Make ProfileSyncService's WeakPtrFactory the last member

Make ProfileSyncService's WeakPtrFactory the last member

This will ensure that WeakPtrs are invalidated before any other members'
destructors are invoked. This is also one of the last classes needing to
change to allow us to turn on a Clang check for WeakPtrFactory member order.
Please see the bug for more details.

BUG=303818
Committed: https://crrev.com/bd319a43669b5dff4ac77246df7a154026e12bdd
Cr-Commit-Position: refs/heads/master@{#308015}

[dmichael (off chromium), 2014-12-11 19:08:26]

dmichael@chromium.org changed reviewers:
+ stanisc@chromium.org

[dmichael (off chromium), 2014-12-11 19:08:27]

Please let me know if you would rather this be structured differently. For
example, we could use an Init method on each Controllers. The current way makes
the distinction between the two callbacks clearer for BackupRollbackController,
but I'm open to other ideas.

Thanks!

[dmichael (off chromium), 2014-12-11 19:28:02]

CC anujk.sharma

[dmichael (off chromium), 2014-12-11 21:08:59]

Hold off, I forgot something obvious. I'll switchover to an Init method.

[chromium-reviews, 2014-12-11 21:27:28]

How about changing PSS's startup_controller_ and
backup_rollback_controller_ members to scoped_ptrs to delay construction
until after the weak ptr factory has been created?  In PSS's ctor body you
would .reset(new ...) them.  The advantage of this approach is that it
doesn't force construct-then-init semantics on the controllers (no worries
about forgetting to call Init before other methods).

On Thu, Dec 11, 2014 at 1:08 PM, <dmichael@chromium.org> wrote:
>
> Hold off, I forgot something obvious. I'll switchover to an Init method.
>
> https://codereview.chromium.org/795003003/
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[dmichael (off chromium), 2014-12-11 21:30:19]

Good suggestion :) New patch up.

On Thu, Dec 11, 2014 at 2:27 PM, 'Nick Maniscalco' via Chromium-reviews <
chromium-reviews@chromium.org> wrote:

> How about changing PSS's startup_controller_ and
> backup_rollback_controller_ members to scoped_ptrs to delay construction
> until after the weak ptr factory has been created?  In PSS's ctor body you
> would .reset(new ...) them.  The advantage of this approach is that it
> doesn't force construct-then-init semantics on the controllers (no worries
> about forgetting to call Init before other methods).
>
> On Thu, Dec 11, 2014 at 1:08 PM, <dmichael@chromium.org> wrote:
>>
>> Hold off, I forgot something obvious. I'll switchover to an Init method.
>>
>> https://codereview.chromium.org/795003003/
>>
>  To unsubscribe from this group and stop receiving emails from it, send an
> email to chromium-reviews+unsubscribe@chromium.org.
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[stanisc, 2014-12-11 21:44:41]

I was trying to suggest the same approach with scoped_ptrs but the review
tool keeps failing to post my feedback.

On Thu, Dec 11, 2014 at 1:30 PM, David Michael <dmichael@chromium.org>
wrote:

> Good suggestion :) New patch up.
>
> On Thu, Dec 11, 2014 at 2:27 PM, 'Nick Maniscalco' via Chromium-reviews <
> chromium-reviews@chromium.org> wrote:
>
>> How about changing PSS's startup_controller_ and
>> backup_rollback_controller_ members to scoped_ptrs to delay construction
>> until after the weak ptr factory has been created?  In PSS's ctor body you
>> would .reset(new ...) them.  The advantage of this approach is that it
>> doesn't force construct-then-init semantics on the controllers (no worries
>> about forgetting to call Init before other methods).
>>
>> On Thu, Dec 11, 2014 at 1:08 PM, <dmichael@chromium.org> wrote:
>>>
>>> Hold off, I forgot something obvious. I'll switchover to an Init method.
>>>
>>> https://codereview.chromium.org/795003003/
>>>
>>  To unsubscribe from this group and stop receiving emails from it, send
>> an email to chromium-reviews+unsubscribe@chromium.org.
>>
>
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[stanisc, 2014-12-11 22:40:12]

lgtm

[dmichael (off chromium), 2014-12-11 22:41:44]

The CQ bit was checked by dmichael@chromium.org

[commit-bot: I haz the power, 2014-12-11 23:11:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/795003003/20001

[stanisc, 2014-12-11 23:35:13]

lgtm

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[commit-bot: I haz the power, 2014-12-12 01:05:47]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2014-12-12 01:06:41]

Patchset 2 (id:??) landed as
https://crrev.com/bd319a43669b5dff4ac77246df7a154026e12bdd
Cr-Commit-Position: refs/heads/master@{#308015}