Digital PDFs
Documents
Guest
Register
Log In
EK-1184A-TM-PR2
May 1985
338 pages
Original
14MB
view
download
OCR Version
13MB
view
download
Document:
PDP-11/84 System Installation and Technical Reference Manual
Order Number:
EK-1184A-TM
Revision:
PR2
Pages:
338
Original Filename:
OCR Text
EK-1184A-TM-PR2 PRELIMINARY PDP-11/84 - SYSTEM INSTALLATION AND Y_—YTY® an I 1} TECHNICAL REFERENCE MANUAL I Jaiilaliltlall EK-1184A-TM-PR2 PRELIMINARY PDP-11/84 SYSTEM INSTALLATION AND TECHNICAL REFERENCE MANUAL Prepared by Educational Services of Digital Equipment Corporation First Edition, May 1985 C>Qigital Equipment Corporation 1985. All Rights Reserved. Printed in U.S.A. oses and ment is for informationalnotpurp The material in this docu be construed out notice; it should ion. is subject to change with Digital orat Digital EquipmentonsiCorp as a commitment by bility for any errors ion assumes no resp ~ Equipment Corporat that may appear in this document. assumes no responsibilitythatforis the Digital Equipment Corporation not use or reliability of its software on equipment supplied by Digital. NOTICE: o frequency uses, and may emit dradi This equipment generates, ly with been tested and foun to comp energy. The equipment has utin g device pursuant to Subpart J s A comp the limits for a Clas provigde resonable s, which are designed tointe Rule FFC of 15 Part of ence when radio frequency Operrfer protection against such n of this atio ronment. a commercial envi operated in resi in which ence rfer dential area may cause inte equipment in a measures take ired to nse may be requ case the user at his own expe . to correct the interference The following Corporation: d]ifa]i[t]a] I DEC DECmate DECUS DECwriter DIBOL LSI-11 MASSBUS are trademarks Micro/PDP-11 MicroVaxX PDP P/0S Professional Q-Bus Rainbow RSTS of Digital RSX RT UNIBUS VAX Equipment VAXcluster VMS VT Work Processor TABLE OF CONTENTS PAGE CHAPTER 1 SYSTEM INTRODUCTION . . o . INTRODUCTION .1.1 . . . SYSTEM COMPONENTS AND VARIATIONS 1.2 1.2.1 KDJ11-BF Processor Module . . . . . . . KTJ11-B UNIBUS Adapter ModulE 1.2.2 1.2.3 MSV11-JB/JC Memory Module . . . . . . o . . . UNIBUS Terminator 1.2.4 1.2.5 Monitor and Distribution Module. . . 1.2.6 Minimum Load Module . . . . . 1.2.7 Power Supply . . . . . . . . . . Console Serial Line Board 1.2.8 . . . . . Backplane Assembly . 1.2.9 . . . . . Front Panel Assembly 1.2.10 . . . . Controller Power Cabinet 1.2.11 . e . . . Cabinet Blower Unit. 1.2.12 1.2.,13 Box Cooling Fans . . . . . . Additional Expansion and Memory Options. 1.2.14 1.3 SYSTEM SPECIFICATIONS. . . . . . . o . . . . RELATED DOCUMENTS. 1.4 CHAPTER 2 1-1 1-2 1-5 1-6 1-6 1-6 1-6 1-6 1-6 1-7 1-7 1-8 1-8 1-8 1-8 1-8 1-9 1-14 o . . 2-1 2-1 2-2 SITE PREPARATION AND INSTALLATION . o . . . . SHIPPING SPECIFICATIONS 2.2 o o . UNPACKING INSTRUCTIONS. 2.3 . . . . Cabinet Unpacking 2,3.1 . . . . . Box Unpacking 2.3.2 2.4 SYSTEM INSTALLATION . o . . Cabinet Mechanical Installation. 2.4.1 . Box Mechanical Installation. 2.4.2 . Console Serial Line Hookup . 2.4.3 . . SITE PREPARATION . 2.1 Cabinet Site Preparation 2.1.1 . Box Site Preparation 2.1.2 o . . . . . . o . o o o . . 2.4.4 Cabinet Switch Settings. . . . 2.4.5 Box Switch Settings. . . . . 2.4.6 Cabinet Power Hookup . . . . 2.4.7 Box Power Hookup . . . o 2.5 SYSTEM CONTROLS AND INDICATORS . . 2.5.1 Front Panel. . . . . o o 2.5.2 Console Serial Line Distribution Board 2.5.3 Serial Communications Port . . . 2.5.4 Baud Rate Select Switch. . . . 2.6 SYSTEM HARDWARE CONFIGURATION, o o 2.6.1 KDJ11-BF Module Configuration . . 2.6.2 KTJ11-B Module Configuration . o 2.6.3 . . . . . . . . . . . . . . . . . . Monitor and Distribution Module iii . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Configuratio 2-4 2-5 2-5 2-5 2-5 2-6 2-6 2-8 2-9 2-11 2-13 2-13 2-14 2-14 2-17 2-17 2-17 2-18 2-19 2-19 2-20 -4 -] o L] e o ] o 2-28 2-31 3-37 o 2-31 2-32 2-33 2-34 L . ] Cabinet SPC Cable Routing 2=27 L4 Box SPC Cable Routing . . [} . . ® SPC Backplane Locations SPC MODULE INSTALLATION ] Expansion Backplane Installation NPG and BG Jumper Lead Routing Backplane Power Connections . 2-23 2-24 2=25 -} . L] EXPANSION BACKPLANE INSTALLATION 2-20 (-] Minimum Load Module o e . [ NN MSV11-JB/JC Module Configuration CABINET BATTERY UP UNIT INSTALLATION BOX BATTERY BACK UP UNIT INSTALLATION [ [} [ L] ] [ [ ° [] L] L] L] [] [ e [ [ L) [] ® [ o o . o o o & o o . Program Interrupt Request Register . o Configuration and Dlsplay Reglster . o CPU Error Register . . Maintenance Register. o Page Control Register o STACK LIMIT PROTECTION. . . . o . . o . . . ° o o . o o Boot and Dlagnostlc Controller Status Regi . o Line Frequency Clock and Status R‘aglstero KERNEL PROTECTION . . . TRAP AND INTERRUPT SERVICE PRIORITIES iv o Processor Status Word o ADDITIONAL CPU REGISTER DESCRIPTION s — [ . o o o . . 0 . o o . o o Hit/Miss Register o o . . o . (D o . . o KDJ11-BF Cache Operatlons KDJ11-BF Cache Organization . o . Cache Control Register er Memory System Error Regist 0 o ° o ¢ KDJ11-BF CACHE o o . . . o o6 0 Memory Relocation o . o o o . e 6 . Memory Management Register O. Memory Management Register 1. Memory Management Register 2. Memory Management Register 3, Physical Address Construction. o o ) ,eoooooeooevooooeo o 6 06 o o . . e OJO UL Page Descriptor Registers L] o [ . . . (] o UNIBUS Device Interrupt Requests PMI Data Transfer Address Cycle . PMI Data Transfer Protocol PMI Data In Cycles (DATI) and PMI Block Mode Data In Cycles ® o . L) . o [} . . . O "W N+ W+ L] [ L. ] ® e e . . . PMI Data Out Cycles . . MEMORY MANAGEMENT . ers Regist Page Address [ ® L] [} L] [ ] L} [} ® ] L J [ ] L] [ [ R R R GG EGES KRG RV RV [] [ ] oo~Joanese WK L DMA Requests . Pa . . PMI BUS DESCRIPTION PMI BUS Acquisition L] DN INTRODUCTION [ L) (] L} [ [ [ ] [ [ e o [ ] Ld [ ] [ L WWWWWWWWwWwioNNNDDNDN N e L] ) ® L) . [ . L] [} [ [ [ ] [ wwuwuwwwuuuwwwwwwww CHAPTER 3 FUNCTIONAL DESCRIPTION o W - Read MAPPING 4 [ . [ [ [ 3-71 3-76 * Configuration Register AND CONFIGURATION REGISTERS MODE ROM PROGRAMMING DESCRIPTIONS ¢ & ¢ & LJ © ® e LJ [] © © e e © 6 o8 © f} L] [J [] [] LJ » [] L] [J [4 Ld L] L L] o © g © o © & & 4 ©® o © ¢ ©® ¢ © S g € o © o &6 4 & ¢ © @& © 8 o © ¢ o e ® e ® & e & & © o [ TM o e L] . [ ) [] ] Ed L4 © e List o Command Setup Command DIAGNOSTIC ERROR MESSAGES BOOTSTRAP PROGRAMS Bootstrap [ o 8 o Setup * Command Command e Setup Setup [ Setup Command Command Command Command [J Command Setup Setup &6 Command Command ¢ Command Setup & Command Setup o Setup Setup DESC RIPTIONS & COMMAND Command Command ¢ MODE Setup o . . o . Command & Command e Map Test s Command List Command Setup Command Setup - 3-79 3-79 Command Setup N 3-76 3=-77 Register . Setup .1 DIAGNOSTIC COMMAND N o [ 3-65 3-67 Boot W ¢ Y [.] © BOOTSTRAP AND Help e 6 (4 Diagnostic Controller Status Diagnostic Data Register. Diagnostic DATI NPR Cycles Diagnostic DATO NPR Cycles @ g4 [ J F) ® ° .1 .2 .3 .4 s O ® © OO0 o6 Operations 3-62 3-63 3-63 3-64 DIAGNOSTIC & o ¢ o ® [) UNIBUS Mapping Registers. Optional UNIBUS Memory . H N M WO @ L ® L] Cache Memory 3-58 3-59 3-59 3-60 3-61 ® Enable/Disable. Write Operations ° DMA SETUP @ 3-56 3-57 3-58 L] [] ® > W+ DMA Cache DMA Cache UNIBUS ADUTLEWN I I N 3-54 Register. Y COUNTER INTRODUCTION I 3-54 . [ PMG DIALOG N Status UNIT KTJ11-BF CACHE OPERATIONS . KTJ11-B Cache Organization 4.1 I LINE o ¢« o o o WM Receiver CHAPTER C SERIAL Receiver Data Buffer. Transmitter Receiver Status Reglster. Transmitter Data Buffer Register. Break Response KERNEL/SUPERVISOR/USER MODE DESCRIPTIONS L] B BB WWWWRNNRNRNDNFO: BB WWWWWwWLwWwWwuwWwwuwwwuwuwwwww w CONSOLE [ o o . . . . o . . . . © [ L) [ L] [) L] [ [ © r L] [) © o o o e ® L] ° L] L) 4.6.5 4.6.6 ROM Data Organization 4.6.7 o . Jll1 MICRO ODT . 4.7 / (ASCII 057) Slash . 4,7.1 <CR> (ASCII 015) Carr iage Return 4.7.2 KLF> (ASCII 012) Line Feed 4.7.3 d nat $ (ASCII 122) Internal Register Desig 4,7.4 o -~ G (ASCII 107) GO 4.7.5 P (ASCII 120) Proceed 4,7.6 Binary Dump ~S (Control-Shift-S) 4,7.7 [ Single ROM Programs . Multiple ROM Programs . Program Headers . © ° ROM Addresses ROM Formats . © o . . o L] Boot ROM Installation . 000000.‘.0.0.00. BOOT ROM FACILITY (M9312 compatible) . [ L) o . Format General Rules For EEP ROM User Bocts . 4.6 4.6.1 4.6.2 4.6.3 4.6.4 EEPROM o [] wN [ v U =N e 4. CHAPTER 5 SYSTEM MAINTENANCE INTRODUCTION . DIAGNOSTIC TYPES .1 o ° . o CONSOLE TERMINAL ERROR MESSAGE FORMAT . . Console Error Message Description . . . . . o o Unexpected Trap and MMU Error Code Descriptions e2 Boot Program Error Codes/Messages SYSTEM TROUBLESHOOTING AIDS .3 . s © [ ] [] [ [ KTJ11-B UBA Module Mimimum Load Module [ . . © FIELD REPLACEABLE UNITS L ] * o L. MSV11-JB/JC Memory Module L] KDJ11-BF CPU Module . o o . . . o Front Panel MDM Module . L] LJ L ® © MODULE REPLACEMENT/REMOVAL PROCEDURES L] General Module Removal/Replacement CPU Module Removal/Replacement POWER SUPPLY REMOVAL/REPLACEMENT L] * [ 2 L] L] Cabinet Power Supply Removal/Replacement L[] L © ® [} -~ O O O [] © L] . Box Power Supply Removal/Replacement CABINET BLOWER REMOVAL/REPLACEMENT BOX FAN REMOVAL/REPLACEMENT FRONT PANEL REMOVAL/REPLACEMENT [] ® 5.12 [ ° [] N [ [) Cabinet Front Panel Removal/Replacement Box Front Panel Removal/Replacement . . . CIRCUIT BREAKER REMOVAL/REPLACEMENT ment place al/Re Remov er Cabinet Circuit Break 1.1 Box Circuit Breaker Removal/Replacement . 5.11.2 [ CABINET POWER CONTROLLER REMOVAL/REPLACEMENT SLU INTERFACE ASSEMBLY REMOVAL/REPLACEMENT 5.13 Cabinet SLU Assembly Removal/Replacement 5.13.1 Box SLU Assembly Removal/Replacement 5.13.2 . CPU BACKPLANE REMOVAL/REPLACEMENT . 5.14 .5.14.1 Cabinet Backplane Removal \Y i © . . © © L4 ® [) © e L] e . 5.14.2 Box Backplane Removal 5.15 CABINET PERIPHERAL ACCESS . . . . . . . . . . . APPENDIX A CPU INSTRUCTION TIMING o . . . . APPENDIX B PDP-11/84 and HARDWARE/SOFTWARE DIFFER@NCES APPENDIX C PINOUT BACKPLANE TEST CONNECTORS J18 AND J19 . . . . . . . . APPENDIX F V7. - V6.0 ROM CODE DIFFERENCES . . . . . . APPENDIX D BACKPLANE PIN ASSIGNMENTS . APPENDIX E SYSTEM INTERCONNECT DIAGRAM APPENDIX G MULTI BOOT CONTROL TRANSFER . APPENDIX H BOOT AND CONFIGURATION REGISTER MODIFICATION APPENDIX I FLOATING POINT INSTRUCTION TIMING . APPENDIX J SET-UP PARAMETER WORKSHEETS vii . . . . . . CHAPTER 1 SYSTEM INTRODUCTION 1.1 INTRODUCTION The A Series PDP-11/84 (PDP-11/84-A) is a high performance point microprocessor with floating uctio computer containing a J11 ssor n instr 1 executes the PDP-1 accelerator (FPA). The proce n'’s ratio Corpo ment on Digital Equip set. The system operates t memory addressing capabiltiy. 18-bit UNIBUS with a 22-bi The system is available in two configurations: a. ion packaged in a PDP-11X84: a kernel system configurat of the top portion The 42-inch cabinet. (See Figure 1-1.) lling insta for cabinet provides a 10.5-inch enclosure peripherals. b. m configuration PDP-11/84: an expansion box kernel syste packaged in a 10.5-inch rackmountable enclosure. (See Figure 1-2.) NOTE The "PDP-11/84" system this manual designation as used 1in implies that the context applies to both cabinet and box configurations. — Tt 11 X11% »i INTRODUCTION CY SYSTEM FIGURE FIGURE 1-1 1-2 PDP-11X84 CABINET PRODUCT PDP-11/84 1-2 BOX PRODUCT SYSTEM -INTRODUCTION 1.2 SYSTEM COMPONENTS AND VARIATIONS cabinet and box product is shown The system block diagramnelforsysthe ' tem consists of: in Figure 1-3. The ker 1. A KDJ11-BF processor module (CPU) -JC 2 MB V11l 2. An MsV11-JB 1 Mb ECC), memory module(s), or an MsSv ECC memory module(s A KTJ11-B UNIBUS Adapter Module (UBA) , A Monitor and Distribution module (MDM), and 5. One or more Minimum Load Modules (MLM). 3. 4., vate Memory through the high-speedbit Pri The modules communicate usin data lines. g 22-bit address/l16Interconnect (PMI) bus rd munication standa g the EIA RS-232 com A console port supeportin B 11KDJ the into sol terminal to be connected enables a con processor module. and the UNIBUS. n The erfaces to the PMI bus The KTJ11-B UBA intts the ications betwee ress/data commun por all addUNI UBA module sup In . ns) tio BUS periherals the(opCPU all processor memoryUBAand the of end serves as a terminator for addition, the UNIBUS. CONSOLE TERMINAL i | KDJ11-BF cPu. B | | | | -8 ]|1 | xTat UNIBUS ADAPTER MEMORY |, | (UBA) msvit-s : . : o oMl BUS @ UNIBUS MEMORY | | | <::>PER|PHERAL l UNIBUS Lfop'q 1/84-P SYSTEM CABINET AND 80X FIGURE 1-3 PDP-11/84 SIMPLIFIED BLOCK DIAGRAM 1-3 | _] SYSTEM INTRODUCTION The basic cabinet and box components are shown in Figures 1-4 and 1-5. The following subsections briefly describe the basic system components. | — FRONT BEZELIBLANK) POWER CONTROLLER | - CARD CAGE COVER I\ I| POWER SUPPLY CIRCUIT BREAKER CARD CAGE FIGURE 1-5 BASIC BOX HARDWARE 1-4 COMPONENTS SYSTEM INTRODUCTION Table 1-1 specifies the system variations. TABLE 1-1 A-SERIES PRODUCT VARIATIONS 11/84-AA KDJ11-BF, MSV1l1l-JB 1 MB 11/84-AB KDJ11-BF, MSV11l-JB 1 MB 11/84-BA Same as 11/84-AA except MSV11-JC 2 MB 11/84-BB Sahe as 11/84-AB except MSV11-JC 2 MB 11X84-AA KDJ11-BF, MVS1ll1l-JB (JD) 11X84-AB KDJ11-BF, MSV11-JB (JD) 1 MB 11X84-BA Same as 11X84-AA except MSV11-JC 2 MB 11X84-BB Same as 11X84-AB except MSV11-JC 2 M 10.5-inch Box, 10.5-inch Box, 120 Vac 240 Vac 40-inch cabinet, 40-inch cabinet, 120 Vac 1 MB 240 Vac 1.2.1 KDJ11-BF Processor Module The KDJ11-BF (M8190-AE) is a quad-height CPU complete functionality of a PDP-11 processor. module having the The KTJ11-B UNIBUS Adapter Module allows the CPU to interface with Digital's UNIBUS. The module features a: J1ll microprocessor, FPA, 22-bit memory management, 8KB cache memory, programmable line frequency clock, console serial line unit, an alterable configuration EEPROM, and boot and diagnostic ROMs. In addition the KDJ11-BF has diagnostic information during six red power-up LED's for displaying and bootstrapping. A single green LED indicates dc power to the module. INTRODUCTION SYSTEM 1.2.2 KTJ11-B UNIBUS Adapter Module The KTJ11-B UNIBUS Adapter (M8191) is a hex-height module that interfaces with the KDJ11-BF processor and memory through the PMI. The module contains: the PMI adapter 1logic¢, UNIBUS mapping, 1.2.3 and four M9312 MSV11-JB/JC compatible Memory uses 256K a capacity 1. MSV11l-JB (M8637-B) with 2, MSV11-JC (M8637-C) with a The modules byte/word, the PMI sockets. dynamic RAMs, and is versions: two in ROM Module The quad-height memory module available boot 1MB 2MB capacity provide error correction logic, and supports write double-word read, and block mode read operations over bus. A uncorrectable red error; LED 1indicates the occurance of an a green LED indicates the presence of 5 Vdc power., 1.2.4 UNIBUS The UNIBUS Terminator Terminator characteristic end of 1.2.5 the UNIBUS Monitor (M9302) impedance of and 1is 120 a resistive ohms. and provides the Distribution Module network with a The module terminates one SACK turnaround feature. The Monitor and Distribution Module (MDM) is a guad-height module (M7677) and includes: power supply voltage indicators, voltage test points, fan/blower rotation monitor and nonprocessor dgrant (NPG) jumper l1.2.6 Mimimum Load The selection double-height load module 1.2.7 for the includes Power switches. Module Minimum Load Modules (MLM) =15 Vdc and +5 VBB power two power supply LEDs, provide one for each minimum Each regulator. Supply Dc System power is provided by two power supplies: H7202-KB. a supply regulators. H7202-KA and SYSTEM INTRODUCTION The H7202-KA power supply provides: 60 A at +5 Vdc, 2A at +15 or 240 Vac It also provides 3 A at +12 Vdc for vde, +3 A at =15 Vdc. The fans/blower and up to 15 A at +5 Vdc to the PMI memory. A tion. protec ating overhe and ltage power supply provides overvo either 120 Vac The H7202-KB power supply provides power for additional switch allows the user to primary power operation. expansion) operated at backplanes. overheating power: 120 or 240 protection. select Similar Vac, the to H7202-KA, can be overvoltage features and it (i.e., and The power supply provides the following 32 A at +5 Vdc, 2 A at +15 Vdc, and 3 A at -15 Vdc 1.2.8 Console Serial Line Board an EIA des: The console serial 1line board (54-16058-) provi for communication with the KDJ11-B, a port I/0 RS232-C console I/0 port ten-position rotary switch for selectingionthe switch for selecting a baud rate, and a two-position posit restart mode. 1.2.9 Backplane Assembly 0650-01) is a As shown in Figure 1-6, the backplane assemblyMDM(70-2 through 4 are Module slots 13-module slot backplane. through 11 support hexdedicated to the system kernel. Slots 5 oller s (SPCs). Slot 12 Contr or quad-height Small Peripheral Backplane NPG es. supports only guad-height UNIBUS option12modul d in a DIP mente are imple jumper functions, for slots 5 through switch located on the MDM module. ROWS A c L] f E 0 (MDM) M7677 MOM 1 (XDJ11-8) MB190 2 (MSV11- J) MB837 M7858 3 1 4 IKTJ11-8) MB191 HEX OR QUAD UNIBUS OPTION Pt SLOTS 6 ? HEX OR QUAD UNIBUS OPTION KEX OR QUAD UNIBUS OFTION P2 L] . MEX OR QUAD UNIBUS OPTION ] PRIOR 5 ¢ | MoDiReD UNIBUS 0 "‘&‘:"“:‘s" 1 ‘:}"g"‘l‘fs" 12 | UNiBUS FIGURE MEX OR QUAD UNIBUS OPTION HEX OR GUAD UNIBUS OFTION MEX OR QUAD UNISUS OPTION OUAD UNISUS OFTION l e MTS58 1-6 BACKPLANE ASSEMBLY 1-7 »” re SYSTEM INTRODUCTION 1.2.10 Front The front Panel panel Assembly assembly (70-21888-01) includes switches and indicators for status display and operator control of the system. A keyed rotory switch selects one of four power states, and a toggle switch selects system halt or restart modes. 1In addition, three LEDs display the system status, and a two-digit octal LED displays 1.2.11 The diagnostic Cabinet cabinet error Power status. Controller contains either an 877-D power controller for 120 Vac operation, or an 877-F controller for 240 Vac operation. The unit controls all AC power entering the cabinet. It provides primary power for all the power supplies, and the two AC outlets located 1.2.12 The in the top Cabinet cabinet is peripheral Blower Unit cooled by a enclosure. blower wunit (12-22001-01) mounted below the <card cage. Air is drawn through the top front of cabinet, forced through the card cage by a plenum, through power supply, and out through the rear of the cabinet. The blower supply must be is capable of cooling and an optional 1installed during the optional expansion UNIBUS ‘backplane. The operation to insure kit the the power card cage cover proper module cooling. 1.2.13 The box bezel. directs supply, 1.,2.14 The Box Cooling product Fans is cooled by three fans The fans draw air through the air horizontally through and out the rear of the box. Additional Expansion system supports the and the front the front bezel and a the card cage and plenum power Memory mounted behind Options following options: Battery Back Up Unit a. H7231-E, Cabinet b. H7231-F, Box Battery Back Up Unit c. DD1l11-CK, 4-slot Backplane (PDP-11X84) d. DD11-DK, 9-slot Backplane (PDP-11X84) 1-8 (requires (requires M7677-YA) M7677-YA) ON EM ' INTRODUCTI SYST 1.3 SYSTEM SPECIFICATIONS The following tables list the PDP-11/84 system specifications. Table 1-2 through Table 1-4 1list the cabinet specifications; specifications. Table 1-5 through Table 1-7 1list the box ned in the contai are ns icatio Supported peripheral device specif user's guide associated with that device. TABLE 1-2 CABINET ENVIRONMEMTAL SPECIFICATIONS Temperature: Operating 10 deg C to 40 deg C Nonoperating -40 deg C (storage) Humidity: Operating (50 deg F to 104 deg F) to 66 deg C (-40 deg F to 151 deg F) 108 to 90% with max wet bulb temp. 28 deg C (82 deg F) and a min dew point 2 deg C (36 deg F) non condensing. Vibration Operating Nonoperating (packed for shipment) 5 to 22 Hz:0.01 in DA; 22 to 500 Hz 0.25 Gpk. Sweep rate of 1.0 octave/min. All three axis. Vertical Axis Random Vibration: 0.687 Grms overall from 10 to 200 Hz; duration: 1 hr each axis. Altitude: ft) Operating 0 to 2.4 km (8000 Nonoperating 9.1 km Maximum operating with altitude maximum operating (40 deg C) should be reduced 1.8 deg C/ 1000m (30000 ft) (1 deg F/ 1000 ft) above sea level Shock: Operating Nonoperating (packaged for shipment) 10 Gpk for 10 ms wave, (+3 ms), vertical axis only 1/2 sine Flat drop from a 6 in height, three drops total (vertical direction only) SYSTEM INTRODUCTION TABLE . G TR = G WD CED Overall N0 MO D 4ED SN W 1-3 SR WD CABINET CED GEN TR AP S dimensions GED GED CED ED D MECHANICAL WD G GRS IR GED GED D G TS T WD SPECIFICATIONS D D D R D CH) D D TED WD O G50 T 105.7 cm high X 53.9 cm wide X 76.2 cm long (41.64 in high X 21.25 in wide X 30 Unpacked 150.5 kg (331 1b) Packed 182.2 kg (401 1b) in long) Weight: TABLE 1-4 CABINET ELECTRICAL Characteristics D D D e D G - D D . R IO CED T D SPECIFICATIONS Description R S D D 120 Vac operation: Line voltage D D D G e 93 CED S D D . - 132 D D 47.5-63 Hz Current 13.5 (rms) Power factor Greater and low Start Up Current 100 A, Inrush current 160 A D T O D e T D CH) D D D CED D G single-phase, (120 Vrms Frequency A GRS Vrms, and ground (ac) D max at 2880 BTU 240 ' Vac Line 120 max at 120 Vac, V-A max* operation: 186 - 264 Vrms, single-phase, two-wire and ground (240 Vrms nominal) 47.5-63 Current 6.7 A (ac) Hz (rms) load usec duration (peak) Frequency — Vac 3519 voltage " two-wire than 0.60 at full output input voltage (93) 0.16 5 nominal) duration Power G max at 240 Vac 0.16 usec SYSTEM (Cont) TABLE 1-4 —— D TS A A INTRODUCTION S W VS WP N GED VAR GED AR M W SR W G GNS D G e e Power factor Greater than 0.60 at full output lcad Start Up 50 Inrush and Current low input voltage A, 0.16 160 A current usec (peak) (186 Vac) duration max at 240 Vac, 0.16 duration Power 2880 BTU 3519 Noise Transient: (both line voltages) 1KV peak High-energy * spike containing not than 0.2 W of energy per spike transients Conducted V-A max* Cw-10 noise Including mass storage TABLE KHz to 30 MHz, 3 Vrms devices 1-5 BOX ENVIRONMEMTAL Characteristic SPECIFICATIONS Description Temperature: Operating 5 deg C to 50 deg C (41 deg F to 122 deg F) Nonoperating -40 (storage) deg C to 66 deg C (-40 deg F to 151 deg F) Humidity: Operating more 10% to 95% with max wet bulb temp. 32 deg C (82 deg F) and a min dew point 2 deg C (36 deg F) non condensing. usec SYSTEM TABLE INTRODUCTION 1-5 (Cont) Vibration: Operating 5 to 30 Hz: 0.0l in DA; 30 to 500 Hz: 0.5 Gpk. Sweep rate of 1.0 octave/min. All three axes. Nonoperating (packed for shipment) Vertical Axis Random Vibration: Grms overall from 10 to 200 Hz; duration: 1 hr each. 0.687 Altitude: Operating 0 Nonoperating 9.1 to 2.4 km km (8000 (30000 ft) ft) Shock: Operating Nonoperating 10 Gpk for 10 ms (+3 ms), wave, vertical axis only Flat drops Maximum operating with altitude drop from a total 6 in 1/2 height, sine three (vertical direction only) Maximum operating (40 deg C) be reduced 1.8 deg C/ 1000m should Mechanical: Overall dimensions 47 cm wide X 67.5 cm long X 26 cm high (19 in wide X 27 long X 10.44 in high) Weight: Unpacked 42.75 kg Packed 59 (130 kg (95 1b) 1b) in SYSTEM TABLE 120 Vac Line 1-7 BOX ELECTRICAL SPECIFICATIONS operation: voltage 90 - and 47.5 Current 8.0 Start (ac) factor Up Inrush Current current 132 Vrms, ground Frequency Power INTRODUCTION A 63 (120 single-phase, Vrms Hz (rms) max at 120 Greater than 0.60 and Vac nominal 50 120 A, two-wire nominal) 0.16 usec at Vac full output input voltage load duration 80 A (peak) max at 120 Vac, 0.16 usec duration Power 650 BTU 2218 240 Vac Line W MAX operation: voltage 180 - 264 two-wire Vrms, and single-phase, ground (240 Vrms nominal) Freguency 47.5 Current 5.0 (ac) Power factof Start Inrush Up Current current - A 63 (rms) max at Greater than 0.60 -and Vac 240 A, 0.16 usec 80 A (peak) max 650w MAX 240 at nominal 50 duration Power Hz Vac full ocutput input voltage load duration at 240 Vac, 0.16 usec SYSTEM INTRODUCTION TABLE 1-7 (Cont) Noise transient: (both line woltages) High-energy 1 KV peak spike transients not more than 0.2 W of energy per Conducted noise 1.4 RELATED Table 1-8 containing spike CW-10KHz to 30 MHz, 3Vrms DOCUMENTS lists the related TABLE PDP-11/84 related documents. 1-8 PDP-11/84 RELATED DOCUMENTS PDPl]l Bus Handbook EB-17525-20 PDP-11/84-A Installation and Technical Manual EK-1184A-TM PDP-11/84 Site Preparation, Installation Unpacking, and EK-PDP84-1IN Guide PDP-11/84-A System Field Maintenance Print Set Cabinet: Box: MP-02199 MP-02198 KDJli—B User Guide EK~-KDJ1B-UG MSV11-J EK-MSV1J-UG User Guide Chipkit Handbook EJ-01387-92 DCJ11l EK-DCJ11-UG User Guide EB-26085-41/85 Microsystem Handbook 1-14 SYSTEM Printed copies Digital Equipment 444 Whitney Northboro, ATTN: of the above listed documents may be Corporation Street Massachusetts 01532 Printing énd Circulation Services (NR2/M15) Customer Services Section INTRODUCTION ordered from: CHAPTER SITE PREPARATION 2 AND INSTALLATION IMPORTANT acceptance and installation system Following testing, refer to Appendix J and fill in the Parameter worksheets. Set-up The completed worksheets are to remain with the system to allow duplicating the original set-up parameter selections on KDJ11l-BF replacement a CPU module. 2.1 SITE PREPARATION In order to use a PDP-11/84 system; the space requirement, environmental operating limits and the electrical power, available at the site, should be part of the site preparation. The specifications are listed in the following subsections. 2.1.1 Cabinet Physical Site Space Width Height Depth Weight Environmental Preparation Specifications Requirements: 53.25 cm (21.2 in) 105 cm (41.6 in) 75 cm (31.5 in) 150 kg (331 1b) Operating Limits: a. Operating Temperature: b. Relative Humidity: 10% to 90% with max wet bulb deg C (82 deg F) and min dew point 2 deg C non-condensing temp. (36 deg F) Storage Temperature: deg to 151 F) c. 104 deg deg F) 10 deg C to 40 deg C ( 50 deg =-40 deg 2-1 C to 66 deg C (-40 F to 28 SITE allowable maximum with Altitude: Derating Temperature reduced by 1.8 deg C per be should temperature operating (1 deg every 1000 m the system is operated above sea level. d. F AC INSTALLATION PREPARATION AND per 1000 Electrical Depending on ft) Power Reguirement: the site line voltage, two power controllers are available: 877-D Power Controller: Voltage (93 to 120 132 47.5 Frequency = 877-F Pdwer Controller: Vac Nominal Vrms) 63 HZ to Voltage - 240 Vac Nominal (186 264 Vrms) to Frequency - 47.5 to 63 HZ NOTE A dedicated circuit from the power source is recommended for each system. This circuit should provide an isolated ground path between the receptacle an the power source. The power system should be stable and free from electrical noise. Do not connect any conditioners, office the same The user must receptacle(s). NEMA AC circuit with supply Electrical the for b. for 240V service: 120V service: The power cord provided either a NEMA L5-30P, or 2.1.2 Box Site following Receptacle a. eqgquipment such as air copiers, or coffee pots, on the system. AC electrical and Required: NEMA L5-30R NEMA 6-15R (rated (rated @ @ 30 A). 15 A). is approximately 14 a 6-15P plug attached. Preparation power Specifications feet 1long, with SITE PREPARATION AND to (41 deg F to INSTALLATION Physi&al Space Reguirements: Width Height Depth 47.5 cm (19.0 in) 26 cm (10.4 in) 67.5 cm (27.0 in) 43 kg (95 1b) Weight Environméntal a. Operating Limits: Operating Temperature: 5 deg F) b. Relative Humidity: deg C (90 deg non condensing Ce d. deg and Derating 1000 (90 tg - Frequency m the Power 120 132 with temperature F pqr 1000 ft) Voltage deg dew C to point 66 2 deg C deg C (=40 : operating Electrical minimum =40 F) Tem&erature every 50 deg C 10% to 95% with max wet bulb Stoqage Temperature: 151 AC F) deg C system Altitude: maximum should reduced is above be operated by sea 1.8 temp. (36 deg deg F level. Nominal Vrms) - 47.5 to Voltage - 240 Vac (180 to 264 Vrms) 63 HZ Nominal Frequéncy - 47.5 to 63 HZ NEMA AC EHlectrical % Receptacle - Requirement: NOTE fihen switching the power supply input voltage 240 Vac operation, the power cord must from the 120 Vac (shipped configuration) changed. 1. IZQ Vac service: NEMA 5-15R (rated @ 15A) to also 32 F) to allowable deg Requirement: Vac 122 the be C per (1 deg SITE PRERARATION AND INSTALLATION 2. 240 Vac service: NEMA 6-15R (rated @ 15A) a NEMA 75 inches long, with either grou The power cord provided P isplug nd is attached. If an isolated 5-15 P, or a 6-15 ptacle numbers with "IG". provided, prefix NEMA rece 2.2 SHIPPING SPECIFICATIONS ucts (BOX and SHIPPED information. Prodaine This section provides AS rs as shown cont d boar CABINET are shipped in reinforced card in Figures 2-1 and 2-2. Box Container: Width: Height: Length: Weight: 57.5 cm (23 in) 50.8 cm (20 in) 77.5 cm (31 in) 56 kg (125 1b) FIGURE 2-1 BOX SHIPPING CONTAINER Cabinet Container: Width: Heights: Length Weight 86 cm (34 in) 139 cm (55 in) 107 cm (42 in) 182 kg (401 1b) B FIGURE 2-2 CABINET SHIPPING CONTAINER 2-4 SITE 2.3 UNPACKING PRERARATION AND INSTALLATION INSTRUCTIONS Read the following steps prior to unpacking the <cabinet or box product. 2.3.1 Cabinet Unpacking The wunpacking container. To instructions are 1located open the shipping container inside complete the the shipping following steps. IMPORTANT Read warning labels remove plastic l. Cut 2. Remove top 3. Follow directions 2.3.2 The Box and on outside of container. strapping. cover. packed inside Unpacking wunpacking container. instructions To open the are shipping 1located container inside the shipping complete the following steps. IMPORTANT Read warning labels remove plastic l. Cut 2. Carefully 3. Open 4. Follow 2.4 SYSTEM and cut folded sealing top of directions on outside of container strapping. tape. container. packed inside. INSTALLATION To install a cabinet or box read the appropriate subsection and follow the procedures. The procedures do not cover optional device installation. Option installation is <covered 1in the manual supplied with the device. 2-5 SITE PRERARATION AND INSTALLATION 2.4.1 Cabinet Mechanical Installation To install the cabinet complete the following procedure. 1. Complete the 1inside the 2. After rolling the cabinet down the ramps, position cabinet on a level surface in the operational area. the 3. Reverse step 6 of the unpacking instructions and lower the unpacking shipping container. instructions located four leveling feet. Each foot is lowered until the wheel near the foot - is raised approximately 1/8- to 1/4-inch above the floor surface. the cabinet. 4, Level 5. Raise each top nut and tighten it against the cabinet frame while holding the bottom leveling foot hex nut. This completes the mechanical installation procedure. 2.4.2 Box Mechanical Installation To install a box in an ANSI/EIA standard 19-inch the rack, complete following procedure. 1. Locate and check hardware in the box shipping container. 2. Mount the chassis slide bracket by aligning the rack frame Secure the bracket to the holes with the slide bracket. (no star frame with a nut bracket and three no.l0 screws 2-3.) Figure (See Tighten the screws. washers). 3. Mount the three other chassis slide brackets frame as in step 1. to the rack SITE PRERARATION AND INSTALLATION NUT PLATE A Fs/\-;\ h _. AN “ r";\ FIGURE Mount the 2-3 MOUNTING the restraint bracket CHASSIS holes SLIDE bracket with AND RESTRAINT to the rack frame rack frame holes, chassis slide bracket. the holes above the mounted tighten three no.l0 screws bracket. into the nut bars BRACKET by aligning starting Thread to secure two and the Align the right chassis slide (front) clearance hole with the center hole of the mounted chassis slide bracket. Thread a no.10 screws (with star washer) to secure the chassis slide to the bracket. Do not tighten. Align the mounted washers) two rear bracket. holes chassis securing the slide bracket. Tighten the chassis front slide on Thread screw assembly the chassis two chassis Tighten no.l0 slide the two securing to bracket. Secure the slide screws rear to screws. the front front mounted with star the rear end of the chassis slide the left chassis slide assembly to the 1left and rear mounted chassis slide brackets by repeating 4 through 6. 2-7 the (with front steps SITE PRERARATION AND INSTALLATION placing the 9. Extend the stabilizer bar (if present) before 10. Lower forward: Fully extend both chassis slide assemblies the ing align , blies assem the box assembly onto the slide the box rail. box on the slide assemblies. FIGURE 2-4 11. / AR AR L Q...!QQ...‘......‘ ) o, chassis slide slot with the tab on SECURING BOX INTO CHASSIS SLIDE ASSEMBLY slides by aligning the front Secure the box to the chassis aded box hole. Thread a no.8 clearance hole with the thre screw and tighten. 12. the rear chassis slide Slide the box rearward aligning box bole. Thread a no.8 aded thre clearance hole with the ' screw and tighten. 13. e it to the rack frame Slide the box to the rear and secur screws to the restraint by threading two, no.8 phillips l. Tighten the screws. bracket located on the box rear pane This completes the installation of standard 19-inch rack. 2.4.3 Console Serial Line Hookup the box 1in an ANSI/EIA Install the console serial line as shown in Figure 2-=5. 2-8 SITE PRERARATION AND INSTALLATION TERMINAL - CONSOLE CONSOLE CABLE \\ ] TERMINAL ' r oy \; N Sty CONNECTOR — LU S V., L\ mNNEflOR+—@L — ] J-I - a /| FIGURE 2.4.4 Cabinet 2-5 SERIAL Switches Settings switch settings The following are LINE HOOKUP the required for -<cabinet product. NOTE Switch settings must be verified before power applied to the 1. Set the Forced Dialogue switch to OFF (0). 2. Set 3. Set the Baud Rate switch to match the (see Figure 2-6.) console terminal. 4, Assure the baud that controller rate the is system. of AC the console terminal. power outlet baud matches AC input power requirement. rate the of the power Model 877-D is for 120 Vac operation and 877-F is for 240 Vac operation. The on a label located on the power printed is number model controller. SITE PRERARATION AND INSTALLATION WA FTTITIRGTIRY TR T E TR TR P ; | SALO -ex| 14D = 200 | ~ 4o s = 2400 7 = 300 s = 1200 : - \ { | b BULK HEAQ~ DIL00 [S4 ) I l I 1 { | L] L) L ©° L] © L] . FIGURE 2-6 s CONNECTOR | O O T O g // I { | SCREW (10} sAUO CABINET REARVIEW Turn the power controller circuit breaker to position. See Figure the off (0) 2-6. Set the front panel keylock switch to OFF. (See Figure supply circuit 2-110) power both turn door, front Open the (See Figure 2-7.) breakers to OFF (0). This completes the cabinet switch set up. MODULE " COVER v POWER SUPPLY CIRCUIT BREAKER J‘\ \ifl J SLOWER FIGURE 2-7 CABINET FRONTVIEW 2-10 SITE 2.4.5 INSTALLATION Settings Switch Box PRERARATION AND The following switch settings are reqguired for the box product. 1. Slide the box forward to the mechanical stop. 2. Remove the four top cover screws and lift 3. Set the power supply input AC voltage to match front, supply power the (See Figure cover off. 2-8.) (See voltage. the 2-8.) Figure on located switch, AC power outlet the o o ®? oA = FIGURE 2-8 BOX POWER 4. Reinstall the top cover. 5. Set the Forced 6. Set the baud rate of Set the Baud Rate switch 7. console Dialogue terminal. the See SUPPLY INPUT switch to console to OFF SWITCH (0). terminal. match Figure SELECT 2-9. the baud rate for the N INSTALLATION AND SITE PREPARATIO 4 '\ Yd BULKNEAD """ L ¢ FIGURE 8. Set the 9. Turn the This completes front panel 2-9 keylock circuit breaker the box BOX switch REARVIEW switch to OFF set to OFF. (0). See up. Qecuir BAEAKER FIGURE 2-10 BOX FRONT VIEW 2-12 Figure 2-10. SITE PREPARATION AND INSTALLATION 2.4.6 Cabinet Power Hookup Complete the following procedure. CAUTION Assure that the wall outlet voltage line voltage selected for operation. switch Assure that all system specified in subsection 2.4.4. matches the are as settings NOTE If the power controller turned 1is off (select switch - OFF) all power transfer is inhibited. . Set the power controller select switch to REMOTE. 2 . Plug the AC power cord into the AC power outlet. 3. Turn the power controller circuit breaker to ON 4, Open the front door and turn the main power supply 5. If an expansion backplane is installed, set Figure (1). 2-6. breaker (right breaker) to ON (l). power supply circuit breaker See Figure 2-7. (left breaker) the to ON. Box Power Hookup Complete the following procedure. CAUTION Assure that the wall outlet voltage line voltage selected for operation. Assure that all system switch specified in subsection 2.4.5. matches the are as settings circuit expansion This completes the cabinet AC electrical hook up. 2.4.7 See SITE PREPARATION AND INSTALLATION l. Plug the female end of the AC power receptacle mounted on the rear of the box. 2. Plug the male end of the AC power cord cord 1into the See Figure into the AC 2-9. power outlet. 3. Remove the restraint 4. Slide the Slide the box back box out circuit breaker 5. bracket at the approximately rear of the box. See 6 inches Figure and turn into the rack and replace the restraint electrical hook up. to ON (l). 2-10. the bracket This 2.5 completes SYSTEM The CONTROLS following indicators. conditions. Chapter 2.5.1 The the 5, Front box AND INDICATORS subsections describe the system controls and The functions described are for normal operating If a problem occurs during normal operation refer to System Maintenance. Panel front panel consists of a keylock power switch, HALT Switch, a Start-up Test indicators. Figure 2-11 shows indicators. 4 LED the display, and front panel RESTART/RUN/ POWER and controls ) ofe dilgfilt]a]t enane SECURE PDP-11/84 \_ sTanT.up TEST " sTANDBY CDocon - AESTART D AUN ' = RUN CO sareny HaLT L FIGURE 2-11 FRONT PANEL CONTROLS 2-14 AND INDICATORS RUN and INSTALLATION SITE PRERARATION AND of four power states. ONE select each of the power switch selections. to used switch rotary position four The Keylock Switch is a Table 2-1 lists and describes TABLE 2-1 KEYLOCK POSITION DESCRIPTIONS Power supplies are turned off. DC power to the logic and blower/fan assembly is off. However, AC power into OFF is present. the power supplies ENABLE The ON position. Power supply voltages are present to SECURE Same as the ENABLE position except that the console terminal Halt-On-Break feature and the HALT/RUN/RESTART the logic and blower/fan assembly. switch are disabled. STANDBY Power is supplied to the PMI memory, blower/fans but other voltages are turned off. —-——-—————-——-———— _‘—)-m—-——“-‘——fl-——_-----——-————--————-———-_--— The HALT/RUN/RESTART switch functions are enabled only when the keylock switch 1is in the ENABLED position. Table 2-2 lists the switch positions and their functions. TABLE 2-2 HALT/RUN/RESTART SWITCH FUNCTION POSITION The CPU program is stopped and the incremented content of the program counter is displayed on the console HALT terminal. The CPU enters J1l Micro-ODT. RUN Entering RUN from RESTART enables CPU operations to run. Entering from HALT causes the processor to remain in micro-ODT awaiting a command from the console terminal. RESTART This momentary switch position initiates processor execution of bootstrap program instructions located in the boot ROM according to the set up configuration in the EEPROM., For modifying EEPROM configurations see the EEPROM section - w p aED s was @S (D he =) CED WD CED GER CER P GER AID A D D WD G YED M S in this chapter. R CER W TED I O CHD M D ED B ATR O CED TEO - D D GNC TN GED GED GED CE D G SRR R GUD D D R D CEP = S - SITE PRERARATION AND INSTALLATION Three LEDs are located on the front panel. the LED status and functions. _ TABLE - — . I FRONT PANEL 2-3 W G U NS S W G CE GEE WD SN W M Table describes 2-3 INDICATORS . 5 D CED CED WD D D A O CHD D S OO0 TN O D ) G e wm e = m-—————— -————————-——————-——————--o--&—-am—'—'—na—‘n—- ON RUN OFF The J1l1 processor is fetching and executing instructions. This is the normal condition. The processor is halted or waiting for an interrupt. When the processor is in Micro-ODT the RUN indicator blinks for each console keystroke. The RUN LED also turns off during extended DMA activity. DC ON ON OFF BATTERY ON DC power is available to the logic and all voltages are within specified levels. DC voltages are not available to the logic or voltages are present but not within tolerances. Battery is present and charged to 80% or great- er Slow Blink (1 Hz) Fast Blink (10 Hz) OFF ' capacity. Battery is at less than 80% capacity and is charging. The ac power has failed; the battery is discharging, but the memory content remains valid. Battery is either fully discharged or not pre- sent in the system. Memory content will not be preserved —— A if ac power fails. = SN P R S D SED THI GHE GED IS G TS D G GED WD W The two-digit Start-up Test LED error diagnostic System Maintenance. codes. W G M P D D D MR GO WD SER ED TED WD IO AT KNP D D D W SN G TR S e sEm o display provides the start-up The codes are described in Chapter 5, SITE PRERARATION AND INSTALLATION 2.5.2 Console Serial Line Distribution Board to mounted 1is The console serial line distribution board (SLU) a includes SLU The products. both of panel back inside the select switch serial communications port connector, a baud rate position for mounting SLU and a Forced Dialogue mode switch. The 2-9. and 2-12 Figures in shown are the cabinet and- box 8AUD /-Wfl‘ S \\ BULK HEAQ~ e SCAEW (104 ' { || { ) FORCED DIALOG PITCH sLu CONMECTOR t Lt FIGURE 2-12 CABINET BACKPANEL VIEW 2.5.3 Serial Communications Port Figures A serial line connector is located on the back panel. It or. 2-12 and 2-9 show the physical location of the connect duplex full compliant RS449 RS232-C, EIA provides an communications 1link between the CPU and console terminal. Note that 20 ma applications are not supported. 2.5.4 Baud Rate Select Switch ng The baud rate switch enables the operator to set the operati rate baud The system. and l termina console the between rate baud is set for one of eight possible baud rates. Table 2-4 lists the switch positions and their corresponding baud rates. 2-17 SITE INSTALLATION PRERARATION AND TABLE 2-4 BAUD RATE SWITCH DESCRIPTIONS - SWITCH POSITION BAUD RATE 0 38400 1 19200 2 9600 3 4800 4 2400 5 1200 6 600 7 300 2.6 SYSTEM HARDWARE CONFIGURATION system the n Distributio and Monitor the to dedicated Slot MDM is backplane. Kkernel. system the comprise 4 through 1 Slots module (M7677). through 12 support most UNIBUS compatible small 5 Slots peripheral controllers (SPC). The following subsection describe Figure 2-13 shows the location of the modules within configuration details for the kernel. ROWS A 1 F (MOM) M7677 MOM - . 3 o) c 8 ) R (XDJ11-8) MB190D (MSV11 . JI MBB37 2 M7556 3 (KTJ11-8) MB191 S HEX OR QUAD UNIBUS OPTION P1 SLOTS 6 HEX OR QUAD UNIBUS OPTION P2 7 HEX OR QUAD UNIBUS OPTION P3 8 HEX OR QUAD UNIBUS OPTION Pa FIGURE P 3 M&%“'JZD HEX OR QUAD UNIBUS OPTION 10 ”&?’B‘fs" HEX OR QUAD UNIBUS OPTION pe 11| MODIFIED HEX OR QUAD UNIBUS OPTION P? 12| UNBYS 2-13 QUAD UNIBUS OPTION H9277-A BACKPLANE 2-18 ‘ M7556 MODULE PRIOR e LOCATIONS SITE PRERARATION AND INSTALLATION 2.6.1 KDJ11-BF Module Configuration The KDJ11-BF module has three jumper groupings and one DIP switch pack for hardware configuration. The jumpers should be installed as shown in the Figure 2-14. The DIP switches should be- set to OFF. DCOK LED 1P SWITCH FOR BAUD AATE SELECT AND (GREEN) BOOT STRAP CONTROL OIAGNOSTIC /8 OFF CONNECTOR FOR CABLE TO BAUD RATE SELECT AND TWO-DIGIT DISPLAY N //78 [/ &Q CONNECTOR wio - (A~SERIES} H-A SOCKET :w—‘_FPJ40—PIN (P-SERIES) TP110{Jo TP10 £53 {H1 BYTE} o] @ DCJ11 CPU HYBRID 15:08 €80 (LO BYTE) /'EEPROM (SEE NOTE V) | wao ARRAY €35 TPa0O[)0" o TP42 . @ CONTROL CONNECTOR [53] {SID:II/ \- ROM REMOTE SWITCH I ——a FOR CABLE TO ~—_| CONSOLE SLU LEDS (RED) GATE TP41 ARRAY (SEE NOTE 2) w20 - 10 TP210TP22 TP200{ ] I L I NOTES: 1. WHEN 24-PIN EEPROM IS USED, INSERT PIN ONE OF EPAOM IN PIN 3 OF SOCKET. 2. WHEN 2X EEPROM 1S USED, TP40 IS CONNECTED TO TP41. WHEN 8K EEPROM IS USED. TP41 (S CONNECTED TO TP42. FIGURE 2-14 KDJ11-B JUMPER AND DIP SWITCH LOCATION 2.6.2 KTJ11-B Module Configuration The UBA module does not have any hardware jumpers or switches for sockets for the addition of M9312 four has It socket ROM the for 2-15 Figure See ROMs. configuration. compatible user locations. the The M9312 compatible user ROMs are installed with pin one of chip toward the left edge of the component side of the module. W31 TYPE OFTION ROM SOCKF TS (4) 2] eres [~ ROM SOCKET 1 e 17773000~ 17713178 3 f1e 11773400~ 17773878 72 ¢« tue fe ADOWESS 17773700-17713376 17779000 V7773178 Ered €143 #1142 | Il FIGURE 2-15 n__ n J___ 1 I KTJ11-B ROM SOCKET LOCATIONS 2-19 SITE PRERARATION 2.6.3 AND Monitor and INSTALLATION Distribution Module Configuration The guad-height Monitor and Distribution Processour Grant This Module (MDM) has an eight switch DIP pack. The eight switches correspond to backplane slots 5 through 12 and ‘are wired to the UNIBUS Non (NPG) line. eliminates the wire wrapping of backplane pins for non DMA SPCs. For non DMA SPCs the switch turned ON (toward module handles). For UNIBUS DMA devices turned is switch OFF. An audible alarm is mounted on failure 2-17 for of the a the module and sounds box fan or the cabinet blower occurs. switch location Na and numbering. A N (3o D5 D4 ::]” {20 PINI D3I D2 D :N;|J‘ - OV Qo000 Oksumne O.__ AUDIBLE ALARM sLov NUMBER 12 ) NPG ow 00000000 —1 swiTCH ) PACK "7 D1 02 - 5 (MAINPOVER SUPPL Y +SVE8 AND « 12V OLOWE R+ ANS) O4 19 . SOWE N Q1P v) 5V IF XPANSI <15V (€ XPANSI)'. POWE R SUPPL Y 03 - <15V MAIN PUE R SUPPLY FIGURE 2-16 MDM DIP SWITCH LOCATION 2.6.4 MSV11-JB/JC Memory Module Configuration The guad-height fiéfibry module provides: a. A red LED to indicate uncorrectable-errors b. A green LED to indicate c. Two switch packs for starting and CSR address selection d. Four See Figure factory-set 2-17 for the presence of +5 VBB jumpers. LED, switch pack, 2-20 and jumper a See Figure VOLTAGE TEST POINTS . |g gsv (‘.80 -58’3/ T when 1is the layout. SITE PRERARATION AND _J 1 4 - INSTALLATION T — 1 @m% O 02 w4 W3 03 LIl T FIGURE The l - 2-17 MSV11-JB/JC B I LAY, ) LED/SWITCH PACK/JUMPER LAYOUT starting address is configured using switch pack S1, switches 8 Table 2-5 lists the switch settings, starting addresses and decimal number. TABLE SWITCH 1 23 2-5 SETTING* 456 78 MSV11-JX STARTING ADDRESS STARTING DECIMAL DECIMAL ADDRESS Kwords Kbytes (Octal) 00000000 01 000O0O00O0 04000000 100000O00O 11000000 * o 1] hn 00000O0O00O Switch Switch The CSR address 4, The base base plus 2. 0 0 512 1024 10000000 1024 2048 14000000 1536 3072 : on off is configured using switch pack S2, switches 1 address is 17772100. Each successive address is the Table 2-6 lists all 2-21 sixteen possible CSR addresses. SITE PRERARATION TABLE AND 2-6 INSTALLATION MSV11-JB/JC. CSR ADDRESS SELECTIONS CSR [\ 4 S G A G D P ADDRESS P — VD CED P WO AP T R OO D (OCTAL) D ) TR D G R S D 17772100 17772102 17772104 17772106 17772110 17772112 17772114 17772116 17772120 17772122 17772124 17772126 17772130 17772132 17772134 17772136 configurations jumper (MSV11-JC) factory-set The jumpers TABLE - ) D D O D D D CED G ED SER are 2-7 WD GED GED GND 1MB (MSV11-JB) the Assure different. specified in Table 2-7. for modules memory as are MSV11-JB/JC CRD G GER N D R W A D IR WD SR JUMPER D R IR D R D and 2MB that the CONFIGURATIONS GO0 G D TI0 e AND OE) CER D CED G e e e — — MSV11-JB Wl OouT 256K W2 IN Half-populated Vertical W3,wW4 Dynamic Reserved RAMs for module future use MSV11-3JC ouT 256K ouT Fully Vertical D D X D G D WD CET N WS R WED G CER G N GO G GES R R D R TEN G O R D Dynamic populated Reserved e D T D D CED D O Rams D G D for I D D G module future O GED D OO0 G0 SED AR eme use e TGS D SITE 2.6.5 The Minimum MLM An loads MLM to are under is used a to provide the following inserted ensure AND INSTALLATION Module modules regulator Qe Load PRERARATION in CPU minimum minimum power supply conditions: backplane current slot drain of 3 2 (rows C and D) from the +5 VBB E and F) the -15 Vdc A regulator. An MLM is inserted to ensure regulator.. If an a in CPU mimimum expansion installed, an backplane (rows of 1 A from the backplane current drain backplane slot 12 of A 1 (DD11-CK MLM 1is inserted E and F) to ensure -15 Vdc regulator. (rows from or DD11-DK) in the last slot a mimimum current is of the drain NOTE An MLM is not required in (CPU or expansion) if the minimum 1 A of -15 When not required, the Load the the V. last backplane installed Modules options must be two LEDs slot draw removed from the indicate the backplane. As shown in Figure 2-18, presence of +5 and VBB the -15 MLM has 1 nd 8 1 FIGURE 2-18 to Vdc. 8 1 02 RED GREEN ] MINIMUM LOAD 2-23 [ LAY, ) MODULE LAYOUT SITE PRERARATION AND INSTALLATION 2.7 EXPANSION BACKPLANE INSTALLATION ?wo types of expansion backplanes are available for in the cabinet. installation They are: 1. DD11-CK 4-slot backplane 2. DDl11-DK 9-slot backplane Both backplanes are shown in Figure 2-19, 1:l ) | QuUAD wEiHY N . /A/ . ouoon /4 WODULES / m::mm N DD11-0K BACKPLANE TN — e, AN ) N/ VYV / ' / // V4 W/ A/ ) [ /\/ Sy e 7Zra 1 1 [ 1| FIGURE 2-19 EXPANSION BACKPLANE SLOT ASSIGNMENTS the UNIBUS ectors contain allbeginnin The standard UNIBUS A conn g of the the and B of slot 1 are connections. - Rows DD11-DK BCll-A the by pied and should be occu UNIBUS DD11-CK and UNIBUS in cable. the -DK, or of slot 4 eof slot Rows A and B of slot 9of ofthetheUNIBDD11 s Thes e. plan back the on US DD11-CK are the end or inat term should be occupied by the BC1ll-A UNIBUS out cable or a module (M9302 or M9312). SITE 2.7.1 To Expansion install following the expansion the AC power circuit Remove the release Lower the breaker back the backplane from the OFF (0). cover door panel the left front- by lifting cabinet up panel the side and two Remove metal the not two LEXAN insert(s) assembly, power using Remove Remove INSTALLATION a 4 mm pérform controller (5/32) by hex after side from by (plastic) unscrewing cover the head the =-viewed bottom. setting wrench 10 from See unscrewing phillips to screws. the figure the cabinet 2-20. four shoulder screws. cover behind. over the backplane and the Discard the metal insert, it 1is reinstalled. Position the expansion align the two tapped four tapped screw holes backplane through screw holes for the for the DD1l1-DK. the front DD11-CK, or NOTE The backplane harness a lug attached mounting screw. If the fasteners. bulkhead screws AND Installation steps. Remove the Backplane PRERARATION there modules, jumpers slots. are a includes must sufficient install from that the be number modules backplane a pins . ground installed of after lead with under the NPG jumper removing CAl-CBl1 for NPG all and the S T e -——y - 63;7 : 47, »//é% SITE PRERARATION AND INSTALLATION FIGURE 2-20 EXPANSION BACKPLANE MOUNTING are supplied with the Install the two/four 8-32 screws that s. backplane. Do not tighten the screw Install the backplane wiring harness . connectors cabinet power distribution connectors 10. Install two hex modules in each end slot of the into the backplane to align the slots. 