| blob | e7bf3979facb95005fdbae1070013d3953a1c822 |
1 This document describes how to use the kernel's L2TP drivers to
2 provide L2TP functionality. L2TP is a protocol that tunnels one or
3 more sessions over an IP tunnel. It is commonly used for VPNs
4 (L2TP/IPSec) and by ISPs to tunnel subscriber PPP sessions over an IP
5 network infrastructure. With L2TPv3, it is also useful as a Layer-2
6 tunneling infrastructure.
8 Features
9 ========
11 L2TPv2 (PPP over L2TP (UDP tunnels)).
12 L2TPv3 ethernet pseudowires.
13 L2TPv3 PPP pseudowires.
14 L2TPv3 IP encapsulation.
15 Netlink sockets for L2TPv3 configuration management.
17 History
18 =======
20 The original pppol2tp driver was introduced in 2.6.23 and provided
21 L2TPv2 functionality (rfc2661). L2TPv2 is used to tunnel one or more PPP
22 sessions over a UDP tunnel.
24 L2TPv3 (rfc3931) changes the protocol to allow different frame types
25 to be passed over an L2TP tunnel by moving the PPP-specific parts of
26 the protocol out of the core L2TP packet headers. Each frame type is
27 known as a pseudowire type. Ethernet, PPP, HDLC, Frame Relay and ATM
28 pseudowires for L2TP are defined in separate RFC standards. Another
29 change for L2TPv3 is that it can be carried directly over IP with no
30 UDP header (UDP is optional). It is also possible to create static
31 unmanaged L2TPv3 tunnels manually without a control protocol
32 (userspace daemon) to manage them.
34 To support L2TPv3, the original pppol2tp driver was split up to
35 separate the L2TP and PPP functionality. Existing L2TPv2 userspace
36 apps should be unaffected as the original pppol2tp sockets API is
37 retained. L2TPv3, however, uses netlink to manage L2TPv3 tunnels and
38 sessions.
40 Design
41 ======
43 The L2TP protocol separates control and data frames. The L2TP kernel
44 drivers handle only L2TP data frames; control frames are always
45 handled by userspace. L2TP control frames carry messages between L2TP
46 clients/servers and are used to setup / teardown tunnels and
47 sessions. An L2TP client or server is implemented in userspace.
49 Each L2TP tunnel is implemented using a UDP or L2TPIP socket; L2TPIP
50 provides L2TPv3 IP encapsulation (no UDP) and is implemented using a
51 new l2tpip socket family. The tunnel socket is typically created by
52 userspace, though for unmanaged L2TPv3 tunnels, the socket can also be
53 created by the kernel. Each L2TP session (pseudowire) gets a network
54 interface instance. In the case of PPP, these interfaces are created
55 indirectly by pppd using a pppol2tp socket. In the case of ethernet,
56 the netdevice is created upon a netlink request to create an L2TPv3
57 ethernet pseudowire.
59 For PPP, the PPPoL2TP driver, net/l2tp/l2tp_ppp.c, provides a
60 mechanism by which PPP frames carried through an L2TP session are
61 passed through the kernel's PPP subsystem. The standard PPP daemon,
62 pppd, handles all PPP interaction with the peer. PPP network
63 interfaces are created for each local PPP endpoint. The kernel's PPP
64 subsystem arranges for PPP control frames to be delivered to pppd,
65 while data frames are forwarded as usual.
67 For ethernet, the L2TPETH driver, net/l2tp/l2tp_eth.c, implements a
68 netdevice driver, managing virtual ethernet devices, one per
69 pseudowire. These interfaces can be managed using standard Linux tools
70 such as "ip" and "ifconfig". If only IP frames are passed over the
71 tunnel, the interface can be given an IP addresses of itself and its
72 peer. If non-IP frames are to be passed over the tunnel, the interface
73 can be added to a bridge using brctl. All L2TP datapath protocol
74 functions are handled by the L2TP core driver.
