ip6_output.c

嵌入式系统设计与实验教材二源码linux内核移植与编译

/*
 *	IPv6 output functions
 *	Linux INET6 implementation 
 *
 *	Authors:
 *	Pedro Roque		<[e-mail]>	
 *
 *	$Id: ip6_output.c,v 1.33 2001/09/20 00:35:35 davem Exp $
 *
 *	Based on linux/net/ipv4/ip_output.c
 *
 *	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:
 *	A.N.Kuznetsov	:	airthmetics in fragmentation.
 *				extension headers are implemented.
 *				route changes now work.
 *				ip6_forward does not confuse sniffers.
 *				etc.
 *
 *      H. von Brand    :       Added missing #include <linux/string.h>
 *	Imran Patel	: 	frag id should be in NBO
 */

#include <linux/config.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/string.h>
#include <linux/socket.h>
#include <linux/net.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <linux/in6.h>
#include <linux/route.h>

#include <linux/netfilter.h>
#include <linux/netfilter_ipv6.h>

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

#include <net/ipv6.h>
#include <net/ndisc.h>
#include <net/protocol.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/rawv6.h>
#include <net/icmp.h>

static __inline__ void ipv6_select_ident(struct sk_buff *skb, struct frag_hdr *fhdr)
{
	static u32 ipv6_fragmentation_id = 1;
	static spinlock_t ip6_id_lock = SPIN_LOCK_UNLOCKED;

	spin_lock_bh(&ip6_id_lock);
	fhdr->identification = htonl(ipv6_fragmentation_id);
	if (++ipv6_fragmentation_id == 0)
		ipv6_fragmentation_id = 1;
	spin_unlock_bh(&ip6_id_lock);
}

static inline int ip6_output_finish(struct sk_buff *skb)
{

	struct dst_entry *dst = skb->dst;
	struct hh_cache *hh = dst->hh;

	if (hh) {
		read_lock_bh(&hh->hh_lock);
		memcpy(skb->data - 16, hh->hh_data, 16);
		read_unlock_bh(&hh->hh_lock);
	        skb_push(skb, hh->hh_len);
		return hh->hh_output(skb);
	} else if (dst->neighbour)
		return dst->neighbour->output(skb);

	kfree_skb(skb);
	return -EINVAL;

}

/* dev_loopback_xmit for use with netfilter. */
static int ip6_dev_loopback_xmit(struct sk_buff *newskb)
{
	newskb->mac.raw = newskb->data;
	__skb_pull(newskb, newskb->nh.raw - newskb->data);
	newskb->pkt_type = PACKET_LOOPBACK;
	newskb->ip_summed = CHECKSUM_UNNECESSARY;
	BUG_TRAP(newskb->dst);

	netif_rx(newskb);
	return 0;
}


int ip6_output(struct sk_buff *skb)
{
	struct dst_entry *dst = skb->dst;
	struct net_device *dev = dst->dev;

	skb->protocol = __constant_htons(ETH_P_IPV6);
	skb->dev = dev;

	if (ipv6_addr_is_multicast(&skb->nh.ipv6h->daddr)) {
		if (!(dev->flags&IFF_LOOPBACK) &&
		    (skb->sk == NULL || skb->sk->net_pinfo.af_inet6.mc_loop) &&
		    ipv6_chk_mcast_addr(dev, &skb->nh.ipv6h->daddr)) {
			struct sk_buff *newskb = skb_clone(skb, GFP_ATOMIC);

			/* Do not check for IFF_ALLMULTI; multicast routing
			   is not supported in any case.
			 */
			if (newskb)
				NF_HOOK(PF_INET6, NF_IP6_POST_ROUTING, newskb, NULL,
					newskb->dev,
					ip6_dev_loopback_xmit);

			if (skb->nh.ipv6h->hop_limit == 0) {
				kfree_skb(skb);
				return 0;
			}
		}

		IP6_INC_STATS(Ip6OutMcastPkts);
	}

	return NF_HOOK(PF_INET6, NF_IP6_POST_ROUTING, skb,NULL, skb->dev,ip6_output_finish);
}


