, _page is nulled out.
It's OK to avoid doing this work, in any case, because closing a page
tears down the TextIndicator anyway.
- 3:40 PM Changeset in webkit [179808] by
-
- 3 edits in trunk/Source/WebCore
Tweak inline playback controls to match system spec
https://bugs.webkit.org/show_bug.cgi?id=141375
<rdar://problem/19760754>
Reviewed by Sam Weinig.
Rework the UI of the inline media controls on iOS, to
better match the system specification. I've batched a
few changes into one patch because many of them are
inter-dependent, and not very aggressive. Changes are:
- updated artwork for the buttons.
- separate artwork for normal and active states.
- background images are now explicitly sized and positioned in the middle of the element, allowing audio and video to use the same glyphs even though the elements are different sizes.
- use plus-darker blend mode on the button glyphs.
- rearranged some of the rules to group things in a logical order.
- time should front-pad a "0" character, if less than 10.
- no need for an "active" class on the Airplay button (although I won't be surprised if this changes back).
- Modules/mediacontrols/mediaControlsiOS.css:
(::-webkit-media-controls):
(video::-webkit-media-controls-wireless-playback-picker-button.active): Deleted.
(audio::-webkit-media-controls-wireless-playback-picker-button.active): Deleted.
(audio::-webkit-media-controls-play-button:active): Deleted.
(audio::-webkit-media-controls-play-button.paused): Deleted.
(video::-webkit-media-controls-timeline): Deleted.
- Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.updateWirelessPlaybackStatus): No need
for the "active" class.
(ControllerIOS.prototype.formatTime): Pad with a leading zero.
- 3:22 PM Changeset in webkit [179807] by
-
- 9 edits9 adds50 deletes in trunk
Make SVGUseElement work without creating any SVGElementInstance objects
https://bugs.webkit.org/show_bug.cgi?id=141374
Reviewed by Sam Weinig.
Source/WebCore:
- dom/ElementIterator.h: Changed the * and -> operators to be const.
There is no need for the iterator itself to be modified just to dereference it.
- dom/TypedElementDescendantIterator.h: Added DoubleTypedElementDescendantIterator.
This allows callers to call descendantsOfType on two elements, as long as the caller
can guarantee that both have the same number of descendants of that type. It's handy
for walking a tree of cloned elements to set up something between each original and
its clone. In the future we might instead change the cloning machinery so it can do
this work as we clone, and if so, we could consider deleting this.
- svg/SVGElement.cpp:
(WebCore::SVGElement::correspondingElement): Made this const.
(WebCore::SVGElement::invalidateInstances): Got rid of the rule that said "this can
only be done for an element in a document", since it's useful to do this on an element
that has just been removed from a document. Removed the "updateStyleIfNeeded" call
here now that the other changes make it no longer needed. Removed an unimportant
assertion that we only invalidate use elements that are in a document; that's not
a necessary restriction. Streamlined the logic a bit.
- svg/SVGElement.h: Made correspondingElement const.
- svg/SVGUseElement.cpp:
(WebCore::SVGUseElement::insertedInto): Removed an assertion about
m_targetElementInstance since that's gone now.
(WebCore::SVGUseElement::svgAttributeChanged): Changed code that transfers
size attributes to the shadow tree to use shadowTreeTargetClone instead of
m_targetElementInstance.
(WebCore::SVGUseElement::clearResourceReferences): Removed code to detach
m_targetElementInstance, and also the call to removeAllTargetReferencesForElement,
because we no longer use those.
(WebCore::SVGUseElement::buildPendingResource): Moved the code to build the
shadow tree in here and deleted the buildShadowAndInstanceTree function.
Also changed logic so that we use a pending resource any time the target is not
a valid one. That helps us correctly handle cases where we initially have an
invalid target, but later get a value one
(WebCore::SVGUseElement::buildShadowAndInstanceTree): Deleted. The code here
was greatly simplified and moved into buildPendingResource.
(WebCore::SVGUseElement::buildInstanceTree): Deleted.
(WebCore::SVGUseElement::hasCycleUseReferencing): Deleted. Cycles are now
detected by the new isValidTarget function and so there's no need for a
separate explicit check for a cycle.
(WebCore::associateClonesWithOriginals): Added. Helper that makes
functions that build the shadow tree simpler and easier to read.