76 Each tunnel and session within a tunnel is assigned a unique tunnel_id
77 and session_id. These ids are carried in the L2TP header of every
78 control and data packet. (Actually, in L2TPv3, the tunnel_id isn't
79 present in data frames - it is inferred from the IP connection on
80 which the packet was received.) The L2TP driver uses the ids to lookup
81 internal tunnel and/or session contexts to determine how to handle the
82 packet. Zero tunnel / session ids are treated specially - zero ids are
83 never assigned to tunnels or sessions in the network. In the driver,
84 the tunnel context keeps a reference to the tunnel UDP or L2TPIP
85 socket. The session context holds data that lets the driver interface
86 to the kernel's network frame type subsystems, i.e. PPP, ethernet.
88 Userspace Programming
89 =====================
91 For L2TPv2, there are a number of requirements on the userspace L2TP
92 daemon in order to use the pppol2tp driver.
94 1. Use a UDP socket per tunnel.
96 2. Create a single PPPoL2TP socket per tunnel bound to a special null
97 session id. This is used only for communicating with the driver but
98 must remain open while the tunnel is active. Opening this tunnel
99 management socket causes the driver to mark the tunnel socket as an
100 L2TP UDP encapsulation socket and flags it for use by the
101 referenced tunnel id. This hooks up the UDP receive path via
102 udp_encap_rcv() in net/ipv4/udp.c. PPP data frames are never passed
103 in this special PPPoX socket.
105 3. Create a PPPoL2TP socket per L2TP session. This is typically done
106 by starting pppd with the pppol2tp plugin and appropriate
107 arguments. A PPPoL2TP tunnel management socket (Step 2) must be
108 created before the first PPPoL2TP session socket is created.
110 When creating PPPoL2TP sockets, the application provides information
111 to the driver about the socket in a socket connect() call. Source and
112 destination tunnel and session ids are provided, as well as the file
113 descriptor of a UDP socket. See struct pppol2tp_addr in
114 include/linux/if_ppp.h. Note that zero tunnel / session ids are
115 treated specially. When creating the per-tunnel PPPoL2TP management
116 socket in Step 2 above, zero source and destination session ids are
117 specified, which tells the driver to prepare the supplied UDP file
118 descriptor for use as an L2TP tunnel socket.
120 Userspace may control behavior of the tunnel or session using
121 setsockopt and ioctl on the PPPoX socket. The following socket
122 options are supported:-
124 DEBUG - bitmask of debug message categories. See below.
125 SENDSEQ - 0 => don't send packets with sequence numbers
126 1 => send packets with sequence numbers
127 RECVSEQ - 0 => receive packet sequence numbers are optional
128 1 => drop receive packets without sequence numbers
129 LNSMODE - 0 => act as LAC.
130 1 => act as LNS.
131 REORDERTO - reorder timeout (in millisecs). If 0, don't try to reorder.
133 Only the DEBUG option is supported by the special tunnel management
134 PPPoX socket.
136 In addition to the standard PPP ioctls, a PPPIOCGL2TPSTATS is provided
137 to retrieve tunnel and session statistics from the kernel using the
138 PPPoX socket of the appropriate tunnel or session.
140 For L2TPv3, userspace must use the netlink API defined in
141 include/linux/l2tp.h to manage tunnel and session contexts. The
142 general procedure to create a new L2TP tunnel with one session is:-
144 1. Open a GENL socket using L2TP_GENL_NAME for configuring the kernel
145 using netlink.
147 2. Create a UDP or L2TPIP socket for the tunnel.
149 3. Create a new L2TP tunnel using a L2TP_CMD_TUNNEL_CREATE
150 request. Set attributes according to desired tunnel parameters,
151 referencing the UDP or L2TPIP socket created in the previous step.
153 4. Create a new L2TP session in the tunnel using a
154 L2TP_CMD_SESSION_CREATE request.
156 The tunnel and all of its sessions are closed when the tunnel socket
157 is closed. The netlink API may also be used to delete sessions and
158 tunnels. Configuration and status info may be set or read using netlink.
