Merge branch 'master' of /repos/git/net-next-2.6
[safe/jmp/linux-2.6] / net / ipv6 / netfilter / ip6_tables.c
index f63fb86..4332f45 100644 (file)
@@ -7,17 +7,8 @@
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
- *
- * 19 Jan 2002 Harald Welte <laforge@gnumonks.org>
- *     - increase module usage count as soon as we have rules inside
- *       a table
- * 06 Jun 2002 Andras Kis-Szabo <kisza@sch.bme.hu>
- *      - new extension header parser code
- * 15 Oct 2005 Harald Welte <laforge@netfilter.org>
- *     - Unification of {ip,ip6}_tables into x_tables
- *     - Removed tcp and udp code, since it's not ipv6 specific
  */
-
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 #include <linux/capability.h>
 #include <linux/in.h>
 #include <linux/skbuff.h>
 #include <linux/poison.h>
 #include <linux/icmpv6.h>
 #include <net/ipv6.h>
+#include <net/compat.h>
 #include <asm/uaccess.h>
 #include <linux/mutex.h>
 #include <linux/proc_fs.h>
+#include <linux/err.h>
 #include <linux/cpumask.h>
 
 #include <linux/netfilter_ipv6/ip6_tables.h>
 #include <linux/netfilter/x_tables.h>
+#include <net/netfilter/nf_log.h>
 
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Netfilter Core Team <coreteam@netfilter.org>");
 MODULE_DESCRIPTION("IPv6 packet filter");
 
-#define IPV6_HDR_LEN   (sizeof(struct ipv6hdr))
-#define IPV6_OPTHDR_LEN        (sizeof(struct ipv6_opt_hdr))
-
 /*#define DEBUG_IP_FIREWALL*/
 /*#define DEBUG_ALLOW_ALL*/ /* Useful for remote debugging */
 /*#define DEBUG_IP_FIREWALL_USER*/
@@ -64,7 +55,7 @@ MODULE_DESCRIPTION("IPv6 packet filter");
 do {                                                           \
        if (!(x))                                               \
                printk("IP_NF_ASSERT: %s:%s:%u\n",              \
-                      __FUNCTION__, __FILE__, __LINE__);       \
+                      __func__, __FILE__, __LINE__);   \
 } while(0)
 #else
 #define IP_NF_ASSERT(x)
@@ -85,44 +76,38 @@ do {                                                                \
 
    Hence the start of any table is given by get_table() below.  */
 
-#if 0
-#define down(x) do { printk("DOWN:%u:" #x "\n", __LINE__); down(x); } while(0)
-#define down_interruptible(x) ({ int __r; printk("DOWNi:%u:" #x "\n", __LINE__); __r = down_interruptible(x); if (__r != 0) printk("ABORT-DOWNi:%u\n", __LINE__); __r; })
-#define up(x) do { printk("UP:%u:" #x "\n", __LINE__); up(x); } while(0)
-#endif
-
 /* Check for an extension */
-int 
+int
 ip6t_ext_hdr(u8 nexthdr)
 {
-        return ( (nexthdr == IPPROTO_HOPOPTS)   ||
-                 (nexthdr == IPPROTO_ROUTING)   ||
-                 (nexthdr == IPPROTO_FRAGMENT)  ||
-                 (nexthdr == IPPROTO_ESP)       ||
-                 (nexthdr == IPPROTO_AH)        ||
-                 (nexthdr == IPPROTO_NONE)      ||
-                 (nexthdr == IPPROTO_DSTOPTS) );
+       return ( (nexthdr == IPPROTO_HOPOPTS)   ||
+                (nexthdr == IPPROTO_ROUTING)   ||
+                (nexthdr == IPPROTO_FRAGMENT)  ||
+                (nexthdr == IPPROTO_ESP)       ||
+                (nexthdr == IPPROTO_AH)        ||
+                (nexthdr == IPPROTO_NONE)      ||
+                (nexthdr == IPPROTO_DSTOPTS) );
 }
 
 /* Returns whether matches rule or not. */
-static inline int
+/* Performance critical - called for every packet */
+static inline bool
 ip6_packet_match(const struct sk_buff *skb,
                 const char *indev,
                 const char *outdev,
                 const struct ip6t_ip6 *ip6info,
                 unsigned int *protoff,
-                int *fragoff, int *hotdrop)
+                int *fragoff, bool *hotdrop)
 {
-       size_t i;
        unsigned long ret;
-       const struct ipv6hdr *ipv6 = skb->nh.ipv6h;
+       const struct ipv6hdr *ipv6 = ipv6_hdr(skb);
 
-#define FWINV(bool,invflg) ((bool) ^ !!(ip6info->invflags & invflg))
+#define FWINV(bool, invflg) ((bool) ^ !!(ip6info->invflags & (invflg)))
 
        if (FWINV(ipv6_masked_addr_cmp(&ipv6->saddr, &ip6info->smsk,
-                                      &ip6info->src), IP6T_INV_SRCIP)
-           || FWINV(ipv6_masked_addr_cmp(&ipv6->daddr, &ip6info->dmsk,
-                                         &ip6info->dst), IP6T_INV_DSTIP)) {
+                                      &ip6info->src), IP6T_INV_SRCIP) ||
+           FWINV(ipv6_masked_addr_cmp(&ipv6->daddr, &ip6info->dmsk,
+                                      &ip6info->dst), IP6T_INV_DSTIP)) {
                dprintf("Source or dest mismatch.\n");
 /*
                dprintf("SRC: %u. Mask: %u. Target: %u.%s\n", ip->saddr,
@@ -131,34 +116,25 @@ ip6_packet_match(const struct sk_buff *skb,
                dprintf("DST: %u. Mask: %u. Target: %u.%s\n", ip->daddr,
                        ipinfo->dmsk.s_addr, ipinfo->dst.s_addr,
                        ipinfo->invflags & IP6T_INV_DSTIP ? " (INV)" : "");*/
-               return 0;
+               return false;
        }
 
-       /* Look for ifname matches; this should unroll nicely. */
-       for (i = 0, ret = 0; i < IFNAMSIZ/sizeof(unsigned long); i++) {
-               ret |= (((const unsigned long *)indev)[i]
-                       ^ ((const unsigned long *)ip6info->iniface)[i])
-                       & ((const unsigned long *)ip6info->iniface_mask)[i];
-       }
+       ret = ifname_compare_aligned(indev, ip6info->iniface, ip6info->iniface_mask);
 
        if (FWINV(ret != 0, IP6T_INV_VIA_IN)) {
                dprintf("VIA in mismatch (%s vs %s).%s\n",
                        indev, ip6info->iniface,
                        ip6info->invflags&IP6T_INV_VIA_IN ?" (INV)":"");
-               return 0;
+               return false;
        }
 
-       for (i = 0, ret = 0; i < IFNAMSIZ/sizeof(unsigned long); i++) {
-               ret |= (((const unsigned long *)outdev)[i]
-                       ^ ((const unsigned long *)ip6info->outiface)[i])
-                       & ((const unsigned long *)ip6info->outiface_mask)[i];
-       }
+       ret = ifname_compare_aligned(outdev, ip6info->outiface, ip6info->outiface_mask);
 
        if (FWINV(ret != 0, IP6T_INV_VIA_OUT)) {
                dprintf("VIA out mismatch (%s vs %s).%s\n",
                        outdev, ip6info->outiface,
                        ip6info->invflags&IP6T_INV_VIA_OUT ?" (INV)":"");
-               return 0;
+               return false;
        }
 
 /* ... might want to do something with class and flowlabel here ... */
@@ -171,77 +147,71 @@ ip6_packet_match(const struct sk_buff *skb,
                protohdr = ipv6_find_hdr(skb, protoff, -1, &_frag_off);
                if (protohdr < 0) {
                        if (_frag_off == 0)
-                               *hotdrop = 1;
-                       return 0;
+                               *hotdrop = true;
+                       return false;
                }
                *fragoff = _frag_off;
 
                dprintf("Packet protocol %hi ?= %s%hi.\n",
-                               protohdr, 
+                               protohdr,
                                ip6info->invflags & IP6T_INV_PROTO ? "!":"",
                                ip6info->proto);
 
                if (ip6info->proto == protohdr) {
                        if(ip6info->invflags & IP6T_INV_PROTO) {
-                               return 0;
+                               return false;
                        }
-                       return 1;
+                       return true;
                }
 
                /* We need match for the '-p all', too! */
                if ((ip6info->proto != 0) &&
                        !(ip6info->invflags & IP6T_INV_PROTO))
-                       return 0;
+                       return false;
        }
-       return 1;
+       return true;
 }
 
 /* should be ip6 safe */
-static inline int 
+static bool
 ip6_checkentry(const struct ip6t_ip6 *ipv6)
 {
        if (ipv6->flags & ~IP6T_F_MASK) {
                duprintf("Unknown flag bits set: %08X\n",
                         ipv6->flags & ~IP6T_F_MASK);
-               return 0;
+               return false;
        }
        if (ipv6->invflags & ~IP6T_INV_MASK) {
                duprintf("Unknown invflag bits set: %08X\n",
                         ipv6->invflags & ~IP6T_INV_MASK);
-               return 0;
+               return false;
        }
-       return 1;
+       return true;
 }
 
 static unsigned int
-ip6t_error(struct sk_buff **pskb,
-         const struct net_device *in,
-         const struct net_device *out,
-         unsigned int hooknum,
-         const struct xt_target *target,
-         const void *targinfo)
+ip6t_error(struct sk_buff *skb, const struct xt_target_param *par)
 {
        if (net_ratelimit())
-               printk("ip6_tables: error: `%s'\n", (char *)targinfo);
+               printk("ip6_tables: error: `%s'\n",
+                      (const char *)par->targinfo);
 
        return NF_DROP;
 }
 
-static inline
-int do_match(struct ip6t_entry_match *m,
-            const struct sk_buff *skb,
-            const struct net_device *in,
-            const struct net_device *out,
-            int offset,
-            unsigned int protoff,
-            int *hotdrop)
+/* Performance critical - called for every packet */
+static inline bool
+do_match(struct ip6t_entry_match *m, const struct sk_buff *skb,
+        struct xt_match_param *par)
 {
+       par->match     = m->u.kernel.match;
+       par->matchinfo = m->data;
+
        /* Stop iteration if it doesn't match */
-       if (!m->u.kernel.match->match(skb, in, out, m->u.kernel.match, m->data,
-                                     offset, protoff, hotdrop))
-               return 1;
+       if (!m->u.kernel.match->match(skb, par))
+               return true;
        else
-               return 0;
+               return false;
 }
 
 static inline struct ip6t_entry *
@@ -250,24 +220,136 @@ get_entry(void *base, unsigned int offset)
        return (struct ip6t_entry *)(base + offset);
 }
 