11. Tighten the 8-32 screws installed in step 8. 12, Remove the hex modules from the backplane. 13. Replace the LEXAN (plastic) cover on the backplane. 14. Replace the side panel using the four two phillips head screws. 2-26 shoulder bolts and SITE 15. Replace above the 16. the side shoulder shoulder bolts. Close the mounting 17. outer the rear panel by bolts. panel PRERARATION AND aligning Lower bulkhead screws. the and the INSTALLATION two cover brackets brackets tighten the onto captive . Close the front door and lock using the hex key. This completes the installation of a DD11-CK or DD11-DK optional backplane. 2.7.2 The Expansion NPG line is Backplane the NPG/BG UNIBUS Jumper grant line Lead for Routing devices performing data transfers without PDP-11/84 processor intervention. Continuity of the NPG line is provided by wirewraps or jumpers on the expansion backplane. When an NPR device is placed wire from pin CAl to pin routing of the NPG signal Figure the 2-21. DD11-DK Grant (slot 1 in priority has a slot, the corresponding CBl of that slot must be through the backplane decreases from priority and highest slot slot jumper removed. The 1is shown 1in 1 9 to slot has 9 in lowest). NOTE If The bus routed an NPR device is removed jumper wire grant lines (BG4 through through each slot in slot 1 slot decreases from from CAl to to row CBl from must BG7) D. for Grant a be slot, the reconnected. non NPR priority devices for each 9. NOTE A bus grant jumper card G727 in row D, or G7273 in row C and D must be installed in all unoccupied SPC slots. If an SPC slot is left open, bus grant continuity will system will not operate. be lost and the are level L g ré LJ L .’ N/ M h - - RO WO, - - \} NIVNN - - g4 1 SITE PRERARATION AND INSTALLATION NOTE Routing wire NPG only. backplanes DIP switch on the expansion for are wraps has NPG routing CPU backplane® MDM. FIGURE 2-21 NPG JUMPER LEADS ROUTING 2.7.3 Backplane Power Connections a wire thruogh backplane expansion the Power is supplied to power distribution board with the power the connecting harness of set to a The power wires run from the backplane supply. Mate~N-Lok connectors wired directly into the distribution board. The power harness from the DD11-DK contains two 15-pin Mate-N-Lok The DDl1-CK 6-pin Mate-N-Lok connector. and one connectors, The connector. 6-pin one and connector 15-pin one backplane has signal the and 2-22 Figure in shown are locations pin connector and (DD11-CK) 2-8 assignments for each pin are listed in Table Table 2-9 (DD1l1-CK). s 00t ~ imi lak . k\w’oo" . 'oc>at; Yo o of . -\"O o of o of o & i) o P L~ n . NJQ(D&:/ M 0 O [0 |) B 8§ o CTOR ) W M CONmECTOR FIGURE 2-22 BACKPLANE CONNECTOR DESIGNATIONS SITE PRERARATION AND TABLE 2-8 DD11-CK POWER CONNECTOR SIGNAL ASSIGNMENTS N L B WD OWJOULD WN K 15-Pin 15 Mate-N-Lok +5 V +15 V Spare +5 V Connector 14 18 connected) Spare (not connected) (not connected) (not connected) ‘ 6-Pin Red 18 l4 14 Green Black Black 14 18 18 = Red Blue Brown White - Mate-N-Lok GND LTC (line clock) DC LO AC LO Spare (not connected) Spare (not connected) Red Gray Orange 14 Spare (not +15 B Ground Ground Spare Spare +5 B -15 V Spare -15 B INSTALLATION 18 18 18 18 - - Connector Black Brown Violet Yellow = = SITE PRERARATION AND 2-9 DD11-DK POWER CONNECTOR SIGNAL ASSIGNMENTS TABLE o e M G S S S D D INSTALLATION M WU WED M GED S G D EN D SR I D D W D S WD WS NS CH WM WP W 15-Pin Mate-N-Lok +15 V +15 V GED WD S WP CHD GED G MED WD M MED R WD CH) GNP DR G Connector 1 Red 14 18 Gray Orange Spare 14 Red Ground 14 Black Ground 14 Black Ground 14 Black 14 Red 18 Brown +5 V Spare Spare Spare Spare +5 (not (not (not (not connected) connected) connected) connected) B Spare (not connected) -5V Spare (not connected) 15-Pin +5 Mate-N-Lok Connector Spare Spare +15 2 14 VvV (not Orange (not connected) B Spare Red connected) Spare +5 V 14 Red 18 White (not connected) Ground 14 Black Ground 14 14 Black 18 Blue 18 Green Ground Spare Spare -15 V Spare -15 B (not (not connected) connected) (not connected) 6-Pin Mate-N-Lok Black Connector Black GND LTC (line DC LO AC Lo Spare Spare D D D CED WER TRR D NS I emm e e D GID G GES D R G clock) Brown Violet Yellow (not (not GED N G connected) connected) AP SD G A GAR AR AEC D D GED A AN D S D G S R D D WD WD e D D SND R CH D 5D D ) e D G D D S - SITE PRERARATION AND 2.7.4 SPC Backplane INSTALLATION Locations The small peripheral control sections (C, D, E and F) well as power 1lines as UNIBUS the all contain collectively voltages (+5 V, +15 V and -15 V). These sections can be used by hex- or quad-height peripheral devites. the SPC backplane modules containing Appendix D shows the the control logic for pin designations for connectors. Appendix D also shows the pin designations of the standard and modified UNIBUS connectors. The modified UNIBUS differs from the standard UNIBUS in that certain pins have been redesignated. 2.8 SPC MODULE INSTALLATION CPU backplane slots 5 through 12 support the installation of UNIBUS SPC modules.. Backplane slots 5 through 11 support both hex- and quad-height SPC modules; slot 12 supports quad-height SPC modules onlv. connector, Row A of slot Hex-height SPC modules 12 occupy all supports four the UNIBUS out rows of the cable Dbackplane while gquad-height occupy rows A through C. Quad-height SPC modules, when installed, occupy the same backplane rows as the system CPU and To install an memory modules. SPC module, the two handles steps. Grasp 2. Install the module in the backplane slot by sliding it into 3. When card by following l. the the module perform the mounted at the top. cage guides. the module is about three quarters installed, grasp the handles (toward the module center) and swing them upwards, away from the module (see Figure 2-23). FIGURE 2-23 SPC MODULE INSTALLATION 2-31 SITE 4. PRERARATION AND INSTALLATION and press This action seats the module into Continue to slide the module into the hackplane the handles downward. the backplane and secures it in the card cage. This completes the installation procedure for an SPC module. 2.8.1 Cabinet SPC Cable Routing Use the following directions for proper cable routing. 1. slot, cable the and connector the into plugged the SPC cables are After installing an SPC in the appropriate backplane is routed to the bulkhead assembly. 2. connectors mounted on the edge of the module, are routed toward the right side All SPC cables that plug into front of the card cage as shown in Figure 2-24. When the connector is mounted on the top of the module, the 1is routed up and over the cable hanger assembly cable located above the card cage. 4. If the connector is mounted near the module handle the installed cable 1is routed around the front of the card cage. FIGURE 2-24 CABINET SPC CABLE ROUTING 2-32 SITE Located 2.8.2 Use 1, on the cable clamps A is bar to hang Box SPC the mounted the SPC Cable following After right used side PRERARATION AND INSTALLATION of the <card cage are plastic the SPC cable to the card cage. horizontally behind the card cage and used to secure cables that are for proper routed to the bulkhead. Routing directions installing the SPC is routed to The back panel and top cables the right an are SPC in bulkhead has the plugged two the assembly. from routing. appropriate into bulkhead referenced cable backplane connector and the assembly the areas, module Cables are that (installed bulkhead. routed plug in See into toward connectors backplane) Figure FIGURE 2-25 the are 2-25. BOX SPC bottom left rear. Cables that plug into ‘connectors mounted on the the slot cable lower left handle bulkhead. of mounted at the module routed to the upper right CABLE ROUTING top SITE PRERARATION AND INSTALLATION 2.9 CABINET BATTERY BACK UP UNIT INSTALLATION To install the BBU complete the following procedure. CAUTION 1lbs; 42 is The weight of the BBU lifting positioning the BBU requires two people. and 1. Unpack the BBU and installation kit. 2. Remove the BBU from the shipping container and place it on 3. Set the front panel TOY (Time of Year) switch to OFF. See a flat Figure surface. 2-26. : I I TOY FIGURE 4. I VOLTAGE ON/QFF SELECT Con Co BBU 2-26 PANEL FRONT Set the BBU VOLTAGE SELECT switch to match the site 1line voltage. 5. Set the rear BBU AC VOLTAGE SELECT to match the front panel VOLTAGE SELECT switch setting. BBU AC INPUT CONNECTOR (422) See Figure FAULT INTERFACE CONNECTOR (J20) {FAILSAFE JUMPER} / / @ @\ 88U INPUT AC VOLTAGE SELECT GROUND STUD FOR DC POWER OUTPUT USER INTERIACE COMNECTOR (J18) SWITCH =——240 11§5— DC POWER QUTPUT CONNECTOR (J9) POWER CONTROLLER BUS CONNECTOR (J19} FIGURE 2-27 BBU 2-34 REAR PANEL 2-27. SITE Open the front Turn off the and rear power PRERARATION cabinet supply doors and power AND using the INSTALLATION hex controller key. circuit breakers. Unplug the AC power cord from Remove the cabinet right up from both sides. See the outlet. side panel by Figure 2-28, lifting it straight 14 SHOULDER SCREW (4) e EMI SHIELD = = SCREW (2) L SIDE PANEL e 13 FIGURE 2-28 SIDE 10. Remove the two phillips securing the EMI panel to 11, Carefully screws lift and protruding line from up the PANEL REMOVAL head and four shoulder the cabinet frame. the BBU cabinet 2-35 enclosure frame. with the Figure See screws four 2-29. SITE PRERARATION AND INSTALLATION CARD CAGE /‘ussemsw U-NUT REVAINER " W/10-32 SCREW (4) \Nzx E®- g | LINE FILTER 10-32 KEPNUT (4) BATTERY BACKUP 877-0/F POWER CONTROLLER (H7231-€) UNIT NIT (H7231- FIGURE 2-29 MOUNTING THE EBU 12. 13. Position the BBU on the four mounting screws and against the cabinet frame. Install and installation to the frame. slide it the (from tighten four 10-32 hex nuts BBU the ing kit) on the mounting screws, secur 14. Remove all cables from the installation kit. 15. Install the two-position keyed jumper into mating connector 16. Plug one end of the 17. (J20) on the BBU rear panel. keyed cable (7020396), wire, into the BBU mating connector (J9). with ground Place the Remove the hex nut on the BBU ground stud. replac e and ground wire from the cable on the stud, and tighten the hex nut. 18. Plug the other end of the keyed cable, with ground into the line filter bracket connector marked BB-J1l. the cable ground wire on the stud. 19. wire, Place Install a 10-32 hex nut (from the installation kit) on stud. Tighten the nut to secure the ground wire. 2-36 the SITE PRERARATION AND INSTALLATION 20. Plug one end of the other keyed cable (7008288) into the mating BBU power controller bus connector (J19). Plug the other end of the cable into the 877-D/F Power. Controller connector marked either J8 or J9. 21, Plug the. (1700730) the top ten position connector of the signal <cable into the mating polarized connector located at right of the card cage assembly. Attach the cable ground wire to the stud with the 10-32 nut provided. Secure the wire by tightening the nut. 23, Plug the other connector screws 24 . end (J18) on the (D-sub) on of the the BBU. cable Tighten into the the hex mating self-retaining connector. Plug the appropriate power cord (120V or 240V) into the BBU and plug the other. end of the cord into the remaining UNSWITCHED outlet on the 877-D/F power controller. 25, Check each installed cable to ensure that none damaged during reinstallation of the EMI shield. 26 . Reverse steps side panel. This completes 2.10 BOX the BATTERY 6 through 8 installation BACK UP UNIT to reinstall the EMI will shield be and procedure. INSTALLATION To install the BBU complete the following procedure. CAUTION The weight of the BBU 1is 42 positioning the unit requires Unpack the BBU and Remove the BBU from a surface. flat Set the front installation See panel 1bs; 1lifting two people. the TOY kit. shipping Figure 2=37 container 2-30. switch and to OFF. and place it on SITE PRERARATION AND INSTALLATION I I TOY FIGURE 4. VOLTAGE ON/OFF SELECT - —m BBU 2-30 FRONT PANEL site line Set the rear BBU AC VOLTAGE SELECT switch to match See Figure 2-31. front panel VOLTAGE SELECT switch. the Set the VOLTAGE switch SELECT to match the voltage. 5. 88U AC INPUT CONNECTOR (J22) FAULT INTERFACE CONNECTOR (J20) (FAILSAFE JUMPER) {ooo} 1 /e GROUND STUD FOR DC POWER OUTPUT BBU INPUT AC VOLTAGE SELECT USER INTERFACE CONNECTOR (J18) SWITCH =240 DC POWER QUTPUT 15— CONNECTOR (J9} POWER CONTROLLER BUS CONNECTOR (J19) FIGURE 2-31 BBU REAR PANEL 6. Follow the directions supplied with the rack and 7. Turn the box circuit breaker off. secure the BBU FIGURE to the 2-32 carefully rack. See Figure 2-32. CIRCUIT BREAKER LCCATION 2-38 SITE PRERARATION AND INSTALLATION 8. Unplug the AC power cord from the AC outlet. 9. the that secure Loosen the four captive screws See Figure 2-33. cover, and remove the cover. box top FIGURE 2-33 TOP COVER REMOVAL 10. Remove the blank rear bulkhead panel Bl by loosening the two phillips head screws that secure it to the bulkhead. See Figure 2-34. BULKMEAD A I suikneao—=1"/ FIGURE 2-34 BULKHEAD LOCATION 2-39 SITE PRERARATION AND INSTALLATION 11. Remove the BBU panel from the installation kit. 12. Route the two attached cables through the bulkhead 13. Install the BBU connector panel (shown in Figure that was occupied by panel Bl. the opening 2-35) 1in slot MDM opening and tighten the two flat-head screws bulkhead into the bulkhead frame. ] FIGURE 2-35 BBU CONNECTOR PANEL 14, Label and remove through slot 4. the cables from the module 15. Remove the five modules from the backplane. 16. Plug the signal cable connector (10 position) 1into the backplane PC board mating connector (J12). See Figure 2‘36 ° 8ACKPLANE FIGURE 2-36 BBU CABLES 2-40 AND LOCATIONS SITE 17. 18, PRERARATION AND INSTALLATION Plug the other BBU cable connector (3 position) into the panel-mounted mating connector located behind the rear of the power supply. : Reinstall the five modules insuring that they are properly .in the backplane. 19, Plug into 20, Remove 21, Install the two position keyed jumper into mating (J20) on the BBU rear panel. See Figure 2-37, the cables the their remaining two 88U AC INPUT CONNECTOR (J22) module cables seated connectors. from the installation kit. connector FAULT INTERFACE CONNECTOR (J20) (FAILSAFE JUMPER) ;I @ . {ooo} [erel=] : BBU INPUT GROUND STUO FOR usehmenucs AC VOLTAGE SELECT DC POWER QUTPUT SWITCH =—240 ns— CONNECTOR (J18) DC POWER OUTPUT CONNECTOR (J9) POWER CONTROLLER BUS CONNECTOR (J19) FIGURE BBU 2-37 REAR PANEL NOTE BBU connector J19 22, 23. Plug one end of the wire, cable (7020396), (J9). with ground the Place stud. ground BBU the Remove the hex nut on and replace and stud, the on cable the from wire ground hex nut. Plug the other end of into the keyed cable, with the BBU bulkhead panel connector Jl. ground wire 25. keyed into the BBU mating connector tighten 24, is not used with this system. on the ground wire, Place the cable stud. Install a 10-32 hex nut (from the installation kit) Tighten the nut to secure the ground wire. stud. on the SITE PRERARATION AND INSTALLATION 26 . Plug the signal cable (1700730) into the mating polarized connector located on the BBU bulkhead panel. Attach the cable ground wire to the stud with the 10-32 hex nut provided, 27. : Plug the -other end (D-sub) of the cable into the mating Tighten the self-retaining screws on connector on the BBU. the 28. and *ighten the nut. connector. Peplace the top cover of the box, and tighten the captive screws. 29. Plug the appropriate power cord (120V or 240V) to the BBU the cord into the AC power and plug the other end of outlet. This completes the installation of the BBU. CHAPTER FUNCTIONAL 3.1 3 DESCRIPTION INTRODUCTION The PDP-11/84 functional block diagram is shown Figure in 3-1. KDJ11l-BF CPU, MSV11-JB/JC ECC memory and three modules are: The J11 a has module KDJ11l-BF The adapter. UNIBUS KTJ11-B Private memory, cache 8KB an FPA, integral microprocessor with Memory Interconnect (PMI) EEPROM/ROM, registers, arbitration configuration 1logic, memory registers, management a SLU (serial line unit), six red diagnostic LEDs and a green +5V power LED. PMI a 1into divided The KTJ11-B UNIBUS Adapter (UBA) module is The handshake logic enables data a UNIBUS section. and section PMI The devices. UNIBUS and PMI transfers to occur between section has diagnostic registers and PMI arbitration/interface UNIBUS cache, DMA 32-word a has section The UNIBUS logic. M9312-type for sockets interface logic. logic, mapping arbitration and wuser ROMs and UNIBUS The MSV11-JB memory module has a 1MB capacity, while the MSV11l-JdC has a 2MB capacity. A maximum of two memory modules are allowed in the system configuration. The modules the PMI KTJ11-B DATBI) DATO, transfer and module. can DATOB) UNIBUS PMI data using devices read the wuse be word or byte Data transfers (i.e., DATI,DATIP, PMI write operations mode. NOTE The PDP-11/84 UNIBUS Power slightly different from most Appendix B for protocol between the Handshaking Logic on the operations, be word or block mode. can PMI. up protocol PDP-lls. Refer description. is to ( and i.e., FUNCTIONAL DESCRIPTION All communications between through standard configured on the UNIBUS system. éLL devices No ROM HAND DIAG | __| SHAKE DATA ; t PMI | cache MEMORY INTERFACE INTERFACE b |, : + ' ‘ PMIARE occur may be PMI ARB AND MISC INTERFACE REG.S gsze Lodic REG ] _J FPI11 FPA UBA the devices KTJ11-8 MODULE MEMORY --= - EPROM and OQBus MSV11-J MODULE KDJ11-8F MODULE SSTSOLE UNIBUS protocol. : t ] .. : ! § UNIBUS ’r AND INTERFACE o UNIBUS e [ - MAPPING.. ROM PMI pCTAL CONSOLE DISPLAY TERMINAL MA-15296 FIGURE 3-1 PDP-11/84 FUNCTIONAL BLOCK DIAGRAM 3.2 PMI BUS DESCRIPTION path The PMI bus provides a high-performance communicationKTJ11B the and , Memory between the KDJ11-BF (CPU), MSV11-JB/JC PMI to unique are that s (UBA). The PMI consists of 14 signal protocol. The signal lines are described in Table 3-1. The KDJ11-BF CPU module is also used in QBUS systems and the therefore some of the signals retain their QBUS names, but in is (UBA) B KTJ11a when change s signal these functionality of the system. Data and address information is multiplexed and uses the same data/address lines as QOBUS protocol. FUNCTIONAL B IR I BDAL <21-00> P NP WD W PMI 3-1 TABLE SIGNAL LINE DESCRIPTIONS SmS w0 e = GNP AP M GEE SR G SMD GED GED WIS WD NS GES D AT W WD GEN GED WS CAD SN SRR AP TE WS SR GED GES GHP UL GHD GHR G GMD GED QA TR GLP TP GER GNP MDE GHD b eSS 22 multiplexed bidirectional Data/Address During DESCRIPTION Lines. the address phase of a data transfer cycle, the PMI master gates address information onto these During the data phase of the cycle the slave lines. (DATI) or the master (DATO) gates data onto BDAL<K15-00> and parity error/control information onto BDALK17-16>. BBS7 Bank 7 Select (I/O Page Select). When the PMI master gates an address onto BDAL<21-00>, it asserts BBS7 to reference the I/0 Page (including the I1/0 Page addresses reserved as non-existent memory). When BBS7 is asserted, BDAL<K12-00> specify the I/O page address, BRPLY Negation of BBS7 selects the memory address space. Reply This signal is asserted by the UBA as a slave during the PMI DATO(B) vector BDIN Data DATI response cycle and during the interrupt cycle. Input This signal is used by PMI protocol during UNIBUS interrupt The CPU asserts BDIN after gating the grant cycles. The UBA latches the interrupt priority onto BDAL<03-00>. interrupt priority on the leading edge of BDIN, BIACKI Interrupt Acknowledge This signal is used by PMI protocol during UNIBUS interrupt When the UBA receives the assertion of BIACK grant cycles. (from the CPU), signals. it asserts one of the UNIBUS grant (BGn) FUNCTIONAL DESCRIPTION TABLE 3-1 (Cont) nfi——__-—_—————-—---_—_— ——-m--n-——————-————-—_-—-———-—‘-‘!-fl—fl_d —u---:—-_—_——-_———_-—_ ‘d—m—_--:————————--——-——-——-———-—m——‘au-u- Power OK This signal is asserted and negated by the UBA in response to AC LO on the UNIBUS following standard UNIBUS power-up/ down protocol. UNIBUS devices and/or PMI memory may, during power-up, prolong the assertion of AC LO/BPOK. The following signals are asserted (low) and negated (high) by the PMI master: PMI PBCYC Bus Cycle The PMI master asserts this signal at the start of a PMI cycle and negates this signal at the end of that cycle. PBYT PMI Byte BWTBT Write and Indication Wwhen the PMI master gates an address onto BDAL<21-00>, it asserts or negates these signals to indicate what type of data transfer will occur during the next bus cycle: BWTBT | H H L H L H L L PMI PBLKM PBYT Block | Bus Cycle Type DATI or DATBI DATIP DATO DATOB Mode When a PMI master wishes to read more than two words of data, it uses both PBCYC and PBLKM to control the It timing of the Block Mode Data In (DATBI) cycle. DATBI the of start the at asserts both PBCYC and PBLKM cycle. It negates PBLKM after reading two data words _ and then reasserts PBLKM (unless the next two words words, two last the reading After cycle). will end the the PMI master negates PBCYC (PBLKM is already negated). - o = o e GHD <O WIS CED G WP D S CHD G WD a5 - e D TED G D aTD W IO D G WER MR W WD SIS G MO W D SND NP WIS WD CH G S s - D D D O e D R D O W D O OO FUNCTIONAL TABLE PMI Write 3-1 DESCRIPTION (Cont) Strobe The PMI master asserts this signal after gating data onto BDAL<15-00>. The PMI slave latches the data into its write buffer on the leading edge of the PWISTB pulse. QSACK The UBA asserts this signal on the PMI in response to from SACK DMR the UNIBUS. The UBA asserts this signal on the PMI in response to NPR from the UNIBUS or when the UBA is performing a DMA cycle its own behalf. in DMG The CPU asserts this signal when PMI mastership has been granted to the UBA in response to a DMG. QBR7-4 The UBA asserts one of these signals in response to one of the BR7-4 lines being asserted on the UNIBUS during interrupt request cycles. The following signals are asserted and negated by the PMI slave: PSSEL PMI Slave Selected The PMI slave (CPU or Memory only) asserts this signal whenever it decodes a valid address on BDALK21-00>. NOTE : PHBPAR When PUBMEM is asserted the PMI slave does not PUBMEM is respond to PMI control signals. UNIBUS that e indicat to UBA the by d asserte UBA does The ed. address being is space memory The CPU ignores the not assert PSSEL. assertion of PSSEL if PUBMEM is asserted. PMI High Byte Data Parity This signal is generated by PMI memory during DATI and DATBI cycles and provides odd parity for the high byte data PLBPAR (on BDALK15-08>). PMI Low Byte Data Parity This signal is generated by PMI memory during DATI and DATBI cycles and provides even partiy for the low byte data (on BDAL<07-00>). FUNCTIONAL DESCRIPTION TABLE PRDSTB 3-1 (Cont) Read Strobe PMI The PMI slave asserts and negates this line to control data transfers during DATI and DATBI cycles. The PMI master latches the first word of the received data on the negating edge of this signal. The PMI master latches the second data word a specified time after this signal PSBFUL is negated. PMI Slave Buffer Full A PMI slave asserts PSBFUL during a write cycle, indicating that it's Write Buffer is full, and that it can not respond to another cycle regquest. The new PMI master may gate an address onto BDAL<Z21-00> while PSBFUL is asserted, but it must not assert PBCYC until PSBFUL negated. is The following signals are used for communication between the CPU and the UNIBUS adapter. PMI Memory modules do not use these signals. PMAPE PMI UNIBUS Map Enable The CPU module asserts this signal if Memory Management Register 3 (MMR3) bit 5 is set and negates this signal if MMR<O05> is clear. The UBA module enables the UNIBUS Map if PMAPE is asserted and disables the UNIBUS Map if if PUBSYS PMAPE PMI is negated. UNIBUS System This signal is a static signal which indicates whether the system is a UNIBUS system or a QBUS system and is asserted by the UBA. The CPU follows PMI protocol for data transfers whether PSSEL is asserted or not. PUBMEM PMI UNIBUS - Memory This signal line is asserted by the UBA to indicate The that the UNIBUS memory space is being addressed. UBA asserts PUBMEM during the assertion of PBCYC. NOTE When PUBMEM is asserted the PMI slave does not PSSEL is respond to PMI control signals. . The CPU addressed when memory asserted by PMI ignores the assertion of PSSEL if PUBMEM is asserted. The UBA does not assert PSSEL. FUNCTIONAL DESCRIPTION TABLE 3-1(Cont) ————-——-——-——--—-——4--—— ———-——-——_-———————-————-———-——-———————-—————— PUBTMO - PMI UNIBUS Timeout This signal is asserted by the UBA in response to any of the following conditions: PBSY a. A non-existent memory timeout occurs whens. the UBA sends an address out on the Unibu b. When a SACK timeout occurs during an c. device has been When an interrupting UNIBUS but does not interrupt cycle. granted UNIBUS mastership, execute an interrupt transaction. PMI Busy or UBA) ed by the PMI masterted(CPU This signal is assert the PMI by nega is and tership when it gains PMI masuis CPU is The ip. ersh mast PMI hes master when it reling power-up. the PMI master on 3.2.1 PMI BUS Acquisition PMI Bus master and the CPU is the default In the PDP11/84 systrem(UBA er. This is US UNIB ult ) is the defan the UBA ismast pte the UNIBUS Ada not requestings ion at power up.es Whe always the conditthe CPU arbitrat to become PMI master and hold the PMI bus, PBUSY asserted. of the devices ones the in the past, wfien none Unlike other PDPll's the UBA arbitrat to UNIBUS are requesting use of theassebus, . rted become UNIBUS master and holds BBSY responding to a DMA PMI mastership when The CPU relinquishes rrup UBA. Once the CPU has t cycle from the rega inte request or an rol the PMI, it canmet: in PMI mastership relinquished cont owinofg cond itions are only when the foll 1. OSACK has been negated for 75ns minimum. 2. PBSY has been negated for Ons minimum. 3=-7 FUNCTIONAL DESCRIPTION ences a memory or I/0 page When the CPU, as PMI master, referresp as the slave on the PMI address on the UNIBUS, the UBA of onds the data transfer as bus while controlling the UNIBUS side ) master. CPU issues a DMG (DMA Grant) The UBA becomes PMI master when the or . The UBA may accept the DMG the or performs an interrupt cycle at er mast S both PMI and UNIBU interrupt gra. t, thus becoming may pass the DMG or interrupt grant it ely, nativ Alter same time. UNIBUS on to a requesting UNIBUS device, which would then become master. ws: Mastership of the UNIBUS and/or PMI bus is requested as follo an NPR me UNIBUS master through 1. A UNIBUS device can beco During sfers over the UNIBUS. onds request and control data, tran as the UBA is PMI master and resp these data transfers device accesses a PMI memory UNIBUS slave if the UNIBUS location or a UBA 1I/0 page location, a PMI location. I/O page a BR7-4 me UNIBUS master through 2. A UNIBUS device can becomast data rol cont can ce devi the reguest. As UNIBUS traner, UBA the s case both In s. sfer and/or interrupt vector US slav e. The device may perform an will respond as UNIB access a PMI memory location, interrupt vector cycle or a UBA 1/0 page location. PMI I/O page location or a at the same PMI and UNIBUS master PMI 3. The UBA can becomeandboth and UNIBUS As interrupt requests. time through DMG PMI. ct access to the master, the UBA has dire 3.2.2 DMA Requests can ugh an NPR request to esthewithUBA, A UNIBUS DMA device, throDATI UBA the cycl B DATO and P, DATO perform UNIBUS DATI, ion of the data transfer. controlling the PMI port NOTE Unlike previous PDPll systems, the CPU unconditional does not necessarily give programmed priority to DMA regquests. It 4)can to be give itself SETUP mode Chapter (see ing, wait r afte ests priority over over DMA requ time, to perform a of th for this programmed leng est. requ t rrup inte an r hono memory transfer or to When placed transfers UBA itself cycle in test mode becomes complete The following master of the PMI control of the PMI. PMI protocol arbitrating a flow and is DESCRIPTION UBA perform the without a regquesting UNIBUS is the requesting device and has when diagnostic FUNCTIONAL the DMA the DMA UNIBUS simultaneously and the CPU and UBA (NPR) from a observed Non-Processor can device. 1In this case when performing a by Request UNIBUS device: PMI BUS UNIBUS l. The UNIBUS device asserts NPR. 2. If the UBA is UNIBUS master it negates BBSY after removing address, data and control information from the bus. 3. The UBA asserts on the 4. The 5. The PMI DMR (DMA Reqguest) bus. CPU bus arbitration logic asserts DMG (DMA Grant) on the PMI after receiving DMR and 75ns minimum after the negation of QOSACK from a previous PMI bus transaction. UBA receives the assertion of DMG. 6. The UBA asserts NPG Processor Grant). 7. The the SACK 8. The Note: UBA asserts QSACK on the PMI. The UBA asserts PUBTMO instead of QSACK (indicating No SACK timeout) if 10 the DMA SACK is not us after it received within asserts BGn on UNIBUS. The CPU then cancels the cycle and resumes arbitration. (Non- requesting device with highest priority asserts and negates NPR. FUNCTIONAL DESCRIPTION UNIBUS PMI BUS 9- The CPU arbitration logic receives QSACK and negates DMG. Note: Because the UBA provides the No SACK Timeout function on the PMI, the CPU always asserts DMG untill it receives QSACK or UBTMO. 10. The UBA negates NPG on the UNIBUS. 11. The UBA asserts PBSY, after receivind the negation of PBSY from the previous PMI cycle, and becomes PMI master. 12. After receiving the negation of BBSY from the previous bus master the UNIBUS device asserts BBSY and negates SACK. S master through an NPR When a UNIBUS DMA device becomes USUNIBU B DATI, DATIP, DATO and DATO UNIB request, it can perform sses 1/0 PMI a , tion PMI memory loca cycles. If the device acceS I/O aPage location located on the UBA, UNIBU a or Page 1location e. For PMI memory and PMI I/0 the UBA responds as the UNIBasUS slav the PMI master, controls the PMI Page accesses, the UBA portion of the data transfer. Data transfer cycles are described in section 3.2.5. 13. The UBA negates QSACK. 14. The CPU resumes arbitration 75 ns minimum after receiving the negation of QSACK. 15. The UNIBUS device removes address, data, and control information from the bus and negates BBUSY. 16. After the PMI slave or the UBA has removed all data and control information from the bus, the UBA negates PBSY. _-anmgaum-pcpa D D D ED D R D D D R D T WD CED O D ——-—-—_-n—-'es——--‘a@-—-—n—-:—————————-——-—q-m—— FUNCTIONAL 3.2.3 UNIBUS CPU The and arbitrating e e o Device UBA inter- PMI e Interrupt observe rupt DESCRIPTION Requests following the protocol flow when L S requests: BUS v e o UNIBUS i B l. 2. The CPU receives the appropriate request level on QBR7-4. 3. The CPU arbitration T The UNIBUS device asserts the appropriate interrupt request line BR7-4. logic asserts one of the four lines, DALK03-00>, to indicate the level of the granted interrupt. 4. The CPU asserts BDIN 150 ns minimum after gating DAL<K03-00> onto the bus. 5. The CPU asserts BIAK 225 ns minimum 6. 7. after it asserts BDIN. The UBA latches DAL<K03-00> on the asserting edge of BDIN. The CPU receives the assertion of IACK on the PMI. Note: The UBA compares the interrupt level being granted with its own interrupt level and can block the grant. In this case the UBA performs an interrupt or data transfer cycle and has complete control of the PMI and the UNIBUS simultaneously. In this case the BGn line would not be asserted on the UNIBUS. 8. The UBA asserts UNIBUS Grant DAL<03>= BG7, the selected (BGn) line. DAL<02> = BGS, etc. 9., If the UBA was the UNIBUS master it removes address, data and control from BBSY. the bus and information negates FUNCTIONAL DESCRIPTION PMI BUS UNIBUS ___________________________________ I_...-_..._._...__... ___..__..._.._———_..-—— 10. The highest ing device assertion priority requestreceives the of BGn and asserts SACK. l11. 12. The Note: 15. UBA asserts The device negates BRn. QSACK. The UBA asserts PUBTMO (indicating No SACK timeout) if SACK is not received on the UNIBUS within 10 us after it asserts BGn. When the CPU receives the assertion of PUBTMO it cancels the interrupt cycle and resumes arbitration. 13. The UBA negates BGn. 14, After receiving the negation of BBSY from the previous bus master the UNIBUS device asserts BBSY and negates SACK. After the UBA receives the negation of PBSY from the previous PMI cycle the UBA asserts PBSY and negates QSACK. 16. it's The UBA now has control of the PMI bus and may initiate a PMI data transfer cycle(s) and/or an interrupt cycle. NOTE The CPU resumes NPR arbitration 75 ns after the BR resume not does but QSACK of negation PSW and arbitration until it has updated the PC to complete the interrupt cycle or has aborted the interrupt request. FUNCTIONAL PMI BUS DESCRIPTION UNIBUS __________________________________ B DR e R R R R R TR When a UNIBUS device becomes UNIBUS master through an- interrupt request, it can perform Interrupt vector cycles or UNIBUS DATI, DATIP, DATO and DATOB cycles. If the device accesses a PMI memory location, a PMI 1I/0 Page location or a UNIBUS I/0 Paqge location, the UBA responds as UNIBUS slave. For PMI memory and PMI 1I/0 Page accesses, the UBA as the PMI master, controls the PMI portion of the data transfer. BDIN and BIACK being asserted does not effect the data transfer. Data transfer cycles are described in section 3.2.5. The following sequence describes the interrupt transfer cycle: 17. The interrupting device, as bus master, gates its vector onto the data lines and asserts INTR. 18. The UBA, BRPLY on as the PMI master, asserts PMI. 19. The UBA as slave receives the assertion of INTR and latches the interrupt vector. 20. The UBA asserts SSYN. 21. The UBA gates the vector onto the DAL 22. lines. The CPU latches the interrupt vector 200 ns minimum after receiving BRPLY. 23, The CPU negates BDIN and BIAK. 24. The UBA receives the negation of BIACK and negates 26. The UBA negates BRPLY. PBSY. 25. After receiving SSYN, the device removes it's vector from the data lines and negates INTR and BBSY. : FUNCTIONAL DESCRIPTION 3.2.4 PMI Data Transfer Address Cycle iately after addressing phase of the PMI cycle starts immed ted PBSY on asser has and rship maste PMI d the CPU or UBA has gaine The the PMI. The PMI bus acquisition phase is described in 3.2.1. MASTER 1. The address SLAVE is gated out on BDAL<21:00>, and BS7 is asserted if the address is in the I/O page. The signal lines BWIBT and PBYT are asserted to indicate the cycle to be performed: BWTBT | | PBYT Bus Cycle Type ——ac- ———————a—————————————————————- DATI or DATB H H DATIP DATO DATOB L H L H L L 2. When a valid address is decoded by a slave, it responds as a. The UBA asserts PUBMEM if the address UNIBUS b. PBCYC is follows: is UNIBUS memory, Or I/0 page. PMI memory and the CPU assert PSSEL. asserted. How the cycle proceeds is dependent on whether the CPU or UBA is master, and what the response was from the slave as follows: a. When the CPU is PMI master: If PSSEL is asserted and PUBMEM is negated, the CPU proceeds with a PMI memory cycle. If PSSEL is negated, the CPU performs a PMI cycle with the UBA responding as the PMI slave. b. When the UBA is PMI masters:, If PSSEL is negated, then the UBA aborts the PMI cycle and does not respond as UNIBUS slave. —n—————-—————u--—:—wn —————_-—-————-——————-————————c—a—:———qawuafln——— FUNCTIONAL DESCRIPTION 3.2.5 PMI Data Transfer Protocol , the data transfer Following the data transfer address cycle The transfer of data on the PMI can be grouped cycle begins. fer cycles: into 3 general types of PMI data trans 1. The Data In (DATI) and Data In Pause (DATIP) cycles. 2. The Block Data In (DATBI) cycle. 3. The Data Out (DATO) are used to read one or two words. This is used to to sixteen words. and Byte Out Data of read up (DATOB) . These data transfer cycles are used to write a single word or byte. The following subsections describe each These the cycles. 3.2.6 PMI Data In Cycles (DATI) and (DATIP) , a PMI master When accessing the PMI memory address space two words of data. or the DATI(P) cycle to read either Sonememor y, a UNIBU or page 1/0 r accessing eithe PMI master uses When reads single words only. DATI cycle except that The PMI DATIP cycle is identical to the that the next cycle indic to PBYT is asserted with the address cycle) ate be a data out will nt (immediately following the curre cycle to the same address. The following is a description of a DATI(P) data transfer cycle: PMI SLAVE PMI MASTER PBCYC is asserted during the address- ing phase of the cycle (described in subsection 3.2.4). 1. Data from the specified address is gated onto the bus. 2. If the slave is PMI memory, PHBPAR and PLBPAR are generated and gated onto the bus. Note: These parity bits are - e en wn e - — . S R I D M D GED D W wmp D = PR— ———-—--————--—-_-—_—---——_----———-----—- FUNCTIONAL DESCRIPTION PMI SLAVE _-_..._,..,_ PMI MASTER ______________ I._....__,_._......__._. ._._...-’_._ _____________________ generated only for memory locations which are cached on the CPU (i.e. PMI memory). 3, PRDSTB is asserted. 4, PRDSTB is negated. The first data word, with PHBPAR and PLBAR, are latched on the negating edge of PRDSTB. If only one word is to be read, PBCIC is negated and the cycle ends. If two words are to be read, PBCYC remains asserted. If a read-modifywrite (DATIP) is being performed a DATO cycle will take place here. The DATO(B) cycle is described in section 3.2.8. 6. The second data word 1is gated onto the bus 80 ns maximum after the negation of 7. PRDSTB. PHBPAR and PLBPAR are gener- ated for the second data word and are gated onto the bus 100 ns after the negation of PRDSTB. The second data word, along with PHBPAR and PLBAR, are received 145 ns maximum after the negating edge of PRDSTB. PBCYC is negated and the cycle ends. If a read-modify- write (DATIP) is being performed, PRCYC remains asserted, and a DATO cycle is performed here. The DATO(B) cycle is described in section 4.2.9. 9, Data is removed from the bus after receiving the negation of PBCYC. FUNCTIONAL 3.2.7 PMI Block When accessing PMI Block of data. accessing The PMI the Mode master following Cycles memory In address (DATBI) can start DATBI to data =zero 04-01 and are flow the master all equal to describes the a read PMI up terminates one. during the uses the sixteen words cycle when even word boundaries. 01 and 00 on This must the transfer when PMI SLAVE cycle. MASTER e asserted D e T T T S the specified addressing cycle. The addressing in section 3.2.4. 1. asserted. is master to transfers DATBI phase of the is described PBLKM space cycle master does -not use the DATBI the I/O page or UNIBUS memory. PMI ____________ B is In only and does not cross sixteen word at the transfer start, address bits both equal address bits PBCYC Data PMI Data A PMI either boundaries means that The Mode DESCRIPTION phase 2. Data from address bus. 3. If the Note: gated slave PHBPAR and is gated These is PLBPAR parity PMI are onto generated onto the memory, generated the bus. bits are memory locations which are cached on the CPU (i.e. PMI memory). 6. The first data and PLBAR, is negating edge word, with latched on of PRDSTB. only for 4, PRDSTB is asserted. 5. PRDSTB is negated data bus. is removed The second onto the and the from the word is PHBPAR the 7. after the data bus 80 ns negation gated maximum of PRDSTB. FUNCTIONAL DESCRIPTION PMI MASTER m ____________________________________ Jeemmm 8. PMI SLAVE e ———m———————————— PHBPAR and PLBPAR are generated for the second word and gated onto the bus 100 ns maximum after the negation of PRDSTB. 9. The second data word, with PHBPAR and PLBPAR, is received 145 ns maximum after the negating edge of PRDSTB. 10. PBLKM is negated after latching the second data word. 11. The second data word is removed from the bus after receiving the negation of PBLKM. 12. Data transfer cycles continue in the same manner as steps 1 thru 11 above until two words are left to be transferred. The last two words are trans- ferred with the same timing, but the signal PBLKM is not asserted by the master. 13. PBCYC is negated after latching the last data word. 14. Data is removed from the bus after recieving the negation of PBCYC. 3.2.8 PMI Data Out Cycles The PMI master uses the PMI Data Out (DATO or. DATOB) transfer a single word or byte to a PMI slave The following flow describes a DATO(B) cycle: cycles to FUNCTIONAL PMI MASTER PMI DESCRIPTION SLAVE PBCYC is asserted during the addressing phase of the cycle. The addressing phase is described in section 4.2.5. 1. PBCYC gated 2. PWTSTB 5. PWTSTB 6. PBCYC 3.3 is asserted and onto the bus. is is is MEMORY data is asserted. 3. The assertion received and latched in. of PWTSTB is the data is 4, PSBUFL is asserted. 7. PSBUFL is negated. negated. negated. MANAGEMENT Memory management is used to relocate a 16-bit virtual address, if necessary, and transmit the 22-bit physical address to cache memory, or PMI memory. Address modification is the PMI function of memory management. The modification of addresses is called relocation because it consists of adding a fixed constant to a virtual address to create a physical address. Memory management also allows the user to protect one section of memory from access by programs located in another section. Memory management divides memory into individual sections called pages. Each page has a protection or access key associated with it that With defines the type of access allowed on that particular page. the memory management unit, a page can be Kkeyed nonresident (memory neither readable nor writable), no write operations to memory, or read/write. These two types of protection, in association with other features, enable the user to develop a secure operating system. It is often desirable to load a program into one area of physical memory memory, and for execute it as example, if when it were 3-19 located several in another area of user programs are FUNCTIONAL DESCRIPTION When any one program 1is simultaneously stored 1in memory. processor as if it were the running, it must be accessed by g at 0. This process is beginnin s located in the set of addresse called relocation. When the processor accesses virtual bus address 0, a base address is added to it and the relocated 0 location of the program is accessed. Typically this base address is added to all references while the program is running. A different base address is used in memory. for each of the other programs Memory management specifies relocation on a page basis, which allows a large program to be loaded into nonadjacent pages in memory. This capability eliminates the need to shuffle programs It also minimizes unusable memory to accommodate a new one. to be loaded into a specific users more allowing fragments, thus memory size. A program and its data can-occupy as many as 16 pages 1in the The size of each page may vary and can be any multiple memory. This feature allows 1in length. of 32 words up to 4096 words small areas of memory to be protected (stacks, buffers, etc.), and also allows the last page of program, exceeding 4K words, be the of adequate to length to protect and relocate the remainder of program. As a result, fixed-length the memory fragmentation problem inherent with The base address of each page pages is eliminated. space, address can be any multiple of 32 words in the physical The variable page thus ensuring efficient use of PMI memory. run at changed dynamically length also allows the pages to be time. use for Memory management provides three separate sets of pages These modes. user and supervisor, kernel, processor's in the isolating sets of pages increase system protection by physically supervisor programs and the kernel from service user programs program. The service programs are also separated from the kernel program. time the reduce sets greatly register relocation Separate of sets three The mapping. necessary to switch context between that system operating an designing in registers also aid the user more 1is and modular, is communications, has clearly defined easily debugged and maintained. of The virtual bus address space is further divided, within each and user pages, into instruction space supervisor, kernel, the I space contains code, that is, and data space (I and D space). word that is part of the program such as instructions, index any that D space contains information words and immediate operands. 3-20 FUNCTIONAL DESCRIPTION can be modified, such as data buffers. cate data and management can reloaddr By using this feature, esmemory ess values. with separate base instruction referenc program of 64K words ible to have a user Therefore it is poss . consisting of 32K of instructions and 32K of data Address registers consists of s 48(PDRPage The memory management Page and s), Descriptor Register registers four Registers (PARs), 48 are s (MMRO-3) . These g subsections Memory Management Register owin located on the KDJ11l-BF module. The foll » describe each of the registers. 3.3.1 Page Address Registers Address (PARs) contain the 16-bit Page The Page Address Registers spe page. the of ess the base addr cifies Field (PAF). The PAR are s ster regi e -are read/write. Thes ruction. Thei (See figure 3-2.) All bits r inst ole start or a RESET not affected by iscons UNDEFINED. state at power up FIGURE 3-2 PAGE ADDRESS REGISTER FORMAT 3.3.2 Page Descriptor Registers on relative tain informatitro r Registers (PDR,s) con The Page Descripto l. These con access e length and ansion, pagect to page exp a RESET or rt sta console not aff edat by registers are The unused All . NED EFI UND power up is state instruction. zero ir is mat for er ist reg The n. not be writteare bits read asure 3-3;andbitcandes 3-2. provided in Table criptions show in Fig L BYPASS CACHE BEDDEERE PAGEYLENGTH _J FIELD PAGE TEN WRIT Accest|é:&maou. EXPANSION DIRECTION FIGURE 3-3 PAGE DESCRIPFTOR REGISTER FORMAT 3-21 FUNCTIONAL DESCRIPTION TABLE Bypass 3-2 Cache (R/W) 14:8 Page Length Field (R/W) PAGE DESCRIPTOR REGISTER BIT DESCRIPTIONS This bit implements a conditional cache bypass mechanism. If set, references to the selected virtual page will bypass the cache. A cache bypass causes the cache location to be invalidated whenever a read or write hit occurs. This field specifies the block numbe which defines the boundary of the current page. The block number of the virtual address is compared against the Page Length Field to detect length errors.’ An error occurs when expanding upwards if the block number Field, and is greater than the Page Length when expanding down if the block number is less than the Page Length Field. Page Written (RO) or not this page whether indicates bit This into) since written (i.e. has been modified (1 is loaded was PDR or PAR the either affirmative). It is useful in applications which involve disk swapping and memory which pages It is used to determine overlays. have been modified and hence must be saved in their new and which pages have not been modified and can simply be overlaid. This bit is reset to 0 whenever either the PDR or the associated PAR is written into. Expansion Direction (R/W) This bit specifies in which direction the page expands. If ED=0 the page expands upwards from block number 0 to include blocks with higher addresses; if ED=1, the page expands downwards from block number 127 to include blocks with lower addresses. Upward expansion is usually used for program space while downward expansion Access Control is used for stack space. This field contains the acces rights to this particular pace. The access codes or "keys" specify the manner in which a page may be accessed and whether or not a given access ———' --‘—----—---a-—--"“--——-—-—-—---‘-—-——--‘_--—-----—-—-———--v FUNCTIONAL TABLE - S G S G I D D I G GER TED CED IR NED GNP G D GND CHD GNE GED GED should G GED CED R D 00 01 10 11 Memory Memory Management Management M result operation. 3.3.3 3-2 D D in The (Cont) GER GED GER an D D GED R AED GED abort access GED GAS G GED the codes 0 GED R R D b W e e STID N WD SN GED NN D S e current are: Non-resident - abort Read only - abort on Not used - abort all Read/write Register Register GHD DESCRIPTION all accesses writes accesses 0 (MMRO), at address 17 777 572, contains error flags, the page number whose reference caused the abort, and various other status flags. MMRO is cleared at power up, by a console start, and by a RESET instruction. Figure 3-4 shows the register format; Table 3-3 contains the bit descriptions. ABORT: READ ONLY ABORT: PAGE LENGTH ABORT: NON-RESIDENT PROCESSOR MOOE PAGE SPACE PAGE NUMBER ENABLE RELOCATION FIGURE 3-4 MEMORY MANAGEMENT REGISTER 0 FORMAT FUNCTIONAL DESCRIPTION TABLE BIT 15 14 13 3-3 MEMORY MANAGEMENT REGISTER NAME Abort Non Resident (R/WO O BIT DESCRIPTIONS FUNCTION Bit 15 a page is set by attempting to access with an Access Control Field key egqual to 0 or 2. It is also set by attempting to use memory relocation with a mode (PS<15:14>) of 2. Abort Page Length This bit is set a location in a (R/W) number by attempting to access page with a block <12:6>) (virtual address bits that is outside the area authorized by the Page Length Field of the Page Descriptor Register for that page. Abort - This Read write in Only" pages Onlv (R/W) bit is a set by attempting "Read Onlv" page. have access keys to "Read- of 1. NOTE: Bits <15:13> can be set by an explicit write; however such an action does not cause an abort. Whether set explicitly or by an abort, bits <«15:13> cause memory management to freeze the contents of MMRO <6:1>, MMR1l, and MMR2. The status registers remain frozen until MMRO <15:13> are cleared by an explict write or any initialization sequence. 6:5 Processor Mode (RO) 4 Page Space (RO) These bits indicate the processor mode (kernel/supervisor/user/illegal) associated with the page causing the abort (kernel = 00, supervisor = 01, user = 11, illegal = 10). 1If the illegal mode is specified, an abort is ¢generated and bit <15> is set. This D) abort Number bit indicates associated (0 = I 3:1 Page (RO) 0 Enable This Relocation (R/W) set to 1, all When bit 0 is the address with the page space, 1 These three bits the page causing bit allows = D (I page number relocation. When addresses are relocated. set to 0, memory manage- ment is inoperative not relocated. and or the space). contain the the abort. address space causing addresses are of FUNCTIONAL DESCRIPTION 3.3.4 Memory Management Register Memory Management Register 1 1 (MMR1l) at address 17777574 records any auto increment or decrement of the general purpose-registers. This register supplies necessary information needed to recover from a memory management abort. MMR1l is read only. Its state at Figure 3-5 shows the register format. power up is UNDEFINED. VIRTUALADDRESS FIGURE 3.3.5 Memory 3-5 MEMORY. MANAGEMENT REGISTER Management Register L Figure 3-6 shows L T — AMOUNT CHANGED {2'S COMPLEMENT) : the S~ FORMAT 2 Memory Management Register 2 (MMR2), at loaded with the virtual address at instruction fetch. MMR2 is read only. UNDEFINED. 