WebKit-https.git
20 months ago[Modern Media Controls] Turn media/modern-media-controls/ios-inline-media-controls...
graouts@webkit.org [Thu, 1 Feb 2018 20:30:04 +0000 (20:30 +0000)]
[Modern Media Controls] Turn media/modern-media-controls/ios-inline-media-controls back on
https://bugs.webkit.org/show_bug.cgi?id=182390

Reviewed by Eric Carlson.

Update tests to match the modern-media-controls design and expectations.

* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-button-padding-expected.txt: Removed.
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-button-padding.html: Removed.
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-buttons-styles-expected.txt: Removed.
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-buttons-styles.html: Removed.
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-constructor-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-constructor.html:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-controls-bar-styles-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-controls-bar-styles.html:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-layout-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-layout.html:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-status-label-enabled-hidden-controls-bar-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-status-label-enabled-hidden-controls-bar.html:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-status-label-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-status-label.html:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-tight-padding-expected.txt: Removed.
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-tight-padding.html: Removed.
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-time-control-styles-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-time-control-styles.html:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-dropping-controls-expected.txt:
* media/modern-media-controls/ios-inline-media-controls/ios-inline-media-dropping-controls.html:
* platform/ios/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed, rebaseline imported/w3c/web-platform-tests/service-workers/service-worker...
cdumez@apple.com [Thu, 1 Feb 2018 19:47:28 +0000 (19:47 +0000)]
Unreviewed, rebaseline imported/w3c/web-platform-tests/service-workers/service-worker/performance-timeline.https.html

LayoutTests/imported/w3c:

* web-platform-tests/service-workers/service-worker/performance-timeline.https-expected.txt:

LayoutTests:

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[RenderTreeBuilder] Introduce RenderTreeBuilder to takeChild()
zalan@apple.com [Thu, 1 Feb 2018 19:42:32 +0000 (19:42 +0000)]
[RenderTreeBuilder] Introduce RenderTreeBuilder to takeChild()
https://bugs.webkit.org/show_bug.cgi?id=182373
<rdar://problem/37101484>

Reviewed by Antti Koivisto.

This is in preparation for moving mutation code out of takeChild.

No change in functionality.

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::takeChild):
* rendering/RenderBlock.h:
* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::takeChild):
* rendering/RenderBlockFlow.h:
* rendering/RenderButton.cpp:
(WebCore::RenderButton::takeChild):
* rendering/RenderButton.h:
* rendering/RenderElement.cpp:
(WebCore::RenderElement::takeChild):
(WebCore::RenderElement::removeAndDestroyChild):
(WebCore::RenderElement::destroyLeftoverChildren):
* rendering/RenderElement.h:
* rendering/RenderFullScreen.cpp:
(WebCore::RenderFullScreen::wrapExistingRenderer):
(WebCore::RenderFullScreen::unwrapRenderer):
* rendering/RenderGrid.cpp:
(WebCore::RenderGrid::takeChild):
* rendering/RenderGrid.h:
* rendering/RenderMenuList.cpp:
(RenderMenuList::takeChild):
* rendering/RenderMenuList.h:
* rendering/RenderMultiColumnFlow.cpp:
(WebCore::RenderMultiColumnFlow::processPossibleSpannerDescendant):
* rendering/RenderObject.cpp:
(WebCore::RenderObject::removeFromParentAndDestroy):
* rendering/RenderRuby.cpp:
(WebCore::RenderRubyAsInline::takeChild):
(WebCore::RenderRubyAsBlock::takeChild):
* rendering/RenderRuby.h:
* rendering/RenderRubyRun.cpp:
(WebCore::RenderRubyRun::takeChild):
* rendering/RenderRubyRun.h:
* rendering/RenderTableRow.cpp:
(WebCore::RenderTableRow::collapseAndDestroyAnonymousSiblingRows): Deleted.
* rendering/RenderTableRow.h:
* rendering/svg/RenderSVGContainer.cpp:
(WebCore::RenderSVGContainer::takeChild):
* rendering/svg/RenderSVGContainer.h:
* rendering/svg/RenderSVGInline.cpp:
(WebCore::RenderSVGInline::takeChild):
* rendering/svg/RenderSVGInline.h:
* rendering/svg/RenderSVGRoot.cpp:
(WebCore::RenderSVGRoot::takeChild):
* rendering/svg/RenderSVGRoot.h:
* rendering/svg/RenderSVGText.cpp:
(WebCore::RenderSVGText::takeChild):
* rendering/svg/RenderSVGText.h:
* rendering/updating/RenderTreeBuilder.cpp:
(WebCore::RenderTreeBuilder::collapseAndDestroyAnonymousSiblingRows):
(WebCore::RenderTreeBuilder::removeFromParentAndDestroyCleaningUpAnonymousWrappers):
* rendering/updating/RenderTreeBuilder.h:
* rendering/updating/RenderTreeBuilderFirstLetter.cpp:
(WebCore::RenderTreeBuilder::FirstLetter::updateStyle):
(WebCore::RenderTreeBuilder::FirstLetter::createRenderers):
* rendering/updating/RenderTreeBuilderList.cpp:
(WebCore::RenderTreeBuilder::List::updateItemMarker):
* rendering/updating/RenderTreeBuilderMultiColumn.cpp:
(WebCore::RenderTreeBuilder::MultiColumn::createFragmentedFlow):
(WebCore::RenderTreeBuilder::MultiColumn::destroyFragmentedFlow):
* rendering/updating/RenderTreeBuilderRuby.cpp:
(WebCore::RenderTreeBuilder::Ruby::insertChild):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[WPE][GTK] Make RunLoop::TimerBase robust to its own deletion inside its source callback
mcatanzaro@igalia.com [Thu, 1 Feb 2018 19:41:11 +0000 (19:41 +0000)]
[WPE][GTK] Make RunLoop::TimerBase robust to its own deletion inside its source callback
https://bugs.webkit.org/show_bug.cgi?id=182271

Reviewed by Carlos Garcia Campos.

RunLoopTimer::fired executes the user's callback, which could delete the RunLoopTimer
itself. But the source callback is not prepared to handle this case. We can detect it
easily, because TimerBase's destructor will call g_source_destroy(), which confusingly
removes the GSource from its GMainContext without actually destroying the GSource. Then we
can check if the GSource is still attached using g_source_is_destroyed().

* wtf/glib/RunLoopGLib.cpp:
(WTF::RunLoop::TimerBase::TimerBase):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoFollowup test expectation adjustment from r227947.
jlewis3@apple.com [Thu, 1 Feb 2018 19:07:48 +0000 (19:07 +0000)]
Followup test expectation adjustment from r227947.

Unreviewed test expectations.

* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION (r219342): Scaled HTML widget is not responding to a clicks outside the...
dino@apple.com [Thu, 1 Feb 2018 18:59:38 +0000 (18:59 +0000)]
REGRESSION (r219342): Scaled HTML widget is not responding to a clicks outside the body
https://bugs.webkit.org/show_bug.cgi?id=182394
<rdar://problem/34840816>

Reviewed by Simon Fraser.

Source/WebCore:

If a scale < 1 is applied to the page, then the visual viewport will be bigger
than the layout viewport. Our hit testing code would then ignore any hits
that were outside the layout viewport.

The fix is to only apply a hit testing clip if the page is scaling up, not down.

Update the existing fast/dom/elementFromPoint-scaled-scrolled.html test.

* page/FrameView.cpp:
(WebCore::FrameView::layoutViewportToAbsoluteRect const): Deleted. This helper is
no longer used, and it would have probably been more confusing to have it accept
a flag to ignore the scale if it is less than 1.
* page/FrameView.h:
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::hitTest): No need to take the layout rect, remove the origin,
and pass it to a helper that added the origin back. The only thing the helper was
doing for us was applying a scale factor, which we only want to do if it was
scaling up.

LayoutTests:

Add a test for a scaled down page.

* fast/dom/elementFromPoint-scaled-scrolled-expected.txt:
* fast/dom/elementFromPoint-scaled-scrolled.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed, rolling out r227958 and r227972.
jlewis3@apple.com [Thu, 1 Feb 2018 18:31:06 +0000 (18:31 +0000)]
Unreviewed, rolling out r227958 and r227972.
https://bugs.webkit.org/show_bug.cgi?id=182393

This caused a consistent crash on macOS. (Requested by
mlewis13 on #webkit).

Reverted changesets:

"Update imagebitmap tests."
https://bugs.webkit.org/show_bug.cgi?id=182335
https://trac.webkit.org/changeset/227958

"[WPE] Update test expectations for r227958"
https://bugs.webkit.org/show_bug.cgi?id=182391
https://trac.webkit.org/changeset/227972

Patch by Commit Queue <commit-queue@webkit.org> on 2018-02-01

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[WPE] Update test expectations for r227958
Ms2ger@igalia.com [Thu, 1 Feb 2018 18:01:28 +0000 (18:01 +0000)]
[WPE] Update test expectations for r227958
https://bugs.webkit.org/show_bug.cgi?id=182391

Unreviewed test gardening.

* platform/wpe/imported/w3c/web-platform-tests/2dcontext/imagebitmap/createImageBitmap-drawImage-expected.txt: Removed.
* platform/wpe/imported/w3c/web-platform-tests/2dcontext/imagebitmap/createImageBitmap-invalid-args-expected.txt:
* platform/wpe/imported/w3c/web-platform-tests/2dcontext/imagebitmap/createImageBitmap-origin.sub-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoBubbles intermittently disappear from bot watchers dashboard
aakash_jain@apple.com [Thu, 1 Feb 2018 17:30:26 +0000 (17:30 +0000)]
Bubbles intermittently disappear from bot watchers dashboard
https://bugs.webkit.org/show_bug.cgi?id=182085

Reviewed by Alexey Proskuryakov.

* BuildSlaveSupport/build.webkit.org-config/public_html/dashboard/Scripts/BuildbotIteration.js:
(BuildbotIteration.prototype.failureLogURL): failureLogURL doesn't depend on step in Buildbot 0.9.
Return the Build Page URL for Buildbot Buildbot 0.9. Earlier code was also doing the same using a
confusing if (!this._firstFailedStep.logs) statement which is always false for Buildbot 0.9.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago-Wreturn-type warning in DFGObjectAllocationSinkingPhase.cpp
mcatanzaro@igalia.com [Thu, 1 Feb 2018 17:27:06 +0000 (17:27 +0000)]
-Wreturn-type warning in DFGObjectAllocationSinkingPhase.cpp
https://bugs.webkit.org/show_bug.cgi?id=182389

Reviewed by Yusuke Suzuki.

Fix the warning.

As a bonus, remove a couple unreachable breaks for good measure.

* dfg/DFGObjectAllocationSinkingPhase.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoStructured cloning a Symbol should throw
utatane.tea@gmail.com [Thu, 1 Feb 2018 17:22:35 +0000 (17:22 +0000)]
Structured cloning a Symbol should throw
https://bugs.webkit.org/show_bug.cgi?id=182380

Reviewed by Darin Adler.

Source/WebCore:

Test: js/dom/post-message-symbol.html

* bindings/js/SerializedScriptValue.cpp:
(WebCore::CloneSerializer::dumpIfTerminal):
Structured cloning should throw "DataCloneError" DOMException if it finds Symbol.
Remove unused `isNumber()` case. It is done in dumpImmediate.
(WebCore::CloneSerializer::serializeUndefined): Deleted.
(WebCore::CloneSerializer::serializeBoolean): Deleted.
(WebCore::CloneSerializer::serializeNumber): Deleted.
Remove unused functions.

LayoutTests:

* js/dom/post-message-symbol-expected.txt: Added.
* js/dom/post-message-symbol.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[Modern Media Controls] Turn media/modern-media-controls/macos-fullscreen-media-contr...
commit-queue@webkit.org [Thu, 1 Feb 2018 17:20:24 +0000 (17:20 +0000)]
[Modern Media Controls] Turn media/modern-media-controls/macos-fullscreen-media-controls back on
https://bugs.webkit.org/show_bug.cgi?id=182385

Patch by Antoine Quint <graouts@apple.com> on 2018-02-01
Reviewed by Eric Carlson.

Update tests to match the modern-media-controls design and expectations.

* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-containers-styles-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-containers-styles.html:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-styles-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-styles.html:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-constructor-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-constructor.html:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-controls-bar-styles-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-controls-bar-styles.html:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-right-container-margin.html:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-time-control-styles-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-time-control-styles.html:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-volume-styles-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-volume-styles.html:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoWe no longer need to queue service worker jobs until the connection to the service...
cdumez@apple.com [Thu, 1 Feb 2018 17:10:18 +0000 (17:10 +0000)]
We no longer need to queue service worker jobs until the connection to the service worker process has been established
https://bugs.webkit.org/show_bug.cgi?id=182375

Reviewed by Youenn Fablet.

We no longer need to queue service worker jobs until the connection to the service worker process
has been established. We initially did this to work around the fact that registrations restored
from disk would not have an active worker until the service worker process had been established.
However, this issue has been fixed in r227696.

This is basically a revert of r227220, which is no longer needed after r227696.

No new tests, initial fix was covered by an API test that still passes.

* workers/service/server/SWServer.cpp:
(WebCore::SWServer::clearAll):
(WebCore::SWServer::clear):
(WebCore::SWServer::scheduleJob):
(WebCore::SWServer::serverToContextConnectionCreated):
* workers/service/server/SWServer.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[GTK] Shift + mouse scroll should scroll horizontally
carlosgc@webkit.org [Thu, 1 Feb 2018 16:04:09 +0000 (16:04 +0000)]
[GTK] Shift + mouse scroll should scroll horizontally
https://bugs.webkit.org/show_bug.cgi?id=181629

Reviewed by Michael Catanzaro.

Source/WebCore:

We currently turn vertical scroll into horizontal when scrolling over the horizontal scrollbar. When Shift key is
pressed, we still want to scroll in the scrollbar direction when scrolling over a scrollbar, so we need to swap
directions in both scrollbars depending on whther the Shift key is pressed or not.

* page/EventHandler.cpp:
(WebCore::EventHandler::shouldSwapScrollDirection const): Renamed.
(WebCore::EventHandler::handleWheelEvent): Use the new name.
(WebCore::EventHandler::shouldTurnVerticalTicksIntoHorizontal const): Deleted.
* page/EventHandler.h:
* platform/PlatformWheelEvent.h:
(WebCore::PlatformWheelEvent::copySwappingDirection const): Swap the direction of the event.
(WebCore::PlatformWheelEvent::copyTurningVerticalTicksIntoHorizontalTicks const): Deleted.
* platform/glib/EventHandlerGLib.cpp:
(WebCore::EventHandler::shouldSwapScrollDirection const): Take into account whether the Shift key is present.
(WebCore::EventHandler::shouldTurnVerticalTicksIntoHorizontal const): Deleted.