+/* All zeroes == unconditional rule. */
+/* Mildly perf critical (only if packet tracing is on) */
+static inline bool unconditional(const struct ip6t_ip6 *ipv6)
+{
+       static const struct ip6t_ip6 uncond;
+
+       return memcmp(ipv6, &uncond, sizeof(uncond)) == 0;
+}
+
+#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
+    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
+/* This cries for unification! */
+static const char *const hooknames[] = {
+       [NF_INET_PRE_ROUTING]           = "PREROUTING",
+       [NF_INET_LOCAL_IN]              = "INPUT",
+       [NF_INET_FORWARD]               = "FORWARD",
+       [NF_INET_LOCAL_OUT]             = "OUTPUT",
+       [NF_INET_POST_ROUTING]          = "POSTROUTING",
+};
+
+enum nf_ip_trace_comments {
+       NF_IP6_TRACE_COMMENT_RULE,
+       NF_IP6_TRACE_COMMENT_RETURN,
+       NF_IP6_TRACE_COMMENT_POLICY,
+};
+
+static const char *const comments[] = {
+       [NF_IP6_TRACE_COMMENT_RULE]     = "rule",
+       [NF_IP6_TRACE_COMMENT_RETURN]   = "return",
+       [NF_IP6_TRACE_COMMENT_POLICY]   = "policy",
+};
+
+static struct nf_loginfo trace_loginfo = {
+       .type = NF_LOG_TYPE_LOG,
+       .u = {
+               .log = {
+                       .level = 4,
+                       .logflags = NF_LOG_MASK,
+               },
+       },
+};
+
+/* Mildly perf critical (only if packet tracing is on) */
+static inline int
+get_chainname_rulenum(struct ip6t_entry *s, struct ip6t_entry *e,
+                     const char *hookname, const char **chainname,
+                     const char **comment, unsigned int *rulenum)
+{
+       struct ip6t_standard_target *t = (void *)ip6t_get_target(s);
+
+       if (strcmp(t->target.u.kernel.target->name, IP6T_ERROR_TARGET) == 0) {
+               /* Head of user chain: ERROR target with chainname */
+               *chainname = t->target.data;
+               (*rulenum) = 0;
+       } else if (s == e) {
+               (*rulenum)++;
+
+               if (s->target_offset == sizeof(struct ip6t_entry) &&
+                   strcmp(t->target.u.kernel.target->name,
+                          IP6T_STANDARD_TARGET) == 0 &&
+                   t->verdict < 0 &&
+                   unconditional(&s->ipv6)) {
+                       /* Tail of chains: STANDARD target (return/policy) */
+                       *comment = *chainname == hookname
+                               ? comments[NF_IP6_TRACE_COMMENT_POLICY]
+                               : comments[NF_IP6_TRACE_COMMENT_RETURN];
+               }
+               return 1;
+       } else
+               (*rulenum)++;
+
+       return 0;
+}
+
+static void trace_packet(struct sk_buff *skb,
+                        unsigned int hook,
+                        const struct net_device *in,
+                        const struct net_device *out,
+                        const char *tablename,
+                        struct xt_table_info *private,
+                        struct ip6t_entry *e)
+{
+       void *table_base;
+       const struct ip6t_entry *root;
+       const char *hookname, *chainname, *comment;
+       unsigned int rulenum = 0;
+
+       table_base = private->entries[smp_processor_id()];
+       root = get_entry(table_base, private->hook_entry[hook]);
+
+       hookname = chainname = hooknames[hook];
+       comment = comments[NF_IP6_TRACE_COMMENT_RULE];
+
+       IP6T_ENTRY_ITERATE(root,
+                          private->size - private->hook_entry[hook],
+                          get_chainname_rulenum,
+                          e, hookname, &chainname, &comment, &rulenum);
+
+       nf_log_packet(AF_INET6, hook, skb, in, out, &trace_loginfo,
+                     "TRACE: %s:%s:%s:%u ",
+                     tablename, chainname, comment, rulenum);
+}
+#endif
+
+static inline __pure struct ip6t_entry *
+ip6t_next_entry(const struct ip6t_entry *entry)
+{
+       return (void *)entry + entry->next_offset;
+}
+
 /* Returns one of the generic firewall policies, like NF_ACCEPT. */
 unsigned int
-ip6t_do_table(struct sk_buff **pskb,
+ip6t_do_table(struct sk_buff *skb,
              unsigned int hook,
              const struct net_device *in,
              const struct net_device *out,
              struct xt_table *table)
 {
+#define tb_comefrom ((struct ip6t_entry *)table_base)->comefrom
+
        static const char nulldevname[IFNAMSIZ] __attribute__((aligned(sizeof(long))));
-       int offset = 0;
-       unsigned int protoff = 0;
-       int hotdrop = 0;
+       bool hotdrop = false;
        /* Initializing verdict to NF_DROP keeps gcc happy. */
        unsigned int verdict = NF_DROP;
        const char *indev, *outdev;
        void *table_base;
        struct ip6t_entry *e, *back;
        struct xt_table_info *private;
+       struct xt_match_param mtpar;
+       struct xt_target_param tgpar;
 
        /* Initialization */
        indev = in ? in->name : nulldevname;
@@ -278,104 +360,106 @@ ip6t_do_table(struct sk_buff **pskb,
         * things we don't know, ie. tcp syn flag or ports).  If the
         * rule is also a fragment-specific rule, non-fragments won't
         * match it. */
+       mtpar.hotdrop = &hotdrop;
+       mtpar.in      = tgpar.in  = in;
+       mtpar.out     = tgpar.out = out;
+       mtpar.family  = tgpar.family = NFPROTO_IPV6;
+       mtpar.hooknum = tgpar.hooknum = hook;
 
-       read_lock_bh(&table->lock);
-       private = table->private;
        IP_NF_ASSERT(table->valid_hooks & (1 << hook));
-       table_base = (void *)private->entries[smp_processor_id()];
+
+       xt_info_rdlock_bh();
+       private = table->private;
+       table_base = private->entries[smp_processor_id()];
+
        e = get_entry(table_base, private->hook_entry[hook]);
 
        /* For return from builtin chain */
        back = get_entry(table_base, private->underflow[hook]);
 
        do {
+               struct ip6t_entry_target *t;
+
                IP_NF_ASSERT(e);
                IP_NF_ASSERT(back);
-               if (ip6_packet_match(*pskb, indev, outdev, &e->ipv6,
-                       &protoff, &offset, &hotdrop)) {
-                       struct ip6t_entry_target *t;
-
-                       if (IP6T_MATCH_ITERATE(e, do_match,
-                                              *pskb, in, out,
-                                              offset, protoff, &hotdrop) != 0)
-                               goto no_match;
-
-                       ADD_COUNTER(e->counters,
-                                   ntohs((*pskb)->nh.ipv6h->payload_len)
-                                   + IPV6_HDR_LEN,
-                                   1);
-
-                       t = ip6t_get_target(e);
-                       IP_NF_ASSERT(t->u.kernel.target);
-                       /* Standard target? */
-                       if (!t->u.kernel.target->target) {
-                               int v;
-
-                               v = ((struct ip6t_standard_target *)t)->verdict;
-                               if (v < 0) {
-                                       /* Pop from stack? */
-                                       if (v != IP6T_RETURN) {
-                                               verdict = (unsigned)(-v) - 1;
-                                               break;
-                                       }
-                                       e = back;
-                                       back = get_entry(table_base,
-                                                        back->comefrom);
-                                       continue;
-                               }
-                               if (table_base + v != (void *)e + e->next_offset
-                                   && !(e->ipv6.flags & IP6T_F_GOTO)) {
-                                       /* Save old back ptr in next entry */
-                                       struct ip6t_entry *next
-                                               = (void *)e + e->next_offset;
-                                       next->comefrom
-                                               = (void *)back - table_base;
-                                       /* set back pointer to next entry */
-                                       back = next;
-                               }
+               if (!ip6_packet_match(skb, indev, outdev, &e->ipv6,
+                   &mtpar.thoff, &mtpar.fragoff, &hotdrop) ||
+                   IP6T_MATCH_ITERATE(e, do_match, skb, &mtpar) != 0) {
+                       e = ip6t_next_entry(e);
+                       continue;
+               }
 
-                               e = get_entry(table_base, v);
-                       } else {
-                               /* Targets which reenter must return
-                                   abs. verdicts */
-#ifdef CONFIG_NETFILTER_DEBUG
-                               ((struct ip6t_entry *)table_base)->comefrom
-                                       = 0xeeeeeeec;
-#endif
-                               verdict = t->u.kernel.target->target(pskb,
-                                                                    in, out,
-                                                                    hook,
-                                                                    t->u.kernel.target,
-                                                                    t->data);
+               ADD_COUNTER(e->counters,
+                           ntohs(ipv6_hdr(skb)->payload_len) +
+                           sizeof(struct ipv6hdr), 1);
 
-#ifdef CONFIG_NETFILTER_DEBUG
-                               if (((struct ip6t_entry *)table_base)->comefrom
-                                   != 0xeeeeeeec
-                                   && verdict == IP6T_CONTINUE) {
-                                       printk("Target %s reentered!\n",
-                                              t->u.kernel.target->name);
-                                       verdict = NF_DROP;
-                               }
-                               ((struct ip6t_entry *)table_base)->comefrom
-                                       = 0x57acc001;
+               t = ip6t_get_target(e);
+               IP_NF_ASSERT(t->u.kernel.target);
+
+#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
+    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
+               /* The packet is traced: log it */
+               if (unlikely(skb->nf_trace))
+                       trace_packet(skb, hook, in, out,
+                                    table->name, private, e);
 #endif
-                               if (verdict == IP6T_CONTINUE)
-                                       e = (void *)e + e->next_offset;
-                               else
-                                       /* Verdict */
+               /* Standard target? */
+               if (!t->u.kernel.target->target) {
+                       int v;
+
+                       v = ((struct ip6t_standard_target *)t)->verdict;
+                       if (v < 0) {
+                               /* Pop from stack? */
+                               if (v != IP6T_RETURN) {
+                                       verdict = (unsigned)(-v) - 1;
                                        break;
+                               }
+                               e = back;
+                               back = get_entry(table_base, back->comefrom);
+                               continue;
+                       }
+                       if (table_base + v != ip6t_next_entry(e) &&
+                           !(e->ipv6.flags & IP6T_F_GOTO)) {
+                               /* Save old back ptr in next entry */
+                               struct ip6t_entry *next = ip6t_next_entry(e);
+                               next->comefrom = (void *)back - table_base;
+                               /* set back pointer to next entry */
+                               back = next;
                        }
-               } else {
 
-               no_match:
-                       e = (void *)e + e->next_offset;
+                       e = get_entry(table_base, v);
+                       continue;
+               }
+
+               /* Targets which reenter must return
+                  abs. verdicts */
+               tgpar.target   = t->u.kernel.target;
+               tgpar.targinfo = t->data;
+
+#ifdef CONFIG_NETFILTER_DEBUG
+               tb_comefrom = 0xeeeeeeec;
+#endif
+               verdict = t->u.kernel.target->target(skb, &tgpar);
+
+#ifdef CONFIG_NETFILTER_DEBUG
+               if (tb_comefrom != 0xeeeeeeec && verdict == IP6T_CONTINUE) {
+                       printk("Target %s reentered!\n",
+                              t->u.kernel.target->name);
+                       verdict = NF_DROP;
                }
+               tb_comefrom = 0x57acc001;
+#endif
+               if (verdict == IP6T_CONTINUE)
+                       e = ip6t_next_entry(e);
+               else
+                       /* Verdict */
+                       break;
        } while (!hotdrop);
 
 #ifdef CONFIG_NETFILTER_DEBUG
-       ((struct ip6t_entry *)table_base)->comefrom = NETFILTER_LINK_POISON;
+       tb_comefrom = NETFILTER_LINK_POISON;
 #endif
-       read_unlock_bh(&table->lock);
+       xt_info_rdunlock_bh();
 
 #ifdef DEBUG_ALLOW_ALL
        return NF_ACCEPT;
@@ -384,19 +468,8 @@ ip6t_do_table(struct sk_buff **pskb,
                return NF_DROP;
        else return verdict;
 #endif
-}
 
-/* All zeroes == unconditional rule. */
-static inline int
-unconditional(const struct ip6t_ip6 *ipv6)
-{
-       unsigned int i;
-
-       for (i = 0; i < sizeof(*ipv6); i++)
-               if (((char *)ipv6)[i])
-                       break;
-
-       return (i == sizeof(*ipv6));
+#undef tb_comefrom
 }
 
 /* Figures out from what hook each rule can be called: returns 0 if
@@ -409,10 +482,9 @@ mark_source_chains(struct xt_table_info *newinfo,
 
        /* No recursion; use packet counter to save back ptrs (reset
           to 0 as we leave), and comefrom to save source hook bitmask */
-       for (hook = 0; hook < NF_IP6_NUMHOOKS; hook++) {
+       for (hook = 0; hook < NF_INET_NUMHOOKS; hook++) {
                unsigned int pos = newinfo->hook_entry[hook];
-               struct ip6t_entry *e
-                       = (struct ip6t_entry *)(entry0 + pos);
+               struct ip6t_entry *e = (struct ip6t_entry *)(entry0 + pos);
 
                if (!(valid_hooks & (1 << hook)))
                        continue;
@@ -423,30 +495,39 @@ mark_source_chains(struct xt_table_info *newinfo,
                for (;;) {
                        struct ip6t_standard_target *t
                                = (void *)ip6t_get_target(e);
+                       int visited = e->comefrom & (1 << hook);
 
