GitLab

Upstream all BUILD.gn changes to enable compilation of the code
that has been upstreamed by the previous ten CLs and changes to
browser_state_keyed_service_factories.mm to use the new factory.

Update the default provider factory to use chromium provider (it
will be merged with ChromeBrowserProvider eventually).

BUG=653086
R=pinkerton@chromium.org, rohitrao@chromium.org

Committed: https://chromium.googlesource.com/chromium/src/+/1e23576749d95a02d22b2ff4277eaf132dce4831
Review-Url: https://codereview.chromium.org/2589843002
Cr-Commit-Position: refs/heads/master@{#439601}

This CL makes the code in MediaPermission go through
PermissionManager, which it didn't before (there was a TODO for this). It also slightly simplifies the logic in PermissionManager for querying the permission status.

Review-Url: https://codereview.chromium.org/2579593002
Cr-Commit-Position: refs/heads/master@{#439600}

This case can happen when UrlRequest.start() throws an exception.

BUG=675629
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_cronet_tester

Review-Url: https://codereview.chromium.org/2585203003
Cr-Commit-Position: refs/heads/master@{#439599}

BUG=

Review-Url: https://codereview.chromium.org/2588093002
Cr-Commit-Position: refs/heads/master@{#439598}

CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2584993002
Cr-Commit-Position: refs/heads/master@{#439597}

BUG=none

Review-Url: https://codereview.chromium.org/2592483002
Cr-Commit-Position: refs/heads/master@{#439596}

These changes were suggested by tengs@ in https://codereview.chromium.org/2564653004/ after it had already been submitted.

BUG=672263

Review-Url: https://codereview.chromium.org/2587783003
Cr-Commit-Position: refs/heads/master@{#439595}

VPX MemoryPool is refcounted and it'll be destroyed by
by whatever thread holds a reference last.

Currently registering the memorypool dump provider happens
in the memorypool ctor while unregistering it happens in the dtor.

This can cause problems since unregistering it should happen
on the same thread where it was registered first.

This CL moves the register/unregister of memory pool dump
provider to VpxVideoDecoder that is always called from media thead.

BUG=637409

Review-Url: https://codereview.chromium.org/2580223002
Cr-Commit-Position: refs/heads/master@{#439594}

In the case of not accepting all devices in the options that are given
to WebBluetooth requestDevice(), knowing the number of devices that are
in the chooser is helpful for researching some methods to distinguish
devices with the same name. This CL adds UMA for the number of devices
that are in the chooser when a device is paired.

BUG=674195

Review-Url: https://codereview.chromium.org/2577183002
Cr-Commit-Position: refs/heads/master@{#439593}

This was caused because some inspector protocol messages were being dropped with PlzNavigate with cross-process navigations. The scenario is that
1) FrameHostHolder::DispatchProtocolMessage was being called before the navigation happens. It adds the message to sent_messages_.
2) AboutToNavigate(NavigationHandle* navigation_handle) is called
3) FrameHostHolder::SendMessageToClient is called since the first renderer has responded to the protocol message from 1). It removes the sent message from sent_messages_ and forwards the reply using SendMessageToClient.
4) ReadyToCommitNavigation is called, which only then creates the pending FrameHostHolder
5) DidFinishNavigation is called which sets current_ to pending_.

This manifested in hangs with telemetry, since in cross-process navigations the Page.navigate command was being recieved incorrectly in the first renderer and it was sending a frameId from that renderer. The catapult code was waiting for a navigation event for that frameId which would never come (it would get it from the new process instead).

There were two bugs here with PlzNavigate, compared to the non-PlzNavigate case.
1 The first FrameHostHolder (i.e. current_ during the navigation) was not being paused. This was leading to responses to messages that were sent to the first renderer to respond to the inspector client. Since the response arrived, sent_messages_ was being cleared and we never got to send the message to the new renderer.
2) Even with the above fixed, we had to keep track of messages that were sent to the first renderer while the FrameHostHolder is (now) suspended. This didn't need to happen in the non-PlzNavigate case because the pending FrameHostHolder is created immediately (in the equivalent to step 2 above). At that point in time, SetPending ends up calling Reattach which then grabs the sent_messages_ from the current_ FrameHostHolder. The Page.navigate message was therefore being sent correctly to the second process. When the first process replies, it's ignored since it goes into pending_messages_ buffer. With PlzNavigate, there's a period of time between 2) and 4) where we know we have a navigation but we don't have the RenderFrameHost yet. So we need to emulate the non-PlzNavigate behavior by holding on to sent messages during that window, so that we can send them to the new FrameHostHolder once it commits.

