Context Navigation

← Previous Period
Next Period →

Timeline

Dec 9, 2017:

6:53 PM Changeset in webkit [225729] by Darin Adler

1 edit
2 adds in trunk/LayoutTests

Add test demonstrating leaks that happen when we create reference cycles with DOM objects
https://bugs.webkit.org/show_bug.cgi?id=180323

Reviewed by Filip Pizlo.

fast/dom/reference-cycle-leaks-expected.txt: Added.
fast/dom/reference-cycle-leaks.html: Added.

1:42 PM Changeset in webkit [225728] by Yusuke Suzuki

4 edits in trunk/Source/WTF

[WTF] Remove RELAXED_CONSTEXPR
https://bugs.webkit.org/show_bug.cgi?id=180624

Reviewed by JF Bastien.

All of our required compilers support relaxed constexpr in C++14.
We can drop RELAXED_CONSTEXPR macro in our tree.

wtf/Compiler.h:
wtf/Expected.h:

(std::experimental::fundamentals_v3::expected::operator*):
(std::experimental::fundamentals_v3::expected::value):
(std::experimental::fundamentals_v3::expected::error):

wtf/Unexpected.h:

(std::experimental::fundamentals_v3::unexpected::value):
(std::experimental::fundamentals_v3::unexpected::value const):

12:34 PM Changeset in webkit [225727] by Yusuke Suzuki

4 edits in trunk

Fix WTF::Hasher tuple expansion with variadic args
https://bugs.webkit.org/show_bug.cgi?id=180623

Reviewed by JF Bastien.

Source/WTF:

We expand a tuple with ..., and use this in a function's argument list.
And in this argument list, we call add() for each value. This will be
expanded as follows.

[] (...) { }((add(hasher, std::get<i>(values)), 0)...);

will become,

[] (...) { }((add(hasher, std::get<0>(values)), 0), (add(hasher, std::get<1>(values)), 0), ...);

However, the evaluation order of the C++ argument is unspecified. Actually,
in GCC environment, this add() is not called in our expected order. As a
result, currently we have test failures in TestWTF.

This patch just uses variadic templates to expand tuples, and call add() in
our expected order. This patch fixes an existing failure of TestWTF in GCC environment.

wtf/Hasher.h:

(WTF::Hasher::computeHash):
(WTF::addArgs):
(WTF::addTupleHelper):
(WTF::add):

Tools:

TestWebKitAPI/Tests/WTF/Hasher.cpp:

(TestWebKitAPI::TEST):

12:11 PM Changeset in webkit [225726] by Yusuke Suzuki

2 edits in trunk/Source/WTF

Use relaxed constexpr for StringHasher
https://bugs.webkit.org/show_bug.cgi?id=180622

Reviewed by JF Bastien.

Now VC++ is updated and all the WebKit compilers support C++14 relaxed constexpr.
This patch uses relaxed constexpr for StringHasher constexpr implementation

wtf/text/StringHasher.h:

(WTF::StringHasher::addCharactersAssumingAligned):
(WTF::StringHasher::Converter):
(WTF::StringHasher::computeHashAndMaskTop8Bits):
(WTF::StringHasher::computeHash):
(WTF::StringHasher::computeLiteralHash):
(WTF::StringHasher::computeLiteralHashAndMaskTop8Bits):
(WTF::StringHasher::defaultConverter):
(WTF::StringHasher::avalancheBits):
(WTF::StringHasher::finalize):
(WTF::StringHasher::finalizeAndMaskTop8Bits):
(WTF::StringHasher::avoidZero):
(WTF::StringHasher::calculateWithRemainingLastCharacter):
(WTF::StringHasher::calculateWithTwoCharacters):
(WTF::StringHasher::processPendingCharacter const):
(WTF::StringHasher::StringHasher): Deleted.
(WTF::StringHasher::avalancheBits3): Deleted.
(WTF::StringHasher::avalancheBits2): Deleted.
(WTF::StringHasher::avalancheBits1): Deleted.
(WTF::StringHasher::avalancheBits0): Deleted.
(WTF::StringHasher::calculateWithRemainingLastCharacter1): Deleted.
(WTF::StringHasher::calculateWithRemainingLastCharacter0): Deleted.
(WTF::StringHasher::calculate1): Deleted.
(WTF::StringHasher::calculate0): Deleted.
(WTF::StringHasher::calculate): Deleted.
(WTF::StringHasher::computeLiteralHashImpl): Deleted.

11:48 AM Changeset in webkit [225725] by fpizlo@apple.com

11 edits
4 adds in trunk/Source/JavaScriptCore

InferredType should not use UnconditionalFinalizer
https://bugs.webkit.org/show_bug.cgi?id=180456

Reviewed by Saam Barati.

This turns InferredStructure into a cell so that we can unconditionally finalize them without
having to add things to the UnconditionalFinalizer list. I'm removing all uses of
UnconditionalFinalizers and WeakReferenceHarvesters because the data structures used to manage
them are a top cause of lock contention in the parallel GC. Also, we don't need those data
structures if we use IsoSubspaces, subspace iteration, and marking constraints.

JavaScriptCore.xcodeproj/project.pbxproj:
Sources.txt:
heap/Heap.cpp:

(JSC::Heap::finalizeUnconditionalFinalizers):

heap/Heap.h:
runtime/InferredStructure.cpp: Added.

(JSC::InferredStructure::create):
(JSC::InferredStructure::destroy):
(JSC::InferredStructure::createStructure):
(JSC::InferredStructure::visitChildren):
(JSC::InferredStructure::finalizeUnconditionally):
(JSC::InferredStructure::InferredStructure):
(JSC::InferredStructure::finishCreation):

runtime/InferredStructure.h: Added.
runtime/InferredStructureWatchpoint.cpp: Added.

