[PATCH][RFC] etherip: Ethernet-in-IPv4 tunneling

[Lennert Buytenhek <buytenh@wantstofly.org>]

Hi,

After struggling with various userland VPN solutions for a while (and
failing to make IPSEC tunnel mode do what I want), I decided to just
implement ethernet-in-IP tunneling in the kernel and let IPSEC transport
mode handle the rest.

There appeared to be an RFC for ethernet-in-IP already, RFC 3378, so I
just implemented that.  It's very simple -- slap a 16-bit header (0x3000,
which is 4 bits of etherip version number and 12 bits of padding) onto
the beginning of the ethernet packet, and then wrap it in an IP packet.

Below is what I came up with, against the latest Fedora Core 3 kernel,
which is 2.6.10-something.  It survives some fairly basic testing between
a number of different machines, UP and SMP.  (Corresponding iproute2
patch is available from http://www.wantstofly.org/~buytenh/etherip/ )

Notes:
- daddr=0 tunnel mode is meaningless for generic ethernet tunneling,
  so I didn't implement that.  Packets are just dropped on the floor
  if daddr==0 at the time of sending, which is the default mode for
  the etherip0 device.

Issues and TODO:
- Implement MULTICAST(daddr) tunnel mode, seems useful to have.
- Perhaps we should always present a MTU=1500 device to the user and
  deal with fragmentation issues ourselves.
- Don't take TTL of outer packet from inner packet.
- Figure out what to do with DF.
- Check whether ECN bits are correctly {en,de}capsulated.
- Check out iffy-looking '2 * sizeof(struct etheriphdr)' construct
  (same problem in ip_gre.c?)

Comments?  I would like to see this upstream when the remaining issues
have been sorted out.


cheers,
Lennert


diff -urN linux-2.6.10.orig/include/linux/in.h linux-2.6.10/include/linux/in.h
--- linux-2.6.10.orig/include/linux/in.h	2005-01-12 21:44:31.000000000 +0100
+++ linux-2.6.10/include/linux/in.h	2005-01-12 21:44:28.000000000 +0100
@@ -39,6 +39,7 @@
 
   IPPROTO_ESP = 50,            /* Encapsulation Security Payload protocol */
   IPPROTO_AH = 51,             /* Authentication Header protocol       */
+  IPPROTO_ETHERIP = 97,		/* Ethernet-in-IP tunneling (rfc 3378)	*/
   IPPROTO_PIM    = 103,		/* Protocol Independent Multicast	*/
 