Source/WebKit:

Swap scroll direction when Shift is pressed for consistency with GtkScrolledWindow.

* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseScrollEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[GTK] Problem with Washington Post images
carlosgc@webkit.org [Thu, 1 Feb 2018 15:55:14 +0000 (15:55 +0000)]
[GTK] Problem with Washington Post images
https://bugs.webkit.org/show_bug.cgi?id=181421

Reviewed by Carlos Alberto Lopez Perez.

Source/WebCore:

This is because Washington Post is using the user agent to decide the image formats it serves. In the case of
chromium the images are served as webp, for firefox jpeg is used and in our case it's assuming we are safari and
it's providing jp2 images that we don't support. Add a user agent quirk to pretend to be chromium for
washingtonpost.com.

* platform/UserAgentQuirks.cpp:
(WebCore::urlRequiresChromeBrowser):

Tools:

Add test case for the new user agent quirk.

* TestWebKitAPI/Tests/WebCore/UserAgentQuirks.cpp:
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[SOUP] Ensure domain is valid when converting a WebCore Cookie to Soup
carlosgc@webkit.org [Thu, 1 Feb 2018 15:53:42 +0000 (15:53 +0000)]
[SOUP] Ensure domain is valid when converting a WebCore Cookie to Soup
https://bugs.webkit.org/show_bug.cgi?id=182328

Reviewed by Michael Catanzaro.

soup_cookie_parse() adds the initial '.' to the domain if missing before creating the SoupCookie, but
soup_cookie_new() allows for domain to be a hostname that needs to match exactly. When converting a WebCore
Cookie into a SoupCookie we always want the domain to be considered as such and not as a hostname, so we need to
prepend the '.' if missing.

Fixes: imported/w3c/webdriver/tests/cookies/add_cookie.py::test_add_domain_cookie

* platform/network/soup/CookieSoup.cpp:
(WebCore::Cookie::toSoupCookie const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[RenderTreeBuilder] Move RenderTableRow::collapseAndDestroyAnonymousSiblingRows to...
zalan@apple.com [Thu, 1 Feb 2018 15:52:33 +0000 (15:52 +0000)]
[RenderTreeBuilder] Move RenderTableRow::collapseAndDestroyAnonymousSiblingRows to RenderTreeBuilder
https://bugs.webkit.org/show_bug.cgi?id=182374
<rdar://problem/37102005>

Reviewed by Antti Koivisto.

It's only called from RenderTreeBuilder.

No change in functionality.

* rendering/RenderTableRow.cpp:
(WebCore::RenderTableRow::collapseAndDestroyAnonymousSiblingRows): Deleted.
* rendering/RenderTableRow.h:
* rendering/updating/RenderTreeBuilder.cpp:
(WebCore::RenderTreeBuilder::removeFromParentAndDestroyCleaningUpAnonymousWrappers):
* rendering/updating/RenderTreeBuilderTable.cpp:
(WebCore::RenderTreeBuilder::Table::collapseAndDestroyAnonymousSiblingRows):
* rendering/updating/RenderTreeBuilderTable.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoRewrite fast/events/scroll-in-scaled-page-with-overflow-hidden.html to conform with...
fred.wang@free.fr [Thu, 1 Feb 2018 14:03:45 +0000 (14:03 +0000)]
Rewrite fast/events/scroll-in-scaled-page-with-overflow-hidden.html to conform with CSSOM View
https://bugs.webkit.org/show_bug.cgi?id=182287

Patch by Frederic Wang <fwang@igalia.com> on 2018-02-01
Reviewed by Antonio Gomes.

Per the CSSOM View specification, it is wrong to use document.body.scrollTop to retrieve the
vertical offset of the viewport in this test, because the body is potentially scrollable.
This commit relies on window.scrollY instead so that the test still works after bug 5991.

* fast/events/scroll-in-scaled-page-with-overflow-hidden-expected.txt:
* fast/events/scroll-in-scaled-page-with-overflow-hidden.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION(r227893): fast/events/touch/touch-stale-node-crash.html and other tests...
carlosgc@webkit.org [Thu, 1 Feb 2018 13:58:22 +0000 (13:58 +0000)]
REGRESSION(r227893): fast/events/touch/touch-stale-node-crash.html and other tests crash
https://bugs.webkit.org/show_bug.cgi?id=182350

Reviewed by Carlos Alberto Lopez Perez.

Ensure events synthesized from touch gestures have a valid window, screen and device.

* UIProcess/API/gtk/WebKitWebViewBase.cpp:
* UIProcess/gtk/GestureController.h: Add virtual destructor to GestureControllerclient.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[Modern Media Controls] Turn media/modern-media-controls/media-controls back on
commit-queue@webkit.org [Thu, 1 Feb 2018 13:38:01 +0000 (13:38 +0000)]
[Modern Media Controls] Turn media/modern-media-controls/media-controls back on
https://bugs.webkit.org/show_bug.cgi?id=182377

Patch by Antoine Quint <graouts@apple.com> on 2018-02-01
Reviewed by Dean Jackson.

Update tests to match the modern-media-controls design and expectations, moving 3 tests under
media/modern-media-controls/macos-inline-media-controls since they are testing functionality
that is now specific to inline media controls.

* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-bar-always-ltr-expected.txt: Renamed from LayoutTests/media/modern-media-controls/media-controls/media-controls-controls-bar-always-ltr-expected.txt.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-bar-always-ltr.html: Renamed from LayoutTests/media/modern-media-controls/media-controls/media-controls-controls-bar-always-ltr.html.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-placard-expected.txt: Added.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-placard.html: Added.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-shows-start-button-expected.txt: Renamed from LayoutTests/media/modern-media-controls/media-controls/media-controls-start-button-expected.txt.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-shows-start-button.html: Renamed from LayoutTests/media/modern-media-controls/media-controls/media-controls-start-button.html.
* media/modern-media-controls/media-controls/media-controls-appear-when-focus-expected.txt:
* media/modern-media-controls/media-controls/media-controls-appear-when-focus.html:
* media/modern-media-controls/media-controls/media-controls-constructor-expected.txt:
* media/modern-media-controls/media-controls/media-controls-constructor.html:
* media/modern-media-controls/media-controls/media-controls-placard-compressed-metrics-expected.txt:
* media/modern-media-controls/media-controls/media-controls-placard-compressed-metrics.html:
* media/modern-media-controls/media-controls/media-controls-placard-expected.txt: Removed.
* media/modern-media-controls/media-controls/media-controls-placard.html: Removed.
* platform/ios/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoQueue a microtask when a waitUntil() promise is settled
cdumez@apple.com [Thu, 1 Feb 2018 11:50:47 +0000 (11:50 +0000)]
Queue a microtask when a waitUntil() promise is settled
https://bugs.webkit.org/show_bug.cgi?id=182372
<rdar://problem/37101019>

Reviewed by Mark Lam.

LayoutTests/imported/w3c:

Reaseline WPT test now that all checks are passing.

* web-platform-tests/service-workers/service-worker/extendable-event-async-waituntil.https-expected.txt:

Source/JavaScriptCore:

Export a symbol so it can be used in WebCore.

* runtime/JSGlobalObject.h:

Source/WebCore:

Queue a microtask when a waitUntil() promise is settled, as per:
- https://w3c.github.io/ServiceWorker/#dom-extendableevent-waituntil (step 5)

Otherwise, we decrement m_pendingPromiseCount too quickly and it may cause
following calls to waitUntil() to throw when they shouldn't.

No new tests, rebaselined existing test.

* workers/service/ExtendableEvent.cpp:
(WebCore::ExtendableEvent::addExtendLifetimePromise):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUpdate imagebitmap tests.
Ms2ger@igalia.com [Thu, 1 Feb 2018 11:16:09 +0000 (11:16 +0000)]
Update imagebitmap tests.
https://bugs.webkit.org/show_bug.cgi?id=182335

Unreviewed test gardening.

LayoutTests/imported/w3c:

* web-platform-tests/2dcontext/imagebitmap/common.sub.js: Renamed from LayoutTests/imported/w3c/web-platform-tests/2dcontext/imagebitmap/common.js.
* web-platform-tests/2dcontext/imagebitmap/createImageBitmap-drawImage-expected.txt:
* web-platform-tests/2dcontext/imagebitmap/createImageBitmap-drawImage.html:
* web-platform-tests/2dcontext/imagebitmap/createImageBitmap-invalid-args-expected.txt:
* web-platform-tests/2dcontext/imagebitmap/createImageBitmap-invalid-args.html:
* web-platform-tests/2dcontext/imagebitmap/createImageBitmap-origin.sub-expected.txt: Added.
* web-platform-tests/2dcontext/imagebitmap/createImageBitmap-origin.sub.html: Added.
* web-platform-tests/2dcontext/imagebitmap/w3c-import.log:
* web-platform-tests/common/namespaces.js: Added.
* web-platform-tests/common/w3c-import.log:
* web-platform-tests/images/pattern.mp4: Added.
* web-platform-tests/images/pattern.svg: Added.
* web-platform-tests/images/w3c-import.log:

LayoutTests:

* platform/gtk/imported/w3c/web-platform-tests/2dcontext/imagebitmap/createImageBitmap-drawImage-expected.txt: Removed.
* platform/gtk/imported/w3c/web-platform-tests/2dcontext/imagebitmap/createImageBitmap-invalid-args-expected.txt:
* platform/gtk/imported/w3c/web-platform-tests/2dcontext/imagebitmap/createImageBitmap-origin.sub-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[GTK] MiniBrowser settings are just a list of names
carlosgc@webkit.org [Thu, 1 Feb 2018 09:44:39 +0000 (09:44 +0000)]
[GTK] MiniBrowser settings are just a list of names
https://bugs.webkit.org/show_bug.cgi?id=182005

Reviewed by Sergio Villar Senin.

The default size we are using is too small to show the values column, and the window is not resizable because
GObjectClass::constructed() is not called in the parent class.

* MiniBrowser/gtk/BrowserSettingsDialog.c:
(browser_settings_dialog_init): Use 600x400 as default size.
(browserSettingsDialogConstructed): Chain up to parent constructed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoInvalidate style for sibling combinators accurately on class change
antti@apple.com [Thu, 1 Feb 2018 09:01:01 +0000 (09:01 +0000)]
Invalidate style for sibling combinators accurately on class change
https://bugs.webkit.org/show_bug.cgi?id=182336

Reviewed by Zalan Bujtas.

Source/WebCore:

Use Style::Invalidator to invalidate only those elements that may be affected by a class
change for sibling combinators and nth pseudo classes.

* css/RuleFeature.cpp:

Add new AllSiblings MatchElement to use for nth pseudo classes with subselectors.

(WebCore::isSiblingOrSubject):

Add a helper.

(WebCore::RuleFeatureSet::computeNextMatchElement):
(WebCore::RuleFeatureSet::computeSubSelectorMatchElement):

Compute and propage MatchElement::AllSiblings.

* css/RuleFeature.h:
* dom/Node.cpp:
(WebCore::Node::updateAncestorsForStyleRecalc):

Don't need to test for childrenAffectedByPropertyBasedBackwardPositionalRules anymore (an oddly named bit for nth pseudo classes).

* style/StyleInvalidator.cpp:
(WebCore::Style::Invalidator::invalidateStyleWithMatchElement):

Invalidate only the potentially affected elements.
The old code would just unconditionally invalidate the current element. This would propagate to descedants of siblings via
affectedByPreviousSibling bits. That mechanism can be removed when everything has been switched to accurate invalidation.

LayoutTests:

Adapt to progressions.

* fast/css/direct-adjacent-style-update-optimization-expected.txt:
* fast/css/direct-adjacent-style-update-optimization.html:
* fast/css/indirect-adjacent-style-update-optimization-expected.txt:
* fast/css/indirect-adjacent-style-update-optimization.html:
* fast/css/nth-last-child-of-style-update-optimization.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed. Remove WebInspectorUI watchlist and remove myself from 2 watchlists more
carlosgc@webkit.org [Thu, 1 Feb 2018 08:52:43 +0000 (08:52 +0000)]
Unreviewed. Remove WebInspectorUI watchlist and remove myself from 2 watchlists more

Patch by Andres Gomez <agomez@igalia.com> on 2018-01-31

* Scripts/webkitpy/common/config/watchlist:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUse different debug red colors for different contexts
simon.fraser@apple.com [Thu, 1 Feb 2018 07:28:10 +0000 (07:28 +0000)]
Use different debug red colors for different contexts
https://bugs.webkit.org/show_bug.cgi?id=182362

Reviewed by Tim Horton.
Source/WebCore:

Pure red is used elsehwere in the system as a debug color indicator, so use different
shades of red for WebKit in the two places where we paint a reddish wash in debug builds,
so they are identifiable.

* page/FrameView.cpp:
(WebCore::FrameView::paintContents):

Source/WebKit:

Pure red is used elsehwere in the system as a debug color indicator, so use different
shades of red for WebKit in the two places where we paint a reddish wash in debug builds,
so they are identifiable.

* Shared/RemoteLayerTree/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::drawInContext):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[Modern Media Controls] Turn media/modern-media-controls/macos-inline-media-controls...
graouts@webkit.org [Thu, 1 Feb 2018 07:12:09 +0000 (07:12 +0000)]
[Modern Media Controls] Turn media/modern-media-controls/macos-inline-media-controls back on
https://bugs.webkit.org/show_bug.cgi?id=182338

Reviewed by Eric Carlson.

Update tests to match the modern-media-controls designs and expectations. Certain tests didn't make any sense anymore
and were removed, for instance tests related to the compact mode.

* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-audio-background-expected.txt: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-audio-background.html: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-buttons-styles-expected.txt: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-buttons-styles.html: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-compact-buttons-styles-expected.txt: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-compact-buttons-styles.html: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-compact-controls-bar-styles-expected.txt: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-compact-controls-bar-styles.html: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-compact-expected.txt: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-compact.html: Removed.
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-controls-bar-styles-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-controls-bar-styles.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-layout-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-layout.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-resize-with-hidden-controls-bar-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-resize-with-hidden-controls-bar.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-start-button-style-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-start-button-style.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-status-label-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-status-label.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-time-control-styles-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-time-control-styles.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-volume-slider-visibility-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-volume-slider-visibility.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-volume-styles-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-volume-styles.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-dropping-controls-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-dropping-controls.html:
* platform/ios/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[CMake] Make JavaScriptCore headers copies
don.olmstead@sony.com [Thu, 1 Feb 2018 06:50:23 +0000 (06:50 +0000)]
[CMake] Make JavaScriptCore headers copies
https://bugs.webkit.org/show_bug.cgi?id=182303

Reviewed by Alex Christensen.

