GitLab

Review URL: https://codereview.chromium.org/362923002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281699 0039d316-1c4b-4281-b951-d872f2087c98

Reject push streams which exceed the limit.

For now, the browser will send RST_STREAM with REFUSED_STREAM status.
This mimics mod_spdy behavior.

BUG=170544, 377538

R=jgraettinger@chromium.org

TEST=SpdySessionTest.PushedStreamShouldNotCountToClientConcurrencyLimit
     SpdySessionTest.RejectPushedStreamExceedingConcurrencyLimit
     SpdySessionTest.IgnoreReservedRemoteStreamsCount

Review URL: https://codereview.chromium.org/367963003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281698 0039d316-1c4b-4281-b951-d872f2087c98

TBR=nick@chromium.org
BUG=none

Review URL: https://codereview.chromium.org/376833003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281697 0039d316-1c4b-4281-b951-d872f2087c98

This contains fixes for the following sorts of issues:
* Signedness mismatch

BUG=81439
TEST=none

Review URL: https://codereview.chromium.org/371153002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281696 0039d316-1c4b-4281-b951-d872f2087c98

Migration of code from ContentViewCore to NavigationController to avoid ContentViewCore abstractions.

Review URL: https://codereview.chromium.org/369503003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281695 0039d316-1c4b-4281-b951-d872f2087c98

Corrects the regression of [a-z] keys for non-US layouts would have wrong VKEY code if CAPSLOCK is on.

For example, Ctrl+A with Caps used to work for FR keyboard, but after cl 357613002, it doesn't work. The root cause is that cl 357613002 tries the remove the modifier states for the xkeyevent for its keysym for [a-z0-9]  keys but only considered the shift/altgr states, not capslock & other states. 

BUG=386066
TEST=Verified on Pixel device.

Review URL: https://codereview.chromium.org/376463002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281694 0039d316-1c4b-4281-b951-d872f2087c98

Several tests were intended to verify muxed content behavior, but
are not actually using muxed clusters. This is a precursor change
that makes the use of muxed clusters more explicit. There are
only slight changes to some of the tests to make them properly reflect
the original intent. A followup change will actually create proper
muxed clusters and make any necessary updates to the test expectations.

Review URL: https://codereview.chromium.org/361303003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281693 0039d316-1c4b-4281-b951-d872f2087c98

Delegated renderer disabled for layout tests (as on other platforms).

BUG=314190
TBR=piman

Review URL: https://codereview.chromium.org/374643002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281692 0039d316-1c4b-4281-b951-d872f2087c98

BUG=392037
TEST=base_unittests --gtest_filter=ThreadLocalStorageTest.*

Review URL: https://codereview.chromium.org/379503002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281691 0039d316-1c4b-4281-b951-d872f2087c98

One Windows and Mac, continue to use a whitelist of app ids.

BUG=353449
TEST=none
R=tengs@chromium.org

Review URL: https://codereview.chromium.org/286203003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281690 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281689 0039d316-1c4b-4281-b951-d872f2087c98

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog_blink.html?url=/trunk&range=177624:177637&mode=html
TBR=pilgrim@chromium.org,yutak@chromium.org
BUG=

Review URL: https://codereview.chromium.org/372183002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281688 0039d316-1c4b-4281-b951-d872f2087c98

I'm only 50% sure I got everything right when adding the WPR archive etc, so any guidance appreciated.

BUG=391197

Review URL: https://codereview.chromium.org/367283002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281687 0039d316-1c4b-4281-b951-d872f2087c98

Insertion text handles should be hidden when the underlying text changes. As
such, add a hook to notify the AdapterInputConnection embedder of an IME event 
when text is deleted, in turn hiding the insertion handle.

BUG=389444

Review URL: https://codereview.chromium.org/351403002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281686 0039d316-1c4b-4281-b951-d872f2087c98

BUG=391520
TBR=caitkp@chromium.org

Review URL: https://codereview.chromium.org/378763002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281685 0039d316-1c4b-4281-b951-d872f2087c98

According to our style guide, in unit tests, the tested file's header should
be first.

