Skip to contentSkip to conversation

sctp: Implement RFC6951: UDP Encapsulation of SCTP

PATCH Superseded Xin Long 6 participants 25 messages Oct 13, 2020 – Oct 19, 2020

Subsystems (4)

v3 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Xin Long

author maintainer
#0 lore raw

v3 cover sctp: Implement RFC6951: UDP Encapsulation of SCTP

Description From the RFC:

The Main Reasons:

o To allow SCTP traffic to pass through legacy NATs, which do not provide native SCTP support as specified in [BEHAVE] and [NATSUPP].

o To allow SCTP to be implemented on hosts that do not provide direct access to the IP layer. In particular, applications can use their own SCTP implementation if the operating system does not provide one.

Implementation Notes:

UDP-encapsulated SCTP is normally communicated between SCTP stacks using the IANA-assigned UDP port number 9899 (sctp-tunneling) on both ends. There are circumstances where other ports may be used on either end, and it might be required to use ports other than the registered port.

Each SCTP stack uses a single local UDP encapsulation port number as the destination port for all its incoming SCTP packets, this greatly simplifies implementation design.

An SCTP implementation supporting UDP encapsulation MUST maintain a remote UDP encapsulation port number per destination address for each SCTP association. Again, because the remote stack may be using ports other than the well-known port, each port may be different from each stack. However, because of remapping of ports by NATs, the remote ports associated with different remote IP addresses may not be identical, even if they are associated with the same stack.

Because the well-known port might not be used, implementations need to allow other port numbers to be specified as a local or remote UDP encapsulation port number through APIs.

Patches:

This patchset is using the udp4/6 tunnel APIs to implement the UDP Encapsulation of SCTP with not much change in SCTP protocol stack and with all current SCTP features keeped in Linux Kernel.

1 - 4: Fix some UDP issues that may be triggered by SCTP over UDP. 5 - 7: Process incoming UDP encapsulated packets and ICMP packets. 8 -10: Remote encap port's update by sysctl, sockopt and packets. 11-14: Process outgoing pakects with UDP encapsulated and its GSO. 15-16: Add the part from draft-tuexen-tsvwg-sctp-udp-encaps-cons-03.
17: Enable this feature.

Tests:

- lksctp-tools/src/func_tests with UDP Encapsulation enabled/disabled:

Both make v4test and v6test passed.

- sctp-tests with UDP Encapsulation enabled/disabled:

repeatability/procdumps/sctpdiag/gsomtuchange/extoverflow/ sctphashtable passed. Others failed as expected due to those "iptables -p sctp" rules.

- netperf on lo/netns/virtio_net, with gso enabled/disabled and with ip_checksum enabled/disabled, with UDP Encapsulation enabled/disabled:

No clear performance dropped.

v1->v2:
- Fix some incorrect code in the patches 5,6,8,10,11,13,14,17, suggested by Marcelo.
- Append two patches 15-16 to add the Additional Considerations for UDP Encapsulation of SCTP from draft-tuexen-tsvwg-sctp-udp-encaps-cons-03.
v2->v3:
- remove the cleanup code in patch 2, suggested by Willem.
- remove the patch 3 and fix the checksum in the new patch 3 after talking with Paolo, Marcelo and Guillaume.
- add 'select NET_UDP_TUNNEL' in patch 4 to solve a compiling error.
- fix __be16 type cast warning in patch 8.
- fix the wrong endian orders when setting values in 14,16.

Xin Long (16):
udp: check udp sock encap_type in __udp_lib_err
udp6: move the mss check after udp gso tunnel processing
udp: support sctp over udp in skb_udp_tunnel_segment
sctp: create udp4 sock and add its encap_rcv
sctp: create udp6 sock and set its encap_rcv
sctp: add encap_err_lookup for udp encap socks
sctp: add encap_port for netns sock asoc and transport
sctp: add SCTP_REMOTE_UDP_ENCAPS_PORT sockopt
sctp: allow changing transport encap_port by peer packets
sctp: add udphdr to overhead when udp_port is set
sctp: call sk_setup_caps in sctp_packet_transmit instead
sctp: support for sending packet over udp4 sock
sctp: support for sending packet over udp6 sock
sctp: add the error cause for new encapsulation port restart
sctp: handle the init chunk matching an existing asoc
sctp: enable udp tunneling socks

 include/linux/sctp.h         |  20 ++++++
 include/net/netns/sctp.h     |   8 +++
 include/net/sctp/constants.h |   2 +
 include/net/sctp/sctp.h      |   9 ++-
 include/net/sctp/sm.h        |   4 ++
 include/net/sctp/structs.h   |  14 ++--
 include/uapi/linux/sctp.h    |   7 ++
 net/ipv4/udp.c               |   2 +-
 net/ipv4/udp_offload.c       |   3 +
 net/ipv6/udp.c               |   2 +-
 net/ipv6/udp_offload.c       |   8 +--
 net/sctp/Kconfig             |   1 +
 net/sctp/associola.c         |   4 ++
 net/sctp/ipv6.c              |  44 +++++++++----
 net/sctp/offload.c           |   6 +-
 net/sctp/output.c            |  22 +++----
 net/sctp/protocol.c          | 148 +++++++++++++++++++++++++++++++++++++++----
 net/sctp/sm_make_chunk.c     |  21 ++++++
 net/sctp/sm_statefuns.c      |  52 +++++++++++++++
 net/sctp/socket.c            | 116 +++++++++++++++++++++++++++++++++
 net/sctp/sysctl.c            |  60 ++++++++++++++++++
 21 files changed, 503 insertions(+), 50 deletions(-)
Signature · 1 line
2.1.0
author maintainer
#1 lore raw

PATCH v3 1/16 udp: check udp sock encap_type in __udp_lib_err

There is a chance that __udp4/6_lib_lookup() returns a udp encap sock in __udp_lib_err(), like the udp encap listening sock may use the same port as remote encap port, in which case it should go to __udp4/6_lib_err_encap() for more validation before processing the icmp packet.

This patch is to check encap_type in __udp_lib_err() for the further validation for a encap sock.

---
 net/ipv4/udp.c | 2 +-
 net/ipv6/udp.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
net/ipv4/udp.c +1 -1
index 09f0a23..ca04a8a 100644
@@ -702,7 +702,7 @@ int __udp4_lib_err(struct sk_buff *skb, u32 info, struct udp_table *udptable)
 	sk = __udp4_lib_lookup(net, iph->daddr, uh->dest,
 			       iph->saddr, uh->source, skb->dev->ifindex,
 			       inet_sdif(skb), udptable, NULL);
-	if (!sk) {
+	if (!sk || udp_sk(sk)->encap_type) {
 		/* No socket for error: try tunnels before discarding */
 		sk = ERR_PTR(-ENOENT);
 		if (static_branch_unlikely(&udp_encap_needed_key)) {
net/ipv6/udp.c +1 -1
index 29d9691..cde9b88 100644
@@ -560,7 +560,7 @@ int __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 
 	sk = __udp6_lib_lookup(net, daddr, uh->dest, saddr, uh->source,
 			       inet6_iif(skb), inet6_sdif(skb), udptable, NULL);
-	if (!sk) {
+	if (!sk || udp_sk(sk)->encap_type) {
 		/* No socket for error: try tunnels before discarding */
 		sk = ERR_PTR(-ENOENT);
 		if (static_branch_unlikely(&udpv6_encap_needed_key)) {
Signature · 1 line
2.1.0
author maintainer
#2 lore raw

PATCH v3 2/16 udp6: move the mss check after udp gso tunnel processing

For some protocol's gso, like SCTP, it's using GSO_BY_FRAGS for gso_size. When using UDP to encapsulate its packet, it will return error in udp6_ufo_fragment() as skb->len < gso_size, and it will never go to the gso tunnel processing.

So we should move this check after udp gso tunnel processing, the same as udp4_ufo_fragment() does.

v1->v2:
- no change.
v2->v3:
- not do any cleanup.

