Merge branch 'linus' into cont_syslog
[safe/jmp/linux-2.6] / net / ipv6 / icmp.c
index e3a8e27..ce79929 100644 (file)
@@ -5,8 +5,6 @@
  *     Authors:
  *     Pedro Roque             <roque@di.fc.ul.pt>
  *
- *     $Id: icmp.c,v 1.38 2002/02/08 03:57:19 davem Exp $
- *
  *     Based on net/ipv4/icmp.c
  *
  *     RFC 1885
 #include <linux/socket.h>
 #include <linux/in.h>
 #include <linux/kernel.h>
-#include <linux/sched.h>
 #include <linux/sockios.h>
 #include <linux/net.h>
 #include <linux/skbuff.h>
 #include <linux/init.h>
 #include <linux/netfilter.h>
+#include <linux/slab.h>
 
 #ifdef CONFIG_SYSCTL
 #include <linux/sysctl.h>
 #include <net/ip6_route.h>
 #include <net/addrconf.h>
 #include <net/icmp.h>
+#include <net/xfrm.h>
+#include <net/inet_common.h>
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
 
-DEFINE_SNMP_STAT(struct icmpv6_mib, icmpv6_statistics) __read_mostly;
-
 /*
  *     The ICMP socket(s). This is the most convenient way to flow control
  *     our ICMP output as well as maintain a clean interface throughout
@@ -77,42 +75,47 @@ DEFINE_SNMP_STAT(struct icmpv6_mib, icmpv6_statistics) __read_mostly;
  *
  *     On SMP we have one ICMP socket per-cpu.
  */
-static DEFINE_PER_CPU(struct socket *, __icmpv6_socket) = NULL;
-#define icmpv6_socket  __get_cpu_var(__icmpv6_socket)
+static inline struct sock *icmpv6_sk(struct net *net)
+{
+       return net->ipv6.icmp_sk[smp_processor_id()];
+}
 
-static int icmpv6_rcv(struct sk_buff **pskb);
+static int icmpv6_rcv(struct sk_buff *skb);
 
-static struct inet6_protocol icmpv6_protocol = {
+static const struct inet6_protocol icmpv6_protocol = {
        .handler        =       icmpv6_rcv,
-       .flags          =       INET6_PROTO_FINAL,
+       .flags          =       INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
 };
 
-static __inline__ int icmpv6_xmit_lock(void)
+static __inline__ struct sock *icmpv6_xmit_lock(struct net *net)
 {
+       struct sock *sk;
+
        local_bh_disable();
 
-       if (unlikely(!spin_trylock(&icmpv6_socket->sk->sk_lock.slock))) {
+       sk = icmpv6_sk(net);
+       if (unlikely(!spin_trylock(&sk->sk_lock.slock))) {
                /* This can happen if the output path (f.e. SIT or
                 * ip6ip6 tunnel) signals dst_link_failure() for an
                 * outgoing ICMP6 packet.
                 */
                local_bh_enable();
-               return 1;
+               return NULL;
        }
-       return 0;
+       return sk;
 }
 
-static __inline__ void icmpv6_xmit_unlock(void)
+static __inline__ void icmpv6_xmit_unlock(struct sock *sk)
 {
-       spin_unlock_bh(&icmpv6_socket->sk->sk_lock.slock);
+       spin_unlock_bh(&sk->sk_lock.slock);
 }
 
-/* 
+/*
  * Slightly more convenient version of icmpv6_send.
  */
-void icmpv6_param_prob(struct sk_buff *skb, int code, int pos)
+void icmpv6_param_prob(struct sk_buff *skb, u8 code, int pos)
 {
-       icmpv6_send(skb, ICMPV6_PARAMPROB, code, pos, skb->dev);
+       icmpv6_send(skb, ICMPV6_PARAMPROB, code, pos);
        kfree_skb(skb);
 }
 
@@ -129,9 +132,9 @@ void icmpv6_param_prob(struct sk_buff *skb, int code, int pos)
 
 static int is_ineligible(struct sk_buff *skb)
 {
-       int ptr = (u8*)(skb->nh.ipv6h+1) - skb->data;
+       int ptr = (u8 *)(ipv6_hdr(skb) + 1) - skb->data;
        int len = skb->len - ptr;
-       __u8 nexthdr = skb->nh.ipv6h->nexthdr;
+       __u8 nexthdr = ipv6_hdr(skb)->nexthdr;
 
        if (len < 0)
                return 1;
@@ -151,15 +154,14 @@ static int is_ineligible(struct sk_buff *skb)
        return 0;
 }
 
