[IPV4]: IGMPv3 support, with help from Vinay Kulkarni

include/linux/igmp.h

@@ -32,13 +32,60 @@ struct igmphdr
 	__u32 group;
 };
 
+/* V3 group record types [grec_type] */
+#define IGMPV3_MODE_IS_INCLUDE		1
+#define IGMPV3_MODE_IS_EXCLUDE		2
+#define IGMPV3_CHANGE_TO_INCLUDE	3
+#define IGMPV3_CHANGE_TO_EXCLUDE	4
+#define IGMPV3_ALLOW_NEW_SOURCES	5
+#define IGMPV3_BLOCK_OLD_SOURCES	6
+
+struct igmpv3_grec {
+	__u8	grec_type;
+	__u8	grec_auxwords;
+	__u16	grec_nsrcs;
+	__u32	grec_mca;
+	__u32	grec_src[0];
+};
+
+struct igmpv3_report {
+	__u8 type;
+	__u8 resv1;
+	__u16 csum;
+	__u16 resv2;
+	__u16 ngrec;
+	struct igmpv3_grec grec[0];
+};
+
+struct igmpv3_query {
+	__u8 type;
+	__u8 code;
+	__u16 csum;
+	__u32 group;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8 qrv:3,
+	     suppress:1,
+	     resv:4;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u8 resv:4,
+	     suppress:1,
+	     qrv:3;
+#else
+#error "Please fix <asm/byteorder.h>"
+#endif
+	__u8 qqic;
+	__u16 nsrcs;
+	__u32 srcs[0];
+};
+
 #define IGMP_HOST_MEMBERSHIP_QUERY	0x11	/* From RFC1112 */
 #define IGMP_HOST_MEMBERSHIP_REPORT	0x12	/* Ditto */
 #define IGMP_DVMRP			0x13	/* DVMRP routing */
 #define IGMP_PIM			0x14	/* PIM routing */
 #define IGMP_TRACE			0x15
-#define IGMP_HOST_NEW_MEMBERSHIP_REPORT 0x16	/* New version of 0x11 */
+#define IGMPV2_HOST_MEMBERSHIP_REPORT	0x16	/* V2 version of 0x11 */
 #define IGMP_HOST_LEAVE_MESSAGE 	0x17
+#define IGMPV3_HOST_MEMBERSHIP_REPORT	0x22	/* V3 version of 0x11 */
 
 #define IGMP_MTRACE_RESP		0x1e
 #define IGMP_MTRACE			0x1f
@@ -68,6 +115,7 @@ struct igmphdr
 
 #define IGMP_ALL_HOSTS		htonl(0xE0000001L)
 #define IGMP_ALL_ROUTER 	htonl(0xE0000002L)
+#define IGMPV3_ALL_MCR	 	htonl(0xE0000016L)
 #define IGMP_LOCAL_GROUP	htonl(0xE0000000L)
 #define IGMP_LOCAL_GROUP_MASK	htonl(0xFFFFFF00L)
 
@@ -79,6 +127,18 @@ struct igmphdr
 #include <linux/skbuff.h>
 #include <linux/in.h>
 
+struct ip_sf_socklist
+{
+	unsigned int		sl_max;
+	unsigned int		sl_count;
+	__u32			sl_addr[0];
+};
+
+#define IP_SFLSIZE(count)	(sizeof(struct ip_sf_socklist) + \
+	(count) * sizeof(__u32))
+
+#define IP_SFBLOCK	10	/* allocate this many at once */
+
 /* ip_mc_socklist is real list now. Speed is not argument;
    this list never used in fast path code
  */
@@ -88,12 +148,28 @@ struct ip_mc_socklist
 	struct ip_mc_socklist	*next;
 	int			count;
 	struct ip_mreqn		multi;
+	unsigned int		sfmode;		/* MCAST_{INCLUDE,EXCLUDE} */
+	struct ip_sf_socklist	*sflist;
+};
+
+struct ip_sf_list
+{
+	struct ip_sf_list	*sf_next;
+	__u32			sf_inaddr;
+	unsigned long		sf_count[2];	/* include/exclude counts */
+	unsigned char		sf_gsresp;	/* include in g & s response? */
+	unsigned char		sf_oldin;	/* change state */
+	unsigned char		sf_crcount;	/* retrans. left to send */
 };
 
 struct ip_mc_list
 {
 	struct in_device	*interface;
 	unsigned long		multiaddr;
+	struct ip_sf_list	*sources;
+	struct ip_sf_list	*tomb;
+	unsigned int		sfmode;
+	unsigned long		sfcount[2];
 	struct ip_mc_list	*next;
 	struct timer_list	timer;
 	int			users;
@@ -103,13 +179,31 @@ struct ip_mc_list
 	char			reporter;
 	char			unsolicit_count;
 	char			loaded;
+	unsigned char		gsquery;	/* check source marks? */
+	unsigned char		crcount;
 };
 
