netns xfrm: ipcomp support
[safe/jmp/linux-2.6] / net / ipv4 / ip_input.c
index 63ab523..c29de98 100644 (file)
@@ -5,12 +5,10 @@
  *
  *             The Internet Protocol (IP) module.
  *
- * Version:    $Id: ip_input.c,v 1.55 2002/01/12 07:39:45 davem Exp $
- *
  * Authors:    Ross Biro
  *             Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
  *             Donald Becker, <becker@super.org>
- *             Alan Cox, <Alan.Cox@linux.org>
+ *             Alan Cox, <alan@lxorguk.ukuu.org.uk>
  *             Richard Underwood
  *             Stefan Becker, <stefanb@yello.ping.de>
  *             Jorge Cwik, <jorge@laser.satlink.net>
 #include <linux/netlink.h>
 
 /*
- *     SNMP management statistics
- */
-
-DEFINE_SNMP_STAT(struct ipstats_mib, ip_statistics) __read_mostly;
-
-/*
  *     Process Router Attention IP option
  */
 int ip_call_ra_chain(struct sk_buff *skb)
@@ -160,6 +152,7 @@ int ip_call_ra_chain(struct sk_buff *skb)
        struct ip_ra_chain *ra;
        u8 protocol = ip_hdr(skb)->protocol;
        struct sock *last = NULL;
+       struct net_device *dev = skb->dev;
 
        read_lock(&ip_ra_lock);
        for (ra = ip_ra_chain; ra; ra = ra->next) {
@@ -168,12 +161,12 @@ int ip_call_ra_chain(struct sk_buff *skb)
                /* If socket is bound to an interface, only report
                 * the packet if it came  from that interface.
                 */
-               if (sk && inet_sk(sk)->num == protocol &&
+               if (sk && inet_sk(sk)->inet_num == protocol &&
                    (!sk->sk_bound_dev_if ||
-                    sk->sk_bound_dev_if == skb->dev->ifindex)) {
+                    sk->sk_bound_dev_if == dev->ifindex) &&
+                   net_eq(sock_net(sk), dev_net(dev))) {
                        if (ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)) {
-                               skb = ip_defrag(skb, IP_DEFRAG_CALL_RA_CHAIN);
-                               if (skb == NULL) {
+                               if (ip_defrag(skb, IP_DEFRAG_CALL_RA_CHAIN)) {
                                        read_unlock(&ip_ra_lock);
                                        return 1;
                                }
@@ -196,8 +189,10 @@ int ip_call_ra_chain(struct sk_buff *skb)
        return 0;
 }
 
-static inline int ip_local_deliver_finish(struct sk_buff *skb)
+static int ip_local_deliver_finish(struct sk_buff *skb)
 {
+       struct net *net = dev_net(skb->dev);
+
        __skb_pull(skb, ip_hdrlen(skb));
 
        /* Point into the IP datagram, just past the header. */
@@ -205,25 +200,26 @@ static inline int ip_local_deliver_finish(struct sk_buff *skb)
 
        rcu_read_lock();
        {
-               /* Note: See raw.c and net/raw.h, RAWV4_HTABLE_SIZE==MAX_INET_PROTOS */
                int protocol = ip_hdr(skb)->protocol;
-               int hash;
-               struct sock *raw_sk;
-               struct net_protocol *ipprot;
+               int hash, raw;
+               const struct net_protocol *ipprot;
 
        resubmit:
-               hash = protocol & (MAX_INET_PROTOS - 1);
-               raw_sk = sk_head(&raw_v4_htable[hash]);
-
-               /* If there maybe a raw socket we must check - if not we
-                * don't care less
-                */
-               if (raw_sk && !raw_v4_input(skb, ip_hdr(skb), hash))
-                       raw_sk = NULL;
+               raw = raw_local_deliver(skb, protocol);
 
-               if ((ipprot = rcu_dereference(inet_protos[hash])) != NULL) {
+               hash = protocol & (MAX_INET_PROTOS - 1);
+               ipprot = rcu_dereference(inet_protos[hash]);
+               if (ipprot != NULL) {
                        int ret;
 