(JSC::InferredStructureWatchpoint::fireInternal):

runtime/InferredStructureWatchpoint.h: Added.
runtime/InferredType.cpp:

(JSC::InferredType::visitChildren):
(JSC::InferredType::willStoreValueSlow):
(JSC::InferredType::makeTopSlow):
(JSC::InferredType::set):
(JSC::InferredType::removeStructure):
(JSC::InferredType::InferredStructureWatchpoint::fireInternal): Deleted.
(JSC::InferredType::InferredStructureFinalizer::finalizeUnconditionally): Deleted.
(JSC::InferredType::InferredStructure::InferredStructure): Deleted.

runtime/InferredType.h:
runtime/VM.cpp:

(JSC::VM::VM):

runtime/VM.h:

8:58 AM Changeset in webkit [225724] by Konstantin Tokarev

14 edits in trunk

[python] Replace print >> operator with print() function for python3 compatibility
https://bugs.webkit.org/show_bug.cgi?id=180611

Reviewed by Michael Catanzaro.

Source/JavaScriptCore:

Scripts/make-js-file-arrays.py:

(main):

Tools:

CygwinDownloader/cygwin-downloader.py:

(download_url_to_file):

Scripts/webkitpy/common/system/profiler.py:

(Perf.profile_after_exit):

Scripts/webkitpy/common/version_check.py:
Scripts/webkitpy/layout_tests/lint_test_expectations.py:

(main):

Scripts/webkitpy/layout_tests/run_webkit_tests.py:

(main):

Scripts/webkitpy/layout_tests/servers/run_webkit_httpd.py:

(run_server):

Scripts/webkitpy/tool/commands/analyzechangelog.py:

(ChangeLogAnalyzer._print_status):

Scripts/webkitpy/tool/commands/queries.py:

(BugsToCommit.execute):
(PatchesInCommitQueue.execute):
(PatchesToCommitQueue.execute):
(PatchesToReview._print_report):
(WhatBroke._print_builder_line):
(WhatBroke._print_blame_information_for_builder):
(WhatBroke.execute):
(ResultsFor._print_layout_test_results):
(ResultsFor.execute):
(FailureReason._print_blame_information_for_transition):
(FailureReason._explain_failures_for_builder):
(FailureReason._builder_to_explain):
(FailureReason.execute):
(FindFlakyTests._find_failures):
(FindFlakyTests._print_statistics):
(FindFlakyTests._walk_backwards_from):
(execute):
(PrintExpectations.execute):
(PrintBaselines.execute):
(PrintBaselines._print_baselines):
(FindResolvedBugs.execute):

Scripts/webkitpy/tool/commands/rebaseline.py:

(AbstractParallelRebaselineCommand._run_webkit_patch):
(AbstractParallelRebaselineCommand._rebaseline):

Scripts/webkitpy/tool/servers/gardeningserver.py:

(GardeningHTTPRequestHandler.rebaselineall):

Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:

(GardeningServerTest.disabled_test_rebaselineall.run_command):
(GardeningServerTest):

2:41 AM Changeset in webkit [225723] by rniwa@webkit.org

4 edits
2 adds in trunk

iOS: Crash in Document::updateLayout() via Document::processViewport
https://bugs.webkit.org/show_bug.cgi?id=180619
<rdar://problem/35717575>

Reviewed by Zalan Bujtas.

Source/WebCore:

The crash is caused by modern media controls updating the layout in the middle of insertedIntoAncestor
via HTMLMediaElement::setControllerJSProperty inside Document::pageScaleFactorChangedAndStable.

Fixed the crash by delaying the work to update the viewport configuration until didFinishInsertingNode
since updating the viewport configuration results in a lot of related code running in response,
and making sure all that code never tries to execute an author script is not attainable in the short term,
and a maintenance nightmare in the long term.

Test: media/ios/viewport-change-with-video.html

html/HTMLMetaElement.cpp:

(WebCore::HTMLMetaElement::insertedIntoAncestor):
(WebCore::HTMLMetaElement::didFinishInsertingNode): Added.

html/HTMLMetaElement.h:

LayoutTests:

Added a regression test for the crash.

media/ios/viewport-change-with-video-expected.txt: Added.
media/ios/viewport-change-with-video.html: Added.

Dec 8, 2017:

11:39 PM Changeset in webkit [225722] by commit-queue@webkit.org

3 edits in trunk/Tools

[Win] The way to detect Windows 10 is wrong
https://bugs.webkit.org/show_bug.cgi?id=179344
<rdar://problem/35562264>

Patch by Basuke Suzuki <Basuke Suzuki> on 2017-12-08
Reviewed by Alex Christensen.

Python2.7 doesn't return correct information on Windows 10.
Use platform.win32_ver() instead of sys.getwindowsversion().

Scripts/webkitpy/common/system/platforminfo.py:

(PlatformInfo._win_version):
(PlatformInfo._win_version_str):

Scripts/webkitpy/common/system/platforminfo_unittest.py:

(fake_sys):
(fake_sys.FakeSysModule):
(fake_platform):
(fake_platform.FakePlatformModule.win32_ver):
(TestPlatformInfo.test_os_name_and_wrappers):
(TestPlatformInfo.test_os_version):
(TestPlatformInfo.test_display_name):
(TestPlatformInfo.test_total_bytes_memory):

9:40 PM Changeset in webkit [225721] by commit-queue@webkit.org

3 edits in trunk/Tools

[WinCairo] Fix runtime error after r225229
https://bugs.webkit.org/show_bug.cgi?id=180614

Patch by Basuke Suzuki <Basuke Suzuki> on 2017-12-08
Reviewed by Alex Christensen.

The list of fallback versions doesn't match with version name mapping introduced recently.
It should be safe by creating the list automatically from the name map for consistency.

Scripts/webkitpy/common/version_name_map.py:

(VersionNameMap.init):
(VersionNameMap.to_name):
(VersionNameMap.from_name):
(VersionNameMap):
(VersionNameMap.names):
(VersionNameMap.mapping_for_platform):

Scripts/webkitpy/port/win.py:

(WinCairoPort):
(WinCairoPort.default_baseline_search_path):

9:36 PM Changeset in webkit [225720] by commit-queue@webkit.org

2 edits in trunk/Source/WebKit

Service Worker should use a correct user agent
https://bugs.webkit.org/show_bug.cgi?id=180566
<rdar://problem/35926295>

Patch by Youenn Fablet <youenn@apple.com> on 2017-12-08
Reviewed by Chris Dumez.

