GitLab

Whenever TileManager reassigns memory it should call
SetIsLikelyToRequireADraw on it's client. This happens when it checks if
there are more tiles that need to be rasterized after all raster tasks
are done. Furthermore when synchronously rasterizing tiles TileManager
should call SetIsLikelyToRequireADraw(false) on it's client. It's
necessary to update the client (LayerTreeHostImpl) otherwise it could
end up drawing again and again without any real work being done.

BUG=439275

Review URL: https://codereview.chromium.org/1035043002

Cr-Commit-Position: refs/heads/master@{#322637}

This includes a minor hack to use properties.additional_properties
directly in SetProperties and CreateConfiguraiton until we can add
proper types.

BUG=470262

Review URL: https://codereview.chromium.org/1038873004

Cr-Commit-Position: refs/heads/master@{#322636}

close() fails in UdpSocketLibevent. This change will allow to detect
if this problem is caused by memory corruption or not.

BUG=461246

Review URL: https://codereview.chromium.org/1036773003

Cr-Commit-Position: refs/heads/master@{#322635}

Allows the surfaceless path (used for Ozone/Freon) to queue up more
work while waiting for the currently displayed buffer to be released
by the hardware. In turn this means slightly better frame rates on
WebGL apps as we're able to better hit the vsync deadline.

BUG=467135
TESTS=Ran on stout with Freon and verified that the FPS on aquarium increased to similar values as X11

Review URL: https://codereview.chromium.org/1038263002

Cr-Commit-Position: refs/heads/master@{#322634}

Use a const ref to avoid the expensive operation of incrementing the ref
count of RefCountedThreadSafe VideoProviderClientImpl.

BUG=336733

Review URL: https://codereview.chromium.org/1039783002

Cr-Commit-Position: refs/heads/master@{#322633}

Some smartcards do not properly handle the output length (either bounds
checking or returning the output size) when the output buffer is non-NULL.
Work around those by calling twice despite already having a pre-allocated
buffer.

BUG=470204

Review URL: https://codereview.chromium.org/1037123002

Cr-Commit-Position: refs/heads/master@{#322632}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/939253e..135b7ec

CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
TBR=hcm@google.com

Review URL: https://codereview.chromium.org/1039363002

Cr-Commit-Position: refs/heads/master@{#322631}

Goal - Improve quality of service (QOS) in chromium when QUIC is
enabled. For some users when QUIC is enabled, their access to QUIC
based servers doesn't work well. This could be due to either high loss
rate while using UDP or service providers blacklisting ports 80/443 for
QUIC traffic.

+ One simple approach (which would detect breakage quickly) would be
  to look at the packet loss (received) when the crypto handshake
  is confirmed. When packet loss is "really bad", say over
  50% for 5 handshakes in a row, we will disable QUIC.

+ For a connection is packet loss is higher than the packet loss
  threshold, then we mark that port as recently broken. This would
  require crypto handshake confirmation.

+ Added two parameters - number of "lossy" connections (defaults to 0)
  and packet loss threshold (defaults to 1000%) params to determine
  when to disable QUIC. These parameters will be controlled by finch
  in the next CL.

R=rch@chromium.org, asvitkine@chromium.org,mpearson@chromium.org

Review URL: https://codereview.chromium.org/1025573002

Cr-Commit-Position: refs/heads/master@{#322630}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/bb055b9..53944e9

TBR=pkasting@chromium.org,yutak@chromium.org

Review URL: https://codereview.chromium.org/1039723003

Cr-Commit-Position: refs/heads/master@{#322629}

BUG=471098
TEST=Shaka player works.

Review URL: https://codereview.chromium.org/1042543002

Cr-Commit-Position: refs/heads/master@{#322628}

The buffering state change is the last callback fired by
VideoRendererImpl, therefore we should wait on this rather than the
paint callback.

BUG=470517

Review URL: https://codereview.chromium.org/1042473002

Cr-Commit-Position: refs/heads/master@{#322627}

This fixes middle-button scrolling on devices with no mousewheel. The
code is identical to existing code in the X11 build (xf86-input-cmt).

BUG=471286

Review URL: https://codereview.chromium.org/1037323002

Cr-Commit-Position: refs/heads/master@{#322626}

The message logging can be enabled by setting cvox.LibLouis.DEBUG to
true.

R=dtseng@chromium.org
BUG=none

Review URL: https://codereview.chromium.org/1028583002

Cr-Commit-Position: refs/heads/master@{#322625}

This CL adds if conditions to check if the notification comes from
the target BluetoothAdapter.

BUG=441581
TEST=device_unittests --gtest_filter=*BluetoothAudioSink*

Review URL: https://codereview.chromium.org/1038213004

Cr-Commit-Position: refs/heads/master@{#322624}

This patch enables javascript file injection for
<webview>.executeScript({file: ...}).

BUG=434081

Review URL: https://codereview.chromium.org/1004253002

Cr-Commit-Position: refs/heads/master@{#322623}

This cl is last partial cl from https://codereview.chromium.org/775143003/
for easy review.
Until now, BeginFrame is initiated by each renderer process.
With this cl, BeginFrame is scheduled from parent(browser) cc scheduler
to child(renderer) cc scheduler.
For more detailed information, please see this (http://goo.gl/D1Qxrr)

Also, --enable-begin-frame-scheduling is no longer needed because all platform except Mac
uses it as a default.

