Merge bk://kernel.bkbits.net/davem/net-2.5

into home.transmeta.com:/home/torvalds/v2.5/linux

15 files changed, 385 insertions, 135 deletions

diff --git a/include/linux/igmp.h b/include/linux/igmp.h
index 2cc2472d2eb2..8144cabdad06 100644
--- a/include/linux/igmp.h
+++ b/include/linux/igmp.h
@@ -191,7 +191,7 @@ struct ip_mc_list
          (IGMPV3_MASK((value) >> (nbmant), nbexp) + (nbexp))))
 
 #define IGMPV3_QQIC(value) IGMPV3_EXP(0x80, 4, 3, value)
-#define IGMPV3_MRC(value) IGMPV3_EXP(0x8000, 12, 3, value)
+#define IGMPV3_MRC(value) IGMPV3_EXP(0x80, 4, 3, value)
 
 extern int ip_check_mc(struct in_device *dev, u32 mc_addr, u32 src_addr, u16 proto);
 extern int igmp_rcv(struct sk_buff *);
@@ -199,10 +199,12 @@ extern int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr);
 extern int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr);
 extern void ip_mc_drop_socket(struct sock *sk);
 extern int ip_mc_source(int add, int omode, struct sock *sk,
-		struct ip_mreq_source *mreqs);
-extern int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf);
+		struct ip_mreq_source *mreqs, int ifindex);
+extern int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf,int ifindex);
 extern int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
 		struct ip_msfilter *optval, int *optlen);
+extern int ip_mc_gsfget(struct sock *sk, struct group_filter *gsf,
+		struct group_filter *optval, int *optlen);
 extern int ip_mc_sf_allow(struct sock *sk, u32 local, u32 rmt, int dif);
 extern void ip_mr_init(void);
 extern void ip_mc_init_dev(struct in_device *);
diff --git a/include/linux/net.h b/include/linux/net.h
index fc6b655dcdc3..8b012430f49d 100644
--- a/include/linux/net.h
+++ b/include/linux/net.h
@@ -140,6 +140,9 @@ struct net_proto_family {
 	struct module	*owner;
 };
 
+extern int	     net_family_get(int family);
+extern void	     net_family_put(int family);
+
 struct iovec;
 
 extern int	     sock_wake_async(struct socket *sk, int how, int band);
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 1a5c99c5afa2..1af054ea6385 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -59,7 +59,7 @@ extern struct rt6_info		*rt6_lookup(struct in6_addr *daddr,
 					    struct in6_addr *saddr,
 					    int oif, int flags);
 
-extern struct rt6_info		*ndisc_get_dummy_rt(void);
+extern struct rt6_info		*ip6_dst_alloc(void);
 
 /*
  *	support functions for ND
diff --git a/net/core/dev.c b/net/core/dev.c
index b6f627be7b08..9c041e1be73e 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2582,6 +2582,16 @@ int register_netdevice(struct net_device *dev)
 	if ((ret = kobject_register(&dev->kobj)))
 		goto out_err;
 	
+	/* Fix illegal SG+CSUM combinations. */
+	if ((dev->features & NETIF_F_SG) &&
+	    !(dev->features & (NETIF_F_IP_CSUM |
+			       NETIF_F_NO_CSUM |
+			       NETIF_F_HW_CSUM))) {
+		printk("%s: Dropping NETIF_F_SG since no checksum feature.\n",
+		       dev->name);
+		dev->features &= ~NETIF_F_SG;
+	}
+
 	/*
 	 *	nil rebuild_header routine,
 	 *	that should be never called and used as just bug trap.
diff --git a/net/core/sock.c b/net/core/sock.c
index 8c85f0d79a86..29fdc583abe3 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -589,8 +589,10 @@ static kmem_cache_t *sk_cachep;
  */
 struct sock *sk_alloc(int family, int priority, int zero_it, kmem_cache_t *slab)
 {
-	struct sock *sk;
-       
+	struct sock *sk = NULL;
+
+	if (!net_family_get(family))
+		goto out;
 	if (!slab)
 		slab = sk_cachep;
 	sk = kmem_cache_alloc(slab, priority);
@@ -602,14 +604,16 @@ struct sock *sk_alloc(int family, int priority, int zero_it, kmem_cache_t *slab)
 			sock_lock_init(sk);
 		}
 		sk->slab = slab;
-	}
-
+	} else
+		net_family_put(family);
+out:
 	return sk;
 }
 
 void sk_free(struct sock *sk)
 {
 	struct sk_filter *filter;
+	const int family = sk->family;
 
 	if (sk->destruct)
 		sk->destruct(sk);
@@ -624,6 +628,7 @@ void sk_free(struct sock *sk)
 		printk(KERN_DEBUG "sk_free: optmem leakage (%d bytes) detected.\n", atomic_read(&sk->omem_alloc));
 
 	kmem_cache_free(sk->slab, sk);
+	net_family_put(family);
 }
 
 void __init sk_init(void)
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 265eb0604436..680ab028ad69 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -103,6 +103,7 @@
 
 #define IP_MAX_MEMBERSHIPS 20
 
+#ifdef CONFIG_IP_MULTICAST
 /* Parameter names and values are taken from igmp-v2-06 draft */
 
 #define IGMP_V1_Router_Present_Timeout		(400*HZ)
@@ -126,13 +127,12 @@
 #define IGMP_V2_SEEN(in_dev) ((in_dev)->mr_v2_seen && \
 		time_before(jiffies, (in_dev)->mr_v2_seen))
 
