GitLab

NOTRY=true

Review URL: https://codereview.chromium.org/576563002

Cr-Commit-Position: refs/heads/master@{#294948}

BUG=

Review URL: https://codereview.chromium.org/575603002

Cr-Commit-Position: refs/heads/master@{#294947}

The CancelIo() operation does not synchronously cancel IO operations,
we must wait for them to return before returning to avoid writes to
resources which may have already been destructed.

See this blog post for more information:

http://blogs.msdn.com/b/oldnewthing/archive/2011/02/02/10123392.aspx

BUG=413494
TEST=manual inspect of cancellation states.

Review URL: https://codereview.chromium.org/570713003

Cr-Commit-Position: refs/heads/master@{#294946}

https://chromium.googlesource.com/chromium/blink/+log/1d4eb25ceba65e3e78786bce1251dc248d20af19..fbd4f37d728e462a9cbc3482156ad7e8c2c1d0b5
TBR=loislo@chromium.org,dpranke@chromium.org

Review URL: https://codereview.chromium.org/564663004

Cr-Commit-Position: refs/heads/master@{#294945}

Determine how many extension icons to show on each row of the overflow menu based on the width of other menu items.

Screenshot: http://imgur.com/mPQAbxs
BUG=412503

Review URL: https://codereview.chromium.org/553233002

Cr-Commit-Position: refs/heads/master@{#294944}

When in the component build multiple threads (apps) will use the same
view manager client lib and could fight over the current GL context, so
always make our context current in BitmapUploader.

When shutting down abrubtly (i.e. when losing a connection to an app),
the client will not necessarily properly unref all resources it uses.
DCHECKing here isn't helpful, so just skip over resources that are not
fully unreffed.

R=ben@chromium.org

Review URL: https://codereview.chromium.org/569173007

Cr-Commit-Position: refs/heads/master@{#294943}

BUG=352293

Review URL: https://codereview.chromium.org/566413002

Cr-Commit-Position: refs/heads/master@{#294942}

This means Chrome will no longer register itself as the handler for the
bogus "chromehtml" URL protocol. (Regular file associations should not
register the class name as a URL protocol.)

BUG=413051

Review URL: https://codereview.chromium.org/559253002

Cr-Commit-Position: refs/heads/master@{#294941}

This patch was generated by running the rewrite_scoped_refptr clang tool
on a ChromeOS build.

BUG=110610

Review URL: https://codereview.chromium.org/556293004

Cr-Commit-Position: refs/heads/master@{#294940}

BUG=412882
TEST=components_unittests
R=tengs@chromium.org

Review URL: https://codereview.chromium.org/562763004

Cr-Commit-Position: refs/heads/master@{#294939}

Metrics already exist but were never added to the inputs.

R=stevenjb@chromium.org

Review URL: https://codereview.chromium.org/571233002

Cr-Commit-Position: refs/heads/master@{#294938}

We figured out which part caused the problom, so these logs are no longer needed.

BUG=393331
TEST=bots
TBR=kbr@chromium.org

Review URL: https://codereview.chromium.org/551803006

Cr-Commit-Position: refs/heads/master@{#294937}

Program the NIC to wake-on-packet for packets arriving from
the GCM server.  Chrome notifies the connection manager (shill)
by calling D-Bus methods.  The return status indicates if the
operation succeeded or not (depending on hardware support and
the type of connection).

Keywords: wowlan, wake-on-lan, wake-on-wlan

BUG=360295
TEST=verified dbus methods are called, and packets are sent on the
TEST=connection passed to the methods.  Also verified that shill
TEST=programs the NIC correctly, and packets wake up the device.

Review URL: https://codereview.chromium.org/409883006

Cr-Commit-Position: refs/heads/master@{#294936}

R=asvitkine@chromium.org,isherman@chromium.org
BUG=404791

Review URL: https://codereview.chromium.org/523273003

Cr-Commit-Position: refs/heads/master@{#294935}

Revert of ash: Add checks for supervised users. (patchset #5 id:80001 of https://codereview.chromium.org/561713002/)

Reason for revert:
This is breaking NotificationApiTest.TestUserGesture, which assumes it starts with an empty notification list.

Somehow this is causing the browser test to be in a state where there is already a notification in the list regarding a supervised user.