---
 net/ipv6/udp_offload.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
net/ipv6/udp_offload.c +4 -4
index 584157a..aa602af 100644
@@ -28,10 +28,6 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb,
 	int tnl_hlen;
 	int err;
 
-	mss = skb_shinfo(skb)->gso_size;
-	if (unlikely(skb->len <= mss))
-		goto out;
-
 	if (skb->encapsulation && skb_shinfo(skb)->gso_type &
 	    (SKB_GSO_UDP_TUNNEL|SKB_GSO_UDP_TUNNEL_CSUM))
 		segs = skb_udp_tunnel_segment(skb, features, true);
@@ -48,6 +44,10 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb,
 		if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4)
 			return __udp_gso_segment(skb, features);
 
+		mss = skb_shinfo(skb)->gso_size;
+		if (unlikely(skb->len <= mss))
+			goto out;
+
 		/* Do software UFO. Complete and fill in the UDP checksum as HW cannot
 		 * do checksum of UDP packets sent as multiple IP fragments.
 		 */
Signature · 1 line
2.1.0
author maintainer
#3 lore raw

PATCH v3 3/16 udp: support sctp over udp in skb_udp_tunnel_segment

For the gso of sctp over udp packets, sctp_gso_segment() will be called in skb_udp_tunnel_segment(), we need to set transport_header to sctp header.

As all the current HWs can't handle both crc checksum and udp checksum at the same time, the crc checksum has to be done in sctp_gso_segment() by removing the NETIF_F_SCTP_CRC flag from the features.

Meanwhile, if the HW can't do udp checksum, csum and csum_start has to be set correctly, and udp checksum will be done in __skb_udp_tunnel_segment() by calling gso_make_checksum().

Thanks to Paolo, Marcelo and Guillaume for helping with this one.

v1->v2:
- no change.
v2->v3:
- remove the he NETIF_F_SCTP_CRC flag from the features.
- set csum and csum_start in sctp_gso_make_checksum().

---
 net/ipv4/udp_offload.c | 3 +++
 net/sctp/offload.c     | 6 +++++-
 2 files changed, 8 insertions(+), 1 deletion(-)
net/ipv4/udp_offload.c +3 -0
index e67a66f..4d31255 100644
@@ -49,6 +49,7 @@ static struct sk_buff *__skb_udp_tunnel_segment(struct sk_buff *skb,
 	__skb_pull(skb, tnl_hlen);
 	skb_reset_mac_header(skb);
 	skb_set_network_header(skb, skb_inner_network_offset(skb));
+	skb_set_transport_header(skb, skb_inner_transport_offset(skb));
 	skb->mac_len = skb_inner_network_offset(skb);
 	skb->protocol = new_protocol;
 
@@ -67,6 +68,8 @@ static struct sk_buff *__skb_udp_tunnel_segment(struct sk_buff *skb,
 				      (NETIF_F_HW_CSUM | NETIF_F_IP_CSUM))));
 
 	features &= skb->dev->hw_enc_features;
+	/* CRC checksum can't be handled by HW when it's a udp tunneling packet. */
+	features &= ~NETIF_F_SCTP_CRC;
 
 	/* The only checksum offload we care about from here on out is the
 	 * outer one so strip the existing checksum feature flags and
net/sctp/offload.c +5 -1
index 74847d6..9f6f818 100644
@@ -27,7 +27,11 @@ static __le32 sctp_gso_make_checksum(struct sk_buff *skb)
 {
 	skb->ip_summed = CHECKSUM_NONE;
 	skb->csum_not_inet = 0;
-	gso_reset_checksum(skb, ~0);
+	/* csum and csum_start in gso cb may be needed to do the udp
+	 * checksum when it's a udp tunneling packet.
+	 */
+	SKB_GSO_CB(skb)->csum = (__force __wsum)~0;
+	SKB_GSO_CB(skb)->csum_start = skb_headroom(skb) + skb->len;
 	return sctp_compute_cksum(skb, skb_transport_offset(skb));
 }
 
Signature · 1 line
2.1.0
author maintainer
#4 lore raw

PATCH v3 4/16 sctp: create udp4 sock and add its encap_rcv

This patch is to add the functions to create/release udp4 sock, and set the sock's encap_rcv to process the incoming udp encap sctp packets. In sctp_udp_rcv(), as we can see, all we need to do is fix the transport header for sctp_rcv(), then it would implement the part of rfc6951#section-5.4:

"When an encapsulated packet is received, the UDP header is removed.
Then, the generic lookup is performed, as done by an SCTP stack whenever a packet is received, to find the association for the received SCTP packet"

Note that these functions will be called in the last patch of this patchset when enabling this feature.

v1->v2:
- Add pr_err() when fails to create udp v4 sock.
v2->v3:
- Add 'select NET_UDP_TUNNEL' in sctp Kconfig.

---
 include/net/netns/sctp.h     |  5 +++++
 include/net/sctp/constants.h |  2 ++
 include/net/sctp/sctp.h      |  2 ++
 net/sctp/Kconfig             |  1 +
 net/sctp/protocol.c          | 43 +++++++++++++++++++++++++++++++++++++++++++
 5 files changed, 53 insertions(+)
include/net/netns/sctp.h +5 -0
index d8d02e4..3d10bef 100644
@@ -22,6 +22,11 @@ struct netns_sctp {
 	 */
 	struct sock *ctl_sock;
 
+	/* udp tunneling listening sock. */
+	struct sock *udp4_sock;
+	/* udp tunneling listening port. */
+	int udp_port;
+
 	/* This is the global local address list.
 	 * We actively maintain this complete list of addresses on
 	 * the system by catching address add/delete events.
include/net/sctp/constants.h +2 -0
index 122d9e2..b583166 100644
@@ -286,6 +286,8 @@ enum { SCTP_MAX_GABS = 16 };
 				 * functions simpler to write.
 				 */
 
+#define SCTP_DEFAULT_UDP_PORT 9899	/* default udp tunneling port */
+
 /* These are the values for pf exposure, UNUSED is to keep compatible with old
  * applications by default.
  */
include/net/sctp/sctp.h +2 -0
index 4fc747b..bfd87a0 100644
@@ -84,6 +84,8 @@ int sctp_copy_local_addr_list(struct net *net, struct sctp_bind_addr *addr,
 struct sctp_pf *sctp_get_pf_specific(sa_family_t family);
 int sctp_register_pf(struct sctp_pf *, sa_family_t);
 void sctp_addr_wq_mgmt(struct net *, struct sctp_sockaddr_entry *, int);
+int sctp_udp_sock_start(struct net *net);
+void sctp_udp_sock_stop(struct net *net);
 
 /*
  * sctp/socket.c
net/sctp/Kconfig +1 -0
index 39d7fa9..5da599f 100644
@@ -11,6 +11,7 @@ menuconfig IP_SCTP
 	select CRYPTO_HMAC
 	select CRYPTO_SHA1
 	select LIBCRC32C
+	select NET_UDP_TUNNEL
 	help
 	  Stream Control Transmission Protocol
 
net/sctp/protocol.c +43 -0

61 lines not on this page · raw message

Signature · 1 line
2.1.0
author maintainer
#5 lore raw

PATCH v3 5/16 sctp: create udp6 sock and set its encap_rcv

This patch is to add the udp6 sock part in sctp_udp_sock_start/stop(). udp_conf.use_udp6_rx_checksums is set to true, as:

"The SCTP checksum MUST be computed for IPv4 and IPv6, and the UDP
checksum SHOULD be computed for IPv4 and IPv6"

says in rfc6951#section-5.3.

v1->v2:
- Add pr_err() when fails to create udp v6 sock.
- Add #if IS_ENABLED(CONFIG_IPV6) not to create v6 sock when ipv6 is disabled.

