WebKit-https.git
2 years agoUndefined subroutine in svn-create-patch
jbedard@apple.com [Thu, 2 Feb 2017 18:33:29 +0000 (18:33 +0000)]
Undefined subroutine in svn-create-patch
https://bugs.webkit.org/show_bug.cgi?id=167742

Unreviewed build-fix after r211543.

* Scripts/VCSUtils.pm: fixSVNPatchForAdditionWithHistory was not exported.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211570 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Crash] com.apple.WebKit.WebContent at WebKit: WebKit::WebPage::fromCorePage()
cdumez@apple.com [Thu, 2 Feb 2017 18:31:54 +0000 (18:31 +0000)]
[Crash] com.apple.WebKit.WebContent at WebKit: WebKit::WebPage::fromCorePage()
https://bugs.webkit.org/show_bug.cgi?id=167738
<rdar://problem/30229990>

Reviewed by Andreas Kling.

Source/WebCore:

Upon destruction of a Page, we destroy the BackForwardClient, which is supposed
to keep track of HistoryItems associated to this particular page and remove them
from the PageCache. Given the crash trace, the issue seems to be that some
HistoryItems associated with the Page sometimes linger in the PageCache *after*
the Page has been destroyed, which leads to crashes later on when pruning the
PageCache.

In order to make the process more robust, this patch refactors the code so that
the Page is now in charge of removing all its associated HistoryItems from the
PageCache instead of relying on the BackForwardClient. Also, instead of having
the Page keep track of which HistoryItems are associated with it (which is
error prone), we now scan all PageCache entries instead to find which ones are
associated with the Page. While this is in theory slower, this is much safer
and in practice not an issue because the PageCache usually has 3-5 entries.

No new tests, could not reproduce.

* history/CachedPage.cpp:
(WebCore::CachedPage::CachedPage):
* history/CachedPage.h:
(WebCore::CachedPage::page):
* history/PageCache.cpp:
(WebCore::PageCache::removeAllItemsForPage):
* history/PageCache.h:
* page/Page.cpp:
(WebCore::Page::~Page):

Source/WebKit/mac:

The BackForwardClient no longer needs to worry about removing HistoryItems
from the PageCache now that WebCore takes care of it.

* History/BackForwardList.mm:
(BackForwardList::close):

Source/WebKit/win:

The BackForwardClient no longer needs to worry about removing HistoryItems
from the PageCache now that WebCore takes care of it.

* BackForwardList.cpp:
(BackForwardList::close):

Source/WebKit2:

The BackForwardClient no longer needs to worry about removing HistoryItems
from the PageCache now that WebCore takes care of it.

* WebProcess/WebPage/WebBackForwardListProxy.cpp:
(WebKit::WebBackForwardListProxy::addItemFromUIProcess):
(WebKit::WebBackForwardListProxy::addItem):
(WebKit::WebBackForwardListProxy::close):
* WebProcess/WebPage/WebBackForwardListProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211569 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago<rdar://problem/30323148> Webkit Nightly on 10.10 broken
andersca@apple.com [Thu, 2 Feb 2017 17:52:44 +0000 (17:52 +0000)]
<rdar://problem/30323148> Webkit Nightly on 10.10 broken

Follow-up fix.

* Platform/IPC/mac/ConnectionMac.mm:
(IPC::Connection::open):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211568 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed. Fix jhbuild dependencies build.
calvaris@igalia.com [Thu, 2 Feb 2017 17:46:20 +0000 (17:46 +0000)]
Unreviewed. Fix jhbuild dependencies build.

* gtk/patches/gst-plugins-good-0004-qtdemux-add-context-for-a-preferred-protection.patch:
Patch updated.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211567 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago<rdar://problem/30323148> Webkit Nightly on 10.10 broken
andersca@apple.com [Thu, 2 Feb 2017 17:39:57 +0000 (17:39 +0000)]
<rdar://problem/30323148> Webkit Nightly on 10.10 broken

Try to fix the 10.10 build.

* Platform/IPC/mac/ConnectionMac.mm:
(IPC::Connection::open):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211566 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoHandle m_processLauncher being null in ChildProcessProxy::processIdentifier()
andersca@apple.com [Thu, 2 Feb 2017 17:37:16 +0000 (17:37 +0000)]
Handle m_processLauncher being null in ChildProcessProxy::processIdentifier()
https://bugs.webkit.org/show_bug.cgi?id=167713
rdar://problem/28896113

Reviewed by Dan Bernstein.

This can happen if the process has been explicitly terminated.

* UIProcess/ChildProcessProxy.h:
(WebKit::ChildProcessProxy::processIdentifier):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211565 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoColumn progression wrong after enabling pagination on RTL document
antti@apple.com [Thu, 2 Feb 2017 17:06:25 +0000 (17:06 +0000)]
Column progression wrong after enabling pagination on RTL document
https://bugs.webkit.org/show_bug.cgi?id=167733
<rdar://problem/29012252>

Reviewed by Zalan Bujtas.

Source/WebCore:

Column progression depends on document direction but was not updated when direction changed.

Test: fast/multicol/pagination/pagination-dynamic-rtl.html

* rendering/RenderBox.cpp:
(WebCore::RenderBox::styleDidChange):

    Update column styles if document direction changes.

LayoutTests:

* fast/multicol/pagination/pagination-dynamic-rtl-expected.html: Added.
* fast/multicol/pagination/pagination-dynamic-rtl.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211564 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMarking http/tests/contentextensions/block-everything-unless-domain.html as flaky...
ryanhaddad@apple.com [Thu, 2 Feb 2017 17:04:42 +0000 (17:04 +0000)]
Marking http/tests/contentextensions/block-everything-unless-domain.html as flaky on macOS.
https://bugs.webkit.org/show_bug.cgi?id=167653

Unreviewed test gardening.

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211563 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[EME][GStreamer] Fixed decryptor selection
calvaris@igalia.com [Thu, 2 Feb 2017 16:54:48 +0000 (16:54 +0000)]
[EME][GStreamer] Fixed decryptor selection
https://bugs.webkit.org/show_bug.cgi?id=167588

Reviewed by Žan Doberšek.

Source/WebCore:

GStreamer selected a decryptor automatically but we need to be
able to tell it which decryptor we want because one file can be
encrypted with more than one system and the one we want to use can
depend on what the user tells us.

Now when the demuxer is about to select the demuxer, it runs a
GstContext query with the events, which are forwarded through the
EME api to the application, which will answer the the
generateKeyRequest and then we can instruct the demuxer to select
the one we tell it, not the one that is selected automatically. If
the demuxer has already a preferred decryptor, the codepath will
be similar to the older one.

Something that is also fixed is the report of the keyneeded event,
which will contain all pssh boxes as requested by the spec.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::handleMessage): Intercept
the event and handle it.
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::MediaPlayerPrivateGStreamerBase):
Initialize the CDM session.
(WebCore::MediaPlayerPrivateGStreamerBase::~MediaPlayerPrivateGStreamerBase):
Unlock the protection condition just in case and set CDM to null.
(WebCore::extractEventsAndSystemsFromMessage): Access the events
and accepted key systems coming at the message from the demuxer.
(WebCore::MediaPlayerPrivateGStreamerBase::handleSyncMessage):
Handle the protection event not as an element event but as a
context request. It also concatenates all the initdatas in case
there is more than one. The event is sent in the main thread and
the current one blocks until a timeout is hit or the
generateKeyRequest is called.
(WebCore::MediaPlayerPrivateGStreamerBase::handleProtectionEvent):
Check if the event was already handled and if not, run the need
key event.
(WebCore::MediaPlayerPrivateGStreamerBase::receivedGenerateKeyRequest):
Notify the that the query arrived so that we can unlock the
demuxer thread.
(WebCore::keySystemIdToUuid): Translates ids to uuids, meaning
the ones coming from JS to the ones coming from the media engine.
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
Attributes and methods.
* platform/graphics/gstreamer/eme/WebKitClearKeyDecryptorGStreamer.cpp:
(webkit_media_clear_key_decrypt_class_init):
(webKitMediaClearKeyDecryptorRequestDecryptionKey): Deleted.
* platform/graphics/gstreamer/eme/WebKitClearKeyDecryptorGStreamer.h:
Added the ui and uuid.
* platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.cpp:
(webkitMediaCommonEncryptionDecryptSinkEventHandler): Do not rely
on the subclass, just forward the event to the pipeline inside a message.
* platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.h:
Remove unused function signature.
* platform/graphics/gstreamer/mse/AppendPipeline.cpp:
(WebCore::appendPipelineNeedContextMessageCallback): Renamed from
appendPipelineElementMessageCallback.
(WebCore::AppendPipeline::AppendPipeline): Connect to the
need-context message instead of the element one.
(WebCore::AppendPipeline::dispatchPendingDecryptionKey):
Dispatches the drm-cypher message if there is one pending.
(WebCore::AppendPipeline::dispatchDecryptionKey): Dispatches the
drm-cypher.
(WebCore::AppendPipeline::handleNeedContextSyncMessage): Handles
the need-context event instead of the element one.
(WebCore::AppendPipeline::connectDemuxerSrcPadToAppsinkFromAnyThread):
Dispatches the key to the pipeline if there's any pending.
(WebCore::appendPipelineElementMessageCallback): Deleted.
(WebCore::AppendPipeline::handleElementMessage): Deleted.
* platform/graphics/gstreamer/mse/AppendPipeline.h: New methods
and attributes.
* platform/graphics/gstreamer/mse/MediaPlayerPrivateGStreamerMSE.cpp:
(WebCore::MediaPlayerPrivateGStreamerMSE::dispatchDecryptionKey):
Use the dispatchDecryptionKey method of the AppendPipeline.

Tools:

This patches are pending at https://bugzilla.gnome.org/show_bug.cgi?id=770107

* gtk/jhbuild.modules: Added patches to gstreamer and
gst-plugins-good packages.
* gtk/patches/gst-plugins-good-0004-qtdemux-add-context-for-a-preferred-protection.patch: Added.
* gtk/patches/gstreamer-0001-protection-added-function-to-filter-system-ids.patch: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211562 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoLayoutTest media/modern-media-controls/buttons-container/buttons-container-buttons...
graouts@webkit.org [Thu, 2 Feb 2017 16:30:28 +0000 (16:30 +0000)]
LayoutTest media/modern-media-controls/buttons-container/buttons-container-buttons-property.html is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=167371

Unreviewed test gardening.

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211561 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoSkip fast/visual-viewport/ios/bottom-bar-with-keyboard.html on ios-simulator.
ryanhaddad@apple.com [Thu, 2 Feb 2017 16:08:34 +0000 (16:08 +0000)]
Skip fast/visual-viewport/ios/bottom-bar-with-keyboard.html on ios-simulator.

Unreviewed test gardening.

* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211560 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, rolling out r211529.
ryanhaddad@apple.com [Thu, 2 Feb 2017 16:04:29 +0000 (16:04 +0000)]
Unreviewed, rolling out r211529.

This change caused the test to fail on macOS.

Reverted changeset:

"Clean up css3/filters/backdrop/backdrop-filter-with-clip-
path.html expected results."
http://trac.webkit.org/changeset/211529

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211559 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoLayoutTest media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscre...
commit-queue@webkit.org [Thu, 2 Feb 2017 14:59:34 +0000 (14:59 +0000)]
LayoutTest media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-styles.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=167372

Unreviewed test gardening.

Patch by Antoine Quint <graouts@apple.com> on 2017-02-02

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211558 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoLayoutTest media/modern-media-controls/media-controller/media-controller-resize.html...
commit-queue@webkit.org [Thu, 2 Feb 2017 14:26:22 +0000 (14:26 +0000)]
LayoutTest media/modern-media-controls/media-controller/media-controller-resize.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=164571

Unreviewed test gardening.

Patch by Antoine Quint <graouts@apple.com> on 2017-02-02

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211557 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoLayoutTest media/modern-media-controls/rewind-button/rewind-button.html is a flaky...
commit-queue@webkit.org [Thu, 2 Feb 2017 14:22:04 +0000 (14:22 +0000)]
LayoutTest media/modern-media-controls/rewind-button/rewind-button.html is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=167352

Unreviewed test gardening.

Patch by Antoine Quint <graouts@apple.com> on 2017-02-02

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211556 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRemoved unused EventHandler members
commit-queue@webkit.org [Thu, 2 Feb 2017 10:38:00 +0000 (10:38 +0000)]
Removed unused EventHandler members
https://bugs.webkit.org/show_bug.cgi?id=167723

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2017-02-02
Reviewed by Yusuke Suzuki.

* page/EventHandler.cpp:
(WebCore::EventHandler::clear):
(WebCore::EventHandler::handleWheelEvent):
* page/EventHandler.h:
Remove unused members.

* editing/FrameSelection.cpp:
* html/HTMLTextAreaElement.cpp:
* rendering/RenderObject.cpp:
* rendering/RenderSlider.cpp:
Remove unnecessary includes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211555 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWeb Inspector: Upgrade Esprima to the latest one to support dynamic import
utatane.tea@gmail.com [Thu, 2 Feb 2017 10:31:18 +0000 (10:31 +0000)]
Web Inspector: Upgrade Esprima to the latest one to support dynamic import
https://bugs.webkit.org/show_bug.cgi?id=167698

Reviewed by Joseph Pecoraro.

Source/WebInspectorUI:

* UserInterface/External/Esprima/esprima.js:
Update to Esprima@5307e30 (4.0.0-dev).