-#ifdef CONFIG_MULTICAST
 static void igmpv3_add_delrec(struct in_device *in_dev, struct ip_mc_list *im);
-#endif
 static void igmpv3_del_delrec(struct in_device *in_dev, __u32 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);
 int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 	int sfcount, __u32 *psfsrc, int delta);
@@ -770,11 +770,18 @@ static void igmp_heard_query(struct in_device *in_dev, struct igmphdr *ih,
 			in_dev->mr_v2_seen = jiffies +
 				IGMP_V2_Router_Present_Timeout;
 		}
+		/* cancel the interface change timer */
+		in_dev->mr_ifc_count = 0;
+		if (del_timer(&in_dev->mr_ifc_timer))
+			atomic_dec(&in_dev->refcnt);
+		/* clear deleted report items */
 		igmpv3_clear_delrec(in_dev);
 	} else if (len < 12) {
 		return;	/* ignore bogus packet; freed by caller */
 	} else { /* v3 */
 		max_delay = IGMPV3_MRC(ih3->code)*(HZ/IGMP_TIMER_SCALE);
+		if (!max_delay)
+			max_delay = 1;	/* can't mod w/ 0 */
 		in_dev->mr_maxdelay = max_delay;
 		if (ih3->qrv)
 			in_dev->mr_qrv = ih3->qrv;
@@ -951,7 +958,6 @@ static void igmpv3_add_delrec(struct in_device *in_dev, struct ip_mc_list *im)
 	in_dev->mc_tomb = pmc;
 	write_unlock_bh(&in_dev->mc_lock);
 }
-#endif
 
 static void igmpv3_del_delrec(struct in_device *in_dev, __u32 multiaddr)
 {
@@ -997,7 +1003,23 @@ static void igmpv3_clear_delrec(struct in_device *in_dev)
 		in_dev_put(pmc->interface);
 		kfree(pmc);
 	}
+	/* clear dead sources, too */
+	read_lock(&in_dev->lock);
+	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
+		struct ip_sf_list *psf, *psf_next;
+
+		spin_lock_bh(&pmc->lock);
+		psf = pmc->tomb;
+		pmc->tomb = 0;
+		spin_unlock_bh(&pmc->lock);
+		for (; psf; psf=psf_next) {
+			psf_next = psf->sf_next;
+			kfree(psf);
+		}
+	}
+	read_unlock(&in_dev->lock);
 }
+#endif
 
 static void igmp_group_dropped(struct ip_mc_list *im)
 {
@@ -1030,8 +1052,8 @@ static void igmp_group_dropped(struct ip_mc_list *im)
 
 	igmp_ifc_event(in_dev);
 done:
-	ip_mc_clear_src(im);
 #endif
+	ip_mc_clear_src(im);
 }
 
 static void igmp_group_added(struct ip_mc_list *im)
@@ -1102,7 +1124,7 @@ void ip_mc_inc_group(struct in_device *in_dev, u32 addr)
 	im->crcount = 0;
 	atomic_set(&im->refcnt, 1);
 	spin_lock_init(&im->lock);
-#ifdef  CONFIG_IP_MULTICAST
+#ifdef CONFIG_IP_MULTICAST
 	im->tm_running=0;
 	init_timer(&im->timer);
 	im->timer.data=(unsigned long)im;
@@ -1116,7 +1138,9 @@ void ip_mc_inc_group(struct in_device *in_dev, u32 addr)
 	im->next=in_dev->mc_list;
 	in_dev->mc_list=im;
 	write_unlock_bh(&in_dev->lock);
+#ifdef CONFIG_IP_MULTICAST
 	igmpv3_del_delrec(in_dev, im->multiaddr);
+#endif
 	igmp_group_added(im);
 	if (in_dev->dev->flags & IFF_UP)
 		ip_rt_multicast_event(in_dev);
@@ -1173,7 +1197,9 @@ void ip_mc_down(struct in_device *in_dev)
 	for (i=in_dev->mc_list; i; i=i->next)
 		igmp_group_dropped(i);
 
+#ifdef CONFIG_IP_MULTICAST
 	igmpv3_clear_delrec(in_dev);
+#endif
 
 	ip_mc_dec_group(in_dev, IGMP_ALL_HOSTS);
 }
@@ -1186,12 +1212,12 @@ void ip_mc_up(struct in_device *in_dev)
 
 	ASSERT_RTNL();
 
+	in_dev->mc_tomb = 0;
 #ifdef CONFIG_IP_MULTICAST
 	in_dev->mr_gq_running = 0;
 	init_timer(&in_dev->mr_gq_timer);
 	in_dev->mr_gq_timer.data=(unsigned long) in_dev;
 	in_dev->mr_gq_timer.function=&igmp_gq_timer_expire;
-	in_dev->mc_tomb = 0;
 	in_dev->mr_ifc_count = 0;
 	init_timer(&in_dev->mr_ifc_timer);
 	in_dev->mr_ifc_timer.data=(unsigned long) in_dev;
@@ -1237,6 +1263,12 @@ static struct in_device * ip_mc_find_dev(struct ip_mreqn *imr)
 	struct net_device *dev = NULL;
 	struct in_device *idev = NULL;
 
+	if (imr->imr_ifindex) {
+		idev = inetdev_by_index(imr->imr_ifindex);
+		if (idev)
+			__in_dev_put(idev);
+		return idev;
+	}
 	if (imr->imr_address.s_addr) {
 		dev = ip_dev_find(imr->imr_address.s_addr);
 		if (!dev)
@@ -1282,13 +1314,16 @@ static int ip_mc_del1_src(struct ip_mc_list *pmc, int sfmode,
 		ip_rt_multicast_event(pmc->interface);
 	}
 	if (!psf->sf_count[MCAST_INCLUDE] && !psf->sf_count[MCAST_EXCLUDE]) {
+#ifdef CONFIG_IP_MULTICAST
 		struct in_device *in_dev = pmc->interface;
+#endif
 
