net: Abstract default ADVMSS behind an accessor.

Make all RTAX_ADVMSS metric accesses go through a new helper function,
dst_metric_advmss().

Leave the actual default metric as "zero" in the real metric slot,
and compute the actual default value dynamically via a new dst_ops
AF specific callback.

For stacked IPSEC routes, we use the advmss of the path which
preserves existing behavior.

Unlike ipv4/ipv6, DecNET ties the advmss to the mtu and thus updates
advmss on pmtu updates.  This inconsistency in advmss handling
results in more raw metric accesses than I wish we ended up with.

Signed-off-by: David S. Miller <davem@davemloft.net>

11 files changed, 75 insertions, 33 deletions

diff --git a/drivers/scsi/cxgbi/libcxgbi.c b/drivers/scsi/cxgbi/libcxgbi.c
index be56617..d2ad3d6 100644
--- a/drivers/scsi/cxgbi/libcxgbi.c
+++ b/drivers/scsi/cxgbi/libcxgbi.c
@@ -825,7 +825,7 @@ unsigned int cxgbi_sock_select_mss(struct cxgbi_sock *csk, unsigned int pmtu)
 	unsigned int idx;
 	struct dst_entry *dst = csk->dst;
 
-	csk->advmss = dst_metric(dst, RTAX_ADVMSS);
+	csk->advmss = dst_metric_advmss(dst);
 
 	if (csk->advmss > pmtu - 40)
 		csk->advmss = pmtu - 40;
diff --git a/include/net/dst.h b/include/net/dst.h
index 755ac6c..03a1c3d 100644
--- a/include/net/dst.h
+++ b/include/net/dst.h
@@ -112,10 +112,22 @@ dst_metric_raw(const struct dst_entry *dst, const int metric)
 static inline u32
 dst_metric(const struct dst_entry *dst, const int metric)
 {
-	WARN_ON_ONCE(metric == RTAX_HOPLIMIT);
+	WARN_ON_ONCE(metric == RTAX_HOPLIMIT ||
+		     metric == RTAX_ADVMSS);
 	return dst_metric_raw(dst, metric);
 }
 
+static inline u32
+dst_metric_advmss(const struct dst_entry *dst)
+{
+	u32 advmss = dst_metric_raw(dst, RTAX_ADVMSS);
+
+	if (!advmss)
+		advmss = dst->ops->default_advmss(dst);
+
+	return advmss;
+}
+
 static inline void dst_metric_set(struct dst_entry *dst, int metric, u32 val)
 {
 	dst->_metrics[metric-1] = val;
diff --git a/include/net/dst_ops.h b/include/net/dst_ops.h
index 51665b3..15fb7af 100644
--- a/include/net/dst_ops.h
+++ b/include/net/dst_ops.h
@@ -16,6 +16,7 @@ struct dst_ops {
 
 	int			(*gc)(struct dst_ops *ops);
 	struct dst_entry *	(*check)(struct dst_entry *, __u32 cookie);
+	unsigned int		(*default_advmss)(const struct dst_entry *);
 	void			(*destroy)(struct dst_entry *);
 	void			(*ifdown)(struct dst_entry *,
 					  struct net_device *dev, int how);
diff --git a/net/decnet/af_decnet.c b/net/decnet/af_decnet.c
index 0065e7e..2af15b1 100644
--- a/net/decnet/af_decnet.c
+++ b/net/decnet/af_decnet.c
@@ -829,7 +829,7 @@ static int dn_confirm_accept(struct sock *sk, long *timeo, gfp_t allocation)
 		return -EINVAL;
 
 	scp->state = DN_CC;
-	scp->segsize_loc = dst_metric(__sk_dst_get(sk), RTAX_ADVMSS);
+	scp->segsize_loc = dst_metric_advmss(__sk_dst_get(sk));
 	dn_send_conn_conf(sk, allocation);
 
