CA1165450A

CA1165450A - Data processing system with self testing and configuration mapping capability

Info

Publication number: CA1165450A
Application number: CA000364346A
Authority: CA
Inventors: Subhash C. Varshney; James A. Ryan; David R. Bourgeois
Original assignee: Honeywell Information Systems Inc
Current assignee: Bull HN Information Systems Inc
Priority date: 1979-12-28
Filing date: 1980-11-10
Publication date: 1984-04-10
Also published as: GB8331380D0; GB8401531D0; JPS56101255A; AU6511680A; FR2472783A1; GB2066529A; GB2132394A; GB2066529B; US4334307A; GB2132394B; DE3048614A1

Abstract

ABSTRACT OF THE DISCLOSURE
A data processing system employing firmware for execut-ing a self test routine each time the system goes through the power-up cycle. The self test firmware provides for compilation of a system configuration map during each execution so that configuration and status data is made available for accessing by the system operation firmware and application software. This enables external systems to set appropriate interrupt vectors and levels and to arrange their physical I/O and device handlers so that various devices within the local system can be accessed. The routines performed in the self test operation include a CPU test, a RAM test, a real time clock test, a communication controller loop-back test, a ROM signature calculation, a controller I/O test, a system configuration map compilation, and a status display routine.

Description

i ~ 65~0 BACXGROUND OF T~IE INVENT~ON
This invention pertains to dàta processing systems and, more particularly, to a data processing system having a program-controlled self testing capability and incorporating ' means for generating a system configuration map each time the self test program is exercised.
It is increasingly common in the data processing industry to employ networks of processing units coupled together by communication links. The overall processing capability of the system is distributed among the individual units of the system. This development has been made pos- i j~sible by the advent of large scale integration circuit fabrication techniques, which have led to the r~oady avail-'ability of small, relatively inexpensive microprocessor ',circuit modules which may be conveniently employed at many different locations within the system.
In such distributed data processing systems, it becomes !
jincreasingly important, as the complexity and processing power of each of the individual units is increased, to provide means and methods for automatically verifying the operational functionality of the units. Furthermore, as each unit is designed to provide more flexible operations and therefore is configurable to operate in a variety of different modes, means and methods must be provided to ' indicate the particular configuration and status of each unit to the other units in the system and to provide for updating the configuration and status information each time a unit is placed in a different operating mode.

.~

l 1 65~50 OBJECTS ANI:~ SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide an improved data processing system and improved operating methods therefor.
; A further object is to provide an improved data proces-sing system and methods for operating such system wherein an , individual processing unit within a distributed processing `network has an automatic, pre-programmed self testing capa-bility.
Still a further object is to provide a system of the type described wherein the self testing capability is coupled with means and methods for indicating the operating status and configuration of the unit.
~' Still a further object is to provide a system of the ~type described wherein the status and configuration informa-tion generated for an individual unit is automatically ''updated at periodic intervals. i - , To achieve the foregoing objects and in accordance with ja first aspect of the invention, a data processing system i having processing means and memory means is provided with means for implementing the steps of operating a first test program to determine the functionality of a first section of , the memory means, inhibiting further operation of the system ~if the first memory section is determined to function erron-eously, operating a second test program to determine the functionality of a seco~d section of the memory means, and storing in the first memory section the results obtained during the operation of the second test program.

~ ~ ~i 5 ~ 5 ~) In accordance with an another aspect of the invention, a data processing system having processor means, random access memory ~RAM) means and read only memory ~ROM) means i~ provided with means for implementing the steps of sequen-tially reading stored data out of the storage locations in the ROM memory means, calculating a signature word based on the contents of the data read out from the ROM memory means, and storing the signature word in a predetermined location in the R~M memory means.
` In accordance with a further aspect of the invention, a ` data processing system having processor means, memory means i~ and a plurality of I/O controller means is provided havin~ i apparatus for implementing the steps of issuing an interrupt request signal to a selected one of the I/O controller means lS to test such means, receiving an interrupt response signal from the selected I/O controller means and comparing the ; signal against a predetermined interrupt response value, generating an interrupt status word, including an error flag bit, in response to the result of the comparison operation performed in the last-mentioned step, and storing the interrupt status word in a predetermined storage location of the memory means.
! The accompanying drawings which are incorporated in and constitute a part of the specification, illustrate a preferred embodiment of the invention and together with the description, serve to explain the principles of the invention.

~ 1 6s~lsn BRIEF DESCRIPTION OF THE DRAWINGS
IN THE DRAWINGS:
FIG. 1 is a schematic block diagram illustrating the present invention in a system for controlling an automatic bank teller terminal.
FIG. 2 is a schematic block diagram illustrating further details of the microprocessor controller (MPC) and interface connector of the system shown in Fig. 1.
` FIGS. 3a and 3b, placed side-by-side with Fig. 3a on the left, represent a schematic block diagram of the circuits within the MPC of Fig. 2.
FIGS. 4-15 are flowchart diagrams illustrating the microprogram firmware resident in the MPC for controlling the self test and configuration status functions of the Isystem-FIGS. 16-38 provide a microinstruction listing for the ,firmware shown in the flowchart diagrams of Figs. 4-15.

