[INET]: Introduce tunnel4/tunnel6

Basically this patch moves the generic tunnel protocol stuff out of
xfrm4_tunnel/xfrm6_tunnel and moves it into the new files of tunnel4.c
and tunnel6 respectively.

The reason for this is that the problem that Hugo uncovered is only
the tip of the iceberg.  The real problem is that when we removed the
dependency of ipip on xfrm4_tunnel we didn't really consider the module
case at all.

For instance, as it is it's possible to build both ipip and xfrm4_tunnel
as modules and if the latter is loaded then ipip simply won't load.

After considering the alternatives I've decided that the best way out of
this is to restore the dependency of ipip on the non-xfrm-specific part
of xfrm4_tunnel.  This is acceptable IMHO because the intention of the
removal was really to be able to use ipip without the xfrm subsystem.
This is still preserved by this patch.

So now both ipip/xfrm4_tunnel depend on the new tunnel4.c which handles
the arbitration between the two.  The order of processing is determined
by a simple integer which ensures that ipip gets processed before
xfrm4_tunnel.

The situation for ICMP handling is a little bit more complicated since
we may not have enough information to determine who it's for.  It's not
a big deal at the moment since the xfrm ICMP handlers are basically
no-ops.  In future we can deal with this when we look at ICMP caching
in general.

The user-visible change to this is the removal of the TUNNEL Kconfig
prompts.  This makes sense because it can only be used through IPCOMP
as it stands.

The addition of the new modules shouldn't introduce any problems since
module dependency will cause them to be loaded.

Oh and I also turned some unnecessary pskb's in IPv6 related to this
patch to skb's.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 165 insertions, 126 deletions

diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index 011cca7ae02..e40f7532237 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -235,6 +235,7 @@ config IP_PNP_RARP
 #   bool '    IP: ARP support' CONFIG_IP_PNP_ARP		
 config NET_IPIP
 	tristate "IP: tunneling"
+	select INET_TUNNEL
 	---help---
 	  Tunneling means encapsulating data of one protocol type within
 	  another protocol and sending it over a channel that understands the
@@ -395,7 +396,7 @@ config INET_ESP
 config INET_IPCOMP
 	tristate "IP: IPComp transformation"
 	select XFRM
-	select INET_TUNNEL
+	select INET_XFRM_TUNNEL
 	select CRYPTO
 	select CRYPTO_DEFLATE
 	---help---
@@ -404,14 +405,14 @@ config INET_IPCOMP
 	  
 	  If unsure, say Y.
 
+config INET_XFRM_TUNNEL
+	tristate
+	select INET_TUNNEL
+	default n
+
 config INET_TUNNEL
-	tristate "IP: tunnel transformation"
-	select XFRM
-	---help---
-	  Support for generic IP tunnel transformation, which is required by
-	  the IP tunneling module as well as tunnel mode IPComp.
-	  
-	  If unsure, say Y.
+	tristate
+	default n
 
 config INET_DIAG
 	tristate "INET: socket monitoring interface"
diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile
index 35e5f599909..9ef50a0b9d2 100644
--- a/net/ipv4/Makefile
+++ b/net/ipv4/Makefile
@@ -22,7 +22,8 @@ obj-$(CONFIG_SYN_COOKIES) += syncookies.o
 obj-$(CONFIG_INET_AH) += ah4.o
 obj-$(CONFIG_INET_ESP) += esp4.o
 obj-$(CONFIG_INET_IPCOMP) += ipcomp.o
-obj-$(CONFIG_INET_TUNNEL) += xfrm4_tunnel.o 
+obj-$(CONFIG_INET_XFRM_TUNNEL) += xfrm4_tunnel.o
+obj-$(CONFIG_INET_TUNNEL) += tunnel4.o
 obj-$(CONFIG_IP_PNP) += ipconfig.o
 obj-$(CONFIG_IP_ROUTE_MULTIPATH_RR) += multipath_rr.o
 obj-$(CONFIG_IP_ROUTE_MULTIPATH_RANDOM) += multipath_random.o
diff --git a/net/ipv4/ipip.c b/net/ipv4/ipip.c
index 03d13742a4b..eef07b0916a 100644
--- a/net/ipv4/ipip.c
+++ b/net/ipv4/ipip.c
@@ -114,7 +114,6 @@
 #include <net/sock.h>
 #include <net/ip.h>
 #include <net/icmp.h>
-#include <net/protocol.h>
 #include <net/ipip.h>
 #include <net/inet_ecn.h>
 #include <net/xfrm.h>
@@ -274,7 +273,7 @@ static void ipip_tunnel_uninit(struct net_device *dev)
 	dev_put(dev);
 }
 