* UserInterface/Models/ScriptSyntaxTree.js:
(WebInspector.ScriptSyntaxTree.prototype._gatherIdentifiersInDeclaration.gatherIdentifiers):
(WebInspector.ScriptSyntaxTree.prototype._gatherIdentifiersInDeclaration):
(WebInspector.ScriptSyntaxTree.prototype._recurse):
(WebInspector.ScriptSyntaxTree.prototype._createInternalSyntaxTree):
(WebInspector.ScriptSyntaxTree):
* UserInterface/Workers/Formatter/ESTreeWalker.js:
(ESTreeWalker.prototype._walkChildren):
(ESTreeWalker):
Add new nodes, SpreadProperty, RestProperty, and Import.
SpreadProperty and RestProperty are the part of ES2018 rest and spread properties.
https://github.com/sebmarkbage/ecmascript-rest-spread
Import is dynamic import node. The syntax is similar to Super.
https://github.com/tc39/proposal-dynamic-import

* UserInterface/Workers/Formatter/EsprimaFormatter.js:
(EsprimaFormatter.prototype._handleTokenAtNode):

LayoutTests:

* inspector/formatting/formatting-javascript-expected.txt:
* inspector/formatting/formatting-javascript.html:
* inspector/formatting/resources/javascript-tests/import-expected.js: Added.
(async.load):
* inspector/formatting/resources/javascript-tests/import.js: Added.
(async.load):
* inspector/formatting/resources/javascript-tests/object-array-literal-expected.js:
* inspector/formatting/resources/javascript-tests/object-array-literal.js:
* inspector/formatting/resources/javascript-tests/variable-declaration-expected.js:
* inspector/formatting/resources/javascript-tests/variable-declaration.js:
* inspector/formatting/resources/utilities.js:
(TestPage.registerInitializer.):
(TestPage.registerInitializer):
Fix a bug which occurs when output is not expected one.

* inspector/model/parse-script-syntax-tree-expected.txt:
* inspector/model/parse-script-syntax-tree.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211554 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRemoved unused m_navigationStart from NavigatorGamepad
commit-queue@webkit.org [Thu, 2 Feb 2017 10:30:17 +0000 (10:30 +0000)]
Removed unused m_navigationStart from NavigatorGamepad
https://bugs.webkit.org/show_bug.cgi?id=167722

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2017-02-02
Reviewed by Yusuke Suzuki.

* Modules/gamepad/NavigatorGamepad.cpp:
(WebCore::NavigatorGamepad::NavigatorGamepad):
(WebCore::NavigatorGamepad::from):
* Modules/gamepad/NavigatorGamepad.h:
(WebCore::NavigatorGamepad::navigationStart): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211553 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRemoved unused m_errorHandlingModeReentry from Interpreter
commit-queue@webkit.org [Thu, 2 Feb 2017 09:49:53 +0000 (09:49 +0000)]
Removed unused m_errorHandlingModeReentry from Interpreter
https://bugs.webkit.org/show_bug.cgi?id=167726

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2017-02-02
Reviewed by Yusuke Suzuki.

* interpreter/Interpreter.cpp:
(JSC::Interpreter::Interpreter):
* interpreter/Interpreter.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211552 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoIn iOS, we should take background assertion when accessing localstorage databases.
commit-queue@webkit.org [Thu, 2 Feb 2017 08:33:29 +0000 (08:33 +0000)]
In iOS, we should take background assertion when accessing localstorage databases.
https://bugs.webkit.org/show_bug.cgi?id=165478

Source/WebCore:

Move WebSQLiteDatabaseTrackerClient from WebKitLegacy to WebCore so that it can be accessible from
WebKit1 and WebKit2. Previously, to avoid dependencies on UIKit, WebKitLegacy introduced several
global methods for UIKit to setup the start/end background task blocks on runtime (WebKitSetStartBackgroundTaskBlock,
WebKitSetInvalidWebBackgroundTaskIdentifier and WebKitSetEndBackgroundTaskBlock). Since we have to
move the background task handling to WebCore, to avoid adding WebCore dependencies on UIKit, this
patch introdues a new WebCore class WebBackgroundTaskController which holds the start/end background
task blocks. The existing WebKitSetStartBackgroundTaskBlock and WebKitSetEndBackgroundTaskBlock methods
in WebKit1 will use WebBackgroundTaskController to store the blocks set by UIKit.

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2017-02-02
Reviewed by Brady Eidson.

No new test since this is code refactoring.

* WebCore.xcodeproj/project.pbxproj: Add a new class WebBackgroundTaskController to the project. Also move
    WebSQLiteDatabaseTrackerClient from WebKitLegacy to WebCore.
* platform/ios/WebBackgroundTaskController.h: Use properties to hold the blocks for starting or ending background tasks.
* platform/ios/WebBackgroundTaskController.mm:
(+[WebBackgroundTaskController sharedController]):
(-[WebBackgroundTaskController dealloc]):
(-[WebBackgroundTaskController startBackgroundTaskWithExpirationHandler:]): Start a background task with a expiration handler;
    to start the background task, we will use backgroundTaskStartBlock set up by UIKit.
(-[WebBackgroundTaskController endBackgroundTaskWithIdentifier:]): Call backgroundTaskEndBlack to end a background task.
* platform/ios/WebSQLiteDatabaseTrackerClient.h: Copied from Source/WebKit/ios/Storage/WebSQLiteDatabaseTrackerClient.h.
* platform/ios/WebSQLiteDatabaseTrackerClient.mm: Copied from Source/WebKit/ios/Storage/WebSQLiteDatabaseTrackerClient.mm.
(WebCore::WebSQLiteDatabaseTrackerClient::sharedWebSQLiteDatabaseTrackerClient): Make WebSQLiteDatabaseTrackerClient a singleton.
(WebCore::WebSQLiteDatabaseTrackerClient::WebSQLiteDatabaseTrackerClient):
(WebCore::WebSQLiteDatabaseTrackerClient::~WebSQLiteDatabaseTrackerClient):
(WebCore::WebSQLiteDatabaseTrackerClient::willBeginFirstTransaction): Use a utility class WebDatabaseTransactionBackgroundTaskController
    to schedule database transaction background task.
(WebCore::WebSQLiteDatabaseTrackerClient::didFinishLastTransaction): Use WebDatabaseTransactionBackgroundTaskController to stop
    databas transaction background task.
(transactionBackgroundTaskIdentifierLock: Moved from Source/WebKit/mac/Storage/WebDatabaseManager.mm.
(setTransactionBackgroundTaskIdentifier): Ditto.
(getTransactionBackgroundTaskIdentifier): Ditto.
(+[WebDatabaseTransactionBackgroundTaskController startBackgroundTask]): Ditto.
(+[WebDatabaseTransactionBackgroundTaskController endBackgroundTask]): Ditto.

Source/WebKit:

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2017-02-02
Reviewed by Brady Eidson.

* WebKit.xcodeproj/project.pbxproj: Moved WebSQLiteDatabaseTrackerClient to WebCore.

Source/WebKit/ios:

Move application background task handling code from WebKit to WebCore.

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2017-02-02
Reviewed by Brady Eidson.

* Misc/WebUIKitSupport.h: Remove several methods that only used internally for background
    task handling. They are not needed in WebKit any more since background task handling
    is moved to WebCore and wrapped in WebBackgroundTaskController class.
* Misc/WebUIKitSupport.mm:
(WebKitSetInvalidWebBackgroundTaskIdentifier): Instead of storing the value in a static global
    variable, save it in WebBackgroundTaskController.
(WebKitSetStartBackgroundTaskBlock): Ditto.
(WebKitSetEndBackgroundTaskBlock): Ditto.

Source/WebKit/mac:

Move database transaction background task handling code from WebDatabaseManager to
WebCore's WebSQLiteDatabaseTrackerClient.

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2017-02-02
Reviewed by Brady Eidson.

* Storage/WebDatabaseManager.mm:
* Storage/WebDatabaseManagerInternal.h: Remove a category for background task handling.
* WebCoreSupport/WebApplicationCache.mm:
(+[WebApplicationCache initializeWithBundleIdentifier:]): Use WebCore::WebSQLiteDatabaseTrackerClient.
* WebView/WebView.mm:
(-[WebView _commonInitializationWithFrameName:groupName:]): Ditto.

Source/WebKit2:

Just like in WebKit1, when initializing a WKWebView, initialize the database transaction
tracker client. Also, we should set up the start and end background task blocks here. In
WebKit1, this is done inside UIKit.

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2017-02-02
Reviewed by Brady Eidson.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _initializeWithConfiguration:]):
(-[WKWebView _setUpSQLiteDatabaseTrackerClient]): Set up the start/end background task blocks
    and database transaction tracker client.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211551 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[EME] Implement MediaKeySession::update()
zandobersek@gmail.com [Thu, 2 Feb 2017 06:28:03 +0000 (06:28 +0000)]
[EME] Implement MediaKeySession::update()
https://bugs.webkit.org/show_bug.cgi?id=167636

Reviewed by Xabier Rodriguez-Calvar.

Source/WebCore:

Implement the MediaKeySession::update() method by following the steps as
they are described in the specification.

In order to sanitize the passed-in response data, CDM::sanitizeResponse()
is added. It passes the SharedBuffer object by reference to the CDMPrivate
interface implementor, which returns a SharedBuffer object containing
sanitized response data.

CDMInstance::updateLicense() virtual method is added to perform the license
update for some specific CDMInstance object. After the update the CDMInstance
invokes the callback that's passed to updateLicense(), providing information
about session being closed, changed keys or expiration value, any message
that has to be enqueued, and whether the update was successful.

After that callback is invoked, MediaKeySession::update() goes on to handle
all the provided information in a future task, finally resolving the promise
(or rejecting it beforehand in case of any failure during response handling
or license update).

Three algorithms that can be invoked from MediaKeySession::update() (key
status update, expiration update and session closure) will be implemented
separately. Placeholder methods are provided until then.

MockCDM::sanitizeResponse() and MockCDMInstance::updateLicense() are
implemented for testing purposes. For now only the response sanitization
and sanitized response format are checked there. Key status update,
expiration update and session closure should be tested once the
implementations for those algorithms are added.

Test: media/encrypted-media/mock-MediaKeySession-update.html

* Modules/encryptedmedia/CDM.cpp:
(WebCore::CDM::sanitizeResponse):
* Modules/encryptedmedia/CDM.h:
* Modules/encryptedmedia/CDMInstance.h:
* Modules/encryptedmedia/CDMPrivate.h:
* Modules/encryptedmedia/MediaKeySession.cpp:
(WebCore::MediaKeySession::update):
(WebCore::MediaKeySession::updateKeyStatuses):
(WebCore::MediaKeySession::updateExpiration):
(WebCore::MediaKeySession::sessionClosed):
* Modules/encryptedmedia/MediaKeySession.h:
* testing/MockCDMFactory.cpp:
(WebCore::MockCDM::sanitizeResponse):
(WebCore::MockCDMInstance::updateLicense):
* testing/MockCDMFactory.h:

LayoutTests:

Add the mock-MediaKeySession-update.html test, cotaining a few cases that check
the basic operations of MediaKeySession::update(), focusing on proper promise
resolution and rejection based on the state of the object and the passed-in
response argument. Skip the test on all platforms for now.

* media/encrypted-media/mock-MediaKeySession-update-expected.txt: Added.
* media/encrypted-media/mock-MediaKeySession-update.html: Added.
* platform/efl/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211550 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMake run-benchmark script supports 'config' key in test plan.
dewei_zhu@apple.com [Thu, 2 Feb 2017 03:10:58 +0000 (03:10 +0000)]
Make run-benchmark script supports 'config' key in test plan.
https://bugs.webkit.org/show_bug.cgi?id=167715

Reviewed by Ryosuke Niwa.

Add support for 'config' key in test paln.
'config' key is a dictionary that specifies environment configurations while test is running.
Specifies orientation to be portrait in MotionMark benchmark.

* Scripts/webkitpy/benchmark_runner/README.md:
* Scripts/webkitpy/benchmark_runner/benchmark_runner.py:
(BenchmarkRunner.__init__):
(BenchmarkRunner._run_benchmark):
* Scripts/webkitpy/benchmark_runner/browser_driver/browser_driver.py:
(BrowserDriver.prepare_env):
* Scripts/webkitpy/benchmark_runner/browser_driver/linux_browser_driver.py:
(LinuxBrowserDriver.prepare_env):
* Scripts/webkitpy/benchmark_runner/browser_driver/osx_browser_driver.py:
(OSXBrowserDriver.prepare_env):
* Scripts/webkitpy/benchmark_runner/browser_driver/osx_safari_driver.py:
(OSXSafariDriver.prepare_env):
* Scripts/webkitpy/benchmark_runner/data/plans/motionmark.plan:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211549 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, rolling out r211461.
commit-queue@webkit.org [Thu, 2 Feb 2017 01:57:00 +0000 (01:57 +0000)]
Unreviewed, rolling out r211461.
https://bugs.webkit.org/show_bug.cgi?id=167721

Big regression on kraken (Requested by jfbastien on #webkit).

Reverted changeset:

"OSR entry: delay outer-loop compilation when at inner-loop"
https://bugs.webkit.org/show_bug.cgi?id=167149
http://trac.webkit.org/changeset/211461

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211548 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, fix the nightly open source build.
wenson_hsieh@apple.com [Thu, 2 Feb 2017 01:54:46 +0000 (01:54 +0000)]
Unreviewed, fix the nightly open source build.

Adds more header checks when importing from WebKitAdditions.

* UIProcess/Cocoa/WebPageProxyCocoa.mm:
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
* WebProcess/WebCoreSupport/mac/WebDragClientMac.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211547 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, fix unintended change.
keith_miller@apple.com [Thu, 2 Feb 2017 01:49:20 +0000 (01:49 +0000)]
Unreviewed, fix unintended change.

