GitLab

It is going to be used by V8 sampling profiler.

BUG=406277

Review URL: https://codereview.chromium.org/1007253005

Cr-Commit-Position: refs/heads/master@{#320837}

BUG=413541
TEST=manually tested.

Review URL: https://codereview.chromium.org/1011633002

Cr-Commit-Position: refs/heads/master@{#320836}

This CL implements two P0 browser test cases
- It2MeBrowserTest.CancelShare
- It2MeBrowserTest.VerifyAccessCodeNotReusable

This only behavioral change in this CL is that we will always return INVALID_ACCESS_CODE
for It2Me Connections failed with HOST_IS_OFFLINE.
HOST_IS_OFFLINE suggests that the host portion of the access code
is invalid and we should hide this implementation detail from the user.

BUG=466849,466850

Review URL: https://codereview.chromium.org/1007543003

Cr-Commit-Position: refs/heads/master@{#320835}

BUG=466378

Review URL: https://codereview.chromium.org/1016483006

Cr-Commit-Position: refs/heads/master@{#320834}

BUG=358315
R=rpaquay@chromium.org

Review URL: https://codereview.chromium.org/1012953004

Cr-Commit-Position: refs/heads/master@{#320833}

This CL encapsulates Me2MeConnectFlow as an object.
Ultimately, we would like to create an interface to abstract
connection, error handling on the It2MeFlow and the Me2MeFlow
on the desktop remoting end.

Summary of changes:
1. Move pairingInfo into remoting.Host.
2. Update the field names of remoting.pairingInfo to match
   storage keys.
3. Move code from crd_connect.js into Me2MeConnectFlow.

BUG=466406

Review URL: https://codereview.chromium.org/1008003002

Cr-Commit-Position: refs/heads/master@{#320832}

so that dynamically linked NaCl applications like ARC (App Runtime
for Chrome) start faster.

ARC is a dynamically linked NaCl application which lists more than
70 DSO files in its nmf file. crrev.com/319931 allowed Chrome to
pre-open up to 2 DSO files when it spawns a NaCl plugin process so
that the plugin process can open/read/map the DSO files without
issuing any Chrome IPC. This CL increase the number of DSO files
Chrome can pre-open from 2 to 120+.

BUG=348232
BUG=nativeclient:3802
TEST=ipc_tests, browser_tests (*PackagedAppTest.SuccessfulLoad)
TEST=manually tested non-SFI ARC apps

Review URL: https://codereview.chromium.org/998833002

Cr-Commit-Position: refs/heads/master@{#320831}

Used until chrome switches to the new skia api

BUG=
TBR=robertphilips@chromium.org

Review URL: https://codereview.chromium.org/1014473005

Cr-Commit-Position: refs/heads/master@{#320830}

R=dmazzoni@chromium.org
BUG=None

Review URL: https://codereview.chromium.org/1014663002

Cr-Commit-Position: refs/heads/master@{#320829}

BUG=456424

Review URL: https://codereview.chromium.org/997623002

Cr-Commit-Position: refs/heads/master@{#320828}

BUG=465765

Review URL: https://codereview.chromium.org/1004283003

Cr-Commit-Position: refs/heads/master@{#320827}

The browser compositor contents and renderer contents are each put into separate Surfaces which are aggregated together by a Display.

BUG=449319

Review URL: https://codereview.chromium.org/1001203002

Cr-Commit-Position: refs/heads/master@{#320826}

BUG=None

Review URL: https://codereview.chromium.org/937633002

Cr-Commit-Position: refs/heads/master@{#320825}

Also, delete the 'Experimental' field trial group since it was never
used.

BUG=465174

Review URL: https://codereview.chromium.org/1008303002

Cr-Commit-Position: refs/heads/master@{#320824}

M42 is the last version we ship to ICS devices. We want to make sure
that ICS users only use the most stable bookmark manager. Therefore we
should disable it on ICS now.

BUG=466869

Review URL: https://codereview.chromium.org/1008053002

Cr-Commit-Position: refs/heads/master@{#320823}

CreateNetworkDelegate is called on the IO thread, but SetDataReductionProxyService
is called on the UI thread. There is the possibility for a race to occur in
reading/writing the value of the WeakPtr; we can avoid it by passing a callback
to DataReductionProxyUsageStats (instead of the WeakPtr), since we are guaranteed
that the callback is called on the UI thread.