This fixes telemetry_unittests' testWebPageReplay, gpu_process_launch_tests, hardware_accelerated_feature_tests and the page cyclers with PlzNavigate.

BUG=674730
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2585053002
Cr-Commit-Position: refs/heads/master@{#439592}

Apply an additional overlay to the content_browser manifest, which will allow
internal Chromecast builds to host embedded services and interfaces in the
"content_browser" service.

BUG=internal b/33584507
TEST=All services connect correctly

Review-Url: https://codereview.chromium.org/2562393005
Cr-Commit-Position: refs/heads/master@{#439591}

Contain: paint caused us to lose subpixel anti aliasing in many cases.
Removing it doesn't seem to cause any performance issues.

BUG=none

Review-Url: https://codereview.chromium.org/2583943003
Cr-Commit-Position: refs/heads/master@{#439590}

R=rtoy@chromium.org

Review-Url: https://codereview.chromium.org/2585683002
Cr-Commit-Position: refs/heads/master@{#439589}

We can't fully imitate what CSS would do for a box shadow
on a small box without redrawing the shadow for every height
below a certain threshold. Instead, simply reduce the "elevation"
of the shadows (which controls blur and offset) to something
we can handle.

This only has any effect on active windows that are shorter or
narrower than 100dip, and it maintains a distinction between active
and inactive all the way down to 36dip. It gets rid of the unsightly
seam that our current strategy causes. Sebastien has agreed to this
mitigation (although he might have additional feedback when he plays
with it live, once it's on canary).

BUG=608852

Review-Url: https://codereview.chromium.org/2586653002
Cr-Commit-Position: refs/heads/master@{#439588}

This is the first part of a larger migration of some parts of proximity_auth
into cryptauth.

BUG=672263

Review-Url: https://codereview.chromium.org/2561203002
Cr-Commit-Position: refs/heads/master@{#439587}

Create directories for basic WebAudio nodes and objects and move the
tests into the appropriate directory.

Files were just moved, with automatic update of script paths using

for f in $*
do
  sed -i 's;\(<script.*\)src=";\1src="../;' $f
done

and a few files were manually updated to get appropriate paths for
some audio resources:

AudioBufferSource/audiobuffersource.html
decodeAudioData/decode-audio-data-basic.html
OfflineAudioContext/offlineaudiocontext-detached-no-crash.html
Panner/hrtf-database.html
ScriptProcessor/scriptprocessornode-detached-no-crash.html

and the following needed to be updated due to the path change:

decodeAudioData/decode-audio-data-basic-expected.txt

BUG=647278
TEST=All layout tests still pass

Review-Url: https://codereview.chromium.org/2581463002
Cr-Commit-Position: refs/heads/master@{#439586}

TBR=oshima@chromium.org
BUG=522168

Review-Url: https://codereview.chromium.org/2588923002
Cr-Commit-Position: refs/heads/master@{#439585}

If the floating point number in the matrix can not be represented as
an integer then we should return false so that the calling code does
not try to cast it as one.

For large enough integers this may return an incorrect result still, as
it may return true if the floating point representation has a
fractional part if that floating point representation also is used for
an integer converted to float. But at that point we can not tell which
would be closer to the original intent - using the integer-casted
number or the floating point one, so we leave the result as before.

R=ajuma@chromium.org
BUG=675435

Review-Url: https://codereview.chromium.org/2585263003
Cr-Commit-Position: refs/heads/master@{#439584}

BUG=669682

Review-Url: https://codereview.chromium.org/2583503002
Cr-Commit-Position: refs/heads/master@{#439583}

This CL helps to consistently blacklist across DeclRewriterBase and
UnresolvedRewriterBase (and therefore to avoid/fix
https://crbug.com/584538) by stopping to blacklist "trace" and
"traceImpl" method names.