* runtime/SamplingProfiler.cpp:
(JSC::SamplingProfiler::StackFrame::displayName):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211546 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Mac] Update CARingBuffer class
eric.carlson@apple.com [Thu, 2 Feb 2017 01:48:29 +0000 (01:48 +0000)]
[Mac] Update CARingBuffer class
https://bugs.webkit.org/show_bug.cgi?id=167656
<rdar://problem/30318329>

Unreviewed fix the 32-bit build after r211539.

* platform/audio/mac/CAAudioStreamDescription.cpp:
(WebCore::CAAudioStreamDescription::CAAudioStreamDescription):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211545 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRebaseline compositing/geometry/fixed-in-composited.html for ios-simulator.
ryanhaddad@apple.com [Thu, 2 Feb 2017 01:47:37 +0000 (01:47 +0000)]
Rebaseline compositing/geometry/fixed-in-composited.html for ios-simulator.

Unreviewed test gardening.

* platform/ios-simulator/compositing/geometry/fixed-in-composited-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211544 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoREGRESSION (r210820): svn-create-patch does not emit property change only diff
dbates@webkit.org [Thu, 2 Feb 2017 01:41:24 +0000 (01:41 +0000)]
REGRESSION (r210820): svn-create-patch does not emit property change only diff
https://bugs.webkit.org/show_bug.cgi?id=167169

Reviewed by David Kilzer.

More directly handle the SVN 1.9 or newer syntax change for a diff of a copied or moved file
- a SVN diff header and an empty body. In particular, remove a duplicate leading SVN diff
header from the output of "svn diff" on a file that was copied or moved.

Following r210820 svn-create-patch emits to standard output a diff only if it has at least one
chunk range line (e.g. @@ -0,0 +1,7 @@) as a means to workaround a syntax change in the diff
output of a copied or moved file in SVN version 1.9 or newer. In SVN 1.9 or newer "svn diff"
of a copied or moved file always emits to standard output a leading SVN diff header (an "Index:"
line followed by a separator line) with an empty body; => no chunk range lines. However a diff
of a file with only a property change also does not contain any chunk range lines. Therefore
svn-create-patch no longer emitted to standard output such a diff. Instead of indirectly detecting
a SVN diff header with an empty body by counting the number of chunk range lines in the diff
we should directly test for the presence of a leading SVN diff header with an empty body and
remove such lines from the diff.

* Scripts/VCSUtils.pm:
(fixSVNPatchForAdditionWithHistory): Added.
* Scripts/svn-create-patch:
(generateDiff): Pass the output from "svn diff" to fixSVNPatchForAdditionWithHistory() when
we have a non empty patch for an added file with history (i.e. a copied or moved file).
* Scripts/webkitperl/VCSUtils_unittest/fixSVNPatchForAdditionWithHistory.pl: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211543 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoThe sampling profile should have an option to sample from C frames.
keith_miller@apple.com [Thu, 2 Feb 2017 01:23:37 +0000 (01:23 +0000)]
The sampling profile should have an option to sample from C frames.
https://bugs.webkit.org/show_bug.cgi?id=167614

Reviewed by Saam Barati.

We should be able to use the sampling profiler, at least
internally, to trace C calls.  This patch only modifies the JSC
shell although it would be nice to add it to the Web Inspector in
a future patch.

* runtime/Options.h:
* runtime/SamplingProfiler.cpp:
(JSC::FrameWalker::FrameWalker):
(JSC::FrameWalker::walk):
(JSC::FrameWalker::recordJSFrame):
(JSC::CFrameWalker::CFrameWalker):
(JSC::CFrameWalker::walk):
(JSC::CFrameWalker::isCFrame):
(JSC::CFrameWalker::advanceToParentFrame):
(JSC::CFrameWalker::frame):
(JSC::SamplingProfiler::takeSample):
(JSC::SamplingProfiler::processUnverifiedStackTraces):
(JSC::SamplingProfiler::StackFrame::displayName):
* runtime/SamplingProfiler.h:
(JSC::SamplingProfiler::UnprocessedStackFrame::UnprocessedStackFrame):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211542 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[iOS WK2] With visual viewports, a fixed bottom bar can be clipped out when the keybo...
simon.fraser@apple.com [Thu, 2 Feb 2017 01:12:53 +0000 (01:12 +0000)]
[iOS WK2] With visual viewports, a fixed bottom bar can be clipped out when the keyboard is visible
https://bugs.webkit.org/show_bug.cgi?id=167710
rdar://problem/30100286

Reviewed by Wenson Hsieh.
Source/WebCore:

Add a bit of logging to show when RLC decides to not composite a fixed element which
is out of bounds.

Test: fast/visual-viewport/ios/bottom-bar-with-keyboard.html

* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::requiresCompositingForPosition):

Source/WebKit2:

The unobscuredRectInContentCoordinates passed to -didUpdateVisibleRect:... could project outside
of the bounds of the document, which is OK when rubber-banding, but not when we're in a stable state,
because that can cause fixed elements to get pushed outside the doc. This happened when the keyboard
triggered bottom content insets on the scroll view.

LayoutTests:

* fast/visual-viewport/ios/bottom-bar-with-keyboard-expected.txt: Added.
* fast/visual-viewport/ios/bottom-bar-with-keyboard.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211541 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Mac] Update CARingBuffer class
eric.carlson@apple.com [Wed, 1 Feb 2017 23:49:41 +0000 (23:49 +0000)]
[Mac] Update CARingBuffer class
https://bugs.webkit.org/show_bug.cgi?id=167656

Unreviewed build fix after r211539.

* platform/audio/mac/CAAudioStreamDescription.cpp:
(WebCore::CAAudioStreamDescription::CAAudioStreamDescription): Initialize a variable to fix
release builds.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211540 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Mac] Update CARingBuffer class
eric.carlson@apple.com [Wed, 1 Feb 2017 23:38:04 +0000 (23:38 +0000)]
[Mac] Update CARingBuffer class
https://bugs.webkit.org/show_bug.cgi?id=167656

Reviewed by NOBODY (OOPS!).

Source/WebCore:

API test CARingBufferTest added.

* WebCore.xcodeproj/project.pbxproj: Add headers to framework so they can be used from the API test.

* platform/audio/AudioStreamDescription.h: Add Int32.

* platform/audio/mac/CAAudioStreamDescription.cpp:
(WebCore::CAAudioStreamDescription::CAAudioStreamDescription):
(WebCore::CAAudioStreamDescription::~CAAudioStreamDescription):
(WebCore::CAAudioStreamDescription::format): Support Int32.
(WebCore::CAAudioStreamDescription::operator==): Make inline.
* platform/audio/mac/CAAudioStreamDescription.h:

* platform/audio/mac/CARingBuffer.cpp:
(WebCore::CARingBuffer::CARingBuffer): Move initializers into class declaration.
(WebCore::CARingBuffer::allocate): Use flush method, get allocation info from a CAAudioStreamDescription.
(WebCore::FetchABL): Add a mode parameter to allow samples to replace or mix into destination.
(WebCore::CARingBuffer::flush): New.
(WebCore::CARingBuffer::fetch): Add mode parameter.
(WebCore::CARingBuffer::~CARingBuffer): Deleted.
* platform/audio/mac/CARingBuffer.h:
(WebCore::CARingBuffer::~CARingBuffer):

* platform/graphics/avfoundation/AudioSourceProviderAVFObjC.mm:
(WebCore::AudioSourceProviderAVFObjC::prepare): Use new CARingBuffer constructor.
(WebCore::operator==): Deleted.
(WebCore::operator!=): Deleted.

* platform/mediastream/mac/WebAudioSourceProviderAVFObjC.mm:
(WebCore::WebAudioSourceProviderAVFObjC::prepare): Use new CARingBuffer constructor
(WebCore::operator==): Deleted.
(WebCore::operator!=): Deleted.

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebCore/CARingBuffer.cpp: Added.
(TestWebKitAPI::CARingBufferTest::SetUp):
(TestWebKitAPI::CARingBufferTest::setup):
(TestWebKitAPI::CARingBufferTest::setListDataBuffer):
(TestWebKitAPI::CARingBufferTest::description):
(TestWebKitAPI::CARingBufferTest::bufferList):
(TestWebKitAPI::CARingBufferTest::ringBuffer):
(TestWebKitAPI::CARingBufferTest::capacity):
(TestWebKitAPI::CARingBufferTest::audioBufferListSizeForStream):
(TestWebKitAPI::CARingBufferTest::configureBufferListForStream):
(TestWebKitAPI::TEST_F):
(TestWebKitAPI::MixingTest::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211539 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWebProcess crashes in int WTF::__throw_bad_variant_access<int> when expanding/shrink...
enrica@apple.com [Wed, 1 Feb 2017 23:25:15 +0000 (23:25 +0000)]
WebProcess crashes in  int WTF::__throw_bad_variant_access<int> when expanding/shrinking a block selection.
https://bugs.webkit.org/show_bug.cgi?id=167673
rdar://problem/30229620

Reviewed by Anders Carlsson

This is a speculative fix for a bug that might have been introduced
with http://trac.webkit.org/changeset/208479.
The code in containsRange inline fuction in WebPageIOS.mm has not been
updated when the return value of Range::compareBoundaryPoints was changed
to ExceptionOr<short>.
Since there is already a method containsRange in the Range class that
does the right thing, expandedRangeFromHandle now uses that.

* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::expandedRangeFromHandle):
(WebKit::containsRange): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211538 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoVersioning.
jmarcell@apple.com [Wed, 1 Feb 2017 23:16:20 +0000 (23:16 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211537 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMarking compositing/repaint/fixed-background-scroll.html as flaky on macOS WK1.
ryanhaddad@apple.com [Wed, 1 Feb 2017 23:08:08 +0000 (23:08 +0000)]
Marking compositing/repaint/fixed-background-scroll.html as flaky on macOS WK1.
https://bugs.webkit.org/show_bug.cgi?id=154612

Unreviewed test gardening.

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211536 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoIPC::Connection receive ports should be guarded
andersca@apple.com [Wed, 1 Feb 2017 22:52:02 +0000 (22:52 +0000)]
IPC::Connection receive ports should be guarded
https://bugs.webkit.org/show_bug.cgi?id=167704

Reviewed by Tim Horton.

Guarding receive rights will make sure that they won't be closed accidentally. They are created
with a context pointer and can only be unguarded or destructed with the same context pointer.

* Platform/IPC/mac/ConnectionMac.mm:
(IPC::Connection::platformInvalidate):
Use mach_port_destruct and pass the connection pointer as the context.

(IPC::Connection::platformInitialize):
Guard the server port with the connection pointer as the context.

(IPC::Connection::open):
Use mach_port_construct to create the port which lets us avoid a call to mach_port_set_attributes and setMachPortQueueLength.
Make the port guarded and use the connection pointer as the context.

(IPC::createReceiveSource):
Get rid of this and just duplicate the five lines of code in two places. For the receive port we want to use mach_port_destruct
in our cancel handler.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211535 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRebaselining compositing/geometry/fixed-in-composited.html for macOS after r211523.
ryanhaddad@apple.com [Wed, 1 Feb 2017 22:41:52 +0000 (22:41 +0000)]
Rebaselining compositing/geometry/fixed-in-composited.html for macOS after r211523.
https://bugs.webkit.org/show_bug.cgi?id=167620

Unreviewed test gardening.

* platform/mac/compositing/geometry/fixed-in-composited-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211534 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMarking inspector/debugger/probe-manager-add-remove-actions.html as slow on macOS...
ryanhaddad@apple.com [Wed, 1 Feb 2017 22:41:49 +0000 (22:41 +0000)]
Marking inspector/debugger/probe-manager-add-remove-actions.html as slow on macOS debug.
https://bugs.webkit.org/show_bug.cgi?id=167711

Unreviewed test gardening.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211533 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMarking inspector/worker/debugger-shared-breakpoint.html as flaky.
ryanhaddad@apple.com [Wed, 1 Feb 2017 22:34:08 +0000 (22:34 +0000)]
Marking inspector/worker/debugger-shared-breakpoint.html as flaky.
https://bugs.webkit.org/show_bug.cgi?id=167203

Unreviewed test gardening.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211532 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoSimple line layout: Move TextFragmentIterator::runWidth to ::textWidth.
zalan@apple.com [Wed, 1 Feb 2017 22:25:58 +0000 (22:25 +0000)]
Simple line layout: Move TextFragmentIterator::runWidth to ::textWidth.
https://bugs.webkit.org/show_bug.cgi?id=167705
<rdar://problem/30314393>

Reviewed by Antti Koivisto.

We also don't need to compute logicalHeight for logicalLeft/right since the
cases where it matters (floating boxes) are not supported by simple line layout.

Covered by existing test.

* rendering/SimpleLineLayout.cpp:
(WebCore::SimpleLineLayout::updateLineConstrains):
* rendering/SimpleLineLayoutTextFragmentIterator.cpp:
(WebCore::SimpleLineLayout::nextBreakablePositionInSegment):
(WebCore::SimpleLineLayout::TextFragmentIterator::textWidth):
(WebCore::SimpleLineLayout::TextFragmentIterator::runWidth): Deleted.
* rendering/SimpleLineLayoutTextFragmentIterator.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211531 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoFix compilation error under recent clang update
jer.noble@apple.com [Wed, 1 Feb 2017 22:15:07 +0000 (22:15 +0000)]
Fix compilation error under recent clang update
https://bugs.webkit.org/show_bug.cgi?id=167707

Reviewed by Eric Carlson.

Include TypedArrayInlines.h to fix a "missing symbol" link error when compiling under a recent clang update.

* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211530 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoClean up css3/filters/backdrop/backdrop-filter-with-clip-path.html expected results.
simon.fraser@apple.com [Wed, 1 Feb 2017 22:13:21 +0000 (22:13 +0000)]
Clean up css3/filters/backdrop/backdrop-filter-with-clip-path.html expected results.

https://trac.webkit.org/changeset/201374 added some very confused results for this test.
Just move the -expected.html next to the test, and remove the .txt and .png results.

* css3/filters/backdrop/backdrop-filter-with-clip-path-expected.html: Renamed from LayoutTests/platform/ios-simulator/css3/filters/backdrop/backdrop-filter-with-clip-path-expected.html.
* css3/filters/backdrop/backdrop-filter-with-clip-path-expected.txt: Removed.
* platform/mac/css3/filters/backdrop/backdrop-filter-with-clip-path-expected.png: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211529 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoPatch Details: EWS Bubbles wrap on Patch Details page
joepeck@webkit.org [Wed, 1 Feb 2017 22:08:23 +0000 (22:08 +0000)]
Patch Details: EWS Bubbles wrap on Patch Details page
https://bugs.webkit.org/show_bug.cgi?id=167709

Reviewed by Ryosuke Niwa.

* template/en/default/attachment/edit.html.tmpl:
Match the iframe size in list.html.tmpl.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211528 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoPerformance Timeline: Prepare for Worker support
joepeck@webkit.org [Wed, 1 Feb 2017 22:08:21 +0000 (22:08 +0000)]
Performance Timeline: Prepare for Worker support
https://bugs.webkit.org/show_bug.cgi?id=167633

Reviewed by Ryosuke Niwa.

Remove references to Document/Frame in Performance. Instead use
ScriptExecutionContext and pass in the time origin / SecurityOrigin
where it is relevant. No tests should change this is just refactoring.

* loader/ResourceTimingInformation.cpp:
(WebCore::ResourceTimingInformation::addResourceTiming):
* page/DOMWindow.cpp:
(WebCore::DOMWindow::performance):
* page/Performance.cpp:
(WebCore::Performance::Performance):
(WebCore::Performance::now):
(WebCore::Performance::reduceTimeResolution):
(WebCore::Performance::navigation):
(WebCore::Performance::timing):
(WebCore::Performance::addResourceTiming):
(WebCore::Performance::scriptExecutionContext): Deleted.
* page/Performance.h:
* page/Performance.idl:
* page/PerformanceResourceTiming.cpp:
(WebCore::monotonicTimeToDOMHighResTimeStamp):
(WebCore::passesTimingAllowCheck):
(WebCore::PerformanceResourceTiming::PerformanceResourceTiming):
(WebCore::PerformanceResourceTiming::redirectStart):
(WebCore::PerformanceResourceTiming::redirectEnd):
(WebCore::PerformanceResourceTiming::fetchStart):
(WebCore::PerformanceResourceTiming::domainLookupStart):
(WebCore::PerformanceResourceTiming::domainLookupEnd):
(WebCore::PerformanceResourceTiming::connectStart):
(WebCore::PerformanceResourceTiming::connectEnd):
(WebCore::PerformanceResourceTiming::secureConnectionStart):
(WebCore::PerformanceResourceTiming::requestStart):
(WebCore::PerformanceResourceTiming::responseStart):
(WebCore::PerformanceResourceTiming::responseEnd):
(WebCore::PerformanceResourceTiming::networkLoadTimeToDOMHighResTimeStamp):
(WebCore::monotonicTimeToDocumentMilliseconds): Deleted.
(WebCore::PerformanceResourceTiming::initiatorType): Deleted.
(WebCore::PerformanceResourceTiming::resourceTimeToDocumentMilliseconds): Deleted.
* page/PerformanceResourceTiming.h:
* page/PerformanceUserTiming.cpp:
(WebCore::UserTiming::findExistingMarkStartTime):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211527 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Mac] TestWebKitAPI includes system frameworks incorrectly
ap@apple.com [Wed, 1 Feb 2017 22:02:04 +0000 (22:02 +0000)]
[Mac] TestWebKitAPI includes system frameworks incorrectly
https://bugs.webkit.org/show_bug.cgi?id=167703
rdar://problem/30301498

Reviewed by Ryosuke Niwa.

FRAMEWORK_SEARCH_PATH results in frameworks being treated as user ones, not system.
Make TestWebKitAPI use -iframework instead, like the rest of WebKit does.

Also, reduced the number of subframework search paths, looks like we only need one now.

* TestWebKitAPI/Configurations/Base.xcconfig:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211526 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoREGRESSION (r211379): [ios-simulator] LayoutTest compositing/geometry/fixed-in-compos...
simon.fraser@apple.com [Wed, 1 Feb 2017 21:44:18 +0000 (21:44 +0000)]
REGRESSION (r211379): [ios-simulator] LayoutTest compositing/geometry/fixed-in-composited.html is failing
https://bugs.webkit.org/show_bug.cgi?id=167620
rdar://problem/30309207

The body needs to be taller to allow the scrollBy() to be a stable scroll.

* compositing/geometry/fixed-in-composited.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211523 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoREGRESSION(r205136): {}.toString.call(crossOriginWindow) should not throw
cdumez@apple.com [Wed, 1 Feb 2017 20:13:05 +0000 (20:13 +0000)]
REGRESSION(r205136): {}.toString.call(crossOriginWindow) should not throw
https://bugs.webkit.org/show_bug.cgi?id=167681
<rdar://problem/30301117>

Reviewed by Mark Lam.

LayoutTests/imported/w3c:

Rebaseline W3C test now that we passing one more check. We're still failing
later because {}.toString.call(crossOriginWindow) returns "[object Window]"
instead of "[object Object]". I am planning to fix this in a follow-up.
This is not a regression since we were returning "[object Window]" before
r205136.

* web-platform-tests/html/browsers/origin/cross-origin-objects/cross-origin-objects-expected.txt:

Source/WebCore:

As per https://html.spec.whatwg.org/#crossorigingetownpropertyhelper-(-o,-p-):
"""
If P is @@toStringTag, @@hasInstance, or @@isConcatSpreadable, then return
PropertyDescriptor{ [[Value]]: undefined, [[Writable]]: false, [[Enumerable]]: false, [[Configurable]]: true }.
"""

We now implement this behavior instead of throwing.

Test: http/tests/security/symbols-cross-origin.html

* bindings/js/JSDOMWindowCustom.cpp:
(WebCore::jsDOMWindowGetOwnPropertySlotRestrictedAccess):
* bindings/js/JSLocationCustom.cpp:
(WebCore::JSLocation::getOwnPropertySlotDelegate):

LayoutTests:

Add layout test coverage.

* http/tests/security/symbols-cross-origin-expected.txt: Added.
* http/tests/security/symbols-cross-origin.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211504 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, fix the WebKit nightly open source build
wenson_hsieh@apple.com [Wed, 1 Feb 2017 20:03:48 +0000 (20:03 +0000)]
Unreviewed, fix the WebKit nightly open source build
<rdar://problem/30308635>

We cannot assume all internal SDKs have the latest WebKitAdditions, so we need an explicit header check here.

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211503 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoCorrect "filesHaveSameVolume" predicate
bfulgham@apple.com [Wed, 1 Feb 2017 20:00:18 +0000 (20:00 +0000)]
Correct "filesHaveSameVolume" predicate
https://bugs.webkit.org/show_bug.cgi?id=167696
<rdar://problem/30298722>

Reviewed by David Kilzer.

Source/WebCore:

We are passing %-encoded strings to the underlying operating system's file system APIs.
This doesn't work. Instead, we need to present a decoded version of the file path
that matches what the system APIs expect.

Tested by new TestWebKitAPI Test.

* platform/FileSystem.cpp:
(WebCore::filesHaveSameVolume): Make sure the file paths we give to the underlying
operating system are not percent encoded.
* platform/FileSystem.h: Export 'filesHaveSameVolume' for use by testing system.

Tools:

Add new tests that confirm that the 'filesHaveSamePath' predicate properly handles
percent-escaped path inputs.

* TestWebKitAPI/Tests/WebCore/FileSystem.cpp:
(TestWebKitAPI::FileSystemTest::spaceContainingFilePath):
(TestWebKitAPI::FileSystemTest::bangContainingFilePath):
(TestWebKitAPI::FileSystemTest::quoteContainingFilePath):
(TestWebKitAPI::TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211502 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[mac-wk1] LayoutTest media/modern-media-controls/tracks-support/tracks-support-click...
commit-queue@webkit.org [Wed, 1 Feb 2017 19:54:25 +0000 (19:54 +0000)]
[mac-wk1] LayoutTest media/modern-media-controls/tracks-support/tracks-support-click-track-in-panel.html is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=165319
<rdar://problem/30284104>

Patch by Antoine Quint <graouts@apple.com> on 2017-02-01
Reviewed by Dean Jackson.

Source/WebCore:

Running media/controls/track-menu.html before media/modern-media-controls/tracks-support/tracks-
support-click-track-in-panel.html makes that test time out in all test runs. The root of the issue
is that animations are suspended by media/controls/track-menu.html with a call to
internals.suspendAnimations(), and that state isn't reset with a call to internals.resumeAnimations().
Then, media/modern-media-controls/tracks-support/tracks-support-click-track-in-panel.html fails because
the selection animation for the tracks panel menu item that is clicked never completes and the delegate
to notify that an item in the tracks panel was selected is never fired, which leads to the test failure.

We change Internals::suspendAnimations() and Internals::resumeAnimations() to only affect the current
document, rather than calling into AnimationController::suspendAnimations() which would do just that,
but also set a Frame-wide flag that would prevent further animations from running, even in a subsequent
document load.

* dom/Document.cpp:
(WebCore::Document::prepareForDestruction): Ensure the document that is about to be destroyed is no longer
associated with an AnimationController.
* page/animation/AnimationController.cpp:
(WebCore::AnimationControllerPrivate::ensureCompositeAnimation): Update the animation's suspend state in case
the document its renderer is associated with is suspended. This is required since previously CompositeAnimations
would set their suspend state in their constructor, based on the Frame-wide suspended state, but there is no
document to use as a basis to query its suspended state in that constructor.
(WebCore::AnimationControllerPrivate::animationsAreSuspendedForDocument):
(WebCore::AnimationControllerPrivate::detachFromDocument):
(WebCore::AnimationControllerPrivate::suspendAnimationsForDocument):
(WebCore::AnimationControllerPrivate::resumeAnimationsForDocument):
(WebCore::AnimationControllerPrivate::startAnimationsIfNotSuspended):
(WebCore::AnimationController::animationsAreSuspendedForDocument):
(WebCore::AnimationController::detachFromDocument):
* page/animation/AnimationController.h:
* page/animation/AnimationControllerPrivate.h:
* testing/Internals.cpp:
(WebCore::Internals::animationsAreSuspended):
(WebCore::Internals::suspendAnimations):
(WebCore::Internals::resumeAnimations):

LayoutTests:

Since we've fixed the root cause of this test's flakiness, we no longer need to mark it as flaky.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211501 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed build fix after r211488.
ryanhaddad@apple.com [Wed, 1 Feb 2017 19:44:07 +0000 (19:44 +0000)]
Unreviewed build fix after r211488.

* accessibility/AXObjectCache.h: Removed property svn:executable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211500 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, remove loop
utatane.tea@gmail.com [Wed, 1 Feb 2017 18:34:31 +0000 (18:34 +0000)]
Unreviewed, remove loop
https://bugs.webkit.org/show_bug.cgi?id=167694

It causes timeout.
Original bug can be tested even without this loop.

* stress/arity-fixup-should-not-touch-stack-area-below-sp.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211498 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed build fix.
bfulgham@apple.com [Wed, 1 Feb 2017 18:31:05 +0000 (18:31 +0000)]
Unreviewed build fix.

* html/HTMLMediaElement.h: Removed property svn:executable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211497 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoNULL-deref crash in TextTrack::removeCue()
jer.noble@apple.com [Wed, 1 Feb 2017 18:22:21 +0000 (18:22 +0000)]
NULL-deref crash in TextTrack::removeCue()
https://bugs.webkit.org/show_bug.cgi?id=167615

Reviewed by Eric Carlson.

Source/WebCore:

Test: http/tests/media/track-in-band-hls-metadata-crash.html

Follow-up to r211401. When passing around a reference to an object, the assumption is that
the caller is retaining the underlying object. This breaks down for
InbandDataTextTrack::removeDataCue(), which releases its own ownership of the cue object,
then passes the reference to that object to its superclass to do further remove steps. The
retain count of the cue can thus drop to zero within the scope of
InbandTextTrack::removeCue(). Use "take" semantics to remove the cue from the
m_incompleteCueMap without releasing ownership, and pass a reference to that retained object
on to removeCue(), guaranteeing that the cue will not be destroyed until after the
romeveDataCue() method returns.

* html/track/InbandDataTextTrack.cpp:
(WebCore::InbandDataTextTrack::removeDataCue):

LayoutTests:

* http/tests/media/track-in-band-hls-metadata-crash-expected.txt: Added.
* http/tests/media/track-in-band-hls-metadata-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211495 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[WK1] Allow the drag client to be initialized on platforms that do not support drag...
wenson_hsieh@apple.com [Wed, 1 Feb 2017 18:15:13 +0000 (18:15 +0000)]
[WK1] Allow the drag client to be initialized on platforms that do not support drag and drop
https://bugs.webkit.org/show_bug.cgi?id=167695
<rdar://problem/30305390>

Reviewed by Sam Weinig.

Due to the common codepath for destroying the drag client, we need to ensure that the drag client actually
exists, even on platforms that do not support drag and drop, in order to avoid crashing when the DragController
is destroyed.

