1.\" $NetBSD: route.4,v 1.31 2017/07/03 21:30:58 wiz Exp $ 2.\" 3.\" Copyright (c) 1990, 1991, 1993 4.\" The Regents of the University of California. All rights reserved. 5.\" 6.\" Redistribution and use in source and binary forms, with or without 7.\" modification, are permitted provided that the following conditions 8.\" are met: 9.\" 1. Redistributions of source code must retain the above copyright 10.\" notice, this list of conditions and the following disclaimer. 11.\" 2. Redistributions in binary form must reproduce the above copyright 12.\" notice, this list of conditions and the following disclaimer in the 13.\" documentation and/or other materials provided with the distribution. 14.\" 3. Neither the name of the University nor the names of its contributors 15.\" may be used to endorse or promote products derived from this software 16.\" without specific prior written permission. 17.\" 18.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 19.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 20.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 21.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 22.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 23.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 24.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 25.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 26.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 27.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 28.\" SUCH DAMAGE. 29.\" 30.\" @(#)route.4 8.6 (Berkeley) 4/19/94 31.\" 32.Dd April 11, 2017 33.Dt ROUTE 4 34.Os 35.Sh NAME 36.Nm route 37.Nd kernel packet forwarding database 38.Sh SYNOPSIS 39.In sys/socket.h 40.In net/if.h 41.In net/route.h 42.Ft int 43.Fn socket PF_ROUTE SOCK_RAW "int family" 44.Sh DESCRIPTION 45.Ux 46provides some packet routing facilities. 47The kernel maintains a routing information database, which 48is used in selecting the appropriate network interface when 49transmitting packets. 50.Pp 51A user process (or possibly multiple co-operating processes) 52maintains this database by sending messages over a special kind 53of socket. 54This supplants fixed size 55.Xr ioctl 2 Ns 's 56used in earlier releases. 57Routing table changes may only be carried out by the super user. 58.Pp 59The operating system may spontaneously emit routing messages in response 60to external events, such as receipt of a redirect, or failure to 61locate a suitable route for a request. 62The message types are described in greater detail below. 63.Pp 64Routing database entries come in two flavors: for a specific 65host, or for all hosts on a generic subnetwork (as specified 66by a bit mask and value under the mask. 67The effect of wildcard or default route may be achieved by using 68a mask of all zeros, and there may be hierarchical routes. 69.Pp 70When the system is booted and addresses are assigned 71to the network interfaces, each protocol family 72installs a routing table entry for each interface when it is ready for traffic. 73Normally the protocol specifies the route 74through each interface as a 75.Dq direct 76connection to the destination host 77or network. 78If the route is direct, the transport layer of a protocol family 79usually requests the packet be sent to the same host specified in 80the packet. 81Otherwise, the interface is requested to address the packet to the 82gateway listed in the routing entry (i.e. the packet is forwarded). 83.Pp 84When routing a packet, 85the kernel will attempt to find 86the most specific route matching the destination. 87(If there are two different mask and value-under-the-mask pairs 88that match, the more specific is the one with more bits in the mask. 89A route to a host is regarded as being supplied with a mask of 90as many ones as there are bits in the destination). 