GitLab

This is part of a suite of four patches:
Part 1: https://codereview.chromium.org/1097383006
Part 2: https://codereview.chromium.org/1099943003
Part 3: https://codereview.chromium.org/1100163002
Part 4: <this>

BUG=475141

Review URL: https://codereview.chromium.org/1056633003

Cr-Commit-Position: refs/heads/master@{#326792}

[ozone] Keep the queue of surfaceless buffers inside gl_surface_ozone and only dispatch them when the page flip ack callback for the previous one has been been called.

We previously had 2 queues, one in gl_surface_ozone one in the platform
display controller. Remove one of them since it shouldn't be necessary.

BUG=

Review URL: https://codereview.chromium.org/1091253003

Cr-Commit-Position: refs/heads/master@{#326791}

BUG=478090

Review URL: https://codereview.chromium.org/1086073006

Cr-Commit-Position: refs/heads/master@{#326790}

The remote command framework requires the remote job implementations
to call the callback asynchronously. Post task to the message loop
of current thread to fix this issue.

BUG=480982

Review URL: https://codereview.chromium.org/1102023002

Cr-Commit-Position: refs/heads/master@{#326789}

This is going to allow testing of permissions changes.

BUG=475141

Review URL: https://codereview.chromium.org/1101153002

Cr-Commit-Position: refs/heads/master@{#326788}

This issue landed here: https://codereview.chromium.org/1097553002/
And was reverted here: https://codereview.chromium.org/1072193012/
Because it failed flaky, due to second profile's history service not being initialized.

This CL relands the original, with a modification line 474 of app_controller_mac_browsertest.mm to call WaitForHistoryToLoad.

BUG=477392

Review URL: https://codereview.chromium.org/1102833002

Cr-Commit-Position: refs/heads/master@{#326787}

The VSyncProvider does not guarantee that the callback is ever called.
On Ozone-GBM, the VSyncProvider relies on CRTC information to determine
when the vsync deadline will occurr. Since we're headless, that
information cannot be use, so the callback will never be called.

BUG=471550

Review URL: https://codereview.chromium.org/1060433002

Cr-Commit-Position: refs/heads/master@{#326786}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/2761548..6682798

TBR=kbr@chromium.org

Review URL: https://codereview.chromium.org/1109533002

Cr-Commit-Position: refs/heads/master@{#326785}

BUG=476648,479710

Review URL: https://codereview.chromium.org/1100413002

Cr-Commit-Position: refs/heads/master@{#326784}

Also added a TraceDataEndpoint class that I'm thinking can be used to decouple the creation of the final trace (since there's some special logic in there for adding '{}'s and some specific names like 'traceEvent') from where it gets written. This is used right now by the new BackgroundTracingManager for Slow Reports and Deep Reports, but we could also refactor the existing String and File sinks this way as a follow up CL (WIP here: https://codereview.chromium.org/1002103004/).

Then for Deep Reports we can specify the data sink like so:

TraceController::DisableRecording(
  TraceController::CreateCompressedStringSink(
    TraceController::CreateFileEndpoint(...)));

Review URL: https://codereview.chromium.org/1088673003

Cr-Commit-Position: refs/heads/master@{#326783}

For incremental conversion of iOS code from BrowserContextKeyedServiceFactories
to BrowserStateKeyedServiceFactories, iOS uses cross-dependencies between both
types of factories.

This CL had two virtual methods to KeyedServiceBaseFactory to convert a context
to the correct type (content::BrowserContext or web::BrowserState) and back so
that we can have BrowserStateKeyedServiceFactories registered into a
BrowserContextDependencyManager (temporary).

Implement the method in all derived classes and refactor the implementation
of GetAssociatedPrefRegistry that no longer needs to be virtual.

BUG=459020,478763

Review URL: https://codereview.chromium.org/1090373003

Cr-Commit-Position: refs/heads/master@{#326782}

BUG=

Review URL: https://codereview.chromium.org/1076443002

Cr-Commit-Position: refs/heads/master@{#326781}

In cases where a SpecialStoragePolicy is set, the channel id store could still
incorrectly persist some newly-added IDs to disk if their writes did not commit
before the store was destroyed.  This CL fixes the bug by forcing a commit of
all pending writes before issuing any policy-mandated deletes.