TEST=Not related. Refactoring only.
BUG=248427

Review URL: https://codereview.chromium.org/372963002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281684 0039d316-1c4b-4281-b951-d872f2087c98

This patch just adds the flag. The implementation will be committed later.

BUG=305511
TEST=none

Review URL: https://codereview.chromium.org/374803003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281683 0039d316-1c4b-4281-b951-d872f2087c98

Now there is a but that the bubble is shown even though the window is minimized.
This CL fixes this bug.

BUG=389014
TEST=manual

Review URL: https://codereview.chromium.org/364053004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281682 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281681 0039d316-1c4b-4281-b951-d872f2087c98

Eliminate SyncFileSystemService from SyncProcessRunner to inject its dependency from tests.

BUG=344769

Review URL: https://codereview.chromium.org/377563002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281680 0039d316-1c4b-4281-b951-d872f2087c98

Use the same computation as is used in DelegatedFrameHost.

A separate patch will rename viewport_size.

BUG=391386
TBR=danakj

Review URL: https://codereview.chromium.org/375793002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281679 0039d316-1c4b-4281-b951-d872f2087c98

Pass BookmarkModel* to UpdateStarredStateOfMatches().
StringForURLDisplay() is only used in two places:
 - base_search_provider.cc: This is the only caller which passes true as |check_accept_lang|.  Move the code to read kAcceptLanguages pref to this file.
 - history_url_provider.cc: Pass empty string as |languages| from this file.

Add Profile* for subclasses as a member if needed.

BUG=388515
TEST=git cl try

Review URL: https://codereview.chromium.org/369423003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281678 0039d316-1c4b-4281-b951-d872f2087c98

Keep-alive above 250ms causes PseudoTCP to reset congestion window
between messages, making connection much less responsive.

BUG=391991

Review URL: https://codereview.chromium.org/376763002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281677 0039d316-1c4b-4281-b951-d872f2087c98

We are systematically removing results.Add() callsites and replacing them
with results.AddValue() callsites. I've done scirra.py and sunspider.py,
and nednguyen@ is doing more in the tools/perf directory.

BUG=346958

Review URL: https://codereview.chromium.org/372163002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281676 0039d316-1c4b-4281-b951-d872f2087c98

In the current implementation, ServiceWorkerVersion controls its status based
on event notifications (eg. ServiceWorkerVersion::OnInstallEventFinished).

But the spec algorithm requires to manage the status more flexibly during
registration process. For example, the version should be set to INSTALLED
status after associated to the documents because the page should be able to
see '.waiting' on 'onstatechange' event.

Because of this reason the register job should manage the status of the version.

(Spec) https://slightlyoff.github.io/ServiceWorker/spec/service_worker/index.html


BUG=379012
TEST=content_unittest --gtest_filter=ServiceWorker* 
TEST=run_webkit_tests.py http/tests/serviceworker/waiting.html

Review URL: https://codereview.chromium.org/369063004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281675 0039d316-1c4b-4281-b951-d872f2087c98

After delegated frame is used, this message is not handled anymore
by GpuProcessHostUIShim.
On Aura, this message is handled by BrowserCompositorOutputSurfaceProxy.
On Mac, vsync params updating is triggered not by gpu process
but by browser process itself with DisplayLinkMac.
On Android, browser process uses android system's vsync signal.

R=ajuma@chromium.org, piman@chromium.org, kbr@chromium.org
BUG=NONE

Review URL: https://codereview.chromium.org/369003002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281674 0039d316-1c4b-4281-b951-d872f2087c98

BUG=288383

Review URL: https://codereview.chromium.org/377613003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281673 0039d316-1c4b-4281-b951-d872f2087c98

This seems to be breaking Linux builds:

http://build.chromium.org/p/chromium/builders/Linux/builds/51190
http://build.chromium.org/p/chromium.linux/builders/Linux%20Builder%20%28dbg%29%2832%29/builds/26110


> Disable two yama tests on 3.2 kernels with 32bit userland and 64bit kernel.
> 
> Disabling yama restrictions is broken there.
> 
> BUG=391916
> R=jorgelo@chromium.org
> 
> Review URL: https://codereview.chromium.org/371113003

TBR=thakis@chromium.org

Review URL: https://codereview.chromium.org/376793002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281672 0039d316-1c4b-4281-b951-d872f2087c98

Pulled out of https://codereview.chromium.org/349263004/ at cpu's request.

