[IPV6] MROUTE: Support multicast forwarding.

Based on ancient patch by Mickael Hoerdt
<hoerdt@clarinet.u-strasbg.fr>, which is available at
<http://www-r2.u-strasbg.fr/~hoerdt/dev/linux_ipv6_mforwarding/patch-linux-ipv6-mforwarding-0.1a>.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>

include/linux/Kbuild

@@ -261,6 +261,7 @@ unifdef-y += mempolicy.h
 unifdef-y += mii.h
 unifdef-y += mman.h
 unifdef-y += mroute.h
+unifdef-y += mroute6.h
 unifdef-y += msdos_fs.h
 unifdef-y += msg.h
 unifdef-y += nbd.h

include/linux/ipv6.h

@@ -159,6 +159,9 @@ struct ipv6_devconf {
 	__s32		accept_source_route;
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
 	__s32		optimistic_dad;
+#endif
+#ifdef CONFIG_IPV6_MROUTE
+	__s32		mc_forwarding;
 #endif
 	void		*sysctl;
 };
@@ -190,6 +193,7 @@ enum {
 	DEVCONF_PROXY_NDP,
 	DEVCONF_OPTIMISTIC_DAD,
 	DEVCONF_ACCEPT_SOURCE_ROUTE,
+	DEVCONF_MC_FORWARDING,
 	DEVCONF_MAX
 };
 
@@ -230,6 +234,7 @@ struct inet6_skb_parm {
 #endif
 
 #define IP6SKB_XFRM_TRANSFORMED	1
+#define IP6SKB_FORWARDED	2
 };
 
 #define IP6CB(skb)	((struct inet6_skb_parm*)((skb)->cb))