-                       if (e->comefrom & (1 << NF_IP6_NUMHOOKS)) {
+                       if (e->comefrom & (1 << NF_INET_NUMHOOKS)) {
                                printk("iptables: loop hook %u pos %u %08X.\n",
                                       hook, pos, e->comefrom);
                                return 0;
                        }
-                       e->comefrom
-                               |= ((1 << hook) | (1 << NF_IP6_NUMHOOKS));
+                       e->comefrom |= ((1 << hook) | (1 << NF_INET_NUMHOOKS));
 
                        /* Unconditional return/END. */
-                       if (e->target_offset == sizeof(struct ip6t_entry)
-                           && (strcmp(t->target.u.user.name,
-                                      IP6T_STANDARD_TARGET) == 0)
-                           && t->verdict < 0
-                           && unconditional(&e->ipv6)) {
+                       if ((e->target_offset == sizeof(struct ip6t_entry) &&
+                            (strcmp(t->target.u.user.name,
+                                    IP6T_STANDARD_TARGET) == 0) &&
+                            t->verdict < 0 &&
+                            unconditional(&e->ipv6)) || visited) {
                                unsigned int oldpos, size;
 
+                               if ((strcmp(t->target.u.user.name,
+                                           IP6T_STANDARD_TARGET) == 0) &&
+                                   t->verdict < -NF_MAX_VERDICT - 1) {
+                                       duprintf("mark_source_chains: bad "
+                                               "negative verdict (%i)\n",
+                                                               t->verdict);
+                                       return 0;
+                               }
+
                                /* Return: backtrack through the last
                                   big jump. */
                                do {
-                                       e->comefrom ^= (1<<NF_IP6_NUMHOOKS);
+                                       e->comefrom ^= (1<<NF_INET_NUMHOOKS);
 #ifdef DEBUG_IP_FIREWALL_USER
                                        if (e->comefrom
-                                           & (1 << NF_IP6_NUMHOOKS)) {
+                                           & (1 << NF_INET_NUMHOOKS)) {
                                                duprintf("Back unset "
                                                         "on hook %u "
                                                         "rule %u\n",
@@ -475,8 +556,15 @@ mark_source_chains(struct xt_table_info *newinfo,
                                int newpos = t->verdict;
 
                                if (strcmp(t->target.u.user.name,
-                                          IP6T_STANDARD_TARGET) == 0
-                                   && newpos >= 0) {
+                                          IP6T_STANDARD_TARGET) == 0 &&
+                                   newpos >= 0) {
+                                       if (newpos > newinfo->size -
+                                               sizeof(struct ip6t_entry)) {
+                                               duprintf("mark_source_chains: "
+                                                       "bad verdict (%i)\n",
+                                                               newpos);
+                                               return 0;
+                                       }
                                        /* This a jump; chase it. */
                                        duprintf("Jump rule %u -> %u\n",
                                                 pos, newpos);
@@ -496,162 +584,194 @@ mark_source_chains(struct xt_table_info *newinfo,
        return 1;
 }
 
-static inline int
-cleanup_match(struct ip6t_entry_match *m, unsigned int *i)
+static int
+cleanup_match(struct ip6t_entry_match *m, struct net *net, unsigned int *i)
 {
+       struct xt_mtdtor_param par;
+
        if (i && (*i)-- == 0)
                return 1;
 
-       if (m->u.kernel.match->destroy)
-               m->u.kernel.match->destroy(m->u.kernel.match, m->data);
-       module_put(m->u.kernel.match->me);
+       par.net       = net;
+       par.match     = m->u.kernel.match;
+       par.matchinfo = m->data;
+       par.family    = NFPROTO_IPV6;
+       if (par.match->destroy != NULL)
+               par.match->destroy(&par);
+       module_put(par.match->me);
        return 0;
 }
 
-static inline int
-standard_check(const struct ip6t_entry_target *t,
-              unsigned int max_offset)
+static int
+check_entry(struct ip6t_entry *e, const char *name)
 {
-       struct ip6t_standard_target *targ = (void *)t;
-
-       /* Check standard info. */
-       if (targ->verdict >= 0
-           && targ->verdict > max_offset - sizeof(struct ip6t_entry)) {
-               duprintf("ip6t_standard_check: bad verdict (%i)\n",
-                        targ->verdict);
-               return 0;
+       struct ip6t_entry_target *t;
+
+       if (!ip6_checkentry(&e->ipv6)) {
+               duprintf("ip_tables: ip check failed %p %s.\n", e, name);
+               return -EINVAL;
        }
-       if (targ->verdict < -NF_MAX_VERDICT - 1) {
-               duprintf("ip6t_standard_check: bad negative verdict (%i)\n",
-                        targ->verdict);
-               return 0;
+
+       if (e->target_offset + sizeof(struct ip6t_entry_target) >
+           e->next_offset)
+               return -EINVAL;
+
+       t = ip6t_get_target(e);
+       if (e->target_offset + t->u.target_size > e->next_offset)
+               return -EINVAL;
+
+       return 0;
+}
+
+static int check_match(struct ip6t_entry_match *m, struct xt_mtchk_param *par,
+                      unsigned int *i)
+{
+       const struct ip6t_ip6 *ipv6 = par->entryinfo;
+       int ret;
+
+       par->match     = m->u.kernel.match;
+       par->matchinfo = m->data;
+
+       ret = xt_check_match(par, m->u.match_size - sizeof(*m),
+                            ipv6->proto, ipv6->invflags & IP6T_INV_PROTO);
+       if (ret < 0) {
+               duprintf("ip_tables: check failed for `%s'.\n",
+                        par.match->name);
+               return ret;
        }
-       return 1;
+       ++*i;
+       return 0;
 }
 
-static inline int
-check_match(struct ip6t_entry_match *m,
-           const char *name,
-           const struct ip6t_ip6 *ipv6,
-           unsigned int hookmask,
-           unsigned int *i)
+static int
+find_check_match(struct ip6t_entry_match *m, struct xt_mtchk_param *par,
+                unsigned int *i)
 {
-       struct ip6t_match *match;
+       struct xt_match *match;
        int ret;
 
        match = try_then_request_module(xt_find_match(AF_INET6, m->u.user.name,
-                                       m->u.user.revision),
+                                                     m->u.user.revision),
                                        "ip6t_%s", m->u.user.name);
        if (IS_ERR(match) || !match) {
-               duprintf("check_match: `%s' not found\n", m->u.user.name);
+               duprintf("find_check_match: `%s' not found\n", m->u.user.name);
                return match ? PTR_ERR(match) : -ENOENT;
        }
        m->u.kernel.match = match;
 
-       ret = xt_check_match(match, AF_INET6, m->u.match_size - sizeof(*m),
-                            name, hookmask, ipv6->proto,
-                            ipv6->invflags & IP6T_INV_PROTO);
+       ret = check_match(m, par, i);
        if (ret)
                goto err;
 
-       if (m->u.kernel.match->checkentry
-           && !m->u.kernel.match->checkentry(name, ipv6, match,  m->data,
-                                             hookmask)) {
-               duprintf("ip_tables: check failed for `%s'.\n",
-                        m->u.kernel.match->name);
-               ret = -EINVAL;
-               goto err;
-       }
-
-       (*i)++;
        return 0;
 err:
        module_put(m->u.kernel.match->me);
        return ret;
 }
 
-static struct ip6t_target ip6t_standard_target;
+static int check_target(struct ip6t_entry *e, struct net *net, const char *name)
+{
+       struct ip6t_entry_target *t = ip6t_get_target(e);
+       struct xt_tgchk_param par = {
+               .net       = net,
+               .table     = name,
+               .entryinfo = e,
+               .target    = t->u.kernel.target,
+               .targinfo  = t->data,
+               .hook_mask = e->comefrom,
+               .family    = NFPROTO_IPV6,
+       };
+       int ret;
 
-static inline int
-check_entry(struct ip6t_entry *e, const char *name, unsigned int size,
-           unsigned int *i)
+       t = ip6t_get_target(e);
+       ret = xt_check_target(&par, t->u.target_size - sizeof(*t),
+             e->ipv6.proto, e->ipv6.invflags & IP6T_INV_PROTO);
+       if (ret < 0) {
+               duprintf("ip_tables: check failed for `%s'.\n",
+                        t->u.kernel.target->name);
+               return ret;
+       }
+       return 0;
+}
+
+static int
+find_check_entry(struct ip6t_entry *e, struct net *net, const char *name,
+                unsigned int size, unsigned int *i)
 {
        struct ip6t_entry_target *t;
-       struct ip6t_target *target;
+       struct xt_target *target;
        int ret;
        unsigned int j;
+       struct xt_mtchk_param mtpar;
 
-       if (!ip6_checkentry(&e->ipv6)) {
-               duprintf("ip_tables: ip check failed %p %s.\n", e, name);
-               return -EINVAL;
-       }
-
-       if (e->target_offset + sizeof(struct ip6t_entry_target) >
-                                                               e->next_offset)
-               return -EINVAL;
+       ret = check_entry(e, name);
+       if (ret)
+               return ret;
 
        j = 0;
-       ret = IP6T_MATCH_ITERATE(e, check_match, name, &e->ipv6, e->comefrom, &j);
+       mtpar.net       = net;
+       mtpar.table     = name;
+       mtpar.entryinfo = &e->ipv6;
+       mtpar.hook_mask = e->comefrom;
+       mtpar.family    = NFPROTO_IPV6;
+       ret = IP6T_MATCH_ITERATE(e, find_check_match, &mtpar, &j);
        if (ret != 0)
                goto cleanup_matches;
 
        t = ip6t_get_target(e);
-       ret = -EINVAL;
-       if (e->target_offset + t->u.target_size > e->next_offset)
-                       goto cleanup_matches;
        target = try_then_request_module(xt_find_target(AF_INET6,
                                                        t->u.user.name,
                                                        t->u.user.revision),
                                         "ip6t_%s", t->u.user.name);
        if (IS_ERR(target) || !target) {
-               duprintf("check_entry: `%s' not found\n", t->u.user.name);
+               duprintf("find_check_entry: `%s' not found\n", t->u.user.name);
                ret = target ? PTR_ERR(target) : -ENOENT;
                goto cleanup_matches;
        }
        t->u.kernel.target = target;
 
-       ret = xt_check_target(target, AF_INET6, t->u.target_size - sizeof(*t),
-                             name, e->comefrom, e->ipv6.proto,
-                             e->ipv6.invflags & IP6T_INV_PROTO);
+       ret = check_target(e, net, name);
        if (ret)
                goto err;
 
-       if (t->u.kernel.target == &ip6t_standard_target) {
-               if (!standard_check(t, size)) {
-                       ret = -EINVAL;
-                       goto err;
-               }
-       } else if (t->u.kernel.target->checkentry
-                  && !t->u.kernel.target->checkentry(name, e, target, t->data,
-                                                     e->comefrom)) {
-               duprintf("ip_tables: check failed for `%s'.\n",
-                        t->u.kernel.target->name);
-               ret = -EINVAL;
-               goto err;
-       }
-
        (*i)++;
        return 0;
  err:
        module_put(t->u.kernel.target->me);
  cleanup_matches:
-       IP6T_MATCH_ITERATE(e, cleanup_match, &j);
+       IP6T_MATCH_ITERATE(e, cleanup_match, net, &j);
        return ret;
 }
 