 		/* no more filters for this source */
 		if (psf_prev)
 			psf_prev->sf_next = psf->sf_next;
 		else
 			pmc->sources = psf->sf_next;
+#ifdef CONFIG_IP_MULTICAST
 		if (psf->sf_oldin &&
 		    !IGMP_V1_SEEN(in_dev) && !IGMP_V2_SEEN(in_dev)) {
 			psf->sf_crcount = in_dev->mr_qrv ? in_dev->mr_qrv : 
@@ -1297,6 +1332,7 @@ static int ip_mc_del1_src(struct ip_mc_list *pmc, int sfmode,
 			pmc->tomb = psf;
 			rv = 1;
 		} else
+#endif
 			kfree(psf);
 	}
 	return rv;
@@ -1327,7 +1363,9 @@ int ip_mc_del_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 	}
 	spin_lock_bh(&pmc->lock);
 	read_unlock(&in_dev->lock);
+#ifdef CONFIG_IP_MULTICAST
 	sf_markstate(pmc);
+#endif
 	if (!delta) {
 		if (!pmc->sfcount[sfmode])
 			return -EINVAL;
@@ -1344,10 +1382,13 @@ int ip_mc_del_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 	if (pmc->sfmode == MCAST_EXCLUDE &&
 	    pmc->sfcount[MCAST_EXCLUDE] == 0 &&
 	    pmc->sfcount[MCAST_INCLUDE]) {
+#ifdef CONFIG_IP_MULTICAST
 		struct ip_sf_list *psf;
+#endif
 
 		/* filter mode change */
 		pmc->sfmode = MCAST_INCLUDE;
+#ifdef CONFIG_IP_MULTICAST
 		pmc->crcount = in_dev->mr_qrv ? in_dev->mr_qrv : 
 			IGMP_Unsolicited_Report_Count;
 		in_dev->mr_ifc_count = pmc->crcount;
@@ -1356,6 +1397,7 @@ int ip_mc_del_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 		igmp_ifc_event(pmc->interface);
 	} else if (sf_setstate(pmc) || changerec) {
 		igmp_ifc_event(pmc->interface);
+#endif
 	}
 	spin_unlock_bh(&pmc->lock);
 	return err;
@@ -1393,6 +1435,7 @@ static int ip_mc_add1_src(struct ip_mc_list *pmc, int sfmode,
 	return 0;
 }
 
+#ifdef CONFIG_IP_MULTICAST
 static void sf_markstate(struct ip_mc_list *pmc)
 {
 	struct ip_sf_list *psf;
@@ -1428,6 +1471,7 @@ static int sf_setstate(struct ip_mc_list *pmc)
 	}
 	return rv;
 }
+#endif
 
 /*
  * Add multicast source filter list to the interface list
@@ -1454,7 +1498,9 @@ int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 	spin_lock_bh(&pmc->lock);
 	read_unlock(&in_dev->lock);
 
+#ifdef CONFIG_IP_MULTICAST
 	sf_markstate(pmc);
+#endif
 	isexclude = pmc->sfmode == MCAST_EXCLUDE;
 	if (!delta)
 		pmc->sfcount[sfmode]++;
@@ -1471,14 +1517,17 @@ int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 		for (j=0; j<i; j++)
 			(void) ip_mc_del1_src(pmc, sfmode, &psfsrc[i]);
 	} else if (isexclude != (pmc->sfcount[MCAST_EXCLUDE] != 0)) {
+#ifdef CONFIG_IP_MULTICAST
 		struct in_device *in_dev = pmc->interface;
 		struct ip_sf_list *psf;
+#endif
 
 		/* filter mode change */
 		if (pmc->sfcount[MCAST_EXCLUDE])
 			pmc->sfmode = MCAST_EXCLUDE;
 		else if (pmc->sfcount[MCAST_INCLUDE])
 			pmc->sfmode = MCAST_INCLUDE;
+#ifdef CONFIG_IP_MULTICAST
 		/* else no filters; keep old mode for reports */
 
 		pmc->crcount = in_dev->mr_qrv ? in_dev->mr_qrv : 
@@ -1487,8 +1536,10 @@ int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 		for (psf=pmc->sources; psf; psf = psf->sf_next)
 			psf->sf_crcount = 0;
 		igmp_ifc_event(in_dev);
-	} else if (sf_setstate(pmc))
+	} else if (sf_setstate(pmc)) {
 		igmp_ifc_event(in_dev);
+#endif
+	}
 	spin_unlock_bh(&pmc->lock);
 	return err;
 }
@@ -1530,13 +1581,7 @@ int ip_mc_join_group(struct sock *sk , struct ip_mreqn *imr)
 
 	rtnl_shlock();
 