+                       if (!net_eq(net, &init_net) && !ipprot->netns_ok) {
+                               if (net_ratelimit())
+                                       printk("%s: proto %d isn't netns-ready\n",
+                                               __func__, protocol);
+                               kfree_skb(skb);
+                               goto out;
+                       }
+
                        if (!ipprot->no_policy) {
                                if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
                                        kfree_skb(skb);
@@ -236,16 +232,16 @@ static inline int ip_local_deliver_finish(struct sk_buff *skb)
                                protocol = -ret;
                                goto resubmit;
                        }
-                       IP_INC_STATS_BH(IPSTATS_MIB_INDELIVERS);
+                       IP_INC_STATS_BH(net, IPSTATS_MIB_INDELIVERS);
                } else {
-                       if (!raw_sk) {
+                       if (!raw) {
                                if (xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
-                                       IP_INC_STATS_BH(IPSTATS_MIB_INUNKNOWNPROTOS);
+                                       IP_INC_STATS_BH(net, IPSTATS_MIB_INUNKNOWNPROTOS);
                                        icmp_send(skb, ICMP_DEST_UNREACH,
                                                  ICMP_PROT_UNREACH, 0);
                                }
                        } else
-                               IP_INC_STATS_BH(IPSTATS_MIB_INDELIVERS);
+                               IP_INC_STATS_BH(net, IPSTATS_MIB_INDELIVERS);
                        kfree_skb(skb);
                }
        }
@@ -265,12 +261,11 @@ int ip_local_deliver(struct sk_buff *skb)
         */
 
        if (ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)) {
-               skb = ip_defrag(skb, IP_DEFRAG_LOCAL_DELIVER);
-               if (!skb)
+               if (ip_defrag(skb, IP_DEFRAG_LOCAL_DELIVER))
                        return 0;
        }
 
-       return NF_HOOK(PF_INET, NF_IP_LOCAL_IN, skb, skb->dev, NULL,
+       return NF_HOOK(PF_INET, NF_INET_LOCAL_IN, skb, skb->dev, NULL,
                       ip_local_deliver_finish);
 }
 
@@ -288,28 +283,27 @@ static inline int ip_rcv_options(struct sk_buff *skb)
                                              --ANK (980813)
        */
        if (skb_cow(skb, skb_headroom(skb))) {
-               IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS);
+               IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_INDISCARDS);
                goto drop;
        }
 
        iph = ip_hdr(skb);
+       opt = &(IPCB(skb)->opt);
+       opt->optlen = iph->ihl*4 - sizeof(struct iphdr);
 
-       if (ip_options_compile(NULL, skb)) {
-               IP_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
+       if (ip_options_compile(dev_net(dev), opt, skb)) {
+               IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_INHDRERRORS);
                goto drop;
        }
 
-       opt = &(IPCB(skb)->opt);
        if (unlikely(opt->srr)) {
                struct in_device *in_dev = in_dev_get(dev);
                if (in_dev) {
                        if (!IN_DEV_SOURCE_ROUTE(in_dev)) {
                                if (IN_DEV_LOG_MARTIANS(in_dev) &&
                                    net_ratelimit())
-                                       printk(KERN_INFO "source route option "
-                                              "%u.%u.%u.%u -> %u.%u.%u.%u\n",
-                                              NIPQUAD(iph->saddr),
-                                              NIPQUAD(iph->daddr));
+                                       printk(KERN_INFO "source route option %pI4 -> %pI4\n",
+                                              &iph->saddr, &iph->daddr);
                                in_dev_put(in_dev);
                                goto drop;
                        }
@@ -326,40 +320,51 @@ drop:
        return -1;
 }
 
-static inline int ip_rcv_finish(struct sk_buff *skb)
+static int ip_rcv_finish(struct sk_buff *skb)
 {
        const struct iphdr *iph = ip_hdr(skb);
+       struct rtable *rt;
 
        /*
         *      Initialise the virtual path cache for the packet. It describes
         *      how the packet travels inside Linux networking.
         */
-       if (skb->dst == NULL) {
+       if (skb_dst(skb) == NULL) {
                int err = ip_route_input(skb, iph->daddr, iph->saddr, iph->tos,
                                         skb->dev);
                if (unlikely(err)) {
                        if (err == -EHOSTUNREACH)
-                               IP_INC_STATS_BH(IPSTATS_MIB_INADDRERRORS);
+                               IP_INC_STATS_BH(dev_net(skb->dev),
+                                               IPSTATS_MIB_INADDRERRORS);
                        else if (err == -ENETUNREACH)
-                               IP_INC_STATS_BH(IPSTATS_MIB_INNOROUTES);
+                               IP_INC_STATS_BH(dev_net(skb->dev),
+                                               IPSTATS_MIB_INNOROUTES);
                        goto drop;
                }
        }
 