BUG=460876

Review URL: https://codereview.chromium.org/1012923003

Cr-Commit-Position: refs/heads/master@{#320822}

Resourceless software draw can only have a single root layer
render surface. Since decision to create render surface has
been moved to main thread, this condition no longer holds
and the mismatch causes a crash in LayerIterator.

Instead tweak the logic to only *use* the root layer render
surface, while being possible to to have more than one
render surface in the tree. This means only the root render
layer can be a render target.

BUG=466695

Review URL: https://codereview.chromium.org/1007623002

Cr-Commit-Position: refs/heads/master@{#320821}

This CL encapsulates various objects that maintain the state
of a remoting connection into a struct, remoting.ConnectionInfo.

This will prepares us for moving the creation logic of the
SessionConnector into the application delegates.

BUG=465878

Review URL: https://codereview.chromium.org/1010053002

Cr-Commit-Position: refs/heads/master@{#320820}

Do not retransmit QUIC data for streams that have been reset. Protected
by FLAGS_quic_do_not_retransmit_for_reset_streams.

This is an attempt to reland cr/88179867 now that the underlying bug has
been fixed. Includes tests which crashed on the old CL and pass now.

Merge internal change: 88695223
https://codereview.chromium.org/1009243002/

Move QUIC reads to a more optimized reader.
Guarded by FLAGS_quic_use_optimized_packet_reader. This is not enabled
for the chromium's QUIC server yet.

The new QuicPacketReader is optimized to spend less time in memset and
uses a 16 packet receive buffer.  Saves 1.09 % of CPU in memset alone
according to profiling EndToEndTest.

Merge internal change: 88573234
https://codereview.chromium.org/1008313002/

Rollback of merge internal change: 87346681.

Chromium CL reviewed:
  https://codereview.chromium.org/968923005/

Chomium committed CL: "Land Recent QUIC Changes until 03/02/2015".
  https://codereview.chromium.org/968233004/

This CL was in M43 bit not in M42. It impacted last week's Dev release
and all Canary released after 03/03/2015.

*** Original change description ***

Remove TCP and BBR's max congestion window.  No practical change,
because Chrome's receive window is the limiting factor.

Adds kMaxReportedCwnd and kMaxReorderingForRtt as constants to replace
the max CWND in classes that used to rely on max CWND.

***

Merge internal change: 88545044
https://codereview.chromium.org/1003863006/

Make the SentPacketManager remove pending retransmissions for reset
streams. No behavior change yet. That will be in a flag protected followup.

Minor change to linked_hash_map's erase method to return an iterator
pointing to the new location of the element that followed the last
element erased by the function call (ala internal source code and
std:: c++).

Merge internal change: 88507368
https://codereview.chromium.org/1013583002/

Only changes comments. Replace "CID" with "connection id" in various
comments in quic_connection_test.cc

Merge internal change: 88463135
https://codereview.chromium.org/1013573002/

Provide a MockConnection constructor that takes connection_id,
perspective, and is_secure as arguments.  Used by the internal server
testing code. Clarify debugging log message in quic_dispatcher.cc.

Clarify debugging log message in quic_dispatcher.cc. Improve testing
infrastructure.

Merge internal change: 88461470
https://codereview.chromium.org/1008293002/

Add a kBYTE congestion option to QUIC to negotiate the TcpCubicBytesSender.

Merge internal change: 88328620
https://codereview.chromium.org/1009203003/

Add support for the MIN1 connection option to TcpCubicBytesSender.
This sender is currently unused.

Allows TcpCubicBytesSender to have a min cwnd of 1 packet in bytes.

Merge internal change: 88167032
https://codereview.chromium.org/1008283002/

R=rch@chromium.org

Review URL: https://codereview.chromium.org/1008323002

Cr-Commit-Position: refs/heads/master@{#320819}

BUG=none

TBR=cpu@chromium.org

Review URL: https://codereview.chromium.org/1000203008

Cr-Commit-Position: refs/heads/master@{#320818}

Adds an CanExceedIdleDeadlineIfRequired function to RendererScheduler
to enable IdleTasks to estimate the best time for them to exceed their deadline
if they have a large monolithic task which could not be run otherwise.