-	if (!imr->imr_ifindex)
-		in_dev = ip_mc_find_dev(imr);
-	else {
-		in_dev = inetdev_by_index(imr->imr_ifindex);
-		if (in_dev)
-			__in_dev_put(in_dev);
-	}
+	in_dev = ip_mc_find_dev(imr);
 
 	if (!in_dev) {
 		iml = NULL;
@@ -1638,13 +1683,13 @@ int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr)
 }
 
 int ip_mc_source(int add, int omode, struct sock *sk, struct
-	ip_mreq_source *mreqs)
+	ip_mreq_source *mreqs, int ifindex)
 {
 	int err;
-	struct ip_mreqn	imr;
+	struct ip_mreqn imr;
 	u32 addr = mreqs->imr_multiaddr;
 	struct ip_mc_socklist *pmc;
-	struct in_device *in_dev;
+	struct in_device *in_dev = 0;
 	struct inet_opt *inet = inet_sk(sk);
 	struct ip_sf_socklist *psl;
 	int i, j, rv;
@@ -1656,7 +1701,7 @@ int ip_mc_source(int add, int omode, struct sock *sk, struct
 
 	imr.imr_multiaddr.s_addr = mreqs->imr_multiaddr;
 	imr.imr_address.s_addr = mreqs->imr_interface;
-	imr.imr_ifindex = 0;
+	imr.imr_ifindex = ifindex;
 	in_dev = ip_mc_find_dev(&imr);
 
 	if (!in_dev) {
@@ -1753,7 +1798,7 @@ done:
 	return err;
 }
 
-int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf)
+int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf, int ifindex)
 {
 	int err;
 	struct ip_mreqn	imr;
@@ -1773,7 +1818,7 @@ int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf)
 
 	imr.imr_multiaddr.s_addr = msf->imsf_multiaddr;
 	imr.imr_address.s_addr = msf->imsf_interface;
-	imr.imr_ifindex = 0;
+	imr.imr_ifindex = ifindex;
 	in_dev = ip_mc_find_dev(&imr);
 
 	if (!in_dev) {
@@ -1783,7 +1828,8 @@ int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf)
 	err = -EADDRNOTAVAIL;
 
 	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
-		if (memcmp(&pmc->multi, &imr, sizeof(imr)) == 0)
+		if (pmc->multi.imr_multiaddr.s_addr == msf->imsf_multiaddr &&
+		    pmc->multi.imr_ifindex == imr.imr_ifindex)
 			break;
 	}
 	if (!pmc)		/* must have a prior join */
@@ -1834,9 +1880,6 @@ int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
 
 	if (!MULTICAST(addr))
 		return -EINVAL;
-	if (msf->imsf_fmode != MCAST_INCLUDE &&
-	    msf->imsf_fmode != MCAST_EXCLUDE)
-		return -EINVAL;
 
 	rtnl_shlock();
 
@@ -1852,7 +1895,8 @@ int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
 	err = -EADDRNOTAVAIL;
 
 	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
-		if (memcmp(&pmc->multi, &imr, sizeof(imr)) == 0)
+		if (pmc->multi.imr_multiaddr.s_addr == msf->imsf_multiaddr &&
+		    pmc->multi.imr_ifindex == imr.imr_ifindex)
 			break;
 	}
 	if (!pmc)		/* must have a prior join */
@@ -1882,6 +1926,61 @@ done:
 	return err;
 }
 
+int ip_mc_gsfget(struct sock *sk, struct group_filter *gsf,
+	struct group_filter *optval, int *optlen)
+{
+	int err, i, count, copycount;
+	struct sockaddr_in *psin;
+	u32 addr;
+	struct ip_mc_socklist *pmc;
+	struct inet_opt *inet = inet_sk(sk);
+	struct ip_sf_socklist *psl;
+
+	psin = (struct sockaddr_in *)&gsf->gf_group;
+	if (psin->sin_family != AF_INET)
+		return -EINVAL;
+	addr = psin->sin_addr.s_addr;
+	if (!MULTICAST(addr))
+		return -EINVAL;
+
+	rtnl_shlock();
+
+	err = -EADDRNOTAVAIL;
+
+	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
+		if (pmc->multi.imr_multiaddr.s_addr == addr &&
+		    pmc->multi.imr_ifindex == gsf->gf_interface)
+			break;
+	}
+	if (!pmc)		/* must have a prior join */
+		goto done;
+	gsf->gf_fmode = pmc->sfmode;
+	psl = pmc->sflist;
+	rtnl_shunlock();
+	count = psl ? psl->sl_count : 0;
+	copycount = count < gsf->gf_numsrc ? count : gsf->gf_numsrc;
+	gsf->gf_numsrc = count;
+	if (put_user(GROUP_FILTER_SIZE(copycount), optlen) ||
+	    copy_to_user((void *)optval, gsf, GROUP_FILTER_SIZE(0))) {
+		return -EFAULT;
+	}
+	for (i=0; i<copycount; i++) {
+		struct sockaddr_in *psin;
+		struct sockaddr_storage ss;
+
+		psin = (struct sockaddr_in *)&ss;
+		memset(&ss, 0, sizeof(ss));
+		psin->sin_family = AF_INET;
+		psin->sin_addr.s_addr = psl->sl_addr[i];
+		if (copy_to_user((void *)&optval->gf_slist[i], &ss, sizeof(ss)))
+			return -EFAULT;
+	}
+	return 0;
+done:
+	rtnl_shunlock();
+	return err;
+}
+
 /*
  * check if a multicast source filter allows delivery for a given <src,dst,intf>
  */
@@ -1892,6 +1991,9 @@ int ip_mc_sf_allow(struct sock *sk, u32 loc_addr, u32 rmt_addr, int dif)
 	struct ip_sf_socklist *psl;
 	int i;
 
+	if (!MULTICAST(loc_addr))
+		return 1;
+
 	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
 		if (pmc->multi.imr_multiaddr.s_addr == loc_addr &&
 		    pmc->multi.imr_ifindex == dif)
@@ -1972,8 +2074,6 @@ int ip_check_mc(struct in_device *in_dev, u32 mc_addr, u32 src_addr, u16 proto)
 }
 
