GitLab

The Linux sandbox performance implications that made using
recvmsg()/sendmsg() slower than read()/write() have not applied for a
long time, since we switched from seccomp (v1) to seccomp-bpf.

Review URL: https://codereview.chromium.org/1139703005

Cr-Commit-Position: refs/heads/master@{#330414}

You can now delete things by long pressing suggestions in the dropdown (assuming it's the right kind of suggestion).

BUG=486153

Review URL: https://codereview.chromium.org/1134793004

Cr-Commit-Position: refs/heads/master@{#330413}

Custom menu items where used for WebUI NTP and is no longer used.
This CL removes support for such menus.

BUG=None

Review URL: https://codereview.chromium.org/1142863002

Cr-Commit-Position: refs/heads/master@{#330412}

This patch removes install_verifier accessor from ExtensionSystem.
It can be its own browser context keyed service.
It can be built by its new factory.

Review URL: https://codereview.chromium.org/1127133006

Cr-Commit-Position: refs/heads/master@{#330411}

If one of the core services apps called terminate, the thread won't be valid.

This is split off https://codereview.chromium.org/1139673003

BUG=484234

Review URL: https://codereview.chromium.org/1145473005

Cr-Commit-Position: refs/heads/master@{#330410}

The YouCompleteMe configuration still needs to determine the list of
system include paths. The current chromium.ycm_extra_conf.py logic
requires that the host machine has a 'clang' binary in its PATH that can
yield this information.

This change calculates the system include paths based on the clang++
binary used to compile the target file. It allows the use of wrappers
(E.g. goma), by looking for a clang++ command in the first and second
tokens in the compile command. It's still not perfect since the Clang
library used by ycmd might be different, but it should work better than
the current logic.

BUG=none
NOTRY=true

Review URL: https://codereview.chromium.org/1137303005

Cr-Commit-Position: refs/heads/master@{#330409}

Currently each ContentViewCore creates a KitkatCaptioningBridge on K+ devices.
And all of them listens to the system CaptioningManager.
So whenever the system captioning changes, all ContentViewCores will have to handle it separately.
This change simplifies that process by making KitkatCaptioningBridge a singleton.
All ContentViewCores become listeners to that singleton.
Whenever system captioning changes, a single CaptioningChangeDelegate class will handle the change.
It will then inform all ContentViewCores that are alive.
If all ContentViewCores are GC'ed, KitkatCaptioningBridge will stop listen to system CaptioningManager.

BUG=457850

Review URL: https://codereview.chromium.org/1132053005

Cr-Commit-Position: refs/heads/master@{#330408}

BUG=323304

Review URL: https://codereview.chromium.org/1122383005

Cr-Commit-Position: refs/heads/master@{#330407}

Add flag '/deferredrun' to recovery executable for elevated install of recovery component so we can distinguish from first install.

Also changed version arg format from --version to /version.

BUG=

Review URL: https://codereview.chromium.org/1136943004

Cr-Commit-Position: refs/heads/master@{#330406}

BUG=

Review URL: https://codereview.chromium.org/1132133008

Cr-Commit-Position: refs/heads/master@{#330405}

'unsafety' must be 'unsafely', apparently.

BUG=441605

Review URL: https://codereview.chromium.org/1124173007

Cr-Commit-Position: refs/heads/master@{#330404}

BUG=479533
TBR=dtseng

Review URL: https://codereview.chromium.org/1094103006

Cr-Commit-Position: refs/heads/master@{#330403}

Disable 0RTT if the hostname of the server and origin are different.  Currently
dead code since HttpStreamFactoryImpl::GetAlterantiveServiceFor() refuses to
return an altenative service with hostname different than that of the origin,
but I plan to change this very soon.

While disabling 0RTT in this case is not strictly necessary, it enables us to
move faster with alternative service implementation.  We expect the most
performance gain from pooling to existing connections to different hosts, which
is not hindered by this change.

* Add origin hostname parameter to QuicStreamRequest::Request() method.
* Add server_and_origin_have_same_host parameter to QuicStreamFactory::Create()
  and CreateAuxiliaryJob() methods, and QuicStreamFactory::Job constructor.
* Add server_and_origin_have_same_host_ member to QuicSteamFactory::Job.
* Driveby: rename was_alternate_protocol_recently_broken_ member to
  was_alternative_service_recently_broken_.

BUG=474217

Review URL: https://codereview.chromium.org/1140153003

Cr-Commit-Position: refs/heads/master@{#330402}

SetInfo parses phone numbers and gets rid of invalid ones, so SetRawInfo
should be used instead. This is consistent with AutofillOptionsHandler::SetAddress

BUG=455314

Review URL: https://codereview.chromium.org/1139533007

Cr-Commit-Position: refs/heads/master@{#330401}

BUG=internal b/19560149

Review URL: https://codereview.chromium.org/1127183008

Cr-Commit-Position: refs/heads/master@{#330400}

This ensures that OnProcessLaunched is called on a
UtilityProcessHostImpl even when an in-process utility
thread is in use. The point is to allow Mojo IPC to
utility code even in single-process mode.

BUG=None

Review URL: https://codereview.chromium.org/1138823003

Cr-Commit-Position: refs/heads/master@{#330399}

- TrayCast is the central injection point for the UI code
- TrayCast itself just delegates between two other views, of which only one is active at a time depending on if the device is currently casting
- ScreenCaptureTrayItem is disabled so it will not be active when the device is casting (otherwise there would be two very similar tray items that have the same functionality but slightly different appearance)

BUG=433140

Review URL: https://codereview.chromium.org/1118613003

Cr-Commit-Position: refs/heads/master@{#330398}

Fix ProfileInfoCache behaviour when it tried to download the
non-existing "NothingToDownload" placeholder profile avatar from the
Web, instead of just loading it from the resources.

BUG=486854

Review URL: https://codereview.chromium.org/1127113003

Cr-Commit-Position: refs/heads/master@{#330397}

This moves files into cc/playback/ cc/raster/ cc/tiles/ and cc/output/
and is based on the proposal found in
https://docs.google.com/spreadsheets/d/1wmPOmV9uqd9zNJ5l2TGePH7_vlSvPeXbC-Y3SeGJ_rc/edit#gid=0

R=enne, vmpstr
TBR=piman
BUG=488755

Review URL: https://codereview.chromium.org/1144693002

Cr-Commit-Position: refs/heads/master@{#330396}

The main preferences package already has an owners file, but the test
package and the chrome shell preferences package don't.

Review URL: https://codereview.chromium.org/1140873006

Cr-Commit-Position: refs/heads/master@{#330395}

First, PlatformEventSource is a singleton so it shouldn't be created by PlatformViewportX11. Once we have multiple windows, this will be very clear. Until then, we also have UAFs on destruction of PlatformViewportX11 because PlatformEventSource could be lower in the stack calling a PlatformViewportX11 method which cause both of these objects to be deleted.

Second, ensure that child windows of the PlatformViewport aren't used after the parent window is destructed, otherwise we get X errors. This is done by ensuring that NativeViewportImpl, which owns the PlatformViewport, destructs OnscreenContextProvider first. OnscreenContextProvider will in turn destroy all windows in the CommandBufferDriver that it created.

Lastly, make NativeViewportImpl destroy itself when its platform window is closed (i.e. because the user closed the window). The old implementation in that method didn't do anything.

This is split off https://codereview.chromium.org/1139673003

BUG=484234

Review URL: https://codereview.chromium.org/1130353008

Cr-Commit-Position: refs/heads/master@{#330394}

BUG=None

Review URL: https://codereview.chromium.org/1138003007

Cr-Commit-Position: refs/heads/master@{#330393}

This is so we can prevent other from adding this dependency back.

They will think twice or at least consult with OWNERS if we have this in
place in the DEPS exclusion list.

BUG=134675
TEST=checkdeps
R=sky@chromium.org

Review URL: https://codereview.chromium.org/1138643012

Cr-Commit-Position: refs/heads/master@{#330392}

Simplify the Config class and consolidate init logic.
  Inline GN argument file parsing and make it optional.
  Take an optional build_dir argument for Paths use.

Simplify the Paths class; respect Config's build_dir.
  Only reconstruct build_dir if Config doesn't have one.
  Remove unused methods; rename mojo_runner path member.

Remove unused GN utilities and dart apptest support.

