Re: [PATCH] audit: remove the audit freelist

[Florian Westphal <fw@strlen.de>]

Richard Guy Briggs <rgb@redhat.com> wrote:
> >  static void audit_buffer_free(struct audit_buffer *ab)
> >  {
> > -	unsigned long flags;
> > -
> >  	if (!ab)
> >  		return;
> >  
> >  	kfree_skb(ab->skb);
> > -	spin_lock_irqsave(&audit_freelist_lock, flags);
> > -	if (audit_freelist_count > AUDIT_MAXFREE)
> > -		kfree(ab);
> > -	else {
> > -		audit_freelist_count++;
> > -		list_add(&ab->list, &audit_freelist);
> > -	}
> > -	spin_unlock_irqrestore(&audit_freelist_lock, flags);
> > +	kfree(ab);
> >  }
[..]

> >  	nlh = nlmsg_put(ab->skb, 0, 0, type, 0, 0);
> >  	if (!nlh)
> > -		goto out_kfree_skb;
> > +		goto err;
> >  
> >  	return ab;
> >  
> > -out_kfree_skb:
> > -	kfree_skb(ab->skb);
> > -	ab->skb = NULL;
> 
> Why is the kfree_skb() skipped on error from nlmsg_put()?  I don't see
> much risk in nlmsg_put() failing considering the very simple arguments,
> however the code path is not trivial either.

if nlmsg_put fails we jump to err and ...

> >  err:
> >  	audit_buffer_free(ab);
> >  	return NULL;

... ab->skb gets free'd by audit_buffer_free() here.