Merge branch 'multicast-init-as-INCLUDE-when-join-SSM-INCLUDE-group'

Hangbin Liu says:

====================
multicast: init as INCLUDE when join SSM INCLUDE group

Based on RFC3376 5.1 and RFC3810 6.1, we should init as INCLUDE when join SSM
INCLUDE group. In my first version I only clear the group change record. But
this is not enough as when a new group join, it will init as EXCLUDE and
trigger an filter mode change in ip/ip6_mc_add_src(), which will clear all
source addresses' sf_crcount. This will prevent early joined address sending
state change records if multi source addresses joined at the same time.

In this v2 patchset, I fixed it by directly initializing the mode to INCLUDE
for SSM JOIN_SOURCE_GROUP. I also split the original patch into two separated
patches for IPv4 and IPv6.

Test: test by myself and customer.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

include/linux/igmp.h

@@ -109,6 +109,8 @@ struct ip_mc_list {
 extern int ip_check_mc_rcu(struct in_device *dev, __be32 mc_addr, __be32 src_addr, u8 proto);
 extern int igmp_rcv(struct sk_buff *);
 extern int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr);
+extern int ip_mc_join_group_ssm(struct sock *sk, struct ip_mreqn *imr,
+				unsigned int mode);
 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,

include/net/ipv6.h

@@ -1100,6 +1100,8 @@ void ipv6_sysctl_unregister(void);
 
 int ipv6_sock_mc_join(struct sock *sk, int ifindex,
 		      const struct in6_addr *addr);
+int ipv6_sock_mc_join_ssm(struct sock *sk, int ifindex,
+			  const struct in6_addr *addr, unsigned int mode);
 int ipv6_sock_mc_drop(struct sock *sk, int ifindex,
 		      const struct in6_addr *addr);
 #endif /* _NET_IPV6_H */

net/ipv4/igmp.c

@@ -1200,13 +1200,14 @@ static void igmpv3_del_delrec(struct in_device *in_dev, struct ip_mc_list *im)
 	spin_lock_bh(&im->lock);
 	if (pmc) {
 		im->interface = pmc->interface;
-		im->crcount = in_dev->mr_qrv ?: net->ipv4.sysctl_igmp_qrv;
 		im->sfmode = pmc->sfmode;
 		if (pmc->sfmode == MCAST_INCLUDE) {
 			im->tomb = pmc->tomb;
 			im->sources = pmc->sources;
 			for (psf = im->sources; psf; psf = psf->sf_next)
-				psf->sf_crcount = im->crcount;
+				psf->sf_crcount = in_dev->mr_qrv ?: net->ipv4.sysctl_igmp_qrv;
+		} else {
+			im->crcount = in_dev->mr_qrv ?: net->ipv4.sysctl_igmp_qrv;
 		}
 		in_dev_put(pmc->interface);
 		kfree(pmc);
@@ -1288,7 +1289,7 @@ static void igmp_group_dropped(struct ip_mc_list *im)
 #endif
 }
 
-static void igmp_group_added(struct ip_mc_list *im)
+static void igmp_group_added(struct ip_mc_list *im, unsigned int mode)
 {
 	struct in_device *in_dev = im->interface;
 #ifdef CONFIG_IP_MULTICAST
@@ -1316,7 +1317,13 @@ static void igmp_group_added(struct ip_mc_list *im)
 	}
 	/* else, v3 */
 