* WebView/WebView.mm:
(-[WebView _commonInitializationWithFrameName:groupName:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211492 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoAX: Incorrect range from index and length in text controls when there are newlines
n_wang@apple.com [Wed, 1 Feb 2017 18:12:08 +0000 (18:12 +0000)]
AX: Incorrect range from index and length in text controls when there are newlines
https://bugs.webkit.org/show_bug.cgi?id=167677
<rdar://problem/30300099>

Reviewed by Chris Fleizach.

Source/WebCore:

When there are newlines in text controls, the VisiblePosition that is created
from the index could be wrong. Because the index doesn't include the information
regarding a position's affinity. Fixed it by using CharacterOffset instead, since
it doesn't require that information.

Test: accessibility/mac/range-for-contenteditable-newline.html

* accessibility/AXObjectCache.cpp:
(WebCore::AXObjectCache::traverseToOffsetInRange):
(WebCore::AXObjectCache::nextCharacterOffset):
(WebCore::AXObjectCache::characterOffsetForIndex):

LayoutTests:

* accessibility/mac/range-for-contenteditable-newline-expected.txt: Added.
* accessibility/mac/range-for-contenteditable-newline.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211491 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoios-simulator TestExpectations gardening.
ryanhaddad@apple.com [Wed, 1 Feb 2017 17:54:05 +0000 (17:54 +0000)]
ios-simulator TestExpectations gardening.

Unreviewed test gardening.

* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211489 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoAXObjectCache is missing two functions at link time when !HAVE(ACCESSIBILITY)
timothy@hatcher.name [Wed, 1 Feb 2017 17:46:12 +0000 (17:46 +0000)]
AXObjectCache is missing two functions at link time when !HAVE(ACCESSIBILITY)
https://bugs.webkit.org/show_bug.cgi?id=167691

Reviewed by Chris Fleizach.

* accessibility/AXObjectCache.h:
(WebCore::AXObjectCache::recomputeDeferredIsIgnored): Added empty stub.
(WebCore::AXObjectCache::performDeferredIsIgnoredChange): Added empty stub.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211488 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoHTMLMediaElement.h fails to build when ENABLE(MEDIA_CONTROLS_SCRIPT) is false
timothy@hatcher.name [Wed, 1 Feb 2017 17:37:03 +0000 (17:37 +0000)]
HTMLMediaElement.h fails to build when ENABLE(MEDIA_CONTROLS_SCRIPT) is false
https://bugs.webkit.org/show_bug.cgi?id=167689

Reviewed by Joseph Pecoraro.

* html/HTMLMediaElement.h: Move mediaControlsHost() into the ENABLE(MEDIA_CONTROLS_SCRIPT)
block that is just above it to match m_mediaControlsHost.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211487 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWeb Inspector: Use guaranteed RunLoop instead of RunLoop::current for dispatching...
commit-queue@webkit.org [Wed, 1 Feb 2017 17:34:00 +0000 (17:34 +0000)]
Web Inspector: Use guaranteed RunLoop instead of RunLoop::current for dispatching inspector GC event
https://bugs.webkit.org/show_bug.cgi?id=167683
<rdar://problem/30167791>

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2017-02-01
Reviewed by Timothy Hatcher.

* inspector/agents/InspectorHeapAgent.cpp:
(Inspector::SendGarbageCollectionEventsTask::SendGarbageCollectionEventsTask):
Use RunLoop::main instead of RunLoop::current which may go away.

* runtime/InitializeThreading.cpp:
(JSC::initializeThreading):
Ensure RunLoop::main is initialized when using JSC APIs.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211486 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWeb Inspector: Remove keyboard shortcut to close tab - does not match expectations
commit-queue@webkit.org [Wed, 1 Feb 2017 17:33:56 +0000 (17:33 +0000)]
Web Inspector: Remove keyboard shortcut to close tab - does not match expectations
https://bugs.webkit.org/show_bug.cgi?id=167672

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2017-02-01
Reviewed by Timothy Hatcher.

* UserInterface/Views/TabBrowser.js:
(WebInspector.TabBrowser):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211485 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Readable Streams API] Implement ReadableByteStreamController pull()
commit-queue@webkit.org [Wed, 1 Feb 2017 16:17:52 +0000 (16:17 +0000)]
[Readable Streams API] Implement ReadableByteStreamController pull()
https://bugs.webkit.org/show_bug.cgi?id=167593

Patch by Romain Bellessort <romain.bellessort@crf.canon.fr> on 2017-02-01
Reviewed by Youenn Fablet.

Source/WebCore:

Implemented pull() method for ReadableByteStreamController. Also updated
pendingPullIntos attribute, which was a more complex object than necessary
(an array is enough).

Added 2 tests that allow covering most of new code. Code not yet tested will
become reachable as the rest of the spec is implemented (new tests will then
be added).

* Modules/streams/ReadableByteStreamInternals.js:
(privateInitializeReadableByteStreamController): Updated pendingPullIntos.
(readableByteStreamControllerCancel): Updated pendingPullIntos.
(readableByteStreamControllerClose): Updated pendingPullIntos.
(readableByteStreamControllerHandleQueueDrain): Added.
(readableByteStreamControllerPull): Added.

LayoutTests:

Added 2 tests that allow covering most of new code. Code not yet tested will
become reachable as the rest of the spec is implemented (new tests will then
be added).

* streams/readable-byte-stream-controller-expected.txt:
* streams/readable-byte-stream-controller.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211484 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed. Fix coding style in MiniBrowser
tpopela@redhat.com [Wed, 1 Feb 2017 15:28:08 +0000 (15:28 +0000)]
Unreviewed. Fix coding style in MiniBrowser

* MiniBrowser/gtk/BrowserWindow.c:
(editingCommandCallback):
(insertImageCommandCallback):
(insertLinkCommandCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211483 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoImplement the alwaysRunsAtBackgroundPriority WK2 setting using thread QoS.
akling@apple.com [Wed, 1 Feb 2017 14:55:00 +0000 (14:55 +0000)]
Implement the alwaysRunsAtBackgroundPriority WK2 setting using thread QoS.
<https://webkit.org/b/167387>
<rdar://problem/29711409>

Reviewed by Antti Koivisto.

Source/bmalloc:

Support changing the QoS level of the scavenger thread asynchronously through
a request variable. This is not the most elegant thing in the world, but since
threads are only allowed to change their own QoS class, our options are limited.

* bmalloc/Heap.cpp:
(bmalloc::Heap::concurrentScavenge):
* bmalloc/Heap.h:
(bmalloc::Heap::takeRequestedScavengerThreadQOSClass):
(bmalloc::Heap::setScavengerThreadQOSClass):
* bmalloc/bmalloc.h:
(bmalloc::api::setScavengerThreadQOSClass):

Source/WebKit2:

Remove the old ProcessThrottlerClient implementation of alwaysRunsAtBackgroundPriority
and replace it with WTF::setGlobalMaxQOSClass().

If the setting is enabled, it's passed over to each WK2 child process along with its
bootstrap parameter, and we notify WTF in XPCServiceInitializer(), before anything
too threading related happens.

* Platform/IPC/Connection.cpp:
(IPC::Connection::processIncomingMessage):
* Shared/EntryPointUtilities/mac/XPCService/XPCServiceEntryPoint.h:
(WebKit::XPCServiceInitializer):
* Shared/EntryPointUtilities/mac/XPCService/XPCServiceEntryPoint.mm:
(WebKit::XPCServiceInitializerDelegate::getExtraInitializationData):
* UIProcess/API/APIProcessPoolConfiguration.h:
* UIProcess/ChildProcessProxy.cpp:
(WebKit::ChildProcessProxy::ChildProcessProxy):
(WebKit::ChildProcessProxy::getLaunchOptions):
* UIProcess/ChildProcessProxy.h:
* UIProcess/Databases/DatabaseProcessProxy.cpp:
(WebKit::DatabaseProcessProxy::DatabaseProcessProxy):
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::NetworkProcessProxy):
(WebKit::NetworkProcessProxy::alwaysRunsAtBackgroundPriority): Deleted.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::assertionState):
* UIProcess/ProcessThrottlerClient.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::WebProcessProxy):
(WebKit::WebProcessProxy::alwaysRunsAtBackgroundPriority): Deleted.
* UIProcess/WebProcessProxy.h:

Source/WTF:

Add a new mechanism for overriding the max thread QoS level globally:

    void setGlobalMaxQOSClass(qos_class_t)
    qos_class_t adjustedQOSClass(qos_class_t)

The QoS cap applies to all newly created threads, threads that try to override
their QoS class manually, and also passed down to bmalloc.

* wtf/Threading.cpp:
(WTF::setCurrentThreadIsUserInteractive):
(WTF::setCurrentThreadIsUserInitiated):
(WTF::setGlobalMaxQOSClass):
(WTF::adjustedQOSClass):
* wtf/Threading.h:
* wtf/ThreadingPthreads.cpp:
(WTF::createThreadInternal):
* wtf/cocoa/WorkQueueCocoa.cpp:
(WTF::dispatchQOSClass):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211482 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[GTK] Add an API to add a custom tab into the print dialog
tpopela@redhat.com [Wed, 1 Feb 2017 14:14:36 +0000 (14:14 +0000)]
[GTK] Add an API to add a custom tab into the print dialog
https://bugs.webkit.org/show_bug.cgi?id=151998

Reviewed by Carlos Garcia Campos.

Source/WebKit2:

Add a new create-custom-widget signal to the WebKitPrintOperation. The
signal is emitted before the dialog is displayed and it gives an
opportunity to embed a custom widget in the dialog. You can do so by
creating a new WebKitPrintCustomWidget and returning it from the
create-custom-widget signal handler. The WebKitPrintCustomWidget is
emitting two signals:
  - update - emitted when the currently selected printer is changed,
             to be able to actualize the custom widget based on the
             current printer
  - apply - emitted when the dialog is closed, just before the
            printing will start, to be able e.g. to change content
            based on the custom widget state.

* PlatformGTK.cmake:
* UIProcess/API/gtk/WebKitAutocleanups.h:
* UIProcess/API/gtk/WebKitPrintCustomWidget.cpp: Added.
(webkitPrintCustomWidgetSetProperty):
(webkit_print_custom_widget_class_init):
(webkit_print_custom_widget_new):
(webkit_print_custom_widget_get_widget):
(webkit_print_custom_widget_get_title):
(webkitPrintCustomWidgetEmitCustomWidgetApplySignal):
(webkitPrintCustomWidgetEmitUpdateCustomWidgetSignal):
* UIProcess/API/gtk/WebKitPrintCustomWidget.h: Added.
* UIProcess/API/gtk/WebKitPrintCustomWidgetPrivate.h: Added.
* UIProcess/API/gtk/WebKitPrintOperation.cpp:
(webkitPrintOperationAccumulatorObjectHandled):
(webkit_print_operation_class_init):
(notifySelectedPrinterCallback):
(webkitPrintOperationRunDialog):
* UIProcess/API/gtk/docs/webkit2gtk-4.0-sections.txt:
* UIProcess/API/gtk/docs/webkit2gtk-4.0.types:
* UIProcess/API/gtk/docs/webkit2gtk-docs.sgml:
* UIProcess/API/gtk/webkit2.h:

Tools:

Add a new WebKitPrintOperation/custom-widget test in TestPrinting
that is testing a newly added API.

* TestWebKitAPI/Tests/WebKit2Gtk/TestPrinting.cpp:
(testPrintCustomWidget):
(beforeAll):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211481 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoLoad resources speculatively
antti@apple.com [Wed, 1 Feb 2017 13:38:49 +0000 (13:38 +0000)]
Load resources speculatively
https://bugs.webkit.org/show_bug.cgi?id=167660

Reviewed by Andreas Kling.

* NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::makeEntry):

    Factor to a function.

(WebKit::NetworkCache::Cache::store):
* NetworkProcess/cache/NetworkCache.h:
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.cpp:
(WebKit::NetworkCache::SpeculativeLoad::SpeculativeLoad):

    Support loads where we don't have existing cache entry to validate.

(WebKit::NetworkCache::SpeculativeLoad::didReceiveBuffer):

    Synthesize a NetworkCache::Entry if we can't store it.

(WebKit::NetworkCache::SpeculativeLoad::didFinishLoading):
(WebKit::NetworkCache::SpeculativeLoad::didFailLoading):
(WebKit::NetworkCache::SpeculativeLoad::abort):
(WebKit::NetworkCache::SpeculativeLoad::didComplete):
* NetworkProcess/cache/NetworkCacheSpeculativeLoad.h:
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.cpp:
(WebKit::NetworkCache::constructRevalidationRequest):

    Make having existing cache entry optional.

(WebKit::NetworkCache::SpeculativeLoadManager::retrieveEntryFromStorage):

    Allow validation without validation headers (that is, normal load).

(WebKit::NetworkCache::SpeculativeLoadManager::revalidateSubresource):

    Make having existing cache entry optional.

(WebKit::NetworkCache::canRevalidate):

    Allow speculative loads without validation headers if we have high confidence that the
    page is going to request this resource again. This is based on the time span we have
    seen this resource being loaded on a given page and how much time has elapsed since we
    last loaded it.

    For example if we have seen the resource over the last 10 days we'll speculate that it will
    be required for the next 5 days.

(WebKit::NetworkCache::SpeculativeLoadManager::preloadEntry):
(WebKit::NetworkCache::SpeculativeLoadManager::revalidateEntry): Deleted.
* NetworkProcess/cache/NetworkCacheSpeculativeLoadManager.h:
* NetworkProcess/cache/NetworkCacheSubresourcesEntry.cpp:
(WebKit::NetworkCache::SubresourceInfo::encode):
(WebKit::NetworkCache::SubresourceInfo::decode):

    Encode the firstSeen and lastSeen time stamps.