DETAILED DESCRIPTION ~F EMBODIMENT
Fig. 1 illustrates the present invention in the form of a system for controlling an automatic bank teller terminal.
The system includes a microprocessor controller (MPC) 100 which is coupled via an interface connector 150 to a plurality llof input/output ~I/O) controllers 200, 202, 204, 206, 208, an external memory module 210, and a safe unit 212. The latter includes a safe interface circuit 226, function switches 228, a cash dispenser unit 230, and a depository unit 232. The general organization and function of an exemplary system of this type, as well as the details of one 11~5'~0 type o~ safe unie 212 and cash dispenser and depository apparatus operable therewith, are described in copending Canadian application Serial No. 364,246 entitled "Automatic Cash Depository" by Ronald D. Guibord, Robert G. Yetman and Richard G. Harris, filed November 7, 1980; Serial No. 362,470, "Automatic Note Dispenser With Purge Control" by Ronald D. Guibord, Neil W. Harman and Richard E. Hennessy, filed October 15, 1980; and Serial No. 363,942, "Purge Control For Automatic Note Dispenser" by Ronald D. Guibord, Neil W. Harman and Richard E. Hennessy, filed November 4, 1980.
The system of Figure 1 can be interfaced to any intelligent trans-action controller via a communications link including modem 214, which is coupled to communication controller 200. Printer controller 202 interfaces the MPC to a journal printer 216 and a receipt printer 218 which operate to provide two forms of hardcopy output for the system.
Keyboard controller 204 interfaces the MPC to a remote keyboard unit 220 which may be located, for example, on the customer control panel provided with the teller terminal. A card reader controller 206 functions as an inter-face for a magnetic card reader 222 adapted, for example, to read the magnet-ically encoded data from a customer identification card inserted into the control panel. Controller 208 interfaces the MPC to a cathode ray tube (CRT) display unit 224 also provided on the customer control panel to provide visual instruction data to guide the customer in executing a transaction. An external memory module 210 is coupled to the MPC to provide a storage facility for the operating . - 6 -1 1 65~0 system firmware, as well as to provide additional general purpose data storage capacity. The safe unit 212 i5 coupled to the MPC via the serial communication port of the MPC and enables communication between the microprocessor and the safe unit via interface circuit 226 in accordance with a bit-serial communication link protocol.
Fig. 2 illustrates the ~C inputs and outputs as well as the interface connector 150 in greater detail. The connector employed in the illustrated embodiment includes ten circuit board connection slots. The first six slots, `
151, 152, 153, 154, 155, and 156 are utilized to accommodate the five controllers 200, 202, 204, 206, 208, and the external memory unit 210, respectively. Slots 7-10 are provided for system expansion and are not utilized in the present embodi-; ment. As shown in Fig. 2, an input/output controller (IOC) device selection bus 157 is provided to connect the MPC 100 with the ten slot connectors. Each of the ten lines 01XX-OAXX (designated in accordance with hexidecimal notation) is connected to a specified.contact terminal in each of the ten interface slots. When the MPC performs an I~O operation, one of the ten lines within bus 157 ici enabled to activate the particular I/O device selected. The I/O devices inserted in the interface slots have their top ten terminals connected in common such that the device will be enabled by an IOC
selection signal no matter what slot the device is inserted in.
An IOC interrupt bus 158 also connects the interface slots 1-10 to the MPC 100. Eleven interrupt lines designated ~Rll-IR14 and IR16-IR22 are included within the bus to ~ 1 6~'~50 enable each of the I/O devices to send one or two interrupt level response signals to the MPC. The interrupt response lines are connected to the same pair of contact terminals within each of the interface slots in the cascading fashion illustrated in Fig. 2. This permits the proper allocation of interrupt lines to the different I/O devices while still permitting the flexibility of inserting, with a minimum of restrictions, any I/O device into any interface slot. The restriction imposed by the illustrated interface arrangement is that if any I/O device requires two interrupt levels, the device inserted in the next higher number slot must require ! zero interrupt levels. With the ten slot arrangement illus-; trated, this restriction limits the system to the use of no more than five two-interrupt devices (inserted in slots 1, ~3, 5, 7, and 9, or 2, 4, 6, 8, and 10). Of course, inter-face connector 150 will accommodate up to ten single interrupt devices. As will be described hereinafter, the self test firmware employed with the system provides an interrupt con-tention check to make sure that the above-described interrupt restrictions are properly observed.
The connection between the MPC and the interface con-; nector 150 further includes an 8-line control bus 159, a 20-line address bus 160, and an 8-line data bus 161. The 36 lines included within these three buses are connected in parallel to the same 36 contact terminals within each of the connector slots. ~ single line 162 is connected to the MPC
serial communication port and, as indicated in Fig. 1, is dedicated to the safe unit 212.

