On Sun, Apr 10, 2005 at 05:29:26AM +1000, herbert wrote:
>
> > On Sat, 2005-04-09 at 07:12, Herbert Xu wrote:
> >
> > > ===== net/xfrm/xfrm_policy.c 1.75 vs edited =====
> > > --- 1.75/net/xfrm/xfrm_policy.c 2005-04-01 16:24:20 +10:00
> > > +++ edited/net/xfrm/xfrm_policy.c 2005-04-09 18:02:53 +10:00
> > > @@ -216,8 +216,8 @@
> > >
> > > expired:
> > > read_unlock(&xp->lock);
> > > - km_policy_expired(xp, dir, 1);
> > > - xfrm_policy_delete(xp, dir);
> > > + if (!xfrm_policy_delete(xp, dir))
> > > + km_policy_expired(xp, dir, 1);
> > > xfrm_pol_put(xp);>
> > > }
> > >
> >
> > If the policy was already dead you will still return a 0 from
> > xfrm_policy_delete(). Fixable by policy_kill returning an int.
>
> You're right. The other callers of xfrm_policy_kill needs to
> check it too.
Actually, my code was right :)
xfrm_policy_kill can only be called by the one who removed the
policy from the linked list. Therefore it can never fail.
If this logic is wrong we will be getting fat warnings from
xfrm_policy_kill.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@xxxxxxxxxxxxxxxxxxx>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
|