(WebKit::NetworkCache::SubresourceInfo::SubresourceInfo):
(WebKit::NetworkCache::makeSubresourceInfoVector):
(WebKit::NetworkCache::SubresourcesEntry::SubresourcesEntry):
(WebKit::NetworkCache::SubresourcesEntry::updateSubresourceLoads):
* NetworkProcess/cache/NetworkCacheSubresourcesEntry.h:
(WebKit::NetworkCache::SubresourceInfo::lastSeen):
(WebKit::NetworkCache::SubresourceInfo::firstSeen):
(WebKit::NetworkCache::SubresourceInfo::setNonTransient):
(WebKit::NetworkCache::SubresourceInfo::SubresourceInfo): Deleted.
(WebKit::NetworkCache::SubresourceInfo::setTransient): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211480 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoArityFixup should adjust SP first
utatane.tea@gmail.com [Wed, 1 Feb 2017 11:29:25 +0000 (11:29 +0000)]
ArityFixup should adjust SP first
https://bugs.webkit.org/show_bug.cgi?id=167239

Reviewed by Michael Saboff.

JSTests:

Significantly large arity fixup reliably causes this crash.

* stress/arity-fixup-should-not-touch-stack-area-below-sp.js: Added.

Source/JavaScriptCore:

Arity fixup extends the stack and copy/fill the stack with
the values. At that time, we accidentally read/write stack
space below the stack pointer. As a result, we touch the area
of the stack space below the x64 red zone. These areas are unsafe.
OS may corrupt this space when constructing a signal stack.
The Linux kernel could not populate the pages for this space
and causes segmentation fault. This patch changes the stack
pointer before performing the arity fixup.

* jit/ThunkGenerators.cpp:
(JSC::arityFixupGenerator):
* llint/LowLevelInterpreter32_64.asm:
* llint/LowLevelInterpreter64.asm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211479 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[GStreamer][MSE] qtdemux: Update the tfdt patch to the version finally accepted upstream
eocanha@igalia.com [Wed, 1 Feb 2017 10:59:17 +0000 (10:59 +0000)]
[GStreamer][MSE] qtdemux: Update the tfdt patch to the version finally accepted upstream
https://bugs.webkit.org/show_bug.cgi?id=166529

Reviewed by Xabier Rodriguez-Calvar.

Source/WebCore:

* platform/graphics/gstreamer/mse/AppendPipeline.cpp:
(WebCore::AppendPipeline::AppendPipeline): Don't use the "always-honor-tfdt" property
in qtdemux anymore.

Tools:

* gtk/patches/gst-plugins-good-use-the-tfdt-decode-time.patch:
New patch version accepted in GStreamer upstream.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211478 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[GTK] Install either libxfont1-dev or libxfont-dev as dependency
calvaris@igalia.com [Wed, 1 Feb 2017 10:58:52 +0000 (10:58 +0000)]
[GTK] Install either libxfont1-dev or libxfont-dev as dependency
https://bugs.webkit.org/show_bug.cgi?id=167498

Reviewed by Carlos Garcia Campos.

Depending on the distro version you can need either libxfont1-dev
or libfont-dev to build with jhbuild.

* gtk/install-dependencies:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211477 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[Mac][cmake] Unreviewed speculative buildfix after r211403.
ossy@webkit.org [Wed, 1 Feb 2017 10:46:10 +0000 (10:46 +0000)]
[Mac][cmake] Unreviewed speculative buildfix after r211403.
https://bugs.webkit.org/show_bug.cgi?id=167601

* UIProcess/API/Cocoa/WKWebView.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211476 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, build fix for macOS and iOS ports, part 2
utatane.tea@gmail.com [Wed, 1 Feb 2017 10:42:35 +0000 (10:42 +0000)]
Unreviewed, build fix for macOS and iOS ports, part 2

Annotate the function with static.

* bindings/js/ScriptModuleLoader.cpp:
(WebCore::rejectToPropagateNetworkError):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211475 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoUnreviewed, build fix for macOS and iOS ports
utatane.tea@gmail.com [Wed, 1 Feb 2017 10:32:33 +0000 (10:32 +0000)]
Unreviewed, build fix for macOS and iOS ports

* bindings/js/ModuleFetchFailureKind.h: Renamed from Source/ModuleFetchFailureKind.h.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211474 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoPropagate networking errors correctly for import() operator
utatane.tea@gmail.com [Wed, 1 Feb 2017 10:00:32 +0000 (10:00 +0000)]
Propagate networking errors correctly for import() operator
https://bugs.webkit.org/show_bug.cgi?id=167501

Reviewed by Ryosuke Niwa.

.:

* Source/ModuleFetchFailureKind.h: Added.

Source/WebCore:

We use the promise chain inside the module loader pipeline.
The problem is that,

1. The errors of loading module scripts are propagated through the promise's rejection.
2. Some module related errors like syntax error (when scanning module dependencies) are
   reported at loading time (evaluating time). These errors are propagated through the
   promise rejections and dumped to the console. LoadableModuleScript set catch handler
   for that and print these errors to the console for ScriptElement.
3. Some of errors are already reported to the console. For example, networking errors
   are reported by the CachedModuleLoader.
4. But there is no way to distinguish between (2) and (3) at the catch handler.

Previously, we propagate the special symbol for the already reported errors to distinguish
that. This was OK because these errors cannot be catched by the user code. However,
recent `import()` call needs to expose these errors to the user code. So now, accidentally,
the special purpose symbol is exposed.

Instead of propagating the special symbol, this patch just propagates the errors. But these
error objects are annotated with the private symbol. So these errors can be distinguished
from the other ones.

Currently, we do not have the way to expose the error details to the client of the
CachedScript. So we just report the error with the failed / canceled. This should be
more descriptive error messages in the separate patch[1].

[1]: https://bugs.webkit.org/show_bug.cgi?id=167553

Tests: http/tests/security/mixedContent/import-insecure-script-in-iframe.html
       js/dom/modules/import-not-found-error.html

* WebCore.xcodeproj/project.pbxproj:
* bindings/js/JSDOMPromise.h:
(WebCore::DeferredPromise::resolveWithCallback):
(WebCore::DeferredPromise::rejectWithCallback):
* bindings/js/ScriptController.cpp:
(WebCore::ScriptController::setupModuleScriptHandlers):
* bindings/js/ScriptController.h:
(WebCore::ScriptController::moduleLoaderAlreadyReportedErrorSymbol): Deleted.
(WebCore::ScriptController::moduleLoaderFetchingIsCanceledSymbol): Deleted.
* bindings/js/ScriptModuleLoader.cpp:
(WebCore::rejectToPropagateNetworkError):
(WebCore::ScriptModuleLoader::fetch):
(WebCore::ScriptModuleLoader::notifyFinished):
* bindings/js/WebCoreBuiltinNames.h:

LayoutTests:

* http/tests/security/mixedContent/import-insecure-script-in-iframe-expected.txt: Added.
* http/tests/security/mixedContent/import-insecure-script-in-iframe.html: Added.
* http/tests/security/mixedContent/resources/frame-with-insecure-import.html: Added.
* js/dom/modules/import-not-found-error-expected.txt: Added.
* js/dom/modules/import-not-found-error.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211473 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoTeach cache coders to encode time_points
antti@apple.com [Wed, 1 Feb 2017 07:55:44 +0000 (07:55 +0000)]
Teach cache coders to encode time_points
https://bugs.webkit.org/show_bug.cgi?id=167670

Reviewed by Andreas Kling.

Source/WebKit2:

Encode time_point instead of duration.

* NetworkProcess/cache/NetworkCacheStorage.cpp:
(WebKit::NetworkCache::decodeRecordMetaData):
(WebKit::NetworkCache::Storage::readRecord):
(WebKit::NetworkCache::encodeRecordMetaData):
(WebKit::NetworkCache::Storage::encodeRecord):
(WebKit::NetworkCache::Storage::traverse):

Source/WTF:

* wtf/persistence/Coders.h:
(WTF::Persistence::Coder<std::chrono::system_clock::time_point>::encode):
(WTF::Persistence::Coder<std::chrono::system_clock::time_point>::decode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211472 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRegression (Safari 10.1): Pressing Return in a contenteditable no longer inserts...
wenson_hsieh@apple.com [Wed, 1 Feb 2017 07:46:31 +0000 (07:46 +0000)]
Regression (Safari 10.1): Pressing Return in a contenteditable no longer inserts a line break under certain conditions
https://bugs.webkit.org/show_bug.cgi?id=167525
<rdar://problem/30270210>

Reviewed by Ryosuke Niwa.

Source/WebCore:

Test: fast/events/input-events-insert-newlines-after-mutation.html

Reverts an unintended change made while refactoring code for input events that caused input events to be
dispatched immediately rather than on the scoped queue. Normally, input events are dispatched in
CompositeEditCommand::apply after the end of the scope, but TypingCommands may fire input events *from within*
the scope by calling typingAddedToOpenCommand.

Instead, TypingCommands should always dispatch events
synchronously after the end of the scoped queue in CompositeEditCommand::apply, but this is a riskier change
than we should currently allow, so we should revert to our old behavior for the time being.

* editing/Editor.cpp:

LayoutTests:

Adds a new test covering newline insertion with mutation observers and an input event handler. Also rebaselines
a drag and drop test to account for dispatching input events on the scoped queue.

* fast/events/input-events-drag-and-drop-expected.txt:
* fast/events/input-events-insert-newlines-after-mutation-expected.txt: Added.
* fast/events/input-events-insert-newlines-after-mutation.html: Added.
* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211471 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years ago[WebRTC][WebKit2] Support expanding the sandbox to allow microphone access at process...
bfulgham@apple.com [Wed, 1 Feb 2017 06:07:51 +0000 (06:07 +0000)]
[WebRTC][WebKit2] Support expanding the sandbox to allow microphone access at process launch
https://bugs.webkit.org/show_bug.cgi?id=167669
<rdar://problem/29974333>

Reviewed by Alexey Proskuryakov.

This change implements a temporary workaround to <rdar://problem/29448368> to allow us to do live
testing of WebRTC microphone access. It should be removed when that underlying bug is fixed.

This change adds a new process creation parameter that indicates whether we wish to expand the
sandbox to allow microphone access.

Tested by WebRTC suite (part of another set of bugs).

* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode): Serialize the audio access entitlement handle.
(WebKit::WebProcessCreationParameters::decode): Ditto.
* Shared/WebProcessCreationParameters.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::createNewWebProcess): Create an entitlement handle and pass along to
the child process.
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess): If an audio entitlement (for the microphone)
was desired, expand the sandbox to include access.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211470 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRemoved unused m_nestedCount from RunLoop
commit-queue@webkit.org [Wed, 1 Feb 2017 05:40:17 +0000 (05:40 +0000)]
Removed unused m_nestedCount from RunLoop
https://bugs.webkit.org/show_bug.cgi?id=167674

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2017-01-31
Reviewed by Sam Weinig.

* wtf/RunLoop.h:
* wtf/cf/RunLoopCF.cpp:
(WTF::RunLoop::RunLoop):
(WTF::RunLoop::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211469 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agold: warning: instance method '_setAutofilled:' in category conflicts with same method...
joepeck@webkit.org [Wed, 1 Feb 2017 05:34:31 +0000 (05:34 +0000)]
ld: warning: instance method '_setAutofilled:' in category conflicts with same method from another category
https://bugs.webkit.org/show_bug.cgi?id=167684

Reviewed by Dan Bernstein.

* DOM/DOMHTML.mm:
(-[DOMHTMLInputElement _isAutofilled]): Deleted.
(-[DOMHTMLInputElement _setAutofilled:]): Deleted.
Consolidate in the WebDOMOperationsPrivate category.

* DOM/DOMPrivate.h:
Include WebDOMOperationsPrivate header to get other DOM Private categories.

* DOM/WebDOMOperations.mm:
(-[DOMHTMLInputElement _isAutofilled]):
* DOM/WebDOMOperationsPrivate.h:
Include getter alongside the setter.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211468 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoDisable API test WebKit2.WKWebProcessPlugInRangeHandle
ryanhaddad@apple.com [Wed, 1 Feb 2017 03:56:10 +0000 (03:56 +0000)]
Disable API test WebKit2.WKWebProcessPlugInRangeHandle
https://bugs.webkit.org/show_bug.cgi?id=167594

Unreviewed test gardening.

Disabling this test while we investigate the cause of a timeout on ios-simulator.

* TestWebKitAPI/Tests/WebKit2Cocoa/BundleRangeHandle.mm:
(TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211467 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoSimple line layout: Do not assert on zero length/width trailing whitespace.
zalan@apple.com [Wed, 1 Feb 2017 03:19:23 +0000 (03:19 +0000)]
Simple line layout: Do not assert on zero length/width trailing whitespace.
https://bugs.webkit.org/show_bug.cgi?id=167679

Reviewed by Simon Fraser.

It's valid to have an zero length/width trailing whitespace.

Covered by existing tests.

* rendering/SimpleLineLayout.cpp:
(WebCore::SimpleLineLayout::LineState::removeTrailingWhitespace): Zero length -> nothing to do.
Zero width with length > 0 (font-size: 0) -> remove trailing run.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211466 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMarking gamepad/gamepad-visibility-1.html as flaky.
ryanhaddad@apple.com [Wed, 1 Feb 2017 02:16:55 +0000 (02:16 +0000)]
Marking gamepad/gamepad-visibility-1.html as flaky.
https://bugs.webkit.org/show_bug.cgi?id=167517

Unreviewed test gardening.

