Merge branch 'master' of /repos/git/net-next-2.6
[pandora-kernel.git] / net / bridge / br_netfilter.c
index 4c4977d..6b80ebc 100644 (file)
@@ -3,15 +3,8 @@
  *     Linux ethernet bridge
  *
  *     Authors:
- *     Lennert Buytenhek               <buytenh@gnu.org>
- *     Bart De Schuymer (maintainer)   <bdschuym@pandora.be>
- *
- *     Changes:
- *     Apr 29 2003: physdev module support (bdschuym)
- *     Jun 19 2003: let arptables see bridged ARP traffic (bdschuym)
- *     Oct 06 2003: filter encapsulated IP/ARP VLAN traffic on untagged bridge
- *                  (bdschuym)
- *     Sep 01 2004: add IPv6 filtering (bdschuym)
+ *     Lennert Buytenhek               <buytenh@gnu.org>
+ *     Bart De Schuymer                <bdschuym@pandora.be>
  *
  *     This program is free software; you can redistribute it and/or
  *     modify it under the terms of the GNU General Public License
@@ -204,15 +197,24 @@ static inline void nf_bridge_save_header(struct sk_buff *skb)
                                         skb->nf_bridge->data, header_size);
 }
 
-/*
- * When forwarding bridge frames, we save a copy of the original
- * header before processing.
+static inline void nf_bridge_update_protocol(struct sk_buff *skb)
+{
+       if (skb->nf_bridge->mask & BRNF_8021Q)
+               skb->protocol = htons(ETH_P_8021Q);
+       else if (skb->nf_bridge->mask & BRNF_PPPoE)
+               skb->protocol = htons(ETH_P_PPP_SES);
+}
+
+/* Fill in the header for fragmented IP packets handled by
+ * the IPv4 connection tracking code.
  */
 int nf_bridge_copy_header(struct sk_buff *skb)
 {
        int err;
-       int header_size = ETH_HLEN + nf_bridge_encap_header_len(skb);
+       unsigned int header_size;
 
+       nf_bridge_update_protocol(skb);
+       header_size = ETH_HLEN + nf_bridge_encap_header_len(skb);
        err = skb_cow_head(skb, header_size);
        if (err)
                return err;
@@ -246,27 +248,48 @@ static int br_nf_pre_routing_finish_ipv6(struct sk_buff *skb)
        skb_dst_set(skb, &rt->u.dst);
 
        skb->dev = nf_bridge->physindev;
+       nf_bridge_update_protocol(skb);
        nf_bridge_push_encap_header(skb);
-       NF_HOOK_THRESH(PF_BRIDGE, NF_BR_PRE_ROUTING, skb, skb->dev, NULL,
+       NF_HOOK_THRESH(NFPROTO_BRIDGE, NF_BR_PRE_ROUTING, skb, skb->dev, NULL,
                       br_handle_frame_finish, 1);
 
        return 0;
 }
 
-static void __br_dnat_complain(void)
+/* Obtain the correct destination MAC address, while preserving the original
+ * source MAC address. If we already know this address, we just copy it. If we
+ * don't, we use the neighbour framework to find out. In both cases, we make
+ * sure that br_handle_frame_finish() is called afterwards.
+ */
+static int br_nf_pre_routing_finish_bridge(struct sk_buff *skb)
 {
-       static unsigned long last_complaint;
+       struct nf_bridge_info *nf_bridge = skb->nf_bridge;
+       struct dst_entry *dst;
 
-       if (jiffies - last_complaint >= 5 * HZ) {
-               printk(KERN_WARNING "Performing cross-bridge DNAT requires IP "
-                      "forwarding to be enabled\n");
-               last_complaint = jiffies;
+       skb->dev = bridge_parent(skb->dev);
+       if (!skb->dev)
+               goto free_skb;
+       dst = skb_dst(skb);
+       if (dst->hh) {
+               neigh_hh_bridge(dst->hh, skb);
+               skb->dev = nf_bridge->physindev;
+               return br_handle_frame_finish(skb);
+       } else if (dst->neighbour) {
+               /* the neighbour function below overwrites the complete
+                * MAC header, so we save the Ethernet source address and
+                * protocol number. */
+               skb_copy_from_linear_data_offset(skb, -(ETH_HLEN-ETH_ALEN), skb->nf_bridge->data, ETH_HLEN-ETH_ALEN);
+               /* tell br_dev_xmit to continue with forwarding */
+               nf_bridge->mask |= BRNF_BRIDGED_DNAT;
+               return dst->neighbour->output(skb);
        }
+free_skb:
+       kfree_skb(skb);
+       return 0;
 }
 