   IPPROTO_COMP   = 108,                /* Compression Header protocol */
diff -urN linux-2.6.10.orig/net/ipv4/etherip.c linux-2.6.10/net/ipv4/etherip.c
--- linux-2.6.10.orig/net/ipv4/etherip.c	1970-01-01 01:00:00.000000000 +0100
+++ linux-2.6.10/net/ipv4/etherip.c	2005-01-12 22:09:51.699077165 +0100
@@ -0,0 +1,758 @@
+/*
+ *	Linux NET3:	Ethernet over IP protocol decoder. 
+ *
+ *	Authors: Alexey Kuznetsov (kuznet@ms2.inr.ac.ru)
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ *
+ */
+
+/*
+   This version of net/ipv4/etherip.c created by Lennert Buytenhek
+   by mashing net/ipv4/ip_gre.c and net/ipv4/ipip.c together.
+
+   For comments look at net/ipv4/ip_gre.c
+ */
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <asm/uaccess.h>
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <linux/in.h>
+#include <linux/tcp.h>
+#include <linux/udp.h>
+#include <linux/if_arp.h>
+#include <linux/mroute.h>
+#include <linux/init.h>
+#include <linux/in6.h>
+#include <linux/inetdevice.h>
+#include <linux/igmp.h>
+#include <linux/netfilter_ipv4.h>
+#include <linux/etherdevice.h>
+
+#include <net/sock.h>
+#include <net/ip.h>
+#include <net/icmp.h>
+#include <net/protocol.h>
+#include <net/ipip.h>
+#include <net/arp.h>
+#include <net/checksum.h>
+#include <net/dsfield.h>
+#include <net/inet_ecn.h>
+#include <net/xfrm.h>
+
+#ifdef CONFIG_IPV6
+#include <net/ipv6.h>
+#include <net/ip6_fib.h>
+#include <net/ip6_route.h>
+#endif
+
+struct etheriphdr
+{
+	struct iphdr	iph;
+	u16		version;
+} __attribute__((packed));
+
+#define HASH_SIZE  16
+#define HASH(addr) ((addr^(addr>>4))&0xF)
+
+#define ETHERIP_VERSION		0x3000
+
+static int etherip_fb_tunnel_init(struct net_device *dev);
+static int etherip_tunnel_init(struct net_device *dev);
+static void etherip_tunnel_setup(struct net_device *dev);
+
+static struct net_device *etherip_fb_tunnel_dev;
+
+static struct ip_tunnel *tunnels_r_l[HASH_SIZE];
+static struct ip_tunnel *tunnels_r[HASH_SIZE];
+static struct ip_tunnel *tunnels_l[HASH_SIZE];
+static struct ip_tunnel *tunnels_wc[1];
+
+static struct ip_tunnel **tunnels[4] = { tunnels_wc, tunnels_l, tunnels_r, tunnels_r_l };
+
+static rwlock_t etherip_lock = RW_LOCK_UNLOCKED;
+
+static struct ip_tunnel * etherip_tunnel_lookup(u32 remote, u32 local)
+{
+	unsigned h0 = HASH(remote);
+	unsigned h1 = HASH(local);
+	struct ip_tunnel *t;
+
+	for (t = tunnels_r_l[h0^h1]; t; t = t->next) {
+		if (local == t->parms.iph.saddr &&
+		    remote == t->parms.iph.daddr && (t->dev->flags&IFF_UP))
+			return t;
+	}
+	for (t = tunnels_r[h0]; t; t = t->next) {
+		if (remote == t->parms.iph.daddr && (t->dev->flags&IFF_UP))
+			return t;
+	}
+	for (t = tunnels_l[h1]; t; t = t->next) {
+		if (local == t->parms.iph.saddr && (t->dev->flags&IFF_UP))
+			return t;
+	}
+	if ((t = tunnels_wc[0]) != NULL && (t->dev->flags&IFF_UP))
+		return t;
+	return NULL;
+}
+
+static struct ip_tunnel **etherip_bucket(struct ip_tunnel *t)
+{
+	u32 remote = t->parms.iph.daddr;
+	u32 local = t->parms.iph.saddr;
+	unsigned h = 0;
+	int prio = 0;
+
+	if (remote) {
+		prio |= 2;
+		h ^= HASH(remote);
+	}
+	if (local) {
+		prio |= 1;
+		h ^= HASH(local);
+	}
+	return &tunnels[prio][h];
+}
+
+static void etherip_tunnel_unlink(struct ip_tunnel *t)
+{
+	struct ip_tunnel **tp;
+
+	for (tp = etherip_bucket(t); *tp; tp = &(*tp)->next) {
+		if (t == *tp) {
+			write_lock_bh(&etherip_lock);
+			*tp = t->next;
+			write_unlock_bh(&etherip_lock);
+			break;
+		}
+	}
+}
+
+static void etherip_tunnel_link(struct ip_tunnel *t)
+{
+	struct ip_tunnel **tp = etherip_bucket(t);
+
+	t->next = *tp;
+	write_lock_bh(&etherip_lock);
+	*tp = t;
+	write_unlock_bh(&etherip_lock);
+}
+