R=brianderson@chromium.org, danakj@chromium.org, mithro@mithis.com
BUG=372086

Review URL: https://codereview.chromium.org/1016033006

Cr-Commit-Position: refs/heads/master@{#322622}

base::TerminateBecauseOutOfMemory will call base::::OnNoMemory which is classified by
breakpad server as OOM crash.
We have a lot of variations of crashes which are expected OOM crashes.
Combining them into one class of crashes helps to discard them triaging crashes.

BUG=429763,460995

Review URL: https://codereview.chromium.org/1039043002

Cr-Commit-Position: refs/heads/master@{#322621}

On OS X the declaration of the close function looks like this:

int close(int) __asm("_close" optional extra stuff);

If we do not include unistd.h the close_nocancel.cc translation unit does not
see the asm label. This is technically an ODR violation, but not normally a
significant one, because close will get mangled to _close by the compiler
anyway. However under LLVM LTO the mangled symbol is internally given a
different name to the unmangled symbol, and the linker gets confused by the
multiple declarations with the same mangled name.

This also lets us simplify this code by defining the close function itself
(which will receive its name from the asm label in the declaration) rather
than trying to guess what the extra stuff is and naming the function we
define after that.

BUG=471146
R=thakis@chromium.org,mark@chromium.org

Review URL: https://codereview.chromium.org/1036123003

Cr-Commit-Position: refs/heads/master@{#322620}

BUG=470704

Review URL: https://codereview.chromium.org/1037043002

Cr-Commit-Position: refs/heads/master@{#322619}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/5c3405c..bb055b9

TBR=pkasting@chromium.org,yutak@chromium.org

Review URL: https://codereview.chromium.org/1042493002

Cr-Commit-Position: refs/heads/master@{#322618}

BUG=470741
R=iannucci@chromium.org

Review URL: https://codereview.chromium.org/1015633005

Cr-Commit-Position: refs/heads/master@{#322617}

A number of fixes to please clang_tidy.  There should be no functional
changes.

Putting the construction of a "null" SerializedPacket into a function
rather than constructing one into a local variable everywhere one is
needed.

Renaming class members to end with "_".  Taking the initial "k" off
local variables.

Merge internal change: 89397417
https://codereview.chromium.org/1039473003/

Move test for incoming packets with incorrect connection ID from
QuicConnection::OnPacketHeader to ::OnUnauthenticatedPublicHeader so it
works.

Add tests to verify code handles incoming packets with invalid
connection IDs.

Add DCHECKs to verify that QuicConnection::OnPublicResetPacket and
::OnVersionNegotiationPacket don't get packets with invalid
connection IDs.

Add comment to QuicConnection::ProcessUdpPacket to note packet may have
incorrect connection ID.

Fix b/19710533.

Merge internal change: 89336249
https://codereview.chromium.org/1038293002/

Fix a flaky QUIC end to end test by supplying the correct buffer size as
the initial default.

Merge internal change: 89220759
https://codereview.chromium.org/1038223002/

Initialize slowstart_packets_sent.

Avoid uninitialized memory.

Merge internal change: 89126237
https://codereview.chromium.org/1036123002/

QUIC - minor cleanup while merging the following change.

In Chromium's AddToCache function, we use "www.google.com" instead of
empty path as it was done in the internal CL. All other changes were
already merged by rch.

Make some changes to the QuicInMemory server to clarify the interface.
Specifically that the cache is indexed by host and path, not by method
nor scheme, nor full URL, nor request headers, etc. In preparation for
making the chrome version more similar.

Merge internal change: 89092881
https://codereview.chromium.org/1038203002/

R=rch@chromium.org

Review URL: https://codereview.chromium.org/1041853002

Cr-Commit-Position: refs/heads/master@{#322616}

This change removes SetDrawViewport, instead tracking the current window
viewport in DirectRenderer.

Review URL: https://codereview.chromium.org/1035863002

Cr-Commit-Position: refs/heads/master@{#322615}

Merge internal change: 89221001

R=rch@chromium.org

Review URL: https://codereview.chromium.org/1037133002

Cr-Commit-Position: refs/heads/master@{#322614}

Update tests so they check for formatted exceptions (so they will catch
a bug such as this one).

BUG=444240

Review URL: https://codereview.chromium.org/1041843002

Cr-Commit-Position: refs/heads/master@{#322613}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/1e37d17..939253e

CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
TBR=hcm@google.com

Review URL: https://codereview.chromium.org/1043493002

Cr-Commit-Position: refs/heads/master@{#322612}

You can't assume that there's a root layer always during layout, even if
there's a composite and readback.  WebTestProxyBase::CapturePixelsAsync
iterates through popups (which may get closed between that call and
Layout).  If it's closed, then its root layer goes away.

BUG=470817

Review URL: https://codereview.chromium.org/1038133003

Cr-Commit-Position: refs/heads/master@{#322611}

Fixes hitting the maximum connection limit of 6 per domain when a
media tag is used with preload=metadata.