---
 include/net/netns/sctp.h |  1 +
 net/sctp/protocol.c      | 26 ++++++++++++++++++++++++++
 2 files changed, 27 insertions(+)
include/net/netns/sctp.h +1 -0
index 3d10bef..f622945 100644
@@ -24,6 +24,7 @@ struct netns_sctp {
 
 	/* udp tunneling listening sock. */
 	struct sock *udp4_sock;
+	struct sock *udp6_sock;
 	/* udp tunneling listening port. */
 	int udp_port;
 
net/sctp/protocol.c +26 -0
index 2b7a3e1..49b5d75 100644
@@ -869,6 +869,28 @@ int sctp_udp_sock_start(struct net *net)
 	setup_udp_tunnel_sock(net, sock, &tuncfg);
 	net->sctp.udp4_sock = sock->sk;
 
+#if IS_ENABLED(CONFIG_IPV6)
+	memset(&udp_conf, 0, sizeof(udp_conf));
+
+	udp_conf.family = AF_INET6;
+	udp_conf.local_ip6 = in6addr_any;
+	udp_conf.local_udp_port = htons(net->sctp.udp_port);
+	udp_conf.use_udp6_rx_checksums = true;
+	udp_conf.ipv6_v6only = true;
+	err = udp_sock_create(net, &udp_conf, &sock);
+	if (err) {
+		pr_err("Failed to create the SCTP udp tunneling v6 sock\n");
+		udp_tunnel_sock_release(net->sctp.udp4_sock->sk_socket);
+		net->sctp.udp4_sock = NULL;
+		return err;
+	}
+
+	tuncfg.encap_type = 1;
+	tuncfg.encap_rcv = sctp_udp_rcv;
+	setup_udp_tunnel_sock(net, sock, &tuncfg);
+	net->sctp.udp6_sock = sock->sk;
+#endif
+
 	return 0;
 }
 
@@ -878,6 +900,10 @@ void sctp_udp_sock_stop(struct net *net)
 		udp_tunnel_sock_release(net->sctp.udp4_sock->sk_socket);
 		net->sctp.udp4_sock = NULL;
 	}
+	if (net->sctp.udp6_sock) {
+		udp_tunnel_sock_release(net->sctp.udp6_sock->sk_socket);
+		net->sctp.udp6_sock = NULL;
+	}
 }
 
 /* Register address family specific functions. */
Signature · 1 line
2.1.0
author maintainer
#6 lore raw

PATCH v3 6/16 sctp: add encap_err_lookup for udp encap socks

As it says in rfc6951#section-5.5:

"When receiving ICMP or ICMPv6 response packets, there might not be
enough bytes in the payload to identify the SCTP association that the SCTP packet triggering the ICMP or ICMPv6 packet belongs to. If a received ICMP or ICMPv6 packet cannot be related to a specific SCTP association or the verification tag cannot be verified, it MUST be discarded silently. In particular, this means that the SCTP stack MUST NOT rely on receiving ICMP or ICMPv6 messages. Implementation constraints could prevent processing received ICMP or ICMPv6 messages."

ICMP or ICMPv6 packets need to be handled, and this is implemented by udp encap sock .encap_err_lookup function.

The .encap_err_lookup function is called in __udp(6)_lib_err_encap() to confirm this path does need to be updated. For sctp, what we can do here is check if the corresponding asoc and transport exist.

Note that icmp packet process for sctp over udp is done by udp sock .encap_err_lookup(), and it means for now we can't do as much as sctp_v4/6_err() does. Also we can't do the two mappings mentioned in rfc6951#section-5.5.

---
 net/sctp/protocol.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)
net/sctp/protocol.c +19 -0
index 49b5d75..dd2d9c4 100644
@@ -848,6 +848,23 @@ static int sctp_udp_rcv(struct sock *sk, struct sk_buff *skb)
 	return 0;
 }
 
+static int sctp_udp_err_lookup(struct sock *sk, struct sk_buff *skb)
+{
+	struct sctp_association *asoc;
+	struct sctp_transport *t;
+	int family;
+
+	skb->transport_header += sizeof(struct udphdr);
+	family = (ip_hdr(skb)->version = 4) ? AF_INET : AF_INET6;
+	sk = sctp_err_lookup(dev_net(skb->dev), family, skb, sctp_hdr(skb),
+			     &asoc, &t);
+	if (!sk)
+		return -ENOENT;
+
+	sctp_err_finish(sk, t);
+	return 0;
+}
+
 int sctp_udp_sock_start(struct net *net)
 {
 	struct udp_tunnel_sock_cfg tuncfg = {NULL};
@@ -866,6 +883,7 @@ int sctp_udp_sock_start(struct net *net)
 
 	tuncfg.encap_type = 1;
 	tuncfg.encap_rcv = sctp_udp_rcv;
+	tuncfg.encap_err_lookup = sctp_udp_err_lookup;
 	setup_udp_tunnel_sock(net, sock, &tuncfg);
 	net->sctp.udp4_sock = sock->sk;
 
@@ -887,6 +905,7 @@ int sctp_udp_sock_start(struct net *net)
 
 	tuncfg.encap_type = 1;
 	tuncfg.encap_rcv = sctp_udp_rcv;
+	tuncfg.encap_err_lookup = sctp_udp_err_lookup;
 	setup_udp_tunnel_sock(net, sock, &tuncfg);
 	net->sctp.udp6_sock = sock->sk;
 #endif
Signature · 1 line
2.1.0
author maintainer
#7 lore raw

PATCH v3 7/16 sctp: add encap_port for netns sock asoc and transport

encap_port is added as per netns/sock/assoc/transport, and the latter one's encap_port inherits the former one's by default. The transport's encap_port value would mostly decide if one packet should go out with udp encapsulated or not.

This patch also allows users to set netns' encap_port by sysctl.

v1->v2:
- Change to define encap_port as __be16 for sctp_sock, asoc and transport.

---
 include/net/netns/sctp.h   |  2 ++
 include/net/sctp/structs.h |  6 ++++++
 net/sctp/associola.c       |  4 ++++
 net/sctp/protocol.c        |  3 +++
 net/sctp/socket.c          |  1 +
 net/sctp/sysctl.c          | 10 ++++++++++
 6 files changed, 26 insertions(+)
include/net/netns/sctp.h +2 -0
index f622945..6af7a39 100644
@@ -27,6 +27,8 @@ struct netns_sctp {
 	struct sock *udp6_sock;
 	/* udp tunneling listening port. */
 	int udp_port;
+	/* udp tunneling remote encap port. */
+	int encap_port;
 
 	/* This is the global local address list.
 	 * We actively maintain this complete list of addresses on
include/net/sctp/structs.h +6 -0
index 0bdff38..aa98e7e 100644
@@ -178,6 +178,8 @@ struct sctp_sock {
 	 */
 	__u32 hbinterval;
 
+	__be16 encap_port;
+
 	/* This is the max_retrans value for new associations. */
 	__u16 pathmaxrxt;
 
@@ -877,6 +879,8 @@ struct sctp_transport {
 	 */
 	unsigned long last_time_ecne_reduced;
 
+	__be16 encap_port;
+
 	/* This is the max_retrans value for the transport and will
 	 * be initialized from the assocs value.  This can be changed
 	 * using the SCTP_SET_PEER_ADDR_PARAMS socket option.
@@ -1790,6 +1794,8 @@ struct sctp_association {
 	 */
 	unsigned long hbinterval;
 
+	__be16 encap_port;
+
 	/* This is the max_retrans value for new transports in the
 	 * association.
 	 */
net/sctp/associola.c +4 -0
index fdb69d4..336df4b 100644
@@ -99,6 +99,8 @@ static struct sctp_association *sctp_association_init(
 	 */
 	asoc->hbinterval = msecs_to_jiffies(sp->hbinterval);
 
+	asoc->encap_port = sp->encap_port;
+
 	/* Initialize path max retrans value. */
 	asoc->pathmaxrxt = sp->pathmaxrxt;
 
@@ -624,6 +626,8 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc,
 	 */
 	peer->hbinterval = asoc->hbinterval;
 
+	peer->encap_port = asoc->encap_port;
+
 	/* Set the path max_retrans.  */
 	peer->pathmaxrxt = asoc->pathmaxrxt;
 
net/sctp/protocol.c +3 -0
index dd2d9c4..5b74187 100644
@@ -1359,6 +1359,9 @@ static int __net_init sctp_defaults_init(struct net *net)
 	/* Set udp tunneling listening port to default value */
 	net->sctp.udp_port = SCTP_DEFAULT_UDP_PORT;
 
+	/* Set remote encap port to 0 by default */
+	net->sctp.encap_port = 0;
+
 	/* Set SCOPE policy to enabled */
 	net->sctp.scope_policy = SCTP_SCOPE_POLICY_ENABLE;
 
net/sctp/socket.c +1 -0
index 53d0a41..09b94cd 100644
@@ -4876,6 +4876,7 @@ static int sctp_init_sock(struct sock *sk)
 	 * be modified via SCTP_PEER_ADDR_PARAMS
 	 */
 	sp->hbinterval  = net->sctp.hb_interval;
+	sp->encap_port  = htons(net->sctp.encap_port);
 	sp->pathmaxrxt  = net->sctp.max_retrans_path;
 	sp->pf_retrans  = net->sctp.pf_retrans;
 	sp->ps_retrans  = net->sctp.ps_retrans;
net/sctp/sysctl.c +1 -0
index c16c809..ecc1b5e 100644
@@ -36,6 +36,7 @@ static int rto_alpha_max = 1000;
 static int rto_beta_max = 1000;
 static int pf_expose_max = SCTP_PF_EXPOSE_MAX;
 static int ps_retrans_max = SCTP_PS_RETRANS_MAX;
+static int udp_port_max = 65535;
 
 static unsigned long max_autoclose_min = 0;
 static unsigned long max_autoclose_max @@ -291,6 +292,15 @@ static struct ctl_table sctp_net_table[] = {

