Power Hawk Series 900 PowerMAX OS Version 6.1 Release Notes

Power HawkTM Series 900
PowerMAX OSTM
Version 6.1 Release Notes
February 2004
R
EA
D
TH INS M
IS TA E B
PR LL EF
O IN OR
D G
E
U
CT
0891089-6.1
Disclaimer
The information contained in this document is subject to change without notice. Concurrent Computer Corporation has taken efforts to
remove errors from this document, however, Concurrent Computer Corporation’s only liability regarding errors that may still exist is to
correct said errors upon their being made known to Concurrent Computer Corporation.
License
Duplication of this manual without the written consent of Concurrent Computer Corporation is prohibited. Any copy of this manual
reproduced with permission must include the Concurrent Computer Corporation copyright notice.
Trademark Acknowledgments
Night Hawk is a registered trademark of Concurrent Computer Corporation.
PowerWorks, PowerMAX OS, NightStar, NightView, NightSim, NightProbe, NightTune, NightGraphics
MAXAda, NightBench and ARMS are trademarks of Concurrent Computer Corporation.
Power Hawk, PowerMAXION and TurboHawk are trademarks of Concurrent Computer Corporation.
UNIX is a registered trademark and X/Motif is a trademark of The Open Group.
POSIX is a registered trademark of the Institute of Electronics and Electrical Engineers.
AltiVec and PowerStack II are trademarks of Motorola Corporation.
PowerPC is a trademark of and used under license from International Business Machine Corporation.
The X Window System is a trademark of The Open Group.
OSF/Motif is a registered trademark of the Open Software Foundation, Inc.
Synergy, VYFD, VGM5, VSS4 ASTRix, Raptor DX and STAR are trademarks of Synergy Microsystems, Inc.
SanDisk is a registered trademark of SanDisk Corporation.
Ethernet is a trademark of the Xerox Corporation.
All other trademarks are the property of the respective owners.
© February 2004 Concurrent Computer Corporation – Al l Rights Reserved
Concurrent Computer Corporation
2881 Gateway Drive
Pompano Beach, FL 33069
Note: Information subject to change without notice.
Contents
1.0. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2.0. Power Hawk Series 900 System Overview . . . . . . . . . . . . . . . . . 2
2.1. Functional Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2. Hardware Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3. Power Hawk Series 900 CPU Board Descriptions . . . . . . . . . . . 3
2.3.1. Model 910 Singlel CPU Board . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3.2. Model 920 Dual CPU Board . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.4. Power Hawk Series 900 Specifications . . . . . . . . . . . . . . . . . . . 3
3.0. Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1. PowerMAX OS Software Documentation . . . . . . . . . . . . . . . . . . 6
3.2. Vendor Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.0. Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.1. Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2. Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.1. System Console Terminal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.2. SCSI Interface Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.3. Disk Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.4. CD-ROM Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.5. Ethernet Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.6. High Speed Driver Enhanced (HSDE) Controller . . . . . . . . . . . . 9
4.2.7. MIL-STD-1553 Version 5 (V5) Interface . . . . . . . . . . . . . . . . . . 9
4.2.8. IEEE488 GPIB Bus Interface Controller . . . . . . . . . . . . . . . . . . 9
4.2.9. Asynchronous Communications Multiplexer (MVC) . . . . . . . . . 9
4.2.10. High-Performance Asynchronous Communications
Multiplexer (MVCS) 9
4.2.11. Real-Time Clock and Interrupt Module (RCIM) . . . . . . . . . . . 10
4.2.12. Single-Head PMC Graphics Card . . . . . . . . . . . . . . . . . . . . . 10
4.2.13. Four-Head PMC Graphics Card . . . . . . . . . . . . . . . . . . . . . . . 10
4.2.14. VMIVME-2510B 64-bit TTL I/O Board . . . . . . . . . . . . . . . . 10
4.2.15. VMIVME-3122 High-Performance 16-bit A/D Converter Board
5.0.
6.1
6.0.
PowerMAX OS Version 6.1 Release Notes
10
4.2.16. VMIVME-4140 Intelligent 32/16-Channel Analog Output Board
10
Available Software Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.1. Package Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.1.1. Description of Standard Packages . . . . . . . . . . . . . . . . . . . . . . 19
5.1.2. Description of Drivers Packages . . . . . . . . . . . . . . . . . . . . . . . 20
5.1.3. Description of TCP/IP Networking Packages . . . . . . . . . . . . . . 23
5.1.4. Description of Network File System Packages . . . . . . . . . . . . . 23
5.1.5. Description of Security Package . . . . . . . . . . . . . . . . . . . . . . . 23
5.1.6. Description of Diskless Package . . . . . . . . . . . . . . . . . . . . . . . 24
5.1.7. Description of FBS Packages . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1.8. Description of Software Development Packages . . . . . . . . . . . 24
5.1.9. Description of X Window System Packages . . . . . . . . . . . . . . . 25
5.1.10. Description of NightStar Tools Packages . . . . . . . . . . . . . . . . 26
5.1.11. Description of License Manager Packages . . . . . . . . . . . . . . . 26
5.1.12. Description of Miscellaneous Packages: . . . . . . . . . . . . . . . . . 27
5.1.13. Power Hawk Series 700/900 Diagnostics . . . . . . . . . . . . . . . . 27
System Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.1. Large SCSI Disk Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
III
6.2. System Disk Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.3. Installation Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.3.1. Custom Mode of Installation . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.3.2. Semi-Automatic Mode of Installation . . . . . . . . . . . . . . . . . . . 31
6.4. Installation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6.5. Installing Additional Packages . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6.6. Installing Service Releases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7.0. Rebooting The System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
8.0. General Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
8.1. XFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
8.2. PCI-to-PCI Bridge Configuration . . . . . . . . . . . . . . . . . . . . . . . . 41
9.0. General Information on Release 6.1 . . . . . . . . . . . . . . . . . . . . . . . 42
9.1. Architecture Unification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
9.2. Enhancements in this Release . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
9.2.1. Dynamic Interrupt CPU Assignments . . . . . . . . . . . . . . . . . . . 42
9.2.2. Loosely-Coupled Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
9.3. Known Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
10.0.
11.0.
12.0.
13.0.
14.0.
IV
. . . . . . . . . Optional Products Available with Release 6.1 45
Compatibility Issues Between Systems . . . . . . . . . . . . . . . . . . 46
Service Releases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Direct Software Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Manual Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
PowerMAX OS Version 6.1 Release Notes
Introduction
1.0.
Introduction
This document provides a general overview of PowerMAX OSTM Release Version 6.1 (Release 6.1).
Release 6.1 is designed to provide support for the newly introduced Concurrent Computer Corporation
Power HawkTM Series 900 systems. There are two models in the Series 900 line which are essentially
identical except that the Model 920 system has two CPUs while the Model 910 has only one CPU. Series
900 systems use a Raptor DXTM/VYFD VMEbus single board computer (SBCs) manufactured by
Synergy Microsystems, Inc. (Note - These release notes will refer to the Synergy Raptor DXTM/VYFD
SBC as the “VYFD SBC”.)
PowerMAX OS is based on UNIX® System V Release 4.2 MP with real-time enhancements provided
by Concurrent Computer Corporation.
Release 6.1 is distributed on two CD-ROM discs:
◆
◆
PowerMAX OS Boot CD
PowerWorksTM Standard Products CD
The PowerMAX OS Boot CD contains a bootable mini-kernel, install and restore software. The
PowerWorks Standard Products CD contains the standard software packages that may be installed once
the software on the PowerMAX OS Boot CD has been installed.
An optional products disc (see below) may also be provided if optional products have been previously
ordered by the user.
◆
PowerWorksTM Optional Products CD
A service release disc (see below) containing updates, enhancements, patches, etc, may also be provided
on an “as-released-basis” to users who are currently under a software maintenance contract.
6.1
◆
PowerMAX OS Version 6.1 Release Notes
PowerWorksTM Service Release CD
1
Power Hawk Series 900 System Overview
2.0.
Power Hawk Series 900 System Overview
2.1.
Functional Overview
The Power Hawk Series 900 is Concurrent’s high-performance computer system for real-time data
acquisition, simulation, and industrial systems application. The Model 910 contains a single PowerPCTM
MPC7455 G4 CPU while the Model 920 contains two CPUs. The Model 910 and the Model 920 are
contained on a VMEbus VYFD SBC with up to 1 GB of memory on a single VME card
The MPC7455 is a 3rd generation G4 PowerPC processor which consists of a processor core, 32 KB
separate L1 instruction and data caches, a 256 KB L2 cache, and an internal L3 controller with tags that
support a 2 MB backside L3 cache through a dedicated high-bandwidth interface. The core is a highperformance superscaler design supporting multiple execution units, including four independent units
that execute AltiVec instructions.
The L3 cache is known as backside cache, because it is directly attached to the CPU via a private bus.
The pre-G4 PPC processor Power Hawk systems had to negotiate for the system bus when accessing
cache. This caused cache accesses to compete with memory accesses from I/O devices. This new cache
scheme provides for better determinism in execution and since the cache bus is private, its speed can be
increased as processor clock speeds increase.
AltiVec computational instructions are executed on four independent, pipelined AltiVec execution units.
A maximum of two AltiVec instructions can be issued in order to any combination of AltiVec execution
units per clock cycle, with as many as 10 AltiVec instructions executing concurrently. Depending on
data size, AltiVec vectors are 4, 8 or 16 elements long. There are 32 AltiVec registers each 128-bits wide.
AltiVec technology can be most accurately thought of as a set of registers and execution units added to
the PowerPC architecture in an analogous manner to the addition of floating point units.
The block diagram below is of a Model 920 which is representative of the Model 910 except for the
additional CPU.
Figure 2-1. Model 920 Block Diagram
100BaseT Ethernet
RS-232
RS-422
Model 920 CPU Board
PMC Expander Board
PowerPC
PowerPC
PowerPC
IEEE-488
IEEE-488
Discovery
Controller
SDRAM
Graphics
UltraSCSI
RCIM
CD-ROM
PCI bus
Reflective
Memory
VME64x
Disk
2
Digital
I/O
MIL-STD
1553
IEEE-488
Customer
VME Card
PowerMAX OS Version 6.1 Release Notes
6.1
A/D
D/A
Power Hawk Series 900 System Overview
2.2.
Hardware Description
The Power Hawk 900 Series systems are available in rack mount or table-top options. The
Series 900 chassis is available in 12-slot or 21-slot 6U VME configurations. The chassis
contains integral peripheral bays and autoranging power supplies. A typical Series 900
chassis with SBCs also shown, is illustrated below.
Figure 2-2. Series 900 Chassis
2.3.
Power Hawk Series 900 CPU Board Descriptions
2.3.1.
Model 910 Singlel CPU Board
The Power Hawk Model 910 is a 6U VME CPU module that contains one
MPC7455 PowerPC CPUs with 128 MB to 1GB of main memory, 2MB of L3
cache and up to 128MB of flash memory. Also includes up to three 10/100BaseT
Ethernet ports, two RS232 serial ports, nine real-time clocks, two PMC slots, and
a PSCx Ultra 160/320 SCSI Interface PMC card.
2.3.2.
Model 920 Dual CPU Board
The Power Hawk Model 920 is a 6U VME CPU module that contains two
MPC7455 PowerPC CPUs with 128 MB to 1GB of main memory, 2MB of L3
cache and up to 128MB of flash memory. Also includes up to three 10/100BaseT
Ethernet ports, two RS232 serial ports, nine real-time clocks, two PMC slots, and
a PSCx Ultra 160/320 SCSI Interface PMC card.
2.4.
Power Hawk Series 900 Specifications
Processor Board
6.1
• Model 910 - Single G4 PowerPC MPC7455 G4
• Model 920 - Dual G4 PowerPC MPC7455 G4
PowerMAX OS Version 6.1 Release Notes
3
Power Hawk Series 900 System Overview
Memory
•
•
•
•
128, 256, 512 MB or 1 GB
256 KB L2 and 2 MB L3 Cache per CPU
128 MB User Flash
128 KB NVRAM and clock/calendar
I/O Interfaces
• VME64x (5-row), A32, D64
- PCI bus to VMEbus bridge
• 2 PMC slots, 64-bit 33/66 MHz
• PSCx Ultra 160/320 SCSI Interface PMC card
Processor On-Board I/O
• Up to 3 10/100BaseT Ethernet ports
• 2 RS232 serial ports
Optional I/O Controllers
•
•
•
•
•
•
•
•
16-line asynchronous serial mux with RS-232, RS-422 and RS-485
Single and quad-port 100BaseT and 1 Gbit Ethernet
IEEE-488 GPIB
MIL-STD-1553B with BC, RT and BM functions
Graphics PMC cards
A/D, D/A and digital I/O
Real-Time Clock & Interrupt Module (RCIM)
PMC Expander Modules
Peripherals
•
•
•
•
18, 36, 73 and 146 GB low profile 3.5" fixed and removable disk drives
CD-ROM drive
4mm DAT cartridge tape drive
ASCII terminal consoles
Chassis
•
•
•
•
•
12-slot and 21-slot 6U VME
Integral peripheral bays
Autoranging power supplies
Rack mount and table-top options
Optional ruggedized and ATR packaging
Environmental
4
Operating temperature: 10o to 50o C
Storage temperature: -40 o to 85o C
Relative humidity: 0% to 85%, non-condensing
Altitude: 10,000 feet
PowerMAX OS Version 6.1 Release Notes
6.1
•
•
•
•
Power Hawk Series 900 System Overview
Regulatory
6.1
• Safety: UL and CSA
• EMI/RFI: FCC Class A
PowerMAX OS Version 6.1 Release Notes
5
Documentation
3.0.
Documentation
3.1.
PowerMAX OS Software Documentation
Table 1-1 below lists the PowerMAX OS documentation available from Concurrent. Note
that standalone release notes are available for the various platforms. The corresponding
release notes will be provided with the applicable platform.