+static struct ip_tunnel * etherip_tunnel_locate(struct ip_tunnel_parm *parms, int create)
+{
+	u32 remote = parms->iph.daddr;
+	u32 local = parms->iph.saddr;
+	struct ip_tunnel *t, **tp, *nt;
+	struct net_device *dev;
+	unsigned h = 0;
+	int prio = 0;
+	char name[IFNAMSIZ];
+
+	if (remote) {
+		prio |= 2;
+		h ^= HASH(remote);
+	}
+	if (local) {
+		prio |= 1;
+		h ^= HASH(local);
+	}
+	for (tp = &tunnels[prio][h]; (t = *tp) != NULL; tp = &t->next) {
+		if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr)
+			return t;
+	}
+	if (!create)
+		return NULL;
+
+	if (parms->name[0])
+		strlcpy(name, parms->name, IFNAMSIZ);
+	else {
+		int i;
+		for (i=1; i<100; i++) {
+			sprintf(name, "etherip%d", i);
+			if (__dev_get_by_name(name) == NULL)
+				break;
+		}
+		if (i==100)
+			goto failed;
+	}
+
+	dev = alloc_netdev(sizeof(*t), name, etherip_tunnel_setup);
+	if (!dev)
+		return NULL;
+
+	nt = dev->priv;
+	SET_MODULE_OWNER(dev);
+	dev->init = etherip_tunnel_init;
+	nt->parms = *parms;
+
+	if (register_netdevice(dev) < 0) {
+		free_netdev(dev);
+		goto failed;
+	}
+
+	dev_hold(dev);
+	etherip_tunnel_link(nt);
+	/* Do not decrement MOD_USE_COUNT here. */
+	return nt;
+
+failed:
+	return NULL;
+}
+
+static void etherip_tunnel_uninit(struct net_device *dev)
+{
+	if (dev == etherip_fb_tunnel_dev) {
+		write_lock_bh(&etherip_lock);
+		tunnels_wc[0] = NULL;
+		write_unlock_bh(&etherip_lock);
+	} else
+		etherip_tunnel_unlink((struct ip_tunnel*)dev->priv);
+	dev_put(dev);
+}
+
+
+void etherip_err(struct sk_buff *skb, u32 info)
+{
+#ifndef I_WISH_WORLD_WERE_PERFECT
+/* It is not :-( All the routers (except for Linux) return only
+   8 bytes of packet payload. It means, that precise relaying of
+   ICMP in the real Internet is absolutely infeasible.
+ */
+
+	struct iphdr *iph = (struct iphdr*)skb->data;
+	int type = skb->h.icmph->type;
+	int code = skb->h.icmph->code;
+	struct ip_tunnel *t;
+
+	switch (type) {
+	default:
+	case ICMP_PARAMETERPROB:
+		return;
+
+	case ICMP_DEST_UNREACH:
+		switch (code) {
+		case ICMP_SR_FAILED:
+		case ICMP_PORT_UNREACH:
+			/* Impossible event. */
+			return;
+		case ICMP_FRAG_NEEDED:
+			/* Soft state for pmtu is maintained by IP core. */
+			return;
+		default:
+			/* All others are translated to HOST_UNREACH.
+			   rfc2003 contains "deep thoughts" about NET_UNREACH,
+			   I believe they are just ether pollution. --ANK
+			 */
+			break;
+		}
+		break;
+	case ICMP_TIME_EXCEEDED:
+		if (code != ICMP_EXC_TTL)
+			return;
+		break;
+	}
+
+	read_lock(&etherip_lock);
+	t = etherip_tunnel_lookup(iph->daddr, iph->saddr);
+	if (t == NULL || t->parms.iph.daddr == 0)
+		goto out;
+	if (t->parms.iph.ttl == 0 && type == ICMP_TIME_EXCEEDED)
+		goto out;
+
+	if (jiffies - t->err_time < IPTUNNEL_ERR_TIMEO)
+		t->err_count++;
+	else
+		t->err_count = 1;
+	t->err_time = jiffies;
+out:
+	read_unlock(&etherip_lock);
+	return;
+#endif
+}
+
+static inline void etherip_ecn_decapsulate(struct iphdr *iph, struct sk_buff *skb)
+{
+	if (INET_ECN_is_ce(iph->tos)) {
+		if (skb->protocol == htons(ETH_P_IP)) {
+			IP_ECN_set_ce(skb->nh.iph);
+		} else if (skb->protocol == htons(ETH_P_IPV6)) {
+			IP6_ECN_set_ce(skb->nh.ipv6h);
+		}
+	}
+}
+
+static inline u8
+etherip_ecn_encapsulate(u8 tos, struct iphdr *old_iph, struct sk_buff *skb)
+{
+	u8 inner = 0;
+	if (skb->protocol == htons(ETH_P_IP))
+		inner = old_iph->tos;
+	else if (skb->protocol == htons(ETH_P_IPV6))
+		inner = ipv6_get_dsfield((struct ipv6hdr *)old_iph);
+	return INET_ECN_encapsulate(tos, inner);
+}
+
+int etherip_rcv(struct sk_buff *skb)
+{
+	struct iphdr *iph;
+	struct ip_tunnel *tunnel;
+	struct etheriphdr *ethiph;
+
+	if (!pskb_may_pull(skb, sizeof(struct etheriphdr)))
+		goto out;
+