-extern int ip_check_mc(struct in_device *dev, u32 mc_addr);
+/* V3 exponential field decoding */
+#define IGMPV3_MASK(value, nb) ((nb)>=32 ? (value) : ((1<<(nb))-1) & (value))
+#define IGMPV3_EXP(thresh, nbmant, nbexp, value) \
+	((value) < (thresh) ? (value) : \
+        ((IGMPV3_MASK(value, nbmant) | (1<<(nbmant+nbexp))) << \
+         (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)
+
+extern int ip_check_mc(struct in_device *dev, u32 mc_addr, u32 src_addr, u16 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_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);
+extern int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
+		struct ip_msfilter *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 *);
 extern void ip_mc_destroy_dev(struct in_device *);

include/linux/in.h

@@ -85,6 +85,14 @@ struct in_addr {
 #define IP_MULTICAST_LOOP 		34
 #define IP_ADD_MEMBERSHIP		35
 #define IP_DROP_MEMBERSHIP		36
+#define IP_UNBLOCK_SOURCE		37
+#define IP_BLOCK_SOURCE			38
+#define IP_ADD_SOURCE_MEMBERSHIP	39
+#define IP_DROP_SOURCE_MEMBERSHIP	40
+#define IP_MSFILTER			41
+
+#define MCAST_EXCLUDE	0
+#define MCAST_INCLUDE	1
 
 /* These need to appear somewhere around here */
 #define IP_DEFAULT_MULTICAST_TTL        1
@@ -105,6 +113,24 @@ struct ip_mreqn
 	int		imr_ifindex;		/* Interface index */
 };
 
+struct ip_mreq_source {
+	__u32		imr_multiaddr;
+	__u32		imr_interface;
+	__u32		imr_sourceaddr;
+};
+
+struct ip_msfilter {
+	__u32		imsf_multiaddr;
+	__u32		imsf_interface;
+	__u32		imsf_fmode;
+	__u32		imsf_numsrc;
+	__u32		imsf_slist[1];
+};
+
+#define IP_MSFILTER_SIZE(numsrc) \
+	(sizeof(struct ip_msfilter) - sizeof(__u32) \
+	+ (numsrc) * sizeof(__u32))
+
 struct in_pktinfo
 {
 	int		ipi_ifindex;

include/linux/inetdevice.h

@@ -34,7 +34,17 @@ struct in_device
 	int			dead;
 	struct in_ifaddr	*ifa_list;	/* IP ifaddr chain		*/
 	struct ip_mc_list	*mc_list;	/* IP multicast filter chain    */
+	rwlock_t		mc_lock;	/* for mc_tomb */
+	struct ip_mc_list	*mc_tomb;
 	unsigned long		mr_v1_seen;
+	unsigned long		mr_v2_seen;
+	unsigned long		mr_maxdelay;
+	unsigned char		mr_qrv;
+	unsigned char		mr_gq_running;
+	unsigned char		mr_ifc_count;
+	struct timer_list	mr_gq_timer;	/* general query timer */
+	struct timer_list	mr_ifc_timer;	/* interface change timer */
+
 	struct neigh_parms	*arp_parms;
 	struct ipv4_devconf	cnf;
 };

include/net/ip.h

@@ -79,6 +79,7 @@ extern rwlock_t ip_ra_lock;
 extern void		ip_mc_dropsocket(struct sock *);
 extern void		ip_mc_dropdevice(struct net_device *dev);
 extern int		ip_mc_procinfo(char *, char **, off_t, int);
+extern int		ip_mcf_procinfo(char *, char **, off_t, int);
 