 
-#ifdef CONFIG_IP_MULTICAST
- 
 int ip_mc_procinfo(char *buffer, char **start, off_t offset, int length)
 {
 	off_t pos=0, begin=0;
@@ -1991,7 +2091,9 @@ int ip_mc_procinfo(char *buffer, char **start, off_t offset, int length)
 		if (in_dev == NULL)
 			continue;
 
+#ifdef CONFIG_IP_MULTICAST
 		querier = IGMP_V1_SEEN(in_dev) ? "V1" : "V2";
+#endif
 
 		len+=sprintf(buffer+len,"%d\t%-10s: %5d %7s\n",
 			     dev->ifindex, dev->name, dev->mc_count, querier);
@@ -2049,11 +2151,8 @@ int ip_mcf_procinfo(char *buffer, char **start, off_t offset, int length)
 
 		for (imc=in_dev->mc_list; imc; imc=imc->next) {
 			struct ip_sf_list *psf;
-			unsigned long icount, xcount;
 
 			spin_lock_bh(&imc->lock);
-			icount = imc->sfcount[MCAST_INCLUDE];
-			xcount = imc->sfcount[MCAST_EXCLUDE];
 			for (psf=imc->sources; psf; psf=psf->sf_next) {
 				if (first) {
 					len += sprintf(buffer+len, "%3s %6s "
@@ -2080,33 +2179,6 @@ int ip_mcf_procinfo(char *buffer, char **start, off_t offset, int length)
 					in_dev_put(in_dev);
 					goto done;
 				}
-				icount -= psf->sf_count[MCAST_INCLUDE];
-				xcount -= psf->sf_count[MCAST_EXCLUDE];
-			}
-			if (icount > 0 || xcount > 0) {
-				if (first) {
-					len += sprintf(buffer+len, "%3s %6s "
-						"%10s %10s %6s %6s\n", "Idx",
-						"Device", "MCA", "SRC", "INC",
-						"EXC");
-					first = 0;
-				}
-				len += sprintf(buffer+len, "%3d %6.6s 0x%08x "
-					"%10s %6lu %6lu\n", dev->ifindex,
-					dev->name, ntohl(imc->multiaddr),
-					"NONE", icount, xcount);
-				pos=begin+len;
-				if(pos<offset)
-				{
-					len=0;
-					begin=pos;
-				}
-				if(pos>offset+length) {
-					spin_unlock_bh(&imc->lock);
-					read_unlock(&in_dev->lock);
-					in_dev_put(in_dev);
-					goto done;
-				}
 			}
 			spin_unlock_bh(&imc->lock);
 		}
@@ -2125,5 +2197,3 @@ done:
 	return len;
 }
 
-#endif
-
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 9a4db884dc1f..3f657dcde315 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -1312,6 +1312,6 @@ void __init ip_init(void)
 