#ifdef CONFIG_NETFILTER
static int route6_me_harder(struct sk_buff *skb)
{
	struct ipv6hdr *iph = skb->nh.ipv6h;
	struct dst_entry *dst;
	struct flowi fl;

	fl.proto = iph->nexthdr;
	fl.fl6_dst = &iph->daddr;
	fl.fl6_src = &iph->saddr;
	fl.oif = skb->sk ? skb->sk->bound_dev_if : 0;
	fl.fl6_flowlabel = 0;
	fl.uli_u.ports.dport = 0;
	fl.uli_u.ports.sport = 0;

	dst = ip6_route_output(skb->sk, &fl);

	if (dst->error) {
		if (net_ratelimit())
			printk(KERN_DEBUG "route6_me_harder: No more route.\n");
		return -EINVAL;
	}

	/* Drop old route. */
	dst_release(skb->dst);

	skb->dst = dst;
	return 0;
}
#endif

static inline int ip6_maybe_reroute(struct sk_buff *skb)
{
#ifdef CONFIG_NETFILTER
	if (skb->nfcache & NFC_ALTERED){
		if (route6_me_harder(skb) != 0){
			kfree_skb(skb);
			return -EINVAL;
		}
	}
#endif /* CONFIG_NETFILTER */
	return skb->dst->output(skb);
}

/*
 *	xmit an sk_buff (used by TCP)
 */

int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl,
	     struct ipv6_txoptions *opt)
{
	struct ipv6_pinfo * np = sk ? &sk->net_pinfo.af_inet6 : NULL;
	struct in6_addr *first_hop = fl->nl_u.ip6_u.daddr;
	struct dst_entry *dst = skb->dst;
	struct ipv6hdr *hdr;
	u8  proto = fl->proto;
	int seg_len = skb->len;
	int hlimit;

	if (opt) {
		int head_room;

		/* First: exthdrs may take lots of space (~8K for now)
		   MAX_HEADER is not enough.
		 */
		head_room = opt->opt_nflen + opt->opt_flen;
		seg_len += head_room;
		head_room += sizeof(struct ipv6hdr) + ((dst->dev->hard_header_len + 15)&~15);

		if (skb_headroom(skb) < head_room) {
			struct sk_buff *skb2 = skb_realloc_headroom(skb, head_room);
			kfree_skb(skb);
			skb = skb2;
			if (skb == NULL)
				return -ENOBUFS;
			if (sk)
				skb_set_owner_w(skb, sk);
		}
		if (opt->opt_flen)
			ipv6_push_frag_opts(skb, opt, &proto);
		if (opt->opt_nflen)
			ipv6_push_nfrag_opts(skb, opt, &proto, &first_hop);
	}

	hdr = skb->nh.ipv6h = (struct ipv6hdr*)skb_push(skb, sizeof(struct ipv6hdr));

	/*
	 *	Fill in the IPv6 header
	 */

	*(u32*)hdr = __constant_htonl(0x60000000) | fl->fl6_flowlabel;
	hlimit = -1;
	if (np)
		hlimit = np->hop_limit;
	if (hlimit < 0)
		hlimit = ((struct rt6_info*)dst)->rt6i_hoplimit;

	hdr->payload_len = htons(seg_len);
	hdr->nexthdr = proto;
	hdr->hop_limit = hlimit;

	ipv6_addr_copy(&hdr->saddr, fl->nl_u.ip6_u.saddr);
	ipv6_addr_copy(&hdr->daddr, first_hop);

	if (skb->len <= dst->pmtu) {
		IP6_INC_STATS(Ip6OutRequests);
		return NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev, ip6_maybe_reroute);
	}

	if (net_ratelimit())
		printk(KERN_DEBUG "IPv6: sending pkt_too_big to self\n");
	icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, dst->pmtu, skb->dev);
	kfree_skb(skb);
	return -EMSGSIZE;
}

/*
 *	To avoid extra problems ND packets are send through this
 *	routine. It's code duplication but I really want to avoid
 *	extra checks since ipv6_build_header is used by TCP (which
 *	is for us performace critical)
 */

int ip6_nd_hdr(struct sock *sk, struct sk_buff *skb, struct net_device *dev,
	       struct in6_addr *saddr, struct in6_addr *daddr,
	       int proto, int len)
{
	struct ipv6_pinfo *np = &sk->net_pinfo.af_inet6;
	struct ipv6hdr *hdr;
	int totlen;

	skb->protocol = __constant_htons(ETH_P_IPV6);
	skb->dev = dev;

	totlen = len + sizeof(struct ipv6hdr);

	hdr = (struct ipv6hdr *) skb_put(skb, sizeof(struct ipv6hdr));
	skb->nh.ipv6h = hdr;