-static void ipip_err(struct sk_buff *skb, u32 info)
+static int ipip_err(struct sk_buff *skb, u32 info)
 {
 #ifndef I_WISH_WORLD_WERE_PERFECT
 
@@ -286,21 +285,22 @@ static void ipip_err(struct sk_buff *skb, u32 info)
 	int type = skb->h.icmph->type;
 	int code = skb->h.icmph->code;
 	struct ip_tunnel *t;
+	int err;
 
 	switch (type) {
 	default:
 	case ICMP_PARAMETERPROB:
-		return;
+		return 0;
 
 	case ICMP_DEST_UNREACH:
 		switch (code) {
 		case ICMP_SR_FAILED:
 		case ICMP_PORT_UNREACH:
 			/* Impossible event. */
-			return;
+			return 0;
 		case ICMP_FRAG_NEEDED:
 			/* Soft state for pmtu is maintained by IP core. */
-			return;
+			return 0;
 		default:
 			/* All others are translated to HOST_UNREACH.
 			   rfc2003 contains "deep thoughts" about NET_UNREACH,
@@ -311,14 +311,18 @@ static void ipip_err(struct sk_buff *skb, u32 info)
 		break;
 	case ICMP_TIME_EXCEEDED:
 		if (code != ICMP_EXC_TTL)
-			return;
+			return 0;
 		break;
 	}
 
+	err = -ENOENT;
+
 	read_lock(&ipip_lock);
 	t = ipip_tunnel_lookup(iph->daddr, iph->saddr);
 	if (t == NULL || t->parms.iph.daddr == 0)
 		goto out;
+
+	err = 0;
 	if (t->parms.iph.ttl == 0 && type == ICMP_TIME_EXCEEDED)
 		goto out;
 
@@ -329,7 +333,7 @@ static void ipip_err(struct sk_buff *skb, u32 info)
 	t->err_time = jiffies;
 out:
 	read_unlock(&ipip_lock);
-	return;
+	return err;
 #else
 	struct iphdr *iph = (struct iphdr*)dp;
 	int hlen = iph->ihl<<2;
@@ -344,15 +348,15 @@ out:
 	struct rtable *rt;
 
 	if (len < hlen + sizeof(struct iphdr))
-		return;
+		return 0;
 	eiph = (struct iphdr*)(dp + hlen);
 
 	switch (type) {
 	default:
-		return;
+		return 0;
 	case ICMP_PARAMETERPROB:
 		if (skb->h.icmph->un.gateway < hlen)
-			return;
+			return 0;
 
 		/* So... This guy found something strange INSIDE encapsulated
 		   packet. Well, he is fool, but what can we do ?
@@ -366,16 +370,16 @@ out:
 		case ICMP_SR_FAILED:
 		case ICMP_PORT_UNREACH:
 			/* Impossible event. */
-			return;
+			return 0;
 		case ICMP_FRAG_NEEDED:
 			/* And it is the only really necessary thing :-) */
 			rel_info = ntohs(skb->h.icmph->un.frag.mtu);
 			if (rel_info < hlen+68)
-				return;
+				return 0;
 			rel_info -= hlen;
 			/* BSD 4.2 MORE DOES NOT EXIST IN NATURE. */
 			if (rel_info > ntohs(eiph->tot_len))
-				return;
+				return 0;
 			break;
 		default:
 			/* All others are translated to HOST_UNREACH.
@@ -389,14 +393,14 @@ out:
 		break;
 	case ICMP_TIME_EXCEEDED:
 		if (code != ICMP_EXC_TTL)
-			return;
+			return 0;
 		break;
 	}
 