include/linux/mroute6.h

@@ -0,0 +1,227 @@
+#ifndef __LINUX_MROUTE6_H
+#define __LINUX_MROUTE6_H
+
+#include <linux/types.h>
+#include <linux/sockios.h>
+
+/*
+ *	Based on the MROUTING 3.5 defines primarily to keep
+ *	source compatibility with BSD.
+ *
+ *	See the pim6sd code for the original history.
+ *
+ *      Protocol Independent Multicast (PIM) data structures included
+ *      Carlos Picoto (cap@di.fc.ul.pt)
+ *
+ */
+
+#define MRT6_BASE	200
+#define MRT6_INIT	(MRT6_BASE)	/* Activate the kernel mroute code 	*/
+#define MRT6_DONE	(MRT6_BASE+1)	/* Shutdown the kernel mroute		*/
+#define MRT6_ADD_MIF	(MRT6_BASE+2)	/* Add a virtual interface		*/
+#define MRT6_DEL_MIF	(MRT6_BASE+3)	/* Delete a virtual interface		*/
+#define MRT6_ADD_MFC	(MRT6_BASE+4)	/* Add a multicast forwarding entry	*/
+#define MRT6_DEL_MFC	(MRT6_BASE+5)	/* Delete a multicast forwarding entry	*/
+#define MRT6_VERSION	(MRT6_BASE+6)	/* Get the kernel multicast version	*/
+
+#define SIOCGETMIFCNT_IN6	SIOCPROTOPRIVATE	/* IP protocol privates */
+#define SIOCGETSGCNT_IN6	(SIOCPROTOPRIVATE+1)
+#define SIOCGETRPF	(SIOCPROTOPRIVATE+2)
+
+#define MAXMIFS		32
+typedef unsigned long mifbitmap_t;	/* User mode code depends on this lot */
+typedef unsigned short mifi_t;
+#define ALL_MIFS	((mifi_t)(-1))
+
+#ifndef IF_SETSIZE
+#define IF_SETSIZE	256
+#endif
+
+typedef	__u32		if_mask;
+#define NIFBITS (sizeof(if_mask) * 8)        /* bits per mask */
+
+#if !defined(__KERNEL__) && !defined(DIV_ROUND_UP)
+#define	DIV_ROUND_UP(x,y)	(((x) + ((y) - 1)) / (y))
+#endif
+
+typedef struct if_set {
+	if_mask ifs_bits[DIV_ROUND_UP(IF_SETSIZE, NIFBITS)];
+} if_set;
+
+#define IF_SET(n, p)    ((p)->ifs_bits[(n)/NIFBITS] |= (1 << ((n) % NIFBITS)))
+#define IF_CLR(n, p)    ((p)->ifs_bits[(n)/NIFBITS] &= ~(1 << ((n) % NIFBITS)))
+#define IF_ISSET(n, p)  ((p)->ifs_bits[(n)/NIFBITS] & (1 << ((n) % NIFBITS)))
+#define IF_COPY(f, t)   bcopy(f, t, sizeof(*(f)))
+#define IF_ZERO(p)      bzero(p, sizeof(*(p)))
+
+/*
+ *	Passed by mrouted for an MRT_ADD_MIF - again we use the
+ *	mrouted 3.6 structures for compatibility
+ */
+
+struct mif6ctl {
+	mifi_t	mif6c_mifi;		/* Index of MIF */
+	unsigned char mif6c_flags;	/* MIFF_ flags */
+	unsigned char vifc_threshold;	/* ttl limit */
+	u_short	 mif6c_pifi;		/* the index of the physical IF */
+	unsigned int vifc_rate_limit;	/* Rate limiter values (NI) */
+};
+
+#define MIFF_REGISTER	0x1	/* register vif	*/
+
+/*
+ *	Cache manipulation structures for mrouted and PIMd
+ */
+
+struct mf6cctl
+{
+	struct sockaddr_in6 mf6cc_origin;		/* Origin of mcast	*/
+	struct sockaddr_in6 mf6cc_mcastgrp;		/* Group in question	*/
+	mifi_t	mf6cc_parent;			/* Where it arrived	*/
+	struct if_set mf6cc_ifset;		/* Where it is going */
+};
+
+/*
+ *	Group count retrieval for pim6sd
+ */
+
+struct sioc_sg_req6
+{
+	struct sockaddr_in6 src;
+	struct sockaddr_in6 grp;
+	unsigned long pktcnt;
+	unsigned long bytecnt;
+	unsigned long wrong_if;
+};
+
+/*
+ *	To get vif packet counts
+ */
+
+struct sioc_mif_req6
+{
+	mifi_t	mifi;		/* Which iface */
+	unsigned long icount;	/* In packets */
+	unsigned long ocount;	/* Out packets */
+	unsigned long ibytes;	/* In bytes */
+	unsigned long obytes;	/* Out bytes */
+};
+
+/*
+ *	That's all usermode folks
+ */
+
+#ifdef __KERNEL__
+