5'~) Figs. 3a and 3b provide a circuit schematic showing the individual circuit components employed within MPC 100. For ease of understanding, Figs. 3a and 3b should be viewed to-gether with Fig. 3a on the left. The MPC includes a micro-processor 102 which may be, for example, a conventional model 8088 microprocessor manufactured by the Intel Corporation.
Microprocessor 102 includes an 8-line multiplexed input/output bus 103 for transmitting eight address signals AD0-AD7 and for receiving and transmitting 8-bit parallel data bytes to and from a transceiver circuit 108. The latter is coupled to an 8-line internal data bus 105 which-~is in turn connected to IOC transceiver 142. Data bus 105 allows the transfer of data between the microprocessor 102 and the interrupt controller 124 and memory modules 130, 132, 136, and 138 provided within the MPC as well as between the microprocessor and the I/O devices. Data is transferred to and from the I/O devices coupled to interface connector 150 via the internal data bus 105, transceiver 142, and the IOC data bus 161. Data is transferred to and from the microprocessor and the safe unit 212 via internal data bus 105, serial communication port 128, and the single line serial port bus 162.
An 8-bit address latch circuit 106 is also connected to the multiplex bus 103 and is used to latch the AD0-AD7 outputs from the microprocessor and hold them for providing address signals to the remainder of the system. A control signal A~E is provided as a strobe to latch the signals on bus 103 and to feed them, along with the twelve address outputs MA8-MA19 from the microprocessor, to an address `~ ~ 65'~0 , buffer circuit 114 via the 20-line address bus 113. The twenty outputs from buffer 114 are provided as address signals MAO-MAl9 on the 20-line address bus 111, which feeds the various memory and control circuits of the MPC as well as the I/O devices coupled to the interface connector 150.
The address signals are fed to the I/O devices via address buffer 144 and the IOC address bus 160.
Microprocessor 102 also issues control signals S0-S2 via the 3-line output bus 164. Control signals S0-S2, along with a clock signal generated by oscillator 104, are decoded by decode circuit 116 to provide nine control signals on the 9-line control bus 115. A first of these control signals is the advanced memory write command AMWC which is transferred via buffer circuit 120 and IOC control bus 159 to the external memory module 210 where it is employed to strobe data write operations therein. Memory read command signal ~DRDC is employed within the MPC to strobe memory read operations in`
the RAM modules 130 and 132 and in the PROM modules 136-138.
MRDC is also transferred via buffer 120 and the IOC control bus 159 to the I/O devices to control memory read operations therein. The address latch enable signal ALE is utilized only within the MPC and, as previously mentioned, strobes the address latch 106 to develop the address signals MAO-MA7.
The data enable signal DEN provided on control bus 115 is applied to the IOC data transceiver 142 to condition it for data transfer operations. DEN is also coupled via buffer 120 and the IOC control bus 159 to the I/O devices.
~ data transmit/receive control signal DT/R is fed internally ~ 1fi5`~50 within the MPC to the transceiver 14~ and operates to properly condition the bus driver circuits thereof to perform transmit or receive operations. DT/R is also coupled to the I/O devices via IOC control bus 159.
An input/output read command IROC is provided on con-trol bus 115 and applied internally within the MPC to the interrupt controller 124, serial Com Port }28, and the diagnostic PROM 110 to provide a read enable signals to ~ these circuits during input/output operations. IROC is also `transmitted to the I/O devices via control bus 159. An input/output write command IOWC is also generated for appli-cation to the interrupt controller 124, serial Com Port 128, and diagnostic PROM 110 to enable write operations on the ~input/output ports of the circuits. IOWC is likewise trans-' mitted via the IOC control bus 159.
An advanced input/output write command signal AIOWC is generated and employed on the IOC control bus to set up selected registers in the I/O devices prior to the genera-, tion of IOWC. An interrupt acknowledge signal INTA, also " provided on control bus 115, is applied to the interrupt controller 124 as well as to the I/O devices to acknowledge the receipt of interrupt signals initiated by these devices.
Oscillator 104 feeds the clock signal CLK to the micro-j~processor 104 for the usual timing and control purposes and in addition provides an input to decode circuit 116 to enable the generation of the nine control signals just described. In addition, oscillator 104 feeds the clock signal to real time clock generator circuit 118 which in turn generates an interrupt signal IR8. IR8 is fed to inter.upt controller 124 to provide internal interrupt capability within the MPC.