Addendum to landed patch.
This change was removed from the last version of the patch but proves to be needed by Safari.

UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::establishWorkerContextConnectionToStorageProcess):
Added back setting the user agent at start of service worker in case a page is already created.

7:51 PM Changeset in webkit [225719] by Alan Bujtas

3 edits
2 adds in trunk

Document::updateLayout() could destroy current frame.
https://bugs.webkit.org/show_bug.cgi?id=180525
<rdar://problem/35906836>

Reviewed by Simon Fraser.

Source/WebCore:

Early return when Document::updateLayout() triggers Frame destruction.

Test: fast/frames/crash-when-iframe-is-remove-in-eventhandler.html

dom/TreeScope.cpp:

(WebCore::absolutePointIfNotClipped):

LayoutTests:

fast/frames/crash-when-iframe-is-remove-in-eventhandler-expected.txt: Added.
fast/frames/crash-when-iframe-is-remove-in-eventhandler.html: Added.

7:44 PM Changeset in webkit [225718] by Chris Dumez

7 edits in trunk

ServiceWorkerGlobalScope is a global object and should be marked as [ImplicitThis] in the IDL
https://bugs.webkit.org/show_bug.cgi?id=180615

Reviewed by Brady Eidson.

LayoutTests/imported/w3c:

Rebaseline WPT tests now that more checks are passing.

web-platform-tests/service-workers/service-worker/interfaces-sw.https-expected.txt:
web-platform-tests/workers/interfaces.worker-expected.txt:

Source/WebCore:

ServiceWorkerGlobalScope is a global object and should be marked as [ImplicitThis] in the IDL, similarly to what we do for Window.
This allows a getter to be fallback to the global object as ThisValue when the cast of the ThisValue to the expected type fails.

No new tests, rebaselined existing test.

workers/DedicatedWorkerGlobalScope.idl:
workers/WorkerGlobalScope.idl:
workers/service/ServiceWorkerGlobalScope.idl:

7:03 PM Changeset in webkit [225717] by beidson@apple.com

13 edits in trunk/Source

Delay some service worker operations until after the database import completes.
https://bugs.webkit.org/show_bug.cgi?id=180573

Reviewed by Chris Dumez.

Source/WebCore:

No new tests (Not testable yet).

Right after the storage process launches it starts importing registration records.
During this time, a lot of the decisions we make regarding registrations, fetch, etc.
are invalid, as they rely on those in-memory records being in place.

This patch delays certain things until after the initial origin table import is complete.

workers/service/server/RegistrationStore.cpp:

(WebCore::RegistrationStore::databaseOpenedAndRecordsImported):

workers/service/server/SWOriginStore.h:

workers/service/server/SWServer.cpp:

(WebCore::SWServer::registrationStoreImportComplete): Tell the origin store the initial

import was completed.

(WebCore::SWServer::addRegistrationFromStore):

workers/service/server/SWServer.h:

Source/WebKit:

StorageProcess/ServiceWorker/WebSWOriginStore.cpp:

(WebKit::WebSWOriginStore::importComplete): Tell the Origin Table on all connects that

the import is complete.

(WebKit::WebSWOriginStore::registerSWServerConnection):

StorageProcess/ServiceWorker/WebSWOriginStore.h:

WebProcess/Storage/WebSWClientConnection.cpp:

(WebKit::WebSWClientConnection::mayHaveServiceWorkerRegisteredForOrigin const):
(WebKit::WebSWClientConnection::setSWOriginTableIsImported): Run deferred tasks!
(WebKit::WebSWClientConnection::matchRegistration): If the import isn't complete yet, delay

the match registration task until later.

(WebKit::WebSWClientConnection::runOrDelayTask): Either send the message now or save off

the task to wait until the import is complete.

(WebKit::WebSWClientConnection::getRegistrations): If the import isn't complete yet, delay

the get registrations task until later.

(WebKit::WebSWClientConnection::initializeSWOriginTableAsEmpty): Deleted.

WebProcess/Storage/WebSWClientConnection.h:
WebProcess/Storage/WebSWClientConnection.messages.in:

WebProcess/Storage/WebSWOriginTable.cpp:

(WebKit::WebSWOriginTable::setSharedMemory):

WebProcess/Storage/WebSWOriginTable.h:

(WebKit::WebSWOriginTable::isImported const):
(WebKit::WebSWOriginTable::setIsImported):
(WebKit::WebSWOriginTable::isInitialized const): Deleted.
(WebKit::WebSWOriginTable::initializeAsEmpty): Deleted.

5:05 PM Changeset in webkit [225716] by commit-queue@webkit.org

17 edits
3 adds in trunk

Service Worker should use a correct user agent
https://bugs.webkit.org/show_bug.cgi?id=180566
<rdar://problem/35926295>

Patch by Youenn Fablet <youenn@apple.com> on 2017-12-08
Reviewed by Chris Dumez.

Source/WebCore:

Test: http/wpt/service-workers/useragent.https.html

Make ServiceWorkerFrameLoaderClient return a valid UserAgent.
Pass user agent to ServiceWorkerThread so that navigation.userAgent is correctly initialized.
Allow ServiceWorkerFrameLoaderClient to clean itself when no longer needed.

loader/EmptyFrameLoaderClient.h:
workers/service/context/ServiceWorkerThread.cpp:

(WebCore::ServiceWorkerThread::ServiceWorkerThread):

workers/service/context/ServiceWorkerThread.h:
workers/service/context/ServiceWorkerThreadProxy.cpp:

(WebCore::ServiceWorkerThreadProxy::ServiceWorkerThreadProxy):
(WebCore::ServiceWorkerThreadProxy::frameLoaderClient):

workers/service/context/ServiceWorkerThreadProxy.h:

Source/WebKit:

Add support to set service worker user agent from UIProcess to service worker process.
One user agent is currently supported per service worker process and it can be changed at any given time.
Only new service worker will use the new value.
Once a service worker is launched, it will stay with the same user agent value.