1 address 17 777 576, |is the beginning of each Its state at power up is register b\ REGISTER NUMBER format. e AMOUNT CHANGED {7'S COMPLEMENT) Bl ~~— _J REGISTER NUMBER FIGURE 3-6 MEMORY MANAGEMENT REGISTER 2 FORMAT 3.3.6 Memory Management Register 3 772 516, 17 at address (MMR3), Memory Management Register 3 enables or disables D space, 22-bit mapping, the CSM instruction, and the I/0 map (when applicable). MMR3 is cleared at power up, Figure 3-7 shows by a console start, and by a RESET instruction. s. description the register format; Table 3-4 provides the bit FUNCTIONAL DESCRIPTION Y ot ENABLE UNIBUS MAP ENABLE 22-81T MAPPING ENABLE CSM INSTRUCTION ENABLE KERNEL DATA SPACE ENABLE SUPERVISOR DATA SPACE ENABLE USER DATA SPACE FIGURE 3-7 MEMORY MANAGEMENT REGISTER 3 FORMAT TABLE 3-4 MEMORY MANAGEMENT REGISTER 3 BIT DESCRIPTIONS NAME BIT(S) 15:06 Unused. 2 ===== 5 Enable UNIBUS Map (R/W) 4 Enable 22-bit Mapping (R/W) This bit enables the I/O Map for the UNIBUS Adapter. This bit, when set, selects 22-bit memory addressing. When this bit is clear, 18-bit addressing is selected. (18 or 22-bit addressing is actually enabled only when MMRO bit 0 3 Enable CSM Instruction is set). This bit enables recognition of the Call Supervisor Mode instruction. (R/W) 2:0 Enable Data Space (R/W) These three bits enable Data Space mapping for kernel, supervisor, and user mode, respectively. 3.3.7 Physical Address Construction If UNIBUS map relocation is enabled (MMR3 bit 05 = 1), UNIBUS er pairs address bits <17:13> select one of 31 mapping regist of the t conten (corresponding to octal codes 00 thru 36)., The address UNIBUS to selected mapping register pair |is added bits<12:00> to produce the memory address. If UNIBUS address FUNCTIONAL DESCRIPTION 37), the I/O Page is code (octal ones all are <17:13> bits =zeros,, all Memory address bits <21:18> are set equal selected. bits address UNIBUS to memory address bits <17:00> are identical <17:00>, and BBS7 is asserted. See Figure 3-8. 17 13 UNIBUS ADORESS [ 12 01 00 W | ADDRESS 8ITS 17-13 SELECT ONE OF 31 MAPPING REGISTER PAIRS 21 186 15 : 00 THROUGH 0t t 36 (OCTAL} 00 = ADDER 21 L FIGURE 3.3.8 Memory 3-8 1] 01 PHYSICAL ADDRESS 00 INTERPRETATION Relocation When memory management is enabled, the normal 16-bit direct-byte address is no longer interpreted as a direct physical address (PA) but as a virtual bus address (VBA) containing information to be used in constructing a new 22-bit PA. The information contained in the VBA is combined with relocation information contained in the page address register (PAR) to make a 22-bit PA. Using memory management, memory can be dynamically allocated 1in pages composed of from 1 to 128 blocks of 32 words each. The starting PA for each page is a multiple of 32 words, and each page has a maximum size of 4096 words. Pages may be located anywhere within the PA space. The set of 16 PARs to be used to create the PA is determined by the current mode of operation of the CPU (i.e., kernel, supervisor, or user modes ; ref subsection 3-10). 3.4 KDJ11-BF CACHE The CPU has dual concurrent DMA tag cache memories. They are used to allow a activity, and decrease system access time of instructions and data. The 8KB cache 3-27 is located on the KDJ1l1l-B FUNCTIONAL DESCRIPTION module. UNIBUS memory is not cached. for only occur operations Cache PMI memory cycles, Figure 3-9 is a matrix showing the cache response for both DMA a CPU data transfers from and to the PMI memory space. There are The two cache memory tags referenced in the matrix. heading refers to DMA activity. PMI activity DMA matrix The CPU matrix heading refers to involving the CPU tag. DMA Wt READ WARITE WOROD READ MEMORY INVALIDATE CACRHE-WRITE MEMORY WRITE | INVAUIDATE cru READ MEMORY N/A WRITE N/A N/A WRITE FORCE wss READ MEMORY WRITE MEMORYINVALIDATE CACHE WRITE BCITH CACHE AND WRITE BOTH | mremORY MEMORY MEMORY T MEMORY N/A READ CACHE WRITE READ BYPASS FORCE MISS AND ALLOCATE OATA WRITE BOTH CACHE AND CACHE-WRITE BYPASS READ CACHED | READ MENMORY WRITE MEMORY 8YTE MEMORY miss Wit MISS CACHE AND INO CACHE INVALIOATE READ MEMORY INVALIDATE RaCHE CHANGE MEMORY -NO CACHE CACHE AND READ MEMORY | CHANGE CACHE AND WRITE WRITE MEMORY | READ MEMORY whiTe MEMORY 8 oha ol READ MEMORY | READ MEMORY -NO CACHE CHANGE WRITE MEMORY ~NO CACMHE CHANGE -NO CACHE CHANGE WRITE MEMORY ~NO CACHE CHANG: FIGURE 3-9 CACHE RESPONSE MATRIX Cache Parity Errors affect the Cache Response Matrix in the following ways: 1. Error forces a During DMA Write Cycles, a DMA Tag Parity is invalidated. location Cache Hit response, and the cache 2. During CPU Read Cycles 3. During CPU Write Byte Cycles (Non-Bypass; Non-Force Miss), a CPU Tag Parity Error forces a Cache Hit response, but the (Non-Bypass), Parity Error forces a Cache Miss data is loaded with bad parity. a CPU response. Tag or Data FUNCTIONAL DESCRIPTION 4, During CPU Read Bypass or Write Bypass Cycles, a Data Parity Error forces a Cache Hit Response. location is invalidated. 5, For all Force Miss Cycles, (Non-Bypass) 3.4.1 KDJ11-BF Cycle, Cache The KDJ11-BF cache KDJ11-BF addresses Cache and Parity for is the CPU CPU Tag or The cache Write ignored. Word Operations 1is PMI 1initially flushed memory locations, (emptied). As the the KDJ1ll-BF cache begins to fill with addresses and data. If the KDJ11l-BF addresses PMI memory within an 8K memory space the cache fills, as the addresses are accessed, to the 8K limit of the cache size. This means that for each cache address location the data for that address is a duplicate of the corresponding PMI address's data. As each instruction is accessed by the KDJ1ll-BF it's address |is compared 1in <cache to see if there are any matches. If a match occurs the PMI memory cycle is aborted and the data stored in the cache address 1is wused by the KDJ11-BF. 1If a cache miss occurs the PMI memory cycle is completed. In addition to filling it's cache monitors the PMI address lines for DMA the space memory writes address to PMI memory addresses. If a write into a PMI memory matches a cache address that particular cache address is invalidated. The following cache options are available on the KDJ11l-BF: l. Conditional made to cache bypass - selected virtual pages can be Bit <15> in the PDRs sets this bypass the cache. condition. 2. Unconditional cache bypass - all CPU references to bypass the cache. sets this condition. Bit <9> 3. Flush cache - 4, Lock instruction (ASRB, TSTSET, and WRTCLK) instructions guarantee a cache bypass reference. 3.4.2 KDJ11-BF Cache all Valid bits can be made in the Cache Control register in the cache are cleared. - these Organization TAG dual The KDJ11-BF contains an 8K byte direct map cache with The Store which allows concurrent operations of the CPU and DMA. sections: distinct three 1into subdivided Cache Memory can be Data Store, CPU TAG Store and the DMA TAG Store. FUNCTIONAL DESCRIPTION The KDJ11-BF Cache interprets the CPU (or DMA) shown as description. Figure in 3-10 and Table 3-5 physical contains address bit the || BN CACHE INDEX CACHE TAG BYTE SELECTION FIGURE 3-10 CPU/DMA PHYSICAL ADDRESS INTERPRETATION REGISTER TABLE 3-5 CPU/DMA PHYSICAL ADDRESS INTERPRETATION BIT DESCRIPTIONS S D D SED G 21:13 CED D WD D D D Cache (R/W) D D D CHD D e Tag - 1. During CPU read/write operations, these bits are compared with bits 21:13 of the CPU Cache Tag Register (Figure 3-11) to determine the cache hit/miss status. 2. During DMA read/write operations, these bits are compared with bits 21:13 of the DMA Tag Register (Figure 3-12) to determine the cache hit/miss status. For either CPU or DMA operations a tag hit occurs when the cache tag contents matches the CPU/DMA tag register and the CPU/DMA valid bit 12:01 00 Cache (R/W) Index Byte Selection The High Byte <15:08>. is set. The CPU cache interprets the CPU/DMA physical address directly and selects one of 4096 word cache memory locations. During CPU/DMA write operatons setting this bit selects writing into the high byte cache memory location (Figure 3-13). Parity bit reflects 3-30 odd parity on data bits FUNCTIONAL The Low Byte Parity bit reflects even parity DESCRIPTION on data bits <07:00>, The CPU Tag Parity bit reflects odd parity on CPU Tag bits The DMA Tag Parity bit reflects odd parity on DMA Tag bits <21:13>. <21:13>. The CPU and DMA Tag Valid bits are not DMA Tag parity | 27 20 included in 6 5 the CPU and calculations. 19 18 16 17 1A 1 13 14 15 ? 8 9 10 0 1 2 3 4 [ T T T [ Jelofofofofelolefofofefefe] TAG ¢PU CPU TAG PARITY (ODD) CPU TAG VAIID FIGURE 3-11 CPU CACHE TAG REGISTER FORMAT 27 20 19 18 17 16 15 14 13 12 11 10 9 8 7 olololo]olo 6 0 5 4 olo DMA TAG OMA TAG PARITY (ODO) DMA TAG VALID FIGURE 3-12 DMA TAG REGISTER FORMAT 3 2 1 o]olo O o] FUNCTIONAL DESCRIPTION S — LOW BYTE HIGH BYTE DATA DATA HIGH BYTE PARITY {ODDY LOwW BYTE . PARITY (EVEN) FIGURE 3-13 CPU CACHE DATA ORGANIZATION 3.4.3 Cache Control Register (CCR), at address 17 777 746, controls Two copies of the cache control the operation of the cache. register are kept on the KDJ11-BF. One copy is kept in the chip The chip copy implements bits <10:0> set:the other on the board. as read/write but interprets only bits <9 and 3:2>. This copy is The board used as the source of data when the register is read. copy. write-only ‘a is This >. <10,8,7,6,0 bits copy implements shows the 3=14 Figure accessed. be explicitly can not It The Cache Control Register register format; Table 3-6 contains the bit descriptions. 0 o) 0 0 0 H J WRITE WRONG TAG PARITY UNCONDITIONAL T CACHE BYPASS FLUSH CACHE ENABLE PARITY ERROR ABORT WRITE WRONG DATA PARITY UNINTERPRETED FORCE CACHE MISS ~ DIAGNOSTIC MODE DISABLE CACHE PARITY INTERRUPT FIGURE 3-14 CACHE CONTROL REGISTER FORMAT FUNCTIONAL DESCRIPTION TABLE BIT(S) 3-6 CACHE CONTROL REGISTER BIT DESCRIPTIONS FUNCTION NAME —--——————— ——u—————--—---—-—--——--—--s--———-d———_—-—————-——---—-———-h-— Always 15:11 Unused 10 Write Wrong Tag Parity (R/W) read cleared bits. as When this bit is set, the CPU and Tag Parity bits are both written as wrong parity during all operations which update these bits. A cache tag parity error will thus occur on to that location. 09 Unconditional cache bypass (R/W) 08 Flush Cache (WO) 07 Parity Error Abort (R/W) When this bit Write Wrong Data Parity (R/W) is will not affect set, all the next access references to memory by the CPU will bypass the cache and go directly to main memory. Read or write hits will result in the invalidation of the corresponding cache location; misses Writing a "1" the into cache contents. this bit clears all CPU Tag and DMA Tag Valid bits invalidating the entire contents of the cache. Writing a "0" into this bit has no effect. Flush Cache always reads as zero. The KDJ1l1l-BF requires approximately 1 msec to flush the cache. During the period, DMA activity is possible and CPU activity is suspended. This bit is set for diagnostic purposes only. When it is set, a cache parity error (during a CPU cache read) will cause the CPU to abort the current instruction and trap to parity error vector 114. When this bit is clear, a cache parity error (during a CPU cache read) results in a force miss and data fetch from main memory. The CPU will trap to 114 only if CCR bit <0> is clear. DMA cycle cache parity errors will cause a trap to 114 if CCR <7> is set or if CCR <0> is clear. CCR <7> has no effect on main memory parity errors which always cause the CPU to abort the current instruction and 06 DMA trap to 114. When this bit is set, both the high and low data parity bits are written with wrong parity during all operations which update these bits. This will cause a cache data parity error to occur on the next access to that location. FUNCTIONAL DESCRIPTION TABLE 3-6 (Cont) FUNCTION BIT(S) NAME 03:02 Force Miss When either of these bits is set, CPU reads 01 Diagnostic Mode (R/W) When this bit is set, a 10 usec nonexistant memory timeout during a word write will not (R/W) will be reported as cache misses. cause a nonexistant memory trap and will not set CPU Error Register bit 05. All non-bypass and non-forced miss word writes will allocate the cache irregardless of the nonexistant memory timeout. Disable Cache This bit controls Cache Parity Interrupts when CCR <7> is clear (normal operation). If CCR <7> Parity is clear, a cache parity error (during a CPU Interrupt 00 cache read) results in a force miss and data (R/W) fetch from main memory. The CPU will trap to 114 only if CCR bit <0> is clear. DMA cycle cache parity errors will cause a trap to 114 if CCR <7> is set or if CCR <0> is clear. Table 3-7 summarizes the effect of CCR <7,0> parity on errors during CPU cache reads. Table 3-7 CACHE PARITY ERRORS DURING CPU CYCLES 0 0 Cache Miss and Update Cache; 0 1 Cache Miss and Update cache; 1 X Abort Instruction and Trap to ll4. Interrupt No to 114. Interrupt. Table 3-8 summarizes the effect of CCR <7,0> on errors during DMA writes. DMA Tag Parity FUNCTIONAL 3-8 TABLE PARITY ERRORS DURING DMA CYCLES CACHE Result of Cache Parity Error CCR<O> CCR<7> DESCRIPTION 0 0 Interrupt 0 1 No 1 X Trap The Cache Control Register to 114. Interrupt. to 114. is cleared by the negation of DCOK and The Jl1 ODT It is-not affected by BUS INIT. by a console start. cleared. be to CCR and flushed be to cache command G will cause 3.4.4 Memory System Error Register 777 744, The Memory System Error Register (MSER), at address 17 MSER reflects the status of cache and main memory parity errors. bits 14 and 13 are used by the KDJ11-BF Boot and Diagnostic Figure 3-15 shows the to test the Cache DMA Tag Store. programs register format; Table 3-9 contains the bit descriptions. 15 14 ‘ 13 } 12 1" 10 9 8 0 0 0 o) o 7 6 5 4 3 2 1 0 0 0 0 0 DTS PAR OTS CmP CPU ABORT CACHE HB DATA PARITY ERROR CACHE LB DATA PARITY ERROR CACHE CPU TAG PARITY ERROR CACHE DMA TAG PARITY ERROR FIGURE 3-15 MEMORY SYSTEM ERROR REGISTER FORMAT FUNCTIONAL DESCRIPTION TABLE 3-9 MEMORY SYSTEM ERROR REGISTER BIT DESCRIPTIONS - CPU G P I D G . S W SIS S D D GER GED G RO ADD WD G GED GED CEO WD OO N D NP b G N e This bit is set if a cache or main memory parity error results in an instruction Abort (RO) abort (i.e. only during the demand read cycle). Cache parity errors cause an abort only if CCR <7> is set. Main memory parity errors always DMA 14 Tag Store Comparator (DTS CMP) (RO) In Stand-Alone Mode (BCSR <8> set), this bit indicates the output of the Cache DMA Tag Store Comparator for the previous non-I/0 Page reference with cache miss. When BCSR <8> is clear, DTS CMP reads as a "O". DMA Tag 13 cause an abort. In Stand-Alone Mode (BCSR <8> set), this bit indicates the output of the DMA Tag Store Store Parity (DTS (RO) non—-I1/0 Parity Check Logic. for the previous PAR) Page Reference with cache miss. When BCSR <8> is clear, DTS PAR reads as a "0". These bits always read as 12-08 Unused 07 Cache HB Data Parity Error (R/W) "0". This bit is set if a parity error is detected in the high data byte during a CPU cache read. If CCR <7> is clear, MSER <7> is also set by a low byte parity error and by the set condition of MSER <5> or <4>. Cache 0é Parity LB Data Error (RO) high byte parity error and by the set condition MSER bits <5> or <4>. Cache CPU Tag Parity Error 05 (RO) m o - - e This bit is set if a parity error is detected in the CPU tag field during a CPU cache read. If CCR <7> is clear, MSER <7> is also set by a high or low data byte parity error. Note: - This bit is set if a parity error is detected in the low data byte during a CPU cache read. If CCR <7> is clear, MSER <6> is also set by a - - Cache parity errors are ignored (do not affect MSER <7-5>), if either CCR <3> or <2> (Force Miss) is set or if the CPU Tag Valid bit is clear. D € wH e oD e D o - - G D D SR G D D D WD WD D S TED D S W D N WD N - D D R I D A GED D D S S T O ) S — 0 FUNCTIONAL TABLE 3-9 BIT (Cont) NAME 04 FUNCTION Cache DMA Parity Tag This Error in (RO) bit the is DMA set tag a parity error is field is during a write Cache MSER is parity <4>), set or Unused errors if if are either the These DMA bits ignored CCR <3> or Tag Valid always read zero. parity occur result in depending on 1If CCR 2. trap 1If CCR during abort the following condition <7> (Parity error causes set MSER <15> to which instruction thru <7> Error If CCR <7> error is of Abort) location clear, clear, causes and the and CPU if DMA tag field cause a trap to parity location if to relevant error bits MSER vector location 114. Cache CPU and/or CCR CCR force trap to CCR bits <7> set, <0> is a access may location 114, and <0>: cache parity instruction, to MSER <7: 5> and also force a and to <0> a <7:5>. is which 114 CCR set, is a miss CPU occur <K7> clear, a cache miss, trap thru cache The errors if abort the thru vector Cache a 1is Miss) clear. 114. parity error causes the CPU to relevant error bits MSER <7:5> location 114. 3. detected affect CPU to to trap the CPU to abort the current and the relevant error bit(s) vector is a (Force as Cache errors not is always cause the set MSER <15> and an (do <2> bit Main memory parity errors current instruction, to location 114. l. DMA operation. Note: 03:00 DESCRIPTION cache and does to not during set or by any cache set the vector parity set trap a DMA if CCR the thru cycle <0> is clear. The Memory reference. It is System It is unaffected Error also by a Register cleared RESET on is cleared power instruction. up or by a MSER write console start. FUNCTIONAL DESCRIPTION 3.4.5 Hit/Miss Register This register, at address 17 777 752, indicates whether the six most recent CPU memory references resulted in cache hits or cache misses. The Hit/Miss Register is read only. Its value at power The Hit/Miss Register 1is not affected by up 1s UNDEFINED. The Hit/ Miss Register console start or a RESET instruction. will always read zero when the CPU is in console ODT mode. Figure 3-16 illustrates the register format; Table 3-10 contains the bit descriptions. HIT/MISS REGISTER FORMAT FIGURE 3-16 - TABLE 3-10 HIT/MISS REGISTER BIT DESCRIPTIONS read as zeros. 15:06 Unused Always 05:00 Cache Hit Bits enter from the right (at bit <0>) and are shifted leftward. A set bit indicates a cache hit, a cleared bit indicates a cache miss. 3.5 ADDITIONAL CPU REGISTER DESCRIPTIONS The general CPU Registers include: Two sets of six working registers (RO-R5) Kernel/supervisor/user stack pointers (R6) Program counter (R7). t W Other major registers are described in the following subsections. 38 FUNCTIONAL DESCRIPTION 3.5.1 Processor Status Word 17 777 776, contains The Processor Status Word (PS), at location The PSW 1is on the status of the processor. CONFI information GURATION initialized at power up (depends on . EE ThePROM uction instr RESET options) and is cleared at console start and ter regis the s trate does not affect the PS., Figure 3-17 illus Table 3-11 contains the bit descriptions. 15 14 13 12 1" 10 9 0 0 7 8 6 5 3 4 1 2 0 CARRY PRIORITY SUSPENDED LEVEL INSTRUCTION OVERFLOW - ZERO SET REGISTER| NEGATIVE TRACE TRAP MANAGEMENT MODE PREVIOUS MEMORY CURRENT MEMORY MANAGEMENT MODE . FIGURE 3-17 PROCESSOR STATUS WORD REGISTER TABLE 3-11 PROCESSOR STATUS WORD BIT DESCRIPTIONS BIT NAME 15514 Current Mode (R/W, protected) FUNCTION Current processor mode: 00 = kernel 01 = supervisor 11 = 10 = illegal (traps) user. 13:12 Previous Mode Previous processor mode, same 11 Register Set General register set select: (R/W, protected) (R/W, protected) encoding as current mode. 0 = register set O 1 = register set 1. D G P WD S S WP S GE G D D D IR G WD WD WP W W S W - D D AN G W S WD R SR NP GUY GED AU I W D ) S DW - ED NN A SR G P WS WD FUNCTIONAL TABLE DESCRIPTION 3-11 (Cont) 8 Suspended Instruction(R/W) Reserved 7:5 Priority (R/W, protected) Processor level. 4 Trace Trap Set (R/W, protected) 3:0 Condition Codes 3.5.2 Program The Interrupt Request implements a software 772, Request Program interrupt location request 240, responsibility to exiting. PIRQ is future interrupt force a trace Processor condition (R/W) Interrupt to for use. priority trap. codes. Register Register (PIRQ), interrupt at location facility. 1is When 17 777 program granted, the processor traps through is the interrupt service routine's clear the appropriate bit in PIRQ before cleared at power-up, by a console start and by It the RESET instruction. Table 3-12 contains the Figure bit 3-18 illustrates descriptions. | . PIR7 ) PRIORITY ENCODED VALUE OF BITS<15:9> L the register J PRIORITY ENCODED VALUE OF BITS<15:9> PIR6 PIRS PiR4 PR3 PIR2 PR FIGURE a 3-18 PROGRAM INTERRUPT 3-40 REQUEST REGISTER and FUNCTIONAL DESCRIPTION TABLE 3-12 PROGRAM INTERRUPT REQUEST REGISTER BIT DESCRIPTIONS 15:09 Each bit, when set, provides one of PIR 7-1 seven levels of software interrupt corresponding to interrupt priority levels 7 through 1. 08 07:05 Unused. These three bits are set by the CPU to the encoded value of the highest pending interrupt request (bits 15:09). Piority encoded value of bits <15:09> 04 03:00 Unused. Piority encoded The function of these bits is value of bits to bits 07:05. identical <15:09> 3.5.3 CPU Error Register The Error Register, at address 17 777 766, identifies the source of any abort or trap that caused a trap through location 4. The It is also CPU Error Register is cleared when it is written. It is unaffected by a cleared at power up or by console start. RESET instruction. Figure 3-19 shows the register format; Table 3-13 contains the bit descriptions. 0 0 0 0 0 0 0 0 0 ILLEGAL HALT ——‘ ADDRESS ERROR NONEXISTENT MEMORY 1/Q BUS TIMEQUT YELLOW STACK VIOLATION RED STACK VIOLATION FIGURE 3-19 CPU ERROR REGISTER FORMAT 3-41 0 DESCRIPTION FUNCTIONAL TABLE 3-13 CPU ERROR REGISTER BIT DESCRIPTIONS Illegal HALT Set when execution of a HALT instruction is attempted in user or supervisor mode. Address Error Set when word access to an odd byte (RO) address or an instruction fetch from an Non-existent Set when a reference to main memory Memory times internal register is attempted. (RO) out. - Set when a reference to the I/O page I/0 Bus Timeout times out. (RO) Set on a yellow zone stack overflow Yellow Stack Violation trap. (RO) Set on a red zone stack overflow trap. Red Stack Violation 3.5.4 CONFIGURATION AND DISPLAY REGISTER Configuration The read-only Boot and Diagnostic Register reflects the status of the eight edge-mounted switches at the top All of the switches are also routed KDJ11-BF module to allow them to be asserted of the KDJ11-BF module. on connectors the Switches 1-8 control register bits 7-0 respectively. remotely. NOTE All eight and restricted OFF, switches on setting to the any special of KDJ11l-BF these applications. module switches Refer Appendix H for a more detailed description. are 1is to Data bits 2-0 (switches 6-8) are also connected directly three baud module to select the baud rate. the baud purpose. to rate select lines of the console SLU on the KDJ11-BF These three bits always control console, and are not used for any other the rate Dbaud the normally OFF to allow are Switches 6-8 3-42 FUNCTIONAL DESCRIPTION select switch on the Console Serial Line board (rear of box or cabinet) to select the baud rate. This eliminates the need to gain access to the CPU module to select the baud rate. The only time switches 6-8 would be used to select the baud is if the baud rate switch was not present. for additional information. rate Refer to-Appendix H Data bits 7-3 (switches 1-5) are read by the ROM code to define by the ROM code after power up or some of the actions to be taken a detail description of each of for H restart. Refer to Appendix these bits. Bits 15-8 are always read The value of bits 7-0 depends on the position of the as 2zeros. switches on the CPU module, and any external switch which might Figure 3-20 shows the register format. be connected. X = don't care FIGURE 3-20 BOOT AND DIAGNOSTIC CONFIGURATION REGISTER FORMAT The write-only Boot and Diagnostic Display Register (BDR), at address 17 777 524, allows the Boot Diagnostic programs to light the front panel Start-Up Test LED display and the LEDs on the These display bits are also available on an KDJ11-B module. Bits 05-00 are cleared on power up (all LEDs external connector. Figure 3-21 shows the register negation of DCOK. the by on) format; Table 3-14 contains the bit descriptions. 18 14 13 12 n 10 8 9 7 6 s 4 3 8-817T SWITCH PACK X = dont't care FIGURE 3-21 DISPLAY 3-43 REGISTER 2 1 0 FUNCTIONAL DESCRIPTION TABLE 3-14 DISPLAY REGISTER BIT DESCRIPTIONS 15::06 =—-- Unused 05::00 LED 5-0 These bits enable the boot and diagnostic programs to light the LEDs located at the top of the CPU module. Clearing any of these bits lights the corresponding LED. 3.5.5 Maintenance Register The Maintenance Register, at address 17 777 750, accesses the 16-bit word read by the Jl1 Chip Set (through GPO Code) test BPOK H, read the power up option code, read the halt/trap option bit. Other bits in the maintenance register, not used by Jll microcode, contain information on the module type and system parameters useful to operating system and diagnostic software. Figure 3-22 shows the register format; Table 3-15 contains the bit descriptions. The power up option code is hard wired operation (code 2). The PSW 1is set begins program execution at address for standard bootstrap to 340 and the processor 173000. The Boot and Diagnostic Code, which starts at that location, configures the KDJ11-B and runs stand-alone diagnostics before acting on the user specified power up option stored as part of the EEPROM Configuration Data. Because the J1ll1 Microcode never sees power up option code 3, selecting a start location specified by register bits <15:09>, these Maintenance Register bits are used to specify system parameters. 15 14 13 12 1" 0 0 0 ¢ Q 10 9 8 ? 6 5 4 0 0 1 0 3 2 1 1 ] RESERVED UNIBUS SYSTEM FPA AVAILABLE MODULE TYPE (FIXED) HALT/TRAP OPTION POWER UP OPTION (FIXED) BPOK H FIGURE 3-22 MAINTENANCE 3-44 REGISTER FORMAT 0 FUNCTIONAL DESCRIPTION TABLE 3-15 MAINTENANCE REGISTER BIT DESCRIPTIONS - AP TP GED GHD GED P GED D D GED GRS WED WD Reserved for future expansion. 15:11 Unused 10 Unused Reserved 09 UNIBUS This bit reflects the status of the extrenally applied UNIBUS Adapter Line. A "1" indicates System 08 Read as zeros. for future use. (RO) that the system includes a UNIBUS Adapter. FPA When set, this bit indicates that the FPA is Available available for use. (RO) 07:04 Module Type 03 Halt/Trap 02-01 Power Up (R/W) Code This 4-bit code is hard-wired as a n2v, indicating a KDJ11-BF Module. This read/write bit determines the response of a processor to a Kernel Mode Halt instruction. Setting the bit selects the Trap Option, causing the CPU to trap to location 4. Clearing the bit selects the Halt Option, causing the CPU to halt and enter ODT. This bit is cleared by the negation of DCOK and is set by the Boot and Diagnostic ROM code if the Trap Option is selected by a bit in the Configuration RAM. The Trap Option is not intended for normal use and is reserved for controller applications. This 2-bit code is hard-wired as a "2". At power up, the processor sets the PC to 173000 and sets the PSW to 370. It then starts program execution at location 173000, which is the starting location for the KDJ11-BF Boot and Diagnostic ROM program. These programs test out the KDJ11-BF Module and then implement the user selected power up option specified in the Configuration Data. This bit is set (1) if the PMI BUS signal BPOK H is asserted, indicating that AC Power is okay. FUNCTIONAL DESCRIPTION 3.5.6 Boot The and Diagnostic Controller 17 777 520, is both word the register format; Table and Diagnostic Controller Boot address 3-23 descriptions. programs PMG test from the BCSR response 17765776 alone also of read/write Halt backup on Status and the Break set Counter feature Boot these and programs Diagnostic to access the I/O parameters, to control access ¢¢BCSR), addressable. bit ROM parameters the for to for line clock, enter or at disable the addresses 17765000 - 17773776 and to control Page can use the BCSR to alter PMG enable or disable the Halt on Break to the ROM and and EEPROM memories. PMG CNTY FRC LCIE PMG CNT2 OIS LKS CLK SEL 1 RS3 WE CLK SELO RS3 85 ENB HOB . DIS 65 SA MOODE 3-23 oIS 73 BOOT AND to exit selectively NOT USED FIGURE at Figure contains the and Diagnostic and and ROM's at addresses 17773000 access to the EEPROM memory. Programs which and line clock feature and to Register byte 3-16 Boot status, Grant) Register mode. allows the and/or allows Mastership Console stand BCSR battery (Processor enable The to The Status DIAGNOSTIC CONTROLLER REGISTER FORMAT FUNCTIONAL TABLE BOOT AND DIAGNOSTIC DESCRIPTION 3-16 CONTROLLER REGISTER BIT DESCRIPTIONS Battery Backup Reboot Enable. When set, this bit indicates that batterv backup failed to maintain voltages to the memory system during the previous power failure. When this bit is clear, it indicates that the system does not feature battery backup, or that battery backup maintained voltages during the previous power failure. This signal is received from backplane pin BH1l and latched when DC OK is "1" asserted. or "O". 14 Not used Could be a 13 FRC LCIE Force Line Clock Interrupt Enable. If this bit is set, assertion of the signal selected by BCSR <11,10> (Clock Select Bits 1 and 0) will unconditionally request interrupts. If FRC LCIE is clear, assertion of the selected signal will request interrupts only if the Line Clock Status Register bit <6> (LCIE) is set under program control. FRC LCIE is the cleared by 12 DIS LKS negation of DCOK. Line Clock Status Register Disable. this bit is set, the Line Clock If Status Register (LKS) is disabled. If this bit is clear, LKS is enabled and responds to bus address 17777546. LKS DIS is cleared by the negation of DCOK. 11 10 CLK SEL1 CLK SELZ2 Clock Select Bits 1 and 0. These two bits select the source of the line clock interrupt request: CLK SEL1 CLK SELO Source of Interrupt 0 1 0 1 External On-Board On-Board On-Board LTC Line 50 Hz 60 Hz 800 Hz FUNCTIONAL DESCRIPTION TABLE 3-16 ENB HOB (R/W) (Cont) Enable Halt on Break. When this bit is set, the Console Serial Line Unit Halt on Break feature is enabled. When this bit is clear the feature is disabled. ENB HOB is cleared by the negation of DCOK. 08 SA MODE (R/W) Stand-Alone Mode. When this bit is set, the KDJ11-B operates in stand-alone mode, using its Cache as main memory. External memory and peripherals are all disabled. When SA MODE is £, enabling clear, Stand-Alone Mode is turned of is set MODE SA s. heral perip and external memory by the negation of DCOK. 07 DIS 73 (R/W) Disable 17773000. When this bit is set, response of the 16-bit ROM memory to addresses 17773000 - 17773776 is disabled, allowing the operation of an external ROM that uses those addresses. When DIS 73 is clear, the 16-bit ROMs respond to those addresses, using the high byte of the page controcl register as the most significant address bits. DIS 73 is cleared by the negation of DCOK. 06 DIS 65 (R/W) Disable 17765000. When this bit is set, response of the Boot and Diagnostic 16-bit and 8-bit ROM memory to addresseS 17765000 17765776 is disabled, allows the operation of external ROM which uses those addresses. When DIS 65 is clear, the ROM memory selected by BCSR <5> responds to those addresses, using the low byte of the page control register as the most significant address bits. DIS 65 is cleared by the negation of DCOK. 05 RS3 65 (R/W) ROM Socket 3 at 17765000. This bit selects whether there is a 16-bit ROM in ROM sockets one and two or there is an 8-bit ROM in ROM socket three responds to addresses 17765000 17765776 (assuming that BCSR <4> is clear). If RS3 65 is set, the 8-bit ROM is selected. If RS3 65 is clear the 16-bit ROM is selected. In either case, the low byte of the page control register provides the most significant address bits. RS3 65 is cleared by the negation of DCOK. 3-48 FUNCTIONAL TABLE D GNP I SED P W T D D VED D WD O S (Cont) IR GED IR CEP WD GES AP D WED G WYP GUD NI GED GNP WD GED WEP NS VRS NP D GED Gl D GO GNP D CED GNP WL GH) GNP CED D SUD WE CEN GED W W WS G e ROM socket 3 Write Enable. If BCSR <6> (DIS 65) RS3 WE 04 N W 3-16 DESCRIPTION is clear, and if BCSR <5> and <4> RS3 WE) are both set, (RS3 65 and then the program can write access ROM socket 3 which typically contains an EEPROM. RS3 WE is cleared by Power Up and by Bus Initialize. 03 Unused This bit always reads as "0". 02 01 PMG CNT2 PMG CNT1 Processor Mastership Grant Count bits 2, and 0. These three bits enable the PMG Counter and PMG CNTO 00 select the length of time for PMG Counter overflow. When enabled, the PMG Counter begins counting when the KDJ11-BF must access an I/0O Page location or external memory. overflow causes Counter the KDJ11-BF to suppress all DMA Requests and give the processor bus mastership during the next DMA arbitration cycle. When the PMG Counter is disabled, the processor is blocked from bus mastership as long as DMA Requests are pending. All three bits are cleared by the negation of DCOK. PMG CNT2 PMG CNT1 PMG CNTO Count Time 0 0 0 (Disabled)* 0 1 1 0 -0 1 1 0 0 0 1 1 1 1 1 0 1 0 1 0 1 0.4 0.8 1.6 3.2 6.4 12.8 25.6 usec usec usec usec usec usec usec * The PMG count of 0 (Disabled) is not recommended for most typical systems, and is reserved for special 3.5.7 Page Control applications. Register a 1is 522, 777 17 addrress at Register, Control The Page . addressable word and byte both is that register read-write the 3-17 contains Figure 3-24 shows the register format:; Table bit descriptions. FUNCTIONAL DESCRIPTION HIGH BYTE FIGURE 3-24 LOW BYTE PAGE CONTROL REGISTER FORMAT TABLE 3-17 PAGE CONTROL REGISTER BIT DESCRIPTIONS FUNCTION BIT(S) NAME 15 Not 14:09 High Byte (R/W) ' These six bits provide the most significant ROM address bits when the 16-bit ROM sockets are accessed by bus addresses 17773000 - 08:07 Not Always 06:01 Low Byte (R/W) used used read Always zero. as 17773776. read as zeros. These six bits provide the most significant ROM (or EEPROM) address bits when the l6-bit (or EEPROM) or the 8-bit ROM sockets are accessed by bus addresses 17765000 - 17765776. 00 —— — S P 3.5.8 Not AES P AR G used Always R SR D M GO WS G MED GED R I YD D T Line read as ) NS A G SR A GED WD GID TS G Frequency Clock zero. D WP CEO R D W WD WD U and Status D M I S R ER SR GO SR GED GED WD GE GEG GMD T SUS MR mE m E S GO Register The Line Clock provides the system with timing information at determined by the UNIBUS LTC line or by the one intervals fixed Boot of the on-board KDJ1l1l-BF frequency signals as programmed by and Diagnostic Controller Status Register bits 11 and 10, Typically, LTC cycles at the AC 1line frequency, producing intervals of 16.7 msec (60 Hz line) or 20.0 msec (50 Hz 1line). The three on-board frequencies are 50 Hz, 60 Hz and 800 Hz. The Clock Status Register Clock interrupts to be control. Alternatively, (LKS), at address 17777546, enabled line 3-50 allows Line and disabled wunder program clock interrupts can be FUNCTIONAL DESCRIPTION unconditionally enabled by setting BCSR <13> (FRC LCIE). Program recognition of the Clock Status Register can be disabled by setting BCSR <12> (LKS DIS). The normal KDJ11l-BF configuration is FRC LCIE and LKS DIS both clear. These bits are set up by the Boot and Diagnostic ROM programs from -the KDJ11-BF Configuration Data. Figure 3-25 shows the register format; Table 3-18 contains the bit descriptions. FIGURE 4-25 CLOCK STATUS REGISTER FORMAT TABLE 3-18 CLOCK STATUS REGISTER BIT DESCRIPTIONS read as zero. 15:08 Unused. Always 07 LCM (R/W) Line Clock Monitor. This bit is set by the leading edge of the external BEVENT line (or of one of the three on-board clock frequencies) and by Bus Initialize. LCM is cleared automatically on processor interrupts acknowledge. It is also cleared by writes to the LKS with bit <7> = "0". 06 LCIE (R/W) Line Clock Interrupt Enable -- This bit, when set, causes the set condition of LCM (LKS <7>) to initiate a program interrupt request at a priority level of 6. When LCIE is clear, line clock interrupts are disabled. LCIE is cleared by Power Up and by Bus INIT. LCIE is held set INIT. LCIE is held set when BCSR <13> (FRC LCIE) 05:00 Unused Always is set. read as zeros. FUNCTIONAL DESCRIPTION 3.6 STACK LIMIT PROTECTION The KDJ11-BF checks kernel stack references against a fixedis limit less of 400(8). If the virtual address of the stack reference the current than 400(8), a yellow stack trap occurs at the end of instruction. stack A stack trap can only occur in kernel mode ard only on a throug h nce refere 5 or 4 mode a as reference, which is defined R6, or a JSR, trap, or interrupt stack push. In addition, the J1l1 checks for kernel stack aborts during of these interrupt, trap, and abort seguences. I1f, during one initia tes sequences, a kernal stack push causes an abort, the Jller bit <2>, a red zone stack trap by setting CPU Error Registr (R6) and loading virtual address 4 into the kernel stack pointe PC and PS trapping thru location 4 in kernel data space-. The old tively. are saved in kernel data space locations 0 and 2 respec NOTE The J-11 treatment of yellow stack trap 1is The 11/70 includes a identical to the 11/44. inclusive stack limit register, and a more Jll's The reference. stack of definition definition of a red stack trap is unigue. 3.7 KERNEL PROTECTION In order to interference, mechanisms: a. against system protect the kernel operating ion protect ng followi the rates incorpo BF KDJ11lthe In kernel mode, HALT, RESET, and SPL execute as specified. In supervisor or user mode, HALT <causes a trap through location 4, while RESET and SPL are treated as NOPs. and PS In supervisor or d. All trap and interrupt vector references are classified as e. Kernel stack references b. In kernel mode, RTI and RTT can <7:5> freely. <15:11> PS alter In supervisor or user mode, RTI and RTT can only set PS <15:11> and cannot alter PS <7:5>., c. In kernel mode, MTPS can alter PS <7:5>. user mode, MTPS cannot alter PS <7:5>. irrespective of the memory man-= references, kernel space trap or interrupt. the of time agement mode at the are checked for stack overflow. Supervisor and user stack references are not checked. 3-52 FUNCTIONAL DESCRIPTION 3.8 TRAP AND INTERRUPT SERVICE PRIORITIES In both traps and interrupts, the currently executing program 1is which is interrupted and a new program, the starting address ofed. The execut is , specified by the trap or interrupt vector is V vector a through pts hardware process for traps and interru identical: PS --> temp 1 lsave PS, PC in temporaries 0 --» PS <15:14> | force kernel mode M[V] --> PC M[V+2] =-=-> PS templ<15:14> --> PS<13:12> SP-2 --> SP templ --> M[SP] ifetch PC from vector, data space 1 fetch PS from vector, data space Iset previous mode !selected by new PS !push old PS on stack, data space PC --> SpP-2 temp --> e 2 SP tpush old PC on stack, data space temp2 -=-> M[SP] 1go execute next instruction NOTE If an abort occurs during either the vector fetch "or the stack push, the PS and PC are restored to their original state (i.e. to the state prior to trap sequence execution). The priority order for traps and interrupts is as follows: red stack address trap error memory management violation timeout/non-existent memory parity error trace (T-bit) trap yellow stack trap power fail floating point PIRQ 7 PIRQ 6 PIRQ 5 PIRQ 4 trap interrupt level 7 interrupt level 6 interrupt level 5 interrupt level 4 PIRQ 3 PIRQ 2 PIRQ 1 halt line FUNCTIONAL DESCRIPTION NOTE The halt line is given highest priority when processor hangs 3.9 CONSOLE SERIAL The console serial -the up. LINE UNIT line provides the KDJ11-BF processor with a The console serial the console terminal., for interface serial serial line is full duplex. It provides an RS-423 EIA interface which is also RS-232C compatible. ‘ This serial line interface is based on the DC319 Digital Link For additional (DLART). Transmitter Receiver Asynchronous 1 under Chapter in 1listed details refer to Chipkit handbook additional documents. insure that the console serial receive and The user should identical and are determined by three are rates baud transmit switches settings that are mounted on top of the KDJ11-BF Module, via the external SLU distribution board and baud or remotely, rate switch. The switch settings should remain in the off position. settings These switch settings, plus five additional switches Facility Diagnostic and Boot the via read (Switches 07-03) may be and Boot the on a is 07 switch If (BCR). Configuration Register Diagnostic Programs assumes that the system does not have a console terminal and uses switches 06-00 to select a limited Setting switch 07 does range of KDJ11-BF and system parameters. not disable the console terminal interface which runs at the baud The settin of these switches, reflected by switches 02-00. rate however, is determined by system configuration considerations. the Receiver There are four console serial line unit registers: Status Transmitter the Buffer, Data Receiver Status Register, the recognition Program Buffer. Data Transmitter the and Register, Each register is these registers can not be disabled. of described in the following subsections. 3.9.1 Receiver Status Register (at Figure 3-26 shows the Receiver Status Register (RCSR) format descriptions. bit the contains 3-19 Table address 17 777 560). FUNCTIONAL 15 14 13 12 0 0 1" 10 9 g o] o 0 7 6 DESCRIPTION 5 4 3 2 1 0 o 1] 0 0 0 0 RCV ACT ——J AX DONE RX IE FIGURE TABLE 15:12 11 3-19 3-26 RECEIVER RECEIVER STATUS Unused. RCV ACT (RO) STATUS Read as REGISTER REGISTER BIT FORMAT DESCRIPTIONS zeros. Receiver Active. This bit is set at the center of the start bit of the serial input data and is cleared at the expected center (per DLART timing) of the stop bit at the end of the serial data. RX DONE is set one bit time after RCV ACT is cleared. zeros. 10:08 Unused. Read 07 RX Receiver Done. This bit is set when an entire character has been received and is ready to be read from the RBUF register. This bit is It automatically cleared when RBUF is read. DONE (RO) is 06 RX IE (R/W) as also cleared Unused. Power Up. Receiver Interrupt Enable. This bit is cleared by Power Up and Bus INIT. If both RCVR DONE and RCVR INT ENB are set a program interrupt is requested. 05:00 by Read as zeros. FUNCTIONAL DESCRIPTION 3.9.2 Receiver Data Buffer (RBUF) format Figure 3-27 shows the Receiver Data Buffer register ins the bit 17 777 ERR \ FRM RCV address (at descriptions. OVR conta RECEIVED DATA BITS BRK ERR 3-20 Table 562). ERR FIGURE 3-27 RECEIVED DATA BUFFER REGISTER FORMAT TABLE 3-20 RECEIVED DATA BUFFER REGISTER BIT DESCRIPTIONS BIT(S) NAME 15 ERR (RO) FUNCTION Error. This bit is set if RBUF <14> or <13> is set. ERR is cleared if these two bits are cleared. This bit cannot generate a program interrupt. 14 13 NOTE: OVR ERR (RO) Overrun Error. This bit is set if a previously received character was not read before being FRM ERR (RO) Framing Error. This bit is set if the present character had no valid stop bit. This bit is overwritten by the present character. used to detect break. Error conditions remain present until the next character is received, at which point, the error bits are updated. The Error bits are not necessarily cleared by Power Up. 12 Unused ‘This bit always reads as "0". 11 RCV BRK (RO) Received Break. This bit is set at the end of a received character for the serial data input 10:08 Unused These bits always read as "0". 07:00 Received These read-only bits contain the last received D - Data Bits D D W D D D T SN S DW remained in the SPACE condition for all 11 bit time. RCV BRK then remains set until the serial data input returns to the MARK condition. character. G M T D D S S D D TH S - - —— s D T CED D TD D A A W P MU W TED N D GED M GHD AN SN WD w0 S e G e S FUNCTIONAL 3.9.3 Transmitter Status DESCRIPTION Register Figure 3-28 shows the Transmitter (at address 17 777 564). descriptions. Status Table Register (XCSR) format 3-21 contains the bit MAINT TX ROY XMIT BRK TX FIGURE TABLE 3-21 3-28 TRANSMIT TRANSMIT IE STATUS STATUS REGISTER REGISTER BIT FORMAT DESCRIPTIONS BIT(S) NAME FUNCTION 15:08 Unused Read 07 TX RDY (RO) Transmitter Ready. This bit is cleared when XBUF is loaded and sets when XBUF can receive TX IE (R/W) zeros. another character. XMT RDY is set by Power Up by and 06 as Bus Transmitter INIT. Interrupt Enable. cleared by Power Up and by Bus TX is RDY and TX requested. are set, a If INIT. program is both interrupt 05:03 Unused Read 02 MAINT (RO) Maintenance. This bit is used to facilitate a maintenance self-test. When MAINT is set, the external serial input is disconnected and the serial output is used as the serial input. This bit is cleared by Power Up and by Bus INIT. 01 Unused Read 00 XMIT BRK. Transmit Break. (R/W) as IE This bit as zeros. zero. serial output XMIT BRK is is When this bit is set, the forced to the SPACE CONDITION. cleared by Power Up and by Bus INIT. FUNCTIONAL DESCRIPTION 3.9.4 Transmitter Data Buffer Register Buffer Register (XBUF) Figure 3-29 shows the Transmitter566.DataTable 3-22 contains the bit 17 777 format, at address: descriptions. 15 14 13 12 1 10 9 8 7 6 5 4 3 2 1 0 XBUF AT FIGURE 3-29 TRANSMITTER DATA BUFFER REGISTER FORM IONS TABLE 3-22 TRANSMITTER DATA BUFFER REGISTER BIT DESCRIPT BIT(S) NAME FUNCION 15:08 Unused Always read as zeros. 07:00 XBUF These eight bits are used to load the (WO) transmited character. 3.9.5 Break Response to Unit may be configured either The KDJ11-BF Console Serial Line brea a when onse n or to have no resp e the processokr perform a halt operatio operation will caus condition is received. A haltdebu . gging technique (ODT) microcodel to halt and enter the octa and Boot the of 9 bit selected via The Halt on Break Option is us During Power up oOr Register. Stat er roll Cont ic Diagnost will always set bit ram prog nostic ROM Restart, the boot and diag k condition if the Brea on Halt the 9 to a 1. This will enableE position. Keylock switch is in the ENABL ved tion at the end of a recei The DLART recognizes a break condi E SPAC the in ined rema character for which the serial. dataTheinput 1line ion gnit Reco Break condition for all 11 bit times t returns to the MARK inpu data al seri the l unti rted remain asse condition. FUNCTIONAL DESCRIPTION 3.10 KERNEL/SUPERVISOR/USER MODE DESCRIPTIONS The PDP-11/84 processor family offers three modes Kernel, Supervisor, and User. of execution, Their use is to enhance the memory and flexibility the increase to protection scheme and nts. environme g ogrammin multi-pr and ng timeshari of functionality Kernel mode is the most privileged of the three modes and allows In an operation system featuring execution of any instruction. is system multi-programming, the wultimate control of the , Typically mode. implemented in code that executes in Kernel g schedulin job s, operation I/O this includes; control of physical resource management. and Memory management mapping and protection allows these executive elements to be protected from inadvertant or malicious tampering If by programs executing in the less privileged processor modes. kernel the only then mode, kernel in mapped only is page I/0 the has access to the memory management registers to re-map or modify the protection. This is because the memory management registers themselves exist in the I/O page. In order for a user program to have sensitive functions performed in its behalf, a request must be made of the executive program, typically in the form of a software trap that vectors the processor into kernel mode. Thus the executive code remains in control and can verify that the function requested is consistent with the operation of the system as a whole. be The supervisor mode is the next most privileged mode, and may programs of execution and for the mapping to provide used requiring protection from them. shareable by users but still s, logical I/O processors, interpreter This might include command or runtime systems. the prohibits and mode privileged least the does RESET as HALT and as such instructions will A multiprogramming operating system Supervisor mode. wuser programs to user mode to of execution restrict typically prevent a single user from having & negative effect on the system The user's virtual address space is set up such that as a whole. that those written are be the only areas of memory that can as protected are users among shared Areas that user. to belong 1is User mode of execution read-only, 3.11 PMG execute only, or for both read and execute access. COUNTER The PMG Counter enables the processor to become PMI master during heavy UNIBUS DMA activity. This allows the processor to limit To change the hog mode control of the PMI during DMA activity. setup console the on section the see counter parameters PMG the 3-59 FUNCTIONAL DESCRIPTION mode. The user can select from seven counter values and one The PMG counter default setting is the Disabled mode, disable. The least PMG with no DMA interruptions from the processcor. mode, is disable counter timeout, with the exception of the counter PMG selection 7. Selection 1 provides the, fastest clock of number per timeout, most number of PMI interrupts counter PMG eight the cycles. Table 3-24 provides a list of selections. Table 3-24 PMG COUNTER TIMEOUT LIST Switch Pos. Count Timeout 0 (Disabled)* 1 2 0.4 usec 0.8 usec 3 1.6 4 5 3.2 6.4 usec usec usec 6 12.8 usec 7 25.6 usec * The PMG count of 0 (Disabled) is not recommended for most typical systems, and is reserved for special applications. 3.12 KTJ11-B CACHE OPERATIONS The KTJ11-B DMA Cache is used to decrease PMI memory read access The cache is divided into four time for UNIBUS DMA devices. sections, with each section capable of storing up to eight addresses. The Initially,the KTJ11-B cache is flushed (i.e., emptied). device DMA UNIBUS a from y boundar octal an on first PMI read causes the KTJ11l-B to read from memory and store that address and the next 7 PMI address locations locations in section A. 1If the next PMI read is one of the addresses stored 1in the cache, a cache hit occurs. If the next PMI read does not match any cache address (cache miss), the KTJ1l-B does a memory read cycle for the requested address - if that address is on an octal boundary - and the next seven PMI memory addresses. These eight words are than stored in -.the next available cache section (i.e., B, C or D). DESCRIPTION FUNCTIONAL The KTJ11-B cache also monitors the PMI address If lines. a the KTJ11-B cache compares the write into an address occurs, When a match (hit) address with it's stored cache addresses. occurs the cache address location is invalidated. 