ipmr.c

Linux内核源代码 为压缩文件 是<<Linux内核>>一书中的源代码

	msg->im_msgtype = assert;
	igmp->code 	=	0;
	skb->nh.iph->tot_len=htons(skb->len);			/* Fix the length */
	skb->h.raw = skb->nh.raw;
        }

	if (mroute_socket == NULL) {
		kfree_skb(skb);
		return -EINVAL;
	}

	/*
	 *	Deliver to mrouted
	 */
	if ((ret=sock_queue_rcv_skb(mroute_socket,skb))<0) {
		if (net_ratelimit())
			printk(KERN_WARNING "mroute: pending queue full, dropping entries.\n");
		kfree_skb(skb);
	}

	return ret;
}

/*
 *	Queue a packet for resolution. It gets locked cache entry!
 */
 
static int
ipmr_cache_unresolved(vifi_t vifi, struct sk_buff *skb)
{
	int err;
	struct mfc_cache *c;

	spin_lock_bh(&mfc_unres_lock);
	for (c=mfc_unres_queue; c; c=c->next) {
		if (c->mfc_mcastgrp == skb->nh.iph->daddr &&
		    c->mfc_origin == skb->nh.iph->saddr)
			break;
	}

	if (c == NULL) {
		/*
		 *	Create a new entry if allowable
		 */

		if (atomic_read(&cache_resolve_queue_len)>=10 ||
		    (c=ipmr_cache_alloc_unres())==NULL) {
			spin_unlock_bh(&mfc_unres_lock);

			kfree_skb(skb);
			return -ENOBUFS;
		}

		/*
		 *	Fill in the new cache entry
		 */
		c->mfc_parent=-1;
		c->mfc_origin=skb->nh.iph->saddr;
		c->mfc_mcastgrp=skb->nh.iph->daddr;

		/*
		 *	Reflect first query at mrouted.
		 */
		if ((err = ipmr_cache_report(skb, vifi, IGMPMSG_NOCACHE))<0) {
			/* If the report failed throw the cache entry 
			   out - Brad Parker
			 */
			spin_unlock_bh(&mfc_unres_lock);

			kmem_cache_free(mrt_cachep, c);
			kfree_skb(skb);
			return err;
		}

		atomic_inc(&cache_resolve_queue_len);
		c->next = mfc_unres_queue;
		mfc_unres_queue = c;

		mod_timer(&ipmr_expire_timer, c->mfc_un.unres.expires);
	}

	/*
	 *	See if we can append the packet
	 */
	if (c->mfc_un.unres.unresolved.qlen>3) {
		kfree_skb(skb);
		err = -ENOBUFS;
	} else {
		skb_queue_tail(&c->mfc_un.unres.unresolved,skb);
		err = 0;
	}

	spin_unlock_bh(&mfc_unres_lock);
	return err;
}

/*
 *	MFC cache manipulation by user space mroute daemon
 */

int ipmr_mfc_delete(struct mfcctl *mfc)
{
	int line;
	struct mfc_cache *c, **cp;

	line=MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);

	for (cp=&mfc_cache_array[line]; (c=*cp) != NULL; cp = &c->next) {
		if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
		    c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr) {
			write_lock_bh(&mrt_lock);
			*cp = c->next;
			write_unlock_bh(&mrt_lock);

			kmem_cache_free(mrt_cachep, c);
			return 0;
		}
	}
	return -ENOENT;
}

int ipmr_mfc_add(struct mfcctl *mfc, int mrtsock)
{
	int line;
	struct mfc_cache *uc, *c, **cp;

	line=MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);

	for (cp=&mfc_cache_array[line]; (c=*cp) != NULL; cp = &c->next) {
		if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
		    c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr)
			break;
	}

	if (c != NULL) {
		write_lock_bh(&mrt_lock);
		c->mfc_parent = mfc->mfcc_parent;
		ipmr_update_threshoulds(c, mfc->mfcc_ttls);
		if (!mrtsock)
			c->mfc_flags |= MFC_STATIC;
		write_unlock_bh(&mrt_lock);
		return 0;
	}

	if(!MULTICAST(mfc->mfcc_mcastgrp.s_addr))
		return -EINVAL;

	c=ipmr_cache_alloc();
	if (c==NULL)
		return -ENOMEM;

	c->mfc_origin=mfc->mfcc_origin.s_addr;
	c->mfc_mcastgrp=mfc->mfcc_mcastgrp.s_addr;
	c->mfc_parent=mfc->mfcc_parent;
	ipmr_update_threshoulds(c, mfc->mfcc_ttls);
	if (!mrtsock)
		c->mfc_flags |= MFC_STATIC;

	write_lock_bh(&mrt_lock);
	c->next = mfc_cache_array[line];
	mfc_cache_array[line] = c;
	write_unlock_bh(&mrt_lock);