Long idle task design doc:
https://docs.google.com/a/chromium.org/document/d/1yBlUdYW8VTIfB-DqhvQqUeP0kf-Ap1W4cao2yQq58Do/edit

BUG=455713

Review URL: https://codereview.chromium.org/969373002

Cr-Commit-Position: refs/heads/master@{#320817}

Substitute slashes for dots in the class name if we are going to pass it
to ClassLoader.loadClass via JNI, instead of calling JNIEnv::FindClass.
This avoids spamming the log with warnings from Dalvik about invalid
class names; it was working correctly anyway due to implementation
details of the VM, but we shouldn't be doing this.

This is not the ideal solution to the problem but is a relatively safe
and isolated fix we can cherrypick to a release branch while we work on
a better one.

BUG=461773
R=rmcilroy

Review URL: https://codereview.chromium.org/1008363004

Cr-Commit-Position: refs/heads/master@{#320816}

Also added a test to test that batch read works when
getMoreData() is called multiple times (i.e when the data is
large).
There are existing test cases for batch read (testInputStreamReadMoreBytesThanAvailable,
testInputStreamReadExactBytesAvailable, and
testInputStreamReadLessBytesThanAvailable)

BUG=398997

Review URL: https://codereview.chromium.org/972213002

Cr-Commit-Position: refs/heads/master@{#320815}

R=sievers@chromium.org
BUG=466725
TEST=trybots

Review URL: https://codereview.chromium.org/1006093002

Cr-Commit-Position: refs/heads/master@{#320814}

BUG=461032

Review URL: https://codereview.chromium.org/997203003

Cr-Commit-Position: refs/heads/master@{#320813}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/d68a550..b83205a

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

Review URL: https://codereview.chromium.org/1010903004

Cr-Commit-Position: refs/heads/master@{#320812}

Revert of Delay font cache building by few seconds to yield for startup resource contention. (patchset #3 id:40001 of https://codereview.chromium.org/992413003/)

Reason for revert:
Suspected to be causing failures on Intel Release GPU bot.

See crbug.com/467706

Original issue's description:
> Delay font cache building by few seconds to yield for startup resource contention.
>
> During first renderer start there are lot of things happening simultaneously some of them are:
> 1. Renderer is going through all font files on the system to create a font collection.
> 2. Renderer loading up startup URL, accessing HTML/JS File cache, net activity etc.
> 3. Extension initialization.
>
> We delay building of this cache mainly to avoid parallel font file loading along with Renderer. Some systems have significant number of font files which takes long time process.
>
> BUG=458751,436195
> R=cpu,scottmg
>
> Committed: https://crrev.com/787e29c61825173adaf3414ff8e3d3c0c26f529b
> Cr-Commit-Position: refs/heads/master@{#320226}

TBR=cpu@chromium.org,scottmg@chromium.org,gab@chromium.org,shrikant@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=458751,436195

Review URL: https://codereview.chromium.org/1009403004

Cr-Commit-Position: refs/heads/master@{#320811}

It'd be useful to diagnose IPC related issues, especially
Chrome OS builds where Release build is the norm.
This shouldn't be noisy as these won't happen in normal use.

R=viettrungluu@chromium.org
TEST=466814
BUG=

Review URL: https://codereview.chromium.org/1003483004

Cr-Commit-Position: refs/heads/master@{#320810}

BUG=461857

Review URL: https://codereview.chromium.org/994193002

Cr-Commit-Position: refs/heads/master@{#320809}

Revert of Revert of MSan: check data is initialized when serializing with base::Pickle. (patchset #1 id:1 of https://codereview.chromium.org/985893002/)

Reason for revert:
The bugs that caused the revert have been fixed - Unreverting.