91If no entry is found, the destination is declared to be unreachable, 92and a routing\-miss message is generated if there are any 93listeners on the routing control socket described below. 94.Pp 95A wildcard routing entry is specified with a zero 96destination address value, and a mask of all zeroes. 97Wildcard routes will be used 98when the system fails to find other routes matching the 99destination. 100The combination of wildcard routes and routing redirects can provide 101an economical mechanism for routing traffic. 102.Pp 103One opens the channel for passing routing control messages 104by using the socket call shown in the synopsis above: 105.Pp 106The 107.Fa family 108parameter may be 109.Dv AF_UNSPEC 110which will provide 111routing information for all address families, or can be restricted 112to a specific address family by specifying which one is desired. 113There can be more than one routing socket open per system. 114.Pp 115Messages are formed by a header followed by a small 116number of sockaddrs (now variable length particularly 117in the 118.Tn ISO 119case), interpreted by position, and delimited 120by the new length entry in the sockaddr. 121An example of a message with four addresses might be an 122.Tn ISO 123redirect: 124Destination, Netmask, Gateway, and Author of the redirect. 125The interpretation of which address are present is given by a 126bit mask within the header, and the sequence is least significant 127to most significant bit within the vector. 128.Pp 129Any messages sent to the kernel are returned, and copies are sent 130to all interested listeners. 131The exception to this is a new address marked as tentative, where copies 132will be sent once Duplicate Address Detection has completed and 133the tentative flag cleared or the duplicated flag set. 134Also, new address messages will also be emitted when other flags on the address 135change such as deprecated and detached. 136The kernel will provide the process ID for the sender, and the 137sender may use an additional sequence field to distinguish between 138outstanding messages. 139However, message replies may be lost when kernel buffers are exhausted. 140.Pp 141The kernel may reject certain messages, and will indicate this 142by filling in the 143.Fa rtm_errno 144field. 145The routing code returns 146.Er EEXIST 147if 148requested to duplicate an existing entry, 149.Er ESRCH 150if 151requested to delete a non-existent entry, 152or 153.Er ENOBUFS 154if insufficient resources were available 155to install a new route. 156In the current implementation, all routing processes run locally, 157and the values for 158.Fa rtm_errno 159are available through the normal 160.Em errno 161mechanism, even if the routing reply message is lost. 162.Pp 163A process may avoid the expense of reading replies to 164its own messages by issuing a 165.Xr setsockopt 2 166call indicating that the 167.Dv SO_USELOOPBACK 168option 169at the 170.Dv SOL_SOCKET 171level is to be turned off. 172A process may ignore all messages from the routing socket 173by doing a 174.Xr shutdown 2 175system call for further input. 176.Pp 177A process can specify which route message types it's interested in by passing 178an array of route messsage types to the 179.Xr setsockopt 2 180call with the 181.Dv RO_MSGFILTER 182option at the 183.Dv PF_ROUTE 184level. 185For example, to only get specific messages: 186.Bd -literal -offset indent 187unsigned char rtfilter[] = { RTM_IFINFO, RTM_IFANNOUNCE }; 188 189if (setsockopt(routefd, PF_ROUTE, RO_MSGFILTER, 190 &rtfilter, (socklen_t)sizeof(rtfilter)) == -1) 191 err(1, "setsockopt(RO_MSGFILTER)"); 192.Ed 193.Pp 194If a route is in use when it is deleted, 195the routing entry will be marked down and removed from the routing table, 196but the resources associated with it will not 197be reclaimed until all references to it are released. 