T.R | Title | User | Personal Name | Date | Lines |
---|
95.1 | | KERNEL::GARNETT | | Fri May 18 1990 15:23 | 411 |
| From: BAHTAT::BAHTAT::MRGATE::"A1::STARTUP IAN" 15-MAY-1990 13:52:19.72
To: @Distribution_List
CC:
Subj: RSTS/E News 2 - May 90
From: NAME: Ian Startup
FUNC: RSTS/E Software Support
TEL: 7845 3102 <STARTUP IAN AT A1 at ERIC at LZO>
To: See Below
RSTS/E Newsletter Issue 2 - May 1990
COPYRIGHT (c) 1990 by Digital Equipment Corporation.
----------------------------------------------------
So who says the PDP-11 is dead. Good news of two new high end processors
the 11/93 and 11/94. Also, ever needed to restore a customers disk and all
its data. No, then think yourself lucky, otherwise read all about RECOVR.COM
and its role in recovering data.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
ANN. PDP-11/90
All Sales Update Europe articles are for Internal Use only. This
proprietary information indicates that unauthorised or inadvertent
disclosure could cause business damage to the Corporation. It can be
distributed to Digital employees, but should not be given to
customers, competitors, vendors, or other persons or organisations
without originator authorisation.
PDP-11 SYSTEMS
PDP-11/90 Family Announcement
Digital's highest performance PDP-11 processor
11/93 Q-Bus system (MicroPDP-11/93)
11/94 UNIBUS system
Upgrades from many previous systems
Approximately:
double the speed of the 11/73
40% faster than 11/83, 11/84
similar price to and replacing 11/83, 11/84
Demonstrating our commitment to PDP-11 customers. Announced May 7th, order
from mid Q4. First shipment July 1990.
INTRODUCTION
The PDP-11/90 family is based on a new 16 bit PDP-11 processor. Available in
2 MB or 4 MB memory configurations, the PDP-11/93 and PDP-11/94 are our
highest performance PDP-11 processors. These systems will provide the extra
power, and security that our PDP-11 users have been asking for.
CPU intensive benchmarks comparing the PDP-11/83 and 11/84 to the PDP-11/93
and 11/94, demonstrate up to a 40% performance improvement.
PRODUCT DESCRIPTION
The PDP-11/90 family provides more functionality than the 11/8x family while
preserving backward compatibility. The new KDJ11-E* CPU used in 11/93, 11/94
and upgrades, provides more power with increased on-board functionality.
based on proven 18Mz J-11 & FPJ-11 chips
2MB or 4MB memory on board
DMA & CPU operate in parallel
8 serial lines on board
battery operated clock and calendar
8KW of EEPROM for user customisation
The KDJ11-E* processor provides a single-quad module; SBC implementation of
the functional capabilities found on four or five modules in current PDP-11
designs. This integration of capabilities delivers a unique form factor
advantage, that is coupled with hardware and software compatibility and
complemented by enhanced performance, to ensure that the KDJ11-E* will satisfy
a wide range of PDP-11 design and application requirements.
Like its predecessors, the KDJ11-E* is based on the DCJ11 microprocessor chip
set and the FPJ11 floating point coprocessor. This CMOS implementation of the
full PDP-11 instruction set and memory management architecture has been
integrated with an enhanced, pipelined memory subsystem, employing 2 or 4 MB
of fast, 70NS parity memory, to deliver improved system performance. These
KDJ11-E* advantages have been further extended by the addition of eight
buffered, programmable, asynchronous serial lines and the integration of a
time-of-year (TOY) clock.
Other KDJ11-E* enhancements include: a 64 Kbyte bootstrap/diagnostic ROM with
100% diagnostic coverage; an eight Kbyte EEPROM for user customisation: a
console interface facility that makes it easy to configure the functional
capabilities of the processor, for optimum performance across a full spectrum
of PDP-11 systems and applications.
FEATURES
18 MHz DCJ11/FPJ11 microprocessor based SBC design implementation.
2 or 4 MB of fast parity memory (4 Mbit, 70 nanosecond DRAMS).
Pipelined, concurrent processing memory subsystem (enhanced compute and
I/O performance).
Eight-line buffered, programmable asynchronous serial interface.
TOY clock; battery powered time-of-year clock.
Private Memory Interconnect (PMI) protocol support (for UNIBUS systems).
64 Kbyte bootstrap/diagnostic ROM facility (100% diagnostic coverage).
8 Kbyte EEPROM facility for application customisation.
Easy-to-use, menu-driven console monitor facility (for setting functional
configuration parameters).
Compact, quad module design that conserves backplane space and consumes
less power for added configuration flexibility.
PDP-11 software compatibility and support for the most current versions
of PDP-11 family operating systems.
PERFORMANCE
The KDJ11-E* delivers more performance and provides more functional
capabilities than any previous PDP-11 processor. In typical compute and I/O
intensive benchmarks, the KDJ11-E* has demonstrated performance improvements
that average approximately 40% higher than current high-end products based on
the KDJ11-BF. Given these results, typical customer applications are expected
to demonstrate similar performance gains.
However, when positioning the KDJ11-E*, please note that systems performance
is a complex function of many application specific and configuration dependent
variables, and the KDJ11-E* performance can be expected to vary accordingly.
In general, the KDJ11-E* has been optimized for enhanced performance in
demanding multiuser, multitasking, compute and I/O intensive applications.
POSITIONING
The product offers continued investment protection for our PDP-11 installed
base that are not yet ready to grow to Digital 32-bit systems, or are
committed to the 16-bit operating systems for the foreseeable future. The
products are also strategic, in that they provide a defensive sales
alternative to an installed base customer considering migration to a
competitor's alternative. With the higher performance of the PDP-11/9x
processors, Digital can provide a simple "We still have your solution, and we
have made it even better" answer to competitive thrusts into our 16-bit
customer base.
While individual application results will vary, the PDP-11/90 family provides
a growth path and continued investment protection for our PDP-11 customers.
These products will stimulate demand in accounts which have remained dormant
to PDP-11 upgrades and will provide growth path and continued investment
protection for Digital's installed base.
Further strategic opportunities may exist, for accounts considering emerging
market opportunities in Eastern Europe, where many PDP-11 systems are
currently installed. All PDP-11 systems are exportable world-wide, provided
other export licencing provisions restricting use and user are met. The 16
bit Micro PDP-11/90 systems can be the basis of an entry level, readily
exportable product, establishing market share that can be grown and eventually
migrated to Digital's 32-bit platforms.
SOFTWARE SUPPORT
The PDP-11/9x processor is compatible with current versions of all PDP-11
operating systems. Although designed to be compatible with older versions of
software, we are not able to test all possible combinations. Therefore latest
versions are recommended:
Operating System Version At SSB
RSX-11M, RSX-11S V4.6 Q2/FY90
RSX-11M-PLUS V4.3 Q3/FY90
VAX Coprocessor/RSX V1.0 Spring 89
RT-11 V5.5 Winter 89
DSM-11 V4.1 Spring 89
RSTS/E V9.7 Summer 87
V10.0 Summer 90
ORDERING INFORMATION
Announcement: May 7th
Orderable: Mid Q4/FY90
First shipment: early Q1/FY91
================================================================================
How To Create And Use System Recovery Media (RECOVR.COM)
Previous to RSTS/E V9, most customers used SAVRES (either to tape or IMAGE
copies to disk) to keep as recovery backups. Unfortunately SAVRES is not the
easiest of products to maintain and so with the advent of V9, a new feature
was added to enable system recovery media to be generated.
BACKUP is now the preferred medium and indeed, since V9.7, SAVRES cannot be
used to create tape copies, and at some future date SAVRES will no longer be
able to create disk images either.
So how do we create tapes that can be used for disaster recovery?
This is a two step procedure. The first is to create BACKUP sets of the
various files necessary to recreate the customers disks. The second is to
create a bootable recovery tape that can be used when we need to restore the
files from the BACKUP sets onto a system where we have no other bootable system
disks. The bootable recovery tape should be dedicated to that purpose i.e
it should not be used to store customer or system data.
The creation of the BACKUP sets is dependent upon the customer and their BACKUP
strategy, some may take complete backups every day others might use an
incremental backup strategy.
e.g.
BACKUP/ACCOUNT/VERIFY [*,*]*.* MS0:BACKUP.BCK
or
BACKUP/ACCOUNT/VERIFY [34,*]*.* MS0:TUESDY.BCK
The bootable recovery tape is created by the use of the DCL command file
RECOVR.COM which is located in [0,1].
When we invoke RECOVR.COM with the following command:
@[0,1]RECOVR
The following information and questions will be asked (these may be slightly
different from version to version)
Do you want the RECOVR.LOG file to be created? <YES>
The output device you specify to be the
minimum recovery kit will be initialized.
It should be physically, but not logically
mounted.
Output device?
Density? <use system default>
Initializing output tape...
Output device initialized.
Building recovery kit...
Recovery kit completed.
After the tape has been generated it will contain the following files
ERR.ERR
SYSINI.COM
BACKUP.TSK
SYSGEN.SIL
INIT.SYS
DCL.RTS
RSX.RTS
IMPORTANT NOTES
===============
If some of the above files do not exist in [0,1] no error message is output,
but it results in a non useable bootable tape. Always check that the files
exist before creating a bootable recovery tape.
When recovering a V9.x system a bootable recovery tape that has been generated
on the same V9.x system must be used, we cannot use a V9.y tape. e.g. a V9.5
bootable recovery tape on a V9.6 system.