160 The L2TP driver also supports static (unmanaged) L2TPv3 tunnels. These
161 are where there is no L2TP control message exchange with the peer to
162 setup the tunnel; the tunnel is configured manually at each end of the
163 tunnel. There is no need for an L2TP userspace application in this
164 case -- the tunnel socket is created by the kernel and configured
165 using parameters sent in the L2TP_CMD_TUNNEL_CREATE netlink
166 request. The "ip" utility of iproute2 has commands for managing static
167 L2TPv3 tunnels; do "ip l2tp help" for more information.
169 Debugging
170 =========
172 The driver supports a flexible debug scheme where kernel trace
173 messages may be optionally enabled per tunnel and per session. Care is
174 needed when debugging a live system since the messages are not
175 rate-limited and a busy system could be swamped. Userspace uses
176 setsockopt on the PPPoX socket to set a debug mask.
178 The following debug mask bits are available:
180 PPPOL2TP_MSG_DEBUG verbose debug (if compiled in)
181 PPPOL2TP_MSG_CONTROL userspace - kernel interface
182 PPPOL2TP_MSG_SEQ sequence numbers handling
183 PPPOL2TP_MSG_DATA data packets
185 If enabled, files under a l2tp debugfs directory can be used to dump
186 kernel state about L2TP tunnels and sessions. To access it, the
187 debugfs filesystem must first be mounted.
189 # mount -t debugfs debugfs /debug
191 Files under the l2tp directory can then be accessed.
193 # cat /debug/l2tp/tunnels
195 The debugfs files should not be used by applications to obtain L2TP
196 state information because the file format is subject to change. It is
197 implemented to provide extra debug information to help diagnose
198 problems.) Users should use the netlink API.
200 /proc/net/pppol2tp is also provided for backwards compaibility with
201 the original pppol2tp driver. It lists information about L2TPv2
202 tunnels and sessions only. Its use is discouraged.
204 Unmanaged L2TPv3 Tunnels
205 ========================
207 Some commercial L2TP products support unmanaged L2TPv3 ethernet
208 tunnels, where there is no L2TP control protocol; tunnels are
209 configured at each side manually. New commands are available in
210 iproute2's ip utility to support this.
212 To create an L2TPv3 ethernet pseudowire between local host 192.168.1.1
213 and peer 192.168.1.2, using IP addresses 10.5.1.1 and 10.5.1.2 for the
214 tunnel endpoints:-
216 # modprobe l2tp_eth
217 # modprobe l2tp_netlink
219 # ip l2tp add tunnel tunnel_id 1 peer_tunnel_id 1 udp_sport 5000 \
220 udp_dport 5000 encap udp local 192.168.1.1 remote 192.168.1.2
221 # ip l2tp add session tunnel_id 1 session_id 1 peer_session_id 1
222 # ifconfig -a
223 # ip addr add 10.5.1.2/32 peer 10.5.1.1/32 dev l2tpeth0
224 # ifconfig l2tpeth0 up
226 Choose IP addresses to be the address of a local IP interface and that
227 of the remote system. The IP addresses of the l2tpeth0 interface can be
228 anything suitable.
230 Repeat the above at the peer, with ports, tunnel/session ids and IP
231 addresses reversed. The tunnel and session IDs can be any non-zero
232 32-bit number, but the values must be reversed at the peer.
234 Host 1 Host2
235 udp_sport=5000 udp_sport=5001
236 udp_dport=5001 udp_dport=5000
237 tunnel_id=42 tunnel_id=45
238 peer_tunnel_id=45 peer_tunnel_id=42
239 session_id=128 session_id=5196755
240 peer_session_id=5196755 peer_session_id=128
242 When done at both ends of the tunnel, it should be possible to send
243 data over the network. e.g.