 	/* Prepare fake skb to feed it to icmp_send */
 	skb2 = skb_clone(skb, GFP_ATOMIC);
 	if (skb2 == NULL)
-		return;
+		return 0;
 	dst_release(skb2->dst);
 	skb2->dst = NULL;
 	skb_pull(skb2, skb->data - (u8*)eiph);
@@ -409,7 +413,7 @@ out:
 	fl.proto = IPPROTO_IPIP;
 	if (ip_route_output_key(&rt, &key)) {
 		kfree_skb(skb2);
-		return;
+		return 0;
 	}
 	skb2->dev = rt->u.dst.dev;
 
@@ -424,14 +428,14 @@ out:
 		    rt->u.dst.dev->type != ARPHRD_TUNNEL) {
 			ip_rt_put(rt);
 			kfree_skb(skb2);
-			return;
+			return 0;
 		}
 	} else {
 		ip_rt_put(rt);
 		if (ip_route_input(skb2, eiph->daddr, eiph->saddr, eiph->tos, skb2->dev) ||
 		    skb2->dst->dev->type != ARPHRD_TUNNEL) {
 			kfree_skb(skb2);
-			return;
+			return 0;
 		}
 	}
 
@@ -439,7 +443,7 @@ out:
 	if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED) {
 		if (rel_info > dst_mtu(skb2->dst)) {
 			kfree_skb(skb2);
-			return;
+			return 0;
 		}
 		skb2->dst->ops->update_pmtu(skb2->dst, rel_info);
 		rel_info = htonl(rel_info);
@@ -453,7 +457,7 @@ out:
 
 	icmp_send(skb2, rel_type, rel_code, rel_info);
 	kfree_skb(skb2);
-	return;
+	return 0;
 #endif
 }
 
@@ -855,39 +859,12 @@ static int __init ipip_fb_tunnel_init(struct net_device *dev)
 	return 0;
 }
 
-#ifdef CONFIG_INET_TUNNEL
 static struct xfrm_tunnel ipip_handler = {
 	.handler	=	ipip_rcv,
 	.err_handler	=	ipip_err,
+	.priority	=	1,
 };
 
-static inline int ipip_register(void)
-{
-	return xfrm4_tunnel_register(&ipip_handler);
-}
-
-static inline int ipip_unregister(void)
-{
-	return xfrm4_tunnel_deregister(&ipip_handler);
-}
-#else
-static struct net_protocol ipip_protocol = {
-	.handler	=	ipip_rcv,
-	.err_handler	=	ipip_err,
-	.no_policy	=	1,
-};
-
-static inline int ipip_register(void)
-{
-	return inet_add_protocol(&ipip_protocol, IPPROTO_IPIP);
-}
-
-static inline int ipip_unregister(void)
-{
-	return inet_del_protocol(&ipip_protocol, IPPROTO_IPIP);
-}
-#endif
-
 static char banner[] __initdata =
 	KERN_INFO "IPv4 over IPv4 tunneling driver\n";
 
@@ -897,7 +874,7 @@ static int __init ipip_init(void)
 
 	printk(banner);
 
-	if (ipip_register() < 0) {
+	if (xfrm4_tunnel_register(&ipip_handler)) {
 		printk(KERN_INFO "ipip init: can't register tunnel\n");
 		return -EAGAIN;
 	}
@@ -919,7 +896,7 @@ static int __init ipip_init(void)
  err2:
 	free_netdev(ipip_fb_tunnel_dev);
  err1:
-	ipip_unregister();
+	xfrm4_tunnel_deregister(&ipip_handler);
 	goto out;
 }
 