3.12.1 KTJ11-B Cache Organization The KTJ11-B DMA Cache contains thirty-two 16-bit data registers, (A, B, C and D) of eight data registers arranged in four sets with each set is a Valid Bit and an d Associate each (000-111). The data registers are located in RAM 18-bit Tag Register. memory as shown in Table 3-24. located in the KTJ11-B are presented in Figure 3-30. The Tag Registers and Valid Bits Refer to the format Gate Array. TABLE 3-24 RAM MEMORY DATA REGISTER LOCATIONS RAM Address RAM Address Register Register 00 01 02 03 04 05 06 07 Set A Set A Set A Set A Set A Set A Set A Set A Register 0 Register 1 Register 2 Register 3 Register 4 Register 5 Register 6 Register 7 20 21 22 23 24 25 26 27 Set C Set C Set C Set C Set C Set C Set C Set C Register 0 Register 1 Register 2 Register 3 Register 4 Register 5 Register 6 Register 7 10 11 12 13 14 15 16 17 Set B Set B Set B Set B Set B Set B Set B Set B Register Register Register Register Register Register Register Reqgister O 1 2 3 4 5 6 7 30 31 32 33 34 35 36 37 Set D Set D Set D Set D Set D Set D Set D Set D Register Register Register Register Register Register Register Register —— TAG REGISTER O 1 2 3 4 5 6 7 UNUSED FIGURE 3-30 VALID BIT AND TAG REGISTER FORMAT (One of Four) 3-61 FUNCTIONAL DESCRIPTION physical (or CPU) DMA the interprets The KTJ11-B DMA Cache 3-25 Table and register the 1illustrates 3-31 Figure address. contains 21 20 L the bit descriptions. 18 18 17 16 T TT LTI 14 15 13 N I 1" — 5 4 3 2 I I T I T TITT] 10 9 8 TAG FIELD 7 6 ) INDEX FIELD BYTE SELECTION FIGURE 3-31 TABLE 3-25 PHYSICAL ADDRESS UBA PHYSICAL ADDRESS INTERPRETATION INTERPRETATION 21:04 Tag Field These 18 bits comprise a field that corresponds to the bits in the Tag Registers. 03:01 Index Field These three bits indicate if the address is on an even eight-word boundary (index=0) and points to one of eight data registers in a set (A-D). 00 Byte Selection Selects high or low byte write operations. This bit has no effect during DMA operations. 3.12.2 DMA Cache Enable/Disable The KTJ11-B Memory Configuration Register (KMCR) (described 1in subsection 3.13.3) contains both control and diagnostic status bits for the KTJ11-B DMA Cache. When bit 06 of the KMCR is cleared, or when bit 05 of Memory Management Register 3 is <cleared (i.e when the UNIBUS Map is disabled), then the DMA Cache is disabled and initialized to its power up condition. All four Valid Bits are cleared. Set A is the Next Available Set, followed by Set B, Sef: C, and Set D. The thirty-two data registers and the four tag registers are not cleared and contain random information. 3-62 FUNCTIONAL DESCRIPTION Memory and 06 When KTJ11-B Memory Configuration Register Dbit is Cache DMA the set, both are 05 bit 3 Register Management enabled and operates as described in the (See Figure following subsections. physical address 3-32.) 3.12.3 DMA Cache Write Operations the Dgring QPU and DMA write operations, tag bit for each f;eld is checked against the Tag Register and Valid Hit has Cache DMA a r of the four Sets to determine whethe caused which Set the then d, If a Cache Hit has occure occur;ed. is Bit Valid its and Set able the hit becomes the Next Avail cleared. Otherwise, the DMA Cache is not affected. " Y Im MEMORY UNIBUS [ " onieus | ggfics i MEMORY MSV11-R cru UNIBUS KDJ11-8 FIGURE 3-32 PDP-11/84 CACHE DIAGRAM 3.12.4 DMA Cache Read Operations the memory, the physical addrdess During all DMA reads from PMI for bit Vali and ster Regi nst the Tag tag field 1is checked agai has Hit e Cach DMA a her whet e rmin each of the four Sets to dete ical address index field is checked for occurred. Also, the phys an even 8-word boundary (index 0). and if the index field does If a DMA Cache Hit has not occurre d, essed memory location not egqual zero, then the content of thee addr is not affected. is gated onto the UNIBUS. The DMA Cach if the index field not occurred, and take If a DMA Cache Hit has foll place: owing operations equals zero, then the ed 1into the Tag ess Tag Field 1is load 1. The physical addrndin g to the Next Available Set. Register correspo 1is gated 2. The content of the addressed PMI memory location onto the UNIBUS. 3-63 FUNCTIONAL 3. DESCRIPTION The content of the addressed main memory location and of the seven succeeding memory locations is stored in the data registers of the Next Available Set. 4. If all eight data registers are successfully 1loaded, the valid Bit corresponding to the Next Available Set is set and that Set becomes the Least Available Set. 5. If a parity error occurs while reading one of the memory locations, the corresponding Valid Bit is cleared and the Set remains the Next Available Set. NOTE The KTJ11l-B contains no parity error indication. Since the offending memory location is not stored in the DMA Cache, the DMA device will receive any and when it if indications error parity n. locatio memory that s accesse specifically If a DMA Cache Hit has occurred, then the following operations take place: 1. The Set (A-D) whose Tag Register caused the hit, along with the physical address index field, selects a data register whose contents is gated onto the UNIBUS. 2. If the Index Field equals 7 (the last cdata register in the set has been read), the corresponding Valid Bit is cleared, and that data register set becomes the Next Available Set. 3. If the Index Field does not equal 7, the Valid Bit remains set, but the data register Set becomes the Least Available Set. 3.13 UNIBUS MAPPING The UNIBUS Map is the interface between the UNIBUS and PMI It responds as a slave to UNIBUS signals and is used to memory. The convert 18-bit UNIBUS addresses to 22-bit memory addresses. signal additional an by d accompanie is address 22-bit memory line, BBS7 L. The assertion of BBS7 L disables the PMI address decoding and selects the I/O Page. UNIBUS address space is 256 KB of which the top 8KB addresses always reference the I/0O page. The lower 248KB of UNIBUS address space can be used by the UNIBUS map to reference physical memory. (See Figure 3-33.) FUNCTIONAL DESCRIPTION 777 777 /0 PAGE . (8K BYTES) 760 000 757 777 TO UNIBUS MAP (248K BYTES) 000 000 18-8IT UNIBUS ADDRESSES FIGURE 3-33 UNIBUS ADDRESS SPACE gement Register rammed, via Memory Manaled The UNIBUS Map can be toprog or relocation run with relocation enab Three (MMR3) bit 05, disabled. = 0), the UNIBUS Map abled (MMR3 bit 05 bits If relocation is dising 8) to the UNIBUS =zeros (address memory21-1addr appends four lead ess. Memory g the 22-bit address, thus prod0ucin bits 17-00. ess addr US identical to UNIB address bits 17-0bits are signal is L BBS7 the , 3 are all ones If memory address g the 17-1 ' I/O Page. asserted, selectin the UNIBUS Map (MMR3 bit 05 = 1), If relocation is enabled select one of 31 mapping ess Dbits 17-13 tocode decodes UNIBUS addr s 00 thru 36). UNIBThe onding to octal er pair register pairs (corresp US is added to ist cted mapping regmemo content of the sele US UNIB If ess. ry addr uce the address bits 12-00 3 toareprod is Page 1/0 the 37), code all ones (octal ry address bits 17-1 ry memo , rted asse 1is al to the memo selected. The BBS7 L sign tical to UNIBUS address bits 17-00 and iden are 0 17-0 address bits are not asserted. memory address bits 21-18 3.13.1 UNIBUS Mapping Registers pairs of which only 32 mapping register on. The UNIBUS Map contains Figures 3-34 for address relocati ster Seepair 31 are actually e used s may be . The mapping regi ~and 3-35, and Tabl or 3-26 ' indirectly: accessed directly FUNCTIONAL DESCRIPTION 1. accessed are Direct Access - The mapping registers Each . addresses Page I/O their through individually Register, Address Hi a of consists pair mapping register 21-16 and a Lo relocation address Dbits which contains address bits n relocatio contains which Register, Address 1 5-0 l © 2. enabled, 1is Indirect Access - When UNIBUS Map Relocation UNIBUS address bits 17-13 select the appropriate mapping 18-bit UNIBUS register pair to be used in relocating the address. 15 14 13 0 0 0 12 1 10 9 8 7 6 0 0 0 0 0 0 0 ] = L ; 0 1 2 3 4 5 RELOCATION ADDRESS BITS 21-16 FIGURE 3-34 HI-ADDRESS REGISTER FORMAT i5 i4 13 12 11 10 9 8 7 6 5 4 3 2 1 0 Q J . . RELOCATION ADDRESS BITS 15-01 FIGURE 3-35 LO-ADDRESS REGISTER FORMAT TABLE 3-26 UNIBUS MAP REGISTER PAIRS UNIBUS ADDRESSES MAPPED VIA REGISTER REGISTER PAIR NO. I/0 PAGE ADDRESSES HI REGISTER LO REGISTER 0 1 2 3 17 17 17 17 770 770 770 770 200 204 210 214 17 17 17 17 770 770 770 770 202 206 212 216 000 020 040 060 000 000 000 000 - 017 037 057 077 777 777 777 777 4 5 6 7 17 17 17 17 770 770 770 770 220 224 230 234 17 17 17 17 770 770 770 770 222 226 232 236 100 120 140 160 000 000 000 000 - 117 137 157 177 777 777 777 777 10 11 12 13 ' 17 770 240 17 770 244 17 770 250 17 770 254 17 770 242 17 770 246 17 770 252 17 770 256 3266 PAIR 200 000 - 217 777 220 000 - 237 777 240 000 - 257 777 260 000 - 277 777 FUNCTIONAL TABLE 3-26 DESCRIPTION (Cont) ——<--————--—————-—-————---——-—-——————-——-———-————————-——'———-—— 14 15 16 17 17 770 17 770 17 770 17 770 260 264 270 274 17 17 17 17 770 262 770 266 770 272 770 276 300 000 320 000 340 000 360 000 - 317 - 337 - 357 - 377 777 777 777 777 20 21 17 17 770 770 300 304 770 770 302 306 400 420 000 000 - 417 - 437 777 777 22 23 17 770 310 17 770 314 17 17 24 25 26 27 17 17 17 17 770 770 770 770 320 324 330 334 17 17 17 17 770 770 770 770 322 326 332 336 500 520 540 560 000 000 000 000 - 517 537 557 577 777 777 777 777 30 31 32 33 17 17 17 17 770 770 770 770 340 344 350 354 17 17 17 17 770 770 770 770 342 346 352 356 600 620 640 660 000 000 Q00 000 - 617 637 657 677 777 777 777 777 34 35 36 37 * 17 17 17 17 770 770 770 770 360 364 370 374 17 17 17 17 770 770 770 770 362 366 372 376 - * Can be read or written into, 3.13.2 Optional 440 000 - 457 777 460 000 - 477 777 17 770 312 17 770 316 UNIBUS 700 000 720 000 740 000 1I1I/0 Page (No 717 777 737 777 757 777 Relocation) but not used for mapping. Memory to install UNIBUS Memory instead of, or The ability in addition However, it differs somewhat PMI memory, has been preserved. to assigned UNIBUS address space is from previous implementations. memory in 8K byte segments, starting with the segment UNIBUS to below the I/O page and proceeding downward. Those UNIBUS address segments assigned to UNIBUS Memory can not Whenever the CPU to access PMI memory via the I/O Map. used be bv memorv PMT disahle will KTJ1l-B the accesses UNIBUS memory, asserting the PMI UBMEM signal. The KDJ11-B CPU Module does not cache UNIBUS Memory, because it disables its cache when UBMEM is asserted. FUNCTIONAL DESCRIPTION NOTE 18-bit UNIBUS PDP-11/84. The UNIBUS memories address range ONLY of are each supported UNIBUS by the memory module 1is determined by jumpers or switches on that module. The KTJ1ll-B Memory Configuration Register (KMCR), described 1in subsection 3.13.3, within must accurately system. reflect the placement of UNIBUS memory the The KMCR register is cleared by the assertion of DC loaded by the KDJ11l-B boot and diagnostic programs by the KDJ11-B EEPROM Configuration Data. LO and is as specified KMCR <04:00> specify the number of 8K byte address segments, from 0 to 31, assigned to UNIBUS Memory. KMCR <05> specifies the location of the UNIBUS Memory from the viewpoint of the CPU. See Table 3-28., If KMCR <05> is clear (22-bit Mode), the top UNIBUS memory location is 17 757 776. If KMCR <05> is mode), the top UNIBUS memory location is 757 776. If the system has no UNIBUS memory, cleared. 1In this configuration: l. PMI memory, as seen by the locations from 00 000 000 up 2., All UNIBUS DMA devices then CPU, to as access PMI KMCR set <05:00> (18-bit must all be resides 1in contiguous high as 17 757 777. memory thru the UNIBUS Map. If the system contains UNIBUS memory only, all be set. 1In this configuration: 1. UNIBUS memory, ‘locations 2. If to as KMCR resides high as all UNIBUS DMA from address 00 O <05:00> must in contiguous 757 777. devices, up to as resides high as The UNIBUS Map Register Pairs are still accessible read-write registers, but the UNIBUS Map is disabled does not respond to the UNIBUS address range 0 thru 777. the l. the CPU, up UNIBUS memory, as seen by in contiguous 1locations address 757 777. 3. KMCR as seen by from address 0 then system <05> is UNIBUS contains clear memory, segments both (22-bit as assigned PMI seen to memory mode), by 3-68 UNIBUS memory, and if 8K byte then: the UNIBUS and as and 757 CPU, Memory falls by within KMCR the <04:00>. UNIBUS FUNCTIONAL DESCRIPTION Memory addresses are assigned downward starting with 8K Table 3-28 lists the byte segment 17 740 000 - 17 757 777. KMCR bit various the by allocated UNIBUS Address space codes. 1in contiguous resides CPU, the PMI memory, as seen by locations from address 00 000 000 up to as high as the last KTJ11-B The address not assigned to UNIBUS memory. residing memory the response of PMI specifically disables in locations assigned to UNIBUS memory by asserting the PMI UNIBUS Memory line (PUBMEM). section the The UNIBUS DMA devices access PMI memory thru assigned been not has which space address Map UNIBUS of the UNIBUS Each 8K byte segment assigned to to UNIBUS memory. Memory, disable its corresponding UNIBUS Map Register Pair. read-write as accessible still are pairs Disabled Map does not respond to their UNIBUS the registers, but assigned UNIBUS address space. with directly The UNIBUS DMA devices access UNIBUS Memory same the accesses XXX XXX DMA Address address. 18-bit an XXX. XXX 17 address CPU by accessed UNIBUS Memory location NOTE of number the on 1limit The KTJ11-B places no disabled by KMCR be may which pairs register software system typical However, <04:00>. minimum of 5 or & register pairs to a requires a with memory PMI allow DMA devices to access ' moderate degree of efficiency. If the system contains both PMI memory and UNIBUS memory, and KMCR 1. <05> is set (18-bit mode), 1if then: UNIBUS memory, as seen by the CPU, falls within the 8K byte UNIBUS segments assigned to UNIBUS Memory by KMCR <04:00>. 8K with starting downward Memory addresses are assigned the lists 3-27 Table 777. -- 757 000 740 segment byte UNIBUS Address space allocated by the various KMCR bit codes. PMI memory locations as seen by from address the CPU, resides 1in 000 000 up to as high contiguous as the last address not assigned to UNIBUS memory. The KTJ11-B disables the response of PMI memory residing in loca- tions assigned to UNIBUS memory by asserting the PMI UNIBUS Memory line (PUBMEM). FUNCTIONAL 3‘ DESCRIPTION PMI memory, as contiguous locations the last this 1is the seen by the UNIBUS DMA devices, from address 000 000 up resides to as high as address not assigned to UNIBUS memory. Because an 18-bit system, system software does not enable UNIBUS Map. The UNIBUS DMA devices access UNIBUS Memory directly, the same 18-bit addresses used by the CPU. TABLE 3-27 REGISTER SELECTION OF UNIBUS MEMORY UNIBUS 04 KMCR 03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 OKB 8KB 16KB 24KB 32KB 40KB 48KB 56KB XX XX XX XX XX XX XX 740 720 700 660 640 620 600 000 000 000 000 000 000 000 - XX XX XX XX XX XX XX 757 757 757 757 757 757 757 777 777 777 777 777 777 777 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 64KB 72KB 80KB 88KB 96KB 104KB 112KB 120KB XX XX XX XX XX XX XX XX 560 540 520 500 460 440 420 400 000 000 000 000 000 000 000 000 - XX 757 XX 757 XX 757 XX 757 XX 757 XX 757 XX 757 XX 757 777 777 777 777 777 777 777 777 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 1 128KB 136KB 144KB 152KB XX XX XX XX 360 340 320 300 000 000 000 000 - XX 757 XX 757 XX 757 XX 757 777 777 777 777 1 1 1 1 0 0 0 0 1 1 1 1 0 0 1 1 0 1 0 1 160KB 168KB XX6KB 184KB XX XX XX XX 260 240 220 200 000 C00 000 €000 - XX 757 - XX 757 - XX 757 - XX 757 777 777 777 777 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 192KB 200KB 208KB 216KB 224KB 232KB 240KB 248KB XX 160 XX 140 XX 120 XX 100 XX 060 XX 040 XX 020 XX 000 000 000 000 €000 000 000 000 000 - XX 757 - XX 757 - XX 757 - XX 757 - XX 757 - XX 757 - XX 757 - XX 757 777 777 777 777 777 777 777 777 D D D R WD O O Register 02 01 a0 (TR XD AT eED G EED oD aff Gl Bits 00 Gl WED GEN GED Memory Size A @) GEd Gl W GEb GNP PP GED GWD UNIBUS Memory Address Range GNP GEP GED GEN GMN VES GED 4 GED il GID GED GED GHL GED GED GAF GND VNS WMD D WS WD oD TIO FUNCTIONAL DESCRIPTION NOTE XX = for 3.13.3 The 777 17 for KMCR Memory KMCR <05> = <05> "1" Configuration KTJ11-B Memory 734, allows = "0" (18-bit (22-bit mode) XX = 00 mode) Register Configuration Register (KMCR), at the KDJ11l-B Boot and Diagnostic address 17 to programs configure the KTJ1l1l-B for the memory within the system. distribution of UNIBUS and Main Additional KMCR bits allow the DMA Cache to be enabled and disabled, provide diagnostic status of the Read Buffer and provide information on the system reboot status. Figure 3-36 shows the register format, and Table 3-28 contains the bit descriptions. OMA CACHE STATUS BITS SELECT STATUS — RBT PLS CA ENB 18-8IT MODE UNIBUS MEMORY SIZE FIGURE 3-36 MEMORY CONFIGURATION REGISTER (KMCR) TABLE BIT(S) 15:09 3-28 MEMORY CONFIGURATION NAME DMA REGISTER BIT DESCRIPTIONS FUNCTION Cache Status (RO) Bits These DMA The seven bits reflect the Cache. KMCR <15> is DMA content of KMCR <14:09> value of the value of the status of the Cache Hit. depends upon KMCR <08> the (Status Select). 08 Status (R/W) Select This bit selects (See Tables the 3-29 and content 3-30.) of KMCR <15-09>. FUNCTIONAL DESCRIPTION TABLE 07 Reboot Pulse (RBT PLS) (RO) 3-28 (Cont) This bit is set by the front panel reboot pulse , which also generates a KTJ1ll-B Power by cleared not is PLS Down/Power Up Cycle. RBT Power KTJ11-B the during the assertion of DC LO Down/Power UP cycle initiated by the front panel reboot pulse, but it is cleared by any other DC LO assertion. 06 05 Cache Enable (CA ENB) (R/W) 18-Bit Mode (R/W) This bit, when set, enables the DMA Cache. When CA ENB is clear, the DMA Cache is disabled. CA ENB is cleared by the assertion of DC LO. When this bit is set, the CPU can access UNIBUS Memory only when address bits <21:18> = 00. When this bit is clear, they can access UNIBUS memory if address bits <21:18> = 17. This bit is cleared by the assertion of DC LO. Write access to this bit is disabled when DCSR <08> (Diagnostic Mode) 04:00 UNIBUS Memory Size is clear. If the system contains main memory only (no UNIBUS memory), these five bits, as well as KMCR <05>, must be cleared. If the system contains UNIBUS memory only (no main memory), then KMCR <05:00> must be set. If the system contains both main memory and UNIBUS memory, KMCR <04:00> indicate the number of 8K byte address segments assigned to UNIBUS memory. As described in section 3.4, UNIBUS memory is assigned downward, starting with the segment below the I/O Page. These bits are cleared by assertion of DC LO. Write access to these bits is disabled when DCSR <08> (Diagnostic Mode) is clear. _—n—u s_@m-.‘m-‘a——-—--— o OO D T - D D e D S W _—‘-’-——‘———-——-——--—--———m-n—:——:@- If KMCR <8> (Status Select = 0, then KMCR <15-08> contain the DMA 2-bit Most Recently Used Set code, and the the Cache Hit bit, four Valid bits. Figure 3-37 shows the field format; Table 3-29 contains the bit descriptions. FUNCTIONAL DESCRIPTION DMA CACHE BIT UNUSED SET A VALID SET B VALID SET C VALID SET D VALID STATUS FIGURE TABLE BIT(S) SELECT 3-37 3-29 STATUS STATUS SELECT SELECT=0 = 0 FIELD FIELD FORMAT DESCRIPTION NAME FUNCTION 15 DMA Cache Hit This bit is updated during all writes to main memory, and all reads from main memory. It is set if a cache hit is detected, and cleared if a cache miss is detected. This bit is cleared when KMCR <6> 1is clear. 14-13 Unused Always 12 Set A Valid read as Reflects the current status of the Valid bit corresponding when KMCR <6> 11 Set B Valid zero. Reflects the corresponding when KMCR <6> to Set A. is clear. current The status to Set B. is clear. The bit is of the bit is cleared Valid 10 Set C Valid Reflects the current status corresponding to Set C. The when KMCR <6> is clear. of the bit ic Valid 09 Set D Valid Reflects of the Valid bit is the corresponding when KMCR <6> If as of current status to Set D. is clear. The bit cleared bit cleared bit cleared KMCR 8> = 1, then KMCR <15-08> contain the DMA Cache Hit bit well as the six bits which determine the relative availability the four sets (i.e., A, B, C, D). 3-73 FUNCTIONAL DESCRIPTION is disabled, and is clear, the DMA cache Set When KMCR <6> (CA ENB) lity A is the Next are set. the six Set Availabi by Setbits D. Set B, Set C, and Available Set, followed e e is enabled. If a DMAwritcach When KMCR <6> is set, the ofDMAthecach to e sets during a CPU or DMA hit is detected for one Set. lble Avai Next the memory, then that set becomes during a DMA cted for one of the sets 1f a DMA cache hit is dete t Available Leas set becomes the read from main memory, them sthat data registers are successfully set' Set. Similarly, if DMAa read cache miss, then that set becomes the a g owin foll ed joad Least Available Set. the Dbit Figure 3-38 shows the field format; Table 3-30 contains descriptions. A TOPS B A TOPS C A TOPS D B TOPS C B TOPS D C TOPS D STATUS SELECT FIGURE 3-38 SELECT STATUS = 1 FIELD FORMAT TABLE 3-30 SELECT STATUS = 1 FIELD DESCRIPTION -——-—————————;—-—————-——-——-_—p ———————--n——_—n-m—a——_n—-—————_————————————.- w—fl—_---—‘-—G——-—Qafiu-fl -—--———-‘—“‘w——--_--———--——_———-—-_—--—-—c—‘- 15 DMA Cache Hit This bit is updated during all writes to main memory, and all DMA reads from main memory. It is set if a cache hit is detected, and cleared if a cache miss is detected. The bit is cleared when KMCR <6> - - ——— - D A IR D D (e D @D N I S A S GRS D e Gy S D IR G S G is clear. O S D D G GO 70 SN O - D O B G O D D G G WD S G YER G e G IO CHP O GOR EA O MO 6O FUNCTIONAL TABLE 14 A Tops B (ATPSB) If ATPSB Set B. is If 3-30 set, ATPSB (Cont) Set is DESCRIPTION A is more clear, available Set B is than more available than when Set A. ATPSB is set when KMCR <6> is clear, Set A becomes the Next Available Set, and when Set B becomes the Least Available Set. ATPSB is cleared when Set B becomes the Next Available Set, and when Set A becomes the Least Available 13 A Tops C (ATPSC) 12 A Tops D (ATPSD) If Set. ATPSC is If ATPSD Set D. is If Tops C (BTPSC) Set A, ATPSD Set A becomes Tops D (BTPSD) If BTPSC when Set and when is Available Available C Tops (CTPSD) D is more available than A is more clear, is set the Set when Next available D is KMCR more <6> Available than available is clear, Set, Set D becomes the Least Available Set. cleared when Set D becomes the Next Set, Set. is and set, when Set B Set is A B becomes Set C cleared Set, Set. the more Next becomes when the Set and when becomes the available C is KMCR Least than more available <6> is clear, Available Set, Least Available C Set becomes B the becomes Set. Next the Least If BTPSD is set, Set B is more available than Set D. If BTPSD is clear, Set D is more available than Set B. BTPSD is set when KMCR <6> is clear, when Set B becomes the Next Available Set, and when BTPSD is 09 A Set C. If BTPSC is clear, Set than Set B. BTPSC is set when Available Available B Set is than BTPSC 10 set, ATPSD Available Available B Set when and when ATPSD is 11 set, Set C. If ATPSC is clear, Set C is more available than Set A. ATPSC is set when KMCR <6> is clear, when Set A becomes the Next Available Set, and when Set C becomes the Least Available Set. ATPSC is cleared when Set C becomes the Next Available Set, and when Set A becomes the Least Available Set. If CTPSD Set D. than when If Set Set Set D becomes the Least Available Set. cleared when Set D becomes the Next Set, Set. is and set, CTPSD when Set is C Set is clear, B becomes more Set the available D is more Least than available C. CTPSD is set when KMCR <6> is clear, C becomes the Next Available Set, and when Set D becomes the Least Available Set. CTPSD is cleared when Set D becomes the Next Available Set, and when Set C becomes the Least Available Set. FUNCTIONAL DESCRIPTION AND DIAGNOSTIC KTJ11-B 3.14 CONFIGURATION REGISTERS KTJ11-B Diagnostic and Configuration Registers are used with tic diagnostic programs to check out the KTJ11-B, both in Diagnos stic Diagno and Boot KDJ11-B The d. disable UNIBUS Mode, with the Programs also use these registers to enable or disable the KTJ11-B DMA Cache and to specify the presence and location of UNIBUS memory. When operating in diagnostic mode, the KTJ11-B can be programmed to perform diagnostic NPR cycles which test out its address and data paths along with the UNIBUS Map. 3.14.1 Diagnostic Controller Status Register Diagnostic Programs use the Diagnostic Controller Status Register (DCSR), at address 17 777 730, to enter and to exit from Diagnostic Mode, to select the source of the Diagnostic Data Register (DDR) and to perform Diagnostic NPR cycles which test the UNIBUS Map along with the KTJ11l-B address and data paths. Figure 3-39 shows the register format; Table 3~31 contains the bit descriptions. 0 b] 0 0 o) 0 0 0 o | DATI GO DONXM ERR DIAGNOSTIC MODE DNPR DONE BOOT ROM OIS DOR SELECT FIGURE 3-39 DIAGNOSTIC CONTROLLER STATUS REGISTER FORMAT TABLE 3-31 DIAGNOSTIC CONTROLLER STATUS REGISTER BIT DESCRIPTIONS 15 DNXM ERR Diagnostic Non-Existant Memory Error register. This bit is cleared at the start of a diagnostic NPR cycle and set if there is a non-existant memory timeout during that cycle. DNXM ERR is also cleared when DCSR <08> (Diagnostic Mode) is cleared. FUNCTIONAL TABLE 3-31 DESCRIPTION (Cont) --—-———-—————-—--————— —————-—-————-——.—————— ——————-——-——-——————-—— Unused These Diagnostic When Mode (R/W) bits always Done This bit NPR cycles to the DNPR Done is of Diagnostic a 03 Boot When bits ROM this Select GO (WO) 3.14.2 Diagnostic Diagnostic a are Done "1" in no is bit Diagnostic cleared 00, and Data Register INIT or by Bus by by a any (DDR). by completion of the NPR Cycle. is read as zero. set, response UBA When by this to the bit those is cleared, addresses. assertion of DC the UBA boot This bit is LO. The DDR Select bits are the cleared by INIT. Writing a "1" into this bit sets up a Diagnostic Data-In NPR Cycle and clears DCSR bit 07. The NPR cycle is actually initiated by the next CPU read cycle which accesses the PMI. That cycle provides the address used in the NPR cycle. The data fetched during that cycle is loaded into the Diagnostic Data Reglster (DDR) . Data Register Programs use the 777 732, along Diagnostic Data Register (DDR), at with the Diagnostic Controller Status (DCSR), to perform Diagnostic NPR cycles and to monitor state of various UNIBUS data, .address and control signals. address Register the there DNPR set bit operations. DATI for normal assertion of the These two bits select the contents of Diagnostic Data Register during read Bus 00 configured by Diagnostic always responds cleared (R/W) is set boot ROM at addresses 177773000 - 177773776 is disabled, allowing operation of any external ROM which uses those addresses on the UNIBUS. DDR set when write These 02:01 is pending. DCSR with Unused (R/W) is to 06:04 ROM KTJ11-B bit LO. write Disable zero. this bit is set, the UNIBUS is disabled the KTJll-B is configured for Diagnostic Mode. When this bit is clear, the UNIBUS is DC DNPR as and enabled and the operation. This 07 read -- 17 3-77 FUNCTIONAL DESCRIPTION Diagnostic NPR cycles test out the UNIBUS Map along with many of the KTJ1l1l-B address and data paths. During Diagnostic cycles, DCSR <02:01> are set egual to 0, thus selecting Diagnostic NPR Register. Following a Diagnostic Data-In NPR the NPR Cycle, the Diagnostic NPR Register contains the transferred data which can then be read through the DDR. Diagnostic programs set up a Diagnostic Data-Out NPR cycle by writing the data to be transferred into the DDR. Figure 3-40 shows the register format; Table 3-32 contains the bit descriptions. All writes to the DDR access information accessed during on DCSR <02:01>. 14 15 12 13 o1 { o | o1 | on the Diagnostic read operations 9 8 o1 | o1 | on 0/1 1 10 7 6 S 4 3 NPR Register. The from the DDR depends 2 0 1 0 L_I Al6 AV7 co a PB SSYN MSYN FIGURE 3-40 DIAGNOSTIC DATA REGISTER FORMAT Cdntents of the DDR when Select Code = 11 TABLE 3-32 DIAGNOSTIC DATA REGISTER CONTENT DESCRIPTIONS " DDR SELECT BITS CONTENT OF DIAGNOSTIC DATA REGISTER Bit 02 Bit 01 0 0 Diagnostic NPR Register 0 1 UNIBUS Data Lines D15-00 1 0 UNIBUS Address Lines Al5-00%* 1 1 UNIBUS Address Lines Al7-16 and various UNIBUS Control Lines * NOTE: Asserted address line Al6 during the Diagnostic UNIBUS Address Lines read operation, a parity error abort. 3-78 may cause FUNCTIONAL 3.14.3 DESCRIPTION Diagnostic DATI NPR Cycles The execution procedure for diagnostic DATI cycles is as follows: Diagnostic mode with 1. The KTJ11l-B must be running in 2. The diagnostic program writes a 1 into DCSR bit 00. Select bits (DCSR 02-01) = O. DDR NOTE At this point, PMI bus 3. I/0 page any UNIBUS memory read access, Or read or write access results in a timeout. The diagnostic program writes the test data pattern into The KTJ11l-B latches address the target memory location. bits Al17-00 of this cycle. 4. The KTJ11-B then executes its diagnostic DATI NPR cycle, storing the fetched data in the Diagnostic Data register. The address used in this cycle in produced by the UNIBUS The 18-bit address using the latched 18-bit address. Map, may be used directly (UNIBUS Map Relocation disabled) or it may be relocated to produce a 22-bit address (UNIBUS Map . enabled) 5. The diagnostic program verifies that register contains the correct data. the Data Diagnostic 3.14.4 Diagnostic DATO NPR Cycles The execution procedure for diagnostic DATO NPR cycles is as follows: 1. The KTJ11-B must be running in Diagnostic mode. 2. The diagnostic program loads the data into the Diagnostic Data register. NPR cycle the for Loading this register primes the KTJ11-B for a diagnostic NPR cycle. NOTE At this point, non a bus any UNIBUS memory read access, Or PMI I/O page read or write access results in timeout. 3-79 FUNCTIONAL DESCRIPTION The KTJ11-B latches KDJ11-B external address write bits to memory Al7-00 address from space. the next The KTJ11-B then executes its diagnostic DATO NPR cycle, The using the data stored in the Diagnostic Data register. Map, UNIBUS address.used in this cycle is produced by the using the latched 18-bit address. The diagnostic program verifies The that the target memory If it wishes to check contains the correct data. location before so do must it register, Data Diagnostic the alter would which operation write external an performing the contents of that register. CHAPTER BOOTSTRAP 4.1 INTRODUCTION The CPU contains programs (ROM up or restart various changed The two AND ROMs DIAGNOSTIC (read ROM PROGRAMMING only memories) which also data contains in an the ROMs EEPROM is permanent to be taken registers are Parameters in program in reguire the can also be the user used the at to power up or restart, be configured. EEPROM can be and the cannot (electrically programmable read only memory). The EEPROM is parameters which the ROM program uses to determine are UBA stores code) used to test the CPU, UBA and memory at power and to allow the starting of the user's software on devices. The by the user. CPU 4 and changed how under be eraseable used what to store actions various CPU control of and a ROM called Setup mode. Setup mode does not to remove the CPU or UBA modules. The EEPROM to store customer bootstrap programs. The diagnostic ROM program is automatically started by the CPU each time the system 1is powered up or restarted by use of the RESTART switch on the front panel. The ROM program will run tests selected by parameters 1in the EEPROM. After testing is complete, parameters in the EEPROM taken next by the ROM program. In a typical example the starts a program from referred to as booting ROM determine program what action automatically After the user's software entered again wuntil the In is some cases after testing complete 4-1 the ROM program to 1loads the user’'s disk or tape. This is a program and this mode will be to as automatic boot mode. the ROM program will not be powered up or restarted. is be and commonly referred is started system |is enters a BOOTSTRAP mode AND which next by way DIAGNOSTIC allows of the ROM PROGRAMMING user keyboard to select commands what action entered is through to be the taken console terminal. This mode will be refered to as Dialog mode. the EEPROM determine the tests to be run, general mode to be entered after testing is complete and final configuration of certain registers on the CPU and parameters module 'In the before the system cases the ROM program enters some The the the UBA in selections in the software EEPROM. is This started. Dialog occurs mode if regardless the user types of CTRL 'C at the console terminal during testing, or the boot sequence, or anytime the FORCE DIALOG switch is turned on. Generally, this is done by the user to allow changes to be made to the parameters in not the EEPROM, or to previously selected The FORCE DIALOG allow the user by the EEPROM. switch allows to boot a wuser to the device which was wunconditionally override the selections in the EEPROM. This override is provided because there are certain modes which cannot be aborted because the ROM program executes the modes too gquickly and is not able to monitor the console terminal for a CTRL C typed by the user. NOTE All user input is ignored at the console keyboard until the "Testing 1in progress - Please wait" message is typed out by the ROM code. The description of the commands for the ROM code assume that the EPROMs installed are at Version 7.0 (V7.0). Earlier PDP-11/84s contain EPROMS with V6.0 ROM code. The version of the ROM code is typed out each time Setup mode is enetred from Dialog mode and is displayed at the upper right corner of the printout. It 1is not necessary to version number. version numbers. Socket Location on (M8190) CPU E116 (low E117 (high Differences Appendix The the CPU following Part byte) byte) between V7.0 module to lists the Number V7.0 determine ROM part Part the ROM code numbers and Number V6.0 23-116E5-00 23-077E5-00 23-117E5-00 23-078E5-00 and V6.0 ROM code are described 1in F. following program or remove The illustrations would restarting of print out the CPU. on are the examples console of messages terminal during the power ROM up BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Figure 4-1 shows an example of a typical system booting automatic boot mode. In this case the user's software and was booted from device DU unit 0. wup 1in is RTI11 in progress - Please wait Testing Memory Size is 1024 K Bytes 9 Step memory test Step 1 2 3 456 789 Starting automatic boot Starting system RT-11FB (S) from DUO V05.0 FIGURE 4-1 AUTOMATIC BOOT MODE EXAMPLE The messages that follow the line "Starting system from DUQ" come by the ROM generated not are and booted software the from all At this point the ROM program is not executing and program. actions are determined by the user's Figure 4-2 shows an example of a software. typical system powering up, running the internal diagnostics and then entering dialog mode. to 1is action The ROM program waits for the user to select what occur next. Testing in progress - Please Memory Size is 1024 K Bytes 9 Step memory Step 1 2 3 test 7 456 wait 89 Commands are Help, Boot, List, Setup, Map and Type a command then press the RETURN Kkey: FIGURE 4.2 DIALOG MODE Dialog mode 4-2 SYSTEM COMMAND allows the POWERUP IN DESCRIPTIONS user to: DIALOG MODE Test. EXAMPLE BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING a. Boot a device b. List boot programs available for the user c. Execute ROM resident tests d. Provide a map of all memory and I/O page locations e. Enter setup mode. When dialog mode is entered the ROM program prints out the message shown in Figure 4-3 at the console terminal and waits for the user to select a command. Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: FIGURE 4-3 DIALOG MODE COMMANDS when dialog mode is entered the user has six commands The from. to choose six commands are listed in the command line for user convenience. The user may obtain a brief description of each command by typing H followed by pressing the RETURN key or by typing ? only. All of the commands may be executed by typing only the first letter of the command followed by pressing the RETURN key. For example, the Map command can be invoked by typing either M or Ma or Map followed by pressing the RETURN key. On input all lower case letters are converted to upper case and leading spaces and tabs are ignored. If Use the DELETE key to delete the previous character typed. code ROM the video is EEPROM the in on the terminal type selecti will erase the previous character on the screen when the DELETE key is depressed. If the terminal type is hardcopy the ROM code The will use slashes "/" to identify all deleted characters. typing by user may at any time delete the entire command line CONTROL (CTRL) U. NOTE Typing a CTRL U depressing the (or CTRL CTRL key depressing the CTRL U (or R) is performed by R) while simultaneously key. The user can also type CTRL R which will retype the command line. 4-4 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING CTRL R 1is normally used when the terminal type is hardcopy to clear up command lines where the DELETE key has been wused. For both CTRL R and CTRL U the ROM code will print out a short prompt first. Figure 4-4 shows an example of CTRL U being typed. Neither shows CTRL U or CTRL R is echoed by the ROM code. an example of CTRL R use. Figure 4-5 ) NOTE following the in user inputs All examples are examples and underlined. The RETURN text as key is specified in the <CR>. Commands are Help, Boot, List, Setup, Map Type a command then press the RETURN key: KDJ11-B and Test. B DX5 CTRL U > FIGURE 4-4 CONTROL U EXAMPLE Figure 4-4 shows an example of CTRL U being typed to clear up the command line selection is Commands without are Help, Type a command KDJ11-B retyping all of 1it. The terminal type hardcopy. >B Boot, List, then press the Setup, Map and RETURN key: Test. B DX/X/Ul CTRL R DUl FIGURE 4-5 CTRL R EXAMPLE Input is limited to 16 characters and spaces. There are no cases where any of the commands would need more than 16 characters. 1If the user types more than 16 characters the ROM code will delete all of the input and retype the KDJ11-B prompt and wait for input. Figure 4-6 character is Commands Type a KDJ11-B shows an an example. equivalent are Help, command to Boot, then press typing CTRL List, Setup, the RETURN U Typing Map key: and the Test. 12345678901234567 > FIGURE 4-6 CTRL U EQUIVALENT EXAMPLE 4-5 seventeenth BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING The ROM code will ignore character, or the printable character echoed as Note that code is any space or tab typed second tab or space typed in between. All tabs are prior to a in a row without a converted to and spaces. these rules accepting also input If be an invalid typed out input is and more an example an of Commands Type a are Invalid entry Commands are Type a at any time -the ROM user. will shows entry. Help, Boot, then Help, command the received an "invalid entry" message input will be requested. Figure 4-7 invalid command apply generally from List, Setup, Map and the RETURN key: MP List, Setup, Map and the RETURN key: press Boot, then press Test. <CR> Test. FIGURE 4~-7 INVALID ENTRY EXAMPLE 4.2.1 Help Command This command commands. typing ? command. types out a brief description of all available It can be executed by either typing H <CR>, only. Dialog mode is restarted at the end of Figure 4-8 shows an example of the Help command or by this being executed. Commands Type a are Help, command Boot, then Command Description Help Boot Type Load and List List boot Setup Enter Map Test Map memory Continuous Commands Type a are this the Setup, Map RETURN key: message a program start from and H a Test. <CR> device programs Setup mode Help, command List, press then and I/O page self test - Type Boot, List, press the Setup, RETURN CTRL Map key: FIGURE 4-8 HELP COMMAND DISPLAY 4-6 C and to exit Test. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Boot 4.2.2 Command be This command allows a device to bootstrapped. The command If the device arguments are the device name and the unit number. it. If the for user the prompt will name is left off the program 1left off the program assumes that unit” zero was 1is number unit desired. The unit number ranges from 0 0 to 255(10) depending on The device name is a one or two the device and the boot program. In most cases the letter mnemonic which describes the device. device name is two letters. and when typing the Boot command the user may either type B <CR> or switches, optional then type the device name, unit number and type B followed by a space and then type the number and optional device name, unit switches. The three optional switches used with the boot command are: /A Request to allow the user to type in a non standard CSR for unit address for the controller. /0 The unit number is octal instead numbers greater than of decimal 7. /U If the boot exists in the base ROM and also on the UBA, the base ROM boot and use the boot from the UBA override board or M9312 module. The format when using a switch is to type the device name and unit number followed by / and the switches. When there is more than one switch, use only one slash. - When the user types the Boot command without an argument, the ROM code will prompt the user for additional information with the following message: : Enter device name and unit number then press the RETURN key: the ROM code would 1list the At this point if the user typed ? boot programs available and then retype the "enter device name and unit number" message and wait for a selection. Wwhen the ROM code has a device name it searches for the first boot program with the same device name. The ROM code looks for matches in the following order. The /U switch effectively tells the ROM code not to look for the device name in the EEPROM or the CPU ROM, but go directly to the UBA ROMs or the M9312 present. lst area 2nd to search area to search EEPROM CPU ROM code 4-7 module if BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING area 3rd to search = UBA module if present 4th area to search = M9312 module boot any first, searched Note that since the EEPROM is always as a -name device same the with EEPROM the into 1loaded 1is that This boot. that replace boot in the CPU ROM will effectively replace a CPU ROM boot by effectively to user a allow would loading a boot in the EEPROM with the same name. Table 4-1 describes how the ROM code interprets user input. TABLE USER 4-1 ROM CODE ACTION ROM CODE ACTION INPUT 5oL Boot DLO B DL Boot DLI B DUS8 Boot DU unit 8 B DU10/0 Boot DU unit 8 B DU1O /A Address = 17760400 Boot DUl1l0 with non standard CSR address of 17760400 B DU 3/U Boot DU3 using UBA or M9312 B DU1l1l/UO Boot DU unit number 9 using UBA or B DU 10: Boot DU1l0 B DU11/U/0O rom boot M9312 BDUO ROM boot instead of CPU ROM. Invalid format will cause invalid entry B DUO instead of CPU ROM code. error message Invalid format. No space allowed in the device name DU. Invalid format. There must be a space between the Boot command and the device name. If the user types a colon after the wunit number it will be .s The single letter device name of B ignored (e.g., B DLl:) g non DEC boot devices on the supportin of method a s implement UNIBUS. The letter B causes the ROM code to transfer control to the address contained in location 17773024 of a ROM on the UNIBUS s location 17773024 is not odd. if the addresin 4-8 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING When the the CPU ROM passes control CPU ROMs and ROMs UBA the a unit number and Rl will contain will RO disabled, be will table. the translation by passed was address an contain 0 unless ROM the odd is UNIBUS the on 17773024 location in the address If If program will type out an invalid device message. from addresses all to respond not does device 17773776 the ROM program will also type out an UNIBUS the 17773000 to invalid message. device . which a module The single letter device name of B is used when to a similar 17773024 a switch pack that responds at address has address starting the Usually M9312 module is used in the system. would be set in the switch pack on the desired program the of module. Figure 4-9 shows shows an example of DL2 being booted boot using the command. Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: B DL2 <CR> Trying DL2 Starting system RT-11FB (S) .SET QUIE .R TT from V05.01 DATIME Date? [dd-mmm-yy]? FIGURE 4,2.3 List DL2 4-9 DL2 BOOT EXAMPLE Command The user enters this command by typing the letter L <CR>. This command will print out a 1list of all available boot programs found in the CPU ROM, the CPU EEPROM, or any M9312 type ROMs located on the UBA or an M9312 module 1if present. The information listed is the device name, allowable unit number range, source of the boot program and a short device description. The device name is normally a two letter mnemonic. In some cases the name may be a single letter. The device name must always be letters from A to Z. At input,.the ROM program always converts all lower case letters to upper case. The unit number range is the allowable range of unit numbers that is valid for a particular boot program. The 4-9 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING range varies from 0 to 255, depending on the device. If the unit number range information is blank, the ROM code will assume the range limit is O to 255. The unit number range for M9312 type ROMs is always left blank. The The source lists where the actual boot program is located. the of outside the on description is intended to be the name DL of name device a for device to be booted. An example would be the description would be RL0O2. Dialog mode is restarted at the completion of the 1list command. The mnemonic for each ROM found on either the UBA or the M9312 If will be checked against a list of mnemonics in the ROM code. the mnemonic matches an item in this list the ROM code will print If no match 1is found the out a description of that device. description will be left blank for that mnemonic. In order for a M9312 type ROM to be listed it must be in a M9312 type format as described in subsections 4.6.3 thru 4.6.6. Figure 4-10 shows an example of a screen display for the list command. Commands are Hélp, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: Device Unit name DU numbers Source 0-255 CPU ROM RD51, RD52, RX50, RC25, RA80, RA81, RA6O0 RLOZ DL DX 0-3 0-1 CPU ROM CPU ROM DD DK MU 0-1 0-7 0-255 CPU ROM CPU ROM CPU ROM TU5S8 RKO05 TK50, TU8l 0-1 CPU ROM <CR> Device type RLO1l, RXO01 DY L RX02 Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: FIGURE 4-10 LIST COMMAND DISPLAY BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING NOTE Figure 4-10 is an example and may not represent the exact screen display. 4.2.4 Setup Command all This command enables the user to 1list and/or change command This s. parameter boot parameters in the EEPROM including and all of the programmable parameters are discussed in detail in Subsection 4.3 4.2.5 Map Command d typing M <CR>. This command will The user enters this commanby try to identify all memory in the system and then map all to locations in the I/O page. Memory is mapped from location 0 for mapped not is Memory s. the 1I/0 page in 1,024 byte increment every location. The routine will try to identify the size of each memory, the CSR address for each memory if applicable, the CSR type (ECC or Parity) and the general bus type. It is important to note that if two memories share some common addresses or have CSR's with the same address the map command will not work properly. During mapping of memory if two or more memories are present they are not contiguous the descriptions with a blank line. ROM : code will separate and their After all memory is mapped the ROM code will wait for the user to press the RETURN key to continue the map, which will then typed out all addresses in the I/O page that respond. The I/O page map In addition, all goes from addresses 17760000 to 17777776. addresses that respond that are on the KDJ11-B or on the KTJ11-B are provided with a short description. There is no des- cription for addresses that respond and are on the external bus, with the exception of memory CSR's, if present. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING To command. Dialog mode is restarted at completion of the map help prevent the data shown from being scrolled off the screen on video terminals the ROM code will wait for the user to press <CR> The ROM code always the data might overflow the screen. anytime at least 24 lines of 80 coloumn assumes the terminal can display data. Figure 4-11 shows an example of a map command printout. Commands are Help, Boot, List, Setup, Map and Test. M <CR> Type a command then press the RETURN key: Memory Map | Starting Address Ending Size in CSR CSR Bus 1024 17772100 Parity PMI address K Bytes address 00000000 - 03777776 type Press the RETURN key when ready to continue <CR> I/0 page Map Starting Ending Address address 17765000 17765776 17770200 17770376 CPU 17772100 17772150 Memory 17772152 17772200 17772300 17772276 17772376 17772516 17773000 17774400 17773776 17774406 17777520 17777524 17777546 17777560 17777572 17777600 17777730 17777744 ROM or Unibus Map CSR Supervisor I and D PDR/PAR's Kernel I and D PDR/PAR's MMR3 CPU ROM BCSR, or UBA ROM PCR, BCR/BDR Clock CSR Console SLU MMRO,1,2 User I and D 17777566 17777576 17777676 =17777734 17777752 DCSR, MSER, DDR, CCR, 17777766 17777772 CPU Error PIRQ 17777776 PSW Commands Type a are Help, command then EEPROM List, Setup, Map the RETURN key: press FIGURE PDR/PAR's KMCR MREG, Hit/Miss 4-11 MAP COMMAND and DISPLAY Test. type BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.2.6 Test This command tests in runs 30 all 1is Command causes a the ROM code continous applicable complete: tests If and the console. print out errors if At the time the test loop the total number of 1loops any. also type a the test most code test the by at the power test up 70 after error typing is exited and the after the loop general loop number of starts restarts occurs The user may exit run ROM then an error may to The entered. The user loop. and test routine CTRL C at is the the ROM code will total number of test command and if the test is applicable the ROM code will loop on that specific test only until an error occurs or CTRL C is type. If the test number selected is not a loopable test the general test loop will be entered and all loopable tests will be run. NOTE CTRL C is console Figure 4-12 command by user aborts not echoed by the ROM code on the terminal. shows an example of “the user entering typing T <CR> which will run all loopable the testing sequence after four passes by the test tests. typing CTRL The C. Commands Type a are Continuous Passes Total Errors Commands a are = Boot, then self Total Type Help, command press test - List, the Type Setup, Map RETURN key: CTRL C to exit and Test. T <KCR> CTRL C 4 0 Help, command FIGURE Boot, then press 4-12 List, the Setup, Map and Test. RETURN key: TEST COMMAND EXAMPLE Figure 4-13 shows an example of the user looping on only The user aborts the test loop by typing CTRL C after 202 test 60. passes. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Commands Type a are Looping on test 60 Total Passes = 202 Errors = 0 Type SETUP a are the Type CTRL - Help, List, press Boot, Setup, RETURN C Setup, Map key: FIGURE 4-13 LOOP-ON-TEST entered by and T Test. 60 CTRL RETURN press is exit the then COMMAND to Map key: List, command MODE Setup mode Boot, then Total Commands 4.3 Help, command and <CR> C - Test. EXAMPLE DESCRIPTIONS typing S <CR> in dialog mode. Setup mode allows the user to list or change most ¢f the parameters in the EEPROM. Setup mode also allows changes to any bootstrap programs stored in the EEPROM. Setup mode has fifteen commands. After power up or restart and the completion of all code loads the first 105 bytes of the EEPROM beginning at location 2000. This area in memory is as the Setup Table. The setup TABLE contains parameters except The EEPROM may system. The the EEPROM resident contain various first 105 bytes boot tests the ROM into memory referred to all of the programs. types of information for is information needed by the the ROM code to configure the KDJ11-B (CPU) and the KTJ11-B (UBA) and to determine the boot device, test selections and modes. Other information in the EEPROM could be user bootstrap programs and a foreign language file. Setup mode allows changes to the first 105 bytes and to the user bootstrap programs. The foreign language area if present cannot be changed in setup mode. When setup mode is first entered it "types out a commands and provides a short description of Figure 4-14 shows an example of setup mode being dialog mode after the user types S <CR>. 1list of all each command. entered from BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: S <CR> KDJ11-B ROM V7.0 KDJ11-B Setup mode Command Description 1 Exit 2 List/change parameters in the Setup table 3 List/change boot translations in the Setup table 6 List/change the Automatic boot selections in the table Reserved List/change the switch boot selections in the table 8 Initialize the Setup table 11 12 Delete an EEPROM boot Load an EEPROM boot into memory 14 Save boot into the EEPROM 4 5 7 9 10 13 15 List boot programs save the Setup table into the EEPROM Load EEPROM data into the Setup table an EEPROM boot Edit/create Enter ROM ODT Type a command then press the RETURN key: FIGURE 4-14 SETUP MODE COMMAND DESCRIPTIONS NOTE The version number of the ROM code is printed out This at the beginning of the Setup mode message.versi on is code ROM manual revision assumes the ) 7.0 (V7.0). iption of each The following paragraphs provide a detailed descr r followed To execute a command, type the command numbetype command. CTRL C may by pressing the RETURN key. At any time the n user of ning begin the to retur to return to dialog mode or CTRL Z to setup mode. NOTE Never terminate a change of any parameter with CTRL C or CTRL 2. If this is done the change is Always use the terminating ignored and lost. character RETURN CTRL C or CTRL Z. after 4-15 any change and then use BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING the at or Z, CTRL typing by restarted 1is When setup mode Commands 2 thru 15, the ROM code will print out a completion of short command message instead of the full list of commands. The user may either type in a new command now, or press <CRY to list command short the shows 4-15 Figure the full command menu. ‘ ) message. KDJ11=-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4.3.]1 Setup Command 4-15 SHORT COMMAND MESSAGE 1 This is the exit command for the set up mode and returns the user to Dialog mode. Dialog mode is also entered if CTRL C is typed. 4,3.2 Setup Command 2 This command prints out the current status of various parameters and allows the user to change them if desired. When setup mode command 2 is entered parameter in the ROM code prints out the current status of all parameters, repeats the first parameter, and waits for user input. The user can type <CR>s to position the program at the desired parameter to be changed. The user can also go directly to the parameter by typing the letter to the left of the the first list. NOTE After changing any parameters in the Setup Table, Command 9 (Save) should be executed. To change a parameter the user types in the new value and <CR>. Typing <CR>, Line feed or . will cause the ROM code to proceed to the next parameter. Typing Or - will cause the ROM code to proceed to the previous parameter. Any of these characters can be used to change a value. Figure 4-16 shows an example of command 2 being entered. This example also shows the values of the parameters if the "initialize setup table" command 8, is executed in setup mode. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING ' °* KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: 2 <CR> List/change parameters in the Setup table A - ANSI Video terminal B - Power up C - Restart O=Dialog, 0=Dialog, D - Ignore battery (1) - 0=No, l=Yes (1l)=Automatic, 2=0DT, 3=24 (l)=Automatic, 2=0DT, 3=24 E - PMG 0-(7) l=.4us, 2=.8, F - Disable clock CSR G - Force clock interrupts H - Clock I J K LL M O - P Q R S 0=No, . 0=No, 2=Dis 173, 0=No, 0=No, Disable Setup mode Disable all . testing - Enable UNIBUS memory - Disable UBA ROM - Enable UBA cache (1) - Enable 18 bit mode test (1) CTRL Z to exit or press ANSI Video terminal the RETURN (1) FIGURE 0=No, 4-16 COMMAND 2 =1 =1 = § =7 =0 = 0 = A\ l=Yes l=Yes 3=Both l=Yes l=Yes =1 =0 =0 =0 = 0 0=No, List/change parameters in the Setup table Type l=yes 4=3.2,...7=25.6 0=No, l=Yes 0=No, l=Yes O=Power supply, 1=50Hz, 2=60Hz, 3=800Hz - Enable ECC test (1) = Disable long memory test - Disable ROM 0=No, 1=Dis 165, - Enable trap on Halt - Allow alternate boot block N - 0=No, 3=1.6, = 0 0=No, l=Yes l=Yes =0 0=No, 0=No, 0=No, 0=No, l=Yes l=Yes l=Yes l=Yes =] =0 =1 = 0 key for l=Yes No = , =0 change 0 New = EXAMPLE NOTE If 124 KW of UNIBUS memory 1is present, the last two parameters will not be present (Enable UBA cache and enable 18-bit mode). When this condition occurs UBA cache is always disabled and 18-bit mode is forced unconditionally. ' The following parameter A - ANSI paragraphs specified Video in terminal a detail Figure description of each Command 2 4-16. present: When set to 1 this indicates that the console terminal is an ANSI video terminal. When 0 it indicates that the console terminal is ha;d copy oOr non %NSI compatible-. When video terminal 1is 4-17 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING selected the DELETE key will erase screen. The ROM space, then backspace terminal is identifies power up code selected deleted accom- to plishes the previous character on the this the console and the DELETE key <characters by using if ANSI video terminal is by sending terminal. a backspace, When hardcopy 1is wused the ROM code the slash character. At selected fthe ROM code a VT52 is will the cursor at line 9 an ANSI screen clear and then position send ROM The video terminal parameter is used only by the coloumn 1. This information is not used by the operating system. code. NOTE VTS52's are not ANSI compatible. If present you must set this parameter to 0 to prevent the Clear Screen command from disabling the B Power VT52. up mode and C restart mode: started When the ROM code is These are two separate parameters. it checks a status bit to determine if the unit is powering up or if the front panet RESTART switch was activated. The ROM code then uses the appropriate mode selected. There are four choices for the power up mode and the restart mode. The user may define the the action taken by the ROM code at power up or restart to be the same or different. 0 - Dialog mode: At completion of the diagnostics, dialog mode 1is entered. Dialog mode is also entered any time the force dialog switch is set to the on position regardless of the modes selected here. (See subsection 4.2 for a description of Dialog mode.) 1 - Automatic mode: At completion of the diagnostics the ROM code enters an automatic boot routine that will try to boot a previously selected device or device's. The device's are selected in the EEPROM (refer subsection 4.3.4). The list of devices can be from 1 to 6 devices 1long. Each device 1is tried until a successful boot occurs or there are no more devices to try. The default list of devices 1is A, DLO, MSO and MUO 1in this order. "A" is a special single letter mnemonic which causes the ROM code to boot the first disk MSCP device that it can. Removable media devices are tried before fixed media devices. 2 - ODT mode: At completion of a very limited set of tests the ROM code executes a halt instructicn and passes control to J11 micro ODT (refer to subsection 4.7). If the user types P at this point without changing any registers the ROM code will contine normal testing and then enter dialog mode. This mode would normally be used in debug environments. The ROM code will not change any 1locations in memory before entering ODT mode. 4-18 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 3 - 24 mode: At completion of a limited set of tests the ROM code loads the PSW with the contents of location 26 and then transfers control to the address 1located 1in 1location 24, This mode is used when non volatile memory is present and power fail recovery is desired. The ROM code will not change any locations in memory before executing mode 24. ‘D - Ignore battery: This parameter mode 1is set meaning that is used to 24 only when the current power up or restart (3). Normally this parameter is set to 0 the memory battery ok signal must be present in order to execute mode 24. If this parameter is set to 1 mode 24 is executed regardless of the status of the battery. At power up, 1f the mode selected 1is mode 24 and the ignore battery parameter is set to 0 and the battery status indicates that voltages were not maintained, the ROM code will ignore the power up selection and use selection 1is the also mode restart 24 the selection. ROM code will If the default restart to dialog mastership count mode. E - PMG This in is count: parameter the BCSR. disabled. sets The When the value range is set, the of the processor 0 to 7. When set to count value enables zero the the counter KDJ11-B to suppress DMA requests and give the processor bus mastership during the next DMA arbitration <cycle after the counter overflows. The processor will only take the bus for one cycle before relinquishing control of the bus to a regquesting device. The following table shows the time needed for the counter to overflow for the different wvalues of the PMG count. This parameter is normally set to 7. vValue Time 0 * 1 2 3 4 5 6 7 for counter to overflow Disabled* - 0.4 0.8 1.6 3.2 6.4 12.8 25.6 usec usec usec usec usec usec usec The PMG count of 0 (Disabled) for most typical systems, and special applications. is is not recommended reserved for BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Disable F - Clock CSR: When set to 1 this parameter disables the clock CSR at address When set to 0 the clock CSR 1is enabled. This 17777546. . parameter is normally set to O. G - Force Clock Interrupts: when set to 1 the clock will unconditionally request interrupts when the processor priority is 5 or less. When set to 0 the clock can request interrupts only if the clock CSR is enabled, clock CSR bit 6 1is 1 and the processor priority is 5 or less. This parameter is normally set to 0. NOTE If the command parameter Force Clock Interrupts is selected the user should always disable the clock CSR since the CSR has no control over the clock. H - Clock Select: This parameter determines the source of the The choices are listed below. 1 2 3 I - to be used. Source Value 0 <clock Clock sourced from backplane pin BR1l. The power supply normally drives this Clock sourced on the Clock sourced on the Clock sourced on the Enable ECC signal at 50 or 60 Hz. KDJll-B at 50 Hz KDJ11-B at 60 Hz KDJ11-B at 800 Hz Test: When set to 1 this parameter enables the ECC memory test to be run on any ECC memorys present except UNIBUS memory. If the system contains a mix of ECC and non ECC memory the ROM code will run the ECC tests only on the ECC memories. The ROM code uses bit 4 of the memory CSR to determine if the memory is ECC or If bit 4 is a read/write bit and can be written as a 1l parity. wWhen this and a 0, the ROM code assumes the memory 1is ECC. . This bypassed always is test ECC the 0 to parameter 1is set parameter is normally set to 1 even when ECC memory is not This parameter would be reset if an ECC memory was present. installed whose ECC hamming code does not match that of an MS11-P type memory. BOOTSTRAP J - Disable When set data to test Long 1 AND DIAGNOSTIC ROM PROGRAMMING Memory Test: this parameter bypasses the memory for all memory above 256 K bytes. address shorts data test is run on all parameter is normally set to 0. address shorts When set to 0, available memory. - . the This NOTE If the long memory test is disabled and parity memory exists above 256 K bytes it is very likely that memory will contain parity errors after power K - Disable up. ROM: This parameter allows the user to selectively disable all or part of the ROM code after the selected device has been booted. Normally the ROM code on the CPU responds to two 256 word pages in the I/0 page. One page responds to address's from 17773000 to 17773777, the other page responds to address's from 17765000 to 17765777. up or Both of restart. these.pages After a device are automatically enabled at power is booted, pages may be disabled by the ROM code. the var- set to iations of this parameter. one or both The following This parameter of 1is 0. Value Rom 0 None 1 2 17765000-17765777 3 17765000-17765777 Pages and Disabled. 17773000-17773777 17773000-17773777 NOTE If the ROM code is booting directly from a M9312 type boot ROM located on either the UBA module or the M9312 module, the ROM code will automatically disable the CPU ROM in the 17773xxx address range and will enable the ROMs on the board which were selected for booting. This action 1is taken regardless of the status of the disable UBA ROM parameter (Q) and the disable ROM parameter (K). L. - Enable Trap on Halt: these table lists normally BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING If this parameter is set to 1 the processor will trap to location 4 if a halt instruction is executed in kernel mode. If this parameter is set to 0 the processor will enter Jll micro ODT if a halt instruction is executed in kernel mode. This parameter is ) ' normally set to 0. M - Allow Alternate Boot Block: After the boot block of a device is loaded into memory the ROM code looks at word locations 0 and 2 to see if the device looks If the data is not correct, the ROM code will type out bootable. an error message indicating that the media is not bootable. When this parameter is set to 1 the ROM code looks for location 0 be If this parameter is set to 0 th ROM code any non zero number. looks for location 0 to be a value of 240 to 277 and for location 2 to be 400 to 777. This parameter is normally O but may have to be changed to 1 to allow some user's operating systems to boot properly. N - Disable Setup Mode: enter mode dialog in setup mode from dialog mode. The command lines command the types user If the will not show the setup command. If this parameter is set to 1 the user will not be able to the response will be invalid command. If Force Dialog mode 1is selected, regardless enabled unconditionally parameter. unauthorized This entry allows parameter into Setup mode. is mode setup then of the value of this prevent to users some This assumes the user has the force dialog switch under some type of physical control. O - Disable all testing: If this parameter is set and Force Dialog mode is the ROM not program will bypass virtually all testing. selected, No location in memory will be changed unless the selected boot program makes This is a special parameter which should not be used a change. user It has been provided for cases where the unless necessary. user the when and up, response at power immediate almost needs needs the contents of memory to be left unaltered. NOTE If all exists testing then it 1is disabled and parity memory is very likely that memory will contain parity errors after power up . BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING P - Enable UNIBUS Memory Test: If this parameter is a 1 then any available UNIBUS memory will be tested. When 0 UNIBUS memory is not tested. This parameter is normally a 1. ‘Q - the by taken Disable This UBA parameter boot. normal If UNIBUS memory code. ROM is not present then no action ROM: is copied to bit this when bit 3 in the is 1, DCSR of the UBA after address range of 17773000 to 17773776. the UBA ROMs are enabled. ignored. in the UBA When this bit is is normally 0. boot either a UBA or M9312 ROM boot, to tries This bit a the UBA ROMs are disabled. This allows other ROM boards on the UNIBUS to show up ROM |is When a 1 user this parameter 1is ‘ NOTE M9312 If the ROM code is booting directly from a the module, M9312 the located on ROM boot type ROM CPU the disable ROM code will automatically in the 17773xxx address range and the ROMs on the The ROMs on the M9312 module will be UBA module. enabled This action 1is taken regardless of the and (Q) status of the Disable UBA ROM parameter the R - ROM parameter (K). UBA Cache: Enable When a 1, Disable this parameter causes the UBA cache to be enabled and to be tested by the ROM code. 1If a failure occurs during testing of the UBA cache then it will be disabled. When 0 the UBA cache is always disabled and is not tested. This parameter is normally a l. S - Enable 18 bit Mode: This bit is copied occurs. This this 4.3.3 to bit 5 of the KMCR on the causes 18 bit addressing only. Setup parameter Command is normally a When a 0, UBA. When a 1 22 bit addressing O. 3 This command prints out the current contents of the translation table and allows the translation table to be changed. The translation table is used to allow devices to be booted using non ~ 4-23 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING When the ROM program enters the boot standard CSR addresses. routine, RO contains the unit number and R2 contains the device The ROM code tries to find a match in the name (mnemonic). If no translation table for the device name and unit number. address CSR default the use match 1is found the boot program will for the device. If a match is found the translation table will define the CSR address to be used. .Figure 4-17 is an example of this command. KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: 3 <CR> List/change boot translations in the Setup table TT1 blank TT2 TT3 TT4 TTS TT6 TT7 TT8 TT9 blank blank blank blank blank blank blank blank . Type CTRL Z to exit or press the RETURN key for No change TT1 blank Device name = FIGURE 4-17 SETUP COMMAND 3 EXAMPLE The ROM code is now waiting for the user to enter a new device If the user does not desire to change any items in the name. translation table The mode prompt. entry by pressing types the user CSR address. he/she would Type CTRL 2 to return to the setup user may skip over any entry and go to the next To enter a new device or change an entry <CR>. in the new device name, the unit number and the system that In the example shown in Figure 4-18, the user has a the standard at contrcller UDA50 a has one RA80 and a RA60 using a KLESI-U with RC25 a has also The user address of 172150. the same share KLESI-U the and UDA50 the Since controller. standard CSR address one of them must be set to respond to a 1In this example the KLESI interface is set to different address. The RQZS has a unit number plug set respond to address 17760500. 4-24 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING for units 4 and 5. The RA80 is unit 0 and the RA60 is unit 1 and 2. Since the RC25's interface is at a non standard CSR address and there are two unit numbers there will be two entries in the translation table for units TT1 blank Device name Unit number CSR address DU4 TT1 TT2 = = 4 and 5. ' DU <CR> 4 <CR> = 17760500 <CR> address 17760500 blank Device name Unit number CSR address DU5S5 TT2 TT3 Device FIGURE = DU <CR> 5 <CR> 17760500 <CR> address 17760500 = blank name 4-18 = CTRL 2 TWO-ENTRY TRANSLATION TABLE EXAMPLE The translation table also provides a means of handling multiple For example if the user RL02 controllers. as such controllers the had two RL0O2 controllers with six drives of which 2 where on second controller at address 17760400 the translation table could first the on be would 0-3 Drives be set up to handle this. controller at the standard address and would not require any be would controller second the on 1 Drives 0 and entries. labeled as drives 4 and 5 and entered into the translation table. Since RL0O2 controllers only recognize unit numbers from 0-3 the unit numbers 4 and 5 would have to be translated to unit numbers the into entries Figure 4-19 shows an example of the 0 and 1. translation table. ‘ blank name number = TT1 DL4 TT2 blank DLO Device name Unit number CSR address TT2 DL5 TT3 Device 17760400 address = = DL <CR> 5 ] <CR> = DL1 = DL <CR> 4 0 <CR> on Unit CSR address unan TT1 Device <CR> 17760400 17760400 <CR> address 17760400 blank = name CTRL 2 FIGURE 4-19 UNIT NUMBER TRANSLATION EXAMPLE | 4-25 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.3.4 Setup Command 4 This command allows the user to select the devices to be-tried in the automatic boot sequence. The user creates a small list that defines the devices and the order in which they are to be tried. One entry is needed to define a device and its unit number. If the same device is used more than once with different unit numbers, then one entry is needed for each unit number. NOTE The selections for this command use the in the EEPROM as command 6 locations same that follows. When Command 4 is executed the ROM code will prompt the user for a device name. The user would then type in either the single or double letter mnemonic associated with the device to be selected. The ROM code will then prompt for the unit number. The ROM code will continue prompting for all six entries in the table. Figure 4-20 shows an example of command 4. In this example the user adds the boot for the RX02 unit 1 (DY) by replacing the Exit The exit name (E) with DY and typing in the unit number next. name is typed for the next entry. BOOTSTRAPAAND DIAGNOSTIC ROM PROGRAMMING KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: 4 <CR> List/change the Automatic boot selections in the Setup table A = B E L Disk MSCP automatic boot = External ROM boot = Exit automatic boot = Loop continously Boot Boot Boot 1 2 3 = Boot 4 = MUO Boot.5 = E 6 = blank Boot = = A DLO MSO Type CTRL Z to exit or press the RETURN key for No change Boot 1 Device Boot 2 Device Boot 3 = A name = <CR> = <CR> =-DLO name = MSO Device name = <CR> Boot 4 Device = MUO name = <CR> = DY = = E 0 Boot 5 Device = E name <CR> =1 <CR> Unit number Boot € = blank Device name Unit number <CR> <CR> KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-20 SETUP COMMAND 4 EXAMPLE BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Table 4-2 lists names the and the four special associated ROM TABLE A mnemonic device 4-2 ROM CODE ACTION The ROM code will boot the first bootable disk MSCP device it can find. The ROM code will try removable media units first, B single-letter action. then fixed media units. This mnemonic causes the ROM code to check for a UNIBUS ROM board in address range of 17773000 to 17773776. If the ROM exists and location 17773024 is not odd the ROM code will disable the internal CPU ROMs and the UBA ROMs and jump to the location specified in location 17773024 of the UNIBUS ROM board. Usually this board M9301 or user-supplied equivalent. would be a M9312, E The only purpose of this mnemonic is to indicate to the ROM code that there are no other devices to try in the list. This is used when there are five or less devices in the list. It follows the last device in the list to be tried. If all six entries are filled in the list then this mnemonic is not needed, the list will terminate automatically after trying the last entry. When this mnemonic is reached the ROM code will restart the boot sequence from the beginning and print error messages for each device that fails to boot as they are tried. After the second pass through the list without successfully booting a device the ROM code will enter dialog mode. L This mnemonic will also mark the end of the list but when this mnemonic is reached the ROM code will restart the boot sequence at the beginning of the list and continue trying to boot each device in the list until either a successful boot has occurred or the sequence is terminated by the user typing CTRL C. All messages are suppressed when the ROM code the boot list. boot error is looping on The action taken by the ROM code for the four single letter mnemonics A, B, E and L applies only to automatic boot mode with the exception of B which can also be executed from the dialog mode boot command. The dialog mode boot command would treat the 4-28 BOOTSTRAP AND DfAGNOSTIC ROM PROGRAMMING single letter mnemonics of A, E and L as invalid devices if they were used as the device name in the boot command. If the user creates a bootstrap program and loads it into the EEPROM, that program should not be given a device name of either A, B,E or L All other single letter since these are already defined. mnemonics are for use. free 4.3.5 Setup Command 5 If this command This command is reserved and is not used. changes are made. no and d entered setup mode will be restarte 4.3.6 Setup Command 1is 6 This command allows the user to define the value of three of the eight switches at the edge of the CPU module to boot specific possible This command defines six of the eight devices. combinations of switches 2-4, the other two combinations have a fixed definition that cannot be changed. Refer to Appendix H for additional information. Command 6 is not normally switches might be defined is: used. The only time that these and a. If the system had cabling between J3 on the CPU module b. The user desired to define six positions such that each position causes the ROM code to enter Automatic boot mode a remote 8-position rotory switch, and after testing is completed, and attempt to boot only one device which was defined by this command in Setup mode. Normally the three switches are off. 1If automatic boot mode is selected, the ROM code will use the list defined by command 4 in Setup mode and try to boot each item in the list one at a time until all selections in the list have been attempted. When switches 2-4 are set to one of the six combinations shown in the e EXample and Force Dialog mode is not selected the ROM code will enter the auto boot mode and attempt to boot device selected by this commmand. the only the ROM code will print out the normal error message and dialog mode. NOTE The six selections in the table are stored in the EEPROM as the six boot the of area same selections decsribed in Command 4. 4-29 one If the boot is unsuccessful enter BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Figure 4-21 shows an example of setup mode command 6 with three of the six possible positions being defined to select DUO, DUl and DU2 and the other three all KDJ11-B Setup mode Press the RETURN key Type a command List/change Switches Switches Switches Switches Switches Switches Type Switches Device the 2,3,4 2,3,4 2,3,4 2,3,4 2,3,4 2,3,4 CTRL Z then to for Help press the switch on on on off off off 2,3,4 boot on off off on on off exit or on name being defined to RETURN key: selections off on off on off on press on select 6 in DLO. <CR> the Setup table = DUQ = DU1 = DU2 = DLO = DLO = DLO the off RETURN = key for No change DUO = FIGURE 4-21 SETUP COMMAND 6 EXAMPLE 4.3.7 Setup Command 7 This command performs the exact same in the List command in Dialog mode. setup mode for user convenience, function as The command see detailed description of this command. the completion of this command. 4.,3.8 Setup This command Command the List command is duplicated in Subsection 4.2.3 Setup mode is contents of the for a restarted at 8 initializes the current setup table in memory to the default values. This command does not affect the contents of the EEPROM itself. The setup save command 9 must be executed 1in order to save the setup table into the EEPROM. This command only affects parameters associated with commands 2 to 6 of Setup mode and does not affect any data that is not in the first 105 bytes of the EEPROM. The following Command 8 All is items list the value of the parameters after entered: parameters are set which are to set listed 0 to under setup command with the exception of 1. (See Figure 4-16.) 4-30 A, 2 of B, setup C, I, P mode and R BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING - .- All are entries cleared The automatic boot will be set in the translation table and will list as blank. to A, selection DLO, MSO, list MUO, E, under setup wunder blank. setup ' command 3 command 4 Since Setup Command 6 shares the same area as Command 4 it's list of parameter values will be identical to that of Command 4. To enter this command the user must type 8 <CR>. After the ROM code prompt, the user must type 1 <CR>. Command 9 (Save) should be executed after Command 8 if the user wishes to retain the defaults in the EEPROM. Figure 4-22 shows an example of Command 8 execution, KDJ11-B Press Type Setup mode the RETURN key for Help a command then press the Initialize the Are you sure ? Type a command RETURN key: 8 <CR> l1=Yes. 0=No, then press the RETURN key: 1 <CR> Setup KDJ11-B Setup mode Press the RETURN key Type a command FIGURE 4.3.9 Setup Command table for Help the RETURN key: then press 4-22 SETUP COMMAND 8 EXAMPLE 9 This commands copies the current contents of the setup table in memory into the EEPROM. The command should be executed after any changes are made to the EEPROM in Setup mode. This is the only that writes anything into the first 105(10) bytes of the command EEPROM. When saving data into the EEPROM the ROM code will only write the locations that need to be written. This command will always write a new and correct checksum into the EEPROM unless a failure occurs. If a location cannot be written the ROM code will try once more and then report the error. It takes approximately 15 ms to write each location. If command 9 is entered and no changes have been made to the setup table, the ROM <code will print out a message saying no change were made and then restart Setup mode. If changes are to prompt the user to make sure they will code ROM made the be desire to make the changes. Figure~” 4-23 shows an example of Command 9. 4-31 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN Save the Setup table Are you sure ? the 9 <CR> the RETURN key: 1 <CR> into the EEPROM 0=No, l=Yes Type a ¢ommand then press Writing key: EEPROM KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-23 SETUP COMMAND 9 EXAMPLE 4.3.10 Setup Command 10 This command will restore the setup table in memory with the values actually stored in-‘the EEPROM. This command allows the user to restore the setup table after making some temporary changes. It is also used to load the actual data from the EEPROM into the setup table if an error occurred during the EEPROM checksum tests. When an error occurs during the EEPROM checksum tests, the ROM code assumes the data is bad and loads a set of default values into the setup table and uses them. 1In this case the user could load the actual data and then verify the data is OK before trying to save it back into the EEPROM. Figure 4-24 shows an example of Command 10. KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: 10 <CR> Load EEPROM data into the Setup table Are you sure ? 0=No, l=Yes Type a command then press the RETURN key: KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-24 SETUP COMMAND 10 EXAMPLE 4-32 1 <CR> BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.3.11 Setup Command 11 If this This command allows the user to delete an EEPROM boot. command is executed the ROM code will prompt the usef for the device name of the EEPROM boot to be deleted. After the device name is typed- in the ROM <code will look for the first boot program in the EEPROM with that device name and delete it, 1if If there are any boot programs following the deleted found. code will automatically move all of these wuse the space made available by the deleted program the ROM programs up to program. (See Figure 4-25.) KDJ11-B Setup mode Press the RETURN key for Help Type a command then.press the RETURN key: an Delete 11 <CR> EEPROM boot Type CTRL Z to exit or press the RETURN key for No change = CC name Device ? 0=No, Are you sure KDJ11-B Setup mode <KCR> l=Yes Type a command then press the RETURN key: 1 <CR> Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-25 SETUP COMMAND 11 EXAMPLE 4,.3.12 Setup Command 12 This command is used to copy an EEPROM boot program into memory. command is executed, the ROM code prompts the user for the When 1in 1loaded be to boot program EEPROM the device name of the then be examined and/or edited using program can The memory. Figure 4-26 shows an example of Command 12. Setup Command 13. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING ‘ for Help KDJ11-B Setup mode Press the RETURN key Type a command then press the RETURN key: Load EEPROM boot an 12 <CR> into memory Type CTRL Z to exit or press the RETURN key for No change = name Device Are you sure ? CC <CR> 0=No, l=Yes Type a command then press the RETURN key: KDJ11-B Press 1 <CR> Setup mode the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-26 SETUP COMMAND 12 EXAMPLE 4.3.13 Setup Command 13 This command is used to either create a new EEPROM boot program or to edit a program previously loaded with Command 12 above. Command 13 allows the user to change the device name, the device description, the allowable unit number range, the beginning and ending addresses of the program in memory, and the start address of the program. When these changes are complete the ROM code enters ROM ODT which is is a ROM code version of J11 Micro ODT. When this command for EEPROM the in space le availab the list will it entered first bootstrap programs Figure 4-27 shows an example of Command 13. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: 13 <CR> Edit/create an EEPROM boot Type CTRL Z to exit or press the RETURN key for No change 1410 Bytes free in the EEPROM = AA New = EA <CR> = 000600 New = 10000 <CR> Last byte address = 000615 New = 10177 <CR> start address = 000600 New = 10000 <CR> Device name Beginning address Highest Unit number = 3 | New = 255 <CR> Device Description = EA BOOT New = RM02,RM03 <CR> Enter ROM ODT xxxxxx/ = open word location xxxxxx if address even, byte if odd = close location RETURN . or LF = close location and open next close location and open previous ROM ODT> ROM ODT> 010000/000000 012705 <CR> 010002/000000 101 <CR> ROM ODT> 010004/000000 12706 <CR> ROM ODT> 010006/000000 1000 <CR> etc. ‘ Type CTRL 2 exit back to the setup mode menu. FIGURE 4-27 SETUP COMMAND 13 EXAMPLE program in The beginning address is the first location of the of the last byte of ‘memory. The last byte address is the address s of data + code used in memory. If in doubt, use the last addres it will since 5 for this value. Do not use a much larger number waste EEPROM space. The start address is the address that the ROM code willsame©pass as The start address does not have to be the control to. range the in value a and the load address but it must be even defined by the load and ending addresses. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING The highest unit number defines the allowable range of valid unit If the value is set to 3 the allowable numbers for this device. Tf a unit range is 0 to 3. The highest range is 0 to 255. then a range in not is it and time number is typed in at boot occur. will invalid unit number error The device description is an optional but recommended description The maximum length of this name is 11 of the device name. be the name that characters or spaces. The name should normally is physically marked on the outside of the device (i.e. RLO2). 4.3.14 Setup Command 14 boot program saving a boot program into memory the device name of the program This command allows the user to save the existing located in memory into the EEPROM. This is the only command that The other commands' only actually writes a boot into the EEPROM. When change a copy of the boot program that resides in memory. must not match the name of an existing program in the EEPROM. If the program name already exists the user must delete that program If two or first or change the name of the program to be saved. name same the with EEPROM the into written more programs were an shows 4-28 Figure used. and found be only the first one would example of Command 14. KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: Save boot into the 14 <CR> EEPROM Type CTRL Z to exit or press the RETURN key for No change Are you sure ? 0=No, l=Yes Type a command then press the RETURN key: Writing the EEPROM - Please wait KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-28 SETUP COMMAND 14 EXAMPLE 1 <CR> BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING executed where Figure 4-29 shows an example of Command 14 being the data in the setup table matches the data in the EEPROM, thus no changes were made. KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: into the already in Save boot Boot is No 14 <CR> EEPROM the EEPROM changes made KDJ11-B Setup mode . Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4.3.15 Setup Command 4-29 SETUP COMMAND code will open EXAMPLE 15 This command puts the user into ROM ODT ROM 14 up the address (see Figure 4-30). The defined by the beginning address of the program. ROM ODT is not the same as Jll micro The only purpose of ROM ODT is to allow the user to create ODT. or edit a small bootstrap program to be stored in the EEPROM. are In ROM ODT the only allowable addresses that can be examined other Any from 0-28 KW (0-00157776). memory of addresses the addresses and any attempt to accessed the I/O page or any registers is not allowed. Table 4-3 lists the ROM ODT commands. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING TABLE 4-3 ROM ODT COMMANDS Prints contents of specified location or ifof no address is defined then print contents the last location that was opened. If location opened is an odd number then print out only the contents of the byte. If location is even then mode is word, if location is odd then mode is byte. Leading zeroes are assumed. Only bits 15 through zero of the address are used. RETURN LINE FEED <CR> Closes an open location <LF> Closes an open location and then opens o Period the next location. If word, increment address by 2, if byte, increment address by 1. Alternate character for line feed. This command is useful when the terminal is a VT2xx series terminal. It is also convenient to use with the keypad. Up Closes an open location and fthen opens the previous location. If in word mode then decrement by 2, if byte decrement arror by - Minus 1. Alternate character for Up arror. This command is useful when the terminal is a VT2xx series terminal. It is also convenient to use with the keypad. DELETE Delete Deletes the previous character typed. Exit ROM ODT and return to setup mode. The following -paragraphs present examples of ROM ODT use. EXAMPLE 1 Location 200 is opened. location contents. 202 is It is then closed with no changes and opened, which is then closed after changing it's BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING ROM ODT > 200/ ROM ODT > 000200/100000 <LF>. ROM ODT > 000202/003333 44 <CR> ODT ROM EXAMPLE > 2 locations Byte location 1001 is opened. It is then closed and chang ed and 1002 and 1003 are opened. Location 1003's data is then closed. ROM ODT > 1001/ ROM ODT > 001001/101 <LF> ROM ODT > 001002/104 <LF> ROM ODT > 001003/113 141 <CR> ODT ROM > EXAMPLE 3 The user attempts to open location 170000 which page, and not allowed. is in the 1I/0 ROM ODT > 77770000/ ODT ROM EXAMPLE > 4 Location 150000 is opened and then closed. by typing / only. ROM ODT > 150000/ ROM ODT > 150000/032737 <CR> ROM ODT > / ROM ODT > 150000/032737 It is then reopened BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: Enter ROM 15 <CR> ODT Type CTRL Z to exit or press the RETURN key for No change xxxxxx/ = open word location xxxxxx if address even, byte if odd = close location RETURN = close location and open next LF . or close location and open previous ROM ODT> 010000,/000000 ROM ODT> 010002/000000 ROM ODT> 010004/000000 ROM ODT> CTRL 012705 <CR> 101 <CR> 12706 <CR> Z KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: FIGURE 4-30 SETUP COMMAND 15 EXAMPLE 4.4 DIAGNOSTIC ERROR MESSAGES information are When an error occurs, the test number and other of the current number the always The error number is printed. test number that the ROM code is running. The test numbers corresponding error messages are described in Chapter 5. 4.5 BOOTSTRAP and PROGRAMS Bootstrap programs may be found in various areas of the system. The ROM code contains bootstrap programs for the following UNIBUS devices: Device Name DU* DL DX DY DD DK MU** Device Type RX50, RC25, RLO1, RX01 RX02 TUS8 RKOS TU81 RA80, RA81, RLO2 4-40 RA60 BOOTSTRAP AND DIAGNOSTIC ROM' PROGRAMMING * DU refers to a general purpose boot program for disk MSCP devices. ** MU refers to a general purpose boot program fortape MSCP devices. For users who have devices not covered by the ROM code bootstrap the ROM code also supports the use of M9312 type boot ROMs list, which are typically shipped with the controller module for devices which can be booted. These ROMs are currently used in all existing UNIBUS PDP 11 products. An example of this would be a RKO7 disk drive which uses a M9312 ROM to allow the drive to be booted. Table 4-4 lists the M9312 type boot ROMs currently available. The user may install the ROM in either the UBA module or an optional M9312 module. The ROM code will list and boot any M9312 Subsection 4.5.1 contains type ROM located in either module. additional information on M9312 type ROMs. The ROM code also allows users to install bootstrap programs for The user new devices or for custom boot programs in the EEPROM. using by EEPROM the into programs can install machine language loaded is program the Once 15. through 11 Setup mode commands into the EEPROM, it will be avail- able to the user at any time. in the assembles EEPROM boot programs in memory and then starts program at a start address defined by the boot program. EEPROM may contain more than one boot program depending on the The the for the There is one important difference between boot programs The EEPROM is an eight bit device which means EEPROM and others. they are the programs cannot be executed out of the EEPROM as always code ROM The type ROMs. ROM code or M9312 from the size of the programs. When a boot program is requested, the ROM code searches boot program in the following sequence: 1. Search the EEPROM on the CPU first. 2. Search the ROM code con the CPU next. 3. Search the ROM sockets on 4. Search the ROM sockets on the M9312 board next, the UBA next. if present. BOOTSTRAP.AND DIAGNOSTIC ROM PROGRAMMING 4.5.1 Bootstrap List Table 4-4 describes the M9312-type boot ROMs that are available. These ROMs are used on all UNIBUS processors. Some of the ROMs listed are not required because the base ROM code in the 'CPU ROM contains bootstraps for some devices. Many of the devices listed are old and today, However, they generally not available are included because many systems contain In general, the ROMs needed to boot a device are these devices. For example, shipped with the interface for the device itself. r the ROM controlle RL11l an and RLO2 an purchased if a customer would come with the RL11l controller. The CPU ROM code uses a two-letter M9312 ROM to identify the ROM. than one bootstrap program. in each mnemonic contained Some of the ROMs contain more Some programs require more than ROM to fully implement a bootstrap. one TABLE 4-4 AVAILABLE M9312 TYPE ROMS SUPPORTED DEVICES CT 23-761A9-00 TU60 cassette tape drive DK DT 23-756A9-00 23-756A9-00 RKO3, RKO5 disk drives TU55, TU56 tape drives Note: The RKO0S5 boot is also in the DM 23=-752A9-00 DP 23-755A9-00 DB 23-755A9-00 CPU ROM RK06, RKO7 disk drives RP02, RP03 disk drives. This ROM mustbe installed in the KTJ11-B. RP04, RPO5, RPO6, RMO2, RMO03 disk drives. This ROM must be installed in the KTJ11-B. DS 23-759A9-00 RS03, RS04 disk drives MM 23-757A9-00 TUl6, TEl6, TU45, TM02, TMO3, TU77 MS 23-764A9-00 TS04, TS1ll, TU80, TS0S5 tape drives MT 23-758A9-00 TUl0, TEl10, TS03 tape drives PR 23-760A9-00 PCO5 High speed paper reader TT 23-760A9-00 tape drives Low spedéd paper reader (Teletype) 4-42 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING XE 23-E22A9-00 DECnet DEUNA ethernet interface XL 23-926A9-00 23-927A9-00 23-92849-00 DL11-E (DECnet DDCMP) NOTE: Three ROMs are required 23-862A9-00 23-863A9-00 23-864A9-00 DMC11, DMR11l (DECnet DDCMP) NOTE: Three ROMs are required to implement this bootstrap. XU 23-868A9-00 23-869A9-00 23-870A9-00 DUll (DECnet DDCMP) NOTE: Three ROMs are reqguired to implement this bootstrap. XW 23-865A9-00 XM implement to DUPl1]l this bootstrap. (DECnet DDCMP) ‘NOTE :Three ROMs are erguired 23-866A9-00 23-867A9-00 implement to this bootstrap. NOTE In V6.0 and identifies which routine the UBA or the M9312 will not code, v7.0 ROM ROMs on the identify boot ROMs which use more than of this, these from started boots will not list and can not be code without using a small EEPROM base ROM the ROMs. the to control boot program to transfer in a future corrected be problem will This correctly one for ROM release of the the CPU boot. ROM code. instructions Refer to Appendix F for the EEPROM to allow the ROM boot on the UBA or which is not compatible for The ROMs boot listed to CPU ROM to handle set up a multi the M9312, or any ROM with the M9312 ROM format ROMs. in Table 4-5 are similar bootstrap programs are module. Because generally located not needed because in the base ROM on the CPU BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING TABLE 4-5 CPU ROM BOOT PROGRAMS —_————_c-———cn——-——— -:-smcxau————_—---n——--—-—:—c-————--—————————-q-- MNEMONIC oo SUPPORTED DEVICES DEC“PART NUMBER drivee tapedg 23-765m9-00 TUS8 cartri DK 23-756A9-00 RKO5 disk drive DL 23-751A9-00 RLO1, RL02 disk drives DU 23-767A9-00 (General boot for all Disk DX 23-753A9-00 RX01 floppy disk drive DY 23-81129-00 RX02 floppy disk drive 4.5.2 Format EEPROM The first 105 bytes parameters the ROM for code selections MSCP devices) RA80, RA81, RA60, RC25, RX50 disk drives and to the of the EEPROM stores the Dbase determine the power up/restart the list of devices to try to boot. following the first 105 bytes are four bytes which the use hardware CPU, the UBA, and the information needed by for any purpose such as storing a serial number. bytes are never used by the ROM code. mode, test Immediately user may These four The optional bootstrap programs follow immediately after these four bytes. The EEPROM may also contain translations for some of the ROM code messages for local language requirements for non English users. The local language text, if present always starts at the end of the EEPROM. Figure 4-31 1illustrates the EEPROM layout. A BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING EEPROM size = Start of 2048 bytes. ====-———-c====——eo—————————————=——- EEPROM Base ) area. CPU and UBA Hardware parameters Boot device information 105 Bytes Translation table 4 Bytes Reserved for customer use only* variable length Optional bootstrap # 1 Selection expand towards beginning Vvariable Length | | End of EEPROM information optional Foreign language text or UFD area if no foreign language == =s— oo o- === ~—=-c-mememe—m—mee——mms— * The four bytes reserved for customer use may be accessed as follows: 1. Insure that bit 6 is reset in the CPU BCSR'at 17777520. 2. Insure bit 5 is set in the CPU BCSR at 17777520. 3. Bit 4 in the CPU BCSR at 17777520 must be set to write the EEPROM. It need not be set when reading the EEPROM. 4. The low byte of the PCR must be 0 at 17777522. 5. The four bytes of data can now be read or written at addresses 17765322 to 17765330. Each byte is accessed in a 16-bit word where the data is in bits 7 thru 0. Bits 15 thru 8 are not driven by the CPU and must be ignored since their value will change. FIGURE 4-31 EEPROM LAYOUT If a serial number 1is written to the four customer-reserved the format could be to write up to a 24-bit number in the bytes, 4. first three bytes and write a checksum for the number in byte the to up is format actual the and suggestion is only a This user. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING When data is written to the EEPROM the user must wait at least 10 ms after the write cycle for each byte before proceeding. The FEPROM can not be written more than 10,000 times. ~After any write the data should always be checked after 10 ms to verify that it was written correctly. NOTE It is strongly recommended that bit 4 cf the BCSR the user is writing to the set unless not be EEPROM, and that it be reset as soon as writes This bit does not have to be set are complete. to read the EEPROM. 4.5.3 General Rules For EEPROM User Boots EFPROM boots are assembled into the lower 28 K words of memory by the ROM code. If there are no errors such as checksum errors the ROM code will pass ccntrol to the program according to the starting address in the boot program. At the start of an EEPROM boot the following is true: 1. Memory management is disabled and 22 bit mode is off. 2. RO contains the unit number. 