-static int sysctl_icmpv6_time __read_mostly = 1*HZ;
-
-/* 
- * Check the ICMP output rate limit 
+/*
+ * Check the ICMP output rate limit
  */
-static inline int icmpv6_xrlim_allow(struct sock *sk, int type,
+static inline int icmpv6_xrlim_allow(struct sock *sk, u8 type,
                                     struct flowi *fl)
 {
        struct dst_entry *dst;
+       struct net *net = sock_net(sk);
        int res = 0;
 
        /* Informational messages are not limited. */
@@ -170,19 +172,20 @@ static inline int icmpv6_xrlim_allow(struct sock *sk, int type,
        if (type == ICMPV6_PKT_TOOBIG)
                return 1;
 
-       /* 
+       /*
         * Look up the output route.
         * XXX: perhaps the expire for routing entries cloned by
         * this lookup should be more aggressive (not longer than timeout).
         */
-       dst = ip6_route_output(sk, fl);
+       dst = ip6_route_output(net, sk, fl);
        if (dst->error) {
-               IP6_INC_STATS(IPSTATS_MIB_OUTNOROUTES);
+               IP6_INC_STATS(net, ip6_dst_idev(dst),
+                             IPSTATS_MIB_OUTNOROUTES);
        } else if (dst->dev && (dst->dev->flags&IFF_LOOPBACK)) {
                res = 1;
        } else {
                struct rt6_info *rt = (struct rt6_info *)dst;
-               int tmo = sysctl_icmpv6_time;
+               int tmo = net->ipv6.sysctl.icmpv6_time;
 
                /* Give more bandwidth to wider prefixes. */
                if (rt->rt6i_dst.plen < 128)
@@ -197,7 +200,7 @@ static inline int icmpv6_xrlim_allow(struct sock *sk, int type,
 /*
  *     an inline helper for the "simple" if statement below
  *     checks if parameter problem report is caused by an
- *     unrecognized IPv6 option that has the Option Type 
+ *     unrecognized IPv6 option that has the Option Type
  *     highest-order two bits set to 10
  */
 
@@ -205,7 +208,7 @@ static __inline__ int opt_unrec(struct sk_buff *skb, __u32 offset)
 {
        u8 _optval, *op;
 
-       offset += skb->nh.raw - skb->data;
+       offset += skb_network_offset(skb);
        op = skb_header_pointer(skb, offset, sizeof(_optval), &_optval);
        if (op == NULL)
                return 1;
@@ -221,33 +224,31 @@ static int icmpv6_push_pending_frames(struct sock *sk, struct flowi *fl, struct
        if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
                goto out;
 
-       icmp6h = (struct icmp6hdr*) skb->h.raw;
+       icmp6h = icmp6_hdr(skb);
        memcpy(icmp6h, thdr, sizeof(struct icmp6hdr));
        icmp6h->icmp6_cksum = 0;
 
        if (skb_queue_len(&sk->sk_write_queue) == 1) {
-               skb->csum = csum_partial((char *)icmp6h,
+               skb->csum = csum_partial(icmp6h,
                                        sizeof(struct icmp6hdr), skb->csum);
                icmp6h->icmp6_cksum = csum_ipv6_magic(&fl->fl6_src,
                                                      &fl->fl6_dst,
                                                      len, fl->proto,
                                                      skb->csum);
        } else {
-               u32 tmp_csum = 0;
+               __wsum tmp_csum = 0;
 
                skb_queue_walk(&sk->sk_write_queue, skb) {
                        tmp_csum = csum_add(tmp_csum, skb->csum);
                }
 