 /* This requires some explaining. If DNAT has taken place,
- * we will need to fix up the destination Ethernet address,
- * and this is a tricky process.
+ * we will need to fix up the destination Ethernet address.
  *
  * There are two cases to consider:
  * 1. The packet was DNAT'ed to a device in the same bridge
@@ -280,62 +303,29 @@ static void __br_dnat_complain(void)
  * call ip_route_input() and to look at skb->dst->dev, which is
  * changed to the destination device if ip_route_input() succeeds.
  *
- * Let us first consider the case that ip_route_input() succeeds:
- *
- * If skb->dst->dev equals the logical bridge device the packet
- * came in on, we can consider this bridging. The packet is passed
- * through the neighbour output function to build a new destination
- * MAC address, which will make the packet enter br_nf_local_out()
- * not much later. In that function it is assured that the iptables
- * FORWARD chain is traversed for the packet.
+ * Let's first consider the case that ip_route_input() succeeds:
  *
+ * If the output device equals the logical bridge device the packet
+ * came in on, we can consider this bridging. The corresponding MAC
+ * address will be obtained in br_nf_pre_routing_finish_bridge.
  * Otherwise, the packet is considered to be routed and we just
  * change the destination MAC address so that the packet will
  * later be passed up to the IP stack to be routed. For a redirected
  * packet, ip_route_input() will give back the localhost as output device,
  * which differs from the bridge device.
  *
- * Let us now consider the case that ip_route_input() fails:
+ * Let's now consider the case that ip_route_input() fails:
  *
  * This can be because the destination address is martian, in which case
  * the packet will be dropped.
- * After a "echo '0' > /proc/sys/net/ipv4/ip_forward" ip_route_input()
- * will fail, while __ip_route_output_key() will return success. The source
- * address for __ip_route_output_key() is set to zero, so __ip_route_output_key
+ * If IP forwarding is disabled, ip_route_input() will fail, while
+ * ip_route_output_key() can return success. The source
+ * address for ip_route_output_key() is set to zero, so ip_route_output_key()
  * thinks we're handling a locally generated packet and won't care
- * if IP forwarding is allowed. We send a warning message to the users's
- * log telling her to put IP forwarding on.
- *
- * ip_route_input() will also fail if there is no route available.
- * In that case we just drop the packet.
- *
- * --Lennert, 20020411
- * --Bart, 20020416 (updated)
- * --Bart, 20021007 (updated)
- * --Bart, 20062711 (updated) */
-static int br_nf_pre_routing_finish_bridge(struct sk_buff *skb)
-{
-       if (skb->pkt_type == PACKET_OTHERHOST) {
-               skb->pkt_type = PACKET_HOST;
-               skb->nf_bridge->mask |= BRNF_PKT_TYPE;
-       }
-       skb->nf_bridge->mask ^= BRNF_NF_BRIDGE_PREROUTING;
-
-       skb->dev = bridge_parent(skb->dev);
-       if (skb->dev) {
-               struct dst_entry *dst = skb_dst(skb);
-
-               nf_bridge_pull_encap_header(skb);
-
-               if (dst->hh)
-                       return neigh_hh_output(dst->hh, skb);
-               else if (dst->neighbour)
-                       return dst->neighbour->output(skb);
-       }
-       kfree_skb(skb);
-       return 0;
-}
-
+ * if IP forwarding is enabled. If the output device equals the logical bridge
+ * device, we proceed as if ip_route_input() succeeded. If it differs from the
+ * logical bridge port or if ip_route_output_key() fails we drop the packet.
+ */
 static int br_nf_pre_routing_finish(struct sk_buff *skb)
 {
        struct net_device *dev = skb->dev;
@@ -379,11 +369,6 @@ static int br_nf_pre_routing_finish(struct sk_buff *skb)
                                        skb_dst_set(skb, (struct dst_entry *)rt);
                                        goto bridged_dnat;
                                }
-                               /* we are sure that forwarding is disabled, so printing
-                                * this message is no problem. Note that the packet could
-                                * still have a martian destination address, in which case
-                                * the packet could be dropped even if forwarding were enabled */
-                               __br_dnat_complain();
                                dst_release((struct dst_entry *)rt);
                        }
 free_skb:
@@ -392,12 +377,11 @@ free_skb:
                } else {
                        if (skb_dst(skb)->dev == dev) {
 bridged_dnat:
-                               /* Tell br_nf_local_out this is a
-                                * bridged frame */
-                               nf_bridge->mask |= BRNF_BRIDGED_DNAT;
                                skb->dev = nf_bridge->physindev;
+                               nf_bridge_update_protocol(skb);
                                nf_bridge_push_encap_header(skb);
-                               NF_HOOK_THRESH(PF_BRIDGE, NF_BR_PRE_ROUTING,
+                               NF_HOOK_THRESH(NFPROTO_BRIDGE,
+                                              NF_BR_PRE_ROUTING,
                                               skb, skb->dev, NULL,
                                               br_nf_pre_routing_finish_bridge,
                                               1);
@@ -417,8 +401,9 @@ bridged_dnat:
        }
 
        skb->dev = nf_bridge->physindev;
+       nf_bridge_update_protocol(skb);
        nf_bridge_push_encap_header(skb);
-       NF_HOOK_THRESH(PF_BRIDGE, NF_BR_PRE_ROUTING, skb, skb->dev, NULL,
+       NF_HOOK_THRESH(NFPROTO_BRIDGE, NF_BR_PRE_ROUTING, skb, skb->dev, NULL,
                       br_handle_frame_finish, 1);
 
        return 0;
@@ -437,6 +422,10 @@ static struct net_device *setup_pre_routing(struct sk_buff *skb)
        nf_bridge->mask |= BRNF_NF_BRIDGE_PREROUTING;
        nf_bridge->physindev = skb->dev;
        skb->dev = bridge_parent(skb->dev);
+       if (skb->protocol == htons(ETH_P_8021Q))
+               nf_bridge->mask |= BRNF_8021Q;
+       else if (skb->protocol == htons(ETH_P_PPP_SES))
+               nf_bridge->mask |= BRNF_PPPoE;
 
        return skb->dev;
 }
@@ -535,7 +524,8 @@ static unsigned int br_nf_pre_routing_ipv6(unsigned int hook,
        if (!setup_pre_routing(skb))
                return NF_DROP;
 
-       NF_HOOK(PF_INET6, NF_INET_PRE_ROUTING, skb, skb->dev, NULL,
+       skb->protocol = htons(ETH_P_IPV6);
+       NF_HOOK(NFPROTO_IPV6, NF_INET_PRE_ROUTING, skb, skb->dev, NULL,
                br_nf_pre_routing_finish_ipv6);
 
        return NF_STOLEN;
@@ -607,8 +597,9 @@ static unsigned int br_nf_pre_routing(unsigned int hook, struct sk_buff *skb,
        if (!setup_pre_routing(skb))
                return NF_DROP;
        store_orig_dstaddr(skb);
+       skb->protocol = htons(ETH_P_IP);
 
-       NF_HOOK(PF_INET, NF_INET_PRE_ROUTING, skb, skb->dev, NULL,
+       NF_HOOK(NFPROTO_IPV4, NF_INET_PRE_ROUTING, skb, skb->dev, NULL,
                br_nf_pre_routing_finish);
 
        return NF_STOLEN;
@@ -655,8 +646,10 @@ static int br_nf_forward_finish(struct sk_buff *skb)
        } else {
                in = *((struct net_device **)(skb->cb));
        }
+       nf_bridge_update_protocol(skb);
        nf_bridge_push_encap_header(skb);
-       NF_HOOK_THRESH(PF_BRIDGE, NF_BR_FORWARD, skb, in,
+
+       NF_HOOK_THRESH(NFPROTO_BRIDGE, NF_BR_FORWARD, skb, in,
                       skb->dev, br_forward_finish, 1);
        return 0;
 }
@@ -707,6 +700,10 @@ static unsigned int br_nf_forward_ip(unsigned int hook, struct sk_buff *skb,
        /* The physdev module checks on this */
        nf_bridge->mask |= BRNF_BRIDGED;
        nf_bridge->physoutdev = skb->dev;
+       if (pf == PF_INET)
+               skb->protocol = htons(ETH_P_IP);
+       else
+               skb->protocol = htons(ETH_P_IPV6);
 
        NF_HOOK(pf, NF_INET_FORWARD, skb, bridge_parent(in), parent,
                br_nf_forward_finish);
@@ -744,59 +741,10 @@ static unsigned int br_nf_forward_arp(unsigned int hook, struct sk_buff *skb,
        return NF_STOLEN;
 }
 
-/* PF_BRIDGE/LOCAL_OUT ***********************************************
- *
- * This function sees both locally originated IP packets and forwarded
- * IP packets (in both cases the destination device is a bridge
- * device). It also sees bridged-and-DNAT'ed packets.
- *
- * If (nf_bridge->mask & BRNF_BRIDGED_DNAT) then the packet is bridged
- * and we fake the PF_BRIDGE/FORWARD hook. The function br_nf_forward()
- * will then fake the PF_INET/FORWARD hook. br_nf_local_out() has priority
- * NF_BR_PRI_FIRST, so no relevant PF_BRIDGE/INPUT functions have been nor
- * will be executed.
- */
-static unsigned int br_nf_local_out(unsigned int hook, struct sk_buff *skb,
-                                   const struct net_device *in,
-                                   const struct net_device *out,
-                                   int (*okfn)(struct sk_buff *))
-{
-       struct net_device *realindev;
-       struct nf_bridge_info *nf_bridge;
-
-       if (!skb->nf_bridge)
-               return NF_ACCEPT;
-
-       /* Need exclusive nf_bridge_info since we might have multiple
-        * different physoutdevs. */
-       if (!nf_bridge_unshare(skb))
-               return NF_DROP;
-
-       nf_bridge = skb->nf_bridge;
-       if (!(nf_bridge->mask & BRNF_BRIDGED_DNAT))
-               return NF_ACCEPT;
-
-       /* Bridged, take PF_BRIDGE/FORWARD.
-        * (see big note in front of br_nf_pre_routing_finish) */
-       nf_bridge->physoutdev = skb->dev;
-       realindev = nf_bridge->physindev;
-
-       if (nf_bridge->mask & BRNF_PKT_TYPE) {
-               skb->pkt_type = PACKET_OTHERHOST;
-               nf_bridge->mask ^= BRNF_PKT_TYPE;
-       }
-       nf_bridge_push_encap_header(skb);
-
-       NF_HOOK(PF_BRIDGE, NF_BR_FORWARD, skb, realindev, skb->dev,
-               br_forward_finish);
-       return NF_STOLEN;
-}
-
 #if defined(CONFIG_NF_CONNTRACK_IPV4) || defined(CONFIG_NF_CONNTRACK_IPV4_MODULE)
 static int br_nf_dev_queue_xmit(struct sk_buff *skb)
 {
-       if (skb->nfct != NULL &&
-           (skb->protocol == htons(ETH_P_IP) || IS_VLAN_IP(skb)) &&
+       if (skb->nfct != NULL && skb->protocol == htons(ETH_P_IP) &&
            skb->len > skb->dev->mtu &&
            !skb_is_gso(skb))
                return ip_fragment(skb, br_dev_queue_push_xmit);
@@ -820,21 +768,7 @@ static unsigned int br_nf_post_routing(unsigned int hook, struct sk_buff *skb,
        struct net_device *realoutdev = bridge_parent(skb->dev);
        u_int8_t pf;
 