Table 1-1. PowerMAX OS Software Documentation
Manual Name
Pub. Number
PowerMAX OS Programming Guide
0890423-080
Character User Interface Programming
0890424-000
Device Driver Programming
0890425-070
STREAMS Modules and Drivers
0890426-020
User’s Guide
0890428-010
System Administration Volume 1
0890429-080
System Administration Volume 2
0890430-090
Compilation Systems Volume 1 (Tools)
0890459-050
Compilation Systems Volume 2 (Concepts)
0890460-050
PowerMAX OS Real-Time Guide
0890466-080
PowerMAX OS Documentation Overview
0890470-110
PowerMAX OS Guide to Real-Time Services
0890479-100
Power Hawk 900 Series Console Reference Manual
0830060-000
Power Hawk 900 Series PowerMAX OS Version 6.1 Release Notes
0891089-6.1
Power Hawk 900 Series Diskless Systems Administrator’s Manual
0891090-000
Copies of the Concurrent documentation can be ordered by contacting the Concurrent
Software Support Center. The toll-free number for calls within the continental United
States is 1-800-245-6453. For calls outside the continental United States, the number is
1-954-283-1822 or 1-305-931-2408.
Additionally, the manuals are available in pdf format on the Concurrent Computer
Corporation web site at www.ccur.com.
6
PowerMAX OS Version 6.1 Release Notes
6.1
Standalone product release notes are sometimes provided with software products. The
release notes you receive will be at the software revision level that matches the associated
software product level.
Documentation
3.2.
Vendor Documentation
Synergy commercial off-the-shelf (COTS) documentation applicable to the various
Synergy Single Board Computers (SBC) are listed below. You may contact your local
Synergy sales office to purchase Synergy manuals not provided with the system or visit the
Synergy web site at: http://www.synergymicro.com for more information. See the table
below for a list of Synergy manual names and applicable document numbers.
6.1
Manual Name
Synergy
Document
Number
Raptor DX VMEbus Dual G4, Dual PMC & StarFabric
PowerPC Single Board Computer User Guide
02-0426/UG-VYFD-01
Synergy Microsystem STAR User Guide
03-0072/UG-STAR-1.1
Synergy Microsystem ASTRix User Guide
03-0076/UG-ASTX-1.1
PowerMAX OS Version 6.1 Release Notes
7
Prerequisites
4.0.
Prerequisites
4.1.
Software
Release 6.1 or later.
4.2.
Hardware
CPU
Model 910 - single CPU SBC with minimum of 128 MB of memory.
or
Model 920 - dual CPU SBC with minimum of 128 MB of memory
4.2.1.
System Console Terminal
Video display terminal such as a Wyse 150, vt100, or comparable device
connected to Serial Port A.
4.2.2.
SCSI Interface Controller
Minimum requirements are:
•
•
•
4.2.3.
At least one Synergy PSCx Ultra 160/320 SCSI Interface card.
At least one supported SCSI disk drive for system files and swap space.
At least SCSI CD-ROM device for software installation.
Disk Drive
As stated above, each system must have an appropriate system drive. The smallest
disk drive supported for the installation disk (that contains the PowerMAX OS
executable) is 2GB.
The installation disk is supported on the PSCx SCSI/PMC card.
4.2.4.
CD-ROM Drive
A supported SCSI CD-ROM device connected to the PSCx SCSI/PMC card
controller is required for installation of system software.
4.2.5.
Ethernet Controllers
4.2.5.1.
PCI/PMC DEC Ethernet Controller
8
PowerMAX OS Version 6.1 Release Notes
6.1
Additional networking connections are possible with selected PMCbased DEC 21040, 21140A, 21142 and 21143 controller chips.
Contact your Concurrent representative to determine if a particular
card is supported.
Prerequisites
4.2.5.2.
Interphase 4207 Eagle
VME-based Ethernet may be provided via the Interphase 4207 Eagle
Ethernet Controller which supplies AUI connections. The minimum
revision level of the ethernet controller board (PN 2010221) is Rev E.
4.2.5.3.
Interphase 4221 Condor
VME-based Ethernet may be provided via the Interphase 4221
Condor Ethernet Controller which supplies AUI connections. The
minimum revision level of the ethernet controller board (PN 2010316)
is Rev A.
4.2.5.4.
Galileo GT64260 Ethernet Controller
Ethernet may be provided with 1 to 3 on-board GT64260 Ethernet
controller ports.
4.2.6.
High Speed Driver Enhanced (HSDE) Controller
VME-based HSD interface is provided by the High Speed Driver Enhanced
(HSDE) controller. The minimum revision level of the HSDE controller board
(PN 1573300) is Rev P.
4.2.7.
MIL-STD-1553 Version 5 (V5) Interface
VME-based MIL-STD-1553 Version 5 (V5) interface is provided by the 1553
controller. ABI-V5-1 is a Single Channel board, and ABI-V5-2 is a Dual Channel
board.
4.2.8.
IEEE488 GPIB Bus Interface Controller
VME-based IEEE GPIB Bus Interface is provided by the IEEE488 controller. The
minimum revision level of the IEEE488 controller board (PN 2010313) is Rev -.
4.2.9.
Asynchronous Communications Multiplexer (MVC)
The 16-Line Asynchronous Communications Multiplexer with industry standard
parallel port provides 16-ports of high-performance asynchronous serial
communications. Supports 4-port RS-232D and RS-422 line control modules.
Ports can run at speeds up through 38,400 baud.
4.2.10.
High-Performance Asynchronous Communications
Multiplexer (MVCS)
6.1
Supports individual line rates up to 1.5 Mbps, aggregate throughput of 48 Mbps,
and independent baud rates for each port. Can be configured with up to four of any
mix of four-port RS-232, RS-422 and RS-485 serial line modules. An optional 16port I/O transition panel with RJ-45 connectors and internal cable is available.
PowerMAX OS Version 6.1 Release Notes
9
Prerequisites
4.2.11.
Real-Time Clock and Interrupt Module (RCIM)
The RCIM is a multi-function PCI mezzanine card (PMC) that provides functions
ideally suited for time-critical applications. Up to sixteen CPU boards, in the
same or different chassis, can be linked via the RCIM’s synchronization cable.
The RCIM can be installed on any system with an available PMC slot. Refer to the
Real-Time Clock & Interrupt Module User’s Guide for more information.
4.2.12.
Single-Head PMC Graphics Card
High-performance, single-head PMC graphics card. Includes a Cirrus Logic
CL-GD7555 64-bit graphics controller, 4 MB display memory, two PS/2
keyboard/mouse ports, VGA BIOS, and 3-foot adapter cable. Supports true 24-bit
color at resolutions up to 1280 x 1024 using Xlib.
4.2.13.
Four-Head PMC Graphics Card
High-performance, four-head PMC graphics card. Includes four Cirrus Logic CLGD7555 64-bit graphics controllers, 4 MB display memory per head, eight PS/2
keyboard/mouse ports, VGA BIOS, and 3-foot adapter cables. Supports true 24bit color at resolutions up to 1280 x 1024 using Xlib.
4.2.14.
VMIVME-2510B 64-bit TTL I/O Board
The direction of each 8-bit port is individually programmable. Separate board
address decoding for control and data registers. Built-in-Test logic for fault
detection and isolation. Compatible with VMIC's family of Intelligent I/O
Controllers (VMIVME-90xx series).
4.2.15.
VMIVME-3122 High-Performance 16-bit A/D Converter Board
Supports 64 channels of differential or single-ended wide range analog inputs.
4.2.16.
VMIVME-4140 Intelligent 32/16-Channel Analog Output Board
Provides 32 high quality analog output channels with 12-bit resolution. Each
output has a dedicated D/A Converter (DAC) assigned to it.
4.2.17.
PBX3 3-slot PMC Expansion Board
10
PowerMAX OS Version 6.1 Release Notes
6.1
The PBX3 Expansion board provides 3 single-width PMC slots at 66 MHz, and
connects to the SBC through a stackable PMC card or adapter. The stacking
design provides a PMC expansion solution without giving up the use of an onboard PMC slot.
Prerequisites
6.1
(This page left intentionally blank)
PowerMAX OS Version 6.1 Release Notes
11
Available Software Packages
5.0.
Available Software Packages
This section contains brief descriptions of all the available software packages. Note that
your complement of available software may be less than that shown in Table 1-2 and
entirely depends on the optional software packages you purchased.
The availability of a software package for ALL currently supported platforms is shown in
Table 1-2 by a “y” (yes) or “n” (no) in the appropriate column.
Package dependencies are also specified in Table 1-2. Packages with dependencies must be
installed after the packages they depend on.
When installing the optional package(s) from the system installation menu, you must also
remember to select all required dependency package(s). The installation scripts will install
the packages in the proper order.
Some optional packages are relocatable, that is, objects may be installed in an alternative
directory other than root. When installing a relocatable package, the user will be prompted
for an alternate installation path.
12
PowerMAX OS Release 6.1 Release Notes
6.1
Note that all packages are dependent on base.
Available Software Packages
Table 1-2. Software Packages
Power
Hawk
Series 700
Power
Hawk
Series 900
Base System
y
y
base
Advanced Commands
y
y
cmds
Open Source
y
y
open-source
Domestic Encryption
Utilities
y
y
crypt
Fortran Libraries
y
y
hf77libs
International Encryption Utilities
y
y
crypt-int
Kernel Debugger
y
y
kdb
Network Support Utilities
y
y
nsu
OA&M
y
y
oam
Online Manual Pages
y
y
man
Printer Support
y
y
lp
Software Packaging
Tools
y
y
softint
Terminfo Utilities
y
y
terminf
Condor Ethernet Driver
y
y
cnd
CD-ROM Driver
y
y
cdfs
DR11W
Driver
y
n
dr11w
Eagle Ethernet Driver
y
y
egl
Generic Floppy Driver
y
y
gf
Package Description
Package
Name
Package
Dependencies
Standard:
lp,nsu
cmds
Drivers:
nsu
nsu
6.1
Drivers (Continue):
PowerMAX OS Release 6.1 Release Notes
13
Available Software Packages
Table 1-2. Software Packages (Cont.)
Power
Hawk
Series 700
Power
Hawk
Series 900
y
y
gs
High Performance
Serial Driver(HPS)
y
n
hps
HSDE
y
y
hsde
Peregrine FDDI Driver
y
n
pg
X.25 Driver
y
n
ix25
1553 V2 ABI Driver
y
n
1553drv
1553 V2 ABI Libraries
y
n
1533lib
1553 V5 ABI Driver
y
y
1553v5drv
1553 V5 ABI Libraries
y
y
1533v5lib
NCR SCSI Driver
y
y
ncr
DEC Ethernet Driver
y
y
dec
nsu
GT4260 Ethernet Driver
n
y
gte
nsu
Symbios Fast Ethernet
Driver
y
n
sym
nsu
Generic SCSI Device
Driver
14
Package
Name
Package
Dependencies
nsu
1553drv
1553v5drv
PowerMAX OS Release 6.1 Release Notes
6.1
Package Description
Available Software Packages
Table 1-2. Software Packages (Cont.)
Power
Hawk
Series 700
Power
Hawk
Series 900
Intelligent Bus Interface
Module (IBIM)
y
n
ibim
Asynchronous
Communications
Multiplexer
y
y
mvc
High-Performance
Asynchronous
Communications
Multiplexor
y
y
mvcs
National Instruments
GPIB
User-Level Device
Driver (GPIB)
y
y
ngpib
VMIC High Speed Data
(HSD) Driver
y
n
vhsd
V2510 User Level
Driver
y
y
v2510-510
V3122 User Level
Driver
y
y
v3122-510
V4140 User Level
Driver
y
y
v4140-510
Internet Utilities
y
y
inet
nsu
Commands Networking
Extension
y
y
netcmds
lp,inet
Remote Procedure
Calls Utilities
y
y
rpc
inet
Package Description
Package
Name
Package
Dependencies
Drivers (Continue):
TCP/IP Networking:
6.1
Network File System:
PowerMAX OS Release 6.1 Release Notes
15
Available Software Packages
Table 1-2. Software Packages (Cont.)
Power
Hawk
Series 700
Power
Hawk
Series 900
Network File
System Utilities
y
y
nfs
nsu, inet,
rpc, dfs
Distributed File
System Utilities
y
y
dfs
inet
y
y
audit
Package Description
Package
Name
Package
Dependencies
Security:
Auditing
Frequency-Based Scheduler (FBS):
Frequency-Based
Scheduler and
Performance Monitor
y
y
fbs
Frequency-Based
Scheduler
Manual Pages
y
y
fbsman
y
y
diskless
Concurrent Compilation
System
y
y
hc
Fortran 77 Compilation
System
y
y
hf77
C++ Compilation
System
y
y
c++
MAXAda Compilation
System
y
y
MAXAda
analyze
AXI for MAXAda
y
y
MAXaxi
MAXAda, x11
Closely-Coupled Systems: (Series 700 only)
Loosely-Coupled Systems: (Series 700/900)
Diskless
cmds, nfs, netcmds,
dec, sym and ncr
Software Development:
16
PowerMAX OS Release 6.1 Release Notes
6.1
Software Development (Continue):
Available Software Packages
Table 1-2. Software Packages (Cont.)
Package Description
Power
Hawk
Series 700
Power
Hawk
Series 900
y
y
See CrossDev
package
y
y
x11ipc
Cross-Development
Package Set
Package
Name
Package
Dependencies
See CrossDev
package
Window System:
X Window System
with OSF/Motif:
x11
x11ipc
x11progs
x11ipc, x11
x11dev
x11ipc, x11,
x11progs
x11pman
x11dman
Metro-X X11R6
Server
y
y
metroess
X11R6 Fonts
y
y
xfonts
Metro Link OpenGL
y
y
metrogl
NightStar
y
y
nstar
NightBench
y
y
nbench
x11
NightProbe
y
y
nprobe,
nprobeserv
x11, nstar
NightSim
y
y
nsim,
nsimserver
x11, nstar
NightTrace
y
y
ntrace,
ntracelog
x11, nstar
NightTune
y
y
ntune
x11
NightView
y
y
NightView
x11
NightStar Tools:
6.1
NightStar Tools (Continue):
PowerMAX OS Release 6.1 Release Notes
17
Available Software Packages
Table 1-2. Software Packages (Cont.)
