icmp.c

ipv6地址转换器

/*
 *	Internet Control Message Protocol (ICMPv6)
 *	Linux INET6 implementation
 *
 *	Authors:
 *	Pedro Roque		<[e-mail]>
 *
 *	$Id: icmp.c,v 1.21.2.1 1999/05/19 22:07:36 davem Exp $
 *
 *	Based on net/ipv4/icmp.c
 *
 *	RFC 1885
 *
 *	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.
 */

/*
 *	Changes:
 *
 *	Andi Kleen		:	exception handling
 *	Andi Kleen			add rate limits. never reply to a icmp.
 *					add more length checks and other fixes.
 */

#define __NO_VERSION__
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/init.h>

#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/icmpv6.h>

#include <net/ip.h>
#include <net/sock.h>

#include <net/ipv6.h>
#include <net/checksum.h>
#include <net/protocol.h>
#include <net/raw.h>
#include <net/rawv6.h>
#include <net/transp_v6.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/icmp.h>

#include <asm/uaccess.h>
#include <asm/system.h>

struct icmpv6_mib icmpv6_statistics;

/*
 *	ICMP socket for flow control.
 */

struct socket *icmpv6_socket;

int icmpv6_rcv(struct sk_buff *skb, unsigned long len);

static struct inet6_protocol icmpv6_protocol = 
{
	icmpv6_rcv,		/* handler		*/
	NULL,			/* error control	*/
	NULL,			/* next			*/
	IPPROTO_ICMPV6,		/* protocol ID		*/
	0,			/* copy			*/
	NULL,			/* data			*/
	"ICMPv6"	       	/* name			*/
};

struct icmpv6_msg {
	struct icmp6hdr		icmph;
	__u8 			*data;
	struct in6_addr		*daddr;
	int			len;
	__u32			csum;
};



/*
 *	getfrag callback
 */

static int icmpv6_getfrag(const void *data, struct in6_addr *saddr, 
			   char *buff, unsigned int offset, unsigned int len)
{
	struct icmpv6_msg *msg = (struct icmpv6_msg *) data;
	struct icmp6hdr *icmph;
	__u32 csum;

	/* 
	 *	in theory offset must be 0 since we never send more 
	 *	than IPV6_MIN_MTU bytes on an error or more than the path mtu
	 *	on an echo reply. (those are the rules on RFC 1883)
	 *
	 * 	Luckily, this statement is obsolete after
	 *	draft-ietf-ipngwg-icmp-v2-00           --ANK (980730)
	 */

	if (offset) {
		csum = csum_partial_copy((void *) msg->data +
					 offset - sizeof(struct icmp6hdr), 
					 buff, len, msg->csum);
		msg->csum = csum;
		return 0;
	}

	csum = csum_partial_copy((void *) &msg->icmph, buff,
				 sizeof(struct icmp6hdr), msg->csum);

	csum = csum_partial_copy((void *) msg->data, 
				 buff + sizeof(struct icmp6hdr),
				 len - sizeof(struct icmp6hdr), csum);

	icmph = (struct icmp6hdr *) buff;

	icmph->icmp6_cksum = csum_ipv6_magic(saddr, msg->daddr, msg->len,
					     IPPROTO_ICMPV6, csum);
	return 0; 
}


/* 
 * Slightly more convenient version of icmpv6_send.
 */
void icmpv6_param_prob(struct sk_buff *skb, int code, void *pos)
{
	int offset = (u8*)pos - (u8*)skb->nh.ipv6h; 
	
	icmpv6_send(skb, ICMPV6_PARAMPROB, code, offset, skb->dev);
	kfree_skb(skb);
}

/*
 * Figure out, may we reply to this packet with icmp error.
 *
 * We do not reply, if:
 *	- it was icmp error message.
 *	- it is truncated, so that it is known, that protocol is ICMPV6
 *	  (i.e. in the middle of some exthdr)
 *	- it is not the first fragment. BTW IPv6 specs say nothing about
 *	  this case, but it is clear, that our reply would be useless
 *	  for sender.
 *
 *	--ANK (980726)
 */

static int is_ineligible(struct ipv6hdr *hdr, int len)
{
	u8 *ptr;
	__u8 nexthdr = hdr->nexthdr;

	if (len < (int)sizeof(*hdr))
		return 1;

	ptr = ipv6_skip_exthdr((struct ipv6_opt_hdr *)(hdr+1), &nexthdr, len - sizeof(*hdr));
	if (!ptr)
		return 0;
	if (nexthdr == IPPROTO_ICMPV6) {
		struct icmp6hdr *ihdr =	(struct icmp6hdr *)ptr;
		return (ptr - (u8*)hdr) > len || !(ihdr->icmp6_type & 0x80); 
	}
	return nexthdr == NEXTHDR_FRAGMENT;
}

int sysctl_icmpv6_time = 1*HZ; 