Source/JavaScriptCore:

* CMakeLists.txt:
* PlatformGTK.cmake:
* PlatformJSCOnly.cmake:
* PlatformMac.cmake:
* PlatformWPE.cmake:
* PlatformWin.cmake:
* shell/CMakeLists.txt:
* shell/PlatformWin.cmake:

Source/WebCore:

No new tests. No change in behavior.

* CMakeLists.txt:
* PlatformMac.cmake:
* PlatformWPE.cmake:
* PlatformWin.cmake:
* PlatformWinCairo.cmake:

Source/WebKit:

* CMakeLists.txt:
* Scripts/generate-forwarding-headers.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoReplace tryLargeMemalignVirtual with tryLargeZeroedMemalignVirtual and use it to...
sbarati@apple.com [Thu, 1 Feb 2018 05:36:40 +0000 (05:36 +0000)]
Replace tryLargeMemalignVirtual with tryLargeZeroedMemalignVirtual and use it to allocate large zeroed memory in Wasm
https://bugs.webkit.org/show_bug.cgi?id=182064
<rdar://problem/36840132>

Reviewed by Geoffrey Garen.

Source/bmalloc:

This patch replaces the tryLargeMemalignVirtual API with tryLargeZeroedMemalignVirtual.
By doing that, we're able to remove the AllocationKind enum. To zero the memory,
tryLargeZeroedMemalignVirtual uses mmap(... MAP_ANON ...) over previously mmapped
memory. This both purges the any resident memory for the virtual range and ensures
that the pages in the range are zeroed. Most OSs should implement this by taking a
page fault and zero filling on first access. Therefore, this API is returning pages
that will result in page faults on first access. Hence, the name 'virtual' in the API.
This API differs from the old API in that users of it need not call madvise themselves.
The memory is ready to go.

* bmalloc.xcodeproj/project.pbxproj:
* bmalloc/AllocationKind.h: Removed.
* bmalloc/DebugHeap.cpp:
(bmalloc::DebugHeap::memalignLarge):
(bmalloc::DebugHeap::freeLarge):
* bmalloc/DebugHeap.h:
* bmalloc/Heap.cpp:
(bmalloc::Heap::splitAndAllocate):
(bmalloc::Heap::tryAllocateLarge):
(bmalloc::Heap::allocateLarge):
(bmalloc::Heap::shrinkLarge):
(bmalloc::Heap::deallocateLarge):
* bmalloc/Heap.h:
* bmalloc/IsoPage.cpp:
(bmalloc::IsoPageBase::allocatePageMemory):
* bmalloc/VMAllocate.h:
(bmalloc::vmZeroAndPurge):
* bmalloc/VMHeap.cpp:
(bmalloc::VMHeap::tryAllocateLargeChunk):
* bmalloc/VMHeap.h:
* bmalloc/bmalloc.cpp:
(bmalloc::api::tryLargeZeroedMemalignVirtual):
(bmalloc::api::freeLargeVirtual):
(bmalloc::api::tryLargeMemalignVirtual): Deleted.
* bmalloc/bmalloc.h:

Source/JavaScriptCore:

This patch switches WebAssembly Memory to always use bmalloc's
zeroed virtual allocation API. This makes it so that we don't
dirty the memory to zero it. It's a huge compile time speedup
on WasmBench on iOS.

* wasm/WasmMemory.cpp:
(JSC::Wasm::Memory::create):
(JSC::Wasm::Memory::~Memory):
(JSC::Wasm::Memory::addressIsInActiveFastMemory):
(JSC::Wasm::Memory::grow):
(JSC::Wasm::commitZeroPages): Deleted.

Source/WTF:

* wtf/Gigacage.cpp:
(Gigacage::tryAllocateZeroedVirtualPages):
(Gigacage::freeVirtualPages):
(Gigacage::tryAllocateVirtualPages): Deleted.
* wtf/Gigacage.h:
* wtf/OSAllocator.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoLayout Test fast/events/beforeunload-dom-manipulation-crash.html is crashing
pvollan@apple.com [Thu, 1 Feb 2018 04:18:38 +0000 (04:18 +0000)]
Layout Test fast/events/beforeunload-dom-manipulation-crash.html is crashing
https://bugs.webkit.org/show_bug.cgi?id=181204
<rdar://problem/36256274>

Reviewed by Ryosuke Niwa.

Source/WebCore:

When a frame element is moved in the DOM tree during the execution of a beforeunload handler,
the frame will be detached when removed from its previous position in the DOM tree. When being
detached, an attempt will also be made to stop the load by calling FrameLoader::stopAllLoaders().
However, this method will return early when executed in a beforeunload handler, since navigation
is not allowed then. The end result is a detached frame which will continue to load, and hitting
asserts in DocumentLoader::dataReceived(), and DocumentLoader::notifyFinished(). It should be
possible to stop a frame load, even when executing a beforeunload handler.

No new tests. Covered by the existing test fast/events/beforeunload-dom-manipulation-crash.html.

* history/PageCache.cpp:
(WebCore::PageCache::addIfCacheable): Fix a failing API test by allowing scripts to be executed
under the PageCache::prune method.
* loader/FrameLoader.cpp:
(WebCore::FrameLoader::isStopLoadingAllowed const):
(WebCore::FrameLoader::stopAllLoaders):
* loader/FrameLoader.h:
* svg/graphics/SVGImage.cpp:
(WebCore::SVGImage::~SVGImage): Disable scripts disallowed assertions in this scope, since it is
safe in this context.

Tools:

Implement 'testRunner.forceImmediateCompletion()' for WK1.

* DumpRenderTree/TestRunner.cpp:
(forceImmediateCompletionCallback):
(TestRunner::staticFunctions):

LayoutTests:

* fast/events/beforeunload-dom-manipulation-crash.html: Make it clear that the
frame element is a child of the 'del' element.
* fast/events/beforeunload-dom-manipulation-crash-expected.html:
* platform/mac-wk1/TestExpectations: Unskip test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227948 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoinline-block baseline not computed correctly for vertical-lr
jfernandez@igalia.com [Thu, 1 Feb 2018 01:56:52 +0000 (01:56 +0000)]
inline-block baseline not computed correctly for vertical-lr
https://bugs.webkit.org/show_bug.cgi?id=170176

Reviewed by Manuel Rego Casasnovas.

Source/WebCore:

When computing the baseline position of inline-block elements we use the
InlineFlow logicalTop and the FontMetrics ascent. The issue comes from
the fact that these units are incompatible. The logicalTop of a
vertical-lr element is offset to the left edge, while the ascent is the
distance from the right edge.

We need to either use logical value for the FontMetrics ascent so we can
compute the correctly the baselines of vertical-lr elements, or just using
the logicalBottom for these cases.

The approach based on a logicalAscent API for FontMetrics would require
a lot of work because inline-block logic assumes everything is vertical-rl
and at some point, flips the elements along the block-axis in case of
vertical-lr mode.

While it'd be desirable to get rid of this flipping logic, this patch tries
first the simpler approach of using logicalBottom, which aligns with the
currently implemented logic.

Tests: fast/inline-block/baseline-vertical-01.html
       fast/inline-block/baseline-vertical-02.html
       fast/inline-block/baseline-vertical-03.html
       fast/inline-block/baseline-vertical-04.html
       fast/inline-block/baseline-vertical-05.html
       fast/inline-block/baseline-vertical-06.html
       fast/inline-block/baseline-vertical-07.html
       fast/inline-block/baseline-vertical-08.html

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::placeBoxesInBlockDirection):
* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::inlineBlockBaseline const):

LayoutTests:

Tests to evaluate the baseline alignment in vertical modes.
Additionally, several tests were rebaselined.

Some of the new tests are marked as Failure for the ios-sim
platform because of pixel rounding errors in the absolute
positioned elements used as reference.

Finally, there are color differences in the border-styles-vertical-lr-expected.png
caused by changes in the gtk+ platform. Those differences were not noticeable until
now that the patch causes diffs in the expected.txt files.

* fast/inline-block/baseline-vertical-01-expected.html: Added.
* fast/inline-block/baseline-vertical-01.html: Added.
* fast/inline-block/baseline-vertical-02-expected.html: Added.
* fast/inline-block/baseline-vertical-02.html: Added.
* fast/inline-block/baseline-vertical-03-expected.html: Added.
* fast/inline-block/baseline-vertical-03.html: Added.
* fast/inline-block/baseline-vertical-04-expected.html: Added.
* fast/inline-block/baseline-vertical-04.html: Added.
* fast/inline-block/baseline-vertical-05-expected.html: Added.
* fast/inline-block/baseline-vertical-05.html: Added.
* fast/inline-block/baseline-vertical-06-expected.html: Added.
* fast/inline-block/baseline-vertical-06.html: Added.
* fast/inline-block/baseline-vertical-07-expected.html: Added.
* fast/inline-block/baseline-vertical-07.html: Added.
* fast/inline-block/baseline-vertical-08-expected.html: Added.
* fast/inline-block/baseline-vertical-08.html: Added.
* fast/text/emphasis-avoid-ruby-expected.png:
* fast/text/emphasis-avoid-ruby-expected.txt:
* fast/text/emphasis-overlap-expected.png:
* fast/text/emphasis-overlap-expected.txt:
* platform/gtk/fast/backgrounds/background-leakage-transforms-expected.png:
* platform/gtk/fast/backgrounds/background-leakage-transforms-expected.txt:
* platform/gtk/fast/writing-mode/border-styles-vertical-lr-expected.png:
* platform/gtk/fast/writing-mode/border-styles-vertical-lr-expected.txt:
* platform/ios-simulator/TestExpectations:
* platform/ios/fast/backgrounds/background-leakage-transforms-expected.png: Added.
* platform/ios/fast/backgrounds/background-leakage-transforms-expected.txt:
* platform/ios/fast/writing-mode/border-styles-vertical-lr-expected.png:
* platform/ios/fast/writing-mode/border-styles-vertical-lr-expected.txt:
* platform/mac/fast/backgrounds/background-leakage-transforms-expected.png:
* platform/mac/fast/backgrounds/background-leakage-transforms-expected.txt:
* platform/mac/fast/writing-mode/border-styles-vertical-lr-expected.txt:
* platform/win/fast/writing-mode/text-orientation-basic-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoMake run-perf-tests work with StyleBench and re-enable it
rniwa@webkit.org [Thu, 1 Feb 2018 01:42:34 +0000 (01:42 +0000)]
Make run-perf-tests work with StyleBench and re-enable it
https://bugs.webkit.org/show_bug.cgi?id=182322

Reviewed by Antti Koivisto.

This patch updates the test harness to report new metrics of StyleBench.

Replaced the total time by the geometric mean of the time spent in each suite, and added scores
as a metric in addition to time. Also, updated the PerfTestRunner harness and the results template
to support "Score" metric with "pt" as its unit.

* Skipped: Unskipped the test.
* StyleBench/resources/benchmark-report.js:
(window.benchmarkClient.willStartFirstIteration..createTest): Added the unit as an optional argument
to report the score in addition to time. Also report the geomean of time instead of the total time.
(window.benchmarkClient.willStartFirstIteration): Report the geomean of time.
(window.benchmarkClient.didRunSuites): Record the entire measuredValues instead of just tests.
(window.benchmarkClient.didFinishLastIteration): Report the scores.
* resources/results-template.html: Added "Score" as a metric and "pt" as its unit.
* resources/runner.js:
(PerfTestRunner..finish): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed GTK+ test gardening.
aboya@igalia.com [Thu, 1 Feb 2018 01:40:09 +0000 (01:40 +0000)]
Unreviewed GTK+ test gardening.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUpdate ReadMe.md
mcatanzaro@igalia.com [Thu, 1 Feb 2018 01:38:34 +0000 (01:38 +0000)]
Update ReadMe.md
https://bugs.webkit.org/show_bug.cgi?id=182314

Reviewed by Alex Christensen.

Add download link for Epiphany Technology Preview. Improve build instructions for GTK. Add
build instructions for WPE. Add instructions to run MiniBrowser on Linux. Miscellaneous
typographical adjustments.

* ReadMe.md:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed, rolling out r227942.
achristensen@apple.com [Thu, 1 Feb 2018 00:58:52 +0000 (00:58 +0000)]
Unreviewed, rolling out r227942.

r227875 should not have been rolled out.

Reverted changeset:

"Unreviewed, rolling out r227875."
https://bugs.webkit.org/show_bug.cgi?id=182357
https://trac.webkit.org/changeset/227942

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed, rolling out r227875.
mcatanzaro@igalia.com [Thu, 1 Feb 2018 00:08:01 +0000 (00:08 +0000)]
Unreviewed, rolling out r227875.
https://bugs.webkit.org/show_bug.cgi?id=182357

Missing cross-platform TestController implementation

Reverted changeset:

"Add callbacks to testRunner.statisticsSetShouldPartitionCookiesForHost() and testRunner.statisticsUpdateCookiePartitioning()"
https://bugs.webkit.org/show_bug.cgi?id=181958
https://trac.webkit.org/changeset/227875

Source/WebKit:

* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsUpdateCookiePartitioning):
(WKWebsiteDataStoreSetStatisticsShouldPartitionCookiesForHost):
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _resourceLoadStatisticsUpdateCookiePartitioning]):
(-[WKWebsiteDataStore _resourceLoadStatisticsSetShouldPartitionCookies:forHost:]):
(-[WKWebsiteDataStore _resourceLoadStatisticsUpdateCookiePartitioning:]): Deleted.
(-[WKWebsiteDataStore _resourceLoadStatisticsSetShouldPartitionCookies:forHost:completionHandler:]): Deleted.
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):
(WebKit::WebResourceLoadStatisticsStore::logUserInteraction):
(WebKit::WebResourceLoadStatisticsStore::logNonRecentUserInteraction):
(WebKit::WebResourceLoadStatisticsStore::scheduleCookiePartitioningUpdate):
(WebKit::WebResourceLoadStatisticsStore::scheduleCookiePartitioningUpdateForDomains):
(WebKit::WebResourceLoadStatisticsStore::scheduleClearPartitioningStateForDomains):
(WebKit::WebResourceLoadStatisticsStore::mergeWithDataFromDecoder):
(WebKit::WebResourceLoadStatisticsStore::clearInMemory):
(WebKit::WebResourceLoadStatisticsStore::updateCookiePartitioning):
(WebKit::WebResourceLoadStatisticsStore::updateCookiePartitioningForDomains):
(WebKit::WebResourceLoadStatisticsStore::clearPartitioningStateForDomains):
* UIProcess/WebResourceLoadStatisticsStore.h:

Tools:

* WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::didReceiveMessageToPage):
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::statisticsUpdateCookiePartitioning):
(WTR::TestRunner::statisticsSetShouldPartitionCookiesForHost):
(WTR::TestRunner::statisticsCallDidSetPartitionOrBlockCookiesForHostCallback): Deleted.
* WebKitTestRunner/InjectedBundle/TestRunner.h:
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didSetPartitionOrBlockCookiesForHost): Deleted.
* WebKitTestRunner/TestInvocation.h:
* WebKitTestRunner/cocoa/TestControllerCocoa.mm:
(WTR::TestController::statisticsUpdateCookiePartitioning):
(WTR::TestController::statisticsSetShouldPartitionCookiesForHost):

LayoutTests:

* http/tests/resourceLoadStatistics/add-blocking-to-redirect-expected.txt:
* http/tests/resourceLoadStatistics/add-blocking-to-redirect.html:
* http/tests/resourceLoadStatistics/add-partitioning-to-redirect-expected.txt:
* http/tests/resourceLoadStatistics/add-partitioning-to-redirect.html:
* http/tests/resourceLoadStatistics/non-prevalent-resources-can-access-cookies-in-a-third-party-context.html:
* http/tests/resourceLoadStatistics/partitioned-and-unpartitioned-cookie-deletion.html:
* http/tests/resourceLoadStatistics/partitioned-and-unpartitioned-cookie-with-partitioning-timeout.html:
* http/tests/resourceLoadStatistics/partitioned-cookies-with-and-without-user-interaction-expected.txt:
* http/tests/resourceLoadStatistics/partitioned-cookies-with-and-without-user-interaction.html:
* http/tests/resourceLoadStatistics/remove-blocking-in-redirect-expected.txt:
* http/tests/resourceLoadStatistics/remove-blocking-in-redirect.html:
* http/tests/resourceLoadStatistics/remove-partitioning-in-redirect-expected.txt:
* http/tests/resourceLoadStatistics/remove-partitioning-in-redirect.html:
* http/tests/resourceLoadStatistics/third-party-cookie-with-and-without-user-interaction.html:
* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoFollow-up to r227939.
bfulgham@apple.com [Thu, 1 Feb 2018 00:03:05 +0000 (00:03 +0000)]
Follow-up to r227939.
https://bugs.webkit.org/show_bug.cgi?id=182354
<rdar://problem/37046844>

Make sure the correct sandbox is used on iOS as well.

* StorageProcess/ios/StorageProcessIOS.mm:
(WebKit::StorageProcess::initializeSandbox):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoFix some ARM64_32 build failures.
mark.lam@apple.com [Thu, 1 Feb 2018 00:00:45 +0000 (00:00 +0000)]
Fix some ARM64_32 build failures.
https://bugs.webkit.org/show_bug.cgi?id=182356
<rdar://problem/37057690>

Reviewed by Michael Saboff.

* wtf/MathExtras.h:
(WTF::dynamicPoison):
* wtf/text/ASCIIFastPath.h:
(WTF::copyLCharsFromUCharSource):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION(r220094): com.apple.WebKit.Storage lost its sandbox
bfulgham@apple.com [Wed, 31 Jan 2018 23:56:11 +0000 (23:56 +0000)]
REGRESSION(r220094): com.apple.WebKit.Storage lost its sandbox
https://bugs.webkit.org/show_bug.cgi?id=182354
<rdar://problem/37046844>

Reviewed by Ryosuke Niwa.

The Database process was renamed to Storage, but it's sandbox was not updated.

* Configurations/WebKit.xcconfig: Update for sandbox rename.
* DerivedSources.make: Ditto.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Databases.sb: Removed.
* Resources/SandboxProfiles/ios/com.apple.WebKit.Storage.sb: Copied from Resources/SandboxProfiles/ios/com.apple.WebKit.Databases.sb.
* StorageProcess/mac/com.apple.WebKit.Databases.sb.in: Removed.
* StorageProcess/mac/com.apple.WebKit.Storage.sb.in: Copied from StorageProcess/mac/com.apple.WebKit.Databases.sb.in.
* WebKit.xcodeproj/project.pbxproj: Update for sandbox rename.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoShould chose the best match during 'route' if there are multiple matches.
dewei_zhu@apple.com [Wed, 31 Jan 2018 23:41:37 +0000 (23:41 +0000)]
Should chose the best match during 'route' if there are multiple matches.
https://bugs.webkit.org/show_bug.cgi?id=182326

Reviewed by Ryosuke Niwa.

r227749 made a change that 'analysisCategoryPage' will be added before 'analysisTaskPage'.
As route names for both pages starts with 'analysis', whichever added first will be chosen.
For a route like 'analysis/task/1'. As a result, 'analysisCategoryPage' will be chosen and
this is not expected behavior. Adding the logic on the cases when route name does not extact
match the route name, always choose the longest mathcing route name.

Also modernized the code of 'page-router.js' to use const & let instead of var.

Added a browser test to guard against this bug.

* browser-tests/index.html: Import 'page-router-tests.js'.
* browser-tests/page-router-tests.js: Added unit test to guard against this bug.
* public/v3/pages/page-router.js:
(PageRouter.prototype.route): Added logic to find best matching in the case of inexact match.
(PageRouter.prototype.pageDidOpen):
(PageRouter.prototype._updateURLState):
(PageRouter.prototype._serializeToHash):
(PageRouter.prototype._deserializeFromHash):
(PageRouter.prototype._serializeHashQueryValue):
(PageRouter.prototype._deserializeHashQueryValue):
(PageRouter.prototype._countOccurrences):
(PageRouter):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoBitmapImage::drawPattern() may not draw a complete frame even after all the data...
commit-queue@webkit.org [Wed, 31 Jan 2018 23:23:35 +0000 (23:23 +0000)]
BitmapImage::drawPattern() may not draw a complete frame even after all the data is received
https://bugs.webkit.org/show_bug.cgi?id=182277

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2018-01-31
Reviewed by Simon Fraser.

Source/WebCore:

BitmapImage::drawPattern() needs to destroy the incomplete decoded frame
before trying to draw it as a pattern.

Test: http/tests/images/draw-pattern-slow-load-large-image.html

* loader/cache/CachedImage.cpp:
(WebCore::CachedImage::updateBufferInternal): We need to disable CachedImage
data buffering for testing. This simulates slow network where intervals
between data chunks can last for seconds.
* loader/cache/CachedImage.h:
* platform/graphics/BitmapImage.cpp:
(WebCore::BitmapImage::drawPattern): Destroy the incomplete decoded frame
before drawing this frame as a pattern. We do not destroy incomplete decoded
frame once new data is received because it may be drawn by async image
drawing while waiting for the newer frame to finish decoding.
* testing/Internals.cpp:
(WebCore::Internals::setForceUpdateImageDataEnabledForTesting):
* testing/Internals.h:
* testing/Internals.idl:

LayoutTests:

* http/tests/images/draw-pattern-slow-load-large-image-expected.html: Added.
* http/tests/images/draw-pattern-slow-load-large-image.html: Added.
* http/tests/resources/load-and-stall.php: Add a new argument to allow
repeating the stall till the end of the file. This simulates real slow
network where sending data and stalling happens constantly and not only
a single time.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION (r227340): Layout Test fast/workers/worker-cloneport.html is a flaky failure
cdumez@apple.com [Wed, 31 Jan 2018 23:22:31 +0000 (23:22 +0000)]
REGRESSION (r227340): Layout Test fast/workers/worker-cloneport.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=182007
<rdar://problem/37005504>

Reviewed by Ryosuke Niwa.

Increase the test's timeout. The test sometimes takes more than 1 second to run and would
log a FAIL line, even though the test eventually succeeds.

* fast/workers/worker-cloneport.html:
* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoAdd a release assertion to ensure timers are deleted in the right thread
rniwa@webkit.org [Wed, 31 Jan 2018 23:08:44 +0000 (23:08 +0000)]
Add a release assertion to ensure timers are deleted in the right thread
https://bugs.webkit.org/show_bug.cgi?id=182351

Reviewed by David Kilzer.

Added a relese assertion in ~TimerBase that the current thread is the one in which the timer was created.

We use canAccessThreadLocalDataForThread for this purpose since the condition is more complicated
when WebThread is being used.

* platform/Timer.cpp:
(WebCore::TimerBase::~TimerBase): Added the assertion.
* platform/Timer.h:
(WebCore::TimerBase::m_thread): Always store the current thread in a timer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION(r227223): http/tests/resourceLoadStatistics/clear-in-memory-and-persistent...
mcatanzaro@igalia.com [Wed, 31 Jan 2018 22:43:29 +0000 (22:43 +0000)]
REGRESSION(r227223): http/tests/resourceLoadStatistics/clear-in-memory-and-persistent-store-one-hour.html, http/tests/resourceLoadStatistics/grandfathering.html timing out on GTK, WPE
https://bugs.webkit.org/show_bug.cgi?id=182222

Reviewed by Alex Christensen.

Source/WebKit:

Add callbacks to notify when resource load statistics deletion is complete.

* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsClearInMemoryAndPersistentStore):
(WKWebsiteDataStoreStatisticsClearInMemoryAndPersistentStoreModifiedSinceHours):
* UIProcess/API/C/WKWebsiteDataStoreRef.h:

Tools:

Use the new API to notify when resource load statistics deletion is complete.

* WebKitTestRunner/TestController.cpp:
(WTR::TestController::statisticsClearThroughWebsiteDataRemovalCallback):
(WTR::TestController::statisticsClearInMemoryAndPersistentStore):
(WTR::TestController::statisticsClearInMemoryAndPersistentStoreModifiedSinceHours):

LayoutTests:

* platform/gtk/TestExpectations:
* platform/wpe/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoChange SWServer::claim to set the iterator value as soon as possible
commit-queue@webkit.org [Wed, 31 Jan 2018 22:40:10 +0000 (22:40 +0000)]
Change SWServer::claim to set the iterator value as soon as possible
https://bugs.webkit.org/show_bug.cgi?id=182337

Patch by Youenn Fablet <youenn@apple.com> on 2018-01-31
Reviewed by Chris Dumez.

No change of behavior, this is a "let's go safe way" fix.

* workers/service/server/SWServer.cpp:
(WebCore::SWServer::claim):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoVersioning.
jmarcell@apple.com [Wed, 31 Jan 2018 22:18:15 +0000 (22:18 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoBuild fix for CLoop after r227874.
mark.lam@apple.com [Wed, 31 Jan 2018 22:15:55 +0000 (22:15 +0000)]
Build fix for CLoop after r227874.
https://bugs.webkit.org/show_bug.cgi?id=182155
<rdar://problem/36286266>

Not reviewed.

Just needed support for lea of a LabelReference in cloop.rb (just like those
added for arm64.rb and x86.rb).

* offlineasm/cloop.rb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[RenderTreeBuilder] Move RenderObject::removeFromParentAndDestroyCleaningUpAnonymousW...
zalan@apple.com [Wed, 31 Jan 2018 22:00:16 +0000 (22:00 +0000)]
[RenderTreeBuilder] Move RenderObject::removeFromParentAndDestroyCleaningUpAnonymousWrappers to RenderTreeBuilder.
https://bugs.webkit.org/show_bug.cgi?id=182348
<rdar://problem/37082837>

Reviewed by Antti Koivisto.

All the callsites are in the RenderTreeUpdater by now.

No change in functionality.

* rendering/RenderObject.cpp:
(WebCore::isAnonymousAndSafeToDelete): Deleted.
(WebCore::findDestroyRootIncludingAnonymous): Deleted.
(WebCore::RenderObject::removeFromParentAndDestroyCleaningUpAnonymousWrappers): Deleted.
* rendering/RenderObject.h:
* rendering/updating/RenderTreeBuilder.cpp:
(WebCore::isAnonymousAndSafeToDelete):
(WebCore::findDestroyRootIncludingAnonymous):
(WebCore::RenderTreeBuilder::removeFromParentAndDestroyCleaningUpAnonymousWrappers):
* rendering/updating/RenderTreeBuilder.h:
* rendering/updating/RenderTreeUpdater.cpp:
(WebCore::RenderTreeUpdater::tearDownRenderers):
(WebCore::RenderTreeUpdater::tearDownTextRenderer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoOccasional null deref under WebPageProxy::updateBackingStoreDiscardableState()
timothy_horton@apple.com [Wed, 31 Jan 2018 21:39:49 +0000 (21:39 +0000)]
Occasional null deref under WebPageProxy::updateBackingStoreDiscardableState()
https://bugs.webkit.org/show_bug.cgi?id=182349
<rdar://problem/27822258>

Reviewed by Simon Fraser.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::updateBackingStoreDiscardableState):
m_drawingArea can be null during process launch and relaunch.

Cocoa ports don't use the result of setBackingStoreIsDiscardable(),
but they do run this code to keep it up to date.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoRelease assertion in Performance::resourceTimingBufferFullTimerFired when the resourc...
rniwa@webkit.org [Wed, 31 Jan 2018 21:07:36 +0000 (21:07 +0000)]
Release assertion in Performance::resourceTimingBufferFullTimerFired when the resource timing buffer is shrunk
https://bugs.webkit.org/show_bug.cgi?id=182319
<rdar://problem/36904312>

Reviewed by Chris Dumez.

Source/WebCore:

The crash was caused by a wrong release assertion. Handle author scripts shrinking the resource timing buffer
while resourcetimingbufferfull event is being dispatched.

Also fixed a bug that a superflous resourcetimingbufferfull event will be fired when new resource timing entries
are queued while resourcetimingbufferfull event is being dispatched.

Test: http/tests/performance/performance-resource-timing-resourcetimingbufferfull-queue-resource-entry.html
      http/tests/performance/performance-resource-timing-resourcetimingbufferfull-shrinking-buffer-crash.html

* page/Performance.cpp:
(WebCore::Performance::resourceTimingBufferFullTimerFired):

LayoutTests:

Added regression tests for shrinking the resoruce timing buffer and queuing a new resource timing entry while
resourcetimingbufferfull event is being dispatched.

* http/tests/performance/performance-resource-timing-resourcetimingbufferfull-queue-resource-entry-expected.txt: Added.
* http/tests/performance/performance-resource-timing-resourcetimingbufferfull-queue-resource-entry.html: Added.
* http/tests/performance/performance-resource-timing-resourcetimingbufferfull-shrinking-buffer-crash-expected.txt: Added.
* http/tests/performance/performance-resource-timing-resourcetimingbufferfull-shrinking-buffer-crash.html: Added.

