igmp.c

linux 内核源代码

/*
 *	Linux NET3:	Internet Group Management Protocol  [IGMP]
 *
 *	This code implements the IGMP protocol as defined in RFC1112. There has
 *	been a further revision of this protocol since which is now supported.
 *
 *	If you have trouble with this module be careful what gcc you have used,
 *	the older version didn't come out right using gcc 2.5.8, the newer one
 *	seems to fall out with gcc 2.6.2.
 *
 *	Version: $Id: igmp.c,v 1.47 2002/02/01 22:01:03 davem Exp $
 *
 *	Authors:
 *		Alan Cox <[e-mail]>
 *
 *	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.
 *
 *	Fixes:
 *
 *		Alan Cox	:	Added lots of __inline__ to optimise
 *					the memory usage of all the tiny little
 *					functions.
 *		Alan Cox	:	Dumped the header building experiment.
 *		Alan Cox	:	Minor tweaks ready for multicast routing
 *					and extended IGMP protocol.
 *		Alan Cox	:	Removed a load of inline directives. Gcc 2.5.8
 *					writes utterly bogus code otherwise (sigh)
 *					fixed IGMP loopback to behave in the manner
 *					desired by mrouted, fixed the fact it has been
 *					broken since 1.3.6 and cleaned up a few minor
 *					points.
 *
 *		Chih-Jen Chang	:	Tried to revise IGMP to Version 2
 *		Tsu-Sheng Tsao		E-mail: [e-mail] and [e-mail]
 *					The enhancements are mainly based on Steve Deering's
 * 					ipmulti-3.5 source code.
 *		Chih-Jen Chang	:	Added the igmp_get_mrouter_info and
 *		Tsu-Sheng Tsao		igmp_set_mrouter_info to keep track of
 *					the mrouted version on that device.
 *		Chih-Jen Chang	:	Added the max_resp_time parameter to
 *		Tsu-Sheng Tsao		igmp_heard_query(). Using this parameter
 *					to identify the multicast router version
 *					and do what the IGMP version 2 specified.
 *		Chih-Jen Chang	:	Added a timer to revert to IGMP V2 router
 *		Tsu-Sheng Tsao		if the specified time expired.
 *		Alan Cox	:	Stop IGMP from 0.0.0.0 being accepted.
 *		Alan Cox	:	Use GFP_ATOMIC in the right places.
 *		Christian Daudt :	igmp timer wasn't set for local group
 *					memberships but was being deleted,
 *					which caused a "del_timer() called
 *					from %p with timer not initialized\n"
 *					message (960131).
 *		Christian Daudt :	removed del_timer from
 *					igmp_timer_expire function (960205).
 *             Christian Daudt :       igmp_heard_report now only calls
 *                                     igmp_timer_expire if tm->running is
 *                                     true (960216).
 *		Malcolm Beattie :	ttl comparison wrong in igmp_rcv made
 *					igmp_heard_query never trigger. Expiry
 *					miscalculation fixed in igmp_heard_query
 *					and random() made to return unsigned to
 *					prevent negative expiry times.
 *		Alexey Kuznetsov:	Wrong group leaving behaviour, backport
 *					fix from pending 2.1.x patches.
 *		Alan Cox:		Forget to enable FDDI support earlier.
 *		Alexey Kuznetsov:	Fixed leaving groups on device down.
 *		Alexey Kuznetsov:	Accordance to igmp-v2-06 draft.
 *		David L Stevens:	IGMPv3 support, with help from
 *					Vinay Kulkarni
 */

#include <linux/module.h>
#include <asm/uaccess.h>
#include <asm/system.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/jiffies.h>
#include <linux/string.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/skbuff.h>
#include <linux/inetdevice.h>
#include <linux/igmp.h>
#include <linux/if_arp.h>
#include <linux/rtnetlink.h>
#include <linux/times.h>

#include <net/net_namespace.h>
#include <net/arp.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <net/route.h>
#include <net/sock.h>
#include <net/checksum.h>
#include <linux/netfilter_ipv4.h>
#ifdef CONFIG_IP_MROUTE
#include <linux/mroute.h>
#endif
#ifdef CONFIG_PROC_FS
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#endif

#define IP_MAX_MEMBERSHIPS	20
#define IP_MAX_MSF		10

#ifdef CONFIG_IP_MULTICAST
/* Parameter names and values are taken from igmp-v2-06 draft */

#define IGMP_V1_Router_Present_Timeout		(400*HZ)
#define IGMP_V2_Router_Present_Timeout		(400*HZ)
#define IGMP_Unsolicited_Report_Interval	(10*HZ)
#define IGMP_Query_Response_Interval		(10*HZ)
#define IGMP_Unsolicited_Report_Count		2


#define IGMP_Initial_Report_Delay		(1)