245 # ping 10.5.1.1
248 Sample Userspace Code
249 =====================
251 1. Create tunnel management PPPoX socket
253 kernel_fd = socket(AF_PPPOX, SOCK_DGRAM, PX_PROTO_OL2TP);
254 if (kernel_fd >= 0) {
255 struct sockaddr_pppol2tp sax;
256 struct sockaddr_in const *peer_addr;
258 peer_addr = l2tp_tunnel_get_peer_addr(tunnel);
259 memset(&sax, 0, sizeof(sax));
260 sax.sa_family = AF_PPPOX;
261 sax.sa_protocol = PX_PROTO_OL2TP;
262 sax.pppol2tp.fd = udp_fd; /* fd of tunnel UDP socket */
263 sax.pppol2tp.addr.sin_addr.s_addr = peer_addr->sin_addr.s_addr;
264 sax.pppol2tp.addr.sin_port = peer_addr->sin_port;
265 sax.pppol2tp.addr.sin_family = AF_INET;
266 sax.pppol2tp.s_tunnel = tunnel_id;
267 sax.pppol2tp.s_session = 0; /* special case: mgmt socket */
268 sax.pppol2tp.d_tunnel = 0;
269 sax.pppol2tp.d_session = 0; /* special case: mgmt socket */
271 if(connect(kernel_fd, (struct sockaddr *)&sax, sizeof(sax) ) < 0 ) {
272 perror("connect failed");
273 result = -errno;
274 goto err;
275 }
276 }
278 2. Create session PPPoX data socket
280 struct sockaddr_pppol2tp sax;
281 int fd;
283 /* Note, the target socket must be bound already, else it will not be ready */
284 sax.sa_family = AF_PPPOX;
285 sax.sa_protocol = PX_PROTO_OL2TP;
286 sax.pppol2tp.fd = tunnel_fd;
287 sax.pppol2tp.addr.sin_addr.s_addr = addr->sin_addr.s_addr;
288 sax.pppol2tp.addr.sin_port = addr->sin_port;
289 sax.pppol2tp.addr.sin_family = AF_INET;
290 sax.pppol2tp.s_tunnel = tunnel_id;
291 sax.pppol2tp.s_session = session_id;
292 sax.pppol2tp.d_tunnel = peer_tunnel_id;
293 sax.pppol2tp.d_session = peer_session_id;
295 /* session_fd is the fd of the session's PPPoL2TP socket.
296 * tunnel_fd is the fd of the tunnel UDP socket.
297 */
298 fd = connect(session_fd, (struct sockaddr *)&sax, sizeof(sax));
299 if (fd < 0 ) {
300 return -errno;
301 }
302 return 0;
304 Internal Implementation
305 =======================
307 The driver keeps a struct l2tp_tunnel context per L2TP tunnel and a
308 struct l2tp_session context for each session. The l2tp_tunnel is
309 always associated with a UDP or L2TP/IP socket and keeps a list of
310 sessions in the tunnel. The l2tp_session context keeps kernel state
311 about the session. It has private data which is used for data specific
312 to the session type. With L2TPv2, the session always carried PPP
313 traffic. With L2TPv3, the session can also carry ethernet frames
314 (ethernet pseudowire) or other data types such as ATM, HDLC or Frame
315 Relay.
317 When a tunnel is first opened, the reference count on the socket is
318 increased using sock_hold(). This ensures that the kernel socket
319 cannot be removed while L2TP's data structures reference it.
321 Some L2TP sessions also have a socket (PPP pseudowires) while others
322 do not (ethernet pseudowires). We can't use the socket reference count
323 as the reference count for session contexts. The L2TP implementation
324 therefore has its own internal reference counts on the session
325 contexts.
327 To Do
328 =====
330 Add L2TP tunnel switching support. This would route tunneled traffic
331 from one L2TP tunnel into another. Specified in
332 http://tools.ietf.org/html/draft-ietf-l2tpext-tunnel-switching-08
334 Add L2TPv3 VLAN pseudowire support.
336 Add L2TPv3 IP pseudowire support.
338 Add L2TPv3 ATM pseudowire support.
340 Miscellaneous
341 =============
343 The L2TP drivers were developed as part of the OpenL2TP project by
344 Katalix Systems Ltd. OpenL2TP is a full-featured L2TP client / server,
345 designed from the ground up to have the L2TP datapath in the
346 kernel. The project also implemented the pppol2tp plugin for pppd
347 which allows pppd to use the kernel driver. Details can be found at
348 http://www.openl2tp.org.