BUG=486220,488128
TEST=android_run_mandoline.py, android_mojo_shell.py, and apptest_runner.py work as expected.
R=sky@chromium.org

Review URL: https://codereview.chromium.org/1144673003

Cr-Commit-Position: refs/heads/master@{#330391}

The TC for input time from OutputE2ETest.Input will be enabled again
after adding AX TIME role.
The related CL:
https://codereview.chromium.org/1123313004/
https://codereview.chromium.org/1132963002/

BUG=480861

Review URL: https://codereview.chromium.org/1124383009

Cr-Commit-Position: refs/heads/master@{#330390}

I saw this error on a try run: http://build.chromium.org/p/tryserver.chromium.linux/builders/android_chromium_gn_compile_rel/builds/89243/steps/compile/logs/stdio

Review URL: https://codereview.chromium.org/1139423002

Cr-Commit-Position: refs/heads/master@{#330389}

When the "--disable-background-networking" command-line switch is
specified, extensions updates shouldn't be downloaded.

BUG=489244

Review URL: https://codereview.chromium.org/1140873004

Cr-Commit-Position: refs/heads/master@{#330388}

BUG=None
R=jam@chromium.org

Review URL: https://codereview.chromium.org/1127203004

Cr-Commit-Position: refs/heads/master@{#330387}

This is because they have an unowned pointer to cc::SurfaceManager.

Also ensure we don't have a double delete of display_ in DisplayImpl's destructor. This occurred because when display_ was deleted, it would end up calling back into  OutputSurfaceLost() which would attempt to delete display__ again.

This is split off https://codereview.chromium.org/1139673003

BUG=484234

Review URL: https://codereview.chromium.org/1131933006

Cr-Commit-Position: refs/heads/master@{#330386}

Saw a build error in http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_gn_chromeos_rel/builds/38730/steps/compile/logs/stdio

Review URL: https://codereview.chromium.org/1131113007

Cr-Commit-Position: refs/heads/master@{#330385}

Otherwise when Browser::OnViewManagerDisconnected gets called, Browser::ui_ will get deleted twice.

This is split off https://codereview.chromium.org/1139673003

BUG=484234

Review URL: https://codereview.chromium.org/1137873004

Cr-Commit-Position: refs/heads/master@{#330384}

Which of UIWebView and WKWebView to use can be controller by command line
flags, finch and availability (iOS version).

BUG=429756

Review URL: https://codereview.chromium.org/1144803002

Cr-Commit-Position: refs/heads/master@{#330383}

This is split off https://codereview.chromium.org/1139673003

BUG=484234

Review URL: https://codereview.chromium.org/1142083002

Cr-Commit-Position: refs/heads/master@{#330382}

This should have already been closed before we get to the destructor. Otherwise we get double deletes since the Destroy() call can end up calling the delegate which will delete this object.

Also in the X11 implementation, don't call delegate's OnClosed() before null'ing xwindow_, otherwise Destroy() can be called multiple times.

This is split off https://codereview.chromium.org/1139673003

BUG=484234

Review URL: https://codereview.chromium.org/1132313003

Cr-Commit-Position: refs/heads/master@{#330381}

This was a good experiment, but there's nobody currently driving this iteration, and the code is intrusive, accumulating technical debt around it.

BUG=370696

Review URL: https://codereview.chromium.org/1144463003

Cr-Commit-Position: refs/heads/master@{#330380}

BUG=452726

Review URL: https://codereview.chromium.org/1132123003

Cr-Commit-Position: refs/heads/master@{#330379}

PushMessagingAppIdentifier::FindByAppId was DCHECKing when called from
PushMessagingServiceImpl::CanHandle to see if an app_id belongs to
Push, since it assumes that it would only be called with valid Push
app_ids.

This patch makes it test whether the app_id is a Push Messaging app_id
before performing any validation DCHECKs.

BUG=489304

Review URL: https://codereview.chromium.org/1140873005

Cr-Commit-Position: refs/heads/master@{#330378}

BUG=none
TEST=none

Review URL: https://codereview.chromium.org/1134753004

Cr-Commit-Position: refs/heads/master@{#330377}

This addresses a TODO and removes an ugly constant from NetLog.  On
the down side, TimeTicks::UnixEpoch() is cached, which may result in
less accuracy (And all bets are off if Chrome is suspended to disk).

BUG=488660

Review URL: https://codereview.chromium.org/1127133008

Cr-Commit-Position: refs/heads/master@{#330376}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/f6e36c6..d804fe3

TBR=kbr@chromium.org,yhirano@chromium.org,hiroshige@chromium.org

Review URL: https://codereview.chromium.org/1142833002

Cr-Commit-Position: refs/heads/master@{#330375}