(WebCore::associateReplacementCloneWithOriginal): Added. Helper to
make associateReplacementClonesWithOriginals simple.
(WebCore::associateReplacementClonesWithOriginals): Added. Helper that
makes functions that build the shadow tree simpler and easier to read.
(WebCore::SVGUseElement::buildShadowTree): Call associateClonesWithOriginals
since associateInstancesWithShadowTreeElements no longer does this.
(WebCore::SVGUseElement::isValidTarget): Added. Covers all the different
reasons a target might not be valid: type of element, reference cycles, and
also "not in document" (refactored in here; not sure when that can happen
in practice, might be possible to remove it later).
(WebCore::SVGUseElement::expandUseElementsInShadowTree): Add checks for
documents that are still loading; this used to be checked when building the
instance tree. Added calls to associateReplacementClonesWithOriginals and
associateClonesWithOriginals; that used to be done by later in the
associateInstancesWithShadowTreeElements function. Use isValidTarget so
we handle cycles as well as invalid target types.
(WebCore::SVGUseElement::expandSymbolElementsInShadowTree): Added a call to
associateReplacementClonesWithOriginals, since we can no longer do that in
associateInstancesWithShadowTreeElements.
(WebCore::SVGUseElement::associateInstancesWithShadowTreeElements): Deleted.
(WebCore::SVGUseElement::instanceForShadowTreeElement): Deleted.
(WebCore::SVGUseElement::invalidateDependentShadowTrees): Removed a comment
that simply restated the name of the function.
- svg/SVGUseElement.h: Removed instanceForShadowTreeElement,
buildShadowAndInstanceTree, detachInstance, buildInstanceTree,
hasCycleUseReferencing, associateInstancesWithShadowTreeElements,
instanceForShadowTreeElement, and m_targetElementInstance. Added isValidTarget.
LayoutTests:
Results changed on some tests that expected the old "remove all content if a cycle is detected"
behavior from the <use> element. The new behavior is to inhibit cycles, but render everything
else, which is much easier to implement correctly and also makes logical sense. Changed all
those tests to be reference tests, which makes sense since they are focusing on what gets
rendered in these complex cases, and the expected results are a lot easier to understand in
SVG form than they were in txt/png form. This also means we can remove a lot of platform-specific
results since reference tests aren't sensitive to small platform differences in rendering.
- platform/efl/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.png: Removed.
- platform/efl/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.png: Removed.
- platform/efl/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.png: Removed.
- platform/efl/svg/hixie/error/017-expected.png: Removed.
- platform/gtk/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.png: Removed.
- platform/gtk/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.png: Removed.
- platform/gtk/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.png: Removed.
- platform/gtk/svg/custom/use-on-disallowed-foreign-object-3-expected.png: Removed.
- platform/gtk/svg/custom/use-on-disallowed-foreign-object-3-expected.txt: Removed.
- platform/gtk/svg/custom/use-recursion-1-expected.png: Removed.
- platform/gtk/svg/custom/use-recursion-1-expected.txt: Removed.
- platform/gtk/svg/custom/use-recursion-2-expected.png: Removed.
- platform/gtk/svg/custom/use-recursion-2-expected.txt: Removed.
- platform/gtk/svg/custom/use-recursion-3-expected.png: Removed.
- platform/gtk/svg/custom/use-recursion-3-expected.txt: Removed.
- platform/gtk/svg/custom/use-recursion-4-expected.png: Removed.
- platform/gtk/svg/custom/use-recursion-4-expected.txt: Removed.
- platform/gtk/svg/hixie/error/017-expected.png: Removed.
- platform/ios-sim-deprecated/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-on-disallowed-foreign-object-3-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-recursion-1-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-recursion-2-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-recursion-3-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-recursion-4-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/hixie/error/017-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-on-disallowed-foreign-object-3-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-recursion-1-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-recursion-2-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-recursion-3-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-recursion-4-expected.txt: Removed.
- platform/mac-mountainlion/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.txt: Removed.
- platform/mac-mountainlion/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.txt: Removed.
- platform/mac-mountainlion/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-on-disallowed-foreign-object-3-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-recursion-1-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-recursion-2-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-recursion-3-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-recursion-4-expected.txt: Removed.
- platform/mac-mountainlion/svg/hixie/error/017-expected.txt: Removed.
- platform/mac/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.png: Removed.
- platform/mac/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.txt: Removed.
- platform/mac/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.png: Removed.