-static inline int
+static bool check_underflow(struct ip6t_entry *e)
+{
+       const struct ip6t_entry_target *t;
+       unsigned int verdict;
+
+       if (!unconditional(&e->ipv6))
+               return false;
+       t = ip6t_get_target(e);
+       if (strcmp(t->u.user.name, XT_STANDARD_TARGET) != 0)
+               return false;
+       verdict = ((struct ip6t_standard_target *)t)->verdict;
+       verdict = -verdict - 1;
+       return verdict == NF_DROP || verdict == NF_ACCEPT;
+}
+
+static int
 check_entry_size_and_hooks(struct ip6t_entry *e,
                           struct xt_table_info *newinfo,
                           unsigned char *base,
                           unsigned char *limit,
                           const unsigned int *hook_entries,
                           const unsigned int *underflows,
+                          unsigned int valid_hooks,
                           unsigned int *i)
 {
        unsigned int h;
 
-       if ((unsigned long)e % __alignof__(struct ip6t_entry) != 0
-           || (unsigned char *)e + sizeof(struct ip6t_entry) >= limit) {
+       if ((unsigned long)e % __alignof__(struct ip6t_entry) != 0 ||
+           (unsigned char *)e + sizeof(struct ip6t_entry) >= limit) {
                duprintf("Bad offset %p\n", e);
                return -EINVAL;
        }
@@ -664,16 +784,22 @@ check_entry_size_and_hooks(struct ip6t_entry *e,
        }
 
        /* Check hooks & underflows */
-       for (h = 0; h < NF_IP6_NUMHOOKS; h++) {
+       for (h = 0; h < NF_INET_NUMHOOKS; h++) {
+               if (!(valid_hooks & (1 << h)))
+                       continue;
                if ((unsigned char *)e - base == hook_entries[h])
                        newinfo->hook_entry[h] = hook_entries[h];
-               if ((unsigned char *)e - base == underflows[h])
+               if ((unsigned char *)e - base == underflows[h]) {
+                       if (!check_underflow(e)) {
+                               pr_err("Underflows must be unconditional and "
+                                      "use the STANDARD target with "
+                                      "ACCEPT/DROP\n");
+                               return -EINVAL;
+                       }
                        newinfo->underflow[h] = underflows[h];
+               }
        }
 
-       /* FIXME: underflows must be unconditional, standard verdicts
-           < 0 (not IP6T_RETURN). --RR */
-
        /* Clear counters and comefrom */
        e->counters = ((struct xt_counters) { 0, 0 });
        e->comefrom = 0;
@@ -682,27 +808,34 @@ check_entry_size_and_hooks(struct ip6t_entry *e,
        return 0;
 }
 
-static inline int
-cleanup_entry(struct ip6t_entry *e, unsigned int *i)
+static int
+cleanup_entry(struct ip6t_entry *e, struct net *net, unsigned int *i)
 {
+       struct xt_tgdtor_param par;
        struct ip6t_entry_target *t;
 
        if (i && (*i)-- == 0)
                return 1;
 
        /* Cleanup all matches */
-       IP6T_MATCH_ITERATE(e, cleanup_match, NULL);
+       IP6T_MATCH_ITERATE(e, cleanup_match, net, NULL);
        t = ip6t_get_target(e);
-       if (t->u.kernel.target->destroy)
-               t->u.kernel.target->destroy(t->u.kernel.target, t->data);
-       module_put(t->u.kernel.target->me);
+
+       par.net      = net;
+       par.target   = t->u.kernel.target;
+       par.targinfo = t->data;
+       par.family   = NFPROTO_IPV6;
+       if (par.target->destroy != NULL)
+               par.target->destroy(&par);
+       module_put(par.target->me);
        return 0;
 }
 
 /* Checks and translates the user-supplied table segment (held in
    newinfo) */
 static int
-translate_table(const char *name,
+translate_table(struct net *net,
+               const char *name,
                unsigned int valid_hooks,
                struct xt_table_info *newinfo,
                void *entry0,
@@ -718,7 +851,7 @@ translate_table(const char *name,
        newinfo->number = number;
 
        /* Init all hooks to impossible value. */
-       for (i = 0; i < NF_IP6_NUMHOOKS; i++) {
+       for (i = 0; i < NF_INET_NUMHOOKS; i++) {
                newinfo->hook_entry[i] = 0xFFFFFFFF;
                newinfo->underflow[i] = 0xFFFFFFFF;
        }
@@ -731,7 +864,7 @@ translate_table(const char *name,
                                newinfo,
                                entry0,
                                entry0 + size,
-                               hook_entries, underflows, &i);
+                               hook_entries, underflows, valid_hooks, &i);
        if (ret != 0)
                return ret;
 
@@ -742,7 +875,7 @@ translate_table(const char *name,
        }
 
        /* Check hooks all assigned */
-       for (i = 0; i < NF_IP6_NUMHOOKS; i++) {
+       for (i = 0; i < NF_INET_NUMHOOKS; i++) {
                /* Only hooks which are valid */
                if (!(valid_hooks & (1 << i)))
                        continue;
@@ -758,17 +891,19 @@ translate_table(const char *name,
                }
        }
 
+       if (!mark_source_chains(newinfo, valid_hooks, entry0))
+               return -ELOOP;
+
        /* Finally, each sanity check must pass */
        i = 0;
        ret = IP6T_ENTRY_ITERATE(entry0, newinfo->size,
-                               check_entry, name, size, &i);
-
-       if (ret != 0)
-               goto cleanup;
+                               find_check_entry, net, name, size, &i);
 
-       ret = -ELOOP;
-       if (!mark_source_chains(newinfo, valid_hooks, entry0))
-               goto cleanup;
+       if (ret != 0) {
+               IP6T_ENTRY_ITERATE(entry0, newinfo->size,
+                                  cleanup_entry, net, &i);
+               return ret;
+       }
 
        /* And one copy for every other CPU */
        for_each_possible_cpu(i) {
@@ -776,9 +911,6 @@ translate_table(const char *name,
                        memcpy(newinfo->entries[i], entry0, newinfo->size);
        }
 
-       return 0;
-cleanup:
-       IP6T_ENTRY_ITERATE(entry0, newinfo->size, cleanup_entry, &i);
        return ret;
 }
 
@@ -816,9 +948,12 @@ get_counters(const struct xt_table_info *t,
        /* Instead of clearing (by a previous call to memset())
         * the counters and using adds, we set the counters
         * with data used by 'current' CPU
-        * We dont care about preemption here.
+        *
+        * Bottom half has to be disabled to prevent deadlock
+        * if new softirq were to run and call ipt_do_table
         */
-       curcpu = raw_smp_processor_id();
+       local_bh_disable();
+       curcpu = smp_processor_id();
 
        i = 0;
        IP6T_ENTRY_ITERATE(t->entries[curcpu],
@@ -831,41 +966,57 @@ get_counters(const struct xt_table_info *t,
                if (cpu == curcpu)
                        continue;
                i = 0;
+               xt_info_wrlock(cpu);
                IP6T_ENTRY_ITERATE(t->entries[cpu],
                                  t->size,
                                  add_entry_to_counter,
                                  counters,
                                  &i);
+               xt_info_wrunlock(cpu);
        }
+       local_bh_enable();
 }
 
-static int
-copy_entries_to_user(unsigned int total_size,
-                    struct xt_table *table,
-                    void __user *userptr)
+static struct xt_counters *alloc_counters(struct xt_table *table)
 {
-       unsigned int off, num, countersize;
-       struct ip6t_entry *e;
+       unsigned int countersize;
        struct xt_counters *counters;
        struct xt_table_info *private = table->private;
-       int ret = 0;
-       void *loc_cpu_entry;
 
        /* We need atomic snapshot of counters: rest doesn't change
           (other than comefrom, which userspace doesn't care
           about). */
        countersize = sizeof(struct xt_counters) * private->number;
-       counters = vmalloc(countersize);
+       counters = vmalloc_node(countersize, numa_node_id());
 
        if (counters == NULL)
-               return -ENOMEM;
+               return ERR_PTR(-ENOMEM);
 
-       /* First, sum counters... */
-       write_lock_bh(&table->lock);
        get_counters(private, counters);
-       write_unlock_bh(&table->lock);
 
-       /* choose the copy that is on ourc node/cpu */
+       return counters;
+}
+
+static int
+copy_entries_to_user(unsigned int total_size,
+                    struct xt_table *table,
+                    void __user *userptr)
+{
+       unsigned int off, num;
+       struct ip6t_entry *e;
+       struct xt_counters *counters;
+       const struct xt_table_info *private = table->private;
+       int ret = 0;
+       const void *loc_cpu_entry;
+
+       counters = alloc_counters(table);
+       if (IS_ERR(counters))
+               return PTR_ERR(counters);
+
+       /* choose the copy that is on our node/cpu, ...
+        * This choice is lazy (because current thread is
+        * allowed to migrate to another cpu)
+        */
        loc_cpu_entry = private->entries[raw_smp_processor_id()];
        if (copy_to_user(userptr, loc_cpu_entry, total_size) != 0) {
                ret = -EFAULT;
@@ -876,8 +1027,8 @@ copy_entries_to_user(unsigned int total_size,
        /* ... then go back and fix counters and names */
        for (off = 0, num = 0; off < total_size; off += e->next_offset, num++){
                unsigned int i;
-               struct ip6t_entry_match *m;
-               struct ip6t_entry_target *t;
+               const struct ip6t_entry_match *m;
+               const struct ip6t_entry_target *t;
 
                e = (struct ip6t_entry *)(loc_cpu_entry + off);
                if (copy_to_user(userptr + off
@@ -920,116 +1071,235 @@ copy_entries_to_user(unsigned int total_size,
        return ret;
 }
 
-static int
-get_entries(const struct ip6t_get_entries *entries,
-           struct ip6t_get_entries __user *uptr)
+#ifdef CONFIG_COMPAT
+static void compat_standard_from_user(void *dst, void *src)
 {
-       int ret;
-       struct xt_table *t;
-
-       t = xt_find_table_lock(AF_INET6, entries->name);
-       if (t && !IS_ERR(t)) {
-               struct xt_table_info *private = t->private;
-               duprintf("t->private->number = %u\n", private->number);
-               if (entries->size == private->size)
-                       ret = copy_entries_to_user(private->size,
-                                                  t, uptr->entrytable);
-               else {
-                       duprintf("get_entries: I've got %u not %u!\n",
-                                private->size, entries->size);
-                       ret = -EINVAL;
-               }
-               module_put(t->me);
-               xt_table_unlock(t);
-       } else
-               ret = t ? PTR_ERR(t) : -ENOENT;
+       int v = *(compat_int_t *)src;
 
-       return ret;
+       if (v > 0)
+               v += xt_compat_calc_jump(AF_INET6, v);
+       memcpy(dst, &v, sizeof(v));
 }
 
-static int
-do_replace(void __user *user, unsigned int len)
+static int compat_standard_to_user(void __user *dst, void *src)
 {
-       int ret;
-       struct ip6t_replace tmp;
-       struct xt_table *t;
-       struct xt_table_info *newinfo, *oldinfo;
-       struct xt_counters *counters;
-       void *loc_cpu_entry, *loc_cpu_old_entry;
+       compat_int_t cv = *(int *)src;
 
-       if (copy_from_user(&tmp, user, sizeof(tmp)) != 0)
-               return -EFAULT;
+       if (cv > 0)
+               cv -= xt_compat_calc_jump(AF_INET6, cv);
+       return copy_to_user(dst, &cv, sizeof(cv)) ? -EFAULT : 0;
+}
 
-       /* overflow check */
-       if (tmp.size >= (INT_MAX - sizeof(struct xt_table_info)) / NR_CPUS -
-                       SMP_CACHE_BYTES)
-               return -ENOMEM;
-       if (tmp.num_counters >= INT_MAX / sizeof(struct xt_counters))
-               return -ENOMEM;
+static inline int
+compat_calc_match(struct ip6t_entry_match *m, int *size)
+{
+       *size += xt_compat_match_offset(m->u.kernel.match);
+       return 0;
+}
 
-       newinfo = xt_alloc_table_info(tmp.size);
-       if (!newinfo)
-               return -ENOMEM;
+static int compat_calc_entry(struct ip6t_entry *e,
+                            const struct xt_table_info *info,
+                            void *base, struct xt_table_info *newinfo)
+{
+       struct ip6t_entry_target *t;
+       unsigned int entry_offset;
+       int off, i, ret;
 
