Subj : Re: GC question
To   : netscape.public.mozilla.jseng
From : lyg
Date : Fri May 28 2004 11:11 am


"MA" <sukhisoul@yahoo.com> wrote in message
news:69bf8050.0405270519.af6863@posting.google.com...
> >
> >    Why? is the JS_MaybeGC never be called? Why the created AppObj and
its
> > properties is not freed when I JS_RemoveRoot()?
> >    what is wrong ?
> >
>
> All the objects you create are global objects. According to my
understanding,
> global objects are never GC'ed. So, you need to manually free them, not
> rely on GC to do that.
>
>
   because when i open a new document, i need a "root object" to manager the
objects created in the new document, I have to create many global objects
for the root objects. when I access the opened document, i just fetch the
rooted "root object" and JS_SetGlobalObject(). In other words, at the same
time, there is only one object as the global object and the others are
rooted not to GCed. When i want close the document, I will
JS_RemoveRoot(rootedObj) to let it GCed. Now it
is not the global object, why not be GCed?

> Secondly, you are passing a pointer to an object when calling
> JS_AddRoot. This is wrong. Pass a "pointer to object-pointer" when
> calling JS_AddRoot. This, of course, has nothing to do with your
> problem.
> >
   I think it is right to use JS_AddRoot(cx, &obj); not like JS_AddRoot(cx,
obj);

   the follow is cited from jsapi.h
  /**/
      A JS GC root is a pointer to a JSObject *, JSString *, or jsdouble *
that
 * itself points into the GC heap (more recently, we support this extension:
 * a root may be a pointer to a jsval v for which JSVAL_IS_GCTHING(v) is
true).
 *
 * Therefore, you never pass JSObject *obj to JS_AddRoot(cx, obj).  You
always
 * call JS_AddRoot(cx, &obj), passing obj by reference.  And later, before
obj
 * or the structure it is embedded within goes out of scope or is freed, you
 * must call JS_RemoveRoot(cx, &obj).


> MA

.