BUG=
TEST=

Review URL: https://codereview.chromium.org/1015233002

Cr-Commit-Position: refs/heads/master@{#326780}

I forgot the WebAudio case when I increased the silence lead-in;
increase that time by five seconds also.

BUG=480798,477653

Review URL: https://codereview.chromium.org/1105853002

Cr-Commit-Position: refs/heads/master@{#326779}

Cr-Commit-Position: refs/heads/master@{#326778}

BUG=470746,479171
TEST=manual

Review URL: https://codereview.chromium.org/1063753004

Cr-Commit-Position: refs/heads/master@{#326777}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/cb3d77a..2761548

TBR=kbr@chromium.org

Review URL: https://codereview.chromium.org/1093973003

Cr-Commit-Position: refs/heads/master@{#326776}

Because base/strings/string_split.h defines:
typedef std::vector<std::pair<std::string, std::string> > StringPairs;

BUG=412250

Review URL: https://codereview.chromium.org/1092923006

Cr-Commit-Position: refs/heads/master@{#326775}

The CL does the following:

1. Replace GuestViewManager::FromBrowserContext with GuestViewManager::CreateWithDelegate.
2. Rename GuestViewManager::FromBrowserContextIfAvailable to GuestViewManager::FromBrowserContxt.
3. Creates a GuestViewManagerDelegate interface and ExtensionsGuestViewManagerDelegate.
4. GuestViewManager is created with a particular delegate implementation.
5. GuestView permission checks and registration is moved to the ExtensionsGuestViewDelegate.
6. owner_extension_id() is renamed to owner_host().

After this patch the following classes can be moved to Components (once GuestViewEvent is decoupled from EventRouter):

1. GuestViewManager
2. GuestViewManagerDelegate
3. GuestViewBase
4. GuestViewMessageFilter
5. GuestView

BUG=444869
TBR=noms@chromium.org for chrome/browser/profiles, nkostylev@chromium.org chrome/browser/chromeos/login, thestig@chromium.org for chrome/browser/printing.

Review URL: https://codereview.chromium.org/1066563006

Cr-Commit-Position: refs/heads/master@{#326774}

Packed relative relocations do not currently participate in
relocation of shared RELRO pages into a service process's load
addresses.  This undermines RELRO sharing, by reducing the
number of swapped RELRO pages (often to zero).

Fix by processing packed relative relocations a second time when
relocating shared RELRO pages.  Also, tidy up and rationalize
some duplicated code through use of templates.

BUG=385553

Review URL: https://codereview.chromium.org/1099253002

Cr-Commit-Position: refs/heads/master@{#326773}

https://folio.googleplex.com/login-m44/Redlines

BUG=478449
TEST=manual

Review URL: https://codereview.chromium.org/1100623003

Cr-Commit-Position: refs/heads/master@{#326772}

Reland of https://codereview.chromium.org/1087133005/ (Add a WinHeap dump provider to the memory profiler.)

Reason for reland: A flaky test has been removed. (WinHeapDumpProviderTest.GetHeapInformation)

Add a WinHeap dump provider to the memory profiler.

TBR=nduca@chromium.org
BUG=466141

Review URL: https://codereview.chromium.org/1083003003

Cr-Commit-Position: refs/heads/master@{#326771}

Review URL: https://codereview.chromium.org/1104813003

Cr-Commit-Position: refs/heads/master@{#326770}

This CL adds extension icons in the install dialog as well as the "was installed" bubble.

Followup to https://codereview.chromium.org/855513002/

XIB changes in ExtensionInstallPromptBundle.xib:
* convert the "bundle items" NSTextField (filled with a multiline string) to an NSView (filled with dynamically created per-item subviews)
* add an NSImageView in the top right to hold the bundle item

XIB changes in ExtensionInstalledBubbleBundle.xib:
* convert the "installed" and "failed" NSTextFields (filled with multiline strings) to NSViews (filled with dynamically created per-item subviews)

BUG=448724

Review URL: https://codereview.chromium.org/816223008

Cr-Commit-Position: refs/heads/master@{#326769}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/fa1e205..cb3d77a

TBR=kbr@chromium.org

Review URL: https://codereview.chromium.org/1093503005

Cr-Commit-Position: refs/heads/master@{#326768}

Inject PrefService, prefs::kNtpMostVisitedURLsBlacklist and
HistoryService through the constructor of TopSitesImpl.