.proc_handler = proc_dointvec,
},
{
+ .procname = "encap_port",
+ .data = &init_net.sctp.encap_port,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ .extra1 = SYSCTL_ZERO,
+ .extra2 = &udp_port_max,
+ },
+ {
.procname = "addr_scope_policy",
.data = &init_net.sctp.scope_policy,
.maxlen = sizeof(int),

Signature · 1 line
2.1.0
author maintainer
#8 lore raw

PATCH v3 8/16 sctp: add SCTP_REMOTE_UDP_ENCAPS_PORT sockopt

This patch is to implement:

rfc6951#section-6.1: Get or Set the Remote UDP Encapsulation Port Number

with the param of the struct:

  struct sctp_udpencaps {
    sctp_assoc_t sue_assoc_id;
    struct sockaddr_storage sue_address;
    uint16_t sue_port;
  };

the encap_port of sock, assoc or transport can be changed by users, which also means it allows the different transports of the same asoc to have different encap_port value.

v1->v2:
- no change.
v2->v3:
- fix the endian warning when setting values between encap_port and sue_port.

---
 include/uapi/linux/sctp.h |   7 +++
 net/sctp/socket.c         | 114 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 121 insertions(+)
include/uapi/linux/sctp.h +7 -0
index 28ad40d..cb78e7a 100644
@@ -140,6 +140,7 @@ typedef __s32 sctp_assoc_t;
 #define SCTP_ECN_SUPPORTED	130
 #define SCTP_EXPOSE_POTENTIALLY_FAILED_STATE	131
 #define SCTP_EXPOSE_PF_STATE	SCTP_EXPOSE_POTENTIALLY_FAILED_STATE
+#define SCTP_REMOTE_UDP_ENCAPS_PORT	132
 
 /* PR-SCTP policies */
 #define SCTP_PR_SCTP_NONE	0x0000
@@ -1197,6 +1198,12 @@ struct sctp_event {
 	uint8_t se_on;
 };
 
+struct sctp_udpencaps {
+	sctp_assoc_t sue_assoc_id;
+	struct sockaddr_storage sue_address;
+	uint16_t sue_port;
+};
+
 /* SCTP Stream schedulers */
 enum sctp_sched_type {
 	SCTP_SS_FCFS,
net/sctp/socket.c +114 -0

138 lines not on this page · raw message

Signature · 1 line
2.1.0
author maintainer
#9 lore raw

PATCH v3 9/16 sctp: allow changing transport encap_port by peer packets

As rfc6951#section-5.4 says:

"After finding the SCTP association (which
includes checking the verification tag), the UDP source port MUST be stored as the encapsulation port for the destination address the SCTP packet is received from (see Section 5.1).

When a non-encapsulated SCTP packet is received by the SCTP stack, the encapsulation of outgoing packets belonging to the same association and the corresponding destination address MUST be disabled."

transport encap_port should be updated by a validated incoming packet's udp src port.

We save the udp src port in sctp_input_cb->encap_port, and then update the transport in two places:

1. right after vtag is verified, which is required by RFC, and this allows the existent transports to be updated by the chunks that can only be processed on an asoc.

2. right before processing the 'init' where the transports are added, and this allows building a sctp over udp connection by client with the server not knowing the remote encap port.

3. when processing ootb_pkt and creating the temporary transport for the reply pkt.

Note that sctp_input_cb->header is removed, as it's not used any more in sctp.

v1->v2:
- Change encap_port as __be16 for sctp_input_cb.

---
 include/net/sctp/sm.h      |  1 +
 include/net/sctp/structs.h |  7 +------
 net/sctp/ipv6.c            |  1 +
 net/sctp/protocol.c        | 11 ++++++++++-
 net/sctp/sm_make_chunk.c   |  1 +
 net/sctp/sm_statefuns.c    |  2 ++
 6 files changed, 16 insertions(+), 7 deletions(-)
include/net/sctp/sm.h +1 -0
index 5c491a3..a499341 100644
@@ -380,6 +380,7 @@ sctp_vtag_verify(const struct sctp_chunk *chunk,
         if (ntohl(chunk->sctp_hdr->vtag) = asoc->c.my_vtag)
                 return 1;
 
+	chunk->transport->encap_port = SCTP_INPUT_CB(chunk->skb)->encap_port;
 	return 0;
 }
 
include/net/sctp/structs.h +1 -6
index aa98e7e..81464ae 100644
@@ -1120,14 +1120,9 @@ static inline void sctp_outq_cork(struct sctp_outq *q)
  * sctp_input_cb is currently used on rx and sock rx queue
  */
 struct sctp_input_cb {
-	union {
-		struct inet_skb_parm	h4;
-#if IS_ENABLED(CONFIG_IPV6)
-		struct inet6_skb_parm	h6;
-#endif
-	} header;
 	struct sctp_chunk *chunk;
 	struct sctp_af *af;
+	__be16 encap_port;
 };
 #define SCTP_INPUT_CB(__skb)	((struct sctp_input_cb *)&((__skb)->cb[0]))
 
net/sctp/ipv6.c +1 -0
index 8a58f42..a064bf2 100644
@@ -1053,6 +1053,7 @@ static struct inet_protosw sctpv6_stream_protosw = {
 
 static int sctp6_rcv(struct sk_buff *skb)
 {
+	memset(skb->cb, 0, sizeof(skb->cb));
 	return sctp_rcv(skb) ? -1 : 0;
 }
 
net/sctp/protocol.c +10 -1
index 5b74187..0d16e5e 100644
@@ -843,6 +843,9 @@ static int sctp_ctl_sock_init(struct net *net)
 
 static int sctp_udp_rcv(struct sock *sk, struct sk_buff *skb)
 {
+	memset(skb->cb, 0, sizeof(skb->cb));
+	SCTP_INPUT_CB(skb)->encap_port = udp_hdr(skb)->source;
+
 	skb_set_transport_header(skb, sizeof(struct udphdr));
 	sctp_rcv(skb);
 	return 0;
@@ -1139,9 +1142,15 @@ static struct inet_protosw sctp_stream_protosw = {
 	.flags      = SCTP_PROTOSW_FLAG
 };
 
+static int sctp4_rcv(struct sk_buff *skb)
+{
+	memset(skb->cb, 0, sizeof(skb->cb));
+	return sctp_rcv(skb);
+}
+
 /* Register with IP layer.  */
 static const struct net_protocol sctp_protocol = {
-	.handler     = sctp_rcv,
+	.handler     = sctp4_rcv,
 	.err_handler = sctp_v4_err,
 	.no_policy   = 1,
 	.netns_ok    = 1,
net/sctp/sm_make_chunk.c +1 -0
index 9a56ae2..21d0ff1 100644
@@ -2321,6 +2321,7 @@ int sctp_process_init(struct sctp_association *asoc, struct sctp_chunk *chunk,
 	 * added as the primary transport.  The source address seems to
 	 * be a better choice than any of the embedded addresses.
 	 */
+	asoc->encap_port = SCTP_INPUT_CB(chunk->skb)->encap_port;
 	if (!sctp_assoc_add_peer(asoc, peer_addr, gfp, SCTP_ACTIVE))
 		goto nomem;
 
net/sctp/sm_statefuns.c +2 -0
index c669f8b..8edab15 100644
@@ -6268,6 +6268,8 @@ static struct sctp_packet *sctp_ootb_pkt_new(
 	if (!transport)
 		goto nomem;
 
+	transport->encap_port = SCTP_INPUT_CB(chunk->skb)->encap_port;
+
 	/* Cache a route for the transport with the chunk's destination as
 	 * the source address.
 	 */
Signature · 1 line
2.1.0
author maintainer
#10 lore raw

PATCH v3 10/16 sctp: add udphdr to overhead when udp_port is set

sctp_mtu_payload() is for calculating the frag size before making chunks from a msg. So we should only add udphdr size to overhead when udp socks are listening, as only then sctp can handle the incoming sctp over udp packets and outgoing sctp over udp packets will be possible.