Original issue's description:
> ash: Add checks for supervised users.
>
> Propagate the user's supervised state to the ash system
> so that it becomes available for views
>
> https://chromereviews.googleplex.com/79527013/
>
> BUG=394417
>
> Committed: https://crrev.com/8616ac4cf7bd77f53dc001eb9d8d61c6c975eaaa
> Cr-Commit-Position: refs/heads/master@{#294821}

TBR=antrim@chromium.org,derat@chromium.org,jennyz@chromium.org,bauerb@chromium.org,treib@chromium.org,merkulova@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=394417

Review URL: https://codereview.chromium.org/575653002

Cr-Commit-Position: refs/heads/master@{#294934}

This reverts commit f51d7453.

The UI is breaking with certain extensions and users have no
ability to work around this. Putting back behind a flag and
trunk-only restriction.

BUG=414497,474476
CC=jleichtling@chromium.org,amineer@chromium.org
TBR=kalman@chromium.org,fsamuel@chromium.org

Review URL: https://codereview.chromium.org/571213003

Cr-Commit-Position: refs/heads/master@{#294933}

* Changes the include order of YASM
Default includes need to be put to the end for YASM.

In particular there are several versions of x86inc.asm in the tree and
libvpx requires its own version. This change in order allows libvpx to
include its own version.

* Roll libvpx to include BUILD.gn.

libvpx now builds on x86 posix.

R=brettw@chromium.org, scherkus@chromium.org, tomfinegan@chromium.org

Review URL: https://codereview.chromium.org/535613002

Cr-Commit-Position: refs/heads/master@{#294932}

Otherwise keyed services see different profiles in unit_tests at
different times since the incognito->GetOriginalProfile() changes. This
is causing release code to pick up unnecessary workarounds to keep
unit_tests green in try runs.

The fix is achieved by adding `TestingProfile::Builder::BuildIncognito`,
which differs from `Build` by requiring an |original_profile| argument,
which will own the incognito profile (so no need to return scoped_ptr).
This simplifies a lot of call sites, since they no longer need to do a
scoped_ptr dance of their own.

BUG=277296

Review URL: https://codereview.chromium.org/556173002

Cr-Commit-Position: refs/heads/master@{#294931}

This patch was generated by running the rewrite_scoped_refptr clang tool
on a ChromeOS build.

BUG=110610

Review URL: https://codereview.chromium.org/568413004

Cr-Commit-Position: refs/heads/master@{#294930}

This patch adds a bit more information into traces in the activation
path.

BUG=414292
R=reveman@chromium.org

Review URL: https://codereview.chromium.org/569413002

Cr-Commit-Position: refs/heads/master@{#294929}

Removes the spammy warning message, and replaces it with a sanity-
check that the queue has not grown past 10 seconds' worth of
packets.  If the sanity-check trips, a process crash dump will be
sent to make us aware, via crash reports, that this is a problem
in-the-wild.

Review URL: https://codereview.chromium.org/539323002

Cr-Commit-Position: refs/heads/master@{#294928}

Currently, when a scroll starts touch selection handles are hidden and
then re-shown after scroll ends. But, if scroll ends in a fling, handles
are not re-shown, because there hadn't been a way to determine end of
the fling. Now that there is an IPC message notifying end of flings, we
can re-show handles after a fling.

BUG=413927

Review URL: https://codereview.chromium.org/569863003

Cr-Commit-Position: refs/heads/master@{#294927}

Related changes:
 - Created a DriWindowManager to keep track of the platform windows.
   These are then used to perform the bounds clipping for the cursor.
 - Moved delegate manager to DriWindowDelegateManager. In a follow-up CL
   the delegate manager in the browser process will be removed in favor
   of holing on to the DriWindowDelegates on the DriWindow.

BUG=none
NOTRY=true

Review URL: https://codereview.chromium.org/556073003

Cr-Commit-Position: refs/heads/master@{#294926}

BUG=414162
TBR=dsinclair

Review URL: https://codereview.chromium.org/577493002

Cr-Commit-Position: refs/heads/master@{#294925}

Replaced the test for validation of requests and responses
with one that relies on the Connection et al classes to set
up the validators and deliver messages.

BUG=413303

Review URL: https://codereview.chromium.org/551293004

Cr-Commit-Position: refs/heads/master@{#294924}

Review URL: https://codereview.chromium.org/566863005

Cr-Commit-Position: refs/heads/master@{#294923}

Modifications from upstream:
- renamed class as it doesn't make much sense to test an "abstract" class.
- made this into an E2E test so we can test against the real ChromeVox message strings and not have to fake out a lot of objects.

Review URL: https://codereview.chromium.org/571643004

Cr-Commit-Position: refs/heads/master@{#294922}

Also catch 401 errors.