- platform/mac/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.txt: Removed.
- platform/mac/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.png: Removed.
- platform/mac/svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.txt: Removed.
- platform/mac/svg/custom/use-on-disallowed-foreign-object-3-expected.png: Removed.
- platform/mac/svg/custom/use-on-disallowed-foreign-object-3-expected.txt: Removed.
- platform/mac/svg/custom/use-recursion-1-expected.png: Removed.
- platform/mac/svg/custom/use-recursion-1-expected.txt: Removed.
- platform/mac/svg/custom/use-recursion-2-expected.png: Removed.
- platform/mac/svg/custom/use-recursion-2-expected.txt: Removed.
- platform/mac/svg/custom/use-recursion-3-expected.png: Removed.
- platform/mac/svg/custom/use-recursion-3-expected.txt: Removed.
- platform/mac/svg/custom/use-recursion-4-expected.png: Removed.
- platform/mac/svg/custom/use-recursion-4-expected.txt: Removed.
- platform/mac/svg/hixie/error/017-expected.png: Removed.
- platform/mac/svg/hixie/error/017-expected.txt: Removed.
- svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.svg: Added. Made this be a reference test,
and made it expect more of the recursion to work.
- svg/W3C-SVG-1.2-Tiny/struct-use-recursion-01-t-expected.txt: Removed.
- svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.svg: Added. More of the same.
- svg/W3C-SVG-1.2-Tiny/struct-use-recursion-02-t-expected.txt: Removed.
- svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.svg: Added. More of the same.
- svg/W3C-SVG-1.2-Tiny/struct-use-recursion-03-t-expected.txt: Removed.
- svg/custom/use-on-disallowed-foreign-object-3-expected.svg: Added. More of the same.
- svg/custom/use-recursion-1-expected.svg: Added. More of the same.
- svg/custom/use-recursion-2-expected.svg: Added. More of the same.
- svg/custom/use-recursion-3-expected.svg: Added. More of the same.
- svg/custom/use-recursion-4-expected.svg: Added. More of the same.
- svg/hixie/error/017-expected.txt: Removed.
- svg/hixie/error/017-expected.xml: Added. More of the same.
- svg/in-html/defs-after-use.html: Updated incorrect bug number in this test.
- 3:21 PM Changeset in webkit [179806] by
-
- 3 edits in trunk/LayoutTests
Skip fast/parser/document-open-in-unload.html on all WK2 platforms
Tracked by:
[WK2] fast/parser/document-open-in-unload.html makes the following test crash
<http://webkit.org/b/98345>
- platform/mac-wk2/TestExpectations: Move Skip expectation from here...
- platform/wk2/TestExpectations: ...to here with updated bug number.
- 3:20 PM Changeset in webkit [179805] by
-
- 2 edits in trunk/LayoutTests
Update section headers for mac-wk2/TestExpectations
- platform/mac-wk2/TestExpectations:
- 2:06 PM Changeset in webkit [179804] by
-
- 12 edits3 adds in trunk/Source
[WK2] Add logging to validate the network cache efficacy (Part 1)
https://bugs.webkit.org/show_bug.cgi?id=141269
<rdar://problem/19632080>
Reviewed by Antti Koivisto.
Source/WebCore:
Export an extra symbol.
- WebCore.exp.in:
Source/WebKit2:
Add console logging to validate the network cache efficacy. This will
tell us if how the network cache satisties requests, in particular:
- Request cannot be handled by the cache
- Entry was not in the cache but is no longer there (pruned)
- Entry is in the cache but is not usable
- Entry is in the cache and is used.
This patch introduces a SQLite-based network cache statistics storage
that is used to store requests we have seen before, and query if we
have seen a request before. The storage is lightweight as it only
stores hashes in the database, in a background thread.
The statistics cache is initially bootstapped from the network disk
cache so that we have data initially and get as accurate statistics
as possible from the start.
To maintain an acceptable level of performance, we have a hard limit
on the number of unique requests that are retained set to 100000.
Diagnostic logging for this will be added in a follow-up patch.
- 12:22 PM Changeset in webkit [179803] by
-
- 2 edits in trunk/LayoutTests
[iOS] Gardening: Some MathML tests crash in RenderMathMLOperator::advanceForGlyph() or boundsForGlyph()
Tracked by: <http://webkit.org/b/141371>
- platform/ios-simulator-wk2/TestExpectations: Mark tests as
crashing.