-       /* choose the copy that is on our node/cpu */
-       loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
-       if (copy_from_user(loc_cpu_entry, user + sizeof(tmp),
-                          tmp.size) != 0) {
-               ret = -EFAULT;
-               goto free_newinfo;
-       }
+       off = sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);
+       entry_offset = (void *)e - base;
+       IP6T_MATCH_ITERATE(e, compat_calc_match, &off);
+       t = ip6t_get_target(e);
+       off += xt_compat_target_offset(t->u.kernel.target);
+       newinfo->size -= off;
+       ret = xt_compat_add_offset(AF_INET6, entry_offset, off);
+       if (ret)
+               return ret;
 
-       counters = vmalloc(tmp.num_counters * sizeof(struct xt_counters));
-       if (!counters) {
-               ret = -ENOMEM;
-               goto free_newinfo;
+       for (i = 0; i < NF_INET_NUMHOOKS; i++) {
+               if (info->hook_entry[i] &&
+                   (e < (struct ip6t_entry *)(base + info->hook_entry[i])))
+                       newinfo->hook_entry[i] -= off;
+               if (info->underflow[i] &&
+                   (e < (struct ip6t_entry *)(base + info->underflow[i])))
+                       newinfo->underflow[i] -= off;
        }
+       return 0;
+}
 
-       ret = translate_table(tmp.name, tmp.valid_hooks,
-                             newinfo, loc_cpu_entry, tmp.size, tmp.num_entries,
-                             tmp.hook_entry, tmp.underflow);
-       if (ret != 0)
-               goto free_newinfo_counters;
+static int compat_table_info(const struct xt_table_info *info,
+                            struct xt_table_info *newinfo)
+{
+       void *loc_cpu_entry;
 
-       duprintf("ip_tables: Translated table\n");
+       if (!newinfo || !info)
+               return -EINVAL;
+
+       /* we dont care about newinfo->entries[] */
+       memcpy(newinfo, info, offsetof(struct xt_table_info, entries));
+       newinfo->initial_entries = 0;
+       loc_cpu_entry = info->entries[raw_smp_processor_id()];
+       return IP6T_ENTRY_ITERATE(loc_cpu_entry, info->size,
+                                 compat_calc_entry, info, loc_cpu_entry,
+                                 newinfo);
+}
+#endif
+
+static int get_info(struct net *net, void __user *user, int *len, int compat)
+{
+       char name[IP6T_TABLE_MAXNAMELEN];
+       struct xt_table *t;
+       int ret;
+
+       if (*len != sizeof(struct ip6t_getinfo)) {
+               duprintf("length %u != %zu\n", *len,
+                        sizeof(struct ip6t_getinfo));
+               return -EINVAL;
+       }
+
+       if (copy_from_user(name, user, sizeof(name)) != 0)
+               return -EFAULT;
+
+       name[IP6T_TABLE_MAXNAMELEN-1] = '\0';
+#ifdef CONFIG_COMPAT
+       if (compat)
+               xt_compat_lock(AF_INET6);
+#endif
+       t = try_then_request_module(xt_find_table_lock(net, AF_INET6, name),
+                                   "ip6table_%s", name);
+       if (t && !IS_ERR(t)) {
+               struct ip6t_getinfo info;
+               const struct xt_table_info *private = t->private;
+#ifdef CONFIG_COMPAT
+               struct xt_table_info tmp;
+
+               if (compat) {
+                       ret = compat_table_info(private, &tmp);
+                       xt_compat_flush_offsets(AF_INET6);
+                       private = &tmp;
+               }
+#endif
+               info.valid_hooks = t->valid_hooks;
+               memcpy(info.hook_entry, private->hook_entry,
+                      sizeof(info.hook_entry));
+               memcpy(info.underflow, private->underflow,
+                      sizeof(info.underflow));
+               info.num_entries = private->number;
+               info.size = private->size;
+               strcpy(info.name, name);
+
+               if (copy_to_user(user, &info, *len) != 0)
+                       ret = -EFAULT;
+               else
+                       ret = 0;
+
+               xt_table_unlock(t);
+               module_put(t->me);
+       } else
+               ret = t ? PTR_ERR(t) : -ENOENT;
+#ifdef CONFIG_COMPAT
+       if (compat)
+               xt_compat_unlock(AF_INET6);
+#endif
+       return ret;
+}
+
+static int
+get_entries(struct net *net, struct ip6t_get_entries __user *uptr, int *len)
+{
+       int ret;
+       struct ip6t_get_entries get;
+       struct xt_table *t;
+
+       if (*len < sizeof(get)) {
+               duprintf("get_entries: %u < %zu\n", *len, sizeof(get));
+               return -EINVAL;
+       }
+       if (copy_from_user(&get, uptr, sizeof(get)) != 0)
+               return -EFAULT;
+       if (*len != sizeof(struct ip6t_get_entries) + get.size) {
+               duprintf("get_entries: %u != %zu\n",
+                        *len, sizeof(get) + get.size);
+               return -EINVAL;
+       }
+
+       t = xt_find_table_lock(net, AF_INET6, get.name);
+       if (t && !IS_ERR(t)) {
+               struct xt_table_info *private = t->private;
+               duprintf("t->private->number = %u\n", private->number);
+               if (get.size == private->size)
+                       ret = copy_entries_to_user(private->size,
+                                                  t, uptr->entrytable);
+               else {
+                       duprintf("get_entries: I've got %u not %u!\n",
+                                private->size, get.size);
+                       ret = -EAGAIN;
+               }
+               module_put(t->me);
+               xt_table_unlock(t);
+       } else
+               ret = t ? PTR_ERR(t) : -ENOENT;
+
+       return ret;
+}
+
+static int
+__do_replace(struct net *net, const char *name, unsigned int valid_hooks,
+            struct xt_table_info *newinfo, unsigned int num_counters,
+            void __user *counters_ptr)
+{
+       int ret;
+       struct xt_table *t;
+       struct xt_table_info *oldinfo;
+       struct xt_counters *counters;
+       const void *loc_cpu_old_entry;
+
+       ret = 0;
+       counters = vmalloc_node(num_counters * sizeof(struct xt_counters),
+                               numa_node_id());
+       if (!counters) {
+               ret = -ENOMEM;
+               goto out;
+       }
 
-       t = try_then_request_module(xt_find_table_lock(AF_INET6, tmp.name),
-                                   "ip6table_%s", tmp.name);
+       t = try_then_request_module(xt_find_table_lock(net, AF_INET6, name),
+                                   "ip6table_%s", name);
        if (!t || IS_ERR(t)) {
                ret = t ? PTR_ERR(t) : -ENOENT;
                goto free_newinfo_counters_untrans;
        }
 
        /* You lied! */
-       if (tmp.valid_hooks != t->valid_hooks) {
+       if (valid_hooks != t->valid_hooks) {
                duprintf("Valid hook crap: %08X vs %08X\n",
-                        tmp.valid_hooks, t->valid_hooks);
+                        valid_hooks, t->valid_hooks);
                ret = -EINVAL;
                goto put_module;
        }
 
-       oldinfo = xt_replace_table(t, tmp.num_counters, newinfo, &ret);
+       oldinfo = xt_replace_table(t, num_counters, newinfo, &ret);
        if (!oldinfo)
                goto put_module;
 
        /* Update module usage count based on number of rules */
        duprintf("do_replace: oldnum=%u, initnum=%u, newnum=%u\n",
                oldinfo->number, oldinfo->initial_entries, newinfo->number);
-       if ((oldinfo->number > oldinfo->initial_entries) || 
-           (newinfo->number <= oldinfo->initial_entries)) 
+       if ((oldinfo->number > oldinfo->initial_entries) ||
+           (newinfo->number <= oldinfo->initial_entries))
                module_put(t->me);
        if ((oldinfo->number > oldinfo->initial_entries) &&
            (newinfo->number <= oldinfo->initial_entries))
                module_put(t->me);
 
-       /* Get the old counters. */
+       /* Get the old counters, and synchronize with replace */
        get_counters(oldinfo, counters);
+
        /* Decrease module usage counts and free resource */
        loc_cpu_old_entry = oldinfo->entries[raw_smp_processor_id()];
-       IP6T_ENTRY_ITERATE(loc_cpu_old_entry, oldinfo->size, cleanup_entry,NULL);
+       IP6T_ENTRY_ITERATE(loc_cpu_old_entry, oldinfo->size, cleanup_entry,
+                          net, NULL);
        xt_free_table_info(oldinfo);
-       if (copy_to_user(tmp.counters, counters,
-                        sizeof(struct xt_counters) * tmp.num_counters) != 0)
+       if (copy_to_user(counters_ptr, counters,
+                        sizeof(struct xt_counters) * num_counters) != 0)
                ret = -EFAULT;
        vfree(counters);
        xt_table_unlock(t);
@@ -1039,9 +1309,54 @@ do_replace(void __user *user, unsigned int len)
        module_put(t->me);
        xt_table_unlock(t);
  free_newinfo_counters_untrans:
-       IP6T_ENTRY_ITERATE(loc_cpu_entry, newinfo->size, cleanup_entry,NULL);
- free_newinfo_counters:
        vfree(counters);
+ out:
+       return ret;
+}
+
+static int
+do_replace(struct net *net, void __user *user, unsigned int len)
+{
+       int ret;
+       struct ip6t_replace tmp;
+       struct xt_table_info *newinfo;
+       void *loc_cpu_entry;
+
+       if (copy_from_user(&tmp, user, sizeof(tmp)) != 0)
+               return -EFAULT;
+
+       /* overflow check */
+       if (tmp.num_counters >= INT_MAX / sizeof(struct xt_counters))
+               return -ENOMEM;
+
+       newinfo = xt_alloc_table_info(tmp.size);
+       if (!newinfo)
+               return -ENOMEM;
+
+       /* choose the copy that is on our node/cpu */
+       loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
+       if (copy_from_user(loc_cpu_entry, user + sizeof(tmp),
+                          tmp.size) != 0) {
+               ret = -EFAULT;
+               goto free_newinfo;
+       }
+
+       ret = translate_table(net, tmp.name, tmp.valid_hooks,
+                             newinfo, loc_cpu_entry, tmp.size, tmp.num_entries,
+                             tmp.hook_entry, tmp.underflow);
+       if (ret != 0)
+               goto free_newinfo;
+
+       duprintf("ip_tables: Translated table\n");
+
+       ret = __do_replace(net, tmp.name, tmp.valid_hooks, newinfo,
+                          tmp.num_counters, tmp.counters);
+       if (ret)
+               goto free_newinfo_untrans;
+       return 0;
+
+ free_newinfo_untrans:
+       IP6T_ENTRY_ITERATE(loc_cpu_entry, newinfo->size, cleanup_entry, net, NULL);
  free_newinfo:
        xt_free_table_info(newinfo);
        return ret;
@@ -1049,20 +1364,11 @@ do_replace(void __user *user, unsigned int len)
 
 /* We're lazy, and add to the first CPU; overflow works its fey magic
  * and everything is OK. */
-static inline int
+static int
 add_counter_to_entry(struct ip6t_entry *e,
                     const struct xt_counters addme[],
                     unsigned int *i)
 {
-#if 0
-       duprintf("add_counter: Entry %u %lu/%lu + %lu/%lu\n",
-                *i,
-                (long unsigned int)e->counters.pcnt,
-                (long unsigned int)e->counters.bcnt,
-                (long unsigned int)addme[*i].pcnt,
-                (long unsigned int)addme[*i].bcnt);
-#endif
-
        ADD_COUNTER(e->counters, addme[*i].bcnt, addme[*i].pcnt);
 