Power
Hawk
Series 700
Power
Hawk
Series 900
y
y
Nviewp
Élan License Manager5
y
y
elan5lm
Élan License Manager4
y
y
elan4lm
elan5lm
Distributed XFS File
System
y
y
xfsd
inet, rpc
Virtual Partition
y
y
vp
Release 9.7
y
n
n/a
Release 9.8
y
y
n/a
Package Description
NightView Debugger
Support
Package
Name
Package
Dependencies
License Manager:
Miscellaneous:
18
PowerMAX OS Release 6.1 Release Notes
6.1
Power Hawk Diagnostic:
Available Software Packages
5.1.
Package Descriptions
The following pages contain a brief description of ALL of the packages available on Power
Hawk systems running the PowerMAX OS. Refer to Table 1-2 to determine if any given
package listed is applicable to your particular system.
5.1.1.
Description of Standard Packages
Base System (base)
The Base System package provides the base set of commands and system
utilities.
Printer Support (lp)
Although some printing capabilities are provided in the base package, more
advanced printing capabilities, and a wider range of printers, are included in the
Printer Support package.
Network Support Utilities (nsu)
The Network Support Utilities package provides the basis on which networking
capabilities are built.
Terminfo Utilities (terminf)
The Terminfo Utilities Package provides support for a wide variety of terminals
beyond those provided in the base package.
Advanced Commands (cmds)
The Advanced Commands package provides the remaining user and
administrative commands.
hf77libs (hf77libs)
This package provides runtime libraries for the Fortran 77 Compilation System.
OA&M (oam)
The Operations Administration and Maintenance package provides a characterbased, menu-oriented interface to a wide variety of advanced, server-oriented
administrative tasks.
Software Packaging Tools (softint)
This package provides tools to support the development process and includes a
variety of archive libraries as well as tools to create and modify packages.
Kernel Debugger (kdb)
The Kernel Debugger package provides a tool to assist in the porting and
debugging of kernel modules and drivers by allowing the developer to examine
and control a running kernel.
Domestic Encryption Utilities (crypt)
The Domestic Encryption Utilities package supports the encryption of files and
other data. This package is for distribution in the United States.
International Encryption Utilities (crypt-int)
6.1
Same as above but for international distribution.
PowerMAX OS Release 6.1 Release Notes
19
Available Software Packages
Online Manual Pages (man)
System manual pages (man pages) provided in an on-line format for viewing using
the man command.
5.1.2.
Description of Drivers Packages
Condor Ethernet Driver (cnd)
This package supports the Condor Ethernet 4211 VME board. Up to 6 Condor
Ethernet boards are supported.
CD-ROM File System (cdfs)
The CD-ROM File System package provides read-only access to file systems on
SCSI CD-ROM devices. ISO-9660 and High Sierra formats are supported.
Eagle Ethernet Driver (egl)
This package supports the Interphase Ethernet 4207 Eagle VME board. Up to 6
Eagle Ethernet boards are supported.
Generic Floppy Device Driver (gf)
The gr driver is a generic CD-ROM device driver. This driver supports the SCSI
Floppy device connected to the internal NCR/Symbios Controller..
Generic SCSI Device Driver (gs)
The gs driver package supports the following SCSI devices: SCSI-based CPUs,
scanners, printers, media changers, and communication devices.
High Performance Serial Driver (hps)
This package supports the High Performance Serial adapter, a VME board
providing 16 asynchronous serial ports running up to 38400 baud, and 1 optional
Centronics parallel printer port. Up to eight HPS adapters are supported.
Intelligent Bus Interface Module (ibim)
This driver package supports the IBIM module. The IBIM module is a data
acquisition module that acts as a host platform for a variety of analog and digital
interface modules. The IBIM can support up to four daughter boards.
NCR/Symbios SCSI Driver (ncr)
This driver package supports the internal SCSI controller chip and the controller
chips on separate PMC or PCI cards. This package also provides for automatic
configuration of the controller during system initialization (for example, no
Sadapters(4) file changes required). Up to seven disks and/or tape drives
may be connected to this SCSI bus.
DEC 21x4x Ethernet Driver (dec)
This driver package supports the Ethernet controller chip on the Power Hawk
series systems along with DEC 21x4x controller chips on separate PMC or PCI
cards. Supported chips include the 21040, 21140A, 21142 and 21143. This
package also provides for automatic configuration of the controller during system
initialization (e.g., no Sadapters(4) file changes required).
20
PowerMAX OS Release 6.1 Release Notes
6.1
100baseT (except on the 21040), 10baseT, BNC and AUI connections are
supported.
Available Software Packages
Symbios Fast Ethernet Driver (sym)
This driver package supports the Symbios 53C885 Ethernet controller chip on the
Power Hawk Series 700 SBCs.
Although the Symbios Ethernet hardware is not available on Power Hawk
This package must be installed on Power Hawk Series 900 SBCs before the
Diskless package may be successfully installed on Series 900 SBCs.
Galileo/Marvell Discovery Ethernet Ports (gte)
This driver package supports the Galileo/Marvell GT64260 Discovery system
controller chip in PowerHawk 900 series systems. The GT6420 supports up to
three independent 10/100Mb Fast Ethernet ports. Each port is fully compliant
with the IEEE 802.3 and 802.3u standards and integrates the MAC function and
a dual speed RMII interface.
High Speed Data Enhanced Channel Driver (hsde)
This package supports the HSDE Channel Interface. The HSDE provides highspeed, 32-bit parallel bidirectional link for transferring control, status and data
between the Power Hawk 700/900 System and an external device using the Encore
HSD Interface Model 9132 protocol.
1553 V2 ABI Driver (1553drv)
This package provides a user-level device driver for the Version 2 (V2) Advanced
Bus Interface (ABI) MIL-STD-1553 Adapter.
1553 V2 ABI Libraries (1553lib)
This package provides program interfaces that can be used within an application
program. This interface is the same as those provided by the manufacturer of the
1553 board, SBS Engineering, Inc.
1553 V5 ABI Driver (1553v5drv)
This package provides a user-level device driver for the Version 5 (V5) Advanced
Bus Interface (ABI) MIL-STD-1553 Adapter board manufactured by SBS
Engineering, Inc.
1553 V5 ABI Libraries (1553v5lib)
This package provides program interfaces which can be used within an application
program to control the ABI-VI MIL-STD-1553 Adapter. This set of interfaces is
an enhanced version of the set provided by the manufacturer of the board, SBS
Engineering, Inc.
Motorola MVME300 IEEE488 Interface Driver (mvme300)
This package supports Motorola's MVME300 IEEE 488 bus interface controller.
Up to 8 MVME300 controllers are supported. Available on Night Hawk systems
only.
Ikon DR11W Driver (dr11w)
This package supports the Ikon 10089 DR11W emulator board. Up to 16 DR11W
boards are supported (8 in the Primary I/O bus, 8 in the Secondary I/O bus).
6.1
X.25 Driver (ix25)
PowerMAX OS Release 6.1 Release Notes
21
Available Software Packages
This package supports the X.25 high speed synchronous communication module
that runs X.25/LAPB/HDLC.
Asynchronous Communications Multiplexer (mvc)
This package supports the Asynchronous Communications Multiplexer (MVC)
adapter board. The 16-Line Asynchronous Communications Multiplexer with
industry standard parallel port provides 16-ports of high-performance
asynchronous serial communications. Supports 4-port RS-232D and RS-422 line
control modules. Ports can run at speeds up through 38,400 baud.
High-Performance Asynchronous Communications Multiplexer(mvcs)
This package supports the High-Performance Asynchronous Communications
Multiplexer (MVCS). The MVCS adapter board supports individual line rates up
to 1.5 Mbps, aggregate throughput of 48 Mbps, and independent baud rates for
each port. Can be configured with up to four of any mix of four-port RS-232, RS422 and RS-485 serial line modules.
National Instruments GPIB User-Level Device Driver (ngpib)
The ngpib user-level device driver provides programmers with a configuration
utility and library functions to access and control a National Instruments General
Purpose Interface Bus (GPIB) 1014D board. This interface is a dual ported, high
performance interface to the IEEE-488 bus, and is capable of controlling two
independent GPIB bus configurations.
vhsd Driver (vhsd)
This driver provides support for the VMIC VMIVME-5620 Intelligent HSD
Emulator on the Power Hawk Series 700/900 systems.
V2510 Driver (v2510)
This driver provides support for the CE2510 64-bit TTL I/O board.
V3122 Driver (v3122)
This driver provides support for the CE3122 64/32/16-Channel, 16-Bit High
Performance Analog to Digital converter board.
V4140 Driver (v4140)
This driver provides support for the CE4140 intelligent 32/16-channel, 12-bit
analog output board with D/A per channel and auto-calibration
5.1.3.
Description of TCP/IP Networking Packages
Internet Utilities (inet)
The inet package includes the software needed to run the TCP/IP network and
tools such as ftp, telnet, and rcp.Administrative software for setting up the
network is also included.
Commands Networking Extension (netcmds)
22
PowerMAX OS Release 6.1 Release Notes
6.1
The Commands Networking Extension Package extends the functionality of
several basic commands by supporting the means to share printers across a
network, and use additional transport mechanisms for the sending and receiving
of electronic mail.
Available Software Packages
Remote Procedure Calls Utilities (rpc)
The Remote Procedure Calls Utilities package supports the remote execution
facility.
Network Time Protocol (xntp)
xntpd is a complete implementation of the Network Time Protocol (NTP)
version 2 standard as defined by RFC 1119, and also retains compatibility with
version 1 servers as defined by RFC 1059. xntpd does all computations in fixed
point arithmetic and is entirely free of floating point code. The computations
done in the protocol and clock adjustment code are carried out with high precision
and with attention to the details which might introduce systematic bias into the
integrations, to try to maintain an accuracy suitable for synchronizing with even
the most precise external time source. Xntp is part of the open-source package.
5.1.4.
Description of Network File System Packages
Network File System Utilities (nfs)
The Network File System Utilities package supports the means to transparently
share resources across a network with other computers running the Network File
System.
Distributed File System Utilities (dfs)
The dfs utilities package provides a simple user interface for performing
networked operations such as advertising local resources and accessing remote
resources.
5.1.5.
Description of Security Package
Auditing (audit)
The Auditing package provides auditing facilities allowing a system administrator
or security auditor to record and report all security-related events that occur on the
system.
5.1.6.
Description of Diskless Package
Diskless (diskless)
This package provides support to configure and control multiple Power Hawk
single-board computers (SBC) in a common VME chassis (referred to as a
closely-coupled system). Diskless SBCs in this chassis can be booted over the
VMEbus by a single server SBC in the rack, where the server SBC is attached to
system disk(s) that contain the system object images for all diskless SBCs in the
chassis.
A closely-coupled system consists of multiple Power Hawk 700 Series SBCs
residing in the same chassis.
6.1
On Power Hawk 720/740 closely-coupled systems, an additional PCI-to-PCI (P0)
bus also connects all SBCs in the chassis, and this P0Bus provides a high-speed
connection for inter-SBC communications, including a point-to-point network
interface.
PowerMAX OS Release 6.1 Release Notes
23
Available Software Packages
This package also supports multiple SBCs in one or more remote VME chassis
connected via Ethernet (referred to as a loosely-coupled system). Diskless SBCs
in these remote VME chassis can boot themselves via TFTP over the Ethernet,
with a single Power Hawk server SBC providing the downloadable object images
required for each diskless SBC.
Power Hawk Series 900 systems are supported in loosely-coupled configurations,
but they are NOT supported in closely-coupled configurations. The Synergybased loosely-coupled configurations may contain a mix of Series 700 and 900
systems.
5.1.7.
Description of FBS Packages
Frequency-Based Scheduler and Performance Monitor (fbs)
This package provides kernel support for the Frequency-Based Scheduler and
Performance Monitor and Real-Time Command Processor.
Frequency Based Scheduler Manual Pages (fbsman)
This package provides the man pages associated with the Frequency-Based
Scheduler and Performance Monitor.
5.1.8.
Description of Software Development Packages
Concurrent C Compilation System (hc)
This package provides the hc C compiler. This compiler offers ANSI C
compliance and support for pre-ANSI C, together with other extensions.
Fortran 77 Compilation System (hf77)
This package provides the hf77 Fortran compiler and runtime libraries. This
compilation system offers Fortran 77 and MIL-STD 1753 compliance, together
with many popular extensions and a cross-reference tool with interface checking.
C++ Compilation System (c++)
The 6.0 version of c++ utilizes Edison Design Group's C++ front end and
Concurrent's proprietary Common Code Generator (CCG) technology to produce
highly optimized object code tailored to the architecture of the target system. The
6.0 release combines into a single compiler what were previously separate C and
C++ compilers. This release does not require un-installing previous versions of
the C and C++ compilers.
MAXAda Compilation System (MAXAda)
This package consists of the MAXAdaTM tool set for the development of Ada
programs on Concurrent computers under the PowerMAX OS environments.
MAXAda processes the Ada language as specified by the Reference Manual for
the Ada Programming Language, ANSI/ISO/IEC-8652:1995.
Ada to X Interface for MAXAda (MAXaxi)
This package consists of the Ada X Window system interface for MAXAda.
24
PowerMAX OS Release 6.1 Release Notes
6.1
Cross-Development Environment
Available Software Packages
Consists of packages that contain all the header files and libraries required for
cross development. The Cross-Development Environment allows users to
develop applications from any supported Concurrent real-time computer systems
targeting a specific PowerMAX OS release and architecture.
5.1.9.
Description of X Window System Packages
X Window System Version 11, Release 6 (x11)
Includes OSF/Motif commands, libraries and header files. X Window System
sub-package names and short descriptions follow:
x11ipc
x11
x11progs
x11dev
x11pman
x11dman
the libraries for ICE and ktalk only
all other libraries including runtime support files they reference
X client programs xdm, mwm, application-default files, etc.