3. Rl will be 0 if no address was passed by the translation table or the /A switch in the boot command. Rl will be an alternate address for the program to use if the translation table matches the boot device's name and unit number with an entry in the translation table or the /A switch was used in the boot command. 4. The ROM code loads a trap handler for timouts to location 4 The program is loaded into memory starting at in memory. location 1000 if the start address of the EEPROM boot program 1is 10000 or greater. The program is loaded into memory starting at location 17600 if the start address of the EEPROM boot program is 0 to 7776. The rom code will load the address of the timout handler into location 4 as long as the EEPROM boot does not already occupy location 4 If a timout occurs during the boot program and the itself. timout handler is entered the handler will restart the rom code with the non existent controller message in R5 and the rom code will assume that the value in Rl is the address of the controller that timed out. BOOTSTRAP AND DIAGNOSTIC ROM- PROGRAMMING if The EEPROM may use memory management and restart the ROM code Kernel to only the EEPROM program restricts the use of MMUmust not use any boot The EEPROM Rs 0-3 and 7. Instruction PAR/PD of the other PAR/ PDRs if the ROM code is to be restarted. The In bitO ="0). ROM code should be re- started with MMU OFF (MMRO on the is it 4, locati ps overla the case where the EEPROM boot responsibility of the EEPROM boot to handle timeouts. It is the responsibility of the user's program to boot the device and check the boot block for a bootable secondary boot program. It is recommended that EEPROM boot programs not be located 1in memory between addresses 2000 to 2300, 16000 to 16040, and 20000 programs is to 40000. The recommended location for EEPROM boot cal address (Physi point memory addresses above the g8Kk-word . 00040000) The following paragraphs describe the way user-written boot programs should handle errors or success. It is recommended that This will the user write programs that adhere to these rules. event that the allow the user to receive meaningful messages in However, the user's boot errors occur in the boot program. program need not return control if it is not desired. Once the EEPROM complete control 1is boot started the CPU. of the user program boot has The user's program would restart the ROM code for one of the following three reasons: 1. An error occurred attempting to boot a device and the ROM code 1is restarted to type out a general error message. This allows the automatic boot mode to try another item for booting. To reenter the ROM code for error message printing, the load RS5 with the error message user's boot program would - 4 of the BCSR are set to O, 7 bits that sure make desired, with the MMU off. 165762 @ JMP a execute then and The following list specifies the octal value of the error ‘message selection codes and the text of the message printed. At the time the ROM code 1is restarted, RO should still contain the unit number and Rl should contain the address of the controller. 270 = Drive 273 = not ready 271 = Non bootable media present 272 = No disk present or drive unloaded No tape present 274 = Non existent controller 275 = Non existent drive 4-47 BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING = Invalid unit number = Invalid device = Controller error = Drive error 276 277 300 301 2. the boot 1If the user's program monitors the keyboard during is typed. C CTRL if code ROM return control to the it could reentered be would code The ROM This is an optional feature. same way as described in 1 above with the exception that the R5 should be an octal value not equal to 270 to 301 or 1. 3. successful The boot is and the ROM code 1is temporarily print out the "Starting system from " message. to restarted returns code After the message printout is complete the ROM To reenter the ROM code to user's program. the to control code would wuser’'s the print the "Starting system message" R5 with the number 1, make sure bits 7 - 4 of the BCSR load PC, JSR a are 0 and then restart the ROM code by executing @ 165762 with MMU off. When the ROM code has completed typing the message it returns the user's boot at the instruction following the to control existing compatible with all user's program should the code, ROM the versions of Reset bit 11 JSR instruction. do the At this point, to be following: (register set 1 select) in the PSW Reset the display register by writing 000077 777 to address 17 524, Clear MMR3 (17 772 516) to make sure 22-bit mode is off. 4.6 BOOT ROM FACILITY (M9312 compatible) The KTJ11-B Boot ROM Facility allows the user to install M9312 compatible boot programs written for UNIBUS devices which are not directly supported by the KDJ11-B boot programs. ROM programs which run on the M9312 should work on the KTJ11-B. The M9312 compatible boot programs are implemented in from one to four 512 X 4 bit ROM's. Each ROM contains 64, l16-bit words of accessible the last code which are located in the first half of the ROM. 256 4-bit ROM locations are not used. The KDJ11-B CPU Module can be configured to boot the system from its self-contained boot programs, from the KTJ1ll-B Boot one of ROM facility, or a boot ROM option which resides on the UNIBUS. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.6.1 Boot ROM Installation Digital - Equipment meet must M9312 compatible Dboot ROMs Corporation purchase specification 23-000A9-01 for 512 X 4 tri—-state PROMs. The ROM's must be encoded per the” Digital Equipment Corporation specification K-SP-M9312-0-8. programs. boot more or one contain A single ROM may ROM's. four to up require may program boot Alteratively, a single Dboot each keep to d encourage been have s programme However, ROM program within a single ROM. Table 4-6 presents the IC location of each ROM socket along its addres with - range. TABLE 4-6 ROM LOCATIONS AND ADDRESSES ——————--————— «--—-——-——--———--———--—-——-———-————————- ROM Number IC Location Addresses T T B145 2 El44 17 773 000 thru 17 773 176 3 E143 17 773 400 thru 17 773 576 4 E142 17 773 17 773 200 thru 17 773 376 600 thru 17 773 776 4.6,2 ROM Addresses 17 773 000 - 17 773 776 thru The KTJ11-B Boot ROM logic responds to addresses 17 773 000 subdivided are locations As shown in Table 4-7 these 17 773 776. the into four, 64-word segments, each of which addresses one of by operation read a to responds logic ROM The four ROM sockets. decoding a l16-bit data word from four successive locations in the If an empty ROM socket is accessed, 4 bit ROM. X 512 selected the data word 4.6.3 ROM read is typically 161777. Formats This section summarizes the format M9312 Specification K-SP-M9312-0-8. information contained Two types of in ROM formats the are identified: the format for one or more programs contained ia single ROM, and the format for one boot program contained in two or more ROMs. Specific information 1is given on headers and on the format of data within the ROM's. 4-49 the program BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.6.4 Single ROM Programs ROM's which contain one or more programs in a single ROM conform to the following format: 1. Each boot program must begin with a program header block as described in subsection 4.5.2. The header block for the first (or only) program must start at ROM Word Address O. 2. Word Address 24 of the ROM must remain reserved and be 3. Word Address 26 of the ROM must remain reserved and be set (This location is required for systems which to 000340. reboot via the M9312 module and is not used by KTJ11-B to set 173000. systems which reboot via the CPU module) . Word Address 176 of the ROM must be a CRC-16 word 4., previous 63 words (omitting location 24 octal). 4.6.5 Multiple for the ROM Programs Some boot programs require more than a single ROM. The first ROM would be formatted as described in Chapter 3. The continuation ROM(s) would have the following format: 1. The first word of each continuation ROM must contain 177776 2. Word Address 24 of the ROM must remain reserved and be 3. set Word Address 26 of the ROM must remain reserved and be which (This 1location 1is required for systems to 340, M9312 module and is not wused by KTJ11-B the via reboot octal. to set 173000, systems which reboot via the CPU module) . 4. word for the Word Address 176 of the ROM must be a CRC-16 locati on 24). previous 63 words (but omitting the word in 4.6.6 Program Header As stated in section 4.6.4, the beginning of must as contain a header section. follows: each boot program This header section is formatted BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING The first word contains an ASCII identifier which consists of two characters with a zero parity bit. The high and low bytes contain the first and second characters respectively. The characters are used by- the KDJ11-B Boot and Diagnostic The ROM programs to search for the selected boot program. contain bytes these of both that KTJ11-B ROM codes requires characters with ASCII octal values of 101 to 132 or 141 172 to (A-2 or a-2). The second word contains an offset from its address to the start of the next program header. If the ROM contains only one header, the second word, located in ROM Address 2, contains 176, which points to start of the next ROM. The third word address is the Power-up entry point for unit zero, used by systems containing the M9312 to disable a the boot branch to diagnostics prior to running point. entry this use not do systems KTJ11-B program. The fourth word address is an alternative entry point for unit zero, used by systems containing the M9312 to enable a boot branch to diagnostics before running the KTJ11-B systems do not use this entry point. The fifth word 0, contains indicating instruction located in the previous word. wunit 0 program for the The sixth word address is the entry point used by systems for unit numbers other than zero. containing the M9312 KDJ11-B uses this entry point after first loading the unit number (zero or non-zero) into RO and then setting the PSW Carry Bit (disabling the branch to M9312 diagnostics). The seventh word contains the address of the Control/Status Register of the device to be booted. It is used by the instruction located in the previous word. an instruction which saves the PC The eighth word contains branch to diagnostics on the next which in R4 for systems uses this entry point when a non KDJ11-B The instruction. standard CSR address is passed in Rl. The ninth word contains an instruction which branches to M9312 diagnostics if the PSW carry bit is clear. KTJ11-B systems always set the carry bit. an The tenth word contains l start of the boot program. > 10. 51 unconditional branch to the BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.6.7 ROM Data Organization Each 512 X 4 bit ROM contains 64,.16-bit words of accessible code which are located in the first half of the ROM. Each l6sbit word is stored in four successive ROM locations. Whenever the KTJ11-B ucts the ROM logic is addressed by a Data-In operation, it constr 4-7 Table ns. locatio ROM 16-bit word from the appropriate presents the four nibbles. location and polarity of bits within each group of Note that bits 12, 11, and 10 are inverted, and bits 08 and 00 are not stored in the expected order. TABLE 4-7 ROM DATA ORGANIZATION Data Data Data Data 03 07 11%* 02 06 10* 01 05 09 08 04 00 Bjt 2 Bit 3 Nibble Number 0 Nibble Number 1 Nibble Number 2 14 15 Nibble Number 3 Bit 1 13 Bit O 12* * pata Word Bits 12, 11 and 10 are stored inverted 4.7 J11 MICRO ODT J11 Micro ODT is entered anytime the CPU is halted by: 1. 2. 3. Placing the front panel toggle switch in the HALT position, Executing a HALT instruction, if the halt mode is enabled, and the system is in kernel, or is setup to Pressing the Break key, Iif the terminal panel keylock front the generate a Break character, and switch is set to ENABLE. J11 Micro ODT allows the wuser to examine and/or change any page memory space. location in the 22-bit CPU memory space Or I/0 start a program, In addition, J11 Micro ODT allows the user to: step a program when to reinitiate program execution and to single positi on. Table 4-8 the front panel toggle switch is in the HALT i > summarizes the J11 micro ODT commands. 52 BOOTSTRAP TABLE Slash 4-9 J11 MICRO AND ODT n/ DIAGNOSTIC COMMAND Opens ROM' PROGRAMMING SUMMARY the specified location (n) and its contents. n octal Carriage Line Return Feed , number. <CR> Closes an <LF> Closes an and then Register $n or Rn Opens a open location. open location the next opens contiguous Internal outputs is an location. specific processor register (n). n is an integer from 0 to 7 or character S. Processor Status Word S | Opens the Processor Status Designator Register. $ or Must follow R command. Go G Proceed P Resumes Control-Shift-S Manufacturing Binary the Dump Starts program the execution. program execution. use only. The following paragraphs provide a detail description of each J11 Micro ODT command specified in Table 4-9. 1In most cases, each description is supplied with an example. Note that operator input is underlined. When entering addresses or data, they will When entering entered A be (i.e. filled by addresses leading zero's are not required, ODT. in the 1I/0 page, all 22bits 17776100). ? will be printed whenever: addresses are accessed that error is detected. illegal characters result in a are timeout, or must be entered, a parity BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.7.1 / (ASCII 057) Slash I/O device This command is used to open a memory location, word status processor or register, processor internal register, to specify a register, and must be proceeded by octal digits location, or a tegister designator. (i.e. 'In response to /, ODT prints the contents of that location to location six characters) and waits for either data 33 for that [/ The <LF>. or <CR> (i.e.) be entered or a valid close command the of contents the display to may be entered again immediately previously opened location. Examples: . @1000/ 012737 <CR> @100/ 000200 7422 <CR> ;Open memory location 00000100 ;and deposit data (7422) ;close the location. and snew data. - (ASCII 015) s The contents (012737) are ;displayed. <CR> closes the :location without modification. ;Re-open the location and deposit @/ 007422 6422 <CR> 4.7.2 <CR> ;Open memory location 00001000. Carriage Return location's If a This command is used to close an open location. <CR> with the precede should user the changed, contents are to be If no change is desired, <CR> closes the location the new data. without altering contents. see previous examples. Examples: 4.7.3 its <LF> (ASCII 012) Line Feed is location the open except This command is the same as <CR> Memory opened. is location next contiguous the and closed are registers processor and 2 by incremented addresses are If the PS is opened it is closed and no new incremented by one. location is opened. BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING Examples: ;Location 1000 is opened, the @1000/ 012737 <LF> <LF>. with ;then closed sThe <LF> caused the next 00001002 100200 0O <LF> are ; The next location is opened 00001004 176100 <CR> and :to examine the contents sthen closed with <CR>. Internal Register Designator or R (ASCII 122) (ASCII 044) the In ;location to be opened and ;contents to be displayed. ;this case the contents ;changed the operator. 4.7.4 $ and ;contents are displayed, Either character when followed by a register number 0 to 7 or the If S will open that specific processor register. designator PS number last more than one number is typed after the R or §, the typed will be used. NOTE be cannot PS the of <4>) (bit The trace bit modified by the user. This is because ODT uses the T bit for single-stepping. The register set used with the R command is determined by PS<11>. If PS<11> is set to a 1, register set one will be used. PS<11>=0 register set zero will be used. The SP (R6) that is used is determined by PS<15:14>. 4.7.5 This G (ASCII command entered 107) is Go used immediately to start before program the G. execution This the LOAD ADDRESS and START switch consoles. The PC (R7) is loaded with if no data is entered), the following zero: PS, System Error Status Register. If the HALT G the Register, command position, reentered typed MMRO<K15:13,0>, and address in Cache is the typed is issued the PC in 7777773000G MMR3, Cache flushed with system will be to the last would be it CPU the front be as location Register, and panel switch bits, Memory Floating is Go i.e. Point initialized. in the ODT will command 173000G. to other PDP11 will be used cleared to UNIBUS The sixteen read a equivalent initialized, displayed. 4-55 Error Register, and the will is sequence on the address (0 registers are PIRQ, Control at function be truncates 1if the user Since the BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING is disabled by the Go command, the memory management unit starting address is always in the lower 28 K words of memory (0 157776) or the I/O page. Examples: @1000G :The program is started at location 1000. @1000G ; The program is started with the Halt switch @1000 4,7.6 P (ASCII son. The CPU initializes registers and then ;halts without executing the first instruction. :The PC is displayed and then the ODT prompt. 120) Proceed This command is used to resume execution of a program and corresponds to the CONTINUE switch on other PDP11 consoles. Program execution resumes at the address pointed to by the PC The next instruction is fetched and executed, outstanding (R7) . interrupts will be serviced. switch in the If the P command is issued with the front panel instruction of HALT position it 1is recognized at the end will (R7) PC the of execution and ODT is reentered. The content ruction single-inst can In this fashion, the user be printed. on the console step through a program.and obtain a PC "trace" terminal. Example: @R7/002464 1000 <CR> @P - ¢sR7 (PC) is opened and the scontents displayed. The new saddress is entered in R7. ; The proceed command is issued - sand the program continues at @p 001004 :The proceed command is issued swith front panel switch in the sHalt position. The PC is ap ; 001010 : - @ ;location 1000. ;displayed. ;Etc. ; BOOTSTRAP AND DIAGNOSTIC ROM PROGRAMMING 4.7.7 S (Control-shift-S) Binary Dump 1is oses by manufacturing and from This command is used for testThepurp ived command is normally rece not a normal user command. It is .” inal term ole not Dby the cons another computer and ole cons the from ed issu be command NOT recommended that thethis code 23 I ASCI the back es echo console ODT terminal because oard to lockup, thus preventing data from keyb the e caus may and on the screen. being displayed use command from a terminal beca There is no reason to issue this ive rece to nded inte the terminal is and it dumps the binary data efficently is intended to more the and comm The the ASCII data. "/" and g usin to display a portion of the memory as compared <LF> commands. inals by ly entered on many term This command can be accident LL" key cases by the " NO SCRO CTRL s or 'in many gene typing CTRL S, cond code. 23 I rate the ASCI itions normally since all these , it exit to r and, in orde ly enters this commterm— If the user accidentthe a "a" type and - inal user reset the is recommended that is ODT ole cons the that minimum of three times in order to insurecomm and protocol 1is as ready to accept commands ' again. The follows: ands 1. After a prompt character, ODT receives a CTRL S comm and echoes it. 2. of the serial 1line must The host system at the other end ting rpreted by ODT as the star send two 8-bit bytes s inte byte t firs The ed. echo address. These two byte are not seco bits s ifie spec byte nd the specifies Dbits <15:08> andess. 16> <21: bits ess addr Bus <07:00> of the startina addr to icted restr is nd are always forced to 0, the DUMP commaThe starting address the first 32K words of address space. may be even or odd. 3. been received, ODT After the second address byte has line, starting at the al seri outputs 10 bytes to the t is finished outpu the When address previously specified. ODT prints <CR>, <LF>, Q. CHAPTER SYSTEM 5.1 MAINTENANCE INTRODUCTION This chapter error removal 5.2 message and contains troubleshooting interpretation, information, diagnostic and field replaceable unit (FRU) replacement procedures. TYPES DIAGNOSTIC The system supports three types 1. 5 DECX1ll - provides each option and of diagnostic programs: system tests to check the interaction of 1isolate system failures to the subsystem component 2. XXDP+ - provides tests that check localize hardware failures to the 1individual options function level. and 3. ROM Resident - CPU ROM-resident Start-up diagnostics test various functions specific to the system modules and PMI bus, and isolate failures to the module or option level. The.following XXDP+ diagnostic programs are available to test various functions on the three kernel modules. To initiate an XXDP+ diagnostic the system must first successfully complete the start-up diagnostics. OKDA?? (KDJ1l1l-BF) OKTA?? (KTJ1l1l-B) MAINTENANCE SYSTEM is time (Run (MSV11-J) VMJA?? minutes depending on memory size.) approximately The "??" at the end of the diagnostic name assures latest diagnostic version is executed when called. To load and execute a diagnostic, 1issue the For example: followed by the diagnostic name. . R RUN (R) 60 - 20 that the command <KCR> OKDA?? a during executed are The start-up diagnostics, listed below, of evaluation quick a provide They system power up or restart. the three kernel modules. The start-up diagnostics check the following system functions: Processor: Verify CPU and FPA functionality and test Disables communications between the CPU cache and the CPU. (PMI) the main all data paths between system devices and but memory, the not CPU cache memory. Tests the main memory using the PMI protocol. Main Memory: PMI its adapter, UNIBUS the Tests Adapter: UNIBUS and control paths with the addressing and communication, UNIBUS disabled. A failure at any point during these diagnostics halts the testing and: 1. Displays an error code and an error message on the 2. Displays terminnal (if connected). an error code on the front panel console START-UP TEST display. 3. Displays the error code in the CPU module diagnostic LEDs. Normally, the system displays the same error information in all three locations. If the console terminal is not working, refer to the START-UP TEST display. If neither 1location 1is working, refer 5.3 to the CPU module CONSOLE TERMINAL LEDs. ERROR MESSAGE FORMAT A typical console error message is shown in Figure 5-1. SYSTEM Testing in Memory 9 Step progress - Please Size is 1024 memory test K Bytes Step Error 1 3 45 6 7 wait 8 46 Memory See 2 CSR Error troubleshooting Error PC= 173436 documentation PCR page= 15 Program listing address= RO = 060000 R1 = 052525 R4 = 100000 R2 = 172100 R5 = 040000 R3 R6 = 172300 Par3 Command 1 Rerun Loop 3 Map a The a. command Console console An test memory and I/O page then the RETURN error 5-1 probable An "See Error PC ERROR message code 5-2 contains following octal number codes, test this is that failed. lists the the error causes. lists the - this error is a messages and documentation" of unexpected the traps, 5-3 of start-up descriptions, description their and of descriptions. specifies ROM program the the message. line - this address line PC=), the page number in the in information: one-line address case EXAMPLE the - troubleshooting (Error the DISPLAY Description the address to reference listing address=) In key: MESSAGE test description error. Table d. 5-1 error the A press Error Message error Table C. 010000 test diagnostic b. 172344 = on FIGURE 5.3.1 = 015436 Description 2 Type MAINTENANCE error (PCR the error page=), and 1listing address (Program is the SYSTEM MAINTENANCE address following unexpected the instruction which caused the trap. The content of R0O-R6 of register set 0, and the content of kernel PAR 3. The tests do not use register set 1. Register set 1 is used mainly by ROM code support routines. For some tests the system displays the expected data, and the received data failing address, (bad data). the A command line which describes user-selectable commands. To execute a command enter the associated command number (e.g., 1, 2, etc.) followed by RETURN. The commands are: 1. 2. Rerun the test. If the test passes the ROM code will continue testing. If all other tests pass, the ROM code will display the total number of errors and enter dialog mode regardless of the mode selection in the EEPROM. Loop on test. This option causes the ROM code to continuously loop on the test that failed. These loops are generally very large and are not intended to be used as scope loops. The test runs until an error occurs or end-of-test has been reached. 1In either case, the test is started again and continues to loop until the user types CTRL C at the console. At this point, the ROM code will display the total number of errors and the total number of successful passes if any. Both the error counter and the pass counter have a maximum value of 65535. If either counter reaches this value the count will not overflow to zero, it will stay at 3. this value. Map memory and I/0O page. This command is normally used if a memory error occurs. If a memory is not configured properly the MAP command may point to where the memory actually is. In a multi memory ' system in which one of the memory fails this command can be used as a method of physically identifying the failing memory, if it has a CSR. This command is only available when the bus on. The command is not available for tests 76 through 56. 1is turned (or codes) SYSTEM MAINTENANCE 4. This command allows the user Advance to the next test. and continue. This command test failing to bypass the generally that are errors for up show only will non fatal. considered If the error is fatal and field service would 1like to bypass the error it is possible by typing CTRL O, 4 and RETURN, and the command will be executed. CAUTION: wuser all unless bypassed Errors should not be otected. write—pr or removed been has software At this point the ROM code flushes it's input buffer of any previously typed characters and waits for input from the user. TABLE 5-1 ERROR CODES, TEST DESCRIPTIONS AND PROBABLE CAUSES PROBABLE CAUSE ERROR CODE TEST DESCRIPTION 77 Initially set to this value on up. power The halt switch is ON at power up. BGn or NPG line A UNIBUS is open and SACK is being asserted by the bus. Check MDM module. All grant cards must Dbe installed. Terminator is faulty. Power supply 76 First CPU 75 Turn MMU on. and 74 CPU MMU M8190 Run MMU M8190 tests, tests register tests. Turn on PMI and look at the UBA RESTART bit. Then turn off PMI. M8190, M8191 is faulty. SYSTEM MAINTENANCE TABLE 5-1 (Cont) PROBABLE CAUSE TEST ——u—m—‘-———_———m_—n@———————-— m:-xmu—n——-_—a-sena-—-_gn--——————-———-——————-——— Power up NOT A FAILURE - Selected to ODT. Mode is ODT in EEPROM and the system is in (J11)ODT M8191 72 Power up to 24/26. M8190, 71 EEPROM checksum tests. EEPROM accidentally written, restore data using Set up mode commands, verify W40 installed on M1890 module. 70 CPU ROM checksum and PCR M8190 Miscellaneous CPU and EIS M8190 tests. 67 tests. 66 Console SLU test 1 check for each register's M8190 response 65 Console SLU test 2 transmit and receive data M8190 Console SLU test 3 - M8190 in maintenance mode. 64 check interrupts and errors in maintenance . mode 63 Test MMU aborts. M8190 62 Stand—-alone mode CPU M8190 61 cache tests. Clock test. M8190 Clock from power supply 60 Floating Point Processor. M8190 Unused. —-—u—_mw——————_——nuw —————————————————————————————— q-cu—c-—bwam—: 5-6 SYSTEM TABLE ERROR TEST CODE DESCRIPTION 55 5-1 (Cont) PROBABLE" CAUSE Exit stand-alone mode. Check location of address 17760000for timeout. M8190 UBA M8191 register response test, check UNIBUS through DDR for hung lines. M8191 UNIBUS failure M8190 54 Memory size 53 Check memory location 52 51 0 - 4K test. UNIBUS present failure at PMI memory test. PMI memory/M8190 PMI M8190 O. word memory Cache testing using memory . 50 Memory tests test byte 1 - Data 47 Memory parity/ECC 46 Memory address/shorts tests. test. 45 44 UBA ROM UBA map path 43 UBA 42 UBA 41 PMI memory PMI memory test. M8191 registers data M8191 unmapped cycles test. mapped diagnostic diagnostic test. UBA floating test using nostic memory response test. cycles PMI tests. address/data mapped cycles. diag- M8191 M8191 M8191 MAINTENANCE SYSTEM MAINTENANCE TABLE 5-1 37 PROBABLE CAIJSE TEST DESCRIPTION ERROR CODE 40 (Cont) UBA address overflow test. M819l1 M8191 UBA cache data test. PMI memory 36 UBA cache LRU (Least M8191 35 UBA cache floating M8191 Recently Used) test. address test in TAG store. 34 UBA cache parity error M8191 33 UNIBUS memory data path UNIBUS memory detection testo. test. M8191 32 UNIBUS memory parity UNIBUS memory 31 UNIBUS memory address/ UNIBUS memory logic test. shorts test. NOTE s 25, 22 and 6, codes With the exception of error codelem indicators and are 30 through 1 are bootstrap prob not diagnostic errors. , tape, etc.) Some bootstrap problems (e.g.., diskrs may be indimay be corrected by the user. Othe cators of errors in the device being bootstraped 30 Test Exit Routine 27 Unused 26 Unused wacflm--Qn -am@@-nw@wufl‘gwa‘g@QGGfi SYSTEM MAINTENANCE TABLE ERROR CODE 25 5-1 TEST DESCRIPTION - Air mover and voltage regulator test. ED W WED N WD D M CED WA TR D G WIS CHR TS CHE A WL D MR S W W W G e Cabinet blower Box fans regulator module H7213 MLM module not installed No memory module(s) in system. 24 Unused 23 XON not received after XOFF, To correct, type CTRL Q at the console. Console terminal not ready due to XOFF signal received from terminal while attempting to print a message. This is normally not considered a failure, because the condition could be operator error. If the oper- ator has typed CTRL S without following with CTRL Q or the terminal is not ready, (e.g. could have run out of paper). 22 21 Console SLU transmit ready bit Drive not M8190 set. error. The device that the user is attempting to boot is displaying an error code in its error register. Verify that media is in good condition and 20 Controller error. bootable. The UNIBUS controller for the device the customer is attempting to boot is displaying an error code: SYSTEM MAINTENANCE PROBABLE TEST CAUSE a-s———c—:—cs—e-—————————— -n———————--nm—_-_———»—-—-nu-ne—c:—————-—————:———- in its control and status register (CSR). Ensure that the NPG jumper was removed if the device is a DMA controller. Consult the devices technical manual for more information. 17 Invalid The mnemonic typed in for the boot device is either incorrect or the boot rom for that device is not installed .Go to the dialog mocde and "LIST" the device. valid unit 16 Invalid 15 Non-existent 14 Non-existent controller devices. The unit number after the mnemonic is not within acceptable range for that device. See that devices technical manual for help. number. The drive number the user drive. is trying to boot not on the 11/84. The controller No 13 No 12 No disk present or drive is not loaded correctly. 11 - cm - on wo e present. Non-bootable media the o - o caD is - s o e oo D " an D D TR D e D D No media de- D ) D addressed installed. in drive or LOAD button not the in. The bootstrap data from the device does not conform to the boot block specifications. Ensure that media is bootable. in D tape drive drive. B e the vice the user is trying to to boot from is not on the UNIBUS or is incorrectly. tape for from is B I D D D CED XD ORI O WP W) WD P IO D WD WD CED D Eb GEP GNP GED GHR ek D D W W SYSTEM MAINTENANCE TABLE 5-1 (Cont) PROBABLE TEST ERROR CAUSE DESCRIPTION CODE bootable. Change Setup mode to accept non-standard boot blocks. 10 Drive not ready. No media present in the drive or the disk drive has not completed its spin up function. 6 Console disabled.- 5 Unused-. 4 Dialog mode. Self explanatory. The system is in dialog mode and waiting for input from the console terminal to rewind. 3 UBA ROM boot in progress. May take a few seconds. 2 EEPROM boot in progress. May take a few seconds. 1 CPU ROM boot in progress. Blank May take a up to 5 minutes for some devices. Program control has been transfered to: a secondary boot, an EEPROM boot, UBA/M9312 boot. or a SYSTEM MAINTENANCE TABLE 5-2 START-UP DIAGNOSTIC ERROR MESSAGE DESCRIPTIONS DESCRIPTION ) ERROR MESSAGE M8190 CPU Cache Error CPU cache logic error M8190 FP Error CPU Floating point error M8190 CPU ROM Error CPU ROM logic or checksum error M8190 EEPROM checksum CPU EEPROM logic or checksum error Error M8190 clock Error CPU clock logic or power supply clock : error M8190 CPU Error Other CPU errors UNIBUS signal Error A UNIBUS signal is always asserted No memory in location 0O Memory failed or is addressed Memory Error General memory test errors Memory CSR Error Memory errors during parity or ECC Other UBA errors UBA Error - This Unexpected trap to D T is a general error message that occurs during any unexpected location O testing UBA Cache error Error M8191 UBA Cache M8191 incorrectly traps. The address of follows this message. D D D D O S D WD o0 ATD CED D T e D D S D WA D D T e T D ) D D D D G the trap D 0 e G S0 S0 D W G SR e e s o wes o S e 5.3.2 Unexpected Trap and MMU Error Code Descriptions Figure 5-2 shows an example of the error code and message The error number of displayed when an unexpected trap occurs. unexpected traps is always the current test number plus 100. NOTE Operator input is wunderlined examples. 5-12 in the following SYSTEM MAINTENANCE In the example the error code (or test number) is 62. The actual The Start-up Test display will error code is read as 162. two-digit display. a only is it display only 62 since Unexpected traps are always considered fatal errors. Testing in progress - Please wait Error 162 Unexpected trap to location 250 MMU See troubleshooting documentation Updated PC= 173436 RO = 101365 R4 = 101367 PCR page= 15 Program listing address= 015436 Command Description 1 Rerun test Loop on test 2 R2 = 177746 R6 = 172276 Rl = 076410 RS = 000250 Type a command then press FIGURE 5-2 R3 = 177744 Par3 = 052400 the RETURN key: UNEXPECTED TRAP ERROR EXAMPLE E letter single For codes 76 and 75, the ROM code displays the message the After 5-3.) Figure (See followed by the test number. only The is displayed, the ROM code will not except any input. option for the user is FIGURE 5-3 to restart the system or repair the problem. 5.3.3 EXAMPLE OF TEST ERROR Boot Program Error Codes/Messages Error codes 21 through 10 with the boot programs (described in Table 5-1) are associated for disks, tapes, and DECNET devices. 5f13 SYSTEM MAINTENANCE ROM-resident Dboot These errors are applicable for all areCPUwritt to pass these programs, and any EEPROM boots thats 14, 16, en and 17 apply to errors back to the CPU ROM. Only error UBA or M9312-type ROM boots. am Figures 5-4 and 5-5 show examples of an error. from a boot progr when the BOOT command is used in Dialog mode Test. Commands are Help, Boot, List, Setup, Map and <CR> DL1 B key: N RETUR the press Type a command then Trying DL1 Message 12 f No disk present, or drive is not loaded correctly Description Command Reboot 1 Go to Dialog mode 2 Type a command then press the RETURN key: FIGURE 5-4 BOOT PROGRAM ERROR EXAMPLE Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: B DL3 <CR> Trying DL3 Message 15 Non existent drive Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: FIGURE 5-5 BOOT ERROR MESSAGE EXAMPLE sequence all boot When the ROM code enters the automatic boot pass through the list first the on essed error messages are suppr booted on the ly ssful succe is e devic no of boot devices. If boot sequence first pass the ROM code will restart the automaticand display all and try to boot all of the selected devices again applicable error messages. SYSTEM MAINTENANCE When the ROM code has failed to boot any of the devices automatic boot list, Dialog mode is entered. the in selected Figure 5-6 shows an example of a boot error display when the automatic boot sequence failed to find a bootable device and Dialog mode 1is entered. Testing in progress - Please wait Memory Size 9 Step Step is memory 1 1024 K Bytes test 2 3456 7 89 Starting automatic boot Trying DUO No disk present, or drive is not loaded Trying DU2 Trying DU3 Drive not ready Drive Error Trying DUl Trying DLO Non bootable media in the drive No disk present, or drive is not loaded Commands are Help, Boot, List, Setup, Map and Test. Type a command then press the RETURN key: FIGURE 5-6 AUTOMATIC BOOT ERROR MESSAGE EXAMPLE 5.4 SYSTEM TROUBLESHOOTING AIDS The system provides the following troubleshooting aids: 1. LED monitors 2. Voltage test points 3. Monitor logic with audible alarm. Figure The LEDs are located on the front panel and the modules. Figure 5-8 5-7 shows the front panel monitor LED labeled DC ON. and test points The LED. shows the location of each module monitor logic with alarm are located on the MDM module. 5.4.1 Front Panel The front panel indicator DC ON monitors the DC output voltages the main power supply.(See Figure 5=7.) When the front panel on keylock switch is in the ENABLE, SECURE, or STANDBY positions, ac power is supplied to the power supply. 5-15 SYSTEM MAINTENANCE the If the DC ON LED is OFF, the probable cause is one of supply regulators. regulator Each . located on the MDM module. ~ " N 1] dlilgli power LED monitor has a specific uane R SECURE STANOSY PDP-11/84 C - RESTANT Docon b nun L Lt !_1 C O satreny . \ STANT-UP TEST (@ LT FIGURE 5-7 SYSTEM FRONT PANEL 5.4.2 MDM Module The MDM module provides the following troubleshooting aids: voltages 1. LEDs that monitor most power supply 2. Test points for checking power supply voltages. 2., Blower/fan rotation monitoring logic. (see Figure 2-8). Figure 5-9 shows the location.of the MDM module LEDs test and LEDs D1 through D5 each monitor one or two power supply points. Loss of a voltage turns the associated LED off. voltages. checked The tolerance for each voltage should be within +/- 10%, If a voltage is found on the test points located above the LEDs. to be out of tolerance or not present, one of the power supply regulators specified in Table 5-3 is the probable fault. If the The rotation monitor logic indirectly checks the +12 Vdc. monitor the pulse, sed rotation-ba a send to fail blower/fans logic causes an audible alarm to sound and powers down the system one minute later. ~ SYSTEM MAINTENANCE TABLE 5-3 POWER SUPPLY REGULATOR FAULT ISOLATION H7200 in H7202-KA H7213 in H7202-KA H7211 in +5V EXPANSION POWER SUPPLY H7200 in H7202-KB +/-15V EXPANSION POWER SUPPLY H7211 in H7202-KB +5V MAIN POWER SUPPLY +5VBB AND +12V BLOWER/FAN +/-15V MAIN POWER SUPPLY H7202-KA CARD CAGE TOP FOR CABINET REAR FOR BOX 3 4 2 1 MOM " MDM (\7677) o A\ wnnnnnnn-nn D1 = +5V (MAIN POWER SUPPLY! D2 = +5V BB AND + 12V (BLOWER FANS! D3 = +15V AND - 15V (MAIN POWER SUPPLY} D4 = +5V (EXPANSION PQWER SUPPLY) 05 = +15V and - 15V (EXPANSION POWER SUPPLY} D1= ERROR LED =/ DIAGNOSTIC LEDs Mo L DCOX POWER LED (GREEN:! cd - DIAGNOSTIC LEDs 1 /‘ -~ L= IGREEN) MINIMUM LOAD MODULE (M7556) TM|} 1. ||| el s | LS8 D2 = +5V BB POWER 02= .15V (OFF)- D1 = +5V B8 MAIN POWER SUPPLY (RED! MSV11-J8:JC MEMORY (MB637-8A/CA) KDJ11-8F (MB190} f—MINIMUM LOAD MODULE (M7558) —02= <15V (GREEN) /.-—-01 = +5V MAIN POWER SUPPLY (RED) U FIGURE WYY YY el 5-8 MDM MODULE LED LAYOUT SYSTEM MAINTENANCE MDM Module Notes ¢ontains system the if an Q. An M7677-YA MDM must be used b. I1f the MDM is the suspected problem, check the voltage test . H7231-E BBU. points the and NPG module replacement. configuration prior to pack switch : ’ (See Figure 5-9.) Loss of a voltage turns off the associated LED. the check If an LED indicates 1loss of a voltage, ement. replac tor requla corresponding test point prior to switch The setting of the NPG (non processor grant) select switch NPG Each . status NPG pack is used to select Figure 12. h throug 5 slot ane backpl corresponds to a CPU 5-9 shows the location and slot number of the each NPG switch. For non DMA devices a CPU backplane slot the NPG switch should be in the ON position. A common NPG problem occurs when DMA devices are installed with the NPG switch ON (arbitration mechanism is bypassed). This causes an error code 20 when attempting to device, a indicating error. controller problem, turn OFF the NPG switch for that slot. . 18: -ouv c.;o 58106/ (10 P St[:" ’ {20 iy 05 D4 03 g;] " 02 OV go0Qo0 Oksu NUTE O& AUDIULE ALARM SLOV NUMSER 12 5 08000008 or¢ NG o Ng PACK wum;n ROTE -5 IMAIN FOVEH SUPPLY! 01 U A 't ANS) 07 - +§VEB ARD « 12V (BLOWE 03 + +15V MAIN POLIE R SUPRLY 04 <5V tF RPANSIUL PUWE R SLIPBL V) 09 1%V (£ XPANSI()'. POWE R SUWPLY) FIGURE 5-9 MDM MODULE LAYOUT 5-18 boot that To correct the SYSTEM MAINTENANCE 5.4.3 KDJ11-BF CPU Module The KDJ11-BF module provides: 1. A green POWER OK LED, indicating that dc power to 2. Six red error code LEDs, which correspond to diagnostic error codes (see Figure 5-10). module the CPU is present. front In addition the error codes appear on the the Start-up panel Start-up Test display. For example, if an error code 51 occurs during the start up diagnostics, 51 appears on the front panel display and the corresponding CPU module LEDs light (that is, 1 (LSB), 4, and 6). suspected If the KDJ11-BF module is the module replacement, assure that: problem and prior to 1. The module jumper configurations are as specified in Figure 2. The 5-10, and are OFF. DIP switches CONNECTOR . FOR CABLE TO BAUD AATE SELECT 0IP SWITCH FO BAUO RATE SELECT AND BOOT STRAP CONTROL §/8 OFF AND TWO-DIGIT DISPLAY QE*\ 4N CONNECTOR FOR CABLE TO ] CONSOLE SLU \ : ROM {HY BYTE) =] 15: wio P16 i (LO BYTE) =] EEPRAOM 7 REMOTE SWITCH 1 CONTROL CONNECTOR (A-SERIES) :fl\. FPII1-A 40-PIN SOCKET (P-SERIES) D €53 .. | ¢) DCJ1Y CPU HVPR'D £80 | GATE | TP40 00 o TPa2 : / Q? , - Jo TP10 wao (SEE NOTE 1) DIAGNOSTIC LEDS (RED} LS8 ! [=5] I/llfl'l(ll/ 5:08 07:00 / | DCOK LED (GAEEN) _ TPat €3 GATE . ARRAY (SEE NOTE 2} - w20 O (I o | TP200{J0 TP210TP22 I B NOTES: 1. WHEN 24-PIN EEPROM IS USED. INSERT PIN ONE OF EPROM IN PIN 3 OF SOCKET. 2. WHEN 2K EEPROM (S USED. TPAQ IS CONNECTED TO TPAY. WHEN 8K EEPROM IS USED. TP41 IS CONNECTED TO TP42. FIGURE 5-10 KDJ11-BF JUMPER AND SWITCH PACK LOCATIONS 5.4.4 MSV11-JB/JC Memory Module The quad-height memory module provides: a. A red LED to indicate uncorrectable errors 5-19 MAINTENANCE SYSTEM b. A green LED to indicate the presence of +5 Vdc c. Two switch packs for starting and CSR address selection d. Four factory-set jumpers. LEDs, both check problem, If the module is the suspected the settings, and jumper configurations prior to module See Figure 5-11 for LED, switch pack, and Jumper switch pack replacement. layout. = - | A\ T we w3 03 i oo FIGURE The 1 - 5-11 MSV11-JB/JC LED/SWITCH PACK/JUMPER LAYOUT starting address is configured using switch pack S1l, switches 8 Table 5-4 lists the switch settings, starting addresses and decimal number. The top sixteen entries apply to the 1MB (MSV11-JB) memory and the bottom sixteen entries apply to the 2MB (MSV11-JC) memory. TABLE - e e — D D D W CED D WD D D D SWITCH SETTING* 1 23456 78 0 000O0O0CO0O 0000O0O0O01 0 000O0O0OT1O0 00000011 0000O01O00O 00000101 D 5-4 G T D D MSV11-JB/JC D e T T G e D e e SN S RO STARTING D R I W STARTING ADDRESS (OCTAL) 00000000 00040000 00100000 00140000 02000000 00240000 D R W ADDRESS oD S R D D GND D SED D I DECIMAL (K bytes) 0 8 16 24 32 40 TR D XD e S SYSTEM MAINTENANCE | TABLE'5—4 (Cont) SWITCH SETTING* STARTING 00000110 00000111 00001000 00001001 00001010 00001011 00001100 00001101 00001110 00001111 0 000X XXX 0001 XXZXX 0010XXXX 0011 XXZXX 01 00XXXX 0101 XXXX 0110XXXX 0111 XXXX 1000XXXX 00300000 00340000 | 004000005 00440000 00500000 | 00540000 00600000 ; 00640000 ! 00700000 | 00740000 ! 00000000-00740000 ADDRESS (OCTAL) 1 2 7 4 56 7 8 1011XXXX 1100XZXXX 1101 XXXX 1110XXXX 1111XXXX 13000000-13740000 14000000+14740000 15000000-15740000 16000000+16740000 17000000 17740000 = = Switch on Switch off X = Switch 1152-1272 1280-1400 12000000-+12740000 1010XXXX 1 48 56 64 72 80 88 96 104 112 120 000-120 128-248 256-376 384-504 512-632 640-760 768-888 896-1016 1024-1144 05000000-05740000 1 001X XXX 0 (K bytes) 00100000-01740000 02000000-02740000 03000000-03740000 04000000-04740000 06000000+06740000 07000000-+07740000 10000000+10740000 ©11000000+11740000 * DECIMAL 1408-1528 1536-1656 1664-1784 1792-1912 1920-2040 can be either on or off The CSR address is configured using switch pack $2, switches 1 - Each successive address is the The base address is 17772100. 4. possible CSR addresses. sixteen all lists Table 5~5 base plus 2. TABLE CSR ADDRESS SELECTIONS 5-5 MSV11-JB/JC G S2 1 WP SED TED WD GEN U IR GED GED GRS MR AND ED SETTING 2 3 4 0000 0001 0010 0011 N - GED WED D M) AP WEE GNP GNP D WD B CSR ADDRESS (OCTAL) 17772100 17772102 17772104 17772106 SYSTEM MAINTENANCE TABLE S2 SETTING 1 2 3 4 5-5 (Cont) CSR ADDRESS (OCTAL) 17772110 17772112 17772114 17772116 17772120 17772122 17772124 17772126 17772130 17772132 17772134 17772136 0 1 0 1 0 1 0 1 0 1 0 1 1MB ({MSV11-JB) The dumper configurations for the Assure memory modules are different. (MSV11-JC) Table 5-6. factory-set jumpers are as specified in JUMPER(S) 2MB the JUMPER CONFIGURATIONS TABLE 5-6 MSV11-JB/JC MODULE and that POSITION DESCRIPTION meviloge T wl OUuT 256K Dynamic RAMs w2 IN Half-populated module Vertical Reserved for future use w3,w4 MSV11-JC wl ouT 256K Dynamic Rams w2 ouT Fully populated module Vertical Reserved for future use W3,wW4 5.4.5 KTJ11-B Module support M9312 compatible The module provides four sockets to cted problem, check the ROMs suspe the If this module is ROMs. Figure 5-12. and their orientation as shown in 5-22 MAINTENANCE SYSTEM 0312 TYPE OPTION ROM SOCKETS i4} i | mOM L1 Ewes 2 177173200 17773378 - 17773576 17773400 Eve3 19 ] e ADDAESS 17773000- 17773176 £r144 i3 i SOCKET 17773600-177737%¢ e Pin NO.1 R R R R FIGURE 5-12 KTJ11-B ROM SOCKET LOCATIONS 5.4.6 Minimum Load Module The MLM modules provide minimum under the following conditions: e cl supply regulator An MLM is inserted in CPU backplane slot 3 (rows C ensure a regulator. b. power An MLM is minimum inserted to ensure regulator. If an a in minimum expansion installed, an backplane (rows of current MLM E drain CPU backplane current ? drain backpflane of 2 slot of 370 (DD11-CK A 12 mA from (rows from or and D) and F) the E the DD11-DK) is !inserted in F) to ensure a minimum current and 1loads +5VBB =15 V is the last slot of the 370 mA from the -15 Vdc regulator. drain NOTE An MLM is not required in the (CPU or expansion) if the the minimum lA of -15 Vdc When not required, the load last backplane installed modules options must be two LEDs slot draw removed from the indicate the backplane. As shown in Figure 5-13, presence of +5 and VBB the -15 MLM has Vdc. 5-23 to SYSTEM MAINTENANCE RED GREEN o ] CYRIY FIGURE 5-13 MINIMUM LOAD MODULE LAYOUT 5.5 FIELD REPLACEABLE UNITS Table 5-7 lists the field replaceable units (FRU) . TABLE 5-7 FRU DESCRIPTIONS —---n—m—‘m_——m o - m@ma-amma———pm—-—-————————————————— PART NUMBER DESCRIPTION M7677-YA* Monitor Distribution Module (MDM) M8191 KTJ11-B, UBA Module M8190~AE KDJ11-BF, CPU - FPA Module M8637-BA MSV11-JB, 1 Mb ECC Memory Module M8637-CA MSV11-JC, 2 Mb ECC Memory Module M7556 Minimum Load Module M9302 Terminator Module 70-20650-01 CPU Backplane muacauwamamgmamua@ - mau@gmummwmmnmmu——————c-—--—m— SYSTEM MAINTENANCE TABLE 5-7 (Cont) ————————--------———----—--—-—-—-— H7202-KA Power Suppiy H7202-KB Power Supp;y 54-16508 Console SL? Board H7211-B** +15, H7213-D** +12V, H7200~-C** +5V Regulaior Module 70-21888-01 Front Panei Assembly 12-22001-01 Blower 12-22271-02 Fan 877-D Power Cont?oller 120V 877-F Power ContLoller 240V 70-21116-01 Circuit breaker assembly H7231-E* Cabinet Battery H7231-F* Box DD11-CK 4-slot Backplane DDll -DK 9-slot Backplane -15V Regulator Module +5V Regulator Module OPTIONS: * be installed the BBU option. Specifies 5.6 MODULE The following Set-up H7231-E/F subsections Table and up Unit if the Unit system contains minimum etch revision. REMOVAL/REPLACEMENT removal up Back M7677-YA must ** module Battery Back PROCEDURES provide replacement selection the general procedures. procedure. 5-25 module, and the CPU Included is the CPU SYSTEM MAINTENANCE 5.6.1 General Module Removal/Replacement module) To remove any module listed in Table 5-7 (except the CPU use the following procedure. CAUTION static sensitive. Modules are 1. 2., Always wear a properly when handling modules. 3. Modules they anytime must their shipping 1. be placed removed on a static mat a backplane or from static bags. Open the cabinet front and rear doors using the hex key, or slide 2. are connected ground strap the box out of the rack. Turn either: a. breakers b. power The cabinet power supply and The box OFF, breaker OFF. circuit the AC power cord from the outlet. 3. Remove 4. Assure the ground strap is properly conneéted. S. Remove all 6. Pull cables circuit controller or from the module and label each one. the module handles out and slide the module from the backplane. 7. Place the module on the static mat. This completes the removal of a module. reverse the To reinstall a module procedure. 5.6.2 CPU Module Removal/Replacement replacement The case. special a Replacing the CPU module is module Setup Features must be confirmed, and if necessary revised to the original CPU parameters and selections. 5=26 SYSTEM MAINTENANCE have should user the installation, During the initial system eets supplied recorded the setup feature selections on the J worksh this manual. of ix in the Installation Guide or in Append original CPU the of ions select the Retrieve this form. Compare module (as specified on the Confirm and/or revise the procedure. ' features setup the with worksheets) defaults of the replacement module. using the factory-set following NOTE Dialog and Setup mode commands are described 1in subsections 4.2 and 4.3. 1. properly connected ground strap remove the module from the backplane as specified 1in Wearing a defective subsection 5.6.1. 2. Ensure that the replacement module jumper configurations and DIP switch settings are as specified in subsection 5.4.3. the replacement CPU module as in described 3. 1Install 4. Set the FORCED DIALOGUE switch to ON. 5. Power up the system, and ensure that it passes the subsection 5.6.1. startup diagnostics. 7. On successful completion of the startup diagnostics, the (See Figure 4-3.) system enters Dialog mode. S <CR> on the console Enter Setup mode by typing: 8. If the customer chose the factory defaults perform steps 9. Type: 8 <CR> to initialize the setup table to the factory default values. Answer 1 <CR> to the ROM code prompt: Are (See Figure 4-22.) you sure? O0=No, l=Yes. 10. Type: 9 <CR> to copy the 11. Type: 2 <CR> to display the Setup Table parameters. 