ServiceWorker: Get registration info and its version attributes in one lock operation

ServiceWorker: Get registration info and its version attributes in one lock operation

Registration info and its version attributes owned by SWProviderContext can be
accessed from both the main thread and the worker thread, so they are protected
by a lock operation. However, getter functions for those info and attributes are
separated and returned values could be in an invalid state due to interleaved
operations.

This CL merges getter functions (registration() and GetVersionAttributes()) into
GetRegistrationInfoAndVersionAttributes() and avoids such an unexpected state.

BUG=437677
TEST=should pass all existing tests

Committed: https://crrev.com/0b705112ae022e84c89e4f71b4ea6c75835e0d30
Cr-Commit-Position: refs/heads/master@{#313470}

[nhiroki, 2015-01-28 05:30:39]

Patchset #2 (id:20001) has been deleted

[nhiroki, 2015-01-28 05:30:46]

Patchset #1 (id:1) has been deleted

[nhiroki, 2015-01-28 05:53:55]

nhiroki@chromium.org changed reviewers:
+ ksakamoto@chromium.org

[nhiroki, 2015-01-28 05:54:59]

Hi, can you review this? Thanks!

https://codereview.chromium.org/885443006/diff/40001/content/renderer/service...
File content/renderer/service_worker/embedded_worker_context_client.cc (right):

https://codereview.chromium.org/885443006/diff/40001/content/renderer/service...
content/renderer/service_worker/embedded_worker_context_client.cc:430: //
Register a registrationi and its version attributes with the dispatcher
registrationi -> registration

I'll fix this later.

[Kunihiko Sakamoto, 2015-01-28 06:38:55]

https://codereview.chromium.org/885443006/diff/40001/content/child/service_wo...
File content/child/service_worker/web_service_worker_provider_impl.cc (right):

https://codereview.chromium.org/885443006/diff/40001/content/child/service_wo...
content/child/service_worker/web_service_worker_provider_impl.cc:52: if
(!context_->IsAssociatedWithRegistration())
I wonder if this and GetRegistrationInfoAndVersionAttributes() below should be
atomic as well?

[nhiroki, 2015-01-28 07:36:58]

Patchset #2 (id:60001) has been deleted

[nhiroki, 2015-01-28 07:40:42]

Patchset #2 (id:80001) has been deleted

[nhiroki, 2015-01-28 07:42:27]

Thank you! Updated.

https://codereview.chromium.org/885443006/diff/40001/content/child/service_wo...
File content/child/service_worker/web_service_worker_provider_impl.cc (right):

https://codereview.chromium.org/885443006/diff/40001/content/child/service_wo...
content/child/service_worker/web_service_worker_provider_impl.cc:52: if
(!context_->IsAssociatedWithRegistration())
On 2015/01/28 06:38:55, Kunihiko Sakamoto wrote:
> I wonder if this and GetRegistrationInfoAndVersionAttributes() below should be
> atomic as well?

Good point. Done.

https://codereview.chromium.org/885443006/diff/40001/content/renderer/service...
File content/renderer/service_worker/embedded_worker_context_client.cc (right):

https://codereview.chromium.org/885443006/diff/40001/content/renderer/service...
content/renderer/service_worker/embedded_worker_context_client.cc:430: //
Register a registrationi and its version attributes with the dispatcher
On 2015/01/28 05:54:58, nhiroki (slow) wrote:
> registrationi -> registration
> 
> I'll fix this later.

Done.

[Kunihiko Sakamoto, 2015-01-28 07:50:39]

lgtm

[nhiroki, 2015-01-28 08:06:16]

The CQ bit was checked by nhiroki@chromium.org

[commit-bot: I haz the power, 2015-01-28 08:06:38]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/885443006/100001

[commit-bot: I haz the power, 2015-01-28 10:08:31]

Committed patchset #2 (id:100001)

[commit-bot: I haz the power, 2015-01-28 10:09:31]

Patchset 2 (id:??) landed as
https://crrev.com/0b705112ae022e84c89e4f71b4ea6c75835e0d30
Cr-Commit-Position: refs/heads/master@{#313470}