Cleanup WorkerThreadableWebSocketChannel's logic to wait for the main thread

Cleanup WorkerThreadableWebSocketChannel's logic to wait for the main thread

- Pass WorkerGlobalScope as a reference until we set it to RefPtr
- Wrap m_syncHelper with a function hasTerminatedPeer() for readability
- Stop holding reference on WorkerGlobalScope in
  WorkerThreadableWebSocketChannel
- Move ASSERT(m_syncHelper) to waitForMethodCompletion()
- Return a value indicating failure when waitForMethodCompletion()
  returns because of shutdown
- Make the methods not calling waitForMethodCompletion() also return
  without doing anything if the peer is already terminated
- Move task posting code into waitForMethodCompletion()
- More comments for readability

BUG=none

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=173608

[yhirano, 2014-05-07 08:05:17]

https://codereview.chromium.org/265713004/diff/1/Source/modules/websockets/Wo...
File Source/modules/websockets/WorkerThreadableWebSocketChannel.cpp (right):

https://codereview.chromium.org/265713004/diff/1/Source/modules/websockets/Wo...
Source/modules/websockets/WorkerThreadableWebSocketChannel.cpp:495:
RefPtr<Bridge> protect(this);
Can you tell me why this move is needed? I think posting a task can't delete
|this|.

https://codereview.chromium.org/265713004/diff/1/Source/modules/websockets/Wo...
Source/modules/websockets/WorkerThreadableWebSocketChannel.cpp:497: if
(!waitForMethodCompletion())
[optional] We could have postTaskAndWait for this pattern.

[tyoshino (SeeGerritForStatus), 2014-05-08 05:02:59]

https://codereview.chromium.org/265713004/diff/1/Source/modules/websockets/Wo...
File Source/modules/websockets/WorkerThreadableWebSocketChannel.cpp (right):

https://codereview.chromium.org/265713004/diff/1/Source/modules/websockets/Wo...
Source/modules/websockets/WorkerThreadableWebSocketChannel.cpp:495:
RefPtr<Bridge> protect(this);
On 2014/05/07 08:05:18, yhirano wrote:
> Can you tell me why this move is needed? I think posting a task can't delete
> |this|.

As discussed offline, it looks we no longer need these protects. But for this
CL, let's keep them and just let me make their position consistent.

https://codereview.chromium.org/265713004/diff/1/Source/modules/websockets/Wo...
Source/modules/websockets/WorkerThreadableWebSocketChannel.cpp:497: if
(!waitForMethodCompletion())
On 2014/05/07 08:05:18, yhirano wrote:
> [optional] We could have postTaskAndWait for this pattern.

Done.

[yhirano, 2014-05-08 05:09:41]

lgtm

Can "Create RefPtr<Bridge> for protection before calling postTaskToLoader" be
deleted from the description?

[tyoshino (SeeGerritForStatus), 2014-05-08 05:34:03]

On 2014/05/08 05:09:41, yhirano wrote:
> lgtm
> 
> Can "Create RefPtr<Bridge> for protection before calling postTaskToLoader" be
> deleted from the description?

Removed.

[tyoshino (SeeGerritForStatus), 2014-05-08 05:34:12]

The CQ bit was checked by tyoshino@chromium.org

[commit-bot: I haz the power, 2014-05-08 05:34:37]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/tyoshino@chromium.org/265713004/20001

[commit-bot: I haz the power, 2014-05-08 07:17:42]

Change committed as 173608