⚠ Archived content — this site is no longer maintained.   Current WebKit documentation is at docs.webkit.org.

Timeline



Apr 30, 2017:

10:02 PM Changeset in webkit [216004] by jmarcell@apple.com
  • 7 edits in trunk/Source

Versioning.

9:58 PM Changeset in webkit [216003] by jmarcell@apple.com
  • 1 copy in tags/Safari-604.1.20

Tag Safari-604.1.20.

5:44 PM Changeset in webkit [216002] by Chris Dumez
  • 18 edits
    6 deletes in trunk/Source

Drop support for legacy ClientRect / ClientRectList
​https://bugs.webkit.org/show_bug.cgi?id=171418

Reviewed by Sam Weinig.

Source/WebCore:

Drop support for legacy ClientRect / ClientRectList types now we use
DOMRect everywhere.

  • CMakeLists.txt:
  • DerivedSources.cpp:
  • DerivedSources.make:
  • WebCore.xcodeproj/project.pbxproj:
  • dom/ClientRect.cpp: Removed.
  • dom/ClientRect.h: Removed.
  • dom/ClientRect.idl: Removed.
  • dom/ClientRectList.cpp: Removed.
  • dom/ClientRectList.h: Removed.
  • dom/ClientRectList.idl: Removed.
  • dom/DOMAllInOne.cpp:
  • dom/Element.cpp:
  • html/HTMLMediaElement.cpp:
  • html/track/VTTRegion.cpp:
  • page/DragController.cpp:

Source/WebInspectorUI:

Drop ClientRectList-related code.

  • UserInterface/Models/NativeFunctionParameters.js:

Source/WebKit2:

GTK build fix.

  • WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMClientRect.cpp:

(WebKit::kit):
(WebKit::core):
(WebKit::wrapClientRect):
(webkit_dom_client_rect_constructed):

  • WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMClientRectList.cpp:

(ClientRectList::create):
(ClientRectList::items):
(ClientRectList::ClientRectList):

  • WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMClientRectListPrivate.h:
  • WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMClientRectPrivate.h:
  • WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMElement.cpp:

(webkit_dom_element_get_bounding_client_rect):
(webkit_dom_element_get_client_rects):

5:15 PM Changeset in webkit [216001] by Alan Bujtas
  • 5 edits
    2 adds in trunk

Initialize a new layout state while bailing out of simple line layout only when needed.
​https://bugs.webkit.org/show_bug.cgi?id=171481
<rdar://problem/31906382>

Source/WebCore:

'-webkit-border-fit: lines' requires inline box tree.
We normally bail out of simple line layout early when we see this property.
When this property is dynamically set and we already have a simple line layout context,
RenderBlockFlow::ensureLineBoxes() takes care of the switch by calling layoutLineBoxes().
This patch makes sure that we create the layout state for this (forced)line layout only when
we don't yet have one.

Reviewed by Antti Koivisto.

Test: fast/multicol/simple-line-layout-switch-to-normal-layout-while-inside-layout.html

  • rendering/RenderBlockFlow.cpp:

(WebCore::PaginatedLayoutStateMaintainer::PaginatedLayoutStateMaintainer):
(WebCore::PaginatedLayoutStateMaintainer::~PaginatedLayoutStateMaintainer):
(WebCore::RenderBlockFlow::ensureLineBoxes):

  • rendering/RenderView.cpp:

(WebCore::RenderView::pushLayoutStateForPaginationIfNeeded):
(WebCore::RenderView::pushLayoutStateForPagination): Deleted.

  • rendering/RenderView.h:

LayoutTests:

Reviewed by Antti Koivisto.

  • fast/multicol/simple-line-layout-switch-to-normal-layout-while-inside-layout-expected.txt: Added.
  • fast/multicol/simple-line-layout-switch-to-normal-layout-while-inside-layout.html: Added.
4:39 PM Changeset in webkit [216000] by mitz@apple.com
  • 2 edits in trunk/Tools

[Cocoa] Have check-webkit-style advise against use of [get…Class() alloc]
​https://bugs.webkit.org/show_bug.cgi?id=171486

Reviewed by Sam Weinig.

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

(check_soft_link_class_alloc): Added. Looks for [get…Class() alloc] and suggests

alloc…Instance() instead.

(check_style): Invoke new check.

2:24 PM Changeset in webkit [215999] by jmarcell@apple.com
  • 2 edits in tags/Safari-604.1.19.1/Source/WebCore

Cherry-pick r215990. rdar://problem/31906525

2:23 PM Changeset in webkit [215998] by jmarcell@apple.com
  • 7 edits in tags/Safari-604.1.19.1/Source

Versioning.

2:19 PM Changeset in webkit [215997] by beidson@apple.com
  • 4 edits in trunk

More fixing after r215991
​https://bugs.webkit.org/show_bug.cgi?id=171483

Source/WebKit2:

  • NetworkProcess/cocoa/NetworkProcessCocoa.mm:

(WebKit::NetworkProcess::syncAllCookies):

Tools:

  • TestWebKitAPI/Tests/WebKit2Cocoa/WebsiteDataStoreCustomPaths.mm:

(TEST):

2:14 PM Changeset in webkit [215996] by jmarcell@apple.com
  • 1 copy in tags/Safari-604.1.19.1

New tag.

2:04 PM Changeset in webkit [215995] by mitz@apple.com
  • 2 edits in trunk/Source/WebCore

Tried to fix the iOS build after r215992

  • platform/ios/WebVideoFullscreenControllerAVKit.mm:
1:55 PM Changeset in webkit [215994] by beidson@apple.com
  • 4 edits in trunk/Source

Fix r215991
​https://bugs.webkit.org/show_bug.cgi?id=171483

Unreviewed.

Source/WebCore:

  • platform/spi/cf/CFNetworkSPI.h:

Source/WebKit2:

  • NetworkProcess/cocoa/NetworkProcessCocoa.mm:

(WebKit::NetworkProcess::syncAllCookies):

1:53 PM Changeset in webkit [215993] by mitz@apple.com
  • 2 edits in trunk/Source/WebCore

Tried to fix the iOS build after r215992

  • platform/ios/WebVideoFullscreenInterfaceAVKit.mm:

(WebAVPlayerLayerView_pictureInPicturePlayerLayerView):
(WebVideoFullscreenInterfaceAVKit::setupFullscreen):

1:47 PM Changeset in webkit [215992] by mitz@apple.com
  • 13 edits in trunk/Source

[Cocoa] Replaces uses of [get…() alloc] with alloc…Instance()
​https://bugs.webkit.org/show_bug.cgi?id=171485

Reviewed by Geoffrey Garen.

