Factor out setIndexedDBClientCreateFunction from blink::initialize

Factor out setIndexedDBClientCreateFunction from blink::initialize

This CL moves setIndexedDBClientCreateFunction to WorkerContentSettingsClient's constructor
and WebLocalFrame's constructor.

BUG=593269
Committed: https://crrev.com/3056d79df9ba4fe776a42455ebb1f495bd582ad2
Cr-Commit-Position: refs/heads/master@{#382538}

[haraken, 2016-03-22 06:26:31]

haraken@chromium.org changed reviewers:
+ sigbjornf@opera.com

[haraken, 2016-03-22 06:26:32]

PTAL

Ideally I want to remove IDBClientImpl, which is used just to overcome the
disalllowed dependency from modules/ to web/. ChromeClient is a mechanism to
overcome the disallowed dependency but we cannot use ChromeClient for
IDBClientImpl because IDBClientImpl is used for workers (ChromeClient is only
for the main thread). So I gave up removing IDBClientImpl in this CL.

Another idea is to move setIndexedDBClientCreateFunction to
ModulesIntializer.initialize().

[haraken, 2016-03-22 06:26:56]

BTW, this is the final patch of the initialization/shutdown clean-up.

[sof, 2016-03-22 07:19:53]

How about removing IndexedClientImpl entirely and have the indexeddb module do
the necessary init (if needed..) in ModulesInitializer::initialize() instead?

I could be overlooking some detail though.. workable?

[haraken, 2016-03-22 07:22:47]

On 2016/03/22 07:19:53, sof wrote:
> How about removing IndexedClientImpl entirely and have the indexeddb module do
> the necessary init (if needed..) in ModulesInitializer::initialize() instead?
> 
> I could be overlooking some detail though.. workable?

At first I tried that but it's not easy. IndexedClientImpl is used to bypass the
disallowed dependency from modules/ to web/ (to call allowIndexedDB). If we
remove the IndexedClientImpl, we need some alternative way to bypass the
dependency. We can use ChromeClient for the main thread but we don't yet have an
alternative way for worker threads.

[sof, 2016-03-22 07:46:04]

Thanks, now I understand the dep. constraint - could you add a comment next to
*Impl to explain?

lgtm

[haraken, 2016-03-22 08:33:38]

On 2016/03/22 07:46:04, sof wrote:
> Thanks, now I understand the dep. constraint - could you add a comment next to
> *Impl to explain?

Done.

[haraken, 2016-03-22 08:33:43]

The CQ bit was checked by haraken@chromium.org

[haraken, 2016-03-22 08:33:43]

The patchset sent to the CQ was uploaded after l-g-t-m from sigbjornf@opera.com
Link to the patchset: https://codereview.chromium.org/1823953002/#ps40001
(title: " ")

[commit-bot: I haz the power, 2016-03-22 08:33:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1823953002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1823953002/40001

[commit-bot: I haz the power, 2016-03-22 09:36:40]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-03-22 09:37:57]

Description was changed from

==========
Factor out setIndexedDBClientCreateFunction from blink::initialize

This CL moves setIndexedDBClientCreateFunction to WorkerContentSettingsClient's
constructor
and WebLocalFrame's constructor.

BUG=593269
==========

to

==========
Factor out setIndexedDBClientCreateFunction from blink::initialize

This CL moves setIndexedDBClientCreateFunction to WorkerContentSettingsClient's
constructor
and WebLocalFrame's constructor.

BUG=593269

Committed: https://crrev.com/3056d79df9ba4fe776a42455ebb1f495bd582ad2
Cr-Commit-Position: refs/heads/master@{#382538}
==========

[commit-bot: I haz the power, 2016-03-22 09:37:58]

Patchset 3 (id:??) landed as
https://crrev.com/3056d79df9ba4fe776a42455ebb1f495bd582ad2
Cr-Commit-Position: refs/heads/master@{#382538}

[haraken, 2016-03-30 09:34:06]

A revert of this CL (patchset #3 id:40001) has been created in
https://codereview.chromium.org/1847513002/ by haraken@chromium.org.

The reason for reverting is: This caused threading race:

https://bugs.chromium.org/p/chromium/issues/detail?id=599011#c1
.