-               tmp_csum = csum_partial((char *)icmp6h,
+               tmp_csum = csum_partial(icmp6h,
                                        sizeof(struct icmp6hdr), tmp_csum);
-               tmp_csum = csum_ipv6_magic(&fl->fl6_src,
-                                          &fl->fl6_dst,
-                                          len, fl->proto, tmp_csum);
-               icmp6h->icmp6_cksum = tmp_csum;
+               icmp6h->icmp6_cksum = csum_ipv6_magic(&fl->fl6_src,
+                                                     &fl->fl6_dst,
+                                                     len, fl->proto,
+                                                     tmp_csum);
        }
-       if (icmp6h->icmp6_cksum == 0)
-               icmp6h->icmp6_cksum = -1;
        ip6_push_pending_frames(sk);
 out:
        return err;
@@ -263,7 +264,7 @@ static int icmpv6_getfrag(void *from, char *to, int offset, int len, int odd, st
 {
        struct icmpv6_msg *msg = (struct icmpv6_msg *) from;
        struct sk_buff *org_skb = msg->skb;
-       __u32 csum = 0;
+       __wsum csum = 0;
 
        csum = skb_copy_and_csum_bits(org_skb, msg->offset + offset,
                                      to, len, csum);
@@ -273,39 +274,66 @@ static int icmpv6_getfrag(void *from, char *to, int offset, int len, int odd, st
        return 0;
 }
 
+#if defined(CONFIG_IPV6_MIP6) || defined(CONFIG_IPV6_MIP6_MODULE)
+static void mip6_addr_swap(struct sk_buff *skb)
+{
+       struct ipv6hdr *iph = ipv6_hdr(skb);
+       struct inet6_skb_parm *opt = IP6CB(skb);
+       struct ipv6_destopt_hao *hao;
+       struct in6_addr tmp;
+       int off;
+
+       if (opt->dsthao) {
+               off = ipv6_find_tlv(skb, opt->dsthao, IPV6_TLV_HAO);
+               if (likely(off >= 0)) {
+                       hao = (struct ipv6_destopt_hao *)
+                                       (skb_network_header(skb) + off);
+                       ipv6_addr_copy(&tmp, &iph->saddr);
+                       ipv6_addr_copy(&iph->saddr, &hao->addr);
+                       ipv6_addr_copy(&hao->addr, &tmp);
+               }
+       }
+}
+#else
+static inline void mip6_addr_swap(struct sk_buff *skb) {}
+#endif
+
 /*
  *     Send an ICMP message in response to a packet in error
  */
-void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info, 
-                struct net_device *dev)
+void icmpv6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info)
 {
+       struct net *net = dev_net(skb->dev);
        struct inet6_dev *idev = NULL;
-       struct ipv6hdr *hdr = skb->nh.ipv6h;
+       struct ipv6hdr *hdr = ipv6_hdr(skb);
        struct sock *sk;
        struct ipv6_pinfo *np;
        struct in6_addr *saddr = NULL;
        struct dst_entry *dst;
+       struct dst_entry *dst2;
        struct icmp6hdr tmp_hdr;
        struct flowi fl;
+       struct flowi fl2;
        struct icmpv6_msg msg;
        int iif = 0;
        int addr_type = 0;
        int len;
-       int hlimit, tclass;
+       int hlimit;
        int err = 0;
 
-       if ((u8*)hdr < skb->head || (u8*)(hdr+1) > skb->tail)
+       if ((u8 *)hdr < skb->head ||
+           (skb->network_header + sizeof(*hdr)) > skb->tail)
                return;
 
        /*
-        *      Make sure we respect the rules 
+        *      Make sure we respect the rules
         *      i.e. RFC 1885 2.4(e)
         *      Rule (e.1) is enforced by not using icmpv6_send
         *      in any code that processes icmp errors.
         */
        addr_type = ipv6_addr_type(&hdr->daddr);
 
-       if (ipv6_chk_addr(&hdr->daddr, skb->dev, 0))
+       if (ipv6_chk_addr(net, &hdr->daddr, skb->dev, 0))
                saddr = &hdr->daddr;
 
        /*
@@ -314,8 +342,8 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
 
        if ((addr_type & IPV6_ADDR_MULTICAST || skb->pkt_type != PACKET_HOST)) {
                if (type != ICMPV6_PKT_TOOBIG &&
-                   !(type == ICMPV6_PARAMPROB && 
-                     code == ICMPV6_UNK_OPTION && 
+                   !(type == ICMPV6_PARAMPROB &&
+                     code == ICMPV6_UNK_OPTION &&
                      (opt_unrec(skb, info))))
                        return;
 
@@ -342,7 +370,7 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
                return;
        }
 
-       /* 
+       /*
         *      Never answer to a ICMP packet.
         */
        if (is_ineligible(skb)) {
@@ -350,6 +378,8 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
                return;
        }
 