Source/WebCore:

  • platform/graphics/avfoundation/objc/CDMSessionAVContentKeySession.mm:

(WebCore::CDMSessionAVContentKeySession::contentKeySession):

  • platform/graphics/avfoundation/objc/MediaPlaybackTargetPickerMac.mm:

(WebCore::MediaPlaybackTargetPickerMac::devicePicker):

  • platform/ios/ValidationBubbleIOS.mm:

(-[WebValidationBubbleTapRecognizer initWithPopoverController:]):
(WebCore::ValidationBubble::ValidationBubble):

  • platform/ios/WebVideoFullscreenControllerAVKit.mm:

(WebVideoFullscreenControllerContext::setUpFullscreen):

  • platform/ios/WebVideoFullscreenInterfaceAVKit.mm:

(WebAVPlayerLayerView_pictureInPicturePlayerLayerView):
(WebVideoFullscreenInterfaceAVKit::setupFullscreen):

  • platform/mac/WebVideoFullscreenInterfaceMac.mm:

(-[WebVideoFullscreenInterfaceMacObjC setUpPIPForVideoView:withFrame:inWindow:]):

  • platform/network/ios/PreviewConverter.mm:

(WebCore::PreviewConverter::PreviewConverter):

  • rendering/RenderThemeIOS.mm:

(WebCore::iconForAttachment):

Source/WebKit/mac:

  • WebView/WebView.mm:

(-[WebUITextIndicatorData initWithImage:textIndicatorData:scale:]):
(-[WebUITextIndicatorData initWithImage:scale:]):

Source/WebKit2:

  • Platform/mac/MenuUtilities.mm:

(WebKit::menuItemForTelephoneNumber):
(WebKit::menuForTelephoneNumber):

1:24 PM Changeset in webkit [215991] by beidson@apple.com
  • 23 edits in trunk

Updates to _WKWebsiteDataStoreConfiguration cookie storage location SPI.
<rdar://problem/31906397> and ​https://bugs.webkit.org/show_bug.cgi?id=171483

Reviewed by Geoff Garen (and kind of Andy Estes).
Source/WebCore:

No new tests (Covered by API test).

  • platform/spi/cf/CFNetworkSPI.h:

Source/WebKit2:

Updates include:
-Taking a file path as the config argument instead of a directory.
-Proper sandbox extension handling.
-Proper and thorough API test.

  • NetworkProcess/NetworkProcess.cpp:

(WebKit::NetworkProcess::syncAllCookies):

  • NetworkProcess/NetworkProcess.h:
  • NetworkProcess/NetworkProcess.messages.in:
  • NetworkProcess/cocoa/NetworkProcessCocoa.mm:

(WebKit::NetworkProcess::syncAllCookies):

  • NetworkProcess/mac/RemoteNetworkingContext.mm:

(WebKit::RemoteNetworkingContext::ensureWebsiteDataStoreSession):

  • Shared/WebsiteDataStoreParameters.cpp:

(WebKit::WebsiteDataStoreParameters::encode):
(WebKit::WebsiteDataStoreParameters::decode):

  • Shared/WebsiteDataStoreParameters.h:
  • UIProcess/API/Cocoa/WKProcessPool.mm:

(-[WKProcessPool _syncNetworkProcessCookies]):

  • UIProcess/API/Cocoa/WKProcessPoolPrivate.h:
  • UIProcess/API/Cocoa/WKWebsiteDataStore.mm:

(-[WKWebsiteDataStore _initWithConfiguration:]):

  • UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.h:
  • UIProcess/API/Cocoa/_WKWebsiteDataStoreConfiguration.mm:

(-[_WKWebsiteDataStoreConfiguration _cookieStoragePath]):
(-[_WKWebsiteDataStoreConfiguration _setCookieStoragePath:]):
(-[_WKWebsiteDataStoreConfiguration _cookieStorageDirectory]): Deleted.
(-[_WKWebsiteDataStoreConfiguration _setCookieStorageDirectory:]): Deleted.

  • UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::syncNetworkProcessCookies):

  • UIProcess/WebProcessPool.h:
  • UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:

(WebKit::WebsiteDataStore::parameters):

  • UIProcess/WebsiteData/WebsiteDataStore.cpp:

(WebKit::WebsiteDataStore::resolveDirectoriesIfNecessary):

  • UIProcess/WebsiteData/WebsiteDataStore.h:

(WebKit::WebsiteDataStore::resolvedCookieStoragePath):
(WebKit::WebsiteDataStore::resolvedCookieStorageDirectory): Deleted.

Tools:

  • TestWebKitAPI/Tests/WebKit2Cocoa/WebsiteDataStoreCustomPaths.mm:

(TEST):

12:52 PM Changeset in webkit [215990] by mitz@apple.com
  • 2 edits in trunk/Source/WebCore

<rdar://problem/31906525> WebCore-7604.1.19 has failed to build

  • platform/ios/WebItemProviderPasteboard.mm:

(-[WebItemProviderPasteboard setItemsUsingRegistrationInfoLists:]): Resolved type ambiguity

by using allocUIItemProviderInstance() instead of [getUIItemProviderClass() alloc].

11:12 AM Changeset in webkit [215989] by jdiggs@igalia.com
  • 8 edits
    2 adds in trunk

[ATK] Implement support for aria-autocomplete
​https://bugs.webkit.org/show_bug.cgi?id=171167

Reviewed by Chris Fleizach.

Source/WebCore:

Expose ATK_STATE_SUPPORTS_AUTOCOMPLETION in the state set, and the value
of aria-autocomplete as an object attribute, for elements which indicate
they support autocompletion.

Test: accessibility/gtk/aria-autocomplete.html

  • accessibility/AccessibilityObject.cpp:

(WebCore::AccessibilityObject::supportsARIAAutoComplete):
(WebCore::AccessibilityObject::ariaAutoCompleteValue):

  • accessibility/AccessibilityObject.h:
  • accessibility/atk/WebKitAccessibleWrapperAtk.cpp:

(webkitAccessibleGetAttributes):
(setAtkStateSetFromCoreObject):

  • html/HTMLAttributeNames.in:

Tools:

Add "AXSupportsAutoCompletion" attribute so the presence of
ATK_STATE_SUPPORTS_AUTOCOMPLETION can be tested.

  • WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:

(WTR::AccessibilityUIElement::boolAttributeValue):

LayoutTests:

  • accessibility/gtk/aria-autocomplete-expected.txt: Added.
  • accessibility/gtk/aria-autocomplete.html: Added.
11:11 AM Changeset in webkit [215988] by jdiggs@igalia.com
  • 14 edits
    2 adds in trunk

[ATK] ARIA treegrid role should be exposed as ATK_ROLE_TREE_TABLE; not ATK_ROLE_TABLE
​https://bugs.webkit.org/show_bug.cgi?id=171170

Reviewed by Chris Fleizach.

Source/WebCore:

AccessibilityTable::roleValue() was returning GridRole for both grid and treegrid.
Change this so that GridRole is returned for grid and TreeGridRole is returned for
treegrid. Add TreeGridRole to, or call isTable() in, methods where the behavior
should be the same as GridRole.