6. terminal. and 10. Answer: (See Figure 4-14.) Otherwise begin at step 1ll. Setup Table 1 <CR> to the ROM code prompt. into the 9 EEPROM. (See Figure 4-23,) (See Figure 4-16.) Compare the replacement module parameters to the original selections; revise as required. 5-27 SYSTEM MAINTENANCE 12. Type: 3 <CR> to display the Translation Table parameters. (See Figure 4-17.) Compare the parameters to the original selections; 13. Type: 4 <CR> to display the automatic boot selections. (See Figure 4-20.) Compare the selections to the original:; revise as 14. revise as required. required. Type: 6 <CR> to display the CPU switch boot selections. (See Figure 4-21.) Compare the selections to the original; revise as required. 15. Type: 9 <CR> to copy the revised parameters and selections into the EEPROM. (See Figure 4-23.) Answer: 1 <CR> to the ROM 16. code Type: prompt. 1 <CR> to exit Setup mode and return to Dialog mode. This completes the CPU module replacement and setup procedure. 5.7 POWER SUPPLY REMOVAL/REPLACEMENT Both the cabinet and box products have a H7202-KA power supply The regulator boards and containing three regulator boards. regulator boards are the that power supplies are FRUs. Assure not defective prior to replacing a power supply. 5.7.1 Cabinet Power Supply Removal/Replacement To remove a power supply regulator supply, use the following procedure. board or the main power 1. Open the front and rear doors using the hex key. 2. Turn the power supply and power controller circuit breakers OFF. 3. Remove the AC power cord from the outlet. 4. Loosen the two captive screws holding the blower assembly. S. Slide 6. Slide the blower assembly out and up to remove it from the blower assembly out about disconnect the blower motor power cable. cabinet assembly. See Figure 5-14. four inches,and the SYSTEM MAINTENANCE POWER SUPPLY CIRCUIT BREAKER PLASTIC RETAINERS FIGURE 5-14 BLOWER ASSEMBLY REMOVAL Push the power supply tray lock of slide) tray the in. lock the tray handle until the second tray edge engages. See 5-15. N\ Figure left (located on the Slide the tray out by pulling on *e OOOOOOOOOO[ 0.0 COws snacx¢ w J 0 0 2 © O & OO 4 FIGURE 5-15 CABINET POWER SUPPLY REMOVAL Loosen and remove the power supply hold down located near the top left edge of the power supply. Using a phillips head screwdriver and bus wires. remove the bracket four 5 VDc SYSTEM MAINTENANCE 10. Remove 11. Using a 3/8-inch 12, Pull the power the two ribbon cables wrench supply and remove the the forward AC input ground and cable. lug. remove it £from the cover of cabinet. 13. Loosen the three captive the power supply and screws remove the holding the top cover. See Figure 5-16. -5 1V & *12V0C uTED VD H711) VDL SYAND OFFS / L] |Yo FIGURE 5-16 POWER SUPPLY REGULATOR REMOVAL 14. or the communications (H7213) To remove the memory/fan option (H7211) regulators, gently lift them out by grasping each corner standoff and lift up. 15, supply over To remove the H7200 regulator, turn the power loosen regulator, the supporting while and (open side down) the six phillips head screws securing it to the chassis. 16. Remove the regulator from the chassis. This completes the removal of the regulators boards To reinstall, reverse the above procedure. supply. and power 5.7.2 Box Power Supply Removal To remove a power supply regulator following procedure. 1. Turn the circuit breaker OFF. 5-30 or power supply, use the SYSTEM Unplug the AC power Remove the box top screws and lifting Loosen the cover. See from <c¢over the the by cover outlet. removing Figure the four on the power and lift to captive off. two phillips head screws Slide the cover backwards supply remove it. 5-17. FIGURE To cord MAINTENANCE remove 5-17 either communications out by grasping option the POWER SUPPLY the memory/fan (H7211) REMOVAL (H7213) or the regulators, gently lift them two corner standoffs and lift up. To remove the H7200 regulator, the power supply must be removed from the box. Loosen the four screws that secure the power supply to the chassis shelves. Lift the power supply, out of the box. .See Figure 5-18. STAND OGPS ®) FIGURE 5-18 POWER SUPPLY 5-31 REGULATOR REMOVAL SYSTEM 7. MAINTENANCE After removal, turn the power supply over (open side down) and while supporting the regulator, loosen the six phillips head screws securing it to the chassis. 8. Remove the regulator from the chassis. NOTE If the power supply 1is replaced, set the AC 120/240 voltage select switch to match the outlet AC voltage. This completes the removal procedure for regulators. To reinstall reverse the above procedure. the box regulators power and cupply power and supply., 5.8 CABINET BLOWER REMOVAL/REPLACEMENT To remove the blower assembly use the following procedure. 1. Open the front and rear doors using the hex key. 2. Turn the power supply and power contrcller circuit breakers 3. Remove the AC power cord from the outlet. 4. Loosen the two captive screws holding the blower assembly, slide the blower assembly out abcut four inches, and OFF. disconnect the blower motor power cable. 5. Slide the blower assembly out while lifting up, and it from the cabinet. FIGURE 5-19 See Fiqure 5-19. BLOWER ASSEMBLY 5-32 REMOVAL remove SYSTEM MAINTENANCE This completes the removal procedure for cabinet blower assembly. Reinstall the blower assembly in the reverse order. 5.9 BOX FAN REMOVAL/REPLACEMENT There are three fans used in the box product. Two fans cool the module cardcage and the third cools the power supply. To remove any one of the fans 1. Turn the circuit 2. Unplug 3. the use the following procedure. breaker off. AC power cord Remove the four screws Remove the bezel from from the from the the outlet. rear of the box flange. box. 4, Loosen the six captive of the fans. Lift off screws from the metal grid the metal grid. 5. Loosen the two phillips screws securing the fan mounting position on the plenum. See Figure 5-20. 6. Unplug the fan power cable and remove the fan. in front to its a new SCREWS (4) FIGURE 5-20 BOX FAN REMOVAL CAUTION When installing a points toward the fan insure plenum. the airflow When installing a replacement fan, tighten mounting screws snug. DO NOT OVERTIGHTEN. This fan, completes the fan reverse the above removal procedure. procedure. 5-33 To arrow the reinstall SYSTEM MAINTENANCE 5.10 Front Panel Removal/Replacement cabinet The box and front are panels identical, different removal and replacement procedures. but have 5.10.1 Cabinet Front Panel Removal/Replacement To remove the cabinet front panel use the following procedure. 1. Open the front and rear doors using the hex key. 2. Turn the power supply and power controller circuit bhreakers OFF. 3. Remove the AC power cord from the outlet. 4. Disconnect the cable from the front Figure panel assembly. See 5-21. /’/flk 74k MOUNTING HARDWARE F T iN — m— CANLE = FIGURE 5-21 CABINET FRONT PANEL REMOVAL 5. Remove the four 3/8-inch nuts holding the the 6. ¢to front panel panel assembly. door. Remove the front panel off the standoffs. This completes the removal of the cabinet front To reinstall the front panel, reverse the above procedure. SYSTEM 5.10.2 To Box remove Turn Front the Panel box the the AC Remove the front the box Remove from panel supply Unplug screws Removal/Replacement front power power the use the circuit from the cable following breaker procedures. OFF. outlet. bezel by bezel rear loosening that is plugged the four Phillips side. (see Figure 5-22). the MAINTENANCE and Pull removing the four the away from into bezel the front panel assembly. Loosen the and front the to front the panel head screws securing chassis. assembly. ‘\! / Remove remove panel 2 SEZ © 8 flfii Lfaur Z4-852 2411 £22gl) g , \ ~. TM~ FIGURE 5-22 FRONT PANEL REMOVAL This completes the removal of the front panel. box front panel, reverse the above procedure. To reinstall the SYSTEM MAINTENANCE 5.11 CIRCUIT BREAKER REMOVAL/REPLACEMENT ers for the power supply on (s). Both products have circuit break the mbly is located externally The box circuit breaker asse lower breaker is located inside right side. The cabinet circuit : one for the main power supply kers brea right side and has two and one for the expansion supply. 5.11.1 Cabinet Circuit Breaker Removal/Replacement remove To procedure. 1. the circuit assembly use breaker following the Open the front and rear doors using the hex key. 2. Turn the power supply and power controller circuit breakers 3. Remove the AC power cord from the outlet. 4, . OFF Unplug the cords connected to the assembly. See Figure 5-23. POWER SUPPLY CIACUIT SREAKER / PLASTIC RETAINERS FIGURE 5-23 CABINET CIRCUIT BREAKER REMOVAL 5. 6. ng in on the Remove the blower power connector ngby tthepushi connector loose. connector release clips, and pulli it breaker assembly Loosen the two screws securing the circu e the assembly. to the the cabinet support and remov This completes the removal of the circuit breaker assembly. 5=-36 SYSTEM To reinstall the assembly, reverse the above MAINTENANCE procedure. Ensure that the power cable plugged into the unswitched power controller inserted outlet is 5.11.2 Box Circuit To remove the into Jl. Breaker Removal/Replacement «circuit breaker assembly, use the following ‘ procedures. l. Turn the circuit breaker off. 2. Unplug 3. A4, through Remove the rear 4 bulkhead sections marked Al See Figure 5-24. each section has two flathead screws. the AC power cord from the outlet. SCREWS (4} CQIRCUIT BREAKER FIGURE 4, Remove box. 5. Reach 6. the 5-24 four the CIRCUIT screws inside the release the three Remove BOX securing box cable circuit BREAKER through clips breaker the the inside REMOVAL breaker assembly bulkhead along the through the to the access and box wall. rear bulkhead access. 7. Remove the the four screws securing the wires circuit breaker. Label each wire. on the back of This completes the replacement procedure for the circuit breaker assembly. To reinstall the <c¢ircuit breaker, reverse the procedure. 5-37 SYSTEM 5.12 To MAINTENANCE CABINET POWER CONTROLLER REMOVAL/REPLACEMENT remove the 877 power controller assembly, wuse the following procedure. Open the front and rear doors using the hex key. Turn the power supply and power controller circuit breakers OFF. Remove the AC power cord from the outlet. Lift the left side panel straight up from both sides. See Figure 5-25. careful, the outside panel is heavy. [ Be L SHOULDER SCREW (4) et EMI SHIELD %S{:REW 121 4! FIGURE Remove the EMI the 5-25 CABINET SIDE PANEL REMOVAL two phillips and shield to the Label each power plug and plugs. Loosen See the Figure 10 its screws securing side. receptacle, and remove the 5-26, phillips controller to the four shoulder cabinet head screws cabinet bulkhead. Grasp the CONTROLLER metal power cord the controller from the cabinet rear. 5-38 securing the $ee Figure 5-27. restraint and power remove % T | N P L TTTIBVYY g SYSTEM MAINTENANCE FIGURE 5-26 POWER CONTROLLER REMOVAL = SIDE VIEW 827.0/F CONTROLLER AEAR VIEW FIGURE 5-27 POWER CONTROLLER REMOVAL - REAR VIEW This completes the removal procedure for power the controller. To reinstall the controller reverse the above procedure. 5.13 SLU INTERFACE ASSEMBLY REMOVAL/REPLACEMENT Both products contain an SLU Interface different removal and replacement procedures. Assembly, Dbut use 5.13.1 Cabinet SLU Assembly Removal To remove the SLU assembly use the following procedure. 1. Open the front and rear doors using the hex key. 2. Turn the power supply and power controller circuit breakers OFF. 5-39 SYSTEM MAINTENANCE 3. Remove the AC power 4. Loosen the ten cord captive from the outlet. screws securing the bulkhead to the frame. 5. Open the 6. Unplug 7. Loosen and Unplug Figure by pulling down the SLU cable connector 8. bullhead remove to the the cable 5-28. (console the from the terminal) two hex top. from the standoffs connector. securing the cross member. from the SLU assembly connector. See | stz scngms 54 FIGURE 5-28 |L e SCREW (21 CABINET SLU ASSEMBLY REMOVAL 9, Hold the SLU assembly while removing the 10. Remove the SLU assembly from the cabinet. two secure the SLU assembly to the cross member. This completes the removal of the SLU assembly. SLU assembly reverse the above procedure. screws that To reinstall the SYSTEM MAINTENANCE 5.13.2 Box SLU Assembly Removal/Replacement To remove the SLU assembly, use the following procedure. 1. Turn the circuit breaker off. 2. Unplug the AC power cord from the outlet. 3. Remove the cable plfigged into the SLU connector. 4. Remove the two hex standoffs securing the connector to the 5. Remove the four screws on the top remove the back panel. and cover, cover. 6. Remove the connector plugged into the SLU assembly board. 7. Remove the two assembly mounting screws from 8. Remove the hex standoffs from the SLU connectors. the box. See Figure 5-29,. the rear of / L * - ) |3 . e [] 1000 fr * AL e W (R pe— r—/‘flt;ll'l . " L AW i I FIGURE 9. 5-29 BOX SLU ASSEMBLY REMOVAL Remove the SLU assembly from the box. This completes the removal of the SLU assembly. assembly reverse the above procedure. To reinstall the 5.14 CPU BACKPLANE REMOVAL/REPLACEMENT Depending on the number of options installed in the system, a backplane replacement can be a time consuming task. It is recommended that you replace the backplane ONLY if is clearly known to be the faulty FRU. 5-41 SYSTEM MAINTENANCE 5.14.1 Cabinet Removal remove the CPU backplane, the rear doors and front use following procedure. using the hex key. Open the Turn the power supply and power controller circuit breakers . OFF Remove the AC power cord from the outlet. Remove and label all the module cables. Remove all cabinet number. backplane slot modules Remove the right side panel. each 1label and its with See Figure 5-30. . L SHOULDER SCREW (4) L 1 — / 1 L\ . 7 EMI SHIELD SCREW (2) e Z \\ " SIDE PANEL \ \\. To Backplane FIGURE 5-30 SIDE PANEL REMOVAL Remove the two phillips head and four shoulder securing the EMI panel to the cabinet frame. Remove the left Remove the four shoulder the cabinet side screws panel. screws frame. 5-42 securing the EMI panel to SYSTEM 10. MAINTENANCE Loosen the ten screws securing the bulkhead to the See Figure 5-31. frame and lower the bulkhead. cabinet SAUD /mcn BULK WEAQ,~ SCREW (10) te [ * ettt FIGURE 5-31 CABINET REAR VIEW 11. Remove four the black plastic retainers plastic lexan cover over the space backplane location. See Figure 5-32. for cru SACKPLANE SCREW (4} METAL PANEL METAL cry P. BACKPLAKE FIGURE 5-32 CABINET 5-43 BACKPLANE REMOVAL securing the the ‘ expansion SYSTEM 12, MAINTENANCE If an expansion connectors. cabinet 130 backplane Remove is the remove the power left side access. Disconnect the power cables Remove the six screws securing the metal panel over rear access to the backplane. Remove the metal panel. the 16. Remove 17. cable to the card clamps the four on the expansion backplane. 15. the the the the rear access for covering 14, Locsen 1installed lexan cover through the metal panel backplane. flathead screws securing the backplane Remove the backplane by pulling the it toward the rear and twisting it through the side panel access. cabinet This completes the removal procedure for the backplane. reinstall the backplane reverse the above procedure. 5.14.2 To Box Backplane remove 1:0 Turn the To Removal/Replacement backplane, the ¢to cage. power use circuit Unplug the AC power Remove the screws the following procedure. breaker OFF. cord from the securing the outlet. top cover, and remove the cover. Unplug all the module cables and label each with its module number. Unplug all the modules Remove the four number. cables. Unplug labeling 1/4-inch the nuts the module securing two backplane cables. with the ' its four See Figure slot power 5-33. SYSTEM FIGURE BACKPLANE card cage. 8. Slide the card cage to the rear of the box. Lift and the card cage to the rear and remove from the box. slide Turn the card cage screws. located over Remove This and the at REMOVAL Remove the mounting screws BOX 7. 9. two 5-33 MAINTENANCE the rear remove the of the four backplane backplane. completes the removal procedure for the CPU backplane. reinstall the backplane reverse the above procedure. 5.17 CABINET PERIPHERAL To ACCESS Always extend the front stabilizer bar before sliding an. option out of the top portion of the cabinet. The bar keeps the the cbainet from tipping forward. (See Figure 5-38.) SYSTEM MAINTENANCE FIGURE 5-38 STABILIZER BAR EXTENSION APPENDIX A CPU INSTRUCTION TIMING CPU INSTRUCTION TIMING A.1l INTRODUCTION The execution time of an instruction a. The b. The mode of addreséing used, and c. The In type for type general, of the instruction memory total The tables in this referenced. execution section on: executed, being instruction fetch/execute calculation/fetch time. depends time time can be 1is the plus used the to sum of the operand(s) calculate the base address length of an instruction in terms of microcycles (MC). Tables A-1 thru A-8 list the standard and floating—-point instructions, their Op Code listing, and execution times (MC). The EXECUTION MC column specifies the number of microcycles required to fetch/execute the base instruction. The R/W column specifies the number of read microcycles (R) and write microcycles (W) in the EXECUTION MC column. Any remaining microcycles are non I/O (NIO). If the instruction operands, a destination table) TABLE or and Fl1 The numbers involves reference DEST is made TABLE. the calculation/fetch to a in the The separate last tables of table column, referenced are one (a wusually S1, or more source Dl or labeled thru D6, thru F5. The source/destination tables specify the number of microcycles the source/destination calculation/fetch requires, and how many of these are read or write microcycles. As before, any remaining microcycles are NIO. contained in the tables are based on the assumptions that: a. b. A memory read A memory write must last must a last minimum a of minimum four of CLK eight periods, CLK and c. Any An NIO lasts four CLK periods (no periods, DMA). wait states caused by slower memory or a DMA transfer must be to the total instruction time. If wait states are required, the first wait state of a nonstretched read or NIO cycle will 1last four <clock periods, and can continue 1in increments of two clock periods. Further wait states for stretched cycles occur in increments of two clock periods. added Floating-point instruction execution times are given The actual execution time will vary depending on the A-2 as a type range. of data CPU being The operated following EXAMPLE 1: How long does INSTRUCTION TIMING on. examples illustrate a MOV RO,@ 2044 how to use instruction the tables. last? 1: Step From Table A-2 , the execution time for the MOV base instruction is found to be 1 MC, or four CLK periods. This consists of one read and no write microcycles (R/W column). Depending on the type of memory in the system, the microcycle may be stretched. If so, the microcycle lasts at least eight CLK periods and may be stretched thereafter in increments of two CLK periods. Step To 2: find (RO), mode 0 operand Step the operand refer to register 0 calculation/fetch time for the source operand Table S-1. From Table S-1, it is shown that a calculate/fetch takes 0 MC. Note that the is already available to the DCJ1l (in the register file). 3: To find the operand calculation/fetch time for the destination operand (the contents of memory location 2044), refer to Table D-3. Table D-3, specifies that a mode 3 register 7 calculate/ fetch requires microcycle). 3 MC Note (i.e., that the one read microcycle remaining and one write microcycle 1is an NIO account. If microcycle. The type of memory in the system must be taken into the read <cycle is stretched, the stretched cycle lasts at least eight CLK periods and may be stretched thereafter in increments of two CLK periods. The write microcycle lasts at least eight CLK periods and may be stretched 1in increments of two CLK periods. Step For 4: a determination microcycles. 16 CLK periods In if of this no the minimum example, microcycle It time is 1 required, + 0 stretching + 3, or occurs). total 4 MC wup the (which is INSTRUCTION TIMING CPU EXAMPLE The 2: source show 2 register for one of How long Step As and 7 operations. these. does instruction From Table in is 2 Table 14 floating—-point This 2000 example column shows instruction a instructions for certain mode timing calculation last? A-8, the base instruction time for the CLRD MC. F-2, the register PRECISION). However, add read 1 calculation/fetch 7 This instruction memory Step CLRD for the microcycle 2: mode base an in 1: specified Step destination tables a negative number reference) means that 1 is MC time for the shown as (-1 should be operand under subtracted from the time. MC for cycles to the memory account write for. operation. There are no (assumes no ' 3: Total up the microcycles: cycle 14 =1 +1 = 14 MC minimum stretching). TABLE A-1 SINGLE OPERAND INSTRUCTIONS TIMING OP MNEMONIC INSTRUCTION CODE EXECUTION SOURCE DEST TABLE LISTING MC R/W TABLE General CLR(B) Clear 0050DD 1 I/0 - D3 COM(B) INC(B) DEC(B) NEG(B) Complement (1l's) Increment Decrement Negate (2s complement) Test 0051DD 0052DD 0053DD 1 1 1 I/0 I/0 I/0 - D4 D4 D4 0054DD 0057DD 1 1 I/0 I/0 - D4 D4 TST(B) (a DOUBLE - TABLE A-1 TIMING INSTRUCTION CPU (Cont) Rotate and Shift 0060DD 0061DD 1 1 /0 I/0 - 0062DD 1 1/0 - 0055DD 0056DD 0067DD 1 1 1 (low bit interlocked) 0072DD interlocked 0073DD ROR(B) Rotate right ROL(B) Rotate left ASR(B) Arithmetic SWAB shift right Swap bytes D4 - D4 I/0 I/0 I1/0 - D4 D4 D3 5 1/1 - D4 4 1/1 - D4 1 0003DD D4 D4 /0 Multiple-Precision ADC(B) Add carry SBC(B) Subtract carry Sign extend SXT Multiprocessing TSTSET Test and set WRTLCK Write TABLE A-2 DOUBLE OPERAND INSTRUCTIONS TIMING MNEMONIC INSTRUCTION OP CODE EXECUTION SOURCE DEST TABLE MC R/W TABLE 01SSDD 02SSDD 06SSDD 16SSDD 1 1 1 1 1/0 1/0 1/0 1/0 S-1 S-1 S-1 S-1 D3 D2 D4 D4 03SSDD 1 1/0 -1/0 1/0 S-1 S-=1 S-=1 D2 1/0 == D1 - D1 LISTING General MOV(B) Move CMP(B) Compare Add ADD Subtract SUB _ Logical BIT(B) BIC(B) BIS(B) Bit test (AND) Bit clear Bit set (OR) 04SSDD 05S8SDD 1 1 D4 D4 Register MUL Multiply 0704SS DIV Divide 071RSS 22 (Notes 34 5,11 1/0 INSTRUCTION TIMING CPU TABLE A-2 (Cont) shift automatically Arith shift combined Exclusive (OR) (Notes 6,7,12) 072RSS 4 1/0 - D1 073RSS 5} (Note 1l 1/0 - D1 1/0 - D4 074RDD TABLE A-3 . GEN S S G S GED A S UM GED GED GED TED GED GED AED GEL D GND e GEE W G van il 13) BRANCH INSTRUCTIONS D G W I N S D W S I CHED GRS GED G G D D R IR TS W G G R D D D I S S G - WD - TIMING MNMONIC INSTRUCTION BRANCH BRANCH BRANCH OP NOT TAKEN CODE TAKEN LISTING MC R/W MC 000400 001000 001400 100000 100400 102000 102400 103000 103400 2 2 2 2 2 2 2 2 2 1/0 1/0 1/0 1/0 1/0 1/0 1/0 1/0 1/0 4 4 4 4 4 4 4 4 4 2/0 2/0 2/0 2/0 2/0 2/0 020000 002400 003000 2 2 2 1/0 1/0 1/0 4 4 4 2/0 2/0 2/0 003400 2 1/0 4 2/0 101000 101400 103000 103400 2 2 2 2 1/0 1/0 1/0 1/0 4 4 4 4 0 77RNN 3 1/0 5 BRANCHES BR Branch BNE Br BEQ Br BPL Br BMI Br BVC Br BVS Br BCC BCS Br Br SIGNED if if if if if if if if (unconditional) not equal (to 0) equal (to 0) plus minus overflow is clear overflow is set carry is clear carry is set CONDITIONAL BRANCHES Br if BLT Br BGT Br if less than (0) if greater than (0) BLE Br if BGE (to greater or equal 0) less or equal (to 0) BRANCHES UNSIGNED CONDITIONAL BHI Br if BLOS if lower or same Br if higher or same Br if lower Subtract 1 and branch BHIS BLO SOB higher Br (if /= 0) CPU INSTRUCTION TIMING TABLE A-4 JUMP and SUBROUTINE TIMING MNEMONIC INSTRUCTION JMP Jump RTS Return JSR MARK Jump to subroutine from subroutine Stack cleanup - OP CODE EXECUTION MC R/W DST 0001DD - -- D5 D6 (Note 4) 00020R 0064NN 5 10 3/0 3/0 -- (Note 14) LISTING 004RDD - - TABLE TIMING MNMONIC INSTRUCTION EMT Emulator trap TRAP Trap BPT Breakpoint trap IOT 1Input/output trap RTI Return from interrupt RTT Return from interrupt OP CODE LISTING EXECUTION 104000--104377 104400--104777 000003 000004 000002 000006 MC R/W 20 20 20 20 9 9 4/2 4/2 4/2 4/2 4/0 4/0 TABLE A-6 CONDITION CODE OPERATORS TIMING MNEMONIC INSTRUCTION CLC CLV CLZ CLN CCC SEC Clear C Clear V Clear 2 Clear N Clear all CC bits Set C OP CODE LISTING 000241 000242 000244 000250 000257 000261 EXECUTION MC R/W 3 3 3 3 3 3 1/0 1/0 1/0 1/0 1/0 1/0 CPU INSTRUCTION TIMING TABLE SEV SEZ SEN SCC A-6 000262 000264 - 000270 000277 Set V Set 2Z Set N Set all CC bits 1/0 1/0 1/01/0 3 3 3 3 TABLE A-7 MISCELLANEQOUS INSTRUCTIONS TIMING MNEMONIC HALT OP CODE LISTING INSTRUCTION EXECUTION MC 000000 Halt R/W DEST TABLE - - -- - -- WAIT Wait for interrupt 000001 - NOP SPL MFPI MTPI MFPD MTPD MTPS (No operation) Set priority level to N Move from previous instr space Move to previous instr space Move from previous data space Move to previous data space Move byte to PSW PS <- (svc) 000240 00023N 0056DD 1065SS 1066DD 1064SS 3 7 5 3 5 3 8 1/0 1/0 1/1 2/0 1/1 2/0 1/0 D-1 D-3 D-1 D-3 D-1 <~ PS <7:0> 1067DD 1 1/0 D-3 <- proc code 000007 2 1/0 - RESET Reset external bus MFPS MFPT CSM Move byte from PSW (dst) Move from processor (R0<7:0>) 000005 28 0070DD Call to supervisor mode 3/3 TABLE A-8 FLOATING-POINT INSTRUCTIONS TIMING MNEMONIC INSTRUCTION ABSD ABSF ADDD ADDF CFCC Make Absolute Make Absolute Add Add Copy Floating EXECUTION MC OP CODE LISTING 1706 fdst fdst 1706 172 (AC) fsvc 172 (AC) fsve A-8 MIN 23 19 41 31 TYPICAL 48 35 NON MODE 0 MAX TABLE 24 20 119 102 FF FF D-1 CPU CMPD Compare CMPF Compare DIVD Divide Divide ILd & C DIVF LDCDF & F to D Ld & C from 173 173 174 174 (AC (AC (AC (AC + 4) + 4) + 4) + 4) 177 (AC + 4) 26 177 (AC + 4) 21 167 to D 177 (AC) src 42 LDCIF Id & C Integer to F 177 (AC) src 36 LDCLD ILd & C Long Integer to D 177 (AC) src 52 LDCLF Ld 177 172 (AC) src (AC + 4) 26 44 16 176 172 (AC + 4) (AC + 4) 17 12 17 18 13 Status 1701 src 6 6 Separate 171 MODD FPP Program Multiply and MULF Integer and Fraction Multiply Multiply NEGD Negate NEGE Negate MODF MULD SETD SETF Set Floating Set Floating Double Mode Set SETL Set Mode 170011 Integer 170002 Mode Long Integer + 4) 171 (AC + 4) 171 (AC) fsrc 171 (AC) fsrc 1707 fdst 1707 fdst 170001 Mode SETI (AC 170012 202 217 268 82 94 115 173 165 56 22 o | LDFPS ! LDF Exponent 61 23 ) 1) LDEXP F ) T to W LDD Load Load Load Load 18 19 6 6 14 Long W C tx) & Integer PPTT Integer ] Ld C 170000 1704 fdst 1704 fdst B2 LDCID from F Dto LDCFD Codes Condition Clear Clear —_ CLRD CLRF (Cont) = TABLE A-8 INSTRUCTION TIMING CPU TIMING INSTRUCTION (Cont) TABLE A-8 STCDF STCDI "STCDL : St & C from D to F St & C from to Integer & C from D Long Integer STCFD St & C from F to D STCFI St & C from F to Integer STCFL St & C from F to Long Integer St to STST SUBD SUBF TSTD TSTF Store (AC) fdst 17 176 (AC) fdst 176 (AC) 176 (AC) 175 175 STD Store STEXP Store Exponent STF Store STFPD 176 174 175 174 20 F-2 26 38 F-5 fdst 26 54 F-5 fdst 19 20 F-2 (AC + 4) ' (AC + 4) (AC) £fdst (AC) dst (AC) fdst 23 35 F-5 23 51 F-5 12 16 8 12 16 8 F-2 F-5 F-2 9 9 FPP Program Status Store FPP Status Subtract Subtract Test Test 1702 dst 1703 dst 7 173 (AC) fsrc 173 (AC) fsrc 1705 fdst 1705 fdst 47 37 11 9 SOURCE MICROCODE MODE REGISTER CYCLES D 55 41 7 F-5 122 104 12 10 F-1 F-1 F-1 F-1 READ MEMORY CYCLES 0 0==7 0 0 1 2 2 3 3 4 4 0==7 0--6 7 0--6 7 0--6 7 2 2 1 4 '3 3 6 1 1 1 2 2 1 2 (Note 1) 5 5 0--6 7 5 8 2 3 (Note 1) 6 7 0==7 0-=7 4 6 2 3 D G L S I D TR G D D CER SEP F-5 ' SOURCE - - D GID VD GFD PR GED SN CN0 GED CAD GED GED 4D WD Y~ WD GES WED GED SED SND T GER TEP YED D CED CHD WER GNb CND D CPU TABLE D—-1 DESTINATION ADDRESS: DEST INATION MODE DESTINATION REGISTER 0 1 2 2 3 3 4 4 5 5 6 0--7 0==7 0--6 7 0--6 7 0--6 7 O=--F¢ 7 0-=7 7 0-=7 READ-ONLY MICROCODE CYCLES INSTRUCTION TIMING SINGLE OPERAND READ MEMORY CYCLES 0 2 2 1 4 3 3 0 1 1 1 2 2 7 5 9 4 2 2 3 2 6 3 DESTINATION DESTINATION MICROCODE READ MODE REGISTER CYCLES CYCLES 2) (Note 3) MEMORY 0 0~=7 0 0 1 2 2 3 3 4 4 5 5 6 7 0==7 0--6 7 0--6 7 0--6 7 0--6 7 0==7 0=-=7 3 3 2 5 4 4 8 6 10 5 7 1 1 1 2 2 1 2 2 3 2 3 A-11 (Note (Note 2) (Note 3) CPU TIMING INSTRUCTION TABLE D-3 DESTINATION ADDRESS TIMES: DESTINATION DESTINATION MICROCODE MODE REGISTER CYCLES 0 0 1 1 2 2 3 3 4 0--6 7 0--6 7 0--6 7 0--6 7 0--6 4 7 5 5 6 7 D-4 DESTINATION ADDRESS DESTINATION DESTINATION REGISTER TIMES: READ CYCLES READ WRITE 0 1 0 1 0 1 1 1 0 0 0 1 1 1 1 1 1 1 1 1 1 2 1 2 1 1 1 1 2 6 2 6 4 3 3 7 5 9 4 6 MODE - MEMORY 0 5 0--6 7 - 0-=7 0=-=7 TABLE WRITE-ONLY MODIFY | MICROCODE WRITE MEMORY CYCLES CYCLES Read WRITE 0 0--6 0 0 0 0 1 7 0--6 5 3 1 1 0 1 2 2 7 0--6 7 7 3 7 2 1 2 1 1 3 3 4 0--6 7 0--6 5 4 4 2 2 1 1 1 1 4 5 5 7 0--6 7 8 6 10 2 2 3 1 1 1 6 7 0-=7 0-=7 5 7 2 3 1 1 T P S WD TED XD NS WD WHD D GER CED SED AER CED CER CED SR CED WD R R AN A-12 CED WS D CED €0 RO G CED WD Gl «ED MO CID mED W GEP 1 eI GED W) 5 CMD CED GED GED GNP (Note 2) (Note 3) YER GER TED GED Gmb ES b TABLE DESTINATION ADDRESS TIMES: JMP DESTINATION DESTINATION MICROCODE MEMORY MODE REGISTER CYCLES READ - D D-5 e CEn G D D - - - e - G D TED N TR IR G CEN whm O WD R CES G A T CWD - D D D D D D D G D TD IR CID D CE WD D D D O D D O T O COID D D TIMING INSTRUCTION CPJ CYCLES WRITE D SRR WS CED A CED GND SR GED AED TG D CED CHD GED CUR € (TN G50 I CED LED GO GO W R0 S GO D CER <WD A D OB CED D 4D D CED D CED GED D D O O A D O €D XD 0 OO0 D CEp M S O CED €5 €ED €00 WD WED D D G S WD WP GED GED GED MOD I RS GNP OND G D D eI W GID eED C GID D ofD WD G0 P a5 en G D OwD s e DESTINATION DESTINATION MICROCODE MEMORY MODE REGISTER CYCLES READ WRITE 1 1 0==7 9 2 2 0-=7 10 2 1 3 3 4 0==6 7 0==7 10 ) 10 3 3 2 1 1 1 1 5 0=-=7 11 3 6 0--6 10 3 1 6 7 9 3 1 7 0-=7 12 4 1 D D N D S wED D D R XD D D D D D D D FLOATING O MICROCODE MEMORY MODE REGISTER O G D e e R D D D D D D O Wi G i) ST WD D D ) D D D D D SOURCE ) - - D D O D 1--7 e S D D D AT D CER ) ) ) Cm D D G D TYD S D e MEMORY €D OIS CYCLES GED D - Gn G TR D G W D F=-1 w TABLE D CYCLES A-13 G READ . D O S R S5 WRITE e SR D 0 G D aTD e CPU INSTRUCTION TIMING TABLE F-1 7 0--6 7 0--7 0--7 0--7 2 3 3 4 5 6 7 0--7 1 4 3 4 5 4 6 1 3 3 -2 3 3 4 0 0 0. o 0 0 0 4 4 1 5 5 4 5 5 6 0 0 0 0 0 0 0 0 DOUBLE PRECISION MICROCODE MODE 1 2 2 3 3 4 5 6 7 5 5 0 (Note 15) 6 5 6 7 6 8 0-=7 0--6 7 - 0=-=6 7 0--7 0==-7 0-=7 0==7 1 2 2 3 3 4 5 6 7 : MEMORY REGISTER 0--7 0--6 7 0--6 7 0==7 0-=7 0-=7 0-=7 MEMORY CYCLES READ 0 WRITE 0 0 0 1 1 0 1 1 2 2 2 1 2 2 2 2 2 2 0 5 0 5 (-1) .(Note 15) O 1 6 4 4 1 4 3 3 1 4 3 4 5 4 6 DOUBLE PRECISION 1 2 2 3 0-=7 0--6 7 0--6 A-14 INSTRUCTION TIMING CPU TABLE F-2(Cont) MICROCODE MODE MEMORY REGISTER MEMORY (Note MICROCODE MODE 1 2 2 3 3 4 5 6 7 WRITE READ CYCLES 15) MEMORY MEMORY CYCLES READ WRITE 0==7 0--6 7 0--6 7 0==7 O0==7 Q0==7 0==7 9 9 (-2) 10 9 10 11 10 12 4 4 1 5 5 4 5 5 6 4 4 1 4 4 4 4 4 4 REGISTER (Note 15) CPU INSTRUCTION TIMING TABLE D D RS G S G S F-4 — INTERGER - W D S = SOURCE € GmS N En . A MODES G N D G G G 1--7 WD R G S Y —— - - — - MICROCODE’ MEMORY MEMORY MODE REGISTER CYCLES READ WRITE 1 0 1 1 0 0 INTERGER 1 0-=7 2 2 2 0--6 7 2 0 3 3 0--6 7 3 2 4 5 6 7 2 2 0==7 0=-=7 0=--7 0==7 0 0 3 4 3 5 1 2 2 3 0 0 .0 0 LONG (Note 15) INTERGER 1 0==7 4 2 2 2 0 0--6 7 4 2 0 3 0--6 3 0 5 0 4 5 7 0==7 0=-=7 1 3 4 5 6 6 7 0--7 0--7 5 (Note 15) 3 2 3 3 4 7 MICROCODE MEMORY MEMORY MODE REGISTE CYCLES 1 0-=7 2 2 0--6 2 2 2 3 0 READ 0 0 0 0 0 WRITE INTERGER 7 0--6 A-16 0 0 0 1 1 1 1 1 INSTRUCTION TIMING CPU TABLE F-5 (Cont) 3 4 5 6 7 LONG 7 0-=7 0--7 0==7 0--7 2 3 4 3 5 1 0 1 1 2 1 1 1 1 1 0-=7 0--6 7 0--6 7 0--7 0=-=7 0-=7 0==7 4 4 2 5 4 5 6 5 7 0 0 0 1 1 0 1 1 2 2 2 1 2 2 2 2 2 2 INTERGER 1 2 2 3 3 4 5 6 7 A.2 SOURCE AND DESTINATION TABLE NOTES 1. Subtract 2 MC and one read if both source if or PC, autodecrement modes READ-MODIFY-WRITE mode 07 or 17 1is used. and destination WRITE-ONLY or 47 mode destination READ-MODIFY-~WRITE READ-ONLY and For ns. operatio READ 3 perform actually references bookkeeping purposes, one of the READs is accounted for 1in the EXECUTE, FETCH TIMING. 57 mode destination READ-MODIFY-WRITE READ-ONLY and For 4 READ operations. perform actually references bookkeeping purposes, one of the READs is accounted for in the EXECUTE, FETCHING TIMING. Subtract 1 MC if the link register is PC. Add 1 MC if the source operand is negative. Subtract 1 MC if the source mode is not zero. a. Add 1 MC if the quotient is even. b. Add 2 MC if overflow occurs. CPU INSTRUCTION TIMING c. Add a Add 5 MC and 1 read if the PC is used as destination register, but only if source mode 47 or 57 is not used. 1 MC per shift. Add 1 MC if source operand <15:6> is not zero. 10. Subtract 1 MC if one shift only. 11. Add 4 MC and 1 read if the PC 1is used as a destination register, but only if source mode 47 or 57 is not used. 12, pDivide by zero executes in 5 MC (see note 6). 13. Timing for no shift. Add 1 MC if a left shift. (Notes 8, 9, 11 apply.) Add 2 MC for a right shift. (Notes 8, 10, 11 apply.) 14, Add one MC if a register other than R7 is used. 150 Mode 27 references only access single word excution time 1listed has been accurately compute the total execution time. A-18 operands. compensated The 1in order to APPENDIX PDP-11/84 B HARDWARE/SOFTWARE DIFFERENCES PDP-11/84 B.l HARDWARE/SOFTWARE UNIBUS POWER UP DIFFERENCES PROTOCOL DIFFERENCES The Unibus Power up protocol on different than on most PDP-lls. With most PDP-11 systems, the PDP-11/84 systems (See Figure B-l.) Unibus signal 1is INTIALIZE held asserted for a minimum of 10 milliseconds of DC LINE LOW (DC LO L) on power up. after On is PDP-11/84 for systems, a minimum of power up. This the Unibus signal INIT after the 16 microseconds difference wil not affect L slightly (INIT the )L is negation held asserted negation of DCLO L on any system operations. / DC Power ok | | | DCLO L / | INIT L XXXXXXXXXXXXX\ | / l | | | 5 us min. Most XXXXXXXX FIGURE B.2 PDP-11/84 - PDP-11/84-based = B-~1 Cache 0 Switch Data POWER 11/44 it Register Register general o SPL, MTPS, Y---- PROTOCOL TIMING UP may not the contain PDP-11/44 the in following certain PDP-11/44 (17777754) contain register MFPS, DIFFERENCES replace does DIFFERENCES (17777570). PDP-11/84-based products present in the 11/44: Dual 16 us min. 10 ms min. HARDWARE products o | | | undefined applitions. However, hardware features: o = = FDP-11/84 UNIBUS PDF-11ls | )--- the following functionality set TSTSET, WRTLCK B-2 instructions. not PDP-11/84 HARDWARE/SOFTWARE DIFFERENCES implemented on the PDP-11/44. used for testing by the CPU ROM The following registers are not the registers are Primarily, ROM code, other diagnostics or system configuration by the CPU software. level They are not normally used by the system code. o Boot and Diagnostic controller register (17777520) o ROM page control register (17777522) o Configuration and Display register (17777524) o Diagnostic controller register o Diagnostic Data register o Memory configuration registér (17777734) (17777730) (17777732) and any DMA transfers may differences between the DMA transfers may not occur between UNIBUS registers located on the CPU. DMA peripehrals transfers may only occur between the UNIBUS peripherals and the UBA. occur between UNIBUS periph erals and the addresses of the also ROM sockets located on the UBA (17 773 000~ 17 773 776). Table B-1 summarizes the hardware PDP-11/44 and the PDP-11/84-based products. TABLE B-1 11/84-11/44 DIFFERENCES DIFFERENCES FUNCTION ADDRESS Added register set select 17777776 PS 17777772 PIRQ 17777766 CPU Error 17777754 Cache Data Not implemented 17777752 Hit/Miss No difference 17777750 Maintenance Hardware differences 17777746 Cache Control Hardware differences 17777744 Memory Error Hardware differences bit<11> | No difference Unibus monitoring bits not | implemented (see Chapter 3) (see Chapter 3) (see Chapter 3) PDP-11/84 HARDWARE/SOFTWARE TABLE — e — n-—-—— 17777676 DIFFERENCES B-1 -——-—-—-———— (Cont) ————-—— User Data PAR to - e d -—-————'————-— -—-—-——-—.4-—— No difference 17777760 17777656 to User 17777640 PAR Instruction "No didfference No difference No difference 17777636 to User Data to User Instruction 17777600 PDR 17777576 MMR2 No difference 17777574 MMR1 No difference 17777572 MMRO PDR 17777620 17777616 Eliminated maintenance mode 17777570 Switch 17772516 MMR3 Register Not implemented No difference No difference 17772376 to Kernel Data to Kernel Instruction No difference 17772340 PAR Kernel Data No difference Kernel Instruction No difference PAR 17772360 17772356 17772336 to PDR 17772320 17772316 to - PDP-11/84 TABLE B-1 HARDWARE/SOFTWARE DIFFERENCES (Cont) 17772300 17772276 to Supervisor Data PAR No difference Supervisor No difference Supervisor Data PDR No difference Supervisor No difference 17772260 17772256 to 17772240 17772236 to 17772220 17772216 to 17772200 PDP-11/84 - B.3 11/70 HARDWARE DIFFERENCES applications. The PDP-11/84 may replace the PDP-11/70 in certain hardware PDP-11/70 following the <contain not does it However, features: o Stack Limit Register (17777774) o Micro Break o System o System Size Registers o Physical Error Address Registers o Switch Register ID Register Register (17777770) (17777764) (17777760, 17777762) (17777740, (17777570). The DCJll-based products contain the following present in the 17777742) functionality not 11/70: o MTPS, MFPS, o Bypass cache MFPT, bit CSM, in TSTSET, WRTLCK instruction PDRs. The following registers are not implemented c¢nt the PDP-11/44. for testing by the CPU ROM wused are registers the Primarily, code, code. other diagnostics or system configuration by the CPU ROM They are not normally used by the system level software. B-5 PDP-11/84 DMA HARDWARE/SOFTWARE o Boot and Diagnostic o ROM page o Configuration o Diagnostic controller ¢ Diagnostic Data o Memory control controller register and not register register register register (17777520) (17777522) Display configuration transfers may DIFFERENCES (17777524) (17777730) (17777732) register occur between (17777734) UNIBUS peripehrals and any registers located on the CPU. DMA transfers may only occur between the UNIBUS peripherals and the UBA. DMA transfers may also occur between UNIBUS peripherals and the addresses of the ROM sockets located ont the UBA (17 773 000 - 17 773 776). Table B-2 summarizes the hardware differences PDP-11/70 and PDP-11/84-based products. TABLE ADDRESS 17777776 B-2 11/84 - 11/70 DIFFERENCES FUNCTION PS between DIFFERENCES Added suspended instruction bit Limit Not <8>. 17777774 Stack 17777772 PIRQ 17777770 Micro 17777766 CPU 17777764 System ID Not implemented. 17777760 System Size Not implemented. 17777752 Hit/Miss No 17777750 Maintenance Hardware No Break Error implemented. difference. Net No implemented. difference. difference. (see 17777746 Cache Control Hardware (see differences Chapter 3) differences Chapter 3) the PDP-11/84 TABLE B-2 DIFFERENCES (Cont) Hardware differences (see Chapter 3) Error 17777744 Memory 17777742 High 17777740 Low Error User Data User Instruction User Data User Instruction Error HARDWARE/SOFTWARE Not implemented. Address Not implemented. PAR No difference. No difference. Address 17777676 to 17777660 17777656 to PAR 17777640 17777636 to Added bypass cache, eliminated access flags and access modes other than 0, 2, and 6. PDR 17777620 17777616 to PDR 17777600 Added bypass cache, eliminated access flags and access modes other than 0, 2, and 6. 17777576 MMR2 No difference. 17777574 MMR1 No difference. 17777572 MMRO Eliminated traps, maintenarnce mode, and instruction complete. 17777570 Switch 17772516 MMR3 Not Register Added 17772376 to implemented. Kernel Data PAR Kernel Instruction CSM enable No difference. No difference. 17772360 17772356 to 17772340 PAR bit <3>. PDP-11/84 HARDWARE/SOFTWARE DIFFERENCES TABLE 17772336 to 17772320 Kernel B-2 (Cont) : . Added bypass cache, Data PDR eliminated access flag and access modes 2, 17772316 Kernel to 17772300 Instruction PDR and Supervisor Data .PAR 0, eliminated access flag and 2, 17772276 than Added bypass cache, access to other 6. and modes other than 0, 6. No difference. 17772260 17772256 to Supervisor 17772240 PAR 17772236 Instruction Supervisor Data PDR to 17772220 No difference. Added bypass cache, access modes other eliminated access flag and 2, and than O, 6. 17772216 to Supervisor Instruction PDR 17772200 Added bypass cache, eliminated access flag and access 2, B.4 SOFTWARE modes 6. other than 0, DIFFERENCES Table B-3 summarizes language level) PDP-11 and family. the programming differences between the DCJ1ll and other (at the assembly processors in the PROCESSORS ITEM 1. OPR %R, (R) +; OPR %R, — (R) using the same register as both source and destination: contents of R are incremented (decremenied) by 2 before being used as the source operand. OPR %R, {R) +; OPR %R, — (R) using the same register as bolh register and destination: initial contents of R are used as the source operand. 2. OPR %R, @ (R) +; OPR %R, @ — (R) using the same register as both source and destination: contents of R are incre- mented (decremented) by 2 before being used as the source operand. OPR %R, @ (R) +; OPR %R, @ — (R) using the same register as both source and destination: initial contents of R are used as the source operand. 3.0PRPC, X (R); OPRPC, @ X (R); OPR PC, @ A; OPR PC, A: location A will conain the PC of OPR + 4. OPRPC, X (R); OPRPC, @ X (R), OPR PC A OPRADC @ A: location A will contain the PC of OPR + 2. 4. JMP (R) + or JSR reg, (R) +: contents of R are incremented by 2, then used as the new PC address. JMP (R) + or JSR reg, (R) +: initial contents of B are used as the new PC. 23/24 44 04 34 LSIN 05/10 15/20 35/40 45 70 60 J-11 T-11 VAX ITEM 23/24 44 04 34 LSI1 05/10 15/20 35/40 45 J-11 70 5. JMP %R or JSR reg, %R traps to 10 {Megal instruction). JMP %R or JSR reg, %R traps to 4 (ilegal instruction). 6. SWAB does not change V. SWAB clears V. 7. Register addresses (177700-177717) are valid program addresses when used by CPU. Register addresses (177700-177717) time oul when used as a program Register addresses (177700-177717) time out when used as an address by SOB, MARK, RTT, SXT instructions® ASH, ASHC, DIV, MUL, XOR Floating Point Instructions in base machine. MFPT Instruction. The external option KE11-A provides MUL, DIV, SHIFT operation in the same data format. * AT T instruction is available in 11/04 but is ditferent than other implementations. ' Register addresses (177700-177717) are handied as regular memory addresses in the 170 page. 2 All but MARK. x 8. Basic instructions noted in PDP-11 processor handbook. > CPU or console. > 0T-d address by the CPU. Can be addressed under console operation. T-11 VAX ITEM 23/24 44 04 34 LS4 05/10 15/20 35/40 45 70 60 J-11 The KE11-E (Expansion Instruction Set) provides the instructions MUL, DIV, ASH, and ASHC. These new instructions are 11/45 compatible. The KE11-F (Floating Instruction Set) adds unique stack ordered oriented point instructions: FADD, FSUB, FMUL, FDIV. SPL Instruction T1-d CSM Instruction 9. Power fail during RESET instruction is not recognized until after the instruction is finished (70 milliseconds). RESET instruction consists of 70 millisecond pause with INIT occurring during first 20 milliseconds. Power fail immediately ends the RESET instruction and traps it an INIT is in progress. A minimum INIT of 1 microsecond occurs if instruction aborted. PDP11-04/34/44 are similar with no minimum INIT time. Power fail acts the same as 11/45 (22 milliseconds with about 300 nanoseconds minimum). Power fail during RESET fetch is fatal with no power down sequence. > MFP, MTP instructions > The KEV-11 adds EIS/FIS instructions T-11 VAX ITEM 23/24 44 04 34 LSI11 05/10 15/20 35/40 45 70 J-11 T-1 VAX RESET instruction consists of 10 microseconds of INIT folowed by a 90 microsecond pause. Reset instruction consists of a minimum 8.4 microseconds folowed by a minimum 100 nanosecond pause. Power fail not recognized until the instruction completes. 10. No RTT instructlion HATT sets the “T" bit, the “T" bit trap ¢T-d occurs alter the instruction following RTT. 11. It RTI sets “T" bit, “T" bit trap is acknowledged alter instruction following RTI. It RTI sets “T" bit, “T" bit trap is acknowledged immediately following RTI. 12. If an interrupl occurs during an instruction that has the “T" bit sel, the “T" bit trap is acknowledged before the interrupt. If an interrupt occurs during an instruction and the “T" bit is set, the interrupt is acknowledged betore “T" bit trap. 13. “T" bit trap will sequence out of WAIT instruction. “T" bit trap will not sequence out of WAIT instruction. Waits until an interrupt. 'interrupts nol visible to VAX compatibility mode. NA! 23/24 ITEM 44 04 34 LS 05/10 15720 35/40 45 70 60 J-11 -1 VAX X 14. Explicit reference (direct access) lo PS can load “T” bit. Consoie can also load “T" bit. Only implicit references (RT!, RTT, traps and interrupts) can load “T" bil. Console cannot load “T" bit. -2 Odd address/non-existent references using the stack pointer cause a fatal trap. On bus error in trap service, new stack created at 0/2. 16. The first instruclion in an interrupt routine will not be executed if another interrupt occurs at a higher priorily level than assumed by the first interrupt. Qinnla annaral nuurnn ictar implemented. Dual general purpose register set implemented. } Odd address/non-existent references using SP do not trap. 20dd address aboris 1o native mode. » 17 H 9 The first interrupt in an interrupt service is guaranteed to be executed. = £ET-4d 15. Odd address/non-existent references using the SP cause a HALT. This is a case of double bus error with the second error occurring in the trap servicing the tirst error. Odd address trap not implemented in LSI-11, 11/23 or 11/24. ITEM 18. PSW address, 177776, not imple- mented: must use instructions MTPS (move to PS) and MFPS (move from PS). PSW address implemented, MTPS and MFPS not implemented. PSW address and MTPS and MFPS implemented. 19. Only one interrupt level (BR4) exists. Four interrupt levels exist. F1-d 20. Stack overflow not implemented. Some soit of stack overflow implemented. 21. Odd address trap not implemented. Odd address trap implemented. 22. FMUL and FDIV instructions implicity use R6 (one push and pop); hence R6 : must be set up correcily. not do ons FMUL and FDIV instructi implicitly use R6. 23. Due to their execution time, EIS instructions can abort because of a device interrupt. EIS instructions do not abort because of a device interrupt. 24. Due to their execution time, FIS instructions can abort because of a device interrupt. 3 Can relerence PSW only lrom native mode. 23/24 44 04 34 LSit1 05/10 15/20 35740 45 70 60 J-1 -1 VAX -3 ITEM 23/24 44 04 34 LS 05/10 16/20 35/40 45 70 60 J-1 T-1 25. Due to their execution time, FP11 instructions can abort because of a device interrupt*® FP1t instructions do not abort because of a device interrupt. 26. EIS instructions do a DATIP and DATO bus sequence when letching source operand. E1S instructions do a DATI bus sequence when felching source operand. 217. MOV instruction does just a DATO S1-4 bus sequence for the last memory cycle. MOV instruction does a DATIP and DATO bus sequence for the last memory cycle. -2 28. If PC contains non-existent memory and a bus error occurs, PC will have been incremented. i PC contains non-existent memory address and a bus error occurs, PC will -3 be unchanged. 29. It register contains non-existent -3 memory address in mode 2 and a bus efror occurs, register will ba incramented. Same as above but register is unchanged. X X * Integral Hloating point assumed on 11/23 and 11/24; FP11E assumed for 11/60. X N implementation dependent. 2 MOV instruction does a DATI and a DATO bus sequence for lasl memory cycle. 3Does not suppont bus errors. VAX ITEM 30. if register contains an odd value in. mode 2 and a bus errer occurs, registaer will be incremented. It register contains an odd value in mode 2 and a bus error occurs, register will be unchanged. 31. Condition codes restored to original values after FIS interrupt abort (EIS doesn't abort on 35/40). Condition codes that are restored after EIS/FIS interrupt abort are indelerminate. 32. Opcodes 075040 through 075377 91-d unconditionally trap to 10 as reserved opcodes. If KEV-11 option is present, opcodes 75040 through 07533 periorm a memory read using the register specified by the low order 3 bits as a pointer. If the register contents are a non-existent address, a trap to 4 occuwrs. If the register contents are an existent address, a trap {o 10 occurs. 33. Opcodes 210 thru 217 trap 10 10 as reserved instructions. Opcodes 210 thru 217 are used as a maintenance instruction. 3Does not support bus errors. 4 Unprediclable. ! Traps to native mode. 23/24 44 04 34 LSIN 05/10 16/20 35/40 45 70 J-1 T-11 -3 VAX 23/24 iTEM 44 04 34 LSIN 05/10 15/20 35/40 45 70 J-1 T-11 VAX 34. Opcodes 75040 thwu 75777 trap to 10 as reserved instructions. it KEV-11 options is present, opcodes 75040 thru 75577 can be used as escapes to user microcode. It no user microcode exists, a trap to 10 occurs. 35. Opcodes 170000 thru 177777 trap to 10 as reserved instructions. Opcodes 170000 thru 1777717 are implemented as floating point instructions. Opcodaes 170000 thru 177777 can be used as escapes 1o user microcode. I no user microcode exists, a trap to 10 Opcode 076600 used for maintenance. +1 36. CLR and SXT do just a DATO sequence for the last bus cycie. -2 CLR and SXT do DATIP-DATO sequence for the last bus cycle. 37. MEM MGT maintenance mode MMRO bl 8 is implemented. . MEM MGT mialtonancs mods MMNO bR 8 is not implemented. 38. PS<15:12>, non-kernel mode, nonkernei stack pointer and MTPx and MFPx instructions exist even when MEM MGT is not configured. ~1 Traps (o nalive mode. + Unprediclable. 2CLA and SXT do DATI-DATO. hLY LT—-d OCCurs. ITEM PS<15:12>, non-kerne!l mode, nonkernel stack pointer, and MTPx and MFPx instructions exist only when MEM MGT is configured. 39. Current mode PS bits <15:14> set to 01 or 10 will cause a MEM MGT trap upon any memory reference. Current mode PS bits <15:14> set 10 10 will be treated as kernel mode (00) and not cause a MEM MGT trap. Current mode PS bits <15:14> set to 10 will cause a MEM MGT trap upon any 81-¢€ memory reference. 40. MTPS in user mode will cause MEM MGT trap if PS address 177776 not mapped. It mapped, PS <7:5> and <3:0> affected. MTPS in non-user mode will not cause MEM MGT trap and will only affect PS <3:0> regardiess of whether PS address 177776 is mapped. 41. MFPS in user mode will cause MEM MGT if PS address 177776 not mapped. It mapped, PS <7.0> are accessed. MTPS in user mode will not trap regardless of whether PS address 177776 is mapped. 1 Unprediclable. 