~ 1fi5~l5~) The ~C is provid d with a R~M chip 130 including 256 bytes of random access memory. In addition, the MPC has lKB
of additional random access memory in a second RAM chip 132.
A diagnostic programmable read only memory (PROM) chip 110 is also provided within the MPC to store the self test and configuration mapping control firmware, which is described hereinafter in detail. RAMs 130 and 132 and PROM 110 are controlled by a memory chip select circuit 134 which receives the eight address signals MA8-MA15 from address bus 111 and provides an enabling output to R~M 130, RAM 132, or PROM
110, depending on thc area of memory being accessed by MA8-;MA15. R~M chip 130 is addressed by address signals MA0-MA6 ': I
which are applied from address bus 111 to the address inputs 'A of the RAM chip. RAM 132 is addressed by MA0-MA9. The diagnostic PROM 110 is addressed via the multiplex bus 103 by AD0-AD7 applied directly from microprocessor 102. As previously mentioned, read commands MRDC are applied to RAMs , 130 and 132 and the read command IORC is applied to the PROM
110. Data read from RAMs 130 and 132 is coupled via the internal data bus 105 and transceiver 108 to the microprocessor, while data read from PROM 110 is transmitted to the micro-processor via multiplex bus 103.
The MPC also includes eight 2XB PROM modules 136-138, 'which are used to store the executive firmware employed by the system. PROMs 136-138 a~e enabled by the outputs supplied by a PROM chip select circuit 140. The latter is enabled by IO/MEM fed to its ~ating input G and is further controlled by address inputs MA11-MA13 supplied by address bus 111.
The output of chip select circuit 140 selects one of the 3 ~ 6~t~0 eight PROMs. Each PROM is supplied at its address input A
with address signals MA0-MA10 to select a particular byte location therein. The MRDC read command is also supplied to the PROMs 136-138 as a read strobe signal.
An MPC device select decoder 122 is provided to gener-ate enabling inputs to the diagnostic PROM 110, interrupt controller 124, and serial Com Port 128. Decoder 12~ dècodes the MA4-MA7 address inputs from address bus 111 and is gated by an output from I/O device select decoder 126. The latter decodes the MA8-MAll address signals and is gated by IO/MEM
which is supplied from the microprocessor control bus 164.
The sixteen outputs from decoder 126 represent the gating input to decoder 122 and the fi~teen IOC select bus signals I/O 0lXX-I/O AXX.
Serial Com Port 128 is a programmable communication chip circuit which receives input address data via the MA0-MA2 address signals and an enable signal from MPC device select decoder 122. The serial Com Port 128 receives 8-bit parallel data bytes from the internal data bus 105 and transmits them in bit-serial fashion to the safe unit 212 via the serial port data bus 162. Conversely, Com Port circuit 128 receives bit-serial data on line i62 and transmits it in 8-bit parallel byte form to internal data bus 105.
The circuit 128 utilizes the read and write control signals IORC and IOWC and generates the interrupt signals IR9 and IR10 which are transmit and receive interrupts, respectively.
Interrupt controller 124 is enabled by the MA0 signal from address bus 111 and responds to interrupt signals IR8-IR22 received from the real time cloc~ generator 118, serial `I 1 ~S `~50 ?

Com Port 128, and the I/O devices via IOC interrupt bus 158 to generate the interrupt signal INTR to microprocessor lQ2.
The microprocessor in turn generates the interrupt-acknowl-edge signal INTA to interrupt controller 124 and to the I/O
S devices to signal that the microprocessor is ready to process the interrupt. Thereafter, interrupt controller 124 transmits an interrupt vector to the microprocessor via internal data bus 105. The IR vector is a 32 bit word which is stored within the interrupt controller and tells the processor what 10 memory location in the RAM circuits stores the instructions for servicing the particular interrupt.
Address signals MA0-MA7 are utilized to provide operat-ing instructions to the I/O devices when the latter are enabled by the IOC select signals I/O 01XX - I/O 0AXX.
15 Thus, each of the I/O devices connected to interface con-nector 150 has the capability of decoding ~0-MA7 to provide up to 256 control functions.
~- Diagnostic PROM 110 stores the diagnostic firmware resident within the MPC for performing the self test and 20 configuration mapping operations provided by the present invention. The system hardware is arranged to cause the microprocessor 102 to issue the hexidecimal address FFFF0 during the power-up cycle when the system is initially turned on. When FFFF0 is applied to the address bus 111 and 25 a memory read command is issued, the first instruction from the firmware resident in diagnostic PROM 110 is fed to the microprocessor and-the system commences the self test routine.
The latter may also be called by non-resident diagnostic -software provided from the host controller via com controller 1 ~ 65'~50 200. The purpose of the self test routine is to provide a go/no-go test for the system. Additionally, the routine creates a system configuration map. The pass/fail status of the system, after being exercised by the self test routine, appears visually on an LED display 112 which is driven through the diagnostic PROM 110. There are nine LEDs within display 112 that are stepped to indicate the particular self test subroutine being exercised and are extinguished only after the entire rountine has been completed successfully.
In case of failure, a particular failure code is displayed by the LEDs to alert the operator to the cause of the failure.
The self test routine operates with maximum reliability since it is contained in a dedicated diagnostic PROM which is coupled directly to the microprocessor 102 via the multi-plex bus 103. This bus architecture ensures the execution ~of the self test microprogram as long as at least the micro-processor and the diagnostic PROM are functional. The self test strategy for failure detection operates such that any failure within the MPC or the availability of less than 16K
of random access memory within external memory module 210 is considered to be a catastrophic failure and terminates system operation with the point of failure being encoded on the LED display 112. In such a case the self test firmware hangs up in a loop which continues to repeat the failing test. If the program detects non-catastrophic failures, approriate error flags are set and status indicators gen-erated and entered into a configuration table in MPC ROM 132 and the system continues in operation.

~ 1 65l~50 The self test routine perorms the following operations in order:
1. Microprocessor Test

2. Random Access Memory Test

3. Real Time Clock Test

4. MPC Loop-Back Communications Test

5. ROM Signature Calculation

6. Controller I/O Test

7. System Configuration Creation Figs. 4-15 illustrate the self test microprogram in flowchart diagram form. Figs. 16-38 disclose the micro-program instructions encoded in Intel 8088 source code, which is well known to those skilled in the art. As pre-viously indicated, microprocessor 102 may be an Intel 8088 microprocessor chip and accordingly is capable of executing the microprogram illustrated in connection with Figs. 4-38.
.