Only one bootable recovery tape is needed per RSTS/E version i.e. we do not
have to go through both stages every time. We only need to create the bootable
recovery tape once (or twice to be on the safe side) usually when we have
created a new version of RSTS/E.
So we now have two or more tapes that can be used should we ever need to
restore a customers file system, the important question now, is how to use
these tapes.
If we have a disk that is bootable with the various system files necessary,
then the customer only has to boot that disk and restore direct from the
BACKUP sets whichever files they need to. The complications come when there
is no disk that we can boot e.g. a single disk system where the disk has
been blasted.
In these circumstances follow the procedure below:
1) Create minimal system disk
Boot the bootable recovery tape
RSTS/E V9.7 (MS0) INIT V9.7-08
Today's Date?
Current time?
Installing RSTS on a new system disk <Yes>
- If you wish to DSKINT the pack then answer yes, you
will then go into the DSKINT dialogue and miss the next
question. This is normally the route we would go through,
unless we suspect that there may be data left intact on
the customers disk.
- If you say no you will be asked the next question.
Are you updating an existing RSTS system disk <Yes>
- If you say no then you are taken back to OPTION
_ If you say yes then you are asked which disk you wish to make
bootable. This route will bring down the files described
above, but no more. The disk could then be checked for any
data important to the customer.
Copying required system files
Enabling only console, disks, and tapes.
RSTS/E V9.7 (MS0) INIT V9.7-08
Start timesharing? <Yes>
...
...
...
...
devices disabled
You may now proceed to recover your system
by using the RESTORE command.
2) Restore the files.
At this stage the disk is booted and we are at timesharing, the customer can
then proceed to restore the files from their BACKUP sets.
e.g. restore/ver/acc/replace ms0:backup.bck [*,*]*.*
Note the /replace switch, this must be used. Otherwise the next time the system
is re-booted instead of the system coming up mounting the various disks,
setting up logicals etc, the message
You may now proceed to recover your system
by using the RESTORE command.
will be printed. If you should get this message you will have to restore the
file [0,1]SYSINI.COM from the customers BACKUP set.
e.g. restore/ver/replace/select=[0,1]sysini.com ms0:backup.bck [0,1]sysini.com
To Distribution List:
PETER MILLINGTON @LZO,
ALAN BURROWS @BIO,
CHAS. MOORE @BIO,
SIMON HALL @BIO,
LAWRENCE @TYNE @VAXMAIL,
RICHARD JONES @NLC,
ROY ARMITAGE @LZO,
JOHN CRAMPIN @BIO,
MIKE NORRIS @OLO,
BIRD @TYNE @VAXMAIL,
MATTHEW MUIRHEAD @BIO,
BILL ARTHUR @UVO,
STEVE GIBBONS @OLO,
MARTIN LIVINGSTONE @OLO,
GARNETT @KERNEL @VAXMAIL,
BRIAN LOXTON @BIO,
NEIL BOND @BIO,
LES AYLESBURY @UBO,
BARNARD @SEDSWS @VAXMAIL,
HORNBY_K @UBOHUB @VAXMAIL,
HAYTHORNED @ERIC @VAXMAIL,
ROGER BUCKLE @UBO,
VIC TAYLOR @WLO,
DAVE SCOBBIE @OLO,
PAUL RUSHWORTH @LZO,
HUGH CLELAND @EDO,
DAVID PAYNE @BSO,
GURSARN BANSAL @BIO,
ERIC BORKALA @LZO,
ANDY RYMER @LZO,
IAN STARTUP @LZO
|
95.2 | RSTS PATCHES FOR 11-93/94 | KERNEL::GARNETT | | Sat Jun 23 1990 23:13 | 211 |
|
RSTS/E Newsletter Issue 3 - June 1990
COPYRIGHT (c) 1990 by Digital Equipment Corporation.
----------------------------------------------------
Contained in this months newsletter are details of two patches which are
needed for the new 11/93 and 11/94 processors to run on RSTS/E V9.7. Both
patches MUST be installed.
NOTE
The Time-Of-Year clock included in the 11/93 and 11/94 processors is
not supported in RSTS/E V9.7; this support is included in V10.0.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
PARITY MEMORY HANDLING FOR MICROPDP-11/93 AND PDP-11/94 SYSTEMS
The on-board parity memory of the MicroPDP-11/93 and PDP-11/94 processors
is not compatible with other PDP-11 parity memory. As such, RSTS/E
software will not operate properly with these new processors.
RSTS/E V10.0 includes new code to properly handle this new type of parity
memory. The feature patch provided by this article and article 3.1.12
modifies RSTS/E V9.7's handling of parity memory errors so that V9.7 can
run on the new MicroPDP-11/93 and PDP-11/94 processors. Installing this
patch and the patch included in article 3.1.12 will cause RSTS/E to crash
whenever any memory parity error occurs, even one that would normally be
considered "recoverable."
Customers purchasing MicroPDP-11/93 or PDP-11/94 systems are encouraged to
upgrade their RSTS/E software to V10.0 to allow recovery from certain
memory parity errors. The feature patches provided by this article and
article 3.1.12 are only intended as a temporary solution until RSTS/E V10.0
is installed.
These feature patches are intended only for RSTS/E systems running on the
MicroPDP-11/93 and PDP-11/94 processors. However, they may also be
installed on other processors with ADDRESS type parity memory CSRs, such as
the MicroPDP-11/73, MicroPDP-11/83, and PDP-11/84.
These feature patches must NOT be installed on any system using NO ADDRESS
type parity memory CSRs. To determine if the memory on your system is NO
ADDRESS parity memory, use the Memory table suboption PARITY of the DEFALT
option of INIT.SYS. If the parity CSR address displayed is followed by the
text "(NA)", then the memory on your system is NO ADDRESS parity memory;
this feature patch and the feature patch in article 3.1.12 should NOT be
installed.
As noted above, full support for the MicroPDP-11/93 and PDP-11/94 is
included in RSTS/E V10.0 and later.
NOTE
You must also install Feature Patch 3.1.12F to support
the MicroPDP-11/93 and PDP-11/94 processors in RSTS/E
V9.7.
RSTS/E Software Dispatch, August 1990
RSTS/E V9.7 Seq 1.1.16 F
Initialization
INIT.SYS Program Patches 1 of 2
PROCEDURE:
1. This feature patch changes the parity memory handling in the RSTS/E
INIT.SYS program so that the MicroPDP-11/93 and PDP-11/94 processors
can be used. Since patching the distribution medium is not
recommended, the patch must be installed every time the COPY option is
used to copy INIT.SYS from the distribution medium. Any copy of the
patched INIT.SYS will propagate the feature.
2. The patch described in Step 4 below should be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? INIT.SYS
3. This patch can be installed using ONLPAT, the on-line patching program,
but this is not recommended:
RUN $ONLPAT
Command File Name? <CR> (RETURN for manual patch installation)
File to patch? INIT.SYS
File found in account [0,1]
4. The patch is as follows:
Base address? SATBUF
Offset address? 3614
Base Offset Old New?
051000 003614 105767 ? 4737
051000 003616 022664 ? PATCH
051000 003620 100003 ? 1003
051000 003622 012725 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? PATCH
Offset address? 0
Base Offset Old New?
044306 000000 000000 ? 122737
044306 000002 000000 ? 5
044306 000004 000000 ? JCPTYP
044306 000006 000000 ? 1002
044306 000010 000000 ? 5037
044306 000012 000000 ? 177746
044306 000014 000000 ? 207
044306 000016 ?????? ? ^Z (CTRL/Z for new offset)
RSTS/E Software Dispatch, August 1990
RSTS/E V9.7 Seq 1.1.16 F
Initialization
INIT.SYS Program Patches 2 of 2
Offset address? ^Z (CTRL/Z for new base)
Base address? DEFALT
Offset address? 11156
Base Offset Old New
110002 011156 001363 ? <CR> (no change; verify only)
110002 011160 104400 ? 4737
110002 011162 140600 ? DEFPAT+20
110002 011164 016710 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? DEFPAT
Offset address? 20
Base Offset Old New?
146614 000020 000000 ? 122737
146614 000022 000000 ? 5
146614 000024 000000 ? JCPTYP
146614 000026 000000 ? 1402
146614 000030 000000 ? 104400
146614 000032 000000 ? 140600
146614 000034 000000 ? 5315
146614 000036 000000 ? 62716
146614 000040 000000 ? 4
146614 000042 000000 ? 207
146614 000044 ?????? ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? $$0101
Offset address? 0
Base Offset Old New?
151264 000000 ?????? ? Q!100000
151264 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
RSTS/E Software Dispatch, August 1990
RSTS/E V9.7 Seq 3.1.12 F
Monitor
Monitor Patches 1 of 1
PROCEDURE:
1. This feature patch changes the parity memory handling in the RSTS/E
V9.7 monitor so that the MicroPDP-11/93 and PDP-11/94 processors can be
used. It may be installed in any target monitor SIL to be used on
MicroPDP-11/93 or PDP-11/94 systems.
2. The patch described in Step 4 below can be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? <CR> (RETURN for installed monitor SIL)
3. This patch can be installed using ONLPAT, the on-line patching program.
Note, however, that if the patch is applied online to the installed
monitor SIL, it will not take effect until the system is re-booted.
