[SCTP] Add icmpv6 handler to SCTP.

include/net/sctp/constants.h

@@ -138,12 +138,10 @@ typedef enum {
  */
 
 typedef union {
-
 	sctp_cid_t chunk;
 	sctp_event_timeout_t timeout;
 	sctp_event_other_t other;
 	sctp_event_primitive_t primitive;
-
 } sctp_subtype_t;
 
 #define SCTP_SUBTYPE_CONSTRUCTOR(_name, _type, _elt) \
@@ -421,9 +419,9 @@ typedef enum {
 
 /* Reasons to retransmit. */
 typedef enum {
-	SCTP_RETRANSMIT_T3_RTX,
-	SCTP_RETRANSMIT_FAST_RTX,
-	SCTP_RETRANSMIT_PMTU_DISCOVERY,
+	SCTP_RTXR_T3_RTX,
+	SCTP_RTXR_FAST_RTX,
+	SCTP_RTXR_PMTUD,
 } sctp_retransmit_reason_t;
 
 /* Reasons to lower cwnd. */

include/net/sctp/sctp.h

@@ -130,7 +130,7 @@ extern struct sctp_pf *sctp_get_pf_specific(sa_family_t family);
 extern int sctp_register_pf(struct sctp_pf *, sa_family_t);
 
 /*
- * sctp_socket.c
+ * sctp/socket.c
  */
 extern int sctp_backlog_rcv(struct sock *sk, struct sk_buff *skb);
 extern int sctp_inet_listen(struct socket *sock, int backlog);
@@ -139,7 +139,7 @@ extern unsigned int sctp_poll(struct file *file, struct socket *sock,
 		poll_table *wait);
 
 /*
- * sctp_primitive.c
+ * sctp/primitive.c
  */
 extern int sctp_primitive_ASSOCIATE(sctp_association_t *, void *arg);
 extern int sctp_primitive_SHUTDOWN(sctp_association_t *, void *arg);
@@ -148,14 +148,14 @@ extern int sctp_primitive_SEND(sctp_association_t *, void *arg);
 extern int sctp_primitive_REQUESTHEARTBEAT(sctp_association_t *, void *arg);
 
 /*
- * sctp_crc32c.c
+ * sctp/crc32c.c
  */
 extern __u32 sctp_start_cksum(__u8 *ptr, __u16 count);
 extern __u32 sctp_update_cksum(__u8 *ptr, __u16 count, __u32 cksum);
 extern __u32 sctp_end_cksum(__u32 cksum);
 
 /*
- * sctp_input.c
+ * sctp/input.c
  */
 extern int sctp_rcv(struct sk_buff *skb);
 extern void sctp_v4_err(struct sk_buff *skb, u32 info);
@@ -170,9 +170,16 @@ extern void __sctp_unhash_endpoint(sctp_endpoint_t *);
 extern sctp_association_t *__sctp_lookup_association(const union sctp_addr *,
 						     const union sctp_addr *,
 						     struct sctp_transport **);
-
+extern struct sock *sctp_err_lookup(int family, struct sk_buff *,
+				    struct sctphdr *, struct sctp_endpoint **, 
+				    struct sctp_association **, 
+				    struct sctp_transport **);
+extern void sctp_err_finish(struct sock *, struct sctp_endpoint *, 
+			    struct sctp_association *);
+extern void sctp_icmp_frag_needed(struct sock *, struct sctp_association *,
+				  struct sctp_transport *t, __u32 pmtu);
 /*
- * sctp_hashdriver.c
+ * sctp/hashdriver.c
  */
 extern void sctp_hash_digest(const char *secret, const int secret_len,
 			     const char *text, const int text_len,
@@ -184,9 +191,7 @@ extern void sctp_hash_digest(const char *secret, const int secret_len,
 
 
 #ifdef TEST_FRAME
-
 #include <test_frame.h>
-
 #else
 
 /* spin lock wrappers. */
@@ -312,7 +317,6 @@ static inline void sctp_sysctl_register(void) { return; }
 static inline void sctp_sysctl_unregister(void) { return; }
 #endif
 
