Re: Question about (or bug in?) the kobject implementation

[Greg KH <greg@kroah.com>]

On Wed, Mar 03, 2004 at 05:11:02PM -0500, Alan Stern wrote:
> On Wed, 3 Mar 2004, Greg KH wrote:
> 
> > On Fri, Feb 27, 2004 at 11:02:34PM -0500, Alan Stern wrote:
> > > We're actually discussing two different questions here.
> > > 
> > >     A.	Is it okay to call kobject_add() after calling kobject_del() -- 
> > > 	this was my original question.
> > 
> > No, this is not ok.  It might happen to work, but it is not valid.
> 
> I want to understand _why_ it is not valid.  Can you explain please?
> 
> From what you said earlier, I got the impression that calling _add() after 
> _del() is illegal because it runs the risk that the refcount may be 0 and 
> the object may be gone.

Yes, that is the risk.

> But if you have a separate valid reference, that can't happen.  Would
> it be legal then, or is there more to it?

Hm, it probably would work, hence the current working USB code :)
But I really don't want to "special case" anything here.  So it's easier
to say, "just don't do that".

thanks,

greg k-h