Add public method TopSitesFactory::BuildTopSites() that can
be used by tests to avoid code duplication.

BUG=390953,479174

Review URL: https://codereview.chromium.org/1080403003

Cr-Commit-Position: refs/heads/master@{#326767}

R=jochen@chromium.org
TBR=kalman@chromium.org
BUG=

Review URL: https://codereview.chromium.org/1085733006

Cr-Commit-Position: refs/heads/master@{#326766}

Please take a look at my post here

https://groups.google.com/a/chromium.org/forum/?fromgroups#!topic/chromium-dev/pIQjhkVl4cE

I am not sure about build/common.gypi
Please correct me if I am wrong.

PS:
This is just to start, there are more places to refactor.

BUG=

Review URL: https://codereview.chromium.org/1090213002

Cr-Commit-Position: refs/heads/master@{#326765}

Regarding the latest code on linux,
It was always setting by default language even if value of language was existed.
So we need to add routine for setting language via "spd_set_language".

BUG=289548

Review URL: https://codereview.chromium.org/1043933002

Cr-Commit-Position: refs/heads/master@{#326764}

BUG=429756

Review URL: https://codereview.chromium.org/1107583002

Cr-Commit-Position: refs/heads/master@{#326763}

This brings in the output streaming decode support.

BUG=472009

Review URL: https://codereview.chromium.org/1061993007

Cr-Commit-Position: refs/heads/master@{#326762}

Stats are updated and logged continuously when echo cancellation is enabled. Being enabled is no guarantee for actually running since it, for example, is enabled by default when opening the microphone.
All these invalid data clutter the UMA histogram.

The improvement is a check for the first frame containing an echo state before start logging. This can only happen when the echo cancellation is actually running.

BUG=450193

Review URL: https://codereview.chromium.org/1081063003

Cr-Commit-Position: refs/heads/master@{#326761}

This patch introduces the basic browser tests of download notification.

BUG=480489
TEST=run newly added browsertest

Review URL: https://codereview.chromium.org/1047243002

Cr-Commit-Position: refs/heads/master@{#326760}

Test that we're exposing the cursor position correctly
and that the value is replaced with bullets.

Depends on: https://codereview.chromium.org/1009543009/

BUG=84250

Review URL: https://codereview.chromium.org/1019633003

Cr-Commit-Position: refs/heads/master@{#326759}

When NonUIDataTypeController was stopped while in the ASSOCIATING
state, the models were not stopped even though this state is only
accessible when the models are already loaded.

BUG=479737

Review URL: https://codereview.chromium.org/1096093004

Cr-Commit-Position: refs/heads/master@{#326758}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/5894a94..fa1e205

TBR=rjwright@chromium.org,kbr@chromium.org

Review URL: https://codereview.chromium.org/1104793002

Cr-Commit-Position: refs/heads/master@{#326757}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/18297c96..5b5f7c18

Please follow these instructions for assigning/CC'ing issues:
https://code.google.com/p/v8-wiki/wiki/TriagingIssues

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

Review URL: https://codereview.chromium.org/1102883002

Cr-Commit-Position: refs/heads/master@{#326756}

The most important change here is getting rid of code in
renderer_accessibility.cc that recreated the subtree whenever
AXSelectedChildrenChanged was fired. This is no longer
necessary because Blink now fires the right events on each
individual option in a menu list or list box.

The rest of the changes just make sure the exact attributes
and events match the spec and match what Firefox and
Safari do when interacting with these elements, including
adding some good tests for those.

BUG=323462

Review URL: https://codereview.chromium.org/1057083002

Cr-Commit-Position: refs/heads/master@{#326755}

tested anywhere. Removed those from all related files. Recursive is
used only with NO_SORT so removed the option to use it in
combination with sorting.

Review URL: https://codereview.chromium.org/1096323003

Cr-Commit-Position: refs/heads/master@{#326754}

Originally SetVersionInternal() was quite simple function, but that has been
puzzuled as the codebase has evolved.

This CL decomposes the function for readability and simplifies lifetime
management for a version to be set, that is, passes a refcounted version
as an argument instead of making temporary protect object.

BUG=n/a
TEST=n/a

Review URL: https://codereview.chromium.org/1106723002

Cr-Commit-Position: refs/heads/master@{#326753}