+       mip6_addr_swap(skb);
+
        memset(&fl, 0, sizeof(fl));
        fl.proto = IPPROTO_ICMPV6;
        ipv6_addr_copy(&fl.fl6_dst, &hdr->saddr);
@@ -360,10 +390,9 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
        fl.fl_icmp_code = code;
        security_skb_classify_flow(skb, &fl);
 
-       if (icmpv6_xmit_lock())
+       sk = icmpv6_xmit_lock(net);
+       if (sk == NULL)
                return;
-
-       sk = icmpv6_socket->sk;
        np = inet6_sk(sk);
 
        if (!icmpv6_xrlim_allow(sk, type, &fl))
@@ -390,24 +419,53 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
                goto out_dst_release;
        }
 
-       if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
+       /* No need to clone since we're just using its address. */
+       dst2 = dst;
+
+       err = xfrm_lookup(net, &dst, &fl, sk, 0);
+       switch (err) {
+       case 0:
+               if (dst != dst2)
+                       goto route_done;
+               break;
+       case -EPERM:
+               dst = NULL;
+               break;
+       default:
                goto out;
+       }
 
+       if (xfrm_decode_session_reverse(skb, &fl2, AF_INET6))
+               goto relookup_failed;
+
+       if (ip6_dst_lookup(sk, &dst2, &fl2))
+               goto relookup_failed;
+
+       err = xfrm_lookup(net, &dst2, &fl2, sk, XFRM_LOOKUP_ICMP);
+       switch (err) {
+       case 0:
+               dst_release(dst);
+               dst = dst2;
+               break;
+       case -EPERM:
+               goto out_dst_release;
+       default:
+relookup_failed:
+               if (!dst)
+                       goto out;
+               break;
+       }
+
+route_done:
        if (ipv6_addr_is_multicast(&fl.fl6_dst))
                hlimit = np->mcast_hops;
        else
                hlimit = np->hop_limit;
        if (hlimit < 0)
-               hlimit = dst_metric(dst, RTAX_HOPLIMIT);
-       if (hlimit < 0)
-               hlimit = ipv6_get_hoplimit(dst->dev);
-
-       tclass = np->tclass;
-       if (tclass < 0)
-               tclass = 0;
+               hlimit = ip6_dst_hoplimit(dst);
 
        msg.skb = skb;
-       msg.offset = skb->nh.raw - skb->data;
+       msg.offset = skb_network_offset(skb);
        msg.type = type;
 
        len = skb->len - msg.offset;
@@ -421,44 +479,43 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
 
        err = ip6_append_data(sk, icmpv6_getfrag, &msg,
                              len + sizeof(struct icmp6hdr),
-                             sizeof(struct icmp6hdr),
-                             hlimit, tclass, NULL, &fl, (struct rt6_info*)dst,
-                             MSG_DONTWAIT);
+                             sizeof(struct icmp6hdr), hlimit,
+                             np->tclass, NULL, &fl, (struct rt6_info*)dst,
+                             MSG_DONTWAIT, np->dontfrag);
        if (err) {
+               ICMP6_INC_STATS_BH(net, idev, ICMP6_MIB_OUTMSGS);
                ip6_flush_pending_frames(sk);
                goto out_put;
        }
        err = icmpv6_push_pending_frames(sk, &fl, &tmp_hdr, len + sizeof(struct icmp6hdr));
 
-       if (type >= ICMPV6_DEST_UNREACH && type <= ICMPV6_PARAMPROB)
-               ICMP6_INC_STATS_OFFSET_BH(idev, ICMP6_MIB_OUTDESTUNREACHS, type - ICMPV6_DEST_UNREACH);
-       ICMP6_INC_STATS_BH(idev, ICMP6_MIB_OUTMSGS);
-
 out_put:
        if (likely(idev != NULL))
                in6_dev_put(idev);
 out_dst_release:
        dst_release(dst);
 out:
-       icmpv6_xmit_unlock();
+       icmpv6_xmit_unlock(sk);
 }
 