This sets both navigator.userAgent and User-Agent header name used for fetch within a service worker.
Compute the service worker process user agent by picking the last user agent set for a web page.

UIProcess/ServiceWorkerProcessProxy.cpp:

(WebKit::ServiceWorkerProcessProxy::start):
(WebKit::ServiceWorkerProcessProxy::setUserAgent):

UIProcess/ServiceWorkerProcessProxy.h:
UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::setApplicationNameForUserAgent):

UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::establishWorkerContextConnectionToStorageProcess):
(WebKit::WebProcessPool::createWebPage):
(WebKit::WebProcessPool::updateServiceWorkerUserAgent):

UIProcess/WebProcessPool.h:
WebProcess/Storage/WebSWContextManagerConnection.cpp:

(WebKit::WebSWContextManagerConnection::installServiceWorker):
(WebKit::WebSWContextManagerConnection::setUserAgent):
(WebKit::WebSWContextManagerConnection::removeFrameLoaderClient):

WebProcess/Storage/WebSWContextManagerConnection.h:
WebProcess/Storage/WebSWContextManagerConnection.messages.in:

LayoutTests:

http/wpt/service-workers/useragent-worker.js: Added.

(async.doTest):

http/wpt/service-workers/useragent.https-expected.txt: Added.
http/wpt/service-workers/useragent.https.html: Added.

4:00 PM Changeset in webkit [225715] by Simon Fraser

9 edits
2 adds in trunk

Wrong caret position for input field inside a fixed position parent on iOS 11
https://bugs.webkit.org/show_bug.cgi?id=176896
rdar://problem/33726145

Reviewed by Tim Horton.
Source/WebCore:

In r219668 I added code to compute a layout viewport rect in the web process, so that
after programmatic scrolling, getBoundingClientRect() would return the correct values.
However, that computation sometimes used a different visual viewport than the UI process,
resulting in a different layout viewport being set. This would happen when the keyboard
was visible, and the combination of this and zooming when focusing an input would result
in a state where the scrolling tree contained notes computed with the bad layout viewport.
This could cause apparently offset fixed elements, and bad caret positioning if those fixed
elements contained the focused input.

Fix by passing to the web process the same visual viewport rect that the UI process is using,
namely "unobscuredContentRectRespectingInputViewBounds". This was already being set in
VisibleContentRectUpdateInfo but wasn't encoded/decoded, so fix that. Set it as an optional<>
on FrameView when different from the normal visual viewport, and return it from
visualViewportRect().

Some other minor logging changes.

Test: fast/visual-viewport/ios/caret-after-focus-in-fixed.html

page/FrameView.cpp:

(WebCore::FrameView::setVisualViewportOverrideRect):
(WebCore::FrameView::updateLayoutViewport):
(WebCore::FrameView::visualViewportRect const):

page/FrameView.h:
page/scrolling/mac/ScrollingTreeFixedNode.mm:

(WebCore::ScrollingTreeFixedNode::updateLayersAfterAncestorChange):

Source/WebKit:

Some other minor logging changes.

Shared/VisibleContentRectUpdateInfo.cpp:

(WebKit::VisibleContentRectUpdateInfo::encode const):
(WebKit::VisibleContentRectUpdateInfo::decode):
(WebKit::operator<<):

Shared/VisibleContentRectUpdateInfo.h:

(WebKit::VisibleContentRectUpdateInfo::VisibleContentRectUpdateInfo):

WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::updateVisibleContentRects):

LayoutTests:

Test that focuses an input inside position:fixed, then moves focus to the next
input. This was the most reliable way I could find of triggering the bug.
The test dumps the caret rect.

fast/visual-viewport/ios/caret-after-focus-in-fixed-expected.txt: Added.
fast/visual-viewport/ios/caret-after-focus-in-fixed.html: Added.

4:00 PM Changeset in webkit [225714] by Simon Fraser

3 edits
2 adds in trunk

When the iPhone keyboard is up, sometimes we never commit a stable update and re-show the caret
https://bugs.webkit.org/show_bug.cgi?id=180498

Reviewed by Tim Horton.

Source/WebKit:

When the keyboard is showing, we would think that the page was in a rubber-banding state
because contentOffsetBoundedInValidRange() would always clamp the content offset to a different
value.

This happened because scrollView.contentInset don't change when the keyboard is showing,
but UIKit actually consults scrollView.adjustedContentInset, which does. If we use
scrollView.adjustedContentInset in this computation, we'll get a correct answer.

Also rewrote the clamping logic to be more similar to what UIKit does internally when computing
min/max content offset.

UIProcess/API/Cocoa/WKWebView.mm:

(contentOffsetBoundedInValidRange):

LayoutTests:

Test that completes once a stable update is received after showing the keyboard.

fast/visual-viewport/ios/stable-update-with-keyboard-expected.txt: Added.
fast/visual-viewport/ios/stable-update-with-keyboard.html: Added.

3:53 PM Changeset in webkit [225713] by dino@apple.com

1 edit in trunk/Source/WebCore/html/OffscreenCanvas.cpp

Hopefully fix Windows build

3:32 PM Changeset in webkit [225712] by Alan Bujtas

23 edits
1 copy in trunk/Source/WebCore

Make inline box objects IsoHeap allocated.
https://bugs.webkit.org/show_bug.cgi?id=180556
<rdar://problem/35923629>

Reviewed by Filip Pizlo.

rendering/EllipsisBox.cpp:
rendering/EllipsisBox.h:
rendering/InlineBox.cpp:
rendering/InlineBox.h:
rendering/InlineElementBox.cpp:
rendering/InlineElementBox.h:
rendering/InlineFlowBox.cpp:
rendering/InlineFlowBox.h:
rendering/InlineTextBox.cpp:
rendering/InlineTextBox.h:
rendering/RootInlineBox.cpp:
rendering/RootInlineBox.h:
rendering/TrailingFloatsRootInlineBox.h:
rendering/svg/SVGInlineFlowBox.cpp:
rendering/svg/SVGInlineFlowBox.h:
rendering/svg/SVGInlineTextBox.cpp:
rendering/svg/SVGInlineTextBox.h:
rendering/svg/SVGRootInlineBox.cpp:
rendering/svg/SVGRootInlineBox.h:

3:13 PM Changeset in webkit [225711] by Chris Dumez

12 edits in trunk/Source

Clearing all Website Data should remove service worker registrations on disk
https://bugs.webkit.org/show_bug.cgi?id=180558

Reviewed by Youenn Fablet.

