GitLab

Cr-Commit-Position: refs/heads/master@{#434883}

This fixes and re-enables NativeWidgetMacTest.WindowModalSheet with some
additional test expectations. It was reverted in r433028 when it started
failing.

Fundamentally, the problem is that calling orderOut: on an NSWindow with
an attached sheet results in utterly broken behavior in AppKit,
documented in http://crbug.com/667602.

AppKit's broken behaviour with orderOut and regular child windows is
already handled by BridgedNativeWidget by recreating the parent/child
relationships. However, it can't do this for sheets. It's not possible
to "reattach" a sheet once it becomes detached. And even giving a sheet
a parentWindow results in graphical glitches (http://crbug.com/605098).
Sheets use -[NSWindow sheetParent], not -[NSWindow parentWindow].

Note that -[NSApp hide:] works. It's just -[NSWindow orderOut:] that's
broken. But we need the triggers from -[NSApp hide:] to be consistent
with the different things that need to happen for sheets.

Add checks to BridgedNativeWidget to detect attempts to orderOut: a
window with an attached sheet. Different solutions need to be sought for
these, such as what Chrome already does for the certificate viewer sheet
when we make it tab-modal.

Update DCHECKs in WidgetOwnerNSWindowAdapter to properly support sheets.
This allows NativeWidgetMacTest.WindowModalSheet to be re-enabled.

Updates BridgedNativeWidget::OnVisibilityChanged() to properly fix
http://crbug.com/605098. This bug still manifests in Chrome, but only
after a call to [NSApp hide:]. Re-showing a window improperly attaches
the parent as a parentWindow (rather than only sheetParent), and this is
the cause of the graphical glitches.

BUG=666503, 605098, 667602

Review-Url: https://codereview.chromium.org/2521753003
Cr-Commit-Position: refs/heads/master@{#434882}

There was a TODO comment which said that having a parameter to
set memory state at renderers startup. Still we don't have the
parameter but we already call OnStateChange() immediately after
renderer is ready to communicate.
(See MemoryCoordinatorImpl::OnChildAdded()).
This is effectively the same so remove the TODO comment.

Review-Url: https://codereview.chromium.org/2532953005
Cr-Commit-Position: refs/heads/master@{#434881}

This enum result is recorded for verifying HW codec is used on ARC++,
or if there is any initialization error.

BUG=661864
TEST=test it on Elm and make sure the flag shows correct value in
chrome://histograms/Media.ArcGpuVideoDecodeAccelerator.InitializeResult

Review-Url: https://codereview.chromium.org/2506363003
Cr-Commit-Position: refs/heads/master@{#434880}

BUG=602688
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2525393002
Cr-Commit-Position: refs/heads/master@{#434879}

BUG=669323
R=dgozman

Review-Url: https://codereview.chromium.org/2536073002
Cr-Commit-Position: refs/heads/master@{#434878}

Many WebGL tests use this function.

BUG=668223
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2532543002
Cr-Commit-Position: refs/heads/master@{#434877}

Some code paths are unguarded for the case where we get events while shutting
down/detaching a Frame. With this patch we just early-out.

BUG=655306

Review-Url: https://codereview.chromium.org/2540573002
Cr-Commit-Position: refs/heads/master@{#434876}

It is redundant as we have the field in ResourceResponse.

BUG=666217

Review-Url: https://codereview.chromium.org/2521483002
Cr-Commit-Position: refs/heads/master@{#434875}

As per Sebastien,
- more padding for "No more users" case.
- update text styles for both "Add user" and "No more users".

Also,
- fix active user name font weight (Medium->Normal)
- don't make the "no more users" label a button (it already didn't do
  anything when you clicked it, but there's no sense in making it a
  button).

BUG=667942

Review-Url: https://codereview.chromium.org/2535003002
Cr-Commit-Position: refs/heads/master@{#434874}

Original CL was reverted due to breaking webkit_tests in
chromium.webkit/builders/WebKit%20Mac10.9/builds/40056

This is due to a problem that already bit me :-)
build infra script to zip up mojom files wasn't updated,
ongoing in https://chromium-review.googlesource.com/c/414565/

Original CL description ------------------------------
ImageCapture: move image_capture.mojom to media/capture/mojo/

This CL moves image_capture.mojom from media/mojo/interfaces/ to
media/capture/mojo/ where it belongs logically speaking, and
updates BUILD.gns and include paths.

image_capture.mojom was in media/mojo/interfaces because
when it landed, media/capture/mojo was not existing.