+EXPORT_SYMBOL(icmpv6_send);
+
 static void icmpv6_echo_reply(struct sk_buff *skb)
 {
+       struct net *net = dev_net(skb->dev);
        struct sock *sk;
        struct inet6_dev *idev;
        struct ipv6_pinfo *np;
        struct in6_addr *saddr = NULL;
-       struct icmp6hdr *icmph = (struct icmp6hdr *) skb->h.raw;
+       struct icmp6hdr *icmph = icmp6_hdr(skb);
        struct icmp6hdr tmp_hdr;
        struct flowi fl;
        struct icmpv6_msg msg;
        struct dst_entry *dst;
        int err = 0;
        int hlimit;
-       int tclass;
 
-       saddr = &skb->nh.ipv6h->daddr;
+       saddr = &ipv6_hdr(skb)->daddr;
 
        if (!ipv6_unicast_destination(skb))
                saddr = NULL;
@@ -468,17 +525,16 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
 
        memset(&fl, 0, sizeof(fl));
        fl.proto = IPPROTO_ICMPV6;
-       ipv6_addr_copy(&fl.fl6_dst, &skb->nh.ipv6h->saddr);
+       ipv6_addr_copy(&fl.fl6_dst, &ipv6_hdr(skb)->saddr);
        if (saddr)
                ipv6_addr_copy(&fl.fl6_src, saddr);
        fl.oif = skb->dev->ifindex;
        fl.fl_icmp_type = ICMPV6_ECHO_REPLY;
        security_skb_classify_flow(skb, &fl);
 
-       if (icmpv6_xmit_lock())
+       sk = icmpv6_xmit_lock(net);
+       if (sk == NULL)
                return;
-
-       sk = icmpv6_socket->sk;
        np = inet6_sk(sk);
 
        if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst))
@@ -487,7 +543,7 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
        err = ip6_dst_lookup(sk, &dst, &fl);
        if (err)
                goto out;
-       if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
+       if ((err = xfrm_lookup(net, &dst, &fl, sk, 0)) < 0)
                goto out;
 
        if (ipv6_addr_is_multicast(&fl.fl6_dst))
@@ -495,13 +551,7 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
        else
                hlimit = np->hop_limit;
        if (hlimit < 0)
-               hlimit = dst_metric(dst, RTAX_HOPLIMIT);
-       if (hlimit < 0)
-               hlimit = ipv6_get_hoplimit(dst->dev);
-
-       tclass = np->tclass;
-       if (tclass < 0)
-               tclass = 0;
+               hlimit = ip6_dst_hoplimit(dst);
 
        idev = in6_dev_get(skb->dev);
 
@@ -510,31 +560,28 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
        msg.type = ICMPV6_ECHO_REPLY;
 
        err = ip6_append_data(sk, icmpv6_getfrag, &msg, skb->len + sizeof(struct icmp6hdr),
-                               sizeof(struct icmp6hdr), hlimit, tclass, NULL, &fl,
-                               (struct rt6_info*)dst, MSG_DONTWAIT);
+                               sizeof(struct icmp6hdr), hlimit, np->tclass, NULL, &fl,
+                               (struct rt6_info*)dst, MSG_DONTWAIT,
+                               np->dontfrag);
 
        if (err) {
+               ICMP6_INC_STATS_BH(net, idev, ICMP6_MIB_OUTMSGS);
                ip6_flush_pending_frames(sk);
                goto out_put;
        }
        err = icmpv6_push_pending_frames(sk, &fl, &tmp_hdr, skb->len + sizeof(struct icmp6hdr));
 
-        ICMP6_INC_STATS_BH(idev, ICMP6_MIB_OUTECHOREPLIES);
-        ICMP6_INC_STATS_BH(idev, ICMP6_MIB_OUTMSGS);
-
-out_put: 
+out_put:
        if (likely(idev != NULL))
                in6_dev_put(idev);
        dst_release(dst);
-out: 
-       icmpv6_xmit_unlock();
+out:
+       icmpv6_xmit_unlock(sk);
 }
 
-static void icmpv6_notify(struct sk_buff *skb, int type, int code, u32 info)
+static void icmpv6_notify(struct sk_buff *skb, u8 type, u8 code, __be32 info)
 {
-       struct in6_addr *saddr, *daddr;
-       struct inet6_protocol *ipprot;
-       struct sock *sk;
+       const struct inet6_protocol *ipprot;
        int inner_offset;
        int hash;
        u8 nexthdr;
@@ -556,9 +603,6 @@ static void icmpv6_notify(struct sk_buff *skb, int type, int code, u32 info)
        if (!pskb_may_pull(skb, inner_offset+8))
                return;
 