- 12:22 PM Changeset in webkit [179802] by
-
- 2 edits in trunk/LayoutTests
REGRESSION (r179391): Remove references to deleted SVG tests
Fixes the following lint warnings:
--lint-test-files warnings:
LayoutTests/platform/ios-simulator-wk2/TestExpectations:412 Path does not exist. svg/custom/use-elementInstance-event-target.svg
LayoutTests/platform/ios-simulator-wk2/TestExpectations:413 Path does not exist. svg/custom/use-elementInstance-methods.svg
LayoutTests/platform/ios-simulator-wk2/TestExpectations:417 Path does not exist. svg/custom/use-instanceRoot-event-listeners.xhtml
- platform/ios-simulator-wk2/TestExpectations: Remove deleted
tests.
- 12:00 PM Changeset in webkit [179801] by
-
- 1 edit in trunk/LayoutTests/ChangeLog
Removing an accidentally committed ChangeLog.
- 11:46 AM TestExpectations edited by
- Fix new-run-webkit-tests to just run-webkit-tests. (diff)
- 11:33 AM Changeset in webkit [179800] by
-
- 1 edit1 add in trunk/LayoutTests
AX: The input element with type="search" has no default focus outline
https://bugs.webkit.org/show_bug.cgi?id=140326
Adding results for Mavericks.
- platform/mac-mavericks/fast/css/focus-ring-exists-for-search-field-expected.txt: Added.
- 11:17 AM Changeset in webkit [179799] by
-
- 2 edits in trunk/LayoutTests
http/tests/security/appcache-in-private-browsing.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=141370
- TestExpectations: Marking as such.
- 11:11 AM Changeset in webkit [179798] by
-
- 3 edits in trunk/LayoutTests
Application cache abort() tests are flaky
https://bugs.webkit.org/show_bug.cgi?id=87633
Moved expectations form Efl to root TestExpectations file. Also, removed Crash
expectation, as no bot hits that now.
Removed an erroneously added expectation for abort-cache-onchecking-resource-404.html -
this test doesn't happen to fail on the bots, although it also doesn't appear to
be very robust.
- TestExpectations:
- platform/efl/TestExpectations:
- 10:47 AM Changeset in webkit [179797] by
-
- 2 edits in trunk/LayoutTests
http/tests/appcache/abort-cache-onchecking-resource-404.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=141368
- TestExpectations: This test is intrinsically unreliable, but at least it checks
that there is no crash is any of the code paths that it takes.
- 12:02 AM Changeset in webkit [179796] by
-
- 3 edits3 adds in trunk
AX: The input element with type="search" has no default focus outline
https://bugs.webkit.org/show_bug.cgi?id=140326
Reviewed by Darin Adler.
Source/WebCore:
The platform RenderTheme takes care of the search field, and that code
was missing a check for whether the element was focused.
Test: fast/css/focus-ring-exists-for-search-field.html
- rendering/RenderThemeMac.mm:
(WebCore::RenderThemeMac::paintSearchField):
LayoutTests:
- fast/css/focus-ring-exists-for-search-field.html: Added.
- platform/mac/fast/css/focus-ring-exists-for-search-field-expected.png: Added.
- platform/mac/fast/css/focus-ring-exists-for-search-field-expected.txt: Added.
Feb 7, 2015:
- 11:54 PM Changeset in webkit [179795] by
-
- 2 edits in trunk/Source/WTF
[ARM] GC sometimes stuck in an infinite loop if parallel GC is enabled
https://bugs.webkit.org/show_bug.cgi?id=141290
Reviewed by Anders Carlsson.
- wtf/Platform.h:
- 9:24 PM Changeset in webkit [179794] by
-
- 2 edits in trunk/LayoutTests
fullscreen/full-screen-plugin.html is very flaky on Yosemite WK2
https://bugs.webkit.org/show_bug.cgi?id=141364
Reviewed by Zalan Bujtas.
Speculative fix. Make sure that the plug-in has loaded before using it.
- fullscreen/full-screen-plugin.html:
- 7:26 PM Changeset in webkit [179793] by
-
- 3 edits in trunk/Tools
[iOS] run-webkit-tests fails due to simulator devices from previous SDK installs being marked as unavailable
<http://webkit.org/b/141365>
Reviewed by Daniel Bates.