/* IGMP_Initial_Report_Delay is not from IGMP specs!
 * IGMP specs require to report membership immediately after
 * joining a group, but we delay the first report by a
 * small interval. It seems more natural and still does not
 * contradict to specs provided this delay is small enough.
 */

#define IGMP_V1_SEEN(in_dev) \
	(IPV4_DEVCONF_ALL(FORCE_IGMP_VERSION) == 1 || \
	 IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 1 || \
	 ((in_dev)->mr_v1_seen && \
	  time_before(jiffies, (in_dev)->mr_v1_seen)))
#define IGMP_V2_SEEN(in_dev) \
	(IPV4_DEVCONF_ALL(FORCE_IGMP_VERSION) == 2 || \
	 IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 2 || \
	 ((in_dev)->mr_v2_seen && \
	  time_before(jiffies, (in_dev)->mr_v2_seen)))

static void igmpv3_add_delrec(struct in_device *in_dev, struct ip_mc_list *im);
static void igmpv3_del_delrec(struct in_device *in_dev, __be32 multiaddr);
static void igmpv3_clear_delrec(struct in_device *in_dev);
static int sf_setstate(struct ip_mc_list *pmc);
static void sf_markstate(struct ip_mc_list *pmc);
#endif
static void ip_mc_clear_src(struct ip_mc_list *pmc);
static int ip_mc_add_src(struct in_device *in_dev, __be32 *pmca, int sfmode,
			 int sfcount, __be32 *psfsrc, int delta);

static void ip_ma_put(struct ip_mc_list *im)
{
	if (atomic_dec_and_test(&im->refcnt)) {
		in_dev_put(im->interface);
		kfree(im);
	}
}

#ifdef CONFIG_IP_MULTICAST

/*
 *	Timer management
 */

static __inline__ void igmp_stop_timer(struct ip_mc_list *im)
{
	spin_lock_bh(&im->lock);
	if (del_timer(&im->timer))
		atomic_dec(&im->refcnt);
	im->tm_running=0;
	im->reporter = 0;
	im->unsolicit_count = 0;
	spin_unlock_bh(&im->lock);
}

/* It must be called with locked im->lock */
static void igmp_start_timer(struct ip_mc_list *im, int max_delay)
{
	int tv=net_random() % max_delay;

	im->tm_running=1;
	if (!mod_timer(&im->timer, jiffies+tv+2))
		atomic_inc(&im->refcnt);
}

static void igmp_gq_start_timer(struct in_device *in_dev)
{
	int tv = net_random() % in_dev->mr_maxdelay;

	in_dev->mr_gq_running = 1;
	if (!mod_timer(&in_dev->mr_gq_timer, jiffies+tv+2))
		in_dev_hold(in_dev);
}

static void igmp_ifc_start_timer(struct in_device *in_dev, int delay)
{
	int tv = net_random() % delay;

	if (!mod_timer(&in_dev->mr_ifc_timer, jiffies+tv+2))
		in_dev_hold(in_dev);
}

static void igmp_mod_timer(struct ip_mc_list *im, int max_delay)
{
	spin_lock_bh(&im->lock);
	im->unsolicit_count = 0;
	if (del_timer(&im->timer)) {
		if ((long)(im->timer.expires-jiffies) < max_delay) {
			add_timer(&im->timer);
			im->tm_running=1;
			spin_unlock_bh(&im->lock);
			return;
		}
		atomic_dec(&im->refcnt);
	}
	igmp_start_timer(im, max_delay);
	spin_unlock_bh(&im->lock);
}


/*
 *	Send an IGMP report.
 */

#define IGMP_SIZE (sizeof(struct igmphdr)+sizeof(struct iphdr)+4)


static int is_in(struct ip_mc_list *pmc, struct ip_sf_list *psf, int type,
	int gdeleted, int sdeleted)
{
	switch (type) {
	case IGMPV3_MODE_IS_INCLUDE:
	case IGMPV3_MODE_IS_EXCLUDE:
		if (gdeleted || sdeleted)
			return 0;
		if (!(pmc->gsquery && !psf->sf_gsresp)) {
			if (pmc->sfmode == MCAST_INCLUDE)
				return 1;
			/* don't include if this source is excluded
			 * in all filters
			 */
			if (psf->sf_count[MCAST_INCLUDE])
				return type == IGMPV3_MODE_IS_INCLUDE;
			return pmc->sfcount[MCAST_EXCLUDE] ==
				psf->sf_count[MCAST_EXCLUDE];
		}
		return 0;
	case IGMPV3_CHANGE_TO_INCLUDE:
		if (gdeleted || sdeleted)
			return 0;
		return psf->sf_count[MCAST_INCLUDE] != 0;
	case IGMPV3_CHANGE_TO_EXCLUDE:
		if (gdeleted || sdeleted)
			return 0;
		if (pmc->sfcount[MCAST_EXCLUDE] == 0 ||
		    psf->sf_count[MCAST_INCLUDE])
			return 0;
		return pmc->sfcount[MCAST_EXCLUDE] ==
			psf->sf_count[MCAST_EXCLUDE];
	case IGMPV3_ALLOW_NEW_SOURCES:
		if (gdeleted || !psf->sf_crcount)
			return 0;
		return (pmc->sfmode == MCAST_INCLUDE) ^ sdeleted;
	case IGMPV3_BLOCK_OLD_SOURCES:
		if (pmc->sfmode == MCAST_INCLUDE)
			return gdeleted || (psf->sf_crcount && sdeleted);
		return psf->sf_crcount && !gdeleted && !sdeleted;
	}
	return 0;
}