* http/tests/performance/performance-resource-timing-resourcetimingbufferfull-shrinking-buffer-crash-expected.txt: Added.
* http/tests/performance/performance-resource-timing-resourcetimingbufferfull-shrinking-buffer-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoAdd the support for running StyleBench to run-benchmark
rniwa@webkit.org [Wed, 31 Jan 2018 20:58:20 +0000 (20:58 +0000)]
Add the support for running StyleBench to run-benchmark
https://bugs.webkit.org/show_bug.cgi?id=182323

Reviewed by Antti Koivisto.

Added the plan file and the patch needed to run StyleBench using run-benchmark as of r227863.

* Scripts/webkitpy/benchmark_runner/data/patches/webserver/StyleBench.patch: Added.
* Scripts/webkitpy/benchmark_runner/data/plans/stylebench.plan: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[GTK] Move gir and typelib generation into ENABLE_INTROSPECTION conditional
mcatanzaro@igalia.com [Wed, 31 Jan 2018 20:40:59 +0000 (20:40 +0000)]
[GTK] Move gir and typelib generation into ENABLE_INTROSPECTION conditional
https://bugs.webkit.org/show_bug.cgi?id=182308

Reviewed by Carlos Garcia Campos.

* PlatformGTK.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoAdjusted expectations for fast/forms/searchfield-heights.html.
jlewis3@apple.com [Wed, 31 Jan 2018 19:37:34 +0000 (19:37 +0000)]
Adjusted expectations for fast/forms/searchfield-heights.html.
https://bugs.webkit.org/show_bug.cgi?id=181831

Unreviewed test gardening.

* platform/mac-wk2/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[iOS WK2] API test DataInteractionTests.ExternalSourceDataTransferItemGetFolderAsEntr...
wenson_hsieh@apple.com [Wed, 31 Jan 2018 19:17:56 +0000 (19:17 +0000)]
[iOS WK2] API test DataInteractionTests.ExternalSourceDataTransferItemGetFolderAsEntry is failing after an SDK update
https://bugs.webkit.org/show_bug.cgi?id=182339
<rdar://problem/36041396>

Reviewed by Andy Estes.

This API test wasn't meant to be dependent on the size of an image generated using UIImagePNGRepresentation().
To make this test robust against changes to image encoding, write the raw data to the test folder instead and
check against the length of the data written.

* TestWebKitAPI/Tests/ios/DataInteractionTests.mm:
(testIconImageData):
(runTestWithTemporaryFolder):
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agocom.apple.WebKit.Storage crashing at com.apple.WebCore: WebCore::SWServerRegistration...
commit-queue@webkit.org [Wed, 31 Jan 2018 19:14:51 +0000 (19:14 +0000)]
com.apple.WebKit.Storage crashing at com.apple.WebCore: WebCore::SWServerRegistration::removeClientUsingRegistration
https://bugs.webkit.org/show_bug.cgi?id=182316
<rdar://problem/37025976>

Patch by Youenn Fablet <youenn@apple.com> on 2018-01-31
Reviewed by Chris Dumez.

SWServer.m_clientToControllingWorker and SWServerRegistration.m_clientsUsingRegistration might currently get out of sync.
This is a defensive fix to ensure storage process will not crash in Release.
We keep the ASSERT as this probably means there is an abnormal situation that other patches might fix.

* workers/service/server/SWServerRegistration.cpp:
(WebCore::SWServerRegistration::removeClientUsingRegistration):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[macOS] MiniBrowser isn’t app-sandboxed
mitz@apple.com [Wed, 31 Jan 2018 19:13:31 +0000 (19:13 +0000)]
[macOS] MiniBrowser isn’t app-sandboxed
https://bugs.webkit.org/show_bug.cgi?id=182147

Reviewed by Sam Weinig.

Enable App Sandbox and add a temporary exception allowing read access to all files.

* MiniBrowser/Configurations/Base.xcconfig: Enabled ad-hoc code signing.
* MiniBrowser/Configurations/MiniBrowser.xcconfig: Specified the entitlements file to use.
* MiniBrowser/MiniBrowser.entitlements: Added. Enables App Sandbox with outbound networking
  and read access from /.
* MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj: Added reference to new file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoRemove StorageToWebProcessConnection::removeSWServerConnection
commit-queue@webkit.org [Wed, 31 Jan 2018 19:10:36 +0000 (19:10 +0000)]
Remove StorageToWebProcessConnection::removeSWServerConnection
https://bugs.webkit.org/show_bug.cgi?id=182305

Patch by Youenn Fablet <youenn@apple.com> on 2018-01-31
Reviewed by Chris Dumez.

* StorageProcess/StorageToWebProcessConnection.cpp:
(WebKit::StorageToWebProcessConnection::removeSWServerConnection): Deleted.
* StorageProcess/StorageToWebProcessConnection.h:
* StorageProcess/StorageToWebProcessConnection.messages.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoCanonicalize aquiring the JSCell lock.
keith_miller@apple.com [Wed, 31 Jan 2018 18:57:13 +0000 (18:57 +0000)]
Canonicalize aquiring the JSCell lock.
https://bugs.webkit.org/show_bug.cgi?id=182320

Reviewed by Michael Saboff.

It's currently kinda annoying to figure out where
we aquire the a JSCell's lock. This patch adds a
helper to make it easier to grep...

* bytecode/UnlinkedCodeBlock.cpp:
(JSC::UnlinkedCodeBlock::visitChildren):
(JSC::UnlinkedCodeBlock::setInstructions):
(JSC::UnlinkedCodeBlock::shrinkToFit):
* runtime/ErrorInstance.cpp:
(JSC::ErrorInstance::finishCreation):
(JSC::ErrorInstance::materializeErrorInfoIfNeeded):
(JSC::ErrorInstance::visitChildren):
* runtime/JSArray.cpp:
(JSC::JSArray::shiftCountWithArrayStorage):
(JSC::JSArray::unshiftCountWithArrayStorage):
* runtime/JSCell.h:
(JSC::JSCell::cellLock):
* runtime/JSObject.cpp:
(JSC::JSObject::visitButterflyImpl):
(JSC::JSObject::convertContiguousToArrayStorage):
* runtime/JSPropertyNameEnumerator.cpp:
(JSC::JSPropertyNameEnumerator::visitChildren):
* runtime/SparseArrayValueMap.cpp:
(JSC::SparseArrayValueMap::add):
(JSC::SparseArrayValueMap::remove):
(JSC::SparseArrayValueMap::visitChildren):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[Modern Media Controls] Turn media/modern-media-controls/start-support tests back on
graouts@webkit.org [Wed, 31 Jan 2018 17:35:15 +0000 (17:35 +0000)]
[Modern Media Controls] Turn media/modern-media-controls/start-support tests back on
https://bugs.webkit.org/show_bug.cgi?id=182297

Reviewed by Jon Lee and Eric Carlson.

Source/WebCore:

We identified one cause of flakiness when turning those tests back on and fix it in this patch. There would be
cases where the play/pause button would never actually become visible because we would first set it as one of
the buttons in the left container in the bottom controls bar before we would identify that we should show the
prominent play/pause button, and then once we did, we would add it as a child to the InlineMediaControls. But,
because the actual layout of the ButtonsContainer would happen in a rAF due to being a layout() method, we would
remove the play/pause button from the InlineMediaControls and set it as a child of the left container even
though it wasn't visible.

Performing the layout of a ButtonsContainer should really happen immediately when we set the buttons, so in this
patch we remove the "buttons" property and clients of ButtonsContainer can simply add or remove children which
will be laid out in a row. To support this, we've added two notification methods to LayoutNode, one that indicates
when children have changed, didChangeChildren(), which we override in ButtonsContainer to perform a synchronous
layout. The other method is willRemoveChild() which we also override in ButtonsContainer, this time to reset
the "x" and "visible" properties which we set while performing layout.

This fixes flakiness for media/modern-media-controls/start-support/start-support-click-to-start.html, and maybe
other modern-media-controls tests that could have been affected by this unexpected behavior.

* Modules/modern-media-controls/controls/buttons-container.js:
(ButtonsContainer.prototype.willRemoveChild):
(ButtonsContainer.prototype.didChangeChildren):
(ButtonsContainer.prototype.layout):
(ButtonsContainer.prototype.get buttons): Deleted.
(ButtonsContainer.prototype.set buttons): Deleted.
* Modules/modern-media-controls/controls/inline-media-controls.js:
(InlineMediaControls.prototype.layout):
(InlineMediaControls.prototype._updateBottomControlsBarLabel):
(InlineMediaControls.prototype._addTopRightBarWithMuteButtonToChildren):
(InlineMediaControls):
* Modules/modern-media-controls/controls/layout-node.js:
(LayoutNode.prototype.set children):
(LayoutNode.prototype.addChild):
(LayoutNode.prototype.removeChild):
(LayoutNode.prototype.willRemoveChild):
(LayoutNode.prototype.didChangeChildren):
* Modules/modern-media-controls/controls/macos-fullscreen-media-controls.js:
(MacOSFullscreenMediaControls.prototype.layout):
* Modules/modern-media-controls/media/media-controller.js: Drive-by fix to correctly set the value passed as
data-auto-hide-delay on a <video> element.

LayoutTests:

Update tests to account for the new variable name for the playPauseButton and ensure we only listen to the
fullscreen event once to avoid flakiness in cases where we might exit fullscreen before the test completes.

We also update other tests that tested the behavior of ButtonsContainer.

* media/modern-media-controls/buttons-container/buttons-container-buttons-property-expected.txt:
* media/modern-media-controls/buttons-container/buttons-container-buttons-property.html:
* media/modern-media-controls/buttons-container/buttons-container-constructor-expected.txt:
* media/modern-media-controls/buttons-container/buttons-container-constructor.html:
* media/modern-media-controls/buttons-container/buttons-container-layout-expected.txt:
* media/modern-media-controls/buttons-container/buttons-container-layout.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-rtl-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-rtl.html:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-top-right-controls-bar-hidden-when-mute-button-disabled-expected.txt:
* media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-top-right-controls-bar-hidden-when-mute-button-disabled.html:
* media/modern-media-controls/pip-support/pip-support-click.html:
* media/modern-media-controls/start-support/start-support-click-to-start.html:
* media/modern-media-controls/start-support/start-support-fullscreen.html:
* media/modern-media-controls/start-support/start-support-lowPowerMode-expected.txt:
* media/modern-media-controls/start-support/start-support-lowPowerMode.html:
* platform/ios/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[RenderTreeBuilder] Move RenderMultiColumnFlow::resolveMovedChild to RenderTreeBuilder.
zalan@apple.com [Wed, 31 Jan 2018 16:45:30 +0000 (16:45 +0000)]
[RenderTreeBuilder] Move RenderMultiColumnFlow::resolveMovedChild to RenderTreeBuilder.
https://bugs.webkit.org/show_bug.cgi?id=182315
<rdar://problem/37048160>

Reviewed by Antti Koivisto.

resolveMovedChild() is only called from RenderTreeBuilder.

No change in functionality.

* rendering/RenderFragmentedFlow.h:
* rendering/RenderMultiColumnFlow.cpp:
(WebCore::RenderMultiColumnFlow::resolveMovedChild const): Deleted.
* rendering/RenderMultiColumnFlow.h:
* rendering/updating/RenderTreeBuilder.cpp:
(WebCore::RenderTreeBuilder::resolveMovedChildForMultiColumnFlow):
* rendering/updating/RenderTreeBuilder.h:
* rendering/updating/RenderTreeBuilderBlockFlow.cpp:
(WebCore::RenderTreeBuilder::BlockFlow::insertChild):
* rendering/updating/RenderTreeBuilderInline.cpp:
(WebCore::RenderTreeBuilder::Inline::insertChild):
* rendering/updating/RenderTreeBuilderMultiColumn.cpp:
(WebCore::RenderTreeBuilder::MultiColumn::resolveMovedChild):
* rendering/updating/RenderTreeBuilderMultiColumn.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[GStreamer] Silent WebAudio buffers support
philn@webkit.org [Wed, 31 Jan 2018 16:14:32 +0000 (16:14 +0000)]
[GStreamer] Silent WebAudio buffers support
https://bugs.webkit.org/show_bug.cgi?id=143869

Reviewed by Carlos Garcia Campos.

* platform/audio/gstreamer/WebKitWebAudioSourceGStreamer.cpp:
(webKitWebAudioSrcLoop): Flag silent buffers as GAP buffers. This
enables downstream elements to optimize the buffer processing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227902 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[JHBuild] Add gstreamer-vaapi in moduleset
philn@webkit.org [Wed, 31 Jan 2018 16:11:57 +0000 (16:11 +0000)]
[JHBuild] Add gstreamer-vaapi in moduleset
https://bugs.webkit.org/show_bug.cgi?id=182283

Reviewed by Carlos Alberto Lopez Perez.

GStreamer-VAAPI is required for hardware-accelerated video
playback support on Intel and AMD (using radeosi driver)
platforms.

* gstreamer/jhbuild.modules: Add gstreamer-vaapi.
* gtk/install-dependencies: Install libva.
* wpe/install-dependencies: Ditto.
* Scripts/webkitpy/port/gtk.py:
(GtkPort.setup_environ_for_server): Disable va-api support when
testing because it's incompatible with Mesa's softGL driver.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[WPE] Update jhbuild dependencies
cturner@igalia.com [Wed, 31 Jan 2018 15:55:37 +0000 (15:55 +0000)]
[WPE] Update jhbuild dependencies
https://bugs.webkit.org/show_bug.cgi?id=182286

This upgrades the following packages,
glib 2.44.1 -> 2.54.3
glib-networking 2.42.0 -> 2.54.1
libsoup 2.48.0 -> 2.60.3

The initial reason for performing these upgrades was that when
visiting https://youtube.com, WPE was getting TLS certificate
errors. After upgrading glib-networking, these were fixed, but the
upgrade introduced dependencies on newer versions of the other
packages upgraded in this commit.

The upgrade to glib caused a linking error in gstreamer, the
following errors were being logged during linking,

//usr/lib/x86_64-linux-gnu/libpangoft2-1.0.so.0: undefined reference to `hb_glib_script_from_script'
//usr/lib/x86_64-linux-gnu/libpangoft2-1.0.so.0: undefined reference to `hb_glib_get_unicode_funcs'
//usr/lib/x86_64-linux-gnu/libpangoft2-1.0.so.0: undefined reference to `hb_glib_script_to_script'

This was fixed by adding the glib package built in the jhbuild to
harfbuzz.

Finally, the upgrade of glib caused the glib-warning-fix.patch to
no longer apply, so this patch was removed.

Reviewed by Michael Catanzaro.

* wpe/jhbuild.modules: Added dependecies.
* wpe/patches/glib-warning-fix.patch: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed gardening. Update expectations of several WebDriver tests.
carlosgc@webkit.org [Wed, 31 Jan 2018 13:47:44 +0000 (13:47 +0000)]
Unreviewed gardening. Update expectations of several WebDriver tests.