-
 /* Size of Supported Address Parameter for 'x' address types. */
 #define SCTP_SAT_LEN(x) (sizeof(struct sctp_paramhdr) + (x) * sizeof(__u16))
 
@@ -320,19 +324,15 @@ static inline void sctp_sysctl_unregister(void) { return; }
 
 extern int sctp_v6_init(void);
 extern void sctp_v6_exit(void);
-static inline int sctp_ipv6_addr_type(const struct in6_addr *addr)
-{
-	return ipv6_addr_type((struct in6_addr*) addr);
-}
+extern void sctp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			int type, int code, int offset, __u32 info);
 
-#else /* #ifdef defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) */
+#else /* #ifdef defined(CONFIG_IPV6) */
 
-#define sctp_ipv6_addr_type(a) 0
 static inline int sctp_v6_init(void) { return 0; }
 static inline void sctp_v6_exit(void) { return; }
 
-#endif /* #ifdef defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) */
-
+#endif /* #if defined(CONFIG_IPV6) */
 
 /* Map an association to an assoc_id. */
 static inline sctp_assoc_t sctp_assoc2id(const sctp_association_t *asoc)
@@ -546,7 +546,7 @@ struct sctp_sock {
 	struct sock	  sk;
 #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
 	struct ipv6_pinfo *pinet6;
-#endif /* CONFIG_IPV6 || CONFIG_IPV6_MODULE */
+#endif /* CONFIG_IPV6 */
 	struct inet_opt	  inet;
 	struct sctp_opt	  sctp;
 };
@@ -559,7 +559,7 @@ struct sctp6_sock {
 	struct sctp_opt	  sctp;
 	struct ipv6_pinfo inet6;
 };
-#endif /* CONFIG_IPV6 || CONFIG_IPV6_MODULE */
+#endif /* CONFIG_IPV6 */
 
 #define sctp_sk(__sk) (&((struct sctp_sock *)__sk)->sctp)
 

net/sctp/input.c

@@ -205,21 +205,19 @@ int sctp_rcv(struct sk_buff *skb)
 	 */
 	sctp_bh_lock_sock(sk);
 
-	if (sock_owned_by_user(sk)) {
+	if (sock_owned_by_user(sk))
 		sk_add_backlog(sk, (struct sk_buff *) chunk);
-	} else {
+	else
 		sctp_backlog_rcv(sk, (struct sk_buff *) chunk);
-	}
 
 	/* Release the sock and any reference counts we took in the
 	 * lookup calls.
 	 */
 	sctp_bh_unlock_sock(sk);
-	if (asoc) {
+	if (asoc)
 		sctp_association_put(asoc);
-	} else {
+	else
 		sctp_endpoint_put(ep);
-	}
 	sock_put(sk);
 	return ret;
 
@@ -266,10 +264,8 @@ int sctp_backlog_rcv(struct sock *sk, struct sk_buff *skb)
 }
 
 /* Handle icmp frag needed error. */
-static inline void sctp_icmp_frag_needed(struct sock *sk,
-					 sctp_association_t *asoc,
-					 struct sctp_transport *transport,
-					 __u32 pmtu)
+void sctp_icmp_frag_needed(struct sock *sk, struct sctp_association *asoc,
+			   struct sctp_transport *t, __u32 pmtu)
 {
 	if (unlikely(pmtu < SCTP_DEFAULT_MINSEGMENT)) {
 		printk(KERN_WARNING "%s: Reported pmtu %d too low, "
@@ -278,54 +274,38 @@ static inline void sctp_icmp_frag_needed(struct sock *sk,
 		pmtu = SCTP_DEFAULT_MINSEGMENT;
 	}
 
-	if (!sock_owned_by_user(sk) && transport && (transport->pmtu != pmtu)) {
-		transport->pmtu = pmtu;
+	if (!sock_owned_by_user(sk) && t && (t->pmtu != pmtu)) {
+		t->pmtu = pmtu;
 		sctp_assoc_sync_pmtu(asoc);
-		sctp_retransmit(&asoc->outqueue, transport,
-				SCTP_RETRANSMIT_PMTU_DISCOVERY );
+		sctp_retransmit(&asoc->outqueue, t, SCTP_RTXR_PMTUD);
 	}
 }
 