 	prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
@@ -958,7 +958,7 @@ static int __dn_connect(struct sock *sk, struct sockaddr_dn *addr, int addrlen,
 	sk->sk_route_caps = sk->sk_dst_cache->dev->features;
 	sock->state = SS_CONNECTING;
 	scp->state = DN_CI;
-	scp->segsize_loc = dst_metric(sk->sk_dst_cache, RTAX_ADVMSS);
+	scp->segsize_loc = dst_metric_advmss(sk->sk_dst_cache);
 
 	dn_nsp_send_conninit(sk, NSP_CI);
 	err = -EINPROGRESS;
diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
index e2e9268..b8a5c05 100644
--- a/net/decnet/dn_route.c
+++ b/net/decnet/dn_route.c
@@ -110,6 +110,7 @@ static unsigned long dn_rt_deadline;
 
 static int dn_dst_gc(struct dst_ops *ops);
 static struct dst_entry *dn_dst_check(struct dst_entry *, __u32);
+static unsigned int dn_dst_default_advmss(const struct dst_entry *dst);
 static struct dst_entry *dn_dst_negative_advice(struct dst_entry *);
 static void dn_dst_link_failure(struct sk_buff *);
 static void dn_dst_update_pmtu(struct dst_entry *dst, u32 mtu);
@@ -129,6 +130,7 @@ static struct dst_ops dn_dst_ops = {
 	.gc_thresh =		128,
 	.gc =			dn_dst_gc,
 	.check =		dn_dst_check,
+	.default_advmss =	dn_dst_default_advmss,
 	.negative_advice =	dn_dst_negative_advice,
 	.link_failure =		dn_dst_link_failure,
 	.update_pmtu =		dn_dst_update_pmtu,
@@ -245,7 +247,8 @@ static void dn_dst_update_pmtu(struct dst_entry *dst, u32 mtu)
 		}
 		if (!(dst_metric_locked(dst, RTAX_ADVMSS))) {
 			u32 mss = mtu - DN_MAX_NSP_DATA_HEADER;
-			if (dst_metric(dst, RTAX_ADVMSS) > mss)
+			u32 existing_mss = dst_metric_raw(dst, RTAX_ADVMSS);
+			if (!existing_mss || existing_mss > mss)
 				dst_metric_set(dst, RTAX_ADVMSS, mss);
 		}
 	}
@@ -795,12 +798,17 @@ static int dn_rt_bug(struct sk_buff *skb)
 	return NET_RX_DROP;
 }
 
+static unsigned int dn_dst_default_advmss(const struct dst_entry *dst)
+{
+	return dn_mss_from_pmtu(dst->dev, dst_mtu(dst));
+}
+
 static int dn_rt_set_next_hop(struct dn_route *rt, struct dn_fib_res *res)
 {
 	struct dn_fib_info *fi = res->fi;
 	struct net_device *dev = rt->dst.dev;
 	struct neighbour *n;
-	unsigned mss;
+	unsigned int metric;
 
 	if (fi) {
 		if (DN_FIB_RES_GW(*res) &&
@@ -820,10 +828,12 @@ static int dn_rt_set_next_hop(struct dn_route *rt, struct dn_fib_res *res)
 	if (dst_metric(&rt->dst, RTAX_MTU) == 0 ||
 	    dst_metric(&rt->dst, RTAX_MTU) > rt->dst.dev->mtu)
 		dst_metric_set(&rt->dst, RTAX_MTU, rt->dst.dev->mtu);
-	mss = dn_mss_from_pmtu(dev, dst_mtu(&rt->dst));
-	if (dst_metric(&rt->dst, RTAX_ADVMSS) == 0 ||
-	    dst_metric(&rt->dst, RTAX_ADVMSS) > mss)
-		dst_metric_set(&rt->dst, RTAX_ADVMSS, mss);
+	metric = dst_metric_raw(&rt->dst, RTAX_ADVMSS);
+	if (metric) {
+		unsigned int mss = dn_mss_from_pmtu(dev, dst_mtu(&rt->dst));
+		if (metric > mss)
+			dst_metric_set(&rt->dst, RTAX_ADVMSS, mss);
+	}
 	return 0;
 }
 