-	im->crcount = in_dev->mr_qrv ?: net->ipv4.sysctl_igmp_qrv;
+	/* Based on RFC3376 5.1, for newly added INCLUDE SSM, we should
+	 * not send filter-mode change record as the mode should be from
+	 * IN() to IN(A).
+	 */
+	if (mode == MCAST_EXCLUDE)
+		im->crcount = in_dev->mr_qrv ?: net->ipv4.sysctl_igmp_qrv;
+
 	igmp_ifc_event(in_dev);
 #endif
 }
@@ -1381,8 +1388,7 @@ static void ip_mc_hash_remove(struct in_device *in_dev,
 /*
  *	A socket has joined a multicast group on device dev.
  */
-
-void ip_mc_inc_group(struct in_device *in_dev, __be32 addr)
+void __ip_mc_inc_group(struct in_device *in_dev, __be32 addr, unsigned int mode)
 {
 	struct ip_mc_list *im;
 #ifdef CONFIG_IP_MULTICAST
@@ -1394,7 +1400,7 @@ void ip_mc_inc_group(struct in_device *in_dev, __be32 addr)
 	for_each_pmc_rtnl(in_dev, im) {
 		if (im->multiaddr == addr) {
 			im->users++;
-			ip_mc_add_src(in_dev, &addr, MCAST_EXCLUDE, 0, NULL, 0);
+			ip_mc_add_src(in_dev, &addr, mode, 0, NULL, 0);
 			goto out;
 		}
 	}
@@ -1408,8 +1414,8 @@ void ip_mc_inc_group(struct in_device *in_dev, __be32 addr)
 	in_dev_hold(in_dev);
 	im->multiaddr = addr;
 	/* initial mode is (EX, empty) */
-	im->sfmode = MCAST_EXCLUDE;
-	im->sfcount[MCAST_EXCLUDE] = 1;
+	im->sfmode = mode;
+	im->sfcount[mode] = 1;
 	refcount_set(&im->refcnt, 1);
 	spin_lock_init(&im->lock);
 #ifdef CONFIG_IP_MULTICAST
@@ -1426,12 +1432,17 @@ void ip_mc_inc_group(struct in_device *in_dev, __be32 addr)
 #ifdef CONFIG_IP_MULTICAST
 	igmpv3_del_delrec(in_dev, im);
 #endif
-	igmp_group_added(im);
+	igmp_group_added(im, mode);
 	if (!in_dev->dead)
 		ip_rt_multicast_event(in_dev);
 out:
 	return;
 }
+
+void ip_mc_inc_group(struct in_device *in_dev, __be32 addr)
+{
+	__ip_mc_inc_group(in_dev, addr, MCAST_EXCLUDE);
+}
 EXPORT_SYMBOL(ip_mc_inc_group);
 
 static int ip_mc_check_iphdr(struct sk_buff *skb)
@@ -1688,7 +1699,7 @@ void ip_mc_remap(struct in_device *in_dev)
 #ifdef CONFIG_IP_MULTICAST
 		igmpv3_del_delrec(in_dev, pmc);
 #endif
-		igmp_group_added(pmc);
+		igmp_group_added(pmc, pmc->sfmode);
 	}
 }
 
@@ -1751,7 +1762,7 @@ void ip_mc_up(struct in_device *in_dev)
 #ifdef CONFIG_IP_MULTICAST
 		igmpv3_del_delrec(in_dev, pmc);
 #endif
-		igmp_group_added(pmc);
+		igmp_group_added(pmc, pmc->sfmode);
 	}
 }
 
@@ -2130,8 +2141,8 @@ static void ip_mc_clear_src(struct ip_mc_list *pmc)
 
 /* Join a multicast group
  */
-
-int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr)
+static int __ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr,
+			      unsigned int mode)
 {
 	__be32 addr = imr->imr_multiaddr.s_addr;
 	struct ip_mc_socklist *iml, *i;
@@ -2172,15 +2183,30 @@ int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr)
 	memcpy(&iml->multi, imr, sizeof(*imr));
 	iml->next_rcu = inet->mc_list;
 	iml->sflist = NULL;
-	iml->sfmode = MCAST_EXCLUDE;
+	iml->sfmode = mode;
 	rcu_assign_pointer(inet->mc_list, iml);
-	ip_mc_inc_group(in_dev, addr);
+	__ip_mc_inc_group(in_dev, addr, mode);
 	err = 0;
 done:
 	return err;
 }
+
+/* Join ASM (Any-Source Multicast) group
+ */
+int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr)
+{
+	return __ip_mc_join_group(sk, imr, MCAST_EXCLUDE);
+}
 EXPORT_SYMBOL(ip_mc_join_group);
 
+/* Join SSM (Source-Specific Multicast) group
+ */
+int ip_mc_join_group_ssm(struct sock *sk, struct ip_mreqn *imr,
+			 unsigned int mode)
+{
+	return __ip_mc_join_group(sk, imr, mode);
+}
+
 static int ip_mc_leave_src(struct sock *sk, struct ip_mc_socklist *iml,
 			   struct in_device *in_dev)
 {

net/ipv4/ip_sockglue.c

@@ -984,7 +984,7 @@ static int do_ip_setsockopt(struct sock *sk, int level,
 			mreq.imr_multiaddr.s_addr = mreqs.imr_multiaddr;
 			mreq.imr_address.s_addr = mreqs.imr_interface;
 			mreq.imr_ifindex = 0;
-			err = ip_mc_join_group(sk, &mreq);
+			err = ip_mc_join_group_ssm(sk, &mreq, MCAST_INCLUDE);
 			if (err && err != -EADDRINUSE)
 				break;
 			omode = MCAST_INCLUDE;
@@ -1061,7 +1061,7 @@ static int do_ip_setsockopt(struct sock *sk, int level,
 			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);
+			err = ip_mc_join_group_ssm(sk, &mreq, MCAST_INCLUDE);
 			if (err && err != -EADDRINUSE)
 				break;
 			greqs.gsr_interface = mreq.imr_ifindex;