RUN $ONLPAT
Command File Name? <CR> (RETURN for patch installation)
File to patch? <CR> (RETURN for installed monitor SIL)
File found in account [0,1]
4. The patch is as follows:
Module name? RSTS
Base address? RTIISP
Offset address? 4476
Base Offset Old New?
?????? 004476 101524 ? 524
?????? 004500 014400 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? $$0301
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!4000
?????? 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
|
95.3 | RSTS/E V10 | KERNEL::GARNETT | | Tue Sep 25 1990 08:11 | 176 |
|
RSTS/E Newsletter Issue 4 - September 90
At long last the arrival of RSTS/E V10 is on the near horizon, and so to whet
all your appetites, contained in this months newsletter are details of some of
the features of V10.
The latest date of release is planned to be early November, but the dates have
been put back that often that I would not be surprised if it was to be delayed
beyond that.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
ANNOUNCING RSTS/E VERSION 10.0
Digital announces its continued commitment in providing quality PDP-11
software to enhance your investment with the release of RSTS/E V10.0
(Q*430). This latest version supports the new Micro PDP-11/93 and PDP-11/94
processors as well as adding features that enhance the capability and
reliability of RSTS/E.
PRODUCT DESCRIPTION
RSTS/E V10.0 is a new release that updates and supports the hardware and
software features of RSTS/E V9.7. This release contains corrections to
known problems that were found in V9.7 and prior versions, plus the
addition of several new features.
KEY BENEFITS OF V10.0
1) Better productivity, through Command Line Editing and
Command recall
2) Easier system management, through the Operator/Message
Services, improvements to the BACKUP/RESTORE package,
and enhancements to Print/Batch Services.
3) Simpler installation, especially when performing
upgrades from earlier versions.
4) Improved performance and reliability, in areas such
as LOGIN and MSCP disk I/O.
5) Many new features and enhancements
6) Revised documentation plus a new V10.0 Internals and
Data Structures Manual now included in the documentation set.
NEW FEATURES FOR V10.0
o Command Line Editing and Command Recall
V10.0 provides VMS-compatible Command Line Editing and Command
Recall. Command Line Editing allows full editing of your current
command line, while Command Recall allows you to recall your most
recently entered commands. Both features are available from DCL,
from other run-time systems or from within a user program. The
combination of these two features significantly improves
productivity by reducing the keystrokes required to enter commands
or correct typing errors.
o Operator/Message Services - OMS
The Operator/Message Services (OMS) package represents a
significant enhancement to the overall management of RSTS/E
systems. Combined with the Print/Batch Services (PBS) package and
the BACKUP/RESTORE package, OMS functionally replaces the OPSER
package. (The OPSER package will still be included for customers
that need to continue using it.)
o Backup/Image
BACKUP/IMAGE is a new feature that allows for asynchronous copying
of a file structured RSTS/E disk to a newly initialized disk. You
can use this feature to perform disk-to-disk backups or to copy a
disk's data to a newer disk when upgrading your hardware.
A related new command, BACKUP/COPY, allows users to perform
high-speed file or account transfers from one disk to another.
o Improved Terminal Server Support
RSTS/E V10.0 supports host-initiated LAT connections. With this
feature, system programs such as PBS, and user applications can now
access printers, modems or other devices connected to LAT Terminal
Servers.
o PBS Enhancements
Several enhancements to Print/Batch Services (PBS) make the package
easier to use and also help achieve the goal of replacing the OPSER
package.
o DCL Enhancements
Many new DCL commands have been added to support the new features
and capabilities introduced with V10.0. Most enhancements center
around VMS compatibility.
o Extended Logicals
V10.0 lifts several longstanding restrictions around the use of
logicals. Users can now define significantly more user logicals
than before. In addition, both system and user logical names can
now be up to 15 characters long and can include underscore
characters.
o New RSTS/E Internals Manual
The revised V10.0 documentation set includes a new RSTS/E V10.0
Internals Manual that provides excellent technical information
about the internal data structures and operation of the RSTS/E
operating system. The manual will prove invaluable to RSTS/E
technical support people and system programmers.
DOCUMENTATION
V10.0 includes new versions of the most commonly used manuals in the RSTS/E
documentation set.
New binders and dividers are also included with V10.0 to give your RSTS/E
documentation set a new look. The traditional Chinese Red binder color has
been replaced with the Digital standard gray.
To Distribution List:
PETER MILLINGTON @LZO,
ALAN BURROWS @BIO,
CHAS. MOORE @BIO,
SIMON HALL @BIO,
LAWRENCE @TYNE @VAXMAIL,
RICHARD JONES @NLC,
ROY ARMITAGE @LZO,
JOHN CRAMPIN @BIO,
MIKE NORRIS @OLO,
BIRD @TYNE @VAXMAIL,
MATTHEW MUIRHEAD @BIO,
BILL ARTHUR @UVO,
STEVE GIBBONS @OLO,
MARTIN LIVINGSTONE @OLO,
GARNETT @KERNEL @VAXMAIL,
BRIAN LOXTON @BIO,
NEIL BOND @BIO,
LES AYLESBURY @UBO,
BARNARD @SEDSWS @VAXMAIL,
HORNBY_K @UBOHUB @VAXMAIL,
HAYTHORNED @ERIC @VAXMAIL,
ROGER BUCKLE @UBO,
VIC TAYLOR @WLO,
DAVE SCOBBIE @OLO,
PAUL RUSHWORTH @LZO,
HUGH CLELAND @EDO,
DAVID PAYNE @BSO,
GURSARN BANSAL @BIO,
ERIC BORKALA @LZO,
ANDREW RYMER @LZO,
IAN STARTUP @LZO
|
95.4 | | KERNEL::GARNETT | | Tue Jan 29 1991 16:29 | 511 |
|
RSTS/E Newsletter Issue 5 - February 1991
COPYRIGHT (c) 1991 by Digital Equipment Corporation.
----------------------------------------------------
Well, at long last RSTS/E V10 was released on the streets in early January,
with all those nice new features that were discussed in previous issues.
Details here of several patches required to V10, one for problems with
RQDX controllers and the others for problems with systems crashing. I'll
also let you know of any other patches on V10 that crop up which might cause
problems that could be confused with hardware faults.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
Mandatory Patches for RSTS/E V10.0
February 1991
1.1.1 M - INIT MAY FAIL DURING SOME RQDX CONTROLLER ACCESSES
3.1.1 M - DEASSIGN SYS CALL MAY CRASH SYSTEM
3.4.3 M - HOST-INITIATED LAT CONNECTIONS MAY CRASH THE SYSTEM
3.7.1 M - ATTACHING TO A DETACHED JOB MAY CRASH THE SYSTEM
3.7.2 M - DYNAMIC REGION CREATION MAY CRASH SYSTEM
RSTS/E Software Dispatch, January 1991
RSTS/E V10.0 Seq 1.1.1 M
Initialization
INIT.SYS Program Patches 1 of 2
INIT MAY FAIL DURING SOME RQDX CONTROLLER ACCESSES - MANDATORY PATCH
PROBLEM:
During some accesses of disks on RQDX controllers INIT may hang or abort
with an unexpected trap through vector 4. This may occur during the DSKINT
option, during rebuilding of a disk, or after a power failure.
SOLUTION:
The patching procedure detailed below corrects the above problem. This
problem will be corrected in some future release.
PROCEDURE:
1. This is a required patch to the RSTS/E INIT.SYS Program. Since
patching the distribution medium is not recommended, the patch must be
installed every time the COPY option is used to copy INIT.SYS from the
distribution medium. Any copy of the patched INIT.SYS will propagate
the fix.
2. The patch described in Step 4 below can be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? INIT.SYS
3. This patch can also be installed using ONLPAT, the on-line patching
program:
RUN $ONLPAT
Command File Name? <CR> (RETURN for manual patch installation)
File to patch? INIT.SYS
File found in account [0,1]
4. The patch is as follows:
Base address? PATCH
Offset address? 4
Base Offset Old New?
045352 000004 103003 ? 103005
045352 000006 004737 ? 10046
045352 000010 005226 ? 4737
045352 000012 000261 ? GETCLK
RSTS/E Software Dispatch, January 1991
RSTS/E V10.0 Seq 1.1.1 M
Initialization
INIT.SYS Program Patches 2 of 2
045352 000014 000207 ? 12600
045352 000016 000000 ? 261
045352 000020 000000 ? 207
045352 000022 ?????? ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? $$0101
Offset address? 0
Base Offset Old New?
152034 000000 ?????? ? Q!1
152034 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
RSTS/E Software Dispatch, February 1991
RSTS/E V10.0 Seq 3.1.1 M
Monitor
Monitor Patches 1 of 2
DEASSIGN SYS CALL MAY CRASH SYSTEM - MANDATORY PATCH
PROBLEM:
The DEASSIGN SYS call, when issued without parameters, causes the command
recall buffer to be corrupted such that recalling a command line may crash
the system. In addition, it does not deassign some logicals as expected.
Since this SYS call is used by the LOGIN program, the crash may occur
shortly after issuing a LOGIN command from a logged-in account.
SOLUTION:
The patching procedure detailed below corrects the above problem. This
problem will be corrected in some future release.
PROCEDURE:
1. This is a required patch to the RSTS/E V10.0 executives. It must be
installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? <CR> (RETURN for installed monitor SIL)
3. This patch can be installed using ONLPAT, the on-line patching program.
Note, however, that if the patch is applied online to the installed
monitor SIL, it will not take effect until the system is re-booted.
