GitLab

BUG=682244

Review-Url: https://codereview.chromium.org/2717363003
Cr-Commit-Position: refs/heads/master@{#460648}

Varun owns rolling out a new storage layer for safe browsing, which is
currently taking over in stable.  So these should all be removed
soonish.

BUG=705809,701209

Review-Url: https://codereview.chromium.org/2787563002
Cr-Commit-Position: refs/heads/master@{#460647}

The initial CL leaked DOM objects by calling
documentElement()->lazyReattachIfAttached() during Document::shutdown(). This
reland advances m_lifecycle to DocumentLifecycle::Stopping slightly earlier in
Document::shutdown(), and then only triggers
documentElement()->lazyReattachIfAttached() if the state is earlier than
DocumentLifecycle::Stopping.

Partial stack of the DOM leak:
  ...
  documentElement()->lazyReattachIfAttached();
  blink::Document::willChangeFrameOwnerProperties()\n
  blink::HTMLFrameOwnerElement::setWidget()\n
  blink::FrameView::dispose()\n
  blink::Document::shutdown()\n
  blink::FrameLoader::prepareForCommit()\n
  blink::FrameLoader::commitProvisionalLoad()\n
  blink::DocumentLoader::commitIfReady()\n
  blink::DocumentLoader::processData()

> Most of this CL is plumbing for FrameOwnerProperties. Interesting changes:
>   * LayoutView can no longer have children if its FrameOwner is "DisplayNone".
>   * A local frame is "DisplayNone" if and only if it has no widget [which should
>     correspond to the actual display:none style].
>   * The DisplayNone property is propagated to remote frames via
>     FrameOwnerProperties.
>
> Spec discussion: https://github.com/whatwg/html/issues/1813

BUG=650433

Review-Url: https://codereview.chromium.org/2743053003
Cr-Commit-Position: refs/heads/master@{#460646}

The SmartText selection is controlled by ContextSelectionClient.
This class inherits from SelectionClient and is owned by
SelectionPopupController.

The processing is triggered by two events: the long press and the end
of the selection handle drag. The long press starts the "suggest and
classify" request that determines the type (meaning) of the selection
and might suggest a new selection boundaries. The end of drag event
starts a "classify" request that only determines the type of the selection.

Upon receiving a request ContextSelectionClient retrieves the
surrounding text by IPC that uses blink::WebSurroundingText API
and then calls ContextSelectionProvider to do the actual classification.

The classification result is delivered asynchronously to
SelectionPopupController which adjusts the existing selection by IPC
and updates the action mode menu.

BUG=685420

Review-Url: https://codereview.chromium.org/2740103006
Cr-Commit-Position: refs/heads/master@{#460645}

* Reset stream if HTTP/2 response ":status" header is missing,
  does not start with a number, or has extra text after the number.
* Modify unittests to document behavior.
* Obsolete histogram.

This is mostly a revert of https://crrev.com/2555183002.

BUG=672065

Review-Url: https://codereview.chromium.org/2779763005
Cr-Commit-Position: refs/heads/master@{#460644}

Add a D-Bus config file letting chronos own the
org.chromium.NetworkProxyService service name and letting
all processes make calls to it.

BUG=703217

Review-Url: https://codereview.chromium.org/2783823004
Cr-Commit-Position: refs/heads/master@{#460643}

TBR=kojii@chromium.org

Review-Url: https://codereview.chromium.org/2784943003
Cr-Commit-Position: refs/heads/master@{#460642}

The DownloadNotificationService tells the SystemDownloadNotifier to
unbind from it when it has no notifications to show and wants to shut
itself down.  This isn't needed because the SystemDownloadNotifier will
unbind when it is a safe time for the service to die.  This patch
removes the Observer method that tells the SystemDownloadNotifier to
unbind itself.

BUG=703212

Review-Url: https://codereview.chromium.org/2771353002
Cr-Commit-Position: refs/heads/master@{#460641}

Replacing - with _.
Chromium project prefers using _ for separating words in path names.

BUG=685886
TBR=kochi@chromium.org

Review-Url: https://codereview.chromium.org/2780363002
Cr-Commit-Position: refs/heads/master@{#460640}

PaymentOptions, suitable for use in components/core. This will allow
forthcoming profile filtering and deduplication options (and anything
else which depends on these values) to be written cross-platform, and
reduces the bleeding of Mojo into views logic that doesn't really need
to know about it.

Usage details:
- Desktop: This CL changes PaymentRequestSpec to implement this
  interface. It removes the options() accessor, which is no longer
  necessary.
- iOS: the iOS version of payment_request.h already has a PaymentOptions
  class which corresponds to, and can be trivially modified to
  implement, this interface.
- Android: A small utility class implementing this interface is
  forthcoming. The needed data is present as accessible members of
  PaymentRequestImpl.