Test: accessibility/gtk/interface-table.html

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::selectedChildren):

  • accessibility/AccessibilityTable.cpp:

(WebCore::AccessibilityTable::roleValue):

  • accessibility/atk/WebKitAccessibleWrapperAtk.cpp:

(atkRole):
(getInterfaceMaskFromObject):

  • accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:

(-[WebAccessibilityObjectWrapper _accessibilityTableAncestor]):
(-[WebAccessibilityObjectWrapper _accessibilityTraitsFromAncestors]):

  • accessibility/mac/WebAccessibilityObjectWrapperMac.mm:

(createAccessibilityRoleMap):

Tools:

Add "AXInterfaceTable" attribute to test which elements implement AtkTable.

  • WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:

(WTR::AccessibilityUIElement::boolAttributeValue):

LayoutTests:

Add interface-table.html to verify that all roles which have tabular functionality
implement AtkTable. Update other tests to reflect that the ARIA treegrid role maps
to our internal TreeGridRole AccessibilityRole.

  • accessibility/gtk/interface-table-cell-expected.txt: Updated.
  • accessibility/gtk/interface-table-expected.txt: Added.
  • accessibility/gtk/interface-table.html: Added.
  • accessibility/roles-computedRoleString.html: Updated.
  • platform/gtk/accessibility/aria-selected-expected.txt: Updated.
  • platform/gtk/accessibility/roles-exposed-expected.txt: Updated.
  • platform/mac/accessibility/roles-computedRoleString-expected.txt: Updated.
11:11 AM Changeset in webkit [215987] by rniwa@webkit.org
  • 28 edits in trunk/Websites/perf.webkit.org

Add the support for scheduling a A/B testing with a patch.
​https://bugs.webkit.org/show_bug.cgi?id=171209

Reviewed by Chris Dumez.

Added the support for creating a custom test group with a patch applied.

First, each repository in a repository group has a boolean indicating whether a given repository can have
a patch applied or not. When any configuration in a test group contains a patch, we create build requests
without a test specified in order to "build" those patches. These build requests have negative order numbers
to differentiate them from regular build requests. We can't simply build ones with patches since there could
be differences in SDK, build options, etc... when patches are applied.

The JSON format for commit sets returned by /api/build-requests have been changed from using an array of
commit IDs to an array of dictionaries indicate commit and acceptsPatch boolean. /api/update-triggerable now
uses a dictionary with two keys: repository and acceptsPatch to specify a set of repositories associated with
a repository group, and /privileged-api-create-test-group uses a dictionary with two keys: revision and patch
instead of a revision string to specify commit sets.

Furthermore, the syncing script's configuration have been updated to use a dictionary of repository names to
an options dictionary instead of an array of repositories names. For now, the only supported option is
acceptsPatch but will be extended when we add the support for rolling back system components.
e.g. {"WebKit": {acceptsPatch: true}, "macOS": {}} instead of ["WebKit", "macOS"]

On the UI side, InstantFileUploader has been changed to accept only one file by default, and added a new method
allowMultipleFiles() to allow multiple files to be selected for custom roots. Also replaced the input element
with type=file by a button with a custom label to show labels such as "Apply a patch" or "Add a new root"
instead of the generic label like "choose a file".

  • init-database.sql: Added trigrepo_accepts_patch to triggerable_repositories to indicate whether a given

repository can have a patch applied or not. Made request_test optional in build_requests for when a build
request is created to build patches. Such a build request have a negative request_order. Updated the related
constraints accordingly.

  • public/admin/triggerables.php: Added the support for updating whether a given repository can have a patch

applied in each repository group. Only show the repositories in the repository group for this purpose since
there is no way to accept a patch on a repository without it being a part of the group.
(generate_repository_form): Now takes the markup for checkboxes instead of generating one itself.
(generate_repository_checkboxes): Now takes an array of repositories to generate checkboxes. The checkbox is
shown when the repository ID exists as a key in this array, and is checked when its value is true. The new
capability to skip repositories not in the array is used to hide repositories not associated with the group
in the list of checkboxes to indicate a repository accepts a patch.

  • public/api/update-triggerable.php:

(main): Now updates the description and acceptsRoots states of each repository group, and sets acceptsPatch
boolean for each repository in the group if set in the update.
(validate_repository_groups): Use a reference to $repository_groups in order to set repository_id_list, which
contains an array of repository IDs to find the existing repository group that matches the set via
RepositoryGroupFinder's find_by_repositories. Also added a various validations for acceptsRoots, a dictionary
specifying repository and acceptsPatch.

  • public/include/build-requests-fetcher.php:

(BuildRequestsFetcher::fetch_commits_for_set_if_needed): Instead of returning an array of commit IDs as
"commits", it now returns an array of dictionaries with "commit" and "patch" keys specifying the commit ID
and the patch file's ID respectively as "revisionItems".
(BuildRequestsFetcher::add_uploaded_file): Added. Extracted from fetch_commits_for_set_if_needed. Used to
add either a patch file or a custom root file in the list of uploaded files in the result.

  • public/include/manifest-generator.php:

(fetch_triggerables): Each element in repository group's "repositories" field is now an array of dictionaries
with "repository" and "acceptsPatch" as keys.

  • public/include/repository-group-finder.php:

(RepositoryGroupFinder::construct): Added a map for boolean indicating whether a given repository group
allows a patch on a repository. Used in /privileged-api/create-test-group.
(RepositoryGroupFinder::accepts_patch): Added.
(RepositoryGroupFinder::populate_map): Build up the map for acceptsPatch boolean per repository per group.

  • public/privileged-api/create-test-group.php:

(main): Fixed a bug that we were not explicitly checking for a duplicate test group name (with a test). Create
build requests to "build" patches if there is any patch file specified.
(commit_sets_from_revision_sets): Updated to take a dictionary with "revision" and "patch" as keys to specify
a revision and a patch if any instead of just a revision string for each repository. Also validate that each
repository is allowed to have a patch once the repository group has been found for the set of repositories.
(ensure_commit_sets):

  • public/v3/components/custom-analysis-task-configurator.js:

(CustomAnalysisTaskConfigurator): Added _patchUploaders as an instance variable, which is a dictionary of
configuration names to a map of InstantFileUploader's used to upload a patch. Also renamed _fileUploaders to
_customRootUploaders for clarity.
(CustomAnalysisTaskConfigurator.prototype.setCommitSets):
(CustomAnalysisTaskConfigurator.prototype.didConstructShadowTree.createRootUploader): Added.
(CustomAnalysisTaskConfigurator.prototype.didConstructShadowTree):
(CustomAnalysisTaskConfigurator.prototype._ensurePatchUploader): Added. Creates an instant file uploader for
patches. We only allow a single patch per repository.
(CustomAnalysisTaskConfigurator.prototype._computeCommitSet): Include a patch in the commit set as needed.
(CustomAnalysisTaskConfigurator.prototype._buildRevisionTable): Show the patch file uploader for repositories
which can have patches in the current repository group.
(CustomAnalysisTaskConfigurator.cssTemplate): Show borders between every rows instead of just between tbody's
now that each row can have a patch file uploader.