Source/WebCore:

Clear service worker registrations on disk in addition to the ones in memory.

workers/service/server/RegistrationDatabase.cpp:

(WebCore::v1RecordsTableSchema):
(WebCore::v1RecordsTableSchemaAlternate):
(WebCore::databaseFilename):
Make sure these always get called from the background thread since they use
a static string.

(WebCore::RegistrationDatabase::RegistrationDatabase):
Call importRecordsIfNecessary() instead of openSQLiteDatabase(). importRecordsIfNecessary()
only calls openSQLiteDatabase() if the database file exists, to avoid creating a database
file unnecessarily.

(WebCore::RegistrationDatabase::databasePath const):
New method which returns the database file path.

(WebCore::RegistrationDatabase::openSQLiteDatabase):

(WebCore::RegistrationDatabase::importRecordsIfNecessary):
New methods which imports records if the database file exist. It the database file does
not exist, it does not create it.

(WebCore::RegistrationDatabase::pushChanges):
Call completion handler when changes are pushed.

(WebCore::RegistrationDatabase::clearAll):
Close the database if it is open, then remove the database files.

(WebCore::RegistrationDatabase::doPushChanges):
If the database is not already open, we now open it when trying to write changes for
the first time.

workers/service/server/RegistrationDatabase.h:
workers/service/server/RegistrationStore.cpp:

(WebCore::RegistrationStore::clearAll):
(WebCore::RegistrationStore::flushChanges):

workers/service/server/RegistrationStore.h:

workers/service/server/SWServer.cpp:

(WebCore::SWServer::clearAll):
(WebCore::SWServer::clear):
Also clear the database.

workers/service/server/SWServer.h:

workers/service/server/SWServerWorker.cpp:

(WebCore::SWServerWorker::terminate):
Only call SWServer::terminateWorker() if the worker is running. Otherwise, we hit
an assertion when clearing a registration would worker was already terminated.

Source/WebKit:

StorageProcess/StorageProcess.cpp:

(WebKit::StorageProcess::deleteWebsiteData):
(WebKit::StorageProcess::deleteWebsiteDataForOrigins):

3:11 PM Changeset in webkit [225710] by dino@apple.com

1 edit in trunk/Source/WebCore/html/OffscreenCanvas.cpp

Attempted build fix.

3:03 PM Changeset in webkit [225709] by Joseph Pecoraro

15 edits
1 copy in trunk/Source

ServiceWorker Inspector: Various issues inspecting service worker on mobile.twitter.com
https://bugs.webkit.org/show_bug.cgi?id=180520
<rdar://problem/35900764>

Reviewed by Brian Burg.

Source/JavaScriptCore:

inspector/protocol/ServiceWorker.json:

Include content script content in the initialization info.

Source/WebCore:

inspector/agents/worker/ServiceWorkerAgent.cpp:

(WebCore::ServiceWorkerAgent::getInitializationInfo):

inspector/agents/worker/ServiceWorkerAgent.h:

Add initial script content to initialization so we always at least have main resource content.

Source/WebInspectorUI:

UserInterface/Main.html:
UserInterface/Test.html:

New files.

UserInterface/Controllers/SourceMapManager.js:

(WI.SourceMapManager.prototype._loadAndParseSourceMap):

UserInterface/Models/SourceMapResource.js:

(WI.SourceMapResource.prototype.requestContentFromBackend):
A ServiceWorker inspector doesn't have a main frame, fall back to an
empty frameIdentifier, it is ignored by the Service Worker network agent.

UserInterface/Base/Main.js:

Handle a Service Worker inspector where there is no frame. This can
search the main resource's resource collection.

UserInterface/Controllers/DebuggerManager.js:

(WI.DebuggerManager.prototype.scriptDidParse):

UserInterface/Controllers/FrameResourceManager.js:

(WI.FrameResourceManager.prototype.resourceRequestDidFailLoading):
In Service Workers the resources won't have a parent frame.

(WI.FrameResourceManager.prototype._addNewResourceToFrameOrTarget):
(WI.FrameResourceManager.prototype._processServiceWorkerInitializationInfo):
In service worker initialization fallback to using the script content
as the main resource if we didn't get a Script that actually includes it.

UserInterface/Models/LocalScript.js:

(WI.LocalScript):
(WI.LocalScript.prototype.requestContentFromBackend):
A way to create a local Script with content.

UserInterface/Views/SourceCodeTextEditor.js:

(WI.SourceCodeTextEditor.prototype.get _supportsDebugging):
Disallow breakpoints in a LocalScript since we have no way to tell the
backend where to set breakpoints.

UserInterface/Models/TextRange.js:

(WI.TextRange.fromText):
Add a way to get a TextRange from static text.

2:57 PM Changeset in webkit [225708] by Joseph Pecoraro

7 edits in trunk/Source/WebCore

ServiceWorker Inspector: Should be able to see image content from CacheStorage.add(url) network request
https://bugs.webkit.org/show_bug.cgi?id=180506

Reviewed by Brian Burg.