 /*
  *	Functions provided by ip.c

net/ipv4/ip_output.c

@@ -1312,5 +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
 }

net/ipv4/ip_sockglue.c

@@ -610,9 +610,67 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char *optval, int opt
 			}
 
 			if (optname == IP_ADD_MEMBERSHIP)
-				err = ip_mc_join_group(sk,&mreq);
+				err = ip_mc_join_group(sk, &mreq);
 			else
-				err = ip_mc_leave_group(sk,&mreq);
+				err = ip_mc_leave_group(sk, &mreq);
+			break;
+		}
+		case IP_MSFILTER:
+		{
+			struct ip_msfilter *msf;
+
+			if (optlen < IP_MSFILTER_SIZE(0))
+				goto e_inval;
+			msf = (struct ip_msfilter *)kmalloc(optlen, GFP_KERNEL);
+			if (msf == 0) {
+				err = -ENOBUFS;
+				break;
+			}
+			err = -EFAULT;
+			if (copy_from_user(msf, optval, optlen)) {
+				kfree(msf);
+				break;
+			}
+			err = ip_mc_msfilter(sk, msf);
+			kfree(msf);
+			break;
+		}
+		case IP_BLOCK_SOURCE:
+		case IP_UNBLOCK_SOURCE:
+		case IP_ADD_SOURCE_MEMBERSHIP:
+		case IP_DROP_SOURCE_MEMBERSHIP:
+		{
+			struct ip_mreq_source mreqs;
+			int omode, add;
+
+			if (optlen != sizeof(struct ip_mreq_source))
+				goto e_inval;
+			if (copy_from_user(&mreqs, optval, sizeof(mreqs))) {
+				err = -EFAULT;
+				break;
+			}
+			if (optname == IP_BLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 1;
+			} else if (optname == IP_UNBLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 0;
+			} else if (optname == IP_ADD_SOURCE_MEMBERSHIP) {
+				struct ip_mreqn mreq;
+
+				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);
+				if (err)
+					break;
+				omode = MCAST_INCLUDE;
+				add = 1;
+			} else /*IP_DROP_SOURCE_MEMBERSHIP */ {
+				omode = MCAST_INCLUDE;
+				add = 0;
+			}
+			err = ip_mc_source(add, omode, sk, &mreqs);
 			break;
 		}
 		case IP_ROUTER_ALERT:	
@@ -763,6 +821,20 @@ int ip_getsockopt(struct sock *sk, int level, int optname, char *optval, int *op
 				return -EFAULT;
 			return 0;
 		}
+		case IP_MSFILTER:
+		{
+			struct ip_msfilter msf;
+			int err;
+
+			if (len < IP_MSFILTER_SIZE(0))
+				return -EINVAL;
+			if (copy_from_user(&msf, optval, IP_MSFILTER_SIZE(0)))
+				return -EFAULT;
+			err = ip_mc_msfget(sk, &msf,
+				(struct ip_msfilter *)optval, optlen);
+			release_sock(sk);
+			return err;
+		}
 		case IP_PKTOPTIONS:		
 		{
 			struct msghdr msg;

net/ipv4/route.c

@@ -1790,7 +1790,8 @@ int ip_route_input(struct sk_buff *skb, u32 daddr, u32 saddr,
 
 		read_lock(&inetdev_lock);
 		if ((in_dev = __in_dev_get(dev)) != NULL) {
-			int our = ip_check_mc(in_dev, daddr);
+			int our = ip_check_mc(in_dev, daddr, saddr,
+				skb->nh.iph->protocol);
 			if (our
 #ifdef CONFIG_IP_MROUTE
 			    || (!LOCAL_MCAST(daddr) && IN_DEV_MFORWARD(in_dev))
@@ -2020,7 +2021,7 @@ int ip_route_output_slow(struct rtable **rp, const struct flowi *oldflp)
 		}
 	} else if (res.type == RTN_MULTICAST) {
 		flags |= RTCF_MULTICAST|RTCF_LOCAL;
-		if (!ip_check_mc(in_dev, oldflp->fl4_dst))
+		if (!ip_check_mc(in_dev, oldflp->fl4_dst, oldflp->fl4_src, oldflp->proto))
 			flags &= ~RTCF_LOCAL;
 		/* If multicast route do not exist use
 		   default one, but do not gateway in this case.

net/ipv4/udp.c

@@ -298,6 +298,8 @@ static inline struct sock *udp_v4_mcast_next(struct sock *sk,
 		    ipv6_only_sock(s)					||
 		    (s->bound_dev_if && s->bound_dev_if != dif))
 			continue;
+		if (!ip_mc_sf_allow(sk, loc_addr, rmt_addr, dif))
+			continue;
 		break;
   	}
   	return s;