-       saddr = &skb->nh.ipv6h->saddr;
-       daddr = &skb->nh.ipv6h->daddr;
-
        /* BUGGG_FUTURE: we should try to parse exthdrs in this packet.
           Without this we will not able f.e. to make source routed
           pmtu discovery.
@@ -574,35 +618,46 @@ static void icmpv6_notify(struct sk_buff *skb, int type, int code, u32 info)
                ipprot->err_handler(skb, NULL, type, code, inner_offset, info);
        rcu_read_unlock();
 
-       read_lock(&raw_v6_lock);
-       if ((sk = sk_head(&raw_v6_htable[hash])) != NULL) {
-               while((sk = __raw_v6_lookup(sk, nexthdr, daddr, saddr,
-                                           IP6CB(skb)->iif))) {
-                       rawv6_err(sk, skb, NULL, type, code, inner_offset, info);
-                       sk = sk_next(sk);
-               }
-       }
-       read_unlock(&raw_v6_lock);
+       raw6_icmp_error(skb, nexthdr, type, code, inner_offset, info);
 }
-  
+
 /*
  *     Handle icmp messages
  */
 
-static int icmpv6_rcv(struct sk_buff **pskb)
+static int icmpv6_rcv(struct sk_buff *skb)
 {
-       struct sk_buff *skb = *pskb;
        struct net_device *dev = skb->dev;
        struct inet6_dev *idev = __in6_dev_get(dev);
        struct in6_addr *saddr, *daddr;
        struct ipv6hdr *orig_hdr;
        struct icmp6hdr *hdr;
-       int type;
+       u8 type;
+
+       if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) {
+               struct sec_path *sp = skb_sec_path(skb);
+               int nh;
+
+               if (!(sp && sp->xvec[sp->len - 1]->props.flags &
+                                XFRM_STATE_ICMP))
+                       goto drop_no_count;
 
-       ICMP6_INC_STATS_BH(idev, ICMP6_MIB_INMSGS);
+               if (!pskb_may_pull(skb, sizeof(*hdr) + sizeof(*orig_hdr)))
+                       goto drop_no_count;
 
-       saddr = &skb->nh.ipv6h->saddr;
-       daddr = &skb->nh.ipv6h->daddr;
+               nh = skb_network_offset(skb);
+               skb_set_network_header(skb, sizeof(*hdr));
+
+               if (!xfrm6_policy_check_reverse(NULL, XFRM_POLICY_IN, skb))
+                       goto drop_no_count;
+
+               skb_set_network_header(skb, nh);
+       }
+
+       ICMP6_INC_STATS_BH(dev_net(dev), idev, ICMP6_MIB_INMSGS);
+
+       saddr = &ipv6_hdr(skb)->saddr;
+       daddr = &ipv6_hdr(skb)->daddr;
 
        /* Perform checksum. */
        switch (skb->ip_summed) {
@@ -612,26 +667,23 @@ static int icmpv6_rcv(struct sk_buff **pskb)
                        break;
                /* fall through */
        case CHECKSUM_NONE:
-               skb->csum = ~csum_ipv6_magic(saddr, daddr, skb->len,
-                                            IPPROTO_ICMPV6, 0);
+               skb->csum = ~csum_unfold(csum_ipv6_magic(saddr, daddr, skb->len,
+                                            IPPROTO_ICMPV6, 0));
                if (__skb_checksum_complete(skb)) {
-                       LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 checksum failed [" NIP6_FMT " > " NIP6_FMT "]\n",
-                                      NIP6(*saddr), NIP6(*daddr));
+                       LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 checksum failed [%pI6 > %pI6]\n",
+                                      saddr, daddr);
                        goto discard_it;
                }
        }
 
-       if (!pskb_pull(skb, sizeof(struct icmp6hdr)))
+       if (!pskb_pull(skb, sizeof(*hdr)))
                goto discard_it;
 
-       hdr = (struct icmp6hdr *) skb->h.raw;
+       hdr = icmp6_hdr(skb);
 
        type = hdr->icmp6_type;
 
