[NETFILTER]: Handle NAT module load race

When the NAT module is loaded when connections are already confirmed
it must not change their tuples anymore. This is especially important
with CONFIG_NETFILTER_DEBUG, the netfilter listhelp functions will
refuse to remove an entry from a list when it can not be found on
the list, so when a changed tuple hashes to a new bucket the entry
is kept in the list until and after the conntrack is freed.

Allocate the exact conntrack tuple for NAT for already confirmed
connections or drop them if that fails.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

3 files changed, 32 insertions, 2 deletions

diff --git a/include/linux/netfilter_ipv4/ip_nat_rule.h b/include/linux/netfilter_ipv4/ip_nat_rule.h
index fecd2a06dcd8..73b9552e6a89 100644
--- a/include/linux/netfilter_ipv4/ip_nat_rule.h
+++ b/include/linux/netfilter_ipv4/ip_nat_rule.h
@@ -19,5 +19,10 @@ extern unsigned int
 alloc_null_binding(struct ip_conntrack *conntrack,
                    struct ip_nat_info *info,
                    unsigned int hooknum);
+
+extern unsigned int
+alloc_null_binding_confirmed(struct ip_conntrack *conntrack,
+                             struct ip_nat_info *info,
+                             unsigned int hooknum);
 #endif
 #endif /* _IP_NAT_RULE_H */
diff --git a/net/ipv4/netfilter/ip_nat_rule.c b/net/ipv4/netfilter/ip_nat_rule.c
index 60d70fa41a15..cb66b8bddeb3 100644
--- a/net/ipv4/netfilter/ip_nat_rule.c
+++ b/net/ipv4/netfilter/ip_nat_rule.c
@@ -255,6 +255,27 @@ alloc_null_binding(struct ip_conntrack *conntrack,
         return ip_nat_setup_info(conntrack, &range, hooknum);
 }
 
+unsigned int
+alloc_null_binding_confirmed(struct ip_conntrack *conntrack,
+                             struct ip_nat_info *info,
+                             unsigned int hooknum)
+{
+        u_int32_t ip
+                = (HOOK2MANIP(hooknum) == IP_NAT_MANIP_SRC
+                   ? conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip
+                   : conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip);
+        u_int16_t all
+                = (HOOK2MANIP(hooknum) == IP_NAT_MANIP_SRC
+                   ? conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.all
+                   : conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.src.u.all);
+        struct ip_nat_range range
+                = { IP_NAT_RANGE_MAP_IPS, ip, ip, { all }, { all } };
+
+        DEBUGP("Allocating NULL binding for confirmed %p (%u.%u.%u.%u)\n",
+               conntrack, NIPQUAD(ip));
+        return ip_nat_setup_info(conntrack, &range, hooknum);
+}
+
 int ip_nat_rule_find(struct sk_buff **pskb,
                      unsigned int hooknum,
                      const struct net_device *in,
diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c
index 89db052add81..0ff368b131f6 100644
--- a/net/ipv4/netfilter/ip_nat_standalone.c
+++ b/net/ipv4/netfilter/ip_nat_standalone.c
@@ -123,8 +123,12 @@ ip_nat_fn(unsigned int hooknum,
                 if (!ip_nat_initialized(ct, maniptype)) {
                         unsigned int ret;
 
-                        /* LOCAL_IN hook doesn't have a chain!  */
-                        if (hooknum == NF_IP_LOCAL_IN)
+                        if (unlikely(is_confirmed(ct)))
+                                /* NAT module was loaded late */
+                                ret = alloc_null_binding_confirmed(ct, info,
+                                                                   hooknum);
+                        else if (hooknum == NF_IP_LOCAL_IN)
+                                /* LOCAL_IN hook doesn't have a chain!  */
                                 ret = alloc_null_binding(ct, info, hooknum);
                         else
                                 ret = ip_nat_rule_find(pskb, hooknum,