 #ifdef CONFIG_NET_CLS_ROUTE
-       if (unlikely(skb->dst->tclassid)) {
-               struct ip_rt_acct *st = ip_rt_acct + 256*smp_processor_id();
-               u32 idx = skb->dst->tclassid;
+       if (unlikely(skb_dst(skb)->tclassid)) {
+               struct ip_rt_acct *st = per_cpu_ptr(ip_rt_acct, smp_processor_id());
+               u32 idx = skb_dst(skb)->tclassid;
                st[idx&0xFF].o_packets++;
-               st[idx&0xFF].o_bytes+=skb->len;
+               st[idx&0xFF].o_bytes += skb->len;
                st[(idx>>16)&0xFF].i_packets++;
-               st[(idx>>16)&0xFF].i_bytes+=skb->len;
+               st[(idx>>16)&0xFF].i_bytes += skb->len;
        }
 #endif
 
        if (iph->ihl > 5 && ip_rcv_options(skb))
                goto drop;
 
+       rt = skb_rtable(skb);
+       if (rt->rt_type == RTN_MULTICAST) {
+               IP_UPD_PO_STATS_BH(dev_net(rt->u.dst.dev), IPSTATS_MIB_INMCAST,
+                               skb->len);
+       } else if (rt->rt_type == RTN_BROADCAST)
+               IP_UPD_PO_STATS_BH(dev_net(rt->u.dst.dev), IPSTATS_MIB_INBCAST,
+                               skb->len);
+
        return dst_input(skb);
 
 drop:
@@ -381,10 +386,11 @@ int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt,
        if (skb->pkt_type == PACKET_OTHERHOST)
                goto drop;
 
-       IP_INC_STATS_BH(IPSTATS_MIB_INRECEIVES);
+
+       IP_UPD_PO_STATS_BH(dev_net(dev), IPSTATS_MIB_IN, skb->len);
 
        if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL) {
-               IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS);
+               IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_INDISCARDS);
                goto out;
        }
 
@@ -394,7 +400,7 @@ int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt,
        iph = ip_hdr(skb);
 
        /*
-        *      RFC1122: 3.1.2.2 MUST silently discard any IP frame that fails the checksum.
+        *      RFC1122: 3.2.1.2 MUST silently discard any IP frame that fails the checksum.
         *
         *      Is the datagram acceptable?
         *
@@ -416,7 +422,10 @@ int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt,
                goto inhdr_error;
 
        len = ntohs(iph->tot_len);
-       if (skb->len < len || len < (iph->ihl*4))
+       if (skb->len < len) {
+               IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_INTRUNCATEDPKTS);
+               goto drop;
+       } else if (len < (iph->ihl*4))
                goto inhdr_error;
 
        /* Our transport medium may have padded the buffer out. Now we know it
@@ -424,22 +433,23 @@ int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt,
         * Note this now means skb->len holds ntohs(iph->tot_len).
         */
        if (pskb_trim_rcsum(skb, len)) {
-               IP_INC_STATS_BH(IPSTATS_MIB_INDISCARDS);
+               IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_INDISCARDS);
                goto drop;
        }
 
        /* Remove any debris in the socket control block */
        memset(IPCB(skb), 0, sizeof(struct inet_skb_parm));
 
-       return NF_HOOK(PF_INET, NF_IP_PRE_ROUTING, skb, dev, NULL,
+       /* Must drop socket now because of tproxy. */
+       skb_orphan(skb);
+
+       return NF_HOOK(PF_INET, NF_INET_PRE_ROUTING, skb, dev, NULL,
                       ip_rcv_finish);
 
 inhdr_error:
-       IP_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
+       IP_INC_STATS_BH(dev_net(dev), IPSTATS_MIB_INHDRERRORS);
 drop:
        kfree_skb(skb);
 out:
        return NET_RX_DROP;
 }
-
-EXPORT_SYMBOL(ip_statistics);