+	ethiph = (struct etheriphdr *)skb->nh.raw;
+	if (ethiph->version != htons(ETHERIP_VERSION)) {
+		kfree_skb(skb);
+		return 0;
+	}
+
+	iph = skb->nh.iph;
+
+	read_lock(&etherip_lock);
+	if ((tunnel = etherip_tunnel_lookup(iph->saddr, iph->daddr)) != NULL) {
+		secpath_reset(skb);
+
+		/* Pull etherip header.  */
+		skb_pull(skb, 2);
+		skb->protocol = eth_type_trans(skb, tunnel->dev);
+
+		memset(&(IPCB(skb)->opt), 0, sizeof(struct ip_options));
+
+		tunnel->stat.rx_packets++;
+		tunnel->stat.rx_bytes += skb->len;
+		skb->dev = tunnel->dev;
+		dst_release(skb->dst);
+		skb->dst = NULL;
+		nf_reset(skb);
+		etherip_ecn_decapsulate(iph, skb);
+		netif_rx(skb);
+		read_unlock(&etherip_lock);
+		return 0;
+	}
+	read_unlock(&etherip_lock);
+
+out:
+	return -1;
+}
+
+static int etherip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct ip_tunnel *tunnel = (struct ip_tunnel*)dev->priv;
+	struct net_device_stats *stats = &tunnel->stat;
+	struct iphdr *old_iph = skb->nh.iph;
+	struct iphdr *tiph = &tunnel->parms.iph;
+	u8     tos;
+	u16    df;
+	struct rtable *rt;     		/* Route to the other host */
+	struct net_device *tdev;	/* Device to other host */
+	struct etheriphdr  *ethiph;
+	struct iphdr  *iph;		/* Our new IP header */
+	int    max_headroom;		/* The extra header space needed */
+	int    mtu;
+
+	if (tunnel->recursion++) {
+		stats->collisions++;
+		goto tx_error;
+	}
+
+	/* Need valid non-multicast daddr.  */
+	if (tiph->daddr == 0 || MULTICAST(tiph->daddr))
+		goto tx_error;
+
+	tos = tiph->tos;
+	if (tos&1) {
+		if (skb->protocol == htons(ETH_P_IP))
+			tos = old_iph->tos;
+		tos &= ~1;
+	}
+
+	{
+		struct flowi fl = { .oif = tunnel->parms.link,
+				    .nl_u = { .ip4_u =
+					      { .daddr = tiph->daddr,
+						.saddr = tiph->saddr,
+						.tos = RT_TOS(tos) } },
+				    .proto = IPPROTO_ETHERIP };
+		if (ip_route_output_key(&rt, &fl)) {
+			stats->tx_carrier_errors++;
+			goto tx_error_icmp;
+		}
+	}
+	tdev = rt->u.dst.dev;
+
+	if (tdev == dev) {
+		ip_rt_put(rt);
+		stats->collisions++;
+		goto tx_error;
+	}
+
+	df = tiph->frag_off;
+	if (df)
+		mtu = dst_pmtu(&rt->u.dst) - sizeof(struct etheriphdr);
+	else
+		mtu = skb->dst ? dst_pmtu(skb->dst) : dev->mtu;
+
+	if (skb->dst)
+		skb->dst->ops->update_pmtu(skb->dst, mtu);
+
+	if (skb->protocol == htons(ETH_P_IP)) {
+		df |= (old_iph->frag_off&htons(IP_DF));
+
+		if ((old_iph->frag_off & htons(IP_DF)) &&
+		    mtu < ntohs(old_iph->tot_len)) {
+			icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
+			ip_rt_put(rt);
+			goto tx_error;
+		}
+	}
+#ifdef CONFIG_IPV6
+	else if (skb->protocol == htons(ETH_P_IPV6)) {
+		struct rt6_info *rt6 = (struct rt6_info*)skb->dst;
+
+		if (rt6 && mtu < dst_pmtu(skb->dst) && mtu >= IPV6_MIN_MTU) {
+			if (tiph->daddr || rt6->rt6i_dst.plen == 128) {
+				rt6->rt6i_flags |= RTF_MODIFIED;
+				skb->dst->metrics[RTAX_MTU-1] = mtu;
+			}
+		}
+
+		/* @@@ Is this correct?  */
+		if (mtu >= IPV6_MIN_MTU && mtu < skb->len - 2 * sizeof(struct etheriphdr)) {
+			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, dev);
+			ip_rt_put(rt);
+			goto tx_error;
+		}
+	}
+#endif
+
+	if (tunnel->err_count > 0) {
+		if (jiffies - tunnel->err_time < IPTUNNEL_ERR_TIMEO) {
+			tunnel->err_count--;
+			dst_link_failure(skb);
+		} else
+			tunnel->err_count = 0;
+	}
+
+	max_headroom = LL_RESERVED_SPACE(tdev) + sizeof(struct etheriphdr);
+
+	if (skb_headroom(skb) < max_headroom || skb_cloned(skb) || skb_shared(skb)) {
+		struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom);
+		if (!new_skb) {
+			ip_rt_put(rt);
+  			stats->tx_dropped++;
+			dev_kfree_skb(skb);
+			tunnel->recursion--;
+			return 0;
+		}
+		if (skb->sk)
+			skb_set_owner_w(new_skb, skb->sk);