We will put together a separate changelist that tweaks the Oilpan's
clang plugin so it will be aware of the new name (|Trace|).  This
change can be probably rolled into https://crrev.com/2329463004.

This CL also tweaks a few other remotely related things:

- Unresolved nodes should never resolve to a function, therefore
  we shouldn't be blacklisting "swap" for unresolved nodes
  (and we can rename the function to IsBlacklistedMethodName).

- IsBlacklistedMethodName stops blacklisting "disable".
  - This enables reusing IsBlacklistedMethodName from IsBlacklistedMethod.
  - Blacklisting "disable" for unresolved nodes was always a heuristic
    (i.e. we cannot tell whether the unresolved method overrides
    blink::InspectorAgent::disable).  Flipping this heuristic around
    seems okay.

BUG=584538

Review-Url: https://codereview.chromium.org/2580773003
Cr-Commit-Position: refs/heads/master@{#439582}

https://chromium.googlesource.com/external/github.com/catapult-project/catapult.git/+log/7ca359550f95..38e97e25679c

$ git log 7ca359550..38e97e256 --date=short --no-merges --format='%ad %ae %s'
2016-12-19 jbudorick [devil] Add a constant for forwarder device dynamic port allocation.
2016-12-19 wangyix Apply changes made to net-internals after chromium commit 1be518f2...  to netlog-viewer
2016-12-19 alexandermont Make the atrace agent use the correct version of ADB.

BUG=567947

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.chromium.android:android_optional_gpu_tests_rel
TBR=catapult-sheriff@chromium.org

Review-Url: https://codereview.chromium.org/2590643002
Cr-Commit-Position: refs/heads/master@{#439581}

BUG=665242

Review-Url: https://codereview.chromium.org/2534673002
Cr-Commit-Position: refs/heads/master@{#439580}

No changes other than modifying includes and appropriately forward
declaring the class.

BUG=625684
TBR=dcheng@chromium.org, jochen@chromium.org

Review-Url: https://codereview.chromium.org/2581943003
Cr-Commit-Position: refs/heads/master@{#439579}

Previously we were responding to OnSyncConfigurationCompleted by
invoking our callback to regenerate suggestions. This covered initial
sync setup, along with enabling or disabling any types from sync, in
case Open Tabs was one of the types that was changing. However, when
sync is completely disabled or the user signs out, we don't perform an
configuration, but session data is no longer there. We want to clear
our suggestions, but the hook we were using isn't called.

The approach this CL takes is to switch to using OnStateChanged for
this tracking. This event is called extremely often, for a variety of
things. Because generating suggestions when there are actual sessions
data is not computationally trivial, we only want to invoke the
callback in TabDelegateSyncAdapter when it is reasonable to do so, so
we've added a bool state variable to track the previous state, and only
invoke the callback when we're transitioning between states.

Note that this change does not mean that the current or any existing
NTPs are going to remove their foreign sessions data. By having the
ForeignSessionsSuggestionsProvider provide new suggestions, this
current only seems to affect NTPs that are opened afterwards.

BUG=669041

Review-Url: https://codereview.chromium.org/2578293002
Cr-Commit-Position: refs/heads/master@{#439578}

In this case the alignment enum value is SHELF_ALIGNMENT_BOTTOM_LOCKED.

BUG=666625

Review-Url: https://codereview.chromium.org/2589493007
Cr-Commit-Position: refs/heads/master@{#439577}

BUG=

Review-Url: https://codereview.chromium.org/2578363003
Cr-Commit-Position: refs/heads/master@{#439576}

Using a bug as an opportunity to experiment with TaskScheduler (as well
as confirm the issue is what we think it is: a priority inversion where
a sync IPC from the renderer is blocking on a response from the
BlockingPool).

Only logical difference is that TaskScheduler doesn't have a notion of
named sequences and as such every DOMStorageContextWrapper will have its
own primary and commit sequence instead of sharing one with other
DOMStorageContextWrapper. If anything that's a good thing as it allows
multiple profiles to do their DOMStorage ops in parallel? Don't think
this can conflict..?

BUG=665588, 622400

Review-Url: https://codereview.chromium.org/2576243003
Cr-Commit-Position: refs/heads/master@{#439575}

It's technically more correct to call IsHorizontalAlignment. None of
these cases actually created a bug in practice as tray items aren't
created while the screen is locked and the logout button doesn't show
on locked screens. However, they serve as bad examples in the codebase.