net/ipv6/ipv6_sockglue.c

@@ -729,8 +729,9 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 			struct sockaddr_in6 *psin6;
 
 			psin6 = (struct sockaddr_in6 *)&greqs.gsr_group;
-			retv = ipv6_sock_mc_join(sk, greqs.gsr_interface,
-						 &psin6->sin6_addr);
+			retv = ipv6_sock_mc_join_ssm(sk, greqs.gsr_interface,
+						     &psin6->sin6_addr,
+						     MCAST_INCLUDE);
 			/* prior join w/ different source is ok */
 			if (retv && retv != -EADDRINUSE)
 				break;

net/ipv6/mcast.c

@@ -95,6 +95,8 @@ static int ip6_mc_add_src(struct inet6_dev *idev, const struct in6_addr *pmca,
 			  int delta);
 static int ip6_mc_leave_src(struct sock *sk, struct ipv6_mc_socklist *iml,
 			    struct inet6_dev *idev);
+static int __ipv6_dev_mc_inc(struct net_device *dev,
+			     const struct in6_addr *addr, unsigned int mode);
 
 #define MLD_QRV_DEFAULT		2
 /* RFC3810, 9.2. Query Interval */
@@ -132,7 +134,8 @@ static int unsolicited_report_interval(struct inet6_dev *idev)
 	return iv > 0 ? iv : 1;
 }
 
-int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
+static int __ipv6_sock_mc_join(struct sock *sk, int ifindex,
+			       const struct in6_addr *addr, unsigned int mode)
 {
 	struct net_device *dev = NULL;
 	struct ipv6_mc_socklist *mc_lst;
@@ -179,7 +182,7 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
 	}
 
 	mc_lst->ifindex = dev->ifindex;
-	mc_lst->sfmode = MCAST_EXCLUDE;
+	mc_lst->sfmode = mode;
 	rwlock_init(&mc_lst->sflock);
 	mc_lst->sflist = NULL;
 
@@ -187,7 +190,7 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
 	 *	now add/increase the group membership on the device
 	 */
 
-	err = ipv6_dev_mc_inc(dev, addr);
+	err = __ipv6_dev_mc_inc(dev, addr, mode);
 
 	if (err) {
 		sock_kfree_s(sk, mc_lst, sizeof(*mc_lst));
@@ -199,8 +202,19 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
 
 	return 0;
 }
+
+int ipv6_sock_mc_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
+{
+	return __ipv6_sock_mc_join(sk, ifindex, addr, MCAST_EXCLUDE);
+}
 EXPORT_SYMBOL(ipv6_sock_mc_join);
 
+int ipv6_sock_mc_join_ssm(struct sock *sk, int ifindex,
+			  const struct in6_addr *addr, unsigned int mode)
+{
+	return __ipv6_sock_mc_join(sk, ifindex, addr, mode);
+}
+
 /*
  *	socket leave on multicast group
  */
@@ -646,7 +660,7 @@ bool inet6_mc_check(struct sock *sk, const struct in6_addr *mc_addr,
 	return rv;
 }
 
-static void igmp6_group_added(struct ifmcaddr6 *mc)
+static void igmp6_group_added(struct ifmcaddr6 *mc, unsigned int mode)
 {
 	struct net_device *dev = mc->idev->dev;
 	char buf[MAX_ADDR_LEN];
@@ -672,7 +686,13 @@ static void igmp6_group_added(struct ifmcaddr6 *mc)
 	}
 	/* else v2 */
 
-	mc->mca_crcount = mc->idev->mc_qrv;
+	/* Based on RFC3810 6.1, for newly added INCLUDE SSM, we
+	 * should not send filter-mode change record as the mode
+	 * should be from IN() to IN(A).
+	 */
+	if (mode == MCAST_EXCLUDE)
+		mc->mca_crcount = mc->idev->mc_qrv;
+
 	mld_ifc_event(mc->idev);
 }
 
@@ -770,13 +790,14 @@ static void mld_del_delrec(struct inet6_dev *idev, struct ifmcaddr6 *im)
 	spin_lock_bh(&im->mca_lock);
 	if (pmc) {
 		im->idev = pmc->idev;
-		im->mca_crcount = idev->mc_qrv;
 		im->mca_sfmode = pmc->mca_sfmode;
 		if (pmc->mca_sfmode == MCAST_INCLUDE) {
 			im->mca_tomb = pmc->mca_tomb;
 			im->mca_sources = pmc->mca_sources;
 			for (psf = im->mca_sources; psf; psf = psf->sf_next)
-				psf->sf_crcount = im->mca_crcount;
+				psf->sf_crcount = idev->mc_qrv;
+		} else {
+			im->mca_crcount = idev->mc_qrv;
 		}
 		in6_dev_put(pmc->idev);
 		kfree(pmc);
@@ -831,7 +852,8 @@ static void ma_put(struct ifmcaddr6 *mc)
 }
 
 static struct ifmcaddr6 *mca_alloc(struct inet6_dev *idev,
-				   const struct in6_addr *addr)
+				   const struct in6_addr *addr,
+				   unsigned int mode)
 {
 	struct ifmcaddr6 *mc;
 