* TestExpectations.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoJSC incorrectly interpreting script, sets Global Property instead of Global Lexical...
sbarati@apple.com [Wed, 31 Jan 2018 10:18:28 +0000 (10:18 +0000)]
JSC incorrectly interpreting script, sets Global Property instead of Global Lexical variable (LiteralParser / JSONP path)
https://bugs.webkit.org/show_bug.cgi?id=182074
<rdar://problem/36846261>

Reviewed by Mark Lam.

JSTests:

* stress/jsonp-program-evaluate-path-must-consider-global-lexical-environment.js: Added.
(assert):
(let.func):
(let.o.foo):
(varFunc):

LayoutTests/imported/w3c:

* web-platform-tests/service-workers/service-worker/import-scripts-updated-flag.https-expected.txt:

Source/JavaScriptCore:

This patch teaches the JSONP evaluator about the global lexical environment.
Before, it was using the global object as the global scope, but that's wrong.
The global lexical environment is the first node in the global scope chain.

* interpreter/Interpreter.cpp:
(JSC::Interpreter::executeProgram):
* jsc.cpp:
(GlobalObject::finishCreation):
(shellSupportsRichSourceInfo):
(functionDisableRichSourceInfo):
* runtime/LiteralParser.cpp:
(JSC::LiteralParser<CharType>::tryJSONPParse):
* runtime/LiteralParser.h:

LayoutTests:

* http/tests/security/regress-52192-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227898 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoclean up pushToSaveImmediateWithoutTouchingRegisters a bit
sbarati@apple.com [Wed, 31 Jan 2018 10:04:57 +0000 (10:04 +0000)]
clean up pushToSaveImmediateWithoutTouchingRegisters a bit
https://bugs.webkit.org/show_bug.cgi?id=181774

Reviewed by JF Bastien.

This function on ARM64 was considering what to do with the scratch
register. And conditionally invalidated what was in it. This is not
relevant though, since the function always recovers what was in that
register. This patch just switches it to using dataTempRegister
directly and updates the comment to describe why it can do so safely.

* assembler/MacroAssemblerARM64.h:
(JSC::MacroAssemblerARM64::pushToSaveImmediateWithoutTouchingRegisters):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227897 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[EME] Fix NavigatorEME::tryNextSupportedConfiguration typo
calvaris@igalia.com [Wed, 31 Jan 2018 09:19:56 +0000 (09:19 +0000)]
[EME] Fix NavigatorEME::tryNextSupportedConfiguration typo
https://bugs.webkit.org/show_bug.cgi?id=182324

Reviewed by Žan Doberšek.

* Modules/encryptedmedia/NavigatorEME.cpp:
(WebCore::tryNextSupportedConfiguration): Fix
candidateCofiguration typo.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227895 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed, rolling out r227782.
carlosgc@webkit.org [Wed, 31 Jan 2018 09:17:50 +0000 (09:17 +0000)]
Unreviewed, rolling out r227782.

It didn't fix the problem

Reverted changeset:

"Unreviewed. Try to avoid webdriver tests step timing out in
the bots."
https://trac.webkit.org/changeset/227782

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227894 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION(r227544): [GTK] contextMenuEvent is NULL on CONTEXT_MENU call
carlosgc@webkit.org [Wed, 31 Jan 2018 09:14:33 +0000 (09:14 +0000)]
REGRESSION(r227544): [GTK] contextMenuEvent is NULL on CONTEXT_MENU call
https://bugs.webkit.org/show_bug.cgi?id=182224

Reviewed by Michael Catanzaro.

Move the gestures handling to WebKitWebViewBase. This patch adds GestureControllerClient class, created and
implemented by WebKitWebViewBase and used by GestureController instead of the WebPageProxy. This way we ensure
events are handled consistently.

* UIProcess/API/gtk/PageClientImpl.cpp:
(WebKit::PageClientImpl::doneWithTouchEvent): Cast the GdkEvent since GestureController no longer works wirth
const GdkEvents.
(WebKit::PageClientImpl::zoom): Use webkit_web_view_set_zoom_level() in case of WebKitWebView or
WebPageProxy::setPageZoomFactor() otherwise.
* UIProcess/API/gtk/PageClientImpl.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(ClickCounter::currentClickCountForGdkButtonEvent): Receive a GdkEvent to avoid casts.
(webkitWebViewBaseHandleMouseEvent): Helper function to handle mouse events.
(webkitWebViewBaseButtonPressEvent): Use webkitWebViewBaseHandleMouseEvent.
(webkitWebViewBaseButtonReleaseEvent): Ditto.
(webkitWebViewBaseHandleWheelEvent): Helper function to handle wheel events.
(webkitWebViewBaseScrollEvent): Use webkitWebViewBaseHandleWheelEvent.
(webkitWebViewBaseMotionNotifyEvent): Use webkitWebViewBaseHandleMouseEvent.
(webkitWebViewBaseCrossingNotifyEvent): Ditto.
(webkitWebViewBaseGestureController): Pass the widget and client to GestureController.
* UIProcess/gtk/GestureController.cpp:
(WebKit::GestureController::GestureController): Receives a widget and client now.
(WebKit::GestureController::handleEvent): Remove the const.
(WebKit::GestureController::Gesture::Gesture): Initialize client.
(WebKit::GestureController::Gesture::handleEvent): Remove the const.
(WebKit::GestureController::DragGesture::startDrag): Use the client instead of WebPageProxy.
(WebKit::GestureController::DragGesture::handleDrag): Ditto.
(WebKit::GestureController::DragGesture::handleTap): Ditto.
(WebKit::GestureController::DragGesture::begin): Ignore the const returned by gtk_gesture_get_last_event().
(WebKit::GestureController::DragGesture::update): Ditto.
(WebKit::GestureController::DragGesture::end): Ditto.
(WebKit::GestureController::DragGesture::DragGesture): Receives a widget and client now.
(WebKit::GestureController::SwipeGesture::startMomentumScroll): Use the client instead of WebPageProxy.
(WebKit::GestureController::SwipeGesture::swipe): Ignore the const returned by gtk_gesture_get_last_event().
(WebKit::GestureController::SwipeGesture::SwipeGesture): Receives a widget and client now.
(WebKit::GestureController::ZoomGesture::begin): Start the zoom.
(WebKit::GestureController::ZoomGesture::startZoom): Use the client instead of WebPageProxy.
(WebKit::GestureController::ZoomGesture::handleZoom): Ditto.
(WebKit::GestureController::ZoomGesture::ZoomGesture): Receives a widget and client now.
(WebKit::GestureController::LongPressGesture::longPressed): Use the client instead of WebKitWebView.
(WebKit::GestureController::LongPressGesture::pressed): Ignore the const returned by gtk_gesture_get_last_event().
(WebKit::GestureController::LongPressGesture::LongPressGesture): Receives a widget and client now.
(WebKit::GestureController::Gesture::simulateMouseClick): Deleted.
(WebKit::createScrollEvent): Deleted.
* UIProcess/gtk/GestureController.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227893 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoWeb Automation: cookies returned by automation should have expiry time in seconds
carlosgc@webkit.org [Wed, 31 Jan 2018 09:09:45 +0000 (09:09 +0000)]
Web Automation: cookies returned by automation should have expiry time in seconds
https://bugs.webkit.org/show_bug.cgi?id=182293

Reviewed by Brian Burg.

When creating a WebCore::Cookie from a WebDriver object we convert the given expiry time to milliseconds, but
when creating a WebDriver object from a WebCore::Cookie we are keeping the milliseconds. We should convert to
seconds for consistency, so that the WebDriver always handles seconds.

Fixes: imported/w3c/webdriver/tests/cookies/get_named_cookie.py::test_get_named_cookie

* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::buildObjectForCookie): Convert expiry to seconds.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227891 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoWebDriver: add domains to server config in w3c tests
carlosgc@webkit.org [Wed, 31 Jan 2018 09:08:37 +0000 (09:08 +0000)]
WebDriver: add domains to server config in w3c tests
https://bugs.webkit.org/show_bug.cgi?id=182294

Reviewed by Carlos Alberto Lopez Perez.

Several cookie tests expect server_config to have a domains key.

Fixes: imported/w3c/webdriver/tests/cookies/add_cookie.py::test_add_domain_cookie
       imported/w3c/webdriver/tests/cookies/add_cookie.py::test_add_session_cookie_with_leading_dot_character_in_domain

* Scripts/webkitpy/webdriver_tests/webdriver_w3c_executor.py:
(WebDriverW3CExecutor.__init__):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227881 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed. Update W3C WebDriver imported tests.
carlosgc@webkit.org [Wed, 31 Jan 2018 09:07:43 +0000 (09:07 +0000)]
Unreviewed. Update W3C WebDriver imported tests.

Tools:

Show full pytest failure logs.

pytest truncates long lists when printing a failed assertion.
This is not great for debugging and arguably it should be the
default to show the full diff when comparing lists.

* Scripts/webkitpy/webdriver_tests/pytest_runner.py:
(run):

WebDriverTests:

* imported/w3c/importer.json:
* imported/w3c/tools/wptrunner/requirements_sauce.txt:
* imported/w3c/tools/wptrunner/tox.ini:
* imported/w3c/tools/wptrunner/wptrunner/browsers/sauce.py:
* imported/w3c/tools/wptrunner/wptrunner/executors/executormarionette.py:
* imported/w3c/tools/wptrunner/wptrunner/executors/pytestrunner/runner.py:
* imported/w3c/tools/wptrunner/wptrunner/metadata.py:
* imported/w3c/webdriver/OWNERS:
* imported/w3c/webdriver/tests/__init__.py:
* imported/w3c/webdriver/tests/actions/mouse.py:
* imported/w3c/webdriver/tests/actions/mouse_dblclick.py:
* imported/w3c/webdriver/tests/actions/support/mouse.py:
* imported/w3c/webdriver/tests/cookies/add_cookie.py:
* imported/w3c/webdriver/tests/cookies/delete_cookie.py:
* imported/w3c/webdriver/tests/cookies/get_named_cookie.py:
* imported/w3c/webdriver/tests/element_send_keys/form_controls.py: Added.
* imported/w3c/webdriver/tests/element_send_keys/interactability.py:
* imported/w3c/webdriver/tests/interaction/element_clear.py:
* imported/w3c/webdriver/tests/support/asserts.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227880 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed iOS build fix.
rniwa@webkit.org [Wed, 31 Jan 2018 08:19:35 +0000 (08:19 +0000)]
Unreviewed iOS build fix.

* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227879 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[WinCairo] syntheticBoldOffset makes a font with embedded bitmap fonts shown as doubl...
commit-queue@webkit.org [Wed, 31 Jan 2018 07:28:23 +0000 (07:28 +0000)]
[WinCairo] syntheticBoldOffset makes a font with embedded bitmap fonts shown as double strike in HiDPI
https://bugs.webkit.org/show_bug.cgi?id=182093

Patch by Fujii Hironori <Hironori.Fujii@sony.com> on 2018-01-30
Reviewed by Alex Christensen.

WebKit draws texts doubly by shifting 1px for fonts without bold
fonts but bold font is desired. This synthetic bold method draws
fonts with embedded bitmap fonts look like double strike in HiDPI.

No new tests (Covered by the existing tests).

* platform/graphics/win/FontCacheWin.cpp:
(WebCore::createGDIFont): Overwrite lfWeight if the desired weight
is bold and matched font weight is't bold and in WinCairo port.
Fix wrong indentation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227876 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoAdd callbacks to testRunner.statisticsSetShouldPartitionCookiesForHost() and testRunn...
wilander@apple.com [Wed, 31 Jan 2018 06:01:54 +0000 (06:01 +0000)]
Add callbacks to testRunner.statisticsSetShouldPartitionCookiesForHost() and testRunner.statisticsUpdateCookiePartitioning()
https://bugs.webkit.org/show_bug.cgi?id=181958
https://bugs.webkit.org/show_bug.cgi?id=182072
<rdar://problem/36801804>
<rdar://problem/36845795>

Reviewed by Brent Fulgham.

Because of the asynchronous nature of XPC and cookies,
we need callbacks in these TestRunner functions so that
the layout tests can wait for state changes to finish
before checking test conditions.

Source/WebKit:

* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreStatisticsUpdateCookiePartitioning):
(WKWebsiteDataStoreSetStatisticsShouldPartitionCookiesForHost):
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(-[WKWebsiteDataStore _resourceLoadStatisticsUpdateCookiePartitioning]):
(-[WKWebsiteDataStore _resourceLoadStatisticsUpdateCookiePartitioning:]):
(-[WKWebsiteDataStore _resourceLoadStatisticsSetShouldPartitionCookies:forHost:]):
(-[WKWebsiteDataStore _resourceLoadStatisticsSetShouldPartitionCookies:forHost:completionHandler:]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::resourceLoadStatisticsUpdated):
(WebKit::WebResourceLoadStatisticsStore::logUserInteraction):
(WebKit::WebResourceLoadStatisticsStore::logNonRecentUserInteraction):
(WebKit::WebResourceLoadStatisticsStore::scheduleCookiePartitioningUpdate):
(WebKit::WebResourceLoadStatisticsStore::scheduleCookiePartitioningUpdateForDomains):
(WebKit::WebResourceLoadStatisticsStore::scheduleClearPartitioningStateForDomains):
(WebKit::WebResourceLoadStatisticsStore::mergeWithDataFromDecoder):
(WebKit::WebResourceLoadStatisticsStore::clearInMemory):
(WebKit::WebResourceLoadStatisticsStore::updateCookiePartitioning):
(WebKit::WebResourceLoadStatisticsStore::updateCookiePartitioningForDomains):
(WebKit::WebResourceLoadStatisticsStore::clearPartitioningStateForDomains):
* UIProcess/WebResourceLoadStatisticsStore.h:

Tools:

* WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::didReceiveMessageToPage):
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::statisticsUpdateCookiePartitioning):
(WTR::TestRunner::statisticsSetShouldPartitionCookiesForHost):
(WTR::TestRunner::statisticsCallDidSetPartitionOrBlockCookiesForHostCallback):
* WebKitTestRunner/InjectedBundle/TestRunner.h:
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didSetPartitionOrBlockCookiesForHost):
* WebKitTestRunner/TestInvocation.h:
* WebKitTestRunner/cocoa/TestControllerCocoa.mm:
(WTR::TestController::statisticsUpdateCookiePartitioning):
(WTR::TestController::statisticsSetShouldPartitionCookiesForHost):

LayoutTests:

* http/tests/resourceLoadStatistics/add-blocking-to-redirect-expected.txt:
* http/tests/resourceLoadStatistics/add-blocking-to-redirect.html:
* http/tests/resourceLoadStatistics/add-partitioning-to-redirect-expected.txt:
* http/tests/resourceLoadStatistics/add-partitioning-to-redirect.html:
* http/tests/resourceLoadStatistics/non-prevalent-resources-can-access-cookies-in-a-third-party-context.html:
* http/tests/resourceLoadStatistics/partitioned-and-unpartitioned-cookie-deletion.html:
* http/tests/resourceLoadStatistics/partitioned-and-unpartitioned-cookie-with-partitioning-timeout.html:
* http/tests/resourceLoadStatistics/partitioned-cookies-with-and-without-user-interaction-expected.txt:
* http/tests/resourceLoadStatistics/partitioned-cookies-with-and-without-user-interaction.html:
* http/tests/resourceLoadStatistics/remove-blocking-in-redirect-expected.txt:
* http/tests/resourceLoadStatistics/remove-blocking-in-redirect.html:
* http/tests/resourceLoadStatistics/remove-partitioning-in-redirect-expected.txt:
* http/tests/resourceLoadStatistics/remove-partitioning-in-redirect.html:
* http/tests/resourceLoadStatistics/third-party-cookie-with-and-without-user-interaction.html:
* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227875 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoApply poisoning to TypedArray vector pointers.
mark.lam@apple.com [Wed, 31 Jan 2018 05:23:52 +0000 (05:23 +0000)]
Apply poisoning to TypedArray vector pointers.
https://bugs.webkit.org/show_bug.cgi?id=182155
<rdar://problem/36286266>

Reviewed by JF Bastien.