  • public/v3/components/instant-file-uploader.js:

(InstantFileUploader): Added _fileInput and _allowMultipleFiles as instance variables. We now show a button
in the UI instead of an input with type=file. _fileInput is a hidden input with type=file used inside a click
event of the button to let the user pick a file.
(InstantFileUploader.prototype.allowMultipleFiles): Added. Allows this instance to accept multiple files.
(InstantFileUploader.prototype.didConstructShadowTree): Synthetically click on the hidden input element when
the newly added button element is clicked to open the browser's file picker.
(InstantFileUploader.prototype.render): Hide the button to add a file if this instance can only select one file
and there is already some file being uploaded in this instance.
(InstantFileUploader.htmlTemplate): Replaced the input element with type=file with a button. Its label comes
from the default slot content.

  • public/v3/models/build-request.js:

(BuildRequest): Made the test optional.
(BuildRequest.prototype.isBuild): Returns true if this is a build request for building a patch.
(BuildRequest.prototype.isTest): Returns true if this is a build request for running tests.
(BuildRequest.constructBuildRequestsFromData): Create each commit log here instead of relying on CommitSet's
constructor to construct its commit logs. Also updated per the replacement of an array of commit IDs by
an array of dictionaries with commit and patch properties.

  • public/v3/models/commit-set.js:

(CommitSet): Made _repositoryToCommitMap a real Map object. Also added _repositoryToPatchMap. Also got rid of
the code to instantiate commit logs since that's now done in BuildRequest.constructBuildRequestsFromData.
(CommitSet.prototype.commitForRepository):
(CommitSet.prototype.revisionForRepository):
(CommitSet.prototype.patchForRepository): Added.
(CommitSet.prototype.latestCommitTime): Modernized the code.
(CommitSet.prototype.equals): Modernized the code. Also added the check for patches.
(MeasurementCommitSet): Updated per the change to make _repositoryToCommitMap a real Map.
(CustomCommitSet.prototype.setRevisionForRepository):
(CustomCommitSet.prototype.equals): Added the check for patches.
(CustomCommitSet.prototype.revisionForRepository):
(CustomCommitSet.prototype.patchForRepository): Added.

  • public/v3/models/manifest.js:

(Manifest._didFetchManifest): Updated per the replacement of an array of commit IDs by an array of dictionaries
with commit and patch properties.

  • public/v3/models/repository.js:

(Repository.prototype.ownerId): Renamed from owner for clarity.

  • public/v3/models/test-group.js:

(TestGroup): Modernized the code by using LazilyEvaluatedFunction. Removed _requestsAreInOrder since it's not
necessary anymore with LazilyEvaluatedFunction.
(TestGroup.prototype.addBuildRequest):
(TestGroup.prototype.test): Use the last build request's test since the first few requests could be requests to
build patches.
(TestGroup.prototype.platform): Ditto.
(TestGroup.prototype._lastRequest): Added.
(TestGroup.prototype._orderedBuildRequests): Added.
(TestGroup.prototype.repetitionCount): Only count the build requests for testing (skipping any requests to
build patches).
(TestGroup.prototype.requestedCommitSets): Simply call _computeRequestedCommitSetsLazily.
(TestGroup.prototype._computeRequestedCommitSets): Extracted from requestedCommitSets.
(TestGroup.prototype.requestsForCommitSet):
(TestGroup.prototype.labelForCommitSet): Rewritten. Just compute the label here instead of relying on
_commitSetToLabel since requestedSets is always of the length two at the moment.
(TestGroup._revisionSetsFromCommitSets): Specify both the revision and the patch in the revision set.

  • public/v3/models/triggerable.js:

(TriggerableRepositoryGroup): Added _patchAcceptingSet as an instance variable. Use
sortByNamePreferringOnesWithURL to sort repositories instead of simple sortByName.
(TriggerableRepositoryGroup.prototype.accepts): Added checks for the custom roots and patches.
(TriggerableRepositoryGroup.prototype.acceptsPatchForRepository): Added.

  • server-tests/api-build-requests-tests.js: Updated the test cases per the replacement of an array of commit

IDs by an array of dictionaries with commit and patch properties.

  • server-tests/api-manifest-tests.js: Updated the test case per the name of Repository's owner to ownerId.
  • server-tests/api-update-triggerable.js: Updated the test case per the name of Repository's owner to ownerId,

and added a test case for updating whether a given repository group allows custom roots as well as patches
on repositories via /api/update-triggerable.
(.updateWithOSXRepositoryGroup): Updated the sample syncing script configuration per the format change.
(.refetchManifest): Added.

  • server-tests/privileged-api-create-test-group-tests.js: Updated per the syncing script configuration format

change. Also added a test for creating a test group with a duplicate name, which is expected to fail with
DuplicateTestGroupName, and creating a test group with a patch both when it's allowed and when it's not allowed
in the matching repository group.
(.addTriggerableAndCreateTask): Updated per the format change.

  • server-tests/resources/mock-data.js:

(MockData.addEmptyTriggerable): Added a metric and its configuration to make it appear in the manifest file.
The new test case in api-update-triggerable.js requires this.
(MockData.mockTestSyncConfigWithSingleBuilder): Updated per the syncing script configuration format change.
(MockData.mockTestSyncConfigWithTwoBuilders): Ditto.

  • server-tests/tools-buildbot-triggerable-tests.js: Removed the useless assertions about test configurations,

and added assertions about custom roots and patches in the test case for updateTriggerables.

  • tools/js/buildbot-syncer.js:

(BuildbotSyncer._parseRepositoryGroup): Made each assertion explicitly refer to the specific repository group
to make it more user friendly. Now each repository group uses a dictionary of repository names to its options
in the syncing script configurations. When parsed, we insert it as an array of dictionaries with repository ID
and acceptsPatch boolean specified separately since this is the format /api/update-triggerable expects.

  • tools/js/buildbot-triggerable.js:

(BuildbotTriggerable.prototype.updateTriggerable):

  • unit-tests/build-request-tests.js:

(sampleBuildRequestData): Updated per the commit sets format change in /api/build-requests.

  • unit-tests/buildbot-syncer-tests.js: Updated the existing tests per various format changes and added a couple

of new test cases for the syncing script's configuration validation.
(sampleiOSConfig):
(smallConfiguration):
(createSampleBuildRequest):

  • unit-tests/resources/mock-v3-models.js:

(MockModels.inject): Updated per the repository group format change.

  • unit-tests/test-groups-tests.js:

(sampleTestGroup): Updated per the commit sets format change in /api/build-requests.

7:51 AM Changeset in webkit [215986] by gskachkov@gmail.com
  • 5 edits
    1 add in trunk

Initialize functions too early in an eval
​https://bugs.webkit.org/show_bug.cgi?id=161099

Reviewed by Saam Barati.