- Scripts/webkitpy/xcode/simulator.py:
(Simulator): Add unavailable_version_re precompiled regex.
(Simulator._parse_devices): Check for unavailable versions and
ignore them if found when parsing the output of
xcrun simctl list.
- Scripts/webkitpy/xcode/simulator_unittest.py:
(test_unavailable_devices): Add test with output from
xcrun simctl listwith unavailable runtimes that fails before
the fix.
- 6:43 PM Changeset in webkit [179792] by
-
- 7 edits12 adds in trunk
Add some dictionary lookup tests
https://bugs.webkit.org/show_bug.cgi?id=141355
Reviewed by Darin Adler.
Tests: platform/mac/editing/dictionary-lookup/dictionary-lookup-input.html
platform/mac/editing/dictionary-lookup/dictionary-lookup-inside-selection.html
platform/mac/editing/dictionary-lookup/dictionary-lookup-outside-selection.html
platform/mac/editing/dictionary-lookup/dictionary-lookup-rtl.html
platform/mac/editing/dictionary-lookup/dictionary-lookup.html
- WebCore.exp.in:
Remove an unneeded export.
- editing/mac/DictionaryLookup.h:
Use OBJC_CLASS instead of @class so that this can be included in pure-C++ files.
- testing/Internals.cpp:
(WebCore::Internals::rangeForDictionaryLookupAtLocation):
- testing/Internals.h:
- testing/Internals.idl:
Expose rangeForDictionaryLookupAtHitTestResult fairly directly to JavaScript.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-expected.txt: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-input-expected.txt: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-input.html: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-inside-selection-expected.txt: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-inside-selection.html: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-outside-selection-expected.txt: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-outside-selection.html: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-rtl-expected.txt: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup-rtl.html: Added.
- platform/mac/editing/dictionary-lookup/dictionary-lookup.html: Added.
- platform/mac/editing/dictionary-lookup/lookup-test.js: Added.
(runTest):
Add tests for various cases that we've had trouble with in the past.
- 6:28 PM Changeset in webkit [179791] by
-
- 23 edits in trunk
Add Vector::removeFirstMatching() / removeAllMatching() methods taking lambda functions
https://bugs.webkit.org/show_bug.cgi?id=141321
Reviewed by Darin Adler.
Source/JavaScriptCore:
Use new Vector::removeFirstMatching() / removeAllMatching() methods.
Source/WebCore:
Use new Vector::removeFirstMatching() / removeAllMatching() methods.
Source/WebKit/win:
Use new Vector::removeFirstMatching() / removeAllMatching() methods.
Source/WebKit2:
Use new Vector::removeFirstMatching() / removeAllMatching() methods.
Source/WTF:
Add Vector::removeFirstMatching() / removeAllMatching() methods taking
lambda functions to match the element(s) to remove. This simplifies the
code a bit. Vector::removeAllMatching() is also more efficient than the
manual removal alternative.
- wtf/Vector.h:
Tools:
Use new Vector::removeFirstMatching() / removeAllMatching() methods.
- 5:31 PM Changeset in webkit [179790] by
-
- 1 edit8 deletes in trunk/LayoutTests
Unreviewed gardening.
Remove svg/custom/use-events-crash.svg. It has no value anymore.
See webkit.org/b/141108
- platform/gtk/svg/custom/use-events-crash-expected.png: Removed.
- platform/gtk/svg/custom/use-events-crash-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-events-crash-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-events-crash-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-events-crash-expected.txt: Removed.
- platform/mac/svg/custom/use-events-crash-expected.png: Removed.
- platform/mac/svg/custom/use-events-crash-expected.txt: Removed.
- svg/custom/use-events-crash.svg: Removed.
- 4:53 PM Changeset in webkit [179789] by
-
- 2 edits in trunk/LayoutTests
http/tests/xmlhttprequest/event-listener-gc.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=33342
Reviewed by Darin Adler.
This test relies on a zero-delay timer being a lot faster than fetching from network.
Force a layout before starting the test, because otherwise, the layout can significantly
delay the timer (I've seen 50-60 milliseconds being a common delay in debug builds).
Also, changed the resource URL to avoid Apache error log spew.
- http/tests/xmlhttprequest/event-listener-gc.html:
- 4:44 PM Changeset in webkit [179788] by
-
- 4 edits1 add in trunk/Tools
[iOS] Make Simulator class testable
<http://webkit.org/b/141358>
Rubber-stamped by Darin Adler.
