net/ipv6: separate handling of FIB entries from dst based routes

Last step before flipping the data type for FIB entries:
- use fib6_info_alloc to create FIB entries in ip6_route_info_create
  and addrconf_dst_alloc
- use fib6_info_release in place of dst_release, ip6_rt_put and
  rt6_release
- remove the dst_hold before calling __ip6_ins_rt or ip6_del_rt
- when purging routes, drop per-cpu routes
- replace inc and dec of rt6i_ref with fib6_info_hold and fib6_info_release
- use rt->from since it points to the FIB entry
- drop references to exception bucket, fib6_metrics and per-cpu from
  dst entries (those are relevant for fib entries only)
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/ip6_fib.h

@@ -314,9 +314,7 @@ static inline u32 rt6_get_cookie(const struct rt6_info *rt)
 
 	if (rt->rt6i_flags & RTF_PCPU ||
 	    (unlikely(!list_empty(&rt->rt6i_uncached)) && rt->from))
-		rt = rt->from;
-
-	rt6_get_cookie_safe(rt, &cookie);
+		rt6_get_cookie_safe(rt->from, &cookie);
 
 	return cookie;
 }

include/net/ip6_route.h

@@ -114,8 +114,7 @@ static inline int ip6_route_get_saddr(struct net *net, struct rt6_info *rt,
 				      unsigned int prefs,
 				      struct in6_addr *saddr)
 {
-	struct inet6_dev *idev =
-			rt ? ip6_dst_idev((struct dst_entry *)rt) : NULL;
+	struct inet6_dev *idev = rt ? rt->rt6i_idev : NULL;
 	int err = 0;
 
 	if (rt && rt->rt6i_prefsrc.plen)

net/ipv6/addrconf.c

@@ -916,7 +916,6 @@ void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp)
 		pr_warn("Freeing alive inet6 address %p\n", ifp);
 		return;
 	}
-	ip6_rt_put(ifp->rt);
 
 	kfree_rcu(ifp, rcu);
 }
@@ -1102,8 +1101,8 @@ ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr,
 	inet6addr_notifier_call_chain(NETDEV_UP, ifa);
 out:
 	if (unlikely(err < 0)) {
-		if (rt)
-			ip6_rt_put(rt);
+		fib6_info_release(rt);
+
 		if (ifa) {
 			if (ifa->idev)
 				in6_dev_put(ifa->idev);
@@ -1191,7 +1190,7 @@ cleanup_prefix_route(struct inet6_ifaddr *ifp, unsigned long expires, bool del_r
 		else {
 			if (!(rt->rt6i_flags & RTF_EXPIRES))
 				fib6_set_expires(rt, expires);
-			ip6_rt_put(rt);
+			fib6_info_release(rt);
 		}
 	}
 }
@@ -2375,8 +2374,7 @@ static struct rt6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
 			continue;
 		if ((rt->rt6i_flags & noflags) != 0)
 			continue;
-		if (!dst_hold_safe(&rt->dst))
-			rt = NULL;
+		fib6_info_hold(rt);
 		break;
 	}
 out:
@@ -2687,7 +2685,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
 			addrconf_prefix_route(&pinfo->prefix, pinfo->prefix_len,
 					      dev, expires, flags, GFP_ATOMIC);
 		}
-		ip6_rt_put(rt);
+		fib6_info_release(rt);
 	}
 
 	/* Try to figure out our local address for this prefix */
@@ -3361,7 +3359,7 @@ static int fixup_permanent_addr(struct net *net,
 		ifp->rt = rt;
 		spin_unlock(&ifp->lock);
 
-		ip6_rt_put(prev);
+		fib6_info_release(prev);
 	}
 
 	if (!(ifp->flags & IFA_F_NOPREFIXROUTE)) {
@@ -5636,8 +5634,8 @@ static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
 				ip6_del_rt(net, rt);
 		}
 		if (ifp->rt) {
-			if (dst_hold_safe(&ifp->rt->dst))
 			ip6_del_rt(net, ifp->rt);
+			ifp->rt = NULL;
 		}
 		rt_genid_bump_ipv6(net);
 		break;

net/ipv6/anycast.c

@@ -213,7 +213,7 @@ static void aca_put(struct ifacaddr6 *ac)
 {
 	if (refcount_dec_and_test(&ac->aca_refcnt)) {
 		in6_dev_put(ac->aca_idev);
-		dst_release(&ac->aca_rt->dst);
+		fib6_info_release(ac->aca_rt);
 		kfree(ac);
 	}
 }
@@ -231,6 +231,7 @@ static struct ifacaddr6 *aca_alloc(struct rt6_info *rt,
 	aca->aca_addr = *addr;
 	in6_dev_hold(idev);
 	aca->aca_idev = idev;
+	fib6_info_hold(rt);
 	aca->aca_rt = rt;
 	aca->aca_users = 1;
 	/* aca_tstamp should be updated upon changes */
@@ -274,7 +275,7 @@ int __ipv6_dev_ac_inc(struct inet6_dev *idev, const struct in6_addr *addr)
 	}
 	aca = aca_alloc(rt, addr);
 	if (!aca) {
-		ip6_rt_put(rt);
+		fib6_info_release(rt);
 		err = -ENOMEM;
 		goto out;
 	}
@@ -330,7 +331,6 @@ int __ipv6_dev_ac_dec(struct inet6_dev *idev, const struct in6_addr *addr)
 	write_unlock_bh(&idev->lock);
 	addrconf_leave_solict(idev, &aca->aca_addr);
 
