Remove need for virtual JSObject::unwrappedObject

• Component: New Bugs
• Status: RESOLVED FIXED

[Sam Weinig]

Remove need for virtual JSObject::unwrappedObject

[Sam Weinig]

Created attachment 112738 [details]
Patch

[Sam Weinig]

Created attachment 112996 [details]
Patch

[Sam Weinig]

Created attachment 113000 [details]
Patch

[Filip Pizlo]

Comment on attachment 113000 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=113000&action=review

> Source/JavaScriptCore/runtime/JSGlobalThis.cpp:33
> +ASSERT_CLASS_FITS_IN_CELL(JSGlobalThis);

Is this needed, now that the GC can allocate things of any size?

[Sam Weinig]

(In reply to comment #4)
> (From update of attachment 113000 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=113000&action=review
> 
> > Source/JavaScriptCore/runtime/JSGlobalThis.cpp:33
> > +ASSERT_CLASS_FITS_IN_CELL(JSGlobalThis);
> 
> Is this needed, now that the GC can allocate things of any size?

I don't know, I moved it from JSDOMWindowShell. If it isn't, should we remove all of them?

[Geoffrey Garen]

Comment on attachment 113000 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=113000&action=review

r=me

> Source/JavaScriptCore/runtime/JSGlobalThis.h:51
> +    JSObject* unwrappedObject();

Please make this return JSGlobalObject*, since it's not the case that we're capable of wrapping any kind of JSObject.

> Source/JavaScriptCore/runtime/JSGlobalThis.h:68
> +    WriteBarrier<JSGlobalObject> m_globalVariables;

Please change this name to match its accessor ("unwrappedObject()").

[Sam Weinig]

Committed r98909: <http://trac.webkit.org/changeset/98909>