RUN $ONLPAT
Command File Name? <CR> (RETURN for patch installation)
File to patch? <CR> (RETURN for installed monitor SIL)
File found in account [0,1]
RSTS/E Software Dispatch, February 1991
RSTS/E V10.0 Seq 3.1.1 M
Monitor
Monitor Patches 2 of 2
4. The patch is as follows:
Module name? EMT
Base address? CAL@OVR
Offset address? 3540
Base Offset Old New?
?????? 003540 012701 ? 137
?????? 003542 000010 ? EMTPAT@OVR
?????? 003544 000401 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? EMTPAT@OVR
Offset address? 0
Base Offset Old New?
?????? 000000 000000 ? 12701
?????? 000002 000000 ? 10
?????? 000004 000000 ? 5726
?????? 000006 000000 ? 207
?????? 000010 000000 ? 5764
?????? 000012 000000 ? 30
?????? 000014 000000 ? 1004
?????? 000016 000000 ? 12716
?????? 000020 000000 ? CLRLOG@OVR-76
?????? 000022 000000 ? 12746
?????? 000024 000000 ? CLRLOG@OVR
?????? 000026 000000 ? 207
?????? 000030 ?????? ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? FSS@OVR
Offset address? 6056
Base Offset Old New?
?????? 006056 005764 ? 4737
?????? 006060 000030 ? EMTPAT@OVR+10
?????? 006062 001352 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? ^Z (CTRL/Z for new module)
Module name? RSTS
Base address? $$0301
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!1
?????? 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
RSTS/E Software Dispatch, December 1990
RSTS/E V10.0 Seq 3.4.3 M
Monitor
Terminal Service Patches 1 of 2
HOST-INITIATED LAT CONNECTIONS MAY CRASH THE SYSTEM - MANDATORY PATCH
PROBLEM:
Attempting to establish a LAT connection from your RSTS/E system (a
Host-Initiated Connection) to a busy port on a terminal server may crash
the system.
SOLUTION:
The patch procedure detailed below will correct the above problem.
PROCEDURE:
1. This is a required patch to the RSTS/E V10.0 executive. It must be
installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? <CR> (RETURN for installed monitor SIL)
3. This patch can be installed using ONLPAT, the on-line patching program.
Note, however, that if the patch is applied online to the installed
monitor SIL, it will not take effect until the system is re-booted.
RUN $ONLPAT
Command File Name? <CR> (RETURN for patch installation)
File to patch? <CR> (RETURN for installed monitor SIL)
File found in account [0,1]
4. The patch is as follows:
Module name? LAT
Base address? LATAPP@OVR
Offset address? 2200
Base Offset Old New?
?????? 002200 077067 ? 77072
?????? 002202 000137 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? ^Z (CTRL/Z for new module)
Module name? RSTS
Base address? $$0304
RSTS/E Software Dispatch, December 1990
RSTS/E V10.0 Seq 3.4.3 M
Monitor
Terminal Service Patches 2 of 2
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!4
?????? 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
RSTS/E Software Dispatch, December 1990
RSTS/E V10.0 Seq 3.7.1 M
Monitor
File Processor Patches 1 of 2
ATTACHING TO A DETACHED JOB MAY CRASH THE SYSTEM - MANDATORY PATCH
PROBLEM:
Attempting to attach to a detached job that originated from a LAT
connection may crash the system.
SOLUTION:
The patch procedure detailed below will correct the above problem.
PROCEDURE:
1. This is a required patch to the RSTS/E V10.0 executive. It must be
installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? <CR> (RETURN for installed monitor SIL)
3. This patch can be installed using ONLPAT, the on-line patching program.
Note, however, if the patch is applied online to the installed monitor
SIL, it will take effect immediately unless you have installed the
overlay file. If you have installed the overlay file, then you must
remove and reinstall the file (see Step 5.) for the patch to take
effect.
RUN $ONLPAT
Command File Name? <CR> (RETURN for patch installation)
File to patch? <CR> (RETURN for installed monitor SIL)
File found in account [0,1]
4. The patch is as follows:
Module name? OVR
Base address? LOG3
Offset address? 410
Base Offset Old New?
?????? 000410 016146 ? 10546
?????? 000412 000006 ? 5005
?????? 000414 014302 ? ^Z (CTRL/Z for new offset)
Offset address? 426
RSTS/E Software Dispatch, December 1990
RSTS/E V10.0 Seq 3.7.1 M
Monitor
File Processor Patches 2 of 2
Base Offset Old New?
?????? 000426 012661 ? 12605
?????? 000430 000006 ? 240
?????? 000432 016502 ? ^Z (CTRL/Z for new offset)
Offset address? 750
Base Offset Old New?
?????? 000750 105261 ? 4767
?????? 000752 000006 ? 760-754
?????? 000754 077014 ? ^Z (CTRL/Z for new offset)
Offset address? 760
Base Offset Old New?
?????? 000760 000000 ? 5705
?????? 000762 000000 ? BEQ+2
?????? 000764 000000 ? 105261
?????? 000766 000000 ? 6
?????? 000770 000000 ? 207
?????? 000772 ?????? ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? LOG3SI
Offset address? 0
Base Offset Old New?
?????? 000000 000760 ? 772
?????? 000002 ?????? ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? ^Z (CTRL/Z for new module)
Module name? RSTS
Base address? $$0307
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!1
?????? 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
5. If you have used the $ INSTALL/OVERLAY command then you must issue the
following commands for the patch to take effect:
$ REMOVE/OVERLAY
$ INSTALL/OVERLAY filespec
RSTS/E Software Dispatch, January 1991
RSTS/E V10.0 Seq 3.7.2 M
Monitor
File Processor Patches 1 of 2
DYNAMIC REGION CREATION MAY CRASH SYSTEM - MANDATORY PATCH
PROBLEM:
Use of the UU.RTS directive to create a dynamic region may cause a memory
corruption that will, at some later time, cause the system to crash.
SOLUTION:
The patching procedure detailed below corrects the above problem. This
problem will be corrected in some future release.
PROCEDURE:
1. This is a required patch to the RSTS/E V10.0 executives. It must be
installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using the PATCH
option of INIT.SYS:
Option: PATCH
File to patch? <CR> (RETURN for installed monitor SIL)
3. This patch can be installed using ONLPAT, the on-line patching program.
Note, however, that if the patch is applied online to the installed
monitor SIL, it will take effect immediately (see step 5 below).
RUN $ONLPAT
Command File Name? <CR> (RETURN for patch installation)
File to patch? <CR> (RETURN for installed monitor SIL)
File found in account [0,1]
4. The patch is as follows:
Module name? OVR
Base address? RTS1
Offset address? 440
Base Offset Old New?
?????? 000440 016246 ? 42762
?????? 000442 000006 ? 174000
?????? 000444 042726 ? 6
?????? 000446 174000 ? 240
?????? 000450 001005 ? ^C (up-arrow/C to exit;CTRL/C for INIT)
RSTS/E Software Dispatch, January 1991
RSTS/E V10.0 Seq 3.7.2 M
Monitor
File Processor Patches 2 of 2
File to patch? (same as step 2/3 above)
Module name? OVR
Base address? RTS6
Offset address? 156
Base Offset Old New?
?????? 000156 005767 ? 5737
?????? 000160 ?????? ? TYPCRE
?????? 000162 001431 ? ^Z (CTRL/Z for new offset)
Offset address? 200
Base Offset Old New?
?????? 000200 003422 ? 4767
?????? 000202 110562 ? 524
?????? 000204 000006 ? 240
?????? 000206 032762 ? ^Z (CTRL/Z for new offset)
Offset address? 730
Base Offset Old New?
?????? 000730 000000 ? BLE+3
?????? 000732 000000 ? 10562
?????? 000734 000000 ? 6
?????? 000736 000000 ? 207
?????? 000740 000000 ? 5126
?????? 000742 000000 ? 207
?????? 000744 000000 ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? RTS6SI
Offset address? 0
Base Offset Old New?
?????? 000000 000730 ? 744
?????? 000002 ?????? ? ^Z (CTRL/Z for new offset)
Offset address? ^Z (CTRL/Z for new base)
Base address? ^Z (CTRL/Z for new module)
Module name? RSTS
Base address? $$0307
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!2
?????? 000002 ?????? ? ^C (up-arrow/C to exit;CTRL/C for INIT)
5. If you have used the $ INSTALL/OVERLAY command then you must issue the
following commands for the patch to take effect:
$ REMOVE/OVERLAY
$ INSTALL/OVERLAY filespec
|
95.5 | | KERNEL::GARNETT | | Thu May 02 1991 09:34 | 519 |
|
RSTS/E Newsletter Issue 6 - April 1991
COPYRIGHT (c) 1991 by Digital Equipment Corporation.
----------------------------------------------------
Yet another CRASH patch, and one for the DECnet freaks out there. Again
both patches could be confused with h/w faults so check these patches are in.
Also, I know our disks never get bad blocks, but for those rare occasions that
they do, there's a guide on what should be done when adding a bad block.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
Crash and H/W related Patches for RSTS/E V10.0
February 1991
1.1.1 M - INIT MAY FAIL DURING SOME RQDX CONTROLLER ACCESSES
3.1.1 M - DEASSIGN SYS CALL MAY CRASH SYSTEM
3.4.3 M - HOST-INITIATED LAT CONNECTIONS MAY CRASH THE SYSTEM
3.7.1 M - ATTACHING TO A DETACHED JOB MAY CRASH THE SYSTEM
3.7.2 M - DYNAMIC REGION CREATION MAY CRASH SYSTEM
April 1991
3.4.7 M - COMMAND RECALL CAN CAUSE SYSTEM CRASHES
31.3.1 M - NCP INAPPROPRIATELY PRINTS "Error getting network device DDB"
================================================================================
Procedure for adding bad blocks:
===============================
Because files will have to be created/deleted, the following should only be
carried out in co-operation with the sites system manager.
1) Locating the bad block and file name.
These can usually be found from a ERRDIS report. The block number is
that number reported for the 'Logical Block Number'. Look for the FCB
(File Control Block), if shown then this will contain both the file name
and the account in which it is situated. The disk drive on which the file
resides can be found from the entry for 'Physical Name'.
2) Save the file contents.
If the disk error is fatal, for instance if we PIP the file and the error
'?Data error on device' is displayed, then the file will in all probability
contain corrupt data, the only safe way might be to restore the file
from a previous backup, indeed because files might have interdependencies
more than one file might have to be restored to ensure data integrity
for the complete system.
However provided that the disk error is recoverable the file can be copied
into a new file name which will fit onto a new portion of the disk. The
new file will have to be checked to ensure that it has the same attributes
as the original.
3) Shut the system down and add the bad blocks.
Adding bad blocks can only be done from the offline options. Once the bad
blocks have been added RSTS/E will check the disk and will display the file
which contains the bad block, it will then allow you to delete that file.
4) Bring the system up and copy back the file.
Once the file containing the bad block has been deleted, the system can
be brought up and the copied file renamed to have the same name as the
original.
Note that the system is not completely brought up at this stage, as the
users application programs might fail in unexpected manners due to one
(or more) of their data files being missing.
5) Bring the system completely up.
Once the system is brought up, the user should first thoroughly test the
system and should then take a backup in case of further problems. During
the test phase no new data should be entered onto the system.
Example for RSTS/E V10.0
========================
There may be subtle (and some not so subtle) differences for versions other
than V10, but the basic principle will remain the same for all versions from
V7.0 upwards.
Again it must be re-iterated that the following should only be carried out
in co-operation with the sites system manager.
In this example, commands that you enter are underlined with -'s.
1) Locating the bad block and file name.
=======================================
From ERRDIS we have found that Logical Block Number 4578 is generating
problems, and that the file [30,76]ART01.FIL contains this block, this
file resides on DR1:.
2) Save the file contents.
=========================
PIP DR1:[30,76]ART01.CPY=DR1:[30,76]ART01.FIL/PR
------------------------------------------------
If there is already a file name of ART01.CPY then the error
"?name or account now exists - file [30,76]ART01 .FIL - continuing"
will be printed. In such a case choose a different name for the copy file,
and keep doing this until no error message is printed.
PIP DR1:[30,76]ART01/S
----------------------
Name .Typ Size Prot Access Date Time Clu RTS Pos Op/rr
DR1:[30,76]
ART01 .FIL 62C <124> 27-Apr-91 27-Apr-91 02:31 PM 16 RSX 1139 0/0
RF:FIX FO:SEQ USED:63:0 RECSI:512
CACHE:ON:RAN
ART01 .CPY 62C <124> 29-Apr-91 29-Apr-91 04:33 PM 16 RSX 17759 0/0
RF:FIX FO:SEQ USED:63:0 RECSI:512
CACHE:ON:RAN
Total of 124 blocks in 2 files in DR1:[30,76]
As can be seen both files have exactly the same attributes, so we can
continue.
3) Shut the system down and add the bad blocks.
===============================================
Shutting the system down should only be carried out by the system manager,
to ensure that all users are off the system and that it is brought down in
a controlled manner.
RUN $SHUTUP
-----------
...
...
...
...
16:36 91.04.29 ######## Final Shutdown Phase ########
Please wait for system to re-boot itself...
RSTS V10.0-L WITHDR (DU0:) INIT V10.0-0L
91.04.29 - 16:36
Start timesharing <Yes>? NO
--
Option: <Start> REFRESH
-------
91.04.29 - 16:37
Disk? DR1:
----
Rebuild? NO
--
REFRESH suboption? BADS
----
Bads? ADD
---
Block number? 4578 (As determined from ERRDIS)
----
Really add logical block number to BADB.SYS? YES
---
Block number? <CR> (Press the RETURN key if no more blocks)
----
Bads? <CR> (Press the RETURN key)
----
Disk is being rebuilt - wait ...
[30,76]ART01 .FIL has a bad block.
Delete? YES
---
[30,76]ART01 .FIL will be deleted - Confirm (Y or Control/C)? YES
---
Refresh suboption? ^C (Type Control/C)
--
4) Bring the system up and copy back the file.
=============================================
Option: <Start> <CR> (Press the RETURN key)
----
91.04.29 - 16:38
3 Devices disabled
Proceed with system startup <Yes>? NO
--
$RUN PIP
-------
*ART01.FIL=ART01.CPY/RE
----------------------
*^Z
--
5) Bring the system completely up.
=================================
$@[0,1]START.COM
---------------
Proceed with system startup <Yes>? YES
---
........
........
........
RSTS/E is now on the air !!!
The system should now be tested, to ensure the integrity of the data. New
data should not be entered into the system until this check has been made.
================================================================================
Seq. No. 3.4.7 M
Private Delimiters Cause Problems with Command Recall -
MANDATORY PATCH
RSTS/E V10.0
Monitor
Terminal Service Corrections
Problem
Setting a private delimiter character (or mask) causes improper
behavior when recalling commands. The most severe problems oc-
cur when you set the RETURN character (ASCII 13) as a private
delimiter. For example, if you issue the DCL command SET TERMI-
NAL/DELIMITER=13, the following may result:
o A system crash. If you leave DCL then recall a command upon
returning, or if you issue the DCL RECALL command with a text
or index parameter, the system may crash.
o A disjointed RECALL/ALL display. The display may contain
extraneous characters, may contain commands that you did not
type at that level, or may loop continuously, forcing you to
hit CTRL/C to exit the loop.
Another problem may result from setting any character other than
RETURN as a private delimiter. When recalling a command, the
monitor processes each character of the command. If the com-
mand contains a private delimter, the delimiter will cause the
monitor to send the partial command to the program or runtime
system, satisfying the read. The monitor will then continue to
send the rest of the characters of the recalled command.
This may result in undesirable behavior.
RSTS/E Software Dispatch, June 1991 1/4
Seq. No. 3.4.7 M
For example, suppose you issue the DCL command SET TERMI-
NAL/DELIMITER="E", which makes "E" a private delimiter. If you
then hit uparrow (to recall the SET TERMINAL command), the mon-
itor will recall the "S", and upon encountering the private
delimiter "E", will execute S as a command. The monitor will
continue to send characters to your terminal, executing the com-
mand every time an "E" is encountered, until the last character
of the SET TERMINAL command is processed.
Solution
The patching procedure detailed below corrects the above prob-
lems. These problems will be corrected in some future release.
The monitor will now recall a command either to the end of the
command, or until the first private delimiter is encountered.
In the above example, the monitor will stop sending charac-
ters to your terminal as soon as it sees the private delimiter
(E), displaying but not executing the command S. All charac-
ters following the delimiter in the recalled command will be
ignored. They can be recalled later, once the private delimiter
is removed.
NOTE
Digital recommends that you do not use private delimiters
at the DCL level, since if not properly used, could cause
the unexpected execution of a command before you have a
chance to complete it.
Procedure
1. This is a required patch to the RSTS/E V10.0 executive. It
must be installed in all target monitor SILs.
2/4 RSTS/E Software Dispatch, June 1991
Seq. No. 3.4.7 M
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that if the patch is applied
online to the installed monitor SIL, it will not take effect
until the system is re-booted.
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. The patch is as follows:
Module name? KBD
Base address? TERREC@OVR
Offset address? 762
Base Offset Old New?
131634 000762 010446 ? 4437
131634 000764 010546 ? KBDPAT@OVR+126
131634 000766 013777 ? ^Z CTRL/Z for new offset
Offset address? 1036
Base Offset Old New?
131634 001036 120227 ? 4737
131634 001040 000015 ? CHKDLM@OVR
131634 001042 001463 ? 103436
131634 001044 004737 ? 120227
131634 001046 144526 ? 15
131634 001050 103433 ? 1460
131634 001052 012704 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? KBDPAT@OVR
Offset address? 126
RSTS/E Software Dispatch, June 1991 3/4
Seq. No. 3.4.7 M
Base Offset Old New?
120000 000126 000000 ? 4737
120000 000130 000000 ? CHKDLM@OVR
120000 000132 000000 ? 103403
120000 000134 000000 ? 10546
120000 000136 000000 ? 137
120000 000140 000000 ? TERREC@OVR+766
120000 000142 000000 ? 5726
120000 000144 000000 ? 137
120000 000146 000000 ? TERREC@OVR+1016
120000 000150 000000 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0304
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!100
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
4/4 RSTS/E Software Dispatch, June 1991
Seq. No. 31.3.1 M
NCP SHOW AND LIST COMMANDS SOMETIMES DISPLAY INAPPROPRIATE
INFORMATION - MANDATORY PATCH
DECnet/E Utilities for RSTS/E V10.0
DECnet/E Utilities Corrections
Problem
The LIST KNOWN NODES and SHOW KNOWN NODES commands display
inappropriate information. When the executor node record has
been purged, the first node in the data base is displayed twice.
If no nodes are defined in the data base, the error message
"Error getting network device DDB" is displayed.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
Procedure
1. This is a required patch to NCP.TSK.
2. The patch described in Step 4 below can NOT be installed
using the PATCH option of INIT.SYS.
3. This patch must be installed using ONLPAT, the on-line patch-
ing program:
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? DECNET$:NCP.TSK
File found in account [0,16]
RSTS/E Software Dispatch, May 1991 1/2
Seq. No. 31.3.1 M
4. The patch is as follows:
Base address? NOREIN
Offset address? 610
Base Offset Old New?
122500 000610 001000 ? 1010
122500 000612 000016 ? ^C up-arrow/C to exit
2/2 RSTS/E Software Dispatch, May 1991
|
95.6 | | KERNEL::GARNETT | | Thu May 30 1991 03:47 | 247 |
|
RSTS/E Newsletter Issue 7 - June 1991
COPYRIGHT (c) 1991 by Digital Equipment Corporation.
----------------------------------------------------
One more CRASH patch for your files, some miscellaneous information on new
hardware support, and news of the termination of support for RSTS/E V9.7. Note
that is for V9.7, not for the termination of RSTS/E support itself :-)
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
Crash and H/W related Patches for RSTS/E V10.0
February 1991
1.1.1 M - INIT MAY FAIL DURING SOME RQDX CONTROLLER ACCESSES
3.1.1 M - DEASSIGN SYS CALL MAY CRASH SYSTEM
3.4.3 M - HOST-INITIATED LAT CONNECTIONS MAY CRASH THE SYSTEM
3.7.1 M - ATTACHING TO A DETACHED JOB MAY CRASH THE SYSTEM
3.7.2 M - DYNAMIC REGION CREATION MAY CRASH SYSTEM
April 1991
3.4.7 M - COMMAND RECALL CAN CAUSE SYSTEM CRASHES
31.3.1 M - NCP INAPPROPRIATELY PRINTS "Error getting network device DDB"
June 1991
0.1.2 N - NEW HARDWARE SUPPORT FOR RA91 DISKS
0.1.3 N - NEW HARDWARE SUPPORT FOR DECSERVER 90L
0.1.43 N - RSTS/E V9.7 SUPPORT TERMINATION NOTICE
3.4.9 M - COMMAND RECALL MAY CRASH SYSTEM
================================================================================
--------------------------------------------------------------------------------
Seq. No. 0.1.43 N
RSTS/E Version 9.7 Support Termination Notice
RSTS/E V9.7
RSTS/E V9.7 General Notes
System Notes
Digital's support policy for software products is to maintain
each version for a period of six month's after the release of a
newer version.
RSTS/E V10.0 was officially released on December 6, 1990. There-
fore, support for RSTS/E V9.7 is terminated on June 30, 1991.
Customers must have Version 10.0 software installed on their
systems to be supported after that date.
RSTS/E Software Dispatch, July 1991 1
--------------------------------------------------------------------------------
Seq. No. 0.1.2 N
ANNOUNCING NEW HARDWARE SUPPORT FOR RA92 DISKS
RSTS/E V10.0
RSTS/E V10.0 General Notes
System Notes
RA92 disk drives are now available for use on PDP-11 systems.
RA92 disks are high performance 1.5 gigabyte drives. They fea-
ture static dual porting like other DSA family disks. They use
the same form factor as the existing RA90 disk drive. Two RA92
drives fit in a 10.5 inch high portion of a standard rack (using
physical space equivalent to one RA81). Up to six drives can
be configured in the RA92-C/F/H/J cabinet. Up to eight drives
can be configured in the SA800 storage array. Up to six RA92
drives can be configured with up to eight RA70 drives in the
SA850 storage array cabinet (for a total of over 10 gigabytes in
a single 60 inch high cabinet). The RA92-NA is a single drive
to add to existing SA6xx or SA8xx storage array cabinets. The
drives connect to Q-Bus systems using the KDA50 controller and
to UNIBUS systems using the UDA50 controller.
Configuration Data
RA92 disks are initialized with a clustersize of 64. On systems
that have many small files, RA92 disks are recommended for use
as auxiliary storage, rather than as the system disk. For this
reason, the SA850 is an ideal configuration using the RA70 as
the system disk, and the RA92 as a private disk.
PDP-11s support a maximum of 2 UDA50/KDA50 type controllers.
Each controller can be configured with a maximum of 4 drives.
These drives can consist of any combination of RA60, RA70, RA80,
RA81, RA82, RA90, or RA92 drives.
Where multiple systems are configured, larger numbers of drives
can be configured to share the same storage array cabinet, while
being cabled to different systems.
RSTS/E Software Dispatch, July 1991 1
--------------------------------------------------------------------------------
Seq. No. 0.1.3 N
ANNOUNCING NEW HARDWARE SUPPORT FOR DECSERVER 90L
RSTS/E V10.0
RSTS/E V10.0 General Notes
System Notes
The DECserver 90L is Digital's newest terminal server. Through
it, you can connect up to 8 terminals or printers to your Ether-
net network. The DECserver 90L is NOT downline loaded; therefore
it can be configured in mixed networks, or networks which con-
tain only RSTS/E systems. The server can be mounted in the new
DEChub 90 distribution system providing a maximum of 64 terminal
ports.
The DECserver 90L has a maximum terminal speed of 38.4K b/s and
supports one session per terminal port. The DECserver 90L can
be used for terminals or printers but can not be used for modem
connections. It also offers security capabilities through the
use of password protection.
The DECserver 90L provides the ideal terminal server for RSTS/E
sites upgrading from hardwired terminals to an Ethernet Local
Area Network, since only one session at a time can be used. With
its ROM based microcode (no downline loading), it is also the
ideal terminal server for sites with only RSTS/E systems.
RSTS/E Software Dispatch, July 1991 1
--------------------------------------------------------------------------------
Seq. No. 3.4.9 M
Command Recall May Crash System - MANDATORY PATCH
RSTS/E V10.0
Monitor
Terminal Service Corrections
Problem
Under certain rare circumstances, recalling a command may crash
the system, or result in a corrupted RECALL/ALL display. This
problem can only occur if command line recall is used after
detaching and re-attaching to a job on which command line recall
is not currently available.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
Procedure
1. This is a required patch to the RSTS/E V10.0 executive. It
must be installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that if the patch is applied
online to the installed monitor SIL, it will not take effect
until the system is re-booted.
RSTS/E Software Dispatch, July 1991 1/2
Seq. No. 3.4.9 M
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. The patch is as follows:
Module name? KBD
Base address? KBDVRM@OVR
Offset address? 2674
Base Offset Old New?
120242 002674 105061 ? 4737
120242 002676 000106 ? KBDPAT@OVR+150
120242 002700 012737 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? KBDPAT@OVR
Offset address? 150
Base Offset Old New?
120000 000150 000000 ? 105061
120000 000152 000000 ? 106
120000 000154 000000 ? 137
120000 000156 000000 ? TERREC@OVR+2424
120000 000160 ?????? ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0304
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!400
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
2/2 RSTS/E Software Dispatch, July 1991
--------------------------------------------------------------------------------
|
95.7 | Y | KERNEL::GARNETT | | Wed Jul 17 1991 16:18 | 295 |
|
RSTS/E Newsletter Issue 8 - July 1991
COPYRIGHT (c) 1991 by Digital Equipment Corporation.
----------------------------------------------------
One of the patches released from the States (patch number 3.4.6, released
in June 1991) has been found to be faulty, and could result in a system
hanging or crashing. A new patch has been released to correctly rectify the
original problem.
The patch has been released with two forms of the patching procedure, one for
if the customer had already installed the original version of the patch, the
second for if they had not.
Please check with your customers to ensure that they are aware of the problem.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
Seq. No. 3.4.6 M
Inappropriate Character Echoing - MANDATORY PATCH
RSTS/E V10.0
Monitor
Terminal Service Corrections
*** Supersedes June 1991 RSTS/E V10.0
Software Dispatch Seq. No. 3.4.6 ***
Problem
When the error "?Keyboard wait exhausted" (ERR=15%) is returned
to a program, character echoing sometimes behaves inappropri-
ately. Characters can be echoed more than once on the terminal,
often with the whole command redisplayed when a new character is
typed. This behavior makes it nearly impossible to distinguish
what is truly on the command line.
In other situations, character echoing can stop altogether. The
behavior depends on how the program deals with error 15. If the
program resumes to do some processing, character echoing may
stop. If the program issues another read request, the duplicate
character echoing may result.
The behavior described above can also differ depending on the
/[NO]ANSI and the /SCOPE terminal characteristic settings.
Some known examples of failures are:
o After a full logout on a hardwired terminal which is per-
manently set to /SCOPE, if you type ahead while the logout
information is printed, then hit another character after a
few seconds, characters will be displayed twice.
o If the unsupported ATPK program is run interactively (ATPK
LOGFIL=KB:), character echoing will stop after a few seconds
until a delimiter or the DELETE key is hit.
1/6 RSTS/E Software Dispatch, September 1991
Seq. No. 3.4.6 M
o Any user program which traps error 15, and re-issues a read
request, may see either characters echoed more than once, or
perhaps no echoing of characters at all.