- Scripts/webkitpy/common/system/platforminfo.py:
(PlatformInfo.xcode_simctl_list): Move
xcrun simctl list
command to here from Simulator.refresh() in xcode/simulator.py
so that the output of the command can be mocked.
- Scripts/webkitpy/common/system/platforminfo_mock.py:
(MockPlatformInfo.init): Set self.expected_xcode_simctl_list
to None.
(MockPlatformInfo.xcode_simctl_list): Add method that returns
self.expected_xcode_simctl_list expectation.
- Scripts/webkitpy/xcode/simulator.py: Add missing copyright
and license header.
(Simulator.init): Add optional 'host' parameter to make it
possible to pass in a mock object for testing. Set self._host
to 'host' parameter or create Host() object.
(Simulator.refresh): Call new PlatformInfo.xcode_simctl_list()
method.
- Scripts/webkitpy/xcode/simulator_unittest.py: Add unit test
for current code.
(SimulatorTest):
(SimulatorTest.setUp):
(SimulatorTest._set_expected_xcrun_simctl_list):
(SimulatorTest.test_simulator_device_types):
(test_invalid_device_types_header):
(test_invalid_runtimes_header):
(test_invalid_devices_header):
- 4:32 PM Changeset in webkit [179787] by
-
- 4 edits in trunk/LayoutTests
fast/images/animated-gif-iframe-webkit-transform.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=141323
Reviewed by Alexey Proskuryakov.
Use shouldBecomeEqual() instead of shouldBe() to check for initial test
conditions. This should address the flakiness.
- fast/images/animated-gif-iframe-webkit-transform-expected.txt:
- fast/images/animated-gif-iframe-webkit-transform.html:
- platform/mac/TestExpectations:
- 3:46 PM Changeset in webkit [179786] by
-
- 8 edits1 add in trunk/Source/WebKit2
Add API::HistoryClient and split some things out of API::NavigationClient
https://bugs.webkit.org/show_bug.cgi?id=141264
Reviewed by Darin Adler.
- UIProcess/API/APIHistoryClient.h: Added.
(API::HistoryClient::~HistoryClient):
(API::HistoryClient::didNavigateWithNavigationData):
(API::HistoryClient::didPerformClientRedirect):
(API::HistoryClient::didPerformServerRedirect):
(API::HistoryClient::didUpdateHistoryTitle):
- UIProcess/API/APINavigationClient.h:
(API::NavigationClient::didNavigateWithNavigationData): Deleted.
(API::NavigationClient::didPerformClientRedirect): Deleted.
(API::NavigationClient::didPerformServerRedirect): Deleted.
(API::NavigationClient::didUpdateHistoryTitle): Deleted.
- UIProcess/Cocoa/NavigationState.h:
- UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::createHistoryClient):
(WebKit::NavigationState::HistoryClient::HistoryClient):
(WebKit::NavigationState::HistoryClient::~HistoryClient):
(WebKit::NavigationState::HistoryClient::didNavigateWithNavigationData):
(WebKit::NavigationState::HistoryClient::didPerformClientRedirect):
(WebKit::NavigationState::HistoryClient::didPerformServerRedirect):
(WebKit::NavigationState::HistoryClient::didUpdateHistoryTitle):
(WebKit::NavigationState::NavigationClient::didNavigateWithNavigationData): Deleted.
(WebKit::NavigationState::NavigationClient::didPerformClientRedirect): Deleted.
(WebKit::NavigationState::NavigationClient::didPerformServerRedirect): Deleted.
(WebKit::NavigationState::NavigationClient::didUpdateHistoryTitle): Deleted.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::setHistoryClient):
(WebKit::WebPageProxy::didNavigateWithNavigationData):
(WebKit::WebPageProxy::didPerformClientRedirect):
(WebKit::WebPageProxy::didPerformServerRedirect):
(WebKit::WebPageProxy::didUpdateHistoryTitle):
- UIProcess/WebPageProxy.h:
Add a API::HistoryClient and move the few things that belong on it out of API::NavigationClient.
Adjust accordingly in WebPageProxy and NavigationState.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
(-[WKWebView setNavigationDelegate:]):
(-[WKWebView setUIDelegate:]):
(-[WKWebView _setHistoryDelegate:]):
Lazily push the NavigationState/UIDelegate clients down to WebPageProxy upon
installation of a delegate, so that alternative (C SPI) delegate setters can
be separately created.