No test as this only happens inside a ServiceWorker inspector which
we don't yet have a way to test. In a Page Inspector, the
CacheStorage.add network request already behaves as expected.

inspector/NetworkResourcesData.cpp:

(WebCore::NetworkResourcesData::responseReceived):
(WebCore::shouldBufferResourceData):

inspector/NetworkResourcesData.h:

(WebCore::NetworkResourcesData::ResourceData::forceBufferData const):
(WebCore::NetworkResourcesData::ResourceData::setForceBufferData):
Provide a way to force buffering in NetworkResourceData.

inspector/agents/InspectorNetworkAgent.cpp:

(WebCore::InspectorNetworkAgent::didReceiveResponse):

inspector/agents/InspectorNetworkAgent.h:
inspector/agents/page/PageNetworkAgent.h:
inspector/agents/worker/WorkerNetworkAgent.h:

Enable force buffering in a Worker Network Agent.

2:51 PM Changeset in webkit [225707] by dbates@webkit.org

17 edits in trunk

Remove unnecessary prefix from AutoFillButtonType enumerators
https://bugs.webkit.org/show_bug.cgi?id=180512

Reviewed by Tim Horton.

Source/WebCore:

testing/Internals.cpp:

(WebCore::toAutoFillButtonType):

testing/Internals.h:
testing/Internals.idl:

LayoutTests:

accessibility/auto-fill-crash.html:
accessibility/auto-fill-types.html:
fast/forms/auto-fill-button/hide-auto-fill-button-when-input-becomes-disabled.html:
fast/forms/auto-fill-button/hide-auto-fill-button-when-input-becomes-readonly.html:
fast/forms/auto-fill-button/input-auto-fill-button.html:
fast/forms/auto-fill-button/input-contacts-auto-fill-button.html:
fast/forms/auto-fill-button/input-disabled-auto-fill-button.html:
fast/forms/auto-fill-button/input-readonly-auto-fill-button.html:
fast/forms/auto-fill-button/input-readonly-non-empty-auto-fill-button.html:
fast/forms/auto-fill-button/mouse-down-input-mouse-release-auto-fill-button.html:
fast/forms/auto-fill-button/show-correct-auto-fill-button-when-auto-fill-button-type-changes-expected.html:
fast/forms/auto-fill-button/show-correct-auto-fill-button-when-auto-fill-button-type-changes.html:

2:46 PM Changeset in webkit [225706] by dino@apple.com

5 edits
2 moves
2 adds in trunk

Implement transferToImageBitmap for WebGL offscreen canvas objects
https://bugs.webkit.org/show_bug.cgi?id=180603
<rdar://problem/34147143>

Reviewed by Sam Weinig.

Source/WebCore:

Implement the basic version of creating an ImageBitmap from an
OffscreenCanvas that is using a WebGL context.

Tests: http/wpt/offscreen-canvas/transferToImageBitmap-empty.html

http/wpt/offscreen-canvas/transferToImageBitmap-webgl.html

html/ImageBitmap.cpp:

(WebCore::ImageBitmap::create): Add a new constructor to be used by OffscreenCanvas.
Creates a blank ImageBitmap.

html/ImageBitmap.h:

html/OffscreenCanvas.cpp:

(WebCore::OffscreenCanvas::transferToImageBitmap): Create a new ImageBitmap
and paint the current canvas into it.

LayoutTests:

http/wpt/offscreen-canvas/transferToImageBitmap-empty-expected.txt: Renamed from LayoutTests/http/wpt/offscreen-canvas/transferToImageBitmap-expected.txt.
http/wpt/offscreen-canvas/transferToImageBitmap-empty.html: Renamed from LayoutTests/http/wpt/offscreen-canvas/transferToImageBitmap.html.
http/wpt/offscreen-canvas/transferToImageBitmap-webgl-expected.html: Added.
http/wpt/offscreen-canvas/transferToImageBitmap-webgl.html: Added.

2:45 PM Changeset in webkit [225705] by commit-queue@webkit.org

4 edits in trunk/Source

WebServiceWorkerProvider should use Cancellation error to notify DTL that it cannot handle a fetch
https://bugs.webkit.org/show_bug.cgi?id=180584

Patch by Youenn Fablet <youenn@apple.com> on 2017-12-08
Reviewed by Alex Christensen.

Source/WebCore:

Previously, for each cross origin fetch that is going through a service worker and service worker is not handling the fetch,
we return an AccessControl error so that DocumentThreadableLoader will do preflight and regular load through the network.
This error is wrongly logged in the Inspector.
Change error type to Cancellation so that the Inspector does not log it.

loader/DocumentThreadableLoader.cpp:

(WebCore::DocumentThreadableLoader::didFail):

Source/WebKit:

WebProcess/Network/WebLoaderStrategy.cpp:

(WebKit::WebLoaderStrategy::scheduleLoad):

2:44 PM Changeset in webkit [225704] by Michael Catanzaro

2 edits in trunk/Source/WebInspectorUI

Don't require perl(File::Copy::Recursive)
https://bugs.webkit.org/show_bug.cgi?id=180479

Reviewed by Konstantin Tokarev.

If File::Copy::Recursive is not installed, there is currently a Darwin
fallback to the ditto command. That doesn't exist on Linux, but we can
emulate it by running 'cp -R' on everything in the source directory.
This should probably work everywhere except Windows.

Scripts/copy-user-interface-resources.pl:

(ditto):

2:44 PM Changeset in webkit [225703] by commit-queue@webkit.org

12 edits
3 adds in trunk

Service Worker should use a correct SessionID
https://bugs.webkit.org/show_bug.cgi?id=180585

Patch by Youenn Fablet <youenn@apple.com> on 2017-12-08
Reviewed by Alex Christensen.

Source/WebCore:

Test: http/tests/workers/service/serviceworker-private-browsing.https.html

Store SessionID in SWServer and send it as part of service worker instantiation.

workers/service/server/SWServer.cpp:

(WebCore::SWServer::SWServer):
(WebCore::SWServer::installContextData):
(WebCore::SWServer::runServiceWorker):

workers/service/server/SWServer.h:
workers/service/server/SWServerToContextConnection.h:

Source/WebKit:

Store SessionID in SWServer and send it as part of service worker instantiation.
Use it when creating service worker thread in service worker process.