-       if (type >= ICMPV6_DEST_UNREACH && type <= ICMPV6_PARAMPROB)
-               ICMP6_INC_STATS_OFFSET_BH(idev, ICMP6_MIB_INDESTUNREACHS, type - ICMPV6_DEST_UNREACH);
-       else if (type >= ICMPV6_ECHO_REQUEST && type <= NDISC_REDIRECT)
-               ICMP6_INC_STATS_OFFSET_BH(idev, ICMP6_MIB_INECHOS, type - ICMPV6_ECHO_REQUEST);
+       ICMP6MSGIN_INC_STATS_BH(dev_net(dev), idev, type);
 
        switch (type) {
        case ICMPV6_ECHO_REQUEST:
@@ -650,7 +702,7 @@ static int icmpv6_rcv(struct sk_buff **pskb)
                 */
                if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
                        goto discard_it;
-               hdr = (struct icmp6hdr *) skb->h.raw;
+               hdr = icmp6_hdr(skb);
                orig_hdr = (struct ipv6hdr *) (hdr + 1);
                rt6_pmtu_discovery(&orig_hdr->daddr, &orig_hdr->saddr, dev,
                                   ntohl(hdr->icmp6_mtu));
@@ -698,35 +750,58 @@ static int icmpv6_rcv(struct sk_buff **pskb)
                if (type & ICMPV6_INFOMSG_MASK)
                        break;
 
-               /* 
-                * error of unknown type. 
-                * must pass to upper level 
+               /*
+                * error of unknown type.
+                * must pass to upper level
                 */
 
                icmpv6_notify(skb, type, hdr->icmp6_code, hdr->icmp6_mtu);
-       };
+       }
+
        kfree_skb(skb);
        return 0;
 
 discard_it:
-       ICMP6_INC_STATS_BH(idev, ICMP6_MIB_INERRORS);
+       ICMP6_INC_STATS_BH(dev_net(dev), idev, ICMP6_MIB_INERRORS);
+drop_no_count:
        kfree_skb(skb);
        return 0;
 }
 
+void icmpv6_flow_init(struct sock *sk, struct flowi *fl,
+                     u8 type,
+                     const struct in6_addr *saddr,
+                     const struct in6_addr *daddr,
+                     int oif)
+{
+       memset(fl, 0, sizeof(*fl));
+       ipv6_addr_copy(&fl->fl6_src, saddr);
+       ipv6_addr_copy(&fl->fl6_dst, daddr);
+       fl->proto               = IPPROTO_ICMPV6;
+       fl->fl_icmp_type        = type;
+       fl->fl_icmp_code        = 0;
+       fl->oif                 = oif;
+       security_sk_classify_flow(sk, fl);
+}
+
 /*
- * Special lock-class for __icmpv6_socket:
+ * Special lock-class for __icmpv6_sk:
  */
 static struct lock_class_key icmpv6_socket_sk_dst_lock_key;
 
-int __init icmpv6_init(struct net_proto_family *ops)
+static int __net_init icmpv6_sk_init(struct net *net)
 {
        struct sock *sk;
        int err, i, j;
 
+       net->ipv6.icmp_sk =
+               kzalloc(nr_cpu_ids * sizeof(struct sock *), GFP_KERNEL);
+       if (net->ipv6.icmp_sk == NULL)
+               return -ENOMEM;
+
        for_each_possible_cpu(i) {
-               err = sock_create_kern(PF_INET6, SOCK_RAW, IPPROTO_ICMPV6,
-                                      &per_cpu(__icmpv6_socket, i));
+               err = inet_ctl_sock_create(&sk, PF_INET6,
+                                          SOCK_RAW, IPPROTO_ICMPV6, net);
                if (err < 0) {
                        printk(KERN_ERR
                               "Failed to initialize the ICMP6 control socket "
@@ -735,12 +810,12 @@ int __init icmpv6_init(struct net_proto_family *ops)
                        goto fail;
                }
 