BUG=None.
TEST=None.

Review URL: https://codereview.chromium.org/556003002

Cr-Commit-Position: refs/heads/master@{#294921}

This change adds frame processor tests to cover the behavior recently
landed from bad1c509 that changed "segments" appendMode's partial front
trimming logic.

BUG=408216
R=dalecurtis@chromium.org
TEST=new segments test fails before and passes after bad1c509, new sequence test passes before and after

Review URL: https://codereview.chromium.org/567403004

Cr-Commit-Position: refs/heads/master@{#294920}

This patch was generated by running the rewrite_scoped_refptr clang tool
on a ChromeOS build.

BUG=110610

Review URL: https://codereview.chromium.org/555253005

Cr-Commit-Position: refs/heads/master@{#294919}

Tab.getWebContents() was using JNI to fetch the WebContents object. This
is undesirable for several reasons, including performance impact and
thread access consistency. Remove this and prefer redirecting the query
to ContentViewcore.getWebContents() instead.

This fix is a temporary workaround until all downstream test sites can
be patched.

BUG=398263,414445
TBR=tedchoc@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/571813003

Cr-Commit-Position: refs/heads/master@{#294918}

NPEvent::lParam is defined as uintptr_t which has same width with ULONG on Win32, but they are of different sizes on WIN64. uintptr_t is 8-byte long, while ULONG is still 4-byte long, so it corrupts a 64-bit pointer if casting it to a 32-bit type.

BUG=413535
TEST=square.html (attached in the bug), http://pbskids.org/video (it uses wmode="opaque")
R=ananta@chromium.org, bauerb@chromium.org

Review URL: https://codereview.chromium.org/571583002

Cr-Commit-Position: refs/heads/master@{#294917}

If RendererFreezer::SuspendImminent was called multiple times before it
had a chance to report readiness, it would post multiple tasks to do so
on the current message loop.  However the callback to report readiness
would be cleared after the first task ran and would cause a crash when
the next task tried to run it.  Fix this by not posting new tasks while
a callback is pending and just update the callback instead.

BUG=414396
Signed-off-by: Chirantan Ekbote <chirantan@chromium.org>

Review URL: https://codereview.chromium.org/550093004

Cr-Commit-Position: refs/heads/master@{#294916}

With the current sharding strategy (random), having smaller suites will reduce cycle time.
The Android bots will automatically pick up this config change. If we want to do the desktop bots too, that'd require a master restart.

BUG=396001
TEST=None.
R=tonyg

Review URL: https://codereview.chromium.org/566773003

Cr-Commit-Position: refs/heads/master@{#294915}

R=jam@chromium.org
BUG=406457

Review URL: https://codereview.chromium.org/571133003

Cr-Commit-Position: refs/heads/master@{#294914}

This pulls in the following Native Client changes:

r13747: (dschuff) Fix toolchain build and sandboxed translation tests on FYI bots
r13748: (thakis) Pass -std=gnu++0x in build_nexe.py.

BUG=none
TEST=browser_tests and nacl_integration
CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32,linux_arm_cross_compile,linux_nacl_sdk,linux_chromium_dbg

Review URL: https://codereview.chromium.org/571053004

Cr-Commit-Position: refs/heads/master@{#294913}

which prevent grouping mechanism in MergeNetworkList from working.

The end result is that each Network after MergeNetworkList contains only 1 ip address. With this change, the grouping will happen correctly. WebRTC has logic to only retrieve single IP from each network for connectivity check already.

This CL is a dup of https://codereview.chromium.org/536133003/ which was created under the wrong identity.
BUG=

Review URL: https://codereview.chromium.org/565773003

Cr-Commit-Position: refs/heads/master@{#294912}

Remove stale ChromiumAsyncUrlRequest.java as it is replaced by CronetUrlRequest.java as of https://codereview.chromium.org/520093002/.

BUG=409926
TBR=mmenke@chromium.org

Review URL: https://codereview.chromium.org/569713003

Cr-Commit-Position: refs/heads/master@{#294911}

CQ_EXTRA_TRYBOTS=tryserver.blink:linux_blink_rel,linux_blink_dbg

BUG=

Review URL: https://codereview.chromium.org/575493002

Cr-Commit-Position: refs/heads/master@{#294910}

This patch was generated by running the rewrite_scoped_refptr clang tool
on a ChromeOS build.

BUG=110610

Review URL: https://codereview.chromium.org/567763003

Cr-Commit-Position: refs/heads/master@{#294909}