	/*
	 *	Check to see if we resolved a queued list. If so we
	 *	need to send on the frames and tidy up.
	 */
	spin_lock_bh(&mfc_unres_lock);
	for (cp = &mfc_unres_queue; (uc=*cp) != NULL;
	     cp = &uc->next) {
		if (uc->mfc_origin == c->mfc_origin &&
		    uc->mfc_mcastgrp == c->mfc_mcastgrp) {
			*cp = uc->next;
			if (atomic_dec_and_test(&cache_resolve_queue_len))
				del_timer(&ipmr_expire_timer);
			break;
		}
	}
	spin_unlock_bh(&mfc_unres_lock);

	if (uc) {
		ipmr_cache_resolve(uc, c);
		kmem_cache_free(mrt_cachep, uc);
	}
	return 0;
}

/*
 *	Close the multicast socket, and clear the vif tables etc
 */
 
static void mroute_clean_tables(struct sock *sk)
{
	int i;
		
	/*
	 *	Shut down all active vif entries
	 */
	for(i=0; i<maxvif; i++) {
		if (!(vif_table[i].flags&VIFF_STATIC))
			vif_delete(i);
	}

	/*
	 *	Wipe the cache
	 */
	for (i=0;i<MFC_LINES;i++) {
		struct mfc_cache *c, **cp;

		cp = &mfc_cache_array[i];
		while ((c = *cp) != NULL) {
			if (c->mfc_flags&MFC_STATIC) {
				cp = &c->next;
				continue;
			}
			write_lock_bh(&mrt_lock);
			*cp = c->next;
			write_unlock_bh(&mrt_lock);

			kmem_cache_free(mrt_cachep, c);
		}
	}

	if (atomic_read(&cache_resolve_queue_len) != 0) {
		struct mfc_cache *c;

		spin_lock_bh(&mfc_unres_lock);
		while (mfc_unres_queue != NULL) {
			c = mfc_unres_queue;
			mfc_unres_queue = c->next;
			spin_unlock_bh(&mfc_unres_lock);

			ipmr_destroy_unres(c);

			spin_lock_bh(&mfc_unres_lock);
		}
		spin_unlock_bh(&mfc_unres_lock);
	}
}

static void mrtsock_destruct(struct sock *sk)
{
	rtnl_lock();
	if (sk == mroute_socket) {
		ipv4_devconf.mc_forwarding--;

		write_lock_bh(&mrt_lock);
		mroute_socket=NULL;
		write_unlock_bh(&mrt_lock);

		mroute_clean_tables(sk);
	}
	rtnl_unlock();
}

/*
 *	Socket options and virtual interface manipulation. The whole
 *	virtual interface system is a complete heap, but unfortunately
 *	that's how BSD mrouted happens to think. Maybe one day with a proper
 *	MOSPF/PIM router set up we can clean this up.
 */
 
int ip_mroute_setsockopt(struct sock *sk,int optname,char *optval,int optlen)
{
	int ret;
	struct vifctl vif;
	struct mfcctl mfc;
	
	if(optname!=MRT_INIT)
	{
		if(sk!=mroute_socket && !capable(CAP_NET_ADMIN))
			return -EACCES;
	}

	switch(optname)
	{
		case MRT_INIT:
			if(sk->type!=SOCK_RAW || sk->num!=IPPROTO_IGMP)
				return -EOPNOTSUPP;
			if(optlen!=sizeof(int))
				return -ENOPROTOOPT;

			rtnl_lock();
			if (mroute_socket) {
				rtnl_unlock();
				return -EADDRINUSE;
			}

			ret = ip_ra_control(sk, 1, mrtsock_destruct);
			if (ret == 0) {
				write_lock_bh(&mrt_lock);
				mroute_socket=sk;
				write_unlock_bh(&mrt_lock);

				ipv4_devconf.mc_forwarding++;
			}
			rtnl_unlock();
			return ret;
		case MRT_DONE:
			if (sk!=mroute_socket)
				return -EACCES;
			return ip_ra_control(sk, 0, NULL);
		case MRT_ADD_VIF:
		case MRT_DEL_VIF:
			if(optlen!=sizeof(vif))
				return -EINVAL;
			if (copy_from_user(&vif,optval,sizeof(vif)))
				return -EFAULT; 
			if(vif.vifc_vifi >= MAXVIFS)
				return -ENFILE;
			rtnl_lock();
			if (optname==MRT_ADD_VIF) {
				ret = vif_add(&vif, sk==mroute_socket);
			} else {
				ret = vif_delete(vif.vifc_vifi);
			}
			rtnl_unlock();
			return ret;