For an example where this difference did actually create a bug, see
crbug.com/666625

BUG=none

Review-Url: https://codereview.chromium.org/2588893002
Cr-Commit-Position: refs/heads/master@{#439574}

BUG=670843
R=qyearsley@chromium.org,dpranke@chromium.org

Review-Url: https://codereview.chromium.org/2592473002
Cr-Commit-Position: refs/heads/master@{#439573}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/af5c10c8..6cdbfc6d

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

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

Review-Url: https://codereview.chromium.org/2588643003
Cr-Commit-Position: refs/heads/master@{#439572}

This CL adds support for an als sensor that has "in_intensity_both_raw" naming.
What is more, frequency and offset file names are added as well.

Unittests were changed as well - there is no need to separately delete
a frequency file now in order to receive OnSensorReadingChanged event. A zero
value can be used which means "file must not be created"

BUG=670602

Review-Url: https://codereview.chromium.org/2571613003
Cr-Commit-Position: refs/heads/master@{#439571}

Each BLE device has a limited number of BLE advertisement slots; this class queues devices to which to advertise and provides functionality for moving devices to the end of the queue and requesting the devices which should be advertised.

BUG=672263

Review-Url: https://codereview.chromium.org/2576283002
Cr-Commit-Position: refs/heads/master@{#439570}

This can meature usage rate more accurately.
Count the number of submissions with autofill suggestion shown.

BUG=674678
TEST=AutofillMetricsTest

Review-Url: https://codereview.chromium.org/2575263002
Cr-Commit-Position: refs/heads/master@{#439569}

https://skia.googlesource.com/skia.git/+log/d1c550e71fa7..65869fb64b56

$ git log d1c550e71..65869fb64 --date=short --no-merges --format='%ad %ae %s'
2016-12-19 reed update dox for generateScaledPixels
2016-12-19 robertphillips Guard some uint32_t -> int cases in SkPictureData reading

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
TBR=rmistry@google.com

Review-Url: https://codereview.chromium.org/2590603002
Cr-Commit-Position: refs/heads/master@{#439568}

Build all dependencies of //ios:all on iOS builders. This includes
all the test targets (including the ones that are not yet run as
they first need to be enabled on FYI bots, and the main app that
is not a test target).

BUG=653086

Review-Url: https://codereview.chromium.org/2587123002
Cr-Commit-Position: refs/heads/master@{#439567}

The intent might be ignored if the screen is locked and user just input
pin to unlock it. Instead of simply discard the intent, add a deferred
intent and hold it for a short period of time. Once the USER_PRESENT has
been broadcast, check back the deferred intent and move forward with it.

The other option to handle this scenario would be set the intent aside,
and start an async task to continuously for the screen status, and react
once the screen unlocked. This would be less optimal option since it
requires extra background task while screen is locked.

BUG=455126
TEST=out/Default/bin/run_chrome_public_test_apk -f \*IntentHandlerTest\*
     also verified by file an intent when screen is locked and then
     unlcok screen with pin.

Review-Url: https://codereview.chromium.org/2568843003
Cr-Commit-Position: refs/heads/master@{#439566}

BUG=673520

Review-Url: https://codereview.chromium.org/2586203004
Cr-Commit-Position: refs/heads/master@{#439565}

macOS sierra provides font that are downloaded and placed in a location
separate from the well known /Library/Fonts and
/System/Library/Fonts. We need to add this sandbox option to
dynamically allow access to those fonts.

BUG=662686

Review-Url: https://codereview.chromium.org/2510923003
Cr-Commit-Position: refs/heads/master@{#439564}

BUG=none
R=eugenebut@chromium.org

Review-Url: https://codereview.chromium.org/2585293003
Cr-Commit-Position: refs/heads/master@{#439563}

In XMPP, we rely on the XMPP channel to guarantee the integrity of the
from attribute.  To prevent an attacker from bypassing the from field
check using the LCS channel, we should only handle messages from the
LCS channel from the remoting bot.

BUG=669637

Review-Url: https://codereview.chromium.org/2568133003
Cr-Commit-Position: refs/heads/master@{#439562}