Deferred network connections will be cancelled once reaching the
HAVE_ENOUGH_DATA state for preload=metadata, non-streaming
resources that have not started playback.

A soft cancel is introduced which allows us to preserve the data
cached thus far in the BufferedResourceLoader. Once this cache is
exhausted the loader will fulfill subsequent reads with a cache
miss and continue on as normal.

preload=metadata intentionally does not cache enough data for a
smooth playback experience on slow connections, so the cache miss
here does not increase jank relative to the current status quo.

For similar reasons, since preload=auto does load enough data for
an uninterrupted playback, we can't use this method there or we'll
introduce jank during the cache miss and subsequent catch up.

Note: This does not resolve hung connections if multiple media
elements are actually playing back or if they are served from
a server without range requests.

Note 2: This also fixes a broken media_blink_unittests, which was
missing some Gin dependency.

BUG=162627
TEST=new unittest, passes layout tests.

Review URL: https://codereview.chromium.org/1029763002

Cr-Commit-Position: refs/heads/master@{#322610}

Local profiling suggests that converting std::set to std::vector will save
around 10% of the overall startup time, likely due to the reduced allocs
involved with std::vector.

This change also reduces lookups by traversing the incoming dictionary
and iterating on the available values instead of looking through all
values for all features.

Finally, cleaned up and enforced the testing accessors via code and friend
declarations

BUG=460987

Review URL: https://codereview.chromium.org/1010973013

Cr-Commit-Position: refs/heads/master@{#322609}

The extension profile generator and its subclasses have been in various states
of broken for quite some time. Removing them entirely.

BUG=444230

Review URL: https://codereview.chromium.org/1036373002

Cr-Commit-Position: refs/heads/master@{#322608}

Use the captured description of the touchscreen for Chromebook Pixel to
initialize the mock touchscreen, instead of just faking a couple of
the members and leaving the rest uninitialized.

The press, release, & move tests are merged into one test using real
captured data from the Pixel touchscreen. The other tests just got
their "pressure" values rescaled to the new axis.

Now we can add a test that pases in a non-MT device description.

BUG=461518
TEST=events_unittests

Review URL: https://codereview.chromium.org/1019803006

Cr-Commit-Position: refs/heads/master@{#322607}

Test= FakeVideoCaptureDeviceTest in media_unittests

Review URL: https://codereview.chromium.org/1033733009

Cr-Commit-Position: refs/heads/master@{#322606}

The CL was previously reverted in crrev.com/322355 because it broke downstream
android builds. This is a straight-forward reland of the original issue. A fix
will be landed for downstream once this CL lands.

Original issue's description:
> > precache: Move the java files into //components/precache/android
> >
> > This moves code out of //java/apps/chrome/.../precache into:
> >   . //components/precache/android/: classes that provide information about
> >     device-state, network connectivity etc.
> >   . //chrome/android/.../precache/: precache service and service launchers.
> >
> > BUG=401552

TBR=bengr@chromium.org,yfriedman@chromium.org,thestig@chromium.org,alexclarke@chromium.org
NOPRESUBMIT=true
BUG=401552

Review URL: https://codereview.chromium.org/1036363002

Cr-Commit-Position: refs/heads/master@{#322605}

This tool is based on telemetry and exercises the Smart Lock setup flow. This CL
also introduces the framework that will allow implementing other end-to-end
tests.

BUG=chromium:442728
NOTRY=true

Review URL: https://codereview.chromium.org/1004283002

Cr-Commit-Position: refs/heads/master@{#322604}

Upper case letters for all words is not Android style.

BUG=NONE

Review URL: https://codereview.chromium.org/1029423011

Cr-Commit-Position: refs/heads/master@{#322603}

Reland of "Decrease the lag when switching between different categories in the Cros wallpaper selector".

The original one (https://codereview.chromium.org/1028513003) caused failures in the cros_trunk official builder (see http://master.chrome.corp.google.com:8011/builders/cros%20trunk/builds/31816 for detail).

The issue should fix the broken tests.

BUG=454932

Review URL: https://codereview.chromium.org/1038543002

Cr-Commit-Position: refs/heads/master@{#322602}

BUG=470453

Review URL: https://codereview.chromium.org/1039543003

Cr-Commit-Position: refs/heads/master@{#322601}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/7bc68ec..5c3405c

TBR=pkasting@chromium.org,yutak@chromium.org

Review URL: https://codereview.chromium.org/1039233003

Cr-Commit-Position: refs/heads/master@{#322600}

These are currently DLOG() but we really shouldn't be hiding them. Also
use PLOG() when we know errno will be set.

BUG=none
TEST=build

Review URL: https://codereview.chromium.org/1032233002

Cr-Commit-Position: refs/heads/master@{#322599}

The test appears to be flaky on
https://uberchromegw.corp.google.com/i/clank/builders/instrumentation-hammerhead-clankium-hera

BUG=471211
TBR=peter@chromium.org

Review URL: https://codereview.chromium.org/1036353003

Cr-Commit-Position: refs/heads/master@{#322598}