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211465 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMove slow-running microbenchmarks out of JSTests/microbenchmarks
fpizlo@apple.com [Wed, 1 Feb 2017 01:48:18 +0000 (01:48 +0000)]
Move slow-running microbenchmarks out of JSTests/microbenchmarks
https://bugs.webkit.org/show_bug.cgi?id=167676

Reviewed by Saam Barati.
JSTests:

The microbenchmarks directory is profitable when it's cheap to run. This change moves very slow
tests (>=200ms running time on my machine) and puts them in JSTests/slowMicrobenchmarks instead.
That directory is run only through `slow!; defaultQuickRun` and is not run by run-jsc-benchmarks.
This is a big progression in test running time. But because it frontloads more slow tests, the
tests will seem to run more slowly when you first start the test runner.

* microbenchmarks/default-derived-constructor.js: Removed.
* microbenchmarks/dense-set.js: Removed.
* microbenchmarks/ftl-polymorphic-div.js: Removed.
* microbenchmarks/ftl-polymorphic-mul.js: Removed.
* microbenchmarks/function-bind-inlining.js: Removed.
* microbenchmarks/function-bind-no-inlining.js: Removed.
* microbenchmarks/large-map-iteration-with-additions.js: Removed.
* microbenchmarks/large-map-iteration-with-mutation.js: Removed.
* microbenchmarks/map-get-get-cse.js: Removed.
* microbenchmarks/misc-bugs-847389-jpeg2000.js: Removed.
* microbenchmarks/spread-small-array.js: Removed.
* microbenchmarks/v8-raytrace-with-try-catch-high-frequency-throws.js: Removed.
* slowMicrobenchmarks: Added.
* slowMicrobenchmarks.yaml: Added.
* slowMicrobenchmarks/README: Added.
* slowMicrobenchmarks/default-derived-constructor.js: Copied from JSTests/microbenchmarks/default-derived-constructor.js.
* slowMicrobenchmarks/dense-set.js: Copied from JSTests/microbenchmarks/dense-set.js.
* slowMicrobenchmarks/ftl-polymorphic-div.js: Copied from JSTests/microbenchmarks/ftl-polymorphic-div.js.
* slowMicrobenchmarks/ftl-polymorphic-mul.js: Copied from JSTests/microbenchmarks/ftl-polymorphic-mul.js.
* slowMicrobenchmarks/function-bind-inlining.js: Copied from JSTests/microbenchmarks/function-bind-inlining.js.
* slowMicrobenchmarks/function-bind-no-inlining.js: Copied from JSTests/microbenchmarks/function-bind-no-inlining.js.
* slowMicrobenchmarks/large-map-iteration-with-additions.js: Copied from JSTests/microbenchmarks/large-map-iteration-with-additions.js.
* slowMicrobenchmarks/large-map-iteration-with-mutation.js: Copied from JSTests/microbenchmarks/large-map-iteration-with-mutation.js.
* slowMicrobenchmarks/map-get-get-cse.js: Copied from JSTests/microbenchmarks/map-get-get-cse.js.
* slowMicrobenchmarks/misc-bugs-847389-jpeg2000.js: Copied from JSTests/microbenchmarks/misc-bugs-847389-jpeg2000.js.
* slowMicrobenchmarks/spread-small-array.js: Copied from JSTests/microbenchmarks/spread-small-array.js.
* slowMicrobenchmarks/v8-raytrace-with-try-catch-high-frequency-throws.js: Copied from JSTests/microbenchmarks/v8-raytrace-with-try-catch-high-frequency-throws.js.

Tools:

* Scripts/run-javascriptcore-tests:
(runJSCStressTests): Add slowMicrobenchmarks.yaml.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211464 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMake verifyEdge a RELEASE_ASSERT
fpizlo@apple.com [Wed, 1 Feb 2017 01:46:33 +0000 (01:46 +0000)]
Make verifyEdge a RELEASE_ASSERT
<rdar://problem/30296879>

Rubber stamped by Saam Barati.

* dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211463 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWeb Inspector: Tooltip text should not end in a period
commit-queue@webkit.org [Wed, 1 Feb 2017 01:26:58 +0000 (01:26 +0000)]
Web Inspector: Tooltip text should not end in a period
https://bugs.webkit.org/show_bug.cgi?id=167130

Patch by Devin Rousso <dcrousso+webkit@gmail.com> on 2017-01-31
Reviewed by Brian Burg.

* Localizations/en.lproj/localizedStrings.js:
* UserInterface/Views/BezierEditor.js:
(WebInspector.BezierEditor):
* UserInterface/Views/CSSStyleDeclarationSection.js:
(WebInspector.CSSStyleDeclarationSection.prototype._handleSelectorPaste):
* UserInterface/Views/CSSStyleDeclarationTextEditor.js:
(WebInspector.CSSStyleDeclarationTextEditor.prototype._createTextMarkerForPropertyIfNeeded):
* UserInterface/Views/DOMNodeDetailsSidebarPanel.js:
(WebInspector.DOMNodeDetailsSidebarPanel.prototype._refreshAccessibility.accessibilityPropertiesCallback):
* UserInterface/Views/DebuggerSidebarPanel.js:
(WebInspector.DebuggerSidebarPanel):
* UserInterface/Views/GeneralTabBarItem.js:
(WebInspector.GeneralTabBarItem):
* UserInterface/Views/HeapSnapshotContentView.js:
(WebInspector.HeapSnapshotInstancesContentView):
(WebInspector.HeapSnapshotObjectGraphContentView):
* UserInterface/Views/InlineSwatch.js:
(WebInspector.InlineSwatch):
* UserInterface/Views/ObjectTreePropertyTreeElement.js:
(WebInspector.ObjectTreePropertyTreeElement.prototype._updateEntries):
(WebInspector.ObjectTreePropertyTreeElement.prototype._updateProperties):
* UserInterface/Views/ObjectTreeView.js:
(WebInspector.ObjectTreeView.prototype._updateEntries):
(WebInspector.ObjectTreeView.prototype._updateProperties):
* UserInterface/Views/ResourceTimelineDataGridNode.js:
(WebInspector.ResourceTimelineDataGridNode.prototype._mouseoverRecordBar):
* UserInterface/Views/SpringEditor.js:
(WebInspector.SpringEditor):
* UserInterface/Views/TypeTreeElement.js:
(WebInspector.TypeTreeElement.prototype.onpopulate):
* UserInterface/Views/TypeTreeView.js:
(WebInspector.TypeTreeView.prototype._populate):
* UserInterface/Views/VisualStyleCommaSeparatedKeywordEditor.js:
(WebInspector.VisualStyleCommaSeparatedKeywordEditor):
* UserInterface/Views/VisualStyleDetailsPanel.js:
(WebInspector.VisualStyleDetailsPanel.prototype._generateSection.createOptionsElement):
* UserInterface/Views/VisualStylePropertyEditorLink.js:
(WebInspector.VisualStylePropertyEditorLink.prototype.set linked):
* UserInterface/Views/VisualStylePropertyNameInput.js:
(WebInspector.VisualStylePropertyNameInput):
* UserInterface/Views/VisualStyleSelectorSection.js:
(WebInspector.VisualStyleSelectorSection):
* UserInterface/Views/VisualStyleSelectorTreeItem.js:
(WebInspector.VisualStyleSelectorTreeItem.prototype._updateCheckboxTitle):
(WebInspector.VisualStyleSelectorTreeItem.prototype._updateSelectorIcon):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211462 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoOSR entry: delay outer-loop compilation when at inner-loop
jfbastien@apple.com [Wed, 1 Feb 2017 01:26:00 +0000 (01:26 +0000)]
OSR entry: delay outer-loop compilation when at inner-loop
https://bugs.webkit.org/show_bug.cgi?id=167149

Reviewed by Filip Pizlo.

r211224 was reverted because it caused a massive kraken/ai-astar
regression. This patch instead does the minimally-disruptive
change to fix the original bug as described below, but omits extra
tuning and refactoring which I had before. I'll commit tuning and
refactoring separately, if this sticks. This patch is therefore
very minimal, and layers carefully on top of the complex
spaghetti-logic. The only change it makes is that it uses triggers
to indicate to outer loops that they should compile, which fixes
the immediate bug and seems roughly perf neutral (maybe a small
gain on kraken sometimes, other times a small regression as would
be expected from compiling later).

As of https://bugs.webkit.org/show_bug.cgi?id=155217 OSR
compilation can be kicked off for an entry into an outer-loop,
while executing an inner-loop. This is desirable because often the
codegen from an inner-entry isn't as good as the codegen from an
outer-entry, but execution from an inner-loop is often pretty hot
and likely to kick off compilation. This approach provided nice
speedups on Kraken because we'd select to enter to the outer-loop
very reliably, which reduces variability (the inner-loop was
selected roughly 1/5 times from my unscientific measurements).

When compilation starts we take a snapshot of the JSValues at the
current execution state using OSR's recovery mechanism. These
values are passed to the compiler and are used as way to perform
type profiling, and could be used to observe cell types as well as
to perform predictions such as through constant propagation.

It's therefore desired to enter from the outer-loop when we can,
but we need to be executing from that location to capture the
right JSValues, otherwise we're confusing the compiler and giving
it inaccurate JSValues which can lead it to predict the wrong
things, leading to suboptimal code or recompilation due to
misprediction, or in super-corner-cases a crash.

These effects are pretty hard to measure: Fil points out that
marsalis-osr-entry really needs mustHandleValues (the JSValues
from the point of execution) because right now it just happens to
correctly guess int32. I tried removing mustHandleValues entirely
and saw no slowdowns, but our benchmarks probably aren't
sufficient to reliably find issues, sometimes because we happen to
have sufficient mitigations.

DFG tier-up was added here:
https://bugs.webkit.org/show_bug.cgi?id=112838

* dfg/DFGOperations.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211461 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoMake Dump Render Tree not crash
megan_gardner@apple.com [Wed, 1 Feb 2017 01:04:03 +0000 (01:04 +0000)]
Make Dump Render Tree not crash
https://bugs.webkit.org/show_bug.cgi?id=167667

DumpRenderTree is crashing when running WK2 tests that it's not supposed to be running.
DumpRenderTree uses RunLoop::main.dispatch(), but that crashes because it's not set up on iOS.
So we're making use of RunWebThread, which is the correct thing to do on iOS, and it doesn't crash.
This does not fix any tests, it just fixes the crash.

Reviewed by Simon Fraser.

* DumpRenderTree/TestRunner.cpp:
(TestRunner::callUIScriptCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211460 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoFix the build.
timothy_horton@apple.com [Wed, 1 Feb 2017 00:39:56 +0000 (00:39 +0000)]
Fix the build.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _snapshotRect:intoImageOfWidth:completionHandler:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211459 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoFix debug libwebrtc build after r211439
achristensen@apple.com [Wed, 1 Feb 2017 00:18:35 +0000 (00:18 +0000)]
Fix debug libwebrtc build after r211439
https://bugs.webkit.org/show_bug.cgi?id=167572

* platform/mediastream/libwebrtc/LibWebRTCAudioModule.cpp:
(WebCore::LibWebRTCAudioModule::OnMessage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211458 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWorkaround for simctl install failing to report install failure
jbedard@apple.com [Wed, 1 Feb 2017 00:15:44 +0000 (00:15 +0000)]
Workaround for simctl install failing to report install failure
https://bugs.webkit.org/show_bug.cgi?id=167654

Reviewed by Daniel Bates.

simctl install will sometimes report success even though install has failed.
To workaround this bug, we attempt an app launch after install to ensure
install has succeeded.

* Scripts/webkitpy/xcode/simulator.py:
(Device.install_app): Attempt app launch after install, retry install if launch fails.
(Device.launch_app): Correct comment.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211457 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoSimple line layout: Collect fragments in LineState only when needed for post-processing.
zalan@apple.com [Wed, 1 Feb 2017 00:01:26 +0000 (00:01 +0000)]
Simple line layout: Collect fragments in LineState only when needed for post-processing.
https://bugs.webkit.org/show_bug.cgi?id=167655
<rdar://problem/30290592>

Reviewed by Antti Koivisto.

The fragments we need to keep around for basic line breaking are the last and last non-whitespace fragments.
However in certain cases such as text-align: justify, we have to traverse the entire line
at post-processing which requires all the fragments on the current line.
This patch speeds up the basic line handling by using m_fragment list only for the special cases.
It improves PerformanceTests/Layout/simple-line-layout-non-repeating-text.html by ~7% (local measurement)
median: 114.25 runs/s -> median: 122.16 runs

Covered by existing tests.

* rendering/SimpleLineLayout.cpp:
(WebCore::SimpleLineLayout::LineState::setNeedsAllFragments):
(WebCore::SimpleLineLayout::LineState::hasTrailingWhitespace):
(WebCore::SimpleLineLayout::LineState::lastFragment):
(WebCore::SimpleLineLayout::LineState::expansionOpportunityCount):
(WebCore::SimpleLineLayout::LineState::isEmpty):
(WebCore::SimpleLineLayout::LineState::endPositionForCollapsedFragment):
(WebCore::SimpleLineLayout::LineState::appendFragmentAndCreateRunIfNeeded): Track uncompleted width.
(WebCore::SimpleLineLayout::LineState::revertToLastCompleteFragment):
(WebCore::SimpleLineLayout::LineState::removeTrailingWhitespace):
(WebCore::SimpleLineLayout::removeTrailingWhitespace):
(WebCore::SimpleLineLayout::updateLineConstrains): Set the slow path for text-align: justify.
(WebCore::SimpleLineLayout::createLineRuns):
(WebCore::SimpleLineLayout::closeLineEndingAndAdjustRuns):
(WebCore::SimpleLineLayout::createTextRuns):
* rendering/SimpleLineLayoutTextFragmentIterator.h:
(WebCore::SimpleLineLayout::TextFragmentIterator::TextFragment::isValid):
(WebCore::SimpleLineLayout::TextFragmentIterator::TextFragment::length):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211456 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agocom.apple.WebKit.WebContent.Development crashed in com.apple.WebCore: WebCore::FrameV...
cdumez@apple.com [Tue, 31 Jan 2017 23:49:10 +0000 (23:49 +0000)]
com.apple.WebKit.WebContent.Development crashed in com.apple.WebCore: WebCore::FrameView::didDestroyRenderTree
https://bugs.webkit.org/show_bug.cgi?id=167665
<rdar://problem/30241193>

Reviewed by Andreas Kling.