+#include <linux/skbuff.h>	/* for struct sk_buff_head */
+
+struct net_device;
+struct inet6_dev *ipv6_find_idev(struct net_device *dev);
+
+#ifdef CONFIG_IPV6_MROUTE
+static inline int ip6_mroute_opt(int opt)
+{
+	return (opt >= MRT6_BASE) && (opt <= MRT6_BASE + 10);
+}
+#else
+static inline int ip6_mroute_opt(int opt)
+{
+	return 0;
+}
+#endif
+
+struct sock;
+
+extern int ip6_mroute_setsockopt(struct sock *, int, char __user *, int);
+extern int ip6_mroute_getsockopt(struct sock *, int, char __user *, int __user *);
+extern int ip6_mr_input(struct sk_buff *skb);
+extern int ip6mr_ioctl(struct sock *sk, int cmd, void __user *arg);
+extern void ip6_mr_init(void);
+
+struct mif_device
+{
+	struct net_device 	*dev;			/* Device we are using */
+	unsigned long	bytes_in,bytes_out;
+	unsigned long	pkt_in,pkt_out;		/* Statistics 			*/
+	unsigned long	rate_limit;		/* Traffic shaping (NI) 	*/
+	unsigned char	threshold;		/* TTL threshold 		*/
+	unsigned short	flags;			/* Control flags 		*/
+	int		link;			/* Physical interface index	*/
+};
+
+#define VIFF_STATIC 0x8000
+
+struct mfc6_cache
+{
+	struct mfc6_cache *next;		/* Next entry on cache line 	*/
+	struct in6_addr mf6c_mcastgrp;			/* Group the entry belongs to 	*/
+	struct in6_addr mf6c_origin;			/* Source of packet 		*/
+	mifi_t mf6c_parent;			/* Source interface		*/
+	int mfc_flags;				/* Flags on line		*/
+
+	union {
+		struct {
+			unsigned long expires;
+			struct sk_buff_head unresolved;	/* Unresolved buffers		*/
+		} unres;
+		struct {
+			unsigned long last_assert;
+			int minvif;
+			int maxvif;
+			unsigned long bytes;
+			unsigned long pkt;
+			unsigned long wrong_if;
+			unsigned char ttls[MAXMIFS];	/* TTL thresholds		*/
+		} res;
+	} mfc_un;
+};
+
+#define MFC_STATIC		1
+#define MFC_NOTIFY		2
+
+#define MFC6_LINES		64
+
+#define MFC6_HASH(a, g) (((__force u32)(a)->s6_addr32[0] ^ \
+			  (__force u32)(a)->s6_addr32[1] ^ \
+			  (__force u32)(a)->s6_addr32[2] ^ \
+			  (__force u32)(a)->s6_addr32[3] ^ \
+			  (__force u32)(g)->s6_addr32[0] ^ \
+			  (__force u32)(g)->s6_addr32[1] ^ \
+			  (__force u32)(g)->s6_addr32[2] ^ \
+			  (__force u32)(g)->s6_addr32[3]) % MFC6_LINES)
+
+#define MFC_ASSERT_THRESH (3*HZ)		/* Maximal freq. of asserts */
+
+#endif
+
+#ifdef __KERNEL__
+struct rtmsg;
+extern int ip6mr_get_route(struct sk_buff *skb, struct rtmsg *rtm, int nowait);
+
+#ifdef CONFIG_IPV6_MROUTE
+extern struct sock *mroute6_socket;
+extern int ip6mr_sk_done(struct sock *sk);
+#else
+#define mroute6_socket NULL
+static inline int ip6mr_sk_done(struct sock *sk) { return 0; }
+#endif
+#endif
+
+/*
+ * Structure used to communicate from kernel to multicast router.
+ * We'll overlay the structure onto an MLD header (not an IPv6 heder like igmpmsg{}
+ * used for IPv4 implementation). This is because this structure will be passed via an
+ * IPv6 raw socket, on wich an application will only receiver the payload i.e the data after
+ * the IPv6 header and all the extension headers. (See section 3 of RFC 3542)
+ */
+
+struct mrt6msg {
+#define MRT6MSG_NOCACHE		1
+	__u8		im6_mbz;		/* must be zero		   */
+	__u8		im6_msgtype;		/* what type of message    */
+	__u16		im6_mif;		/* mif rec'd on		   */
+	__u32		im6_pad;		/* padding for 64 bit arch */
+	struct in6_addr	im6_src, im6_dst;
+};
+
+#endif