 #ifdef CONFIG_IP_MULTICAST
 	proc_net_create("igmp", 0, ip_mc_procinfo);
-	proc_net_create("mcfilter", 0, ip_mcf_procinfo);
 #endif
+	proc_net_create("mcfilter", 0, ip_mcf_procinfo);
 }
diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
index a8a561411dea..82ab1fc2faa3 100644
--- a/net/ipv4/ip_sockglue.c
+++ b/net/ipv4/ip_sockglue.c
@@ -631,7 +631,7 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char *optval, int opt
 				kfree(msf);
 				break;
 			}
-			err = ip_mc_msfilter(sk, msf);
+			err = ip_mc_msfilter(sk, msf, 0);
 			kfree(msf);
 			break;
 		}
@@ -670,7 +670,142 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char *optval, int opt
 				omode = MCAST_INCLUDE;
 				add = 0;
 			}
-			err = ip_mc_source(add, omode, sk, &mreqs);
+			err = ip_mc_source(add, omode, sk, &mreqs, 0);
+			break;
+		}
+		case MCAST_JOIN_GROUP:
+		case MCAST_LEAVE_GROUP: 
+		{
+			struct group_req greq;
+			struct sockaddr_in *psin;
+			struct ip_mreqn mreq;
+
+			if (optlen < sizeof(struct group_req))
+				goto e_inval;
+			err = -EFAULT;
+			if(copy_from_user(&greq, optval, sizeof(greq)))
+				break;
+			psin = (struct sockaddr_in *)&greq.gr_group;
+			if (psin->sin_family != AF_INET)
+				goto e_inval;
+			memset(&mreq, 0, sizeof(mreq));
+			mreq.imr_multiaddr = psin->sin_addr;
+			mreq.imr_ifindex = greq.gr_interface;
+
+			if (optname == MCAST_JOIN_GROUP)
+				err = ip_mc_join_group(sk, &mreq);
+			else
+				err = ip_mc_leave_group(sk, &mreq);
+			break;
+		}
+		case MCAST_JOIN_SOURCE_GROUP:
+		case MCAST_LEAVE_SOURCE_GROUP:
+		case MCAST_BLOCK_SOURCE:
+		case MCAST_UNBLOCK_SOURCE:
+		{
+			struct group_source_req greqs;
+			struct ip_mreq_source mreqs;
+			struct sockaddr_in *psin;
+			int omode, add;
+
+			if (optlen != sizeof(struct group_source_req))
+				goto e_inval;
+			if (copy_from_user(&greqs, optval, sizeof(greqs))) {
+				err = -EFAULT;
+				break;
+			}
+			if (greqs.gsr_group.ss_family != AF_INET ||
+			    greqs.gsr_source.ss_family != AF_INET) {
+				err = -EADDRNOTAVAIL;
+				break;
+			}
+			psin = (struct sockaddr_in *)&greqs.gsr_group;
+			mreqs.imr_multiaddr = psin->sin_addr.s_addr;
+			psin = (struct sockaddr_in *)&greqs.gsr_source;
+			mreqs.imr_sourceaddr = psin->sin_addr.s_addr;
+			mreqs.imr_interface = 0; /* use index for mc_source */
+
+			if (optname == MCAST_BLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 1;
+			} else if (optname == MCAST_UNBLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 0;
+			} else if (optname == MCAST_JOIN_SOURCE_GROUP) {
+				struct ip_mreqn mreq;
+
+				psin = (struct sockaddr_in *)&greqs.gsr_group;
+				mreq.imr_multiaddr = psin->sin_addr;
+				mreq.imr_address.s_addr = 0;
+				mreq.imr_ifindex = greqs.gsr_interface;
+				err = ip_mc_join_group(sk, &mreq);
+				if (err)
+					break;
+				omode = MCAST_INCLUDE;
+				add = 1;
+			} else /* MCAST_LEAVE_SOURCE_GROUP */ {
+				omode = MCAST_INCLUDE;
+				add = 0;
+			}
+			err = ip_mc_source(add, omode, sk, &mreqs,
+				greqs.gsr_interface);
+			break;
+		}
+		case MCAST_MSFILTER:
+		{
+			struct sockaddr_in *psin;
+			struct ip_msfilter *msf = 0;
+			struct group_filter *gsf = 0;
+			int msize, i, ifindex;
+
+			if (optlen < GROUP_FILTER_SIZE(0))
+				goto e_inval;
+			gsf = (struct group_filter *)kmalloc(optlen,GFP_KERNEL);
+			if (gsf == 0) {
+				err = -ENOBUFS;
+				break;
+			}
+			err = -EFAULT;
+			if (copy_from_user(gsf, optval, optlen)) {
+				goto mc_msf_out;
+			}
+			if (GROUP_FILTER_SIZE(gsf->gf_numsrc) < optlen) {
+				err = EINVAL;
+				goto mc_msf_out;
+			}
+			msize = IP_MSFILTER_SIZE(gsf->gf_numsrc);
+			msf = (struct ip_msfilter *)kmalloc(msize,GFP_KERNEL);
+			if (msf == 0) {
+				err = -ENOBUFS;
+				goto mc_msf_out;
+			}
+			ifindex = gsf->gf_interface;
+			psin = (struct sockaddr_in *)&gsf->gf_group;
+			if (psin->sin_family != AF_INET) {
+				err = -EADDRNOTAVAIL;
+				goto mc_msf_out;
+			}
+			msf->imsf_multiaddr = psin->sin_addr.s_addr;
+			msf->imsf_interface = 0;
+			msf->imsf_fmode = gsf->gf_fmode;
+			msf->imsf_numsrc = gsf->gf_numsrc;
+			err = -EADDRNOTAVAIL;
+			for (i=0; i<gsf->gf_numsrc; ++i) {
+				psin = (struct sockaddr_in *)&gsf->gf_slist[i];
+
+				if (psin->sin_family != AF_INET)
+					goto mc_msf_out;
+				msf->imsf_slist[i] = psin->sin_addr.s_addr;
+			}
+			kfree(gsf);
+			gsf = 0;
+
+			err = ip_mc_msfilter(sk, msf, ifindex);
+mc_msf_out:
+			if (msf)
+				kfree(msf);
+			if (gsf)
+				kfree(gsf);
 			break;
 		}
 		case IP_ROUTER_ALERT:	
@@ -826,15 +961,37 @@ int ip_getsockopt(struct sock *sk, int level, int optname, char *optval, int *op
 			struct ip_msfilter msf;
 			int err;
 
-			if (len < IP_MSFILTER_SIZE(0))
+			if (len < IP_MSFILTER_SIZE(0)) {
+				release_sock(sk);
 				return -EINVAL;
-			if (copy_from_user(&msf, optval, IP_MSFILTER_SIZE(0)))
+			}
+			if (copy_from_user(&msf, optval, IP_MSFILTER_SIZE(0))) {
+				release_sock(sk);
 				return -EFAULT;
+			}
 			err = ip_mc_msfget(sk, &msf,
 				(struct ip_msfilter *)optval, optlen);
 			release_sock(sk);
 			return err;
 		}
+		case MCAST_MSFILTER:
+		{
+			struct group_filter gsf;
+			int err;
+
+			if (len < GROUP_FILTER_SIZE(0)) {
+				release_sock(sk);
+				return -EINVAL;
+			}
+			if (copy_from_user(&gsf, optval, GROUP_FILTER_SIZE(0))) {
+				release_sock(sk);
+				return -EFAULT;
+			}
+			err = ip_mc_gsfget(sk, &gsf,
+				(struct group_filter *)optval, optlen);
+			release_sock(sk);
+			return err;
+		}
 		case IP_PKTOPTIONS:		
 		{
 			struct msghdr msg;
diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
index 6ac0225988fd..d7c44f39a180 100644
--- a/net/ipv4/netfilter/Kconfig
+++ b/net/ipv4/netfilter/Kconfig
@@ -48,7 +48,7 @@ config IP_NF_IRC
 	  <file:Documentation/modules.txt>.  If unsure, say `Y'.
 