2CLR and SXT do DATI-DATO. 23/24 44 04 34 LSit 05/10 156/20 35/40 45 70 J-11 T-1 VAX ITEM 23/24 44 04 34 LS 05/10 16/20 35/40 45 70 60 J-1 T-11 VAX 42. Programs cannol execute out of internal processor registers. Programs can execute out of internal processor registers. 43. A HALT instruction in user or supervisor mode will trap thru location 4. A HALT instruction in user or supervisor mode will trap thru location 10. -2 44. PDR bit <0> implemented. < 0> not implemented. PDR bit- 61-4 45. PDR bit <7> (any access) implemented. PDR bit <7> (any access) nol implemented. 46. Full PAR <15:0> implemented. Only PAR <11:0> implemented. 47. MMR0O< 12> —trap-memory management—implemented. < 12> not implemented. MMRO0 48. MMR3<2:0> —-D space enable— implemented. MMR3 < 2:0> not implemented. 49. MMR3<5:4> —IOMAP, 22-bit mapping enabied—implemented. MMR3<5:4> not implemented. X VHALT pusi\es PC 8 PSW lo stack, loads PS with 340 and PC with < powerup address> + 40. 2Yraps 10 native mode. 23/24 ITEM 50. MMR3 < 3> —CSM enable— implemented. MMR3 < 3> not implemented. 51. MMR2 tracks instruction fetches and interrupl -vectors. MMR2 tracks only instruction fetches. 52. MFPx %6, MTPx when PS<13:12> = 10 gives unpredictable resulls. MTPx %6, MTPx %6 when PS<13:12> = 10 uses user stack pointer. 55. The ASH 0c-d source left the 31 instruction with a operand of octal 37 times) will decimal register instead of to be left. shifted (shift cause right S6. The ASHC instruction with an octal value of 37 (shift left 31 decimal times) in source operand bite 5:0 and bits 1516 of operand being zero, non the will the register to be shifted instead of left. cause right 44 04 34 LS 05/10 15/20 35/40 45 70 J-11 T-1 VAX APPENDIX C BACKPLANE TEST CONNECTORS BACKPLANE TEST CONNECTORS NEW CONN PIN # SIGNAL Jl8-34 J18-33 Jlg-32 Jlg-31 J18-30 J18-29 Jlg-28 Jl8-27 Jl8-26 J18-25 Jl8-24 J18-23 DCLO ACLO BR7 L BR6 L BRS L BR4 L PA L PB L MSYN L SSYN L ClL CoO L Al7 L Jlg-21 Alé L J18-19 Jlg-18 J18=17 Jlg-16 J18=15 Jl8-14 Al4 Al3 Al2 All AlO AO9 L L L L L L Jig-12 Jig-11 Jlg-10 Jlg-9 Jl8-8 A07 A06 AOS5 AO4 AO3 L L L L L J18-6 J18-5 J18-4 Jlg-3 Jl8-2 Jlg-1 Jl9-20 J19-19 J19-18 J19-17 J19-16 J19-15 Jl9-14 J19-13 J19-12 J19-11 Jl9-10 J19-9 J19-8 J19-7 J19-6 J19-5 J19-4 J19-3 A0l L AOO L BBSY L NPR L GND GND D15 L D14 L D13 L Di2 L D1l L D10 L D09 L D08 L DO7 L D06 L DO5S L DO4 L DO3 L D02 L DOl L DOO L INTR L SACK L J19-1 GND J1l8-22 J18+20 J1l8-13 J1l8-7 Jl19-2 AlS L AO8 L A02 GND L APPENDIX BACKPLANE D PIN ASSIGNMENTS BACKPLANE PIN ASSIGNMENTS O = 2 1 2 1 2 TM 1Y GND Y ARG v ASV TM ASV |ASSYN | asv ABG SV ra GND| ASEL | GND LTC 01s A OUT | BR? AY? L Low TM D14 ASEL | BRE TM 013 ASEL | BRS 0 L D11 012 AN B8R4 1 2 NPG <V NPG mu- . ¢ (1IN} OUT) L ° - € f " ! X L L t L AINT | D10 L) s t D09 L A INT | DOB ENSB L T 00? T LS D04 , HALT | DOS ¢ " " s T v v 14 E 1DE A ROW ROW * ROW ROW L 1 3 q L L A IN M A12 ouTY N GND sSYyN | GND A5 sesy | FO L L L N mMSYN | A6 FOY 002 AC2 (o] 005 D06 A0 AD0 00? A INT | SSYn | CO NPR GND L L 4 L L L v2 L L L L enes L L out 2 A OUTY | BG? t are L A1l L 008 8G? an 003 ouT t TM aINT | 8GC AIN AOUT | INTR | #O° AINT | 8G6 A OUT | ADE t ASEL | AaBR ~NT L so Ensa | SO L ow t HIGH L L L t $01 ~1 A AINT s 0" L2 M2 00 t a out T 8GS L $0 L HaLY | ool TM 8G5 AD® TM 8G4 ASEL | astL | FO1 ] GND | DO3 GND 8Ga GNOD ASEL | GND 15/+8 | DO2 TM ABGC ADG AOL AT | ABR ABG A0S a03 aint | ron L0 L REO CRY L rs [T L ¢ L L AC Lo L ASSYN {m n ouTt $0 out Y out a0 L 3 L L 1)) A0?7 ABR L out el ASEL | FO: 1) L2 . o ? L L M2 sOn "2 SACH t a ou? insa| sy e 8% SPC Backplanc Pin Assignments L7 BACKPLANE PIN ASSIGNMENTS MODIFIED UNIBUS PiN DESIGNATIONS STANDARD UNISUS PIN DESIGNATIONS 4 A s ¢ ° € ; H s « L v ~ . " g ! 2] 1 INIT {*S5V |BGé “ L INTR{GND 18GS o L D00 IGND {BR5 L L D02 DOV |GND L L L L D06 |DOS DE |AC L L Lot DOt D07 a0t L L L 2 1 2 "N a INIT | BV GND s INTR}| prYY: GNO | ‘ BRa L DO4 |DO3 |GND . a e A AROW now /RON ROW 8Ge o ¢ R ] I . pC Lot A00 L N L TP L 000 | GND 8 002 | DO L L DO+ | DO3 L L 006 | D05 L . 006 | DO? L L L L D10 | D09 | AD3 | AD2 D10 |DO0% (AD) AQ2 J D12 D11 jADS AD4 « D12P | D1V | ADS Dwa |D13 ja0? L v Jo PA D15 (AD9 L L L GND [PB (AT AD6 L A0S L A0 D1a| D13| AO7 | AO6 L L]t L - tA 015 A0S | AD6 L t L L N ean]| P8 | A1) AYO L L t L L L L L L L . L L | oL L L L L) Lt L L o0 A L L eAR |B8SY| A12 | AV2 L GND {88SY| A1) A2 L L L Ala L A6 § L L cH GND [SACK] A1S L L GND [nPr | A2 GND |BR7 |GND L v NPG |BRE ISSYN v 8G? »W | L [GND L L Al4 L A6 L (3] L L co co t L < (CORE [MSYN° | GND NOTE: D INDICATES A REDESIGNATED PIN Figure Standard and Modified Backplane Pin Assignments D-3 ~&8wt.-P~eNY]w!O«®HN!mTJD-Zu&urn*Pya8 2M0oST-eto~t*.Mo3-u._EEREN]m~5 _182Sane)QLI..FIln.lw?“im-.rQ.eun|M-.?c_n—9Mc.inh,.PF5---”L1H.w aa&W2[&2)41Wo(%08]o"w&T\t1h-oFL-PTv.a-eht*04]-*®a=>¢=T.d.28Rd[‘e|yI2LH+Lb.m=1i+].MsR>e£eiOlL"Po@-z-gt]?—“lL[k°ve9®tPt=.-®-hL-R'I@_‘n2|—TaL1x@e--a'za]gT®o9l®LP]=x9?.z3=<[YR'vv—E+~le|%ae|msPBm2a"uT2o-zlR's°!8?®oL]=o®Xez[e®-na4rL"_F«=J1t“e¥G2a3Rei34wle-oAt.a8—Y&3o5SAS“0t2]nI6YRY!.ea1I|n%]I1T1710-M8LIT1AAT\A;TlRT8L-8rmT=3M]]1-$lo-1m[liLj]mw-N¢3=e-P€-"1,l.et—~|’y3o*"=P§.J(”f&AmMlP2mOSaugnmiM"ndM"|iPRe.-1M==[*4 :eL[po-PRo—-dr+tOAL|i*iLne1Ml~(ad:lIe-|~Faelg=-rPe—0by)=+Ic*=e-~Pydo-IIyT:>_a.(1&"»&822“-2wWgwWoou0OAGnelltiYomT6¥y1«8T0uT 8”21~-po=Mm-m—-k3steJln¥1 ..f|l-woul2(8a]y2Swdn:3]" [Rae]M-a®TM.TT— S{:5Sl[a®I 1a[Tt .-Tty ,umn-ud.o-a.0hifl m_ -v '(,&2LQYg2d" [ AoTot L=’ .L*% a=2® ITonweA0S14 s o3 o 3 -iTa=lzd[Ttcola—s]- [|e-=ylde] alg> -Ete]- oY)tAY 31o Ls =-*§ =.T R.m—-T«=i-t+.+ M1b o.fi -.C == ——t *Aia)2k4TIRawwoAYoo8snL1018~o1 2-M =——v¢i= | ys a3r02Oong" .o* ! 2% at. 2.~}— 1¢wB- AT b e_ T 82eMyi- *o '. '. '. " N6&2ase8 1 o] cS®k .Cal*l P[c 1=yl .coe?l T| 3vw[OAyS ] L. Bee e e el £ : ASIGNMENTSPINBACKPLANE z Y4ol ) " =* L Lk L TM pre -* =k 2 2 e 2 = 10 e & = = ., JoNe A p°—— 1 —] - < 3 ROW D A¥n COW 4 |t|wn-coTulfl]gJm_lSyLI u_wz-w..a-z-m..Y .aP'«° .[«°. .[°t[ .®«P[ .ot .o? .ot e¥ ».ae2— 43 —fo 21r-¢3|'feM— F-4) 0i [ } . . .2 .' .P .2 L4 ]| -]gL.e]T-k][S"832[T«[§l~}M.M(A~¢QB].i-La.[tl]fot.¥o8Xh"olau€tPud...[c|)(Ae]LT*.-e]-rt3[-loC«"o8On[lh~d*€]nd[.l|1Mo(Ar'Gl]L[-e»]Le}t][t.-"o"«-!P4h]k*NbSa_d ° @ - efes QO e o £ e © f o L] w [} & w s a0 o , o nana [.-}'lI. .'‘. . "«R® -.°otf€.o¥«t2*Lao]o["htdo«"'*i-«O”.[-.2u".°»>.[ &NAg=]- - Lo-[=]1 C.‘(rt 43LxSl 2—(M18p—1-f¢4-e93) ei.Pl-.1. |n“MwB5a3it.-A¢e &-o«MbO¢t291|0Q—(P3BIO@—e-p.94—0Q -.oul!l-3JoNn!lr0t9.—a[asS-e—olar—un2me,d9tco B-8lpuPy ESTHPIN slpuRy POgl[+ovpteLo m(‘flh?rr- — Y 1ALk 1AeYx . I 00 ROU C -~ 1 7g83 (B2 - 1l — | - 1 | - k3 2 ] 1o 1 - 1l =T 308Vr)-E0 Oo—ef—| PI}(3O-140~Oo—t—|2IO13-90O[t| M21(gi3G6]0bO——e £ eaM2l- 1-61I-09RANY L ]203M2 ALY <0884L9i-eG ROU B a2B 00 1AR ~&2a2A2e8g1808e T11 ~A 1w qY 3 -6l 4L ROU A ROW D VA-W6 3 - 3NVIdAOVd T A ROU B ROW C ROU NId W D 4 te = ay = » 2 O vg (vaifii‘}l.Cn(l-{JltfluFfiSI.u.q E‘l‘...LD(INJIlHLCI'.SIUV.I 1HH ‘um'fii1."i1"Nl‘l1ufl1s'.“fl RN ril-..cnll~< g [‘:‘MLCk.(.'M1.‘I‘."~-..1vfl‘l ele i .| lf;l11cu(irfl1-i]\1dfllilud .::: ° Ffi“UGSESESRRRSEARAR R OE LR <Afl1flusluv.J T BITHHIE a2 ERaRREsTNadNnil 1 (1T 11T ] 1T :r_:j:;_' - i o i | HHHHH IR "“fl“{*“'—] 5?1',%13"1 HHARRARIRE] ] : A i ..J..P.‘.,.:.._,.J-J_...Jf - 2 s o :.u 7508 BaaRSRERCRTIIC T 4 T ANER p ik eibiiitiiteessti € L L L EReR - L3 S T s 1_4..'_. IJJJJJJII—J::::::: :;.n: B8K & eReubindaaoicoiiil oo ERe B Etétbbbbécebbbbiese BE c.l:;} =] - . ole ] Et}fl:fl( 1149 1195 o [ 1 r Kk & I I l I [EEE(E[e]eee eg T HHERH } ] e 1l 1 S e el | S 1 i 222 .lnc 3 EEE:'E1'\II ?Ei—‘ -i—-:.'lr.:.{r:.u']!¥ E11qac } Ci2 ] F;l1(11610‘-1IJ!]4L]'{I~4{!4IuT]v’ sy pr croes v iMWf)E!I.Cfl(lflita'1»4"&':'!11} E '...co]'I'I“,iI.lLI'lI“-{...{..u";l ee==tiil fl~1flqgluv.] L .'II.' o0 l '." ; EEEEEEEE - - .r“.'” B ] | | B Hult,{-,'l-fll-ftm e i 3 7 . Lo L& k A ; <3 ?"; i i $3sso5: i EEEGERE BE ,” = .g—?g HH ek SINSWNOISSV ROU M P 28im-¢QoE—[t202=90on=l-.lo—ea3Tr t9[ otye 2 ] .-8822819u¢:.71mtnwmBiLt"]- oPy W=YnLs]gyo= I5T<] : >* G > «L2> = — -] . 24 o = |oLx=tT fi ; J £ P = £ Y 1 901G ~82 I- = Ik et F = S|2= E et|W =R1= ,CeT=wacmtupe L4 z LY 04 o = £ -t S TLINE il A4 | jas} a ”~~ 3INVIdIOVE i - W 9w ‘?T:r qio € f*&'i ?m.fl wid A 9 "iFJ ‘yq 1 q [ k (3 1 4 In fl E u I» & E &|.v] it Y I T 1T zxe || | (2322 ||| || “!$'ui if111t]flh (h1]lfll1f 11311 ‘11 ¢ €« ! * ¢ E T 3 = o ie E 2343 R 3 F k [ 4] Blllgl | lg HE 01 i1z | |ll HEREE. a : : c sl g e .ih'q CRCRVRLRRRL R il s ] e 17 TS SO &3 EINEESITT SETIELLY SHE g & bhgkebibRELE cookestoos phekt | ;p; lles IR X e :: ;5 ';111ch°(ruJ’x;!rnt1pasrluv] ! ___J L AL EEE| || |RER|R b R GRS ele asa|||laaala Uled 2 el | B A Fre pox |3 11 452 :ch h q 0 % GECRIELE® rla -'l S rPJT:] ' ¢y ::‘: nElE 158s|2 | aoo|a I53)) o6a oo ~e kex| || [e2R|R l'll1hdntr1J-tmh]«qnlhltqv [ ” _ g 1 il | IS‘l.i ‘qulfll"P ‘Hfllcn{( ! 14111 2 [} )24 el | ] |9v9% |3 l‘1 h o q ¥| 1 1 q R '.F Alals fr Iy | 31 ] #2| (PCULRL 18] | (2 . ¢y R . : £le gl |k k £ Fip 134 [d g % gt 1 I } ? 4 ? ? } P{ 1Tt 1 > 32 2 v - ce g ketich Rt P o i L % Siop BRUBED E ks 3 B R NId | 2871 O1R-mt 0 TM SINIWNDISSY ROUW F ROW E , APPENDIX E SYSTEM INTERCONNECT DIAGRAM B Xrmyion Rg b v mriniragiies l LR S 1904 . e ou-ta haonse: Lo 740V CnE - 1AL (7620106 01} C-& ASSV el I LD pwa cTAL S L S?D‘t)b W3 Lf-52 Ay €ORD SET) (ac Lf-P2 Hl e - | - jorr0 izovlzull [+ ] | Cnz on t (RED)+ 59 VOB — COMM REG 2= 3 T "'"‘I 8P | ' wveerst-e) e e —— ' IBLUE) ~ 15y 22 . cm:.( L3V 10RN O 13 0RAN) > 1219 | | - | == f‘ P . L 2 | & ki —r1%voc g < s RECEPIACLES) | A} k= inesv Lroam3] <je> s e < < n-s4 117806 11~ 041 { yL s d sus fim — 3 2 . e =1 s Pwh ROST L 12 uo’ossmwli :: :; &1 v UNSW - P > " 1120006 19) 120V i 0K (AC CORD SET) C:tz b - ot eo”"roooretbt I 1702039 261 | | q o8-P22 [_r 3 ' I - 8P P4 s cam ) 10 POWER ¢ 1AL urell SwKHED RELCEPIAC, E5) (1000288 3ri(DIC PWR cr_mluo& BUS (ABLE) | | (17008658 7)120V —> - (EaTehne fr . LOADS) (100083 12 J2e0V R {aC coRD SE1) \-l K | 1w (1221374 @3 - o _ DUPLER £ 10 THASSY. 248v(1221574- 94) H1SOVDC L TTERY BACKUP UNIT - 15evD( 1 | | = ps-pa) o ne -1 B M 3 e e B AGST RIn—F<] 9 Prt OVERADL RIN —1 <« 1 Pr OVERAIE —1- <] T -1y - —1.o fan SaFf £NA { a8V ENA L‘F R :i I <z T T 1reer30-01 yL rt ss-p2s | 10 s — \ IJF 7 | (g |I Y~ | | | ( 88 P20 l 10020609 81} B Lo X, N BBU OPTION lNSlALLiD__ CONTROL SIGNALS {10 B8-23) | ) ' . LN - il | T - _J T — 1 y 21 T &1 NOD LA |L —}-<] __Q pPwaInI@l L T T 8 88U RAS! [>1—— T e8-M 2 3 q soy aval, min—ff f T 2¢ BA1 SIS RIn—L <] 2 12 '8 r -1 20 g; ov vsv 1 88yToyavai —f<] At — ¢ _ e- t——« b a stares ___ — se-pa F'W 2;::‘(;:: o o o oo o T T —Y———Y(— 0P t (DEC PWR CONTROL nm?!ln 1) (ou 150V OUTPUTI | TM (10 4 ! | o 20 20 N3 SUNDBY L (1700083 ) 240V Jpc e ) INPUTIOUTPUT CONTAOL SiGNALS 2 10 AIR FART | GND . . @ L] FAL PULLLP AIRSTanDey DESEL <asETY 128 ) ' [ | 3 42 S Pwi sreaBsT o 1 3 V. T (10 P3-p3) B . ! g 8BU ENA 3 s 4 REv Pwh iniBll L : 1) o412 1 <E $001 Ena auUs RN SIAND BY L 2::‘:? (10 Ps-p2) > v« FANDRYRI7T WL ATV il e S16 RN 126 U e o Py ' 2 Aco 1 2t H [} W7eect2-011 3 23 AUl SEwse ¢+ —1 > LYC UNSW P2 [ Aln [§ |} 21.,2 3¢ s AETUAN FOR FANS ] AN MOVER fl‘_;. : N ! Nesscore—— 10866901} -1 — I| BLR - P2 : 4 l"'h’ Iyt u nee T {10 PS- PG} > 432169 gorrs 200¢h2s) | 30/68n2 | tuswiicnen L) V2 ] e i sy s I 5o | guu S. AN —T—} : T 654321 (SWICHED necEPILES] TMM | loLn) ¢12.3 AIN ——— (1217830-02) 1790003 - 34) 120V i se Pwn MODULE - 11760083 35) 240V (¢ +3v RETURN / ps FILIER ‘55,77 4L 977 POWER CONTROL 4o\, tv outPuT 110 PE-PR) P (o et prgR N o 15-30°) ‘ (uiERMAML NEMA LS- ~Jnas | n )6 F ‘ijg I (nEwa 6-13P) ToowAm poran n7202kA POWLR SUPPLY LFOR uA PWA SUPFLY) (K] . . ) Ay PONT 0N LrwTRAL_F r . {ALD WIRE 31700734 Py > 34 SVRIN . o4b8H (BLK WIRESHINOT ' 10 PWR SUP K -JyE— PS-PY 10 PWR SUP K- 16 - PS-PE —0 ; 1 . [j:-—+5v¢ _ 8P -7 1170075001 amy » 1L rS-P2 © PR W --u<—-—,E] 2 ke ' ; Nd-P3 ROTAION PUSE L1 12VDC AIR MOVER o) :nzm-- oo ROt —IQIRTRI gy 8P- P16 1> AL B-MU- H927T7 & DBY UTILIZATION:I W} FOR BACKPLANE MODUWLE e 2] a: +SVAIN INTEMOCK 1a8 () ' vEL) u: o reR e RIS 9 “eer2e(1i"eer201) sP P13 g §«:i 431 AN s B WP00726 - 94) sr-£13 m 8N 9 (AR MOVER INTERLOCH 148) I . foun) > (1700732i 91} - 10 h BNIERLOCK SWITCH) > /-t . x0P - P T nree12%-01}) (A MOVER 1234356709K] wrerr (wow) I RFRARR H I| Lvys | | ne E 3 & 1 LOIST Lve CABLENS] e —— .~ —— S aon ag e s § . . “':lsvi >venD, L car "A 3 . v 43v —o D) VBB —O \YELj ¢80 TP2TPY TPIGND —o (BLx ; K - s F 7 L~ se > <} UE e gl = [19 a3 ’:noeg e F: £ :‘33 - ¥ 3 2‘-33 Sg :u: 68 % T 777 111 [¥1%" f:.B janannsi " 20 E €XP PWR RQST | 2 E - ENE H ot on ot pa[ . = GND L ) ' 39 v Il re sart svarus - Consont e-: c.g:;c'xttto(u 3; 2——L L pe—pi el RGST QL L Stanoey k- Pwd 9e B AN & B ED L. Ty I‘_le s16 E—— LED P4 L 03 L e 1(0 t?10 jctpe—1—1EDO2 L 19 - LED 08 L 20 o1~ i - : 10 32 On RDAI-B (CPY) 171320’ cPU-42 123456 78900112131413% ¥ 2 [ e jlreenza -, b s - P2 YY)F (DTM »] [ (rozioes-o1) |! | [_;_ . nuo %0 443? [ FRONTPANEL ASSY 412381 vma(Dpna | P K< : 5 . | ; 4 s w1 s =4 [ POKAIOR LED'S = A5232 CONSOLE INTENIACE CABLE B 5 FANT PWR N2 pwR nh5 RE PWR SuP "wmi { | T T T2 T T[yhy 23456 70 919111213 1613%617 81928 : e < 9\ MOM- P | g T- nor. pust 4 SE € PN AOY. PR 4 12.3v l| - ne pe—— e et 1 2 v E e —— 32 b= : ' ‘__..r__ ' s Stv | swivos - E ‘!2 ke +3v (D L. -*F:_} SAY Te - P14 " otn —}mize BAUD RATE bt S i oo ° 3 .. L= ' l_.l |§l | ._L -—— < 2 e it ! x 2 s— I 4,50 ng oF.,2 roocane so-rns <14s 10 pc-pe (DEC POm CONTARGL BUS CABLT) MOM-PA [) oo3 la 1 nreernzy en e? ' =4 — 18 jee ] 28 e f— 1) 19 20 E l‘mm pE4+— \ i19 % fe " -z-—r > Y] wE (1> € < | 1?7 6 12) fe = 20 "9 e4 A g 2: ! 2 FORCED DLG it SwvCe (®])s2 s ? L "3 ox 7020650 91 W3277-4 OPTION) BACKPLANE ASSY 2 s ? e O 10 pum SuP - n-26 «— s " .(.'4& 2 : 3E ’> s>~ . " 0AN] ' S S 21% RIN G 15416058 81 com crum sol. . L ssaves 0w - ' TONSOLE DIST BOARD 3 A $9 e = ’-3: ;OSV H ‘P 1= 10 49 On ROJIH-B (CPU) = .-—:|_ H? - | s4 rs -, 418 21—- CH2 OX ] (1700724-83) ) ____ __ — — —_— | ExPmNDIn OPIION WSIALLLD PS | ' Mp wings O "8 P2 o— 'lrs re I "08 P : o————————0 1m0 82) ‘ (17007%1-01} £-3 r | i ———— ! ¢ { 4321 ""‘)L ' (s :. 130 E_A‘_' LK. i POTRNECT 10R%) (o Lo L -—Ii[_—'"— > , I ] MIn | ] | o ‘2‘ " ey “wh_sdev |RE S = ! SwR e voc T —— . —— — —— e t —— ———— L__ N )] i |1 Hoose-ca® -; . . T A\ mvecas | . = o 14 N 3 POy Ao nic ' o 4 1138 voc e o s— STamoBr | ng $1aN0O8. ¢ tnd $tANDRY - B L s — Bg - [T T T — o —— —_— ...' T s e e hreens o)) . ©DISY L CABLL) 118206010 € DAasSY. (41PN POMIONY Uit BECIPLL L) 11700003 244 1/0v (1 70000) D240V 1M (0MD 341) peiv W o ,E 3 ) b ; - b « | - * 1 . 52 " ; . —. 114!: s - 6 PO BUNDLE 4 » i tap vunu.lo!ll . (NP Cny On Atw geomin s ol " o3 > S sus ? N\ = ! ) . (yo0m o 3K 0 ') N LEHE ——— :.:1 - fi 166n"00 Pum subiLY) 10 POowtn (ONIAOL H o8 sk 508 —A b wll | 0y —_—— _l 10 oP-Pu § o +—— 9 (1reenc-o2) \ e S E ' e 1 | sEh t l sargty 1 - . 3 T [l 1] H LR} ® ': Ei__r_—1: el " Puiy 40013 P com S dele . i> Py o= . HER= ry " es a1 AW PaY ] e an | s A 6 T 3 ) 3 b 7 M s L¢ c-—l (3T ¢ sg # AND P2 ARG WD Tod Shast —~7 ‘__j |l i .—-——T--—-—l Srocs - 13 H ’ + '5;: =H ba F =1 > an tants brati l | u08- 3 | os AoNL ) ' : ey 3 MOS! l-amliy [ P } “OD Ens ¢ . —19° i l "20 mu —.p:"u 1938 [— . , 2 [ g 3 o M sun lsast | a nojuo- K g ALk fna BuS by’ 1 | —r———— 1 g L — ek o o 9001 3 S MUY Uy R g PR | TM1 =d i | Y $ =5 Pt DCw L sie 1 ! Sn0en ~\ s lc ] $Eqe [imeerer e ), s 3 On siset | ) (O Jlooreoe | bt 1] = t:l: | H Nl SEnSL ¢ — | ] &7 B a ] | m . !Q__._ 230N | ) | __I._» T IO Pt Sag T ; | 3 s | ' ! < ' | +H 2 | i L1} 170046690 | I. ——— e e "? 12| ! ok r— NOTE: J3AND 14 ANE 00 4 *sy 418y i i ANOT wsE OF 6se12 s atn SLs wints L‘ Turzn coun ate woound | mF-sg ni-r g 11706734 0cne} _________ : ¥y osv O————0 | rsuuuussvr—,)) 2 o — e MO [ (189234 850N II”H *3v PeA DIST (3616292) i 00y § i IsPANDER | nlle® . we | ni20s 58 POWIR SuPALY ] | 82 5 V.7 - V.6 APPENDIX F ROM DIFFERENCES COPE V.7 - V.6 ROM CODE F.1 INTRODUCTION The command DIFFERENCES descriptions for the ROM code assume that the EPROMs installed are at Version 7.0 (V7.0) Earlier PDP-11/84s contain EPROMS with V6.0 ROM code. The version of the ROM code is typed out each time Set up mode is entered from Dialog mode and 1is displayed in the upper right corner of the printout. It 1is not necessary to remove the CPU module to determine the ROM code version number. The following lists the ROM part numbers and version numbers. Socket on CPU El116 E117 F.2 V6.0 BOOT ROM Location (M8190) (Low (High SUPPORT code Part Byte) Byte) FOR TAPE does not Number V7.0 Part 23-116E5-00 contain DEVICES (TUS81) a in built tape MSCP the ROM TU81 tape drive. The TU81 could be booted if tape MSCP boot was installed in the UBA board or the EEPROM. V7.0 ROM The device F.3 V7.0 In V7.0 code has name DISABLE in a is built tape MSCP bootstrap for a M9312 written type into boot program for the has been added allows the TUS8l. MU. SETUP Setup mode parameters in V6.0 23-077E5-00 23-078E5-00 23-117E5-00 MSCP Number command MODE PARAMETER another parameter (2). This command user to to the disable entry into Setup mode if force Dialog mode is not selected. This command was added to prevent unauthorized entry into Setup mode. This <change assumes that the force Dialog mode switch 1is controlled or that switch 5 on the KDJ11l-B CPU is "ON" to prevent unauthorized When all access Setup mode references to will cause V6.0 Setup mode F.4 V7.0 an is Setup disabled the invalid DISABLE In V7.0 in Parameters to can ALL mode. and Setup the ROM command are command always be TESTING response entered code in Dialog mode eliminated. Typing Setup from from is the Dialog ROM code. In mode. PARAMETER Setup mode Command 2. another parameter has been added to When set, this parameter disables all memory and cache test- ing if force Dialog 1is not set. Force Dialog causes all testing to be run. F-2 F.5 B V.6 - V.7 ROM CODE DIFFERENCES FOR ROM BOOTS MNEMONIC In V7.0 under the B mnemonic for ROM boots the address located at on the Unibus must be an even type device M9312 a on 173024 In data. address of the check only This is the address only. odd. be could but greater or 165000 be must address V6.0 the F.6 EDIT/CREATE COMMAND In V7.0 in the edit/create command boots, the highest unit number entry is now decimal. an user had to type in was which EEPROM In V6.0 the converted to value. decimal F.7 number octal for mode Setup in TRAP TO LOCATION 4 on either In V6.0 - for UNIBUS systems only - if a ROM is found in the found not is mnemonic the and module, M9312 the UBA or the will code ROM the s, description device for table Up Look to location 4 due to an error in the V6.0 ROM trap unexpectedly This problem only occurs if the List command is executed. code. This problem does not affect the Boot command. The ROM can still be booted even if there is command. list 1in the corrected in problem a The system does not hang and the user may reenter This problem has been Dialog mode by typing <CR>. V700. F.8 DISK For V6.0 MSCP BOOT AUTO ROUTINE the boot (device name A), in the MSCP auto boot will Only drives attached try fixed media units from 0 to 7. program then it will try to boot removable media from units 0 to 7, to the at the standard disk MSCP address (172150) are trieaq. controller The MSCP auto boot does not support unit numbers above 7, and the will hang if the controller has a unit number greater boot auto than 7 that responds. the boot program will For V7.0 in the MSCP auto boot, remov- able media units media from 0 from to 255. try to boot wunits 0 to 255, then it will try fixed For unit, each the Dboot program present Dbefore to boot using the standard disk MSCP address, if this attempts fails the boot program will attempt the same urit number from the first floating continuing to the disk MSCP next unit device if it 1is number. address at be would The first floating controller if present, The main advantage no devices from 160010 to 160330. if 160334 F-3 V.7 - ROM V.6 CODE DIFFERENCES of V7.0 is that it allows a user to add a second disk MSCP device making any entries into the translation table as long as without the controller address is set according to the floating CSR address F.9 rules. DISK MSCP When trying BOOT DIFFERENCES to boot a DU device wusing the Dialog mode Boot the V7.0 code will automatically try the first floating command, controller also if the standard controller reports an error of type as long as the standard controller exists (no timeout). any If an error occurs on both controllers the V7.0 ROM code will error messages for each controller starting with the out print standard address first. Nonexistent error messages are not typed out unless the unit 1is If the second controller does non- existent on both controllers. not exist at the proper floating address the ROM code will print out only messages associated with the standard controller. If the translation table is used, or the /A switch is wused then only one controller is tried regardless of the existance of two or more controllers. COMMAND F.10 INITIALIZE V7.0 has been changed PMG value count for the v7.0 ROM F.11l MEMORY such that the initialize value to 7 as opposed to 0 PMG count is 7 for all command for V6.0. sets the The recommended KDJ11-B's with both V6.0 and code. TESTING In V7.0 code all consecutive memory starting from location 0 1is written at least once at power up unless all testing has been disabled. In V6.0 memory above 248 KB may not be written if the long memory test is disabled or CTRL C is typed. F.l12 POWER UP OR RESTART MODE SET TO 3 For V6.0, the ROM code checks for the presence of Unibus and sets up the KMCR accordingly. Before emulating recovery trap through location 24, the ROM code: 1. Reads and saves the contents of location 24, a memory power V.7 - V.6 ROM CODE DIFFERENCES 2. Executes a quick read/write test on location 24, and 3. Restores the orginal contents of location 24. the When the test is successfully completed the ROM code loads ion locat the and jumps to contents of location 26 into the PSW ion 24 1is tested, ROM locat Since 24. ion locat specified in memory cannot be present in the lower portion of memory. For V7.0, the ROM code does not check for Unibus memory and assumes that when mode 24 was selected that the system had the final configuration of memory already installed. Location 24 1is not tested, and it is possible to have ROM in the lower portion of memory. The ROM code loads the con- tents of location 26 into the PSW and jumps to the location specified in location 24. F.13 POWER UP SET TO 3 WITH BATTERY BACKUP For v6.0 if: is 3 at power up, 1. The selected mode 2. The battery indicates that the voltages are lost, and 3, The Ignore Battery function is not set, fthen 4. Go to Dialog mode regardless of the restart mode selection. For V7.0 if: is 3 at power up, 1. The selected mode 2. The battery indicates that the voltages are lost, and 3. The Ignore Battery function is not set, then 4. Execute the Restart mode selection if it F.14 otherwise go to Dialog mode. is not mode 3, ENABLING HALT ON BREAK V6.0 ROM code will not enable the Halt on Break bit in the BCSR until either one break has been received and discarded, any valid character has been received except XON, or the ROM code has given up control of the CPU. This was done to allow the ROM code to ignore the break that often comes from certain terminals when they are powered up. V.7 - V.6 In V7.0 ROM the CODE Halt "Testing 1in Since halt on environment, In either keylock F.15 CTRL on this bit 1is set Please wait generally enabled feature the switch R Break progress break is case, the DIFFERENCES AND Halt is CTRL in U was on not Break the needed bit SECURE immediately after the message" is printed out. only in a single-user and does has not been have removed. any effect if position. ECHOING In V6.0 during keyboard input CTRL R and CTRL U are echoed as R U. 1In V7.0 these inputs are not echoed. The functions still work the same as before. These functions are not echoed because the wup arror "" <character 1is not always available on all terminals. and F.16 In SETUP V7.0, was COMMAND Command originally console when in Setup included terminal the 5 5 wuser to be changed mode to allow has deleted. different automatically from been English character selected text to by Special characters used in other representations fallback Setup mode command is F.1l7 In V7.0, the BOOT SEQUENCE text, since ASCII in the ROM code or local all text characters 1langauges are represented in standard ASCII. In V7.0 the description for Command typed, it is ignored. AUTOMATIC command sets the local text to English. The command is not required printed on the screen uses only the standard which are generally available on all terminals. using The 5 is: Reserved. If by in the MESSAGE ROM code will print out a message indicating when boot sequence 1is started when auto boot mode is message indicates that all tests are complete, and the ROM code is starting the auto boot sequence. In V7.0 the ROM code will print out the name and unit number of the device booted after the starting system message. the automatic selected. This The following V6.0 example: examples illustrate the V7.0 and V6.0 messages. V.7 Testing Memory in progress - Please size 1is 512 K Bytes 9 Step memory Step 1 2 Starting V7.0 V.6 ROM CODE DIFFERENCES wait test 345 6 7 89 system example: Testing in Memory size 9 Step memory Step 1 2 progress 1s 3 512 6 automatic Starting system BOOT single - Please K Bytes 7 8 wait test 45 Starting F.18 A - COMMAND 9 boot from LIST DUO ADDITION letter mnemonic (L) has been added to the list in V7.0. L causes the automatic Loot continously loop until one of the selected successfully booted. Normally, the last device in boot command sequence to devices is the auto boot table is followed with the mnemonic E which terminates the table. If none of the previous devices where bootable the ROM code will print out an error message and request input before proceeding. If L at follows the last device the ROM code will restart the beginning and continously try every device in the the until one is booted or the operator types CTRL C to table table abort the sequence. V6.0 ROMs do not contain this feature. However, it can be implemented by writting a small EEPROM booct to emulate the feature. The feature is useful for fault tolerant booting for a system The must following loaded V6.0 that into roms. is continously the source the EEPROM that This program is try code until and allows not a successful description the needed loop boot of function for Vv7.0. the to occurs. program work for V.7 - V.6 ROM CODE DIFFERENCES .=10000 iProgram is relocatable ;address, START: tstb @#177560 bpl 108 movb bic @#177562,r5 #177600,r5 cmp any characters been typed yNo-Go exit back to auto- boot ;Yes-Check the character ;Get the character from the RBUF ;Clear off all bits above bit 07 ;1Is 208 the character ;1 Yes-Then ;r5> set sboot 10¢%: another sHas r5,#3 beqg to mov #301,r5 movb #100,@#177611 return to sLoad r5 will CTRL C ROM code with will cause the which sequence ;This sand 3 a to to be aborted. with value make fake it ? for drive out the restart the ROM error code auto boot ;Sequence 20S: bic #760,Q#177520 jmp @#165762 ;Make sure :the BCSR sReturn to ;If r5 ; boot sabort ;mode. the ROMs are selected the ROM code. 301 then restart the auto sequence. If r5 is 3 then the sequence and go to Dialog is The following is an example showing the program being loaded into the EEPROM. It assumes there is not already a boot with a device name (mnemonic) of L. If there was, the user would have to delete or rename that boot before saving the new program. KDJ11-B Press Type Setup the a mode RETURN command key then for Help press the Edit/create an EEPROM Type CTRL to exit or press 1410 Bytes free in the EEPROM Z Device name Beginning address Last byte address Start address Highest Unit number RETURN key: 13 boot the RETURN key for New New New = L = 10000 = 10047 = AA = = 000600 000615 = 000600 = 3 F-8 in No New = 10000 New = 377 change V.7 - V.6 ROM CODE DIFFERENCES Device Description = EA ROM ODT> 010000/000000 105737 ROM ODT> 010002/000000 010004/000000 010006/000000 010010/000000 010012/000000 010014/000000 010016/000000 010020/000000 010022/000000 010024/000000 010026/000000 010030/000000 010032/000000 010034/000000 010036/000000 010040/000000 010042/000000 010044/000000 010046,/000000 010050/000000 177560 ROM OoDT> ROM oDT> ROM ODT> ROM oDT> ROM oDT> ROM oDT> ROM ODT> ROM OoDT> ROM onT> ROM oDT> ROM OoDT> ROM OoDT> ROM OoDT> ROM CoT> ROM oDT> ROM oDT> ROM OoDT> ROM oDT> ROM oDT> BOOT 100007 113705 177562 42705 177600 22705 3 1405 12705 301 112737 100 177611 42737 760 177520 . 137 165762 ~2Z (exit) KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: Save boot Are you into sure ? the 14 EEPROM 0=No, l=Yes Type a command then press the RETURN key: 1 Writing the EEPROM - Please wait KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: F.19 LOCAL LANGUAGE SUPPORT Local language translations are not supported in V6.0. supports local language translations. Only V7.0 "F.20 ADDITIONAL MAP COMMAND FEATURE The MAP command in V7.0 has an F-9 additional feature. It will v.7 - V.6 ROM CODE DIFFERENCES determine the speed of the J11 crystal. This is done by counting the number of SOB instructions that can be executed out of cache during one 20 ms cycle of the internal DLART clock. The value is compared against a table of standard values, and 1if it is within 0.1% of any of these then that value is printed out. If the value does not match, then the actual value is printed 15.206, 17, 18, 19 and 20. The standard values are: out . any errors have occurred during testing the speed will not calculated. If be APPENDIX MULTI G BOOT ROM CONTROL TRANSFER MULTI G.l CONTROL TRANSFER BOOT ROM INTRODUCTION In V6.0 and V7.0 of the ROM code, the routine which identifies boot identify correctly not will M9312 the or UBA the on ROMs ROMs which use more than one ROM -for the boot. Because of this, not list and can not be started from the base will boots these ROM code without using a small EEPROM boot program to transfer This problem will will be corrected in any to the ROMs. control future G.2 releases TRANSFER of the CONTROL CPU ROM code. PROGRAM The following is the simpified program used to transfer control to any UBA ROM or M9312 ROM boot if needed. The starting address in location 010020 may have to be changed depending on the ROM and the socket it is located in. bcsr = 177520 decsr = 177730 010000 010002 052737 000200 010004 177520 010006 010010 042737 000010 010012 177730 010014 010016 010020 000261 000137 173012 bis #200,@#bcsr bic #10,@#dcsr ; s : ; ;s H sec jmp @#173012 ' Note: The ROM is on M9312 042737 If to 052737. module then make sure UBA ROMs are enabled ; ; s+ ¢+ disable diagnostics go start boot for ROMs in sockets 1-3 of UBA. Change 16/173012 to 173212 ; 2—-4. ; Note: disable CPU ROMs in 173nnn address range 1f ROMs change are 010006 in address = 1730nn for socket 1 = 1732nn for socket 2 address = 1734nn for socket 3 address = 1736nn for socket 4 previous program DMCl11/DMR11, DUP11l, will work DUll numbers if the user types in description as follows when for and the DL11-E the the DECNET with the ROM sockets from If ROM is not in socket 1 then address in 10020 be adjusted by 200 for each socket as follows: address UBA must boots for following part correct device name and device program is being loaded into the G-2 MULTI BOOT EEPROM under Setup mode of the CPU ROM code. numbers start with 23ROM PART NUMBERS 86229, 865A9, 868A9, 926A9, 863A9, 866A9, 869A9, 927A9, 864A9 B867A9 870A9 928A9 (e.g., 23-86A9-00). DEVICE NAME XM XW XU XL ROM CONTROL TRANSFER Note that all part DEVICE DESCRIPTION DMC11/DMR11 DUP11 DU1l1l DL11-E The same general program could be used for any multi ROM boot or ROM boot which does not follow the M9312 ROM format single any The starting address may have to be adjusted to start standards. the boot. the of an example 1is following The G.3 EEPROM LOAD EXAMPLE DMC11/DMR11 DECNET the for EEPROM the into loaded being program boot ROMs. MULTI BOOT ROM CONTROL TRANSFER KDJ11-B Setup mode Press the RETURN key for Help Type a command then press the RETURN key: 13 an Edit/create free Bytes 1410 to exit or press Z Type CTRL EEPROM boot in address Beginning Last byte address Start address Highest Unit number ROM 010000/000000 010002/000000 010004/000000 010006/000000 010010/000000 010012/000000 010014/000000 010016/000000 010020/000000 010022/000000 oDT> ROM OoDT> ROM ODT> ROM ODT> ROM ODT> ROM ODT> ROM OoDT> ROM oDT> ROM oDT> ROM OoDT> KDJ11-B Press XM New 10000 000615 000600 3 New New New 10021 10000 15 New DMR11/DMC11 052737 000200 177520 042737 000010 177730 000261 000137 173012 ~2 Setup mode into Are you sure Writing the ? the change New the RETURN key for Help boot No 000600 Type a command then press the RETURN key: Save for AA Description Device RETURN key the EEPROM name Device the 14 EEPROM 0=No, l1=Yes Type a command then press the RETURN key: EEPROM - Please wailt 1 APPENDIX H CONFIGURATION REGISTER MODIFICATION CONFIGURATION REGISTER MODIFICATION H.l INTRODUCTION and . Diagnostic Figure H-1 shows the format of the Boot ons to-external connecti the and Configuration Register (BCR) without modified be to values switches which allow the register of the 15-8 bits Since module. CPU the to gaining accessregister are not driven they may be read as 1 or 0. In order for a register bit value to be remotely controled, the switch which controls that bit on the KDJ11-B module must be OFF Whian a to allow control to be passed to the external switch. grounded is line nding correspo the ON, is KDJ11-B module switch which causes the register bit to always be read as a 0 regardless of the position of any external switch. PDP-11/84 on Bits 7-4 <Switches 1-4> are not connected remotely The bits must be set at the KDJ11-B module switch pack systems. unless a custom cable is built and connected to an external switch by the user. Register bits 1 5 9 0 1 2 3 1 1 1 1 1 4 Register bits 7-4 <Switches 1-4> to J3 on the KDJ11=-B module. There is no connection to J3 on PDP-11/84 5 6 7 8 4 '3 2 1 0 also connect systems. By connecting J3 to s P external switches it would be possible to remotely control these four bits in custom applications. Register bit 3 <Switch 5> also connects to the Force Dialog mode switch on the rear panel by way of module. J2 Register bits also connect select on the <Switches 2-0 to switch on the the panel by way of J2 KDJ11-B module. X dDm m e KDJ11-B 6-8> BAUD rate rear on don't care FIGURE H-1 BOOT AND p e e e L DL bl DD Dl Sttt the = DIAGNOSTIC CONFIGURATION REGISTER CONFIGURATION REGISTER MODIFICATION The 8-position switch pack is mounted at the handle end of the 6-8 select the baud rate for the DLART chip. Switches KDJ11-B. read-only BCR at the Switches 1-8 correspond to bits 07-00 in address H.2 ROM 17777524. INTERPRETATION CODE The following paragraphs describe how the ROM code interprets the The ROM code only looks at BCR when it is read. the of values register bits 7:3 <switches 1-5>. NOTE a when that assumes discussion The following pulled is it to connected line switch is OFF, the If an external device is HIGH by the CPU module. grounding considered switch a 1line then the switch 1is ON. When using an device external to ocontrol the selections, any switch on the CPU that goes to an to external switch should be OFF to pass control the external device. Normally switches 1-5 are OFF and the EEPROM contents what action is to be taken at power up or restart. When off, Dialog Switch 5 unconditionally forces the ROM code at the completion of the selected tests. mode if cannot occur Switch 5 determine to enter Auto boot is OFF. and all 1is disabled the console If Switches 5 and 1 are ON, be cannot mode Dialog 1is suppressed. output to the console will program the console, the at occurs input any If entered. transmit an error message to the console indicating the console is disabled.- If Switch 5 is ON and Switches 2-4 are not equal to octal 0 or 7, then Auto boot mode is selected with the device and unit number to be booted determined by a table in the EEPROM and the value in switches 2-4 (1-6). system. PDP-11/84 There are default values in the table for the they user the of needs the meet not do devices default these If The selections can be can be changed in setup mode to any value. The or any EEPROM boots. ROM boots UBA boots, ROM standard CONFIGURATION selections are REGISTER MODIFICATION changed using setup Command 6. Refer to subsection 4.3.4 for more information on Setup Command Setup Command 4. information on for more Refer to subsection 4.3.6 6. Interpretation of the BCR by the X = Don't care 0 = ON 1 = OFF 12345¢6 738 76 54 3210 ROM code. Switch number on KDJ11-B module Register bit OCTAL ACTION POWER UP Console enabled. Unconditionally | X XXX1XXX TAKEN AT transfer control to dialog after running tests. (FORCE DIALOG) The Force Dialog switch - located at the rear of the box cabinet - drives bit 3 LOW (0) when Dialog switch is ON bit 3 is high (1) KDJ11-B module is For the following, baud rate. the as long as Switch 5 on the OFF. the console is enabled and switches Auto selections Switch Boot 1 boot 6-8 select mode is automatically selected for to Switch Boot 6 (SB n) from Setup mode 6. 1 1110XXX 11100XXX 11010XXX 11000XXX 10110ZXXX 10100XXX 1 0010XXX 1 0000XXX 36X 34X 32X 30X 26X 24X 22X 20X Dispatch according to EEPROM Auto boot 6th selection from Autp boot 5th selection from Auto boot 4th selection from Auto boot 3rd selection from Auto boot 2nd selection from Auto boot lst selection from Power up to ODT immediately. Setup Setup Setup Setup Setup Setup Command Command Command Command Command Command We N0 ) B0, Command or When the Force (o )% e )W the it is OFF. CONFIGURATION REGISTER MODIFICATION Dispatch according to EEPROM has four pocssible modes. Auto mode, Dialog Setup mode Command 4, boot Halt mode and or ODT enter (power fail recovery). location 24/26 For the following the console is disabled and Auto used. not boot mode They are trying 1 to 6 devices defined by through vector switches 6-8 are automatically selected for all 1is -selections. 2 3 7 6 5 6 45 4 3 = X ON 7 8 = Don't care Switch number on KDJ11-B module Reg ister 0 ACTION TAKEN AT POWER UP OCTAL 16X 14X 12X 10X 06X 04X 02X 00X 01110XXX 01100XXX 01010 XXX 01000XXX 00110 XXX 00100XXX 00010 X XX 0 00 0O0XXX bit boot Auto Auto boot Run stand according 6th Auto boot 5th Auto boot 4th Auto boot 3rd Auto boot 2nd Auto boot lst to Command 4 Command from Setup Command from Setup Command from Setup Command from Setup Command from Setup Command Setup selection selection selection selection selection selection from alone mode tests Setup in a loop at power up. The following list specifies the default values for SB 1 to SB if the EEPROM is initialized in setup mode. The user may 6 change the values to any they desire by using setup mode Command 6. SB 1 A MSCP Automatic SB SB SB 2 3 4 DLO MSO MUO SB SB 5 6 E RLO1, RLO2 TS11, TU8O TU8l1, TKS50 (Not set) (Not set) boot sequence Table H~-1 shows how to select the baud rate for the Console SLU Switches 6-8 on the select switch. Rate switch select KDJ11-B module must be OFF to allow the Baud rate when using the Baud to correctly select the Baud rate. DN 1 0 TN = OFF AN 1l CONFIGURATION REGISTER MODIFICATION TABLE H-1 Baud rate Selected 38400 19200 the Oor H-2 BCR position bits 2:0 0 1 000 0 01 9600 4800 @ @ mem==———em=mmm——- 2 3 010 011 2400 2 m=m—e—- 4 ————e——- 5 600 @ @ mmemm——— 6 100 101 110 300 @ mee—e———- 7 111 to select if the be Register —==——-—mmeee- how to SELECTION Switch KDJ11-B module has shows RATE @ 1200 Table BAUD the Baud removed due to TABLE H-2 KDJ11-B (1 baud rate rate select with failure. = SWITCH OFF 0 = SELECTION ON) KDJ11-B BAUD DATA READ MODULE RATE FROM BCR SWITCH 6 7 REGISTER 210 8 0O 0 O 38400 0060 0 0 0 1 1 O 19200 0 9600 01 010 0 1 1 4800 011 1 1 1 0 0 1 O 1 O 2400 1200 600 101 110 1 1 1 300 111 100 the switch is switches not on present APPENDIX FLOATING POINT 1 INSTRUCTION TIMING FLOATING POINT INSTRUCTION TIMING Since the FPJ1l is a coprocessor operating in parallel with the J11 chip set, the calculation of floating point instruction times for Jll1 systems (usinf the FPJ11l option) must take this parallel processing into account. TERM - DEFINITION FPJ1l1l cycle Two clock periods J1l nonstretched cycle Two FPJ1ll cycles J11l read cycle (110 ns @ (220 ns J11 nonstretched cycle Dependent on read @ 18 Mhz) 18 Mhz) if cache hit. access time of systemif cache miss the minimum is two J11 nonstretched cycles, after which the J1l1 streches in 1/2-cycle increments " J11 write cycle Instruction Decode until MCONT is asserted. Dependent on write access time of system(two J11 until MCONT) cycles + 1/2 cycles A decode/prefetch cycle followed by a MOV microinstruction that allows the FPJ1l to assert DMR prior to the start of the next microinstruction (INPR for REG. mode).This time equals two nonstretched cycles if the prefetch is a cache hit, otherwise nonstretched plus Address Calculation Time Transfer Time read cycle. J1ll time required to calculate the address of the operand. This time is dependent on the addressing mode of the instruction, the frequency of the system clock, and whether any indirect data required is present in the cache. (See Argument one Table I-1.) J11 time required to load or store floating point operands. This time is one nonstretched cycle (address relocation ucycle) plus one read cycle per l16-bit word read from memory for load class instructions, or one nonstrechted plus one write cycle per 1l6-bit word to memory for store class instructions. FLOATING POINT INSTRUCTION TIMING DEFINITIONS INPR (Cont) (featemp,TEMP) J11 support code microinstruction execute for all FPJ1l instructions. Moves the PC of the previous FPJll to a TEMP register in case instruction that instruction resulted in a floating point (FP) exception. If the FPJ1l is still executing the previous instruction when the Jll reaches it's INPR microinstruction the FPJll asserts STALL causing the J1l11 INPR ucycle to stretch. The J11 then WAITs for the FPJ1ll to deasset STALL signalling the SYSTEM INTERFACE to assert MCONT before ex- ecuting the next microinstruction (OUTR) . WAIT RESYNC J11 time waiting for the completion by the FPJ1l of the previous FP instruction. For Load Class or REG mode instructions, this time from when the J11 INPR cycle streches at the trailing edge of MALE until the FPJ1ll deasserts STALL. This time equals zero if a stall was not required or if the FPJ1ll deasserted the STALL signal after the INPR cycle began but prior to the trailing edge of MALE. Although the WAIT time for the latter case is zero, RESYNC time is required. For store class instructions the WAIT time equals the time between the assertion of SCTL (i.e., when the SYSTEM INTERFACE is ready to execute the first write cycle of a FP store) and the assertion of FPA-RDY (data ready) by the FPJll. For Load Class and REG mode instruc- tions the time required to continue a strechted INPR. This is the time for the SYSTEM INTERFACE to recognize the deassertion of STALL and assert MCONT, plus the time required for the J11 to synchronize MCONT and FLOATING DEFINITIONS POINT INSTRUCTION TIMING (Cont) advance to the next microinstruction. Store class instructions normally do not have RSYNC time since the -waiting in a stretched WRITE and the continuation time is WRITE c¢ycle. However, if the is executing a previous Jll cycle part FPJ1ll MODF/D or DIVD, the FPJ1l will to stretch a non-I/0 cycle prior to the first bus WRITE. This allows the SYSTEM assert INTERFACE limiting to the worst when waiting case a WAIT for and STALL in service case FPJ1l1l DMA time order DMA thus latency output. RESYNC In the stretched non-I/0 cycle added to the effective execution of the Store class instruction. OUTR (PC, FEATEMP), FPE Last J11 less there FP support is an instruction. this associated with TESTPLA is is time microinstruction FPE from Saves the address un- previous of PC in decode FP in- FEATEMP. PRDC SYNC Time required by FPJ1l to struction tion and begin execution after receiving PRDC. This time equals two or three FPJ1l cycles depending upon synchronization. PRDC SYNC is not added to FPJ1l instruction execution times when the FPJ1l is executing a previous FP in- struction Floating Point Execution Time Time at the required instruction arguments. by once For assertion FPJ1l it to has Store of PDRC. complete received class a FP all instructions floating point execution time includes the time from the start of the instruction until the FPJ1l asserts FPA-RDY indicating the first 16-bit word is available for output. (See Table I-2.) Effective Execution Time Load Total J11 time required to execute a FP instruction. class Instruction + Argument + OUTR Decode Transfer + Address + INPR + Calculation WAIT + RSYNC FLOATING POINT INSTRUCTION TIMING DEFINITIONS (Cont) Instruction Decode + INPR + WAIT + RSYNC REG mode + OUTR Instruction Decode + Address Calculation Store class + INPR + Argument Transfer + WAIT + OUTR Load class instructions require input data and deposit results to REG mode instructions are FP the destination FP accumulator. accumulator to FP accumulator. Execution of a Load class FP instruction by the FPJ1l parallel operation Jll with occurs in and can be overlapped.(See Figure I-1) Store class instructions can be overlapped by the Jll as the "FPJ1l will complete a previously started Load class or REG mode instruction. store the instruction and then continue to Execution of the store class instruction must be completed before thus eliminating further the result can be stored to memory, parallel processing for store class FP instructions.(See Figure . -2) TABLE I-1 ADDRESS CALCULATIONS TIMES MODE LOAD CLASS 0 1 2 3 4 5 0 3 3 3 + RD* 4 3 + RD 3 3 2 2 3 4 6 7 27 37 67 77 - D T D D D T S G = J11 0 3 2 2 + RD 4 3 + RD + RDI + RDI + RDI + RD D S T D N R D D G O D D D D A Istream request Read Cycle I-5 + RDI + RDI + RD 2 3 2 1 2 4 + RDI** + RDI + RD D I GED D D GED G * RDI = J11 STORE CLASS + RDI + RDI + RDI + RD G D ) D D D 5 X @b e FLOATING POINT INSTRUCTION TIMING TABLE I-2 FPJ11l (18MHz = INSTRUCTION TIMES lllns cycle) INSTRUCTION MIN CYCLES TYP CYCLES MAX CYCLES STRETCH CYCLES 18MHZ TYP(uS) ADDF/SUBF 7 9 19 5 1.0 ADDD/SUBD 7 9 30 5 1.0 MULF 15 15 16 11 1.7 MULD 26 26 27 22 2.9 DIVF 17 24 30 25 2.7 DIVD 33 48 62 57 5.4 MODF 28 34 43 15 3.7 MODD 39 45 71 26 5.0 CMPF /D 3 4 6 2 0.4 LDF /D 3 3 3 0 0.3 LDEXP 2 3 2 0 0.2 LDCIF/D 10 10 10 3 1.1 LDCLF /D 10 10 10 3 1.1 LDCFD 4 4 4 1 0.4 LDCDF 4 4 8 1 0.4 STF/D 3 3 3 o 0.3 STCFI 8 10 13 1 1.1 STCFL 8 12 16 1 1.3 STCFD 4 4 4 0 0.4 STCDF 6 6 6 1 0.7 STEXP 5 5 5 0 0.6 3 3 3 0 0.3 4 4 5 0 0.4 TSTF/D,LDFPS STFPS ,CFCC,SET ABSF/D,NEGF/D FLOATING POINT INSTRUCTION TIMING NOTE: Stretch cycles indicate the number of cycles out of max cycles that a data dependant stretch of one additional cycle could occur with probability less than 1% for each additional cycle. FLOATING POINT INSTRUCTION TIMING Jl1 FPJ11 Load class instruction is prefetched. This occurrs during previous instruction execution !PRDC SYNC instruction Sun !Prefetch next Su=m !Instruction Decode IAddress fase Qum - Sum 1 Calculation !Argument Transfer ! FPJ11l loads operands | ! ! ! _ if any if any f=n feum RSYNC s O WAIT {FPJ11 lexecution !starts Jun {INPR ! - Swo {OUTR faup ! instruction O=w next FPJ11l —-——- only stalls instruction The FPJ1l is can if next Gmn Gum !Decode FP and REG mode! overlap the loading of operands for subsequent load class instructions ! ! ! ! i § --FPJ11l execution unit done FIGURE I-1l J11/FPJ11 INTERACTION I-8 FOR LOAD CLASS INSTRUCTIONS FLOATING POINT INSTRUCTION TIMING FPJ1l1 J1l1 Store class instruction is prefetched. This occurrs during previous instruction execution tInstruction Decode iPrefetch next instruction . . IPRDC SYNC i ! ! i f Jom {Address Calculation OJems G IFPJ1]l starts lexecutioon gom i Qe ! FPJ1l places operands 1in output buffer sets FPA_RDY ! INPR ! - !Argument Transfer J11 waits during first write if FPA-RDY not asserted completes argument transfer O=o 1J11l i next instruction Oems 1Decode Bem> Ot — Geany; 1OUTR ! FIGURE I-2 J11/FPJ11 INTERACTION FOR STORE CLASS INSTRUCTIONS APPENDIX SET-UP J PARAMETER WORKSHEETS SET-UP PARAMETER WORKSHEETS PURPOSE the set-up initial system The purpose of the following worksheets is to record parameter selections contained in the EEPROM of the KDJ11l-BF CPU module.- The worksheets are ¢to be filled at out installation, or revised when parameter selections are changed. future This information will be wused for programming any The worksheets are to remain with the replagement CPU module. system for future reference. Remove the worksheets from the appendix required. and them fill out as Store the worksheets with other system docunentation. Use a pen to fill out the CURRENT columns and pencil for the columns. NOTE Use Set up Command 1 to exit set-up mode. Use Set-up Command 9 to copy the contents of 'set-up table in memory into the EEPROM. Command 9 should be executed are made. after any the changes NEW SET-UP PARAMETER WORKSHEETS 0=No, A ANSI B - Power up 0O=Dialog, (l)=Automatic, 2=0DT, 3=24 O=Dialog, (1l)=Automatic, 2=0DT, 3=24 Video terminal C - Restart D - Ignore E - PMG F - Disable G - Force H - Clock O=Power ECC 0-(7) 1l=.4us, clock ! - Disable long K - Disable ROM trap supply, 0=No, l=Yes 0=No, l=Yes 2=60Hz, 3=800Hz 0=No, l1=Yes 0=No, l1=Yes 2=Dis 173 4 3=Both 1=Yes mode 0=No, l1=Yes testing 0=No, 1=Yes 0=No, 1=Yes 0=No, 1=Yes 0=No, 1=Yes 0=No, l1=Yes Allow on N - Disable Setup O - Disable all P - Enable UNIBUS Q Disable UBA Halt alternate 18 165, 0=No, 0=No, - Enable 1=Dis 4=3,2,...7=25.6 1=Yes M - test l1=Yes 0=No, Enable UBA 1=50Hz, (1) memory - Enable 3=1.6, CSR test L R - 2=.8, interrupts — clock J S 0=No, battery Enable - (1) boot memory ROM cache bit (1) mode block test (1) SET-UP PARAMETER WORKSHEETS TT1 Device name Unit # CSR Address TT2 Device name Unit # CSR Address TT3 Device Unit # CSR name Address TT4 Device Unit CSR name # Address TTS Device name Unit # CSR Address TT6 Device name Unit # CSR Address TT7 Device name Unit # CSR Address TT8 Device Unit # name CSR Address TTO Device name Unit # CSR Address = = = SET-UP PARAMETER WORKSHEETS CURRENT Boot 1 Device name = Boot 2 Device name = Device name = Boot 4 Device name = Boot 5 Device name = Boot 6 Device name = Boot 3 NEW SET-UP PARAMETER WORKSHEETS (D ED EED GE) R TED GED GED SED R MR D SED CHD GED AID N S GED GED G D D G GIp IR G SN N GED IR CEN GER GED GND D U TED WED WWR GO0 WA N R GER WAD AEP N 43D GED WA NS VEP GED D GMD GND NS IS Switches 2, 3, 4 Switches 2, 3, 4 Switches 2, 3, 4 Switches 2, 3, 4 Switches 2, 3, 4 Switches 2, 3, 4 . G D - D IR G G D SED CED D A CEp GED GV D MND NG WD W G WD GES WED GRS WM b
Home
Privacy and Data
Site structure and layout ©2025 Majenko Technologies