[Sync] Consolidate DataTypeController methods that post on the backend thread

[Sync] Consolidate DataTypeController methods that post on the backend thread

Make NonFrontendDataTypeController:{Start,Stop}AssociationAsync()
and NewNonFrontendDataTypeController::StopLocalService{,Async}()
private non-virtual methods and make them use the new method
NonFrontendDataTypeController::PostTaskOnBackendThread(from_here, task).

This is a prerequisite for fixing the intermittent crash in
NewNonFrontendDataTypeController::StartAssociation() on the
migration integration tests.

Fix a leak introduced by 111794 (see bug 105667).

BUG=107743, 105667
TEST=


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=120167

[akalin, 2012-01-31 04:14:59]

+zea for review

[Nicolas Zea, 2012-01-31 23:21:01]

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
File chrome/browser/sync/glue/autofill_data_type_controller.cc (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
chrome/browser/sync/glue/autofill_data_type_controller.cc:37: return
BrowserThread::PostTask(BrowserThread::DB, FROM_HERE, task);
FROM_HERE vs from_here.

Although using from_here, we'll lose the datatype information. But using
FROM_HERE, we don't know what the source of the task is. Up to you which you
prefer, as long as it's consistent. I think I'm partial to maintaining the task
source (e.g. from_here).

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
File chrome/browser/sync/glue/autofill_profile_data_type_controller.h (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
chrome/browser/sync/glue/autofill_profile_data_type_controller.h:45: //
NewNonFrontendDataTypeController implementation.
remove extra leading space

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/new_no...
File chrome/browser/sync/glue/new_non_frontend_data_type_controller_unittest.cc
(right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/new_no...
chrome/browser/sync/glue/new_non_frontend_data_type_controller_unittest.cc:95:
return BrowserThread::PostTask(BrowserThread::DB, FROM_HERE, task);
FROM_HERE vs from_here

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
File chrome/browser/sync/glue/non_frontend_data_type_controller.h (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
chrome/browser/sync/glue/non_frontend_data_type_controller.h:76: //
successfully, false otherwise.
move successfully to previous line

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
File chrome/browser/sync/glue/non_frontend_data_type_controller_unittest.cc
(right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
chrome/browser/sync/glue/non_frontend_data_type_controller_unittest.cc:80:
return BrowserThread::PostTask(BrowserThread::DB, FROM_HERE, task);
FROM_HERE vs from_here

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/typed_...
File chrome/browser/sync/glue/typed_url_data_type_controller.cc (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/typed_...
chrome/browser/sync/glue/typed_url_data_type_controller.cc:74: this, task)),
I think this should fix the typed url datatype controller leak. If so, perhaps
mention bug 105667 in the commit message.

[akalin, 2012-02-01 02:50:06]

PTAL

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
File chrome/browser/sync/glue/autofill_data_type_controller.cc (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
chrome/browser/sync/glue/autofill_data_type_controller.cc:37: return
BrowserThread::PostTask(BrowserThread::DB, FROM_HERE, task);
On 2012/01/31 23:21:01, nzea wrote:
> FROM_HERE vs from_here.
> 
> Although using from_here, we'll lose the datatype information. But using
> FROM_HERE, we don't know what the source of the task is. Up to you which you
> prefer, as long as it's consistent. I think I'm partial to maintaining the
task
> source (e.g. from_here).

Oops, I did mean to use from_here.  Done.

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
File chrome/browser/sync/glue/autofill_profile_data_type_controller.h (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/autofi...
chrome/browser/sync/glue/autofill_profile_data_type_controller.h:45: //
NewNonFrontendDataTypeController implementation.
On 2012/01/31 23:21:01, nzea wrote:
> remove extra leading space

Done.

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/new_no...
File chrome/browser/sync/glue/new_non_frontend_data_type_controller_unittest.cc
(right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/new_no...
chrome/browser/sync/glue/new_non_frontend_data_type_controller_unittest.cc:95:
return BrowserThread::PostTask(BrowserThread::DB, FROM_HERE, task);
On 2012/01/31 23:21:01, nzea wrote:
> FROM_HERE vs from_here

Done.

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
File chrome/browser/sync/glue/non_frontend_data_type_controller.h (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
chrome/browser/sync/glue/non_frontend_data_type_controller.h:76: //
successfully, false otherwise.
On 2012/01/31 23:21:01, nzea wrote:
> move successfully to previous line

Done.

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
File chrome/browser/sync/glue/non_frontend_data_type_controller_unittest.cc
(right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/non_fr...
chrome/browser/sync/glue/non_frontend_data_type_controller_unittest.cc:80:
return BrowserThread::PostTask(BrowserThread::DB, FROM_HERE, task);
On 2012/01/31 23:21:01, nzea wrote:
> FROM_HERE vs from_here

Done.

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/typed_...
File chrome/browser/sync/glue/typed_url_data_type_controller.cc (right):

http://codereview.chromium.org/9264062/diff/1/chrome/browser/sync/glue/typed_...
chrome/browser/sync/glue/typed_url_data_type_controller.cc:74: this, task)),
On 2012/01/31 23:21:01, nzea wrote:
> I think this should fix the typed url datatype controller leak. If so, perhaps
> mention bug 105667 in the commit message.

Okay.  I'm having trouble getting heapchecker to work locally so I'll update the
patchset later with the suppression removal.

[Nicolas Zea, 2012-02-01 18:08:36]

lgtm

[commit-bot: I haz the power, 2012-02-02 04:46:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/akalin@chromium.org/9264062/10027

[commit-bot: I haz the power, 2012-02-02 06:11:16]

Change committed as 120167