StorageProcess/ServiceWorker/WebSWServerToContextConnection.cpp:

(WebKit::WebSWServerToContextConnection::installServiceWorkerContext):

StorageProcess/ServiceWorker/WebSWServerToContextConnection.h:
StorageProcess/StorageProcess.cpp:

(WebKit::StorageProcess::swServerForSession):

WebProcess/Storage/WebSWContextManagerConnection.cpp:

(WebKit::WebSWContextManagerConnection::installServiceWorker):

WebProcess/Storage/WebSWContextManagerConnection.h:
WebProcess/Storage/WebSWContextManagerConnection.messages.in:

LayoutTests:

http/tests/workers/service/serviceworker-private-browsing-worker.js: Added.

(async):

http/tests/workers/service/serviceworker-private-browsing.https-expected.txt: Added.
http/tests/workers/service/serviceworker-private-browsing.https.html: Added.

2:41 PM Changeset in webkit [225702] by commit-queue@webkit.org

13 edits in trunk

FetchResponse should keep unfiltered ResourceResponse so that it can be used in Service Worker
https://bugs.webkit.org/show_bug.cgi?id=179641
<rdar://problem/35923570>

Patch by Youenn Fablet <youenn@apple.com> on 2017-12-08
Reviewed by Alex Christensen.

LayoutTests/imported/w3c:

web-platform-tests/service-workers/service-worker/fetch-request-css-cross-origin-mime-check.https-expected.txt:
web-platform-tests/service-workers/service-worker/fetch-response-taint.https-expected.txt:

Source/WebCore:

Covered by existing rebased tests.

FetchResponse will now store an unfiltered response.
If it needs to expose it to JavaScript, it will create a filtered response lazily.
This allows service worker to send back to web pages, opaque responses containing every information.

Updating Document::initSecurityContext so that any document loaded with a response whose tainting is Opaque gets a unique origin.
This ensures cross-origin checks to work if service worker returns such a response on a same origin URL.

Updated SubresourceLoader to check cross origin service worker responses based on their tainting.

Modules/fetch/FetchResponse.cpp:

(WebCore::FetchResponse::create):
(WebCore::FetchResponse::error):
(WebCore::FetchResponse::redirect):
(WebCore::FetchResponse::FetchResponse):
(WebCore::FetchResponse::clone):
(WebCore::FetchResponse::url const):
(WebCore::FetchResponse::filteredResponse const):
(WebCore::FetchResponse::BodyLoader::didReceiveResponse):
(WebCore::FetchResponse::resourceResponse const):

Modules/fetch/FetchResponse.h:

(WebCore::FetchResponse::create): Deleted.

dom/Document.cpp:

(WebCore::Document::initSecurityContext):

loader/SubresourceLoader.cpp:

(WebCore::SubresourceLoader::willSendRequestInternal):
(WebCore::SubresourceLoader::checkResponseCrossOriginAccessControl):

Source/WebKit:

WebProcess/Storage/ServiceWorkerClientFetch.cpp:

(WebKit::ServiceWorkerClientFetch::didReceiveResponse): In case of opaque redirected response, handle it as a regular response.

LayoutTests:

TestExpectations:

2:28 PM Changeset in webkit [225701] by sbarati@apple.com

4 edits in trunk/Source/bmalloc

Enable gigacage on iOS with a 32GB runway and ensure it doesn't break WasmBench
https://bugs.webkit.org/show_bug.cgi?id=178557

Reviewed by Mark Lam.

bmalloc/Algorithm.h:

(bmalloc::isPowerOfTwo):

bmalloc/Gigacage.cpp:
bmalloc/Gigacage.h:

2:23 PM Changeset in webkit [225700] by achristensen@apple.com

4 edits in trunk/Source/WebKit

Remove unused code in WebPageGroup
https://bugs.webkit.org/show_bug.cgi?id=180604

Reviewed by Youenn Fablet.

UIProcess/WebPageGroup.cpp:

(WebKit::WebPageGroup::userContentController):
(WebKit::WebPageGroup::createNonNull): Deleted.
(WebKit::WebPageGroup::preferencesDidChange): Deleted.

UIProcess/WebPageGroup.h:
UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::WebProcessPool):

2:05 PM Changeset in webkit [225699] by commit-queue@webkit.org

5 edits in trunk

ApplicationManifestParser should strip whitespace from the raw input
https://bugs.webkit.org/show_bug.cgi?id=180539
rdar://problem/35915075

Patch by David Quesada <david_quesada@apple.com> on 2017-12-08
Reviewed by Joseph Pecoraro.

Source/WebCore:

Modules/applicationmanifest/ApplicationManifestParser.cpp:

(WebCore::ApplicationManifestParser::parseManifest):

Tools:

Added an API test for parsing manifests with surrounding whitespace.

Also drive-by fix ApplicationManifestParserTest.Display. Earlier versions of the
manifest spec explicitly stated that the "display" value should be treated as if
it were run through String.prototype.trim(), which allowed for the really weird
edge case of an array containing one string. This behavior was lost when I changed
ApplicationManifestParser's JSON parsing from using JavaScriptCore to WTF's JSON
parsing. Update the unit test accordingly.

TestWebKitAPI/Tests/WebCore/ApplicationManifestParser.cpp:

(TEST_F):

1:56 PM Changeset in webkit [225698] by Konstantin Tokarev

63 edits in trunk

[python] Replace print operator with print() function for python3 compatibility
https://bugs.webkit.org/show_bug.cgi?id=180592

Reviewed by Michael Catanzaro.