R=gab@chromium.org
TBR=brettw@chromium.org

BUG=391508, 141141, 141517, 142787, 140882, 165760, 163713

Review URL: https://codereview.chromium.org/375683002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281670 0039d316-1c4b-4281-b951-d872f2087c98

480x270 is 16:9 aspect.

BUG=none
TEST=manually tested
R=fukino@chromium.org

Review URL: https://codereview.chromium.org/371953002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281669 0039d316-1c4b-4281-b951-d872f2087c98

Users of WebRTC reported that certain corporate networks that restrict the DNS resolution
to certain domains. In which case DNS resolve might fail in those conditions.

In that case proxy can resolve the addresses by using CONNECT methods.

Chrome TCP P2P sockets already handles the proxy scenarios, so by just providing the
hostname to TCP sockets should solve these problamatic scenarios.

After socket moves to a connected state, resolved address (peer address) must be reported
back to the renderer process. It's certainly needed for sending packets, but for validation
purposes its required.

BUG=https://code.google.com/p/webrtc/issues/detail?id=3384

R=sergeyu@chromium.org,jiayl@chromium.org
TBR=cevans@chromium.org
CC=juberti@chromium.org

Review URL: https://codereview.chromium.org/365903007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281668 0039d316-1c4b-4281-b951-d872f2087c98

BUG=305852

Review URL: https://codereview.chromium.org/313123003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281667 0039d316-1c4b-4281-b951-d872f2087c98

Disabling yama restrictions is broken there.

BUG=391916
R=jorgelo@chromium.org

Review URL: https://codereview.chromium.org/371113003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281666 0039d316-1c4b-4281-b951-d872f2087c98

If an app/extension requests permission to access more than one type of devices (e.g. USB, Bluetooth and Serial) we should show one message listing the transport method(s).

Currently we show a an install-time warning for each type of device the extension/app requests permission for. This is very excessive and this CL solves this problem by showing a single message that include the transport method.

Currently we have three transport methods: Bluetooth, Serial and USB. We could show the following messages depending on the requested permissions:

"Access your hardware devices connected via USB and Bluetooth" (how is this different
than kHid?)
"Access your hardware devices connected via USB and Serial"
"Access your hardware devices connected via Bluetooth and Serial"

and also keep the individual messages if only one category of hardware is
requested.

BUG=384353
R=jww, meacer, kalman

Review URL: https://codereview.chromium.org/336313009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281665 0039d316-1c4b-4281-b951-d872f2087c98

Replace the CHECK call in the PatchServiceManagerCalls function which IAT patches DirectWrite with a DCHECK.

The IAT patches fails with AppCompat if the OpenSCManager intercept is enabled. This is based on crashes we have seen in canary.

In any case that is an unsupported configuration as far as Chrome is concerned because we are compatible with all supported operating
systems. If the IAT patching fails dwrite will fallback to the old behavior of connecting to the font cache service.

BUG=391426
TBR=jam

Review URL: https://codereview.chromium.org/378723002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281664 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None
TEST=None
TBR=erg

Review URL: https://codereview.chromium.org/375703005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281663 0039d316-1c4b-4281-b951-d872f2087c98

Pass SearchTermsData and URLRequestContextGetter to TemplateURLFetcher instead of Profile.
Replace TemplateURLFetcherCallbacks with base::Callback.

BUG=371535
TEST=unit_tests

Review URL: https://codereview.chromium.org/367413003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281661 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281660 0039d316-1c4b-4281-b951-d872f2087c98

BUG=374086,391951,391953,391957,391960
TEST=gpu waterfall
TBR=kbr@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/373883002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281659 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281658 0039d316-1c4b-4281-b951-d872f2087c98