See also crrev.com/2775553004 for further context. That CL will be
updated to pass PaymentOptionsProvider instead of PaymentRequestSpec.

BUG=706117

Review-Url: https://codereview.chromium.org/2774233005
Cr-Commit-Position: refs/heads/master@{#460639}

Currently ContentVerifyJob never sees DoneReading() for (legitimate
or not) 0-byte files. This means 0-byte files (legitimate or corrupted)
never goes through content verification process.
Override URLRequestJob::DoneReading() in URLRequestExtensionJob to
detect the end of 0-byte files so that ContentVerifyJob knows about
these 0-byte files.

Add a test to ensure that URLRequestExtensionJob will see 0-byte files.
ContentVerifyJobUnittest.DeletedAndMissingFiles already makes sure
that 0-byte read will fail content verification.

BUG=703888,703892
Test=With content verification turned on (e.g. with
--extension-content-verification=enforce_strict flag), install an
extension from webstore and modify a JS content inside that extension
to 0 byte file. Content verification should fail for that extension.
If the extension is part of system policy, the extension will be
reinsalled.

Review-Url: https://codereview.chromium.org/2779083005
Cr-Commit-Position: refs/heads/master@{#460638}

A few hundred crash reports in rendering path during execution of
AudioClock::ComputeBufferedMediaDurationMicros. Locks and various
contracts should protect us, so I assume one of the contracts is
being broken and hope to find out where.

AudioClock's buffered_ field is only modified during Render()
and during destruction. Perhaps its being destroyed while still
in-use.

BUG=674856

Review-Url: https://codereview.chromium.org/2780693004
Cr-Commit-Position: refs/heads/master@{#460637}

Add support for OfflineItems showing up in the DownloadNotificationService.
This consists of a few changes:
- Building a glue class between the OfflineContentAggregator and the
SystemDownloadNotifier.
- Adding some properties to DownloadInfo to support the properties
OfflineItem needs ('isTransient' and 'isOpenable').
- Update the SharedPrefs to store the isTransient property.

BUG=691805

Review-Url: https://codereview.chromium.org/2771803004
Cr-Commit-Position: refs/heads/master@{#460636}

On Android O we need the REQUEST_INSTALL_PACKAGE to be able to install
APKs.  Adding the permission to our manifest and adding it to be set by
default on developer builds.  Has no effect on pre-M builds.

BUG=706570

Review-Url: https://codereview.chromium.org/2779233004
Cr-Commit-Position: refs/heads/master@{#460635}

The respondWith() method is used by the payment app to provide a
PaymentAppResponse when the payment successfully completes.

This CL includes a content-side implementation and browser test. Especially, add
a callback to InvokePaymentApp() in PaymentAppProvider. The callback is used to
pass a response data from payment app using respondWith().

Related Spec Link:
  https://w3c.github.io/webpayments-payment-apps-api/#idl-def-paymentrequestevent

See the other CLs in this series:
  [1/3] https://codereview.chromium.org/2715663002/ (RespondWithObserver)
  [2/3] https://codereview.chromium.org/2705293010/ (blink side)
  [3/3] This patch

BUG=661608
TEST=payment_app_browsertest.cc

Review-Url: https://codereview.chromium.org/2718013004
Cr-Commit-Position: refs/heads/master@{#460634}

Cr-Commit-Position: refs/heads/master@{#460633}

BUG=701654

Review-Url: https://codereview.chromium.org/2782643004
Cr-Commit-Position: refs/heads/master@{#460632}

We need to do this before the calls to GetSvnRevision, because they run
svn.

BUG=706552
R=thakis@chromium.org,krasin@chromium.org

Review-Url: https://codereview.chromium.org/2784053003
Cr-Commit-Position: refs/heads/master@{#460631}

Now that the align-content and justify-content CSS properties are
adapted to the new baseline-position CSS values syntax we can adapt the
shorthand that controls such properties to the new syntax as well.

In addition, this patch makes public some of the utility functions
defined in the CSSPropertiesAlignmentUtils class, so that they can be
shared between the different parsing and computed style logic of the
different alignment related CSS properties.

BUG=703584

Review-Url: https://codereview.chromium.org/2762623002
Cr-Commit-Position: refs/heads/master@{#460630}

The tests are:
WebNavigationApiTest.ForwardBack
LoginPromptBrowserTest.SupplyRedundantAuths
LoginPromptBrowserTest.CancelRedundantAuths

I'll investigate these to see if they're real failures or just test issues.