net/ipv6/Kconfig

@@ -209,3 +209,10 @@ config IPV6_SUBTREES
 
 	  If unsure, say N.
 
+config IPV6_MROUTE
+	bool "IPv6: multicast routing (EXPERIMENTAL)"
+	depends on IPV6 && EXPERIMENTAL
+	---help---
+	  Experimental support for IPv6 multicast forwarding.
+	  If unsure, say N.
+

net/ipv6/Makefile

@@ -11,6 +11,8 @@ ipv6-objs :=	af_inet6.o anycast.o ip6_output.o ip6_input.o addrconf.o \
 		exthdrs.o datagram.o ip6_flowlabel.o inet6_connection_sock.o
 
 ipv6-$(CONFIG_SYSCTL) = sysctl_net_ipv6.o
+ipv6-$(CONFIG_IPV6_MROUTE) += ip6mr.o
+
 ipv6-$(CONFIG_XFRM) += xfrm6_policy.o xfrm6_state.o xfrm6_input.o \
 	xfrm6_output.o
 ipv6-$(CONFIG_NETFILTER) += netfilter.o

net/ipv6/addrconf.c

@@ -412,7 +412,7 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 	return ndev;
 }
 
-static struct inet6_dev * ipv6_find_idev(struct net_device *dev)
+struct inet6_dev * ipv6_find_idev(struct net_device *dev)
 {
 	struct inet6_dev *idev;
 
@@ -3547,6 +3547,9 @@ static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
 	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
 #endif
+#ifdef CONFIG_IPV6_MROUTE
+	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
+#endif
 }
 
 static inline size_t inet6_if_nlmsg_size(void)
@@ -4094,6 +4097,16 @@ static struct addrconf_sysctl_table
 			.proc_handler   =       &proc_dointvec,
 
 		},
+#endif
+#ifdef CONFIG_IPV6_MROUTE
+		{
+			.ctl_name	=	CTL_UNNUMBERED,
+			.procname	=	"mc_forwarding",
+			.data		=	&ipv6_devconf.mc_forwarding,
+			.maxlen		=	sizeof(int),
+			.mode		=	0644,
+			.proc_handler	=	&proc_dointvec,
+		},
 #endif
 		{
 			.ctl_name	=	0,	/* sentinel */

net/ipv6/af_inet6.c

@@ -61,6 +61,9 @@
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
+#ifdef CONFIG_IPV6_MROUTE
+#include <linux/mroute6.h>
+#endif
 
 MODULE_AUTHOR("Cast of dozens");
 MODULE_DESCRIPTION("IPv6 protocol stack for Linux");
@@ -953,6 +956,9 @@ static int __init inet6_init(void)
 	err = icmpv6_init();
 	if (err)
 		goto icmp_fail;
+#ifdef CONFIG_IPV6_MROUTE
+	ip6_mr_init();
+#endif
 	err = ndisc_init();
 	if (err)
 		goto ndisc_fail;

net/ipv6/ip6_input.c

@@ -29,6 +29,7 @@
 #include <linux/netdevice.h>
 #include <linux/in6.h>
 #include <linux/icmpv6.h>
+#include <linux/mroute6.h>
 
 #include <linux/netfilter.h>
 #include <linux/netfilter_ipv6.h>
@@ -236,36 +237,84 @@ int ip6_mc_input(struct sk_buff *skb)
 	hdr = ipv6_hdr(skb);
 	deliver = ipv6_chk_mcast_addr(skb->dev, &hdr->daddr, NULL);
 
+#ifdef CONFIG_IPV6_MROUTE
 	/*
-	 *	IPv6 multicast router mode isnt currently supported.
+	 *      IPv6 multicast router mode is now supported ;)
 	 */
-#if 0
+	if (ipv6_devconf.mc_forwarding &&
-	if (ipv6_config.multicast_route) {
+	    likely(!(IP6CB(skb)->flags & IP6SKB_FORWARDED))) {
-		int addr_type;
+		/*
+		 * Okay, we try to forward - split and duplicate
+		 * packets.
+		 */
+		struct sk_buff *skb2;
+		struct inet6_skb_parm *opt = IP6CB(skb);
 
-		addr_type = ipv6_addr_type(&hdr->daddr);
+		/* Check for MLD */
+		if (unlikely(opt->ra)) {
+			/* Check if this is a mld message */
+			u8 *ptr = skb_network_header(skb) + opt->ra;
+			struct icmp6hdr *icmp6;
+			u8 nexthdr = hdr->nexthdr;
+			int offset;
 
-		if (!(addr_type & (IPV6_ADDR_LOOPBACK | IPV6_ADDR_LINKLOCAL))) {
+			/* Check if the value of Router Alert
-			struct sk_buff *skb2;
+			 * is for MLD (0x0000).
-			struct dst_entry *dst;
+			 */
+			if ((ptr[2] | ptr[3]) == 0) {
+				if (!ipv6_ext_hdr(nexthdr)) {
+					/* BUG */
+					goto discard;
+				}
+				offset = ipv6_skip_exthdr(skb, sizeof(*hdr),
+							  &nexthdr);
+				if (offset < 0)
+					goto discard;
 