Note that we can't do this according to transport->encap_port, as different transports may be set to different values, while the chunks were made before choosing the transport, we could not be able to meet all rfc6951#section-5.6 recommends.

v1->v2:
- Add udp_port for sctp_sock to avoid a potential race issue, it will be used in xmit path in the next patch.

---
 include/net/sctp/sctp.h    | 7 +++++--
 include/net/sctp/structs.h | 1 +
 net/sctp/socket.c          | 1 +
 3 files changed, 7 insertions(+), 2 deletions(-)
include/net/sctp/sctp.h +5 -2
index bfd87a0..86f74f2 100644
@@ -578,10 +578,13 @@ static inline __u32 sctp_mtu_payload(const struct sctp_sock *sp,
 {
 	__u32 overhead = sizeof(struct sctphdr) + extra;
 
-	if (sp)
+	if (sp) {
 		overhead += sp->pf->af->net_header_len;
-	else
+		if (sp->udp_port)
+			overhead += sizeof(struct udphdr);
+	} else {
 		overhead += sizeof(struct ipv6hdr);
+	}
 
 	if (WARN_ON_ONCE(mtu && mtu <= overhead))
 		mtu = overhead;
include/net/sctp/structs.h +1 -0
index 81464ae..80f7149 100644
@@ -178,6 +178,7 @@ struct sctp_sock {
 	 */
 	__u32 hbinterval;
 
+	__be16 udp_port;
 	__be16 encap_port;
 
 	/* This is the max_retrans value for new associations. */
net/sctp/socket.c +1 -0
index 26e464b..8e1dcfb 100644
@@ -4928,6 +4928,7 @@ static int sctp_init_sock(struct sock *sk)
 	 * be modified via SCTP_PEER_ADDR_PARAMS
 	 */
 	sp->hbinterval  = net->sctp.hb_interval;
+	sp->udp_port    = htons(net->sctp.udp_port);
 	sp->encap_port  = htons(net->sctp.encap_port);
 	sp->pathmaxrxt  = net->sctp.max_retrans_path;
 	sp->pf_retrans  = net->sctp.pf_retrans;
Signature · 1 line
2.1.0
author maintainer
#11 lore raw

PATCH v3 11/16 sctp: call sk_setup_caps in sctp_packet_transmit instead

sk_setup_caps() was originally called in Commit 90017accff61 ("sctp:
Add GSO support"), as:

"We have to refresh this in case we are xmiting to more than one
transport at a time"

This actually happens in the loop of sctp_outq_flush_transports(), and it shouldn't be tied to gso, so move it out of gso part and before sctp_packet_pack().

---
 net/sctp/output.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
net/sctp/output.c +7 -6
index 1441eaf..fb16500 100644
@@ -508,12 +508,6 @@ static int sctp_packet_pack(struct sctp_packet *packet,
 					sizeof(struct inet6_skb_parm)));
 		skb_shinfo(head)->gso_segs = pkt_count;
 		skb_shinfo(head)->gso_size = GSO_BY_FRAGS;
-		rcu_read_lock();
-		if (skb_dst(head) != tp->dst) {
-			dst_hold(tp->dst);
-			sk_setup_caps(sk, tp->dst);
-		}
-		rcu_read_unlock();
 		goto chksum;
 	}
 
@@ -593,6 +587,13 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	}
 	skb_dst_set(head, dst);
 
+	rcu_read_lock();
+	if (__sk_dst_get(sk) != tp->dst) {
+		dst_hold(tp->dst);
+		sk_setup_caps(sk, tp->dst);
+	}
+	rcu_read_unlock();
+
 	/* pack up chunks */
 	pkt_count = sctp_packet_pack(packet, head, gso, gfp);
 	if (!pkt_count) {
Signature · 1 line
2.1.0
author maintainer
#12 lore raw

PATCH v3 12/16 sctp: support for sending packet over udp4 sock

This patch does what the rfc6951#section-5.3 says for ipv4:

"Within the UDP header, the source port MUST be the local UDP
encapsulation port number of the SCTP stack, and the destination port MUST be the remote UDP encapsulation port number maintained for the association and the destination address to which the packet is sent (see Section 5.1).

Because the SCTP packet is the UDP payload, the length of the UDP packet MUST be the length of the SCTP packet plus the size of the UDP header.

The SCTP checksum MUST be computed for IPv4 and IPv6, and the UDP checksum SHOULD be computed for IPv4 and IPv6."

Some places need to be adjusted in sctp_packet_transmit():

1. For non-gso packets, when transport's encap_port is set, sctp checksum has to be done in sctp_packet_pack(), as the outer udp will use ip_summed = CHECKSUM_PARTIAL to do the offload setting for checksum.

2. Delay calling dst_clone() and skb_dst_set() for non-udp packets until sctp_v4_xmit(), as for udp packets, skb_dst_set() is not needed before calling udp_tunnel_xmit_skb().

then in sctp_v4_xmit():

1. Go to udp_tunnel_xmit_skb() only when transport->encap_port and net->sctp.udp_port both are set, as these are one for dst port and another for src port.

2. For gso packet, SKB_GSO_UDP_TUNNEL_CSUM is set for gso_type, and with this udp checksum can be done in __skb_udp_tunnel_segment() for each segments after the sctp gso.

3. inner_mac_header and inner_transport_header are set, as these will be needed in __skb_udp_tunnel_segment() to find the right headers.

4. df and ttl are calculated, as these are the required params by

     udp_tunnel_xmit_skb().

5. nocheck param has to be false, as "the UDP checksum SHOULD be computed for IPv4 and IPv6", says in rfc6951#section-5.3.

v1->v2:
- Use sp->udp_port instead in sctp_v4_xmit(), which is more safe.

---
 net/sctp/output.c   |  9 +++------
 net/sctp/protocol.c | 41 ++++++++++++++++++++++++++++++-----------
 2 files changed, 33 insertions(+), 17 deletions(-)
net/sctp/output.c +3 -6
index fb16500..6614c9f 100644
@@ -514,8 +514,8 @@ static int sctp_packet_pack(struct sctp_packet *packet,
 	if (sctp_checksum_disable)
 		return 1;
 
-	if (!(skb_dst(head)->dev->features & NETIF_F_SCTP_CRC) ||
-	    dst_xfrm(skb_dst(head)) || packet->ipfragok) {
+	if (!(tp->dst->dev->features & NETIF_F_SCTP_CRC) ||
+	    dst_xfrm(tp->dst) || packet->ipfragok || tp->encap_port) {
 		struct sctphdr *sh  			(struct sctphdr *)skb_transport_header(head);
 
@@ -542,7 +542,6 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	struct sctp_association *asoc = tp->asoc;
 	struct sctp_chunk *chunk, *tmp;
 	int pkt_count, gso = 0;
-	struct dst_entry *dst;
 	struct sk_buff *head;
 	struct sctphdr *sh;
 	struct sock *sk;
@@ -579,13 +578,11 @@ int sctp_packet_transmit(struct sctp_packet *packet, gfp_t gfp)
 	sh->checksum = 0;
 
 	/* drop packet if no dst */
-	dst = dst_clone(tp->dst);
-	if (!dst) {
+	if (!tp->dst) {
 		IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTNOROUTES);
 		kfree_skb(head);
 		goto out;
 	}
-	skb_dst_set(head, dst);
 
 	rcu_read_lock();
 	if (__sk_dst_get(sk) != tp->dst) {
net/sctp/protocol.c +30 -11

55 lines not on this page · raw message

Signature · 1 line
2.1.0
author maintainer
#13 lore raw

PATCH v3 13/16 sctp: support for sending packet over udp6 sock

This one basically does the similar things in sctp_v6_xmit as does for udp4 sock in the last patch, just note that:

1. label needs to be calculated, as it's the param of

     udp_tunnel6_xmit_skb().

2. The 'nocheck' param of udp_tunnel6_xmit_skb() is false, as required by RFC.

v1->v2:
- Use sp->udp_port instead in sctp_v6_xmit(), which is more safe.

