|
|
1.1 root 1: .TH IMP 4 "27 July 1983"
2: .UC 4
3: .SH NAME
4: imp \- 1822 network interface
5: .SH SYNOPSIS
6: .B pseudo-device imp
7: .SH DESCRIPTION
8: The
9: .I imp
10: interface, as described in BBN Report 1822, provides access to an
11: intelligent message processor normally used when participating
12: in the Department of Defense ARPA network. The network interface
13: communicates through a device controller, usually an ACC LH/DH or
14: DEC IMP-11A, with the IMP. The interface is \*(lqreliable\*(rq and
15: \*(lqflow-controlled\*(rq by the host-IMP protocol.
16: .PP
17: To configure IMP support, one of
18: .IR acc (4)
19: and
20: .IR css(4)
21: must be included.
22: The network number on which the interface resides is specified
23: at boot time using the SIOCSIFADDR ioctl. The host number is
24: discovered through receipt of NOOP messages from the IMP.
25: .PP
26: The network interface is always in one of four states: up, down,
27: initializing, or going down. When the system is
28: booted, the interface is marked down. If the hardware controller
29: is successfully probed, the interface enters the initializing
30: state and transmits three NOOP messages to the IMP. It then waits
31: for the IMP to respond with two or more NOOP messages in reply.
32: When it receives these messages it enters the up state. The going
33: down state is entered only when notified by the IMP of an impending
34: shutdown. Packets may be sent through the interface only while it
35: is in the up state. Packets received in any other state are dropped
36: with the error ENETDOWN returned to the caller.
37: .SH DIAGNOSTICS
38: \fBimp%d: leader error\fP. The IMP reported an error in a leader
39: (1822 message header). This causes the interface to be reset and
40: any packets queued up for transmission to be purged.
41: .PP
42: \fBimp%d: going down in 30 seconds\fP.
43: .br
44: \fBimp%d: going down for hardware PM\fP.
45: .br
46: \fBimp%d: going down for reload software\fP.
47: .br
48: \fBimp%d: going down for emergency reset\fP.
49: The Network Control Center (NCC) is manipulating the IMP. By convention
50: these messages are reported to all hosts on an IMP.
51: .PP
52: \fBimp%d: reset (host %d/imp %d)\fP. The host has received a
53: NOOP message which caused it to reset its notion of its
54: current address. This normally occurs at boot time, though
55: it may also occur while the system is running (for example, if
56: the IMP-controller cable is disconnected, then reconnected).
57: .PP
58: \fBimp%d: host dead\fP. The IMP has noted a host, to which a
59: prior packet was sent, is not up.
60: .PP
61: \fBimp%d: host unreachable\fP. The IMP has discovered a host,
62: to which a prior packet was sent, is not accessible.
63: .PP
64: \fBimp%d: data error\fP. The IMP noted an error in data
65: transmitted. The host-IMP interface is reset and the host
66: enters the init state (awaiting NOOP messages).
67: .PP
68: \fBimp%d: interface reset\fP. The reset process has been
69: completed.
70: .PP
71: \fBimp%d: marked down\fP. After receiving a \*(lqgoing down in
72: 30 seconds\*(rq message, and waiting 30 seconds, the host has
73: marked the IMP unavailable. Before packets may be sent to the
74: IMP again, the IMP must notify the host, through a series of
75: NOOP messages, that it is back up.
76: .PP
77: \fBimp%d: can't handle af%d\fP. The interface was handed
78: a message with addresses formatting in an unsuitable address
79: family; the packet was dropped.
80: .SH SEE ALSO
81: intro(4N),
82: inet(4F),
83: acc(4),
84: css(4)
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.