-			dst = skb->dst;
+				if (nexthdr != IPPROTO_ICMPV6)
+					goto discard;
 
-			if (deliver) {
+				if (!pskb_may_pull(skb, (skb_network_header(skb) +
-				skb2 = skb_clone(skb, GFP_ATOMIC);
+						   offset + 1 - skb->data)))
-				dst_output(skb2);
+					goto discard;
-			} else {
+
-				dst_output(skb);
+				icmp6 = (struct icmp6hdr *)(skb_network_header(skb) + offset);
-				return 0;
+
+				switch (icmp6->icmp6_type) {
+				case ICMPV6_MGM_QUERY:
+				case ICMPV6_MGM_REPORT:
+				case ICMPV6_MGM_REDUCTION:
+				case ICMPV6_MLD2_REPORT:
+					break;
+				default:
+					/* Bogus */
+					goto discard;
+				}
+				deliver = 1;
+				goto out;
 			}
+			/* unknown RA - process it normally */
+		}
+
+		if (deliver)
+			skb2 = skb_clone(skb, GFP_ATOMIC);
+		else {
+			skb2 = skb;
+			skb = NULL;
+		}
+
+		if (skb2) {
+			skb2->dev = skb2->dst->dev;
+			ip6_mr_input(skb2);
 		}
 	}
 #endif
-
+out:
 	if (likely(deliver)) {
 		ip6_input(skb);
 		return 0;
 	}
+discard:
 	/* discard */
 	kfree_skb(skb);
 

net/ipv6/ip6_output.c

@@ -55,6 +55,7 @@
 #include <net/icmp.h>
 #include <net/xfrm.h>
 #include <net/checksum.h>
+#include <linux/mroute6.h>
 
 static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
 
@@ -137,8 +138,9 @@ static int ip6_output2(struct sk_buff *skb)
 		struct inet6_dev *idev = ip6_dst_idev(skb->dst);
 
 		if (!(dev->flags & IFF_LOOPBACK) && (!np || np->mc_loop) &&
-		    ipv6_chk_mcast_addr(dev, &ipv6_hdr(skb)->daddr,
+		    ((mroute6_socket && !(IP6CB(skb)->flags & IP6SKB_FORWARDED)) ||
-					&ipv6_hdr(skb)->saddr)) {
+		     ipv6_chk_mcast_addr(dev, &ipv6_hdr(skb)->daddr,
+					 &ipv6_hdr(skb)->saddr))) {
 			struct sk_buff *newskb = skb_clone(skb, GFP_ATOMIC);
 
 			/* Do not check for IFF_ALLMULTI; multicast routing

net/ipv6/ipv6_sockglue.c

@@ -33,6 +33,7 @@
 #include <linux/sockios.h>
 #include <linux/net.h>
 #include <linux/in6.h>
+#include <linux/mroute6.h>
 #include <linux/netdevice.h>
 #include <linux/if_arp.h>
 #include <linux/init.h>
@@ -118,6 +119,9 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 
 	valbool = (val!=0);
 
+	if (ip6_mroute_opt(optname))
+		return ip6_mroute_setsockopt(sk, optname, optval, optlen);
+
 	lock_sock(sk);
 
 	switch (optname) {
@@ -790,6 +794,9 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
 	int len;
 	int val;
 
+	if (ip6_mroute_opt(optname))
+		return ip6_mroute_getsockopt(sk, optname, optval, optlen);
+
 	if (get_user(len, optlen))
 		return -EFAULT;
 	switch (optname) {

net/ipv6/raw.c

@@ -53,6 +53,7 @@
 #if defined(CONFIG_IPV6_MIP6) || defined(CONFIG_IPV6_MIP6_MODULE)
 #include <net/mip6.h>
 #endif
+#include <linux/mroute6.h>
 