-	dst_hold(&aca->aca_rt->dst);
 	ip6_del_rt(dev_net(idev->dev), aca->aca_rt);
 
 	aca_put(aca);
@@ -358,7 +358,6 @@ void ipv6_ac_destroy_dev(struct inet6_dev *idev)
 
 		addrconf_leave_solict(idev, &aca->aca_addr);
 
-		dst_hold(&aca->aca_rt->dst);
 		ip6_del_rt(dev_net(idev->dev), aca->aca_rt);
 
 		aca_put(aca);

net/ipv6/ip6_fib.c

@@ -170,6 +170,7 @@ struct rt6_info *fib6_info_alloc(gfp_t gfp_flags)
 void fib6_info_destroy(struct rt6_info *f6i)
 {
 	struct rt6_exception_bucket *bucket;
+	struct dst_metrics *m;
 
 	WARN_ON(f6i->rt6i_node);
 
@@ -201,6 +202,10 @@ void fib6_info_destroy(struct rt6_info *f6i)
 	if (f6i->fib6_nh.nh_dev)
 		dev_put(f6i->fib6_nh.nh_dev);
 
+	m = f6i->fib6_metrics;
+	if (m != &dst_default_metrics && refcount_dec_and_test(&m->refcnt))
+		kfree(m);
+
 	kfree(f6i);
 }
 EXPORT_SYMBOL_GPL(fib6_info_destroy);
@@ -714,7 +719,7 @@ static struct fib6_node *fib6_add_1(struct net *net,
 			/* clean up an intermediate node */
 			if (!(fn->fn_flags & RTN_RTINFO)) {
 				RCU_INIT_POINTER(fn->leaf, NULL);
-				rt6_release(leaf);
+				fib6_info_release(leaf);
 			/* remove null_entry in the root node */
 			} else if (fn->fn_flags & RTN_TL_ROOT &&
 				   rcu_access_pointer(fn->leaf) ==
@@ -898,12 +903,32 @@ static void fib6_purge_rt(struct rt6_info *rt, struct fib6_node *fn,
 			if (!(fn->fn_flags & RTN_RTINFO) && leaf == rt) {
 				new_leaf = fib6_find_prefix(net, table, fn);
 				atomic_inc(&new_leaf->rt6i_ref);
+
 				rcu_assign_pointer(fn->leaf, new_leaf);
-				rt6_release(rt);
+				fib6_info_release(rt);
 			}
 			fn = rcu_dereference_protected(fn->parent,
 				    lockdep_is_held(&table->tb6_lock));
 		}
+
+		if (rt->rt6i_pcpu) {
+			int cpu;
+
+			/* release the reference to this fib entry from
+			 * all of its cached pcpu routes
+			 */
+			for_each_possible_cpu(cpu) {
+				struct rt6_info **ppcpu_rt;
+				struct rt6_info *pcpu_rt;
+
+				ppcpu_rt = per_cpu_ptr(rt->rt6i_pcpu, cpu);
+				pcpu_rt = *ppcpu_rt;
+				if (pcpu_rt) {
+					fib6_info_release(pcpu_rt->from);
+					pcpu_rt->from = NULL;
+				}
+			}
+		}
 	}
 }
 
@@ -1099,7 +1124,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt,
 		fib6_purge_rt(iter, fn, info->nl_net);
 		if (rcu_access_pointer(fn->rr_ptr) == iter)
 			fn->rr_ptr = NULL;
-		rt6_release(iter);
+		fib6_info_release(iter);
 
 		if (nsiblings) {
 			/* Replacing an ECMP route, remove all siblings */
@@ -1115,7 +1140,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt,
 					fib6_purge_rt(iter, fn, info->nl_net);
 					if (rcu_access_pointer(fn->rr_ptr) == iter)
 						fn->rr_ptr = NULL;
-					rt6_release(iter);
+					fib6_info_release(iter);
 					nsiblings--;
 					info->nl_net->ipv6.rt6_stats->fib_rt_entries--;
 				} else {
@@ -1183,9 +1208,6 @@ int fib6_add(struct fib6_node *root, struct rt6_info *rt,
 	int replace_required = 0;
 	int sernum = fib6_new_sernum(info->nl_net);
 