BUG=705559
TBR=nasko@chromium.org

Review-Url: https://codereview.chromium.org/2783053003 .
Cr-Commit-Position: refs/heads/master@{#460629}

I'll investigate these to see if rr460581 made it more flaky.

BUG=651844,705559
TBR=nasko@chromium.org

Review-Url: https://codereview.chromium.org/2782363002 .
Cr-Commit-Position: refs/heads/master@{#460628}

BUG=702640

Review-Url: https://codereview.chromium.org/2787633002
Cr-Commit-Position: refs/heads/master@{#460627}

This CL adds an indicator for dragging and dropping bookmarks. This
indicator will highlight folders that can be dragged onto and add a line
above/below items for dragging between items. This affects both the list
and the sidebar, and bookmark elements in both are now draggable.

BUG=692843
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2746363013
Cr-Commit-Position: refs/heads/master@{#460626}

- Model file is not included, as it will be delivered through component updater.
- UI continues to work as before when the model file is missing.
- Textures included for each button state.
- Conform to DD UX specs pending.
- Refactored rendering-classes inheritance to be able to generalize them to non quads.

BUG=644562

Review-Url: https://codereview.chromium.org/2775283004
Cr-Commit-Position: refs/heads/master@{#460625}

Since the WebLocalFrame* parameter being passed is already available
from WebViewPlugin::WebViewHelper, these should always match. This is
pre-work to remove the (now redundant) parameter since it's already
available from inside the method.

BUG=361765

Review-Url: https://codereview.chromium.org/2778393002
Cr-Commit-Position: refs/heads/master@{#460624}

The single tap sent during the VR screen tap tests wasn't always
getting registered since device response can be a bit slow during
the VR transition. For the Cardboard case, wait on Javascript to say
it received another click event. For the Daydream case, wait on VrShell
saying its parent has consumed the event before checking.
BUG=

Review-Url: https://codereview.chromium.org/2768583002
Cr-Original-Commit-Position: refs/heads/master@{#460452}
Committed: https://chromium.googlesource.com/chromium/src/+/3e2ce51e88f19f0cc5196284f580583ddd4dbd9a
Review-Url: https://codereview.chromium.org/2768583002
Cr-Commit-Position: refs/heads/master@{#460623}

- On WARNING condition, request a backgrounded child process to purge memory
- On CRITICAL condition, try to discard a tab first. If it failed, try to
  purge memory from all child processes. It it failed, try to purge memory
  from the browser process
- Use somewhat conservative heuristics for purging
  - Browser: suppress purging for 2 mins after a purge
  - Renderer: after purging, suppress the next purging until the renderer
    goes foreground -> background then remains backgrounded for 30 secs

BUG=696844

Review-Url: https://codereview.chromium.org/2763933002
Cr-Commit-Position: refs/heads/master@{#460622}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/04f54bf6..dc1b150e

Please follow these instructions for assigning/CC'ing issues:
https://github.com/v8/v8/wiki/Triaging%20issues

Please close rolling in case of a roll revert:
https://v8-roll.appspot.com/
This only works with a Google account.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel

TBR=hablich@chromium.org,machenbach@chromium.org,kozyatinskiy@chromium.org,vogelheim@chromium.org

Review-Url: https://codereview.chromium.org/2784093002
Cr-Commit-Position: refs/heads/master@{#460621}

We've shrunk the OffscreenCanvas gpu pixel tests' width dimension in
https://codereview.chromium.org/2774423006/. That worked for all other
Android bots except Nexus 5. So we need to further shrink the width
dimension.

TBR=zmo@chromium.org
TBR_REASON=Test suppression on Nexus 5
NOTRY=true
BUG=706647
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2781183003
Cr-Commit-Position: refs/heads/master@{#460620}

Revert of macOS: Fix a bug in logic to get all memory regions. (patchset #1 id:1 of https://codereview.chromium.org/2786733004/ )

Reason for revert:
This failed ProcessMetricsMemoryDumpProviderTest.NoDuplicateRegions on "Mac ASan 64 Tests (1)".

https://findit-for-me.appspot.com/waterfall/failure?url=https://build.chromium.org/p/chromium.memory/builders/Mac%20ASan%2064%20Tests%20%281%29/builds/28499

Original issue's description:
> macOS: Fix a bug in logic to get all memory regions.
>
> The logic was accidentally double-counting the size of a memory region, causing
> some memory regions to get skipped.
>
> BUG=706547
>
> Review-Url: https://codereview.chromium.org/2786733004
> Cr-Commit-Position: refs/heads/master@{#460585}
> Committed: https://chromium.googlesource.com/chromium/src/+/f955ebf64b34ec9cbdcb137c00bc9bba1867d071