        (*i)++;
@@ -1070,53 +1376,87 @@ add_counter_to_entry(struct ip6t_entry *e,
 }
 
 static int
-do_add_counters(void __user *user, unsigned int len)
+do_add_counters(struct net *net, void __user *user, unsigned int len,
+               int compat)
 {
-       unsigned int i;
-       struct xt_counters_info tmp, *paddc;
-       struct xt_table_info *private;
+       unsigned int i, curcpu;
+       struct xt_counters_info tmp;
+       struct xt_counters *paddc;
+       unsigned int num_counters;
+       char *name;
+       int size;
+       void *ptmp;
        struct xt_table *t;
+       const struct xt_table_info *private;
        int ret = 0;
-       void *loc_cpu_entry;
+       const void *loc_cpu_entry;
+#ifdef CONFIG_COMPAT
+       struct compat_xt_counters_info compat_tmp;
 
-       if (copy_from_user(&tmp, user, sizeof(tmp)) != 0)
+       if (compat) {
+               ptmp = &compat_tmp;
+               size = sizeof(struct compat_xt_counters_info);
+       } else
+#endif
+       {
+               ptmp = &tmp;
+               size = sizeof(struct xt_counters_info);
+       }
+
+       if (copy_from_user(ptmp, user, size) != 0)
                return -EFAULT;
 
-       if (len != sizeof(tmp) + tmp.num_counters*sizeof(struct xt_counters))
+#ifdef CONFIG_COMPAT
+       if (compat) {
+               num_counters = compat_tmp.num_counters;
+               name = compat_tmp.name;
+       } else
+#endif
+       {
+               num_counters = tmp.num_counters;
+               name = tmp.name;
+       }
+
+       if (len != size + num_counters * sizeof(struct xt_counters))
                return -EINVAL;
 
-       paddc = vmalloc(len);
+       paddc = vmalloc_node(len - size, numa_node_id());
        if (!paddc)
                return -ENOMEM;
 
-       if (copy_from_user(paddc, user, len) != 0) {
+       if (copy_from_user(paddc, user + size, len - size) != 0) {
                ret = -EFAULT;
                goto free;
        }
 
-       t = xt_find_table_lock(AF_INET6, tmp.name);
+       t = xt_find_table_lock(net, AF_INET6, name);
        if (!t || IS_ERR(t)) {
                ret = t ? PTR_ERR(t) : -ENOENT;
                goto free;
        }
 
-       write_lock_bh(&t->lock);
+
+       local_bh_disable();
        private = t->private;
-       if (private->number != tmp.num_counters) {
+       if (private->number != num_counters) {
                ret = -EINVAL;
                goto unlock_up_free;
        }
 
        i = 0;
        /* Choose the copy that is on our node */
-       loc_cpu_entry = private->entries[smp_processor_id()];
+       curcpu = smp_processor_id();
+       xt_info_wrlock(curcpu);
+       loc_cpu_entry = private->entries[curcpu];
        IP6T_ENTRY_ITERATE(loc_cpu_entry,
                          private->size,
                          add_counter_to_entry,
-                         paddc->counters,
+                         paddc,
                          &i);
+       xt_info_wrunlock(curcpu);
+
  unlock_up_free:
-       write_unlock_bh(&t->lock);
+       local_bh_enable();
        xt_table_unlock(t);
        module_put(t->me);
  free:
@@ -1125,8 +1465,441 @@ do_add_counters(void __user *user, unsigned int len)
        return ret;
 }
 
+#ifdef CONFIG_COMPAT
+struct compat_ip6t_replace {
+       char                    name[IP6T_TABLE_MAXNAMELEN];
+       u32                     valid_hooks;
+       u32                     num_entries;
+       u32                     size;
+       u32                     hook_entry[NF_INET_NUMHOOKS];
+       u32                     underflow[NF_INET_NUMHOOKS];
+       u32                     num_counters;
+       compat_uptr_t           counters;       /* struct ip6t_counters * */
+       struct compat_ip6t_entry entries[0];
+};
+
 static int
-do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
+compat_copy_entry_to_user(struct ip6t_entry *e, void __user **dstptr,
+                         unsigned int *size, struct xt_counters *counters,
+                         unsigned int *i)
+{
+       struct ip6t_entry_target *t;
+       struct compat_ip6t_entry __user *ce;
+       u_int16_t target_offset, next_offset;
+       compat_uint_t origsize;
+       int ret;
+
+       ret = -EFAULT;
+       origsize = *size;
+       ce = (struct compat_ip6t_entry __user *)*dstptr;
+       if (copy_to_user(ce, e, sizeof(struct ip6t_entry)))
+               goto out;
+
+       if (copy_to_user(&ce->counters, &counters[*i], sizeof(counters[*i])))
+               goto out;
+
+       *dstptr += sizeof(struct compat_ip6t_entry);
+       *size -= sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);
+
+       ret = IP6T_MATCH_ITERATE(e, xt_compat_match_to_user, dstptr, size);
+       target_offset = e->target_offset - (origsize - *size);
+       if (ret)
+               goto out;
+       t = ip6t_get_target(e);
+       ret = xt_compat_target_to_user(t, dstptr, size);
+       if (ret)
+               goto out;
+       ret = -EFAULT;
+       next_offset = e->next_offset - (origsize - *size);
+       if (put_user(target_offset, &ce->target_offset))
+               goto out;
+       if (put_user(next_offset, &ce->next_offset))
+               goto out;
+
+       (*i)++;
+       return 0;
+out:
+       return ret;
+}
+
+static int
+compat_find_calc_match(struct ip6t_entry_match *m,
+                      const char *name,
+                      const struct ip6t_ip6 *ipv6,
+                      unsigned int hookmask,
+                      int *size, unsigned int *i)
+{
+       struct xt_match *match;
+
+       match = try_then_request_module(xt_find_match(AF_INET6, m->u.user.name,
+                                                     m->u.user.revision),
+                                       "ip6t_%s", m->u.user.name);
+       if (IS_ERR(match) || !match) {
+               duprintf("compat_check_calc_match: `%s' not found\n",
+                        m->u.user.name);
+               return match ? PTR_ERR(match) : -ENOENT;
+       }
+       m->u.kernel.match = match;
+       *size += xt_compat_match_offset(match);
+
+       (*i)++;
+       return 0;
+}
+
+static int
+compat_release_match(struct ip6t_entry_match *m, unsigned int *i)
+{
+       if (i && (*i)-- == 0)
+               return 1;
+
+       module_put(m->u.kernel.match->me);
+       return 0;
+}
+
+static int
+compat_release_entry(struct compat_ip6t_entry *e, unsigned int *i)
+{
+       struct ip6t_entry_target *t;
+
+       if (i && (*i)-- == 0)
+               return 1;
+
+       /* Cleanup all matches */
+       COMPAT_IP6T_MATCH_ITERATE(e, compat_release_match, NULL);
+       t = compat_ip6t_get_target(e);
+       module_put(t->u.kernel.target->me);
+       return 0;
+}
+
+static int
+check_compat_entry_size_and_hooks(struct compat_ip6t_entry *e,
+                                 struct xt_table_info *newinfo,
+                                 unsigned int *size,
+                                 unsigned char *base,
+                                 unsigned char *limit,
+                                 unsigned int *hook_entries,
+                                 unsigned int *underflows,
+                                 unsigned int *i,
+                                 const char *name)
+{
+       struct ip6t_entry_target *t;
+       struct xt_target *target;
+       unsigned int entry_offset;
+       unsigned int j;
+       int ret, off, h;
+
+       duprintf("check_compat_entry_size_and_hooks %p\n", e);
+       if ((unsigned long)e % __alignof__(struct compat_ip6t_entry) != 0 ||
+           (unsigned char *)e + sizeof(struct compat_ip6t_entry) >= limit) {
+               duprintf("Bad offset %p, limit = %p\n", e, limit);
+               return -EINVAL;
+       }
+
+       if (e->next_offset < sizeof(struct compat_ip6t_entry) +
+                            sizeof(struct compat_xt_entry_target)) {
+               duprintf("checking: element %p size %u\n",
+                        e, e->next_offset);
+               return -EINVAL;
+       }
+
+       /* For purposes of check_entry casting the compat entry is fine */
+       ret = check_entry((struct ip6t_entry *)e, name);
+       if (ret)
+               return ret;
+
+       off = sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);
+       entry_offset = (void *)e - (void *)base;
+       j = 0;
+       ret = COMPAT_IP6T_MATCH_ITERATE(e, compat_find_calc_match, name,
+                                       &e->ipv6, e->comefrom, &off, &j);
+       if (ret != 0)
+               goto release_matches;
+
+       t = compat_ip6t_get_target(e);
+       target = try_then_request_module(xt_find_target(AF_INET6,
+                                                       t->u.user.name,
+                                                       t->u.user.revision),
+                                        "ip6t_%s", t->u.user.name);
+       if (IS_ERR(target) || !target) {
+               duprintf("check_compat_entry_size_and_hooks: `%s' not found\n",
+                        t->u.user.name);
+               ret = target ? PTR_ERR(target) : -ENOENT;
+               goto release_matches;
+       }
+       t->u.kernel.target = target;
+
+       off += xt_compat_target_offset(target);
+       *size += off;
+       ret = xt_compat_add_offset(AF_INET6, entry_offset, off);
+       if (ret)
+               goto out;
+
+       /* Check hooks & underflows */
+       for (h = 0; h < NF_INET_NUMHOOKS; h++) {
+               if ((unsigned char *)e - base == hook_entries[h])
+                       newinfo->hook_entry[h] = hook_entries[h];
+               if ((unsigned char *)e - base == underflows[h])
+                       newinfo->underflow[h] = underflows[h];
+       }
+
+       /* Clear counters and comefrom */
+       memset(&e->counters, 0, sizeof(e->counters));
+       e->comefrom = 0;
+
+       (*i)++;
+       return 0;
+
+out:
+       module_put(t->u.kernel.target->me);
+release_matches:
+       IP6T_MATCH_ITERATE(e, compat_release_match, &j);
+       return ret;
+}
+
+static int
+compat_copy_entry_from_user(struct compat_ip6t_entry *e, void **dstptr,
+                           unsigned int *size, const char *name,
+                           struct xt_table_info *newinfo, unsigned char *base)
+{
+       struct ip6t_entry_target *t;
+       struct xt_target *target;
+       struct ip6t_entry *de;
+       unsigned int origsize;
+       int ret, h;
+
+       ret = 0;
+       origsize = *size;
+       de = (struct ip6t_entry *)*dstptr;
+       memcpy(de, e, sizeof(struct ip6t_entry));
+       memcpy(&de->counters, &e->counters, sizeof(e->counters));
+
+       *dstptr += sizeof(struct ip6t_entry);
+       *size += sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);
+
+       ret = COMPAT_IP6T_MATCH_ITERATE(e, xt_compat_match_from_user,
+                                       dstptr, size);
+       if (ret)
+               return ret;
+       de->target_offset = e->target_offset - (origsize - *size);
+       t = compat_ip6t_get_target(e);
+       target = t->u.kernel.target;
+       xt_compat_target_from_user(t, dstptr, size);
+
+       de->next_offset = e->next_offset - (origsize - *size);
+       for (h = 0; h < NF_INET_NUMHOOKS; h++) {
+               if ((unsigned char *)de - base < newinfo->hook_entry[h])
+                       newinfo->hook_entry[h] -= origsize - *size;
+               if ((unsigned char *)de - base < newinfo->underflow[h])
+                       newinfo->underflow[h] -= origsize - *size;
+       }
+       return ret;
+}
+
+static int compat_check_entry(struct ip6t_entry *e, struct net *net,
+                             const char *name, unsigned int *i)
+{
+       unsigned int j;
+       int ret;
+       struct xt_mtchk_param mtpar;
+
+       j = 0;
+       mtpar.net       = net;
+       mtpar.table     = name;
+       mtpar.entryinfo = &e->ipv6;
+       mtpar.hook_mask = e->comefrom;
+       mtpar.family    = NFPROTO_IPV6;
+       ret = IP6T_MATCH_ITERATE(e, check_match, &mtpar, &j);
+       if (ret)
+               goto cleanup_matches;
+
+       ret = check_target(e, net, name);
+       if (ret)
+               goto cleanup_matches;
+
+       (*i)++;
+       return 0;
+
+ cleanup_matches:
+       IP6T_MATCH_ITERATE(e, cleanup_match, net, &j);
+       return ret;
+}
+
+static int
+translate_compat_table(struct net *net,
+                      const char *name,
+                      unsigned int valid_hooks,
+                      struct xt_table_info **pinfo,
+                      void **pentry0,
+                      unsigned int total_size,
+                      unsigned int number,
+                      unsigned int *hook_entries,
+                      unsigned int *underflows)
+{
+       unsigned int i, j;
+       struct xt_table_info *newinfo, *info;
+       void *pos, *entry0, *entry1;
+       unsigned int size;
+       int ret;
+
+       info = *pinfo;
+       entry0 = *pentry0;
+       size = total_size;
+       info->number = number;
+
+       /* Init all hooks to impossible value. */
+       for (i = 0; i < NF_INET_NUMHOOKS; i++) {
+               info->hook_entry[i] = 0xFFFFFFFF;
+               info->underflow[i] = 0xFFFFFFFF;
+       }
+
+       duprintf("translate_compat_table: size %u\n", info->size);
+       j = 0;
+       xt_compat_lock(AF_INET6);
+       /* Walk through entries, checking offsets. */
+       ret = COMPAT_IP6T_ENTRY_ITERATE(entry0, total_size,
+                                       check_compat_entry_size_and_hooks,
+                                       info, &size, entry0,
+                                       entry0 + total_size,
+                                       hook_entries, underflows, &j, name);
+       if (ret != 0)
+               goto out_unlock;
+
+       ret = -EINVAL;
+       if (j != number) {
+               duprintf("translate_compat_table: %u not %u entries\n",
+                        j, number);
+               goto out_unlock;
+       }
+
+       /* Check hooks all assigned */
+       for (i = 0; i < NF_INET_NUMHOOKS; i++) {
+               /* Only hooks which are valid */
+               if (!(valid_hooks & (1 << i)))
+                       continue;
+               if (info->hook_entry[i] == 0xFFFFFFFF) {
+                       duprintf("Invalid hook entry %u %u\n",
+                                i, hook_entries[i]);
+                       goto out_unlock;
+               }
+               if (info->underflow[i] == 0xFFFFFFFF) {
+                       duprintf("Invalid underflow %u %u\n",
+                                i, underflows[i]);
+                       goto out_unlock;
+               }
+       }
+
+       ret = -ENOMEM;
+       newinfo = xt_alloc_table_info(size);
+       if (!newinfo)
+               goto out_unlock;
+
+       newinfo->number = number;
+       for (i = 0; i < NF_INET_NUMHOOKS; i++) {
+               newinfo->hook_entry[i] = info->hook_entry[i];
+               newinfo->underflow[i] = info->underflow[i];
+       }
+       entry1 = newinfo->entries[raw_smp_processor_id()];
+       pos = entry1;
+       size = total_size;
+       ret = COMPAT_IP6T_ENTRY_ITERATE(entry0, total_size,
+                                       compat_copy_entry_from_user,
+                                       &pos, &size, name, newinfo, entry1);
+       xt_compat_flush_offsets(AF_INET6);
+       xt_compat_unlock(AF_INET6);
+       if (ret)
+               goto free_newinfo;
+
+       ret = -ELOOP;
+       if (!mark_source_chains(newinfo, valid_hooks, entry1))
+               goto free_newinfo;
+
+       i = 0;
+       ret = IP6T_ENTRY_ITERATE(entry1, newinfo->size, compat_check_entry,
+                                net, name, &i);
+       if (ret) {
+               j -= i;
+               COMPAT_IP6T_ENTRY_ITERATE_CONTINUE(entry0, newinfo->size, i,
+                                                  compat_release_entry, &j);
+               IP6T_ENTRY_ITERATE(entry1, newinfo->size, cleanup_entry, net, &i);
+               xt_free_table_info(newinfo);
+               return ret;
+       }
+
+       /* And one copy for every other CPU */
+       for_each_possible_cpu(i)
+               if (newinfo->entries[i] && newinfo->entries[i] != entry1)
+                       memcpy(newinfo->entries[i], entry1, newinfo->size);
+
+       *pinfo = newinfo;
+       *pentry0 = entry1;
+       xt_free_table_info(info);
+       return 0;
+
+free_newinfo:
+       xt_free_table_info(newinfo);
+out:
+       COMPAT_IP6T_ENTRY_ITERATE(entry0, total_size, compat_release_entry, &j);
+       return ret;
+out_unlock:
+       xt_compat_flush_offsets(AF_INET6);
+       xt_compat_unlock(AF_INET6);
+       goto out;
+}
+
+static int
+compat_do_replace(struct net *net, void __user *user, unsigned int len)
+{
+       int ret;
+       struct compat_ip6t_replace tmp;
+       struct xt_table_info *newinfo;
+       void *loc_cpu_entry;
+
+       if (copy_from_user(&tmp, user, sizeof(tmp)) != 0)
+               return -EFAULT;
+
+       /* overflow check */
+       if (tmp.size >= INT_MAX / num_possible_cpus())
+               return -ENOMEM;
+       if (tmp.num_counters >= INT_MAX / sizeof(struct xt_counters))
+               return -ENOMEM;
+
+       newinfo = xt_alloc_table_info(tmp.size);
+       if (!newinfo)
+               return -ENOMEM;
+
+       /* choose the copy that is on our node/cpu */
+       loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
+       if (copy_from_user(loc_cpu_entry, user + sizeof(tmp),
+                          tmp.size) != 0) {
+               ret = -EFAULT;
+               goto free_newinfo;
+       }
+
+       ret = translate_compat_table(net, tmp.name, tmp.valid_hooks,
+                                    &newinfo, &loc_cpu_entry, tmp.size,
+                                    tmp.num_entries, tmp.hook_entry,
+                                    tmp.underflow);
+       if (ret != 0)
+               goto free_newinfo;
+
+       duprintf("compat_do_replace: Translated table\n");
+
+       ret = __do_replace(net, tmp.name, tmp.valid_hooks, newinfo,
+                          tmp.num_counters, compat_ptr(tmp.counters));
+       if (ret)
+               goto free_newinfo_untrans;
+       return 0;
+
+ free_newinfo_untrans:
+       IP6T_ENTRY_ITERATE(loc_cpu_entry, newinfo->size, cleanup_entry, net, NULL);
+ free_newinfo:
+       xt_free_table_info(newinfo);
+       return ret;
+}
+
+static int
+compat_do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user,
+                      unsigned int len)
 {
        int ret;
 