@@ -939,7 +916,7 @@ static void __exit ipip_destroy_tunnels(void)
 
 static void __exit ipip_fini(void)
 {
-	if (ipip_unregister() < 0)
+	if (xfrm4_tunnel_deregister(&ipip_handler))
 		printk(KERN_INFO "ipip close: can't deregister tunnel\n");
 
 	rtnl_lock();
diff --git a/net/ipv4/tunnel4.c b/net/ipv4/tunnel4.c
new file mode 100644
index 00000000000..0d7d386dac2
--- /dev/null
+++ b/net/ipv4/tunnel4.c
@@ -0,0 +1,113 @@
+/* tunnel4.c: Generic IP tunnel transformer.
+ *
+ * Copyright (C) 2003 David S. Miller (davem@redhat.com)
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/netdevice.h>
+#include <linux/skbuff.h>
+#include <net/protocol.h>
+#include <net/xfrm.h>
+
+static struct xfrm_tunnel *tunnel4_handlers;
+static DEFINE_MUTEX(tunnel4_mutex);
+
+int xfrm4_tunnel_register(struct xfrm_tunnel *handler)
+{
+	struct xfrm_tunnel **pprev;
+	int ret = -EEXIST;
+	int priority = handler->priority;
+
+	mutex_lock(&tunnel4_mutex);
+
+	for (pprev = &tunnel4_handlers; *pprev; pprev = &(*pprev)->next) {
+		if ((*pprev)->priority > priority)
+			break;
+		if ((*pprev)->priority == priority)
+			goto err;
+	}
+
+	handler->next = *pprev;
+	*pprev = handler;
+
+	ret = 0;
+
+err:
+	mutex_unlock(&tunnel4_mutex);
+
+	return ret;
+}
+
+EXPORT_SYMBOL(xfrm4_tunnel_register);
+
+int xfrm4_tunnel_deregister(struct xfrm_tunnel *handler)
+{
+	struct xfrm_tunnel **pprev;
+	int ret = -ENOENT;
+
+	mutex_lock(&tunnel4_mutex);
+
+	for (pprev = &tunnel4_handlers; *pprev; pprev = &(*pprev)->next) {
+		if (*pprev == handler) {
+			*pprev = handler->next;
+			ret = 0;
+			break;
+		}
+	}
+
+	mutex_unlock(&tunnel4_mutex);
+
+	synchronize_net();
+
+	return ret;
+}
+
+EXPORT_SYMBOL(xfrm4_tunnel_deregister);
+
+static int tunnel4_rcv(struct sk_buff *skb)
+{
+	struct xfrm_tunnel *handler;
+
+	for (handler = tunnel4_handlers; handler; handler = handler->next)
+		if (!handler->handler(skb))
+			return 0;
+
+	kfree_skb(skb);
+	return 0;
+}
+
+static void tunnel4_err(struct sk_buff *skb, u32 info)
+{
+	struct xfrm_tunnel *handler;
+
+	for (handler = tunnel4_handlers; handler; handler = handler->next)
+		if (!handler->err_handler(skb, info))
+			break;
+}
+
+static struct net_protocol tunnel4_protocol = {
+	.handler	=	tunnel4_rcv,
+	.err_handler	=	tunnel4_err,
+	.no_policy	=	1,
+};
+
+static int __init tunnel4_init(void)
+{
+	if (inet_add_protocol(&tunnel4_protocol, IPPROTO_IPIP)) {
+		printk(KERN_ERR "tunnel4 init: can't add protocol\n");
+		return -EAGAIN;
+	}
+	return 0;
+}
+
+static void __exit tunnel4_fini(void)
+{
+	if (inet_del_protocol(&tunnel4_protocol, IPPROTO_IPIP))
+		printk(KERN_ERR "tunnel4 close: can't remove protocol\n");
+}
+
+module_init(tunnel4_init);
+module_exit(tunnel4_fini);
+MODULE_LICENSE("GPL");
diff --git a/net/ipv4/xfrm4_tunnel.c b/net/ipv4/xfrm4_tunnel.c
index b08d56b117f..2d670935c2b 100644
--- a/net/ipv4/xfrm4_tunnel.c
+++ b/net/ipv4/xfrm4_tunnel.c
@@ -26,64 +26,6 @@ static int ipip_xfrm_rcv(struct xfrm_state *x, struct xfrm_decap_state *decap, s
 	return 0;
 }
 
