[FileAPI] Clean up WebFileSystemImpl before Blink shutdown

[FileAPI] Clean up WebFileSystemImpl before Blink shutdown

WebFileSystemImpl should not outlive V8 instance, since it may have references to V8.
This CL ensures it deleted before Blink shutdown.

BUG=369525
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=269345

[tzik, 2014-05-08 05:19:22]

PTL

[tzik, 2014-05-08 06:07:14]

On 2014/05/08 05:19:22, tzik wrote:
> PTL

Sorry, let me cancel the review request. This is not enough to fix the problem.

[tzik, 2014-05-08 06:42:00]

Updated. PTL.

[tzik, 2014-05-08 07:15:17]

Adding James as the content/renderer OWNER.
Please take a look to the change.

[kinuko, 2014-05-08 07:44:08]

James looks to be ooo

https://codereview.chromium.org/270633009/diff/20001/content/child/fileapi/we...
File content/child/fileapi/webfilesystem_impl.cc (right):

https://codereview.chromium.org/270633009/diff/20001/content/child/fileapi/we...
content/child/fileapi/webfilesystem_impl.cc:347:
g_webfilesystem_tls.Pointer()->Set(NULL);
Don't we do this in WebFileSystemImpl dtor?

https://codereview.chromium.org/270633009/diff/20001/content/renderer/render_...
File content/renderer/render_thread_impl.cc (right):

https://codereview.chromium.org/270633009/diff/20001/content/renderer/render_...
content/renderer/render_thread_impl.cc:566:
WebFileSystemImpl::DeleteThreadSpecificInstance();
Should we do this in ChildThread::Shutdown() rather than here?  (For worker
process case, though we're deprecating it)

[tzik, 2014-05-08 08:05:30]

> James looks to be ooo
Indeed.

John, please take a look as a content/ OWNER?

https://codereview.chromium.org/270633009/diff/20001/content/child/fileapi/we...
File content/child/fileapi/webfilesystem_impl.cc (right):

https://codereview.chromium.org/270633009/diff/20001/content/child/fileapi/we...
content/child/fileapi/webfilesystem_impl.cc:347:
g_webfilesystem_tls.Pointer()->Set(NULL);
On 2014/05/08 07:44:09, kinuko wrote:
> Don't we do this in WebFileSystemImpl dtor?

Ah, yes, it's done in it. Reverted.

https://codereview.chromium.org/270633009/diff/20001/content/renderer/render_...
File content/renderer/render_thread_impl.cc (right):

https://codereview.chromium.org/270633009/diff/20001/content/renderer/render_...
content/renderer/render_thread_impl.cc:566:
WebFileSystemImpl::DeleteThreadSpecificInstance();
On 2014/05/08 07:44:09, kinuko wrote:
> Should we do this in ChildThread::Shutdown() rather than here?  (For worker
> process case, though we're deprecating it)

Done.

[kinuko, 2014-05-08 08:20:47]

lgtm

[jam, 2014-05-08 16:30:17]

lgtm

[tzik, 2014-05-08 16:51:58]

The CQ bit was checked by tzik@chromium.org

[commit-bot: I haz the power, 2014-05-08 16:55:04]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/tzik@chromium.org/270633009/40001

[commit-bot: I haz the power, 2014-05-09 17:04:13]

Change committed as 269345