date and time created 86/02/17 17:59:27 by sklower
[unix-history] / usr / src / share / man / man4 / man4.vax / dh.4
CommitLineData
90a856fd
KM
1.\" Copyright (c) 1980 Regents of the University of California.
2.\" All rights reserved. The Berkeley software License Agreement
3.\" specifies the terms and conditions for redistribution.
4.\"
0ac17531 5.\" @(#)dh.4 6.1 (Berkeley) %G%
90a856fd 6.\"
0ac17531 7.TH DH 4 ""
90a856fd
KM
8.UC 4
9.SH NAME
a72e3b93 10dh \- DH-11/DM-11 communications multiplexer
90a856fd 11.SH SYNOPSIS
a72e3b93 12.B "device dh0 at uba0 csr 0160020 vector dhrint dhxint"
90a856fd 13.br
a72e3b93 14.B "device dm0 at uba0 csr 0170500 vector dmintr"
90a856fd
KM
15.SH DESCRIPTION
16A dh-11 provides 16 communication lines; dm-11's may be optionally
17paired with dh-11's to provide modem control for the lines.
18.PP
19Each line attached to the DH-11 communications multiplexer
20behaves as described in
21.IR tty (4).
22Input and output for each line may independently
23be set to run at any of 16 speeds;
24see
25.IR tty (4)
26for the encoding.
27.PP
28Bit
29.I i
30of flags may be specified for a dh to say that a line is not properly
31connected, and that the line should be treated as hard-wired with carrier
32always present. Thus specifying ``flags 0x0004'' in the specification of dh0
33would cause line ttyh2 to be treated in this way.
34.PP
a72e3b93
KM
35The dh driver normally uses input silos
36and polls for input at each clock tick (10 milliseconds)
37rather than taking an interrupt on each input character.
90a856fd
KM
38.SH FILES
39/dev/tty[hi][0-9a-f]
40.br
41/dev/ttyd[0-9a-f]
42.SH "SEE ALSO"
43tty(4)
44.SH DIAGNOSTICS
45\fBdh%d: NXM\fR. No response from UNIBUS on a dma transfer
46within a timeout period. This is often followed by a UNIBUS adapter
47error. This occurs most frequently when the UNIBUS is heavily loaded
48and when devices which hog the bus (such as rk07's) are present.
49It is not serious.
50.PP
a72e3b93 51\fBdh%d: silo overflow\fR. The character input silo overflowed
90a856fd
KM
52before it could be serviced. This can happen if a hard error occurs
53when the CPU is running with elevated priority, as the system will
54then print a message on the console with interrupts disabled. If the
55Berknet
90a856fd
KM
56is running on a
57.I dh
58line at high speed (e.g. 9600 baud), there is only 1/15th of a second of
59buffering capacity in the silo, and overrun is possible. This may
60cause a few input characters to be lost to users and a network
61packet is likely to be corrupted, but the network will recover.
62It is not serious.