Oilpan: fix build after r188805, part 2.

Oilpan: fix build after r188805, part 2.

TBR=oilpan-reviews
BUG=451288
NOTRY=true

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=188865

[sof, 2015-01-23 08:43:42]

Please take a look.

[sof, 2015-01-23 08:43:46]

The CQ bit was checked by sigbjornf@opera.com

[commit-bot: I haz the power, 2015-01-23 08:44:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/871873002/1

[commit-bot: I haz the power, 2015-01-23 08:44:26]

Committed patchset #1 (id:1) as
https://src.chromium.org/viewvc/blink?view=rev&revision=188865

[haraken, 2015-01-23 09:30:02]

LGTM

[sof, 2015-01-23 10:11:17]

On 2015/01/23 09:30:02, haraken wrote:
> LGTM

Making FetchManager::Loader a ContextLifecycleObserver (i.e., r188805) is
triggering a number of fetch-related crashes with Oilpan enabled now. Due to
loading not being cancelled.

Trying to understand why this is not a problem non-Oilpan, it seems like it
ought to be there too.

(Did not plan to do bugfixing in this area today, but will address. Bots have
gained some extra red.)

[sof, 2015-01-23 11:01:58]

On 2015/01/23 10:11:17, sof wrote:
> On 2015/01/23 09:30:02, haraken wrote:
> > LGTM
> 
> Making FetchManager::Loader a ContextLifecycleObserver (i.e., r188805) is
> triggering a number of fetch-related crashes with Oilpan enabled now. Due to
> loading not being cancelled.
> 
> Trying to understand why this is not a problem non-Oilpan, it seems like it
> ought to be there too.
> 
> (Did not plan to do bugfixing in this area today, but will address. Bots have
> gained some extra red.)

An OwnPtr<> taking care of prompt finalization&cancellation on the non-Oilpan
side; added an explicit dispose() step via
https://codereview.chromium.org/872703002/