Source/JavaScriptCore:

The TypeArray's vector pointer is now poisoned.  The poison value is chosen based
on a TypeArray's jsType.  The JSType must be between FirstTypedArrayType and
LastTypedArrayType.  At runtime, we enforce that the index is well-behaved by
masking it against TypedArrayPoisonIndexMask.  TypedArrayPoisonIndexMask (16) is
the number of TypedArray types (10) rounded up to the next power of 2.
Accordingly, we reserve an array of TypedArrayPoisonIndexMask poisons so that we
can use index masking on the index, and be guaranteed that the masked index will
be within bounds of the poisons array.

1. Fixed both DFG and FTL versions of compileGetTypedArrayByteOffset() to not
   do any unnecessary work if the TypedArray vector is null.

   FTL's cagedMayBeNull() is no longer needed because it is only used by
   compileGetTypedArrayByteOffset(), and we need to enhance it to handle unpoisoning
   in a TypedArray specific way.  So, might as well do the work inline in
   compileGetTypedArrayByteOffset() instead.

2. Removed an unnecessary null-check in DFGSpeculativeJIT's compileNewTypedArrayWithSize()
   because there's already a null check above it that ensures that sizeGPR is
   never null.

3. In LLInt's _llint_op_get_by_val, move the TypedArray length check before the
   loading of the vector for unpoisoning and uncaging.  We don't need the vector
   if the length is 0.

Implementation notes on the need to null check the TypeArray vector:

1. DFG::SpeculativeJIT::jumpForTypedArrayIsNeuteredIfOutOfBounds() does not need a
   m_poisonedVector null check because the function is a null check.

2. DFG::SpeculativeJIT::compileGetIndexedPropertyStorage() does not need a
   m_poisonedVector null check because it is followed by a call to
   cageTypedArrayStorage() which assumes that storageReg cannot be null.

3. DFG::SpeculativeJIT::compileGetTypedArrayByteOffset() already has a
   m_poisonedVector null check.

4. DFG::SpeculativeJIT::compileNewTypedArrayWithSize() does not need a vector null
   check because the poisoning code is preceded by a sizeGPR null check, which
   ensures that the storageGPR (vector to be poisoned) is not null.

5. FTL's compileGetIndexedPropertyStorage() does not need a m_poisonedVector null
   check because it is followed by a call to caged() which assumes that the
   vector cannot be null.

6. FTL's compileGetTypedArrayByteOffset() already has a m_poisonedVector null check.

7. FTL's compileNewTypedArray() does not need a vector null check because the
   poisoning code is preceded by a size null check, which ensures that the
   storage (vector to be poisoned) is not null.

8. FTL's speculateTypedArrayIsNotNeutered() does not need a
   m_poisonedVector null check because the function is a null check.

9. IntrinsicGetterAccessCase::emitIntrinsicGetter()'s TypedArrayByteOffsetIntrinsic
   case needs a null check so that it does not try to unpoison a null vector.

10. JIT::emitIntTypedArrayGetByVal() does not need a vector null check because
    we already do a length check even before loading the vector.

11. JIT::emitFloatTypedArrayGetByVal() does not need a vector null check because
    we already do a length check even before loading the vector.

12. JIT::emitIntTypedArrayPutByVal() does not need a vector null check because
    we already do a length check even before loading the vector.

13. JIT::emitFloatTypedArrayPutByVal() does not need a vector null check because
    we already do a length check even before loading the vector.

14. LLInt's loadTypedArrayCaged() does not need a vector null check because its
    client will do a TypedArray length check before calling it.

* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::checkArray):
* dfg/DFGNode.h:
(JSC::DFG::Node::hasArrayMode):
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::jumpForTypedArrayIsNeuteredIfOutOfBounds):
(JSC::DFG::SpeculativeJIT::compileGetIndexedPropertyStorage):
(JSC::DFG::SpeculativeJIT::compileGetTypedArrayByteOffset):
(JSC::DFG::SpeculativeJIT::compileNewTypedArrayWithSize):
* ftl/FTLAbstractHeapRepository.h:
* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileGetIndexedPropertyStorage):
(JSC::FTL::DFG::LowerDFGToB3::compileGetTypedArrayByteOffset):
(JSC::FTL::DFG::LowerDFGToB3::compileNewTypedArray):
(JSC::FTL::DFG::LowerDFGToB3::speculateTypedArrayIsNotNeutered):
(JSC::FTL::DFG::LowerDFGToB3::cagedMayBeNull): Deleted.
* jit/IntrinsicEmitter.cpp:
(JSC::IntrinsicGetterAccessCase::emitIntrinsicGetter):
* jit/JITPropertyAccess.cpp:
(JSC::JIT::emitIntTypedArrayGetByVal):
(JSC::JIT::emitFloatTypedArrayGetByVal):
(JSC::JIT::emitIntTypedArrayPutByVal):
(JSC::JIT::emitFloatTypedArrayPutByVal):
* llint/LowLevelInterpreter.asm:
* llint/LowLevelInterpreter64.asm:
* offlineasm/arm64.rb:
* offlineasm/x86.rb:
* runtime/CagedBarrierPtr.h:
* runtime/JSArrayBufferView.cpp:
(JSC::JSArrayBufferView::JSArrayBufferView):
(JSC::JSArrayBufferView::finalize):
(JSC::JSArrayBufferView::neuter):
* runtime/JSArrayBufferView.h:
(JSC::JSArrayBufferView::vector const):
(JSC::JSArrayBufferView::offsetOfPoisonedVector):
(JSC::JSArrayBufferView::poisonFor):
(JSC::JSArrayBufferView::Poison::key):
(JSC::JSArrayBufferView::offsetOfVector): Deleted.
* runtime/JSCPoison.cpp:
(JSC::initializePoison):
* runtime/JSCPoison.h:
* runtime/JSGenericTypedArrayViewInlines.h:
(JSC::JSGenericTypedArrayView<Adaptor>::estimatedSize):
(JSC::JSGenericTypedArrayView<Adaptor>::visitChildren):
(JSC::JSGenericTypedArrayView<Adaptor>::slowDownAndWasteMemory):
* runtime/JSObject.h:

Source/WTF:

1. Added the ability to poison a CagedPtr.

2. Prevent CagedPtr from being implicitly instantiated, and add operator= methods
   instead.  This is because implicitly instantiated CagedPtrs with a poisoned
   trait may silently use a wrong poison value.

* wtf/CagedPtr.h:
(WTF::CagedPtr::CagedPtr):
(WTF::CagedPtr::get const):
(WTF::CagedPtr::operator=):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227874 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoREGRESSION(r227550): Resource timing API is disabled on macOS
rniwa@webkit.org [Wed, 31 Jan 2018 05:15:59 +0000 (05:15 +0000)]
REGRESSION(r227550): Resource timing API is disabled on macOS
https://bugs.webkit.org/show_bug.cgi?id=182318

Reviewed by Chris Dumez.

Resource timing API should be enabled by default all versions of macOS we support now.

* Shared/WebPreferences.yaml:
* Shared/WebPreferencesDefaultValues.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227873 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[Win] Warning fix.
commit-queue@webkit.org [Wed, 31 Jan 2018 04:47:43 +0000 (04:47 +0000)]
[Win] Warning fix.
https://bugs.webkit.org/show_bug.cgi?id=177007

Patch by Fujii Hironori <Hironori.Fujii@sony.com> on 2018-01-30
Reviewed by Yusuke Suzuki.

Source/JavaScriptCore:

* interpreter/StackVisitor.cpp:
(JSC::StackVisitor::Frame::dump const):
Changed the type of locationRawBits from unsigned to uintptr_t.
* runtime/IntlNumberFormat.cpp:
(JSC::IntlNumberFormat::createNumberFormat):
Initialize 'style' to avoid potentially uninitialized local variable warning.

Source/WTF:

A 32 bit integer should not be casted to a pointer directly in 64
bit Windows. It should be casted to a uintptr_t beforehand.

Rules for Using Pointers (Windows)
https://msdn.microsoft.com/library/windows/desktop/aa384242

* wtf/ThreadingWin.cpp (InvalidThread): Casted 0xbbadbeef to uintptr_t.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227872 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[webkitpy] Config file for apache is copied twice.
commit-queue@webkit.org [Wed, 31 Jan 2018 04:27:41 +0000 (04:27 +0000)]
[webkitpy] Config file for apache is copied twice.
https://bugs.webkit.org/show_bug.cgi?id=182131

Patch by Basuke Suzuki <Basuke.Suzuki@sony.com> on 2018-01-30
Reviewed by Daniel Bates.

* Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
(LayoutTestApacheHttpd.__init__):
(LayoutTestApacheHttpd._copy_apache_config_file):
(LayoutTestApacheHttpd._get_apache_config_file_path): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227871 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoFetchResponse should support ConsumeData callback on chunk data is received
gskachkov@gmail.com [Wed, 31 Jan 2018 04:23:50 +0000 (04:23 +0000)]
FetchResponse should support ConsumeData callback on chunk data is received
https://bugs.webkit.org/show_bug.cgi?id=181600
<rdar://problem/36932547>

Reviewed by Alex Christensen.

LayoutTests/imported/w3c:

* web-platform-tests/service-workers/service-worker/fetch-event-respond-with-body-loaded-in-chunk.https-expected.txt: Added.
* web-platform-tests/service-workers/service-worker/fetch-event-respond-with-body-loaded-in-chunk.https.html: Added.
* web-platform-tests/service-workers/service-worker/resources/fetch-event-respond-with-body-loaded-in-chunk-worker.js: Added.

Source/WebCore:

Test: imported/w3c/web-platform-tests/service-workers/service-worker/fetch-event-respond-with-body-loaded-in-chunk.https.html

Apply same read-by-chunk strategy for body coming from network as for body coming from ReadableStream.
Refactor to have one consumeBodyReceivedByChunk method for both cases.
This allows streaming data from a service worker to WebProcess as would be done by NetworkProcess.

* Modules/cache/DOMCache.cpp:
(WebCore::DOMCache::addAll):
(WebCore::DOMCache::put):
* Modules/fetch/FetchResponse.cpp:
(WebCore::FetchResponse::BodyLoader::didSucceed):
(WebCore::FetchResponse::BodyLoader::didReceiveData):
(WebCore::FetchResponse::BodyLoader::consumeDataByChunk):
(WebCore::FetchResponse::consumeBody):
(WebCore::FetchResponse::consumeBodyReceivedByChunk):
* Modules/fetch/FetchResponse.h:
* workers/service/context/ServiceWorkerFetch.cpp:
(WebCore::ServiceWorkerFetch::processResponse):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227870 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoWKWebView layout is sometimes wrong after rotation on iPhone X
timothy_horton@apple.com [Wed, 31 Jan 2018 03:02:59 +0000 (03:02 +0000)]
WKWebView layout is sometimes wrong after rotation on iPhone X
https://bugs.webkit.org/show_bug.cgi?id=182304
<rdar://problem/34158671>

Reviewed by Simon Fraser.

* Platform/spi/ios/UIKitSPI.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _computedContentInset]):
(-[WKWebView _scrollViewSystemContentInset]):
(activeMinimumLayoutSize):
It turns out that it is not always safe to look at the safe area insets of
children from inside layoutSubviews, even after the call to super.

Instead, make use of the fact that WKScrollView and WKWebView have identical
coordinate spaces, and map WKWebView's safe area insets into the WKScrollView.
It's safe to use the scroll view's affected-edges and contentScrollInset,
because those aren't updated at the same outside-of-layout time that
safe area insets are.

We could alternatively move all calls to activeMinimumLayoutSize outside
of layoutSubviews, but that seems like a larger and riskier change.

All attempts to write a test have failed; this depends heavily on use of
autolayout and the mechanism by which the system updates system-owned
safe area insets during device rotation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227869 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoMove http/wpt/service-workers/clone-opaque-being-loaded-response.https.html to use...
commit-queue@webkit.org [Wed, 31 Jan 2018 01:35:34 +0000 (01:35 +0000)]
Move http/wpt/service-workers/clone-opaque-being-loaded-response.https.html to use HTTP
https://bugs.webkit.org/show_bug.cgi?id=182202
<rdar://problem/36912402>

Patch by Youenn Fablet <youenn@apple.com> on 2018-01-30
Reviewed by Chris Dumez.
Follow-up to ensure subresource load is also done in HTTP and not HTTPS

* http/wpt/service-workers/clone-opaque-being-loaded-response-worker.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227867 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoJSExports.h should be included as <JavaScriptCore/JSExportMacros.h>
don.olmstead@sony.com [Wed, 31 Jan 2018 00:53:48 +0000 (00:53 +0000)]
JSExports.h should be included as <JavaScriptCore/JSExportMacros.h>
https://bugs.webkit.org/show_bug.cgi?id=182312

Reviewed by Michael Catanzaro.

Source/WebCore:

No new tests. No change in behavior.

* config.h:

Source/WebKit:

* config.h:

Source/WebKitLegacy/cf:

* WebCoreSupport/WebInspectorClientCF.cpp:

Source/WebKitLegacy/mac:

* WebKitPrefix.h:

Tools:

* DumpRenderTree/config.h:
* TestWebKitAPI/config.h:
* WebKitTestRunner/config.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227866 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoMarked fast/workers/worker-cloneport.html as flaky on macOS Release.
jlewis3@apple.com [Wed, 31 Jan 2018 00:46:02 +0000 (00:46 +0000)]
Marked fast/workers/worker-cloneport.html as flaky on macOS Release.
https://bugs.webkit.org/show_bug.cgi?id=182007

Unreviewed test gardening.