		/*
		 *	Manipulate the forwarding caches. These live
		 *	in a sort of kernel/user symbiosis.
		 */
		case MRT_ADD_MFC:
		case MRT_DEL_MFC:
			if(optlen!=sizeof(mfc))
				return -EINVAL;
			if (copy_from_user(&mfc,optval, sizeof(mfc)))
				return -EFAULT;
			rtnl_lock();
			if (optname==MRT_DEL_MFC)
				ret = ipmr_mfc_delete(&mfc);
			else
				ret = ipmr_mfc_add(&mfc, sk==mroute_socket);
			rtnl_unlock();
			return ret;
		/*
		 *	Control PIM assert.
		 */
		case MRT_ASSERT:
		{
			int v;
			if(get_user(v,(int *)optval))
				return -EFAULT;
			mroute_do_assert=(v)?1:0;
			return 0;
		}
#ifdef CONFIG_IP_PIMSM
		case MRT_PIM:
		{
			int v;
			if(get_user(v,(int *)optval))
				return -EFAULT;
			v = (v)?1:0;
			rtnl_lock();
			if (v != mroute_do_pim) {
				mroute_do_pim = v;
				mroute_do_assert = v;
#ifdef CONFIG_IP_PIMSM_V2
				if (mroute_do_pim)
					inet_add_protocol(&pim_protocol);
				else
					inet_del_protocol(&pim_protocol);
#endif
			}
			rtnl_unlock();
			return 0;
		}
#endif
		/*
		 *	Spurious command, or MRT_VERSION which you cannot
		 *	set.
		 */
		default:
			return -ENOPROTOOPT;
	}
}

/*
 *	Getsock opt support for the multicast routing system.
 */
 
int ip_mroute_getsockopt(struct sock *sk,int optname,char *optval,int *optlen)
{
	int olr;
	int val;

	if(optname!=MRT_VERSION && 
#ifdef CONFIG_IP_PIMSM
	   optname!=MRT_PIM &&
#endif
	   optname!=MRT_ASSERT)
		return -ENOPROTOOPT;

	if(get_user(olr, optlen))
		return -EFAULT;

	olr=min(olr,sizeof(int));
	if(put_user(olr,optlen))
		return -EFAULT;
	if(optname==MRT_VERSION)
		val=0x0305;
#ifdef CONFIG_IP_PIMSM
	else if(optname==MRT_PIM)
		val=mroute_do_pim;
#endif
	else
		val=mroute_do_assert;
	if(copy_to_user(optval,&val,olr))
		return -EFAULT;
	return 0;
}

/*
 *	The IP multicast ioctl support routines.
 */
 
int ipmr_ioctl(struct sock *sk, int cmd, unsigned long arg)
{
	struct sioc_sg_req sr;
	struct sioc_vif_req vr;
	struct vif_device *vif;
	struct mfc_cache *c;
	
	switch(cmd)
	{
		case SIOCGETVIFCNT:
			if (copy_from_user(&vr,(void *)arg,sizeof(vr)))
				return -EFAULT; 
			if(vr.vifi>=maxvif)
				return -EINVAL;
			read_lock(&mrt_lock);
			vif=&vif_table[vr.vifi];
			if(VIF_EXISTS(vr.vifi))	{
				vr.icount=vif->pkt_in;
				vr.ocount=vif->pkt_out;
				vr.ibytes=vif->bytes_in;
				vr.obytes=vif->bytes_out;
				read_unlock(&mrt_lock);

				if (copy_to_user((void *)arg,&vr,sizeof(vr)))
					return -EFAULT;
				return 0;
			}
			read_unlock(&mrt_lock);
			return -EADDRNOTAVAIL;
		case SIOCGETSGCNT:
			if (copy_from_user(&sr,(void *)arg,sizeof(sr)))
				return -EFAULT;