OPERATION
Fig. 4 shows the basic self test control loop which, as previously described, is initiated during the system power up cycle and proceeds through a number of self test subroutines. In the process of execution it compiles con-figuration tables which provide system status and configuration information for use by the operating system and by remote units in the processing network which must from time to time interface with the MPC and its various I/O devices. As shown in Fig. 4, the first step A in the self test control loop sets LED display 112 to indicate 0000 0001, which tells the operator that the system is executing the initial test of the routine, the CPU test. The control loop then sequentially t 1 65~150 cyclas through the various testing subroutines B, C, D, E, F, G, and H.
After creating the system configuration tables in step H, the system executes step I wherein the LED display is either set to provide the appropriate error indication, if failures occurred, or is turned off to signal the pass condition. Thereafter, branching step K is executed to determine whether the system is to continue in the diagnostic mode, in which case the Y branch is employed to exit to a diagnostic monitor subroutine L. If the system is to proceed to its regular operation, step K is exited via the N
branch, LED No. 4 is turned on in step M and the system exits to the operational firmware in step N to initiate normal operations.
The microprogram instructions included in the self test control loop of Fig. 4 are shown in detail beginning with the legend "Self Test Begins Here" in Fig. 18 and continuing to the legend "End Of Self Test Control Loop" shown in Fig.
lg. The control loop begins with a series of seven initiali-20- zation instructions beginning with the instruction ORG
OF800~ and ending with the instruction MOV SP,S~ACKH. The subsequent instructions within the control loop are designated by the letters A-N in Figs. 18 and 19, corresponding to the similar designations employed in Fig. 4. The information 2~ provided in Figs. 16, 17, and the top of Fig. 18 provide the necessary definitions for understanding the program. The actual instructions appear in columns two and three of the listing. As indicated in Fig. 18, the first instruction in step A is MOV AL,LEDCPU. This loads register AL with a `~ ~ 65~t~0 data word representing the CPU test pattern "LEDCPU". As shown in Fig. 16, "LEDCPU" equals 01 (0000 0001 in binary notation). On completion of step A, the program executes JMP CPUTST, which is the initial instruction of the CPU test subroutine and directs the program to jump to the location of that subroutine.

CPU Test The CPU test subroutine B is illustrated in the flow-chart diagram of Fig. 5 and described in detail by the microprogram instructions listed in Figs. 19-22 and desig-nated BO-~6. The latter designations correspond to the . i similar designations BO-B6 applied to the flowchart diagram of Fig. 5. Instruction group Bl tests the system data buses -by transmitting patterns of alternating l's and 0's over the buses and verifying the received data to determine if an error condition exists. If no errors are detected, the routine advances to the instruction group B2 to test the various addressing and data transfer modes of the micro-processor. Next, instruction group B3 is executed to test the control transfer capability of the microprocessor first through verification of conditional control transfer opera-tions and then through verification of iteration control transfer operations.
Next, the arithmetic and logic unit of the micropro-cessor is tested through execution of instructions B4. As shown at the top of Fig. 21 instructions are executed for testing the increment, decrement, rotate right, rotate left, not, and addition operations. The final operations executed in the CPU test subroutine are performed through the execution ~ t fi5'~0 of instructions B5 which test the ability of the micropro-cessor to perform certain data string scanning and searching operations. Upon successful completion of the microprocessor tests, the program executes instruction B6 JMP SELFLP2 which is a jump to instruction SELFLP2 in the self test control loop. As shown in Fig. 18, SELFL~2 is the initial instruction of the random access memory test subroutine C. Since this instruction is JMP MEMTST calling for a jump to MEMTST, the `program jumps to instruction C0 shown in Fig. 22 to initiate the RAM TEST subroutine. If an error was detected during the CPU tests, the program does a jump to CPUERR which in turn jumps to CPUHLT in the control loop (Fig. 18), which flags the detection of the CPU error.
.
Random Access Memory Test The RAM TEST subroutine, illustrated in Fig. 6, includes the instruction groups C0-C15 which are described in detail in Figs. 22-26. In this subroutine, the functionality of the MPC static RAM 132 is first tested and if successful operation is verified, the circuits within external memory l!
module 210 are checked. Referring to Fig. 6, after the LED
~display is set to indicate 0000 0010 in step Cl, the MPC RA~q TEST subroutine C2 is entered to check the operation of the MPC RAM 132. The subroutine C2 includes the four operations C2-0 through C2-3 shown in Fig. 7. The microprogram instruc-tions which implement these procedures are shown at the top of Fig. 25. Operation C2-2 is the further subroutine RAMTST
which is illustrated in the flow diagram of Fig. 7a and includes the steps C2-2-0 through C2-2-8. These latter instructions are in turn described in detail in Fig. 26.

~ ~ ~54~) !

The proceduxe used to verify the functionality of the MPC RA~l is contained in the RAMTST subroutine of Fig. 7a.
After the parameters have been properly initialized in step C2-2-1, the program writes, in each location of the RAM, test data determined by subtracting the least significant bit from the most significant bit of the address of the test location. Thereafter, this data is verified in a read operation in step C2-2-3 and if any error is detected, a logical exclusive OR operation is performed between the expected data and the detected data. Any failed RAM bits " are thus indlcated by the logic state 1 and an accumulative failure statistic is obtained by logically ORing the indi-vidual occurrences detected during the test run. The error detection and accumulation function is performed in step C2-2-4. Thereafter, the same sequence of tests is repeated with the test data complemented. This is done is steps C2-2-5 through C2-2-7.
If any errors are detected during the C2 subroutine, it ,. i is considered to be a catastrophic failure and when the program advances to step C3 (Fig. 6) it loops back to step Cl and the MPC RAM TEST procedure is repeated. This loop reiterates so long as an error is detected and the self test program hangs up with the LED display set at 0000 0010.
Assuming that the MPC RAM checks out satisfactorily, the program advances to step~C4 to initiate testing of the external RAM memory. This is done in the TFST 16KB OF IOC
RAM subroutine C9, which is shown in further detail in Fig.

8 and which, as noted, includes RAMTST subroutine C9-2. The instructions included in the subroutine C9-2 are shown in 1 1 6~50 the middle of Fig. 25. It is noted that instruction C9-2 is JMP RAMTST, which is a jump to the RAMTST subroutine shown in Fig. 26 and Fig. 7a, which is the same subroutine previously used to test the MPC RAM. Thus, the ~MSB-LSB) test pattern is verified for each address twice, first in a direct and then in a complemented mode. This is done for each 16K of external RAM. After each 16K RAM block is checked, the re~ults are recorded in step C10 (Fig. 6) for use in the configuration table.
The specific instructions included in step C10 are given in Fig. 23. The program enters a 4 byte word into the table (located in MPC RAM 132) for each 16K byte of external RAM. The first two bytes contain an address point-ing to the beginning of the 16K byte block, the third byte i5 , records the IOC interface (slot) number of the associated memory controller, and the fourth byte records the results of the RAM test. This latter byte is 0 if no failures were ., i detected, otherwise it contains the accumulated failure bits l detected for the 16K memory block. Appropriate table pointer and size data is then entered in the self test control section of the MPC RAM 132. This is done for each 16K block ', of external memory tested.
After each external RAM block within a given interface slot is tested, as described above, the program advances to ; steps C12 and C13 (Fig. 6) and searches for additional inter-face slots that contain external memory to be checked. The -same procedures are repeated for each memory-controller located. After all external memory has been checked, the program executes branch instruction C14 to determine if any ~ I G5~5~

16K block of external memory was found to be good. If no good block flag is found to be set ~instruction group C14 -Fig. 24) the program reverts to step C4 and repeats the entire IOC RAM test cycle again. This loop reiterates so long as no good 16KB block of IOC RAM is located. This condition is therefore a catastrophic failure condition which hangs up the self test progxam in a pattern generation mode with the appropriate error indication being displayed on the LEDs. If at least one good 16K block of RAM was found, the program exits step C14 through the Y branch and advances to step C15, which as shown in Fig. 24 is a jump to SELFLP3. This returns the program to the control loop (Fig.
18), whereupon the instruction CALL RTCTST is executed to initiate the real time clock test subroutine D.

i5 Real Time Clock Test The Real Time Cloc~ Test subroutine D is shown in Figs.

9 and 27. (the IOC RAM TEST description given on page 2-4 of the functional specification should be inspected for description that can be added to the above part of the specification).
The real time clock test determines if the real time clock resident within the ~C is operational. The real time clock generator 118 (Fig. 3a) is designed to interrupt the microprocessor 60 times a second (approximately every 17 msec.). RTC test ins~ructions Dl-D4 test to see whether the interrupt runction is operabie. Instructions D5-D7 test for the occurrence of premature interrupts while instructions D8-D10 test to make sure that no interrupt can occur when 1 1 t ~

the real time clock is disabled. Failure of the real time clock in any of the three modes tested represents a cata-strophic failure condition and thus if the program exits step D4 via the N branch or if it exits either of branching steps D7 or D10 via the Y branch it returns to step Dl and repeats the real time clock test. The program thus hangs ' up in this loop so long as any error is detected and further operation of the system is effectively inhibited with an appropriate error display being presented on the LEDs.
,j If the real time clock test is executed successfully, the program exits via the N branch of step D10 and enters exit step Dll which, as indicated in Fig. 27, is a return ` (RET) instruction. This causes the program to revert to the self test control loop (Fig. 18), which is entered at SELFLP4 for execution of the instruction CALL LPBKTST. This call instruction initiates the MPC Loop-Back Communication Test subroutine E.
, Loop-Back Communication Test 1. :
, The MPC Loop-Back Communication Test subroutine E is described in detail in the flowchart diagram of Pig. 10 and the instruction listing of Pigs. 28 and 29. This test is designed to verify the capability of the Com Controller 200 ,which may be, for example, a Signetics model 2651 program-';mable communications interface module.
The MP~ communications test subroutine E operates to put the Com Controller in a local loop-back mode wherein the transmitter and receiver are tied together. By transmitting and receiving predetermined bit patterns and checking the I J 65 ~ 5 0 results, the program ascertains whether or not the controller is operating properly. As shown in Fig. 10, the first step of the subroutine, El, changes the LED display to indicate 0000 0100, signalling that the loop-back communications test is running.
In steps E2 and E3 the Com Controller is configured for local loop-back operation with the interrupts disabled. In steps E4 and E5 a logical 1 data pattern is transmitted with rotation of the 1 bit to all bit positions. Thereafter a complementary data pattern is transmitted with the logical 0 being rotated to all bit positions. The data patterns received are verified for accuracy and any failed data bits detected are accumulated in memory. Since an error in this test is not considered to be catastrophic, detected errors are flagged and saved and the system is allowed to continue the testing functions.
On completion of the loop-back communication test, the program executes exit instruction E7 (Fig. 28) which is a return to the basic control loop (Fig. 18). This executes ~CALL SIGNTRE, which is the initial instruction of the ROM
~Signature Calculations subroutine F.