-/*
- * This routine is called by the ICMP module when it gets some
- * sort of error condition.  If err < 0 then the socket should
- * be closed and the error returned to the user.  If err > 0
- * it's just the icmp type << 8 | icmp code.  After adjustment
- * header points to the first 8 bytes of the sctp header.  We need
- * to find the appropriate port.
- *
- * The locking strategy used here is very "optimistic". When
- * someone else accesses the socket the ICMP is just dropped
- * and for some paths there is no check at all.
- * A more general error queue to queue errors for later handling
- * is probably better.
- *
- */
-void sctp_v4_err(struct sk_buff *skb, __u32 info)
+/* Common lookup code for icmp/icmpv6 error handler. */
+struct sock *sctp_err_lookup(int family, struct sk_buff *skb,
+			     struct sctphdr *sctphdr,
+			     struct sctp_endpoint **epp, 
+			     struct sctp_association **app, 
+			     struct sctp_transport **tpp)
 {
-	struct iphdr *iph = (struct iphdr *)skb->data;
-	struct sctphdr *sh = (struct sctphdr *)(skb->data + (iph->ihl <<2));
-	int type = skb->h.icmph->type;
-	int code = skb->h.icmph->code;
-	union sctp_addr saddr, daddr;
-	struct inet_opt *inet;
+	union sctp_addr saddr;
+	union sctp_addr daddr;
+	struct sctp_af *af;
 	struct sock *sk = NULL;
-	sctp_endpoint_t *ep = NULL;
-	sctp_association_t *asoc = NULL;
-	struct sctp_transport *transport;
-	int err;
+	struct sctp_endpoint *ep = NULL;
+	struct sctp_association *asoc = NULL;
+	struct sctp_transport *transport = NULL;
 
-	if (skb->len < ((iph->ihl << 2) + 8)) {
-		ICMP_INC_STATS_BH(IcmpInErrors);
-		return;
+	*app = NULL; *epp = NULL; *tpp = NULL;
+
+	af = sctp_get_af_specific(family);
+	if (unlikely(!af)) {
+		return NULL;
 	}
 
-	saddr.v4.sin_family = AF_INET;
-	saddr.v4.sin_port = ntohs(sh->source);
-	memcpy(&saddr.v4.sin_addr.s_addr, &iph->saddr, sizeof(struct in_addr));	
-	daddr.v4.sin_family = AF_INET;
-	daddr.v4.sin_port = ntohs(sh->dest);
-	memcpy(&daddr.v4.sin_addr.s_addr, &iph->daddr, sizeof(struct in_addr));	
+	/* Initialize local addresses for lookups. */
+	af->from_skb(&saddr, skb, 1);
+	af->from_skb(&daddr, skb, 0);
 
 	/* Look for an association that matches the incoming ICMP error 
 	 * packet.
@@ -338,13 +318,12 @@ void sctp_v4_err(struct sk_buff *skb, __u32 info)
 		 */ 
 		ep = __sctp_rcv_lookup_endpoint(&daddr);
 		if (!ep) {
-			ICMP_INC_STATS_BH(IcmpInErrors);
-			return;
+			return NULL;
 		}
 	}
 
 	if (asoc) {
-		if (ntohl(sh->vtag) != asoc->c.peer_vtag) {
+		if (ntohl(sctphdr->vtag) != asoc->c.peer_vtag) {
 			ICMP_INC_STATS_BH(IcmpInErrors);
 			goto out;
 		}
@@ -353,12 +332,90 @@ void sctp_v4_err(struct sk_buff *skb, __u32 info)
 		sk = ep->base.sk;
 
 	sctp_bh_lock_sock(sk);
+
 	/* If too many ICMPs get dropped on busy
 	 * servers this needs to be solved differently.
 	 */
 	if (sock_owned_by_user(sk))
 		NET_INC_STATS_BH(LockDroppedIcmps);
 