When destroying a PageCache entry, we would call Document::prepareFromDestruction()
which would destroy the render tree and then call FrameView::didDestroyRenderTree().
We would sometimes hit the ASSERT(m_widgetsInRenderTree.isEmpty()) assertion in
didDestroyRenderTree(). This indicates that there are still RenderWidgets associated
with the FrameView even though we claim we already destroyed the render tree.

The reason for this is that, in the PageCache context, the top document's frame is
reused on navigation and may not be associated with a new document that is currently
displayed, has a render tree (and possible RenderWidgets). The solution proposed is
to stop calling FrameView::didDestroyRenderTree() in Document::prepareFromDestruction()
if the frame's document is not |this|. This happens when the document is destroyed
while in PageCache. We do not want to mess with the frame / frameview in this case
since they are no longer associated to the present document anymore.

No new tests, already covered by existing tests which currently crash flakily.

* dom/Document.cpp:
(WebCore::Document::destroyRenderTree):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211455 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoRemove unused/rarely used bindings helpers
commit-queue@webkit.org [Tue, 31 Jan 2017 23:38:47 +0000 (23:38 +0000)]
Remove unused/rarely used bindings helpers
https://bugs.webkit.org/show_bug.cgi?id=167648

Patch by Sam Weinig <sam@webkit.org> on 2017-01-31
Reviewed by Anders Carlsson.

* bindings/js/JSDOMBinding.cpp:
* bindings/js/JSDOMBinding.h:
(WebCore::jsString): Deleted.
(WebCore::jsStringOrUndefined): Deleted.
(WebCore::finiteInt32Value): Deleted.
(WebCore::toJS): Deleted.
Remove unused functions. The toJS overrides can be replaced with their
modern JSDOMConvert counterparts.

* bindings/js/JSDOMExceptionHandling.h:
(WebCore::toJS): Deleted.
(WebCore::toJSNewlyCreated): Deleted.
Remove unused functions.

* bindings/js/JSDOMIterator.h:
Add more specific include of JSDOMConvert.h to fix compile.

* bindings/js/JSDOMPromise.cpp:
(WebCore::DeferredPromise::reject):
Create the Symbol directly, instead of using the toJS override.

* bindings/js/JSWorkerCustom.cpp:
(WebCore::constructJSWorker):
Use JSDOMConvert toJS function.

* bindings/js/JSXMLHttpRequestCustom.cpp:
(WebCore::JSXMLHttpRequest::retrieveResponse):
Use JSDOMConvert toJS functions.

* bindings/js/ScriptController.h:
Add missing forward declaration of URL.

* bindings/js/ScriptModuleLoader.cpp:
(WebCore::ScriptModuleLoader::resolve):
Create the Symbol directly, instead of using the toJS override.

* html/HTMLPlugInImageElement.cpp:
(WebCore::HTMLPlugInImageElement::didAddUserAgentShadowRoot):
Use JSDOMConvert toJS functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211454 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWebItemProviderPasteboard should be robust when UIItemProvider fails to initialize...
wenson_hsieh@apple.com [Tue, 31 Jan 2017 23:28:22 +0000 (23:28 +0000)]
WebItemProviderPasteboard should be robust when UIItemProvider fails to initialize an object
https://bugs.webkit.org/show_bug.cgi?id=167663
<rdar://problem/30294240>

Reviewed by Tim Horton.

Adds a nil check for the result of -createObjectOfClass:error:.

* platform/ios/WebItemProviderPasteboard.mm:
(-[WebItemProviderPasteboard valuesForPasteboardType:inItemSet:]):
(-[WebItemProviderPasteboard _tryToCreateAndAppendObjectOfClass:toArray:usingProvider:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211453 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWeb Inspector: add CSS color keyword entries for all grey/gray variations
commit-queue@webkit.org [Tue, 31 Jan 2017 23:24:24 +0000 (23:24 +0000)]
Web Inspector: add CSS color keyword entries for all grey/gray variations
https://bugs.webkit.org/show_bug.cgi?id=167668

Patch by Devin Rousso <dcrousso+webkit@gmail.com> on 2017-01-31
Reviewed by Matt Baker.

* UserInterface/Models/Color.js:
(WebInspector.Color.Keywords):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211452 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoWeb Inspector: Network tab: provide a way to preserve log on navigation
commit-queue@webkit.org [Tue, 31 Jan 2017 23:22:56 +0000 (23:22 +0000)]
Web Inspector: Network tab: provide a way to preserve log on navigation
https://bugs.webkit.org/show_bug.cgi?id=166908

Patch by Devin Rousso <dcrousso+webkit@gmail.com> on 2017-01-31
Reviewed by Matt Baker.

* Localizations/en.lproj/localizedStrings.js:

* UserInterface/Base/Setting.js:
(WebInspector.settings):
Add clearNetworkOnNavigate global setting.

* UserInterface/Views/NetworkGridContentView.css:
(.content-view.network-grid > .data-grid .preserved:not(.selected) .cell-content):
(.content-view.network-grid > .data-grid .preserved:not(.selected) .cell-content .timeline-record-bar > .segment):
* UserInterface/Views/NetworkGridContentView.js:
(WebInspector.NetworkGridContentView):
(WebInspector.NetworkGridContentView.prototype._mainResourceDidChange):
* UserInterface/Views/NetworkSidebarPanel.css:
(.sidebar > .panel.navigation.network .tree-outline > .preserved:not(.selected) > :not(.status)):
(.sidebar > .panel.navigation.network .tree-outline > .preserved:not(.selected) > .icon):
* UserInterface/Views/NetworkSidebarPanel.js:
(WebInspector.NetworkSidebarPanel):
(WebInspector.NetworkSidebarPanel.prototype._mainResourceDidChange):
Add styling to preserved network records.

* UserInterface/Controllers/TimelineManager.js:
(WebInspector.TimelineManager.prototype._mainResourceDidChange):
Only reset the persistent network timeline if the clearNetworkOnNavigate setting is true.

* UserInterface/Views/SettingsTabContentView.js:
(WebInspector.SettingsTabContentView.prototype.initialLayout):
(WebInspector.SettingsTabContentView):
Add checkbox and spacers for new setting.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211451 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoAdding ios-simulator result file for fast/shadow-dom/copy-shadow-tree.html.
ryanhaddad@apple.com [Tue, 31 Jan 2017 22:44:54 +0000 (22:44 +0000)]
Adding ios-simulator result file for fast/shadow-dom/copy-shadow-tree.html.

Unreviewed test gardening.

* platform/ios-simulator/fast/shadow-dom/copy-shadow-tree-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211450 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoVariable used after it's been WTFMove'd
krollin@apple.com [Tue, 31 Jan 2017 22:40:50 +0000 (22:40 +0000)]
Variable used after it's been WTFMove'd
https://bugs.webkit.org/show_bug.cgi?id=167658

Reviewed by Chris Dumez.

Rather than WTFMoving the variable as part of passing it to a function
that takes an RValue-ref, call copyRef to provide the RValue-ref. This
preserves the original variable for later use in the function.

No new tests -- this issue was found by inspection. It's not apparent
what the side-effect of this bug is or how a test would verify it.

* loader/SubresourceLoader.cpp:
(WebCore::SubresourceLoader::didReceiveDataOrBuffer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211449 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoThe mutator should be able to perform increments of GC work
fpizlo@apple.com [Tue, 31 Jan 2017 22:31:24 +0000 (22:31 +0000)]
The mutator should be able to perform increments of GC work
https://bugs.webkit.org/show_bug.cgi?id=167528

Reviewed by Keith Miller and Geoffrey Garen.

Source/JavaScriptCore:

The cool thing about having a concurrent and parallel collector is that it's easy to also make
it incremental, because the load balancer can also hand over work to anyone (including the
mutator) and since the collector is running concurrently anyway, the mutator can usually rely
on the balancer having some spare work.

This change adds a classic work-based incremental mode to the GC. When you allocate K bytes,
you have to do Options::gcIncrementScale() * K "bytes" of draining. This is ammortized so that
it only happens in allocation slow paths.

On computers that have a lot of CPUs, this mode is not profitable and we set gcIncrementScale
to zero. On such computers, Riptide was already performing great because there was no way that
one mutator thread could outpace many GC threads. But on computers with fewer CPUs, there were
problems having to do with making the collector progress quickly enough so that the heap
doesn't grow too much. The stochastic scheduler actually made things worse, because it relies
a lot on the fact that the GC will simply be faster than the mutator anyway. The old scheduler
claimed to address the problem of GC pace, but it used a time-based scheduler, which is not as
precise at keeping pase as the new work-based incremental mode.

In theory, the work-based mode guarantees a bound on how much the heap can grow during a
collection just because each byte allocated means some number of bytes visited. We don't try
to create such a theoretical bound. We're just trying to give the collector an unfair advantage
in any race with the mutator.

Turning on incremental mode, the stochastic scheduler, and passive draining in combination with
each other is a huge splay-latency speed-up on my iPad. It's also a CDjs progression. It does
regress splay-throughput, but I think that's fine (the regression is 11%, the progression is
3x).

* heap/Heap.cpp:
(JSC::Heap::Heap):
(JSC::Heap::~Heap):
(JSC::Heap::markToFixpoint):
(JSC::Heap::updateObjectCounts):
(JSC::Heap::endMarking):
(JSC::Heap::finalize):
(JSC::Heap::didAllocate):
(JSC::Heap::visitCount):
(JSC::Heap::bytesVisited):
(JSC::Heap::forEachSlotVisitor):
(JSC::Heap::performIncrement):
(JSC::Heap::threadVisitCount): Deleted.
(JSC::Heap::threadBytesVisited): Deleted.
* heap/Heap.h:
* heap/MarkStack.cpp:
(JSC::MarkStackArray::transferTo):
* heap/MarkStack.h:
* heap/SlotVisitor.cpp:
(JSC::SlotVisitor::didStartMarking):
(JSC::SlotVisitor::clearMarkStacks):
(JSC::SlotVisitor::appendToMarkStack):
(JSC::SlotVisitor::noteLiveAuxiliaryCell):
(JSC::SlotVisitor::donateKnownParallel):
(JSC::SlotVisitor::drain):
(JSC::SlotVisitor::performIncrementOfDraining):
(JSC::SlotVisitor::didReachTermination):
(JSC::SlotVisitor::hasWork):
(JSC::SlotVisitor::drainFromShared):
(JSC::SlotVisitor::drainInParallelPassively):
(JSC::SlotVisitor::donateAll):
(JSC::SlotVisitor::correspondingGlobalStack):
* heap/SlotVisitor.h:
* heap/SlotVisitorInlines.h:
(JSC::SlotVisitor::reportExtraMemoryVisited):
(JSC::SlotVisitor::forEachMarkStack):
* heap/SpaceTimeMutatorScheduler.cpp:
(JSC::SpaceTimeMutatorScheduler::log):
* heap/StochasticSpaceTimeMutatorScheduler.cpp:
(JSC::StochasticSpaceTimeMutatorScheduler::log):
* jsc.cpp:
(GlobalObject::finishCreation):
(functionHeapCapacity):
* runtime/Options.cpp:
(JSC::overrideDefaults):
* runtime/Options.h:

Source/WTF:

We want dataLog to be locked even if you're not logging to a file!

* wtf/DataLog.cpp:
(WTF::initializeLogFileOnce):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211448 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoLayoutTest media/modern-media-controls/media-controller/media-controller-auto-hide...
ryanhaddad@apple.com [Tue, 31 Jan 2017 22:22:22 +0000 (22:22 +0000)]
LayoutTest media/modern-media-controls/media-controller/media-controller-auto-hide-mouse-enter-over-controls-bar.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=167266

Unreviewed test gardening.

Marking this test as flaky again.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211447 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2 years agoApple Pay button does nothing on nytimes.com
andersca@apple.com [Tue, 31 Jan 2017 22:03:59 +0000 (22:03 +0000)]
Apple Pay button does nothing on nytimes.com
https://bugs.webkit.org/show_bug.cgi?id=167664
rdar://problem/30273885

Reviewed by Sam Weinig.

Source/WebCore:

Separate line item validation into convertAndValidateTotal and convertAndValidate. The former
works on totals (which require an amount), and the latter works on regular line items that do not require an amount
if the item type is "pending".

* Modules/applepay/ApplePaySession.cpp:
(WebCore::convertAndValidateTotal):
(WebCore::convertAndValidate):
Only check the amount if the type is not pending.

(WebCore::ApplePaySession::completeShippingMethodSelection):
(WebCore::ApplePaySession::completeShippingContactSelection):
(WebCore::ApplePaySession::completePaymentMethodSelection):
Call convertAndValidateTotal for totals.

LayoutTests:

Add a new test.

* http/tests/ssl/applepay/ApplePaySession-expected.txt:
* http/tests/ssl/applepay/ApplePaySession.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@211446 268f45cc-cd09-0410-ab3c-d52691b4dbfc