ROM Signature Calculations ~i The ROM Signature Calculations subroutine is described in detail in connection with.the flowchart diagram of Fig.
11 and the instruction listing of Figs. 30-32. The signature calculation subroutine is provided as a means for verifying the MPC PROMs, which store the system executive firmware.
The calculation subxoutine calculates a signature for each lK byte within the PROMs. The signature is calculated in the form of an eight digit binary coded decimal (BCD) word which should have the same value as was determined from the paper tape data used in burning the PROM. A signature value of 7616~070 results for any non-existent lX byte within the PROMs. No error conditions are generated in connection with the ROM signature calculation subroutine since the sys-tem as described herein does not perform a comparison of the cal-culated signature with the known signature value.
In the first step Fl of the subroutine, the LED display is set to indicate 0000 0101 to signal that the signature calculation subroutine is running. Thereafter, in step F2 the processor is set to address the beginning of the initial lK byte in the PROM. In step F3 a signature word is generated by reading out the contents of the PROM and accumulating the data bits therein in accordance with a prescribed signature calculation procedure. The procedure is described at the top of Fig. 31 and is implemented by first setting the signature value to 00000000 (BCD). Thereafter, the signa-;ture value is multiplied by 2 and a first data byte is read out of the PROM and added to the total. If this operation results in a carry being generated out of the most signifi-cant digit of the total, then a BCD 1 is added. This value ; then becomes the signature value and the steps B through D
are repeated to continue the calculation operation.
When 1024 bytes have been read from the PROM and thus accumulated, the program advances to step F4 where the calculated signature word is saved in the configuration table. As indicated in Fig. 30, this procedure is performed i l 6~ 0 by entering the 2 byte offset address of the PROM into the signature save area and then the 4 byte signature word (eight BCD digits) is entered. Thereafter the signature pointer value is moved to the signature table area;
Next, in step F5 the PROM address pointer is incremented by lK to prepare for generation of the next signature word.
; When signatures have been calculated for each lK block of `~PROM memory, the program advances to exit step F7 and executes a return to the self test control loop (Fig. 18) and initiates subroutine G by executing call instruction CALL IOTST.

Controller I/O Test The Controller I/O Test is described in the flowchart diagrams of Figs. 12-14 and by the program instruction list-ing shown in Figs. 33, 34, 35, and the top of Fig. 36. The controller I/O test determines the controller self test ; status and then verifies the interrupt response of the con-troller. As shown in Fig. 12, the IOC interface slot number is set to 0 in step Gl and thereafter the LED display is set in step G2 to indicate the slot number and the number 0110, which indicates that the controller I/O test is being run.
Next, the controller self test status subroutine G3 is `~executed. This subroutine is shown in the flowchart diagram of Fig. 13 and in the instruction listing of Fig. 34. In , , step G3-1, the program reads the identification number of the first controller and saves it in the configuration table along with the I/O slot number. In step G3-2 it ascertains whether the controller has a self test capability. If the answer to this is yes then a self test command is issued and i ~ 6~50 the system waits for the controller to finish the test. If errors are detected during the controller self test operation, they are indicated by appropriate flags.
Thereafter, in step G3-3 the controller status is saved S in the save area of memory for the configuration table. Two bytes of çontroller status configuration data are entered in ', the save area adjacent the previously entered IOC slot number byte and the controller ID byte. Thereafter, the program advances to step G3-4 and executes a return instruc-tion which initiates step G4 (Fig. 12) by calling the Interrupt ~esponse and Contention Test subroutine.
; ~he Interrupt Response and Contention Test subroutine is shown in the flowchart diagram of Fig. 14 and in the instruction listing disclosed in Fig. 35 and the top of Fig.
lS ,;36. This tests verifies the interrupt response for each controller and determines whether there is any interrupt contention between adjacent controllers. Interrupt conten-`
tion results when a controller requiring an interrupt level 'is configured in an interface slot following a controller ' requiring two interrupt levels. Detection of an interrupt contention results in the program turning on LED 12 to signal the interrupt contention condition. The program (Fig. 14) first assigns an interrupt vector in step G4-1 and ; then issues an interrupt request to the selected controller slot in G4-2. In step G4-3 the designated interrupt level is checked to see if a response was received as expected and if no response or an erroneous response occurs, the program branches out of G4-3 via the N branch to step G4-6, where an appropriate error flag is set.

