775365 - Fill out the rooting API

Status: RESOLVED FIXED

(Core :: JavaScript Engine, defect)

Description

[Steve Fink [:sfink] [:s:]]

 - add external type JSHandleString
 - add external type JSMutableHandleValue
 - allow converting a MutableHandle -> Handle
 - add MutableHandle::fromMarkedLocation()
 - make a non-const AudoVectorRooter::handleAt() that returns a MutableHandle

Comment 1

[Steve Fink [:sfink] [:s:]]

Attachment: augment rooting API

Comment 2

[David Mandelin [:dmandelin]]

Did we ever come to a conclusion about whether the typedefs are actually good and/or what their names should be?

Comment 3

[Steve Fink [:sfink] [:s:]]

(In reply to David Mandelin [:dmandelin] from comment #2)
> Did we ever come to a conclusion about whether the typedefs are actually
> good and/or what their names should be?

Well, there are two categories of typedefs, for external and internal use. For internal use, I don't think we've settled on the Handle<JSObject*> vs HandleObject question, and I don't want to stall work on making that decision.

For external use, the typedefs are necessary for C users of the API, which we could pull the plug on at any time. Once we do that, it's mostly the same question as the internal use (though I could imagine wanting typedefs only for external users.)

Comment 4

[Steve Fink [:sfink] [:s:]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/b20344afee94

Comment 5

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/b20344afee94

Comment 6

[David Mandelin [:dmandelin]]

(In reply to Steve Fink [:sfink] from comment #3)
> (In reply to David Mandelin [:dmandelin] from comment #2)
> > Did we ever come to a conclusion about whether the typedefs are actually
> > good and/or what their names should be?
> 
> Well, there are two categories of typedefs, for external and internal use.
> For internal use, I don't think we've settled on the Handle<JSObject*> vs
> HandleObject question, and I don't want to stall work on making that
> decision.

Sure.

> For external use, the typedefs are necessary for C users of the API, which
> we could pull the plug on at any time. Once we do that, it's mostly the same
> question as the internal use (though I could imagine wanting typedefs only
> for external users.)

Yeah. I suppose jsd probably still wants to work for now, or something.