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 770f704..8099733 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -139,6 +139,7 @@ static unsigned long expires_ljiffies;
  */
 
 static struct dst_entry *ipv4_dst_check(struct dst_entry *dst, u32 cookie);
+static unsigned int	 ipv4_default_advmss(const struct dst_entry *dst);
 static void		 ipv4_dst_destroy(struct dst_entry *dst);
 static struct dst_entry *ipv4_negative_advice(struct dst_entry *dst);
 static void		 ipv4_link_failure(struct sk_buff *skb);
@@ -155,6 +156,7 @@ static struct dst_ops ipv4_dst_ops = {
 	.protocol =		cpu_to_be16(ETH_P_IP),
 	.gc =			rt_garbage_collect,
 	.check =		ipv4_dst_check,
+	.default_advmss =	ipv4_default_advmss,
 	.destroy =		ipv4_dst_destroy,
 	.ifdown =		ipv4_dst_ifdown,
 	.negative_advice =	ipv4_negative_advice,
@@ -383,8 +385,7 @@ static int rt_cache_seq_show(struct seq_file *seq, void *v)
 			(__force u32)r->rt_gateway,
 			r->rt_flags, atomic_read(&r->dst.__refcnt),
 			r->dst.__use, 0, (__force u32)r->rt_src,
-			(dst_metric(&r->dst, RTAX_ADVMSS) ?
-			     (int)dst_metric(&r->dst, RTAX_ADVMSS) + 40 : 0),
+			dst_metric_advmss(&r->dst) + 40,
 			dst_metric(&r->dst, RTAX_WINDOW),
 			(int)((dst_metric(&r->dst, RTAX_RTT) >> 3) +
 			      dst_metric(&r->dst, RTAX_RTTVAR)),
@@ -1798,6 +1799,19 @@ static void set_class_tag(struct rtable *rt, u32 tag)
 }
 #endif
 
+static unsigned int ipv4_default_advmss(const struct dst_entry *dst)
+{
+	unsigned int advmss = dst_metric_raw(dst, RTAX_ADVMSS);
+
+	if (advmss == 0) {
+		advmss = max_t(unsigned int, dst->dev->mtu - 40,
+			       ip_rt_min_advmss);
+		if (advmss > 65535 - 40)
+			advmss = 65535 - 40;
+	}
+	return advmss;
+}
+
 static void rt_set_nexthop(struct rtable *rt, struct fib_result *res, u32 itag)
 {
 	struct dst_entry *dst = &rt->dst;
@@ -1823,11 +1837,7 @@ static void rt_set_nexthop(struct rtable *rt, struct fib_result *res, u32 itag)
 
 	if (dst_mtu(dst) > IP_MAX_MTU)
 		dst_metric_set(dst, RTAX_MTU, IP_MAX_MTU);
-	if (dst_metric(dst, RTAX_ADVMSS) == 0)
-		dst_metric_set(dst, RTAX_ADVMSS,
-			       max_t(unsigned int, dst->dev->mtu - 40,
-				     ip_rt_min_advmss));
-	if (dst_metric(dst, RTAX_ADVMSS) > 65535 - 40)
+	if (dst_metric_raw(dst, RTAX_ADVMSS) > 65535 - 40)
 		dst_metric_set(dst, RTAX_ADVMSS, 65535 - 40);
 
 #ifdef CONFIG_NET_CLS_ROUTE
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 4fc3387..f401102 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1436,7 +1436,7 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
 
 	tcp_mtup_init(newsk);
 	tcp_sync_mss(newsk, dst_mtu(dst));
-	newtp->advmss = dst_metric(dst, RTAX_ADVMSS);
+	newtp->advmss = dst_metric_advmss(dst);
 	if (tcp_sk(sk)->rx_opt.user_mss &&
 	    tcp_sk(sk)->rx_opt.user_mss < newtp->advmss)
 		newtp->advmss = tcp_sk(sk)->rx_opt.user_mss;
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 97041f2..2d39066 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -119,9 +119,13 @@ static __u16 tcp_advertise_mss(struct sock *sk)
 	struct dst_entry *dst = __sk_dst_get(sk);
 	int mss = tp->advmss;
 