X program development tools, header files, imake, static libs, etc.
man pages for all the application level programs.
man pages for the libraries and program development tools
Metro-X Enhanced X11R6 server set and OpenGL (metroess, xfonts & metrogl)
Jointly developed with Metro Link, Inc., these packages provide native graphics
capabilities on systems equipped with appropriate display and input devices.
Specific packages include the following:
5.1.10.
metroess
Metro Links Enhanced X11R6 Server Set includes a powerful,
configurable, host based X11R6 server derived from the
X11R6/XFree86 sample servers.
xfonts
Local X11 system fonts, which are required for server installations
that do not use a font server.
metrogl
OpenGL version 1.2 for Metro-X. Includes OpenGL client libraries
(libGL, libGLX, libGLU), header files, sample clients, and GLX
server extension for rendering OpenGL commands with the
Metro-X X11R6 server.
Description of NightStar Tools Packages
NightStar (nstar)
This package provides a NightStar daemon that initiates individual NightStar tool
servers and daemons on behalf of a (possibly remote) NightStar client tool.
NightBench (nbench)
NightBench is a graphical user interface to MAXAda, a tool set for the
development of Ada programs running under the OS.
NightProbe (nprobe, nprobeserv))
The NightProbe Data Monitoring package provides a real-time graphical tool for
monitoring, recording, and altering program data within one or more executing
programs without significantly affecting the execution of those programs.
6.1
NightSim (nsim, nsimserver)
PowerMAX OS Release 6.1 Release Notes
25
Available Software Packages
This package provides a graphical tool for building and monitoring real-time
applications that require scheduled process execution patterns. NightSim provides
a GUI to the Frequency-Based Scheduler (FBS) and the Performance Monitor
facilities.
NightTrace (ntrace, ntracelog))
This package provides a graphical tool for analyzing dynamic behavior of multi
process and/or multiprocessor applications.
NightTune (ntune)
This package provides an interactive, graphical tuning and monitoring
environment specifically designed for Concurrent’s real-time computer systems.
NightView (NightView)
This package provides a general-purpose, source-level debugger for C, C++,
Fortran and Ada with support for multiple processes.
NightView (Nviewp)
This package provides a graphical, non intrusive, source-level monitoring and
debugging tool specifically designed for real-time applications.
5.1.11.
Description of License Manager Packages
Élan License Manager (elan5lm)
This package contains the license manager daemon and reporting commands
required by license-managed applications, including all NightStar tools. This
version is Y2K compliant.
Élan License Manager (elan4lm)
This package is an optional ‘old’ (version 4) license manager daemon that can run
in parallel with the new (elan5lm) license manager daemon on the same server
machine in order to support serving licenses to both ‘old’ and ‘new’ tools at the
same time.
5.1.12.
Description of Miscellaneous Packages:
Distributed XFS File Systems (xfsd)
The xfsd distributed file system package provides shared access to a single file
system from multiple host systems on a multi-initiator disk bus. Unlimited read
access to files in the shared file system is supported. Only limited write access to
files in the shared file system is supported. A network connection between
systems that access a shared file system is required for communication and
synchronization of file system structures.
Virtual Partition (vp)
26
PowerMAX OS Release 6.1 Release Notes
6.1
This package provides a pseudo device driver that performs disk striping (RAID
level 0) or mirrored partitioning (RAID level 1).
Available Software Packages
5.1.13.
Power Hawk Series 700/900 Diagnostics
Software diagnostics on CD-ROM for a single Power Hawk Series 700/900
system. Includes Power I/O and Power AT diagnostics.
6.1
(This page left intentionally blank)
PowerMAX OS Release 6.1 Release Notes
27
System Installation
6.0.
System Installation
The PowerMAX OS operating system is installed as software packages using the Software Packaging
Tools. Two installation modes, Custom and Semi-Automatic, are now available. Refer to Installation
Modes on page 31 for additional details.
NOTE
A system may be booted using an installation CD-ROM disk.
However, only tape is supported as the backup (restore) media.
This means you can boot from a CD-ROM disc and once booted
from disc, you can restore from a backup tape. However. just
before the base package is installed, some tools used for the
RESTORE function are removed to free up memory. If you need
to use the RESTORE function after executing through many
levels of the INSTALL menus you must reboot first from disc.
6.1.
Large SCSI Disk Support
Changes were introduced in a previous PowerMAX OS release to correct the support for
SCSI disks larger than 8 Gigabytes. These corrections changed the location of the
Geometry Block on these large disks. As a result, disks larger than 8 GB formatted using
the format(1m) or format(8) standalone utility prior to Release 4.2 cannot be used
to install release Release 6.1 without being reformatted.
If you have a disk larger than 8 GB that was formatted using the Release 4.1 or earlier
version of format(), you must do the following:
1. Backup any data needed on any of the partitions before installation.
2. During the installation, exercise the option to format the disk and rewrite
the partition information (Geometry Block).
3. Finish the Release 6.1 installation.
4. Restore any data backed up in step 1 above.
6.2.
System Disk Configuration
The system disk configuration, shown in Table 1-3, is suggested for installing the system
using an ufs root file system type.
6.1
Table 1-3. ufs root file system
PowerMAX OS Version 6.1 Release Notes
partition
file
Disk < 9GB Default
Partition size
Disk > 9GB Default
Partition size
0
root
200 MB
500 MB
1
swap
500 MB
2000 MB
2
usr
500 MB
2000 MB
29
System Installation
Table 1-3. (Cont.)ufs root file system
partition
file
Disk < 9GB Default
Partition size
Disk > 9GB Default
Partition size
3
var
500 MB
2000 MB
4
----
<remainder>
<remainder>
boot
1024 KB
1024 KB
1
6
The system disk configuration shown in Table 1-4 is suggested for installing the system
using an xfs root file system type.
Table 1-4.
xfs root file system
partition
file
Disk < 9 GB Default
Partition Size
Disk > 9GB Default
Partition size
0
/stand
100 MB
100 MB
1
swap
500 MB
2000 MB
2
usr
500 MB
2000 MB
3
var
500 MB
2000 MB
4
----
<remainder>
<remainder>
51
root
200 MB
500 MB
62
boot
1024 KB
1024 KB
1. The xfs root must be on partition 5.
2. Partition is applicable to Power Hawk Series 700/900 boot disks only.
Partition sizes may be increased and new partitions may be added, but the above
assignments of file systems to partitions and minimum partition sizes must be maintained.
During system installation, you will be given the option of running the format(1M)
command. You should choose this option to format the system disk and to select partition
sizes.
The format(1M) command “partition default” automatically selects the above partition
sizes for a ufs root file system.
Note that the format(1M) “partition default” option should not be used if an xfs root
file system is selected as the default partitions automatically selected are not appropriate
for the xfs file system.
The root and usr file systems should only contain system files and are not expected to
grow much after system installation. The var file system contains system crashfiles, log(s)
and temporary files. It is strongly recommended that user files be restricted from these file
systems.
30
PowerMAX OS Version 6.1 Release Notes
6.1
Swap space needs should be carefully calculated. Adding too little swap space will result
in unnecessary out-of-memory conditions for your applications. Adding too much swap
space will result in too much of your system’s memory being locked down for swap space
System Installation
management. The total amount of swap space should be at least 1.5 times the size of
physical memory. An initial swap partition is provided on the system disk. If this partition
is insufficient, it is recommended that additional swap partitions be added, preferably on
other disks.
The total amount of swap space for MAXada development systems should be at least 3
times the size of physical memory. This larger swap space compared to runtime systems
is required because of the virtual memory expansion of the compiler.
Every Gigabyte of swap space results in 4 Megabytes of physical memory being used for
swap space management. This rule demonstrates that it would be impractical to create a 9
Gigabyte swap device on a 64 Megabyte system, as this would result in more than half (36
Megabytes) of physical memory being utilized by the kernel for swap space management.
Note that partition four is left unused. This partition may be:
1. Redistributed to make the other system partitions bigger.
2. Used for additional swap space.
3. Used for user files (for example a home file system).
Use the format(1M) “?” command for help with format commands. Refer to the
format(1M) manual page at the end of these release notes. Note that non-system disks
will need to be initialized once the system is re-booted for new disk. This includes running
format(1M) to format and partition the disk and newfs(1M) to initialize the file
systems. Additional steps include creating a mount point directory, adding the appropriate
information to /etc/vfstab and adding new entries to the Device Database
(DDB). See the System Administration Manual for information on disk formatting and
partitioning.
6.3.
Installation Modes
Early in the system installation, the operator will be asked whether to perform a custom or
semi-automatic system installation. A description of each mode is provided in the
following paragraphs.
6.3.1.
Custom Mode of Installation
In custom mode, the operator is prompted during the installation in order to
specify the values for configurable items. Custom mode is recommended if the
predetermined values assigned in the semi-automatic mode are not appropriate for
your site, and/or you want the option of installing only certain packages during
system installation.
6.3.2.
Semi-Automatic Mode of Installation
In semi-automatic mode, the operator still specifies the basic configuration of the
system however, the remainder of the installation is done with a predetermined set
of responses. Note that in semi-automatic mode, all the additional products on the
release distribution media are automatically installed. The operator does not have
the option of installing only specific packages.
6.1
Refer to Table 1-5 for a description of the configurable items and the values that
they will be assigned on a semi-automatic installation.
PowerMAX OS Version 6.1 Release Notes
31
System Installation
Semi-automatic mode can be used if the configurable values are set appropriately
for the given site and all packages are being installed. Otherwise, custom mode
should be used.
NOTE
Refer to Table 1-2 to determine if a given package listed in
Table 1-5 is supported for your particular system.
Table 1-5. Assigned Installation Values, Semi-Automatic Mode
base
Automatic Installation Value
Host nodename
Root password
License key - number of users
License key - number of processors
Obtained during initial installation
“” (null password)
no.
no.
Install man pages?
Start license manager during system boot?
yes
yes
oam
sysadm password
“” (null password)
cnd
Number of cnd adapters on VME bus
1
Number of dr11w adapters on VME bus
1
egl
Number of egl adapters on VME bus
1
hps
Number of hps adapters on VME bus
Configure real-time driver?
1
no
hsde
Number of hsde adapters on VME bus
1
Number of pg adapters on VME bus
1
Number of 1553V5-ABI adapters
1
mvc
Number of mvc adapters on VME bus
Configure real-time driver?
1
no
inet
Configure TCP listener?
yes
MAXAda
Install directory for MAXAda
standard location relative to root directory
MAXaxi
Install directory for Ada X Interface (AXI)
for MAXAda
standard location relative to root directory
x11ipc
x11
x11progs
x11dev
x11pman
x11dman
Install header files and static libraries?
Install man pages?
Start xdm during system boot?
yes
yes
yes
Install directory for NightStar
standard location relative to root directory
nbench
Install directory for NightBench
standard location relative to root directory
nprobe
Install directory for NightProbe
standard location relative to root directory
elan5lm
dr11w
pg
1553v5drv
nstar
32
Configurable Item
PowerMAX OS Version 6.1 Release Notes
6.1
Package
Name
System Installation
Table 1-5. Assigned Installation Values, Semi-Automatic Mode (Cont.)
Package
Name
Automatic Installation Value
nsim
Install directory for NightSim
standard location relative to root directory
ntrace
Install directory for NightTrace
standard location relative to root directory
ntune
Install directory for NightTune
standard location relative to root directory
NightView
Install directory for NightView
standard location relative to root directory
Install directory for Nviewp
standard location relative to root directory
Nviewp
6.1
Configurable Item
PowerMAX OS Version 6.1 Release Notes
33
System Installation
6.4.
Installation Procedure
It is recommended that the user read the Power Hawk Series 900 Console Reference
Manual (Pubs No. 0830060) beforehand in order to be familiar with the Series 900 console
commands used throughout these release notes.
The Power Hawk Model 920 platform uses two monitor ROM programs called STAR (Self
Test And Run) and ASTRix (Advanced Self Test Reboot Initialize eXecute) that provide
basic board initialization, configuration and diagnostic capability. The STAR utility will
execute whenever the system is powered up or is reset. STAR and ASTRix are used to
configure the basic system and to load the initial boot image from distribution media. The
individual commands used in STAR and ASTRix are documented in the Synergy
Microsystem STAR User Guide (03-0072/UG-STAR-1.0), and the Synergy Microsystem
ASTRix User Guide (03-0076/UG-ASTX-1) respectively.
Note
The PCI Memory Space layout, which is setup with the STAR
config command, MUST be set to the proper values before a PowerMAX OS kernel can be successfully booted. The STAR config
command should be used to verify that the PCI Memory parameters are set to the following default values:
PCI 0 Start Addr (ex. 0x80000000): 0x80000000 =
PCI 0 Size (ex. 0x40000000 (1GB)): 0x30000000 =
PCI 1 Start Addr (ex. 0x80000000): 0xC0000000 =
PCI 1 Size (ex. 0x10000000 (256MB)): 0x30000000 =
Please consult the Power Hawk Series 900 Console Reference
Manual (Pubs No. 0830060) section "PCI Memory Space Configuration", for more information on this subject.
STAR/ASTRix commands are used to boot the installation image from the distribution
media. The time to complete the entire installation will vary depending on the packages
selected to be installed and the type of drive being used.
It may be necessary to use operator intervention if the PowerMAX OS Console Software
has been previously loaded into flash memory. If the “STAR0>” prompt does not appear
on the console terminal at power up, it will be necessary to power cycle or reset the board
and use <CTRL>c to obtain a STAR prompt. and then enter the following commands to
load the PowerMAX OS installation image.
STAR0>config
...
Change the
Enable "post" script:
Y =
34
PowerMAX OS Version 6.1 Release Notes
6.1
to N
STAR0> astrix
Y: CPU 1 standing by.
System Installation
Synergy Microsystems ASTRix Rev: 1.01.04PRE1 Oct 17 1929
08:58:19
astrix> scsi-drivers
Using scsi/scsi_mod.o
Using scsi/sd_mod.o
Using cdrom/cdrom.o
Using scsi/sr_mod.o
Using scsi/sym53c8xx.o
Loaded SCSI modules.