PerformanceTests:

JSBench/harness.py:

Source/JavaScriptCore:

Scripts/generateYarrUnicodePropertyTables.py:

(openOrExit):
(verifyUCDFilesExist):
(Aliases.parsePropertyAliasesFile):
(Aliases.parsePropertyValueAliasesFile):

Scripts/make-js-file-arrays.py:

(main):

generate-bytecode-files:

Source/WebCore/PAL:

AVFoundationSupport.py:

Tools:

BuildSlaveSupport/wait-for-SVN-server.py:

(getLatestSVNRevision):
(waitForSVNRevision):

Scripts/download-latest-github-release.py:

(find_latest_release):
(main):

Scripts/update-wasm-gcc-torture.py:

(update_lkgr):
(untar_torture):
(list_js_files):
(waterfall_known_failures):

Scripts/update-webkit-wincairo-libs.py:
Scripts/webkitpy/benchmark_runner/benchmark_runner.py:

(BenchmarkRunner.show_results):

Scripts/webkitpy/benchmark_runner/http_server_driver/simple_http_server_driver.py:

(SimpleHTTPServerDriver.fetch_result):

Scripts/webkitpy/benchmark_runner/run_benchmark.py:

(list_benchmark_plans):

Scripts/webkitpy/bindings/main.py:

(BindingsTests.generate_from_idl):
(BindingsTests.generate_supplemental_dependency):
(BindingsTests.detect_changes):
(BindingsTests.run_tests):
(BindingsTests.main):

Scripts/webkitpy/codegen/main.py:

(BuiltinsGeneratorTests.generate_from_js_builtins):
(BuiltinsGeneratorTests.detect_changes):
(BuiltinsGeneratorTests.single_builtin_test):
(BuiltinsGeneratorTests.run_test):
(BuiltinsGeneratorTests.run_tests):
(BuiltinsGeneratorTests.main):

Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:

(MockBugzilla.fetch_attachment):

Scripts/webkitpy/common/net/buildbot/buildbot.py:

(Builder._fetch_revision_to_build_map):

Scripts/webkitpy/common/system/crashlogs.py:

(CrashLogs._find_newest_log_win):

Scripts/webkitpy/common/system/outputcapture_unittest.py:

(OutputCaptureTest.test_output_capture_scope):

Scripts/webkitpy/common/system/profiler.py:

(GooglePProf.profile_after_exit):
(Perf.profile_after_exit):

Scripts/webkitpy/common/system/user.py:

(User.prompt_with_multiple_lists):
(User.prompt_with_list):
(User._warn_if_application_is_xcode):
(User.edit_changelog):

Scripts/webkitpy/inspector/main.py:

(InspectorGeneratorTests.generate_from_json):
(InspectorGeneratorTests.detect_changes):
(InspectorGeneratorTests.run_tests):
(InspectorGeneratorTests.main):

Scripts/webkitpy/layout_tests/controllers/manager.py:

(Manager._print_expectation_line_for_test):
(Manager._print_expectations_for_subset):
(Manager._print_expectations_for_subset.if):

Scripts/webkitpy/layout_tests/servers/run_webkit_httpd.py:

(run_server):

Scripts/webkitpy/port/config_standalone.py:

(main):

Scripts/webkitpy/port/darwin_testcase.py:

(DarwinTest.test_spindump.logging_run_command):
(DarwinTest.test_sample_process.logging_run_command):

Scripts/webkitpy/port/ios.py:

(IOSPort.clean_up_test_run):

Scripts/webkitpy/port/ios_device_unittest.py:

(IOSDeviceTest.test_spindump.logging_run_command):
(IOSDeviceTest.test_sample_process.logging_run_command):

Scripts/webkitpy/port/ios_simulator_unittest.py:

(IOSSimulatorTest.test_xcrun.throwing_run_command):

Scripts/webkitpy/port/leakdetector_unittest.py:

(test_count_total_bytes_and_unique_leaks.mock_run_script):

Scripts/webkitpy/port/mac_unittest.py:

(MacTest.test_xcrun.throwing_run_command):

Scripts/webkitpy/style/checkers/contributors.py:

(ContributorsChecker.check):

Scripts/webkitpy/tool/commands/abstractlocalservercommand.py:

(AbstractLocalServerCommand.execute):

Scripts/webkitpy/tool/commands/adduserstogroups.py:

(AddUsersToGroups.execute):

Scripts/webkitpy/tool/commands/analyzechangelog.py:

(AnalyzeChangeLog._generate_jsons):
(AnalyzeChangeLog.execute):
(ChangeLogAnalyzer._set_filename):
(ChangeLogAnalyzer.analyze):

Scripts/webkitpy/tool/commands/bugfortest.py:

(BugForTest.execute):

Scripts/webkitpy/tool/commands/bugsearch.py:

(execute):

Scripts/webkitpy/tool/commands/findusers.py:

(FindUsers.execute):

Scripts/webkitpy/tool/commands/gardenomatic.py:

(GardenOMatic.execute):

Scripts/webkitpy/tool/commands/rebaseline.py:

(RebaselineTest.execute):

Scripts/webkitpy/tool/commands/rebaselineserver.py:

(RebaselineServer._prepare_config):

Scripts/webkitpy/tool/commands/setupgitclone.py:

(SetupGitClone.execute):
(SetupGitClone._get_username_and_email):

Scripts/webkitpy/tool/commands/suggestnominations.py:

(SuggestNominations._print_nominations):
(SuggestNominations._print_counts):

Scripts/webkitpy/tool/commands/upload.py:

(CommitMessageForCurrentDiff.execute):
(CreateBug.prompt_for_bug_title_and_comment):

Scripts/webkitpy/tool/multicommandtool.py:

(HelpCommand.execute):

Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:

(TestGardeningHTTPRequestHandler._serve_text):
(TestGardeningHTTPRequestHandler._serve_json):

Scripts/webkitpy/tool/steps/addsvnmimetypeforpng.py:

(AddSvnMimetypeForPng.run):

Scripts/webkitpy/tool/steps/suggestreviewers.py:

(SuggestReviewers.run):

Scripts/webkitpy/w3c/test_importer.py:

(TestImporter.update_tests_options):

Scripts/webkitpy/webdriver_tests/webdriver_test_runner_selenium.py:

(WebDriverTestRunnerSelenium.run):

TestResultServer/model/jsonresults_unittest.py:
gtk/ycm_extra_conf.py:

(FlagsForFile):

lldb/lldb_webkit.py:

(btjs):

LayoutTests:

html5lib/generate-test-wrappers.py:

(_remove_stale_tests):

http/tests/websocket/tests/hybi/bad-handshake-crash_wsh.py: