Trailing-Edge
-
PDP-10 Archives
-
AP-D543V_SB
-
mig603.mem
There is 1 other file named mig603.mem in the archive. Click here to see a list.
DECsystem-10
Monitor Installation Guide
6.03 Monitor
June 1977
This manual describes the steps you must
take to install the 6.03 monitor.
Digital Equipment Corporation. Maynard, Massachusetts
Page 2
Revision: June 1977
The information in this document is subject to change without notice
and should not be construed as a commitment by Digital Equipment
Corporation. Digital Equipment Corporation assumes no responsibility
for any errors that may appear in this manual.
The software described in this document is furnished under a license
and may be used or copied only in accordance with the terms of such
license.
Digital Equipment Corporation assumes no responsibility for the use or
reliability of its software on equipment that is not supplied by
DIGITAL.
Copyright (C) 1973, 1977 by Digital Equipment Corporation
The postage prepaid READER'S COMMENTS form on the last page of this
document requests the user's critical evaluation to assist us in
preparing future documentation.
The following are trademarks of Digital Equipment Corporation:
DIGITAL UNIBUS
DEC DECsystem-10
PDP DECtape
DECUS DDT
KA10 KL10
KI10
Page 3
CONTENTS
Page
PREFACE
CHAPTER 1 BUILDING THE FRONT-END FILE SYSTEM
(KL ONLY) 1-1
CHAPTER 2 READIN BOOTM 2-1
CHAPTER 3 START THE MONITOR 3-1
CHAPTER 4 RESTORE SELECTED FILES 4-1
CHAPTER 5 COPY MONITOR SUPPORT PROGRAMS 5-1
CHAPTER 6 WRITE BOOTS 6-1
CHAPTER 7 MERGE MODIFICATIONS 7-1
CHAPTER 8 CREATE A MONITOR WITH MONGEN 8-1
8.1 INTRODUCTION 8-1
8.2 MONGEN ORGANIZATION 8-1
8.3 EXECUTING MONGEN 8-2
8.3.1 Mongen Dialogue Format 8-3
8.3.2 Generating a New Monitor 8-4
8.4 HDWGEN 8-4
8.5 TTYGEN
8.6 NETGEN
8.7 FGEN
8.8 PRODUCING RELOCATABLE BINARY FILES
COMMON.REL, COMMOD.REL, AND COMDEV.REL
8.8.1 Assembling the Configuration Files
8.8.2 Assembly Error Messages from
COMMON
8.9 DESCRIPTION OF CONFIGURATION FILES
8.9.1 Description of COMMON.MAC
8.9.2 Description of COMMOD.MAC
8.9.3 Description of COMDEV.MAC
8.10 HDWGEN EXAMPLE
8.11 TTYGEN EXAMPLE
8.12 NETGEN EXAMPLE
8.13 FGEN EXAMPLE
8.14 DECIMAL DEFAULT VALUES
Page 4
8.14.1 Symbols Defined in COMMON
8.14.2 Symbols Defined in COMDEV
8.14.3 Symbols Defined in COMMOD
8.15 OCTAL DEFAULT VALUES
8.15.1 Symbols Defined in COMMON
8.15.2 Symbols Defined in COMMOD
8.15.3 Symbols Defined in COMDEV
8.16 SIXBIT DEFAULT VALUES
8.17 NON-STANDARD DEVICE PI ASSIGNMENT
8.18 TERMINAL INTERFACE LINE NUMBERS
8.19 MONGEN ERROR MESSAGES
CHAPTER 9 ASSEMBLE MACRO FILES
CHAPTER 10 LOAD AND SAVE THE MONITOR
CHAPTER 11 MAKE A COPY OF NEW MONITOR
CHAPTER 12 COMMUNICATION SYSTEMS
CHAPTER 13 UPDATE ACCOUNTING FILES
CHAPTER 14 GET NEW MONITOR
CHAPTER 15 COPY NEW MONITOR TO SYS
CHAPTER 16 ONCE DIALOGUE
16.1 INTRODUCTION
16.1.1 Summary of STARTUP Options
16.1.2 Special Considerations
16.1.3 Special Multiprocessing Considerations
16.1.4 Conventions Used in this Chapter
16.2 STARTUP OPTION: QUICK AND NOINITIA
16.3 STARTUP OPTION: GO
16.4 STARTUP OPTION: DESTROY
16.5 STARTUP OPTION: REFRESH
16.6 STARTUP OPTION: UNITID
16.7 STARTUP OPTION: CHANGE
16.8 STARTUP OPTION: LONG
16.9 WHY RELOAD QUESTION
16.10 ERROR MESSAGES
CHAPTER 17 BACKUP
17.1 INTRODUCTION
17.2 FEATURES
17.3 OVERVIEW OF COMMAND FUNCTIONS
17.4 EXAMPLES
17.5 OPERATOR USES
17.6 RESTORING FROM A DISTRIBUTION TAPE
17.7 OBTAINING DIRECTORIES OF BACKUP TAPES
17.8 COMPARING TAPE AND DISK FILES
17.9 CHECKPOINTING LARGE FILES
17.10 BACKUP MESSAGES
Page 5
17.11 CAPACITIES OF DISK MEDIA VERSUS
MAGNETIC TAPE
17.12 BACKUP TAPE FORMAT
CHAPTER 18 BOOTM
Page 6
PREFACE
This manual describes the steps you must take to install the 6.03
monitor. This manual should be used in conjunction with the
DECsystem-10 Operator's Guide. Formerly, some of the information in
this manual was found in various specifications within the
DECsystem-10 Software Notebooks. For example, MONGEN is now Chapter 8
of this manual instead of being a separate specification.
When you receive your DECsystem-10, you will receive a Manufacturing
tape labeled YURMON and the name of your installation. When you
receive your 6.03 monitor, you will receive one 6.03 monitor tape and
one CUSP tape. You also may receive tapes associated with unbundled
products.
The files that make up the 6.03 monitor are listed below.
*.MAC New monitor source files (complete except for unbundled
files).
603.MCO A description of the changes made to the monitor since
the 6.02 monitor release.
603.DDT A DDT patch file that can be applied to the monitor
using the $Y feature of FILDDT to fix problems
concerning 6.03. This file contains those patches
listed in BWR603.RNO.
TOPA10.REL The combined .REL files needed for loading a regular
KA10 monitor.
TOPG10.REL The combined .REL files needed for loading a KALUG
monitor.
TOPI10.REL The combined .REL files needed for loading a regular
KI10 monitor.
TOPH10.REL The combined .REL files needed for loading a KILUG
monitor.
MONGEN.EXE Version 50(135) of the Monitor Generation Program,
Page 7
which is described in Chapter 8.
FGEN.HLP Version 075 of FGEN.HLP, which lists and explains all
the feature test switches that can be set via MONGEN.
CON???.CMD The COMPIL command files that are needed for assembling
monitors.
CMB???.CCL The PIP indirect files needed for making TOP?10.REL.
BWR603.RNO The BEWARE file for the 6.03 monitor, which contains
information needed to load and use the 6.03 monitor.
Read this file before you install the monitor.
BWR603.603 The RUNOFF output from BWR603.RNO.
The above files constitute the standard 6.03 monitor release. In
addition to the above, an installation will also receive any files
associated with the unbundled portions of 6.03 that that installation
is licensed to use.
The following are the unbundled portions of the 6.03 monitor, which
are shipped only to properly licensed installations.
Virtual Memory
DAS78
DA28
DC44
Task to Task Communication
The installation procedures you use are basically the same regardless
of which type of processor your system has (e.g., KA, KI, KL). Any
differences in the installation procedure because of processor type
are indicated within the text of this manual. Read the list below to
determine which classification your installation is; then, start
reading at the chapter number indicated.
Chapter Classification
1 Generating a 6.03 monitor for a KL10 system at a
new installation.
2 Generating a 6.03 monitor for a KA10 or KI10
system at a new installation.
4 Generating a 6.03 monitor for a KA, KI, or KL
system with a currently running 6-series monitor.
CHAPTER 1
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY)
The steps described in this chapter copy the front-end software from
the RSX20-F DECtape to the front-end file space on the public disk.
In performing the steps, you will be using the following front-end
programs. These programs are on the RSX20-F DECtape that you will be
copying onto the disk.
KLI This program loads the KL10 microcode, initializes
the KL10 caches, configures the KL10 memories, and
loads a bootstrap program.
MOU (MOUNT) This program adds (MOUnts) a device to the list of
on-line front-end devices.
UFD This program creates a directory in the front-end
file system on the disk. This program does not
respond with a prompt after you issue a command to
it. Therefore, you must type a CTRL/backslash
after typing the carriage return. Note that the
CTRL/backslash does not echo.
INI This program initializes the front-end file
system. This program does not respond with a
prompt after you issue a command to it.
Therefore, you must type a CTRL/backslash after
typing the carriage return. Note that the
CTRL/backslash does not echo.
PIP This program is a PDP-11 program that copies files
from the DECtape to the disk. The PIP described
in other chapters of this manual is the
DECsystem-10 PIP.
RED (REDIRECT) This program redefines the front-end logical name
SY0: from the DECtape to the disk. This program
does not respond with a prompt after you issue a
command to it. Therefore, you must type a
CTRL/backslash after typing the carriage return.
Note that the CTRL/backslash does not echo.
SAV (SAVE) This program saves the front-end monitor.
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-2
Device names used in this section are PDP-11 formatted device names.
Table 1-1 lists the PDP-11 device names.
Table 1-1
PDP-11 Device Names
PDP-11 Name Device
TT Terminal
DT DECtape
MT Magtape
SY DB0:[5,5], which is the
PDP-11 system area that is
used in the same way
as DSKB:[1,4].
RF Fixed-Head Disk
RP Disk Pack
RK Disk Cartridge
PR Paper-tape Reader
PP Paper-tape Punch
CD Card Reader
LP Line Printer
XY Plotter
In addition to initially installing the front-end file system, you
must install the front-end file system using the steps outlined in
this Chapter if you ever do any of the following:
1. Reinitialize the front-end file system.
2. Change the front-end hardware.
3. Destroy the front-end file system.
The first step is to mount the DECtape labeled RSX20-F. Place the
RSX20-F tape (version 0005J) on one of the front-end (PDP-11) DECtape
drives. Make sure that the Unit Selector Switch is set to 0. This
DECtape drive is called DT0 when you specify it by name.
On the right side of the Unit Selector Switch, set the rightmost
button to REMOTE. Set the leftmost button (on the left side of the
Unit Selector Switch that is set to 0) to WRITE ENABLE.
On the other DECtape drive, set the Unit Selector Switch to 1; mount
the RSX20-F Auxiliary Files DECtape; set the switches for this drive
(DT1) the same as you did for DT0.
Mount the manufacturing tape (labeled YURMON and your installation's
name) on MTA0 (-10 device). If you are not a new installation and
would like to create your own Manufacturing tape, follow the procedure
listed below.
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-3
.GET SYS:BACKUP
.SAVE MTA0:BACKUP
.SAVE MTA0:BACKUP
.START
/TAPE MTA0:
/SAVE DSKB:[1,4]=DSKB:*.*[1,4]
The above procedure saves two copies of BACKUP and the [1,4] disk area
(SYS:).
You can mount this magnetic tape on another magtape drive (other than
MTA0); however, MTA0 is assumed in the following example. Mount a
recently formatted disk pack on RPA0 (or another drive, but RPA0 is
assumed). Whichever drive you choose, it must be dual-ported with the
RH11.
To refer to the "-11" disk pack, you specify a device name in the
following format:
DB unit-number:
where unit-number is 0-7. For example, RPA0 is the equivalent of DB0,
RPA1 is the equivalent of DB1, an RPC3 is the equivalent of DB3.
Set the data switches on the -11 front panel to 000001. Make sure the
ENABLE switch is on; then, push the SW/REG switch. The tape will
begin loading. Refer to Table 14-1 for a description of what occurs
when you press the SW/REG switch.
The following 'dialogue' takes place at the operator terminal between
you, the front-end installer, and the system. Your responses are
those underlined. Press the RETURN key after each response.
RSX-20F V005J 0:06 21-JUN-76
[SY0: REDIRECTED TO DT0:]
[DT0: MOUNTED]
KLI -- VERSION V002Q RUNNING
KLI -- MICROCODE VERSION 131 LOADED
KLI -- % NO FILE - ALL CACHE BEING CONFIGURED
KLI -- ALL CACHES ENABLED
KLI -- % NO FILE - ALL MEMORY BEING CONFIGURED
LOGICAL MEMORY CONFIGURATION:
CONTROLLER
ADDRESS SIZE RQ0 RQ1 RQ2 RQ3 CONTYPE INT
00000000 256K 04 FOR ALL DMA20 4
KLI -- ? FILE 'DT0:BOOT.EXB;0'NOT FOUND
KLI -- ? BOOTSTRAP LOAD FAILED
KLI -- ENTER DIALOGUE [NO,YES,EXIT,BOOT]?
KLI> YES
KLI -- RELOAD MICROCODE [YES,VERIFY,NO]?
KLI> NO
KLI -- RECONFIGURE CACHE [FILE,ALL,YES,NO]?
KLI> NO
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-4
KLI -- CONFIGURE KL MEMORY [FILE,ALL,YES,NO]?
KLI> NO
KLI -- LOAD KL BOOTSTRAP [YES,NO,FILENAME]?
KLI> BOOTM
This example assumes you are booting from magtape.
The following is then printed at the operator terminal.
KLI -- BOOTSTRAP LOADED AND STARTED ;-10 BOOTSTRAP
BOOTM V4(16)
BTM>
After the BTM> prompt, type in a command string terminated by a
carriage return. After performing the command, the bootstrap loader
either restarts itself or transfers to a newly loaded program,
depending on the command you type. You can type either /TM02 or
/TX01, depending on the tape drive used (e.g., for TU70s, type /TX01).
Or, you could type a command in the following format:
structure:file.ext[project,prog] /TM02
/TX01
If you do not specify either /TM02 or /TX01, TM10 is assumed by
default.
In this example /TM02 was typed.
BTM>/TM02
This response begins the loading of the default monitor (i.e.,
DSKB:SYSTEM.EXE[1,4] from drive 0 (MTA0) on the /TM02 magtape
controller). (For more information on BOOTM refer to Chapter 18.) The
following dialogue is then printed on the operator's terminal. You
should type everything that is underlined. Refer to Chapter 3 for a
more detailed description of this dialogue.
RK275A KL10 SYS#1026 01-19-77
WHY RELOAD: SA ;System is stand-alone; refer to Section 16.9
DATE: 21-JAN-77 ;Enter the current date
TIME: 710 ;Type the current time of day, using
;a 24-hour clock.
STARTUP OPTION: REFRESH
;Specify the LONG startup
;option if your installation is a new
;installation or you are
;using blank disk packs.
TYPE STR NAME TO BE REFRESHED (CR IF NONE, ALL IF ALL)
DSKB ;The structure on which the front-end
;file system is to be created.
;Refer to Section 16.8 for
;a description of the LONG
;dialogue.
TO AUTOMATICALLY LOG-IN UNDER [1,2] TYPE "LOGIN"
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-5
.LOGIN
.RUN MTA0:BACKUP ;This executes the BACKUP
/TAPE MTA0: ;program found on the manufacturing
;tape. Refer to Chapter 17.
/RESTORE ;Move files to SYSB
! ;BACKUP's indication that it is busy.
"DONE
?BKPHSG CANNOT GET HIGH SEGMENT BACK ;Ignore this
;error message.
Executing the following procedure creates a file called FE.SYS on RPA0
(DSKB) that will be used by the front-end for its file system.
.R FEFILE
DISK UNIT NAME:RPA0 ;UNIT WHERE DSKB is mounted.
SIZE OF FILE IN BLOCKS (<CRLF> GIVES DEFAULT OF 2000):
[2008 DATA BLOCKS ALLOCATED TO FE.SYS]
[FE.SYS AREA STARTS AT LOGICAL BLOCK 33041.]
[FRONT END FILE CREATED, HOM BLOCKS WRITTEN]
.K/F ;To stop the -10 because the KL10
;must NOT be running
;while installing the
;front-end software.
NOTE
From now on, until you read otherwise,
you are 'talking' only to the -11
front-end, using the -11 command
language and -11 programs. The -10 is
not involved in this communication
process. The RESET command (in the
example below) ensures that the -10 is
not running.
To being communicating with the front-end command parser, type a
CTRL/backslash (which does not echo). If at any time you type an
incorrect response to the parser (i.e., after the prompt PAR%), type a
CTRL/Z followed by a CTRL/backslash. The system retypes the prompt,
after which you can retype your response. Note that some times a
CTRL/backslash takes a few moments to respond.
^\ ;Type a CTRL/backslash.
PAR>SHUT
DECSYSTEM-10 NOT RUNNING
;K/F above kills your job.
;The SHUT command stops
;the -10.
^\ ;Type a CTRL/backslash.
PAR>SET CONSOLE MAINTENANCE
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-6
CONSOLE MODE: MAINTENANCE
PAR%RESET ;KL10 must not be running while installing
;front-end software.
PAR#ST MICROCODE ;Start microcode.
PAR%MCR MOU ;Start the MOUNT program in the front-end.
;MCR is the front-end
;command decoder.
MOU>DT1: ;Mount the tape labeled RSX20-F
;Auxiliary files.
MOU -- MOUNT COMPLETE
MOU>^Z ;Type a CTRL/Z.
^\ ;Type a CTRL/backslash
PAR%MCR PIP ;START PIP.
PIP>TT:=DT0:/LI ;Directory of DT0
DIRECTORY DT0:[5,5]
20-MAR-77 14:28
F11ACP.TSK;41 77. C 21-JUN-76 00:50
PARSER.TSK;37 39. C 21-JUN-76 00:50
TKTN.TSK;45 6. C 21-JUN-76 00:51
MOU.TSK;37 5. C 21-JUN-76 00:51
SETSPD.TSK;41 4. C 21-JUN-76 00:51
KLR.TSK;37 5. C 21-JUN-76 00:52
KLE.TSK;37 23. C 21-JUN-76 00:52
KLX.TSK;43 5. C 21-JUN-76 00:53
KLI.TSK;46 33. C 21-JUN-76 00:53
UA.MCB;150 35. 21-JUN-76 00:53
UB.MCB;150 35. 21-JUN-76 00:54
BT128K.EXB;1 12. 16-MAR-77 10:09
BT256K.EXB;1 12. 16-MAR-77 10:10
BOOTM.EXB;2 34. 16-MAR-77 10:16
TOTAL OF 325. BLOCKS IN 14. FILES
PIP>TT:=DT1:/LI
DIRECTORY DT1:[5,5]
20-MAR-77 14:31
RED.TSK;41 6. C 21-JUN-76 00:07
SAV.TSK;40 12. C 21-JUN-76 00:07
DMO.TSK;37 5. C 21-JUN-76 00:08
T20ACP.TSK;37 8. C 21-JUN-76 00:08
UFD.TSK;37 9. C 21-JUN-76 00:08
INI.TSK;37 23. C 21-JUN-76 00:08
PIP.TSK;1321 56. C 21-JUN-76 00:09
COP.TSK;2 8. C 21-JUN-76 00:09
TOTAL OF 161. BLOCKS IN 9. FILES
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-7
PIP>^Z
^\ ;Type a CTRL/backslash.
PAR%MCR INI
INI>DB0: ;Initialize the front-end file system on DB0.
^\ ;Type a CTRL/backslash.
PAR%MCR MOU ;Start MOUNT program.
MOU>DB0: ;Mount the disk unit.
MOU -- MOUNT COMPLETE
MOU>^Z ;Type a CTRL/Z.
^\ ;Type a CTRL/backslash.
PAR%MCR UFD ;Start the UFD program.
UFD>DB0:[5,5] ;Creating the UFD called [5,5] on the
;-11 disk area (FE.SYS);
;see note below.
^\ ;Type a CTRL/backslash.
PAR%MCR PIP ;Start PIP.
PIP>DB0:=DT0:*.*,DT1:*.* ;To copy all files from the
;DECtape to the front-end file
;system [5,5] on the disk pack on RPA0.
;After approximately 35 minutes,
;the system prints the PIP prompt.
PIP>DB0:BOOT.EXB=DT0: BT128K.EXB
BT256K.EXB
;This copies the BOOTS that
;fits your
;configuration.
PIP>^Z ;Type a CTRL/Z.
^\ ;Type a CTRL/backslash.
PAR%
NOTE
FE.SYS is the logical -11 disk space
area residing on the -10 disk
(DSKB:[1,4],RPA0). FE.SYS contains -11
directory areas and files within these
areas. Note that [5,5] is a directory
area in FE.SYS.
You now reboot the front-end. Set the data switches on the -11 front
panel to 000003. (Refer to Table 14-1.) Make sure the ENABLE switch
is ON; then, push the SW/REG switch.
RSX-20F V00SJ 0:06 21-JUN-76
[SY0: REDIRECTED TO DT0:]
[DT0: MOUNTED]
^\ ;Type a CTRL/backslash to start the parser.
PAR%MCR MOU ;Start the MOUNT program.
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-8
MOU> DB0: ;Mount the disk unit that the -11
;front-end file system (that you just
;installed) resides on (DB0:).
MOU -- MOUNT COMPLETE
MOU>^Z ;Type a CTRL/Z.
^\ ;Type a CTRL/backslash
PAR%MCR RED ;Start the REDIRECT program.
RED>DB0:=SY: ;Make SY: be DB0: instead of DT0:.
;This causes the KL10 to be rebooted from disk
;instead of DECtape (-11 devices).
^\ ;Type a CTRL/backslash.
PAR%MCR SAV ;Start the SAVE program.
SAV>SY:/WB ;Save the front-end monitor on
;SY:(DB0:[5,5]).
;The front-end can now use the files from
;the TOPS-10 disk pack instead of the
;PDP-11 DECtapes.
[DB0: DISMOUNTED]
[DT0: DISMOUNTED]
RSX-20F V005J 2:00 21-JUN-76 ;Messages from the SAVE program
[SY0: REDIRECTED TO DB0:] ;to enable you to bootstrap
[DB0: MOUNTED] ;from DB0: [5,5].
^\ ;Type a CTRL/backslash.
PAR%MCR PIP ;Start PIP.
PIP>TT:=DB0:/LI
DIRECTORY DB0:[5,5]
21-JUN-76 01:05
F11ACP.TSK;41 77. C 21-JUN-76 00:30
PARSER.TSK;37 39. C 21-JUN-76 00:31
TKTN.TSK;45 6. C 21-JUN-76 00:32
MOU.TSK;37 5. C 21-JUN-76 00:33
SETSPD.TSK;41 4. C 21-JUN-76 00:35
KLR.TSK;37 5. C 21-JUN-76 00:36
KLE.RSK;37 23. C 21-JUN-76 00:37
KLX.TSK;43 5. C 21-JUN-76 00:38
KLI.TSK;46 33. C 21-JUN-76 00:39
UA.MCB;150 35. 21-JUN-76 00:41
UB.MCB;150 35. 21-JUN-76 00:43
BT128K.EXB;1 12. 21-JUN-76 00:44
BT256K.EXB;1 12. 21-JUN-76 00:46
BOOTM.EXB;2 34. 21-JUN-76 00:48
KL.CFG;1 1. 21-JUN-76 00:50
RED.TSK;41 6. C 21-JUN-76 00:54
SAV.TSK;40 12. C 21-JUN-76 00:55
DMO.TSK;37 5. C 21-JUN-76 00:56
T20ACP.TSK;37 8. C 21-JUN-76 00:57
UFD.TSK;37 9. C 21-JUN-76 00:58
INI.TSK;37 23. C 21-JUN-76 00:59
PIP.TSK;1321 56. C 21-JUN-76 01:00
BUILDING THE FRONT-END FILE SYSTEM (KL10 ONLY) Page 1-9
COP.TSK;2 8. C 21-JUN-76 01:01
TOTAL OF 453. BLOCKS IN 23. FILES
PIP>^Z ;Type a CTRL/Z.
At this point, RSX20-F can be bootstrapped from the disk by depressing
the switch labeled DISK or, if the front-end disk pack is on a drive
other than RPA0, depress the SW/REG switch. Refer to Chapter 2.
From now on devices and directories specified are DECsystem-10 devices
and directories; unless specified otherwise.
CHAPTER 2
READIN BOOTM
READIN the BOOTM bootstrap and your 6.03 monitor from the new
installation bootstrap tape.
If you have a KL10 system, read Steps 1 and 2. If you have a KA or KI
system, read Steps 3 through 7.
STEP 1 (KL10 only)
Set the ENABLE/DISABLE load switch to ENABLE. This will enable the
other three load switches.
STEP 2 (KL10 only)
Press the upper half of the switch labeled DISK on the KL10 front
panel. This will cause the front-end processor to access the disk on
drive 0 and load the RSX20-F monitor.
The KL10 initialization program (KLINIT) is automatically loaded, and
the default hardware configuration of cache and external memory are
also automatically set up. The bootstrap program (BOOTS) for the
TOPS-10 monitor is then automatically loaded into the central
processor and started. The following is an example of the output you
will receive on the console terminal. You type everything that is
underlined.
RSX-20F V005J 0:16 21-JUN-76
[SY0: REDIRECTED TO DB0:]
[DB0: MOUNTED]
KLI -- VERSION V002J RUNNING
KLI -- MICROCODE VERSION 131 LOADED
KLI -- ALL CACHES ENABLED
LOGICAL MEMORY CONFIGURATION
CONTROLLER
ADDRESS SIZE RQ0 RQ1 RQ2 RQ3 CONTYPE INT
00000000 256 04 FOR ALL DMA20 4
KLI -- BOOTSTRAP LOADED AND STARTED
BOOTS V22(103)
BTS>
^\ ;Type a CTRL/backslash.
READIN BOOTM Page 2-2
PAR>MCR KLI
KLI -- VERSION V0002J RUNNING
KLI -- ENTER DIALOGUE [NO,YES,EXIT,BOOT]?
KLI -- YES
KLI -- RELOAD MICROCODE [YES,VERIFY,NO]?
KLI> NO
KLI> RECONFIGURE CACHE [FILE,ALL,YES,NO]?
KLI> NO
KLI -- CONFIGURE KL MEMORY [FILE,ALL,YES,NO]?
KLI> NO
KLI -- BOOTSTRAP [YES,NO,FILENAME]?
KLI> BOOTM
KLI -- ALL CACHES ENABLED
KLI -- BOOTSTRAP LOADED AND STARTED
BOOTS V4(16)
BTS>
If an error occurs during the KL initialization program, you will
receive an error message preceded by KLI --?. At that point, you are
placed in KLINIT dialogue mode which is described in Appendix A of the
KL Series Operator's Guide. BOOTS outputs the prompt BTS>. After
BOOTS prints its prompt, you must type a file specification,
indicating the monitor you want to read from the BACKUP save set on
the Manufacturing Bootstrap Tape. Type the following command line:
DSKB: YURMON.EXE[1,4]
This response begins the loading of the monitor from the Manufacturing
BOOTSTRAP tape. Then, the ONCE-Only Dialogue starts. Turn to Chapter
4.
STEP 3 (KA/KI Only)
Mount a recently formatted disk pack on a disk drive.
For example, RPA0.
STEP 4 (KA/KI only)
Mount the 6.03 Manufacturing Bootstrap tape containing your monitor on
MTA0, which is write-locked. Drive 0 is required for READIN with TM10
controllers. On TU70's the tape must be mounted on the lowest
numbered ready drive.
If you are not a new installation and you would like to create your
own manufacturing tape, follow the procedure listed below.
.GET SYS:BACKUP
.SAVE MTA0:BACKUP
.SAVE MTA0:BACKUP
.START
/TAPE MTA0:
/SAVE DSKB:[1,4]=DSKB:*.*[1,4]
The above procedure saves two copies of BACKUP and the [1,4] disk area
(SYS:).
READIN BOOTM Page 2-3
STEP 5 (KA/KI only)
For a TM10 controlled magtape set the READIN switches to 340. For a
TU70, set the READIN switches to 220.
STEP 6 (KA/KI only)
In order, press the STOP and RESET switches and set the NXM switch
off; then, press READIN. The tape rewinds and reads in the BOOTM
bootstrap, which then allows you to specify and start the 6.03
monitor.
STEP 7 (KA/KI Only)
BOOTM outputs the prompt characters BTM>. After BOOTM prints its
prompt, you must type a file specification, indicating the monitor you
want to read from the BACKUP save set on the 6.03 Manufacturing
Bootstrap tape. Type the following command line:
DSKB: yurmon.EXE[1,4] /TM02
/TX01
where: yurmon is the name given to your monitor.
You must specify the tape control unit: either TM02 or TX01. The
default, if you do not specify a unit, is TM10.
The command line format to be typed to BOOTM is described in Chapter
19.
This response begins the loading of the default monitor on the 6.03
Manufacturing Bootstrap tape. The following dialogue is then printed
on the operator's terminal. You should type everything that is
underlined. Refer to Chapter 3 for a more detailed description of
this dialogue.
RK277 KI10 SYS#514 2-15-77
WHY RELOAD:SA ;System is stand-alone; refer to Section
;16.9.
DATE: 21-FEB-77 ;Enter the current date
TIME: 7:10 ;Type the current
;time of day, using a 24-hr. clock.
STARTUP OPTION:REFRESH ;Specify the LONG startup
;option if your
;installation is a
;new installation or
;you are using blank
;disk packs.
TYPE STR NAME TO BE REFRESHED (CR IF NONE, ALL IF ALL)
DSKB ;The structure on
;which the monitor
;is to reside.
;Refer to Section 16.8
;for a description of the
READIN BOOTM Page 2-4
;LONG Dialogue.
TO AUTOMATICALLY LOG-IN UNDER [1,2] TYPE "LOGIN"
.LOGIN
.RUN MTA0: BACKUP ;This executes the BACKUP
;program found on the
;manufacturing
;magtape. Refer to
;Chapter 17.
/TAPE MTA0:
/RESTORE ;Move files to DSKB[1,4]
! ;BACKUP's indication
;that it is busy.
"DONE
?BKPHSG CANNOT GET HIGH SEGMENT BACK
;Ignore this error
;message.
Now, turn to Chapter 4.
CHAPTER 3
START THE MONITOR
This chapter briefly describes the system start-up dialogue, which
appeared in the examples in Chapters 1 and 2.
You can start your monitor with the LONG startup option and refresh
the disks. If your disks have already been formatted and refreshed,
respond with the NOINITIA startup option. Note that disk packs must
be formatted by a Field Service formatting program, before you can
attempt to use them with a monitor.
The LONG START-UP Option is not described in this chapter. The
Start-Up Options are a portion of the ONCE-Only dialogue, which is
completely described in Chapter 16.
STEP 1
The ONCE-Only dialogue begins between you and the system. ONCE types:
SYSCHK (Y,N):
If you type Y in response, the system will run a 5-second diagnostic
program that ensures the accessibility of all configured memory and
system devices. A reply of N or a carriage return skips the execution
of the diagnostic program.
STEP 2
ONCE then types the system name (which was specified in MONGEN refer
to Section 8.4, question number 4) followed by the monitor creation
date. For example:
LOAD 3 SYS#10 KL 02-1-77
When ONCE prints the following:
WHY RELOAD:
reply with NEW and press the carriage return. Instead of responding
with NEW, you can respond with one of the below:
OPR NXM CM
START THE MONITOR Page 3-2
PARITY HALT SA
POWER LOOP SCHED
STATIC HUNG OTHER
HARDWARE PM
If you do not reply within 60 seconds, OTHER is assumed.
STEP 3
ONCE prompts with:
DATE:
after which you type the current date; for example: 1-MAR-77. ONCE
then prompts with:
TIME:
after which you type a 4-digit number indicating the current time,
using a 24-hour clock. For example: 2015, indicating 8:15 PM.
STEP 4
ONCE prompts with:
STARTUP OPTION:
after which you respond with LONG if your disks have not been
refreshed. Otherwise, respond with NOINITIA. The NOINITIA option
responds with a . and you must LOGIN as a [1,2] job.
The LONG startup option allows you to explicitly set all system
parameters and to refresh and restructure the file system in a
non-standard way.
The ONCE dialogue, including the LONG startup dialogue is described in
Chapter 16.
CHAPTER 4
RESTORE SELECTED FILES
Before reading this chapter, make sure you are using the latest field
image version of BACKUP. If you are a new installation and you have
just gotten BACKUP from the Manufacturing Bootstrap tape, you should
have the correct BACKUP version. If you have been running a TOPS-10
monitor, be sure you have the proper version of BACKUP on SYS:.
Verify that the version number is the latest by checking it against
the software version number in the most recent DECsystem-10
DISPATCH/BULLETIN.
To install a 6.03 monitor, certain monitor files must be restored at
800 bits/inch from the standard distribution tapes and from the
unbundled distribution tapes. The [10,7] area is used in these
instructions to designate the restore area. But, another
project-programmer number can be substituted for [10,7]. Normally,
these files are restored to the structure DSKB. However, if your
system does not have a DSKB, BACKUP can be restored to any designated
file structure. (Refer to Chapter 17.)
To simplify the restore process, you should restore files on the
monitor tape in the order they appear on the distribution tape. Files
on the monitor tape are normally in alphabetical order. To find out
the appropriate order, print a directory of the tape on the line
printer. To do so, type the following command lines.
.START
/REWIND
/DENSITY 800
/PRINT LPT
The number of files you should restore from the monitor distribution
tape to the disk is determined by the amount of disk space you have
available. The instructions for three different situations are given
below:
1. If your installation has slightly more than 20,000 free disk
blocks
2. If your installation has about 10,000 free disk blocks
3. If your installation has very few free disk blocks.
If you have enough disk space available, you should also restore all
the files from other distribution tapes using the appropriate
project-programmer number. Otherwise, you may need to selectively
restore certain files, because you may need some unbundled files for
your monitor. Also, some of the utility programs, assumed to be on
the monitor distribution tape in the instructions below, may be on the
language distribution tape because of changes in the packaging
procedure.
RESTORE SELECTED FILES Page 4-2
1. If your installation has slightly more than 20,000 free
blocks of disk space, you can restore all files on the 6.03
monitor distribution tape. In this case, the latest versions
of all files are available for use, and you need not worry
about selectively restoring individual files.
The procedure below should be followed, first for the CUSP
tape, then for the 6.03 monitor distribution tape, and last
for each of your unbundled tapes, if you have any.
To restore all files, ensure that the tape is mounted
properly and type to BACKUP:
.R BACKUP
/REWIND
/INTERCHANGE
/DENSITY 800
/RESTORE[10,7]*.*=PIP.EXE
"DONE
/REWIND
/SUPERSEDE ALWAYS
/INTERCHANGE
/RESTORE[10,7]*.*=*.*
"DONE
The purpose of restoring PIP.EXE separately is to protect
yourself. If the disk fills up, you will need PIP to delete
files and then continue restoring. Because all of the
installation batch control files assume that the field image
files are in [10,7], we recommend your using that
project-programmer number.
If you attempt to restore all files and receive the error
message
?LOOKUP/ENTER FAILURE (14)
indicating that the disk storage capacity has been exceeded,
you must restore only selected files as described in
situation 2, which follows.
After restoring all the files, you should print the .DOC
.MAN, .RND, .RNH and .RNO files. They contain useful program
documentation. The .RND, RNH, and .RNO files must first be
expanded with RUNOFF by typing:
.RUN DSK:RUNOFF[10,7]
*filename
*^C
All .RND files become .DOC files, all .RNH files become .HLP
files, and all .RNO files become .MEM files after they have
been through RUNOFF.
2. If your installation has about 10,000 free blocks of disk
space, you can restore all monitor-related files to the disk.
These files enable you to build and start the monitor, to
assemble all monitor source files, to make modifications with
SOUP, and to obtain monitor listings.
To build the monitor you must determine which type of monitor
is best for your configuration and then choose the
RESTORE SELECTED FILES Page 4-3
appropriate .CMD file for your system. (See Chapter
10.) After selecting the file, you should restore it and
PIP.EXE using BACKUP. In the example below, the .CMD file
for the standard KI configuration is used. Type to BACKUP:
/REWIND
/INTERCHANGE
/DENSITY 800
/SUPERSEDE ALWAYS
/RESTORE[10,7]*.*=CONKI.CMD,PIP.EXE
"DONE
/^C
Then, list the contents of CONKI on your terminal using PIP.
Type:
.RUN PIP[10,7]
*TTY:=CONKI.CMD[10,7]
*^C
The monitor related files are printed on your terminal.
Then restore all files printed on your terminal with the
extension .MAC, excluding F???.MAC, HDW???.MAC, TTY???.MAC,
and NET???.MAC, because you must generate these files with
MONGEN, which is described in Chapter 8. There is no need to
repeat files that are duplicated on the TTY listing. Also,
in lines containing an equal sign, only files to the right of
the equal sign need to be restored. For example, for the
line,
COMDEV.RLI=HDWKI.MAC+TTYKI.MAC+NETKI.MAC+COMDEV.MAC
restore only COMDEV.MAC. These files are for the standard KI
monitor, determined by the imbedded 'KI.'
The procedure below should be repeated for the CUSP tape,
then for the 6.03 monitor distribution tape, and last for
each of your unbundled tapes, if you have any.
To restore the necessary files, type:
.RUN DSK:BACKUP[10,7]
/REWIND
/INTERCHANGE
/DENSITY 800
/SUPERSEDE ALL
/RESTORE[10,7]*.*= filenames (separated by commas)
where filenames include BWR603.RNO, COMDEV.MAC, COMMOD.MAC,
COMMON.MAC, MONGEN.EXE, S.MAC, WBOOT.EXE, the appropriate
TOP?10.REL file, and necessary .MAC files from the .CMD file.
Note that these files are distributed on the 6.03 monitor
distribution tapes.
where filenames include COMPIL.EXE, FGEN.HLP, FILEX.EXE,
LINK.EXE, LNKSCN.EXE, LNKLOD.EXE, LNKMAP.EXE, LNKXIT.EXE,
LNKERR.EXE, LNK999.EXE, MACRO.EXE, and RUNOFF.EXE. Note that
these files are on the CUSP tape.
If you are building a monitor from scratch, also restore the
RESTORE SELECTED FILES Page 4-4
following files.
/RESTORE[10,7]*.*=LOGIN.EXE,LOGOUT.EXE,REACT.EXE
If you have any unbundled distribution tapes (e.g., virtual
memory), these tapes must be restored at this time. Refer to
Chapter 10 for a list of the file names contained on the
unbundled distribution tapes.
It is recommended that you restore and print copies of .DOC,
.MAN, .RND, .RNH, and .RNO files if there is sufficient
space. They contain useful program documentation. The .RN?
files must be expanded with RUNOFF by typing:
.R RUNOFF
*filename
*^C
All .RND files become .DOC files, all .RNH files become .HLP
files, and all .RNO files become .MEM files after they have
been run through RUNOFF.
3. If your installation has very few free disk blocks,
selectively restore only essential monitor building files.
With the minimum number of files, you can only build and
start the monitor. You cannot assemble all monitor source
files and you cannot merge your modifications with SOUP. You
may want to free additional disk space to perform these
functions.
To restore only the essential files, mount the 6.03 monitor
tape and follow the procedure below to restore the necessary
files:
.R BACKUP
/REWIND
/INTERCHANGE
/DENSITY 800
/RESTORE[10,7]*.*=BEWARE.603,COMMOD.MAC
/RESTORE[10,7]*.*=COMMON.MAC,COMDEV.MAC,COMPIL.EXE
/RESTORE[10,7]*.*=FGEN.HLP,BACKUP.EXE, MONGEN.EXE
/RESTORE[10,7]*.*=PIP.EXE,S.MAC,TOP?10.REL,W?BOOT.EXE
"DONE
/^C
Then mount the CUSP tape and follow the procedure below to
restore the necessary files:
.START
/REWIND
/INTERCHANGE
/DENSITY 800
/RESTORE[10,7]*.*=BACKUP.EXE,FILEX.EXE
/RESTORE[10,7]*.*=LINK.EXE,LNKSCN.EXE
/RESTORE[10,7]*.*=LNKLOD.EXE,LNKMAP.EXE
/RESTORE[10,7]*.*=LNKXIT.EXE,LNKERR.EXE
/RESTORE[10,7]*.*=LNK999.EXE,MACRO.EXE
/RESTORE[10,7]*.*=PIP.EXE
"DONE
/^C
where TOP?10.REL is the specific .REL file needed for your
RESTORE SELECTED FILES Page 4-5
configuration. (See Chapter 9.)
If you are building a 6.03 monitor from scratch, also restore
the following files:
/RESTORE[10,7]*.*=LOGIN.EXE,LOGOUT.EXE,REACT.EXE
If you have any unbundled distribution tapes (e.g., virtual memory),
these tapes must be restored at this time. Refer to Chapter 10 for a
list of the file names contained on the unbundled distribution tapes.
It is very important that you print and read the last minute
documentation file. You can simply list or print BEWARE.603 or,
assuming the files BWR603.RNO and RUNOFF.EXE have been restored from
the tape, expand the RUNOFF file and then print it.
Expand the file with RUNOFF by typing:
.RUN DSK:RUNOFF[10,7]
*filename
*^C
All .RNO files become .MEM files after they have been run through
RUNOFF.
CHAPTER 5
COPY MONITOR SUPPORT PROGRAMS
If your installation is running a current monitor, you should install
monitor support programs prior to building the 6.03 monitor. If you
have not previously installed the latest versions of monitor-related
programs, or you are building the 6.03 monitor from scratch, you must
now copy monitor support programs to SYS with PIP. Note that ACCT.SYS
must be on SYS or you will not be allowed to LOGIN. You should log in
under [1,2] and type:
.RUN DSK:PIP[10,7]
*SYS:<155>/X=DSK:[10,7]filename,...
where filenames include:
DDT.REL,JOBDAT.REL,*.EXE,*.ATO
*.VMX (for virtual memory monitors only; it is on the VM
unbundled tape)
Some of the files may also have to be copied from [10,6].
Also, if you do not have your system accounting files built, copy
*.SYS specifying protection <157>.
If your installation has file structures of different speeds, put the
following files on the fastest structure (DSKA):
DDT.REL,FORLIB.REL,BASIC.EXE,LIBOL.REL (if using
nonreentrant COBOL)
You may want to put the following files on every file structure:
SYSTAT.EXE,REDALL.EXE,DSKRAT.EXE,DSKLST.EXE,
BACKUP.EXE
because they are needed to track down possible disk hardware problems.
CHAPTER 6
WRITE BOOTS
You now must write BOOTS on disk packs with WTBOOT. WTBOOT is a
generic name for WABOOT, WIBOOT, and WLBOOT, where the second letter
(A,I,L) corresponds to your system processor (KA,KI,KL).
This step is required on all systems to write the latest version of
BOOTS onto the disk packs.
WTBOOT can write BOOTS, a disk bootstrap loader program, on blocks 0
and 4 through 7 of all disk packs in the system. Run WTBOOT, WABOOT,
WIBOOT, or WLBOOT once, the first time a new 6-series monitor is
built. It is necessary to run WTBOOT again only if the disks are
reformatted or if there is a new BOOTS. BOOTS should be written on
each disk pack so that the crash procedure works.
When writing BOOTS on disk packs for a KL10 system with 6.03, it is
important that you use the WLBOOT distributed with the 6.03 monitor.
This is important because previous versions of WLBOOT wrote in block 0
of the disk pack, which contains the front-end file system. Writing
in block 0 wipes out the front-end bootstrap.
To run WTBOOT log in to [1,2] and type:
.R W?BOOT
W?BOOT responds with:
SELECT UNITS?
where you respond with NO or YES. If you respond with NO, the default
BOOTs will be written on all disk packs. If you are using a KL10,
W?BOOT will not write the READIN loaders. If not using a KL10, W?BOOT
will write the default READIN loader on all RS04s and RP0xs. Note
that RP0x indicates RP04, RP05, or RP06.
If you respond with YES, W?BOOT responds with:
UNIT unit-name
You are to respond with the type of BOOTs you want written on each
type of unit in the system. You respond with DP, RP, R2, ZERO, or
SKIP.
DP writes RP02/RP03 BOOTS on the pack mounted on that unit.
RP writes RH10/RP0x BOOTS on the pack mounted on that unit.
R2 writes RH20/RP0x BOOTS on the pack mounted on that unit.
ZERO prevents the monitor from reading BOOTS from the pack
WRITE BOOTS Page 6-2
mounted on that unit.
SKIP prevents W?BOOT from changing the pack mounted on that unit.
If you respond with a carriage return, W?BOOT writes the default
BOOTS. The default is RP02/RP03 BOOTS on RP02/RP03 packs, RH10/RP0x
BOOTS on RP0x packs (i.e., RH10s), RH20/RP0x BOOTS on RH20 packs. For
each RS04 or RP0x, W?BOOT asks you if the READIN loader should be
written in 18- or 22-bit mode format. You respond with NONE, 18, 22,
or carriage return.
NONE if no READIN loader should be written on block 0 (KL10).
22 if the DF10C is normally in 22-bit mode format.
18 if the unit is not on a controller connected to a DF10C, or
if the DF10C is normally in 18-bit mode format.
CR for the default response, which is 18-bit mode on even units
and 22-bit mode on odd units.
When the procedure is completed, the system responds:
EXIT
Note that WTBOOT writes BOOTS only on disk packs that are on line and
not write locked. (For more information on BOOTS, or on loading BOOTS
with a paper tape, refer to the DECsystem-10 Operator's Guide in the
DECsystem-10 Software Notebooks.)
CHAPTER 7
MERGE MODIFICATIONS
Omit this step if you are building an unmodified DEC monitor and do
not intend to merge any modifications developed at your installation.
SOUP (Software Updating Package) is a set of programs used to update
source files. It is designed to simplify the merging of customer
modifications with DEC-supplied files.
The DECsystem-10 Software Updating Package Programmer's Reference
Manual in the DECsystem-10 Software Notebooks describes in detail the
set of programs in the SOUP package and explains how to use them. If
you have several modifications to be merged with many source files,
you may use the DECsystem-10 Software Notebook document entitled
SOUPing Parallel Monitor Developments Together (MULTI) as a guide.
NOTE
Use of SOUP requires that the
appropriate source (.MAC) files have
been restored from the distribution
tape.
Be very familiar with SOUP before
attempting to use it.
CHAPTER 8
GENERATE A MONITOR WITH MONGEN
8.1 INTRODUCTION
The Monitor Generator (MONGEN) is a dialogue program enabling you to
choose the software most appropriate for your installation and to
define your hardware configuration.
The dialogue produces the files needed to build the monitor, and it
then tailors them to your system configuration. Parameter assignments
within these files determine how the monitor data base modules are
assembled. The dialogue consists of questions typed by MONGEN on your
terminal and your answers followed by a carriage return.
Before running MONGEN, you should
1. Become acquainted with the current version of the MONGEN
dialogue.
2. Be thoroughly familiar with your system configuration.
3. Know what changes you wish to make to the monitor currently
running.
8.2 MONGEN ORGANIZATION
The MONGEN dialogue is divided into four sections, each asking
questions about a specific aspect of the system configuration. The
four sections are:
HDWGEN which defines the system hardware configuration
(Section 8.4)
TTYGEN which defines the system terminal configuration
(Section 8.5)
NETGEN which defines the system network configuration (refer
to Section 8.6)
FGEN which defines the system software features and options
(Section 8.7)
The first time you generate the files for a new monitor, you should
answer the questions in each of the four sections. On subsequent
occasions, you can change only one, two, or three of the four
sections. The number of sections you change depends on where your
changes are and which section(s) your changes affect. You need only
answer the questions in the section you are changing.
GENERATE A MONITOR WITH MONGEN Page 8-2
After you answer the questions for the four sections of MONGEN, four
files will be generated - - HDWCNF.MAC, TTYCNF.MAC, NETCNF.MAC, and
F.MAC. These are the default names of the files containing your
MONGEN dialogue. You may change these file names by responding to
question number 2 in HDWGEN, TTYGEN, NETGEN, and/or FGEN.
If, after generating the four files, you wish to redefine one of them,
you can reanswer the questions for that section. Then,you can
assemble this new file along with the other three.
In this chapter, each MONGEN question is described within its
appropriate section.
8.3 EXECUTING MONGEN
To execute the version of MONGEN on SYS:, type the following command:
R MONGEN
To execute the version of MONGEN on your disk area, type the following
command:
RUN MONGEN
MONGEN will respond with
MONGEN FOR 603 MONITORS
/HELP(PROMPT,SHORT,LONG):
MONGEN wishes you to indicate which one of the three following
dialogue modes you wish to use.
MODE EFFECT
SHORT MONGEN types only abbreviated questions on your
terminal. This mode assumes that you need no
explanation of the question and that you are familiar
with the choice of answers. SHORT mode is recommended
only for experienced users of this version of MONGEN.
PROMPT MONGEN types the same abbreviated question as in SHORT
mode, but adds, within parentheses, a choice of
answers. (Note that the first item within the
parentheses is the default answer.) This mode is
recommended for users familiar with the MONGEN
dialogue, but who would prefer to have the choice of
answers typed with each question.
LONG MONGEN types the question, the choice of answers within
parentheses, and an explanation of the question within
square brackets. This mode is recommended for all
first time users of this version of MONGEN.
The mode you choose will be used throughout the dialogue. To change
modes during a MONGEN dialogue, answer a question by typing
/HELP
This answer will change the mode to the next longest, but only for the
current question. To change modes for the rest of the dialogue, type
/HELP:x
GENERATE A MONITOR WITH MONGEN Page 8-3
where x is either SHORT, PROMPT, or LONG (abbreviated S, P, or L).
8.3.1 MONGEN Dialogue Format
MONGEN questions and responses are in the form:
question(possible answers) [explanation]:answer
The possible answer portion of this format description is usually in
one of the following forms:
(min-max) Legal range
(default,min-max) Default response followed by legal range
(Y,N) Yes or No
(a,b,c, ... z) Multiple choice
Type your response directly after the colon. When you finish press
the RETURN key. If you want the default response, press the RETURN
key immediately after the colon. Because of the nature of some MONGEN
questions, possible answers are not offered.
Answers to questions requesting consecutive TTY line number
identification can be typed on one line. For example:
1-10 specifies TTY line numbers 1 through 10.
13 specifies TTY line number 13.
22-24 specifies TTY line numbers 22 through 24.
If you type an illegal or inappropriate response, MONGEN will type an
error message after which you may respond to the question again. The
possible error messages are listed in Section 8.19.
8.3.2 Generating a New Monitor
After typing the command R MONGEN and after you respond to the /HELP
question, MONGEN will type
1^1 WHICH GEN (HDW, TTY, NET, F):
Respond with either HDW, TTY, NET, or F to indicate which section you
wish to configure. For example, if you respond with HDW, MONGEN will
start typing the HDWGEN questions. When the HDWGEN MONGEN dialogue
has been completed, question number one will be retyped requesting you
to type another section name.
------------------
1. Note that all questions within this document are preceded with a
number. These numbers are used for reference only; they are not
printed by MONGEN.
GENERATE A MONITOR WITH MONGEN Page 8-4
8.4 HDWGEN
HDWGEN is the section containing questions about the central
processor(s), disklike storage devices, real-time devices, and
peripheral devices.
2 OUTPUT (DSK:HDWCNF.MAC):
If you accept the file HDWCNF.MAC and device DSK:
as the media for storing your answers to MONGEN,
press the return key. If you do not accept them,
type your preferred device name, file name, and
project/programmer number.
3 DECSYSTEM10 (1040,1050,1055,1070,1077,1080,1088):
Respond with the type of system you are
configuring.
Type of System Response
KA10 CPU (small) 1040
KA10 CPU (large) 1050
dual-KA10 CPUs 1055
KI10 CPU 1070
dual-KI10 CPUs 1077
KL10 CPU 1080
dual-KL10 CPUs 1088
4 SYSTEM NAME:
Respond to this question with your system name in
24 characters or less. This system name is the
'banner' printed when you issue the INITIA
command. For example, KG77A KL10 #1026.
5 CPU0 SERIAL # (1-10000):
Respond with CPU0's serial number, which falls
into the range 1 to 10000. MONGEN will repeat
this question once for each processor in your
system.
5A CPU1 SERIAL #(1-10000):
Respond with CPU1's serial number, which falls
into the range 1 to 10000. (This question is
typed only if you specified in question 3 that you
have a dual-processor system.)
6 # DK10'S ON CPU0 (1,0-2):
If your processor is a KL10, this question will
not be typed unless you answer N to question 8.
Respond with the number of DK10 real-time clocks
for CPU0. The default number is one; the maximum
is two. (The DK10 real-time clock keeps time in
units of 10 microseconds.) If you have a dual
processor system, this question will be repeated.
GENERATE A MONITOR WITH MONGEN Page 8-5
6A # DK10'S ON CPU1 (1,0-2):
Respond with the number of DK10 real-time clocks
for CPU1. (This question is typed only if you
specified in question 3 that you have a
dual-processor system.)
7 EXCLUDE MONITOR OVERHEAD FROM USER RUN TIME (Y,N):
Answer Y (Yes) or N (No). The answer to this
question is independent of whether time accounting
is performed with the line frequency clock (50 or
60 Hertz) or the real-time clock (100000 Hertz).
If you respond with Y, users will receive a more
accurate accounting of their processor usage,
because user runtime is reported independent of
system loading. However, all installations that
charge their customers for processor usage should
be aware that the exclusion of monitor overhead
will decrease the reported user runtime by 10 to
30 percent.
If you respond with N, monitor overhead will be
included in run-time statistics. Whether you
respond with Y or N to this question, monitor
overhead will be reported as a separate statistic
that may be used by the system manager in
determining the amount of CPUtime spent for
overhead.
With EBOX/MBOX accounting (KL10 systems only),
meters are turned off at the beginning of the
overhead period and turned on again at the end.
If your system is not a KL10 (1080 series), jump
to question number 10.
8 EBOX/MBOX TIME ACCOUNTING (Y,N):
This question is typed only if you responded to
question number 3 with 1080 (1088, 1090, 1099).
Answer Y or N. If you respond with Y, run-time
accounting will be computed using the KL10
internal clocks and question number 9 will be
typed. If you respond with N, the next question
typed will be question number 6. For maximum user
run-time reproducibility, you should answer Y to
both this question and question number 9.
9 EXCLUDE PI TIME FROM USER RUN TIME (Y,N):
This question is typed only if you responded with
1080 to question number 3 and Y to question number
7. Answer Y or N. If your response is Y, the
EBOX and the MBOX accounting clocks will be
stopped whenever a Priority Interrupt is in
progress. For maximum user run-time
reproducibility, you should answer Y to both this
question and question number 8.
10 HIGH PRECISION TIME ACCOUNTING (Y,N):
GENERATE A MONITOR WITH MONGEN Page 8-6
This question is typed only if you responded to
question 6 and/or question 6A with a value equal
to or greater than 1. Answer Y or N to this
question. If Y, more precise run-time
measurements will be given because the system will
use 10-microsecond time accounting. This will
more accurately reflect real CPU-time, independent
of context switching. (Therefore, a compute-bound
job is not charged for a whole 60th of a second,
even though an I/O-bound job ran during that time
interval.) If N, question number 11 will be typed.
11 DK10 SOFTWARE (Y,N):
This question is typed only if you responded with
N in response to question number 10 and a value
greater than or equal to 1 in response to question
number 6 and/or question number 6A. To include
the real-time clock service routine, answer Y; to
omit it, answer N. This device is not available
to a user program unless it issues an RTTRP
monitor call. The service routine is used only
for high-precision time accounting.
12 # DATA CHANNELS (2,1-8):
Respond with the number of data channels in your
system (DF10s, DF10Cs, DX10s, or RH20s for disk or
tape). The default is 2; the maximum is 8.
NOTE
Questions 13 through 23 are repeated once
for each data channel you specified in
response to question 12.
13 CHANNEL n TYPE (DF10,DF10C,DX10,RH20):
Respond with the type of channel n (DF10,DF10C or
DX10). This question is asked once for each data
channel that you specified in response to question
number 12. If you respond with DF10 or DF10C,
read questions 14 through 20B. If you respond
with DX10, read questions 21 and 21A. If you
respond with RH20, read questions 22 through 23B.
Note that channel n corresponds to the first,
second, third ... channel you configure. The n
does not necessarily correspond to a hardwired
channel number.
14 # RC10S (0-2):
Respond with the number of controllers for the
RD10 Burroughs disk and RM10B Bryant drums on
channel number n. If you respond with 0, question
number 14A will not be printed.
GENERATE A MONITOR WITH MONGEN Page 8-7
14A # UNITS ON FHx (1-4):
This question is repeated n times (where n is your
response to question number 14.) Each time the
question is typed, x will be incremented to the
next alphabetic in sequence starting with A (e.g.,
FHA, FHB,...). Respond with the number of disks
and/or drums associated with controller FHx.
15 # RH10S FOR RS04'S (0-3):
Respond with the number of controllers for the
RS04 swapping disks on channel number n. If you
respond with 0, question number 15A will not be
typed. Both RS04 and RP04 disk units have RH10
controllers; RS04s have the lower device codes
between the two.
15A # UNITS ON FSx (1-8):
This question is repeated n times (where n is your
response to question number 15). Each time the
question is typed, x will be incremented to the
next alphabetic in sequence starting with A (e.g.,
FSA, FSB,...). Respond with the number of
swapping disks associated with controller FSx on
channel n.
16 # RH10S FOR RP04'S (0-3):
Respond with the number of controllers for RP04
and RP06 disk pack units on channel number n. If
you respond with 0, question number 16A will not
be typed. Both RS04 and RP04 disk units have RH10
controllers; RS04s have the lower device codes of
the two.
16A # UNITS ON RPx (1-8):
This question is repeated n times (where n is your
response to question number 16). Each time the
question is typed, x will be incremented to the
next alphabetic in sequence starting with A (e.g.,
RPA, RPB,...). Respond with the number of disk
packs associated with controller RPx.
17 # RP10S(0-3):
Respond with the number of controllers that are in
the system for RP02 and RP03 disk pack units on
channel number n. If you respond with 0, question
number 17A will not be typed.
17A # UNITS ON DPx (1-8).
This question is repeated n times (where n is your
response to question number 17). Each time the
question is typed, x is incremented to the next
GENERATE A MONITOR WITH MONGEN Page 8-8
alphabetic in sequence starting with A (e.g., DPA,
DPB,...). Respond with the number of disk pack
units associated with controller DPx.
18 # TM10BS (0-2):
Respond with the number of controllers that are in
the system for NRZI-only drives on channel n. If
you respond with 0, question number 18A will not
be typed.
18A # UNITS ON MTx (1-8):
This question is repeated n times (where n is your
response to question number 18). Each time this
question is typed, x is incremented to the next
alphabetic in sequence starting with A (e.g., MTA,
MTB,...). Respond with the number of magtape
units associated with controller MTx.
19 # TC10CS (0-1):
Respond with the number of special systems tape
controllers for TU42s and TU43s on channel n. If
you respond with 0, question number 19A will not
be typed.
19A # UNITS ON MTx (1-8):
This question is repeated n times (where n is your
response to question number 19). Each time this
question is typed, x is incremented to the next
alphabetic in sequence. Respond with the number
of TU42s and TU43s associated with controller MTx.
20 # RH10'S FOR TM02'S (0-2)
Respond with the number of mass-bus tape
controllers for TU16s and TU45s on channel n. If
you respond with 1 or 2, questions 20A and 20B
will be typed.
20A HOW MANY TM02'S ON RH10 #n (1-8)
Respond with the number of subunits (tape drives)
on each unit. This question is repeated once for
each RH10 specified in question 20.
20B HOW MANY DRIVES ON TM02 n (1-8)
Respond with the number of drives on the specified
TM02.
If you have more data channels to configure, refer to question 13,
otherwise, refer to question 24. Questions 14 through 20B will be
repeated if you have to configure more DF10 or DF10C channels.
GENERATE A MONITOR WITH MONGEN Page 8-9
21 # CONTROLLERS (0-1):
Respond with the number of TX01 or TX02
controllers for TU70 tape drives. This question
is typed only when you have configured a DX10 data
channel. If you respond with 1, question 21A is
typed.
21A # UNITS ON MTx (1-8):
Respond with the number of units on TU70 tape
drive.
If you have more data channels to configure, refer to question 13.
Otherwise, refer to question 24. Questions 21 and 21A will be
repeated if you have more DX10 channels to configure.
22 # RH20S for RP04'S, RP06'S (0-1):
Respond with the number of RH20 controllers for
RP04 and RP06 disk pack units on channel n. This
question is typed only if you configure an RH20
channel. If you respond with 1, question 22A is
typed.
22A # UNITS ON RPx (1-8):
Respond with the number of units associated with
controller RPx.
23 # RH20'S FOR TM02'S (0-1):
Respond with the number mass-bus controllers for
TU16s and TU45s on channel n. If you respond with
1, questions 23A and 23B will be typed.
23A HOW MANY TM02'S ON RH20 # x (1-8):
Respond with the number of tape drives (subunits)
associated with RH20 number x.
23B HOW MANY DRIVES ON TM02 x (1-8):
This question is repeated once for each TM02 you
specified in response to question 23A.
If you have more data channels to configure, refer to question 13.
Otherwise, refer to question 24. Questions 22 through 23B will be
repeated if you have more RH20 channels to configure.
24 # TM10AS (0,0-2):
Respond with the number of I/O Bus-type
controllers for NRZI-only drives. If you respond
with 0, question 24A will not be typed.
24A #UNITS ON MTx (1-8):
GENERATE A MONITOR WITH MONGEN Page 8-10
This question is repeated n times (where n is your
response to question 24). Each time this question
is typed, x is incremented to the next alphabetic
in sequence. Respond with the number of magtape
units associated with MTx.
For each magnetic tape you have configured, MONGEN asks if it is a
7-track unit and if it is capable of 6250 bits/inch density.
First, MONGEN prints the following:
SPECIFY WHICH DRIVES (M-N) ARE 7-TRACK DRIVES. [TYPE ONE
NUMBER (M) OR ONE RANGE (M-N) OR ALL ON SEPARATE LINES.
TYPE AN EXTRA CARRIAGE RETURN WHEN THROUGH.]
FOR CONTROLLER MTA
.
.
.
FOR CONTROLLER MTB
.
.
.
After MONGEN has asked the above question for each magnetic tape
drive, it then asks the following question for each tape drive:
SPECIFY WHICH TAPE DRIVES (M-N) ARE CAPABLE OF 6250 BPI
DENSITIES. [TYPE ONE NUMBER (M) OR ONE RANGE (M-N) OR ALL
ON SEPARATE LINES. TYPE AN EXTRA CARRIAGE RETURN WHEN
THROUGH.]
FOR CONTROLLER MTA
.
.
.
FOR CONTROLLER MTB
.
.
.
25 # JOBS (1-n):
Respond with the maximum number of jobs attached
and/or detached. Do not include the null job.
Each job requires approximately 1/4 of 1K core, so
specify only the number of jobs needed.
26 MAX. K OF CORE FOR EACH JOB (0, 0-n):
This question is typed only for KA10-based
systems; question 26A is typed for KI10 and
KL10-based systems. Respond with the maximum
amount of core that any one job may use. This
value is specified in number of 1K core blocks. A
response of 0 indicates all of core.
26A MAX. P OF CORE FOR EACH JOB (256-512):
GENERATE A MONITOR WITH MONGEN Page 8-11
This question is typed only for KI10/KL10 systems.
Respond with the maximum amount of core which any
one job may use. This value is specified in
number of 512-word pages. A response of 0
indicates all of core.
27 # K TOTAL SYSTEM CORE (32-4096):
This question is typed only for KA10-based
systems. Respond with the amount of memory the
system is expected to use. At monitor
initialization time, the operator will be asked if
the actual amount of memory is less than your
response to this question. You should, therefore,
specify the correct amount. If this value is
smaller than the amount of memory you want to put
on line, the monitor will not allow the operator
to put it on line. If more than 256K is desired,
you must set the feature test switch FT22BIT to
-1.
27A #P TOTAL SYSTEM CORE (256-4095):
This question is typed only for KI10/KL10-based
systems. Respond with the amount of memory the
system is expected to use. At monitor
initialization time, the operator will be asked if
the actual amount of memory is less than your
response to this question. You should, therefore,
specify the correct amount. If this value is
smaller than the amount of memory you want to put
on line, the monitor will not allow the operator
to put it on line.
28 CLOCK TICKS PER SECOND (60,50):
Respond with the number of clock ticks per second
(power line frequency). In the U.S., the response
should be 60; in most other countries, the
response should be 50. If your system has a KL10
processor, you may specify either 50 or 60; your
response will determine the frequency of clock
interrupts, which are generated by the KL10
interrupt timer. In KA10- and KI10-based systems,
your response must conform to the actual frequency
of the system power supply.
29 # REAL-TIME DEVICES (0,0-77):
Respond with the number of real-time devices that
can be put on a priority interrupt channel
simultaneously. The default is zero; the maximum
is 77. If your response is 0, question 29A will
be typed.
29A ALLOW JOBS TO BE LOCKED IN CORE (Y,N):
Answer Y or N. A response of Y will allow jobs to
be locked in core; N will not allow jobs to be
GENERATE A MONITOR WITH MONGEN Page 8-12
locked in core and question 30 or 30A will not be
typed.
30 # K MIN GUARANTEED AMONG JOBS NOT LOCKED IN CORE (0, 0-256):
This question is typed only for KA10-based
systems. Respond with a value indicating the
amount of core guaranteed among all jobs that are
not locked in core. The default is 0, which
indicates all of core. If you respond with 0 and
answer Y to question 29A, no jobs will be allowed
to lock in core. The value you specify in
question 30 can be overridden by a SET CORMIN
command.
30A # P MIN GUARANTEED AMONG JOBS NOT LOCKED IN CORE (0,0 - max):
This question is typed only for KI10/KL10-based
systems. Respond with a value indicating the
amount of core guaranteed among all jobs that are
not locked in core. The default is 0, which
indicates all of core. If you respond with 0 and
answer Y to question 29A, no jobs will be allowed
to lock in core. The value you specify in
question 30A can be overridden by a SET CORMIN
command.
31 # HIGH PRIORITY QUEUES (0,0-15):
Respond with the number of high priority queues in
your system. The default is 0; the maximum is
15.
32 CCL COMMANDS TO STAY IN CORE (Y,N):
Answer Y or N. Y allows system programs to pass
commands to one another via core (TMPCOR monitor
call) rather than disk. The preferred answer to
this question is Y.
33 METER (Y,N):
Answer Y or N. Y includes the METER monitor call
software in your system, which enables performance
analysis metering. N omits the METER monitor call
software from your system.
34 SYSCHK (Y,N):
Answer Y or N. Y will allow an initial hardware
integrity check to be performed at monitor
initialization time. N makes it impossible for
this check to be performed.
32 MSGSER (Y,N):
Answer Y or N. Y causes the multiplexed-channel
GENERATE A MONITOR WITH MONGEN Page 8-13
(MPX) software to be loaded. MPX allows more than
one device to be associated with a given I/O
channel. (This feature is required for operation
of TYPESET-10 and/or MCS-10.) N prevents the
loading of the MPX software.
36 PSISER (Y,N):
Answer Y or N. Y causes the PSISER program to be
loaded. PSISER provides the software interrupt
system, which enables users to specify conditions
for which an interrupt is to occur. PSISER is
required for the operation of MCS-10, GALAXY,
and/or TYPESET-10. N prevents the loading of
PSISER.
37 IPCF (Y,N):
Answer Y or N. Y causes IPCF (Inter-Process
Communication Facility) to be loaded. IPCF allows
jobs and/or system processes to communicate with
each other. IPCF is required for the operation of
MCS-10, GALAXY, and/or TYPESET-10. N prevents the
loading of the IPCF software.
38 ENQ/DEQ (Y,N):
Answer Y or N. Y causes the ENQ/DEQ software to
be loaded. This software provides synchronization
primitives used by COBOL and for simultaneous file
updates. N prevents the loading of the ENQ/DEQ
software.
39 # CDRS (1,0-2):
Respond with the number of card readers to be
configured. The default is 1; the maximum is 2.
40 CDP (Y,N):
Respond Y or N. Y will provide software for a
card punch; N will not.
41 DIS (Y,N):
Answer Y or N. Y provides software for a display
device (e.g., VP10, 340, 30, VB10C); N will not.
The display device asked about here is not to be
confused with a display terminal. If you answer
Y, question number 41A will be typed. Question
number 41A will not be typed if you answer with N.
41A TYPE (VP10,340,VB10C):
Respond with the type of display device you have
(VP10, 340, or VB10C). If you have type 30,
answer VP10.
GENERATE A MONITOR WITH MONGEN Page 8-14
42 # TD10S (1,0-2):
Respond with the number of DECtape controllers.
The default is 1; the maximum is 2. If your
response is 0, question 42A will not be typed.
42A # UNITS ON DTx (1-8):
This question will be repeated n times (where n is
your response to question number 39). Each time
this question is typed, x will be incremented to
the next alphabetic in sequence beginning with A
(e.g., DTA, DTB,...). Respond with the number of
DECtape units associated with controller DTx.
43 LPTS (1,0-3):
Respond with the number of line printers supported
by your system. The default number is 1. If you
respond with 0, question 43A will not be typed.
43A LPTn LOWER CASE (Y,N):
Answer Y or N. Y indicates that LPTn has
lower-case capability. N indicates that the line
printer is upper-case only. This question is
repeated n times (where n is your response to
question number 40).
44 PLTS (0,0-2):
Respond with the number of incremental plotters
supported by your system. The default is 0; the
maximum is 2.
45 PTP (Y,N):
Answer Y or N. Y indicates that your system
includes a paper tape punch. The inclusion of a
paper tape punch is standard for KA10/KI10-based
systems, but normally excluded on KL10-based
systems.
46 PTR (Y,N):
Answer Y or N. Y indicates that your system
includes a paper tape reader.
47 # PTYS (20,0-510):
Respond with the number of pseudo-terminals.
(Each operator service routine and batch stream
needs one.) This response will usually equal your
response for question number 25, which asks for
the number of jobs to be attached and/or detached
at one time. If your response is greater than
zero, FTPTYUUO must be set in F.GEN to -1. Refer
GENERATE A MONITOR WITH MONGEN Page 8-15
to Chapter 8.
48 DC44 (Y,N):
Answer Y or N. Y indicates that your system
includes a DC44 front-end. DC44 is a TYPESET-10
front-end that supports a high-speed paper-tape
reader and punch, and an on-line photocomposition
machine. If you answer N, question numbers 48A,
B, C, and D, and E will not be typed. If you
respond with Y, you must set the FTTYPE feature
test switch to -1.
48A DL10 PORT NUMBER FOR DC44 (0,0-7):
Respond with the DC44 channel number (port
number). The default is 0; the maximum is 3.
DL10 port numbers for DC44s should be numbered
sequentially, beginning with 0.
48B # OF PA611RS (0,0-32):
Respond with the number of high-speed paper tape
readers to be supported with the DC44. The
default is 0; the maximum is 32.
48C # OF PA611PS (0,0-32):
Respond with the number of high-speed paper-tape
punches to be supported with the DC44. The
default is 0; the maximum is 32.
48D # OF LPC11S (0,0-6):
Respond with the number of on-line
photocomposition machines to be supported with the
DC44. The default is 0; the maximum is 6.
48E PC11 (Y,N):
Answer Y or N. Y indicates that there is a paper
tape reader and a paper tape punch on the DC44.
49 # OF DA28S (1,0-4):
Respond with the number of inter-processor data
channels for the PDP-8/11/15. The default is 1;
the maximum is 4. If your response is 1 or
greater and the feature test switch FTXTC equals
0, MONGEN assumes that you do not want any DA28s
supported.
49A # OF LINES FOR TTY POOL (8,0-512):
Respond with the number of terminal lines reserved
for use on DA28 channels. The default is 8; the
GENERATE A MONITOR WITH MONGEN Page 8-16
maximum is 512.
50 # DAS78'S (0,0-8):
Respond with the number of DAS78s (bisynchronous
support as used by IBM 360s, 370s, and/or 2780s).
If you respond with a value greater than 0, you
must set the FTDAS78 feature switch to -1.
50A WHICH DL10 PORT IS THE DAS78 CONNECTED TO (0,0-7):
Each DAS78 PDP-11 is connected to a DL10 port.
Respond with the number of the DL10 port.
50B DECIMAL LINES ON THE DAS78 (1-16):
Each DAS78 can support up to 16 bisynchronous
lines for IBM 360s, 370s, and/or 2760s; respond
with the number of lines (in decimal) on this
DAS78.
51 DECIMAL "SYMBOL,VALUE"
52 OCTAL "SYMBOL,VALUE"
53 SIXBIT "SYMBOL,VALUE"
In response to questions 51, 52, and 53, type the
constants at your installation that deviate from
the standard. Unless you specify otherwise,
HDWGEN will define each symbol in accordance with
predefined standard values. A list and an
explanation of the standard symbols and decimal
default values appear in Sections 8.14.1, 8.14.2
and 8.14.3. For example, you could specify 1760
as the number of nanoseconds per memory cycle
(instead of 1000) by typing:
NSPMEM,1760
If you want to respond with more than one entry,
you must type each "symbol, value" on a separate
line, and end each line with a carriage return.
When you finish typing responses to question 51,
type an extra carriage return to force MONGEN to
type question 52. If all the standard values are
acceptable, type only a carriage return in
response to this question.
HDWGEN next types the questions
OCTAL "SYMBOL, VALUE"
and
SIXBIT "SYMBOL, VALUE"
requesting the OCTAL and SIXBIT values that are to
GENERATE A MONITOR WITH MONGEN Page 8-17
be changed from the standard settings. The answer
format is identical to that for the previous
question. Deviations from the list of default
values are the only responses required. If no
changes to the standard are required, you should
press the RETURN key in response to each of the
questions.
Sections 8.15.1 and 8.15.2 list and explain the
standard symbols and octal values; Section 8.16
lists and explains the standard symbols and SIXBIT
values.
54 TYPE "DEVICE-MNEMONIC,PI-CHANNEL,HIGHEST-AC-TO-SAVE"
55 TYPE "DEVICE-MNEMONIC, PI-CHANNEL" FOR SPECIAL DEVICES
56 TYPE "DEVICE-MNEMONIC, PI-CHANNEL, NO-OF-DEVICES"
Questions 54, 55, and 56 request special
information about non-standard peripheral devices
and their associated priority interrupt (PI)
channels. If your installation has no special
devices, answer the three questions with a
carriage return. If your installation has a
special device, turn to Section 8.17.
FILE DSK:HDW04D.MAC CLOSED [HDWGEN FINISHED]
8.5 TTYGEN
MONGEN's TTYGEN section allows you to define your system terminal
configuration. All possible questions are listed below and on
following pages. In order to start the TTYGEN dialogue, you must
respond with TTY to question number 1 described in Section 8.3.2.
2 OUTPUT(DSK:TTYCNF.MAC):
If you accept TTYCNF.MAC and the device DSK: as
the media for storing your answers to MONGEN,
press the RETURN key for your response. If you do
not accept them, type your preferred disk name,
file name, and project/programmer number.
3 HOW MANY DC10S (1,0-2):
Respond with the number of data line scanners to
be supported by your system. The default 1; the
maximum is 2. If you respond with 0, question
numbers 6 and 7 will not be typed. (Refer to
Section 8.18 for a further discussion of DC10s.)
4 HOW MANY DC68S (1,0-2):
Respond with the number of DC68s supported by your
system. The default is 1; the maximum is 2. The
GENERATE A MONITOR WITH MONGEN Page 8-18
DC68 is a PDP-8 680 or 680I communications system.
If you respond with 0, question number 8 will not
be typed.
5 HOW MANY DC76S (1,0-8):
Respond with the number of DC76s supported by your
system. The default is 1; the maximum is 8. The
DC76 is a PDP-11 communications system. If you
respond with 0, question number 9 will not be
typed. This is an unbundled product; therefore,
before typing a carriage return, make sure your
system includes a DC76.
NOTE
For DC68s and DC76s, the monitor does not
need to know the specific type of dataset
control hardware, because correspondence
of terminal lines to dataset controller is
performed within the hardware. When
laying out the terminal line configuration
in TTYGEN, you must specify the total
number of lines available through the DC68
and the total number of lines available
through each DC76. DC76 lines that
require dataset control must be defined as
dataset lines (via question 13).
6 # DC10B [OR 632] 8 LINE DATA GROUPS (1-8):
Respond with a value in the range 1 to 8, where:
1 indicates TTY0-7
2 indicates TTY0-17
3 indicates TTY0-27
4 indicates TTY0-37
5 indicates TTY0-47
6 indicates TTY0-57
7 indicates TTY0-67
8 indicates TTY0-77
NOTE
DC10Cs, DC10Ds, and DC10Hs should be
treated as DC10Bs. Line numbers for
DC10Bs and DC10Es must be dense or
the message SPECIFICATION
ERROR-DSCTAB will be printed at the
terminal when an attempt is made to
assemble the monitor.
7 DC10E DATA SET CONTROL GROUPS (0-7):
Respond with a value in the range 0 to 7.
8 CORRESPONDENCE OF DC10E LINES TO THE DC10B LINES (M-N,P)
GENERATE A MONITOR WITH MONGEN Page 8-19
Respond with m,p for one pair and m-n,p for a
range of pairs (where m is an octal DC10E line,
m-n is the octal range of DC10E lines, and p is
the octal DC10B line).
9 # OCTAL LINES ON DC68, INCLUDING ITS CONSOLE TTY (1-144):
Respond with the total number of lines available
under your system in the range 1 to 144.
10 WHICH DL10 PORT IS THE DC76 CONNECTED TO (0,1-7):
Each PDP-10 is connected to a DL10 port. If there
is only one PDP-11, it is always connected to port
number 0. If there are two PDP-11s, one is
connected to port 0 and the other to port 1.
11 DECIMAL LINES ON DC76 (1-129):
Respond with the number of lines (in decimal) on
the DC76. Each DC76F has 16 lines, and the
console terminal has one line. A DC76 with two
DC76Fs has 33 lines.
12 OPR OCTAL LINE # (CTY,0-n):
Respond with the octal number of device OPR
(operator privileged terminal). The default is
CTY.
Answer the following questions about your terminal lines (m-n). Type
one octal line number (m), one range (m-n), or CTY on separate lines.
Type an extra carriage return when you are finished.
13 DATA SET LINES
Respond with the line numbers of those lines that
are data set lines. Data set lines are a class of
terminals recognized by LOGIN. LOGIN will reset
the lines to computer echoing and hardware tabs.
DC76 lines that require dataset control must be
defined as dataset lines.
14 LINES WITH HARDWARE TABS
Respond with the line numbers of those terminals
with hardware generated tab settings. Most TTY
model 35s and 37s have hardware TABs, and most
model 33s do not. However, since the variety of
features found on model 33, 35, and 37 TTYs is so
great, every system administrator must determine
the physical characteristics of their installation
terminals. It is possible to override MONGEN TAB
settings with the SET TTY TAB and SET TTY NO TAB
commands.
15 REMOTE LINES
GENERATE A MONITOR WITH MONGEN Page 8-20
Respond with the line number of those lines which
are to be considered remote. All data sets are
implicitly remote; however, the installation may
give REMOTE status to any terminal, even those
that are hardwired. In this context, remote is a
software characteristic of the terminal that
imposes certain restrictions. For example, some
project/programmer numbers may not be allowed to
LOGIN at remote terminals. (These terminals are
not to be confused with those terminals located at
a remote station.)
16 LOCAL COPY LINES
Respond with the line numbers of those lines which
are to be considered local copy lines (full
duplex). With local copy lines, echoing is
provided by the terminal instead of by the
computer.
17 HALF DUPLEX LINES
Respond with the line numbers of those lines that
transmit in half-duplex mode. Half-duplex lines
provide two-way, alternate, independent
transmission over the same pair of wires.
NOTE
The term half-duplex is occasionally used
incorrectly. True half-duplex lines are
relatively rare. They include TWX data
sets and some local terminals connected by
DC10C interfaces. Some lines which are
loosely called half-duplex are actually
local copy full-duplex. Question number
15 asks for the identification of local
copy lines. (If you are interested in
reading more about these terms, refer to
Section 8.3 of the specification entitled
SCNSER, Level D Scanner Service in the
DECsystem-10 System Notebooks.)
18 SLAVES
Respond with the line numbers of those lines which
can be classified as slaves. Slave is a software
classification for terminals that cannot LOGIN or
control jobs. These terminals are generally used
to provide I/O for jobs that are controlled from
other terminals. A terminal can be classified as
a slave by any user, but it can be restored to
normal status only by the operator.
19 LINES WITH HARDWARE FORM FEED
Respond with the line numbers of those terminals
which are to have hardware-generated form feeds.
Most model 35 TTYs have hardware-generated form
GENERATE A MONITOR WITH MONGEN Page 8-21
feeds, and most model 33s do not. Since the
variety of features found in terminals is so
great, every system administrator must determine
the hardware characteristics of their installation
terminals. You may assume that no terminals have
hardware-generated form feeds, so that the scanner
service will always output five lines to simulate
form feeds. This can be changed by any user at a
terminal via the SET TTY FORM or SET TTY NO FORM
commands.
20 LINES WHICH RUN INITIA AT STARTUP
Respond with the line numbers of those terminals
that are to run INITIA at start up. INITIA is a
program that performs job initialization
functions. When the monitor starts up, INITIA is
run automatically for those lines specified in
your response to this question. Device OPR should
be specified in your response; remote terminals,
dataset terminals, and remote station terminals
should not be specified to run INITIA. There is a
file read by INITIA, called TTY.INI, which
contains a complete set of parameters for setting
up the teletype configurations. The advantage of
having TTY.INI is that TTY parameters can be
changed without rerunning MONGEN. The INITIA
specification in the DECsystem-10 Software
Notebooks provides further details about TTY.INI.
21 FILLER CLASS CODES (M-N,P):
Respond with the line numbers of those terminals
that require filler characters to standardize
output. P is the filler class code. (Refer to
the description of the SET TTY FILL command in the
DECsystem-10 Operating System Commands Manual for
an explanation of the filler class codes.)
8.6 NETGEN
NETGEN is the MONGEN section which defines your system network
configuration. All possible questions are listed below and on
following pages. In order to start the NETGEN dialogue, you must
respond with NET to question number 1, which is described in Section
8.3.2.
2 OUTPUT (DSK:NETCNF.MAC):
Respond with a carriage return if the file name
NETCNF.MAC and the device DSK: are acceptable to
store your NETGEN dialogue. If they are not
acceptable, respond with the desired file name,
device and/or project/ programmer number.
3 NETWORK SOFTWARE (Y,N):
Answer Y or N. Y indicates that your installation
GENERATE A MONITOR WITH MONGEN Page 8-22
desires the software for remote computers:
DECsystem-10s, PDP-11s, and PDP-15s. A response
of N indicates that the installation does not
desire network capabilities and, therefore, does
not require the associated software. The file
NETCNF.MAC will be closed; NETGEN is finished and
question number 1 will be typed again, allowing
you to configure FGEN. If you respond with Y, you
must also set the FTNET feature test switch to -1.
4 HOW MANY NODES DO YOU WISH TO SUPPORT (3,0-n)
Respond with the number of nodes/stations you wish
to be supported by your system. If you respond
with 0, NETGEN dialogue will terminate and
question number 1 will be printed.
5 HOW MANY DC75NP'S OR DN-87'S ON SYSTEM (1,1-8)
Respond with the number of synchronous front-ends
that are to be supported for networks on your
system. The default response is 1; the maximum
response is 8.
6 TO WHICH DL10 PORT IS THE DC75 or DN87 CONNECTED (0,0-8)
This question is repeated once for each front-end
you specified in question 5. Respond with the
DL10 port number connected to this DC75/DAS85.
The default response is 0; the maximum response
is 8.
7 HOW MANY DN87S's ON THE SYSTEM (1, 0-3):
Respond with the number of DN87S network
front-ends that are connected to DTE20s.
7A TO WHICH DTE20 IS THE DN87S CONNECTED (1, 1-3):
This question is typed once for each DN87S
specified in response to question 7.
8 NODE NUMBER OF CENTRAL SITE (1-77)
Respond with the node number to be associated with
the DECsystem-10 at the central site.
9 NAME OF CENTRAL SITE
Respond with the name of the central site in six
characters or less.
10 # OF REMOTE TTYS (0,0-512)
Respond with the maximum number of TTYs on network
nodes to be handled at any given moment.
GENERATE A MONITOR WITH MONGEN Page 8-23
11 # OF REMOTE CDRS (0,0-n)
Where n is the maximum response to question number
4 above. Respond with the maximum number of card
readers on network nodes to be handled at any
given time.
12 # OF REMOTE LPT'S (0,0-n)
Respond with the maximum number of line printers
on network nodes to be handled at any given time.
13 # OF REMOTE PTR'S (0,0-0)
14 # OF REMOTE PTP'S (0,0-0)
Respond with the maximum number of remote
paper-tape punches on network nodes to be handled
at one time.
15 # OF MTA'S (0,0-0)
17 # OF REMOTE PROCESSES (0,0-n)
Respond with the maximum number of remote
processes that can be connected at any given time.
To answer this question with a value greater than
1, you must set FTTSK to -1.
18 REMOTE DATA ENTRY SOFTWARE (Y,N):
Answer Y or N. Y indicates that you want the
software to support remote data entry terminals,
which are used for MCS-10 applications. If you
respond with Y, FTRDX in FGEN must be set to -1.
19 # OF CONNECTS (46,1-512)
Respond with the maximum number of simultaneous
connections.
8.7 FGEN
The FGEN dialogue simplifies the task of choosing appropriate software
from a wide selection. For example, FGEN allows you to select the
standard DECsystem-10 features for a given application.
If you are choosing all the standard software features, simply agree
during the dialogue to accept the standard symbol settings by typing a
carriage return. Your F.MAC file will then contain all these standard
settings; these, when assembled with the other GEN files, will
produce a monitor with the appropriate routines.
Standard DECsystem-10 software features are included or deleted by
setting feature test switches to -1 or to 0: -1 places the feature in
the monitor, and 0 omits it.
GENERATE A MONITOR WITH MONGEN Page 8-24
In order to set the appropriate feature test switches, you will need
the file FGEN.HLP. MONGEN will first look for FGEN.HLP on the
directory from which MONGEN came, then it will look for it on SYS:.
If FGEN.HLP cannot be found, the message %Can't find FGEN.HLP will be
printed on your terminal.
FGEN allows you to:
1. List the standard settings for a given application
(TINY,KALUG,KILUG,KAFULL,KIFULL,KLFULL).
2. Determine what the feature test switch symbols mean.
3. Alter the feature test switch settings.
4. Add new feature test switches.
2 OUTPUT (DSK:F.MAC):
If you accept F.MAC and the device DSK: as the
media for storing your answers to MONGEN, press
the RETURN key for your response. If you do not
accept them, type your preferred disk name, file
name, and project/programmer number.
3 FEATURE SET(TINY,KALUG,KILUG,KAFULL,KIFULL,KLFULL):
There are six standard feature settings; respond
with one of them. Only these standard settings
have been tested during monitor development and
are fully supported. Although installations may
alter the combination of individual feature test
switches or add new switches during FGEN, such
non-standard switch combinations have not been
tested and are not supported. We wish to
emphasize that you may alter the feature test
setting from the standard or LUG configurations;
however, the new combination is NOT supported by
DIGITAL.
4 STANDARD SETTING (YES,NO,LIST,EXPLAIN):
Respond with one of the words within the
parentheses.
YES Answer YES (the default) if you accept the DECsystem-10
standard feature test switch settings, as identified in
the previous question.
A YES ends FGEN and closes the F.MAC file. You may now
redefine another segment of MONGEN or exit from the
program.
NO A NO allows you to change the setting of individual
switches or add new feature test switches. You can do
this by typing the switch symbol (e.g., FTxxx) and the
value (0 or -1) in the form
switch, value
GENERATE A MONITOR WITH MONGEN Page 8-25
Press the RETURN key twice after the last entry.
LIST LIST allows you to examine the current setting for any
and all of the feature test switches. You may then
either accept or change the current setting.
The standard setting for each feature test switch
varies with the installation's DECsystem-10
application, e.g., TINY, KALUG, KILUG, KAFULL, KIFULL
or KLFULL. You can obtain the default value for an
individual feature test switch by typing LIST <CR> in
answer to the question, and then typing the requested
switch symbol followed by a carriage return. FGEN
returns that switch symbol and its current setting. To
obtain the value of another switch, type the requested
switch, followed by a carriage return. After the last
entry, press the RETURN key twice.
To obtain the current values for all of the feature
test switches you can type LIST <CR> and then type ALL,
followed by a carriage return. FGEN will list all
feature test switch symbols and their values on your
terminal.
EXPLAIN EXPLAIN allows you to examine the meaning of any or all
of the feature test switch symbols in addition to the
current value. You can obtain the default value and
meaning for a feature test switch symbol by responding
to EXPLAIN (or E) in the same way you could respond to
LIST (or L). For example, the dialogue (in prompt
mode) to obtain the default value and meaning for
feature test switches FTTIME and FTSLEE is as follows:
STANDARD SETTING (YES, NO, LIST, EXPLAIN): EXPLAIN <CR>
Only stand settings have been tested by DEC; all other
settings may produce incorrect operation.
Switch (switch to list or explain).
FTTIME <CR>
FTTIME, -1;
;TIME ACCUMULATION
FTSLEE <CR>
FTSLEE, -1;
;SLEEP UUO
After you have completed the dialogue sequence associated with NO,
LIST, or EXPLAIN, FGEN repeats the question:
STANDARD SETTING (YES, NO, LIST, EXPLAIN):
You should answer YES if all of the values are acceptable, NO if you
wish to change any settings, and LIST or EXPLAIN if you wish to obtain
any more default information.
When you answer NO and change the switch settings that you wish to
change, you may have FGEN type each switch and ask for its final
setting. FGEN asks
SET EACH SWITCH (Y,N):
When you type YES, FGEN types
GENERATE A MONITOR WITH MONGEN Page 8-26
FTxxxx,M (ON,OFF,LIST,EXPLAIN,END):
for each switch, where M=-1 if the switch is on and M=0 if the switch
is off. You can then specify whether the switch should be ON or OFF,
or whether to enter LIST or EXPLAIN mode as before. By typing END,
you can specify that no more switches are to be listed and that the
current values are to be used.
5 DO YOU WANT THE VIRTUAL MEMORY FACILITY (YES,NO):
This question will only be typed for KI10- or
KL10- based systems. Answer Y if you want virtual
memory capabilities and have VMSER (which is an
unbundled licensed product).
8.8 PRODUCING RELOCATABLE BINARY FILES COMMON.REL, COMMOD.REL, AND COMDEV.REL
8.8.1 Assembling the Configuration Files
You should assemble the files HDWCNF.MAC, TTYCNF.MAC, NETCNF.MAC,
F.MAC, with S.MAC and COMMON.MAC to produce the relocatable binary
file COMMON.REL, the files HDWCNF.MAC and F.MAC with S.MAC and
COMMOD.MAC to produce COMMOD.REL, and the files HDWCNF.MAC,
TTYCNF.MAC, NETCNF.MAC, F.MAC, with S.MAC and COMDEV.MAC to produce
COMDEV.REL.
To assemble a 6.03 Monitor type
COMPIL F/COMP,S/COMP,HDWCNF+NETCNF+TTYCNF+<COMMON,COMDEV,COMMOD>
It is important that F.MAC precede S.MAC, because S has conditional
assemblies depending on the contents of F. Also, all of the remaining
files must follow S.MAC and precede COMMON.MAC.
8.8.2 Assembly Error Messages from COMMON
One or more of the following messages are typed on your terminal if
your attempt to assemble COMMON is unsuccessful.
1. ?MORE THAN N. TTYS + PTYS EXCEED FIELD OF DDB.^1
The system capacity of 511 TTYs and PTYs has been exceeded.
You must rerun MONGEN, answering questions for fewer TTYs and
PTYs.
2. ?NOT ENOUGH PI'S TO SERVICE THIS CONFIGURATION SUGGEST
EDITING COMMON TO PUT MORE DEVICES ON A SINGLE CHANNEL.
The system capacity of seven priority interrupt channels has
been exceeded. You should assign more devices to each PI
channel by editing INTTAB in the source of COMMON or by
changing the PI assignment of the special devices that have
been added.
------------------
1. Note that all error messages are preceded by a number. These
numbers are used for reference only; they are not printed by COMMON.
GENERATE A MONITOR WITH MONGEN Page 8-27
3. ?FT2REL MUST BE -1 WHEN 2 RELOC REGISTERS EXIST.
4. ?PLEASE ASSEMBLE SOURCES WITH FTRTTRP = -1.
5. ?PLEASE ASSEMBLE SOURCES WITH FTLOCK = -1.
6. ?PLEASE ASSEMBLE SOURCES WITH FTHPQ = -1.
Messages 3 through 6 indicate that you specified incorrect
settings in F.MAC, i.e., you have asked for a feature to be
included with HDWGEN but you did not include the feature in
FGEN. You must rerun FGEN to correct the improper symbol
definitions.
7. ?SPECIFICATION ERROR - DSCTAB.
There is a specification error in the data set control table.
You must correct it by running MONGEN again. Usually this
error occurs when line numbers are not dense on DC10Bs and
DC10Es.
8. %NUMBER OF JOBS REDUCED TO MAX=511.
You specified more than 511 jobs. You should rerun MONGEN to
correct the error.
9. ?512 OR MORE PTY'S EXCEEDS FIELD OF DDB.
You specified more than 512 PTYs. You should rerun MONGEN to
correct the error.
10. ?TRAP OFFSET SWITCH CANNOT BE SET ON A SINGLE PROCESSOR
SYSTEM.
You incorrectly set the CPTOS switch. You must rerun MONGEN
and change the setting of the CPTOS switch to 0 if you have a
single processor system. (Refer to Section 8.15, Octal
Default Values.)
11. ?DL10 MAPPED AREA EXCEEDS 1K.
You made the mapped area too large. You must reduce the size
of the mapped area, or change jumper in DL10 hardware to 8K
and add "DLX8K ==1" in REMCNF.MAC.
12. ? DL10 MAPPED AREA EXCEEDS 8K.
The mapped area is still too large. You must reduce the size
of the mapped area following the procedures described in
message 11.
13. ? DLXLNG IS NOT EQUAL TO LENGTH OF MAPPED AREA.
You have specified DLXNG to be either less than or greater
than the length of the mapped area. You must change it so
that it is equal to the length of the mapped area. (This is
an internal error and should not occur.)
GENERATE A MONITOR WITH MONGEN Page 8-28
8.9 DESCRIPTION OF THE CONFIGURATION FILES
The following configuration files are produced by the MONGEN dialogue
program:
1. HDWCNF.MAC is produced by HDWGEN
2. TTYCNF.MAC is produced by TTYGEN
3. NETCNF.MAC is produced by NETGEN
4. F.MAC is produced by FGEN
All four of these files contain the following information for each of
the questions asked:
1. The question asked by the program as a comment
2. Your response to the question as a comment
3. The MACRO definitions containing the switch or symbol and its
value.
MONGEN always defines all the symbols that appear in the configuration
files, regardless of your answer to a question. Therefore, you do not
have to study the MONGEN program in order to understand what happens
on certain questions. It is sufficient that you only look at the
listings of the configuration files that are assembled with
COMMON.MAC.
8.9.1 Description of COMMON.MAC
Since COMMON.MAC selectively assembles only those items needed for the
defined configuration, it contains the following items:
1. The default symbols that you can override during the MONGEN
dialogue.
2. Special lower core locations (below 400 in the monitor).
3. The monitor startup locations.
4. The EXEC page map for a KI10/KL10 processor.
5. The PDP-10/PDP-11 shared core area on systems with DC75s.
6. CPU data blocks (CDBs) for each processor.
7. Variable data locations for the monitor.
8. The job and high segment tables.
9. The terminal data base.
10. Monitor initialization code to link device data blocks and to
create multiple copies for all multiple devices, with the
exception of disk.
11. Special MACROs to define the PI assignment. These MACROs
create a two-word-per-entry table that contains the
following:
GENERATE A MONITOR WITH MONGEN Page 8-29
a. Device Data Block address
b. Number of devices
c. Priority Interrupt channel for device
d. Interrupt location for device
e. Length of Device Data Block if multiple device
12. Locations set by ONCE-Only dialogue.
13. System error stop code (407 restart).
14. Common subroutine returns.
15. Subroutine to save and restore preserved accumulators.
16. Common byte pointers.
17. The PI channel save and restore routines.
18. Code to handle traps to 40 and 60 (primary CPU) and 140 and
160 (secondary CPU) and the code to handle APR interrupt
entry and exit.
19. Remote communications entry points and tables.
20. Real-time trapping tables.
21. High priority queue UUO code.
22. Scheduler queue definitions and tables.
8.9.2 Description of COMMOD.MAC
You assemble COMMOD.MAC with HDWCNF.MAC, F.MAC and S.MAC to produce
COMMOD.REL. COMMOD selectively assembles the items required for the
defined configuration. It contains the following items:
1. The assembly instructions.
2. The default symbols that you can override during the MONGEN
dialogue.
3. The instructions for writing a new controller routine.
4. The symbol naming conventions.
a. 3-letter prefixes
b. 3-letter suffixes
5. The list of upper and lower limits for various disk
parameters.
6. The data structure description.
7. The core and disk block symbol definitions for the file
system.
GENERATE A MONITOR WITH MONGEN Page 8-30
a. Generalized core blocks
b. Access table
c. BAT blocks
d. Channel Data Blocks
e. Device Data Blocks
f. Home blocks
g. Monitor job tables
h. Controller Data Blocks
i. Logical block numbers within unit
j. Monitor buffer
k. Name block
l. Project-Programmer number block
m. RIB blocks
n. SAT blocks
o. File structure data block
p. System variables
q. UFD blocks
r. Unit data blocks
8.9.3 Description of COMDEV.MAC
COMDEV.MAC contains symbols, code, and data bases associated with
devices (in the following order). The items are listed below.
1. Conversion symbols from the old MONGEN format (XXXXXn) to the
new (M.XXXX).
2. Terminal (TTY) and remote station conversion symbols.
3. MONGEN default symbols.
4. Terminal (TTY) data base.
5. Remote station data base.
6. All other device-specific code and data bases.
7. EXTERNS used to load the proper device routines from the
monitor library.
GENERATE A MONITOR WITH MONGEN Page 8-31
8.10 HDWGEN EXAMPLE
.R MONGEN
MONGEN for 603 monitors
/HELP(PROMPT,SHORT,LONG)]: LONG
Which GEN(HDW,TTY,NET,F)[
HDW to define hardware configuration
TTY to define terminal configuration
NET to define network configuration
F to define software features]: HDW
Output(DSK:<cr>HDWCNF.MAC):<cr>
DECsystem10(1040,1050,1055,1070,1077,1080,1088)[
1040 is small disk system with KA10 cpu
1050 is large disk system with KA10 cpu
1055 is disk system with 2 KA10 cpu's
1070 is disk system with KI10 cpu
1077 is disk system with 2 KI10 cpu's
1080 is disk system with KL10 cpu
1088 is disk system with 2 KL10 cpu's]: 1080<cr>
System name[24 characters or less]: TESTER603<cr>
CPU0 serial #(1-10000): 3333<cr>
Exclude Monitor overhead from user run time(Y,N)[
Overhead is CPU time spent clock queue processing, command
decoding, core shuffling, swapping, and scheduling.
User run time always includes UUO execution and
unless EBOX/MBOX runtime accounting is selected
(KL10 systems only) includes IO interrupt service time.
On KA or KI systems, each CPU must have a DK10]: Y<cr>
EBOX/MBOX runtime accounting?(y,n)[
If EBOX/MBOX runtime accounting is selected in a KL10
based system, user runtime is computed using the KL10
internal accounting clocks]: Y<cr>
Exclude PI time from user runtime?(y,n)[
An answer of "yes" to this question will cause the monitor
to set up the KL10 accounting meters in such a manner that
users will not be charged for cpu time used during interrupts]: Y<cr>
# Data Channels(2,1-8)[DF10s, DF10Cs, DX10s OR RH20'S for disk and tape]: 2<cr>
Channel 0 Type (DF10,DF10C,DX10,RH20): RH20<cr>
On channel # 0:
# RH20S for RP04'S, RP06'S(0-1)[Controllers for RP04, RP06 disk pack
units on channel 0]: 1<cr>
# Units on RPA(1-8): 2<cr>
# RH20's for TM02'S (0-1)[Mass-Bus tape controller for
TU16's and TU45's on channel 0]: 1<cr>
How many TM02's on RH20 # 0 (1-8)[EACH UNIT CAN CONTROL
UP TO 8 SUB-UNITS = TAPE DRIVES]: 2<cr>
How many drives on TM02 0 (1-8): 2<cr>
How many drives on TM02 1 (1-8):2 <cr>
Channel 1 Type (DF10,DF10C,DX10,RH20): DX10<cr>
GENERATE A MONITOR WITH MONGEN Page 8-32
On channel # 1:
# Controllers(0-1)[TX01's or TX02's for TU70 Tape Drives on channel 1]: 1<cr>
# Units on MTB(1-8): 2<cr>
TM10As(0,0-2)[I/O Bus type Controller for NRZI only drives]: 0<cr>
Specify which drives (M-N) are 7 track drives.
[Type one number (M) or one range(M-N) or ALL on separate lines.
Type an extra carriage return when through.]
For controller MTA
<cr>
For controller MTB
<cr>
Specify which tape drives (M-N) are capable of 6250 BPI densities.
[Type one number (M) or one range (M-N) or ALL on separate lines.
Type an extra carriage return when through.]
For controller MTA
<cr>
For controller MTB
<cr>
# DTEs on CPU0(2<cr>,1-4)[BYTE TRANSFER DEVICE USED FOR KL10 to PDP-11 front end communications]: 2<cr>
# Jobs(1-511)[Maximum number attached and detached, not
counting null job]: 80<cr>
Max. PAGES of core For each job(0,0-512)[0 means all of core]: 256<cr>
# K total system core(32-4096)[ONCE-only reports if
less core at startup]: 1056<cr>
Clock ticks per second(60,50)[Power line frequency]: 60<cr>
# Real-time devices(0,0-77)[Max. # which can be put
on PI channels simultaneously]: 0<cr>
Allow jobs to be locked in core(Y,N): Y<cr>
# PAGES min guaranteed among jobs not locked in core(0,0-512)[
minimum free core pool for unlocked jobs, 0 assumes all of core]: 256<cr>
# High priority queues(0,0-15): 2<cr>
CCL commands to stay in core(Y,N)[System programs pass
commands to each other via core(TMPCOR UUO) rather than disk]: Y<cr>
Meter(Y,N)[Performance analysis metering(METER UUO)]: Y<cr>
SYSCHK(Y,N)[Initial Hardware integrity check at ONCE-only time]: N(cr>
MSGSER(Y,N)[Support for device MPX. (more than one device
on an I/O channel). This feature is required for TYPSET-10
and MCS-10]: Y<cr>
PSISER(Y,N)[Advanced programmed software interrupt service -
Support for the PISYS. UUO. This provides an easy
and powerful interrupt method for program to trap asynchronous
events. Required bY<cr> MCS-10]: Y<cr>
IPCF(Y,N)[Inter process communition facility]; Y<cr>
ENQ/DEQ(Y,N)[Sychronization Primitives To-allow
simultaneous file update by multiple co-operating processes]: Y<cr>
GENERATE A MONITOR WITH MONGEN Page 8-33
# CDRs(1,0-2)[Card reader]: 0<cr>
CDP(Y,N)[Card punch]: N
CP10D(Y,N)[Special Systems unbuffered Card Punch]: N<cr>
DIS(Y,N)[Display device(VP10,340,30,VB10C) as distinguished from
display terminals]: N<cr>
# TD10s(1,0-2)[DECtape controls]: 2<cr>
# Units on DTA(1-8): 4<cr>
# Units on DTB(1-8): 2<cr>
LPTs(1,0-3)[Line printers]: 3<cr>
LPT0 Lower case(Y,N)[Does LPT0 have lower case capability]: Y<cr>
LPT1 Lower case(Y,N)[Does LPT1 have lower case capability]: Y<cr>
LPT2 Lower case(Y,N)[Does LPT2 have lower case capability]: N<cr>
PLTS(0,0-2)[Plotters]: 0<cr>
PTP(Y,N)[Paper tape punch]: Y<cr>
PTR(Y,N)[Paper tape reader]: Y<cr>
# PTYs(20,0-510)[Pseudo-terminals - each operator
service program and Batch stream needs one]: 20<cr>
DC44(Y,N)[
The DC44 is a TYPESET-10 front end which supports
the following devices:
PA611R High speed paper tape reader
PA611P High speed paper tape punch
LPC11 Online photocomposition machine]: N<cr>
# of DA28s(0,0-4)[Interprocessor channels for PDP8/11/15]: 0<cr>
# DAS78'S(0,0-8)[IBM 360, 370, and/or 2780 support]: 0<cr>
Decimal "symbol,value"[
For any symbols to be defined.
Type one per line, extra carriage return when through]
<cr>
<cr>
Octal "symbol,value"[
For any symbol to be defined.
Type one per line, extra carriage return when through]
<cr>
<cr>
SIXBIT "symbol,value"[
For any sixbit symbol to be defined.
Type one per line, extra carriage return when through]
<cr>
Type "device-mnemonic,PI-channel" for special devices[
With neither channel AC save routine nor device data block,
the "device-mnemonic" must be 3 characters or less.
GENERATE A MONITOR WITH MONGEN Page 8-34
Type extra carriage return when through.]
<cr>
Type "device-mnemonic,PI-channel,no.-of-devices"[
For special devices with device data blocks.
the "device-mnemonic" must be 3 characters or less.
Type extra carriage return when through.]
<cr>
Type "device-mnemonic,PI-channel,highest-ac-to-save"[
For special devices with channel save routines to save acs up to
the "highest-ac-to-save". "Device" must be 3 char or less.
Type extra carriage return when through.]
<cr>
File DSK:HDWCNF.MAC Closed [HDWGEN finished]
GENERATE A MONITOR WITH MONGEN Page 8-35
8.11 TTYGEN EXAMPLE
Which GEN(HDW,TTY,NET,F)[
HDW to define hardware configuration
TTY to define terminal configuration
NET to define network configuration
F to define software features]: TTY<cr>
Output(DSK:TTYCNF.MAC): <cr>
How many DC10s(1,0-2)[
The DC10 is a data line scanner]: 1<cr>
How many DC68s(1,0-2)[
The DC68 is a PDP-8 680 or 680I communications system]: 1<cr>
How many DC76s(1,0-8)[
The DC76 is a PDP-11 communications system]: 0<cr>
For DC10 0:
# DC10B[ or 632] 8 line data groups(1-8)[
1 is TTY0-7, 2 is TTY0 - 17, ... 8 is TTY0 - 77]: 2<cr>
# DC10E Data set control groups(0-2): 2<cr>
Correspondence of DC10E lines to the DC10B lines(M-N,P)[
Type M,P for one pair and M-N,P for a range of pairs
where M is octal DC10E line, M-N is octal range of DC10E
lines, and P is octal DC10B line]
<cr>
<cr>
FOR DC68 0:
# Octal lines on DC68, including its console TTY(1-144): 20<cr>
OPR octal line #(CTY,0-37)[OPR is privileged operator terminal]:
Answer the following questions about your TTY lines(M-N).
[Type one octal line #(M) or one range(M-N) or CTY on separate
lines. Type extra carriage return when through.]
<cr>
Data set lines[Class of terminal for LOGIN, LOGIN resets line
to computer echoing and no hardware tabs]
<cr>
Lines with hardware tabs[Monitor simulates rest with spaces]
<cr>
Remote lines[Class of terminal for LOGIN, do not confuse
with remote station TTYs]
3-7<cr>
<cr>
Local copy lines[Echoing provided by terminal rather than
by computer. Often (incorrectly) called half duplex]
10-12<cr>
<cr>
Half duplex lines[TWX or half duplex wired scanner(DC10C)]
<cr>
Slaves[No commands may be typed]
GENERATE A MONITOR WITH MONGEN Page 8-36
<cr>
Lines with hardware form feed[Leave out if users
would rather not get form feeds until they do TTY FORM commands]
<cr>
Lines which run INITIA at startup
0-16<cr>
<cr>
Filler class codes(M-N,P)[
Type M,P for one line M with filler class code P or
M-N,P for a range of lines with filler class code P]
<cr>
2741 lines on DC-10 interfaces[]
<cr>
File DSK:TTYCNF.MAC Closed [TTYGEN finished]
8.12 NETGEN EXAMPLE
Which GEN(HDW,TTY,NET,F)[
HDW to define hardware configuration
TTY to define terminal configuration
NET to define network configuration
F to define software features]: NET<cr>
Output(DSK: <cr>NETCNF.MAC): <cr>
Network software(Y,N)[
Software to support remote computers: DECsystem-10's,
PDP-11's, PDP-8's (requires FTNET to be -1)]: Y<cr>
How many nodes do you wish to support(3,0-63)[Maximum]: 3<cr>
How many DC75NP's or DN87's on the system(1,0-8)[Network
front-ends connected to DL10's.]: 1<cr>
For front end number 1:
To which DL10 port is the DC75 or DN87 connected (0,0-7)[]:
How many DN87S's on the system(1,0-3)[Network
front-ends connected to DTE-20's]: 0<cr>
Node number of central site(1-77)[
Unique number identifying DECsystem-10 to network.]: 1<cr>
Name of central site[Six characters or less.]: CENTRAL<cr>
%More than 6 characters.
Name of central site[Six characters or less.]: CENTER<cr>
# of remote TTY's(0,0-512)[
Maximum number of teletypes on network nodes to be
handled at any given time.]: 12<cr>
# of remote CDR's(0,0-63)[
Maximum number of card readers on network nodes to be
handled at any given time.]: 0<cr>
# of remote LPT's(0,0-63)[
Maximum number of line printers on network nodes to be
GENERATE A MONITOR WITH MONGEN Page 8-37
handled at any given time.]: 22<cr>
# of remote PTR's(0,0-0)[
Maximum number of paper tape readers on network nodes to be
handled at any given time.]: 0<cr>
# of remote PTP's(0,0-63)[
Maximum number of paper tape punches on network nodes to be
handled at any given time.]: 0<cr>
# of remote MTA's(0,0-0)[
Maximum number of magnetic tape drives on network nodes to be
handled at any given time.]: 0<cr>
# of remote processes(0,0-128)[
Maximum number of remote processes that can be connected to at
any given time (requires FTTSK to be -1).]: 12<cr>
Remote data entry software(Y,N)[
Software to support remote data entry terminals for MCS-10 applications
(requires FTRDX to be -1)]: Y<cr>
# of connects(46,1-512)[
Maximum number of simultaneous connections.]: 46<cr>
File DSK:NETCNF.MAC Closed [NETGEN finished]
8.13 FGEN EXAMPLE
Which GEN(HDW,TTY,NET,F)[
HDW to define hardware configuration
TTY to define terminal configuration
NET to define network configuration
F to define software features]: F<cr>
Output(DSK: <cr>F.MAC): <cr>
Feature set(TINY,KALUG,KILUG,KAFULL,KIFULL,KLFULL)[
TINY Is minimum subset of features for KA10 timesharing
KALUG Is medium size KA10 monitor with enough features for batch
KILUG Same as KALUG but for KI10 cpu
KAFULL Includes all features of DECsystem10 monitor such
as real-time and extended file system for KA10 cpu
KIFULL Same as KAFUL but for KI10 cpu
KLFULL Same as KIFULL but for KL10 cpu]: KLFULL<cr>
Standard setting(YES,NO,LIST,EXPLAIN)[
Standard values for all feature test switches for your configuration]: YES<cr>
Do you want the virtual memory facility(YES,NO)[VMSER must
be on your distribution tape]: YES<cr>
File DSK:F.MAC Closed [FGEN finished]
Which GEN(HDW,TTY,NET,F)[
HDW to define hardware configuration
TTY to define terminal configuration
NET to define network configuration
F to define software features]:
.
GENERATE A MONITOR WITH MONGEN Page 8-38
8.14 DECIMAL DEFAULT VALUES
This section contains a description of the standard symbols and
default decimal values assumed by the MONGEN program. Section 8.14.1
lists those symbols defined in COMMON; Section 8.14.2 lists those
symbols defined in COMDEV; and Section 8.14.3 lists those symbols
defined in COMCON. All symbols are listed in the order in which they
are defined.
8.14.1 Symbols Defined in COMMON
NSPMEM, 1000
The number of nanoseconds per memory cycle. This symbol is used
to compute the amount of time spent performing core shuffling,
and it is printed by the SYSTAT command. For systems with MB10
memories, the constant should be changed to 1760.
EPL4WD, 17
The number of 4-word blocks in the extended executive push-down
list. When extended it is one-fourth the length of the maximum
executive push-down list. If a push-down list overflow occurs,
this value should be increased.
MONCOR, JOBN*72+EPLLEN (Large Disk Systems)
The monitor reserves a table of at least MINCOR words, or
possibly up to the next 1K boundary, for allocating disk device
data blocks and extended push-down lists. MONCOR is normally
JOBN*90 words, allowing 2.5 open disk files per job. If this
value is too small, a larger value may be supplied. The total
size of the monitor is printed after this space is reserved in
the long ONCE-Only dialogue.
MINCOR, JOBN*55 (Small Disk Systems)
For a small disk system, there are only 1.5 open disk files per
job. Otherwise, the explanation is the same as for large disk
systems above.
CTYDCR, 18
The delay for CTY carriage return during ONCE-Only dialogue. The
default causes an 18-jiffy delay. This allows the monitor to
pause before printing on the next line, allowing the CTY to
execute a carriage return/line feed.
RLDTIM, 112
The time until an auto-reload.
MINMAX, 1024*12
The smallest value allowed for CORMAX after the system begins
execution.
UNIQ1,1 UNIQ2,1 and UNIQ12 ... UNIQ6,1 ... UNIQ16
These represent priority interrupt channels for the exclusive use
of a special device. If the priority interrupt channels are to
be reserved, set the value of UNIQn to 1, where n is the channel
GENERATE A MONITOR WITH MONGEN Page 8-39
number. For example, if a real-time device requires extremely
fast response time, it is necessary to set the BLKI/BLKO pointer
in the channel location. (Channel 7 cannot be reserved, as it is
always used for the clock.)
INDPPN,0
If the value of INDPPN is 0, each programmer number refers to the
same person in every project. If the value of INDPPN is 77777,
programmer numbers may be assigned independently within each
project. This affects only the disk file access protection
mechanism.
M.QSTR,0
The default structure for the QUEUES.
SYSSEG,-1
This function checks files for a flag indicating they are SYS
files. This function also prevents dormant or idle segments from
sitting in core, in addition to achieving fast GETSEGs from SYS:.
CPTOS,0
CPTOS = 0 sets the trap offset for dual processor systems. CPU0
normally traps to location 40, and CPU1 normally traps to
location 140. CPTOS = 0 traps CPU1 to location 40 and CPU0 to
location 140. The value of CPTOS must be set to 0 for
single-processor systems. (KA10 only.)
UFCO, 10
Monitor call fairness count for CPU0 in a multi-processor system.
UFC1, 10
Monitor call fairness count for CPU1 in a multi-processor system.
(CPU1 picks UFC1 jobs that have just completed monitor calls on
CPU0 before picking a job on CPU1.)
M.CBAT, 10
The maximum number of bad memory addresses stored in each CPU
data block on a memory parity error sweep of core.
STDENS,5
The standard magtape density, if the user program does not
override the symbol value with an INIT, OPEN, or SETSTS monitor
call or with the SET DENSITY command. The standard is 6250 bpi
for TU70s.
1 = odd parity + 200 bpi
2 = odd parity + 556 bpi
3 = odd parity + 800 bpi
4 = odd parity + 1600 bpi
5 = odd parity + 6250 bpi
M.BMAX, 13
The initial setting for the maximum number of MPB jobs permitted
to log-in simultaneously. This value initializes the monitor
GETTAB word BATMAX. The operator can change this value during
GENERATE A MONITOR WITH MONGEN Page 8-40
system operation with the OPSER command: SET BATMAX.
M.BMIN, 0
The initial setting for the minimum number of jobs guaranteed for
MPB. This value initializes the monitor GETTAB word BATMIN. The
operator can change this value during system operation with the
OPSER command :SET BATMIN.
JIFSEC.60
The number of clock ticks per second, i.e., jiffies.
DDTRY,4
The number of times the monitor is to try to recover from DECtape
errors.
M.EXE, 1
The symbol which determines the type of file to be created on the
execution of the SAVE command. If M.EXE = 0, normal (i.e., SAV,
HGH, LOW, SHR) files are created; if = 1, .EXE files are
created.
TXTRA,0
The number of extra terminal disk data blocks for slave lines.
LOGSIZE,12
The minimum amount of virtual core required for a user to be
logged-in. This value must be at least as large as the LOGIN
system program, which is currently 12K. If the amount of core is
insufficient, the user will receive a CORE UNAVAILABLE error
message, which includes the virtual amount of core remaining.
M.JMAX, M.JOB
The initial setting for the maximum number of jobs permitted to
LOGIN simultaneously. (This includes all classes of jobs:
batch, interactive, or operator service jobs. Note that operator
service jobs, subjobs under [1,2], or subjobs at OPR or CTY are
not restricted from logging-in even if LOGMAX is exceeded,
because the task may be a critical system function.)
This value, which cannot be less than 1, initializes the monitor
GETTAB word LOGMAX. The operator can change this value during
system operation with the OPSER command :SET LOGMAX.
M.CMCT, 32
The number of times the core tables (CHKTAL) are to be checked
before performing its defined functions, which requires
approximately 5 ms of overhead. This reduces CPU time for CORE,
GETSEG, and RUN monitor calls. The maximum value to which this
symbol may be set is 262,000.
M.XFFA, 0
If M.XFFA is non-zero, the File Daemon will not be called if the
program accessing the file is being run under [1,2] or has the
JACCT bit set.
GENERATE A MONITOR WITH MONGEN Page 8-41
8.14.2 Symbols Defined in COMDEV
MTSIZ, 128
The size of magnetic tape records in 36-bit words (i.e., data
words in the buffer). Users can override this value by using the
SET BLOCK SIZE MTAn: command. Programs can override this value
by building special buffers.
LPTSIZ, 29
The size of the line printer buffer + 2. The value of 29 allows
a full line to be typed on a 132-column line printer in one
buffer. This value should be changed to 26 when using 120-column
line printers.
FLCDEF, 1
The filler class code to use if one is not specified for a TTY
line.
MTELOG, -1
If = -1, errors that require more than one retry are logged. If
= 0, all recoverable errors are logged.
MTDAEM, 10
The number of times per reel to call DAEMON for magtape error
reporting.
8.14.3 Symbols Defined in COMMOD
STRMAX, 14
The maximum number of disk file structures that can be on line
simultaneously. By decreasing this value, one word will be saved
for each decrement of 1. This value is automatically set to 1
for 1040 systems; the value cannot exceed 14.
CCWMAX,10
The maximum length in words of each disk channel command list.
Altering this value will affect efficiency and space. The value
is set to 10 for 1040 systems. CCWMAX is not used for KI10 and
KL10 systems.
SWPMAX,8
The maximum number of disk units that may be used for swapping (1
word per unit). The value is set to 1 for 1040 systems.
SWCLSN,7
The highest class number for swapping; this value cannot exceed
7.
DSKTRY,10
The number of times to try on disk data errors for all file
structures before recalibrating (inclusive of the initial try).
GENERATE A MONITOR WITH MONGEN Page 8-42
SERTRY,3
The number of times to try on search errors before recalibrating
(inclusive of the initial try).
RCLTRY,10
The number of times to recalibrate and try on DSKTRY disk data
errors or on SERTRY search errors (inclusive of the initial try).
The total number of disk tries with recalibration is equal to
DSKTRY * RCLTRY (i.e., 10 * 10 = 100). The total number of
search tries with recalibration is equal to SERTRY *RCLTRY (i.e.,
10 * 3 = 30).
CHVIFP,10
The standard initial fairness count for positioning. The monitor
chooses the nearest request for positioning (CHVFP times -1)
before taking the longest waiting request.
CHVIFT,10
The standard initial fairness count for transfers. The monitor
chooses the shortest latency for a transfer (CHVIFT times -1)
before taking the longest waiting request.
PTRLEN,6
The number of in-core retrieval pointers per file length; this
offsets the size of the device data block.
FIL4WD,9
The number of 8-word blocks allocated per job in a common pool of
monitor free core. The value of FIL4WD should be increased if
the ENTER monitor call's error code 16 is received regularly. To
assign this space the ONCE-Only code multiplies this factor by
the number of jobs. These blocks are used by the Level-D Disk
Service for active, dormant, ACC, NMB, PPB, and UFB blocks. This
pool is permanently reserved for these blocks and is not used for
any other purpose. Another pool is used for variable length core
blocks, such as disk service data blocks and extended exec
push-down lists. The system sets the minimum number of 4-word
core blocks to 50 with 5 or fewer jobs. The value is set to 5
for 1040 systems.
UNVRSF,500
The reciprocal factor of the total disk size. The amount of disk
space is subtracted from the number of free blocks when the
monitor is started up. This area is not available for users'
data. It is a safety factor that ensures there is room to write
the second RIB, etc. The standard reserved for this purpose is
one five-hundredth of the disk space.
MBFN,2
The number of 128-word monitor buffers used for the reading and
writing of non-user data. This value must be at least 2. The
default for MBFN is the number of jobs divided by ten, plus one.
For example, the default for a 50 job system is 6.
LBNHOM,1 LB2HOM,10
GENERATE A MONITOR WITH MONGEN Page 8-43
The standard logical block number on each unit that contains the
HOME block. Note that it is also possible to change this value
for a specific disk unit as follows:
xxxnHM,LBNHOM
xxxnH2,LB2HOM
Where:xxxn is the unit designation, e.g., DPA0HM, 3.
MFDSIZ,8
The number of blocks allocated to the MFD by the refresher in
each file structure. The MFD can be longer than this but there
is a speed advantage to consecutive blocks. Increase this value
if the MFD is normally longer than 8 blocks.
LIMLVL,0
The maximum number of nested SFDs for this monitor. The maximum
value for LIMLVL is 5.
MAXUSI,8
The maximum negative USETI for reading an extended RIB.
8.15 OCTAL DEFAULT VALUES
This section contains a description of the standard symbols and
default octal values assumed by the MONGEN program. Section 8.15.1
lists those symbols defined in COMMON; and Section 8.15.2 lists those
symbols defined in COMMOD. All symbols are listed in the order in
which they are defined.
8.15.1 Symbols Defined in COMMON
A00CVN, 0
The customer version number.
A00MVN, 603
The major version number.
XPANDN,0
The value affects the listing of COMMON. If it is non-zero,
COMMON MACRO expansions are listed.
APRSN,0
The serial number of the arithmetic processor.
DEFDEB,0
Defines the stop code conditions under which the monitor is
automatically reloaded. The default value of 0 indicates that
the monitor is reloaded with the STOP category of stop codes. A
value of 100000000000 indicates that the monitor is reloaded with
JOB stop codes; a value of 200000000000 indicates that the
monitor is reloaded with DEBUG stop codes; and a value of
GENERATE A MONITOR WITH MONGEN Page 8-44
300000000000 indicates that the monitor is reloaded on all types
of stop codes. If the value is 400000000000, auto-reload is
disabled. If the value is 200000000000, the system stops if the
CPU halts. (Refer to the specification STOPCD.RNO.)
M.WCH, JW.WMT
The WATCH bits used to set the initial WATCHing. Because LOGIN
sets JBTWCH, this is useful only for lines that do not need
LOGIN, that is those using INITIA.
bit 19 = 200000 = Time of day started to wait
bit 20 = 100000 = RUN time
bit 21 = 40000 = WAIT time
bit 22 = 20000 = Blocks read
bit 23 = 10000 = Blocks written
bit 24 = 4000 = Version numbers
bit 25 = 2000 = MTA statistics
8.15.2 Symbols Defined in COMMOD
PRVFIL,057
The standard file protection code.
PRVUFD,775
The standard User File Directory (UFD) protection code.
SYRDPR,355000
This value is set by the refresher and it indicates
non-privileged files on SYS: (HOME.SYS, SAT.SYS).
SYNRPR,357000
SYNRPR = 357000 indicates that privileged files are on SYS:
(CRASH.EXE, SNAP.SYS, RECOV.SYS, BADBLK.SYS, SWAP.SYS). Used by
the refresher.
MFDPRV,555000
The standard Master File Directory (MFD) protection code. All
users can READ, LOOKUP, but not CREATE files. Set by the
refresher.
SYSPRV,775000
The SYS User File Directory (UFD) protection code; the
project-programmer number is 1,4. Project 1 members can READ,
CREATE, LOOKUP files; others can READ and LOOKUP only. Set by
the refresher.
BLKMAX,100000
The maximum number of blocks that can be transferred in one disk
file operation (IOWD). This default value is so high that the
effective limit is one cylinder boundary.
PRVSPL,077
The default protection code for spooled files.
GENERATE A MONITOR WITH MONGEN Page 8-45
PRVSYS,155
The default protection code for most files on SYS: (except files
with the .SYS extension).
PRYSYS,157
The default protection code for .SYS file on SYS:.
8.15.3 Symbols Defined in COMDEV
MTDLPT,1
Standard magtape labels are ASCII.
8.16 SIXBIT DEFAULT VALUES
The following list is a description of the standard symbols and
default SIXBIT values assumed by the MONGEN program. (Refer to the
description of the DESTROY startup option in the HDWGEN section.)
M.QSTR,0
Force queues to be placed on a specific STR.
M.SF00,DSKA
The name of the first file structure to contain fixed head disk
units (e.g., FHA, FHB, etc.) when the DESTROY startup option is
invoked. The value (DSKA) may be any 1- to 4-character
alphanumeric name beginning with a letter. The number of fixed
head units in the structure is set by the associated decimal
symbol M.ZF00. Note that the DEC-distributed software sets the
name of the first file structure, which contains all fixed head
units, to DSKA.
M.SFxx,aaaa
The name of the second through sixteenth structure containing
fixed head disk units (where xx =01....15, and aaaa is any
4-character alphanumeric name beginning with a letter). The
number of fixed head units in the structure is set by the
associated decimal symbol M.ZFxx. Note that the DEC-distributed
software does not define these symbols, because it assumes that
all fixed head units are incorporated into one structure, called
DSKA.
M.SD00,DSKB
The name of the first file structure to contain disk pack units
(e.g., DPA, DPB, etc.) when the DESTROY startup option is invoked
by the operator. The value (DSKB) may be any 1- to 4-character
alphanumeric name beginning with a letter. The number of disk
pack units is set by the decimal symbol M.ZD00. Note that the
DEC-distributed software sets the name of the first file
structure, which contains all disk packs, to DSKB.
M.SDxx,aaaa
The name of the second through sixteenth structure containing
disk pack units (where xx =01....15, and aaaa is any 4-character
alphanumeric name beginning with a letter). The number of disk
pack units in the structure is set by the associated decimal
GENERATE A MONITOR WITH MONGEN Page 8-46
symbol M.ZFxx. Note that the DEC-distributed software does not
define these symbols because it assumes that all disk packs are
incorporated into one file structure called DSKB.
M.SSxx,aaaa
The name of the first file structure to contain RH10s/RS04s when
the DESTROY startup option is invoked by the operator. The
number of fixed head units in the structure is set by the
associated decimal symbol M.ZS00.
M.SRxx,aaaa
The name of the first file structure to contain RH10s/RP04s when
the DESTROY startup option is invoked by the operator. The
number of fixed head units in the structure is set by the
associated decimal symbol M.ZR00.
8.17 NON-STANDARD DEVICE PI ASSIGNMENT
Under ordinary circumstances when COMMON is assembled, devices are
assigned to PI channels according to their group priority. (Refer to
Table 8-1.) If you have at your installation a device not listed as a
standard device in Table 9-1 and you have written your own Monitor
Device Service Routine, you must specify the device mnemonic (in 3
characters or less) and designate an appropriate priority interrupt
channel. You must answer all three questions as they apply to your
configuration. The first question
TYPE "DEVICE-MNEMONIC,PI-CHANNEL" FOR SPECIAL DEVICES
requests special device service routines that do not need either a
Channel Save Routine or a Device Data Block. The second question
TYPE "DEVICE-MNEMONIC,PI-CHANNEL, NO.-OF-DEVICES"
requests devices with special service routines that have a Device Data
Block but no Channel Save Routine. The third question
TYPE "DEVICE-MNEMONIC,PI-CHANNEL, HIGHEST-AC-TO-SAVE"
requests devices with special service routines that have a Channel
Save Routine, but no Device Data Block.
Special devices that you added during the HDWGEN dialogue are chained
to the requested channel. To give a device the exclusive use of a
channel, you respond to the "symbol,value" question with
UNIQn,1
where n is the priority interrupt channel to be reserved. (Refer to
the UNIQn,1 entry in Section 8.14.1.)
One or more priority interrupt channels may be reserved for real-time
devices with the RTTRP monitor call. These devices are completely
controlled by user programs and have no specific code loaded with the
monitor. To reserve a priority interrupt channel for use with RTTRP,
you should respond to the "symbol,value" question with
RTCHn,1
where n is the priority interrupt channel to be reserved.
GENERATE A MONITOR WITH MONGEN Page 8-47
(Refer to the RTCHn,1 entry in Section 8.14.1 and to the DECsystem-10
Monitor Calls manual.)
I/O devices are grouped by their relative interrupt speeds. If any
device of a particular group is present, a PI channel is assigned to
that device according to its group priority. Group priorities for
standard devices may be revised by rearranging the devices in INTTAB,
which is in the COMMON source file.
Table 8-1
Device Groups for PI Channel Assignment
Device Mnemonic Group Name
DCB A 136 data control for 270 disk
MTA B TM10n magtape data channel
MTB
DCT C 136 data control for 551 or 516 tape
controls
DTA D TD10 DECtape data channel
DTB
CDR E 461 or CR10 card reader
APR E KI10, KA10, or 166 arithmetic
processor
RTC E DK10 real-time clock (each CPU)
SCN F Terminal scanner
PTR F Paper-tape reader
LPT F Line printer
DTA F DECtape flag channel
MTA F Magtape flag channel
MTB
CTY F Console terminal
NET F Remote devices
DSK G Disk flag channel
PEN G Light pen
PTP G Paper-tape punch
CDP G Card punch
PLT G Plotter
DIS H Display data channel
CLK I Scheduler, clock routines (always
assigned to channel 7)
GENERATE A MONITOR WITH MONGEN Page 8-48
8.18 TERMINAL INTERFACE LINE NUMBERS
The DC10 data line scanner is capable of handling up to a total of 64
lines (physical ports). The lines are monitored by scanners of two
types -- a DC10B or a DC10E. Each DC10B is designed to handle eight
lines of terminal communications, and each DC10E is designed to handle
eight lines of dataset control. When using a terminal through a
dataset, both a DC10B and a DC10E line are required. The DC10E line
interfaces the dataset control signals to system decipherable signals,
and the DC10B line carries the teletype information.
When the monitor assigns terminal numbers to DC10 lines, only the
DC10B lines receive numbers. The reason for this action is that each
DC10E line must be paired with a DC10B line to function properly,
making an additional number unnecessary.
Each installation's system administrator must determine which DC10E
dataset controllers are wired to which DC10B lines. (Refer to Figure
8-2, where line numbers 10 through 17 are DC10B dataset lines that are
hardwired to the DC10E controller lines 20 through 27.)
The sequence of line number assignments for DC76s is:
DC10 0
DC10 1
DC68 0
DC68 1
DC76 0
DC76 1
Remote lines
CTYs
PTYs
GENERATE A MONITOR WITH MONGEN Page 8-49
Figure 8-1 One DC10 with Two DC10Bs
Figure 8-2 One DC10 with One DC10B and One DC10E
GENERATE A MONITOR WITH MONGEN Page 8-50
Figure 8-3 One DC76 with 33 Lines
Figure 8-4 One DC10 with Two DC10Bs and One DC68 with Eight Lines
GENERATE A MONITOR WITH MONGEN Page 8-51
8.19 MONGEN ERROR MESSAGES
?ANSWER MUST BE ONE OR (ANS0, ANS1,...)
Your response to this question was not a legal response. Choose
one of the responses within parentheses.
?ANSWER WITH ONE OF THE CHOICES WITHIN PARENS
You responded to a multiple choice question with an answer that
was not one of the choices. Reanswer the question with one of
the supplied choices typed within parentheses after the question.
?CANNOT BE SAME AS CPUn
When you responded with the serial number of a CPU, your response
matched the response for another CPU. Retype your response and
press the carriage return.
?DEVICE n NOT AVAILABLE
The device you requested is not available at this time. Select
another device or wait until the desired device is ready.
?HIGHEST-AC-TO-SAVE MUST BE N OR LESS
You specified an illegal AC number in response to the question.
Reanswer the question and press the RETURN key.
%MORE THAN 6 CHARACTERS
In answer to question number 9 within the Networks Configuration
dialogue, you responded with a central site name longer than 6
characters. Reanswer the question with a central site name
having fewer than 6 characters.
%MORE THAN 24 CHARACTERS
In answer to question number 4 within the Hardware Configuration
dialogue, you responded with a system name longer than 24
characters. Reanswer the question with a system name having
fewer than 24 characters.
?MUST BE IN RANGE MIN-MAX
You specified an answer that was not within the legal response
range of min to max. Reanswer the question with a value that is
within the legal range.
?NO DEFAULT RANGE ALLOWED, TYPE A NUMBER
In response to a question, you pressed the RETURN key. There is
no default answer for this question. Reanswer the question by
typing a response before pressing the RETURN key.
?NO DEFAULT VALUE ALLOWED
In answer to a question, you pressed the RETURN key. This
question, however, does not have a default answer. Type an
answer in response to the question before you press the RETURN
key.
?NOT AVAILABLE IN x MONITOR
GENERATE A MONITOR WITH MONGEN Page 8-52
A feature was specified that is not available in the monitor you
are generating. Check the response MONGEN typed after you typed
R MONGEN; make sure that you are using the appropriate version
of MONGEN.
?NOT DEV:FILE.EXT P,PN
Your response was not in the legal file specification format;
retype your response and press the RETURN key.
?NOT DEVICE-MNEMONIC, PI-CHANNEL
Your response to the question was not in the desired format.
Retype your response in the correct format.
?NOT DEVICE-MNEMONIC, PI-CHANNEL, HIGHEST-AC-TO-SAVE
Your response to the question was not in the desired format.
Reanswer the question in the correct format.
?NOT DEVICE-MNEMONIC, PI-CHANNEL, NO.-OF-DEVICES
Your response to the question was not in the desired format.
Reanswer the question in the correct format.
?NOT "SWITCH.VALUE" OR KEYWORD
Your response to the question was not legal. Reanswer the
question and press the RETURN key.
?NOT "SYMBOL.VALUE"
Your response to the question was not in the legal format;
reanswer the question and press the RETURN key.
?TYPE ENOUGH TO UNIQUELY DISTINGUISH ANSWER
You abbreviated an answer to a question. The abbreviation was
not long enough to distinguish it from another possible answer.
Reanswer the question without abbreviating the answer.
?TYPE OCTAL LINE NUMBER OR CTY OR CTY1 FOLLOWED BY A CARRIAGE RETURN
Your response to the question was not what MONGEN expected.
Reanswer the question with a TTY line number, the word CTY, or
the word CTY1.
?TYPE OCTAL M-N or M
Type a range of numbers or a single number in response to this
question.
?TYPE OCTAL M-N,P OR M,P
Your response to the question was not in a legal format;
reanswer the question with a range of values before pressing the
RETURN key.
?TYPE ONE ANSWER FOLLOWED BY A CARRIAGE RETURN
You supplied more than one answer to the question. Type one
answer and press the RETURN key.
?TYPE SINGLE NUMBER FOLLOWED BY A CARRIAGE RETURN
GENERATE A MONITOR WITH MONGEN Page 8-53
In response to the question you supplied more than one digit
number, when only one digit is legal. Retype your response and
press the RETURN key.
?UNKNOWN SWITCH
You have specified a switch that is unknown to MONGEN. Refer to
the list of legal switches. Type one of them before pressing the
RETURN key.
CHAPTER 9
ASSEMBLE MACRO FILES
It is always necessary that you assemble F, S, HDWCNF, TTYCNF, NETCNF,
COMMON, COMDEV and COMMOD (and NETPRM and COMNET, if you have
purchased network software) because they are configuration-dependent.
The following command string can be used to assemble these files.
For KL10 systems only, type the following:
.COMPILE DTEPRM
For all systems, type the following:
.COMPILE F/COMP,S/COMP,HDWCNF+NETCNF+TTYCNF+<COMMON,COMDEV,COMMOD>
If you responded Y to the MONGEN question DO YOU WANT NETWORK
SOFTWARE, type the following command line instead of the one above.
.COMPILE F/COMP,S/COMP,NETPRM/COMP,HDWCNF+NETCNF+TTYCNF+<COMMON,
COMDEV,COMNET,COMMOD>
It is important that F.MAC precede S.MAC because S has conditional
assemblies depending on the contents of F.
If you are using a standard combination of feature test switch
settings and DEC-supplied sources, you do not need to assemble the
remainder of the bundled monitor modules because they have already
been assembled and combined in a TOP?10.REL file (see tables on
following pages).
If you are not using a standard combination of feature test switch
settings or you have SOUPed in your own changes, you must assemble all
the bundled monitor sources. The simplest way to do this is to
compile indirectly the appropriate .CMD file (see tables on following
pages). For example, to assemble all the bundled sources for a
standard KI monitor, type:
.COMPILE @CONKI
You must assemble any unbundled software. To assemble monitors with
the virtual memory option, substitute TOPV10.REL (KI10) or TOPW10.REL
(KL10) for TOPI10.REL in the command line.
To assemble monitors with the DC44 option, issue the following
command:
.COMPILE F, S, TYPSER
And then link your monitor in the prescribed fashion.
To assemble monitors with the DAS78 option, issue the following
ASSEMBLE MACRO FILES Page 9-2
command line:
.COMPILE F, S, D78INT
Then, link your monitor in the prescribed fashion. For installation
instructions concerning the -11 portion of the DAS78 and D78SPL,
carefully read D78MNT.RNO, DAS78.RNO, and DAS78.RND found on the
unbundled distribution tape.
To assemble monitors with the XTCSER options, issue the following
command:
.COMPILE F, S, XTCSER
Then, link your monitor in the prescribed fashion.
If you also want to create cross-reference (CREF) listings, assemble
all the sources for your monitor with the CREF switch, i.e., /C. The
simplest way to do this is to assemble the appropriate command files
indirectly to produce .REL files and CREF input files. Then, CREF can
generate the listings. (See the example below.)
The table below lists the different MACRO indirect command files for
both bundled and unbundled software. The intermediate .REL file
extensions are the extensions specified for the .REL files in the .CMD
files.
CONFIGURATION BUNDLED INTERMEDIATE
SOFTWARE .REL FILE
FILENAME EXTENSIONS
KA-LUG CONLUG.CMD *.RLG
KA-Standard CONKA.CMD *.RLA
KI-LUG CONLGI.CMD *.RLH
KI-Standard CONKI.CMD *.RLI
KI-Virtual Memory CONVI.CMD *.RLV
KL-Virtual Memory CONVL.CMD *.RLW
After assembling the monitor sources, you may need to combine certain
modules into a TOP?10.REL file. If you are not using a TOP?10.REL
file from the distribution tape, because you have used nonstandard
feature test switch settings or you have modified the DEC sources, you
must create your own TOP?10.REL file.
To combine the assembled monitor modules and make TOPI10.REL for a KI
system type:
.R PIP
*CMBKI@
.
Note that indirect PIP files exit directly to monitor level
(. prompt). Therefore, you must specify R PIP or START after
specifying an indirect command file name.
To produce the proper TOP?10.REL for your configuration, use the
appropriate PIP indirect command files from the list below.
CONFIGURATION SOFTWARE .REL
ASSEMBLE MACRO FILES Page 9-3
FILENAME CREATED
KA-LUG CMBLUG.CCL TOPG10.REL
KA-Standard CMBKA.CCL TOPA10.REL
KI-LUG CMBLGI.CCL TOPH10.REL
KI-Standard CMBKI.CCL TOPI10.REL
KI-Virtual Memory CMBVI.CCL TOPV10.REL
KL-Virtual Memory CMBVL.CCL TOPW10.REL
EXAMPLE: For a standard KI system you would assemble the source
files, produce a complete set of CREF files (cross referenced
listings) on magtape, produce TOPI10.REL and a listing of the
monitor's global symbols as follows:
.COMPILE/C@CONKI
.ASSIGN MTA LPT
.CREF
.R PIP
*CMBKI@
.
Save your terminal output from the CREF command so you will know the
order of the files on the magtape.
Whenever you want to print the entire set of monitor source files,
mount the tape created in the previous step and type:
.R PIP
*MTAn:(MW)=
*LPT:=MTAn:*.*
If you want to print a single file from the magtape, advance the tape
to the proper position and PIP the file to the line printer. Use your
terminal output that you saved from the CREF command to determine
where the desired file is on the tape.
Then to get a listing of the monitor global symbols, you must run
GLOB. To produce a GLOB listing for a standard KI monitor with
network software and some unbundled software type:
.R GLOB
*COMMON,COMDEV,COMNET,COMMOD
*DSK:TOPI10.GLB=TOPI10$
where $ is an ESCAPE or ALTMODE.
If you assembled the sources with CONKI.CMD, add the extension .RLI to
COMMON,COMDEV,COMNET, and COMMOD in the above command string.
CHAPTER 10
LOAD AND SAVE THE MONITOR
Whether or not you can load the monitor under timesharing using
LINK-10 depends on how much user core is available. In general, if
your installation has about 50K of user core, you should be able to
load the new monitor using LINK-10. If you have significantly less
user core than this, consult your software specialist for details on
loading the monitor.
The following sequence of instructions loads and saves a monitor for a
standard configuration with no unbundled software:
.R LINK
*/NOINITIAL /HASH:6000 YURMON/SAV,YURMON/MAP = /LOCALS -
#/MAXCOR:60K -
#,COMMON,COMDEV,COMMOD,TOP?10 /SEARCH /GO
The filename YURMON is an arbitrary name chosen for the monitor; you
may use any file name. Specify the TOP?10 to correspond to your
system configuration:
TOPG10 for KALUG
TOPA10 for KA standard
TOPH10 for KILUG
TOPI10 for KI standard
TOPV10 for KI-virtual memory
TOPW10 for KL-virtual memory
Note that COMMON, COMDEV, and COMMOD in this example were generated by
the COMPILE commands at the beginning of Chapter 9, so their
extensions are .REL. If an indirect compile was done on a .CMD file,
the extension created by the .CMD file would have to be specified.
For example, if .COM @CONKI was used, you would have to specify
COMMON.RLI, COMDEV.RLI, and COMMOD.RLI.
The TOPI10.REL (or the appropriate TOP?10.REL) file can be the one
from the monitor distribution tape or one created by assembling all
the monitor sources with the proper .CMD file (see Chapter 9) and
combining them with the proper .CCL file (necessary if you are using a
nonstandard combination of feature test switch settings or you have
SOUPed in your own changes). (See Chapter 8.)
If your monitor includes one or more unbundled options (other than
virtual memory), restore the unbundled monitor modules using the
following procedure. Repeat the procedure for each unbundled tape.
.R BACKUP
/TAPE MTxn ;the tape drive with
;the unbundled tape
LOAD AND SAVE THE MONITOR Page 10-2
;on it.
/RESTORE filename ;for example, D78INT.MAC
! ;BACKUP's indication that it
;is busy
"DONE
;repeat above, beginning with
;/TAPE for each unbundled tape
^C ;CTRL/C when you are finished.
Then, you should compile the files on the unbundled tapes. For
example, to compile the D78INT and TSKSER options, you would use the
following command line:
.COMPILE F, S, D78INT, TSKSER
MACRO:F
MACRO:S
MACRO:D78INT
MACRO:TSKSER
Lastly, you should type the following command line to load the monitor
with unbundled software.
.R LINK
*/NOINITIAL/HASH:6000 YURMON/SAV,-
#YURMON/MAP = /LOCALS/MAXCOR:60K-
#COMMON, COMDEV, COMMOD, unbmod,...unbmod-
#TOP?10/SEARCH/GO
Where: unbmod can be COMNET,
TSKSER, TYPSER, XTCSER, and/or
D78INT.
The files included in each of the unbundled software packages are
listed below:
Be sure that the unbundled package name (e.g., XTCSER, TYPSER, or
D78INT) precedes TOP?10. This must be done so that .SYSINI, ONCE,
etc. will be loaded last and, therefore, can be discarded after
initialization.
Using the commands in the above examples, LINK will produce a load map
named YURMON.MAP. The monitor save file will be YURMON.SAVE or
YURMON.HGH and YURMON.LOW. These files must be converted to .EXE
files before they can be loaded with BOOTS. This can be accomplished
by using the FILEX program or the monitor commands GET followed by
NSAVE.
Note that LINK-10 allows you to load and save a program in the same
command string. Also, certain switch settings can speed up the
loading process. MAXCOR:60K is about the minimum core necessary for
loading the monitor; if no parameter is set, however, LINK-10 will
expand to use up to one half the user core. Using a HASH size of 6000
is an upper limit which should work for any system. For the sake of
economy, if you have a smaller system, you should try smaller HASH
values. LINK-10 will work without specifying either of these
parameters; however, they may speed up the loading process if used.
For a complete definition of all switches refer to the LINK-10 manual.
Then, if you have built a two-segment monitor, you must convert your
monitor .HGH and .LOW files to an .EXE file (and your system must
support .EXE files). Type:
.R FILEX
LOAD AND SAVE THE MONITOR Page 10-3
*YURMON.EXE=YURMON.HGH
If you choose to have .EXE file support with a single segment monitor,
type:
.R FILEX
*YURMON.EXE=YURMON.SAV
Virtual Memory
VMSER.MAC The 6.03 monitor source file that implements virtual
memory.
TOPV10.REL The combined .REL files for KI10 virtual memory
monitors.
TOPW10.REL The combined .REL files for KL10 virtual memory
monitors.
CONVI.CMD The command files needed for assembling KI10 virtual
memory monitors.
CMBVI.CCL The command files needed to create the combined .REL
files for KI10 virtual memory systems.
CONVL.CMD The command files needed for assembling KL10 virtual
memory monitors.
CMBVL.CCL The command files needed for creating the combined
KL10 virtual memory .REL files.
MAKPFH.MAC The page fault handler for KI10 and KL10 virtual
MAKPFH.EXE memory monitors.
PFH.VMX
DDT.VMX The debugger to be used with virtual memory systems.
GET.* A user-mode program to .GET virtual memory
GET2.RNO executable programs.
DC44 Option
TYPSER.MAC The 6.03 monitor source file needed for the DC44
interface.
DAS78 Option
The following file is a part of the monitor code.
D78INT.MAC The 6.03 source file for the DAS70 interface.
The following files are part of the PDP-11 code.
DAS78.CTL The control fle needed to assemble the PDP-11 code.
DAS78.P11 The source files for the DAS78 PDP-11 code.
DAS78.BIN The executable DAS78 PDP-11 code.
DAS78.EX The DDT11 file for the PDP-11 code.
The following files are part of QUEUE and QMANGR, for use with
DAS78SPL.
LOAD AND SAVE THE MONITOR Page 10-4
D78SPL.CTL The control file to assemble D78SPL.
D78SPL.MAC The source code for D78SPL.
D78SPL.EXE The executable code for D78SPL.
The following files are used for DAS78 the modified LPTSPL. This
program is used for 2780s, remote batch stations connected to the
DECsystem-10. This modified LPTSPL can also be used to print on the
local line printer.
SPOOL.CTL The control file for assembling a DAS78 LPTSPL.
SPOOL.D78 The source code for DAS78 modified SPOOL.
LPTSPL.EXE The executable code.
The following files support the DAS78 modified SPRINT.
This program is for use when using one 2780 as a Remote Batch Station
to the DECsystem-10. This modified SPRINT will also be used to read
cards from the local card reader.
SPRINT.CTL The control file for assembling a DAS78 SPRINT.
SPRINT.D78 The source code for DAS78 modified SPRINT.
SPRINT.EXE The executable code.
The following files support the PDP-11 Debugging program.
DDT11.MAC Source for DDT11.
DDT11.EXE Executable code.
DDT11.RNO DDT11 document.
The following files support MODEM diagnostics.
DQMODM.HLP A short document on how to use DQMODM.
DQMODM.BIN The executable code.
The following are DOCUMENTS.
DAS78.RNO The DAS78 "Cookbook" - WARNING: Preliminary
version.
DAS78.RND
D78MNT.RNO Installation hints and other useful information.
The following are miscellaneous files contained in the DAS78 option.
C.MAC Needed for assembling QUEUE.
SCNMAC.MAC Needed for assembling QUEUE.
HELPER.REL Needed for LPTSPL, SPRINT, and QUEUE.
QUEUER.REL Needed for LPTSPL, SPRINT, and QUEUE.
SCAN.REL Needed for QUEUE (V4).
LOAD AND SAVE THE MONITOR Page 10-5
WILD.REL Needed for QUEUE (V5).
Note: The MACDLX assembler is required to assemble DAS78.P11, and the
BOOT11 bootstrap is required to load code from the -10 into the -11
memory. Both these programs will be found on the main (bundled) 6.03
monitor distribution tape.
XTCSER Option
The following file is the XTCSER option.
XTCSER.MAC The 6.03 monitor source code for the DA28 driver.
CHAPTER 11
MAKE A COPY OF THE NEW MONITOR
This step is a precaution against losing the newly built monitor. If
the monitor is accidentally destroyed, you need not rebuild it; you
will have a copy in READIN mode available on tape.
Copy BOOTM, BACKUP.EXE, and your newly built monitor to magnetic tape.
You must copy the files at 556 bpi for READIN mode on TM10 controllers
and at 800 or 1600 bpi for TU70s. The entire tape must be written at
one density.
The following steps create a backup monitor tape for a TM10
controller. First, be sure to read the BOOTM description in Chapter
18 on how to create BOOTM.REL and BOOTM.RDI.
.AS MTA0: BACKUP
.SET DENSITY MTA0:556
.SET BLOCKSIZE MTA0:2000
.REWIND MTA0:
.COPY MTA0:=BOOTM.REL
.SET BLOCKSIZE MTA0:128
.COPY MTA0:=BACKUP.EXE
Then, BACKUP your disk area that contains your new monitor by typing:
.R BACKUP
/DENSITY 556
/SAVE DSKB:[ppn]=DSKB:[ppn]*.*
The following steps create a backup monitor tape for TU70s:
.AS MTA0: BACKUP
.REWIND MTA0:
.COPY MTA0:=BOOTM.RDI
.COPY MTA0:=BACKUP.EXE
Then, BACKUP your disk area that contains your new monitor by typing:
.R BACKUP
/SAVE[1,2]*.*=[1,2]*.*
CHAPTER 12
COMMUNICATION SYSTEMS
Prepare for Communications Systems and Remote Stations
1. DC71 - Instructions for installing and usng a DC71
Communications System are in the Remote Station User's Guide
in the DECsystem-10 Software Notebooks.
2. DC72 - Instructions for installing and using a DC72
Communications System are in the Remote Station User's Guide
in the DECsystem-10 Software Notebooks.
3. DC76 - Instructions for installing a DC76 Communications
System are in the Operators Guide in the DECsystem-10
Software Notebooks.
4. Installations that plan to install remote station software
for the first time should contact their software specialist.
CHAPTER 13
UPDATE ACCOUNTING FILES
Under the phased installation concept, if you are running a current
monitor, you have already converted the accounting files ACCT.SYS and
AUXACC.SYS to 6.02 format using the latest version of REACT. If you
have not yet converted the accounting files, LOGIN under [1,2] and
convert to the new format at this time.
If you have no previously running monitor, you must generate the
initial accounting files or augment the Digital Equipment
Corporation-supplied accounting files using REACT.
1. ACCT.SYS - passwords and privileges for all users.
To convert to the new format, simply read the old format
ACCT.SYS with the R command and write it back with the W
command. Be sure the latest version of REACT is on SYS:
Type:
.R REACT
Optional: For help in running REACT type:
*/H
then type:
*R
*W DSKA:[1,4]
CAUTION
Make sure that you are using the latest version of
ACCT.SYS. If you have a copy of ACCT.SYS on several
file structures, make sure each structure has the
latest version.
Also, do not use the passwords supplied in ACCT.SYS.
The passwords are well known and using them could
decrease system security.
2. AUXACC.SYS - disk file structure quotas for all users.
Because device DSK: for any job is defined by the job's file
structure search list, different jobs may have different, and
possibly nonintersecting, definitions of DSK:. Therefore,
UPDATE ACCOUNTING FILES Page 13-2
one job could reference a file as DSK:filename when another
job could not (although that job could possibly reference the
file by specifying the file structure on which the file
exists).
To circumvent this problem, you or the system administrator
can specify all public file structures in the AUXACC entries
for all users. If you do not want a user to have space on a
particular file structure, specify a logged-in quota of 0.
In this case, no UFD is created for the user when he logs in,
but the file structure is included in the user's search list
so that he can always reference other files on that file
structure as device DSK:.
To convert the AUXACC files, type the following to 6.03
REACT:
*A
*W DSKA:[1,4]
and to put a second copy on a slower file structure, type:
*W DSKB:[1,4]
CHAPTER 14
GET NEW MONITOR
You can now get the new monitor from disk with BOOTS or from magtape
with BOOTM.
If you have a KL10 system, begin reading at Step B1. If you have a
KA10 or KI10 system, begin reading below at Step A1. Note that all of
the instructions in this chapter assume that you have a saved copy of
the monitor called SYSTEM.EXE on [1,4] on either DSKA:, DSKB:, ...,
DSK0.
Step A1 READIN BOOTS.
If the monitor was previously running, follow Situation A. If you
have an RP04 and BOOTS was previously written on blocks 0 and 4
through 7 of the disk packs with WTBOOT, follow Situation B.
Otherwise, follow Situation C.
Situation A: Monitor previously running
1. Set the NXM switch to off.
2. Set the memory address switches to 407.
3. Push the START button.
4. BOOTs responds with a carriage return. If there is no
response, try Situations B or C.
5. Type a carriage return to load SYS:SYSTEM.EXE from any disk
(DSKA through DSK0). If some other file, directory, or
structure is desired, type a line in the following format:
structure:file.ext[directory]
BOOTs types a bell and a question mark if it cannot find the
file, or if any errors occur while reading the file.
6. If successful, refer to step C1.
Situation B: RP04 Readin
1. Set the NXM and PAR STOP switch
2. Set the READIN switches to 270. (If you have multiple RH10s,
you may have to set the READIN switches to 274 or 360).
3. On the READIN panel in the RH10, set the toggle switch to
DISK, and set the thumbwheel to the READIN unit number. (If
the DF10C is in KI mode, use an odd-numbered unit. If it is
in KA mode or the RH10 is using a DF10, use an even-numbered
GET NEW MONITOR Page 14-2
unit.)
4. Push the STOP button.
5. Push the RESET button.
6. Push the READIN button.
7. BOOTS responds with a carriage return. If there is no
response, try the procedure for Situation C.
8. Type a carriage return to load SYS:SYSTEM.EXE from any disk
(DSKA through DSK0). If you want a different structure,
file, or directory from the default, type a line in the
following format:
structure:file.ext[directory]
BOOTS types a bell and a question mark if it cannot find the
file or if any errors occur while reading the file.
9. If successful, turn to step C1.
Situation C: Read BOOTS from paper-tape reader
1. Push the STOP button.
2. Push the RESET button.
3. Put the BOOTS tape in the paper tape reader.
4. Set the NXM switch to off.
5. Set the READIN switches to 104.
6. Push the READIN button. The paper tape should begin to move.
If it does not, retry the procedure at least once.
7. BOOTS responds with a carriage return when loaded. If there
is no response, retry the procedure at least once.
8. Type a carriage return to load SYS:SYSTEM.EXE from any disk
(DSKA through DSK0). If you want a different structure,
file, or directory from the default, type a line in the
following format:
structure:file.ext[directory]
BOOTS types a bell and a question mark if it cannot find the
file or if any errors occur while reading the file.
9. If successful, turn to Step C1.
Step A2 Readin a Monitor from Magtape Using BOOTM.
The following instructions describe how you get a KA10 or KI10 monitor
from magtape with BOOTM. These instructions assume that BOOTM has
been written as the first file on the magtape, BACKUP is the second
file, and there is a BACKUP save set containing the monitor to be
loaded.
Note that when BOOTM detects an error, it types a bell and a question
mark followed by an error message. The possible error messages are
listed in Chapter 16.
GET NEW MONITOR Page 14-3
Read BOOTM from magtape. (If you have a dual-magtape system, you must
use MTA for READIN.)
1. Push the STOP button.
2. Push the RESET button.
3. Mount the magtape on MTA unit 0 write-locked. (It must be
mounted on unit 0 for TM10 controllers or on the lowest
numbered ready unit for TU70s.)
4. For a TM10-controlled magtape, set the READIN switches to
340. For a TU70 unit, set the READIN switches to 220.
5. Set the NXM switch to off.
6. Push the READIN switch.
7. The magtape should move and read in BOOTM from the first file
BOOTM outputs the prompt characters BTM>. Then, you must give a file
specification for the monitor you want to read; for example:
DSKB:603.EXE[1,4]
If you type a carriage return in response to BTM>, the default monitor
(SYS:SYSTEM.EXE[1,4]) will be read in. This default file will be
gotten from the magtape on the same unit from which the hardware
READIN was done. If BOOTM was not read by hardware READIN from
magtape, the default for controller is TM10 and the default for unit
is 0.
Turn to Step C1.
Step B1 READIN BOOTS (KL10 Systems)
In most cases you should only have to check to see that the power
light to the right of the load switches is glowing red. Otherwise,
press the black power switch to the POWER ON position.
After powering up the system or deciding that you must do a complete
reload, you have a few alternatives on how you load the console
front-end processor. You can do a standard load from a disk pack or
DECtape, or you can load using the swicth register. In most cases,
you will load from a disk pack because it is much faster than DECtape.
However, if dual-port hardware problems prevent loading from a disk
pack, or if your system manager wants you to use software residing on
a DECtape, you may need to load the system using the DECtape drive.
Also, if you need to load the system using some non-standard hardware
configurations or special software, you will have to use the switch
register. The switch register allows you to load from any disk pack
or DECtape and also enter the KL initialization operator dialogue
(KLINIT). The KLINIT dialogue allows you to take such nondefault
paths as configuring memory yourself, loading a bootstrap program from
any file, and loading a nondefault monitor. The procedures for
loading the front-end processor via a disk pack (Situation A), a
DECtape (Situation B), or the switch register (Situation C) are
described in the next three sections.
Situation A: Loading RSX-20F from Disk
This is the most common way to load the console front end. It assumes
that the system power is ON, that all hardware and software have been
correctly installed, and that the disk pack has been mounted properly.
GET NEW MONITOR Page 14-4
Also, the disk pack containing both the TOPS-10 monitor and the
RSX-20F front-end monitor must be on a dual-ported drive (controller
select switch points to A/B), and the drive must be unit 0 with
respect to both the central and front-end processors.
The operation is as follows:
1. Set the ENABLE/DISABLE load switch to ENABLE.
This will enable the other three load switches.
2. Press the DISK load switch.
This will cause the front-end processor to access the disk on
drive 0 and load the RSX-20F monitor. The KL initialization
program (KLINIT) is loaded and the default hardware
configurations of cache and external memory are set up. The
bootstrap program for the TOPS-10 monitor is then loaded into
the central processor and started. The following is an
example of the output you will receive on the console
terminal.
RSX-20F V006A 0:16 21-JUN-76
[SY0: REDIRECTED TO DB0:]
[DB0: MOUNTED]
KLI -- VERSION V002E RUNNING
KLI -- MICROCODE VERSION 131 LOADED
KLI -- ALL CACHES ENABLED
LOGICAL MEMORY CONFIGURATION:
CONTROLLER
ADDRESS SIZE RQ0 RQ1 RQ2 RQ3 CONTYPE INT
000000 512K 04 FOR ALL DMA20 4
KLI -- BOOTSTRAP LOADED AND STARTED
The first line of output tells you the version and the creation time
and date of the RSX-20F monitor for the front end. The next two lines
tell you that DB0: (the disk pack on drive 0) is the system device
(SY0:) for the front-end tasks. The next eight lines are output by
the KL initialization program. They document the version of KLINIT
that is running and tell you that the KL10 microcode was successfully
loaded, the cache was enabled, external memory was configured, and
that the bootstrap program for the TOPS-10 monitor was loaded and
started.
If an error occurs during the KL initialization program, you will
receive an error message preceded by "KLI -- ?" and you will be placed
in the KLINIT dialogue mode described in Appendix C of the KL Series
Operator's Guide. (The loading of the TOPS-10 monitor is discussed in
Step B2.)
Situation B: Loading RSX-20F from DECtape
This method of loading the console front end should be used if you
cannot load from a disk pack or if you need a particular version of
software that is only on DECtape. It assumes that the system power is
ON, that the proper software exists on DECtape, and that the DECtape
is mounted correctly on Unit 0. Be sure to leave the DECtape mounted
while the system is running.
The operation is as follows:
GET NEW MONITOR Page 14-5
1. Set the ENABLE/DISABLE load switch to ENABLE.
This will enable the other three load switches.
2. Press the DECTAPE load switch.
This will cause the front-end processor to access the DECtape
on drive 0 and load the RSX-20F monitor. The KL
initialization program (KLINIT) is loaded and the default
hardware configurations of cache and external memory are set
up. The bootstrap programs for the TOPS-10 monitor is then
loaded into the central processor and started. The following
is an example of the output you will receive on the console
terminal.
RSX-20F V006A 0:16 21-JUN-76
[SY0: REDIRECTED TO DT0:]
[DT0: MOUNTED]
KLI -- VERSION V002E RUNNING
KLI -- MICROCODE VERSION 131 LOADED
KLI -- ALL CACHES ENABLED
LOGICAL MEMORY CONFIGURATION
CONTROLLER
ADDRESS SIZE RQ0 RQ1 RQ3 CONTYPE INT
000000 512K 04 FOR ALL DMA20 4
KLI -- BOOTSTRAP LOADED AND STARTED
The first line of output tells you the version and the creation time
and date of the RSX-20F monitor for the front end. The next two lines
tell you that DT0: (the DECtape on unit 0) is the system device
(SY0:) for the front-end tasks. The next eight lines are output by
the KL initialization program. They give the version of KLINIT that
is running and tell you that the KL10 microcode was successfully
loaded, the cache was enabled, external memory was configured, and
that the bootstrap program for the TOPS-10 monitor was loaded and
started.
If an error occurs during the KL initialization program, you will
receive an error message preceded by "KLI -- ?" and you will be placed
in the KLINIT dialogue mode described in Appendix C of the KL Series
Operator's Guide. (The loading of the TOPS-10 monitor is discussed in
Section 3.3.)
Situation C: Loading RSX-20F via the Switch Register
You must load the console front end via the switch register if you
need to do any of the following:
1. Enable specific cache or configure external memory yourself
instead of using the default configurations.
2. Load a TOPS-10 bootstrap program from a file with a name
other than BOOT.EXB; for example, BOOTM.EXB containing
BOOTM, which loads a monitor from magnetic tape.
3. Load from a disk pack or DECtape that is not on unit 0.
The software for the front-end processor must reside on a DECtape or
disk pack connected to the front end. If both the front-end software
and the system monitor reside on the same disk pack, the pack must be
mounted on a dual-ported drive and the controller select switch must
be set to A/B.
GET NEW MONITOR Page 14-6
The operation is as follows:
1. Set the ENABLE/DISABLE load switch to ENABLE.
This will enable the other three load switches.
2. Set the appropriate switches (or bits) in the switch
register.
Switch 0 is mandatory when loading via the switch register.
Switches 1 and 2 must be set on if you intend to use the KL
initialization dialogue (KLINIT). If RSX-20F resides on a
disk pack, set switch 7 on; if RSX-20F is on a DECtape,
switch 7 must be off. Switches 8-10 must be set to specify
the drive number of the disk or DECtape. (For a detailed
description of all the bit settings in the switch register,
see Table 14-1.)
Table 14-1
Switch Register Bit Definitions
Bit Meaning
0 If this is set, the remaining bits are interpreted.
You must set this to load via the switch register.
1,2 If both are set, the KL initialization operator
dialogue (KLINIT) is loaded and started. This is the
usual case when loading via the switch register.
If either one is set, the RSX-20F is loaded; no
communication is initiated between the -10 and the
-11 processors at this time.
If both are not set, the system is loaded much like
it is via the DISK or DECTAPE load switch. However,
because other bits are interpreted, you can specify
the unit number of the bootstrap device in bits 8-10.
3-6 Currently not used, and must not be set.
7 If this is set, the bootstrap device is a disk pack
on a dual-ported drive.
If this is not set, the bootstrap device is a DECtape
drive on the front-end processor.
8-10 These three bits allow you to specify the unit number
of the bootstrap device (0 to 7). No bits set
indicate unit 0; bits 9 and 8 set indicate unit 3.
11-14 Currently not used, and must not be set.
15 This indicates the action taken when an I/O error
occurs during the bootstrapping. If this is set, the
operation is retried indefinitely if an error occurs.
If not set (the normal case), a halt occurs after ten
unsuccessful retries.
16,17 Not used, and must not be set.
GET NEW MONITOR Page 14-7
1. Press the SW/REG load switch.
This will cause the front-end processor to access the disk
drive (switch 7 set on) or the DECtape drive (switch 7 set
off) with the unit number as specified in switches 8-10.
The KL initialization program (KLINIT) is loaded and
started and the default hardware configurations of cache
and external memory are set up. The standard bootstrap
program for the TOPS-10 monitor is then loaded into the
central processor and started.
The following is an example of the output you will receive
on the console terminal if switches 0, 7, 8, and 9 are set
on.
RSX-20F V006A 0:16 21-JUN-76
[SY0: REDIRECTED TO DB3:}
[DB3: MOUNTED]
KLI -- VERSION V002E RUNNING
KLI -- MICROCODE VERSION 131 LOADED
KLI -- ALL CACHES ENABLED
LOGICAL MEMORY CONFIGURATION:
CONTROLLER
ADDRESS SIZE RQ0 RQ1 RQ2 RQ3 CONTYPE INT
000000 512K 04 FOR ALL DMA20 4
KLI -- BOOTSTRAP LOADED AND STARTED
The first line of output tells you the version and the
creation time and date of the RSX-20F monitor for the front
end. The next two lines tell you that DB3: (the disk on
unit 3) is the system device (SY0:) for the front end
tasks. If bit 7 had not been set, the DB3: would have
been DT3: for a DECtape. The next eight lines are output
by the KL initialization program. They give the version of
KLINIT that is running and tell you that the KL10 microcode
was successfully loaded, the cache was enabled, external
memory was configured, and that the standard bootstrap
program for the TOPS-10 monitor was loaded and started.
(The loading of the TOPS-10 monitor is discussed in Section
3.3.)
If, in the previous example, you had also set switches 1
and 2 on, the console output would have been as follows:
RSX-20F V006A 0:16 21-JUN-76
[SY0: REDIRECTED TO DB3:]
[DB3: MOUNTED]
KLI -- VERSION V002E RUNNING
KLI -- ENTER DIALOG [NO,YES,EXIT,BOOT]?
KLI>
Switches 1 and 2 set on specify that you wish to enter the
KL initialization program (KLINIT) operator dialogue.
KLINIT has just asked you the first question and is waiting
for an answer. When you have answered all the applicable
questions in the dialogue, the last message issued will be:
KLI -- BOOTSTRAP LOADED AND STARTED
as in the previous example.
(For a complete description of the KLINIT dialogue
GET NEW MONITOR Page 14-8
messages, and examples, refer to Appendix C of the KL
Series Operator's Guide.)
Step B2: Loading The Central Processor (TOPS-10 Monitor)
When the front-end processor has been loaded using any one
of the three methods described in Section 3.2, the system
informs you:
KLI -- BOOTSTRAP LOADED AND STARTED
The actual bootstrap program that has been loaded depends
upon the method of loading the front end.
If you used the DISK load, DECTAPE load, or the SW/REG load
with switches 1 and 2 off, you will have loaded the
bootstrap program found in the file BOOT.EXB. This would
usually be BOOTS, a program to load the monitor from a disk
pack.
If you had used the KLINIT dialog to specify a non-default
BOOT file, you could have loaded a program like BOOT, which
will load the monitor from magnetic tape.
If you are loading the monitor from disk, follow Situation
A. If you are loading the monitor from magtape, follow
Situation B.
Situation A: Loading TOPS-10 from Disk
In response to the message:
KLI -- BOOTSTRAP LOADED AND STARTED
you can press the carriage return key and load the default
monitor from the disk file SYSTEM.EXE or you can specify
another file such as:
DSKB:TSTSYS.EXE
and load a different monitor.
When the TOPS-10 monitor is loaded, it will enter into an
initialization dialogue with you to determine start-up
options and conditions. These are covered in Step C1.
Situation B: Loading TOPS-10 from Magnetic Tape
In order to load the TOPS-10 monitor from a magnetic tape,
the following prerequisites must be met:
1. The magnetic tape containing the monitor should be
mounted on drive 0. If this is not possible, mount it
on any drive but make sure that all other tape drives
are set OFFLINE.
2. The bootstrap program to load a monitor from magnetic
tape, BOOTM, must reside on the front-end load device;
disk or DECtape.
3. Load the console front end via the SW/REG load
procedure. (Situation C) with at least bits 0, 1, and
2 set ON. This will allow you to use the KLINIT
dialog.
GET NEW MONITOR Page 14-9
4. When KLINIT prompts you with:
KLI -- LOAD KL BOOTSTRAP[YES,NO,filename]?
KLI>
answer with the name of the file containing the
magnetic tape bootstrap. For example, if BOOTM was in
the file BOOTM.EXB, respond with:
KLI>BOOTM.EXB
KLINIT will load BOOTM, give you the message:
KLI -- BOOTSTRAP LOADED AND STARTED
and you would then be under control of the bootstrap
program. When BOOTM outputs its prompt characters:
BTM>
you can give the file specification for the monitor you
want to load, such as:
BTM>DSKB:MAGSYS.EXE[1,4]
or you can simply give a carriage return to get the
default file, which is DSKB: SYSTEM.EXE[1,4] from the
same magnetic tape.
When the TOPS-10 monitor is loaded, it enters into an
initialization dialogue with the operator to determine
the start-up options and conditions. These are covered
in Step C1.
STEP C1: TOPS-10 Initialization Dialogue
When the TOPS-10 monitor is loaded, the system checks to
see that the monitor and processor are compatible. If they
are not, the following error message is issued:
?THIS MONITOR WAS BUILT FOR A xxxxxx
AND WILL NOT RUN PROPERLY ON A yyyyyy.
where xxxxxx and yyyyyy can be KA10, KI10 or KL10. If this
message persists, the system administrator should review
the monitor generation procedure in the Software Notebooks.
If the load time diagnostic program SYSCHK was included in
the system software at monitor generation time (determined
during the MONGEN dialogue), the system will prompt with:
SYSCHK (N,Y):
A response of Y runs a 5 second diagnostic program that
ensures the accessibility of all configured memory and
system devices. A reply of N or a carriage return skips
the diagnostic.
The system then types the monitor name, date and version
number; for example:
ABC123 KL10 SYS#1234 03-1-77
When the system prompts:
GET NEW MONITOR Page 14-10
WHY RELOAD:
reply with one of the following acceptable answers:
OPR NXM CM
PARITY HALT SA
POWER LOOP NEW
STATIC HUNG SCHED
HARDWARE PM OTHER
If you do not reply within 60 seconds, OTHER is assumed.
(Refer to Chapter 14 for the meanings and usage of the
above replies.) The reply is stored in the system error
file (ERROR.SYS) and can be retrieved at a later time using
the SYSERR program. For example, a reply of:
WHY RELOAD:SCHED
will record that this particular monitor load was a
scheduled reload.
When the system prompts:
DATE:
enter the numeric day and alphabetic month, in either
order. The month may be abbreviated to any point where it
is still unique. The year is optional. If entered, it
must be either the full four digits or the last two. If
the year is not entered, it is assumed to be the same as in
the monitor creation date. If, for example, the monitor
creation date was June 21, 1976, any of the following
replies would be recorded as September 8, 1976.
DATE:SEP 8 1976
DATE:S 8 76
DATE:8 SEP
When the system prompts:
TIME:
enter a 4 digit time based on a 24-hour clock. For
example,
TIME:2015
represents 8:15 PM.
When the system prompts:
STARTUP OPTION:
reply with one of the following:
QUICK REFRESH LONG
GO UNITED NOINITIA
DESTROY CHANGE
A complete description of each startup option is contained
in Chapter 16. The usual operator's reply is GO to start
the system with a minumim amount of dialogue or QUICK to
start the system immediately without changing any
parameters.
GET NEW MONITOR Page 14-11
The monitor is now ready for timesharing. INITIA is
brought up automatically on systems with OPSER and an
automatic restart file.
If you have a KA10 system, set the NXM switch OFF unless
you are debugging and prefer to have the machine stop and
display the memory address when non-existent memory is
referenced.
CHAPTER 15
COPY NEW MONITOR TO SYS
You should now copy the new monitor to SYS for READIN with BOOTS from
disk.
.COPY SYS:SYSTEM.EXE = 6.03 EXE
CHAPTER 16
ONCE DIALOGUE
In most instances, you should start your monitor with the GO, NOINITIA
or QUICK STARTUP Option. The STARTUP Options are a portion of the
ONCE dialogue, which is described in this chapter.
16.1 INTRODUCTION
The ONCE-Only dialogue (Version 662) is an interactive program used by
the operator at system startup to set or alter a number of important
system parameters. These parameters include the date, time of day,
and most of the disk file structure information.
You respond to a series of questions asked during the ONCE-Only
dialogue by typing your responses, one at a time, in conversational
mode.
Before starting the system you should:
1. Read this chapter to acquaint yourself with the current
version of the ONCE-Only dialogue format.
2. Decide which parameters you wish to set or alter when the
system is started.
3. Be prepared to answer the required ONCE-Only dialogue
questions by familiarizing yourself with your installation's
configuration and the content and format of individual
questions.
The 6.03 version of the ONCE-Only dialogue permits you to answer only
those questions directly related to the parameters that you wish to
define or alter in some way. The rest of the parameters are
automatically set according to a DEC-supplied or an
installation-defined standard. (Some parameter defaults can be
modified by the system programmer at individual installations during
the MONGEN dialogue program.) (Refer to Chapter 8 for a description
of MONGEN.)
16.1.1 Summary of STARTUP Options
You have a choice of eight STARTUP Options. They are summarized here
and then described in the remainder of this chapter.
QUICK To start the system quickly without changing any
parameters. (Refer to Section 16.2.)
ONCE DIALOGUE Page 16-2
NOINITIA To start the system quickly without setting any
new parameters and without running INITIA. (Refer
to Section 16.2.)
GO To check to see if everything is in order and to
start the system with a minimum of dialogue.
(Refer to Section 16.3.)
DESTROY To automatically restructure and refresh all disks
according to a predetermined set of standards.
(Refer to Section 16.4.)
REFRESH To explicitly refresh selected file structures
without changing other system parameters. (Refer
to Section 16.5.)
UNITID To change selected unit IDs without changing other
system parameters. (Refer to Section 16.6.)
CHANGE To explicitly set or change selected file
structure parameters. (Refer to Section 16.7.)
LONG To explicitly set all parameters and restructure
the file system in a non-standard way. (Refer to
Section 16.8.)
You should select the LONG STARTUP Option only if there is no other
alternative. You are encouraged to use one or more of the other seven
STARTUP Options (the simplest combination for their purposes), and to
accept the system standard values whenever possible. Note that the
STARTUP OPTION: question is repeated after you define some of the
parameters. This permits you to define a second set of parameters
(e.g., DESTROY, REFRESH, UNITID, or CHANGE) or to start the system
(QUICK or GO).
The first eight steps described in Section 16.2.2 are common to all
system startups regardless of which option you choose. You must
perform these steps correctly before the ONCE-only dialogue asks you
for the STARTUP Option. The first eight steps are listed in Section
16.2.2 as they apply to all STARTUP Options.
16.1.2 Special Considerations
The software supports the KA10, KI10, and KL10 processors. In
general, the procedures are the same for all systems. The primary
difference between the KA10 and the KI10 is the physical layout of the
console. Where the KA10 console has rocker switches, the KI10 console
has lighted pushbutton switches but the switch functions are the same.
The instructions in this document use the following conventions when
referring to the ON and OFF positions of the switches:
KA10 Console KI10 Console
------------ ------------
Switch ON Front (or bottom)of Lighted
rocker switch is
pressed.
Switch OFF Back (or top) of Not Lighted
rocker switch is
pressed.
ONCE DIALOGUE Page 16-3
In addition, the following special steps should be performed with a
KI10 system before beginning the READIN procedure.
1. Check to see that the MAINT light is OFF. If it is not,
check that the following switches are OFF:
PARITY STOP
FM MANUAL
MEM OVERLAP DIS
MARGIN ENABLE
SINGLE PULSE
SINGLE INSTRUCTION
2. Make sure that the DATA and ADDRESS switches are operating
properly. That is, when the switch is pressed the indicator
is turned either off or on, the opposite to what it had been.
If they do not work properly, make sure that the CONSOLE LOCK
and the CONSOLE DATA LOCK switches are OFF.
3. Under normal circumstances systems with a KI10 processor
should be run with the switches set as follows:
a. NXM STOP switch OFF
b. EXEC PAGING switch ON
c. USER PAGING switch OFF
(These paging switches should be set so that the
addresses examined by the EXAMINE switch are monitor
(EXEC mode) addresses instead of user program (USER mode)
addresses.)
The NXM FLAG and processor clock PIA indicators are located on
different bays.
1. On the KA10 the NXM FLAG and processor clock PIA indicators
(PIA33, PIA34, and PIA35) are located on Bay 1, the left-most
bay, farthest from the console. They are in the second row
of lights, below the register labeled CPA.
2. On the KI10, the NXM FLAG and processor clock PIA indicators
(APR CLK PIA) are located on Bay 3, the same bay as the
console. They are in the bottom row of lights.
16.1.3 Special Multiprocessing Considerations
Under ordinary circumstances, multiprocessing systems do not require
any special startup procedures. CPU1, the secondary processor, waits
and automatically starts after CPU0, the primary processor, is
started. However, if the system is started after a crash, then the
procedure depends upon which CPU crashed.
16.1.4 Conventions Used in this Chapter
1. Errors messages
ONCE-Only error messages are preceded by either a question
mark (?) or a percent sign (%). Messages preceded by a
question mark are fatal unless corrective action is taken.
ONCE DIALOGUE Page 16-4
Messages preceded by a percent sign are warnings.
2. Carriage Return
<CR> indicates that you should type a carriage return.
Because most operator typeins must be terminated by carriage
return, <CR> is used to indicate a blank line and for
emphasis when other terminators, such as altmode, are also
possible.
3. File Structure
STR is an abbreviation for file structure.
ONCE DIALOGUE Page 16-5
STARTUP OPTION: QUICK aND NOINITIA
16.2 STARTUP OPTION: QUICK aND NOINITIA
The QUICK option enables you to start up the system quickly, without
altering any system parameters and without any further dialogue.
After listing off-line disk units, the system starts running.
This option is recommended for startups when speed is important, when
file structure organization and other parameters need not be altered,
or after you have set selected parameters by using another option, and
you are ready to start the system.
The NOINITIA option is identical to QUICK, except that INITIA is not
run.
Note that if ONCE-Only encounters problems in trying to start the
monitor, the LONG dialogue is automatically invoked. (Refer to
Section 16.10 for error messages and to Section 16.8.3 for a
description of the LONG dialogue questions.)
16.2.1 Frequency
The QUICK (or GO) option is invoked everytime the monitor is reloaded
into the machine.
16.2.2 Summary of Steps
1. You get the monitor with BOOTM or BOOTS. (Refer to Chapters
2 and 3.)
2. The system checks to see that the monitor and processor are
compatible. If not, an error message is typed and the
program halts.
3. The system asks if you wish to run the loadtime diagnostic
program, SYSCHK.
SYSCHK(N,Y):
Note that this is asked only if the SYSCHK program has been
included as a part of the system software (as determined by
the system programmer during the MONGEN dialogue, refer to
Chapter 8).
USER RESPONSE: N OR <CR> to skip the diagnostic.
4. The system types the name of the monitor and creation date;
for example:
6.03 SYS #160 3-17-77
5. The system asks for the reason for the reload.
WHY RELOAD:
USER RESPONSE: NEW
You answer with one of the following acceptable answers
depending on your reason. (Refer to Section 16.9 for an
explanation of these answers.)
ONCE DIALOGUE Page 16-6
STARTUP OPTION: QUICK aND NOINITIA
OPR LOOP
PARITY HUNG
POWER PM
STATIC CM
HARDWARE SA
NXM NEW
HALT SCHED
OTHER
6. The system asks for the date.
DATE:
USER RESPONSE: MAR 17 1977
You type the date as the name of the month (January-December)
and the numerical day of the month (1-31) in any order. (The
name of the month may be shortened as long as the
abbreviation is unique.) The year is optional. If included,
the year must be typed as a 4-digit number (1977) or a
2-digit abbreviation (77) following the month and day. If
omitted, the year is assumed to be the same as the year the
monitor was created. Examples of acceptable dates are:
MARCH 17 1977
MAR 17
17 MAR
7. The system asks for the time.
TIME:
USER RESPONSE:0843
You type the time based on a 24-hour clock, e.g.,
0843=8:43 A.M. and 1345=1:45 P.M.
8. The system checks to see if all of memory is accessible.
9. ONCE asks for your choice of
STARTUP OPTION:
USER RESPONSE: QUICK
For a description of the other responses, refer to the list
below:
GO Section 16.3
DESTROY Section 16.4
REFRESH Section 16.5
UNITID Section 16.6
CHANGE Section 16.7
LONG Section 16.8
10. ONCE lists off-line units; for example:
%xxxx IS OFF-LINE
%yyyy IS OFF-LINE
(Note that you have no opportunity to change the status of
these units.)
ONCE DIALOGUE Page 16-7
STARTUP OPTION: QUICK aND NOINITIA
11. ONCE checks to see if there are controllers off line or if
any units are write locked. If so, messages are transmitted
to the operator. Otherwise, if everything is in order, the
monitor starts running the null job and initiates
timesharing, usually via INITIA, OPSER, and an automatic
STARTUP file. If you typed NOINITIA, a . is printed and you
must log in. Otherwise, you are automatically logged in.
16.2.3 Examples
Example 1
SYSCHK (N,Y): Y
MEMORY MAP=
FROM TO SIZE/K
000000 537777 176
CONTROLLER MTA IS UNACCESSIBLE.
6.03 SYSTEM #160 4-12-77
WHY RELOAD: OTHER ;MANUAL EXAMPLE
DATE: 4 MAY
TIME: 0045
STARTUP OPTION: QUICK
%DPA6 IS OFF-LINE
%DPB0 IS OFF-LINE
6.03 SYSTEM #160 4-12-77
DSKN: System #160 Initia types
. the text contained
. in STR.TXT.
.
.LOGIN[1,2] ;This is typed automatically
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
00:45:24(0)
.
00:45:24(0)
.
00:45:32(0)
.
00:45:39(B\L1)
Example 2
BOOTS
6.03
SYSCHK (N,Y): N
6.03 SYSTEM #160 4-19-77
WHY RELOAD:NEW
DATE: APRIL 28
TIME: 1100
ONCE DIALOGUE Page 16-8
STARTUP OPTION: QUICK aND NOINITIA
STARTUP OPTION: QUICK
%DPA4 IS OFF-LINE
%DPA5 IS OFF-LINE
%DPA6 IS OFF-LINE
6.03 SYSTEM #160 11:00:19 CTY
.LOGIN[1,2]
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
.
.
.
Example 3
SYSCHK (N,Y): N
6.03 SYSTEM #160 4-19-77
WHY RELOAD: NEW
DATE: APRIL 20
TIME: 700
STARTUP OPTION: NOINITIA
.
16.3 STARTUP OPTION: GO
When you specify the GO option, ONCE-Only checks to see if everything
is in order, and starts the system with a minimum of dialogue. After
you respond to questions asking if specific off-line units are to be
on line, off line, or down, the system is started.
Note that if ONCE-Only encounters problems in trying to start the
monitor, then the LONG dialogue is automatically invoked. Refer to
Section 16.10 for error messages and to Section 16.8.3 for a
description of the LONG dialogue questions.
16.3.1 Frequency
The GO (or QUICK) option must be invoked every time the monitor is
reloaded into the machine.
16.3.2 Summary of Steps
Steps 1 through 8 are described in Section 16.2.2, because they are
common to all STARTUP Options.
9. ONCE asks for your choice of
STARTUP OPTION:
ONCE DIALOGUE Page 16-9
STARTUP OPTION: GO
USER RESPONSE: GO
10. ONCE lists off-line units, if any, and then asks if off-line
units are to remain off line or not.
%xxxx IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN?
(TYPE #)
USER RESPONSE: 1,2, or 3
11. ONCE checks to see if there are any off-line controllers or
if there are any write-locked units. If so, messages are
transmitted to the operator. Otherwise, if everything is in
order, the monitor starts running the null job and initiates
timesharing, usually via INITIA, OPSER, and an automatic
startup file.
16.3.3 Examples
Example 1
6.03 SYSTEM #160 4-2-77
WHY RELOAD: OTHER ;EXAMPLE
DATE: 4 MAY
TIME: 52
STARTUP OPTION: GO
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
3
%DPB0 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
6.03 SYSTEM #160 00:52:46 CTY
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
00:52:48(0)
.
00:52:49(0)
.
00:52:57(0)
.
Example 2
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 12-19-77
WHY RELOAD: SCHED
DATE: D 28
ONCE DIALOGUE Page 16-10
STARTUP OPTION: GO
TIME: 1120
STARTUP OPTION: GO
%DPA4 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPA5 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
1
6.03 SYSTEM #160 11-21-04 CTY
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
.
.
.
16.4 STARTUP OPTION: DESTROY
The DESTROY option automatically restructures and refreshes all file
structures according to a predetermined set of standards. A minimum
of dialogue is encountered, because the file structure parameters are
determined by the system programmer during the MONGEN dialogue
program. (Refer to Chapter 8.)
The DESTROY option, as the name implies, destroys all existing file
structures, then restructures and refreshes the disk units as directed
by MONGEN. If the system programmer has not altered the values of the
MONGEN symbols, the DESTROY option organizes the file system according
to the "symbol, values" specified by the DEC-supplied software; that
is, all fixed head disks are in the first file structure (called
DSKA), and all disk pack units are in the second file structure
(called DSKB), with 200K of swapping space allocated on each of the
fixed head units, and on the first disk pack unit (as backup). (Refer
to Chapter 8 for a description of the HDWGEN decimal and octal
symbol-values that affect file structure organization.)
The DESTROY option is recommended for new installations starting up
their systems for the first time and for existing installations using
new disk units for the first time.
16.4.1 Frequency
After the file structure parameters have been established, they are
written onto each structure in two places called the HOME blocks.
Once this is done, it need not be done again. When the monitor is
reloaded, it reads the HOME blocks and sets up the disk parameters
automatically.
ONCE DIALOGUE Page 16-11
STARTUP OPTION: DESTROY
16.4.2 Summary of Steps
Steps 1 through 8 are described in Section 16.2.2, because they are
common to all STARTUP Options.
9. ONCE asks for your choice of
STARTUP OPTION:
USER RESPONSE: DESTROY
Existing structures are destroyed, new structures are
created, and all disks are refreshed according to a
predetermined set of values.
A warning message is typed and confirmation is requested.
%WARNING-ALL STRS WILL BE REFRESHED.
PROCEED? (Y or <CR>)
USER RESPONSE: Y
When finished the following is typed
HOME BLOCKS WRITTEN ON ALL UNITS
START SYSTEM? (Y or <CR>)
USER RESPONSE: Y or <CR>
An answer of <CR> returns control to the STARTUP OPTION
question again. A YES answer continues with step 10.
10. ONCE repeats the STARTUP question, to which you should
respond QUICK (see Example 1). ONCE asks you to login.
TO AUTOMATICALLY LOGIN UNDER [1,2] TYPE LOGIN
USER RESPONSE: LOGIN
11. ONCE starts running the null job and initiates timesharing.
12. If a new monitor is being generated for the first time, you
should return to Chapter 4.
16.4.3 Example
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 12-13-77
WHY RELOAD: SA
DATE: 26 DEC
TIME: 1300
STARTUP OPTION: DESTROY
%WARNING-ALL STRS WILL BE REFRESHED.
PROCEED?(Y OR <CR>)
%FHA1 IS OFF-LINE
ONCE DIALOGUE Page 16-12
STARTUP OPTION: DESTROY
%DPA0 IS OFF-LINE
HOME BLOCKS WRITTEN ON ALL UNITS
START SYSTEM? (Y OR <CR>)Y
STARTUP OPTION: Q
TO AUTOMATICALLY LOG-IN UNDER [1,2] TYPE "LOGIN"
LOGIN
.
.
.
16.5 STARTUP OPTION: REFRESH
The REFRESH option allows you to restore certain file structures to
their initial condition (i.e., refresh them) without changing other
system parameters. (Note that only HOME.SYS is required to describe
the file structure. Once these have been defined via DESTROY or LONG,
refreshing individual structures is very simple.)
16.5.1 Frequency
Ordinarily, structures need refreshing every two to three months.
Refreshing may be required when:
1. Disks are full. When you wish to retain only recently
accessed files, you save recently accessed files with BACKUP
before refreshing and then restore them (with BACKUP) after
refreshing.
2. System crashes have left a number of blocks not allocated to
any file. Refer to the DSKRAT documentation for instructions
on determining the number of lost blocks and recovering them
without the need for refreshing the file structure.
3. Fragmentation of free blocks results in inefficient use of
disk space, i.e., when free blocks on the disk are so
scattered that long sequentially written files are using
extended RIBs. (Refer to the DSKLST.RNO specification for
information on interpreting the DSKLST output.)
16.5.2 Summary of Steps
Steps 1 through 8 are described in Section 16.2.2 because they are
common to all STARTUP Options.
9. ONCE asks for your choice of
STARTUP OPTION:
USER RESPONSE: REFRESH
10. ONCE lists any off-line units and asks you to indicate which
off-line units are to remain off-line.
%xxxx IS OFF-LINE
ONCE DIALOGUE Page 16-13
STARTUP OPTION: REFRESH
DO YOU WANT IT TO BE 1)ON-LINE, 2) OFF-LINE, OR 3)DOWN?
(TYPE #)
USER RESPONSE: 1,2, or 3
11. ONCE checks to see if there are any off-line controllers or
if there are any write-locked units. If so, messages are
transmitted to the operator.
12. ONCE asks for the structures to be refreshed.
TYPE STR NAME TO BE REFRESHED (CR IF NONE, ALL IF ALL)
USER RESPONSE: ALL or individual structure name.
When the structures are refreshed, control returns to the
STARTUP OPTION question. You can enter another STARTUP
Option (such as DESTROY, UNITID, CHANGE, or LONG) or you can
start the system (QUICK or GO).
16.5.3 Examples
Example 1
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 12-19-77
WHY RELOAD: NEW
DATE: DEC 28
TIME: 1250
STARTUP OPTION: REFRESH
TYPE STR NAME TO BE REFRESHED (CR IF NONE, ALL IF ALL)
DSKH
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
DSKX
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
DSKX
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
STARTUP OPTION: QUICK
TO AUTOMATICALLY LOG-IN UNDER [1,2] TYPE "LOGIN"
LOGIN
.
.
Example 2
SYSCHK (N,Y):N
6.03 DUAL CPU 11-21-77
WHY RELOAD: SA
DATE: 4 DEC
TIME: 1050
STARTUP OPTION: REF
ONCE DIALOGUE Page 16-14
STARTUP OPTION: REFRESH
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPB0 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%NEED REFRESHING:
DSKA,DSKB,DSKC
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
DSKA
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
STARTUP OPTION: Q
TO AUTOMATICALLY LOG-IN UNDER [1,2] TYPE "LOGIN"
LOGIN
6.03 DUAL CPU 1:06:02
.LOG
.INITIA
6.03 DUAL CPU 01:06:14 CTY
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
.
.
.
16.6 STARTUP OPTION: UNITID
The UNITID option allows you to change selected unit IDs, without
changing any other parameters.
16.6.1 Frequency
After the file structure parameters have been established, they are
written onto each structure in two places called the HOME blocks.
Once this is done, it need not be done again. When the system is
reloaded, the monitor reads the HOME blocks and sets up the disk
parameters automatically.
16.6.2 Summary of Steps
Steps 1 through 8 are described in Section 16.2.2, because they are
common to all STARTUP Options.
9. ONCE asks for your choice of
STARTUP OPTION:
USER RESPONSE: UNITID
ONCE DIALOGUE Page 16-15
STARTUP OPTION: UNITID
10. ONCE lists any off-line units and asks you to indicate which
off-line units are to remain off-line.
%xxxx IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN?
(TYPE #)
USER RESPONSE: 1,2, or 3
ONCE checks to see if there are any off-line controllers or
if there are any write-locked units. If so, messages are
transmitted to the operator.
11. ONCE asks if there are any physical unit names and unit IDs
to be changed.
CHANGE ALL UNIT ID's (Y OR <CR>)
USER RESPONSE: Y OR <CR>
When you specify Y, ONCE lists each unit and asks for the
I.D.:
AFTER EACH UNIT NAME, TYPE THE I.D.
xxx(aaaa):
USER RESPONSE: ABCD
yyyy(bbbb)
USER RESPONSE: EFGH
zzzz(cccc):
USER RESPONSE: IJKL
When you respond with <CR>, ONCE asks, for each unit, the
unit name and its I.D.
TYPE UNIT NAME, A COMMA, AND NEW UNIT ID FOR EACH
DESIRED UNIT.
(EXTRA <CR> WHEN THROUGH)
USER RESPONSE: xxxx, ABCD <CR>
yyyy, EFGH <CR>
<CR>
12. If HOME blocks are to be written, ONCE asks for the physical
unit and then indicates when HOME blocks are written.
TYPE PHYSICAL UNIT TO WRITE HOME BLOCKS (EXTRA CR
WHEN THROUGH)
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
USER RESPONSE: ALL
HOME BLOCKS WRITTEN
When HOME blocks are written, control returns to the STARTUP
Option question. You can enter another option (such as
DESTROY, UNITID, CHANGE, or LONG) or you can start the system
ONCE DIALOGUE Page 16-16
STARTUP OPTION: UNITID
(QUICK or GO).
16.6.3 Examples
Example 1
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 4-19-77
DATE: MAY 28
TIME: 1130
STARTUP OPTION: UNITID
CHANGE ALL UNIT ID'S? (Y OR <CR>)
TYPE UNIT NAME, A COMMA, AND NEW UNIT ID FOR EACH DESIRED UNIT.
(EXTRA <CR> WHEN THROUGH)
DPA4,UNIT4
DPA5,UNIT5
TYPE PHYSICAL UNITS TO WRITE HOME BLOCKS (EXTRA CR WHEN THROUGH)
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
DPA4
DPA5
HOME BLOCKS WRITTEN
STARTUP OPTION: Q
6.03 SYSTEM #160 11:32:43 CTY
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
.
.
.
Example 2
SYSCHK (N,Y): N
6.03 SYSTEM #160 12-02-77
WHY RELOAD: SA
DATE: 4 DEC
TIME: 55
STARTUP OPTION: UNITID
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPB0 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
ONCE DIALOGUE Page 16-17
STARTUP OPTION: UNITID
2
CHANGE ALL UNIT ID'S? (Y OF <CR>)Y
AFTER EACH UNIT NAME, TYPE THE I.D.
FHA0(GLZX):XLZG
FHA1(XZLG):GLZX
DPA0(RP03A):RP03A1
DPA1(RP03B):RP03B1
DPA2(3RPXYZ):RP03C1
DPA3(2RP023):
DPA4(2RP044):
DPA5(2RP042):
DPB1(2RP004):
TYPE PHYSICAL UNITS TO WRITE HOME BLOCKS (EXTRA CR WHEN THROUGH)
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
ALL
HOME BLOCKS WRITTEN
STARTUP OPTION: UNITID
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPB0 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
CHANGE ALL UNIT ID'S (Y OR <CR>) <CR>
TYPE UNIT NAME, A COMMA, AND NEW UNIT ID FOR EACH DESIRED UNIT.
(EXTRA <CR> WHEN THROUGH)
FHA0,FHARD1
FHA1,SAM
TYPE PHYSICAL UNIT TO WRITE HOME BLOCKS
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
BEFORE "HOME" BLOCKS ARE WRITTEN
TYPE STR NAME FOR A LIST OF ITS PARAMETERS(CR IF NONE, ALL IF
ALL)
TYPE PHYSICAL UNIT NAME TO LIST ITS PARAMETERS(CR IF NONE, ALL IF
ALL)
DO YOU WANT TO CHANGE ANY DISK PARAMETERS(CR IF NO)
TYPE PHYSICAL UNIT TO WRITE HOME BLOCKS
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
ALL
HOME BLOCKS WRITTEN
STARTUP OPTION: GO
ONCE DIALOGUE Page 16-18
STARTUP OPTION: UNITID
.
.
.
16.7 STARTUP OPTION: CHANGE
The CHANGE option allows you to set or change selected parameters on
specified structures or units. The parameters are changed, sometimes
causing file structures to be refreshed, but no restructuring
(creating or deleting) of structures is performed. This option asks
you questions about the following:
1. File structure parameters
2. Physical unit parameters
3. Active swapping list
4. System search list
Note that if any changes require HOME blocks to be written or disks to
be refreshed, you are informed and asked for verification.
16.7.1 Frequency
After the file structure parameters have been established, they are
written onto each structure in two places called the HOME blocks.
Once this is completed, it need not be done again. When the monitor
is reloaded, the monitor reads the HOME blocks and sets up the disk
parameters automatically.
16.7.2 Summary of Steps
Steps 1 through 8 are described in Section 16.2.2 because they are
common to all STARTUP Options.
9. ONCE asks for your choice of
STARTUP OPTION:
USER RESPONSE: CHANGE
10. ONCE lists off-line units, if any, and then asks if off-line
units are to remain off-line or not.
%xxxx IS OFF-LINE
DO YOU WANT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN?
(TYPE #)
ONCE checks to see if there are any off-line controllers or
if there are any write-locked units. If so, appropriate
messages are transmitted to the operator terminal.
11. ONCE asks questions about file structure parameters, physical
unit parameters, the active swapping list, and the system
search list. You can accept the standard default answers by
responding to each question with a carriage return. (See
ONCE DIALOGUE Page 16-19
STARTUP OPTION: CHANGE
example 1 in Section 16.7.3.)
Or, you can answer the questions individually, thereby
setting parameters in a non-standard way. If you choose to
set the parameters in a non-standard way, the appropriate
subset of "long dialogue" questions is asked. (See examples
2 and 3 in Section 16.7.3 and the description of the LONG
option, Section 16.8.2.)
When the parameters have been set, control returns to the
STARTUP Option question. You can start the system via quick
or GO or continue with another option.
16.7.3 Examples
Example 1
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 12-19-77
WHY RELOAD: NEW
DATE: DEC 28
TIME: 1230
STARTUP OPTION: CHANGE
%DPA4 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPA5 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #9
2
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
TYPE STR NAME TO CHANGE ITS PARAMETERS(CR IF NONE, ALL IF ALL)
TYPE PHYSICAL UNIT NAME TO CHANGE ITS PARAMETERS(CR IF NONE, ALL
IF ALL)
DO YOU WANT TO CHANGE THE ACTIVE SWAPPING LIST?
DO YOU WANT TO CHANGE THE "SYS" SEARCH LIST?
STARTUP OPTION: QUICK
.
.
.
Example 2
ONCE DIALOGUE Page 16-20
STARTUP OPTION: CHANGE
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 12-19-77
DATE: DEC 28
TIME: 1250
STARTUP OPTION: CHANGE
TYPE STR NAME TO CHANGE ITS PARAMETERS(CR IF NONE, ALL IF ALL)
DSKX
AFTER EACH PRINTING OF CURRENT VALUE, TYPE NEW VALUE OR CR
PARAMETERS WHICH MAY BE CHANGED WITHOUT REFRESHING
# OF CONSECUTIVE BLOCKS TRIED FOR ON OUTPUT = 30
MIN = 1 MAX = 80000
SUM OF BLOCKS GUARANTEED TO USERS = 0
MIN = 0 MAX = 80000
# BLOCKS ALLOWED FOR OVERDRAW PER USER = 500
MIN = 0 MAX = 80000
1000
PARAMETERS WHICH MAY NOT BE CHANGED WITHOUT REFRESHING
K FOR CRASH.SAV = 96
MIN = 0 MAX = 256
0
BLOCKS PER CLUSTER = 5
MIN = 1 MAX = 511
THEREFORE BITS PER CLUSTER ADR. = 13
THEREFORE BLOCKS PER SUPER-CLUSTER = 5
THEREFORE SUPER-CLUSTERS PER UNIT = 8000
BITS PER CLUSTER COUNT = 9
MIN = 1 MAX = 18
THEREFORE BITS PER CHECKSUM = 14
TYPE STR NAME TO CHANGE ITS PARAMETERS(CR IF NONE, ALL IF ALL)
TYPE PHYSICAL UNIT NAME TO CHANGE ITS PARAMETERS(CR IF NONE, ALL
IF ALL)
DO YOU WNAT TO CHANGE THE ACTIVE SWAPPING LIST?
DO YOU WANT TO CHANGE THE "SYS" SEARCH LIST?
TYPE PHYSICAL UNIT TO WRITE HOME BLOCKS
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
DPA4
DPA5
HOME BLOCKS WRITTEN
%NEED REFRESHING:
DSKX
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
ONCE DIALOGUE Page 16-21
STARTUP OPTION: CHANGE
DSKX
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
STARTUP OPTION: Q
TO AUTOMATICALLY LOG-IN UNDER [1,2] TYPE "LOGIN"
Example 3
SYSCHK (N,Y): Y
MEMORY MAP =
FROM TO SIZE/K
000000 537777 176
CONTROLLER MTA IS UNACCESSIBLE
6.03 SYSTEM #160 12-02-77
WHY RELOAD: OTHER ;ONCE EXAMPLE
DATE: 4 DEC 1977
TIME: 46
STARTUP OPTION: CHANGE
%DPA6 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
%DPB0 IS OFF-LINE
DO YOU WANT IT TO BE 1)ON-LINE, 2)OFF-LINE, OR 3)DOWN? (TYPE #)
2
TYPE STR NAME TO CHANGE ITS PARAMETERS (CR IF NONE, ALL IF ALL)
DO YOU WANT TO CHANGE THE ACTIVE SWAPPING LIST?
Y
FOR EACH CLASS TYPED PHYSICAL UNIT NAMES(EXTRA CR WHEN DONE)
CLASS 0
FHA0
FHA1
CLASS 1
DPA0
CLASS 2
DPA0
?UNIT ALREADY IN ACTIVE SWAPPING LIST
DPB0
?UNIT HAS NO SPACE ALLOCATED FOR SWAPPING
DO YOU WANT TO CHANGE THE "SYS" SEARCH LIST?
Y
TYPE STR NAMES FOR "SYS" SEARCH LIST (EXTRA CR WHEN DONE)
DSKN
DSKA
DSKB
DSKC
TYPE PHYSICAL UNIT TO WRITE HOME BLOCKS
(CR IF NONE, ALL IF ALL ;"ALL" IS NORMAL CASE)
ONCE DIALOGUE Page 16-22
STARTUP OPTION: CHANGE
ALL
HOME BLOCKS WRITTEN
STARTUP OPTION: GO
6.03 SYSTEM #160 12-2-77
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
00:49:19(0)
.
00:49:20(0)
.
00:49:28(0)
16.8 STARTUP OPTION: LONG
The LONG option allows you to explicitly set all system parameters,
and to refresh and restructure the entire file system in a
non-standard way. The entire "LONG dialogue", as described in Section
16.8.3, is invoked. Note that the exact sequence of operator dialogue
depends on your answers to the individual questions. That is, the
answers to certain questions cause ONCE-Only to branch in different
directions. Section 16.8.3 follows the most commonly encountered
path.
If ONCE encounters problems in trying to start the monitor, the LONG
dialogue questions may be automatically invoked (with the QUICK and GO
options).
16.8.1 Frequency
As stated in Section 16.1, you should choose the LONG option only when
no other option or combination of options will suffice. For new
systems, the preferred method is for you to specify the DESTROY
option, followed by the CHANGE option (for individual exceptions to
the standard).
16.8.2 Example
This example assumes that you accept all system defaults, i.e., you
respond to all questions with a carriage return. (Section 16.8.3
illustrates a more complex situation.)
BOOTS
6.03
SYSCHK (N,Y):
6.03 SYSTEM #160 3-19-77
DATE: MAR 28
TIME: 1255
STARTUP OPTION: LONG
IN THE FOLLOWING DIALOG, ALL NUMBERS ARE DECIMAL.
TYPE <CR> IF OK, OR A NEW NUMBER TO CHANGE VALUE.
ONCE DIALOGUE Page 16-23
STARTUP OPTION: LONG
# MONITOR BUFFERS = 6
TYPE PHYSICAL UNIT NAME TO LIST # BAD REGIONS(CR IF NONE, ALL IF
ALL)
DSK FILE STRUCTURES(STRS):
DSKA:FHA0(XLZG),FHA1(GLZX)
DSKB:DPA1(ONC501),DPA0(ONC517)
DSKC:DPA2(RP03A)
DSKN:DPA3(2RP023)
DSKX:DPA4(ONC554),DPA4(ONC566)
DSKH:DPA6(2RP)
UNITS IN ACTIVE SWAPPING LIST:
FHA0(0),FHA1(0)
STRS IN "SYS" SEARCH LIST:
DSKA,DSKN,DSKB,DSKC
TYPE STR NAME FOR A LIST OF ITS PARAMETERS(CR IF NONE, ALL IF
ALL)
TYPE PHYSICAL UNIT NAME TO LIST ITS PARAMETERS(CR IF NONE, ALL IF
ALL)
DO YOU WANT TO CHANGE ANY DISK PARAMETERS?(CR IF NO)
TYPE STR NAME TO BE REFRESHED(CR IF NONE, ALL IF ALL)
STARTUP OPTION: QUICK
6.03 SYSTEM #160 12:56:33 CTY
DSKN: System #160 INITIA types the
. text found in
. STR.TXT.
.
.LOGIN[1,2] ;This is typed
;automatically.
.R OPSER
[OPRPAF PROCESSING AUTO COMMAND FILE]
.
.
.
16.8.3 Complete Description of Each "Long Dialogue" Question
NOTE
It is possible for an experienced
operator to stop the typeout of a
lengthy ONCE-Only question. When you
type a character during ONCE-only
typeout, a ^O is echoed and the
remainder of the question is not typed.
(The space character is recommended as
ONCE DIALOGUE Page 16-24
STARTUP OPTION: LONG
the character for you to type to stop
output, because it is usually ignored as
input.)
1. Get the monitor from the disk with BOOTS or from magtape with
BOOTM. Refer to the earlier chapters of this manual.
2. The system compares the type of monitor (1040, 1050, 1055,
1070, 1080, 1090) and the type of processor (KA10, KI10, or
KL10) for compatibility. If incompatible, the following
message is typed and the monitor halts.
THIS MONITOR WAS BUILT FOR A (KA10, KI10, KL10) AND
WILL NOT RUN PROPERLY ON (KI10, KA10, KL10).
3. If the optional loadtime diagnostic SYSCHK is included in the
system software (determined by a system programmer during the
MONGEN dialogue), ONCE asks if you wish to run it at this
time
SYSCHK (N,Y):
USER RESPONSE: N or <CR> to skip the diagnostic
The loadtime diagnostic SYSCHK is a 5-second program that
ensures the accessibility of memory and all devices on a
configuration.
The system types the name of the monitor and the creation
date and asks for the reason for the reload. ONCE types
6.03 SYS #2 3-1-77
WHY RELOAD:
You respond with one of the following acceptable answers or
an abbreviation that uniquely describes it:
OPR
PARITY
POWER
STATIC
HARDWARE
NXM
HALT
LOOP
HUNG
PM
CM
SA
NEW
SCHED
OTHER
(See Section 16.9 for a detailed explanation of the WHY
RELOAD: question.)
4. The system types the creation date and asks for today's date
and time.
DATE: MAR 17 1977
ONCE DIALOGUE Page 16-25
STARTUP OPTION: LONG
Type the date as the name of the month (January - December)
and the numerical day of the month (1-31) in any order. (The
name of the month may be shortened as long as the
abbreviation is unique.) The year is optional. If included,
it must be a 4-digit number (1977) or a 2-digit (77)
abbreviation following the month and day. If omitted, the
creation date of the monitor is assumed. Examples of
acceptable input are:
SEPTEMBER 17 1977
SEPT 17
17 SEPT
S 17
If an incorrect format is detected, the ONCE-Only dialogue
reasks the question in more detail:
PLEASE TYPE TODAY'S DATE AS MM-DD-YY:
NOTE
It is very important that you type in the correct
date. If you type an incorrect date, user files will
be written with the wrong creation and access dates.
The automatic COMPIL, if the source file was created
later than .REL file, will break down. More
seriously, if your installation is purging files by
date created or accessed, users' files could be
incorrectly deleted. If you discover later that you
typed in the wrong date but the system has not
started running the null job yet (Step 32), you may
restart the monitor at 400, reload it, or use the SET
DATE command.
TIME: 0843<CR>
You type time based on a 24-hour clock, e.g., type
0843 <CR> for 8:43 AM, or 1345 <CR> for 1:45 PM.
If an incorrect format is detected, the ONCE-Only
dialogue reasks the question in more detail:
PLEASE TYPE TIME AS HHMM:
As with the date, it is very important that you type
in the correct time. You can change the time after
the monitor has started to timeshare using the SET
DAYTIM command. (Refer to the DECsystem-10 Operating
System Command Manual in the DECsystem10 Software
Notebook.) However, this practice is to be
discouraged because it confuses the creation times of
files and causes the accounting system to make
incorrect charges for connect time. You can reload
or restart the monitor if you detect the error before
Step 31 (null job running).
If, after loading the monitor (Step 0), you are not asked the
reason for the reload and the date and time, perform the
following:
a. Check that the NXM STOP is off. If not (this is a
ONCE DIALOGUE Page 16-26
STARTUP OPTION: LONG
common error and is easily recovered), set if off.
b. Push CONTINUE. ONCE goes back to Step 2. Now you may
respond to date and time queries.
If this procedure does not work, it may be that the monitor
was written without a starting address. Try the following:
a. Set the address switches to 000400.
b. Press STOP, RESET, START (in that order).
c. Go back to Step 1 and respond to DATE, WHY RELOAD, and
TIME queries.
5. ONCE checks to see if all of memory is accessible. The
amount of accessible memory is determined by scanning memory.
If the amount found is less than that specified when the
monitor was built (by a MONGEN question), the following
message is typed.
% MEMORY FROM XXXXXX TO YYYYYY IS OFF LINE
DO YOU WANT IT TO BE 1) ON LINE, OR 2) DOWN?
(TYPE #)
USER RESPONSE: 1 OR 2
If the memory is required, check to see that the memory banks
in question are properly selected and addressed, then type 1
after any adjustments. Memory is rescanned to check for the
required amount.
If the memory is not required, type 2. The memory is then
classified as down or unavailable, and is not used. Note
that on KA10-based systems (1040, 1050, and 1055), all
available memory should be contiguous (all memories
classified as down are at the top of core) for maximum system
performance, e.g., non-contiguous memory limits the maximum
job size and compromises the swapping algorithm.
Nevertheless, it is possible to run on a KA10 with
non-contiguous memory.
Note that after a system power failure, the memories that you
want on-line may not be in a ready condition. If the AW
(await request) light is not ON, open the front door of the
appropriate memory cabinet and push the RESET switch several
times. If the AW light does not go on, go to the rear of the
cabinet and power the memory OFF and then ON. If the light
still does not go on, type 1 anyway (because the indicator
bulb may be burnt out).
If the monitor repeats the message for the same memory,
deselect all ports on that memory and, if contiguous memory
is desired, switch the highest memory down into the deselect
memory. Be sure to check all ports and check for
interleaving.
If the highest memory is interleaved, you must change the
interleaving of the memories adjacent to the highest memory.
6. ONCE asks for the desired STARTUP Option
STARTUP OPTION:
ONCE DIALOGUE Page 16-27
STARTUP OPTION: LONG
You type one of the following, depending upon your needs.
(Refer to Sections 16.2 through 16.8 for a more detailed
explanation of each option.)
QUICK To start the system quickly without changing any
parameters. (Refer to Section 16.2.)
GO To check to see if everything is in order and to
start the system with a minimum of dialogue.
(Refer to Section 16.3.)
DESTROY To automatically restructure and refresh all disks
according to a predetermined set of standards.
(Refer to Section 16.4.)
REFRESH To explicitly refresh selected file structures
without changing other system parameters. (Refer
to Section 16.5.)
UNITID To change selected unit IDs without changing other
system parameters. (Refer to Section 16.6.)
CHANGE To explicitly set or change selected file
structure parameters. (Refer to Section 16.7.)
LONG To explicitly set all parameters and restructure
the file system in a non-standard way. (Refer to
Section 16.8.)
NOINITIA To start the system quickly without setting any
new parameters and without running INITIA (refer
to Section 16.2).
USER RESPONSE: LONG
7. ONCE lists the number of monitor buffers and accepts a
change.
# MONITOR BUFFERS = 6
USER RESPONSE: <CR> to leave at that number
Type n (decimal) to change the number of
buffers
The value you type is not written back onto the disk;
therefore, it lasts only for this load of the monitor. Note
that the value you type must be 2 or greater.
This parameter can be changed permanently by
a. Redefining symbol MBFN during the MONGEN dialogue or
b. Patching the location MBFNUM with EXEC DDT or with
FILDDT and resaving it.
ONCE DIALOGUE Page 16-28
STARTUP OPTION: LONG
8. ONCE reads both HOME blocks from each unit in system.
The ONCE-Only dialogue reads the HOME blocks from all disk
units that the Monitor was generated to handle. If a unit is
write protected, the ONCE-Only dialogue types:
%DPA2 IS WRITE PROTECTED
DO YOU WANT IT TO BE 1) WRITE-ENABLED, OR
2) WRITE-PROTECTED? (TYPE #)
USER RESPONSE: 1
Set the unit properly. If you do not wish to initialize the
HOME blocks, turn the unit off line. After the HOME blocks
have been initialized the first time, the monitor can be
started with the unit write protected.
9. ONCE lists HOME block consistency errors and asks to
initialize. If a disk unit has not been refreshed previously
(true for the first time Monitor startup) or has been written
on by a test or maintenance program, the ONCE-Only dialogue
types:
?FHA0 FIRST HOME BLOCK CONSISTENCY ERROR
?FHA0 SECOND HOME BLOCK CONSISTENCY ERROR
DO YOU WANT TO INITIALIZE THE HOME BLOCKS
ON THIS UNIT?
USER RESPONSE: Y
If only one of the two HOME blocks has a consistency errors,
you type N<CR> and the data in the good HOME block will be
used. If both HOME blocks have consistency errors, you must
dissolve the file structure, redefine, and refresh. You type
Y<CR> to this question and continue with the LONG dialogue.
If there are any off-line controllers, the ONCE-Only dialogue
might type:
%CONTROLLER FHA IS OFF-LINE
DO YOU WANT IT TO BE 1) ON-LINE, OR
2) DOWN? (TYPE #)
USER RESPONSE: 1 or 2
You must change the controller if it was not set properly.
If you type anything other than 2, the monitor tries the
controller again. If you specify that the controller is
down, the monitor does not attempt to use any of the units on
the controller.
ONCE DIALOGUE Page 16-29
STARTUP OPTION: LONG
If there are any on-line controllers, ONCE may type:
%CONTROLLER DPAn WRITE-HEADER LOCKOUT SWITCH
ALLOWS WRITING HEADERS.
DO YOU WANT IT TO BE 1) SET OR
2) IGNORED (TYPE #)
USER RESPONSE: 1
If the write-header lockout switch is OFF, it allows
formatting of disk packs under timesharing. If it is OFF,
turn the switch to ON and type 1. This switch must be ON to
start the monitor.
If you type 2 (ignored), ONCE types
NOT NORMALLY DONE, ARE YOU SURE?
as a warning. You type YES if you are sure, and NO if you
wish to reanswer the previous question.
If there are any off-line units on the controller, the
ONCE-Only dialogue might type:
%FHA1 IS OFF-LINE
DO YOU WANT IT TO BE 1) ON-LINE,
2) OFF-LINE, OR 3) DOWN? (TYPE #)
USER RESPONSE: 1, 2, or 3.
You must change the unit if it has not been set properly. If
you type anything other than 2 or 3, the monitor tries the
unit again. If you specify that the unit is down, the
monitor cannot access it again, unless you issue the ATTACH
command to put the unit off-line. The MOUNT command can be
used to place the unit in the monitor's pool of available
devices. You can also issue the DETACH command to declare a
unit as down, which removes the unit from the monitor's pool
of available devices. You can cause a disk pack unit to be
off-line without powering it down. Simply set the
enable/disable switch to disable the unit. If the monitor is
built to handle more units than are physically attached, no
harm is done. There is only the minor annoyance of answering
this question with 3 every time the Monitor is started (LONG
dialogue). Software unit data blocks (UDBs) can be easily
patched out of the system with EXEC DDT or FILDDT to avoid
the above questions because the unit data blocks (e.g.
RPA3CB) are linked through the LH of relative location
UNISYS. (Refer to PATMON.)
On subsequent startups, one of the following error messages
may occur:
?MORE THAN ONE LAST UNIT IN ACTIVE SWAPPING LIST
?MORE THAN ONE LAST UNIT IN STR XXXX
?NO UNITS IN ACTIVE SWAPPING LIST
ONCE DIALOGUE Page 16-30
STARTUP OPTION: LONG
%LAST UNIT IN ACTIVE SWAPPING LIST NOT FOUND
?CHANGE THE SLAVE OFFSET SWITCH IN BAY 2
?TWO LOGICAL UNIT N'S FOUND IN ACTIVE SWAPPING LIST
%LOGICAL UNIT N MISSING FROM ACTIVE SWAPPING LIST
?LAST UNIT WASN'T FOUND IN STR XXXX
?TWO LOGICAL UNIT N'S FOUND IN STR XXXX
?LOGICAL UNIT N MISSING FROM STR XXXX
?TWO LOGICAL STR N'S FOUND IN "SYS" SEARCH LIST
%LOGICAL STR # N MISSING FROM SYS SEARCH LIST
?NO STR'S IN SYS SEARCH LIST
You must change parameters and/or mount or dismount packs to
get rid of the error condition. It is not an error to have
units that are not in any file structure. So the message
UNITS NOT IN A FILE STRUCTURE:
does not need corrective action. In fact, the swapping unit
in a large system may be entirely dedicated to swapping. In
this case, it is recommended that the swapping unit not be
included in a file structure because the space for the
skeleton file structure can be saved. Note, however, that
the first 12 decimal blocks must still be used for HOME and
BAT blocks.
10. ONCE lists BAT block consistency errors:
?DPA0 FIRST BAT BLOCK CONSISTENCY ERROR
?DPA0 SECOND BAT BLOCK CONSISTENCY ERROR
DO YOU WANT TO INITIALIZE THE BAT BLOCKS ON
THIS UNIT?
USER RESPONSE: N
(Type Y if the BAT block has not been written on by the
mapping option of a maintenance program.)
If only one BAT block has consistency errors, you type N
because the other BAT block is probably all right. If both
BAT blocks have consistency errors, you type Y. ONCE
responds with
NOT NORMALLY DONE, ARE YOU SURE?
USER RESPONSE: Y
ONCE responds with INITIALIZING BAT BLOCKS.
ONCE asks for unit names on which to list the number of bad
regions from the BAT blocks.
ONCE DIALOGUE Page 16-31
STARTUP OPTION: LONG
TYPE PHYSICAL UNIT NAME TO LIST BAD REGIONS
(<CR> IF NONE, ALL IF ALL)
USER RESPONSE: <CR>
Usually there are none, but if you specify a physical name or
ALL, then ONCE types
DPA2 (2RP005)
# BAD REGIONS=xxxxx
#BAD BLOCKS=XXXXX
11. ONCE lists STRs and units for disk packs in them.
The ONCE-Only dialogue types out each file structure name
followed by the physical unit name of each unit in the file
structure in its logical order. The unit ID is typed in
parentheses after each physical unit name. Example:
DSK FILE STRUCTURES (STRS):
DSKA:FHA0(RM10A),FHA1(RD10A)
DSKB:DPA1(DP0KZQ),DPA0(DP2KMR),DPA2(DP1RWP)
If there are no file structures, only the header line is
typed, because you must take action.
12. ONCE lists units not in an STR, if any.
The ONCE-Only dialogue types out each physical unit not in a
file structure followed by the unit ID. This list always
includes the off-line and down units. Parentheses mean the
unit name is zero. Example:
UNITS NOT IN A FILE STRUCTURE:
FHA2(),DPA3(DP35WN)
If there are no units in the file structures, the heading is
not typed.
13. ONCE lists units in the active swapping list.
The ONCE-Only dialogue types out all physical units in the
active swapping list, along with the swapping class of each
unit. Class0, class1 ... classn describe different
varieties of swapping space. Lower number swapping spaces
should contain high-speed devices because all lower class
space is used first.
UNITS IN ACTIVE SWAPPING LIST:
FHA(0),DPA1(1)
If there are no units in the active swapping list, the
ONCE-Only dialogue types:
?NO UNITS IN ACTIVE SWAPPING LIST
ONCE DIALOGUE Page 16-32
STARTUP OPTION: LONG
This typeout is to be expected on the first monitor startup.
This situation must be corrected by including at least one
unit in the active swapping list.
14. ONCE lists STRs in the system search list.
The ONCE-Only dialogue types out all file structure names in
the system search list for device SYS. SYS should include a
fast and a slow file structure. Most .EXE files should be
put on the slower file structure, because the dormant
segments are kept on the faster swapping units. Other active
files such as accounting files should also be considered as
candidates for a faster file structure within SYS.
Example:
STR's IN "SYS SEARCH LIST:
DSKA,DSKB
(There will be none the first time.)
15. ONCE asks for an STR name to list its parameters.
TYPE STR NAME FOR A LIST OF ITS PARAMETERS (CR IF NONE,
ALL IF ALL)
USER RESPONSE: <CR>
16. ONCE asks for a unit name to list its parameters.
TYPE PHYSICAL UNIT NAME TO LIST ITS PARAMETERS (CR IF
NONE, ALL IF ALL)
USER RESPONSE: <CR>
17. ONCE asks if you want to change anything.
BEFORE "HOME" BLOCKS ARE REWRITTEN,
DO YOU WANT TO CHANGE ANY DISK PARAMETERS?
USER RESPONSE: Y
(You wish to define some file structures.)
If you type <CR>, ONCE goes to Step 25 (ask about
refreshing).
18. ONCE asks if you want to dissolve any STRs.
TYPE STR NAME TO BE DISSOLVED (CR IF NONE, ALL IF ALL)
USER RESPONSE: <CR>
(There are none to dissolve.) If you want to redefine a file
structure on a subsequent startup, you must first dissolve
all STRs that have the units you wish to include in the new
STR. If there are no file structures, which is the case the
first time ONCE-Only is run, this question is omitted.
19. ONCE asks for a STR to be defined.
TYPE STR NAME TO BE DEFINED (CR IF NONE)
ONCE DIALOGUE Page 16-33
STARTUP OPTION: LONG
USER RESPONSE: DSKA
(Define the fastest file structure first.)
ONCE asks for the names of units to be in the file structure.
TYPE NAMES OF PHYSICAL UNITS IN STR (ALL IF ALL, EXTRA
CR WHEN DONE)
USER RESPONSE: FHA0 <CR>
FHA1 <CR>
<CR>
If you type a name that is not a physical unit name, this
incorrect name is ignored and you receive the message
?NOT A PHYSICAL UNIT - TRY AGAIN
If you have two units (RM10B or RD10 or a mixture) on the
RC10, follow with an extra carriage return to indicate the
end of file structure DSKA. If there are a small number of
units of a particular type, say two or three, define them to
be in one file structure. If your installation has a large
number of units you can define more than one file structure
for that controller type. Remember that it is possible to
put RM10B and RD10 in the same file structure. You may wish
to leave an RM10B out of any file structure and use it solely
for swapping, thereby eliminating most STR overhead.
If there are any physical units not yet defined to be in a
structure, ONCE asks for another STR to be defined. If not,
then ONCE goes directly to question 20.
TYPE STR NAME TO BE DEFINED (CR IF NONE)
USER RESPONSE: DSKB
(Type in the second fastest file structure name.)
ONCE asks for units to be in this file structure.
TYPE NAMES OF PHYSICAL UNIT IN STR (ALL IF ALL, CR WHEN
DONE)
USER RESPONSE: DPA0 <CR>
DPA1 <CR>
DPA2 <CR>
<CR>
If you have, for example, three units on the RP10.
ONCE asks for another STR to be defined.
TYPE STR NAME TO BE DEFINED (CR IF NONE)
USER RESPONSE: <CR>
If all desired units have been defined to be part of some
file structure.
ONCE DIALOGUE Page 16-34
STARTUP OPTION: LONG
20. ONCE asks if you want to change any STR parameters.
TYPE STR NAME TO CHANGE ITS PARAMETERS (CR IF NONE, ALL
IF ALL)
USER RESPONSE: DSKA
(Because you have not defined any parameters yet, you could
type ALL to define all.)
AFTER EACH PRINTING OF CURRENT VALUE, TYPE NEW VALUE OR
CR
PARAMETERS WHICH MAY BE CHANGED WITHOUT REFRESHING
# OF CONSECUTIVE BLOCKS TRIED FOR ON OUTPUT=n
MIN=1 MAX=m
where n is 10 for RC10 file structures or
30 for RP10 file structures
and m is the size of the STR.
USER RESPONSE: <CR>
SUM OF BLOCKS GUARANTEED TO USERS=0
MIN=0 MAX=m
where m is the number of blocks on the STR. Because
reserve quotas are not implemented for the 603 release,
this number is not used.
USER RESPONSE: <CR>
# BLOCKS ALLOWED FOR OVERDRAW PER USER=n
MIN=0 MAX=m
where n is 200 for RC10 file structures
500 for RP10 file structures
and m is max. number of blocks in STR.
USER RESPONSE: <CR>
The overdraw amount should be as large as a source file
so that editing with TECO can be completed while
exceeding a user's quota, but before the overdraw
amount runs out.
PARAMETERS WHICH MAY NOT BE CHANGED WITHOUT REFRESHING
K FOR CRASH.EXE=n
MIN=0 MAX=4096
n is determined by the larger of the MONGEN-specified
size or the actual size of the machine.
USER RESPONSE: n or <CR>
ONCE DIALOGUE Page 16-35
STARTUP OPTION: LONG
Where n is the core size of the machine on which the
monitor is running. (CRASH.EXE is used by BOOTS to
dump a core image when the monitor crashes.)
BLOCKS PER CLUSTER=n
MIN=1 MAX=511
where n = 1 for RC10 file structures
5 for RP10 file structures. (10 is also supported for
RP10, RHx0, and RP0x file structures.)
USER RESPONSE: <CR>
THEREFORE BITS PER CLUSTER ADR.=13
THEREFORE BLOCKS PER SUPER CLUSTER=m
THEREFORE SUPERCLUSTERS PER UNIT=n
For RP02, m=5 n=8000
For RM10B, m=1, n=2700
For RD10, m=1, n=4000
For RP03, m=5, n=16000
For RP04, m=10, n=15428
For RP06, m=10, n=30780
BITS PER CLUSTER COUNT=9
MIN=1 MAX=18
USER RESPONSE: <CR>
This gives up to 511 clusters in one retrieval pointer.
THEREFORE BITS PER CHECKSUM=14
21. ONCE asks if you want to change any other STR parameters.
TYPE STR NAME TO CHANGE ITS PARAMETERS (CR IF NONE, ALL
IF ALL)
USER RESPONSE: Next STR name to be changed (e.g.,
DSKB)
or
<CR> if none
Repeat question 21. for each STR.