Now mount your drive by typing
(for example) mount /dev/sda1 /mnt .
astrix> boot /dev/cdrom1
Using /lib/modules/2.4.19-Asmp/misc/reboot.o
Synergy Microsystems ASTRix Rev: 1.01.04PRE1 Oct 17 1929
08:58:19
PowerMAX OS Kernel Decompressor
...
(OS is now booting)
General Notes
The cdfs (CD-ROM File System) package (including the
generic CD-ROM driver (gr)) is always automatically installed
following the base package installation.
A system may be booted using an installation CD-ROM disc.
However, only DAT tape is supported as the backup (restore)
media. This means a user can boot from a CD-ROM disc and
once booted from the CD-ROM disc, the user is then able to
restore from a backup tape.
As a side note, just before the base package is installed, some
tools used for the RESTORE function are removed to free up
memory. If you the user wants to use the RESTORE function after
executing many levels of the INSTALL menus you must first
reboot from CD-ROM disc or DAT tape, whichever is applicable.
As the system is brought up, the initial menu will prompt you to choose either the system
software installation program or the file system restore program. At this prompt, choose
submenu 1 INSTALL. (The file system restore program is documented in Chapter 10 of the
System Administration, Volume 2 (Pubs No. 0890430).) Prompts that require user input are
preceded by =>. Most prompts have defaults in parentheses that may be selected by
pressing the “Enter” key. At any prompt, “?” can be entered for help or q to quit
installation. If the user selects to quit installation or if a fatal error occurs, installation will
be suspended by executing a subshell. Three choices are available when exiting the
subshell:
6.1
CONTINUE - repeat last operation
RESTART - start over from the beginning
ABORT
- return to console debugger
PowerMAX OS Version 6.1 Release Notes
35
System Installation
The installation is self-guiding, but the following configuration information is required
from the user:
Installation Mode (choose custom or semi-automatic (see Installation Modes on page 31)
Node name
Timezone
Time/Date
Desired file system types
System disk location (slot and unit number)
Installation media device location
System disk configuration (see System Disk Configuration on page 29)
Note
At this point in the installation sequence, the system disk is initialized and you will be prompted on-screen to insert the PowerWorks Standard Products media.
License key for number of users
License key for number of processors
Following are applicable to custom mode installation only:
Which additional software packages should be installed (see Installing Additional
Packages on page 37)
Configuration information requested by the various packages
If appropriate, any kernel modules from the optional packages you want to deconfigure
(that is, not link with the kernel)
Note
The installation media in use is accessed at various times and
must be kept in the drive during the installation until you are
instructed to install different media.
After all the packages have been installed, you will be given an opportunity, if in the custom
installation mode, to deconfigure kernel modules from optional packages. When
deconfigured, those drivers will not be linked with the kernel. Note that you must consider
package dependencies when deconfiguring drivers. See Table 1-2 for software package
dependency relationships.
At the end of the installation procedure, a kernel for the newly installed system will be built.
36
PowerMAX OS Version 6.1 Release Notes
6.1
When installation is completed the STAR/ASTRix Automatic initialization procedure
contained in the PowerMAX OS Console Reference Manual (0830060-000) should be
completed. This will place the Console Software in flash memory to improve boot time of
the system.
System Installation
6.5.
Installing Additional Packages
In the semi-automatic installation mode all software packages provided on the
PowerWorks Standard Products CD are automatically installed. However, in the custom
installation mode you may decide to delay installation of various packages until after a
basic system configuration is installed.
To install additional packages on an installed system, use the pkgadd(1M) command.
Refer to the pkgadd(1M) manual page at the end of these release notes for more details.
In addition, see the Chapter on “Installing Add-On Software” in the System Administration
Manual Volume 1 (Pubs No. 0890429) for more information.
The following procedure will enable you to install additional packages as you choose.
Note
The system must be configured with the gr (generic CD-ROM
device driver) gr(7) and cdfs (CD-ROM filesystem) modules
and, in addition, an entry must exist for the CD-ROM device
nodes in /etc/conf/node.d/gr.
In the example shown below, the CD-ROM device is /dev/cd/0 and the CDROM disc is mounted under /mnt/cdrom.
1. Insert the PowerWorks Optional Products disc into the CD-ROM
drive and mount the file system as shown below.
mkdir /mnt/cdrom
mount -F cdfs -r /dev/cd/0 /mnt/cdrom
2. Get a list of the packages on the CD-ROM disc and verify that the
package you wish to install is on the CD-ROM disc.
pkginfo -d /mnt/cdrom/pkgs.dstream
3. Multiple packages may be installed at once by specifying more than
one package name, as in:
pkgadd -d /mnt/cdrom/pkgs.dstream nsu lp cmds
4. The package(s) to be installed may be selected from a menu of all the
packages available on the CD-ROM disc by not specifying any package name, as in:
pkgadd -d /mnt/cdrom/pkgs.dstream
6.6.
Installing Service Releases
To install packages on the PowerWorks Service Release CD use the pkgadd(1M)
command.
6.1
Refer to the pkgadd(1M) manual page at the end of these release notes for more details.
In addition, see the Chapter on “Installing Add-On Software” in the System Administration
Manual Volume 1 (Pubs No. 0890429) for more information.
PowerMAX OS Version 6.1 Release Notes
37
System Installation
The following procedure(s) enable you to install the packages on the PowerWorks Service
Release CD as you choose.
Note
The system must be configured with the gr (generic CD-ROM
device driver) gr(7) and cdfs (CD-ROM filesystem) modules
and, in addition, an entry must exist for the CD-ROM device
nodes in /etc/conf/node.d/gr.
In the example shown below, the CD-ROM device is /dev/cd/0 and the CDROM disc is mounted under /mnt/cdrom.
1. Insert the PowerWorks Service Release disc into the CD-ROM drive
and mount the file system as shown below.
mkdir /mnt/cdrom
mount -F cdfs -r /dev/cd/0 /mnt/cdrom
2. Get a list of the packages on the CD-ROM disc and verify that the
package you wish to install is on the CD-ROM disc.
pkginfo -d /mnt/cdrom/pkgs.dstream
3. Multiple packages may be installed at once by specifying more than
one package name, as in:
pkgadd -d /mnt/cdrom/pkgs.dstream nsu base-001 \
base-002 base-003
4. The package(s) to be installed may be selected from a menu of all the
packages available on the CD-ROM disc by not specifying any
package name, as in:
pkgadd -d /mnt/cdrom/pkgs.dstream
5. View the service release README File to get information about the
contents of the service release.
38
PowerMAX OS Version 6.1 Release Notes
6.1
pkginfo -Rd /mnt/cdrom/pkgs.dstream
Rebooting The System
7.0.
Rebooting The System
During the installation, a new kernel is built. When the system installation completes, the system first
halts and then returns to the STAR0> prompt. Enter the following commands to then load and execute
the console processor:
STAR0>config
...
...
Change the
Enable "post" script:
Y =
to Y
STAR0> astrix(loads ASTRix)
* scsi-drivers(loads the SCSI device support)
* boot /dev/sda(loads console processor into memory,
resets the board, and then starts its
execution)
In the above example, the ‘a’ in /dev/sda assumes that the system disk is disk ‘a’ in the ASTRix
environment.
If a kernel was successfully built during the system installation, execute the following commands from
the console (refer to the Series 900 console manual, pubs # 0830060, for information on console
command syntax):
#0>y.
#0>pboot 0
#0>fd dsk(c,u,p,b,l)
#0>fb
If the kernel build fails during the installation, /stand/unix.generic is copied to /stand/unix.
You may use this generic kernel to boot to single-user mode, correct the problem that caused the kernel
build failure, build a new kernel and reboot. Because the generic kernel was not built using your site’s
specific configuration, you should not come up in multi-user mode with the generic kernel. To boot the
generic kernel, execute the following commands from the console:
#0>pboot 3
00000000
#>fd dsk(c,u,p,b,l)
.............................
#0>fb
Set Run Mode
CPU 0 CPU 1
dsk(0,0,0,0,0)/.
Initialize VME
dsk(0,0,0,0,0)/boot
PowerMAX OS Boot Loader
Boot:
/stand/unix.generic
Remain in single-user mode by entering the root password when prompted. Then execute the following
commands:
6.1
# fsck -y /dev/rusr
# mount /dev/usr
# /etc/conf/bin/idbuild -B
PowerMAX OS Version 6.1 Release Notes
<---| for ufs /usr filesystem only
39
Rebooting The System
Note
If the system appears to be hung during the reboot to the new
kernel, power cycle the machine and repeat the steps listed above.
After you have successfully built a kernel, you can bring the system down by executing the command
init 0. The system shutdown/reboot sequence takes care of moving the newly built unix to
/stand/unix. Then follow the above procedure on how to boot a newly installed system.
For information on how to configure and build a kernel, see the Chapter on “Configuring and Building
a Kernel” in the System Administration, Volume 2 (Pubs No. 0890430).
40
PowerMAX OS Version 6.1 Release Notes
6.1
See the following paragraph pertaining to xfs root file system and the use of “non-standard” kernel
names in /stand.
General Notes
8.0.
General Notes
8.1.
XFS
If a new kernel is installed in /stand with a nonstandard name then a hard link needs to
be created. An explanation for this requirement is provided in the following paragraph.
The boot program boots kernels from the file system on partition 0, which it assumes to be
of type ufs. If this file system is also the root, then by convention the kernels are held in
directory /stand within the root. However, if the root is xfs (partition 5), the ufs file
system containing the kernels (and standalones) is mounted on /stand, which in this case
is a directory in the xfs root. Using hard links to the directory /stand/stand ensures
that the conventional boot path can be used.
NOTE
It is important to point out to the system administrator that if
kernels with names other than “unix” are copied to /stand on a
system with an xfs root, they should be given hard links of the
same name in /stand/stand. This will ensure that they can be
booted using the conventional path /stand/unix.xxx. If this
link is missing, they can only be booted using /unix.xxx.
8.2.
PCI-to-PCI Bridge Configuration
6.1
The auto configuration aspects of PCI buses define which local bus addresses are
propagated on each PCI bridge. The actual PCI I/O and Memory usage is dictated
by the PCI cards installed and the resources requested. This function occurs
automatically at system boot time.
PowerMAX OS Version 6.1 Release Notes
41
General Information on Release 6.1
9.0.
General Information on Release 6.1
This section provides a general overview of PowerMAX OSTM Release 6.1. As stated earlier, this
release is designed primarily to provide support for the newly introduced Concurrent Computer
Corporation Power HawkTM Series 900 Systems.
9.1.
Architecture Unification
As the number of hardware platforms on which PowerMAX OS operates
continues to grow, enhancements made in previous releases to simplify
configuration and installation on selected systems continues to be updated to
accommodate new systems.
Listed below are the Board Support Packages (bspxxxx) for the different system
types.
Architecture
Family
Hardware
System Types
Power Hawk Series 700
Power Hawk Series 900
Specific Board
Support Packages
General Board
Support Package
Power Hawk Series 710
(uniprocessor)
bspvgm5
Power Hawk Series 720
(dual processor)
bspvgm5
Power Hawk Series 740
(quad processor)
bspvss4
Power Hawk Model 910
(single processor)
bspvyfd
bspall
Power Hawk Model 920
(dual processor)
bspvyfd
bspall
bspall
The system installation scripts will automatically turn on the appropriate module
(bspxxxx) based upon the type of hardware on which the system is being
installed. It is also possible to manually select one and only one bspxxxx module
and relink a kernel which is appropriate for that particular hardware system type
(within its architecture family).
While the bspall module will function for any of the supported system types
(within its architecture family), the specific board support package module
applicable to your system is preferred as it has been optimized for that system.
9.2.
Enhancements in this Release
9.2.1.
Dynamic Interrupt CPU Assignments
On PowerHawk Series 700 and 900 systems you may now change CPU interrupt
42
PowerMAX OS Version 6.1 Release Notes
6.1
source assignments dynamically, without the need for statically reconfiguring,
rebuilding and rebooting the kernel.
General Information on Release 6.1
Dynamically querying and changing CPU interrupt source assignments may be
done with the new intconfig(1M) utility, or through the use of the new
SET_PIN_CPU and GET_PIN_CPU commands in the syscx(2) system service.
For more information on this feature, you may refer to:
9.2.2.
-
the intconfig(1M) system manual page,
-
and the section “Assigning Interrupts to CPUs” in Chapter 2 of
the PowerMAX OS Real-Time Guide.
the SET_PIN_CPU and GET_PIN_CPU commands in the
syscx(2) system manual page,
Loosely-Coupled Systems
Previous PowerMAX OS releases have supported closely-coupled (CCS) and
loosely-coupled systems (LCS) on Power Hawk Series 600/700 systems.
This 6.1 release adds loosely-coupled (LCS) support for Power Hawk Model 920
systems.
Power Hawk 900 Series may be used as the server or client(s) in a LCS
configuration that contains any mix of Power Hawk Series 700 and Series 900
systems.
The configuration requirements for a Power Hawk Series 900 LCS configuration
are:
-
All Series 700 SBCs should have a Symbios (sym) Ethernet interface
connection to the server SBC, and all Series 900 SBCs should have a
GT64260 (gte) Ethernet interface connection to the server SBC.
-
The server SBC may be a Series 700 system or a Series 900 system
with ethernet access to all of the configured client Series 700/900
SBCs.
-
One SCSI disk drive must be connected to the server SBC for storage
of PowerMAX OS software.
-
One SCSI CD-ROM driver must be connected to the server SBC for
PowerMAX OS software installs and updates.
Note
Release 6.1 does NOT provide any support for closely-coupled
(CCS) configurations for Power Hawk Series 900 systems.
9.3.
Known Problems
6.1
The following is a list of known problems which Concurrent is working to resolve. If a
workaround is available, then it is listed in the Known Workaround column below. These
problems are deemed to have low impact upon the use of the operating system.
PowerMAX OS Version 6.1 Release Notes
43
General Information on Release 6.1
Table 1-6. PowerMAX OS 6.1 Known problems
44
Summary
Known Workaround
586
The system may panic with the follow- Attempt fresh netboots until successful.
ing panic string during a netboot:
PANIC: Reservation Instruction Fault
trap at pc = 0x00193578 in kernel mode
DSISR=0x04000000
This problem is currently under investigation.