+		dev_kfree_skb(skb);
+		skb = new_skb;
+		old_iph = skb->nh.iph;
+	}
+
+	skb->h.raw = skb->nh.raw;
+	skb->nh.raw = skb_push(skb, sizeof(struct etheriphdr));
+	memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
+	dst_release(skb->dst);
+	skb->dst = &rt->u.dst;
+
+	/*
+	 *	Push down and install the etherip header.
+	 */
+
+	ethiph 			=	(struct etheriphdr *)skb->nh.iph;
+
+	iph			=	&ethiph->iph;
+	iph->version		=	4;
+	iph->ihl		=	sizeof(struct iphdr) >> 2;
+	iph->frag_off		=	df;
+	iph->protocol		=	IPPROTO_ETHERIP;
+	iph->tos		=	etherip_ecn_encapsulate(tos, old_iph, skb);
+	iph->daddr		=	rt->rt_dst;
+	iph->saddr		=	rt->rt_src;
+
+	ethiph->version		=	htons(ETHERIP_VERSION);
+
+	if ((ethiph->iph.ttl = tiph->ttl) == 0) {
+		if (skb->protocol == htons(ETH_P_IP))
+			ethiph->iph.ttl = old_iph->ttl;
+#ifdef CONFIG_IPV6
+		else if (skb->protocol == htons(ETH_P_IPV6))
+			ethiph->iph.ttl = ((struct ipv6hdr*)old_iph)->hop_limit;
+#endif
+		else
+			ethiph->iph.ttl = dst_metric(&rt->u.dst, RTAX_HOPLIMIT);
+	}
+
+	nf_reset(skb);
+
+	IPTUNNEL_XMIT();
+	tunnel->recursion--;
+	return 0;
+
+tx_error_icmp:
+	dst_link_failure(skb);
+
+tx_error:
+	stats->tx_errors++;
+	dev_kfree_skb(skb);
+	tunnel->recursion--;
+	return 0;
+}
+
+static int
+etherip_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
+{
+	int err = 0;
+	struct ip_tunnel_parm p;
+	struct ip_tunnel *t;
+
+	switch (cmd) {
+	case SIOCGETTUNNEL:
+		t = NULL;
+		if (dev == etherip_fb_tunnel_dev) {
+			if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p))) {
+				err = -EFAULT;
+				break;
+			}
+			t = etherip_tunnel_locate(&p, 0);
+		}
+		if (t == NULL)
+			t = (struct ip_tunnel*)dev->priv;
+		memcpy(&p, &t->parms, sizeof(p));
+		if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p)))
+			err = -EFAULT;
+		break;
+
+	case SIOCADDTUNNEL:
+	case SIOCCHGTUNNEL:
+		err = -EPERM;
+		if (!capable(CAP_NET_ADMIN))
+			goto done;
+
+		err = -EFAULT;
+		if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
+			goto done;
+
+		err = -EINVAL;
+		if (p.iph.version != 4 || p.iph.protocol != IPPROTO_ETHERIP ||
+		    p.iph.ihl != 5 || (p.iph.frag_off&htons(~IP_DF)))
+			goto done;
+		if (p.iph.ttl)
+			p.iph.frag_off |= htons(IP_DF);
+
+		t = etherip_tunnel_locate(&p, cmd == SIOCADDTUNNEL);
+
+		if (dev != etherip_fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
+			if (t != NULL) {
+				if (t->dev != dev) {
+					err = -EEXIST;
+					break;
+				}
+			} else {
+				if (!p.iph.daddr) {
+					err = -EINVAL;
+					break;
+				}
+
+				t = (struct ip_tunnel*)dev->priv;
+				etherip_tunnel_unlink(t);
+				t->parms.iph.saddr = p.iph.saddr;
+				t->parms.iph.daddr = p.iph.daddr;
+				etherip_tunnel_link(t);
+				netdev_state_change(dev);
+			}
+		}
+
+		if (t) {
+			err = 0;
+			if (cmd == SIOCCHGTUNNEL) {
+				t->parms.iph.ttl = p.iph.ttl;
+				t->parms.iph.tos = p.iph.tos;
+				t->parms.iph.frag_off = p.iph.frag_off;
+			}
+			if (copy_to_user(ifr->ifr_ifru.ifru_data, &t->parms, sizeof(p)))
+				err = -EFAULT;
+		} else
+			err = (cmd == SIOCADDTUNNEL ? -ENOBUFS : -ENOENT);
+		break;
+
+	case SIOCDELTUNNEL:
+		err = -EPERM;
+		if (!capable(CAP_NET_ADMIN))
+			goto done;
+
+		if (dev == etherip_fb_tunnel_dev) {
+			err = -EFAULT;
+			if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
+				goto done;
+			err = -ENOENT;
+			if ((t = etherip_tunnel_locate(&p, 0)) == NULL)
+				goto done;
+			err = -EPERM;
+			if (t->dev == etherip_fb_tunnel_dev)
+				goto done;
+			dev = t->dev;
+		}
+		err = unregister_netdevice(dev);
+		break;
+
+	default:
+		err = -EINVAL;
+	}
+
+done:
+	return err;
+}
+
+static struct net_device_stats *etherip_tunnel_get_stats(struct net_device *dev)
+{
