intro.4n

<B>Digital的Unix操作系统VAX 4.2源码</B>

.\" SCCSID: @(#)intro.4n	8.1	9/11/90
.TH intro 4n
.SH Name
networking \- introduction to networking facilities
.SH Syntax
.nf
.ft B
#include <sys/socket.h>
#include <net/route.h>
#include <net/if.h>
.fi R
.fi
.SH Description
.NXR "intro(4) keyword"
.NXB "network facilities" "introduction"
This section briefly describes the networking facilities
available in the system.
Documentation in this part of Section
4 is broken up into three areas: protocol families, protocols,
and ``network interfaces''.
Entries describing a protocol family are marked ``4f'',
while entries describing protocol use are marked ``4p''.
Hardware support for network interfaces is found
among the standard ``4'' entries.
.PP
All network protocols are associated with a specific protocol family.
.NXR "protocol family" "defined"
A protocol family provides basic services to the protocol
implementation to allow it to function within a specific
network environment.  These services can include 
packet fragmentation and reassembly, routing, addressing, and 
basic transport.  A protocol family can support multiple
methods of addressing, though the current protocol implementations
do not.  A protocol family normally comprises a number
of protocols, one per socket
type.  It is not required that a protocol family support
all socket types.  A protocol family can contain multiple
protocols supporting the same socket abstraction. 
.PP
A protocol supports one of the socket abstractions detailed
in
.MS socket 2 .
.NXR "protocol" "defined"
A specific protocol can be accessed either by creating a
socket of the appropriate type and protocol family or
by requesting the protocol explicitly when creating a socket.
Protocols normally accept only one type of address format,
usually determined by the addressing structure inherent in
the design of the protocol family/network architecture.
Certain semantics of the basic socket abstractions are
protocol-specific.  All protocols are expected to support
the basic model for their particular socket type, but may,
in addition, provide nonstandard facilities or extensions
to a mechanism.  For example, a protocol supporting the
SOCK_STREAM
abstraction may allow more than one byte of out-of-band
data to be transmitted per out-of-band message.
.PP
A network interface is similar to a device interface.
Network interfaces make up the lowest layer of the
networking subsystem, interacting with the actual transport
hardware.  
.NXR "network interface" "defined"
An interface may support one or more protocol
families or address formats.
The SYNTAX section of each network interface
entry gives a sample specification
of the related drivers for use in providing
a system description to
.MS config 8.
The DIAGNOSTICS section lists messages that may appear on the console
and in the system error log file
.PN /usr/adm/syserr/syserr.<hostname> 
due to errors in device operation.
.SH Addressing
.NXR "protocol family" "address format list"
Associated with each protocol family is an address
format.  The following address formats are used by the system:
.EX 0
#define AF_UNIX    1  /* local to host (pipes, portals) */
#define AF_INET    2  /* internetwork: UDP, TCP, etc. */
#define AF_IMPLINK 3  /* arpanet imp addresses */
.EE
.SH Routing
.NXR "routing table" "defined"
The network facilities provide limited packet routing.
A simple set of data structures make up a ``routing table''
used in selecting the appropriate network interface when
transmitting packets.  This table contains a single entry for
each route to a specific network or host.  A user process,
the routing daemon, maintains this data base with the aid
of two socket-specific 
.MS ioctl 2
commands, SIOCADDRT and SIOCDELRT.  The commands allow
the addition and deletion of a single routing
table entry.  Routing table manipulations can
only be carried out by superuser.
.PP
A routing table entry has the following form, as defined
in <net/route.h>:
.NXR(e) "routing table" "entry form"
.EX 0
struct rtentry {
        u_long  rt_hash;
        struct  sockaddr rt_dst;
        struct  sockaddr rt_gateway;
        short   rt_flags;
        short   rt_refcnt;
        u_long  rt_use;
        struct  rtentry *rt_next;
        struct  ifnet *rt_ifp;
};
.EE
with
.I rt_flags
defined from,
.EX 0
#define  RTF_UP      0x1   /* route usable */
#define  RTF_GATEWAY 0x2   /* destination is a gateway */
#define  RTF_HOST    0x4   /* host entry (net otherwise) */
.EE
Routing table entries come in three types: for a specific
host, for all hosts on a specific network, and for any destination
not matched by entries of the first two types (a wildcard route). 
.NXR "routing table" "entry types"
When the system
is booted, each network interface autoconfigured 
installs a routing table entry when it wishes to have packets
sent through it.  Normally, the interface specifies the route
through it is a ``direct'' connection to the destination host
or network.  If the route is direct, the transport layer of
a protocol family usually requests the packet be sent to the
same host specified in the packet.  Otherwise, the interface
may be requested to address the packet to an entity different
from the eventual recipient (that is, the packet is forwarded).
.PP
Routing table entries installed by a user process cannot specify
the hash, reference count, use, or interface fields; these are filled
in by the routing routines.  If
a route is in use when it is deleted (rt_refcnt is nonzero),
the resources associated with it are not
reclaimed until further references to it are released. 