-#ifdef CONFIG_NETFILTER_DEBUG
-       /* Be very paranoid. This probably won't happen anymore, but let's
-        * keep the check just to be sure... */
-       if (skb_mac_header(skb) < skb->head ||
-           skb_mac_header(skb) + ETH_HLEN > skb->data) {
-               printk(KERN_CRIT "br_netfilter: Argh!! br_nf_post_routing: "
-                      "bad mac.raw pointer.\n");
-               goto print_error;
-       }
-#endif
-
-       if (!nf_bridge)
-               return NF_ACCEPT;
-
-       if (!(nf_bridge->mask & (BRNF_BRIDGED | BRNF_BRIDGED_DNAT)))
+       if (!nf_bridge || !(nf_bridge->mask & BRNF_BRIDGED))
                return NF_ACCEPT;
 
        if (!realoutdev)
@@ -849,13 +783,6 @@ static unsigned int br_nf_post_routing(unsigned int hook, struct sk_buff *skb,
        else
                return NF_ACCEPT;
 
-#ifdef CONFIG_NETFILTER_DEBUG
-       if (skb_dst(skb) == NULL) {
-               printk(KERN_INFO "br_netfilter post_routing: skb->dst == NULL\n");
-               goto print_error;
-       }
-#endif
-
        /* We assume any code from br_dev_queue_push_xmit onwards doesn't care
         * about the value of skb->pkt_type. */
        if (skb->pkt_type == PACKET_OTHERHOST) {
@@ -865,24 +792,15 @@ static unsigned int br_nf_post_routing(unsigned int hook, struct sk_buff *skb,
 
        nf_bridge_pull_encap_header(skb);
        nf_bridge_save_header(skb);
+       if (pf == PF_INET)
+               skb->protocol = htons(ETH_P_IP);
+       else
+               skb->protocol = htons(ETH_P_IPV6);
 
        NF_HOOK(pf, NF_INET_POST_ROUTING, skb, NULL, realoutdev,
                br_nf_dev_queue_xmit);
 
        return NF_STOLEN;
-
-#ifdef CONFIG_NETFILTER_DEBUG
-print_error:
-       if (skb->dev != NULL) {
-               printk("[%s]", skb->dev->name);
-               if (realoutdev)
-                       printk("[%s]", realoutdev->name);
-       }
-       printk(" head:%p, raw:%p, data:%p\n", skb->head, skb_mac_header(skb),
-              skb->data);
-       dump_stack();
-       return NF_ACCEPT;
-#endif
 }
 
 /* IP/SABOTAGE *****************************************************/
@@ -901,10 +819,8 @@ static unsigned int ip_sabotage_in(unsigned int hook, struct sk_buff *skb,
        return NF_ACCEPT;
 }
 
-/* For br_nf_local_out we need (prio = NF_BR_PRI_FIRST), to insure that innocent
- * PF_BRIDGE/NF_BR_LOCAL_OUT functions don't get bridged traffic as input.
- * For br_nf_post_routing, we need (prio = NF_BR_PRI_LAST), because
- * ip_refrag() can return NF_STOLEN. */
+/* For br_nf_post_routing, we need (prio = NF_BR_PRI_LAST), because
+ * br_dev_queue_push_xmit is called afterwards */
 static struct nf_hook_ops br_nf_ops[] __read_mostly = {
        {
                .hook = br_nf_pre_routing,
@@ -934,13 +850,6 @@ static struct nf_hook_ops br_nf_ops[] __read_mostly = {
                .hooknum = NF_BR_FORWARD,
                .priority = NF_BR_PRI_BRNF,
        },
-       {
-               .hook = br_nf_local_out,
-               .owner = THIS_MODULE,
-               .pf = PF_BRIDGE,
-               .hooknum = NF_BR_LOCAL_OUT,
-               .priority = NF_BR_PRI_FIRST,
-       },
        {
                .hook = br_nf_post_routing,
                .owner = THIS_MODULE,