+	*epp = ep;
+	*app = asoc;
+	*tpp = transport;
+	return sk;
+	
+out:
+	sock_put(sk);
+	if (asoc)
+		sctp_association_put(asoc);
+	if (ep)	
+		sctp_endpoint_put(ep);
+	return NULL;
+}
+
+/* Common cleanup code for icmp/icmpv6 error handler. */
+void sctp_err_finish(struct sock *sk, struct sctp_endpoint *ep, 
+		     struct sctp_association *asoc)
+{
+	sctp_bh_unlock_sock(sk);
+	sock_put(sk);
+	if (asoc)
+		sctp_association_put(asoc);
+	if (ep)	
+		sctp_endpoint_put(ep);
+}
+
+/*
+ * This routine is called by the ICMP module when it gets some
+ * sort of error condition.  If err < 0 then the socket should
+ * be closed and the error returned to the user.  If err > 0
+ * it's just the icmp type << 8 | icmp code.  After adjustment
+ * header points to the first 8 bytes of the sctp header.  We need
+ * to find the appropriate port.
+ *
+ * The locking strategy used here is very "optimistic". When
+ * someone else accesses the socket the ICMP is just dropped
+ * and for some paths there is no check at all.
+ * A more general error queue to queue errors for later handling
+ * is probably better.
+ *
+ */
+void sctp_v4_err(struct sk_buff *skb, __u32 info)
+{
+	struct iphdr *iph = (struct iphdr *)skb->data;
+	struct sctphdr *sh = (struct sctphdr *)(skb->data + (iph->ihl <<2));
+	int type = skb->h.icmph->type;
+	int code = skb->h.icmph->code;
+	struct sock *sk;
+	sctp_endpoint_t *ep;
+	sctp_association_t *asoc;
+	struct sctp_transport *transport;
+	struct inet_opt *inet;
+	char *saveip, *savesctp;
+	int err;
+
+	if (skb->len < ((iph->ihl << 2) + 8)) {
+		ICMP_INC_STATS_BH(IcmpInErrors);
+		return;
+	}
+
+	/* Fix up skb to look at the embedded net header. */
+	saveip = skb->nh.raw;
+	savesctp  = skb->h.raw;
+	skb->nh.iph = iph;
+	skb->h.raw = (char *)sh;
+	sk = sctp_err_lookup(AF_INET, skb, sh, &ep, &asoc, &transport);
+	/* Put back, the original pointers. */
+	skb->nh.raw = saveip;
+	skb->h.raw = savesctp;
+	if (!sk) {
+		ICMP_INC_STATS_BH(IcmpInErrors);
+		return;
+	}
+	/* Warning:  The sock lock is held.  Remember to call 
+	 * sctp_err_finish!
+	 */
+
 	switch (type) {
 	case ICMP_PARAMETERPROB:
 		err = EPROTO;
@@ -397,13 +454,7 @@ void sctp_v4_err(struct sk_buff *skb, __u32 info)
 	}
 
 out_unlock:
-	sctp_bh_unlock_sock(sk);
-out:
-	sock_put(sk);
-	if (asoc)
-		sctp_association_put(asoc);
-	if (ep)	
-		sctp_endpoint_put(ep);
+	sctp_err_finish(sk, ep, asoc);
 }
 
 /*
@@ -780,8 +831,3 @@ sctp_association_t *__sctp_rcv_lookup(struct sk_buff *skb,
 
 	return asoc;
 }
-
-
-
-
-

net/sctp/ipv6.c

 /* SCTP kernel reference Implementation
  * Copyright (c) 2001 Nokia, Inc.
  * Copyright (c) 2001 La Monte H.P. Yarroll
- * Copyright (c) 2002 International Business Machines, Corp.
+ * Copyright (c) 2002-2003 International Business Machines, Corp.
  *
  * This file is part of the SCTP kernel reference Implementation
  *
@@ -88,17 +88,62 @@ extern struct notifier_block sctp_inetaddr_notifier;
         ntohs((addr)->s6_addr16[6]), \
         ntohs((addr)->s6_addr16[7])
 