+	return &(((struct ip_tunnel*)dev->priv)->stat);
+}
+
+static int etherip_tunnel_change_mtu(struct net_device *dev, int new_mtu)
+{
+	if (new_mtu < 68 || new_mtu > 0xFFF8 - sizeof(struct etheriphdr))
+		return -EINVAL;
+	dev->mtu = new_mtu;
+	return 0;
+}
+
+static void etherip_tunnel_setup(struct net_device *dev)
+{
+	SET_MODULE_OWNER(dev);
+	ether_setup(dev);
+
+	dev->uninit		= etherip_tunnel_uninit;
+	dev->destructor 	= free_netdev;
+	dev->hard_start_xmit	= etherip_tunnel_xmit;
+	dev->get_stats		= etherip_tunnel_get_stats;
+	dev->do_ioctl		= etherip_tunnel_ioctl;
+	dev->change_mtu		= etherip_tunnel_change_mtu;
+
+	dev->hard_header_len	= ETH_HLEN;	//  + sizeof(struct etheriphdr);
+	dev->tx_queue_len	= 0;
+	random_ether_addr(dev->dev_addr);
+
+	dev->iflink		= 0;
+}
+
+static int etherip_tunnel_init(struct net_device *dev)
+{
+	struct net_device *tdev = NULL;
+	struct ip_tunnel *tunnel;
+	struct iphdr *iph;
+
+	tunnel = (struct ip_tunnel*)dev->priv;
+	iph = &tunnel->parms.iph;
+
+	tunnel->dev = dev;
+	strcpy(tunnel->parms.name, dev->name);
+
+	/* Guess output device to choose reasonable mtu and hard_header_len */
+	if (iph->daddr) {
+		struct flowi fl = { .oif = tunnel->parms.link,
+				    .nl_u = { .ip4_u =
+					      { .daddr = iph->daddr,
+						.saddr = iph->saddr,
+						.tos = RT_TOS(iph->tos) } },
+				    .proto = IPPROTO_ETHERIP };
+		struct rtable *rt;
+		if (!ip_route_output_key(&rt, &fl)) {
+			tdev = rt->u.dst.dev;
+			ip_rt_put(rt);
+		}
+	}
+
+	if (!tdev && tunnel->parms.link)
+		tdev = __dev_get_by_index(tunnel->parms.link);
+
+	if (tdev) {
+		dev->hard_header_len = tdev->hard_header_len + sizeof(struct etheriphdr);
+		dev->mtu = tdev->mtu - sizeof(struct etheriphdr);
+	}
+	dev->iflink = tunnel->parms.link;
+
+	return 0;
+}
+
+int __init etherip_fb_tunnel_init(struct net_device *dev)
+{
+	struct ip_tunnel *tunnel = (struct ip_tunnel*)dev->priv;
+	struct iphdr *iph = &tunnel->parms.iph;
+
+	tunnel->dev = dev;
+	strcpy(tunnel->parms.name, dev->name);
+
+	iph->version		= 4;
+	iph->protocol		= IPPROTO_ETHERIP;
+	iph->ihl		= 5;
+
+	dev_hold(dev);
+	tunnels_wc[0]		= tunnel;
+	return 0;
+}
+
+
+static struct net_protocol etherip_protocol = {
+	.handler	=	etherip_rcv,
+	.err_handler	=	etherip_err,
+};
+
+
+/*
+ *	And now the modules code and kernel interface.
+ */
+
+static int __init etherip_init(void)
+{
+	int err;
+
+	printk(KERN_INFO "Ethernet over IPv4 tunneling driver\n");
+
+	if (inet_add_protocol(&etherip_protocol, IPPROTO_ETHERIP) < 0) {
+		printk(KERN_INFO "etherip init: can't add protocol\n");
+		return -EAGAIN;
+	}
+
+	etherip_fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel),
+					     "etherip0", etherip_tunnel_setup);
+	if (!etherip_fb_tunnel_dev) {
+		err = -ENOMEM;
+		goto err1;
+	}
+
+	etherip_fb_tunnel_dev->init = etherip_fb_tunnel_init;
+
+	if ((err = register_netdev(etherip_fb_tunnel_dev)))
+		goto err2;
+out:
+	return err;
+err2:
+	free_netdev(etherip_fb_tunnel_dev);
+err1:
+	inet_del_protocol(&etherip_protocol, IPPROTO_ETHERIP);
+	goto out;
+}
+
+void etherip_fini(void)
+{
+	if (inet_del_protocol(&etherip_protocol, IPPROTO_ETHERIP) < 0)
+		printk(KERN_INFO "etherip close: can't remove protocol\n");
+
+	unregister_netdev(etherip_fb_tunnel_dev);
+}
+
+module_init(etherip_init);
+module_exit(etherip_fini);
+MODULE_LICENSE("GPL");
diff -urN linux-2.6.10.orig/net/ipv4/Kconfig linux-2.6.10/net/ipv4/Kconfig
--- linux-2.6.10.orig/net/ipv4/Kconfig	2005-01-12 21:48:11.000000000 +0100
+++ linux-2.6.10/net/ipv4/Kconfig	2005-01-12 21:53:51.063081266 +0100
@@ -202,6 +202,15 @@
 	  Network), but can be distributed all over the Internet. If you want
 	  to do that, say Y here and to "IP multicast routing" below.
 