198User processes can obtain information about the routing 199entry to a specific destination by using a 200.Dv RTM_GET 201message, 202or by reading the 203.Pa /dev/kmem 204device, or by calling 205.Xr sysctl 3 . 206.Pp 207The messages are: 208.Bd -literal 209#define RTM_ADD 0x1 /* Add Route */ 210#define RTM_DELETE 0x2 /* Delete Route */ 211#define RTM_CHANGE 0x3 /* Change Metrics, Flags, or Gateway */ 212#define RTM_GET 0x4 /* Report Information */ 213#define RTM_LOSING 0x5 /* Kernel Suspects Partitioning */ 214#define RTM_REDIRECT 0x6 /* Told to use different route */ 215#define RTM_MISS 0x7 /* Lookup failed on this address */ 216#define RTM_ONEWADDR 0xc /* Old (pre-8.0) RTM_NEWADDR message */ 217#define RTM_ODELADDR 0xd /* Old (pre-8.0) RTM_DELADDR message */ 218#define RTM_OOIFINFO 0xe /* Old (pre-1.5) RTM_IFINFO message */ 219#define RTM_OIFINFO 0xf /* Old (pre-6.0) RTM_IFINFO message */ 220#define RTM_IFANNOUNCE 0x10 /* iface arrival/departure */ 221#define RTM_IFINFO 0x14 /* iface/link going up/down etc. */ 222#define RTM_OCHGADDR 0x15 /* Old (pre-8.0) RTM_CHGADDR message */ 223#define RTM_NEWADDR 0x16 /* address being added to iface */ 224#define RTM_DELADDR 0x17 /* address being removed from iface */ 225#define RTM_CHGADDR 0x18 /* address properties changed */ 226.Ed 227.Pp 228A message header consists of one of the following: 229.Bd -literal 230struct rt_msghdr { 231 u_short rtm_msglen; /* to skip over non-understood messages */ 232 u_char rtm_version; /* future binary compatibility */ 233 u_char rtm_type; /* message type */ 234 u_short rtm_index; /* index for associated ifp */ 235 int rtm_flags; /* flags, incl kern & message, e.g. DONE */ 236 int rtm_addrs; /* bitmask identifying sockaddrs in msg */ 237 pid_t rtm_pid; /* identify sender */ 238 int rtm_seq; /* for sender to identify action */ 239 int rtm_errno; /* why failed */ 240 int rtm_use; /* from rtentry */ 241 u_long rtm_inits; /* which metrics we are initializing */ 242 struct rt_metrics rtm_rmx; /* metrics themselves */ 243}; 244 245struct if_msghdr { 246 u_short ifm_msglen; /* to skip over non-understood messages */ 247 u_char ifm_version; /* future binary compatibility */ 248 u_char ifm_type; /* message type */ 249 int ifm_addrs; /* like rtm_addrs */ 250 int ifm_flags; /* value of if_flags */ 251 u_short ifm_index; /* index for associated ifp */ 252 struct if_data ifm_data; /* statistics and other data about if */ 253}; 254 255struct ifa_msghdr { 256 u_short ifam_msglen; /* to skip over non-understood messages */ 257 u_char ifam_version; /* future binary compatibility */ 258 u_char ifam_type; /* message type */ 259 u_short ifam_index; /* index for associated ifp */ 260 int ifam_flags; /* value of ifa_flags */ 261 int ifam_addrs; /* like rtm_addrs */ 262 pid_t ifam_pid; /* identify sender */ 263 int ifam_addrflags; /* family specific address flags */ 264 int ifam_metric; /* value of ifa_metric */ 265}; 266 267struct if_announcemsghdr { 268 u_short ifan_msglen; /* to skip over non-understood messages */ 269 u_char ifan_version; /* future binary compatibility */ 270 u_char ifan_type; /* message type */ 271 u_short ifan_index; /* index for associated ifp */ 272 char ifan_name[IFNAMSIZ]; /* if name, e.g. "en0" */ 273 u_short ifan_what; /* what type of announcement */ 274}; 275.Ed 276.Pp 277The 278.Dv RTM_IFINFO 279message uses a 280.Vt if_msghdr 281header, the 282.Dv RTM_NEWADDR , 283.Dv RTM_CHGADDR , 284and 285.Dv RTM_DELADDR 286messages use a 287.Vt ifa_msghdr 288header, 289the 290.Dv RTM_IFANNOUNCE 291message uses a 292.Vt if_announcemsghdr 293header, 294and all other messages use the 295.Vt rt_msghdr 296header. 