- WebKit2.xcodeproj/project.pbxproj:
- 3:17 PM Changeset in webkit [179785] by
-
- 7 edits2 adds12 deletes in trunk
Source/WebCore:
Stop dispatching events to with SVGElementInstance objects as their targets
https://bugs.webkit.org/show_bug.cgi?id=141108
Reviewed by Anders Carlsson.
Test: svg/custom/use-event-retargeting.html
- dom/EventDispatcher.cpp:
(WebCore::eventTargetRespectingTargetRules): Replaced the code that retargeted
events at SVGElementInstance objects with code that retargets them at the use
element instead. Also wrote the code in a simpler way.
LayoutTests:
Stop dispatching events with SVGElementInstance objects as their targets
https://bugs.webkit.org/show_bug.cgi?id=141108
Reviewed by Anders Carlsson.
Many tests are no longer relevant once we aren't doing this any more.
- platform/gtk/svg/custom/use-instanceRoot-event-bubbling-expected.png: Removed.
- platform/gtk/svg/custom/use-instanceRoot-modifications-expected.png: Removed.
- platform/gtk/svg/custom/use-instanceRoot-modifications-expected.txt: Removed.
- platform/ios-sim-deprecated/svg/custom/use-instanceRoot-modifications-expected.txt: Removed.
- platform/ios-simulator/svg/custom/use-instanceRoot-modifications-expected.txt: Removed.
- platform/mac-mountainlion/svg/custom/use-instanceRoot-modifications-expected.txt: Removed.
- platform/mac/svg/custom/use-instanceRoot-event-bubbling-expected.png: Removed.
- platform/mac/svg/custom/use-instanceRoot-modifications-expected.png: Removed.
- platform/mac/svg/custom/use-instanceRoot-modifications-expected.txt: Removed.
- svg/custom/use-instanceRoot-modifications.svg: Removed.
- svg/custom/use-instanceRoot-with-use-removed-expected.txt: Removed.
- svg/custom/use-instanceRoot-with-use-removed.svg: Removed.
- svg/custom/resources/use-instanceRoot-event-bubbling.js: Updated this test to expect
the events to be dispatched with the SVGUseElement as the target. I talked this over with
Sam Weinig and we decided this is good behavior for now, and it almost matches what the
spec says. Might be worth refining later.
- svg/custom/use-instanceRoot-event-bubbling-expected.txt: Updated expected results.
- svg/custom/use-instanceRoot-event-bubbling.xhtml: Tweaked the test a little. It still
could use improvement; it's like half a "repaint test", which is strange.
- svg/custom/use-event-retargeting-expected.txt: Added. Got this test from Blink.
- svg/custom/use-event-retargeting.html: Added. Ditto.
- svg/custom/use-events-crash.svg: Added some more events, a second click, so that we
don't hang with the context menu up when running this. Also converted line endings to
use LF instead of CRLF.
- 2:48 PM Changeset in webkit [179784] by
-
- 6 edits in trunk/Source/WebCore
[Mac] Set -contentsScale on AVPlayerLayer to allow AVPlayer to select the appropriate HLS variant.
https://bugs.webkit.org/show_bug.cgi?id=141354
rdar://problem/19717591
Reviewed by Darin Adler.
AVPlayer will try to determine the correct HLS variant based on the bounds of an AVPlayerLayer.
When not in a layer tree, AVFoundation is not able to determine the correct mapping from logical
units to pixel values. To provide AVPlayer with that scaling value, set -contentsScale based on
both the current device scale and the current page scale.
Since this needs to be set at initialization time, before the AVPlayer is has any AVPlayerItems,
add some plumbing up from MediaPlayer to as the HTMLMediaElement for the appropriate contents
scale.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::mediaPlayerContentsScale):
- html/HTMLMediaElement.h:
- platform/graphics/MediaPlayer.h:
(WebCore::MediaPlayerClient::mediaPlayerContentsScale):
- platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::createAVPlayerLayer):
- platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::GraphicsLayerCA::updateContentsScale):
- 12:21 PM Changeset in webkit [179783] by
-
- 2 edits in trunk/Source/WebCore
ASan complains about plugins/snapshotting/snapshot-plugin-not-quite-blocked-by-image.html
https://bugs.webkit.org/show_bug.cgi?id=141352
rdar://problem/19717490
Reviewed by Anders Carlsson.
