Sync: Refactoring of DEVICE_INFO syncable type - Part 1

Sync: Refactoring of DEVICE_INFO syncable type - Part 1

This reintroduces change 283461 (issue 367153005) that has been reverted due to
memory leaks in some unit tests.
The memory leak was in a mock implementation of ProfileSyncComponentsFactory.
The factory is responsible for creating an instance of LocalDeviceInfoProvider,
which it does on demand and immediately transfer ownership of
LocalDeviceInfoProvider to the caller.
But due to the way mocking was implemented in ProfileSyncComponentsFactoryMock
a mock instance of LocalDeviceInfoProvider was created in the constructor and
if a test never called CreateLocalDeviceInfoProvider(), it would be leaked.

In this change I moved away from using ON_CALL mocking mechanism that resulted
in the leak to a custom implementation that avoids that problem.
See original changes in the patch #1 and the fix in the patch #2.

The original description copied from issue 367153005:

This change introduces a new class LocalDeviceInfoProvider that is responsible
for providing the local device specific DeviceInfo/cache_guid. It initializes
the data asynchronously and allows consumers to get notified when the data
becomes available.

LocalDeviceInfoProvider will allow to remove these responsibilities from
SyncedDeviceTracker / ProfileSyncService and loose coupling between
ProfileSyncService and a number of SyncableService and DataTypeController
derived classes.

LocalDeviceInfoProvider is hosted on the frontend thread. Since it needs
cache_guid to initialize DeviceInfo, which is currently available only on the
backend, I updated SyncBackendHostCore and SyncBackendHostImpl to pass
cache_guid to the frontend via OnBackendInitialized.

For the time being SyncedDeviceTracker remains unchanged
and continues to initialize the local device info too. The entire class will be
removed in the Part 2 of the change which will move SyncedDeviceTracker
functionality to a new SyncableService.

LocalDeviceInfoProvider also replaces SessionsSyncManager::SyncInternalApiDelegate
which was previously used to decouple SessionsSyncManager from the specifics of
of the local device info implementation.
SessionsSyncManager is changed to consume LocalDeviceInfoProvider instead of
SessionsSyncManager::SyncInternalApiDelegate.
SessionDataTypeController now also consumes LocalDeviceInfoProvider to make sure
that SESSIONS type model doesn't start before local device info becomes available.
A very similar approach will be used in the upcoming second part of the change
for DEVICE_INFO DataTypeController and SyncableType.

The change includes new unit tests for LocalDeviceInfoProvider and
SessionDataTypeController.

TBR=maniscalco@chromium.org,sky@chromium.org,zea@chromium.org
BUG=395349
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284561

[stanisc, 2014-07-18 01:14:13]

Richard, please review the change between the original patch (Patch 1) and the
memory leak fix (Patch 2). The rest of the files are exactly the same and have
already been reviewed.

[rlarocque, 2014-07-18 01:19:35]

Thanks for making the changes clear.

LGTM.

You'll need some extra approvals from owners of non-sync code before you can
commit.

[stanisc, 2014-07-18 01:19:41]

I wanted to add that I verified that unit_test memory leaks are gone in
"linux_asan" stdio output. All other failures seem to be unrelated.

[stanisc, 2014-07-18 01:29:11]

I've added original reviewers from the previews change that has been reverted.
Please note that all files in patch set 1 are exactly the same that have already
been reviewed. The only differences are in Patch set 2.
I added the same reviewers to TBR as well.

maniscalco@chromium.org: Please review changes under chrome/browser/ui/sync
(unchanged from the previews CR) 

sky@chromium.org: Please review files under
chrome/browser/extensions/api/
chrome/browser/ui/cocoa/
chrome/browser/ui/toolbar
(All of these files are unchanged from the previous CR). 

zea@chromium.org: Feel free to review all other sync changes.

[sky, 2014-07-18 15:37:18]

Files you asked me to review LGTM

https://codereview.chromium.org/398423002/diff/20001/chrome/browser/extension...
File chrome/browser/extensions/api/sessions/sessions_apitest.cc (right):

https://codereview.chromium.org/398423002/diff/20001/chrome/browser/extension...
chrome/browser/extensions/api/sessions/sessions_apitest.cc:123: 
nit: no newline here.

[stanisc, 2014-07-18 22:37:48]

The CQ bit was checked by stanisc@chromium.org

[commit-bot: I haz the power, 2014-07-18 22:40:33]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/stanisc@chromium.org/398423002/20001

[commit-bot: I haz the power, 2014-07-18 23:42:58]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  linux_chromium_chromeos_clang_dbg on tryserver.chromium
(http://build.chromium.org/p/tryserver.chromium/builders/linux_chromium_chrome...)

[commit-bot: I haz the power, 2014-07-18 23:55:16]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-07-18 23:55:17]

Try jobs failed on following builders:
  linux_chromium_chromeos_clang_dbg on tryserver.chromium
(http://build.chromium.org/p/tryserver.chromium/builders/linux_chromium_chrome...)

[stanisc, 2014-07-21 23:20:03]

The CQ bit was checked by stanisc@chromium.org

[commit-bot: I haz the power, 2014-07-21 23:21:20]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/stanisc@chromium.org/398423002/60001

[commit-bot: I haz the power, 2014-07-22 00:36:29]

Change committed as 284561