-               sk = per_cpu(__icmpv6_socket, i)->sk;
-               sk->sk_allocation = GFP_ATOMIC;
+               net->ipv6.icmp_sk[i] = sk;
+
                /*
                 * Split off their lock-class, because sk->sk_dst_lock
                 * gets used from softirqs, which is safe for
-                * __icmpv6_socket (because those never get directly used
+                * __icmpv6_sk (because those never get directly used
                 * via userspace syscalls), but unsafe for normal sockets.
                 */
                lockdep_set_class(&sk->sk_dst_lock,
@@ -751,39 +826,57 @@ int __init icmpv6_init(struct net_proto_family *ops)
                 */
                sk->sk_sndbuf =
                        (2 * ((64 * 1024) + sizeof(struct sk_buff)));
-
-               sk->sk_prot->unhash(sk);
        }
-
-
-       if (inet6_add_protocol(&icmpv6_protocol, IPPROTO_ICMPV6) < 0) {
-               printk(KERN_ERR "Failed to register ICMP6 protocol\n");
-               err = -EAGAIN;
-               goto fail;
-       }
-
        return 0;
 
  fail:
-       for (j = 0; j < i; j++) {
-               if (!cpu_possible(j))
-                       continue;
-               sock_release(per_cpu(__icmpv6_socket, j));
-       }
-
+       for (j = 0; j < i; j++)
+               inet_ctl_sock_destroy(net->ipv6.icmp_sk[j]);
+       kfree(net->ipv6.icmp_sk);
        return err;
 }
 
-void icmpv6_cleanup(void)
+static void __net_exit icmpv6_sk_exit(struct net *net)
 {
        int i;
 
        for_each_possible_cpu(i) {
-               sock_release(per_cpu(__icmpv6_socket, i));
+               inet_ctl_sock_destroy(net->ipv6.icmp_sk[i]);
        }
+       kfree(net->ipv6.icmp_sk);
+}
+
+static struct pernet_operations icmpv6_sk_ops = {
+       .init = icmpv6_sk_init,
+       .exit = icmpv6_sk_exit,
+};
+
+int __init icmpv6_init(void)
+{
+       int err;
+
+       err = register_pernet_subsys(&icmpv6_sk_ops);
+       if (err < 0)
+               return err;
+
+       err = -EAGAIN;
+       if (inet6_add_protocol(&icmpv6_protocol, IPPROTO_ICMPV6) < 0)
+               goto fail;
+       return 0;
+
+fail:
+       printk(KERN_ERR "Failed to register ICMP6 protocol\n");
+       unregister_pernet_subsys(&icmpv6_sk_ops);
+       return err;
+}
+
+void icmpv6_cleanup(void)
+{
+       unregister_pernet_subsys(&icmpv6_sk_ops);
        inet6_del_protocol(&icmpv6_protocol, IPPROTO_ICMPV6);
 }
 
+
 static const struct icmp6_err {
        int err;
        int fatal;
@@ -810,7 +903,7 @@ static const struct icmp6_err {
        },
 };
 
-int icmpv6_err_convert(int type, int code, int *err)
+int icmpv6_err_convert(u8 type, u8 code, int *err)
 {
        int fatal = 0;
 
@@ -828,7 +921,7 @@ int icmpv6_err_convert(int type, int code, int *err)
        case ICMPV6_PKT_TOOBIG:
                *err = EMSGSIZE;
                break;
-               
+
        case ICMPV6_PARAMPROB:
                *err = EPROTO;
                fatal = 1;
@@ -837,22 +930,37 @@ int icmpv6_err_convert(int type, int code, int *err)
        case ICMPV6_TIME_EXCEED:
                *err = EHOSTUNREACH;
                break;
-       };
+       }
 
        return fatal;
 }
 
+EXPORT_SYMBOL(icmpv6_err_convert);
+
 #ifdef CONFIG_SYSCTL
-ctl_table ipv6_icmp_table[] = {
+ctl_table ipv6_icmp_table_template[] = {
        {
-               .ctl_name       = NET_IPV6_ICMP_RATELIMIT,
                .procname       = "ratelimit",
-               .data           = &sysctl_icmpv6_time,
+               .data           = &init_net.ipv6.sysctl.icmpv6_time,
                .maxlen         = sizeof(int),
                .mode           = 0644,
-               .proc_handler   = &proc_dointvec
+               .proc_handler   = proc_dointvec_ms_jiffies,
        },
-       { .ctl_name = 0 },
+       { },
 };
+
+struct ctl_table * __net_init ipv6_icmp_sysctl_init(struct net *net)
+{
+       struct ctl_table *table;
+
+       table = kmemdup(ipv6_icmp_table_template,
+                       sizeof(ipv6_icmp_table_template),
+                       GFP_KERNEL);
+
+       if (table)
+               table[0].data = &net->ipv6.sysctl.icmpv6_time;
+
+       return table;
+}
 #endif