297.Pp 298The metrics structure is: 299.Bd -literal 300struct rt_metrics { 301 u_long rmx_locks; /* Kernel must leave these values alone */ 302 u_long rmx_mtu; /* MTU for this path */ 303 u_long rmx_hopcount; /* max hops expected */ 304 u_long rmx_expire; /* lifetime for route, e.g. redirect */ 305 u_long rmx_recvpipe; /* inbound delay-bandwidth product */ 306 u_long rmx_sendpipe; /* outbound delay-bandwidth product */ 307 u_long rmx_ssthresh; /* outbound gateway buffer limit */ 308 u_long rmx_rtt; /* estimated round trip time */ 309 u_long rmx_rttvar; /* estimated rtt variance */ 310 u_long rmx_pksent; /* packets sent using this route */ 311}; 312.Ed 313.Pp 314Flags include the values: 315.Bd -literal 316#define RTF_UP 0x1 /* route usable */ 317#define RTF_GATEWAY 0x2 /* destination is a gateway */ 318#define RTF_HOST 0x4 /* host entry (net otherwise) */ 319#define RTF_REJECT 0x8 /* host or net unreachable */ 320#define RTF_DYNAMIC 0x10 /* created dynamically (by redirect) */ 321#define RTF_MODIFIED 0x20 /* modified dynamically (by redirect) */ 322#define RTF_DONE 0x40 /* message confirmed */ 323#define RTF_MASK 0x80 /* subnet mask present */ 324#define RTF_CONNECTED 0x100 /* hosts on this route are neighbours */ 325#define RTF_LLDATA 0x400 /* used by apps to add/del L2 entries */ 326#define RTF_STATIC 0x800 /* manually added */ 327#define RTF_BLACKHOLE 0x1000 /* just discard pkts (during updates) */ 328#define RTF_PROTO2 0x4000 /* protocol specific routing flag */ 329#define RTF_PROTO1 0x8000 /* protocol specific routing flag */ 330#define RTF_SRC 0x10000 /* route has fixed source address */ 331#define RTF_ANNOUNCE 0x20000 /* announce new ARP or NDP entry */ 332#define RTF_LOCAL 0x40000 /* route represents a local address */ 333#define RTF_BROADCAST 0x80000 /* route represents a bcast address */ 334.Ed 335.Pp 336Specifiers for metric values in rmx_locks and rtm_inits are: 337.Bd -literal 338#define RTV_MTU 0x1 /* init or lock _mtu */ 339#define RTV_HOPCOUNT 0x2 /* init or lock _hopcount */ 340#define RTV_EXPIRE 0x4 /* init or lock _expire */ 341#define RTV_RPIPE 0x8 /* init or lock _recvpipe */ 342#define RTV_SPIPE 0x10 /* init or lock _sendpipe */ 343#define RTV_SSTHRESH 0x20 /* init or lock _ssthresh */ 344#define RTV_RTT 0x40 /* init or lock _rtt */ 345#define RTV_RTTVAR 0x80 /* init or lock _rttvar */ 346.Ed 347.Pp 348Specifiers for which addresses are present in the messages are: 349.Bd -literal 350#define RTA_DST 0x1 /* destination sockaddr present */ 351#define RTA_GATEWAY 0x2 /* gateway sockaddr present */ 352#define RTA_NETMASK 0x4 /* netmask sockaddr present */ 353#define RTA_GENMASK 0x8 /* cloning mask sockaddr present */ 354#define RTA_IFP 0x10 /* interface name sockaddr present */ 355#define RTA_IFA 0x20 /* interface addr sockaddr present */ 356#define RTA_AUTHOR 0x40 /* sockaddr for author of redirect */ 357#define RTA_BRD 0x80 /* for NEWADDR, broadcast or p-p dest addr */ 358#define RTA_TAG 0x100 /* route tag */ 359.Ed 360.Pp 361Flags for IPv6 addresses: 362.Bd -literal 363#define IN6_IFF_ANYCAST 0x01 /* anycast address */ 364#define IN6_IFF_TENTATIVE 0x02 /* tentative address */ 365#define IN6_IFF_DUPLICATED 0x04 /* DAD detected duplicate */ 366#define IN6_IFF_DETACHED 0x08 /* may be detached from the link */ 367#define IN6_IFF_DEPRECATED 0x10 /* deprecated address */ 368#define IN6_IFF_NODAD 0x20 /* don't perform DAD on this address 369 * (used only at first SIOC* call) 370 */ 371#define IN6_IFF_AUTOCONF 0x40 /* autoconfigurable address. */ 372#define IN6_IFF_TEMPORARY 0x80 /* temporary (anonymous) address. */ 373.Ed 374.Sh SEE ALSO 375.Xr socket 2 , 376.Xr sysctl 3 377.Sh HISTORY 378Since 379.Nx 8.0 , 380.Cm RTF_CLONED , 381.Cm RTF_CLONING , 382.Cm RTF_LLINFO , 383.Cm RTF_XRESOLVE 384and 385.Cm RTM_RESOLVE 386were obsolete. 387.Cm RTF_CONNECTED 388and 389.Cm RTF_LLDATA 390appeared in 391.Nx 8.0 . 392.Pp 393.Vt ifa_msghdr 394gained the fields ifam_pid and ifam_addrflags in 395.Nx 8.0 . 396