static int
igmp_scount(struct ip_mc_list *pmc, int type, int gdeleted, int sdeleted)
{
	struct ip_sf_list *psf;
	int scount = 0;

	for (psf=pmc->sources; psf; psf=psf->sf_next) {
		if (!is_in(pmc, psf, type, gdeleted, sdeleted))
			continue;
		scount++;
	}
	return scount;
}

static struct sk_buff *igmpv3_newpack(struct net_device *dev, int size)
{
	struct sk_buff *skb;
	struct rtable *rt;
	struct iphdr *pip;
	struct igmpv3_report *pig;

	skb = alloc_skb(size + LL_RESERVED_SPACE(dev), GFP_ATOMIC);
	if (skb == NULL)
		return NULL;

	{
		struct flowi fl = { .oif = dev->ifindex,
				    .nl_u = { .ip4_u = {
				    .daddr = IGMPV3_ALL_MCR } },
				    .proto = IPPROTO_IGMP };
		if (ip_route_output_key(&rt, &fl)) {
			kfree_skb(skb);
			return NULL;
		}
	}
	if (rt->rt_src == 0) {
		kfree_skb(skb);
		ip_rt_put(rt);
		return NULL;
	}

	skb->dst = &rt->u.dst;
	skb->dev = dev;

	skb_reserve(skb, LL_RESERVED_SPACE(dev));

	skb_reset_network_header(skb);
	pip = ip_hdr(skb);
	skb_put(skb, sizeof(struct iphdr) + 4);

	pip->version  = 4;
	pip->ihl      = (sizeof(struct iphdr)+4)>>2;
	pip->tos      = 0xc0;
	pip->frag_off = htons(IP_DF);
	pip->ttl      = 1;
	pip->daddr    = rt->rt_dst;
	pip->saddr    = rt->rt_src;
	pip->protocol = IPPROTO_IGMP;
	pip->tot_len  = 0;	/* filled in later */
	ip_select_ident(pip, &rt->u.dst, NULL);
	((u8*)&pip[1])[0] = IPOPT_RA;
	((u8*)&pip[1])[1] = 4;
	((u8*)&pip[1])[2] = 0;
	((u8*)&pip[1])[3] = 0;

	skb->transport_header = skb->network_header + sizeof(struct iphdr) + 4;
	skb_put(skb, sizeof(*pig));
	pig = igmpv3_report_hdr(skb);
	pig->type = IGMPV3_HOST_MEMBERSHIP_REPORT;
	pig->resv1 = 0;
	pig->csum = 0;
	pig->resv2 = 0;
	pig->ngrec = 0;
	return skb;
}

static int igmpv3_sendpack(struct sk_buff *skb)
{
	struct iphdr *pip = ip_hdr(skb);
	struct igmphdr *pig = igmp_hdr(skb);
	const int iplen = skb->tail - skb->network_header;
	const int igmplen = skb->tail - skb->transport_header;

	pip->tot_len = htons(iplen);
	ip_send_check(pip);
	pig->csum = ip_compute_csum(igmp_hdr(skb), igmplen);

	return NF_HOOK(PF_INET, NF_IP_LOCAL_OUT, skb, NULL, skb->dev,
		       dst_output);
}

static int grec_size(struct ip_mc_list *pmc, int type, int gdel, int sdel)
{
	return sizeof(struct igmpv3_grec) + 4*igmp_scount(pmc,type,gdel,sdel);
}

static struct sk_buff *add_grhead(struct sk_buff *skb, struct ip_mc_list *pmc,
	int type, struct igmpv3_grec **ppgr)
{
	struct net_device *dev = pmc->interface->dev;
	struct igmpv3_report *pih;
	struct igmpv3_grec *pgr;

	if (!skb)
		skb = igmpv3_newpack(dev, dev->mtu);
	if (!skb)
		return NULL;
	pgr = (struct igmpv3_grec *)skb_put(skb, sizeof(struct igmpv3_grec));
	pgr->grec_type = type;
	pgr->grec_auxwords = 0;
	pgr->grec_nsrcs = 0;
	pgr->grec_mca = pmc->multiaddr;
	pih = igmpv3_report_hdr(skb);
	pih->ngrec = htons(ntohs(pih->ngrec)+1);
	*ppgr = pgr;
	return skb;
}