-/* FIXME: Comments. */
-static inline void sctp_v6_err(struct sk_buff *skb,
-			       struct inet6_skb_parm *opt,
-			       int type, int code, int offset, __u32 info)
+/* ICMP error handler. */
+void sctp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+		 int type, int code, int offset, __u32 info)
 {
-	/* BUG.  WRITE ME.  */
+	struct ipv6hdr *iph = (struct ipv6hdr *)skb->data;
+	struct sctphdr *sh = (struct sctphdr *)(skb->data + offset);
+	struct sock *sk;
+	sctp_endpoint_t *ep;
+	sctp_association_t *asoc;
+	struct sctp_transport *transport;
+	struct ipv6_pinfo *np;
+	char *saveip, *savesctp;
+	int err;
+
+	/* Fix up skb to look at the embedded net header. */
+	saveip = skb->nh.raw;
+	savesctp  = skb->h.raw;
+	skb->nh.ipv6h = iph;
+	skb->h.raw = (char *)sh;
+	sk = sctp_err_lookup(AF_INET6, skb, sh, &ep, &asoc, &transport);
+	/* Put back, the original pointers. */
+	skb->nh.raw = saveip;
+	skb->h.raw = savesctp;
+	if (!sk) {
+		ICMP6_INC_STATS_BH(Icmp6InErrors);
+		return;
+	}
+
+	/* Warning:  The sock lock is held.  Remember to call 
+	 * sctp_err_finish!
+	 */
+
+	switch (type) {
+	case ICMPV6_PKT_TOOBIG:
+		sctp_icmp_frag_needed(sk, asoc, transport, ntohl(info));
+		goto out_unlock;
+	default:
+		break;
+	}
+
+	np = inet6_sk(sk);
+	icmpv6_err_convert(type, code, &err);
+	if (!sock_owned_by_user(sk) && np->recverr) {
+		sk->err = err;
+		sk->error_report(sk);
+	} else {  /* Only an error on timeout */
+		sk->err_soft = err;
+	}
+
+out_unlock:
+	sctp_err_finish(sk, ep, asoc);
 }
 
 /* Based on tcp_v6_xmit() in tcp_ipv6.c. */
-static inline int sctp_v6_xmit(struct sk_buff *skb,
-			       struct sctp_transport *transport, int ipfragok)
+static int sctp_v6_xmit(struct sk_buff *skb, struct sctp_transport *transport, 
+			int ipfragok)
 {
 	struct sock *sk = skb->sk;
 	struct ipv6_pinfo *np = inet6_sk(sk);
@@ -110,9 +155,9 @@ static inline int sctp_v6_xmit(struct sk_buff *skb,
 
 	/* Fill in the dest address from the route entry passed with the skb
 	 * and the source address from the transport.
-	 */ 
+	 */
 	fl.fl6_dst = &rt6->rt6i_dst.addr;
-	fl.fl6_src = &transport->saddr.v6.sin6_addr; 
+	fl.fl6_src = &transport->saddr.v6.sin6_addr;
 
 	fl.fl6_flowlabel = np->flow_label;
 	IP6_ECN_flow_xmit(sk, fl.fl6_flowlabel);
@@ -174,7 +219,7 @@ struct dst_entry *sctp_v6_get_dst(sctp_association_t *asoc,
 
 /* Returns the number of consecutive initial bits that match in the 2 ipv6
  * addresses.
- */ 
+ */
 static inline int sctp_v6_addr_match_len(union sctp_addr *s1,
 					 union sctp_addr *s2)
 {
@@ -186,7 +231,7 @@ static inline int sctp_v6_addr_match_len(union sctp_addr *s1,
 		__u32 a1xora2;
 
 		a1xora2 = a1->s6_addr32[i] ^ a2->s6_addr32[i];
-		
+
 		if ((j = fls(ntohl(a1xora2))))
 			return (i * 32 + 32 - j);
 	}
@@ -196,7 +241,7 @@ static inline int sctp_v6_addr_match_len(union sctp_addr *s1,
 