-	if (WARN_ON_ONCE(!atomic_read(&rt->dst.__refcnt)))
-		return -EINVAL;
-
 	if (info->nlh) {
 		if (!(info->nlh->nlmsg_flags & NLM_F_CREATE))
 			allow_create = 0;
@@ -1300,7 +1322,7 @@ int fib6_add(struct fib6_node *root, struct rt6_info *rt,
 			if (pn_leaf == rt) {
 				pn_leaf = NULL;
 				RCU_INIT_POINTER(pn->leaf, NULL);
-				atomic_dec(&rt->rt6i_ref);
+				fib6_info_release(rt);
 			}
 			if (!pn_leaf && !(pn->fn_flags & RTN_RTINFO)) {
 				pn_leaf = fib6_find_prefix(info->nl_net, table,
@@ -1312,7 +1334,7 @@ int fib6_add(struct fib6_node *root, struct rt6_info *rt,
 					    info->nl_net->ipv6.fib6_null_entry;
 				}
 #endif
-				atomic_inc(&pn_leaf->rt6i_ref);
+				fib6_info_hold(pn_leaf);
 				rcu_assign_pointer(pn->leaf, pn_leaf);
 			}
 		}
@@ -1334,10 +1356,6 @@ int fib6_add(struct fib6_node *root, struct rt6_info *rt,
 	     (fn->fn_flags & RTN_TL_ROOT &&
 	      !rcu_access_pointer(fn->leaf))))
 		fib6_repair_tree(info->nl_net, table, fn);
-	/* Always release dst as dst->__refcnt is guaranteed
-	 * to be taken before entering this function
-	 */
-	dst_release_immediate(&rt->dst);
 	return err;
 }
 
@@ -1637,7 +1655,7 @@ static struct fib6_node *fib6_repair_tree(struct net *net,
 				new_fn_leaf = net->ipv6.fib6_null_entry;
 			}
 #endif
-			atomic_inc(&new_fn_leaf->rt6i_ref);
+			fib6_info_hold(new_fn_leaf);
 			rcu_assign_pointer(fn->leaf, new_fn_leaf);
 			return pn;
 		}
@@ -1693,7 +1711,7 @@ static struct fib6_node *fib6_repair_tree(struct net *net,
 			return pn;
 
 		RCU_INIT_POINTER(pn->leaf, NULL);
-		rt6_release(pn_leaf);
+		fib6_info_release(pn_leaf);
 		fn = pn;
 	}
 }
@@ -1763,7 +1781,7 @@ static void fib6_del_route(struct fib6_table *table, struct fib6_node *fn,
 	call_fib6_entry_notifiers(net, FIB_EVENT_ENTRY_DEL, rt, NULL);
 	if (!info->skip_notify)
 		inet6_rt_notify(RTM_DELROUTE, rt, info, 0);
-	rt6_release(rt);
+	fib6_info_release(rt);
 }
 
 /* Need to own table->tb6_lock */
@@ -2261,9 +2279,8 @@ static int ipv6_route_seq_show(struct seq_file *seq, void *v)
 
 	dev = rt->fib6_nh.nh_dev;
 	seq_printf(seq, " %08x %08x %08x %08x %8s\n",
-		   rt->rt6i_metric, atomic_read(&rt->dst.__refcnt),
-		   rt->dst.__use, rt->rt6i_flags,
-		   dev ? dev->name : "");
+		   rt->rt6i_metric, atomic_read(&rt->rt6i_ref), 0,
+		   rt->rt6i_flags, dev ? dev->name : "");
 	iter->w.leaf = NULL;
 	return 0;
 }

net/ipv6/ip6_output.c

@@ -968,7 +968,8 @@ static int ip6_dst_lookup_tail(struct net *net, const struct sock *sk,
 		if (!had_dst)
 			*dst = ip6_route_output(net, sk, fl6);
 		rt = (*dst)->error ? NULL : (struct rt6_info *)*dst;
-		err = ip6_route_get_saddr(net, rt, &fl6->daddr,
+		err = ip6_route_get_saddr(net, rt ? rt->from : NULL,
+					  &fl6->daddr,
 					  sk ? inet6_sk(sk)->srcprefs : 0,
 					  &fl6->saddr);
 		if (err)

net/ipv6/ndisc.c

@@ -1283,7 +1283,7 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 			ND_PRINTK(0, err,
 				  "RA: %s got default router without neighbour\n",
 				  __func__);
-			ip6_rt_put(rt);
+			fib6_info_release(rt);
 			return;
 		}
 	}
@@ -1313,7 +1313,7 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 			ND_PRINTK(0, err,
 				  "RA: %s got default router without neighbour\n",
 				  __func__);
-			ip6_rt_put(rt);
+			fib6_info_release(rt);
 			return;
 		}
 		neigh->flags |= NTF_ROUTER;
@@ -1499,7 +1499,7 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 		ND_PRINTK(2, warn, "RA: invalid RA options\n");
 	}
 out:
-	ip6_rt_put(rt);
+	fib6_info_release(rt);
 	if (neigh)
 		neigh_release(neigh);
 }