/* 
 * Check the ICMP output rate limit 
 */
static inline int icmpv6_xrlim_allow(struct sock *sk, int type,
				     struct flowi *fl)
{
	struct dst_entry *dst;
	int res = 0;

	/* Informational messages are not limited. */
	if (type & 0x80)
		return 1;

	/* Do not limit pmtu discovery, it would break it. */
	if (type == ICMPV6_PKT_TOOBIG)
		return 1;

	/* 
	 * Look up the output route.
	 * XXX: perhaps the expire for routing entries cloned by
	 * this lookup should be more aggressive (not longer than timeout).
	 */
	dst = ip6_route_output(sk, fl);
	if (dst->error) {
		ipv6_statistics.Ip6OutNoRoutes++;
	} else if (dst->dev && (dst->dev->flags&IFF_LOOPBACK)) {
		res = 1;
	} else {
		struct rt6_info *rt = (struct rt6_info *)dst;
		int tmo = sysctl_icmpv6_time;

		/* Give more bandwidth to wider prefixes. */
		if (rt->rt6i_dst.plen < 128)
			tmo >>= ((128 - rt->rt6i_dst.plen)>>5);

		res = xrlim_allow(dst, tmo);
	}
	dst_release(dst);
	return res;
}

/*
 *	an inline helper for the "simple" if statement below
 *	checks if parameter problem report is caused by an
 *	unrecognized IPv6 option that has the Option Type 
 *	highest-order two bits set to 10
 */

static __inline__ int opt_unrec(struct sk_buff *skb, __u32 offset)
{
	u8 *buff = skb->nh.raw;

	return ( ( *(buff + offset) & 0xC0 ) == 0x80 );
}

/*
 *	Send an ICMP message in response to a packet in error
 */

void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info, 
		 struct device *dev)
{
	struct ipv6hdr *hdr = skb->nh.ipv6h;
	struct sock *sk = icmpv6_socket->sk;
	struct in6_addr *saddr = NULL;
	int iif = 0;
	struct icmpv6_msg msg;
	struct flowi fl;
	int addr_type = 0;
	int len;

	/*
	 *	sanity check pointer in case of parameter problem
	 */

	if (type == ICMPV6_PARAMPROB && 
	    (info > (skb->tail - ((unsigned char *) hdr)))) {
		printk(KERN_DEBUG "icmpv6_send: bug! pointer > skb\n");
		return;
	}

	/*
	 *	Make sure we respect the rules 
	 *	i.e. RFC 1885 2.4(e)
	 *	Rule (e.1) is enforced by not using icmpv6_send
	 *	in any code that processes icmp errors.
	 */
	
	addr_type = ipv6_addr_type(&hdr->daddr);

	if (ipv6_chk_addr(&hdr->daddr, skb->dev, 0))
		saddr = &hdr->daddr;

	/*
	 *	Dest addr check
	 */

	if ((addr_type & IPV6_ADDR_MULTICAST || skb->pkt_type != PACKET_HOST)) {
		if (type != ICMPV6_PKT_TOOBIG &&
		    !(type == ICMPV6_PARAMPROB && 
		      code == ICMPV6_UNK_OPTION && 
		      (opt_unrec(skb, info))))
			return;

		saddr = NULL;
	}

	addr_type = ipv6_addr_type(&hdr->saddr);

	/*
	 *	Source addr check
	 */

	if (addr_type & IPV6_ADDR_LINKLOCAL)
		iif = skb->dev->ifindex;

	/*
	 *	Must not send if we know that source is Anycast also.
	 *	for now we don't know that.
	 */
	if ((addr_type == IPV6_ADDR_ANY) || (addr_type & IPV6_ADDR_MULTICAST)) {
		printk(KERN_DEBUG "icmpv6_send: addr_any/mcast source\n");
		return;
	}

	/* 
	 *	Never answer to a ICMP packet.
	 */
	if (is_ineligible(hdr, (u8*)skb->tail - (u8*)hdr)) {
		if (net_ratelimit())
			printk(KERN_DEBUG "icmpv6_send: no reply to icmp error/fragment\n"); 
		return;
	}

	fl.proto = IPPROTO_ICMPV6;
	fl.nl_u.ip6_u.daddr = &hdr->saddr;
	fl.nl_u.ip6_u.saddr = saddr;
	fl.oif = iif;
	fl.fl6_flowlabel = 0;
	fl.uli_u.icmpt.type = type;
	fl.uli_u.icmpt.code = code;

	if (!icmpv6_xrlim_allow(sk, type, &fl)) 
		return; 

	/*
	 *	ok. kick it. checksum will be provided by the 
	 *	getfrag_t callback.
	 */

	msg.icmph.icmp6_type = type;
	msg.icmph.icmp6_code = code;
	msg.icmph.icmp6_cksum = 0;
	msg.icmph.icmp6_pointer = htonl(info);

	msg.data = skb->nh.raw;
	msg.csum = 0;
	msg.daddr = &hdr->saddr;