NOTE
This patch supersedes June 1991 Article 3.4.6M, which
could result in the system either crashing or getting
hung. This would only occur after a job which had detached
continued running and exited a program. The patch pro-
cedure detailed in Step 4b below removes the June 1991
patch, and fixes the above problem.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
NOTE
This solution results in a minor restriction involving
the displaying of the command line after a "?Keyboard wait
exhausted" error occurs. After this error is generated,
if at the next prompt the user types either CTRL/R or
CTRL/T, or if a broadcast to the terminal occurs, the
command line (if any) may not be positioned correctly
on the screen when it is redisplayed. This could cause
difficulty editing a command line. This condition is only
a problem at the interactive level, and may be cleared by
hitting any delimiter such as RETURN or CTRL/C.
2/6 RSTS/E Software Dispatch, September 1991
Seq. No. 3.4.6 M
Procedure
1. This is a required patch to the RSTS/E V10.0 executive. It
must be installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that if the patch is applied
online to the installed monitor SIL, it will not take effect
until the system is re-booted.
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. This patch supersedes a previously published patch. If you
have not installed June 1991 Article 3.4.6M, apply the patch
described in Step 4a, otherwise install the patch described
in Step 4b.
a. Apply this patch only if you have not installed June 1991
Article 3.4.6M:
3/6 RSTS/E Software Dispatch, September 1991
Seq. No. 3.4.6 M
Module name? DSK
Base address? TIM
Offset address? 242
Base Offset Old New?
?????? 000242 142761 ? 407
?????? 000244 000040 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0304
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!40
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
b. Apply this patch only if you have installed June 1991
Article 3.4.6M:
Module name? DSK
Base address? TIM
Offset address? 242
Base Offset Old New?
?????? 000242 152761 ? 407
?????? 000244 000040 ? <CR> no change; verify only
?????? 000246 000105 ? 104
?????? 000250 000404 ? 152761
?????? 000252 000100 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? KBD
Base address? KBDVRM@OVR
Offset address? 2666
Base Offset Old New?
?????? 002666 042761 ? 142761
?????? 002670 020040 ? 40
?????? 002672 000104 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
4/6 RSTS/E Software Dispatch, September 1991
Seq. No. 3.4.6 M
Module name? KIN
Base address? TERINC@OVR
Offset address? 4304
Base Offset Old New?
?????? 004304 042761 ? 142761
?????? 004306 020040 ? 40
?????? 004310 000104 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? TERCLS
Base address? TERCLS@OVR
Offset address? 1246
Base Offset Old New?
?????? 001246 012361 ? 112361
?????? 001250 000044 ? <CR> no change; verify only
?????? 001252 112323 ? 112361
?????? 001254 042713 ? 45
?????? 001256 020100 ? 112361
?????? 001260 000240 ? 103
?????? 001262 042761 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? GEN
Base address? SCH@OVR
Offset address? 3542
Base Offset Old New?
?????? 003542 004737 ? 016704
?????? 003544 120000 ? 052544
?????? 003546 105064 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? GENPAT@OVR
Offset address? 0
Base Offset Old New?
120000 000000 011204 ? 0
120000 000002 132764 ? 0
120000 000004 000040 ? 0
120000 000006 000105 ? 0
120000 000010 001410 ? 0
5/6 RSTS/E Software Dispatch, September 1991
Seq. No. 3.4.6 M
120000 000012 042764 ? 0
120000 000014 020040 ? 0
120000 000016 000104 ? 0
120000 000020 152764 ? 0
120000 000022 000100 ? 0
120000 000024 000104 ? 0
120000 000026 105064 ? 0
120000 000030 000106 ? 0
120000 000032 013704 ? 0
120000 000034 001016 ? 0
120000 000036 000207 ? 0
120000 000040 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
6/6 RSTS/E Software Dispatch, September 1991
|
95.8 | | KERNEL::GARNETT | | Mon Sep 16 1991 13:22 | 293 |
|
RSTS/E Newsletter Issue 9 - Sept 1991
COPYRIGHT (c) 1991 by Digital Equipment Corporation.
----------------------------------------------------
Yet another CRASH/HANG type problem for your information, and a problem
that may be seen by our customers using servers that are running the LAT
V5.2 protocol.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
Crash and H/W related Patches for RSTS/E V10.0
February 1991
1.1.1 M - INIT MAY FAIL DURING SOME RQDX CONTROLLER ACCESSES
3.1.1 M - DEASSIGN SYS CALL MAY CRASH SYSTEM
3.4.3 M - HOST-INITIATED LAT CONNECTIONS MAY CRASH THE SYSTEM
3.7.1 M - ATTACHING TO A DETACHED JOB MAY CRASH THE SYSTEM
3.7.2 M - DYNAMIC REGION CREATION MAY CRASH SYSTEM
April 1991
3.4.7 M - COMMAND RECALL CAN CAUSE SYSTEM CRASHES
31.3.1 M - NCP INAPPROPRIATELY PRINTS "Error getting network device DDB"
June 1991
0.1.2 N - NEW HARDWARE SUPPORT FOR RA91 DISKS
0.1.3 N - NEW HARDWARE SUPPORT FOR DECSERVER 90L
0.1.43 N - RSTS/E V9.7 SUPPORT TERMINATION NOTICE
3.4.9 M - COMMAND RECALL MAY CRASH SYSTEM
September 1991
3.4.12 M - HOST INITIATED CONNECTIONS TO SOME SERVERS MAY FAIL
3.7.6 M - SYSTEM HANGS DUE TO SETTING TERMINAL WIDTH
================================================================================
Seq. No. 3.4.12 M
HOST INITIATED CONNECTIONS TO SOME SERVERS MAY FAIL - MANDATORY
PATCH
RSTS/E V10.0
Monitor
Terminal Service Corrections
Problem
RSTS does not allow Host Initiated connections to terminal
servers, such as the DECserver 90L, that use the LAT V5.2 pro-
tocol. This problem is only seen when attempting to access a
printer or modem device connected to a server running the V5.2
protocol. Users logging onto the RSTS system from these servers
will not experience this problem.
Issuing a "Show Server" command at your terminal server's "Lo-
cal>" prompt will display the version of the LAT protocol used
by that server.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
Procedure
1. This is a required patch to the RSTS/E V10.0 executives. It
must be installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
RSTS/E Software Dispatch, October 1991 1/2
Seq. No. 3.4.12 M
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that if the patch is applied
online to the installed monitor SIL, it will not take effect
until the system is re-booted.
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. The patch is as follows:
Module name? LAT
Base address? LATAPP@OVR
Offset address? 1634
Base Offset Old New?
?????? 001634 001016 ? 101016
?????? 001636 105761 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0304
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!4000
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
2/2 RSTS/E Software Dispatch, October 1991
================================================================================
Seq. No. 3.7.6 M
System Hangs Due to Setting Terminal Width - MANDATORY PATCH
RSTS/E V10.0
Monitor
File Processor Corrections
Problem
Under certain circumstances, the system may hang after setting
the width of a terminal to a smaller value than its current
width. This may happen if the width is changed on a terminal
which is currently in a KB read state. The system hang could
occur after a broadcast to that terminal, or if a command recall
key is pressed on that terminal.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
Procedure
1. This is a required patch to the RSTS/E V10.0 executive. It
must be installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
RSTS/E Software Dispatch, August 1991 1/4
Seq. No. 3.7.6 M
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that this patch will take effect
immediately (see step 5 below).
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. The patch is as follows:
Module name? OVR
Base address? TRM2
Offset address? 122
Base Offset Old New?
?????? 000122 116100 ? 5000
?????? 000124 000021 ? 156100
?????? 000126 042700 ? 21
?????? 000130 177400 ? 160300
?????? 000132 160300 ? 12737
?????? 000134 012737 ? PR5
?????? 000136 034240 ? PS
?????? 000140 177776 ? 110361
?????? 000142 110361 ? 21
?????? 000144 000021 ? 5003
?????? 000146 116103 ? 156103
?????? 000150 000020 ? 20
?????? 000152 042703 ? 160003
?????? 000154 177400 ? 100001
?????? 000156 160003 ? 5003
?????? 000160 100001 ? 110361
?????? 000162 005003 ? 20
?????? 000164 110361 ? 4767
?????? 000166 000020 ? 756-170
?????? 000170 012737 ? ^Z CTRL/Z for new offset
Offset address? 756
2/4 RSTS/E Software Dispatch, August 1991
Seq. No. 3.7.6 M
Base Offset Old New?
?????? 000756 000000 ? 105761
?????? 000760 000000 ? 106
?????? 000762 000000 ? 1405
?????? 000764 000000 ? 160061
?????? 000766 000000 ? 106
?????? 000770 000000 ? 100002
?????? 000772 000000 ? 5061
?????? 000774 000000 ? 106
?????? 000776 000000 ? 207
?????? 001000 ?????? ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? TRM2SI
Offset address? 0
?????? 000000 000756 ? 1000
?????? 000002 ?????? ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0307
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!40
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
5. If you have used the $ INSTALL/OVERLAY command, then you must
issue the following commands for the patch to take effect:
$ REMOVE/OVERLAY
$ INSTALL/OVERLAY <filespec>
Additionally, if you have issued the $ LOAD/OVERLAY=TERMINAL
command, then you must issue the following commands for the
patch to take effect:
RSTS/E Software Dispatch, August 1991 3/4
Seq. No. 3.7.6 M
$ UNLOAD/OVERLAY=TERMINAL
$ LOAD/OVERLAY=TERMINAL
|
95.9 | | KERNEL::GARNETT | | Fri Dec 20 1991 08:44 | 417 |
|
RSTS/E Newsletter Issue 10 - Dec 1991
COPYRIGHT (c) 1991 by Digital Equipment Corporation.