#define AVAILABLE(skb) ((skb) ? ((skb)->dev ? (skb)->dev->mtu - (skb)->len : \
	skb_tailroom(skb)) : 0)

static struct sk_buff *add_grec(struct sk_buff *skb, struct ip_mc_list *pmc,
	int type, int gdeleted, int sdeleted)
{
	struct net_device *dev = pmc->interface->dev;
	struct igmpv3_report *pih;
	struct igmpv3_grec *pgr = NULL;
	struct ip_sf_list *psf, *psf_next, *psf_prev, **psf_list;
	int scount, stotal, first, isquery, truncate;

	if (pmc->multiaddr == IGMP_ALL_HOSTS)
		return skb;

	isquery = type == IGMPV3_MODE_IS_INCLUDE ||
		  type == IGMPV3_MODE_IS_EXCLUDE;
	truncate = type == IGMPV3_MODE_IS_EXCLUDE ||
		    type == IGMPV3_CHANGE_TO_EXCLUDE;

	stotal = scount = 0;

	psf_list = sdeleted ? &pmc->tomb : &pmc->sources;

	if (!*psf_list)
		goto empty_source;

	pih = skb ? igmpv3_report_hdr(skb) : NULL;

	/* EX and TO_EX get a fresh packet, if needed */
	if (truncate) {
		if (pih && pih->ngrec &&
		    AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) {
			if (skb)
				igmpv3_sendpack(skb);
			skb = igmpv3_newpack(dev, dev->mtu);
		}
	}
	first = 1;
	psf_prev = NULL;
	for (psf=*psf_list; psf; psf=psf_next) {
		__be32 *psrc;

		psf_next = psf->sf_next;

		if (!is_in(pmc, psf, type, gdeleted, sdeleted)) {
			psf_prev = psf;
			continue;
		}

		/* clear marks on query responses */
		if (isquery)
			psf->sf_gsresp = 0;

		if (AVAILABLE(skb) < sizeof(__be32) +
		    first*sizeof(struct igmpv3_grec)) {
			if (truncate && !first)
				break;	 /* truncate these */
			if (pgr)
				pgr->grec_nsrcs = htons(scount);
			if (skb)
				igmpv3_sendpack(skb);
			skb = igmpv3_newpack(dev, dev->mtu);
			first = 1;
			scount = 0;
		}
		if (first) {
			skb = add_grhead(skb, pmc, type, &pgr);
			first = 0;
		}
		if (!skb)
			return NULL;
		psrc = (__be32 *)skb_put(skb, sizeof(__be32));
		*psrc = psf->sf_inaddr;
		scount++; stotal++;
		if ((type == IGMPV3_ALLOW_NEW_SOURCES ||
		     type == IGMPV3_BLOCK_OLD_SOURCES) && psf->sf_crcount) {
			psf->sf_crcount--;
			if ((sdeleted || gdeleted) && psf->sf_crcount == 0) {
				if (psf_prev)
					psf_prev->sf_next = psf->sf_next;
				else
					*psf_list = psf->sf_next;
				kfree(psf);
				continue;
			}
		}
		psf_prev = psf;
	}

empty_source:
	if (!stotal) {
		if (type == IGMPV3_ALLOW_NEW_SOURCES ||
		    type == IGMPV3_BLOCK_OLD_SOURCES)
			return skb;
		if (pmc->crcount || isquery) {
			/* make sure we have room for group header */
			if (skb && AVAILABLE(skb)<sizeof(struct igmpv3_grec)) {
				igmpv3_sendpack(skb);
				skb = NULL; /* add_grhead will get a new one */
			}
			skb = add_grhead(skb, pmc, type, &pgr);
		}
	}
	if (pgr)
		pgr->grec_nsrcs = htons(scount);

	if (isquery)
		pmc->gsquery = 0;	/* clear query state on report */
	return skb;
}

static int igmpv3_send_report(struct in_device *in_dev, struct ip_mc_list *pmc)
{
	struct sk_buff *skb = NULL;
	int type;

	if (!pmc) {
		read_lock(&in_dev->mc_list_lock);
		for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
			if (pmc->multiaddr == IGMP_ALL_HOSTS)
				continue;
			spin_lock_bh(&pmc->lock);
			if (pmc->sfcount[MCAST_EXCLUDE])
				type = IGMPV3_MODE_IS_EXCLUDE;
			else
				type = IGMPV3_MODE_IS_INCLUDE;
			skb = add_grec(skb, pmc, type, 0, 0);
			spin_unlock_bh(&pmc->lock);