 config IP_NF_TFTP
-	tristate "TFTP prtocol support"
+	tristate "TFTP protocol support"
 	depends on IP_NF_CONNTRACK
 	help
 	  TFTP connection tracking helper, this is required depending
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 4ab428df1864..681416c6045b 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -120,6 +120,12 @@ int ip6_ra_control(struct sock *sk, int sel, void (*destructor)(struct sock *))
 	return 0;
 }
 
+extern int ip6_mc_source(int add, int omode, struct sock *sk,
+	struct group_source_req *pgsr);
+extern int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf);
+extern int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf,
+	struct group_filter *optval, int *optlen);
+
 
 int ipv6_setsockopt(struct sock *sk, int level, int optname, char *optval, 
 		    int optlen)
@@ -393,12 +399,13 @@ done:
 			break;
 		}
 		psin6 = (struct sockaddr_in6 *)&greq.gr_group;
-		if (optname == IPV6_ADD_MEMBERSHIP)
+		if (optname == MCAST_JOIN_GROUP)
 			retv = ipv6_sock_mc_join(sk, greq.gr_interface,
 				&psin6->sin6_addr);
 		else
 			retv = ipv6_sock_mc_drop(sk, greq.gr_interface,
 				&psin6->sin6_addr);
+		break;
 	}
 	case MCAST_JOIN_SOURCE_GROUP:
 	case MCAST_LEAVE_SOURCE_GROUP:
@@ -414,7 +421,8 @@ done:
 			retv = -EFAULT;
 			break;
 		}
-		if (greqs.gsr_group.ss_family != AF_INET6) {
+		if (greqs.gsr_group.ss_family != AF_INET6 ||
+		    greqs.gsr_source.ss_family != AF_INET6) {
 			retv = -EADDRNOTAVAIL;
 			break;
 		}
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 176963371c72..57877738d350 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -768,7 +768,7 @@ static void mld_clear_delrec(struct inet6_dev *idev)
 		psf = pmc->mca_tomb;
 		pmc->mca_tomb = 0;
 		spin_unlock_bh(&pmc->mca_lock);
-		for (psf=pmc->mca_tomb; psf; psf=psf_next) {
+		for (; psf; psf=psf_next) {
 			psf_next = psf->sf_next;
 			kfree(psf);
 		}
@@ -1042,6 +1042,8 @@ int igmp6_event_query(struct sk_buff *skb)
 		mld_clear_delrec(idev);
 	} else if (len >= 28) {
 		max_delay = MLDV2_MRC(ntohs(mlh2->mrc))*(HZ/10);
+		if (!max_delay)
+			max_delay = 1;
 		idev->mc_maxdelay = max_delay;
 		if (mlh2->qrv)
 			idev->mc_qrv = mlh2->qrv;
@@ -2096,8 +2098,6 @@ static int ip6_mcf_read_proc(char *buffer, char **start, off_t offset,
 			unsigned long icount, xcount, i;
 
 			spin_lock_bh(&imc->mca_lock);
-			icount = imc->mca_sfcount[MCAST_INCLUDE];
-			xcount = imc->mca_sfcount[MCAST_EXCLUDE];
 			for (psf=imc->mca_sources; psf; psf=psf->sf_next) {
 				if (first) {
 					len += sprintf(buffer+len, "%3s %6s "
@@ -2130,36 +2130,6 @@ static int ip6_mcf_read_proc(char *buffer, char **start, off_t offset,
 					in6_dev_put(idev);
 					goto done;
 				}
-				icount -= psf->sf_count[MCAST_INCLUDE];
-				xcount -= psf->sf_count[MCAST_EXCLUDE];
-			}
-			if (icount > 0 || xcount > 0) {
-				if (first) {
-					len += sprintf(buffer+len, "%3s %6s "
-						"%32s %32s %6s %6s\n", "Idx",
-						"Device", "Multicast Address",
-						"Source Address", "INC", "EXC");
-					first = 0;
-				}
-				len += sprintf(buffer+len,"%3d %6.6s ",
-					dev->ifindex, dev->name);
-
-				for (i=0; i<16; i++)
-					len += sprintf(buffer+len, "%02x",
-						imc->mca_addr.s6_addr[i]);
-				len += sprintf(buffer+len, " %32s %6lu %6lu\n",
-					"NONE", icount, xcount);
-				pos = begin+len;
-				if (pos < offset) {
-					len=0;
-					begin=pos;
-				}
-				if (pos > offset+length) {
-					spin_unlock_bh(&imc->mca_lock);
-					read_unlock_bh(&idev->lock);
-					in6_dev_put(idev);
-					goto done;
-				}
 			}
 			spin_unlock_bh(&imc->mca_lock);
 		}
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index c5b38160717d..7688f0ce1e5b 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -427,7 +427,7 @@ static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
 
 	len = sizeof(struct icmp6hdr) + sizeof(struct in6_addr);
 
-	rt = ndisc_get_dummy_rt();
+	rt = ip6_dst_alloc();
 	if (!rt) 
 		return;
 
@@ -524,7 +524,7 @@ void ndisc_send_ns(struct net_device *dev, struct neighbour *neigh,
 		saddr = &addr_buf;
 	}
 
-	rt = ndisc_get_dummy_rt();
+	rt = ip6_dst_alloc();
 	if (!rt) 
 		return;
 
@@ -595,7 +595,7 @@ void ndisc_send_rs(struct net_device *dev, struct in6_addr *saddr,
         int len;
 	int err;
 