Original issue's description:
> Revert of MSan: check data is initialized when serializing with base::Pickle. (patchset #1 id:1 of https://codereview.chromium.org/988693004/)
>
> Reason for revert:
> Found errors w.r.t. IPC messages. Reverting for now while people fix the errors.
>
> Original issue's description:
> > MSan: check data is initialized when serializing with base::Pickle.
> >
> > Serialization implies that the data is going to leave the process one way or
> > another, and it's never a good idea to leak uninitialized bytes. In most cases
> > MSan would catch this later on (e.g. on a write()). There are exceptions to
> > that, however, such as writing to shared memory. It doesn't hurt to have an
> > extra check here.
> >
> > BUG=463906
> > R=thakis@chromium.org
> >
> > Committed: https://crrev.com/2c705b038413d30d7aae5ff5d09882575e93c898
> > Cr-Commit-Position: refs/heads/master@{#319459}
>
> TBR=thakis@chromium.org,earthdok@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=463906
>
> Committed: https://crrev.com/d74906086c38b72bed2017a772c67f3c3fde2709
> Cr-Commit-Position: refs/heads/master@{#319510}

TBR=thakis@chromium.org,earthdok@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=463906

Review URL: https://codereview.chromium.org/1008393004

Cr-Commit-Position: refs/heads/master@{#320808}

Two reasons:
1. This name is a better reflection of what this function does.
2. I want to introduce an IsHotwordHardwareAvailable() that actually
returns whether hotword hardware is available. One that cannot be
overridden with a flag and always returns the real value.

Review URL: https://codereview.chromium.org/1008273003

Cr-Commit-Position: refs/heads/master@{#320807}

Make LoadMonitoringExtensionHostQueue remove itself as an ExtensionHost observer at the correct time.

"Extensions.ExtensionHostMonitoring.NumLoaded" UMA is now correctly recorded.

I also cleaned up ExtensionHost and its Start/Stop events a little to make it
more obvious, but in a way that my previous attempt at this (r320536) won't
cause crashes.

R=yoz@chromium.org
TBR=sky@chromium.org

Review URL: https://codereview.chromium.org/1016473002

Cr-Commit-Position: refs/heads/master@{#320806}

This fixes some files not compiling with the proper flags for component mode.

The device_event_log target in GN is a source set in GN and a component in GYP. This difference was causing some errors. I changed this to be a static library in GYP. It only has 2 files in it so making a whole shared library for this target seems wasteful.

Review URL: https://codereview.chromium.org/1011823002

Cr-Commit-Position: refs/heads/master@{#320805}

BUG=466848

Review URL: https://codereview.chromium.org/1005683003

Cr-Commit-Position: refs/heads/master@{#320804}

browser plugins can detach before we attach again.

BUG=467614
Test=Load a <webview>
Set its style.display = 'block'.
Observer that it succeeds without <webview> going stale or
throwing any error about "attach" failing.

Review URL: https://codereview.chromium.org/1015443002

Cr-Commit-Position: refs/heads/master@{#320803}

Dependent libraries need to be passed in the reverse order that
build_utils.GetSortedTransitiveDependencies returns.

Also, a multi-library resource zip in package_libraries should be sorted
on the numeric values of the directories in the zip, not the string
values.

Review URL: https://codereview.chromium.org/1007043002

Cr-Commit-Position: refs/heads/master@{#320802}

Update the chrome://extensions page to use the developerPrivate api in order
to show the permissions prompt.
Also convert the function to a UIThreadExtensionFunction, and update it to
(if possible) use the new app info dialog.

BUG=461039

Review URL: https://codereview.chromium.org/1008973002

Cr-Commit-Position: refs/heads/master@{#320801}

In the old app-list, the key selection doesn't signal any events
to AppListItemView, so it can't update the background color at
that point.

BUG=467517
R=mgiuca@chromium.org

Review URL: https://codereview.chromium.org/1010853002

Cr-Commit-Position: refs/heads/master@{#320800}

Because ChromeTracingAgent class has global states, when any
exception was raised inside StopTracing method, the active tracing
state of this class is not set to False, making subsequent unittest that
try to StartTracing fails because of this.

Long term fix of this depends on crbug.com/317481 be fixed so that
ChromeTracingAgent no longer have global state.

BUG=459807

Review URL: https://codereview.chromium.org/997193002

Cr-Commit-Position: refs/heads/master@{#320799}

BUG=

Review URL: https://codereview.chromium.org/1018453002

Cr-Commit-Position: refs/heads/master@{#320798}