JSTests:

  • stress/eval-func-decl-with-let-const-class.js: Added.

Source/JavaScriptCore:

Current patch allow to fix problem with scope in function that is
declared within eval. Before scope was set inside Interpretator.cpp and it
was scope where eval is executed, but in this case function would not
see let/const variables and classes declated in eval.
This patch devide declaration and binding in two operation, first just declare
variable with function name, and second bind variable to function with correct
scope

  • bytecompiler/BytecodeGenerator.cpp:

(JSC::BytecodeGenerator::generate):
(JSC::BytecodeGenerator::BytecodeGenerator):

  • bytecompiler/BytecodeGenerator.h:
  • interpreter/Interpreter.cpp:

(JSC::Interpreter::execute):

2:56 AM Changeset in webkit [215985] by jdiggs@igalia.com
  • 6 edits
    2 adds in trunk

[ATK] Emit object:state-changed:busy event when aria-busy value changes
​https://bugs.webkit.org/show_bug.cgi?id=171181

Reviewed by Chris Fleizach.

Source/WebCore:

Emit object:state-changed:busy for AXElementBusyChanged notifications.

Test: accessibility/gtk/aria-busy-changed-notification.html

  • accessibility/atk/AXObjectCacheAtk.cpp:

(WebCore::AXObjectCache::postPlatformNotification):

Tools:

Add handler for AXElementBusyChanged notifications.

  • WebKitTestRunner/InjectedBundle/atk/AccessibilityNotificationHandlerAtk.cpp:

LayoutTests:

New test. Also update the expectations for menu-list-unfocused-notifications.html
because the addition of the handler for AXElementBusyChanged results in that
notification being output in the test results.

  • accessibility/gtk/aria-busy-changed-notification-expected.txt: Added.
  • accessibility/gtk/aria-busy-changed-notification.html: Added.
  • accessibility/gtk/menu-list-unfocused-notifications-expected.txt: Updated.
1:06 AM Changeset in webkit [215984] by gskachkov@gmail.com
  • 49 edits
    13 adds in trunk

[ES6]. Implement Annex B.3.3 function hoisting rules for eval
​https://bugs.webkit.org/show_bug.cgi?id=163208

Reviewed by Saam Barati.

JSTests:

  • stress/eval-func-decl-block-scoping-reassign.js: Added.

(assert):
(throw.new.Error.f):
(throw.new.Error):

  • stress/eval-func-decl-block-with-remove.js: Added.

(assert):
(foo.boo):
(foo):

  • stress/eval-func-decl-block-with-var-and-remove.js: Added.

(assert):
(assertThrow):
(foo):
(boo):
(joo):
(koo):

  • stress/eval-func-decl-block-with-var-sinthesize.js: Added.

(assert):
(assertThrow):
(foo):
(boo):
(hoo):
(joo):
(koo):

  • stress/eval-func-decl-in-block-scope-and-bind-to-top-eval-scope.js: Added.
  • stress/eval-func-decl-in-eval-within-block-with-let.js: Added.

(assert):
(assertThrow):
(foo):
(boo):
(goo):

  • stress/eval-func-decl-in-eval-within-with-scope.js: Added.

(assert):
(assertThrow):
(foo):
(boo):
(boo.let.val2):
(boo.let.val3):

  • stress/eval-func-decl-in-frozen-global.js: Added.

(assert):
(assertThrow):
(throw.new.Error):
(Object.freeze):

  • stress/eval-func-decl-in-global-of-eval.js: Added.

(assert):
(assertThrow):
(bar):
(baz):
(foobar):

  • stress/eval-func-decl-in-global.js: Added.

(assert):
(assertThrow):

  • stress/eval-func-decl-in-if.js: Added.

(assert):

  • stress/eval-func-decl-within-eval-with-reassign-to-var.js: Added.

(assert):
(assertThrow):
(foo):
(boo):
(foobar):
(hoo):
(joo):
(koo):
(loo):

  • stress/eval-func-decl-within-eval-without-reassign-to-let.js: Added.

(assert):
(assertThrow):
(foo):
(boo):
(goo):

  • stress/variable-under-tdz-eval-tricky.js:

(assert):

  • test262.yaml:

Source/JavaScriptCore:

Current patch implements Annex B.3.3 that is related to
hoisting of function declaration in eval.
​https://tc39.github.io/ecma262/#sec-web-compat-evaldeclarationinstantiation
Function declaration in eval should create variable with
function name in function scope where eval is invoked
or bind to variable if it declared outside of the eval.
If variable is created it can be removed by 'delete a;' command.
If eval is invoke in block scope that contains let/const
variable with the same name as function declaration
we do not bind. This patch leads to the following behavior:

function foo() {

{

print(boo); undefined
eval('{ function boo() {}}');
print(boo); function boo() {}

}
print(boo); function boo() {}

}

function foobar() {

{

let boo = 10;
print(boo); 10;
eval('{ function boo() {}}');
print(boo); 10;

}
print(boo) 10

}

function bar() {

{

var boo = 10;
print(boo); 10
eval('{ function boo() {} }');
print(boo); function boo() {}

}
print(boo); function boo() {}

}

function bas() {

{

let boo = 10;
eval(' { function boo() {} } ');
print(boo); 10

}
print(boo); Reference Error

}

Current implementation relies on already implemented
'hoist function in sloppy mode' feature, with small changes.
In short it works in following way: during hoisting of function
with name S in eval, we are looking for first scope that
contains space for variable with name S and if this scope
has var type we bind function there

To implement this feature was added bytecode ops:
op_resolve_scope_for_hoisting_func_decl_in_eval - get variable scope
or return undefined if variable can't be binded there.

There is a corner case, hoist function in eval within catch block,
that is not covered by this patch, and will be fixed in
​https://bugs.webkit.org/show_bug.cgi?id=168184

  • bytecode/BytecodeDumper.cpp:

(JSC::BytecodeDumper<Block>::dumpBytecode):

  • bytecode/BytecodeList.json:
  • bytecode/BytecodeUseDef.h:

(JSC::computeUsesForBytecodeOffset):
(JSC::computeDefsForBytecodeOffset):

  • bytecode/CodeBlock.cpp:

(JSC::CodeBlock::finalizeLLIntInlineCaches):

  • bytecode/EvalCodeBlock.h:

(JSC::EvalCodeBlock::functionHoistingCandidate):
(JSC::EvalCodeBlock::numFunctionHoistingCandidates):

  • bytecode/UnlinkedEvalCodeBlock.h:
  • bytecompiler/BytecodeGenerator.cpp:

(JSC::BytecodeGenerator::BytecodeGenerator):
(JSC::BytecodeGenerator::hoistSloppyModeFunctionIfNecessary):
(JSC::BytecodeGenerator::emitResolveScopeForHoistingFuncDeclInEval):

  • bytecompiler/BytecodeGenerator.h:
  • dfg/DFGAbstractInterpreterInlines.h:

(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

  • dfg/DFGByteCodeParser.cpp:

(JSC::DFG::ByteCodeParser::parseBlock):

  • dfg/DFGCapabilities.cpp:

(JSC::DFG::capabilityLevel):

  • dfg/DFGClobberize.h:

(JSC::DFG::clobberize):

  • dfg/DFGDoesGC.cpp:

(JSC::DFG::doesGC):

  • dfg/DFGFixupPhase.cpp:

(JSC::DFG::FixupPhase::fixupNode):

  • dfg/DFGNode.h:

(JSC::DFG::Node::hasIdentifier):

  • dfg/DFGNodeType.h:
  • dfg/DFGOperations.cpp:
  • dfg/DFGOperations.h:
  • dfg/DFGPredictionPropagationPhase.cpp:
  • dfg/DFGSafeToExecute.h:

(JSC::DFG::safeToExecute):

  • dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::compileResolveScopeForHoistingFuncDeclInEval):

  • dfg/DFGSpeculativeJIT.h:

(JSC::DFG::SpeculativeJIT::callOperation):

  • dfg/DFGSpeculativeJIT32_64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

  • dfg/DFGSpeculativeJIT64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

  • ftl/FTLCapabilities.cpp:

(JSC::FTL::canCompile):

  • ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileResolveScopeForHoistingFuncDeclInEval):

  • interpreter/Interpreter.cpp:

(JSC::Interpreter::execute):

  • jit/JIT.cpp:

(JSC::JIT::privateCompileMainPass):

  • jit/JIT.h:
  • jit/JITOperations.h:
  • jit/JITPropertyAccess.cpp:

(JSC::JIT::emit_op_resolve_scope_for_hoisting_func_decl_in_eval):

  • jit/JITPropertyAccess32_64.cpp:

(JSC::JIT::emit_op_resolve_scope_for_hoisting_func_decl_in_eval):

  • llint/LowLevelInterpreter.asm:
  • parser/Parser.cpp:

(JSC::Parser<LexerType>::parseFunctionDeclarationStatement):

  • parser/Parser.h:

(JSC::Scope::getSloppyModeHoistedFunctions):
(JSC::Parser::declareFunction):

  • runtime/CommonSlowPaths.cpp:

(JSC::SLOW_PATH_DECL):

  • runtime/CommonSlowPaths.h:
  • runtime/EvalExecutable.h:

(JSC::EvalExecutable::numFunctionHoistingCandidates):
(JSC::EvalExecutable::numTopLevelFunctionDecls):
(JSC::EvalExecutable::numberOfFunctionDecls): Deleted.

  • runtime/JSScope.cpp:

(JSC::JSScope::resolve):
(JSC::JSScope::resolveScopeForHoistingFuncDeclInEval):

  • runtime/JSScope.h:

LayoutTests:

  • inspector/runtime/evaluate-CommandLineAPI-expected.txt:
  • inspector/runtime/evaluate-CommandLineAPI.html:
  • js/parser-syntax-check-expected.txt:
  • js/script-tests/parser-syntax-check.js:
12:21 AM Changeset in webkit [215983] by beidson@apple.com
  • 4 edits in trunk/Source/WebKit2

Regression (r215941) Unexpected logging causes perf tests to fail.
<rdar://problem/31907269> and ​https://bugs.webkit.org/show_bug.cgi?id=171482

Reviewed by Alexey Proskuryakov.

  • Shared/mac/SandboxExtensionMac.mm:

(WebKit::resolveAndCreateReadWriteDirectoryForSandboxExtension):

  • UIProcess/Cocoa/WebProcessPoolCocoa.mm:

(WebKit::WebProcessPool::networkingCachesDirectory):
(WebKit::WebProcessPool::webContentCachesDirectory):

  • UIProcess/WebsiteData/WebsiteDataStore.cpp:

(WebKit::WebsiteDataStore::resolveDirectoriesIfNecessary): Only resolve the cookie storage directory

if a custom one has been set.

Apr 29, 2017:

10:02 PM Changeset in webkit [215982] by aakash_jain@apple.com
  • 8 edits in trunk/Tools

Use PEP8 style guide for raising exceptions in webkitpy.
​https://bugs.webkit.org/show_bug.cgi?id=171439

Reviewed by Daniel Bates.

As per ​https://www.python.org/dev/peps/pep-0008/
"When raising an exception in Python 2, use raise ValueError('message')
instead of the older form raise ValueError, 'message'"

  • Scripts/webkitpy/common/net/irc/ircbot.py: Using new style for raising exceptions.
  • Scripts/webkitpy/common/thread/messagepump.py: Same.
  • Scripts/webkitpy/tool/bot/queueengine.py: Same.
  • Scripts/webkitpy/tool/commands/queues.py: Same.
  • Scripts/webkitpy/tool/commands/stepsequence.py: Same.
  • Scripts/webkitpy/tool/multicommandtool.py: Same.
  • Scripts/webkitpy/tool/steps/abstractstep.py: Same.
8:32 PM Changeset in webkit [215981] by commit-queue@webkit.org
  • 9 edits
    3 adds in trunk

Zero out PerformanceResourceTiming properties for cached cross-origin responses without Timing-Allow-Origin
​https://bugs.webkit.org/show_bug.cgi?id=171394

Patch by Joseph Pecoraro <Joseph Pecoraro> on 2017-04-29
Reviewed by Youenn Fablet.

LayoutTests/imported/w3c:

WebKit doesn't create multiple PerformanceResourceTiming entries for
loads that use the same CachedResource. However for revalidation
requests, which may happen for cached entries, we will send new
network requests. These tests cover whether revalidation requests
include timing data or not.

  • web-platform-tests/resource-timing/resources/rt-revalidation-response.py: Added.
  • web-platform-tests/resource-timing/rt-cors-expected.txt:
  • web-platform-tests/resource-timing/rt-cors.js:
  • web-platform-tests/resource-timing/rt-cors.worker-expected.txt:
  • web-platform-tests/resource-timing/rt-revalidate-requests-expected.txt: Added.
  • web-platform-tests/resource-timing/rt-revalidate-requests.html: Added.

Source/WebCore:

  • loader/ResourceTiming.cpp:

(WebCore::ResourceTiming::fromCache):
(WebCore::ResourceTiming::ResourceTiming):

  • loader/ResourceTiming.h:
  • loader/cache/CachedResourceLoader.cpp:

(WebCore::CachedResourceLoader::requestResource):
Include the timing-allow-origin check for cached responses.
Also, avoid including an extra entry for an ongoing cached resource
load, since when that load completes it should be reported.

8:30 PM Changeset in webkit [215980] by commit-queue@webkit.org
  • 6 edits in trunk/Source

Readd assertion removed accidentally in r215955
​https://bugs.webkit.org/show_bug.cgi?id=171466

Patch by Youenn Fablet <​youenn@apple.com> on 2017-04-29
Reviewed by Eric Carlson.