-	rt = ndisc_get_dummy_rt();
+	rt = ip6_dst_alloc();
 	if (!rt) 
 		return;
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index c855580c278a..67eae08db2ca 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -130,12 +130,17 @@ struct fib6_node ip6_routing_table = {
 rwlock_t rt6_lock = RW_LOCK_UNLOCKED;
 
 
-/*	Dummy rt for ndisc */
-struct rt6_info *ndisc_get_dummy_rt()
+/* allocate dst with ip6_dst_ops */
+static __inline__ struct rt6_info *__ip6_dst_alloc(void)
 {
 	return dst_alloc(&ip6_dst_ops);
 }
 
+struct rt6_info *ip6_dst_alloc(void)
+{
+	return __ip6_dst_alloc();
+}
+
 /*
  *	Route lookup. Any rt6_lock is implied.
  */
@@ -640,7 +645,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh)
 	if (rtmsg->rtmsg_metric == 0)
 		rtmsg->rtmsg_metric = IP6_RT_PRIO_USER;
 
-	rt = dst_alloc(&ip6_dst_ops);
+	rt = __ip6_dst_alloc();
 
 	if (rt == NULL)
 		return -ENOMEM;
@@ -1035,9 +1040,7 @@ out:
 
 static struct rt6_info * ip6_rt_copy(struct rt6_info *ort)
 {
-	struct rt6_info *rt;
-
-	rt = dst_alloc(&ip6_dst_ops);
+	struct rt6_info *rt = __ip6_dst_alloc();
 
 	if (rt) {
 		rt->u.dst.input = ort->u.dst.input;
@@ -1181,9 +1184,8 @@ int ip6_pkt_discard(struct sk_buff *skb)
 
 int ip6_rt_addr_add(struct in6_addr *addr, struct net_device *dev)
 {
-	struct rt6_info *rt;
+	struct rt6_info *rt = __ip6_dst_alloc();
 
-	rt = dst_alloc(&ip6_dst_ops);
 	if (rt == NULL)
 		return -ENOMEM;
 
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index d77a9689004b..e972bef42859 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -1052,6 +1052,7 @@ struct proto_ops netlink_ops = {
 struct net_proto_family netlink_family_ops = {
 	.family = PF_NETLINK,
 	.create = netlink_create,
+	.owner	= THIS_MODULE,	/* for consistency 8) */
 };
 
 static int __init netlink_proto_init(void)
@@ -1066,6 +1067,11 @@ static int __init netlink_proto_init(void)
 #ifdef CONFIG_PROC_FS
 	create_proc_read_entry("net/netlink", 0, 0, netlink_read_proc, NULL);
 #endif
+	/* The netlink device handler may be needed early. */ 
+	rtnetlink_init();
+#ifdef CONFIG_NETLINK_DEV
+	init_netlink();
+#endif
 	return 0;
 }
 
@@ -1075,7 +1081,7 @@ static void __exit netlink_proto_exit(void)
        remove_proc_entry("net/netlink", NULL);
 }
 
-module_init(netlink_proto_init);
+subsys_initcall(netlink_proto_init);
 module_exit(netlink_proto_exit);
 
 MODULE_LICENSE("GPL");
diff --git a/net/socket.c b/net/socket.c
index 0d8a2e9d5a04..39a558da28ca 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -69,8 +69,6 @@
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
 #include <linux/wanrouter.h>
-#include <linux/netlink.h>
-#include <linux/rtnetlink.h>
 #include <linux/if_bridge.h>
 #include <linux/init.h>
 #include <linux/poll.h>
@@ -143,6 +141,36 @@ static struct file_operations socket_file_ops = {
 
 static struct net_proto_family *net_families[NPROTO];
 
+static __inline__ void net_family_bug(int family)
+{
+	printk(KERN_ERR "%d is not yet sock_registered!\n", family);
+	BUG();
+}
+
+int net_family_get(int family)
+{
+	struct net_proto_family *prot = net_families[family];
+	int rc = 1;
+
+	barrier();
+	if (likely(prot != NULL))
+		rc = try_module_get(prot->owner);
+	else
+		net_family_bug(family);
+	return rc;
+}
+
+void net_family_put(int family)
+{
+	struct net_proto_family *prot = net_families[family];
+
+	barrier();
+	if (likely(prot != NULL))
+		module_put(prot->owner);
+	else
+		net_family_bug(family);
+}
+
 #if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
 static atomic_t net_family_lockct = ATOMIC_INIT(0);
 static spinlock_t net_family_lock = SPIN_LOCK_UNLOCKED;
@@ -511,7 +539,7 @@ void sock_release(struct socket *sock)
 
 		sock->ops->release(sock);
 		sock->ops = NULL;
-		module_put(net_families[family]->owner);
+		net_family_put(family);
 	}
 
 	if (sock->fasync_list)
@@ -1064,7 +1092,7 @@ int sock_create(int family, int type, int protocol, struct socket **res)
 	sock->type  = type;
 
 	i = -EBUSY;
-	if (!try_module_get(net_families[family]->owner))
+	if (!net_family_get(family))
 		goto out_release;
 
 	if ((i = net_families[family]->create(sock, protocol)) < 0) 
@@ -1953,17 +1981,6 @@ void __init sock_init(void)
 	 *  do_initcalls is run.  
 	 */
 
-
-	/*
-	 * The netlink device handler may be needed early.
-	 */
-
-#ifdef CONFIG_NET
-	rtnetlink_init();
-#endif
-#ifdef CONFIG_NETLINK_DEV
-	init_netlink();
-#endif
 #ifdef CONFIG_NETFILTER
 	netfilter_init();
 #endif