- dom/Document.cpp: (WebCore::Document::ensurePlugInsInjectedScript): This string
is not null terminated.
- 11:50 AM Changeset in webkit [179782] by
-
- 2 edits in trunk/Source/WebKit2
And as a further followup restore the 8bit test too.
- NetworkProcess/cache/NetworkCacheKey.cpp:
(WebKit::hashString):
- 11:45 AM Changeset in webkit [179781] by
-
- 4 edits in trunk/Source/WebKit2
Use longer hashes for cache keys
https://bugs.webkit.org/show_bug.cgi?id=141356
Rubber-stamped by Darin Adler.
Folloup and build fix.
- NetworkProcess/cache/NetworkCacheCoders.h:
- NetworkProcess/cache/NetworkCacheKey.cpp:
(WebKit::hashString):
Use containsOnlyASCII instead of is8Bit so both paths always compute the same hash.
- NetworkProcess/cache/NetworkCacheKey.h:
- 11:19 AM Changeset in webkit [179780] by
-
- 2 edits in trunk/Source/WebKit2
Remove a printf.
- NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::retrieve):
- 11:18 AM Changeset in webkit [179779] by
-
- 8 edits in trunk/Source/WebKit2
Use longer hashes for cache keys
https://bugs.webkit.org/show_bug.cgi?id=141356
Reviewed by Darin Adler.
The current key hashes are 32bit. We should use longer hashes to eliminate collisions.
This patch switches us to using MD5 digests for the cache key hashes. As a result the file names for the cache
entries grow from 8 to 32 character.
Note that we don't need a cryptographic hash (full cache keys are verified against the entries).
MD5 just happens to be fast, convenient and available.
The patch also moves the whole cache hierarchy down to a versioned subdirectory ("WebKitCache/Version 2")
and deletes any old style cache files if they exist.
- NetworkProcess/cache/NetworkCacheCoders.cpp:
(WebKit::NetworkCacheCoder<MD5::Digest>::encode):
(WebKit::NetworkCacheCoder<MD5::Digest>::decode):
- NetworkProcess/cache/NetworkCacheCoders.h:
- NetworkProcess/cache/NetworkCacheKey.cpp:
(WebKit::NetworkCacheKey::NetworkCacheKey):
(WebKit::hashString):
(WebKit::NetworkCacheKey::computeHash):
(WebKit::NetworkCacheKey::hashAsString):
(WebKit::NetworkCacheKey::stringToHash):
- NetworkProcess/cache/NetworkCacheKey.h:
(WebKit::NetworkCacheKey::shortHash):
(WebKit::NetworkCacheKey::toShortHash):
32bit hash to use in the bloom filter.
(WebKit::NetworkCacheKey::hashStringLength):
- NetworkProcess/cache/NetworkCacheStorage.h:
Bump the version.
- NetworkProcess/cache/NetworkCacheStorageCocoa.mm:
(WebKit::traverseCacheFiles):
(WebKit::makeVersionedDirectoryPath):
(WebKit::NetworkCacheStorage::NetworkCacheStorage):
(WebKit::NetworkCacheStorage::initialize):
(WebKit::NetworkCacheStorage::removeEntry):
(WebKit::NetworkCacheStorage::retrieve):
(WebKit::NetworkCacheStorage::store):
(WebKit::NetworkCacheStorage::update):
(WebKit::NetworkCacheStorage::shrinkIfNeeded):
(WebKit::NetworkCacheStorage::deleteOldVersions):
Wipe out the version 1 cache.
- 9:01 AM Changeset in webkit [179778] by
-
- 2 edits in trunk/Tools
run-jsc-stress-tests --remote should use the default ssh port
https://bugs.webkit.org/show_bug.cgi?id=141287
Reviewed by Darin Adler.
- Scripts/run-jsc-stress-tests: Extended URI module to be able to handle ssh scheme with the default 22 port number.
- 12:32 AM Changeset in webkit [179777] by
-
- 2 edits in trunk/LayoutTests
REGRESSION (OS X 10.10.2): http/tests/media/video-query-url.html frequently times out
https://bugs.webkit.org/show_bug.cgi?id=141085
- platform/mac-wk2/TestExpectations: Added an expectation.