Re: [RFC] New kobject/kset/ktype documentation and example code

[Cornelia Huck]

On Thu, 29 Nov 2007 10:47:19 -0500 (EST),
Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote:

> On Thu, 29 Nov 2007, Cornelia Huck wrote:
> 
> > > The only problem arises when a child's driver retains a reference to 
> > > the parent kobject.  If things are done properly, this reference should 
> > > involve incrementing the module count -- which would prevent the module 
> > > from being unloaded in the first place.
> > 
> > This still leaves the possibility that random code may grab a reference
> > once the kobject is present in the tree and lookupable. You gave up the
> > control of the number of references to your object once you made it
> > public.
> 
> I don't agree with this argument.  Code should never grab random
> references without insuring that the owner of the referenced object is
> pinned.  This rule applies to everything, not just kobjects.

Of course. You don't necessarily need prevent module unloading, but pin
the module in memory until the last reference is gone.

> Unfortunately kobjects don't have an owner field.  In practice this 
> means that it isn't possible to pin the owner of some random kobject -- 
> you have to know where the kobject came from or what it's embedded in.  
> All users of kobjects need to work this way.

Yeah, that is what the-patchset-I-have-to-dig-around-for does:
Introduce an owner field in the kobject for pinning the module. (IIRC,
you even did some of the patches?)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/