 #include <net/raw.h>
 #include <net/rawv6.h>
@@ -1135,7 +1136,11 @@ static int rawv6_ioctl(struct sock *sk, int cmd, unsigned long arg)
 		}
 
 		default:
+#ifdef CONFIG_IPV6_MROUTE
+			return ip6mr_ioctl(sk, cmd, (void __user *)arg);
+#else
 			return -ENOIOCTLCMD;
+#endif
 	}
 }
 
@@ -1143,7 +1148,7 @@ static void rawv6_close(struct sock *sk, long timeout)
 {
 	if (inet_sk(sk)->num == IPPROTO_RAW)
 		ip6_ra_control(sk, -1, NULL);
-
+	ip6mr_sk_done(sk);
 	sk_common_release(sk);
 }
 

net/ipv6/route.c

@@ -36,6 +36,7 @@
 #include <linux/route.h>
 #include <linux/netdevice.h>
 #include <linux/in6.h>
+#include <linux/mroute6.h>
 #include <linux/init.h>
 #include <linux/if_arp.h>
 #include <linux/proc_fs.h>
@@ -2106,7 +2107,7 @@ static inline size_t rt6_nlmsg_size(void)
 static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 			 struct in6_addr *dst, struct in6_addr *src,
 			 int iif, int type, u32 pid, u32 seq,
-			 int prefix, unsigned int flags)
+			 int prefix, int nowait, unsigned int flags)
 {
 	struct rtmsg *rtm;
 	struct nlmsghdr *nlh;
@@ -2166,9 +2167,24 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 	} else if (rtm->rtm_src_len)
 		NLA_PUT(skb, RTA_SRC, 16, &rt->rt6i_src.addr);
 #endif
-	if (iif)
+	if (iif) {
-		NLA_PUT_U32(skb, RTA_IIF, iif);
+#ifdef CONFIG_IPV6_MROUTE
-	else if (dst) {
+		if (ipv6_addr_is_multicast(&rt->rt6i_dst.addr)) {
+			int err = ip6mr_get_route(skb, rtm, nowait);
+			if (err <= 0) {
+				if (!nowait) {
+					if (err == 0)
+						return 0;
+					goto nla_put_failure;
+				} else {
+					if (err == -EMSGSIZE)
+						goto nla_put_failure;
+				}
+			}
+		} else
+#endif
+			NLA_PUT_U32(skb, RTA_IIF, iif);
+	} else if (dst) {
 		struct in6_addr saddr_buf;
 		if (ipv6_dev_get_saddr(ip6_dst_idev(&rt->u.dst)->dev,
 				       dst, 0, &saddr_buf) == 0)
@@ -2211,7 +2227,7 @@ int rt6_dump_route(struct rt6_info *rt, void *p_arg)
 
 	return rt6_fill_node(arg->skb, rt, NULL, NULL, 0, RTM_NEWROUTE,
 		     NETLINK_CB(arg->cb->skb).pid, arg->cb->nlh->nlmsg_seq,
-		     prefix, NLM_F_MULTI);
+		     prefix, 0, NLM_F_MULTI);
 }
 
 static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
@@ -2277,7 +2293,7 @@ static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void
 
 	err = rt6_fill_node(skb, rt, &fl.fl6_dst, &fl.fl6_src, iif,
 			    RTM_NEWROUTE, NETLINK_CB(in_skb).pid,
-			    nlh->nlmsg_seq, 0, 0);
+			    nlh->nlmsg_seq, 0, 0, 0);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto errout;
@@ -2303,7 +2319,7 @@ void inet6_rt_notify(int event, struct rt6_info *rt, struct nl_info *info)
 		goto errout;
 
 	err = rt6_fill_node(skb, rt, NULL, NULL, 0,
-				event, info->pid, seq, 0, 0);
+				event, info->pid, seq, 0, 0, 0);
 	if (err < 0) {
 		/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
 		WARN_ON(err == -EMSGSIZE);