ipv4: try to cache dst_entries which would cause a redirect

Not caching dst_entries which cause redirects could be exploited by hosts
on the same subnet, causing a severe DoS attack. This effect aggravated
since commit f88649721268999 ("ipv4: fix dst race in sk_dst_get()").

Lookups causing redirects will be allocated with DST_NOCACHE set which
will force dst_release to free them via RCU.  Unfortunately waiting for
RCU grace period just takes too long, we can end up with >1M dst_entries
waiting to be released and the system will run OOM. rcuos threads cannot
catch up under high softirq load.

Attaching the flag to emit a redirect later on to the specific skb allows
us to cache those dst_entries thus reducing the pressure on allocation
and deallocation.

This issue was discovered by Marcelo Leitner.

Cc: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Marcelo Leitner <mleitner@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: David S. Miller <davem@davemloft.net>

Conflicts:
	include/net/ip.h
	net/ipv4/route.c

Change-Id: I53e4b500a4db2f5fece937a42a3bd810b2640c44

include/net/ip.h

@@ -37,11 +37,12 @@ struct inet_skb_parm {
 	struct ip_options	opt;		/* Compiled IP options		*/
 	unsigned char		flags;
 
-#define IPSKB_FORWARDED		1
-#define IPSKB_XFRM_TUNNEL_SIZE	2
-#define IPSKB_XFRM_TRANSFORMED	4
-#define IPSKB_FRAG_COMPLETE	8
-#define IPSKB_REROUTED		16
+#define IPSKB_FORWARDED		BIT(0)
+#define IPSKB_XFRM_TUNNEL_SIZE	BIT(1)
+#define IPSKB_XFRM_TRANSFORMED	BIT(2)
+#define IPSKB_FRAG_COMPLETE	BIT(3)
+#define IPSKB_REROUTED		BIT(4)
+#define IPSKB_DOREDIRECT	BIT(5)
 };
 
 static inline unsigned int ip_hdrlen(const struct sk_buff *skb)

net/ipv4/ip_forward.c

@@ -107,7 +107,8 @@ int ip_forward(struct sk_buff *skb)
 	 *	We now generate an ICMP HOST REDIRECT giving the route
 	 *	we calculated.
 	 */
-	if (rt->rt_flags&RTCF_DOREDIRECT && !opt->srr && !skb_sec_path(skb))
+	if (IPCB(skb)->flags & IPSKB_DOREDIRECT && !opt->srr &&
+	    !skb_sec_path(skb))
 		ip_rt_send_redirect(skb);
 
 	skb->priority = rt_tos2priority(iph->tos);

net/ipv4/route.c

@@ -2123,9 +2123,10 @@ static int __mkroute_input(struct sk_buff *skb,
 		flags |= RTCF_DIRECTSRC;
 
 	if (out_dev == in_dev && err &&
+	    skb->protocol == htons(ETH_P_IP) &&
 	    (IN_DEV_SHARED_MEDIA(out_dev) ||
 	     inet_addr_onlink(out_dev, saddr, FIB_RES_GW(*res))))
-		flags |= RTCF_DOREDIRECT;
+		IPCB(skb)->flags |= IPSKB_DOREDIRECT;
 
 	if (skb->protocol != htons(ETH_P_IP)) {
 		/* Not IP (i.e. ARP). Do not create route, if it is
@@ -2954,6 +2955,8 @@ static int rt_fill_info(struct net *net,
 	r->rtm_flags	= (rt->rt_flags & ~0xFFFF) | RTM_F_CLONED;
 	if (rt->rt_flags & RTCF_NOTIFY)
 		r->rtm_flags |= RTM_F_NOTIFY;
+	if (IPCB(skb)->flags & IPSKB_DOREDIRECT)
+		r->rtm_flags |= RTCF_DOREDIRECT;
 
 	NLA_PUT_BE32(skb, RTA_DST, rt->rt_dst);