Source/WebCore:

Unreviewed.

  • platform/mediastream/mac/MockRealtimeVideoSourceMac.mm:

(WebCore::MockRealtimeVideoSourceMac::orientationChanged): Improved FIXME after review.

Source/WebKit2:

  • WebProcess/WebPage/WebBackForwardListProxy.cpp:

(WebKit::WebBackForwardListProxy::addItemFromUIProcess):

8:28 PM Changeset in webkit [215979] by aestes@apple.com
  • 7 edits in trunk/Source

[macOS] Picture-in-picture should be disabled in WebPlaybackControlsManager during external media playback
​https://bugs.webkit.org/show_bug.cgi?id=171468
<rdar://problem/31900349>

Reviewed by Eric Carlson.

Source/WebCore:

  • platform/mac/WebPlaybackControlsManager.h: Declared canTogglePictureInPicture as a

readwrite property.

  • platform/mac/WebPlaybackControlsManager.mm: Synthesize canTogglePictureInPicture.

(-[WebPlaybackControlsManager canTogglePictureInPicture]): Deleted.

  • platform/mac/WebPlaybackSessionInterfaceMac.h:
  • platform/mac/WebPlaybackSessionInterfaceMac.mm:

(WebCore::WebPlaybackSessionInterfaceMac::externalPlaybackChanged): Called
-[WebPlaybackControlsManager setCanTogglePictureInPicture:].

Source/WebKit2:

  • UIProcess/Cocoa/WebViewImpl.mm:

(WebKit::WebViewImpl::updateMediaTouchBar): Set canTogglePictureInPicture to YES by default
on WebPlaybackControlsManager.

8:06 PM Changeset in webkit [215978] by Simon Fraser
  • 2 edits in trunk/Source/WebCore

Fix tests broken by r215964.

Fix copy-paste error that caused the wrong paint phase to get passed to the fire writeLayerRenderers()
call, causing extra renderer dumping.

Also remove unused default arguments.

  • rendering/RenderTreeAsText.cpp:

(WebCore::writeLayerRenderers):
(WebCore::writeLayers):

2:25 PM Changeset in webkit [215977] by gskachkov@gmail.com
  • 5 edits in trunk

Deep nesting is leading to ReferenceError for hoisted function
​https://bugs.webkit.org/show_bug.cgi?id=171456

Reviewed by Yusuke Suzuki.

Current patch fix error that appears during hoisting of the function
in block scope. Error happens only when exist some deep scope that lead
to increase scope stack, after which list of the hosted candidates do not
copied to updated scope stack.

Source/JavaScriptCore:

  • parser/Parser.h:

(JSC::Scope::Scope):

LayoutTests:

  • js/function-declaration-statement-expected.txt:
  • js/script-tests/function-declaration-statement.js:

(deepNesting.foo):
(deepNesting):
(deepNestingForFunctionDeclaration.foo):
(deepNestingForFunctionDeclaration):
(eval.deepNestingInEval.foo):
(eval.deepNestingInEval):
(eval.deepNestingForFunctionDeclarationInEval.foo):
(eval.deepNestingForFunctionDeclarationInEval):

2:16 PM Changeset in webkit [215976] by Alan Bujtas
  • 3 edits
    2 adds in trunk

iBooks text can overlap, sometimes columns are shifted splitting words.
​https://bugs.webkit.org/show_bug.cgi?id=171472
<rdar://problem/31096037>

Reviewed by Antti Koivisto.

Source/WebCore:

Instead of just checking if the glyph is taller than the line, we need to ensure that both the
ascent and the descent have enough space (this is for -webkit-line-box-contain: glyph).

Test: fast/text/simple-line-layout-glyph-overflows-line.html

  • rendering/SimpleLineLayout.cpp:

(WebCore::SimpleLineLayout::canUseForText): compute the available space for the ascent/descent
and check them against the ceil-ed(see FontCascade::floatWidthForSimpleText) glyph min/max y.

LayoutTests:

  • fast/text/simple-line-layout-glyph-overflows-line-expected.html: Added.
  • fast/text/simple-line-layout-glyph-overflows-line.html: Added.
11:55 AM Changeset in webkit [215975] by n_wang@apple.com
  • 7 edits in trunk/Source/WebCore

AX: Improve performance of addChildren()/childrenChanged()
​https://bugs.webkit.org/show_bug.cgi?id=171443

Reviewed by Chris Fleizach.

There's a lot of unnecessary work happening when childrenChanged() is being called.
Some of the improvements here:

  1. When childrenChanged() is being called on some element, we are marking its parent chain dirty. However, in the addChild() method we are then clearing each child of that 'dirty' parent, eventually making the entire tree dirty. Added a m_subTreeDirty flag and using the needsToUpdateChildren() check to make sure we are only clearing the right chain without updating the unchanged siblings.
  2. In the addChild() method we are calling accessibilityIsIgnored() on each child and that might lead to going up the parent chain again to get necessary information. Since we are already traversing the tree from top to bottom to add children, added a struct to store the information and pass it to the child so to avoid unnecessary traversal.
  3. Reduced the amount work of ARIA text controls perform when updating its parents in childrenChanged() so that we don't update a big portion of the tree while typing.

No new tests since this didn't change any functionality.

  • accessibility/AccessibilityNodeObject.cpp:

(WebCore::AccessibilityNodeObject::AccessibilityNodeObject):
(WebCore::AccessibilityNodeObject::childrenChanged):
(WebCore::AccessibilityNodeObject::insertChild):
(WebCore::AccessibilityNodeObject::addChildren):
(WebCore::AccessibilityNodeObject::isDescendantOfBarrenParent):

  • accessibility/AccessibilityNodeObject.h:
  • accessibility/AccessibilityObject.cpp:

(WebCore::AccessibilityObject::AccessibilityObject):
(WebCore::AccessibilityObject::defaultObjectInclusion):
(WebCore::AccessibilityObject::setIsIgnoredFromParentDataForChild):

  • accessibility/AccessibilityObject.h:

(WebCore::AccessibilityIsIgnoredFromParentData::AccessibilityIsIgnoredFromParentData):
(WebCore::AccessibilityIsIgnoredFromParentData::isNull):
(WebCore::AccessibilityObject::setNeedsToUpdateSubTree):
(WebCore::AccessibilityObject::needsToUpdateChildren):
(WebCore::AccessibilityObject::setIsIgnoredFromParentData):

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::updateChildrenIfNecessary):
(WebCore::AccessibilityRenderObject::addChildren):

  • accessibility/AccessibilityRenderObject.h:

(WebCore::AccessibilityRenderObject::setRenderObject):
(WebCore::AccessibilityRenderObject::needsToUpdateChildren): Deleted.

11:33 AM Changeset in webkit [215974] by mitz@apple.com
  • 2 edits in trunk/Source/ThirdParty/libwebrtc

[Xcode] libwebrtc SRCROOT includes examples
​https://bugs.webkit.org/show_bug.cgi?id=171478

Reviewed by Tim Horton.