TBR=primiano@chromium.org,erikchen@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=706547

Review-Url: https://codereview.chromium.org/2788453002
Cr-Commit-Position: refs/heads/master@{#460619}

https://skia.googlesource.com/skia.git/+log/291932e8e42e..8f7a9a9bfb51

$ git log 291932e8e..8f7a9a9bf --date=short --no-merges --format='%ad %ae %s'
2017-03-29 msarett 565 color xforms: make sure we clamp if necessary
2017-03-29 mtklein jumper, revert to generating .S files

Created with:
  roll-dep src/third_party/skia


Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel
TBR=stephana@chromium.org

Change-Id: I24f4bc582a1958af5f5945ba574ded9bab5eea45
Reviewed-on: https://chromium-review.googlesource.com/462440Reviewed-by: Skia Deps Roller <skia-deps-roller@chromium.org>
Commit-Queue: Skia Deps Roller <skia-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#460618}

Change log:
https://github.com/google/closure-compiler/compare/b2f2d2f3309c7ee61816e068050651af27bccdfa...f9c738eb53fb5cd1d475871ef16c6d1cca19b815
chrome_extensions.js: a3f41820d1260b0d7ed13acac526b803ac6c1870 -> 04dad454b5797074b803c0bab27f1aa87e19d562

BUG=699302
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation
R=dbeam@chromium.org

Review-Url: https://codereview.chromium.org/2786793002 .
Cr-Commit-Position: refs/heads/master@{#460617}

Changes:
SystemTray::CloseSystemBubble() !activating_ is caught, reported in crbug.com/702267. Widget::IsClosed() indicates a widget is in process of closing, so we use this to do the check in ActivateBubble().
This is speculative fix for the bug.

BUG=702267
TEST=none

Review-Url: https://codereview.chromium.org/2785823002
Cr-Commit-Position: refs/heads/master@{#460616}

R=pkasting@chromium.org
BUG=691895

Review-Url: https://codereview.chromium.org/2783023002
Cr-Commit-Position: refs/heads/master@{#460615}

This CL updates a comment in in test_runner.h to reflect the fact that:
 - imported/ has been renamed to external/.
 - csswg-test/ is being moved to wpt/css/.

Change-Id: I158c6da888b9af8a123047ffe8369e3c88217792
Reviewed-on: https://chromium-review.googlesource.com/462340
Commit-Queue: Quinten Yearsley <qyearsley@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Reviewed-by: Peter Beverloo <peter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#460614}

This change emits a UKM PageLoad metric for the HTTP RTT estimate, and
one for the transport RTT estimate at navigation start.

BUG=706590

Review-Url: https://codereview.chromium.org/2780303002
Cr-Commit-Position: refs/heads/master@{#460613}

Now that a valid WebWidgetClient must always be passed when creating a
WebFrameWidget we can remove some of the redundant m_client == nullptr checks.

m_client can now become nullptr if close() is called on the WebFrameWidget,
but this is generally caught by checking m_layerTreeView.

To be safe if(m_client) checks have been replaced with DCHECK(m_client).

BUG=696895

Review-Url: https://codereview.chromium.org/2775113002
Cr-Commit-Position: refs/heads/master@{#460612}

The current code appeared to work because a DialogDelegate actually
defaults to focusing the default button ("End Task") rather than nothing
(like a default WidgetDelegate). Thus when you press Shift+Escape the
row for the active tab is selected, and "End Task" is focused. This
"fixed" the blue line around the table.

But on ChromeOS it's possible to have no windows open and press
Shift+Escape, and then the Browser row is selected initially. You're not
allowed to end that task, so the button is disabled.
Widget::SetInitialFocus has code to advance focus if the initial focus
View can't actually receive focus, hence the table winds up receiving
focus.

BUG=702489

Review-Url: https://codereview.chromium.org/2784903002
Cr-Commit-Position: refs/heads/master@{#460611}

This change adds a DiscardableHandle class, which acts as the primary
synchronization mechanism for the GPU Discardable Memory system.

This is the first in a sequence of CLs which implement the GPU
Discardable Memory system. This class is currently only created/used by
unit test code.

BUG=706456
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2732223004
Cr-Commit-Position: refs/heads/master@{#460610}

Add a (d)check in data reduction proxy (DRP) code which verifies that
if the CPAT header is present in the request headers, then ECT header
must be present too.

Other checks added: e.g., chrome-proxy header should only
be sent when the resolved proxy is a data saver proxy.

Finally, the CL also modifies few unittests to avoid tripping the
DCHECKs.

BUG=706650

Review-Url: https://codereview.chromium.org/2771413004
Cr-Commit-Position: refs/heads/master@{#460609}