 /* Fills in the source address(saddr) based on the destination address(daddr)
  * and asoc's bind address list.
- */   
+ */
 void sctp_v6_get_saddr(sctp_association_t *asoc, struct dst_entry *dst,
 		       union sctp_addr *daddr, union sctp_addr *saddr)
 {
@@ -432,7 +477,7 @@ static sctp_scope_t sctp_v6_scope(union sctp_addr *addr)
 	return retval;
 }
 
-/* Create and initialize a new sk for the socket to be returned by accept(). */ 
+/* Create and initialize a new sk for the socket to be returned by accept(). */
 struct sock *sctp_v6_create_accept_sk(struct sock *sk,
 				      struct sctp_association *asoc)
 {
@@ -469,11 +514,11 @@ struct sock *sctp_v6_create_accept_sk(struct sock *sk,
 
 	memcpy(newnp, np, sizeof(struct ipv6_pinfo));
 
-	ipv6_addr_copy(&newnp->daddr, &asoc->peer.primary_addr.v6.sin6_addr); 
+	ipv6_addr_copy(&newnp->daddr, &asoc->peer.primary_addr.v6.sin6_addr);
 
 	newinet->sport = inet->sport;
 	newinet->dport = asoc->peer.port;
-	
+
 #ifdef INET_REFCNT_DEBUG
 	atomic_inc(&inet6_sock_nr);
 	atomic_inc(&inet_sock_nr);
@@ -623,11 +668,11 @@ static int sctp_inet6_bind_verify(struct sctp_opt *opt, union sctp_addr *addr)
 /* Fill in Supported Address Type information for INIT and INIT-ACK
  * chunks.   Note: In the future, we may want to look at sock options
  * to determine whether a PF_INET6 socket really wants to have IPV4
- * addresses.  
+ * addresses.
  * Returns number of addresses supported.
  */
 static int sctp_inet6_supported_addrs(const struct sctp_opt *opt,
-				      __u16 *types) 
+				      __u16 *types)
 {
 	types[0] = SCTP_PARAM_IPV4_ADDRESS;
 	types[1] = SCTP_PARAM_IPV6_ADDRESS;

net/sctp/outqueue.c

@@ -357,7 +357,7 @@ void sctp_retransmit(struct sctp_outq *q, struct sctp_transport *transport,
 	__u8 fast_retransmit = 0;
 
 	switch(reason) {
-	case SCTP_RETRANSMIT_T3_RTX:
+	case SCTP_RTXR_T3_RTX:
 		sctp_transport_lower_cwnd(transport, SCTP_LOWER_CWND_T3_RTX);
 		/* Update the retran path if the T3-rtx timer has expired for
 		 * the current retran path.
@@ -365,10 +365,11 @@ void sctp_retransmit(struct sctp_outq *q, struct sctp_transport *transport,
 		if (transport == transport->asoc->peer.retran_path)
 			sctp_assoc_update_retran_path(transport->asoc);
 		break;
-	case SCTP_RETRANSMIT_FAST_RTX:
+	case SCTP_RTXR_FAST_RTX:
 		sctp_transport_lower_cwnd(transport, SCTP_LOWER_CWND_FAST_RTX);
 		fast_retransmit = 1;
 		break;
+	case SCTP_RTXR_PMTUD:
 	default:
 		break;
 	}
@@ -876,7 +877,7 @@ int sctp_outq_flush(struct sctp_outq *q, int rtx_timeout)
 		start_timer = 0;
 		queue = &q->out;
 
-		while (chunk = sctp_outq_dequeue_data(q)) {
+		while ((chunk = sctp_outq_dequeue_data(q))) {
 			/* RFC 2960 6.5 Every DATA chunk MUST carry a valid
 			 * stream identifier.
 			 */
@@ -1570,7 +1571,7 @@ static void sctp_check_transmitted(struct sctp_outq *q,
 
 	if (transport) {
 		if (do_fast_retransmit)
-			sctp_retransmit(q, transport, SCTP_RETRANSMIT_FAST_RTX);
+			sctp_retransmit(q, transport, SCTP_RTXR_FAST_RTX);
 
 		SCTP_DEBUG_PRINTK("%s: transport: %p, cwnd: %d, "
 				  "ssthresh: %d, flight_size: %d, pba: %d\n",