GitLab

BUG=416941

Review URL: https://codereview.chromium.org/598673002

Cr-Commit-Position: refs/heads/master@{#296271}

1. Renamed FakeSocket to FakeStreamSocket and FakeUdpSocket to
   FakeDatagramSocket,
2. Added factories for both stream and datagram fake
   sockets,
3. Move fake sockets to separate files out of fake_session.h

BUG=402993

Review URL: https://codereview.chromium.org/580243003

Cr-Commit-Position: refs/heads/master@{#296270}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/562c87e..b246733

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/597703002

Cr-Commit-Position: refs/heads/master@{#296269}

Increase the #of pings from 9 to 36 for Dev (72 secs of
unresponsiveness).

Increased it to reduce the number of crash dump reports
we receive. Will revert back, when we can auto triage
jank on UI/IO threads.

BUG=352489
R=sky@chromium.org

Review URL: https://codereview.chromium.org/588373002

Cr-Commit-Position: refs/heads/master@{#296268}

This change prepares DeviceInfo class (and all other sync
classes that depend on DeviceInfo and implement DEVICE_INFO
syncable type) to be moved to components/sync_driver library
by removing any dependencies on chrome/browser.

To achieve this all chrome/browser specific code was moved to
LocalDeviceInfoProviderImpl which will stay in browser/sync/glue.
All other classes including DeviceInfo and LocalDeviceInfoProvider
should now depend only on base or components. I'll move
them to components/sync_driver in the next change.

Details:

1) In ui_thread_search_terms_data.cc DeviceInfo::GetLocalDeviceType()
was used just to find out if the device is an Android phone.
Since this code has nothing to do with Sync I replaced these
with direct calls to ui::GetDeviceFormFactor() - that exactly
what DeviceInfo::GetLocalDeviceType() does internally.

2) In sessions_sync_manager.cc - calling DeviceInfo::GetLocalDeviceType()
isn't necessary anymore because the local device type can be
accessed via the owned LocalDeviceInfoProvider instance.
This allowed me to move GetLocalDeviceType() to LocalDeviceInfoProviderImpl
and make it private.

3) DeviceInfo::MakeUserAgentForSyncApi() is called only
from the code that will stay in browser/sync/glue.
I moved the implemented to LocalDeviceInfoProviderImpl class.

BUG=396136

Review URL: https://codereview.chromium.org/594643003

Cr-Commit-Position: refs/heads/master@{#296267}

BUG=none
TBR=pkasting@chromium.org

Review URL: https://codereview.chromium.org/597433002

Cr-Commit-Position: refs/heads/master@{#296266}

, because ThreadSanitizer v2 doesn't need base::subtle to be annotated.
We can't drop the dependency on libdynamic_annotations yet, because some code calls RunningOnValgrind().
Also remove the unused dynamic_annotations.h headers from base/ and add one to content/test/, which uses RunningOnValgrind().

BUG=349861
R=ajwong@chromium.org,phajdan.jr@chromium.org

Review URL: https://codereview.chromium.org/580813002

Cr-Commit-Position: refs/heads/master@{#296265}

R=estade@chromium.org
BUG=none

Review URL: https://codereview.chromium.org/596883002

Cr-Commit-Position: refs/heads/master@{#296264}

Files moved and includes adjusted.

BUG=387072

Review URL: https://codereview.chromium.org/591443002

Cr-Commit-Position: refs/heads/master@{#296263}

BUG=416948

Review URL: https://codereview.chromium.org/600583002

Cr-Commit-Position: refs/heads/master@{#296262}

* Implement detachGLFunctor. The contract here is after detachGLFunctor
  is called, DrawGL is no longer called until the next requestDrawGL.
  Implement just with a mFunctorAttached variable.
* Only create one HardwareView. Each GLSurfaceView creates its own GL
  thread, but we only assume a single render thread. Using SurfaceView
  or hosting multiple AwContents in the same GLSurfaceView is too hard,
  so just do the simplest thing, and create the first view has hardware
  accelerated.

BUG=416981

Review URL: https://codereview.chromium.org/594333002

Cr-Commit-Position: refs/heads/master@{#296261}

This was being dropped on the floor.

Review URL: https://codereview.chromium.org/597683002

Cr-Commit-Position: refs/heads/master@{#296260}

BUG=371692

Review URL: https://codereview.chromium.org/595633002

Cr-Commit-Position: refs/heads/master@{#296259}

BUG=416721
R=brettw@chromium.org

Review URL: https://codereview.chromium.org/593973002

Cr-Commit-Position: refs/heads/master@{#296258}

R=yfriedman@chromium.org
BUG=None

Review URL: https://codereview.chromium.org/599553002

Cr-Commit-Position: refs/heads/master@{#296257}

Added checks for missing arguments in event_sender.cc. Also simplified the uses of x & y in EventSender::GestureEvent.

BUG=415970

Review URL: https://codereview.chromium.org/584203003

Cr-Commit-Position: refs/heads/master@{#296256}

This CL moves RulesRegistryService to //extensions. Since it was dependent on ContentRulesRegistry, we move the essential elements of that interface to //extensions as well, renaming what remains as ChromeContentRulesRegistry.

BUG=352293

Review URL: https://codereview.chromium.org/590573002

Cr-Commit-Position: refs/heads/master@{#296255}

This avoids the majority of the Chromium builders attempting to build
these isolates.

BUG=

Review URL: https://codereview.chromium.org/595663002

Cr-Commit-Position: refs/heads/master@{#296254}

This allows the cast extension to turn off wifi scan while
streaming is in progress, and also set the "media streaming mode"
further experimentation is need to see if these options are actually
useful.

BUG=410500

Review URL: https://codereview.chromium.org/566243005

Cr-Commit-Position: refs/heads/master@{#296253}

On the windows side (and likely chromeos for some devices) we can end
up in a situation where the pixel bounds is not an even multiple of the
scale factor. This means when TiledLayer calculates the scale it ends
up off.

BUG=416655,416775
TEST=added test coverage
R=danakj@chromium.org

Review URL: https://codereview.chromium.org/567743003

Cr-Commit-Position: refs/heads/master@{#296252}

Only whitelisted webUIs could embed webview tag, specifically, they have to be declared in _api_features.json and they have to relax CSP to allow internal browser plugins.

BUG=

Review URL: https://codereview.chromium.org/572633003

Cr-Commit-Position: refs/heads/master@{#296251}

BUG=414766

Review URL: https://codereview.chromium.org/587243004

Cr-Commit-Position: refs/heads/master@{#296250}

When an Android View is invalidated outside of frame dispatch
(animation/input/draw), a sync barrier may be inserted into the shared
UI thread message loop. This effectively blocks dispatch of all Chrome
tasks posted to the browser UI thread for an entire frame (or more if
the View is continually invalidated).

There are currently no easy or even automated ways to avoid this
untimely invalidation, and judicious coding can only take us so far. As
an intermediate assist in debugging the issue, use reflection and a
crude form of MessageQueue inspection to trace known cases where the
MessageQueue is blocked by a sync barrier.

Note that this detection is not perfect, neither is it exact, but it's a
solid proxy for informing traces about such pipeline stalls.

BUG=407133

Review URL: https://codereview.chromium.org/575103002

Cr-Commit-Position: refs/heads/master@{#296249}

ChannelMojo doesn't work on Windows with existing implementaion
and this CL fixes it.

On Windows, ChannelHandle isn't immediately usable: The handle has
to be activated through ConnectNamedPipe() windows API, which is done in
its own Connect() handlshaking phase. ChannelMojo didn't Connect()
underlying channel and took the ChannelHandle over so the handle
wasn't activated.

Instead of hijacking underlying ChannelHandle, this CL actually Connect()s
underlying channel, creates a pipe on the server side, send one side of the
pipe to the client process, and use the pipe for the MessagePipe
initialization.

These initialization task is encapsulated behind new MojoBootstrap class.
ChannelMojo creates MojoBootstrap class to get the PlatformHandle which
is already activated and usable.

BUG=377980
TEST=ipc_mojo_bootstrap_unittest.cc, ipc_channel_mojo_unittest.cc
R=viettrungluu@chromium.org, darin@chromium.org, yzshen@chromium.org

Review URL: https://codereview.chromium.org/553283002

Cr-Commit-Position: refs/heads/master@{#296248}

Revert of Removing browser.Start() and instead starts the browser at the point of creation. (patchset #6 id:120001 of https://codereview.chromium.org/565803002/)

Reason for revert:
investigate build error

Original issue's description:
> Removing browser.Start() and instead starts the browser at the point of creation.
>
> BUG=412126
>
> Committed: https://crrev.com/01c68e8765254d9971296b291ff718b4f6af3004
> Cr-Commit-Position: refs/heads/master@{#295591}
>
> Committed: https://crrev.com/74c6ef17a08754abadb6fb64da14249898285fa8
> Cr-Commit-Position: refs/heads/master@{#296107}

TBR=nednguyen@google.com,slamm@chromium.org,slamm@google.com
NOTREECHECKS=true
NOTRY=true
BUG=412126

Review URL: https://codereview.chromium.org/601473002

Cr-Commit-Position: refs/heads/master@{#296247}

Previous to this change, suffixing a "." to the URL from which a
dictionary was being suggested would allow servers to bypass the
check against a subdomain (e.g. evil.protected.good.com) setting a
dictionary to be used by a superdomain (.good.com).

BUG=389451
R=rsleevi@chromium.org

Review URL: https://codereview.chromium.org/574283006

Cr-Commit-Position: refs/heads/master@{#296246}

Properly initialize |drop_remaining_touches_in_sequence_| after
its addition in https://codereview.chromium.org/586553002/.