  • Configurations/Base.xcconfig: Exclude the Source/webrtc/examples subdirectory from installsrc. Its contents are not used for building any of the targets in the project.
11:03 AM Changeset in webkit [215973] by Yusuke Suzuki
  • 25 edits
    1 copy
    3 moves
    2 adds in trunk/Source

Move WebCore CPUTime to WTF and implement it in all the platforms
​https://bugs.webkit.org/show_bug.cgi?id=171477

Reviewed by Chris Dumez.

Source/WebCore:

Move CPUTime to WTF. And rename getCPUTime to CPUTime::get().

  • CMakeLists.txt:
  • PlatformMac.cmake:
  • WebCore.xcodeproj/project.pbxproj:
  • page/ChromeClient.h:
  • page/PerformanceMonitor.cpp:

(WebCore::PerformanceMonitor::PerformanceMonitor):
(WebCore::PerformanceMonitor::measurePostLoadCPUUsage):
(WebCore::PerformanceMonitor::measurePostBackgroundingCPUUsage):
(WebCore::PerformanceMonitor::measureCPUUsageInActivityState):

  • page/PerformanceMonitor.h:
  • platform/CPUMonitor.cpp:

(WebCore::CPUMonitor::setCPULimit):
(WebCore::CPUMonitor::timerFired):

  • platform/CPUMonitor.h:

Source/WebKit2:

We now hold Seconds in CPUTime instead of int64_t. It finally sends
Seconds through WebKit2 IPC. We do not need to specialize ArgumentCoder
for Seconds since Seconds just holds double.

  • UIProcess/PerActivityStateCPUUsageSampler.cpp:

(WebKit::PerActivityStateCPUUsageSampler::PerActivityStateCPUUsageSampler):
(WebKit::PerActivityStateCPUUsageSampler::reportWebContentCPUTime):
(WebKit::PerActivityStateCPUUsageSampler::loggingTimerFired):

  • UIProcess/PerActivityStateCPUUsageSampler.h:
  • UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::reportWebContentCPUTime):

  • UIProcess/WebProcessPool.h:
  • UIProcess/WebProcessPool.messages.in:
  • WebProcess/WebCoreSupport/WebChromeClient.cpp:

(WebKit::WebChromeClient::reportProcessCPUTime):

  • WebProcess/WebCoreSupport/WebChromeClient.h:

Source/WTF:

We move WebCore::CPUTime to WTF::CPUTime since it is useful even in JSC
and it does not depend on any external libraries.
And we additionally implement WTF::CPUTime in all the platforms: Windows and Unix.

In CPUTime, instead of holding int64_t, we hold MonotonicTime and Seconds.

  • WTF.xcodeproj/project.pbxproj:
  • wtf/CMakeLists.txt:
  • wtf/CPUTime.cpp: Copied from Source/WebCore/platform/CPUTime.cpp.

(WTF::CPUTime::percentageCPUUsageSince):

  • wtf/CPUTime.h: Renamed from Source/WebCore/platform/CPUTime.h.
  • wtf/PlatformEfl.cmake:
  • wtf/PlatformGTK.cmake:
  • wtf/PlatformJSCOnly.cmake:
  • wtf/PlatformMac.cmake:
  • wtf/PlatformWin.cmake:
  • wtf/cocoa/CPUTimeCocoa.mm: Renamed from Source/WebCore/platform/cocoa/CPUTimeCocoa.mm.

(WTF::timeValueToMicroseconds):
(WTF::CPUTime::get):

  • wtf/unix/CPUTimeUnix.cpp: Renamed from Source/WebCore/platform/CPUTime.cpp.

(WTF::timevalToSeconds):
(WTF::CPUTime::get):

  • wtf/win/CPUTimeWin.cpp: Added.

(WTF::fileTimeToSeconds):
(WTF::CPUTime::get):

9:32 AM Changeset in webkit [215972] by Yusuke Suzuki
  • 5 edits in trunk/Source/JavaScriptCore

[JSC] LabelScopePtr is not necessary
​https://bugs.webkit.org/show_bug.cgi?id=171474

Reviewed by Geoffrey Garen.

Originally, LabelScopePtr is introduced because LabelScopes uses Vector<> instead of SegmentedVector<>.
LabelScopePtr holds the pointer to the vector owner and index instead of the pointer to LabelScope directly
since Vector<> can relocate LocalScopes inside it.
The reason why LabelScopes use Vector instead is that there is code copying this vector. SegmentedVector<>
prohibits copying since it is so costly. So, we used Vector<> here instead of SegmentedVector<>.

But the latest code does not have copying code for LabelScopes. Thus, we can take the same design to Label and
RegisterID. Just use SegmentedVector<> and Ref<>/RefPtr<>. This patch removes LabelScopePtr since it is no
longer necessary. And use SegmentedVector for LabelScopes.

  • bytecompiler/BytecodeGenerator.cpp:

(JSC::reclaim):
(JSC::BytecodeGenerator::reclaimFreeRegisters):
(JSC::BytecodeGenerator::newLabelScope):
(JSC::BytecodeGenerator::newLabel):
(JSC::BytecodeGenerator::pushFinallyControlFlowScope):
(JSC::BytecodeGenerator::breakTarget):
(JSC::BytecodeGenerator::continueTarget):
(JSC::BytecodeGenerator::emitEnumeration):

  • bytecompiler/BytecodeGenerator.h:
  • bytecompiler/LabelScope.h:

(JSC::LabelScope::LabelScope):
(JSC::LabelScope::breakTarget):
(JSC::LabelScope::continueTarget):
(JSC::LabelScope::type):
(JSC::LabelScope::name):
(JSC::LabelScope::scopeDepth):
(JSC::LabelScope::ref):
(JSC::LabelScope::deref):
(JSC::LabelScope::refCount):
(JSC::LabelScopePtr::LabelScopePtr): Deleted.
(JSC::LabelScopePtr::operator=): Deleted.
(JSC::LabelScopePtr::~LabelScopePtr): Deleted.
(JSC::LabelScopePtr::operator!): Deleted.
(JSC::LabelScopePtr::operator*): Deleted.
(JSC::LabelScopePtr::operator->): Deleted.
(JSC::LabelScopePtr::null): Deleted.

  • bytecompiler/NodesCodegen.cpp:

(JSC::DoWhileNode::emitBytecode):
(JSC::WhileNode::emitBytecode):
(JSC::ForNode::emitBytecode):
(JSC::ForInNode::emitBytecode):
(JSC::ContinueNode::trivialTarget):
(JSC::ContinueNode::emitBytecode):
(JSC::BreakNode::trivialTarget):
(JSC::BreakNode::emitBytecode):
(JSC::SwitchNode::emitBytecode):
(JSC::LabelNode::emitBytecode):

Note: See TracTimeline for information about the timeline view.