---
 net/sctp/ipv6.c | 43 ++++++++++++++++++++++++++++++++-----------
 1 file changed, 32 insertions(+), 11 deletions(-)
net/sctp/ipv6.c +32 -11
index a064bf2..814754d 100644
@@ -55,6 +55,7 @@
 #include <net/inet_common.h>
 #include <net/inet_ecn.h>
 #include <net/sctp/sctp.h>
+#include <net/udp_tunnel.h>
 
 #include <linux/uaccess.h>
 
@@ -191,33 +192,53 @@ static int sctp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 	return ret;
 }
 
-static int sctp_v6_xmit(struct sk_buff *skb, struct sctp_transport *transport)
+static int sctp_v6_xmit(struct sk_buff *skb, struct sctp_transport *t)
 {
+	struct dst_entry *dst = dst_clone(t->dst);
+	struct flowi6 *fl6 = &t->fl.u.ip6;
 	struct sock *sk = skb->sk;
 	struct ipv6_pinfo *np = inet6_sk(sk);
-	struct flowi6 *fl6 = &transport->fl.u.ip6;
 	__u8 tclass = np->tclass;
-	int res;
+	__be32 label;
 
 	pr_debug("%s: skb:%p, len:%d, src:%pI6 dst:%pI6\n", __func__, skb,
 		 skb->len, &fl6->saddr, &fl6->daddr);
 
-	if (transport->dscp & SCTP_DSCP_SET_MASK)
-		tclass = transport->dscp & SCTP_DSCP_VAL_MASK;
+	if (t->dscp & SCTP_DSCP_SET_MASK)
+		tclass = t->dscp & SCTP_DSCP_VAL_MASK;
 
 	if (INET_ECN_is_capable(tclass))
 		IP6_ECN_flow_xmit(sk, fl6->flowlabel);
 
-	if (!(transport->param_flags & SPP_PMTUD_ENABLE))
+	if (!(t->param_flags & SPP_PMTUD_ENABLE))
 		skb->ignore_df = 1;
 
 	SCTP_INC_STATS(sock_net(sk), SCTP_MIB_OUTSCTPPACKS);
 
-	rcu_read_lock();
-	res = ip6_xmit(sk, skb, fl6, sk->sk_mark, rcu_dereference(np->opt),
-		       tclass, sk->sk_priority);
-	rcu_read_unlock();
-	return res;
+	if (!t->encap_port || !sctp_sk(sk)->udp_port) {
+		int res;
+
+		skb_dst_set(skb, dst);
+		rcu_read_lock();
+		res = ip6_xmit(sk, skb, fl6, sk->sk_mark,
+			       rcu_dereference(np->opt),
+			       tclass, sk->sk_priority);
+		rcu_read_unlock();
+		return res;
+	}
+
+	if (skb_is_gso(skb))
+		skb_shinfo(skb)->gso_type |= SKB_GSO_UDP_TUNNEL_CSUM;
+
+	skb->encapsulation = 1;
+	skb_reset_inner_mac_header(skb);
+	skb_reset_inner_transport_header(skb);
+	skb_set_inner_ipproto(skb, IPPROTO_SCTP);
+	label = ip6_make_flowlabel(sock_net(sk), skb, fl6->flowlabel, true, fl6);
+
+	return udp_tunnel6_xmit_skb(dst, sk, skb, NULL, &fl6->saddr,
+				    &fl6->daddr, tclass, ip6_dst_hoplimit(dst),
+				    label, sctp_sk(sk)->udp_port, t->encap_port, false);
 }
 
 /* Returns the dst cache entry for the given source and destination ip
Signature · 1 line
2.1.0
author maintainer
#14 lore raw

PATCH v3 14/16 sctp: add the error cause for new encapsulation port restart

This patch is to add the function to make the abort chunk with the error cause for new encapsulation port restart, defined on Section 4.4 in draft-tuexen-tsvwg-sctp-udp-encaps-cons-03.

v1->v2:
- no change.
v2->v3:
- no need to call htons() when setting nep.cur_port/new_port.

---
 include/linux/sctp.h     | 20 ++++++++++++++++++++
 include/net/sctp/sm.h    |  3 +++
 net/sctp/sm_make_chunk.c | 20 ++++++++++++++++++++
 3 files changed, 43 insertions(+)
include/linux/sctp.h +20 -0
index 7673123..bb19265 100644
@@ -482,11 +482,13 @@ enum sctp_error {
 	 *  11  Restart of an association with new addresses
 	 *  12  User Initiated Abort
 	 *  13  Protocol Violation
+	 *  14  Restart of an Association with New Encapsulation Port
 	 */
 
 	SCTP_ERROR_RESTART         = cpu_to_be16(0x0b),
 	SCTP_ERROR_USER_ABORT      = cpu_to_be16(0x0c),
 	SCTP_ERROR_PROTO_VIOLATION = cpu_to_be16(0x0d),
+	SCTP_ERROR_NEW_ENCAP_PORT  = cpu_to_be16(0x0e),
 
 	/* ADDIP Section 3.3  New Error Causes
 	 *
@@ -793,4 +795,22 @@ enum {
 	SCTP_FLOWLABEL_VAL_MASK = 0xfffff
 };
 
+/* UDP Encapsulation
+ * draft-tuexen-tsvwg-sctp-udp-encaps-cons-03.html#section-4-4
+ *
+ *   The error cause indicating an "Restart of an Association with
+ *   New Encapsulation Port"
+ *
+ * 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |        Cause Code = 14        |       Cause Length = 8        |
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |   Current Encapsulation Port  |     New Encapsulation Port    |
+ * +-------------------------------+-------------------------------+
+ */
+struct sctp_new_encap_port_hdr {
+	__be16 cur_port;
+	__be16 new_port;
+};
+
 #endif /* __LINUX_SCTP_H__ */
include/net/sctp/sm.h +3 -0
index a499341..fd223c9 100644
@@ -221,6 +221,9 @@ struct sctp_chunk *sctp_make_violation_paramlen(
 struct sctp_chunk *sctp_make_violation_max_retrans(
 					const struct sctp_association *asoc,
 					const struct sctp_chunk *chunk);
+struct sctp_chunk *sctp_make_new_encap_port(
+					const struct sctp_association *asoc,
+					const struct sctp_chunk *chunk);
 struct sctp_chunk *sctp_make_heartbeat(const struct sctp_association *asoc,
 				       const struct sctp_transport *transport);
 struct sctp_chunk *sctp_make_heartbeat_ack(const struct sctp_association *asoc,
net/sctp/sm_make_chunk.c +20 -0
index 21d0ff1..f77484d 100644
@@ -1142,6 +1142,26 @@ struct sctp_chunk *sctp_make_violation_max_retrans(
 	return retval;
 }
 
+struct sctp_chunk *sctp_make_new_encap_port(const struct sctp_association *asoc,
+					    const struct sctp_chunk *chunk)
+{
+	struct sctp_new_encap_port_hdr nep;
+	struct sctp_chunk *retval;
+
+	retval = sctp_make_abort(asoc, chunk,
+				 sizeof(struct sctp_errhdr) + sizeof(nep));
+	if (!retval)
+		goto nodata;
+
+	sctp_init_cause(retval, SCTP_ERROR_NEW_ENCAP_PORT, sizeof(nep));
+	nep.cur_port = SCTP_INPUT_CB(chunk->skb)->encap_port;
+	nep.new_port = chunk->transport->encap_port;
+	sctp_addto_chunk(retval, sizeof(nep), &nep);
+
+nodata:
+	return retval;
+}
+
 /* Make a HEARTBEAT chunk.  */
 struct sctp_chunk *sctp_make_heartbeat(const struct sctp_association *asoc,
 				       const struct sctp_transport *transport)
Signature · 1 line
2.1.0
author maintainer
#15 lore raw

PATCH v3 15/16 sctp: handle the init chunk matching an existing asoc

This is from Section 4 of draft-tuexen-tsvwg-sctp-udp-encaps-cons-03, and it requires responding with an abort chunk with an error cause when the udp source port of the received init chunk doesn't match the encap port of the transport.