.PP
The routing code returns EEXIST if
requested to duplicate an existing entry, ESRCH if
requested to delete a nonexistent entry,
or ENOBUFS if insufficient resources were available
to install a new route.
.PP
User processes read the routing tables through the 
.PN /dev/kmem 
device.
.PP
The
.I rt_use
field contains the number of packets sent along the route.
This value is used to select among multiple
routes to the same destination.  When multiple routes to
the same destination exist, the least used route is selected.
.PP
A wildcard routing entry is specified with a zero
destination address value.  Wildcard routes are used
only when the system fails to find a route to the
destination host and network.  The combination of wildcard
routes and routing redirects can provide an economical
mechanism for routing traffic.
.SH Interfaces
.NXB "network interface"
Each network interface in a system corresponds to a
path through which messages can be sent and received.  A network
interface usually has a hardware device associated with it, though
certain interfaces such as the loopback interface, 
.PN lo ,
do not.
.PP
At boot time, each interface that has underlying hardware support
makes itself known to the system during the autoconfiguration
process.  Once the interface has acquired its address, it is
expected to install a routing table entry so that messages can
be routed through it.  Most interfaces require some part of
their address specified with an SIOCSIFADDR ioctl before they
allow traffic to flow through them.  On interfaces where
the network-link layer address mapping is static, only the
network number is taken from the ioctl; the remainder is found
in a hardware-specific manner.  On interfaces that provide
dynamic network-link layer address mapping facilities (for example,
10Mb/s Ethernets), the entire address specified in the ioctl
is used.
.PP
The following 
.PN ioctl
calls may be used to manipulate network interfaces.  Unless
specified otherwise, the request takes an
.I ifrequest
structure as its parameter.  This structure has the form:
.NXR(e) "ifrequest structure" "form"
.EX 0
struct  ifreq {
  char    ifr_name[16];   /* name of interface (e.g. "ec0") */
  union {
         struct    sockaddr ifru_addr;
         struct    sockaddr ifru_dstaddr;
         short     ifru_flags;
     } ifr_ifru;
#define ifr_addr    ifr_ifru.ifru_addr    /* address */
#define ifr_dstaddr ifr_ifru.ifru_dstaddr /* end of p-to-p link */
#define ifr_flags   ifr_ifru.ifru_flags   /* flags */
};
.EE
.NXB "network interface" "ioctl calls for manipulating"
.TP
SIOCSIFADDR
Set interface address.  Following the address
assignment, the ``initialization'' routine for
the interface is called.
.TP
SIOCGIFADDR
Get interface address.
.TP
SIOCSIFDSTADDR
Set point-to-point address for interface.
.TP
SIOCGIFDSTADDR
Get point-to-point address for interface.
.TP
SIOCSTATE
Read or set ownership and state of a device.
.TP
SIOCSIFFLAGS
Set interface flags field.  If the interface is marked down,
any processes currently routing packets through the interface
are notified.
.TP
SIOCGIFFLAGS
Get interface flags.
.TP
SIOCGIFCONF
Get interface configuration list.  This request takes an
.I ifconf
structure (see SIOCSIFBRDADDR) as a value-result parameter.  The 
.I ifc_len
field should be initially set to the size of the buffer
pointed to by 
.IR ifc_buf .
On return it will contain the length, in bytes, of the
configuration list.
.TP
SIOCGIFNETMASK
Get network address mask.
.TP
SIOCSIFNETMASK
Set network address mask.
.TP
SIOCGIFBRDADDR
Get broadcast address associated with network interface.
.TP
SIOCSIFBRDADDR
Set broadcast address associated with network interface.
.NXE "network interface" "ioctl calls for manipulating"
.NXR(e) "network interface" "associated broadcast address"
.EX 0
/*
 * Structure used in SIOCGIFCONF request.
 * Used to retrieve interface configuration
 * for machine (useful for programs which
 * must know all networks accessible).
 */
struct  ifconf {
        int     ifc_len;    /* size of associated buffer */
        union {
               caddr_t  ifcu_buf;
               struct   ifreq *ifcu_req;
        } ifc_ifcu;
#define ifc_buf ifc_ifcu.ifcu_buf /* buffer address */
#define ifc_req ifc_ifcu.ifcu_req /* array of structures */
};
.EE
The following is the structure used in an SIOCSTATE request
to set device state and ownership.
.EX 0
struct ifstate {
 char    ifr_name[IFNAMSIZ]; /* if name, e.g. "dmv0" */
 u_short if_family;          /* current family ownership */
 u_short if_next_family;     /* next family ownership */
 u_short if_mode:3,          /* mode of device */
         if_ustate:1,        /* user requested state */
         if_nomuxhdr:1,      /* if set, omit mux header */
         if_dstate:4,        /* current state of device */
         if_xferctl:1,       /* xfer control to nxt family */
         if_rdstate:1,       /* read current state */
         if_wrstate:1        /* set current state */
         if_reserved:4;
};
.EE
.SH See Also
socket(2), ioctl(2), intro(4), config(8), routed(8c)
.NXE "network interface"
.NXE "network facilities" "introduction"