BUG=518807
TEST=No code touched, if compile passes, it works
(but there's also content_browsertests).

TBR=xhwang@chromium.org for the removal in media/mojo/interfaces/BUILD.gn

Review-Url: https://codereview.chromium.org/2539543002
Cr-Commit-Position: refs/heads/master@{#434873}

The Linux-specific prlimit() system call combines and extends the
functionality of setrlimit() and getrlimit().  It can be used to both
set and get the resource limits of an arbitrary process.

Since version 2.13, the glibc getrlimit() and setrlimit() wrapper
functions no longer invoke the corresponding system calls, but
instead employ prlimit(), for the reasons described in BUGS.

If new_limit is not NULL, then the rlimit structure to which it points is
used to set new values.  If it is NULL, then prlimit() acts as getrlimit().

So, allow prlimit() with new_limit=NULL, and pid is the current process
(or 0), so the glib implementation of getrlimit can succeed.

BUG=chromium:662450
TEST=boot on ChromeOS w/ sandbox enabled
 No messages like:
   getrlimit(RLIMIT_NOFILE) failed

R=rickyz,jcliang,vapier

Review-Url: https://codereview.chromium.org/2484393004
Cr-Commit-Position: refs/heads/master@{#434872}

* Remove out-of-date comment in OnStreamEnd(); inline |buffer|.
* Remove non-HTTP/2 code path in OnStreamPadding().

Review-Url: https://codereview.chromium.org/2539563002
Cr-Commit-Position: refs/heads/master@{#434871}

Combine ash::TYPE_PLATFORM_APP and TYPE_WINDOWED_APP.
(V1 and V2 apps can be distinguished by IsPlatformApp functions)

Remove ChromeLauncherController::IsWindowedAppInLauncher.
(check for item presence and pinned state by id instead)

Remove unused LauncherItemController::TYPE_WINDOWED_APP.
The only behavior change *may* be the combination of drag groups.
(but afaict, we can already mix V1 and V2 app order on the shelf...)

Update users and tests.

BUG=557406
TEST=Automated; no Chrome OS shelf behavior changes for apps.
R=skuhne@chromium.org,jamescook@chromium.org

Review-Url: https://codereview.chromium.org/2518423003
Cr-Commit-Position: refs/heads/master@{#434870}

BUG=617539

Committed: https://crrev.com/bfbc065caf6187d00c0e549f07ad264bb1614873
Review-Url: https://codereview.chromium.org/2536523002
Cr-Original-Commit-Position: refs/heads/master@{#434612}
Cr-Commit-Position: refs/heads/master@{#434869}

Change http://crrev.com/21d17ca06e742cbfdcaaa54e0f4964248e3fbfff
incorrectly assumed the input channel layout never changes. This
patch fixes that issue by allowing the channel mask to change as
layouts change.

BUG=666813
TEST=updated unittests
TEST=http://storage.googleapis.com/chcunningham-chrome-shared/534301/aac_test.html

Review-Url: https://codereview.chromium.org/2536013002
Cr-Commit-Position: refs/heads/master@{#434868}

This code is no longer used. So remove this.

BUG=608017
TBR=piman@ because presubmit checks after the patch.

Review-Url: https://codereview.chromium.org/2535993002
Cr-Commit-Position: refs/heads/master@{#434867}

Revert of [DevTools] wrap async call frames in pausedDetails (patchset #2 id:20001 of https://codereview.chromium.org/2498263003/ )

Reason for revert:
We need to do it everywhere or nowhere.

Original issue's description:
> [DevTools] wrap async call frames in pausedDetails
>
> Currently CallStackSiderbarPane uses Protocol type directly and do a lot of other evil things.
> Wrapping async call frames is first step on the way to bright future.
>
> BUG=none
> R=lushnikov@chromium.org
>
> Committed: https://crrev.com/92e19e512ea0dcb0ed32bcf0508e6797135f98ea
> Cr-Commit-Position: refs/heads/master@{#432708}

TBR=lushnikov@chromium.org,dgozman@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=none

Review-Url: https://codereview.chromium.org/2532893003
Cr-Commit-Position: refs/heads/master@{#434866}

Remove useless class and struct forward declarations, and add
missing ones in gpu header files.