-	if (dst && dst_metric(dst, RTAX_ADVMSS) < mss) {
-		mss = dst_metric(dst, RTAX_ADVMSS);
-		tp->advmss = mss;
+	if (dst) {
+		unsigned int metric = dst_metric_advmss(dst);
+
+		if (metric < mss) {
+			mss = metric;
+			tp->advmss = mss;
+		}
 	}
 
 	return (__u16)mss;
@@ -2422,7 +2426,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
 
 	skb_dst_set(skb, dst_clone(dst));
 
-	mss = dst_metric(dst, RTAX_ADVMSS);
+	mss = dst_metric_advmss(dst);
 	if (tp->rx_opt.user_mss && tp->rx_opt.user_mss < mss)
 		mss = tp->rx_opt.user_mss;
 
@@ -2556,7 +2560,7 @@ static void tcp_connect_init(struct sock *sk)
 
 	if (!tp->window_clamp)
 		tp->window_clamp = dst_metric(dst, RTAX_WINDOW);
-	tp->advmss = dst_metric(dst, RTAX_ADVMSS);
+	tp->advmss = dst_metric_advmss(dst);
 	if (tp->rx_opt.user_mss && tp->rx_opt.user_mss < tp->advmss)
 		tp->advmss = tp->rx_opt.user_mss;
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 98796b0..d9cb832 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -76,6 +76,7 @@
 
 static struct rt6_info * ip6_rt_copy(struct rt6_info *ort);
 static struct dst_entry	*ip6_dst_check(struct dst_entry *dst, u32 cookie);
+static unsigned int	 ip6_default_advmss(const struct dst_entry *dst);
 static struct dst_entry *ip6_negative_advice(struct dst_entry *);
 static void		ip6_dst_destroy(struct dst_entry *);
 static void		ip6_dst_ifdown(struct dst_entry *,
@@ -103,6 +104,7 @@ static struct dst_ops ip6_dst_ops_template = {
 	.gc			=	ip6_dst_gc,
 	.gc_thresh		=	1024,
 	.check			=	ip6_dst_check,
+	.default_advmss		=	ip6_default_advmss,
 	.destroy		=	ip6_dst_destroy,
 	.ifdown			=	ip6_dst_ifdown,
 	.negative_advice	=	ip6_negative_advice,
@@ -937,8 +939,12 @@ static void ip6_rt_update_pmtu(struct dst_entry *dst, u32 mtu)
 
 static int ipv6_get_mtu(struct net_device *dev);
 
-static inline unsigned int ipv6_advmss(struct net *net, unsigned int mtu)
+static unsigned int ip6_default_advmss(const struct dst_entry *dst)
 {
+	struct net_device *dev = dst->dev;
+	unsigned int mtu = dst_mtu(dst);
+	struct net *net = dev_net(dev);
+
 	mtu -= sizeof(struct ipv6hdr) + sizeof(struct tcphdr);
 
 	if (mtu < net->ipv6.sysctl.ip6_rt_min_advmss)
@@ -990,7 +996,6 @@ struct dst_entry *icmp6_dst_alloc(struct net_device *dev,
 	atomic_set(&rt->dst.__refcnt, 1);
 	dst_metric_set(&rt->dst, RTAX_HOPLIMIT, 255);
 	dst_metric_set(&rt->dst, RTAX_MTU, ipv6_get_mtu(rt->rt6i_dev));
-	dst_metric_set(&rt->dst, RTAX_ADVMSS, ipv6_advmss(net, dst_mtu(&rt->dst)));
 	rt->dst.output  = ip6_output;
 
