896506 - GC: Fix a couple of rooting hazards in XUL code

Status: RESOLVED FIXED

(Core :: XUL, defect)

Description

[Jon Coppeard (:jonco)]

Attachment: Proposed change

This fixes a couple of reported rooting hazards in XUL code.

Is the change to nsXULPDGlobalObject_finalize() safe?  I couldn't see a reason why not to do this (and it doesn't seem to break anything).

Comment 1

[Olli Pettay [:smaug][[e-mail]]]

Comment on attachment 779218 [details] [diff] [review]
Proposed change

nsXULPDGlobalObject *nativeThis ->
nsXULPDGlobalObject* nativeThis

(I know the old! code uses wrong style.
The code is from 1.25 <[e-mail]> 2001-05-08 09:46)

So this is not really fixing any issues, but changes the code so that
tools don't complain about rooting hazard, right?

Comment 2

[Jon Coppeard (:jonco)]

(In reply to Olli Pettay [:smaug] from comment #1)

Yes this is really about silencing warnings so we can get down to zero (or very few) hazards, so that it becomes much more obvious when any real hazards are introduced.

I'll have a look into why the tool thought that the query interface call in the finalizer could cause a GC, I don't think that's possible.

Comment 3

[Jon Coppeard (:jonco)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/9e8ba1191aef

Comment 4

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/9e8ba1191aef