---
 net/sctp/sm_statefuns.c | 50 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 50 insertions(+)
net/sctp/sm_statefuns.c +50 -0
index 8edab15..244a5d8 100644
@@ -87,6 +87,13 @@ static enum sctp_disposition sctp_sf_tabort_8_4_8(
 					const union sctp_subtype type,
 					void *arg,
 					struct sctp_cmd_seq *commands);
+static enum sctp_disposition sctp_sf_new_encap_port(
+					struct net *net,
+					const struct sctp_endpoint *ep,
+					const struct sctp_association *asoc,
+					const union sctp_subtype type,
+					void *arg,
+					struct sctp_cmd_seq *commands);
 static struct sctp_sackhdr *sctp_sm_pull_sack(struct sctp_chunk *chunk);
 
 static enum sctp_disposition sctp_stop_t1_and_abort(
@@ -1493,6 +1500,10 @@ static enum sctp_disposition sctp_sf_do_unexpected_init(
 	if (!sctp_chunk_length_valid(chunk, sizeof(struct sctp_init_chunk)))
 		return sctp_sf_violation_chunklen(net, ep, asoc, type, arg,
 						  commands);
+
+	if (SCTP_INPUT_CB(chunk->skb)->encap_port != chunk->transport->encap_port)
+		return sctp_sf_new_encap_port(net, ep, asoc, type, arg, commands);
+
 	/* Grab the INIT header.  */
 	chunk->subh.init_hdr = (struct sctp_inithdr *)chunk->skb->data;
 
@@ -3392,6 +3403,45 @@ static enum sctp_disposition sctp_sf_tabort_8_4_8(
 
 	sctp_packet_append_chunk(packet, abort);
 
+	sctp_add_cmd_sf(commands, SCTP_CMD_SEND_PKT, SCTP_PACKET(packet));
+
+	SCTP_INC_STATS(net, SCTP_MIB_OUTCTRLCHUNKS);
+
+	sctp_sf_pdiscard(net, ep, asoc, type, arg, commands);
+	return SCTP_DISPOSITION_CONSUME;
+}
+
+/* Handling of SCTP Packets Containing an INIT Chunk Matching an
+ * Existing Associations when the udp encap port is incorrect.
+ *
+ * From Section 4 at draft-tuexen-tsvwg-sctp-udp-encaps-cons-03.
+ */
+static enum sctp_disposition sctp_sf_new_encap_port(
+					struct net *net,
+					const struct sctp_endpoint *ep,
+					const struct sctp_association *asoc,
+					const union sctp_subtype type,
+					void *arg,
+					struct sctp_cmd_seq *commands)
+{
+	struct sctp_packet *packet = NULL;
+	struct sctp_chunk *chunk = arg;
+	struct sctp_chunk *abort;
+
+	packet = sctp_ootb_pkt_new(net, asoc, chunk);
+	if (!packet)
+		return SCTP_DISPOSITION_NOMEM;
+
+	abort = sctp_make_new_encap_port(asoc, chunk);
+	if (!abort) {
+		sctp_ootb_pkt_free(packet);
+		return SCTP_DISPOSITION_NOMEM;
+	}
+
+	abort->skb->sk = ep->base.sk;
+
+	sctp_packet_append_chunk(packet, abort);
+
 	sctp_add_cmd_sf(commands, SCTP_CMD_SEND_PKT,
 			SCTP_PACKET(packet));
 
Signature · 1 line
2.1.0
author maintainer
#16 lore raw

PATCH v3 16/16 sctp: enable udp tunneling socks

This patch is to enable udp tunneling socks by calling sctp_udp_sock_start() in sctp_ctrlsock_init(), and sctp_udp_sock_stop() in sctp_ctrlsock_exit().

Also add sysctl udp_port to allow changing the listening sock's port by users.

Wit this patch, the whole sctp over udp feature can be enabled and used.

v1->v2:
- Also update ctl_sock udp_port in proc_sctp_do_udp_port() where netns udp_port gets changed.
v2->v3:
- call htons() when setting sk udp_port from netns udp_port.

---
 net/sctp/protocol.c |  5 +++++
 net/sctp/sysctl.c   | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 55 insertions(+)
net/sctp/protocol.c +5 -0
index be002b7..79fb4b5 100644
@@ -1469,6 +1469,10 @@ static int __net_init sctp_ctrlsock_init(struct net *net)
 	if (status)
 		pr_err("Failed to initialize the SCTP control sock\n");
 
+	status = sctp_udp_sock_start(net);
+	if (status)
+		pr_err("Failed to initialize the SCTP udp tunneling sock\n");
+
 	return status;
 }
 
@@ -1476,6 +1480,7 @@ static void __net_exit sctp_ctrlsock_exit(struct net *net)
 {
 	/* Free the control endpoint.  */
 	inet_ctl_sock_destroy(net->sctp.ctl_sock);
+	sctp_udp_sock_stop(net);
 }
 
 static struct pernet_operations sctp_ctrlsock_ops = {
net/sctp/sysctl.c +50 -0

68 lines not on this page · raw message

Signature · 1 line
2.1.0
#17 lore raw

Re: [PATCHv3 net-next 16/16] sctp: enable udp tunneling socks

Actually..

On Tue, Oct 13, 2020 at 03:27:41PM +0800, Xin Long wrote:
...

Quoted 2 lines

Also add sysctl udp_port to allow changing the listening sock's port by users.

...

Quoted 4 lines
---
 net/sctp/protocol.c |  5 +++++
 net/sctp/sysctl.c   | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 55 insertions(+)

Xin, sorry for not noticing this earlier, but we need a documentation update here for this new sysctl. This is important. Please add its entry in ip-sysctl.rst.

net/sctp/protocol.c static int __net_init sctp_ctrlsock_init(struct net *net) · line 1474

Quoted 12 lines
net/sctp/protocol.c +3 -0
index be002b7..79fb4b5 100644
@@ -1469,6 +1469,10 @@ static int __net_init sctp_ctrlsock_init(struct net *net)
 	if (status)
 		pr_err("Failed to initialize the SCTP control sock\n");
 
+	status = sctp_udp_sock_start(net);
+	if (status)
+		pr_err("Failed to initialize the SCTP udp tunneling sock\n");

^^^ upper case please.
Nit. There are other occurrences of this.

Quoted 2 lines
net/sctp/protocol.c +1 -0
+
 	return status;

...

net/sctp/protocol.c

Quoted 9 lines
net/sctp/protocol.c +9 -0
+	ret = proc_dointvec(&tbl, write, buffer, lenp, ppos);
+	if (write && ret = 0) {
+		struct sock *sk = net->sctp.ctl_sock;
+
+		if (new_value > max || new_value < min)
+			return -EINVAL;
+
+		net->sctp.udp_port = new_value;
+		sctp_udp_sock_stop(net);

So, if it would be disabling the encapsulation, it shouldn't be calling _start() then, right? Like

		if (new_value)
			ret = sctp_udp_sock_start(net);

Otherwise _start() here will call ..._bind() with port 0, which then will be a random one.

Quoted 19 lines
net/sctp/protocol.c +13 -0
+		ret = sctp_udp_sock_start(net);
+		if (ret)
+			net->sctp.udp_port = 0;
+
+		/* Update the value in the control socket */
+		lock_sock(sk);
+		sctp_sk(sk)->udp_port = htons(net->sctp.udp_port);
+		release_sock(sk);
+	}
+
+	return ret;
+}
+
 int sctp_sysctl_net_register(struct net *net)
 {
 	struct ctl_table *table;
Signature · 1 line
2.1.0
#18 lore raw

RE: [PATCHv3 net-next 16/16] sctp: enable udp tunneling socks

From: Marcelo Ricardo Leitner

Quoted 8 lines

Sent: 15 October 2020 18:43

Actually..

On Tue, Oct 13, 2020 at 03:27:41PM +0800, Xin Long wrote:
...

Quoted 2 lines

Also add sysctl udp_port to allow changing the listening sock's port by users.

I've not read through this change...

But surely the UDP port (or ports) that you need to use will depend on the remote system's configuration.

So they need to be a property of the socket, not a system-wide value.

I can easily imagine my M3UA code connecting to different network providers which have decided to use different UDP port numbers.

David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)

#19 lore raw

Re: [PATCHv3 net-next 16/16] sctp: enable udp tunneling socks

Quoted 9 lines · David Laight

From: Marcelo Ricardo Leitner

Quoted 6 lines

Sent: 15 October 2020 18:43

Actually..

On Tue, Oct 13, 2020 at 03:27:41PM +0800, Xin Long wrote:
...

Quoted 7 lines
Quoted 2 lines
Quoted 2 lines

Also add sysctl udp_port to allow changing the listening sock's port by users.

I've not read through this change...

But surely the UDP port (or ports) that you need to use will depend on the remote system's configuration.

The local encapsulation port can be system wide, the remote encapsulation port is per remote address. See https://tools.ietf.org/html/rfc6951#section-5.1

Best regards
Michael

Quoted 14 lines

So they need to be a property of the socket, not a system-wide value.