 #if 0	/* there's no chance to use these for ndisc */
@@ -1312,8 +1317,6 @@ install_route:
 
 	if (!dst_mtu(&rt->dst))
 		dst_metric_set(&rt->dst, RTAX_MTU, ipv6_get_mtu(dev));
-	if (!dst_metric(&rt->dst, RTAX_ADVMSS))
-		dst_metric_set(&rt->dst, RTAX_ADVMSS, ipv6_advmss(net, dst_mtu(&rt->dst)));
 	rt->dst.dev = dev;
 	rt->rt6i_idev = idev;
 	rt->rt6i_table = table;
@@ -1540,8 +1543,6 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *src,
 	nrt->rt6i_nexthop = neigh_clone(neigh);
 	/* Reset pmtu, it may be better */
 	dst_metric_set(&nrt->dst, RTAX_MTU, ipv6_get_mtu(neigh->dev));
-	dst_metric_set(&nrt->dst, RTAX_ADVMSS, ipv6_advmss(dev_net(neigh->dev),
-							   dst_mtu(&nrt->dst)));
 
 	if (ip6_ins_rt(nrt))
 		goto out;
@@ -1971,7 +1972,6 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
 	rt->rt6i_dev = net->loopback_dev;
 	rt->rt6i_idev = idev;
 	dst_metric_set(&rt->dst, RTAX_MTU, ipv6_get_mtu(rt->rt6i_dev));
-	dst_metric_set(&rt->dst, RTAX_ADVMSS, ipv6_advmss(net, dst_mtu(&rt->dst)));
 	dst_metric_set(&rt->dst, RTAX_HOPLIMIT, -1);
 	rt->dst.obsolete = -1;
 
@@ -2041,7 +2041,6 @@ static int rt6_mtu_change_route(struct rt6_info *rt, void *p_arg)
 {
 	struct rt6_mtu_change_arg *arg = (struct rt6_mtu_change_arg *) p_arg;
 	struct inet6_dev *idev;
-	struct net *net = dev_net(arg->dev);
 
 	/* In IPv6 pmtu discovery is not optional,
 	   so that RTAX_MTU lock cannot disable it.
@@ -2073,7 +2072,6 @@ static int rt6_mtu_change_route(struct rt6_info *rt, void *p_arg)
 	     (dst_mtu(&rt->dst) < arg->mtu &&
 	      dst_mtu(&rt->dst) == idev->cnf.mtu6))) {
 		dst_metric_set(&rt->dst, RTAX_MTU, arg->mtu);
-		dst_metric_set(&rt->dst, RTAX_ADVMSS, ipv6_advmss(net, arg->mtu));
 	}
 	return 0;
 }
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index fee0768..20aa95e 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1521,7 +1521,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
 
 	tcp_mtup_init(newsk);
 	tcp_sync_mss(newsk, dst_mtu(dst));
-	newtp->advmss = dst_metric(dst, RTAX_ADVMSS);
+	newtp->advmss = dst_metric_advmss(dst);
 	tcp_initialize_rcv_mss(newsk);
 
 	newinet->inet_daddr = newinet->inet_saddr = LOOPBACK4_IPV6;
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 6e50ccd..36936c8 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -2361,6 +2361,11 @@ static int xfrm_bundle_ok(struct xfrm_policy *pol, struct xfrm_dst *first,
 	return 1;
 }
 
+static unsigned int xfrm_default_advmss(const struct dst_entry *dst)
+{
+	return dst_metric_advmss(dst->path);
+}
+
 int xfrm_policy_register_afinfo(struct xfrm_policy_afinfo *afinfo)
 {
 	struct net *net;
@@ -2378,6 +2383,8 @@ int xfrm_policy_register_afinfo(struct xfrm_policy_afinfo *afinfo)
 			dst_ops->kmem_cachep = xfrm_dst_cache;
 		if (likely(dst_ops->check == NULL))
 			dst_ops->check = xfrm_dst_check;
+		if (likely(dst_ops->default_advmss == NULL))
+			dst_ops->default_advmss = xfrm_default_advmss;
 		if (likely(dst_ops->negative_advice == NULL))
 			dst_ops->negative_advice = xfrm_negative_advice;
 		if (likely(dst_ops->link_failure == NULL))