WebSWClientConnection should do IPC to StorageProcess if its WebSWOriginTable is not yet initialized

[youenn fablet]

WebSWClientConnection should do IPC to StorageProcess if its WebSWOriginTable is not yet initialized

[youenn fablet]

Created attachment 326877 [details]
Patch

[Chris Dumez]

Comment on attachment 326877 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=326877&action=review

> Source/WebKit/WebProcess/Storage/WebSWOriginTable.h:43
> +    bool isInitialized() const { return m_serviceWorkerOriginTable.sharedMemory(); }

The StorageProcess only sends a SharedMemoryHandle once the store actually contains at least one item. A store can be initialized and have no shared memory, in which case it merely means it is empty. Going to the StorageProcess in such cases is wrong and unnecessarily inefficient.

> LayoutTests/ChangeLog:8
> +        * http/tests/workers/service/basic-unregister.https-expected.txt:

Please explain the test changes.

[youenn fablet]

(In reply to Chris Dumez from comment #2)
> Comment on attachment 326877 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=326877&action=review
> 
> > Source/WebKit/WebProcess/Storage/WebSWOriginTable.h:43
> > +    bool isInitialized() const { return m_serviceWorkerOriginTable.sharedMemory(); }
> 
> The StorageProcess only sends a SharedMemoryHandle once the store actually
> contains at least one item. A store can be initialized and have no shared
> memory, in which case it merely means it is empty. Going to the
> StorageProcess in such cases is wrong and unnecessarily inefficient.

OK, let's make StorageProcess do an IPC to state that the table is empty then.

[youenn fablet]

Created attachment 326889 [details]
Patch

[Chris Dumez]

Comment on attachment 326889 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=326889&action=review

> LayoutTests/ChangeLog:8
> +        Updated tests to use hasServiceWorkerRegistration instead of hasServiceWorkerRegisteredForOrigin.

Since you kill all the call sites to hasServiceWorkerRegisteredForOrigin(), please kill internals.hasServiceWorkerRegisteredForOrigin() too.

[youenn fablet]

> > LayoutTests/ChangeLog:8
> > +        Updated tests to use hasServiceWorkerRegistration instead of hasServiceWorkerRegisteredForOrigin.
> 
> Since you kill all the call sites to hasServiceWorkerRegisteredForOrigin(),
> please kill internals.hasServiceWorkerRegisteredForOrigin() too.

It is removed from Internals.cpp/h/idl.

[Chris Dumez]

Comment on attachment 326889 [details]
Patch

r=me

[WebKit Commit Bot]

Comment on attachment 326889 [details]
Patch

Clearing flags on attachment: 326889

Committed r224832: <https://trac.webkit.org/changeset/224832>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Radar WebKit Bug Importer]

<rdar://problem/35561838>