597
Intermittent errors when booting from
CD.
The problem is currently under investigation. Continued attempts succeed.
600
Package install errors received when
attempting base installation.
The problem is under investigation. If
the SCSI is in PMC Slot 1 and something (anything) else is in PMC Slot 2,
then the problem does not occur.
873
Random NFS Write Errors might occur
on heavy NFS activity.
Mounting the NFS file system with the
rsize=N and wsize=N options may
reduce the random write errors dependent on the value of N. See also
mount_nfs(1M).
723
The console processor will always probe No workaround available.
the 2nd SCSI controller for devices;
even when no devices are expected to be
configured on it. Some PMC cards have
more than one SCSI controller. Probes
for devices will be done for each controller and thereby add time to the initialization process.
996
The ASTRix tftp command is unreliable.
1047
Reading and writing to the RPIC hard- Workaround implemented. Waiting for
ware registers through the GT64260 final fix.
Device Bus is slow.
Use the ASTRix ftp command for manually loading in large files such as netboot images, netboot system gen files
and DRAM images, instead of trying to
use the tftp command.
PowerMAX OS Version 6.1 Release Notes
6.1
Bugzilla
Number
General Information on Release 6.1
Table 1-6. PowerMAX OS 6.1 Known problems (Continued)
1604
PowerMAX OS 6.1 installation on
Power Hawk Series 700 appears to hang
when the transition from the boot CD to
the packages CD occurs.
The problem is currently under investigation
Software Information Bulletin SWB034
applies
Drives that fail:
Plextor PX-40TSi and PX-40TSUWi
Drives that work:
Plextor PX-W4012TS, PlexWriter
40/12/40S, Plextor PX-32TSi, UltraPlex-32 series
6.1
1680
Starting X on a VQPM video board
mounted on PEX3 casued panic
PowerMAX OS Version 6.1 Release Notes
The problem is currently under investigation
45
Optional Products Available with Release 6.1
10.0.
Optional Products Available with Release 6.1
46
PowerMAX OS Version 6.1 Release Notes
6.1
The following option products are currently available with the distribution of PowerMAX OS
Release 6.1.
Compatibility Issues Between Systems
11.0.
Compatibility Issues Between Systems
PowerMAX OS for the Power Hawk Series 600/700/900 and PowerStack II systems, in general, provides
application source and binary compatibility with PowerMAX OS on the Night Hawk HN6200/HN6800,
TurboHawk and PowerMAXION systems. Application code written, compiled and/or linked on one of
these platforms, will operate on the other with the following exceptions:
1.
Applications written for the Night Hawk, TurboHawk and PowerMAXION systems that use the
real-time clocks (RTCs) may require minor modifications to function properly on a Power
Hawk system.
Seven RTCs are available on the Power Hawk Model 620, four 32-bit timers with a 1
microsecond resolution (tick timers) and three 16-bit timers (Z8536 timers) with a 400
nanosecond resolution.
Six RTCs are available on the Power Hawk Model 640 and PowerStack II, three 32-bit timers
with a 1 microsecond resolution (tick timers) and three 16-bit timers (Z8536 timers) with a 400
nanosecond resolution. If an RCIM is installed, four additional RTCS are available with a one
microsecond resolution. (Note - RCIM requires one PMC slot). Note that PMC slots are not
available on the PowerStack III (MTX PCI Series (7-slot) systems.
Three integral RTCs are available on the Power Hawk Series 700, and nine integral RTCs are
available on the Power Hawk Series 900. Installing a Real-Time Clocks and Interrupts Module
(RCIM) will add four additional rtcs. On the Series 700/900, all of the integral rtcs have the
same fixed interrupt level, and each rtc has its own interrupt vector associated with it. The
Series 700/900 rtcs are only accessible in the default mode. This mode provides different
degrees of control over a real-time clock. Default mode enables a user to control the direction
of the clock count, up or down, the clock count value, the resolution per clock count value (1,
10, 100, 1000 or 10,000 microseconds), and whether the clock automatically restarts counting
when the clock count reached zero.
On the Night Hawk, TurboHawk and PowerMAXION, the RTCs have device names in the form
of: /dev/rrtc/0c[x] where x is between 0 and 4. On the Power Hawk Series 600 and
PowerStack II systems, the tick timers have similar names (/dev/rrtc/0c[0-3]) while the
Z8536 timers use /dev/rrtc/1c[0-2]. On the Power Hawk Series 700 systems, the tick
timers have device names in the form of /dev/rrtc/0c[0-3], and the Power Hawk Series 900
systems have tick timers device names in the form of /dev/rrtc/0c[0-8]. If an RCIM is
installed, the four additional RTCs would have the devices names in the form of:
/dev/rrtc/2c[0-3]. (Note - RCIM requires one PMC slot.)
When utilizing one of the tick timers in “default mode” no code changes are required when
moving between systems, only a re-compilation of the application. When switching from the
tick timer to the Z8536 timer, the application will need to be modified to handle the 400
nanosecond resolution of the timers.
The support for “direct mode” is different on each system. Review the rtc(7) man page for
a description of the “direct mode” functionality on each of the systems.
6.1
2.
PowerMAX OS on the Night Hawk, TurboHawk and PowerMAXION systems provides the
ability to generate an address translation to the Interval Timer hardware and read it directly from
the user process. This is done using the /dev/interval_timer device. Such a POSIX-
PowerMAX OS Version 6.1 Release Notes
47
Compatibility Issues Between Systems
format timer does not exist on the Power Hawk Series 600/700/900 and PowerStack II systems
therefore, programs that depend upon this operation will not operate. There are two alternatives
that provide compatible operation.
3.
a
Use the C library routines, such as clock_gettime(). This routine exists
in the shared C library and generates instructions appropriate to the hardware
platform. On the Night Hawk and PowerMAXION, this routine uses the
hardware Interval Timer while on the Power Hawk it uses the 64-bit Time
Base Register. Note that conversion of the Time Base register to POSIX
format takes notably longer than merely reading the Interval Timer register.
b
Use the Time Base Register exclusively for timing. This is a 64-bit register
that increments at 1/4 the clock speed of the processor bus. For example, on
a Power Hawk system that has a processor bus speed of 66.66MHz, it
increments at the rate of 16.65MHz. This is a processor register on the
PPC604 and is directly accessible from user code. However, conversion to
seconds and nanoseconds may be time consuming depending upon the
processor bus speed.
On Night Hawk, TurboHawk and PowerMAXION systems, PowerMAX OS provides a set of
routines to read and/or write the hardware ipl register. This register determines the interrupt
level that the processor is currently running at. The ability to read/write the ipl value is
especially useful for user-level device drivers.
The spl_map() package includes routines to generate virtual mappings to the hardware ipl
register, along with routines to write the register (spl_request() and spl_manage()) and a macro
to do the same (spl_request_macro()). Mappings are done by using the /dev/spl device node.
Power Hawk Series 600/700/900 and PowerStack II systems do not have a hardware ipl register.
Instead a series of hardware registers implement the ipl functionality through a series of bit
masks. Therefore, routines that map directly to the hardware ipl register will not operate on
these systems. /dev/spl does not exist on Power Hawk systems.
48
a
Do not use mmap(3) to map directly to the hardware register(s).
b
Use the spl_map(), spl_unmap() and spl_manage() library
routines for all ipl register reads/writes. In this case, binary compatibility is
possible between Night Hawk, PowerMAXION and Power Hawk platforms.
c
The spl_xxx routines are contained in libud. If the shared version of
libc is used in the executables, then the shared version of libud must also
be used.
d
If the spl_request() function or the spl_request_macro() is
used, the program will need to be recompiled and relinked on the target
machine, as the macro implementation will be different. On the Power Hawk
it will be the same as the C library routine.
e
Do not depend upon hard-coded, platform-specific spl values, as these spl
values may vary between platforms. Note that the spl_manage() calls
make use of generic PL_xx spl values that are compatible across all
PowerMAX platforms. Regardless of which type of spl values you use, an spl
value of 0 allows all interrupts to occur, and higher spl values prevent
increasingly more interrupts from occurring. The spl values are always less
than 255.
PowerMAX OS Version 6.1 Release Notes
6.1
Compatible operation between Power Hawk and the Night Hawk and PowerMAXION systems
can be accomplished using the following techniques.
Compatibility Issues Between Systems
4.
Programs that depend upon the arrangement and size of kernel data structures may not work
properly due to differing sizes of certain items. These would be programs that utilize
/dev/mem or /dev/kmem to read and/or examine kernel structures. These programs should
be recompiled on the target system to ensure functionality.
5.
Power Hawk Model 620 and Model 710 systems are only single processor systems. Programs
that require multiple processors, or make system calls that reference multiple processors, may
not operate properly.
6.
Edge Triggered Interrupts are available on the Night Hawk and PowerMAXION platforms but
not on the Power Hawk Series 600/700/900 and PowerStack II systems. Therefore, there is no
software support on the above noted systems for Edge Triggered Interrupts. If an RCIM is
installed, four Edge Triggered Interrupts are available. These Edge Triggered Interrupts are
compatible with those on the Night Hawk systems. (Note - RCIM requires one PMC slot.)
7.
At the present time there is no support for NVRAM Global Environment Variable (GEV) on
Night Hawk (‘nh’) platforms. The nvram system call always returns ENOSYS on these
platforms and the gev set of commands are not present.
8.
Applications written for the Power Hawk Series 700/900, which utilize the AltiVec instructions
or registers, are not executable on the earlier Night Hawk and Power Hawk systems which are
based on the 604e processor.
9.
The physical CPU addresses that are used to access the physical VME bus address spaces (A32,
A24, A16) differ between Power Hawk Series 700 and 900 systems. Therefore, applications or
user-level device drivers that access VME bus address spaces directly from user-space should
use the vme_address(3C) library routine to obtain the correct physical CPU addresses to use for
creating virtual address translations to VME bus addresses.
6.1
10. On other PowerHawk platforms, kernel tunables define the address range for VME A32 address
space. However, on PowerHawk Series 900 platforms the address range for VME A32 address
space is directly determined from the PCI Memory Space parameters in the STAR config
command. Please consult the Power Hawk Series 900 Console Reference Manual (Pubs No.
0830060) section "PCI Memory Space Configuration", for more information on this subject.
PowerMAX OS Version 6.1 Release Notes
49
Service Releases
12.0.
Service Releases
Service Releases containing updates, enhancements, patches, etc., are periodically released by
Concurrent Computer Corporation on an “as-required-basis”.
Descriptions of available updates and patches to Concurrent Computer Corporation software products
can be examined and ordered on this site: http://www.ccur.com/realtime/rt_si.htm
To find out what patches are installed on your system enter the following command:
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
patch
50
base-001
base-002
base-003
cmds-001
cmds-002
cnd-001
cnd-002
cnd-003
dec-001
dec-002
egl-001
egl-002
fbs-001
fbs-002
hc-001
inet-001
inet-002
inet-003
kdb-001
librt-001
man-001
man-002
(screen output example shown below)
Base System Patch 001 (4.3P1)
Base System Patch 002 (4.3P2)
Base System Patch 003 (4.3P3)
Advanced Commands Patch 001 (4.3P1)
Advanced Commands Patch 002 (4.3P2)
Condor Ethernet Driver Patch 001 (4.3P1)
Condor Ethernet Driver Patch 002 (4.3P2)
Condor Ethernet Driver Patch 003 (4.3P3)
DEC Ethernet Driver Patch 001 (4.3P2)
DEC Ethernet Driver Patch 002 (4.3P3)
Eagle Ethernet Driver Patch 001 (4.3P2)
Eagle Ethernet Driver Patch 002 (4.3P3)
Frequency Based Scheduler Patch 001 (4.3P1)
Frequency Based Scheduler Patch 002 (4.3P2)
Concurrent C Compilation System
Internet Utilities Patch 001 (4.3P1)
Internet Utilities Patch 002 (4.3P2)
Internet Utilities Patch 003 (4.3P3)
Kernel Debugger Patch 001 (4.3P1)
Real-time Libraries Patch 001 (4.3P2)
On-line Manual Pages Patch 001 (4.3P1)
On-line Manual Pages Patch 002 (4.3P2)
PowerMAX OS Version 6.1 Release Notes
6.1
pkginfo -c patch | pg
Direct Software Support
13.0.
Direct Software Support
Software support is available from a central source. If you need assistance or information about your
system, please contact the Concurrent Software Support Center at our toll free number 1-800-245-6453.
For calls outside the continental United States, the number is 1-954-283-1822. The Software Support
Center operates Monday through Friday from 8 a.m. to 5 p.m., Eastern Standard time. You may submit
a request for assistance at any time by using the Concurrent Computer Corporation web site at
http://www.ccur.com/isd_support_contact.asp
6.1
Calling the Software Support Center gives you immediate access to a broad range of skilled personnel
and guarantees you a prompt response from the person most qualified to assist you. If you have a question
requiring on-site assistance or consultation, the Software Support Center staff will arrange for a field
analyst to return your call and schedule a visit.
PowerMAX OS Version 6.1 Release Notes
51
Manual Pages
14.0.
Manual Pages
Manual pages can be found online. The easiest way to access these is by typing man followed by a
manual page name or a command name. Typing man manual, for example, will show online the
manual page that contains a categorized listing of all the current manual names, publication numbers and
latest revision numbers.
On-line manual pages also exist for most software and hardware manuals in this document. These manual
pages provide a description of each manual and also list all related publications, where applicable. To
access, type apropos pubs number that will display the manual page name. Typing man manual page
name will then show on the screen the manual page you are interested in.
52
PowerMAX OS Version 6.1 Release Notes
6.1
Printed copies of the format(1M), pkgadd(1M) and putdev(1M) man pages that may need to be
referenced during installation are provided at the end of these release notes.