I can easily imagine my M3UA code connecting to different network providers which have decided to use different UDP port numbers.

David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)

author maintainer
#20 lore raw

Re: [PATCHv3 net-next 16/16] sctp: enable udp tunneling socks

In reply to Marcelo Ricardo Leitner
Quoted 8 lines · Marcelo Ricardo Leitner

Actually..

On Tue, Oct 13, 2020 at 03:27:41PM +0800, Xin Long wrote:
...

Quoted 2 lines

Also add sysctl udp_port to allow changing the listening sock's port by users.

...

Quoted 8 lines
Quoted 4 lines
---
 net/sctp/protocol.c |  5 +++++
 net/sctp/sysctl.c   | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 55 insertions(+)

Xin, sorry for not noticing this earlier, but we need a documentation update here for this new sysctl. This is important. Please add its entry in ip-sysctl.rst.

no problem, I will add it.

Quoted 10 lines
Quoted 10 lines
net/sctp/protocol.c +0 -0
index be002b7..79fb4b5 100644
@@ -1469,6 +1469,10 @@ static int __net_init sctp_ctrlsock_init(struct net *net)
      if (status)
              pr_err("Failed to initialize the SCTP control sock\n");
 
Quoted 5 lines

net/sctp/protocol.c static int __net_init sctp_ctrlsock_init(struct net *net) · line 1474

Quoted 3 lines
net/sctp/protocol.c +3 -0
+     status = sctp_udp_sock_start(net);
+     if (status)
+             pr_err("Failed to initialize the SCTP udp tunneling sock\n");

^^^ upper case please.
Nit. There are other occurrences of this.

You mean we can remove this log, as it's been handled well in sctp_udp_sock_start()?

Quoted 17 lines
Quoted 3 lines
net/sctp/protocol.c +1 -0
+
      return status;

...

net/sctp/protocol.c

Quoted 9 lines
net/sctp/protocol.c +9 -0
+     ret = proc_dointvec(&tbl, write, buffer, lenp, ppos);
+     if (write && ret = 0) {
+             struct sock *sk = net->sctp.ctl_sock;
+
+             if (new_value > max || new_value < min)
+                     return -EINVAL;
+
+             net->sctp.udp_port = new_value;
+             sctp_udp_sock_stop(net);

So, if it would be disabling the encapsulation, it shouldn't be calling _start() then, right? Like

Quoted 5 lines
                if (new_value)
                        ret = sctp_udp_sock_start(net);

Otherwise _start() here will call ..._bind() with port 0, which then will be a random one.

right, somehow I thought it wouldn't bind with port 0.

Thanks.

Quoted 20 lines
Quoted 20 lines
net/sctp/protocol.c +13 -1
+             ret = sctp_udp_sock_start(net);
+             if (ret)
+                     net->sctp.udp_port = 0;
+
+             /* Update the value in the control socket */
+             lock_sock(sk);
+             sctp_sk(sk)->udp_port = htons(net->sctp.udp_port);
+             release_sock(sk);
+     }
+
+     return ret;
+}
+
 int sctp_sysctl_net_register(struct net *net)
 {
      struct ctl_table *table;
--

2.1.0

author maintainer
#21 lore raw

Re: [PATCHv3 net-next 16/16] sctp: enable udp tunneling socks

Quoted 31 lines · Xin Long
Quoted 8 lines · Marcelo Ricardo Leitner

Actually..

On Tue, Oct 13, 2020 at 03:27:41PM +0800, Xin Long wrote:
...

Quoted 2 lines

Also add sysctl udp_port to allow changing the listening sock's port by users.

...

Quoted 8 lines
Quoted 4 lines
---
 net/sctp/protocol.c |  5 +++++
 net/sctp/sysctl.c   | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 55 insertions(+)

Xin, sorry for not noticing this earlier, but we need a documentation update here for this new sysctl. This is important. Please add its entry in ip-sysctl.rst.

no problem, I will add it.

Quoted 10 lines
Quoted 10 lines
net/sctp/protocol.c +0 -0
index be002b7..79fb4b5 100644
@@ -1469,6 +1469,10 @@ static int __net_init sctp_ctrlsock_init(struct net *net)
      if (status)
              pr_err("Failed to initialize the SCTP control sock\n");
 
Quoted 7 lines
Quoted 5 lines

net/sctp/protocol.c static int __net_init sctp_ctrlsock_init(struct net *net) · line 1474

Quoted 3 lines
net/sctp/protocol.c +3 -0
+     status = sctp_udp_sock_start(net);
+     if (status)
+             pr_err("Failed to initialize the SCTP udp tunneling sock\n");

^^^ upper case please.
Nit. There are other occurrences of this.

You mean we can remove this log, as it's been handled well in sctp_udp_sock_start()?

This one is actually OK :D
I've updated 'udp' with 'UDP' for all code annotations in this patchset.

will post v4.

Thanks.

Quoted 46 lines
Quoted 18 lines
Quoted 4 lines
net/sctp/protocol.c +1 -0
+
      return status;

...

net/sctp/protocol.c

Quoted 9 lines
net/sctp/protocol.c +9 -0
+     ret = proc_dointvec(&tbl, write, buffer, lenp, ppos);
+     if (write && ret = 0) {
+             struct sock *sk = net->sctp.ctl_sock;
+
+             if (new_value > max || new_value < min)
+                     return -EINVAL;
+
+             net->sctp.udp_port = new_value;
+             sctp_udp_sock_stop(net);

So, if it would be disabling the encapsulation, it shouldn't be calling _start() then, right? Like

Quoted 5 lines
                if (new_value)
                        ret = sctp_udp_sock_start(net);

Otherwise _start() here will call ..._bind() with port 0, which then will be a random one.

right, somehow I thought it wouldn't bind with port 0.

Thanks.

Quoted 20 lines
Quoted 20 lines
net/sctp/protocol.c +13 -1
+             ret = sctp_udp_sock_start(net);
+             if (ret)
+                     net->sctp.udp_port = 0;
+
+             /* Update the value in the control socket */
+             lock_sock(sk);
+             sctp_sk(sk)->udp_port = htons(net->sctp.udp_port);
+             release_sock(sk);
+     }
+
+     return ret;
+}
+
 int sctp_sysctl_net_register(struct net *net)
 {
      struct ctl_table *table;
--

2.1.0

maintainer Acked-by on v3 2/16
#22 lore raw

Re: [PATCHv3 net-next 02/16] udp6: move the mss check after udp gso tunnel processing

In reply to Xin Long
Quoted 9 lines · Xin Long

For some protocol's gso, like SCTP, it's using GSO_BY_FRAGS for gso_size. When using UDP to encapsulate its packet, it will return error in udp6_ufo_fragment() as skb->len < gso_size, and it will never go to the gso tunnel processing.

So we should move this check after udp gso tunnel processing, the same as udp4_ufo_fragment() does.

Quoted 6 lines

v1->v2:
- no change.
v2->v3:
- not do any cleanup.

Thanks for revising

maintainer
#23 lore raw
In reply to Xin Long
Quoted 4 lines · On Tue, 13 Oct 2020 15:27:25 +0800 Xin Long wrote:

Description From the RFC:

The Main Reasons:

Quoted 8 lines

o To allow SCTP traffic to pass through legacy NATs, which do not provide native SCTP support as specified in [BEHAVE] and [NATSUPP].

o To allow SCTP to be implemented on hosts that do not provide direct access to the IP layer. In particular, applications can use their own SCTP implementation if the operating system does not provide one.

Marcelo, Neil - please review if you want this one to make 5.10.

Otherwise we can defer until after the merge window.

#24 lore raw
Quoted 9 lines · Jakub Kicinski
Quoted 8 lines · On Tue, 13 Oct 2020 15:27:25 +0800 Xin Long wrote:

Description From the RFC:

The Main Reasons:

o To allow SCTP traffic to pass through legacy NATs, which do not provide native SCTP support as specified in [BEHAVE] and [NATSUPP].

Quoted 6 lines
Quoted 4 lines

o To allow SCTP to be implemented on hosts that do not provide direct access to the IP layer. In particular, applications can use their own SCTP implementation if the operating system does not provide one.

Marcelo, Neil - please review if you want this one to make 5.10.

On the works!

Quoted 1 line

Otherwise we can defer until after the merge window.

Probably not needed.

Thanks,
Marcelo