	*(u32*)hdr = htonl(0x60000000);

	hdr->payload_len = htons(len);
	hdr->nexthdr = proto;
	hdr->hop_limit = np->hop_limit;

	ipv6_addr_copy(&hdr->saddr, saddr);
	ipv6_addr_copy(&hdr->daddr, daddr);

	return 0;
}

static struct ipv6hdr * ip6_bld_1(struct sock *sk, struct sk_buff *skb, struct flowi *fl,
				  int hlimit, unsigned pktlength)
{
	struct ipv6hdr *hdr;
	
	skb->nh.raw = skb_put(skb, sizeof(struct ipv6hdr));
	hdr = skb->nh.ipv6h;
	
	*(u32*)hdr = fl->fl6_flowlabel | htonl(0x60000000);

	hdr->payload_len = htons(pktlength - sizeof(struct ipv6hdr));
	hdr->hop_limit = hlimit;
	hdr->nexthdr = fl->proto;

	ipv6_addr_copy(&hdr->saddr, fl->nl_u.ip6_u.saddr);
	ipv6_addr_copy(&hdr->daddr, fl->nl_u.ip6_u.daddr);
	return hdr;
}

static __inline__ u8 * ipv6_build_fraghdr(struct sk_buff *skb, u8* prev_hdr, unsigned offset)
{
	struct frag_hdr *fhdr;

	fhdr = (struct frag_hdr *) skb_put(skb, sizeof(struct frag_hdr));

	fhdr->nexthdr  = *prev_hdr;
	*prev_hdr = NEXTHDR_FRAGMENT;
	prev_hdr = &fhdr->nexthdr;

	fhdr->reserved = 0;
	fhdr->frag_off = htons(offset);
	ipv6_select_ident(skb, fhdr);
	return &fhdr->nexthdr;
}

static int ip6_frag_xmit(struct sock *sk, inet_getfrag_t getfrag,
			 const void *data, struct dst_entry *dst,
			 struct flowi *fl, struct ipv6_txoptions *opt,
			 struct in6_addr *final_dst,
			 int hlimit, int flags, unsigned length, int mtu)
{
	struct ipv6hdr *hdr;
	struct sk_buff *last_skb;
	u8 *prev_hdr;
	int unfrag_len;
	int frag_len;
	int last_len;
	int nfrags;
	int fhdr_dist;
	int frag_off;
	int data_off;
	int err;

	/*
	 *	Fragmentation
	 *
	 *	Extension header order:
	 *	Hop-by-hop -> Dest0 -> Routing -> Fragment -> Auth -> Dest1 -> rest (...)
	 *	
	 *	We must build the non-fragmented part that
	 *	will be in every packet... this also means
	 *	that other extension headers (Dest, Auth, etc)
	 *	must be considered in the data to be fragmented
	 */

	unfrag_len = sizeof(struct ipv6hdr) + sizeof(struct frag_hdr);
	last_len = length;

	if (opt) {
		unfrag_len += opt->opt_nflen;
		last_len += opt->opt_flen;
	}

	/*
	 *	Length of fragmented part on every packet but 
	 *	the last must be an:
	 *	"integer multiple of 8 octects".
	 */

	frag_len = (mtu - unfrag_len) & ~0x7;

	/* Unfragmentable part exceeds mtu. */
	if (frag_len <= 0) {
		ipv6_local_error(sk, EMSGSIZE, fl, mtu);
		return -EMSGSIZE;
	}

	nfrags = last_len / frag_len;

	/*
	 *	We must send from end to start because of 
	 *	UDP/ICMP checksums. We do a funny trick:
	 *	fill the last skb first with the fixed
	 *	header (and its data) and then use it
	 *	to create the following segments and send it
	 *	in the end. If the peer is checking the M_flag
	 *	to trigger the reassembly code then this 
	 *	might be a good idea.
	 */

	frag_off = nfrags * frag_len;
	last_len -= frag_off;

	if (last_len == 0) {
		last_len = frag_len;
		frag_off -= frag_len;
		nfrags--;
	}
	data_off = frag_off;

	/* And it is implementation problem: for now we assume, that
	   all the exthdrs will fit to the first fragment.
	 */
	if (opt) {
		if (frag_len < opt->opt_flen) {
			ipv6_local_error(sk, EMSGSIZE, fl, mtu);
			return -EMSGSIZE;
		}
		data_off = frag_off - opt->opt_flen;
	}

	if (flags&MSG_PROBE)
		return 0;

	last_skb = sock_alloc_send_skb(sk, unfrag_len + frag_len +
				       dst->dev->hard_header_len + 15,
				       flags & MSG_DONTWAIT, &err);