@@ -849,9 +871,8 @@ static struct ifmcaddr6 *mca_alloc(struct inet6_dev *idev,
 	refcount_set(&mc->mca_refcnt, 1);
 	spin_lock_init(&mc->mca_lock);
 
-	/* initial mode is (EX, empty) */
-	mc->mca_sfmode = MCAST_EXCLUDE;
-	mc->mca_sfcount[MCAST_EXCLUDE] = 1;
+	mc->mca_sfmode = mode;
+	mc->mca_sfcount[mode] = 1;
 
 	if (ipv6_addr_is_ll_all_nodes(&mc->mca_addr) ||
 	    IPV6_ADDR_MC_SCOPE(&mc->mca_addr) < IPV6_ADDR_SCOPE_LINKLOCAL)
@@ -863,7 +884,8 @@ static struct ifmcaddr6 *mca_alloc(struct inet6_dev *idev,
 /*
  *	device multicast group inc (add if not found)
  */
-int ipv6_dev_mc_inc(struct net_device *dev, const struct in6_addr *addr)
+static int __ipv6_dev_mc_inc(struct net_device *dev,
+			     const struct in6_addr *addr, unsigned int mode)
 {
 	struct ifmcaddr6 *mc;
 	struct inet6_dev *idev;
@@ -887,14 +909,13 @@ int ipv6_dev_mc_inc(struct net_device *dev, const struct in6_addr *addr)
 		if (ipv6_addr_equal(&mc->mca_addr, addr)) {
 			mc->mca_users++;
 			write_unlock_bh(&idev->lock);
-			ip6_mc_add_src(idev, &mc->mca_addr, MCAST_EXCLUDE, 0,
-				NULL, 0);
+			ip6_mc_add_src(idev, &mc->mca_addr, mode, 0, NULL, 0);
 			in6_dev_put(idev);
 			return 0;
 		}
 	}
 
-	mc = mca_alloc(idev, addr);
+	mc = mca_alloc(idev, addr, mode);
 	if (!mc) {
 		write_unlock_bh(&idev->lock);
 		in6_dev_put(idev);
@@ -911,11 +932,16 @@ int ipv6_dev_mc_inc(struct net_device *dev, const struct in6_addr *addr)
 	write_unlock_bh(&idev->lock);
 
 	mld_del_delrec(idev, mc);
-	igmp6_group_added(mc);
+	igmp6_group_added(mc, mode);
 	ma_put(mc);
 	return 0;
 }
 
+int ipv6_dev_mc_inc(struct net_device *dev, const struct in6_addr *addr)
+{
+	return __ipv6_dev_mc_inc(dev, addr, MCAST_EXCLUDE);
+}
+
 /*
  *	device multicast group del
  */
@@ -1751,7 +1777,7 @@ static struct sk_buff *add_grec(struct sk_buff *skb, struct ifmcaddr6 *pmc,
 
 		psf_next = psf->sf_next;
 
-		if (!is_in(pmc, psf, type, gdeleted, sdeleted)) {
+		if (!is_in(pmc, psf, type, gdeleted, sdeleted) && !crsend) {
 			psf_prev = psf;
 			continue;
 		}
@@ -2066,7 +2092,7 @@ static void mld_send_initial_cr(struct inet6_dev *idev)
 		if (pmc->mca_sfcount[MCAST_EXCLUDE])
 			type = MLD2_CHANGE_TO_EXCLUDE;
 		else
-			type = MLD2_CHANGE_TO_INCLUDE;
+			type = MLD2_ALLOW_NEW_SOURCES;
 		skb = add_grec(skb, pmc, type, 0, 0, 1);
 		spin_unlock_bh(&pmc->mca_lock);
 	}
@@ -2546,7 +2572,7 @@ void ipv6_mc_up(struct inet6_dev *idev)
 	ipv6_mc_reset(idev);
 	for (i = idev->mc_list; i; i = i->next) {
 		mld_del_delrec(idev, i);
-		igmp6_group_added(i);
+		igmp6_group_added(i, i->mca_sfmode);
 	}
 	read_unlock_bh(&idev->lock);
 }