22. ONCE asks if you want to change any unit parameters.
TYPE PHYSICAL UNIT NAME TO CHANGE ITS PARAMETERS (CR IF
NONE, ALL IF ALL)
USER RESPONSE: ALL
[The first time.] The ONCE-Only dialogue types out one
physical unit name. It repeats the questions for each unit.
AFTER EACH PRINTING OF CURRENT VALUE, TYPE NEW VALUE OR
CR
(Then the monitor types out the first physical device name.)
ONCE DIALOGUE Page 16-36
STARTUP OPTION: LONG
PARAMETERS WHICH MAY NOT BE CHANGED WITHOUT REFRESHING
ZERO UNIT ID-NEW ID NEEDED
USER RESPONSE: RD001
The unit ID may be up to six characters in length. It is
suggested that a unit ID indicate the type of unit. For
example,
RD000, RD001, RD002, RD003, (FOR RD10s)
RM000, RM001, RM002, RM003 (FOR RM10Bs)
2RP000, 2RP001, 2RP002, 2RP003, 2RP004, 2RP005,
2RP006 (FOR RP02s)
Each pack at an installation should have a unique ID.
K FOR SWAPPING ON UNIT=0
MIN=0 MAX=p
For RP02 p=4095. For RM10B or RD10 p=337. For RP03 p=8191.
USER RESPONSE: N
Where n is about 15 times the number of jobs the system is
built for if this is the first drum unit or 0 times for all
other units.
You may wish to allocate some space for swapping on another
unit in case the first unit goes down. A unit will not be
used for swapping unless it is in the active swapping list.
If n is greater than maximum, the message (for RP02).
?CANNOT EXCEED 4095
The maximum (p) may be too large. If the unit is in a file
structure, it is suggested that p-2 be the largest number
specified.
NOTE
There is no requirement that a unit used for swapping
be part of a file structure. By not making it part
of a file structure, the space for the skeleton file
structure can be saved. Note that the HOME and BAT
blocks must still be present (Blocks 1, 2 and 10, 11
decimal).
COMPUTED 1ST LOGICAL BLOCK FOR SWAPPING=YYYYY
1ST LOGICAL BLOCK FOR SWAPPING=YYYYY
MIN=13 MAX=YYYY
USER RESPONSE: <CR>
The ONCE-Only program has computed the best place for
the swapping space by putting it in the middle of the
unit. If there are any bad spots in that region, you
should move the swapping space down by typing a
smaller logical block number. The question is
skipped if no swapping space was allocated.
ONCE DIALOGUE Page 16-37
STARTUP OPTION: LONG
If YYYYY is too low, so that it would interfere with
the HOME and BAT blocks, the ONCE-Only dialogue will
type:
MUST EXCEED 11
If YYYYY is too high, so that it would run past end,
the ONCE-only dialogue will type:
CANNOT EXCEED YYYYY
# SAT BLOCKS ON UNIT=t
MIN=t MAX=223
Where t is the minimum.
For RP02, t=2. For RM10B or RD10, t=1. For RP03,
t=4. For RP04, t=4. For RP06, t=7.
USER RESPONSE: <CR>
Then the ONCE-Only dialogue types
THEREFORE CLUSTERS PER SAT=xxxx
THEREFORE WORDS PER SAT=yyy
If the cluster size is smaller than the recommended size,
more than one SAT block per unit will be required. If you
type a number that is too small, the dialogue responds with
TOO SMALL - MIN =N
and repeats the question.
PARAMETERS WHICH MAY BE CHANGED WITHOUT
REFRESHING
# SAT BLOCKS IN CORE=M
MIN=1 MAX=m
USER RESPONSE: m
The suggested words per SAT block have been chosen so that
each unit has only one or two SAT blocks. It is faster to
have all SAT blocks in memory. If the number of SAT blocks
on this unit is 1, this question is skipped. If ALL had been
typed or another unit is to be changed, ONCE goes back to
beginning of question 18.
23. ONCE lists units in the active swapping list and asks if you
want to make any changes.
UNITS IN ACTIVE SWAPPING LIST:
FHA0 (0), DPA0 (1)
DO YOU WANT TO CHANGE THE ACTIVE SWAPPING
LIST?
USER RESPONSE: Y
ONCE DIALOGUE Page 16-38
STARTUP OPTION: LONG
(You must define an active swapping list if this is initial
system startup.)
FOR EACH CLASS TYPE PHYSICAL UNIT NAMES (CR
WHEN DONE)
CLASS 0
USER RESPONSE: FHA0
If there are more units in Class 0, type FHA1. Type an extra
carriage return when finished with Class 0.
CLASS 1
USER RESPONSE: DPA0 if swapping space is allocated,
or
<CR>
NOTE
1. Class 0, class 1,....class n describe the
different varieties of swapping space. Lower
numbered swapping space should contain higher speed
devices because all lower class space is used first.
NOTE
2. The small monitor (DECsystem-1040) allows only
one file structure and one unit for swapping. After
the question
UNITS IN ACTIVE SWAPPING LIST:
CLASS 0
DPA0
ONCE-Only types
%SWAPPING LIST FULL
With the small monitor this message is expected and
should not alarm you.
24. ONCE lists STRs in the system search list and asks if you
have any changes.
STRS IN "SYS" SEARCH LIST
DSKA,DSKB
DO YOU WANT TO CHANGE THE "SYS" SEARCH LIST?
USER RESPONSE: Y
(You must define a system search list if this is the initial
system startup.)
ONCE DIALOGUE Page 16-39
STARTUP OPTION: LONG
TYPE STR NAMES FOR "SYS" SEARCH LIST (EXTRA CR WHEN
DONE)
USER RESPONSE: DSKA
DSKB
<CR>
The fastest file structure name (e.g., DSKA) followed by a
slower one and an extra carriage return.
25. ONCE asks to list STR parameters.
BEFORE "HOME" BLOCKS ARE WRITTEN
TYPE STR NAME FOR A LIST OF ITS PARAMETERS (CR IF NONE,
ALL IF ALL)
USER RESPONSE: <CR>
You will be able to check after HOME blocks are written. To
check now type structure names or ALL.
26. ONCE asks to list unit parameters.
TYPE PHYSICAL UNIT NAME FOR A LIST OF ITS
PARAMETERS (CR IF NONE, ALL IF ALL)
USER RESPONSE: <CR>
You are able to check later.
27. ONCE asks if you want to change anything.
DO YOU WANT TO CHANGE ANY DISK PARAMETERS?
(CR IF NO)
USER RESPONSE: <CR>
28. Asks for the HOME blocks to be written.
TYPE PHYSICAL UNIT TO WRITE HOME BLOCKS (CR IF NONE,
ALL IF ALL; "ALL" IS NORMAL CASE)
USER RESPONSE: ALL
Under normal circumstances ALL is the only acceptable answer
to this question. <CR> can be used during software
development if you are experimenting and do not want to
change the HOME blocks permanently. (Any other answer can
cause problems.)
ONCE types
HOME BLOCKS WRITTEN
If an error condition exists, ONCE types a message and goes
back to question 8. If there is no error, ONCE lists the
structures defined by the HOME blocks and indicates those
that need refreshing. ONCE then lists the units that are not
in a file structure, the units that are in the swapping list
above, and the STRs in the "SYS" search list. ONCE then
lists the parameters for STRs and physical units if you
request.
ONCE DIALOGUE Page 16-40
STARTUP OPTION: LONG
29. ONCE asks for the STR name to be refreshed.
DO YOU WANT TO CHANGE ANY DISK PARAMETERS
If ONCE determines that one or more structures needs
refreshing, it will type the message:
%NEED REFRESHING:
aaa
where aaa are the STRs that need refreshing.
TYPE STR NAME TO BE REFRESHED (CR IF NONE,
ALL IF ALL)
USER RESPONSE: ALL
(The first time you wish to refresh everything.)
30. ONCE asks you to LOGIN.
TO AUTOMATICALLY LOGIN UNDER [1,2] TYPE LOGIN
USER RESPONSE: LOGIN
31. The Monitor starts running the null job and initiates
timesharing.
32. If the LONG option is run, post the output from questions 10
through 27 on the wall as the latest disk configuration.
33. When building a new monitor return to Chapter 5 for
instructions on restoring files to the disk.
16.9 WHY RELOAD QUESTION
As a part of the loading procedure the ONCE-Only dialogue asks for the
reason for the reload. The question
WHY RELOAD:
is typed on your terminal, just before the questions asking for the
date and time. You must specify your reason for reloading the monitor
by typing one of the answers described in this section, or an
abbreviation that uniquely describes it. If more than one answer
applies, you type the one that appears first in the list. If you wish
to include information related to your answer, e.g., circumstances
surrounding a system crash, you may do so by including the text on the
same line as the answer. The entire string of text is first stored in
the crash AC area and then copied to the log file by DAEMON.
The possible answers to the WHY RELOAD: question fall into four
categories:
System problem
Stand-alone time
New or different monitor
Other
ONCE DIALOGUE Page 16-41
WHY RELOAD QUESTION
1. System Problem
If you are reloading the monitor as a result of a system
problem or crash, you identify the cause of the problem
with one of the following answers.
Answer Meaning
------ -------
OPR Operator error
PARITY Memory parity stop or halt
POWER Power failure on some device
STATIC Static electricity
HARDWARE Hardware malfunction
NXM Non-existent memory stop
HALT STOPCD or halt
LOOP Monitor loop in executive mode
HUNG No response
2. System Stand-Alone
If you are reloading the monitor to get timesharing services
back on the air, you type one of the following answers.
Answer Meaning
------ -------
PM Preventive maintenance. Stand-alone
that was required for regularly
scheduled maintenance procedures.
CM Corrective maintenance. Stand-alone
that was required for unscheduled
maintenance procedures.
SA Other stand-alone projects, such as
programmer debugging session or
reformatting of disk packs.
3. New or Different Monitor
If you are loading a new or a different monitor, you choose
one of the following answers.
Answer Meaning
------ -------
NEW New monitor, e.g., monitor with an added
feature or improvement via a patch.
SCHED A scheduled change from one monitor to
another, e.g., a change from a dual
system monitor to a single system
monitor and back.
ONCE DIALOGUE Page 16-42
WHY RELOAD QUESTION
4. Other
If none of the above reasons applies, you type OTHER.
16.10 ERROR MESSAGES
The following error messages may be typed on your terminal during the
ONCE-Only dialogue.
%ACTIVE SWAPPING LIST FULL
An attempt was made to specify more units to the active
swapping list than the monitor tables can handle. The
current limit is 8. If you need more swapping space, you
should increase the amounts on the eight units already
specified.
?CANNOT EXCEED # BLOCKS IN FILE STRUCTURE = n
The number of disks blocks specified for reserved quotas or
the number of disk blocks specified for the overdraw amount
was too large. You should type in a number less than or
equal to n.
?CANNOT EXCEED # SAT BLOCKS ON UNIT = n
The number of SAT blocks in memory cannot exceed the number
of SAT blocks (n) on the unit. You should type a number
less than or equal to n.
%CONTROLLER DPA HAS WRITE HEADER LOCKOUT SWITCH OFF
DO YOU WANT IT TO BE 1) ON-LINE, 2) OFF-LINE, OR
3) DOWN? TYPE#
?CPU0 FAST AC'S ARE OFF (FM ENB)
In a multiprocessing system, the monitor halts after this
message because it is not possible for a multiprocessing
system to run without the fast ACs (split memory cycle is
enabled). You must enable the fast ACs by performing the
following steps:
1. Push STOP.
2. Turn FM ENB off.
3. Set ADDRESS switches to 400.
4. Push START.
On a single processor system, the above is an advisory
message. The system runs considerably slower without the
fast ACs, so you should enable them as indicated in steps
1-4 above.
ONCE DIALOGUE Page 16-43
ERROR MESSAGES
%DPA IS OFF-LINE
DO YOU WANT IT TO BE 1) ON-LINE OR 2) DOWN? (TYPE#)
Controller DPA (RP10) is off line. You should check the
settings of all switches in RP10 bay. All switches should
be down. After changing the switches, you should type 1.
If you do not want the monitor to use the controller, you
should type 2. This also applies to DPB.
%DPA0 IS OFF-LINE
DO YOU WANT IT TO BE 1) ON-LINE, 2) OFF-LINE, OR
3) DOWN? (TYPE#)
You should check the START/STOP rocker switch and the
ENABLE/DISABLE switch on the individual disk pack unit.
They should be in the normal position with the top of the
switch in. After changing switches, you should type 1. If
you do not want the monitor to use the unit, you should type
2. This message also applies with
DPA1,DPA2,...,DPA7,DPB0,DPB1,...,DPB7.
%DPA0 IS WRITE PROTECTED. DO YOU WANT IT TO BE
1) WRITE-ENABLED, OR 2) WRITE-PROTECTED?
Disk pack unit DPA0 is on line, but is write protected. If
you wish it to remain this way, you should type 2.
Otherwise, you should set the READWRITE/READ ONLY rocker
switch to normal (top of switch in) and then type 1. This
message also applies with DPA1,...,DPA7,DPB0,DPB1,...,DPB7.
?ERROR ON CPU1
?CHANGE THE CPU1 MA TRP OFFSET SWITCH IN BAY1
In a multiprocessing system, the monitor halts after this
message is printed at CTY (CPU0). At the CPU1 console,
behind the door of bay 1, change the setting of the MA TRP
OFFSET switch. Then push CONT on CPU1. (CPU0 is not
affected by this error and continues timesharing.)
?ERROR ON CPU1
?FAST AC'S TURNED OFF - SET FM ENB
In a multiprocessing system, CPU1 halts after this message
is typed at CTY (CPU0). At CPU1 console set FM ENB by
rocking the FM ENB switch (on the maintenance panel behind
the door). Then push CONT at CPU1. (CPU0 is not affected
by this error and continues timesharing.)
%FHA IS OFF-LINE
DO YOU WANT IT TO BE 1) ON-LINE OR 2) DOWN? (TYPE #)
Controller FHA (RC-10) is off line. You should check
settings of all switches in RC-10 bay. All switches should
ONCE DIALOGUE Page 16-44
ERROR MESSAGES
be down. After changing switches, you should type 1. If
you do not want the monitor to use the controller, you
should type 2. Also applies to FHB.
%FHA0 IS OFF-LINE
DO YOU WANT IT TO BE 1) ON-LINE, 2) OFF-LINE, OR
3) DOWN? (TYPE #)
You should check the unit dial selectors. One of them (DISK
A, DISK B, DISK C, or DISK D) should be set to 0. You
should set the switches for all the units you have to 0, 1,
2, or 3. The other units should be OFF. You should not
touch any dials that are dialed to numbers numerically less
than the one just typed out, because the monitor has already
read these units. After changing the switches and dials,
you should type 1. If the unit is temporarily down and will
be fixed while the system runs, you should type 2. In all
other cases, you should type 3. This could apply to
FHA1,...,FHA3,FHB0,...,FHB3.
FIRST BAT BLOCK CONSISTENCY ERROR
The ONCE-Only dialogue has discovered that the first of two
redundant BAT blocks does not contain some of the data
normally expected in a BAT block. This is not a fatal error
because the other BAT block is probably all right. If both
BAT blocks have this error, you should initialize the BAT
blocks. This error may occur if some of the diagnostics are
run.
FIRST BAT BLOCK HARDWARE ERROR
The ONCE-Only dialogue has had a hardware error while
reading the first of two redundant BAT blocks. Because
there is another BAT block, this error is usually not fatal.
The controller status is put in the console lights.
FIRST HOM BLOCK CONSISTENCY ERROR
The ONCE-Only dialogue has discovered that the first of two
redundant HOME blocks does not contain some of the data
normally expected in a HOME block. Therefore, none of the
data should be considered valid. This is not a serious
error since the other HOME block is usually all right. If
both HOME blocks have consistency errors, you have to
dissolve the file structure, redefine it, and refresh it.
FIRST HOM BLOCK HARDWARE ERROR
The ONCE-Only dialogue has had a hardware error while
reading or writing the first of two redundant HOME blocks.
This is not fatal because there is another HOME block. The
controller status is put in the console lights, and the
ONCE DIALOGUE Page 16-45
ERROR MESSAGES
controller is left in its error condition.
?LAST UNIT WASNT FOUND IN STR DSKn
The last unit in file structure is missing. You should
check to see that all the proper packs are mounted and
on-line. If not, you should remount them and restart the
monitor at 140. Otherwise, you have to dissolve the file
structure, redefine it, and then refresh it, thereby
destroying any data already on the unit.
%LOGICAL STR # n MISSING FROM THE "SYS" SEARCH LIST
A file structure is missing from the SYS search list. This
condition need not be corrected, because the monitor will
skip the missing file structure. To avoid the message in
the future, you should change the system search list when
asked.
%LOGICAL UNIT n MISSING FROM ACTIVE SWAPPING LIST
A unit is missing from the active swapping list. This can
happen if a unit is off line or down. This error need not
be corrected since the monitor will order the swapping list
accordingly.
?LOGICAL UNIT n MISSING FROM STR DSKn
A unit is missing from a file structure and must be
remedied. You should check that all proper packs are
mounted and on line. If this is not so, you should add the
proper packs and restart the monitor at 140. Otherwise, you
have to dissolve the file structure, redefine it, and
refresh it, thereby destroying any data already on the unit.
%MEMORY FROM xxxxx TO yyyyy IS OFFLINE
DO YOU WANT IT TO BE 1) ONLINE OR 2) DOWN? (TYPE #)
If the specified memory is supposed to be off line, type 2.
Otherwise, push the reset button at the rear of the
memories. Check to see that the memory is selected and all
required ports are ON. After all switches are in order,
return to the CTY and type 1. If the error message is
repeated, attempt to switch this memory out of the system
and switch the top memory into its place.
?MORE THAN ONE LAST UNIT IN ACTIVE SWAPPING LIST
The active swapping list specified in the disk unit HOME
blocks has more than one unit as the last one. You should
redefine the units in the active swapping list to correct
ONCE DIALOGUE Page 16-46
ERROR MESSAGES
this situation.
?MORE THAN ONE LAST UNIT IN STR DSKn
The file structure has more than one unit specified as the
last unit as recorded in the disk HOME blocks. You should
dissolve the file structure and redefine it.
?NO UNITS IN ACTIVE SWAPPING LIST
None of the on-line units is in the active swapping list.
Because there must be swapping space, you must change the
active swapping list to include a unit that has some
swapping space. If there are no units with swapping space,
you must define swapping space on a unit not in a file
structure. If all units are in file structures, you must
refresh a file structure, define the necessary swapping
space, and redefine the active swapping list.
NOT NORMALLY DONE, ARE YOU SURE?
This warning message is printed when you respond to certain
important questions in an unexpected or non-standard manner.
It allows you to verify your choice of answer or recover
from a typographical error before serious damage is done.
%PROBLEM ON CPU1
CPU1 has halted. Restart it by setting data switches to 400
and pushing STOP, RESET, START. Note that the above message
is printed once a minute until CPU1 is started. You can use
the OPSER command :SET CPU to stop scheduling CPU1 and,
consequently, stop the message.
SAT BLOCK HARDWARE ERROR
The ONCE-Only dialogue has had a hardware error while
reading one of the SAT blocks.
SECOND BAT BLOCK CONSISTENCY ERROR
The ONCE-Only dialogue has discovered that the second of two
redundant BAT blocks does not contain some of the data
normally expected in a BAT block. This is not a fatal error
because the other BAT block is probably all right. If both
BAT blocks have this error, you should initialize the BAT
blocks. This error may occur if some of the diagnostics are
run.
ONCE DIALOGUE Page 16-47
ERROR MESSAGES
SECOND HOM BLOCK CONSISTENCY ERROR
The ONCE-Only dialogue has discovered that the second of two
redundant HOME blocks does not contain some of the data
normally expected in a HOME block. Therefore, none of the
data should be considered valid. This is not a serious
error since the other HOME block is usually all right. If
both HOME blocks have consistency errors, you have to
dissolve the file structures, redefine and refresh.
SECOND HOM BLOCK HARDWARE ERROR
The ONCE-Only dialogue has had a hardware error while
reading or writing the second of two redundant HOME blocks.
This is not fatal since there is another HOME block. The
controller status is put in the console lights, and the
controller is left in its error condition.
?THIS MONITOR WAS BUILT FOR A xxxxx AND
WILL NOT RUN PROPERLY ON A yyyyy.
xxxxx and yyyyy can be PDP-6, KA10, KI10, or KL10. You have
loaded the wrong monitor. You must try another. If, after
several tries, the monitor appears to be the correct one,
then you must rerun MONGEN and change the answer to the
question asking for the type of processor. Refer to Chapter
9, question 3 in Section 9.4.
?TOO SMALL - MIN. #=X
An answer to the ONCE-Only dialogue or a default value is
too small. Type in an answer greater than or equal to X.
?TWO LOGICAL UNIT n's FOUND IN ACTIVE SWAPPING LIST
The active swapping list has more than one unit in the same
position. You must redefine the active swapping list.
?TWO LOGICAL UNIT n's FOUND IN STR DSKn
Two units are marked to be in the same logical position in
the file structure. This happens only if two different file
structures have been given the same name. You should try to
remove the pack that does not belong in the file structure
and then restart the monitor at 140. Otherwise, you have to
dissolve DSKn, redefine it and refresh it.
?TWO LOGICAL STR n's FOUND IN "SYS" SEARCH LIST
Two file structures are marked to be in the same position in
the SYS search list. You should change the SYS search list
when asked. Refreshing is not required.
ONCE DIALOGUE Page 16-48
ERROR MESSAGES
unit
# BAD REGIONS=n
# BAD BLOCKS=n
DO YOU WANT TO INITIALIZE THE BAD BLOCKS ON THIS UNIT?
You should answer with N or a carriage return to leave the
BAT blocks alone on this unit. The only time you should
initialize is the first time the disk is written, since the
blocks contain the accumulated information about bad
sectors. If you answer Y, the ONCE-Only dialogue responds
with NOT NORMALLY DONE, ARE YOU SURE?. Answer Y only if
this important data is to be erased.
?UNIT ALREADY IN ACTIVE SWAPPING LIST
An attempt was made to specify a unit to be in the active
swapping list more than once. You should type a different
unit name to be in the active swapping list. If you have
included the unit name earlier by mistake, you will have
another chance to change the active swapping list.
?UNIT ALREADY IN FILE STRUCTURE
An attempt was made to specify a unit to be in more than one
file structure. You should type a different unit name to be
in this file structure. If you have included the unit in an
earlier file structure by mistake, you will have to dissolve
it.
?UNIT HAS NO SPACE ALLOCATED FOR SWAPPING
An attempt has been made to specify a unit that has no
swapping space allocated to be part of the active swapping
list. The unit is not added to the list. You should do one
of the following:
1. Specify another unit.
2. Type an extra carriage return signifying completion.
3. Define swapping space for a unit not in a file
structure.
4. Change the swapping space for a unit in a file structure
and refresh it.
CHAPTER 17
BACKUP
17.1 INTRODUCTION
BACKUP is a system program that is used to save disk files on magnetic
tape, and later to restore any or all of these files to disk.
Magnetic tape is the medium used for backup storage of disk files and
for transporting files between sites. This chapter outlines some of
the features that BACKUP provides for accomplishing these two tasks.
17.2 FEATURES
BACKUP allows you flexibility in choosing the files to be transferred
between disk and tape. You specify files by the standard file
specification format of dev:filename.ext[project,programmmer].
Wildcards and sub-file directories are fully supported. You may also
select files based on any of the dates/times associated with disk
files.
BACKUP can handle files that are longer than one reel of magtape.
Switches are provided for checkpointing files while the files are
being handled.
If you are backing up the disk file structure, you can recover from a
system crash without starting completely over. You can produce a
directory listing of the tape at the same time you perform a save.
Another operator feature is a set of run-time commands compatible with
the spoolers.
To facilitate transporting files between sites, BACKUP provides an
"interchange" mode of operation that prevents the writing of system
specific overhead information on the tape. Software distribution
tapes for the DECsystem-10 are produced via BACKUP in interchange
mode.
To increase reliability, BACKUP responds to hardware reported tape
write errors by rewriting the data in a repeater record. When the
tape is read later, these repeater records are used instead of the
originals.
These and other features are described more fully in Sections 17.3 and
17.4.
BACKUP Page 17-2
OVERVIEW OF COMMAND FUNCTIONS
17.3 OVERVIEW OF COMMAND FUNCTIONS
BACKUP commands are in the form of verbs. BACKUP prompts with a slash
(/) and has three kinds of verbs:
1. Action
2. Status setting
3. Tape positioning.
Tables 17-1, 17-2 and 17-3 present BACKUP's general functions and the
command verbs that perform them. Verbs are grouped into the three
categories named above. Table 17-4 contains run-time commands, which
may be given during execution of the action verbs.
Table 17-1
Action Verbs
Functions Verbs
Printing directory of a tape PRINT
Saving or restoring disk files SAVE, RESTORE
Verifying agreement of tape and disk files CHECK
Table 17-2
Status Setting Verbs
Functions Verbs
BACKUP:name option (use from SWITCH.INI) OPTION
[do not] use Checkpoints [NO] CPOINT
Density setting (of tape) DENSITY
[do not] Encrypt the magtape [NO] ENCRYPT
Include files according to:
date-time created or modified BEFORE, SINCE
moved or modified MBEFORE, MSINCE
date accessed ABEFORE, ASINCE
length LENGTH
PPN exemptions from above restrictions [NO] EXEMPT
[do not] run in Interchange Mode [NO] INTERCHANGE
[do not] make a Listing file while running [N] LIST
Parity setting PARITY
Protect directories UPROTECTION
BACKUP Page 17-3
OVERVIEW OF COMMAND FUNCTIONS
Restore files from tape SUPERSEDE
Resume at a specified block of initial file RESUME N
Save set name (specify) SSNAME
Sort files or directories alphabetically,
by location, or by directory SORT
Start processing at specified file INITIAL
[do not] Suppress error message prefix MESSAGE [NO] PREFIX
[do not] Suppress first line of error
message MESSAGE [NO] FIRST
[do not] Suppress disk writing during
a RESTORE [NO] WRITE
Suppress filenames and directories SILENCE
[do not] Type filenames or directories
while running [NO] FILES
[NO] DIRECTORIES
Tape unit (use MTB0) TAPE MTB0
[do not] run in USETI mode [NO] USETI
Table 17-3
Tape Positioning Verbs
Functions Verbs
Back up to start of save set SKIP 0 tape list
Skip to end-of-tape mark for each tape EOT
n save sets forward SKIP n tape list
backward SKIP -n tape list
Rewind to start of tape for each tape REWIND
Unload each tape UNLOAD
Table 17-4
Run-Time Commands
Functions Commands
List and explain these commands HELP
Abort current action verb KILL
Continue after a STOP GO
Display current filename and status WHAT
BACKUP Page 17-4
OVERVIEW OF COMMAND FUNCTIONS
Exit from BACKUP when done EXIT
Do not exit from BACKUP when done PAUSE
Reset status settings to defaults when done RESET
Stop temporarily STOP
[do not] Type all directories of files [NO] DIRECTORIES
processed [NO] FILES
Stop typing all directories or files SILENCE
17.3.1 Action Verbs
The action verbs perform I/O and operate on the tape specified by the
last TAPE verb (one of the status setting verbs). A tape must be
specified before using any of the action commands, or an error will
result. In the following lists of commands and switches
spec is the standard file specification:
dev:filename.ext[directories]
date is in the form dd-mm-yy
time is in the form hh:mm:ss
The action verbs are:
Command Action
SAVE spec-list Save the specified disk files on tape
RESTORE spec-list Restore the specified tape files to disk
CHECK spec-list Verify that the tape and disk files agree
[N]PRINT spec Print a directory of the entire tape on spec. N
is an optional prefix meaning narrow. PRINT
produces the effect of a NODIRECTORIES command;
i.e., no user directories will be typed.
The action commands take as an optional argument a list of file
specifications in a format similar to the monitor's COPY command. The
file specifications may contain wildcards and sub-file directories.
For each entry in the list, you may specify both input and output file
specifications (output=input) or just the input specification. This
allows the files to be renamed as they are saved or restored. If no
output specification is given, then the specified files are
transferred without being renamed. Entries in the list are separated
by commas.
If no argument is supplied with an action command, BACKUP will default
a file spec in the following way. The default file spec for all
action verbs is ALL:*.*[PPN,*,*,*,*,*]. This specifies all files on
all UFDs and all SFDs of all file structures with no renaming.
If you are not logged in under [1,2] the default for the SAVE verb is
ALL:*.*[PPN,*,*,*,*,*]=DSK:*.*[PPN,*,*,*,*,*].
For the CHECK and RESTORE verbs the default is
DSK:*.*[PPN,*,*,*,*,*]=ALL:*.*[PPN,*,*,*,*,*].
BACKUP Page 17-5
OVERVIEW OF COMMAND FUNCTIONS
The argument for the PRINT verb is a single file spec. Its default is
LPT:BACKUP.LOG.
Note that specifying any of the file spec parameters overlays only
that parameter and leaves the rest of the defaults standing. Under no
circumstances do the file spec parameters become "sticky" parameters;
e.g., specifying the PPN for one spec does NOT carry over to the next
spec in the list.
17.3.2 Status Setting Verbs
The status setting verbs set a parameter that affects future action
commands. Once a status parameter is set, it remains in effect until
you change it again. The format for date/time arguments is
dd-mm-yy: hh:mm:ss; relative date/times (prefixed by + or -) and
special mnemonic words (YESTERDAY, TODAY, TOMORROW, LOGIN, NOON,
MIDNIGHT) may also be used. BACKUP conforms to the specification in
the Operating System Commands manual and more details on data formats
can be found there.
The complements, formed by preceding the verbs with NO, negate the
commands and sometimes cause alternate actions; their effects are
explained in parentheses.
The status setting verbs are:
Verb Action
ABEFORE date Include only files accessed before
the specified date
[NO]APPEND Append to existing list file
(default) (Complement: do not
supersede).
ASINCE date Include only files accessed since
the specified date
BEFORE date-time Include only files created and last
modified before the specified
date-time
[NO]CPOINT Use checkpoints. (Complement: do
not use checkpoints).
[NO]DATE75 Always accept files with possible
DATE75 problems (default)
(Complement: do not accept these
files)
[NO]DELETE Delete disk files after saving them
(Complement: do not delete)
DENSITY Set tape density (default is system
(200, 556, 800, 1600, 6250) dependent)
[NO]DIRECTORIES Type each user's directory while
running (default) (Complement: do
not type directories at all)
[NO]ENCRYPT Encrypt the magtape (asks for key
BACKUP Page 17-6
OVERVIEW OF COMMAND FUNCTIONS
later) (a key is similar to a
password and it can consist of up
to 30 alphabetic characters)
(Complement: do not encrypt)
[NO]EXEMPT Exempt PPN's of the form [A,*] and
[10,B], where A and B < 7, from
date/time and length restrictions
(default) (Complement: do not
exempt)
[NO]FILES Type each filename while running
(Complement: do not type filename
at all)
INITIAL spec Start processing at spec
[NO]INTERCHANGE Run in interchange mode (In
interchange mode, only the
filenames, extensions, and versions
are written. There is no
information on what UFD a file was
in when it was saved.) (Complement:
run in normal mode, in which UFDs
and device names are written)
LENGTH 1:h Include only files whose length is
between low and high
[NO]LIST spec While running, make a listing file
on spec (default spec is LPT:
BACKUP.LOG) (Complement: do not
make listing file) LIST produces
the effect of a NODIRECTORIES
command, i.e., no user directories
will be typed.
MBEFORE date-time Include only files which have been
last moved or modified before the
specified time
MSINCE date-time Include only files which have been
moved or modified since the
specified time
MESSAGE [NO]PREFIX Suppress error message prefix
MESSAGE [NO]FIRST Suppress first line of error
message text
[NO]MULTIREEL Allow multiple reels during a save
(default; multiple reels are
always permitted on a restore)
(Complement: do not allow multiple
reels during a save)
OPTION NAME Use option BACKUP:name from
SWITCH.INI
PARITY (EVEN,ODD) Set tape parity
[NO]REPEAT Repeat a split file on the
continuation tape (Complement: do
BACKUP Page 17-7
OVERVIEW OF COMMAND FUNCTIONS
not repeat)
RESUME n Resume at block n of initial file
SILENCE Do not type filenames or
directories while running
SINCE date-time Include only files created or
modified since the specified
date-time
SORT DIRECTORIES x Sort directories within each file
structure in order x when saving.
x = ALPHABETICAL LOCATION (by
compressed-file-pointer, which is
related to a file's physical
location on disk), or NONE (by
Master File Directory)
SORT FILES x Sort files within each directory in
order x when saving.
x = ALPHABETICAL, LOCATION, or NONE
(by directory)
SSNAME name Specify the save set name (up to 30
characters; ALL = all save sets on
tape)
SUPERSEDE ALWAYS Always restore file from tape
SUPERSEDE NEVER Restore files from tape unless on
disk
SUPERSEDE OLDER Restore only the new files from
tape (default)
TAPE MTB0 Use tape unit MTB0. If a magtape
drive has the logical name BACKUP,
then the TAPE verb need not be
specified
UPROTECTION NNN Set the protection for created
directories
[NO]USETI Run in USETI mode (speeds up SAVES
with /SINCE, etc.) (Complement: do
not run in USETI)
[NO]WRITE Suppress disk writing during a
RESTORE. (Complement: allow disk
writing during a RESTORE)
17.3.3 Tape Positioning Verbs
BACKUP's tape positioning verbs take immediate effect. They take, as
an argument, a list of tapes. If no tape is specified for a given
command, the last tape declared by a TAPE verb is positioned. The
tape positioning commands are:
Command Action
BACKUP Page 17-8
OVERVIEW OF COMMAND FUNCTIONS
EOT tape list Skip to the end-of-tape mark for
each tape in the list
REWIND tape list Rewind to the beginning of the tape
for each tape in the list
SKIP n tape list Skip n save sets forward for each
tape in the list
SKIP 0 tape list Back up to the start of the current
save set for each tape in the list
SKIP -n tape list Skip n save sets backward for each
tape in the list
UNLOAD tape list Unload each tape from its drive for
each tape in the list.
The following switches may be included within the file specification
list for the action verbs. These switches can be either temporary or
permanent. A temporary switch immediately follows the file to which
it applies. A permanent switch precedes the list of files to which it
applies, or may be typed on a separate line like a status setting
verb.
Switch Action
ABEFORE date (On input file) include only if
accessed before date
ASINCE date (On input file) include only if
accessed since date
BEFORE date-time (On input file) include only if
created before date-time
ERNONE (On input file) give error if no
files match
ERPROTECTION (On input file) give error if there
is a protection failure
ERSUPERSEDE (On output file) do not restore
from tape if on disk
ESTIMATE n (On output file) estimate output
size
LENGTH L:H (On input file) include only if the
file length is between L:H
MBEFORE date-time (On input file) include only if
modified before date-time
MSINCE date-time (On input file) include only if
modified since date-time
OKNONE (On input file) do not give an
error if no files match
OKPROTECTION (On input file) do not give an
error if there is a protection
failure
BACKUP Page 17-9
OVERVIEW OF COMMAND FUNCTIONS
OKSUPERSEDE (On output file) always restore
even if on disk
[NO]PHYSICAL (Input or output) ignore logical
names (Complement: accept logical
names)
PROTECTION nnn (On output file) set the protection
code
SINCE date-time (On input file) include only if
created since the specified
date-time
VERSION v (On output file) set output file
version number
17.3.4 Run-Time Commands
The following run-time commands may be given during the execution of
the action verbs. BACKUP prompts with an exclamation point (!) when
ready to accept a run-time command. (EXIT, HELP and RESET will also
work with a slash (/) as a prompting character.)
The complements, formed by preceding the commands with NO, negate the
commands and sometimes cause alternate actions; their effects are
explained in parentheses.
The run-time commands are:
Commands Action
[NO]DIRECTORIES Start typing every directory
processed (Complement: do not type
these directories)
EXIT Exit from BACKUP when done
[NO]FILES Start typing every file and
directory processed (Complement:
do not type these files)
GO Continue after a STOP
HELP List these commands and
explanations
KILL Abort execution of the current
action verb
PAUSE Do not exit from BACKUP when done
RESET Reset all status settings to their
original defaults when done
SILENCE Stop typing every directory or file
STOP Stop temporarily: can be continued
by GO.
WHAT Display current filename and status
BACKUP Page 17-10
OVERVIEW OF COMMAND FUNCTIONS
BACKUP also supports "/@ file" construction, allowing the use of
indirect command files, like the COMPIL-class commands. All of
BACKUP's verbs and switches, except for the run-time commands, may be
used in an indirect file. (Note that "/verb @ file" is NOT a valid
alternate format.)
17.4 EXAMPLES
The following examples demonstrate the application of the BACKUP
program.
17.4.1 Console User Examples
The console user can execute all of the BACKUP commands on his own
disk area. He can save his disk area or any other files accessible to
him on his own magnetic tape and later restore to his area all his
files or a subset of his files. In the following examples, the user
is logged in under PPN [10,123].
17.4.1.1 Saving a User's Disk Area - To save all files on his own
disk area on a magnetic tape mounted on MTA0, a user can type
.R BACKUP
/TAPE MTA0
/REWIND
/SAVE
!10,123 DSKB
10,123 DSKC
DONE
/UNLOAD
/^C
17.4.1.2 Restoring Selected Files - The following example shows how
to retrieve a specific file and a set of files from a BACKUP tape by
using wildcards.
.R BACKUP
/TAPE MTA1
/REWIND
/RESTORE FOO.BAR, TST???.*
!10,123 DSKB
DONE
/UNLOAD
/^C
17.4.1.3 Renaming Files as They are Transferred - A user may choose
to change a file's name or location path as it is moved between tape
and disk. In the following example, the file name FILEA.MAC on tape
is copied to disk as FILEB.MAC, but its path (device and directory
level) is not changed.
.R BACKUP
/TAPE MTB0
BACKUP Page 17-11
EXAMPLES
/REWIND
/RESTORE FILEB.MAC=FILEA.MAC
!10,123 DSKB
DONE
/UNLOAD
/^C
In the next example, the file FILEC.MAC was located on DSKB in the
user's UFD [10,123] when saved on tape, but it is restored to the
sub-file directory SFD1 on DSKC.