BUG=662195
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2532763002
Cr-Commit-Position: refs/heads/master@{#434865}

This is necessary to prevent the crash (crbug.com/669223) because
the handler object is destructed when the MD settings dialog is
closed.

BUG=669223
TEST=try

Review-Url: https://codereview.chromium.org/2534873003
Cr-Commit-Position: refs/heads/master@{#434864}

- This CL adds AshDevToolsDOMAgent as an observer to all widgets and views.
The WidgetRemovalObserver is specifically to remove the RootView subtree,
while the ViewObserver is to update any other descendant views.
AddViewTree was created so that whenever any |view| is added, it can
be pushed to the frontend.
- Add tests for when views are inserted, removed, rearranged,
removed & inserted.
- Add test to check if a widget owned by a native widget is represented
properly in the DOM.

BUG=648701

Review-Url: https://codereview.chromium.org/2529553002
Cr-Commit-Position: refs/heads/master@{#434863}

add has_element_in_any_list check in ElementAnimations::UpdateActivation during
PushProperities process. This solves a corner case if dom for animation be
removed before animation starts. See bug for details.

BUG=668086
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2527623002
Cr-Commit-Position: refs/heads/master@{#434862}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/dade4fe3..a72be232

Please follow these instructions for assigning/CC'ing issues:
https://github.com/v8/v8/wiki/Triaging%20issues

Please close rolling in case of a roll revert:
https://v8-roll.appspot.com/
This only works with a Google account.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

TBR=hablich@chromium.org,machenbach@chromium.org,littledan@chromium.org,vogelheim@chromium.org

Review-Url: https://codereview.chromium.org/2536903003
Cr-Commit-Position: refs/heads/master@{#434861}

Review-Url: https://codereview.chromium.org/2511033002
Cr-Commit-Position: refs/heads/master@{#434860}

Allow any integer types but always promote the result to unsigned.
This is more consistent with how these operations are normally used.

NOTRY=true

Review-Url: https://codereview.chromium.org/2529413002
Cr-Commit-Position: refs/heads/master@{#434859}

Previously the host binary wasn't rebuilt when building
remoting_dev_me2me_host. This CL adds direct dependency between these
targets to avoid this problem.

Review-Url: https://codereview.chromium.org/2535723003
Cr-Commit-Position: refs/heads/master@{#434858}

This patch gets rid of |DCHECK()|, which is a wrong assumption, in
|reconcileTextDecorationProperties()| since we have a test case against the
assumption.

BUG=663613
TEST=LayoutTests/editing/execCommand/insert-list-and-strikethrough.html

Review-Url: https://codereview.chromium.org/2532873002
Cr-Commit-Position: refs/heads/master@{#434857}

Many subpages show some form of details for a given extension. When the
backing data changes, those changes need to be reflected in the open
subpage. Right now, since the data is just set to the plain JS object,
it's not notified when a change occurs.

Add a subpageItem_ property to extensions page manager so that we can
leverage polymer databinding and link this to the subpages. This way, if
the backing data changes, the open subpage is updated.

Add a test for the same.

BUG=529395
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2526893002
Cr-Commit-Position: refs/heads/master@{#434856}

This puts the crash database in the user data directory, rather than in
the global shared one. In order to accomplish this, the variable
expansion from policy needs to be moved to be suitable for use in
chrome_elf. This code handles variable expansions in a registry key that
can be set to override the user data dir (in either HKLM or HKCU).

We do not need to completely remove the usage of DLLs other than
kernel32 in the code despite the code being used in chrome_elf, as in
all cases, it will not execute until after the DllMain() of chrome_elf
has completed, and we load the used functions via GetProcAddress(), not
by import lib. Currently, chrome_elf is signalled from WinMain() of
chrome to start crash reporting. In the near future it will be started
earlier from a background thread spawned in DllMain(). However, that
thread cannot execute until DllMain has completed, so it's then safe to
cause the other DLL loads that the variable expansion requires.

(Regarding the future thread-based spawning, ref:
https://blogs.msdn.microsoft.com/oldnewthing/20070904-00/?p=25283/ which
explains why the creation of the thread with which we don't synchronize
is OK, because DllMain calls are serialized. But we can discuss that
more in https://codereview.chromium.org/2475863004/ too. )

BUG=565446
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng

Review-Url: https://codereview.chromium.org/2487783002
Cr-Commit-Position: refs/heads/master@{#434855}

The detail view for extensions includes a button to open the extension's
home page (which defaults to the webstore page if an extension doens't
specify one). Hook this up, and add a test.