TBR=rbyers@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/581313003

Cr-Commit-Position: refs/heads/master@{#296245}

Use the ChromeNetworkDelegate::OnProxyFallback() callback to keep track of number of times the data reduction proxies fall back because of network errors. Display a warning to users if there are network errors when connecting to the proxy and no successful requests through the proxy.

BUG=401244

Review URL: https://codereview.chromium.org/568893002

Cr-Commit-Position: refs/heads/master@{#296244}

mojo/aura doesn't use webkit/common/gpu any more.

third_party/wtl is only used on windows and is added in a conditional
if (is_win) section below.

Review URL: https://codereview.chromium.org/599603002

Cr-Commit-Position: refs/heads/master@{#296243}

BUG=none
TBR=ajwong@chromium.org

Review URL: https://codereview.chromium.org/592423004

Cr-Commit-Position: refs/heads/master@{#296242}

This is a possible cause for a failure being seen in the field in the
associated bug; this change is a quick fix for that failure + UMA to
determine if the failure is real.

BUG=416639
R=mef@chromium.org
R=asvitkine@chromium.org

Review URL: https://codereview.chromium.org/597633002

Cr-Commit-Position: refs/heads/master@{#296241}

We recently changed the chrome build so that nacl binaries of both 32 and 64
bit type are not built unless needed.
The sdk had assumed both arches are built as a side effect.

Building both arches in the sdk.

BUG=None
TEST=local
R=binji@chromium.org

Review URL: https://codereview.chromium.org/590523002

Cr-Commit-Position: refs/heads/master@{#296240}

Dismiss the bubble without updating changes to the node.

BUG=329459
R=asvitkine@chromium.org

TEST=
1. Launch chrome, press "Cmd+Shift+D" to "Show Bookmarks Bar"
2. Navigate to any webpage or new tab & press "Cmd+D" to open bookmark window and then press "Enter" key
3. On the same page, again press "Cmd+D" and change the page name in bookmark bubble window
4. Now press "Esc" key and observe
5. On pressing "Esc" key, bubble should get closed without making any changes to the bookmark

Review URL: https://codereview.chromium.org/592323002

Cr-Commit-Position: refs/heads/master@{#296239}

* Changed one to "permission denied" for screen capture.
* Changed one to "no hardware" for platform app.

BUG=416233

Review URL: https://codereview.chromium.org/585263002

Cr-Commit-Position: refs/heads/master@{#296238}

net: manual HSTS addition by request of the site.
Cr-Commit-Position: refs/heads/master@{#296237}

Current LTH Context Lost tests do not test for
impl-side painting, this updates the tests to use
ContentLayer or PictureLayer based on the settings.

Also, SetRequiresHighResToDraw when we recover from lost
context/ new output surface is used, to avoid checkerboards.

BUG=401492, 415155

Review URL: https://codereview.chromium.org/591473003

Cr-Commit-Position: refs/heads/master@{#296236}

Revert of DevTools: Made id unique across different remote browsers (patchset #2 id:20001 of https://codereview.chromium.org/560523004/)

Reason for revert:
Need to delete the contents of chrome_beta_pages; see memory leak at https://build.chromium.org/p/chromium.memory/builders/Linux%20ASan%20LSan%20Tests%20(3)/builds/7887

Original issue's description:
> DevTools: Made id unique across different remote browsers
>
> BUG=338303
>
> Committed: https://crrev.com/7732fd3f76fc31a95f14199f898804e5d80e96e3
> Cr-Commit-Position: refs/heads/master@{#296188}

TBR=dgozman@chromium.org,pfeldman@chromium.org,vkuzkokov@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=338303

Review URL: https://codereview.chromium.org/598663002

Cr-Commit-Position: refs/heads/master@{#296235}

Crashes in net::AssertEventNotSignaled() are hard reproduce, but looks like this is the primary reason.
With crash on CHECK(..._callback_.is_null()) we could identify incorrect code.

BUG=415753

Review URL: https://codereview.chromium.org/580163003

Cr-Commit-Position: refs/heads/master@{#296234}

The gpu_unittests don't use any ANGLE or native graphics API
logic. Thus the isolate shouldn't include any dependency on things
it doesn't use or build. This should clear up runtime errors on the
bots involving gpu_unittests_run and libGLESv2.dll.

BUG=415981

Review URL: https://codereview.chromium.org/594203003

Cr-Commit-Position: refs/heads/master@{#296233}

Redirects field in TopSitesDatabase is encoded as a space-seprated list
of URLs. But data URLS like "data:text/plain,this text  has space" adds
extra space, so that when decoded, non-URLs like "text" and "has" would
trigger DCHECK in URL() downstream, and crash debug Chrome.

This CL validates redirects URLs when they are read. It is not foolproof
(e.g., "data:text/plan,hidden http://extra.url"), but now we can ponder
how to best fix the root cause, without keeping the crashing bug around.

BUG=358034

Review URL: https://codereview.chromium.org/562993002

Cr-Commit-Position: refs/heads/master@{#296232}