+config NET_ETHERIP
+	tristate "IP: ethernet-in-IP tunneling"
+	depends on INET
+	help
+	  Tunneling means encapsulating data of one protocol type within
+	  another protocol and sending it over a channel that understands the
+	  encapsulating protocol. This particular tunneling driver implements
+	  the etherip Ethernet-in-IP protocol as described in RFC 3378.
+
 config IP_MROUTE
 	bool "IP: multicast routing"
 	depends on IP_MULTICAST
diff -urN linux-2.6.10.orig/net/ipv4/Makefile linux-2.6.10/net/ipv4/Makefile
--- linux-2.6.10.orig/net/ipv4/Makefile	2004-12-24 22:34:26.000000000 +0100
+++ linux-2.6.10/net/ipv4/Makefile	2005-01-12 21:48:38.762364033 +0100
@@ -14,6 +14,7 @@
 obj-$(CONFIG_IP_MROUTE) += ipmr.o
 obj-$(CONFIG_NET_IPIP) += ipip.o
 obj-$(CONFIG_NET_IPGRE) += ip_gre.o
+obj-$(CONFIG_NET_ETHERIP) += etherip.o
 obj-$(CONFIG_SYN_COOKIES) += syncookies.o
 obj-$(CONFIG_INET_AH) += ah4.o
 obj-$(CONFIG_INET_ESP) += esp4.o