Manual Pages
6.1
(This page left intentionally blank)
PowerMAX OS Version 6.1 Release Notes
53
54
PowerMAX OS Version 6.1 Release Notes
6.1
Manual Pages
format(1M)
format(1M)
NAME
format - format program for Generic Disk driver disks
SYNOPSIS
/sbin/format [-f |-l |-v] disk
DESCRIPTION
format is a program which is used to format, verify, manage flawmaps, or define partition sizes for disk
drives on HSA, or other SCSI disk controller that employs the Generic Disk driver (gd).
Geometry Blocks
format maintains (in memory) two copies of the disk’s geometry block. The “disk geometry block” is an
exact copy of the geometry block as it exists on the disk. The “current geometry block” is a copy of the
geometry block which may be modified with the “partition” command and then written to disk.
Whenever the “disk geometry block” is read from the disk, the “current geometry block” is initialized to be
the same as the “disk geometry block;” if there is no geometry block on the disk, the “current geometry block”
partitions are initialized to default values. format has several commands which manipulate the “disk
geometry block” and/or the “current geometry block.” For example, the “status” command displays the
partition sizes of the “disk geometry block;” the “partition” command displays the partition sizes of the
“current geometry block.”
RUNNING FORMAT
format may be run either interactively (if no option is specified) or non-interactively (if an option is
specified). Only a single disk may be specified for each invocation of the format program. Multiple disks may
be formatted simultaneously by running multiple copies of format.
“format <disk>” starts the format program in an interactive mode. When format starts, it opens the
raw device for slice 7 of the disk specified. The disk may either be specified as a disk “unit” number or as a
full path to a raw device node for slice 7. format must have read/write access to the device.
In interactive mode, format accepts and executes commands until the program is terminated. A list of
commands may be obtained by typing “?”. Commands may be abbreviated; only enough characters are
required to uniquely identify the command.
“format <option> <disk>” starts the format program in a non-interactive mode. format executes
the single command which is specified by the option. format options are:
-f format the disk (the “format” command)
-v verify the disk (the “verify” command)
After running format, if the disk was formatted or the partition sizes were modified, it is necessary to rebuild the file systems on the disk.
COMMANDS
“? [commandname]”
If the optional parameter is not specified, this command prints a summary of all the commands. If the optional
[commandname] parameter is specified, this command prints information about that specific command,
i.e., command syntax, what the command does, and warnings associated with the command.
“format”
This is a single atomic operation, which may take 10 minutes or more on large disks. When formatting has
successfully completed, the “current geometry block” and the flawmap are written to the drive description
tracks.
“partition [partition#] [size]”
If no parameters are specified, this command prints the partition sizes and partition starting locations of the
“current geometry block”. If the optional parameters are present, this command sets the size of
[partition#] to be [size] units. Typing “partition default” is a special form of the
command, which sets all partitions sizes to default values. The partition start and size are modified only in
the “current geometry block”; the “disk geometry block” is not changed.
53
format(1M)
format(1M)
Partition sizes are always given in the current unit, which may be modified with the “unit” command. The
size actually reserved for a partition is always rounded up to a cylinder boundary. Each time the size of a
partition changes, the starting locations of all subsequent partitions (with larger partition numbers) is also
changed. This is because partition 1 always follows partition 0, etc.
“quit“
This command terminates the program. If the “current geometry block” differs from the “disk geometry
block”, format will not terminate. This means that the “partition” command was used to modify the
geometry block, but the “current geometry block” was never written to disk. At this point it is necessary to
use either the “format” or “write” command to update the geometry block on the disk, or to force a
program exit by typing “q!”.
“read”
This command reads the geometry block from the disk. The “disk geometry block” and “current geometry
block” are initialized. If there is no geometry block on the disk, the “current geometry block” is initialized
with default partition sizes.
“sectorsize [size]”
This command is used to specify an alternate physical sector size to be used when formatting the disk. The
size specified is in bytes. The physical sector size on the disk is not changed until the “format” command
is issued. The default sector size is 512 bytes, unless the disk has already been formatted to a different value.
“status”
This command prints detailed information about the disk, including the partition sizes and a description of
how the disk is organized. Partition sizes are always given in the current unit, which may be modified with
the “unit” command.
“unit [unit]”
This command prints/selects the unit for partition sizes. If no parameters are present, then the current unit is
printed. If the optional [unit] parameter is present, then the current unit becomes [unit]. [unit] may be
any of: “sectors”, “cylinders”, “kbytes”, “megabytes”, or “percent” (percent of disk). Only
the first letter of [unit] is required.
“verify”
This command verifies the disk by reading every sector. A “.” is printed as each cylinder is completed.
“write”
This command writes the “current geometry block” to the disk. The “disk geometry block” is set equal to the
“current geometry block”.
SEE ALSO
gd(7), hsa(7), ise(7), format(8), errpt(1M)
NOTES
The ability to format an IDE drive is not available. IDE drives are formatted at the factory and this cannot be
redone in the field. Similarly, there is no “verify” function.
54
pkgadd(1M)
pkgadd(1M)
NAME
pkgadd - transfer software package or set to the system
SYNOPSIS
pkgadd [-d device] [-r response] [-n] [-q] [-l] [-a admin]
[-p] [pkginst1 [pkginst2 [. . .]]]
pkgadd -s spool [-d device] [-q] [-l] [-p] [pkginst1 [pkginst2 [. . .]]]
DESCRIPTION
pkgadd transfers the contents of a software package or set from the distribution medium or directory to
install it onto the system. A package is a collection of related files and executables that can be independently
installed. A set is made up of a special-purpose package, referred to as a Set Installation Package (SIP), and
a collection of one or more packages that are members of the set. The SIP controls the installation of the set.
pkgadd checks that all packages listed on the command line are on the installation medium. If any of the
packages listed does not exist, no changes are made to the system, that is, none of the listed packages are
installed. Used without the -d option, pkgadd looks in the default spool directory for the package
(/var/spool/pkg). Used with the -s option, it writes the package to a spool directory instead of
installing it.
Error messages are always logged (see -l, below). In addition, when pkgadd terminates, it will send mail
(by default, to root) with all the error messages and a summary of which packages installed completely,
partially, or not at all.
For all appropriate files, the Mandatory Access Control (MAC) and privilege information provided in the
pkgmap file will be set. MAC and privilege can only be set on files installed on a sfs-type file system (that
is, they will be ignored for non-sfs-type file systems). These entries are provided for upward compatibility
with the Enhanced Security Utilities. The following options are available.
-d device
Installs or copies a package/set from device. device can be: (a) the full pathname to a directory, file,
or named pipe (such as /var/tmp); (b) the device identifiers for tape or disk devices (such as /
dev/rmt/* or /dev/dsk/*) [see intro(7)]; (c) a device alias (such as diskette1); or (d)
“-” which specifies packages in datastream format read from standard input. The default device is
the installation spool directory (/var/spool/pkg).
For device identifiers, the device specified (either by pathname or alias), must have an entry in the
device table (/etc/device.tab). If no entry exists in the device table, pkgadd will abort.
Device table entry must have volume attribute set initially via putdev command, otherwise,
pkgadd will abort. Typically usage:
# putdev -a ctape1 cdevice=/dev/rmt/0m volume=”volume=””
(See putdev for additional information.)
A device alias is the unique name by which a device is known. (For example, the alias for a cartridge
tape drive might be ctape1.) The name must be limited in length to 64 characters
(DDB_MAXALIAS) and can contain only alphanumeric characters and/or any of the following
special characters: underscore (_), dollar sign ($), hyphen (-), and period (.). No two devices in the
database can share the same alias.
-r response
Identifies a file or directory, response, which contains the answers to questions posed by a “request
script” during a previous pkgask session conducted in interactive mode [see pkgask(1M)]. When
pkginst is a package, response can be a full pathname or a directory; when pkginst is a SIP, response
must be a directory. For a complete description of request scripts and response files, see your system
administration or software packaging guides.
-n
Installation occurs in non-interactive mode. The default mode is interactive.
55
pkgadd(1M)
pkgadd(1M)
-q
Installation is performed in quiet mode. Only prompts requesting user input and error messages are
displayed on the screen. Error messages are not sent to the standard error output; they are only
logged to /var/sadm/install/logs/pkginst.log.
-a admin
Defines an installation administration file, admin, to be used in place of the default administration
file to specify whether installation checks (such as the check on the amount of space, the system
state, and so on) are done. [For a description of the format of an admin file, see admin(4).] The
token none overrides the use of any admin file, and thus forces interaction with the user. Unless a
full pathname is given, pkgadd looks in the /var/sadm/install/admin directory for the
file. By default, the file default in that directory is used. default specifies that no checking
will be done, except to see if there is enough room to install the package and if there are
dependencies on other packages. The -a option cannot be used if pkginst is a SIP.
-p
Do not give the initial prompt to the user to insert the distribution medium. All other prompts will
continue normally.
pkginst
A short string used to designate a package/set. It is composed of one or two parts: pkg (an
abbreviation for the package/set name) or, if more than one instance of that package exists, pkg plus
inst (an instance identifier). (The term “package instance” is used loosely: it refers to all
instantiations of pkginst, even those that do not include instance identifiers.)
The package name abbreviation pkg is the mandatory part of pkginst. [See pkginfo(1),
pkginfo(4).]
If pkginst is a SIP, the SIP controls installation of the set by using request scripts and preinstall
scripts. The SIP request script, not the package installation tools, is responsible for prompting the
user for responses and taking the appropriate actions. If the request script fails, only the SIP will be
processed. For a complete description of request and preinstall scripts, see your system
administration and/or software packaging guides.
The second part (inst), which is required only if you have more than one instance of the package in
question, is a suffix that identifies the instance. This suffix is either a number (preceded by a period)
or any short mnemonic string you choose. If you don’t assign your own instance identifier when one
is required, the system assigns a numeric one by default. For example, if you have three instances of
the Advanced Commands package and you don’t create your own mnemonic identifiers (such as
old and beta), the system adds the suffixes .2 and .3 to the second and third packages,
automatically.
To indicate all instances of a package, specify ’pkginst .*’, enclosing the command line in single
quotes, as shown, to prevent the shell from interpreting the * character. Use the token all to refer
to all packages available on the source medium.
-s spool
Reads the package into the directory spool instead of installing it.
USAGE
The -r option can be used to indicate a directory name as well as a filename. The directory can contain
numerous response files, each sharing the name of the package with which it should be associated. This would
be used, for example, when adding multiple interactive packages with one invocation of pkgadd. Each
package that had a request script would need a response file. If you create response files with the same name
as the package (for example, package1 and package2) then, after the -r option, name the directory in which
these files reside.
The -n option will cause the installation to halt if any interaction is needed to complete it.
When invoked with no pkginst specified on the command line, pkgadd only displays the names of sets if at
least one SIP exists on the media. Because of this, you shouldn’t include packages on the same media if some
are members of sets and some are not. If you do, the packages which are not members of sets can be installed
only if their pkginst names are provided on the command line.
The pkgadd command checks to see if any of the files in pkginst are already installed on the system and, if
any are, saves this fact before continuing with installation. Later, pkgadd won’t reinstall these files on the
system. If one of the package’s installation scripts removes such a file, the result will be that the file will no
longer be on the system when package installation completes.
56
pkgadd(1M)
pkgadd(1M)
The pkgadd command does not uncompress any files that were already compressed (that is, only those in
“.Z” form) before being processed by pkgmk.
The pkgadd command, when invoked with the -d device option to add packages from the device to the
system, will copy files from the device to the default temporary directory: /var/tmp. If the /var file
system does not contain enough free space to complete this command successfully, then the command user
may set and export the environment variable, TMPDIR, to some temporary directory on a file system that does
contain enough free space and re-execute the command. Once the command completes successfully, the
temporarily extracted files are removed from the system.
FILES
/etc/device.tab
device table
/var/sadm/install/admin/default
default package administration file
/var/sadm/install/logs/pkginst.log
error message log
/var/spool/pkg
default spool directory
/usr/lib/locale/locale/LC_MESSAGES/uxpkg
language-specific message file
[See LANG on environ(5).]
/var/tmp
default location of TMPDIR
REFERENCES
admin(4), compver(4), copyright(4), depend(4), installf(1M), intro(7),
pkgask(1M), pkgchk(1M), pkginfo(1), pkginfo(4), pkgmap(4), pkgparam(1),
pkgrm(1M), putdev(1M), removef(1M), setinfo(4), space(4)
57
pkgadd(1M)
pkgadd(1M)
58
putdev(1M)
putdev(1M)
NAME
putdev - create and update the device database
SYNOPSIS
putdev -a alias [secdev=value][attribute=value[. . .]]
putdev -m device attribute=value[attribute=value[. . .]]
putdev -d device[attribute[. . .]]
putdev -p device attribute=value[,value . . .]
putdev -r device attribute=value[,value . . .]
DESCRIPTION
The putdev command is used to add a new device to the Device Database (DDB), modify an existing
device’s attributes, or remove a device entry from the DDB. It also allows appending new values to attributes
that take value-lists (separated by commas), and removal of specific values from value-lists.
Options
putdev takes the following options:
-a alias
Add a device to the DDB using the specified attributes. The device must be referenced by its alias.
The secdev attribute is defined only if the Enhanced Security Utilities are installed.
-m device
Modify a device entry in the DDB, using the specified attribute values. If a specified attribute does
not exist in the device entry, putdev adds the specified attribute to the entry. It also modifies
attributes that already have a value with the value specified.
-d device
Remove a device entry from the DDB, when executed without the attributes argument. If the
attribute argument is specified, the attribute and its value are deleted from the device entry.
-p device
Append the list of values to the attribute value-list of the device. If the value item is multiply defined
in the input value-list or already defined in the DDB, putdev fails and prints an error message.
-r device
Remove the list of values from the attribute value-list, of the device. The command succeeds, even
if the value has been removed or is not defined for the attribute in the DDB.
alias
Define the alias name of the device, a value which must be unique throughout the DDB. alias is
limited to 64 characters (DDB_MAXALIAS) and should contain only alphanumeric characters and
any of the following special characters: . (period), _ (underscore), $ (dollar sign), and - (hyphen).
secdev
Designate the alias of the secure device that defines all the security attributes, and is used only if the
Enhanced Security Utilities are installed. If secdev is not specified during creation (-a option) or is
deleted (-d option), the current alias is used as the default value of secdev. The validation rules for
secdev are the same as those for alias.