----------------------------------------------------
Contained in this months letter, one problem that may manifest itself in a wide
variety of ways, and another where a MSCP ot TMSCP device may hang. Also
have you ever had an occasion where a high priority job was hanging the system,
and you could not lower its priority; well in this months copy we show how
on a 11/84 running V10 you can change a jobs priority from the console.
Mail can be sent to
ERIC::STARTUP
or IAN STARTUP @LZO
================================================================================
Crash and H/W related Patches for RSTS/E V10.0
February 1991
1.1.1 M - INIT MAY FAIL DURING SOME RQDX CONTROLLER ACCESSES
3.1.1 M - DEASSIGN SYS CALL MAY CRASH SYSTEM
3.4.3 M - HOST-INITIATED LAT CONNECTIONS MAY CRASH THE SYSTEM
3.7.1 M - ATTACHING TO A DETACHED JOB MAY CRASH THE SYSTEM
3.7.2 M - DYNAMIC REGION CREATION MAY CRASH SYSTEM
April 1991
3.4.7 M - COMMAND RECALL CAN CAUSE SYSTEM CRASHES
31.3.1 M - NCP INAPPROPRIATELY PRINTS "Error getting network device DDB"
June 1991
0.1.2 N - NEW HARDWARE SUPPORT FOR RA91 DISKS
0.1.3 N - NEW HARDWARE SUPPORT FOR DECSERVER 90L
0.1.43 N - RSTS/E V9.7 SUPPORT TERMINATION NOTICE
3.4.9 M - COMMAND RECALL MAY CRASH SYSTEM
September 1991
3.4.12 M - HOST INITIATED CONNECTIONS TO SOME SERVERS MAY FAIL
3.7.6 M - SYSTEM HANGS DUE TO SETTING TERMINAL WIDTH
December 1991
3.7.8 M - POSSIBLE MEMORY CORRUPTION
3.10.1 M - SOME TMSCP OPERATIONS MAY RETRY FOREVER
================================================================================
Note a) all input is entered underlined.
b) unless specifically mentioned a carriage return is entered prior to
entering input.
c) The following offsets are applicable for RSTS/E V10. The console
instructions will differ from processor to processor, the instructions
given are for a 11/84.
d) Try this on your own system before attempting it on a customers !!!!
1) turn the key to ENABLE.
2) flick switch from run to halt and back to run.
3) $S/174000 <- determine contents of processor status,
--- we wait until it shows a user job, i.e. the
top two bits will be set. If they are not
both set then type in a P (to proceed) and
then proceed to step 2 above.
3) 1006/000020 <- determine the job number *2, if this value is
----- 0 then type in a P (to proceed) and proceed
to step 2 above. In this case the job number
equates to job 8.
4) 1020/065172 <- this used to point to the jobs JDB2, it
----- now points to location 30(octal) before the
jobs PPN entry in the table JBPPN. So by
adding 30 onto 65172 i.e. 65222 we can find
the jobs PPN.
5) 65222/017115 <- the PPN, in this case equating to 30,77
------
6) 1030/144100 <- determine address of the jobs JCR
-----
7) 1032/005464 <- determine PAR for APR 6 to map the JCR.
-----
8) because of RSTS's mapping mechanism we now have to calculate the real
22 bit address to look at the JCR
546400 the PAR with two 0's appended
+ 4100 the address of JCR minus the first three bits
552500 address which we use to locate JCR
The start of the jobs name is now located 20 on from the start of the JCR.
9) 552520/023364 <- the first 3 letters in radix-50 form, this
------- equates to FIL
10) 552522/060116 <- the second 3 letters in radix-50 form, this
------- equates to OPN
Thus the job which is currently running is [30,77]FILOPN. We can now set
the jobs priority can be set so that it is suspended.
11) the jobs priority is held in the byte 62(octal) from the start of the jobs
JCR (which in this case is 552500), thus look at the address 552562. Note
the jobs runburst is held in the next byte up.
552562/3000 <- this equates to a job running at priority 0,
------- with a runburst of 6.
To suspend the job, it has to be given a priority of -128, consequently in
this case, without pressing the carriage return key, type in 3200, followed
by a carriage return, then type in a P.
This will suspend the job, and the system will hopefully run as normal.
The user can then determine why that job was running in such a state, and
reset the priority so that it does not hang the system.
================================================================================
Seq. No. 3.7.8 M
POSSIBLE MEMORY CORRUPTION - MANDATORY PATCH
RSTS/E V10.0
Monitor
File Processor Corrections
Problem
When the "Create a Job" system call is used to log a job into a
Keyboard Monitor that is less than 3 K-words in size, a timing
condition exists that may cause memory to become corrupted.
This can result in program failures such as traps to 4 or memory
management violations.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
Procedure
1. This is a required patch to the RSTS/E V10.0 executive. It
must be installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that this patch will take effect
immediately (see step 5 below).
RSTS/E Software Dispatch, December 1991 1/3
Seq. No. 3.7.8 M
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. The patch is as follows:
Module name? OVR
Base address? LOG9
Offset address? 536
Base Offset Old New?
?????? 000536 113700 ? 4767
?????? 000540 ?????? ? 212
?????? 000542 004537 ? ^Z CTRL/Z for new offset
Offset address? 754
Base Offset Old New?
?????? 000754 000000 ? 62760
?????? 000756 000000 ? 2
?????? 000760 000000 ? 60
?????? 000762 000000 ? 113700
?????? 000764 000000 ? FIJOB
?????? 000766 000000 ? 207
?????? 000770 ?????? ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? LOG9SI
Offset address? 0
Base Offset Old New?
?????? 000000 000754 ? 770
?????? 000002 ?????? ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0307
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!200
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
2/3 RSTS/E Software Dispatch, December 1991
Seq. No. 3.7.8 M
5. If you have used the $ INSTALL/OVERLAY command, then you must
issue the following commands for the patch to take effect:
$ REMOVE/OVERLAY
$ INSTALL/OVERLAY filespec
RSTS/E Software Dispatch, December 1991 3/3
================================================================================
Seq. No. 3.10.1 M
SOME TMSCP OPERATIONS MAY RETRY FOREVER - MANDATORY PATCH
RSTS/E V10.0
Monitor
Device Driver Corrections
Problem
During some sequences of operations on TMSCP tape drives (TK50
or TU81) or MSCP disk drives (DU type disks) it is possible that
the driver will retry the operation forever. It will appear that
the job is hung in an MU, DU, or MTA wait state, and the job
cannot be killed.
Solution
The patching procedure detailed below corrects the above prob-
lem. This problem will be corrected in some future release.
Procedure
1. This is a required patch to the RSTS/E V10.0 executive. It
must be installed in all target monitor SILs.
2. The patch described in Step 4 below can be installed using
the PATCH option of INIT.SYS:
Option: PATCH
File to patch? <CR> RETURN for installed monitor SIL
3. This patch can be installed using ONLPAT, the on-line patch-
ing program. Note, however, that if the patch is applied
online to the installed monitor SIL, it will not take effect
until the system is re-booted.
RSTS/E Software Dispatch, November 1991 1/3
Seq. No. 3.10.1 M
RUN $ONLPAT
Command File Name? <CR> RETURN for patch installation
File to patch? <CR> RETURN for installed monitor SIL
File found in account [0,1]
4. The patch is as follows:
Module name? MCP
Base address? CPH@OVR
Offset address? 3632
Base Offset Old New?
?????? 003632 005065 ? 12765
?????? 003634 000000 ? 3
?????? 003636 005065 ? 12
?????? 003640 000020 ? 240
?????? 003642 005065 ? <CR> no change; verify only
?????? 003644 000006 ? <CR> no change; verify only
?????? 003646 012765 ? 5015
?????? 003650 000003 ? 5065
?????? 003652 000012 ? 20
?????? 003654 012765 ? ^Z CTRL/Z for new offset
Offset address? 3732
Base Offset Old New?
?????? 003732 004767 ? <CR> no change; verify only
?????? 003734 177674 ? 177704
?????? 003736 013746 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? UQPORT@OVR
Offset address? 776
Base Offset Old New?
?????? 000776 132764 ? 4737
?????? 001000 000100 ? MCPPAT@OVR
?????? 001002 ?????? ? 240
?????? 001004 001020 ? 1005
?????? 001006 062702 ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? MCPPAT@OVR
Offset address? 0
2/3 RSTS/E Software Dispatch, November 1991
Seq. No. 3.10.1 M
Base Offset Old New?
120000 000000 000000 ? 132764
120000 000002 000000 ? 100
120000 000004 000000 ? CMT@OVR+17
120000 000006 000000 ? 1405
120000 000010 000000 ? 12700
120000 000012 000000 ? 3
120000 000014 000000 ? 132764
120000 000016 000000 ? 100
120000 000020 000000 ? CMT@OVR+17
120000 000022 000000 ? 207
120000 000024 ?????? ? ^Z CTRL/Z for new offset
Offset address? ^Z CTRL/Z for new base
Base address? ^Z CTRL/Z for new module
Module name? RSTS
Base address? $$0310
Offset address? 0
Base Offset Old New?
?????? 000000 ?????? ? Q!1
?????? 000002 ?????? ? ^C up-arrow/C to exit;CTRL/C for INIT
RSTS/E Software Dispatch, November 1991 3/3
|