@@ -1135,11 +1908,11 @@ do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
 
        switch (cmd) {
        case IP6T_SO_SET_REPLACE:
-               ret = do_replace(user, len);
+               ret = compat_do_replace(sock_net(sk), user, len);
                break;
 
        case IP6T_SO_SET_ADD_COUNTERS:
-               ret = do_add_counters(user, len);
+               ret = do_add_counters(sock_net(sk), user, len, 1);
                break;
 
        default:
@@ -1150,8 +1923,94 @@ do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
        return ret;
 }
 
+struct compat_ip6t_get_entries {
+       char name[IP6T_TABLE_MAXNAMELEN];
+       compat_uint_t size;
+       struct compat_ip6t_entry entrytable[0];
+};
+
+static int
+compat_copy_entries_to_user(unsigned int total_size, struct xt_table *table,
+                           void __user *userptr)
+{
+       struct xt_counters *counters;
+       const struct xt_table_info *private = table->private;
+       void __user *pos;
+       unsigned int size;
+       int ret = 0;
+       const void *loc_cpu_entry;
+       unsigned int i = 0;
+
+       counters = alloc_counters(table);
+       if (IS_ERR(counters))
+               return PTR_ERR(counters);
+
+       /* choose the copy that is on our node/cpu, ...
+        * This choice is lazy (because current thread is
+        * allowed to migrate to another cpu)
+        */
+       loc_cpu_entry = private->entries[raw_smp_processor_id()];
+       pos = userptr;
+       size = total_size;
+       ret = IP6T_ENTRY_ITERATE(loc_cpu_entry, total_size,
+                                compat_copy_entry_to_user,
+                                &pos, &size, counters, &i);
+
+       vfree(counters);
+       return ret;
+}
+
+static int
+compat_get_entries(struct net *net, struct compat_ip6t_get_entries __user *uptr,
+                  int *len)
+{
+       int ret;
+       struct compat_ip6t_get_entries get;
+       struct xt_table *t;
+
+       if (*len < sizeof(get)) {
+               duprintf("compat_get_entries: %u < %zu\n", *len, sizeof(get));
+               return -EINVAL;
+       }
+
+       if (copy_from_user(&get, uptr, sizeof(get)) != 0)
+               return -EFAULT;
+
+       if (*len != sizeof(struct compat_ip6t_get_entries) + get.size) {
+               duprintf("compat_get_entries: %u != %zu\n",
+                        *len, sizeof(get) + get.size);
+               return -EINVAL;
+       }
+
+       xt_compat_lock(AF_INET6);
+       t = xt_find_table_lock(net, AF_INET6, get.name);
+       if (t && !IS_ERR(t)) {
+               const struct xt_table_info *private = t->private;
+               struct xt_table_info info;
+               duprintf("t->private->number = %u\n", private->number);
+               ret = compat_table_info(private, &info);
+               if (!ret && get.size == info.size) {
+                       ret = compat_copy_entries_to_user(private->size,
+                                                         t, uptr->entrytable);
+               } else if (!ret) {
+                       duprintf("compat_get_entries: I've got %u not %u!\n",
+                                private->size, get.size);
+                       ret = -EAGAIN;
+               }
+               xt_compat_flush_offsets(AF_INET6);
+               module_put(t->me);
+               xt_table_unlock(t);
+       } else
+               ret = t ? PTR_ERR(t) : -ENOENT;
+
+       xt_compat_unlock(AF_INET6);
+       return ret;
+}
+
+static int do_ip6t_get_ctl(struct sock *, int, void __user *, int *);
+
 static int
-do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
+compat_do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
 {
        int ret;
 
@@ -1159,66 +2018,61 @@ do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
                return -EPERM;
 
        switch (cmd) {
-       case IP6T_SO_GET_INFO: {
-               char name[IP6T_TABLE_MAXNAMELEN];
-               struct xt_table *t;
+       case IP6T_SO_GET_INFO:
+               ret = get_info(sock_net(sk), user, len, 1);
+               break;
+       case IP6T_SO_GET_ENTRIES:
+               ret = compat_get_entries(sock_net(sk), user, len);
+               break;
+       default:
+               ret = do_ip6t_get_ctl(sk, cmd, user, len);
+       }
+       return ret;
+}
+#endif
 
-               if (*len != sizeof(struct ip6t_getinfo)) {
-                       duprintf("length %u != %u\n", *len,
-                                sizeof(struct ip6t_getinfo));
-                       ret = -EINVAL;
-                       break;
-               }
+static int
+do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
+{
+       int ret;
 
-               if (copy_from_user(name, user, sizeof(name)) != 0) {
-                       ret = -EFAULT;
-                       break;
-               }
-               name[IP6T_TABLE_MAXNAMELEN-1] = '\0';
-
-               t = try_then_request_module(xt_find_table_lock(AF_INET6, name),
-                                           "ip6table_%s", name);
-               if (t && !IS_ERR(t)) {
-                       struct ip6t_getinfo info;
-                       struct xt_table_info *private = t->private;
-
-                       info.valid_hooks = t->valid_hooks;
-                       memcpy(info.hook_entry, private->hook_entry,
-                              sizeof(info.hook_entry));
-                       memcpy(info.underflow, private->underflow,
-                              sizeof(info.underflow));
-                       info.num_entries = private->number;
-                       info.size = private->size;
-                       memcpy(info.name, name, sizeof(info.name));
-
-                       if (copy_to_user(user, &info, *len) != 0)
-                               ret = -EFAULT;
-                       else
-                               ret = 0;
-                       xt_table_unlock(t);
-                       module_put(t->me);
-               } else
-                       ret = t ? PTR_ERR(t) : -ENOENT;
-       }
-       break;
+       if (!capable(CAP_NET_ADMIN))
+               return -EPERM;
 