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227865 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoWeb Inspector: Replace Object.shallowMerge with ES2018 spread operator
webkit@devinrousso.com [Wed, 31 Jan 2018 00:39:18 +0000 (00:39 +0000)]
Web Inspector: Replace Object.shallowMerge with ES2018 spread operator
https://bugs.webkit.org/show_bug.cgi?id=182219

Reviewed by Brian Burg.

* UserInterface/Base/Utilities.js:
(Object.shallowMerge): Deleted.
* UserInterface/Base/DOMUtilities.js:
(WI.linkifyNodeReference):
* UserInterface/Base/Main.js:
(WI.handlePossibleLinkClick):
(WI.openURL):
(WI.showSourceCodeLocation):
(WI.showOriginalUnformattedSourceCodeLocation):
(WI.showOriginalOrFormattedSourceCodeLocation):
(WI.showOriginalOrFormattedSourceCodeTextRange):
(WI.linkifyLocation):
* UserInterface/Views/DOMTreeElement.js:
(WI.DOMTreeElement.prototype._insertAdjacentHTML):
* UserInterface/Views/WebSocketContentView.js:
(WI.WebSocketContentView.prototype._addRow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227864 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoStyleBench: Generate more mutation steps
rniwa@webkit.org [Wed, 31 Jan 2018 00:31:49 +0000 (00:31 +0000)]
StyleBench: Generate more mutation steps
https://bugs.webkit.org/show_bug.cgi?id=182291

Rubber-stamped by Antti Koivisto

Use zero-based index instead of one-based index.

* StyleBench/resources/tests.js:
(makeSteps):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227863 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUse double-quotes when serializing font-feature-settings
commit-queue@webkit.org [Wed, 31 Jan 2018 00:08:44 +0000 (00:08 +0000)]
Use double-quotes when serializing font-feature-settings
https://bugs.webkit.org/show_bug.cgi?id=182201

Source/WebCore:

According to the CSSOM spec, all strings should be serialized with double-quotes. The feature
tag value was previously serialized with single-quotes; change this to double-quotes to match
the spec and non-WebKit browsers.

Patch by Chris Nardi <cnardi@chromium.org> on 2018-01-30
Reviewed by Myles C. Maxfield.

Updated css3/font-feature-settings-parsing.html, fast/css/inherited-properties-rare-text.html,
and fast/text/font-face-javascript.html.

* css/CSSFontFeatureValue.cpp:
(WebCore::CSSFontFeatureValue::customCSSText const):

LayoutTests:

Update tests with double-quotes instead of single-quotes.

Patch by Chris Nardi <cnardi@chromium.org> on 2018-01-30
Reviewed by Myles C. Maxfield.

* css3/font-feature-settings-parsing-expected.txt:
* css3/font-feature-settings-parsing.html:
* fast/css/inherited-properties-rare-text-expected.txt:
* fast/text/font-face-javascript-expected.txt:
* fast/text/font-face-javascript.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227862 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnreviewed build fix for JSCOnly after r227845.
don.olmstead@sony.com [Tue, 30 Jan 2018 23:13:27 +0000 (23:13 +0000)]
Unreviewed build fix for JSCOnly after r227845.
https://bugs.webkit.org/show_bug.cgi?id=182274

* wtf/PlatformJSCOnly.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227861 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoAdd telemetry to track storage access API adoption
bfulgham@apple.com [Tue, 30 Jan 2018 23:12:03 +0000 (23:12 +0000)]
Add telemetry to track storage access API adoption
https://bugs.webkit.org/show_bug.cgi?id=182197
<rdar://problem/35803309>

Reviewed by Chris Dumez.

Part 2: Add telemetry for the Storage Access API case

Source/WebCore:

Adds a new convenience method to identify origin/page/frame combinations that
have been granted access to the Storage Access API. This is used for debug
logging in the NetworkProcess. It is not used in production builds.

* platform/network/NetworkStorageSession.h:
* platform/network/cf/NetworkStorageSessionCFNet.cpp:
(WebCore::NetworkStorageSession::hasStorageAccessForFrame const): Added.

Source/WebKit:

This change increments a counter when an origin is loaded in a first part context because
it was granted Storage Access API permissions.

* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::logCookieInformation const): Add logging to indicate
loads that happened with the Storage Access API enabled.
* UIProcess/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::requestStorageAccess): Increment counter for
loads using the new API.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227860 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[iOS] API test UIPasteboardTests.DataTransferGetDataWhenPastingPlatformRepresentation...
wenson_hsieh@apple.com [Tue, 30 Jan 2018 23:05:57 +0000 (23:05 +0000)]
[iOS] API test UIPasteboardTests.DataTransferGetDataWhenPastingPlatformRepresentations fails after r223440
https://bugs.webkit.org/show_bug.cgi?id=182307
<rdar://problem/36041485>

Reviewed by Ryosuke Niwa.

Adjusts a failing API test that we forgot to rebaseline after r223440, which ensures that the result of reading
"text/html" from the DataTransfer is sanitized markup. To do this, teach the test harness
(dump-datatransfer-types.html) to ignore inline styles in markup if a flag is set, and then set that flag in
UIPasteboardTests.DataTransferGetDataWhenPastingPlatformRepresentations.

* TestWebKitAPI/Tests/WebKitCocoa/dump-datatransfer-types.html:
* TestWebKitAPI/Tests/ios/UIPasteboardTests.mm:
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227859 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoRelease assert in updateLayout() via AXObjectCache::childrenChanged
rniwa@webkit.org [Tue, 30 Jan 2018 22:47:05 +0000 (22:47 +0000)]
Release assert in updateLayout() via AXObjectCache::childrenChanged
https://bugs.webkit.org/show_bug.cgi?id=182279
<rdar://problem/36994456>

Reviewed by Antti Koivisto.

Source/WebCore:

Disable the assertion in Document::updateLayout and Document::updateStyle* in this particular circumstance as fixing it
would require a large architectural refactoring of the accessibility code.

Test: accessibility/accessibility-object-update-during-style-resolution-crash.html

* accessibility/AXObjectCache.cpp:
(WebCore::AXObjectCache::childrenChanged): Disabled the release assertion here.
* dom/Document.cpp:
(WebCore::Document::isSafeToUpdateStyleOrLayout const): Check LayoutAssertionDisableScope::shouldDisable.
* dom/ScriptDisallowedScope.h:
(WebCore::ScriptDisallowedScope::LayoutAssertionDisableScope): Added.
(WebCore::ScriptDisallowedScope::LayoutAssertionDisableScope::LayoutAssertionDisableScope): Added.
(WebCore::ScriptDisallowedScope::LayoutAssertionDisableScope::~LayoutAssertionDisableScope): Added.
(WebCore::ScriptDisallowedScope::LayoutAssertionDisableScope::shouldDisable): Added.
* page/LayoutContext.cpp:
(WebCore::LayoutContext::layout): Check LayoutAssertionDisableScope::shouldDisable.

LayoutTests:

Added a regression test.

* accessibility/accessibility-object-update-during-style-resolution-crash-expected.txt: Added.
* accessibility/accessibility-object-update-during-style-resolution-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227858 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoSkipping imported/w3c/web-platform-tests/service-workers/service-worker/navigation...
jlewis3@apple.com [Tue, 30 Jan 2018 22:45:09 +0000 (22:45 +0000)]
Skipping imported/w3c/web-platform-tests/service-workers/service-worker/navigation-redirect.https.html.
https://bugs.webkit.org/show_bug.cgi?id=182311

Unreviewed test expectations.

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227857 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[RenderTreeBuilder] Move RenderRubyRun::rubyBaseSafe to RenderTreeBuilder::Ruby
zalan@apple.com [Tue, 30 Jan 2018 22:40:57 +0000 (22:40 +0000)]
[RenderTreeBuilder] Move RenderRubyRun::rubyBaseSafe to RenderTreeBuilder::Ruby
https://bugs.webkit.org/show_bug.cgi?id=182306
<rdar://problem/37041440>

Reviewed by Antti Koivisto.

With all the ruby mutation code moving, only RenderTreeBuilder calls RenderRubyRun::rubyBaseSafe.

No change in functionality.

* rendering/RenderRubyRun.cpp:
(WebCore::RenderRubyRun::rubyText const):
(WebCore::RenderRubyRun::rubyBase const):
(WebCore::RenderRubyRun::firstLineBlock const):
(WebCore::RenderRubyRun::takeChild):
(WebCore::RenderRubyRun::rubyBaseSafe): Deleted.
* rendering/RenderRubyRun.h:
* rendering/updating/RenderTreeBuilderRuby.cpp:
(WebCore::RenderTreeBuilder::Ruby::insertChild):
(WebCore::RenderTreeBuilder::Ruby::rubyBaseSafe):
* rendering/updating/RenderTreeBuilderRuby.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227856 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoMake preserve and restore focus more likely to be symmetrical
megan_gardner@apple.com [Tue, 30 Jan 2018 22:16:08 +0000 (22:16 +0000)]
Make preserve and restore focus more likely to be symmetrical
https://bugs.webkit.org/show_bug.cgi?id=182264
Source/WebKit:

<rdar://problem/36948473>

Reviewed by Tim Horton.

Keep a stack of if we actually increment the focusState, so that
changes to the web content do not result in asymmetric decrements to the focus state.
To work around problems associated with <rdar://problem/37000122>.

* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _restoreFocusWithToken:]):
(-[WKContentView _preserveFocusWithToken:destructively:]):

LayoutTests:

Reviewed by Tim Horton.

Added new test to verify that opening a selection form twice works.

* fast/forms/ios/ipad/select-form-run-twice-expected.txt: Added.
* fast/forms/ios/ipad/select-form-run-twice.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227855 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoStyleBench: Generate more mutation steps
antti@apple.com [Tue, 30 Jan 2018 22:08:11 +0000 (22:08 +0000)]
StyleBench: Generate more mutation steps
https://bugs.webkit.org/show_bug.cgi?id=182291

Reviewed by Ryosuke Niwa.

The test should be doing multiple rounds of mutations in a single document to test what is supposed to test (like it did before r227756).
This also makes test results more stable and less dependent on possible first-time costs.

* StyleBench/resources/benchmark-runner.js:

Adjust correctionFactor to match the new steps count. With 1.5 Firefox is ~72rpm.

* StyleBench/resources/tests.js:
(makeSteps):

Generate 10 rounds of class/element mutation steps.
Give each generated step unique name so reporting works.
Make individual steps do fewer mutations.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227850 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoUnified sources for FEMorphology.cpp seems to break the windows build.
commit-queue@webkit.org [Tue, 30 Jan 2018 21:55:12 +0000 (21:55 +0000)]
Unified sources for FEMorphology.cpp seems to break the windows build.
https://bugs.webkit.org/show_bug.cgi?id=178847

Patch by Fujii Hironori <Hironori.Fujii@sony.com> on 2018-01-30
Reviewed by Yusuke Suzuki.

Fix a compilation error.
platform/graphics/filters/SpotLightSource.cpp(45): error C2027: use of undefined type 'WebCore::FilterEffect'

No new tests (No behavior change)

* Sources.txt: Removed @no-unify for FEMorphology.cpp.
* platform/graphics/filters/SpotLightSource.cpp: Added #include "FilterEffect.h".

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227849 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoTest fix after r227776
mmaxfield@apple.com [Tue, 30 Jan 2018 21:51:18 +0000 (21:51 +0000)]
Test fix after r227776
https://bugs.webkit.org/show_bug.cgi?id=180951

Unreviewed.

* platform/graphics/cocoa/FontCacheCoreText.cpp:
(WebCore::mandatoryAttributesForUserInstalledFonts):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227848 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoCrashTracer: com.apple.WebKit.WebContent at WebCore: WebCore::Document::updateStyleIf...
dino@apple.com [Tue, 30 Jan 2018 20:29:02 +0000 (20:29 +0000)]
CrashTracer: com.apple.WebKit.WebContent at WebCore: WebCore::Document::updateStyleIfNeeded
https://bugs.webkit.org/show_bug.cgi?id=182299
<rdar://problem/36853088>

Removing a misleading comment based on Zalan Bujtas's feedback.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::repaintForPausedImageAnimationsIfNeeded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227846 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months ago[CMake] Make WTF headers copies
don.olmstead@sony.com [Tue, 30 Jan 2018 20:02:17 +0000 (20:02 +0000)]
[CMake] Make WTF headers copies
https://bugs.webkit.org/show_bug.cgi?id=182274

Reviewed by Alex Christensen.

Source/ThirdParty:

* gtest/CMakeLists.txt:

Source/WebDriver:

* CMakeLists.txt:

Source/WTF:

* wtf/CMakeLists.txt:
* wtf/PlatformGTK.cmake:
* wtf/PlatformJSCOnly.cmake:
* wtf/PlatformMac.cmake:
* wtf/PlatformWPE.cmake:
* wtf/PlatformWin.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227845 268f45cc-cd09-0410-ab3c-d52691b4dbfc

20 months agoCrashTracer: com.apple.WebKit.WebContent at WebCore: WebCore::Document::updateStyleIf...
dino@apple.com [Tue, 30 Jan 2018 19:46:38 +0000 (19:46 +0000)]
CrashTracer: com.apple.WebKit.WebContent at WebCore: WebCore::Document::updateStyleIfNeeded
https://bugs.webkit.org/show_bug.cgi?id=182299
<rdar://problem/36853088>

Reviewed by Simon Fraser.

Source/WebCore:

Mostly speculative fix for the case where a scrollTo moves an
animated SVG image into view, causing its animation to restart during
a paint operation. This was causing a release ASSERT, so we now defer
the resumption of the animation into a timer.

Test: svg/animated-svgImage-scroll.html

* rendering/RenderElement.cpp:
(WebCore::RenderElement::repaintForPausedImageAnimationsIfNeeded): Enqueue the
animation if it is an SVGImage.
* svg/graphics/SVGImage.cpp: Add a timer to enqueue animation starts.
(WebCore::SVGImage::SVGImage):
(WebCore::SVGImage::startAnimationTimerFired):
(WebCore::SVGImage::enqueueStartAnimation):
(WebCore::SVGImage::stopAnimation):
* svg/graphics/SVGImage.h:

LayoutTests:

* svg/animated-svgImage-scroll-expected.txt: Added.
* svg/animated-svgImage-scroll.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@227841 268f45cc-cd09-0410-ab3c-d52691b4dbfc