-static struct xfrm_tunnel *ipip_handler;
-static DEFINE_MUTEX(xfrm4_tunnel_mutex);
-
-int xfrm4_tunnel_register(struct xfrm_tunnel *handler)
-{
-	int ret;
-
-	mutex_lock(&xfrm4_tunnel_mutex);
-	ret = 0;
-	if (ipip_handler != NULL)
-		ret = -EINVAL;
-	if (!ret)
-		ipip_handler = handler;
-	mutex_unlock(&xfrm4_tunnel_mutex);
-
-	return ret;
-}
-
-EXPORT_SYMBOL(xfrm4_tunnel_register);
-
-int xfrm4_tunnel_deregister(struct xfrm_tunnel *handler)
-{
-	int ret;
-
-	mutex_lock(&xfrm4_tunnel_mutex);
-	ret = 0;
-	if (ipip_handler != handler)
-		ret = -EINVAL;
-	if (!ret)
-		ipip_handler = NULL;
-	mutex_unlock(&xfrm4_tunnel_mutex);
-
-	synchronize_net();
-
-	return ret;
-}
-
-EXPORT_SYMBOL(xfrm4_tunnel_deregister);
-
-static int ipip_rcv(struct sk_buff *skb)
-{
-	struct xfrm_tunnel *handler = ipip_handler;
-
-	/* Tunnel devices take precedence.  */
-	if (handler && handler->handler(skb) == 0)
-		return 0;
-
-	return xfrm4_rcv(skb);
-}
-
-static void ipip_err(struct sk_buff *skb, u32 info)
-{
-	struct xfrm_tunnel *handler = ipip_handler;
-
-	if (handler)
-		handler->err_handler(skb, info);
-}
-
 static int ipip_init_state(struct xfrm_state *x)
 {
 	if (!x->props.mode)
@@ -111,10 +53,15 @@ static struct xfrm_type ipip_type = {
 	.output		= ipip_output
 };
 
-static struct net_protocol ipip_protocol = {
-	.handler	=	ipip_rcv,
-	.err_handler	=	ipip_err,
-	.no_policy	=	1,
+static int xfrm_tunnel_err(struct sk_buff *skb, u32 info)
+{
+	return -ENOENT;
+}
+
+static struct xfrm_tunnel xfrm_tunnel_handler = {
+	.handler	=	xfrm4_rcv,
+	.err_handler	=	xfrm_tunnel_err,
+	.priority	=	2,
 };
 
 static int __init ipip_init(void)
@@ -123,8 +70,8 @@ static int __init ipip_init(void)
 		printk(KERN_INFO "ipip init: can't add xfrm type\n");
 		return -EAGAIN;
 	}
-	if (inet_add_protocol(&ipip_protocol, IPPROTO_IPIP) < 0) {
-		printk(KERN_INFO "ipip init: can't add protocol\n");
+	if (xfrm4_tunnel_register(&xfrm_tunnel_handler)) {
+		printk(KERN_INFO "ipip init: can't add xfrm handler\n");
 		xfrm_unregister_type(&ipip_type, AF_INET);
 		return -EAGAIN;
 	}
@@ -133,8 +80,8 @@ static int __init ipip_init(void)
 
 static void __exit ipip_fini(void)
 {
-	if (inet_del_protocol(&ipip_protocol, IPPROTO_IPIP) < 0)
-		printk(KERN_INFO "ipip close: can't remove protocol\n");
+	if (xfrm4_tunnel_deregister(&xfrm_tunnel_handler))
+		printk(KERN_INFO "ipip close: can't remove xfrm handler\n");
 	if (xfrm_unregister_type(&ipip_type, AF_INET) < 0)
 		printk(KERN_INFO "ipip close: can't remove xfrm type\n");
 }