			read_lock(&mrt_lock);
			c = ipmr_cache_find(sr.src.s_addr, sr.grp.s_addr);
			if (c) {
				sr.pktcnt = c->mfc_un.res.pkt;
				sr.bytecnt = c->mfc_un.res.bytes;
				sr.wrong_if = c->mfc_un.res.wrong_if;
				read_unlock(&mrt_lock);

				if (copy_to_user((void *)arg,&sr,sizeof(sr)))
					return -EFAULT;
				return 0;
			}
			read_unlock(&mrt_lock);
			return -EADDRNOTAVAIL;
		default:
			return -ENOIOCTLCMD;
	}
}


static int ipmr_device_event(struct notifier_block *this, unsigned long event, void *ptr)
{
	struct vif_device *v;
	int ct;
	if (event != NETDEV_UNREGISTER)
		return NOTIFY_DONE;
	v=&vif_table[0];
	for(ct=0;ct<maxvif;ct++,v++) {
		if (v->dev==ptr)
			vif_delete(ct);
	}
	return NOTIFY_DONE;
}


static struct notifier_block ip_mr_notifier={
	ipmr_device_event,
	NULL,
	0
};

/*
 * 	Encapsulate a packet by attaching a valid IPIP header to it.
 *	This avoids tunnel drivers and other mess and gives us the speed so
 *	important for multicast video.
 */
 
static void ip_encap(struct sk_buff *skb, u32 saddr, u32 daddr)
{
	struct iphdr *iph = (struct iphdr *)skb_push(skb,sizeof(struct iphdr));

	iph->version	= 	4;
	iph->tos	=	skb->nh.iph->tos;
	iph->ttl	=	skb->nh.iph->ttl;
	iph->frag_off	=	0;
	iph->daddr	=	daddr;
	iph->saddr	=	saddr;
	iph->protocol	=	IPPROTO_IPIP;
	iph->ihl	=	5;
	iph->tot_len	=	htons(skb->len);
	ip_select_ident(iph, skb->dst);
	ip_send_check(iph);

	skb->h.ipiph = skb->nh.iph;
	skb->nh.iph = iph;
#ifdef CONFIG_NETFILTER
	nf_conntrack_put(skb->nfct);
	skb->nfct = NULL;
#endif
}

static inline int ipmr_forward_finish(struct sk_buff *skb)
{
	struct dst_entry *dst = skb->dst;

	if (skb->len <= dst->pmtu)
		return dst->output(skb);
	else
		return ip_fragment(skb, dst->output);
}

/*
 *	Processing handlers for ipmr_forward
 */

static void ipmr_queue_xmit(struct sk_buff *skb, struct mfc_cache *c,
			   int vifi, int last)
{
	struct iphdr *iph = skb->nh.iph;
	struct vif_device *vif = &vif_table[vifi];
	struct net_device *dev;
	struct rtable *rt;
	int    encap = 0;
	struct sk_buff *skb2;

	if (vif->dev == NULL)
		return;

#ifdef CONFIG_IP_PIMSM
	if (vif->flags & VIFF_REGISTER) {
		vif->pkt_out++;
		vif->bytes_out+=skb->len;
		((struct net_device_stats*)vif->dev->priv)->tx_bytes += skb->len;
		((struct net_device_stats*)vif->dev->priv)->tx_packets++;
		ipmr_cache_report(skb, vifi, IGMPMSG_WHOLEPKT);
		return;
	}
#endif

	if (vif->flags&VIFF_TUNNEL) {
		if (ip_route_output(&rt, vif->remote, vif->local, RT_TOS(iph->tos), vif->link))
			return;
		encap = sizeof(struct iphdr);
	} else {
		if (ip_route_output(&rt, iph->daddr, 0, RT_TOS(iph->tos), vif->link))
			return;
	}

	dev = rt->u.dst.dev;

	if (skb->len+encap > rt->u.dst.pmtu && (ntohs(iph->frag_off) & IP_DF)) {
		/* Do not fragment multicasts. Alas, IPv4 does not
		   allow to send ICMP, so that packets will disappear
		   to blackhole.
		 */

		IP_INC_STATS_BH(IpFragFails);
		ip_rt_put(rt);
		return;
	}

	encap += dev->hard_header_len;

	if (skb_headroom(skb) < encap || skb_cloned(skb) || !last)