59
putdev(1M)
putdev(1M)
device
Designate the absolute pathname or alias name of the device whose attribute is to be added,
modified, or removed. If device is a pathname, then the attributes of the alias to which it maps are
updated.
attribute
Designate a device attribute to be added, modified, or deleted. This prevents an accidental
modification or deletion of a device’s alias from the DDB.
value
Designate the value to be assigned to a device’s attribute. If any of the values are invalid, then
putdev fails and prints an error message.
Whenever the attributes in a Device Database file are updated, the old version of the file is saved to a file with
the capital letter ‘‘O’’ prefixed to the file name. If there are errors in the modification of device attributes, you
can recover the old versions of the Device Database files.
Attributes
Following are all of the attributes which can be defined for a device:
alias
The unique name by which a device is known. No two devices in the database may share the same
alias name. The name is limited in length to 64 characters (DDB_MAXALIAS) and should contain
only alphanumeric characters and any of the special characters: underscore (_), dollar sign ($),
hyphen (-), and period (.).
bdevice
The absolute pathname to the block special device node associated with the device, if any, with
maximum length of PATH_MAX. This attribute is optional.
bdevlist
A list of additional pathnames of block device special files which map to the same logical or secure
device. Each item in the list is separated by a comma, and each must be an absolute pathname of the
device special file, with a maximum length of PATH_MAX. Since, this attribute takes a list of values,
putdev -p | -r can be used for this attribute. This attribute is optional.
capacity
The capacity of the device or of the typical volume, if removable.
cdevice
The absolute pathname to the character special device node associated with the device, if any, with
maximum length of PATH_MAX. This attribute is optional.
cdevlist
It contains a list of additional pathnames of character device special files mapping to the same logical
or secure device. Each item in the list is separated by a comma, and each must be an absolute
pathname of the device special file, with a maximum length of PATH_MAX. Since, this attribute
takes a list of values, putdev -p | -r can be used for this attribute. This attribute is optional.
cyl
Used by the command specified in the mkfscmd attribute.
desc
A description of any instance of a volume associated with this device (such as floppy diskette).
60
putdev(1M)
putdev(1M)
dpartlist
The list of disk partitions associated with this device. Used only if type=disk. The list should
contain device aliases, each of which must have type=dpart.
dparttype
The type of disk partition represented by this device. Used only if type=dpart. It should be either
fs (for file system) or dp (for data partition).
erasecmd
The command string that, when executed, erases the device.
fmtcmd
The command string that, when executed, formats the device.
fsname
The file system name on the file system administered on this partition, as supplied to the/usr/
sbin/labelit command. This attribute is specified only if type=dpart and
dparttype=fs.
gap
Used by the command specified in the mkfscmd attribute.
mkfscmd
The command string that, when executed, places a file system on a previously formatted device.
mountpt
The default mount point to use for the device. Used only if the device is mountable. For disk
partitions where type=dpart and dparttype=fs, this attribute should specify the location
where the partition is normally mounted.
nblocks
The number of blocks in the file system administered on this partition. Used only if type=dpart
and dparttype=fs.
ninodes
The number of inodes in the file system administered on this partition. Used only if type=dpart
and dparttype=fs.
norewind
The name of the character special device node that allows access to the serial device without
rewinding when the device is closed.
pathname
Defines the pathname to an i-node describing the device (used for non-block or character device
pathnames, such as directories).
removable
This should be set to true to identify a device as a removable one (such as a tape device).
type
A token that represents inherent qualities of the device. Standard types include: 9-track, ctape, disk,
directory, diskette, dpart, and qtape.
volname
The volume name on the file system administered on this partition, as supplied to the /usr/sbin/
labelit command. Used only if type=dpart and dparttype=fs.
volume
A text string used to describe any instance of a volume associated with this device. This attribute
should not be defined for devices which are not removable.
Enhanced Security Attributes
The following security attributes can be defined for a device alias if the Enhanced Security Utilities are
installed:
61
putdev(1M)
putdev(1M)
secdev
The alias name of the physical device or secure device, and is or secure device, and is unique
throughout the Device Database(DDB). This alias name is limited to 64 characters
(DDB_MAXALIAS), and should contain only alphanumeric characters and the following special
characters: "_", "$", "-" or ".". For a secure device alias this attribute’s value is the same as the
device’s alias. For a logical device alias, this attribute’s value is different from the device alias. By
default, secdev is defined to be equal to the device’s alias.
range
The sensitivity Mandatory Access Control (MAC) level range of the device. It should by a hilevellolevel pair, where hilevel and lolevel are both MAC level names or fully qualified levels. The "-"
character is the delimiter between hilevel and lolevel. These levels are stored in the DDB as LIDs,
converted to ASCII characters. The LIDs are validated against the Label Translation
Database, and it is ensured that hilevel dominates lolevel, before they are saved in the DDB. This
attribute must be defined.
state
Determines whether the device is to be used as a private or public device. It can take any one of the
values: private, public, or pub_priv. If it is set as pub_priv, then the device can either be used as
private or public device. If the startup attribute is enabled, then the device is allocated as private, if
the state was set to either private or pub_priv. This attribute must be defined.
mode
Determines the mode of the device. This attribute can take one of the values: static or dynamic. This
attribute must be defined.
.startup\
is a flag (y[es]/n[o]) that indicates whether the device is allocated during startup or not. This attribute
is optional, and startup default value is no.
startup_level
Defines the MAC level at which the device should be set at startup. This can be specified as a level
name or fully qualified level. However, the value is saved in the DDB as an ASCII LID value. This
attribute is optional.
startup_owner
Defines the owner of the device. The value of startup_owner can be specified as the uid or
user name followed by the access permissions. The value must be specified in the format: uid>rwx.
If any of the read, write, or execute access is denied, that field must contain a - (hyphen). The >
character serves as delimiter between the uid or user name and the access permissions. The uid or
user name must be defined on the system (in /etc/passwd), at the time this attribute is defined. This
attribute is optional but must be defined if attribute startup is set to yes.
startup_group
Defines the group to which the device belongs. The value of startup_group can be specified as
the gid or group name followed by the access permissions. The value must be specified in the
format: gid>rwx. If any of the read, write or execute access is denied, that field must contain a (hyphen). The > character serves as delimiter between the gid or group name and the access
permissions. The gid or group name must be defined on the system (in /etc/group), at the time this
attribute is defined. This attribute is optional but must be defined if attribute startup is set to yes.
startup_other
Defines the access permissions for other. The value of startup_other must be specified in the
format: >rwx. If any of the read, write or execute access is denied, then that field must contain a (hyphen). This attribute is optional but must be defined if attribute startup is set to yes.
ual_enable
A flag that enables or disables depending on its value the user authorization list defined in the users
and other attributes. This attribute can take one of the values: y[es] or n[o]. If ‘‘y’’, then the user
authorization list is checked when authorizing an user to use this device. If ‘‘n’’, then no users are
62
putdev(1M)
putdev(1M)
authorized to use this device. This attribute is optional, and value assumed as no if ual_enable
is not defined.
users
The user authorization list that defines the allocation permissions for users. Each item is a uidauthorization or username-authorization pair separated by a > character. The items in the list are
separated by commas. The attribute’s value must be specified in the format: uid1>n,uid2>n,uid3>y.
Each uid or username must be unique in a device entry, and all uids or usernames must be defined
in /etc/passwd, when this attribute is defined. Since this attribute takes a list of values, putdev
-p | -r can be used. This attribute is optional.
other
Defines the authorization permissions for other. This attribute contains only one item and it can take
one of two values: y[es], or n[o]. This attribute is optional, and its value is assumed as no if other
is not defined.
Enhanced Security Usage
The following rules and guidelines should be followed whenusing the putdev command in enhanced security
installations.
The alias names of devices must be valid [see description under NOTICES] and unique throughout
the DDB; otherwise, putdev fails.
The pathnames to device special files in attributes cdevice, bdevice, cdevlist, and devlist must in
attributes cdevice, bdevice, cdevlist, and bdevlist must be absolute pathnames. They cannot be
repeated within an entry or occur in multiple entries. The putdev command checks for uniqueness
of pathnames; otherwise, putdev fails.
Security attributes can be defined for device, or alias only if the system is configured for multilevel
security; otherwise, putdev fails.
The MAC level values for the security level range (hilevel - lolevel) must be valid security level
aliases or fully qualified level names definedin the Level Translation Database (LTDB); otherwise,
putdev fails. If hilevel does not dominate lolevel, then putdev fails.
The secdev attribute requires the following special handling:
The secdev attribute is used to define the essential security attributes of a device, that is, those
attributes required when the Enhanced Security Utilities are installed. secdev must be valid (see
description under Security Attributes) and unique throughout the DDB; otherwise putdev fails.
By default, when adding a new device alias into the Device database, if the secdev attribute is not
defined at the command line, the new device entry is assigned a secdev equal to its alias.
The alias that defines security attributes of a device is called a secure device alias.You can define
other non-security attributes or this alias, if needed. For all secure devices, secdev must have same
value as alias; if secdev is different from alias, the alias is a logical alias.
The security attributes range, state, and mode, must be defined for every secure device alias.
When adding (using -a) or modifying (using -m) a device entry and specifying the secdev attribute
not equal to the alias being added or modified, putdev performs the following checks in the order
specified:
63
putdev(1M)
putdev(1M)
1. If the essential security attributes are being defined for alias, the command fails and displays an
error message. An entry defining the essential security attributes must have the secdev attribute
be equal to its alias.
2. If the essential security attributes are not being defined for alias, and if the specified secdev does
not exist in the Device Database, a warning message is displayed.
3. If the essential security attributes are not being defined for alias, and the specified secdev exists
in the Device Database but does not define the essential security attributes, putdev fails and
displays an error message.
4. If the essential security attributes are not being defined for alias, and the specified secdev exists
in the Device Database and defines the essential security attributes, then the command is
successful.
You should create the secure alias before creating any logical aliases that map to the same secure
alias. Similarly, you should not remove a secure device alias if any logical alias are currently mapped
to that secure alias.
Additional aliases that share the security attributes defined for a secure device can be created by
specifying their secdev to have the same value as the alias of the secure device. If secdev is not
specified, and the essential security attributes are also not specified, then a logical device entry is
created that does not have security attributes.
Special handling of the essential security attributes:
The essential security attributes, mode, state, and range must be created (using -a or -m) and deleted
(using -d) together. Otherwise, putdev fails and issues an error message.
The essential security attributes of a secure alias can be modified (-m) separately after they are
defined.
If the essential security attributes are being deleted from a device entry whose alias is a secdev
attribute for at least another entry in the Device Database, then putdev fails and displays an error
message.
Enhanced Security Example
The following example shows you how to create one secure device (tapedrive1) and two device aliases
(slowtape, fasttape) that map to the secure device. (In the following example, the input is split onto three lines;
you should enter the commands as one line.)
putdev -a tapedrive1 range="SYS_PRIVATE-SYS_PUBLIC" \
state="public" mode="static" startup="n" ual_enable="y" \
users="100>n,101>n" other=">y"
putdev -a slowtape secdev="tapedrive1" cdevice="/dev/tape800"
putdev -a fasttape secdev="tapedrive1" cdevice="/dev/tape1600"
64
putdev(1M)
putdev(1M)
The preceding command sequence creates one secure device alias (tapedrive1) with the specified security
attributes for the tape drive, and two logical device aliases (slowtape and fasttape) with the specified nonsecurity attributes in the DDB.
However, one could create one entry per device with all security attributes specified on the command line:
putdev -a tape1 range="hilevel-lolevel" state="public" mode="static" \
startup="n" ual_enable="y" users="100>n,101>n" other=">y" \
cdevlist="/dev/tape800,/dev/tape1600" desc="tape device"
Another example for adding a disk (say disk number 4) to the Device Database on a machine running with
Enhanced Security is:
putdev -a disk range="SYS_RANGE_MAX-SYS_RANGE_MIN" state="public" \
mode="static" cdevice="/dev/rdsk/4s0" bdevice="/dev/dsk/4s0" \
mountpt="/home" desc="Disk containing the /home"
The DDB can be queried for any alias, or attribute value using the devattr and getdev commands.
Files
/etc/device.tab
/etc/security/ddb/ddb_dsfmap installed by the Enhanced Security Utilities
/etc/security/ddb/ddb_sec installed by the Enhanced Security Utilities
Exit Codes
If putdev is successful, it returns an exit code of 0. Otherwise, it returns one of the following exit codes and
prints the corresponding error message:
1
incorrect usage
USAGE: putdev -a alias [attribute=value] . . .
2
Device Database in inconsistent state - notify administrator
2
Device Database could not be accessed or created
3
alias already exists in Device Database
3
device does not exist in Device Database
?
dsf already exists in Device Database
6
invalid alias or invalid pathname device
4
hilevel does not dominate lolevel in attribute range
4
invalid value for attribute attr
level= level not defined in LTDB
4
essential security attributes (range,state,mode) must be specified together for alias
6
invalid value for attribute attr
user/uid= uid not defined in system
65
putdev(1M)
putdev(1M)
6
invalid value for attribute attr
group/gid= group not defined in system
6
invalid value for attribute attr
invalid permissions specified perm
6
invalid value for attribute attr
invalid delimiter specified in value
6
value multiply defined for attribute attr
6
alias does not define essential security attributes(range,state,mode)
6
alias must be defined with essential security attributes(range,state,mode)
6
cannot specify security attrs for alias and map to another secdev alias
6
Device Database in use. Try again later.
3
alias not defined in Device Database
5
system service not installed
The Enhanced Security Utilities are not installed.
NOTES
Before modification, putdev will back up the most recent copies of the device database files (see Files above),
and place them (with a .old extension) in the same path as the original. In the event that the device database
becomes corrupted, you will need to copy all of the .old files back.
REFERENCES
devattr(1M), getdev(1M)
66