GitLab

When user saves an image, add headers requesting a pass through
from the data reduction proxy. Without this, images will be
saved in webp format which is not supported by most image viewers.
BUG=464436

Review URL: https://codereview.chromium.org/988453002

Cr-Commit-Position: refs/heads/master@{#319681}

Introduce struct AlternativeService and use it in
BrokenAlternateProtol{List,Map} instead of BrokenAlternateProtocolEntry.  This
is the first step towards making an alternative service broken not with respect
to an origin, but by itself.

BUG=392576

Review URL: https://codereview.chromium.org/987813002

Cr-Commit-Position: refs/heads/master@{#319680}

The typedefs in native_web_contents_modal_dialog.h are no longer needed
after http://crrev.com/866263008.
This change removes them, as discussed in that CL.

Review URL: https://codereview.chromium.org/985133002

Cr-Commit-Position: refs/heads/master@{#319679}

Right now, there's a chance that the NSMenu can outlive the underlying menu
model (which is owned by the view controller). Make sure we reset the context
menu before destruction and before creating a new menu to correct this.

BUG=462175

Review URL: https://codereview.chromium.org/967883006

Cr-Commit-Position: refs/heads/master@{#319678}

Revert of [Fallback icons] Change "explicit flow" interface so color hex strings don't use "#". (patchset #4 id:60001 of https://codereview.chromium.org/988863002/)

Reason for revert:
This seems to be causing FallbackIconUrlParserTest unit_tests to fail, see, for example,
https://build.chromium.org/p/chromium.memory/builders/Mac%20ASan%2064%20Tests%20%281%29/builds/7625
https://build.chromium.org/p/chromium.memory/builders/Linux%20Chromium%20OS%20ASan%20LSan%20Tests%20%281%29/builds/7297
https://build.chromium.org/p/chromium.linux/builders/Linux%20Tests%20%28dbg%29%281%29/builds/39924
https://build.chromium.org/p/chromium.linux/builders/Linux%20Tests%20%28dbg%29%281%29%2832%29/builds/10797

Original issue's description:
> [Fallback icons] Change "explicit flow" interface so color hex strings don't use "#".
>
> Design: go/chrome-fallback-icons
>
> Fallback icon "explicit flow" URL used to have '#' in hex colors, e.g.:
>
> chrome://fallback-icon/,#01f,#123456,,/http://www.google.com
>
> Problem: "#" denotes fragment in URLs, so we get rid of it. This causes no ambiguity
> with named color (e.g., "red") because no named color consists of letters a-f only.
> Also adding "ARGB" hex color since this is supported in Skia color parsing.
>
> BUG=455063
>
> Committed: https://crrev.com/8fa91aabefbe9784eca0c169bc2ad2dd92f5d887
> Cr-Commit-Position: refs/heads/master@{#319630}

TBR=jhawkins@chromium.org,huangs@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=455063

Review URL: https://codereview.chromium.org/989133003

Cr-Commit-Position: refs/heads/master@{#319677}

There's no need for AutofillPopUpButton here. None of its functionality
is used.

BUG=448572

Review URL: https://codereview.chromium.org/987503002

Cr-Commit-Position: refs/heads/master@{#319676}

Rename initWithWebContents -> initWithBridge and remove |webContents_|
because it is never used.

BUG=448572

Review URL: https://codereview.chromium.org/983833004

Cr-Commit-Position: refs/heads/master@{#319675}

Pass a pointer instead, as per the style guide. Also makes the
constructor explicit (all callsites already are explicit).

BUG=464816
TBR=blundell@chromium.org,agl,dmichael,garykac

Review URL: https://codereview.chromium.org/982413002

Cr-Commit-Position: refs/heads/master@{#319674}

Same as previous unplug fix, but for the libevdev/gestures path. This
applies to hybrid keyboards with attached mouse/touchpad.

The SYN_DROPPED part is N/A because that is handled by libevdev.

BUG=463002
TEST=Tested unplug by unplugging devices with shift held.
  Shift gets properly released rather than being stuck.
  (NB: Could not reproduce without adding a delay to expose the race).

Review URL: https://codereview.chromium.org/990483002

Cr-Commit-Position: refs/heads/master@{#319673}

Update logic for disabling keyboards on X11 when Maximize Mode is activated.

TEST=Manual testing on device
BUG=chrome-os-partner:37106

Review URL: https://codereview.chromium.org/991933002

Cr-Commit-Position: refs/heads/master@{#319672}

The name help_outline is what the icon library calls it.

BUG=464640
TBR=dfalcantara@chromium.org

Review URL: https://codereview.chromium.org/994683002

Cr-Commit-Position: refs/heads/master@{#319671}

Checking whether credit card scanning is possible can be an expensive
operation. The result of this call is unlikely to change during the
lifetime of a browser process. This patch will query credit card
scanning feature once when the user types in a credit card input field
for the first time. Restarting the browser process is necessary to clear
the cache.

BUG=461944

Review URL: https://codereview.chromium.org/990463002

Cr-Commit-Position: refs/heads/master@{#319670}

Policy extensions weren't installed by the user, and frequently cannot be
uninstalled. Similar to component extensions, we shouldn't synthesize actions
for these extensions.

BUG=463939

Review URL: https://codereview.chromium.org/977943002

Cr-Commit-Position: refs/heads/master@{#319669}

BUG=https://code.google.com/p/chromedriver/issues/detail?id=984

Review URL: https://codereview.chromium.org/883083002

Cr-Commit-Position: refs/heads/master@{#319668}

Before this patch in order to run single test one needs to create
xml file with credentials for every site which can be tested.
This patch fixes this issue.

BUG=369521
R=vabr@chromium.org

Review URL: https://codereview.chromium.org/994573002

Cr-Commit-Position: refs/heads/master@{#319667}

I think, when explicitly declaring a dependency on a target
(like gn_all does on build_utf8_validator_tables), if the target
is unconditionally only built in the host toolset, the dependency
needs to also say that; my initial version of the dependency did
not do that, leading gyp_chromium to fail in obscure ways.

R=thakis@chromium.org
BUG=465198

Review URL: https://codereview.chromium.org/990093002

Cr-Commit-Position: refs/heads/master@{#319666}

Initial attempt found here: https://codereview.chromium.org/983933002/

This is going to be used by https://codereview.chromium.org/975943002/

TEST=LayerOwnerTest.RecreateRootLayerWithNullCompositor
TEST=LayerOwnerTestWithCompositor.RecreateRootLayerWithCompositor

BUG=337596

Review URL: https://codereview.chromium.org/987043004

Cr-Commit-Position: refs/heads/master@{#319665}

This patch should not change any behaviour. It introduces a new version
of onReceviedError that will be called for all resources and provides
more information about the failed request.

BUG=456782

Review URL: https://codereview.chromium.org/992593003

Cr-Commit-Position: refs/heads/master@{#319664}

The operator& is dangerous and makes it unclear what you are doing.
Replace it with a get_pointer() method.

R=Nico
TBR=garykac@chromium.org
BUG=464816

Committed: https://crrev.com/845755856c54aabac7022ee373618468b92f2e61
Cr-Commit-Position: refs/heads/master@{#319565}

Review URL: https://codereview.chromium.org/986563003

Cr-Commit-Position: refs/heads/master@{#319663}

BUG=260022,350202

Review URL: https://codereview.chromium.org/914363003

Cr-Commit-Position: refs/heads/master@{#319662}

Make the chrome://extensions page use chrome.developerPrivate API for
extension error-related functionality.
Also add tests for the api function, and convert it to a
UIThreadExtensionFunction.

Additionally, fix a bug in extensions.js where calling "showOverlay" on an
already-visible hid the overlay (temporarily).

BUG=461039

Review URL: https://codereview.chromium.org/973303002

Cr-Commit-Position: refs/heads/master@{#319661}

Most problems are because of wrong collation or not reading the style guide, so
pointing people to those.

BUG=

Review URL: https://codereview.chromium.org/989533002

Cr-Commit-Position: refs/heads/master@{#319660}

BUG=464040

Review URL: https://codereview.chromium.org/986333002

Cr-Commit-Position: refs/heads/master@{#319659}

BUG=456880

Review URL: https://codereview.chromium.org/954923005

Cr-Commit-Position: refs/heads/master@{#319658}

BUG=165079

Review URL: https://codereview.chromium.org/978383005

Cr-Commit-Position: refs/heads/master@{#319657}

There's a couple possible ways to get keys stuck down:

(1) User unplugs the device while holding a key. A key release is queued
by the kernel but races the unplug event. The unplug arrives first and we
tear everything down without ever dispatching a release event.

(2) Key release gets dropped in-kernel due to queue overruns. We never
get the release, and continue processing without dispatching a release
event. We do get a SYN_DROPPED event to notify us that events were lost.

If stuck, a key can be pressed again to unstick it.

Fix these cases by keeping track of pressed keys for every keyboard
device, and releasing everything during teardown & following queue
overruns.

Also unify with the path added for key filtering. With the additional
state, we can tell that we have an outstanding press and avoid filtering
the corresponding release.

BUG=463002
TEST=
  (1) Tested unplug by unplugging devices with shift held.
      Shift gets properly released rather than being stuck.
      (NB: Could not reproduce without adding a delay to expose the race).
  (2) Tested SYN_DROPPED by suspending chrome with key pressed,
      releasing it, and then pressing more keys to cause an overrun.
      (suspended via kill -STOP <browser>)
  (3) event_unittests

Review URL: https://codereview.chromium.org/985163002

Cr-Commit-Position: refs/heads/master@{#319656}

Since InMemoryURLIndex code is no longer in chrome/browser/history but
instead in chrome/browser/autocomplete, move it out of the history
namespace.

BUG=453790
TBR=jcivelli@chromium.org

Review URL: https://codereview.chromium.org/959343004

Cr-Commit-Position: refs/heads/master@{#319655}

The discardable memory free list might need to be split into
different buckets for good performance. These histograms will
be useful when choosing bucket sizes.

BUG=

Review URL: https://codereview.chromium.org/984803002

Cr-Commit-Position: refs/heads/master@{#319654}

This removes includes of android/ and win/WebSandboxSupport.h, and only
has the platform implementations create it on the platforms that do have
a WebSandboxSupport (Mac, Linux).

If this lands, the empty WebSandboxSupport headers can be removed from Blink
in a followup CL.

BUG=463506

Review URL: https://codereview.chromium.org/976163002

Cr-Commit-Position: refs/heads/master@{#319653}

RTCVideoDecoder: Using I420VideoFrame directly instead of to-be-removed TextureVideoFrame (previously this typeddefed to I420VideoFrame).

BUG=webrtc:1128

Review URL: https://codereview.chromium.org/990913002

Cr-Commit-Position: refs/heads/master@{#319652}

This caused use-after-free crashes when the ThreadedParserReceiver feature is enabled.
R=jam
BUG=277124

Review URL: https://codereview.chromium.org/986643003

Cr-Commit-Position: refs/heads/master@{#319651}

BUG=460514

Review URL: https://codereview.chromium.org/988393002

Cr-Commit-Position: refs/heads/master@{#319650}

Mention CVE-2015-0565.

TBR= mseaborn@chromium.org, shyamsundarr@chromium.org, timwillis@chromium.org
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3944
TEST= none
NOTRY=true
(documentation only change)

Review URL: https://codereview.chromium.org/820673006

Cr-Commit-Position: refs/heads/master@{#319649}

Per blink-dev discussion:
https://groups.google.com/a/chromium.org/d/msg/blink-dev/vCd9TysLAso/UcNQMme-9LsJ

This is relaxing the browser-side checks in order to no longer kill a
renderer process if it tries to open a cross-origin window from a service
worker. It will allow the Blink restriction to be relaxed:
https://codereview.chromium.org/985043002

The browser process now checks whether the renderer process is allowed
access to the given URL, blocking access to special URLs like chrome://.

BUG=457187

Review URL: https://codereview.chromium.org/980383004

Cr-Commit-Position: refs/heads/master@{#319648}

This is part of a suite of thee CLs:
1. https://codereview.chromium.org/968163003
2. <this>
3. https://codereview.chromium.org/968203002

BUG=462255

Review URL: https://codereview.chromium.org/968983002

Cr-Commit-Position: refs/heads/master@{#319647}

This CL lands server change 87003629 by tgreer and change 87977978 by birenroy.

BUG=

Review URL: https://codereview.chromium.org/980393003

Cr-Commit-Position: refs/heads/master@{#319646}

Blink's opacity & LCD logic is being removed - instead this is to be
handled by setting appropriate SkSurfaceProps onto the target canvas.

Unfortunately that's not directly applicable to
BrowserFontResource_Trusted::DrawTextAt/DrawTextToCanvas because the
API allows different opacity flags to be specified for different
calls directed to the same canvas.

We work around by disabling LCD text at the last stage, using a
temporary SkCanvas.

BUG=430617
R=reed@google.com,piman@chromium.org

Review URL: https://codereview.chromium.org/975533004

Cr-Commit-Position: refs/heads/master@{#319645}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/1b110d6..13e51f9

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

Review URL: https://codereview.chromium.org/987073002

Cr-Commit-Position: refs/heads/master@{#319644}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/e08a307..719867b

TBR=jbroman@chromium.org

Review URL: https://codereview.chromium.org/986323002

Cr-Commit-Position: refs/heads/master@{#319643}

Revert of Revert of Roll WebRTC 8649:8651, Libjingle 8639:8651 (patchset #1 id:1 of https://codereview.chromium.org/989253002/)

Reason for revert:
The compile error was not caused by the WebRTC roll.

Original issue's description:
> Revert of Roll WebRTC 8649:8651, Libjingle 8639:8651 (patchset #1 id:1 of https://codereview.chromium.org/987033002/)
>
> Reason for revert:
> Broke compilation on Linux http://build.chromium.org/p/chromium/builders/Linux/builds/59264 with error message: AttributeError: 'Checker' object has no attribute 'run_js_check'
>
> Original issue's description:
> > Roll WebRTC 8649:8651, Libjingle 8639:8651
> >
> > WebRTC 8649:8651
> > Changes: https://chromium.googlesource.com/external/webrtc/trunk/webrtc.git/+log/b4c07a0..61b3feb
> >
> > Libjingle 8639:8651
> > Changes: https://chromium.googlesource.com/external/webrtc/trunk/talk.git/+log/c9f4c13..4138f65
> >
> > TBR=tommi
> >
> > Committed: https://crrev.com/f497d1fbcf1c0919ba260ed64ac18829b3acf7a0
> > Cr-Commit-Position: refs/heads/master@{#319635}
>
> TBR=tommi@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Committed: https://crrev.com/62c6611787693f20f4201411d816d647f77f17d0
> Cr-Commit-Position: refs/heads/master@{#319638}

TBR=tommi@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/994603002

Cr-Commit-Position: refs/heads/master@{#319642}