BUG=529395
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2522303002
Cr-Commit-Position: refs/heads/master@{#434854}

BUG=667208

Review-Url: https://codereview.chromium.org/2527913002
Cr-Commit-Position: refs/heads/master@{#434853}

Moving milestone to M58.

BUG=605661
TEST=messages logged with EME layout tests

Review-Url: https://codereview.chromium.org/2539573002
Cr-Commit-Position: refs/heads/master@{#434852}

Looks like a possible C&P error.  Compile-tested only.

BUG=none
TEST=none
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_nacl_sdk;master.tryserver.chromium.mac:mac_nacl_sdk;master.tryserver.chromium.win:win_nacl_sdk

Review-Url: https://codereview.chromium.org/2540513003
Cr-Commit-Position: refs/heads/master@{#434851}

//base's MRUCache allowed specifying key and value types, like
an underlying std::map<>, but the default base::MRUCache explicitly
specialized the comparator as std::less<>. Rather than require the
caller to overload std::less<> or operator<, allow the comparator
to be specified as a templated argument.

BUG=665735

Review-Url: https://codereview.chromium.org/2532313002
Cr-Commit-Position: refs/heads/master@{#434850}

Adding getCoalescedEvents API to the idl file
of PointerEventInit and getter method to
PointerEvent idl.
Also add the plumbing and creation in EventHandler
layer and PointerEventManager.

BUG=665937

Review-Url: https://codereview.chromium.org/2510133002
Cr-Commit-Position: refs/heads/master@{#434849}

This patch makes |Range::extractContents()| to postpone DOM mutation event
dispatching just before returning value rather than updating C++ variables
which holds |Node*| for every DOM mutation calls, e.g. Node::appendChild(),
Node::removeChild(), etc., to avoid running JavaScript code during in
|extractContents()| for simplicity.

For executing script in |RangeTest|, this patch makes |RangeTest| class to
derived from |EditingTestBase|.

This patch removes
"fast/dom/Range/range-created-in-mutation-event-crash.xhtml" since it causes
script error by calling |Range#getBoundingClientRect()| with |null| value and
coverage of this test is as same as newly added gTest.

This patch is similar to http://crrev.com/199383004 which makes
|Range::deleteContents()| to postpone DOM mutation event.

BUG=642537
TEST=run_webkit_unittets --gtest_filter=extractContentsWithDOMMutationEvent

Review-Url: https://codereview.chromium.org/2532843002
Cr-Commit-Position: refs/heads/master@{#434848}

This patch teaches UISourceCodeFrame to display messages from both
UISourceCodes of PersistenceBinding. This is reasonable, since
there might be different "builders" which add messages to either
of the source codes.

This patch also slightly modifies the handling of decorations:
in case of persistence binding, we prefer network uiSourceCode decorations
over the fileSystem ones: it's hard to merge decorations.

This is also an imperative step towards displaying fileSystem file
by default in the editor.

BUG=649837
R=dgozman

Review-Url: https://codereview.chromium.org/2530483002
Cr-Commit-Position: refs/heads/master@{#434847}

As they evolved to be quite different and the amount of
common code is minimal, it just make sense to define
which member function is used to render borders and which
is used for backgrounds.

BUG=662504

Review-Url: https://codereview.chromium.org/2536883002
Cr-Commit-Position: refs/heads/master@{#434846}

UpdaterState::GetState returns nullptr, if the current platform is
not Windows, and then the test calls at() method on the returned map.

This fixes LTO Linux bot:
https://build.chromium.org/p/chromium.fyi/builders/LTO%20Linux/builds/1560

BUG=615187

Review-Url: https://codereview.chromium.org/2536723004
Cr-Commit-Position: refs/heads/master@{#434845}

This patch makes CompositeEditCommand::moveParagraphs abort in the above
mentioned case, because there is no way to move the contents to a destination
inside itself.

There might be a deeper cause of passing such invalid parameter combination
to the function, which is not covered by this patch.

BUG=585048
TEST=webkit_unit_tests --gtest_filter=TypingCommandTest.insertLineBreakWithIllFormedHTML

Review-Url: https://codereview.chromium.org/2531263002
Cr-Commit-Position: refs/heads/master@{#434844}