.R BACKUP
/TAPE MTB0
/REWIND
/RESTORE DSKC:[10,123,SFD1]=DSKB:FILEC.MAC[10,123]
!10,123 DSKB
DONE
/UNLOAD
/^C
17.5 OPERATOR USES
An operator should periodically save the contents of the disk on
magtape. This provides a backup capability should something
unforeseen happen to the disk.
17.5.1 Saving the Entire Disk
To save the entire contents of DSKB on magtape and concurrently
produce a directory listing, BACKUP is run while logged in under [1,2]
as follows:
.R BACKUP
/TAPE MTA0
/LIST DSK:BACKUP.LOG
/SAVE DSKB:
!1,2 DSKB
1,3
1,4
.
.
.
BACKUP types out each UFD as it begins to save files from that area.
If the tape becomes full before the save is completed, BACKUP stops,
types out the full file identification and block number of the current
file being saved, unloads the magtape, and types the message
$BKPEOT REACHED EOT -- MOUNT NEW TAPE THEN TYPE "GO"
When GO is typed BACKUP continues the save on the new tape. (A file
may be split across tapes.)
BACKUP Page 17-12
OPERATOR USES
17.5.2 Recovering from a System Crash
If the system should crash during a save, for example while BACKUP is
saving files from UFD [10,456], the operator, after bringing the
system up, can instruct BACKUP to start at UFD [10,456] by using the
/INITIAL switch:
.R BACKUP
/TAPE MTA0
/LIST DSK:BACKUP.LOG
/INITIAL DSKB:[10,456]
/SAVE DSKB:
!10,456 DSKB
.
.
.
Note that any status parameters which were in effect before the crash
must be reset, and that the tape should NOT be rewound. Note also
that BACKUP does an append to the existing LIST spec.
17.5.3 Saving Only Recently Created or Modified Files
The procedure for saving only those files created or modified today is
identical to that described in Section 3.2.1, with the additional
inclusion of the /MSINCE status setting switch:
.R BACKUP
/TAPE MTA0
/LIST DSK:BACKUP.LOG
/MSINCE:YESTERDAY
/SAVE DSKB:
!1,2 DSKB
.
.
.
Certain files and disk areas are automatically exempted from date-time
restrictions. Files with the RP.ABU bit (always backup bit) set in
the .RBSTS word of the RIB and PPN's of the form [A,*] and [10,B]
where A and B are less then or equal to 7 are always exempted from
date/time restrictions. This causes all libraries, etc., to always be
saved and restored. (The PPN exemption can be overridden by using the
/NOEXEMPT status setting verb.)
17.5.4 Restoring Only Recently Accessed Files
In the following example the /ASINCE switch is used to restore from
tape only those files whose access date is later than June 1, 1975:
.R BACKUP
/TAPE MTA1
/ASINCE:1-JUN-75
/RESTORE
!
BACKUP Page 17-13
RESTORING FROM A DISTRIBUTION TAPE
17.6 RESTORING FROM A DISTRIBUTION TAPE
BACKUP format distribution tapes are made in interchange mode, and
related files are grouped together in named save sets. The user does
not need to know what UFD the file was in when it was saved. Because
the tapes are made in interchange mode, this information is not even
included on tape. Suppose, for example, a new version of SCAN is on
the current tape. The following commands to BACKUP transfer all files
from only the SCAN save set onto ersatz device DEC:
.R BACKUP
/TAPE MTB0
/INTERCHANGE
/REWIND
/SSNAME SCAN
/RESTORE DEC:=DSK:
The save set name "ALL" can be used in place of SCAN to indicate that
files from all save sets are to be copied.
17.7 OBTAINING DIRECTORIES OF BACKUP TAPES
The command
[N]PRINT spec
prints a directory of the entire tape. The optional prefix N
indicates a narrow listing (72 columns). The default specification is
LPT:BACKUP.LOG. If a line-printer is not available to the user's job,
spooling is not in effect, and the default spec is used, an error will
occur.
The following example produces a disk file directory listing named
BACKUP.LOG.
.R BACKUP
/REWIND
/PRINT DSK:
!
This is an example of the output.
.TYPE BACKUP.LOG
Start of save set SCAN on MTB401
System R5200 SYS #40/2 TOPS-10 monitor 602(13053) APR#40
1600 BPI 9 track 18-Sep-75 13:09:13 BACKUP 2(155) format 1
SCAN REL 58 <055> 31-Mar-75 DSKB: [10,4077]
SCAN RNO 8 <055> 8-Mar-75
SCAN MAC 304 <055> 31-Mar-75
SCN7B REL 59 <055> 4-Sep-75
SCN7B DO 101 <055> 18-Aug-75
SCN7B RNO 97 <055> 18-Aug-75
SCN7B CTL 3 <055> 24-Jul-75
SCN7B MAC 309 <055> 4-Sep-75
SCAN MEM 11 <055> 18-Sep-75
End of save set SCAN on MTB401
System R5200 SYS #40/2 TOPS-10 monitor 602(13053) APR#40
BACKUP Page 17-14
OBTAINING DIRECTORIES OF BACKUP TAPES
1600 BPI 9 track 18-Sep-75 13:09:24 BACKUP 2(155) format 1
Start of save set BACKUP on MTB401
System R5200 SYS #40/2 TOPS-10 monitor 602(13053) APR#40
1600 BPI 9 track 18-Sep-75 13:46:56 BACKUP 2(155) format 1
BACKUP REL 22 <055> 15-Sep-75 DSKB: [10,4077]
BACKRS REL 52 <055> 17-Sep-75
BACKUP SHR 55 <055> 17-Sep-75
BACKUP LOW 34 <055> 17-Sep-75
BACKUP MAC 118 <055> 15-Sep-75
BACKRS MAC 270 <055> 17-Sep-75
BACKUP RNH 19 <055> 15-Sep-75 DSKC: [10,4077]
BACKUP HLP 20 <055> 15-Sep-75
BACKUP RNO 4 <055> 25-Aug-75
BACKUP DOC 4 <055> 25-Aug-75
BACKUP CTL 2 <055> 20-Aug-75
End of save set BACKUP on MTB401
System R5200 SYS #40/2 TOPS-20 monitor 602(13053) APR#40
1600 BPI 9 track 18-Sep-75 13:47:13 BACKUP 2(155) format 1
17.8 COMPARING TAPE AND DISK FILES
The command
CHECK spec-list
Verifies that the tape and disk agree. BACKUP compares the tape files
specified for input to the disk files specified for output, word for
word.
The following example saves all the user's COBOL files on tape and
then verifies that the saved tape files are identical to the disk
files.
.R BACKUP
/TAPE MTA0:
/REWIND
/SAVE *.CBL
/REWIND
/CHECK
!
/
To restore all SHR and LOW files from a tape and then verify that the
files were restored correctly type
/REWIND
/SSNAME ALL
/RESTORE *.SHR,*.LOW
!
/REWIND
/CHECK *.SHR,*.LOW
!
/
BACKUP Page 17-15
CHECKPOINTING LARGE FILES
17.9 CHECKPOINTING LARGE FILES
Installations which maintain exceptionally large files (over 5000
blocks) will want to include the /CPOINT status setting switch when
saving and restoring files. This switch extends BACKUP's system crash
recovery capability to the file block level. During a checkpoint
save, the typeout level is set to type the filenames and checkpoints
as they are passed. To continue from the last checkpoint after a
crash, the /INITIAL switch is used to indicate the file spec and the
/RESUME switch to declare the checkpoint block number. For example:
.R BACKUP
/TAPE MTA0
/INITIAL DSKB:[40,577]DATBAS.DBS
/RESUME 6000
/SAVE
!40,577 DSKB
DATBAS DBS
RESUMING AT CHECKPOINT 6000
7000
8000
.
.
.
Note that the tape should not be rewound to continue a checkpoint
save.
During a checkpoint restore, the disk output file is closed at every
checkpoint and then appended to. The procedure for continuing after a
crash is the same as that for the checkpoint save, with the exception
that the tape must first be rewound.
The default for checkpoints is 1000 blocks.
17.10 BACKUP MESSAGES
17.10.1 Operator Messages
$BKPEOT REACHED EOT -- MOUNT NEW TAPE THEN TYPE "GO"
The tape is full. The operator should mount a new one and then
type GO to continue.
$BKPTWL TAPE WRITE LOCKED -- ADD WRITE RING THEN TYPE "GO"
The tape is write locked. The operator should insert a write
ring and then type GO to continue.
17.10.2 Error Messages
?BKPABC AMBIGUOUS COMMAND
The user has given a command abbreviation that is not unique.
?BKPCOL CAN'T OPEN LISTING DEVICE
The device specified for the LIST command cannot be selected for
output.
BACKUP Page 17-16
BACKUP MESSAGES
?BKPCOM CAN'T OPEN MAGTAPE
The tape device specified cannot be selected for output.
?BKPCSA CAN'T SAVE WITH SAVE SET NAME "ALL"
"ALL" is reserved to mean all SAVE sets on tape; therefore, it
cannot be used to name an individual SAVE set when saving.
?BKPDND DEVICE NOT A DISK
The input device specified for a SAVE is not a disk.
?BKPDNM DEVICE NOT A MAGTAPE
The device specified for the TAPE command or a tape positioning
command is not a magnetic tape drive.
?BKPFSL FILE SPECIFICATION DATA LOST
During the processing of a file specification list, a consistency
check determined that data was lost. If this error occurs,
please send an SPR.
?BKPHSG CANNOT GET HIGH SEGMENT BACK
BACKUP releases its high segment command scanner when performing
I/O to eliminate most of the core while running. The attempt to
restore the high segment after completing the I/O operation has
failed.
?BKPIRC INVALID RUN TIME COMMAND -- TYPE KILL TO ABORT RUN FIRST
The user typed an illegal run time command. Type KILL to abort
the run and return to verb command level. Type HELP to get a
list of valid run time commands.
?BKPKDM KEYS DON'T MATCH -- PLEASE TRY AGAIN
BACKUP asks for the encryption key twice, and compares the keys
for verification. This indicates that the comparison failed.
?BKPLFE LISTING FILE ENTER ERROR
After opening the listing I/O channel, no entry could be made for
the listing file.
?BKPLSI LISTING SPECIFICATION INCORRECTLY FORMATTED
The user has used incorrect formatting or a "wild-card" in
specifying the LIST file. No list specification is created.
?BKPNTS NO TAPE SPECIFIED
While searching for the last tape specification given, BACKUP
could not fine one.
?BKPNZC NEGATIVE AND ZERO CHECKPOINTS ILLEGAL
The user specified a negative or zero argument with the RESUME
command.
?BKPRES REACHED EOT ON A SINGLE REEL SAVE
BACKUP Page 17-17
BACKUP MESSAGES
This message is issued if and when the end of the tape is reached
and the user has specified the /NOMULTIREEL command.
?BKPRTE REACHED TAPE ERROR MAXIMUM
BACKUP will abort the execution of an action verb and return to
command level when a large number of tape I/O errors have
occurred. Currently the error maximum is set to 10. (This is an
assembly parameter.)
?BKPTMI INSUFFICIENT CORE FOR COMMAND
The routine for allocating space for file specifications could
not expand core enough to store the specification.
?BKPTSI TAPE SPECIFICATION INCORRECTLY FORMATTED
The user has used incorrect formatting or a "wild-card" in
specifying the tape device.
17.10.3 Warning Messages
%BKPABT ABORT spec
The transfer of the specified file from tape to disk has been
aborted.
%BKPBTL BLOCK TOO LARGE READING SPEC (BLOCK = n)
The current record read from tape exceeds the buffer size. The
record will be skipped.
%BKPCCM CANNOT COPY MFD FOR STRUCTURE
The program cannot get enough core to copy the Master File
Directory for the indicated structure. Files for this structure
will not be saved, and the program will skip to the next
structure.
%BKPCCR CANNOT COPY UFD/SFD RIB FOR spec
The program cannot get enough core to copy the retrieval
information block for the indicated directory. Files for this
User File Directory/Sub-File Directory will not be saved, and the
program will skip to the next UFD/SFD.
%BKPCCU COPY UFD/SFD FOR spec
The program cannot get enough core to copy the indicated User
File Directory or Sub-File Directory. Files for this UFD or SFD
will not be saved, and the program will skip to the next UFD/SFD.
%BKPCDF CANNOT DELETE FILE error bits (code) spec
The specified file could not be deleted. The error bits and code
returned are listed in the DECsystem-10 Monitor Calls Manual,
Appendix E.
%BKPCDS CHECK DISK FILE SHORTER spec
During a check operation, an end-of-file occurred for the
BACKUP Page 17-18
BACKUP MESSAGES
indicated disk file, even though there is more file data on tape.
%BKPCFD CHECK FILES ARE DIFFERENT spec
The check operation determined that the disk and tape versions of
the idicated file are different.
%BKPCHK CHECKSUM INCONSISTENCY READING spec (BLOCK = n)
During a read, the computed checksum of the current tape record
did not agree with the checksum stored when the record was
written.
%BKPCNF CHECK FILE NOT ON DISK spec
The indicated file could not be found on disk during a check
operation.
%BKPCOD CANNOT OPEN "file structure"
The file structure indicated cannot be selected for I/O. The
transfer of files for this structure will be aborted.
%BKPCTS CHECK TAPE FILE SHORTER spec
During a check operation, and end-of-file occurred for the
indicated tape file, even though there is more file data on disk.
%BKPDIO DISK I/O ERROR error bits DURING spec
A disk I/O error occurred while attempting to read or write the
specified file. The error bits included are those returned by
the monitor GETSTS call. The transfer will be aborted for this
file.
%BKPFEE ENTER error-code spec
File enter error for the indicated file. The error code included
is that returned by the monitor, and the error code abbreviation
is listed in the DECsystem-10 Monitor Calls Manual, Appendix E.
%BKPFLE LOOKUP error-code spec
File lookup error for the indicated file. The error code
included is that returned by the monitor, and the error code
abbreviation is listed in the DECsystem-10 Monitor Calls Manual,
Appendix E.
%BKPFRS FRS TAPES NOT SUPPORTED
On reading a tape, if the tape format seems to correspond to an
FRS tape rather than a BACKUP tape, this message is issued and
BACKUP continues scanning the tape for records in a format which
it understands.
%BKPHSI HEADER FILE SPEC INCONSISTENCY
During a restore operation, a consistency check determined that
the end-of-file record for the current file was missed, and the
current record belongs to another file. The transfer will be
aborted for this file.
%BKPIBL INCORRECT BLOCK LENGTH
BACKUP Page 17-19
BACKUP MESSAGES
During a read, the program encountered a tape record of the wrong
size. The record will be skipped.
%BKPLF LISTING FILE ERROR error bits (code) spec
The listing file is closed and appended to after the processing
of each UFD, so that it will be preserved through a system crash
and recovery procedure. If either the LOOKUP or ENTER UUO for
appending to the listing file fails, this message is issued and
no further output is done to the listing file. The error bits
and code returned are listed in the DECsystem-10 Monitor Calls
Manual, Appendix E.
%BKPNBF NOT BACKUP FORMAT
The current tape record is not in BACKUP format, and the program
will skip to the next tape record.
%BKPNEC NOT ENOUGH CORE
The program cannot get enough core. If this occurs during a
RESTORE, the RESTORE will be aborted. During a Save, the program
will skip to the next structure or User File Directory and
attempt to continue.
%BKPNFF NO FILES FOUND TO MATCH spec
No files were found to match the given file specification.
%BKPRIC RESUME AT INVALID CHECKPOINT ATTEMPTED
This message indicates that the user instructed BACKUP to
continue saving or checking a file at a checkpoint which was
larger than the actual number of blocks in the file.
%BKPROD RESTORE OUTPUT DEVICE IS NOT A DISK
The output device specified for a RESTORE is not a disk.
%BKPSCE SIZE COPY ERROR spec
While transferring a file from tape to disk an error was
encountered because the transferred file was not the same size as
the tape file. If this error occurs, please send an SPR.
%BKPSLE SFD LEVEL EXCEEDED
While attempting to save needed Sub-File Directories the SFD
level was exceeded.
%BKPSNF SAVE SET NOT FOUND name
This message may occur during a RESTORE or CHECK operation and
indicates that the save set named with the last /SSNAME verb
could not be found.
%BKPTHE TAPE HARDWARE ERROR READING/WRITING spec (BLOCK=n)
A hardware error occurred. The transferred file may contain
unreliable data.
%BKPTPE TAPE PARITY ERROR READING/WRITING spec (BLOCK=n)
BACKUP Page 17-20
BACKUP MESSAGES
The hardware detected a parity error. The transferred file may
contain unreliable data.
%BKPUOE UNTRACEABLE OUTPUT ERROR
An error was encountered while trying to move data from disk to
tape.
%BKPURT UNKNOWN RECORD TYPE
While reading the tape, BACKUP found that the type number of the
current tape record is not within the defined range. The program
will skip to the next tape record.
17.11 CAPACITIES OF DISK MEDIA VERSUS MAGNETIC TAPE
The following table illustrates the maximum capacity of various disk
devices and the approximate number of 2400-foot magnetic tapes
required to hold the same number of blocks. The calculations were
made for 9-channel magnetic tape at 800 bpi.
No. of magtapes
Required for same
Disk Device Capacity in Blocks capacity
RD10 (Disk) 4000 .139
RM10B (Drum) 2700 .094
RP02 (Disk Pack) 40000 1.39
RP03 (Disk Pack) 80000 2.78
RP04 (Disk Pack) 154280 5.37
RS04 (Disk) 2048 .071
17.12 BACKUP TAPE FORMAT
Note that Backup is designed for two primary functions: performing
system backup and interchanging files between systems. For the latter
function, Backup provides an "interchange" switch which causes system
dependent data to be ignored and only critical file information to be
written on tape. A restore operation in interchange mode also ignores
system dependent data, allowing the operating system to supply
defaults where necessary. Items not included in interchange mode are
noted in the description which follows.
17.12.1 Tape Record Types
BACKUP tapes are made up of a series of tape records of various types.
Each record is self identifying. All records on the tape are written
at the standard length of 544(10) words, made up of a 32(10) word
header and a 512(10) data area. Even if the data area is not needed,
or is only partially needed it is fully written. All undefined or
unused words are written with zeroes and ignored on read. This
maximizes the probability of reading old tapes. In any case, the tape
format is included in the labels and the save set headers.
BACKUP Page 17-21
BACKUP TAPE FORMAT
The record types are:
1. T$LBL -- tape label used to identify reel ID and destruction
date/time. This record is optional, but if present must be
at the start of the tape.
2. T$BEG -- Beginning of a save set used to identify when the
save set was written and on what device of what system. It
also includes the save set name. This record is mandatory
and must be the first record of the save set.
3. T$END -- end of a save set. This is identical to the T$BEG
record except that it appears at the end.
4. T$FIL -- this is the actual data which has been saved. It is
the only type of record which is encrypted. It is
self-identifying as to the position within the file, but
contains only part of the full path name of the file.
5. T$UFD -- contains the information for each directory. It
gives all information necessary to re-create the directory.
(Not written in interchange mode.)
6. T$EOV -- indicates end of volume (future).
7. T$COM -- comment (ignored).
8. T$CON -- continuation of save set. This is identical to
T$BEG except that it indicates the continuation of the save
set at the start of a new volume. This ensures that each
volume is completely self identifying.
17.12.2 Standard Record Format
Every tape record has the same general format. This consists of a
32(10) word record header followed by one page of data (512(10)
words). All record headers start with the same first twelve words.
The first seven words are:
1. G$TYPE -- Record type as described in the previous section.
This is a small positive integer.
2. G$SEQ -- record sequence number. This is incremented by one
for each record on the tape. If a record is repeated because
of a tape write error, the number of the repeated record is
the same as that of the original.
3. G$RTNM -- relative tape number. This is incremented by one
for each volume.
4. G$FLAG -- various flag bits:
a. GF$EOF -- This flag is set if this is the last tape
record for this disk file. On short files, this can even
be set on the first record of the file.
b. GF$RPT -- this flag is set if this tape record is a
repeat of the previous record. This is set whenever the
record is rewritten because of a tape write error.
BACKUP Page 17-22
BACKUP TAPE FORMAT
c. GF$NCH -- this flag is set if no checksum has been
computed for the tape record.
d. GF$SOF -- this flag is set if this is the first tape
record for this disk file.
5. G$CHK -- checksum of the tape record.
6. G$SIZ -- number of words used for data in this tape record.
7. G$LND -- number of words to skip before the data starts.
The next four words are reserved for future expansion. The twelfth
(last) word in the general section of the record header is reserved
for customer use. The remaining 20 words in the record header vary
for each record type, with the last word of each record header being
reserved for customer use. In interchange mode, customer reserved
words will be written as zero on a save and ignored on a read.
17.12.3 Non-Data Blocks
The data portion of a tape record is primarily for storing file data,
but may be used for saving some overhead information. Any non-data
information written in the data area of a tape record is prefaced with
a control word of the form:
LH = type, RH = length in words including this word.
More than one overhead region can appear. In this case, they follow
each other with no intervening space. The currently defined types for
overhead blocks are:
1. O$NAME -- gives the full path identification of the file
without punctuation. The path components are treated as if
the user gave a quoted representation in "DEC Integrated
Command Language". This block consists of sub-blocks in the
standard order: device, directories (top down), file name,
extension, version, generation. Sub-blocks corresponding to
missing fields in the path specification are omitted. Each
sub-block is in the format:
WORD0: LH = type, RH = length in words including this word.
The rest of the sub-block is the path field in ASCIZ without
leading or imbedded nulls, terminated by at least one null.
Omitted fields will be defaulted. In interchange mode, only
the name, extension and version are written. In interchange
restore, only name, extension and version are used.
sub-block type codes are:
1 = device
2 = name
3 = extension
4 = version
5 = generation
40 = directory (lower directories are 41, 42, ...)
2. O$FILE -- a block containing file attributes. The first
section of this block is a fixed length header area
containing in fixed locations either single word attributes
BACKUP Page 17-23
BACKUP TAPE FORMAT
or byte pointers to ASCIZ string attributes located in the
remaining section. All dates and times are in universal
date/time format. In interchange mode only the critical
attributes (starred) will be written, and the rest of this
block will contain zeros. In the description which follows,
the symbols in brackets represent the RIB data from which the
attribute values will be converted. (If none is given, the
location will be zero.)
a. A$FHLN (*) -- fixed header length in words.
b. A$FLGS -- flags:
1. B$PERM -- permanent (not deletable) [RP.NDL]
2. B$TEMP -- temporary
3. B$DELE -- already deleted
4. B$DLRA -- don't delete for lack of recent access
[RP.ABU}
5. B$NQCF -- not quota checked [RP.NQC]
6. B$NOCS -- does not have valid checksums [RP.ABC]
7. B$CSER -- has checksum error [RP.FCE]
8. B$WRER -- has disk write error [RP.FWE]
9. B$MRER -- had BACKUP read error on RESTORE [RP.BFA]
10. B$DAER -- declared bad by damage assessment [RP.BDA]
c. A$WRIT (*) -- date/time of last write [RB.CRD and RB.CRT]
d. A$ALLS (*) -- allocated size in words [.RBALC]
e. A$MODE (*) -- mode of last write [RB.MOD]
f. A$LENG (*) --length in bytes (1B0 if > 2^35-1) [.RBSIZ]
g. A$BSIZ (*) -- byte size (7 or 36).
h. A$VERS (*) -- version identification (.JBVER format)
[.RBVER]
i. A$PROT -- protection [RB.PRV]. The protection for
directories appears in the directory attribute block
(O$DIRT). For files, the protection word is defined as
four fields of eight bits each with a "5" stored in the
leftmost three bits in order to avoid looking like a byte
pointer:
bits 0-2 "5"
bit 3 reserved for future
bits 4-11 future access
bits 12-19 owner access
bits 20-27 affinity group access
BACKUP Page 17-24
BACKUP TAPE FORMAT
bits 28-35 "world" access
Each file access field is subdivided into bytes which
describe the attribute, write and read (respectively)
protections associated with the file. A description of
the "world" access field follows, with the associated
TOPS-10 protection given in parentheses, if applicable.
The owner and affinity group (project) fields are
similarly defined.
1. PR$SPC (bit 28) -- reserved for special checking.
The rest of the field is special if this bit is set.
2. PR$ATR (bits 29-31) -- the attribute subfield is a
3-bit byte interpreted as follows:
0 -- file is completely hidden.
1 -- file name is visible (7-6).
2 -- file attributes are visible (5-2).
3 -- can change unprotected attributes.
4-5 -- (future)
6 -- can change protection (0).
7 -- can delete the file (1).
3. PR$WRT (bits 32-33) -- the write access subfield is
defined as:
0 -- no write access (7-5).
1 -- append (4).
2 -- write (3).
3 -- superseding generation (2-0).
4. PR$RED (bits 34-35) -- the read access subfield is
defined as:
0 -- no read access (7).
1 -- execute only (6).
2 -- can read the file (5-0).
3 -- (future).
j. A$ACCT -- byte pointer to account string
k. A$NOTE -- byte pointer to annotation string [.RBSPL]
l. A$CRET -- creation date and time of this generation
m. A$REDT -- last read date and time of this generation
[RB.ACD]
n. A$MODT -- monitor set last write date and time [.RBTIM]
BACKUP Page 17-25
BACKUP TAPE FORMAT
o. A$ESTS -- estimated size in words [.RBEST]
p. A$RADR -- requested disk address [.RBPOS]
q. A$FSIZ -- maximum file size in words
r. A$MUSR -- byte pointer to identification of last modifier
s. A$CUSR -- byte pointer identification of creator [.RBAUT]
t. A$BKID -- byte pointer to identification of previous
BACKUP [.RBMTA]
u. A$BKDT -- date and time of last backup
v. A$NGRT -- number of generations to retain
w. A$NRDS - number of opens for read this generation
x. A$NWRT -- number of opens for write this generation
y. A$USRW -- undefined user word [.RBNCA]
z. A$PCAW -- privileged customer word [.RBBCA]
The remainder of this block is reserved for future expansion.
3. O$DIRT -- a block containing directory attributes (not
written in interchange mode). The first section of this
block is fixed length header area containing either directory
attributes or pointers to attributes located in the remaining
section. The symbols in brackets represent the RIB data used
for conversion (the location is zero if none is given). The
directory protection word appears in this block rather than
in the O$FILE block (A$PROT is zero for directories).
a. D$FHLN -- fixed header length in words
b. D$FLGS -- directory flags:
1. DF$FOD -- file only directory
2. DF$AAL -- alpha accounts are legal
3. DF$RLM -- repeat login messages
c. D$ACCT -- account number or ASCII byte pointer to account
string
d. D$PROT -- directory protection [RB.PRV]. The directory
protection word is divided into the same access fields as
the file protection word, A$PROT, but each directory
access field has bits as follows (RIB bits given in
parentheses):
Bit 28 -- reserved for special checking. The rest of the
field is special if this bits is set.
Bits 29-31 -- (future)
Bit 32 -- connect allowed
Bit 33 -- can open files (4)
BACKUP Page 17-26
BACKUP TAPE FORMAT
Bit 34 -- can create generations (2)
Bit 35 -- directory can be read (1)
e. D$FPRT -- default file protection
f. D$LOGT -- date/time of last login in DEC-10 universal
format [RB.CRD and RB.CRT]
g. D$GENR -- default number of generations to keep
h. D$QTF -- first-come-first-served logged-in quota in words
[.RBQTF]
i. D$QTO -- logged out quota in words [.RBQTO]
j. D$ACSL -- list of groups which can access this directory
(see below)
k. D$USRL -- list of groups which this user is in (see
below)
l. D$PRVL -- privilege list (see below)
m. D$PSWD -- ASCII byte pointer to password
The list attribute words given above (D$ACSL, D$USRL, D$PRVL)
may be in any one of the following formats:
a. an ASCII string pointer
b. 5B2 + group (or 5B2 + privilege for D$PRVL)
c. -N,, Relative location of start of list
If in format (c), each word of the list is 5B2 + group (5B2 +
privilege for D$PRVL)
4. O$SYSN -- a block containing the system header line in ASCIZ.
5. O$SSNM -- a block containing the user supplied save set name
in ASCIZ (max of 30 characters). This block is omitted if no
save set name was specified.
17.12.4 Locations in T$LBL Record
This record has no contents in the "data" region. The remaining
locations in the record header are defined as follows:
1. L$DATE -- date/time of labelling in DEC-10 universal format
(i.e., LH=DAYS since 17-NOV-1858, RH=FRACTION of day)
2. L$FMT -- BACKUP tape format (constant = 1).
3. L$BVER -- version of BACKUP writing label in standard .JBVER
format.
4. L$MON -- monitor type (%CNMNT).
BACKUP Page 17-27
BACKUP TAPE FORMAT
5. L$SVER -- system version (%CNDVN).
6. L$APR -- APR processor serial number on which this label was
written (integer).
7. L$DEV -- physical device on which the tape was written in
SIXBIT.
8. L$MTCH -- BTYE (31) 0 (1) 7-track (1) 0 (3) density. Density
is 1=200, 2=556, 3=800, 4=1600, 5=6250.
9. L$RLNM -- REELID in SIXBIT.
10. L$DSTR -- date/time before which the tape cannot be
scratched. Before this time, the only valid operation is to
append.
17.12.5 Locations in T$BEG, T$END, T$CON Records
These save set records all have the same format and are distinguished
by their record types and their location on the tape. All items are
filled in at the time of writing. The data area contains two non-data
blocks, types O$SYSN and O$SSNM. Record header locations following
the first standard twelve words are defined as follows:
1. S$DATE -- date/time of writing this record in universal
format.
2. S$FMT -- BACKUP tape format (constant = 1).
3. S$BVER -- BACKUP version in .JBVER format.
4. S$MON -- monitor type (%CNMNT).
5. S$SVER -- system version (%CNDVN).
6. S$APR -- apr serial number on which written.
7. S$DEV -- physical name of device on which written in SIXBIT.
8. S$MTCH -- BYTE (31) 0 (1) 7-track (1) 0 (3) density. Density
is 1=200, 2=556, 3=800, 4=1600, 5=6250.
17.12.6 Locations in T$UFD Record
This record is not written in interchange mode. When written, the
data portion contains two or three non-data blocks: types O$NAME,
O$FILE (optional) and O$DIRT. Remaining locations in the header
record contain:
1. D$PCHK -- checksum of the O$NAME full path file name block.
2. D$LVL -- directory level: 0=UFD, 1=first SFD, etc.
3. D$STR -- file structure name stored in the following format:
BYTE (7) data type, length in words, ASCII. (Date types are
defined in the T$FIL section.)
BACKUP Page 17-28
BACKUP TAPE FORMAT
17.12.7 Locations in T$FIL Record
The first tape record for a file contains two non-data blocks, types
O$NAME and O$FILE. There is room for two blocks of file data in the
first tape record, and if the file will completely fit in one tape
record, these will be used. If the file is longer than two blocks,
the file will be started in the second tape record, so its pages will
be lined up with tape records. Each tape record identifies the
logical disk word with which it starts. Remaining locations in the
record header are:
1. F$PCHK -- checksum of the full path file name block (O$NAME).
This is just a consistency check for consecutive records of
the file.
2. F$RDW -- relative data word of file of the first data word in
this tape record.
3. F$PTH -- a twelve word block used to store information
suitable for a restoration of the file. This area is big
enough to hold the entire path to a TOPS-10 file in a UFD and
two SFDS. The path information will be stored in the
standard order of device, UFD, first SFD, file name,
extension; with missing fields omitted. The path
information will be stored in the format:
BYTE (7) data type, length in words, ASCII
where data types are defined as:
device = 001
file name = 002
extension = 003
directory = 040
(lower directories = 041,042, ...)
CHAPTER 18
BOOTM
18.1 INTRODUCTION
BOOTM is a bootstrap program used for loading the monitor from a
magnetic tape saved in BACKUP format. BOOTM runs in EXEC mode, and
automatically loads itself into the top 2K of memory. You load BOOTM
either from magnetic tape via the READIN facility or from paper tape.
18.2 OPERATION
DIGITAL supplies BOOTM on either a paper tape or a bootstrap magnetic
tape.
18.2.1 Load From Paper-tape (KA/KI Only)
You should follow the steps outlined below to load BOOTM from paper
tape.
1. Load the BOOTM paper tape into the paper-tape reader.
2. Set the READIN device switches to 104.
3. Press STOP, RESET, and READIN.
4. When the paper tape has been read, BOOTM types:
BOOTM V4(16)
BTM>
The prompt characters, indicating that you can type a
command, are BTM>.
5. Type a command string and press the RETURN key.
6. After BOOTM processes the command you typed, BOOTM either
restarts itself or transfers to the newly loaded program.
The action BOOTM takes depends on the command you type.
BOOTM Page 18-2
OPERATION
18.2.2 Load From Magnetic Tape
You should follow the steps outlined below to load BOOTM from magnetic
tape.
1. Place the BOOTM tape on magnetic tape drive 0 (e.g., MTA0).
If you're using a DX10 controller, you can use any drive.
2. Set the READIN switches to 340. When using a DX10
controller, set the switches to 220.
3. Press STOP, RESET, and READIN.
4. When the magnetic tape has been read, BOOTM types:
BOOTM V4(16)
BTM>
The prompt characters, indicating that you can type a
command, are BTM>.
5. Type a command string and press the RETURN key.
6. After BOOTM processes the command you typed, BOOTM either
restarts itself or transfers to the newly loaded program.
The action BOOTM takes depends on the command you type.
18.3 COMMAND FORMAT
The general format of a BOOTM command is shown below.
structure:file.ext[proj,prog]/switch
where: structure represents a file structure name within the disk
file system (as saved by BACKUP). The default is DSKB.
file.ext is the name and extension of the file involved in the
loading operation. The default is SYSTEM.EXE.
[proj,programmer] is the directory in which the specified file
can be found. The default is [1,4].
/switch is an optional BOOTM switch that specifies what
operation or option BOOTM should perform. The available
options and operations are listed in Section 6.3.1.
18.3.1 Options and Operations
If you do not specify a switch in the command, 1) BOOTM finds the
specified file, 2) clears memory, 3) reads the specified file into
memory, 4) sets the PROGRAM START ADDRESS, and 5) starts the program
at that address. The default file used is DSKB:SYSTEM.EXE[1,4].
BOOTM positions the tape to the second file on the tape before it
starts the program.
The possible switches that you can specify in the BOOTM command line
are listed below.
BOOTM Page 18-3
COMMAND FORMAT
/LOAD BOOTM performs all the operations listed above,
except that it does not start the program (i.e.,
does not perform step 5).
/START:n BOOTM performs all the operations listed above,
except it starts the program at location n. The
default value for n is the start address specified
in the file.
/NOREWIND BOOTM is disabled from repositioning the tape to
the second file, which is assumed to be a copy of
BACKUP.EXE.
/DENSITY:d Set the tape read density to the value of d.
Legal values are 200, 556, 800, and 1600.
/kontroller:u Inform BOOTM what type of tape controller you are
reading from and the unit "u" that has the tape
mounted. Legal values for "kontroller" are TM10,
TX01, TM02, and TC10. If READIN was done from
either a TM10 or a TX01, then it is not necessary
to specify this switch. The default tape drive is
the unit that a READIN was performed on if a TM10
or a TX01 was used. If BOOTM was read in from a
device other than magtape, then the default is
TM10 unit 0. If a TC10 controller is used, it
must be specified by a /TC10 switch. The default
value for u is unit #0. Example: /TX01:2
specifies unit 2 on the TX01 controller.
/REWIND This operation will rewind the tape to load point.
/DENSITY and /kontroller above may be used in
conjunction with this switch; however, no
filename can be given.
/SKIP This operation will skip forward over one file.
Options 5 and 6 above may be used in conjunction
with this command; however, no filename can be
given.
18.4 ERRORS
Whenever BOOTM encounters an error, the following style of error
message appears:
1. The user terminal end-of-line bell is rung;
2. The string ?BTM followed by a 3-character mnemonic followed
by the error message text terminated by a carriage return and
line feed;
3. BOOTM is then restarted.
BOOTM Page 18-4
ERRORS
18.4.1 Error Messages
?BTMIPP - Invalid PROJ,PROG number
?BTMCME - Command error
?BTMISW - Illegal switch or argument
?BTMFNF - File not found
?BTMTSF - Tape mark in save file
?BTMSFI - Save file inconsistent
?BTMNSA - No start address
?BTMNDL - Not an EXE file or directory too long
?BTMPNM - Page not monotonically increasing
?BTMTRE - Tape read error
?BTMCSD - Cannot start DX10
?BTMCSE - Channel synchronization error
?BTMNMS - No magtape status
?BTMFDE - Fatal DX10 error or drive off-line
?BTMNES - Not ending status (TU70 only)
?BTMNCS - Not CU status (TU70 only)
18.5 ASSEMBLY INSTRUCTIONS
The following feature test switches are available in BOOTM:
FTEXE Include EXE file support (default)
FTFRS Include FRS/BACKUP file format (default)
FTTU70 Include TU70 support (default)
FTTC10 Include TC10 support (default)
MAGRIM Assemble in MAGRIM format (default)
PTPSW Make paper tape version (normally off)
CORE Value set to 5000 (default) used to determine where to
readin BOOTM before relocation.
DEBUG Make debug version
Assembly instructions:
1. To make a paper tape version (location independent):
.MAKE BTMPTP.MAC
*IPTPSW==1
MAGRIM==0
$EX$$
.R MACRO
*PTP:=BTMPTP,BOOTM
2. To make a file that can be copied to a magtape for READIN,
just assemble with no special switches.
3. To make a file that can be loaded with DXLD to make a TU70
readin tape:
.MAKE BTMDX.MAC
*IMAGRIM==0
$EX$$
.LOAD DXLD,BTMDX+BOOTM
.ASSIGN DSK OUT
.START ;This produces BOOTM.RDI
BOOTM Page 18-5
EXAMPLES
18.6 EXAMPLES
To retrieve and run a fresh monitor from a BACKUP format magtape
containing DSKB:SYSTEM.EXE(1,4), type a carriage return only.
To rewind a tape on TM10 drive #3 type:
/REWIND/TM10:3
To load RV765 from DSKB:[10,2362] on a BACKUP format tape and start
EXEC DDT type:
RV765[10,2362]/START:401
18.7 SPECIAL 1080 INSTRUCTIONS
There is no readin on KL10 CPUs. See the Operators Guide for
instructions for readin in BOOTM on 1080s. Be sure to specify /TX01,
when reading in a tape on a TU70.