-       case IP6T_SO_GET_ENTRIES: {
-               struct ip6t_get_entries get;
+       switch (cmd) {
+       case IP6T_SO_SET_REPLACE:
+               ret = do_replace(sock_net(sk), user, len);
+               break;
 
-               if (*len < sizeof(get)) {
-                       duprintf("get_entries: %u < %u\n", *len, sizeof(get));
-                       ret = -EINVAL;
-               } else if (copy_from_user(&get, user, sizeof(get)) != 0) {
-                       ret = -EFAULT;
-               } else if (*len != sizeof(struct ip6t_get_entries) + get.size) {
-                       duprintf("get_entries: %u != %u\n", *len,
-                                sizeof(struct ip6t_get_entries) + get.size);
-                       ret = -EINVAL;
-               } else
-                       ret = get_entries(&get, user);
+       case IP6T_SO_SET_ADD_COUNTERS:
+               ret = do_add_counters(sock_net(sk), user, len, 0);
                break;
+
+       default:
+               duprintf("do_ip6t_set_ctl:  unknown request %i\n", cmd);
+               ret = -EINVAL;
        }
 
+       return ret;
+}
+
+static int
+do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
+{
+       int ret;
+
+       if (!capable(CAP_NET_ADMIN))
+               return -EPERM;
+
+       switch (cmd) {
+       case IP6T_SO_GET_INFO:
+               ret = get_info(sock_net(sk), user, len, 0);
+               break;
+
+       case IP6T_SO_GET_ENTRIES:
+               ret = get_entries(sock_net(sk), user, len);
+               break;
+
        case IP6T_SO_GET_REVISION_MATCH:
        case IP6T_SO_GET_REVISION_TARGET: {
                struct ip6t_get_revision rev;
@@ -1253,89 +2107,93 @@ do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
        return ret;
 }
 
-int ip6t_register_table(struct xt_table *table,
-                       const struct ip6t_replace *repl)
+struct xt_table *ip6t_register_table(struct net *net,
+                                    const struct xt_table *table,
+                                    const struct ip6t_replace *repl)
 {
        int ret;
        struct xt_table_info *newinfo;
-       static struct xt_table_info bootstrap
+       struct xt_table_info bootstrap
                = { 0, 0, 0, { 0 }, { 0 }, { } };
        void *loc_cpu_entry;
+       struct xt_table *new_table;
 
        newinfo = xt_alloc_table_info(repl->size);
-       if (!newinfo)
-               return -ENOMEM;
+       if (!newinfo) {
+               ret = -ENOMEM;
+               goto out;
+       }
 
-       /* choose the copy on our node/cpu */
+       /* choose the copy on our node/cpu, but dont care about preemption */
        loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
        memcpy(loc_cpu_entry, repl->entries, repl->size);
 
-       ret = translate_table(table->name, table->valid_hooks,
+       ret = translate_table(net, table->name, table->valid_hooks,
                              newinfo, loc_cpu_entry, repl->size,
                              repl->num_entries,
                              repl->hook_entry,
                              repl->underflow);
-       if (ret != 0) {
-               xt_free_table_info(newinfo);
-               return ret;
-       }
+       if (ret != 0)
+               goto out_free;
 
-       ret = xt_register_table(table, &bootstrap, newinfo);
-       if (ret != 0) {
-               xt_free_table_info(newinfo);
-               return ret;
+       new_table = xt_register_table(net, table, &bootstrap, newinfo);
+       if (IS_ERR(new_table)) {
+               ret = PTR_ERR(new_table);
+               goto out_free;
        }
+       return new_table;
 
-       return 0;
+out_free:
+       xt_free_table_info(newinfo);
+out:
+       return ERR_PTR(ret);
 }
 
-void ip6t_unregister_table(struct xt_table *table)
+void ip6t_unregister_table(struct net *net, struct xt_table *table)
 {
        struct xt_table_info *private;
        void *loc_cpu_entry;
+       struct module *table_owner = table->me;
 
        private = xt_unregister_table(table);
 
        /* Decrease module usage counts and free resources */
        loc_cpu_entry = private->entries[raw_smp_processor_id()];
-       IP6T_ENTRY_ITERATE(loc_cpu_entry, private->size, cleanup_entry, NULL);
+       IP6T_ENTRY_ITERATE(loc_cpu_entry, private->size, cleanup_entry, net, NULL);
+       if (private->number > private->initial_entries)
+               module_put(table_owner);
        xt_free_table_info(private);
 }
 
 /* Returns 1 if the type and code is matched by the range, 0 otherwise */
-static inline int
+static inline bool
 icmp6_type_code_match(u_int8_t test_type, u_int8_t min_code, u_int8_t max_code,
                     u_int8_t type, u_int8_t code,
-                    int invert)
+                    bool invert)
 {
        return (type == test_type && code >= min_code && code <= max_code)
                ^ invert;
 }
 
-static int
-icmp6_match(const struct sk_buff *skb,
-          const struct net_device *in,
-          const struct net_device *out,
-          const struct xt_match *match,
-          const void *matchinfo,
-          int offset,
-          unsigned int protoff,
-          int *hotdrop)
+static bool
+icmp6_match(const struct sk_buff *skb, const struct xt_match_param *par)
 {
-       struct icmp6hdr _icmp, *ic;
-       const struct ip6t_icmp *icmpinfo = matchinfo;
+       const struct icmp6hdr *ic;
+       struct icmp6hdr _icmph;
+       const struct ip6t_icmp *icmpinfo = par->matchinfo;
 
        /* Must not be a fragment. */
-       if (offset)
-               return 0;
+       if (par->fragoff != 0)
+               return false;
 
-       ic = skb_header_pointer(skb, protoff, sizeof(_icmp), &_icmp);
+       ic = skb_header_pointer(skb, par->thoff, sizeof(_icmph), &_icmph);
        if (ic == NULL) {
                /* We've been asked to examine this packet, and we
-                  can't.  Hence, no choice but to drop. */
+                * can't.  Hence, no choice but to drop.
+                */
                duprintf("Dropping evil ICMP tinygram.\n");
-               *hotdrop = 1;
-               return 0;
+               *par->hotdrop = true;
+               return false;
        }
 
        return icmp6_type_code_match(icmpinfo->type,
@@ -1346,31 +2204,31 @@ icmp6_match(const struct sk_buff *skb,
 }
 
 /* Called when user tries to insert an entry of this type. */
-static int
-icmp6_checkentry(const char *tablename,
-          const void *entry,
-          const struct xt_match *match,
-          void *matchinfo,
-          unsigned int hook_mask)
+static bool icmp6_checkentry(const struct xt_mtchk_param *par)
 {
-       const struct ip6t_icmp *icmpinfo = matchinfo;
+       const struct ip6t_icmp *icmpinfo = par->matchinfo;
 
        /* Must specify no unknown invflags */
        return !(icmpinfo->invflags & ~IP6T_ICMP_INV);
 }
 
 /* The built-in targets: standard (NULL) and error. */
-static struct ip6t_target ip6t_standard_target = {
+static struct xt_target ip6t_standard_target __read_mostly = {
        .name           = IP6T_STANDARD_TARGET,
        .targetsize     = sizeof(int),
-       .family         = AF_INET6,
+       .family         = NFPROTO_IPV6,
+#ifdef CONFIG_COMPAT
+       .compatsize     = sizeof(compat_int_t),
+       .compat_from_user = compat_standard_from_user,
+       .compat_to_user = compat_standard_to_user,
+#endif
 };
 
-static struct ip6t_target ip6t_error_target = {
+static struct xt_target ip6t_error_target __read_mostly = {
        .name           = IP6T_ERROR_TARGET,
        .target         = ip6t_error,
        .targetsize     = IP6T_FUNCTION_MAXNAMELEN,
-       .family         = AF_INET6,
+       .family         = NFPROTO_IPV6,
 };
 
 static struct nf_sockopt_ops ip6t_sockopts = {
@@ -1378,25 +2236,47 @@ static struct nf_sockopt_ops ip6t_sockopts = {
        .set_optmin     = IP6T_BASE_CTL,
        .set_optmax     = IP6T_SO_SET_MAX+1,
        .set            = do_ip6t_set_ctl,
+#ifdef CONFIG_COMPAT
+       .compat_set     = compat_do_ip6t_set_ctl,
+#endif
        .get_optmin     = IP6T_BASE_CTL,
        .get_optmax     = IP6T_SO_GET_MAX+1,
        .get            = do_ip6t_get_ctl,
+#ifdef CONFIG_COMPAT
+       .compat_get     = compat_do_ip6t_get_ctl,
+#endif
+       .owner          = THIS_MODULE,
 };
 
-static struct ip6t_match icmp6_matchstruct = {
+static struct xt_match icmp6_matchstruct __read_mostly = {
        .name           = "icmp6",
-       .match          = &icmp6_match,
+       .match          = icmp6_match,
        .matchsize      = sizeof(struct ip6t_icmp),
        .checkentry     = icmp6_checkentry,
        .proto          = IPPROTO_ICMPV6,
-       .family         = AF_INET6,
+       .family         = NFPROTO_IPV6,
+};
+
+static int __net_init ip6_tables_net_init(struct net *net)
+{
+       return xt_proto_init(net, NFPROTO_IPV6);
+}
+
+static void __net_exit ip6_tables_net_exit(struct net *net)
+{
+       xt_proto_fini(net, NFPROTO_IPV6);
+}
+
+static struct pernet_operations ip6_tables_net_ops = {
+       .init = ip6_tables_net_init,
+       .exit = ip6_tables_net_exit,
 };
 
 static int __init ip6_tables_init(void)
 {
        int ret;
 
-       ret = xt_proto_init(AF_INET6);
+       ret = register_pernet_subsys(&ip6_tables_net_ops);
        if (ret < 0)
                goto err1;
 
@@ -1416,7 +2296,7 @@ static int __init ip6_tables_init(void)
        if (ret < 0)
                goto err5;
 
-       printk("ip6_tables: (C) 2000-2006 Netfilter Core Team\n");
+       printk(KERN_INFO "ip6_tables: (C) 2000-2006 Netfilter Core Team\n");
        return 0;
 
 err5:
@@ -1426,7 +2306,7 @@ err4:
 err3:
        xt_unregister_target(&ip6t_standard_target);
 err2:
-       xt_proto_fini(AF_INET6);
+       unregister_pernet_subsys(&ip6_tables_net_ops);
 err1:
        return ret;
 }
@@ -1434,10 +2314,12 @@ err1:
 static void __exit ip6_tables_fini(void)
 {
        nf_unregister_sockopt(&ip6t_sockopts);
+
        xt_unregister_match(&icmp6_matchstruct);
        xt_unregister_target(&ip6t_error_target);
        xt_unregister_target(&ip6t_standard_target);
-       xt_proto_fini(AF_INET6);
+
+       unregister_pernet_subsys(&ip6_tables_net_ops);
 }
 
 /*
@@ -1460,8 +2342,8 @@ static void __exit ip6_tables_fini(void)
 int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset,
                  int target, unsigned short *fragoff)
 {
-       unsigned int start = (u8*)(skb->nh.ipv6h + 1) - skb->data;
-       u8 nexthdr = skb->nh.ipv6h->nexthdr;
+       unsigned int start = skb_network_offset(skb) + sizeof(struct ipv6hdr);
+       u8 nexthdr = ipv6_hdr(skb)->nexthdr;
        unsigned int len = skb->len - start;
 
        if (fragoff)
@@ -1504,9 +2386,9 @@ int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset,
                        }
                        hdrlen = 8;
                } else if (nexthdr == NEXTHDR_AUTH)
-                       hdrlen = (hp->hdrlen + 2) << 2; 
+                       hdrlen = (hp->hdrlen + 2) << 2;
                else
-                       hdrlen = ipv6_optlen(hp); 
+                       hdrlen = ipv6_optlen(hp);
 
                nexthdr = hp->nexthdr;
                len -= hdrlen;