`~ 165~0 If the e~pected interrupt response is detected, the program exits G4-3 via the Y branch to step G4-4 and checks whether the numbers of interrupts received equals p. If yes, the program branches directly to step G4-7 where the interrupt response status is saved in the configuration table. If the number of interrupts received is not equal to 0, the program takes the N branch out of step G4-4 and proceeds to step G4-5, where it determines whether the controller in the preceeding interface slot required two interrupts. If the response to this is no, the program branches to G4-7 to compile the interrupt response status for the configuration table. If the answer determined in step G4-5 is yes, the program advances to step G4-6 and sets an appropriate error flag.
,' In step G4-7 the program enters a single interrupt status byte into the configuration table location for the controller being tested. The interrupt status byte is an 8-bit byte wherein bits 0-5 indicate the interrupt level `, value, bit 6 indicates whether the interrupt test was passed 20 ; or failed and-bit 7 indicates the number of interrupt levels ; for the controller.
! After completing the interrupt response and contention ! test for the selected controller, the program advances to step G5 (Fig. 12) and increments the interface slot number.
When all interface slots have been tested as described above, the program advances to step G7 where LED 12 is turned on if an interrupt test error had been detected. Thereafter, the program advances to exit step G8 which, as shown in Fig. 33, is a return instruction. This returns the program to the i 1 ~5'150 control loop ~Fig. 18) whereupon the instruction CALL SYSCNFG
is executed to initiate the System Configuration Creation su~routine H.

System Configuration Creation The System Configuration Creation subroutine H is described in the flowchart diagram of Fig. 15 and in the instruction listing shown in the middle of Fig. 36. The ~, .. i initial step Hl sets the LED display to indicate 0000 0111, signalling that the system configuration creation subroutine is being run. Thereafter, as shown in Fig. 36, the three instructions H2, H3, and H4 are executed. Each of these instructions is a call instruction followed by a return ; instruction. The final return, instruction H5, causes the program to exit the subroutine and return to the self test control loop to execute the instruction CALL TSTSTS to ; initiate the subroutine I. The latter includes steps I0, Il, and I2 shown in Fig. 4 and described in connection with the program listing of Fig. 37.
In step I0, the program determines whether any test failures were detected during the self test routine. If no failures were detected, the program exits via the N branch ~and executes step I2 to turn off the LEDs of display 112.
! If any failures were detected, the program executes step Il to display the appropriate error pattern on the LEDs. As previously described, nine LEDs are used. LEDs 1-4 display an indication of the self test subroutine which detected the failure condition. The subroutines are indicated on LEDs 1-4 in accordance with the following table:

i 1 6S45~

0001 - CPU Test 0010 - RAM Test 0011 - Real Time Clock Test 0100 - Loop-Back Communication Test 0101 - ROM Signature Calculation 0110 - Controller I/O Test 0111 - System Configuration Map 1000 - Firmware Initialization In Progress LEDs 5-8 display the controller slot number for which the error was detected. A display of 0000 by LEDs 5-8 indi-cates that the error was detected in the MPC during an MPC
; test. LED 12 provides an indication of an interrupt conten-tion error.
Thereafter, the program exits from the self test routine ~5 ! via steps K, L, M, and N. In step K, the program determines whether diagnostic operation is to be continued and if the answer is yes, it exits to diagnostic monitor subroutine L.
If normal system operation is to be initiated, the program exits through the N branch of step K, turns on LED 4 in step ~, and proceeds to execute the application firmware subroutine N.
It is noted t.hat Fig. 38 describes a variety of control subroutines used in the program steps noted on the right of the listing.
It is thus seen that each time the system is powered up, ; the self test routine is executed to verify the functionality of the system and to create a configuration map providing current system configuration and status data. This data is available for access by the operating system and application software to determine where the different I/O controllers exist and what their current status is. The external control systems thus can determine how to assign their own interrupt vectors, levels, physical I/O, and device handlers to enable ~ 1 ~5~50 their own accessing of the controllers and to alter their procedures in the event that parts of the system are indi-cated to be non-functional.
It will be apparent to those skilled in the art that various modification and variations could be made to the embodiment of the invention as hereinabove described without departing from tbe spirit and scope of the invention.

1, i1 I

:' .:

Claims

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A logic test and configuration map control system in a data process-ing system including processor means and memory means comprising, in combina-tion:
first test means responsive to said processor means for determining the functionality of a first section of said memory means;
inhibit means responsive to said first test means for issuing a con-trol signal to said processor means to inhibit further operation of said logic test and configuration map control system if said first memory section is de-termined to function erroneously;
second test means in electrical communication with said first test means and said inhibit means for determining the functionality of a second section of said memory; and means in electrical communication with said second test means for storing in a predetermined location of said first memory section the results of the operation of said second test means indicating the functionality of said second memory section.

2. The logic test and configuration map control system set forth in claim 1 further comprising:
control means for enabling said microprocessor to initiate said first test means each time said data processing system is powered up.

3. A method for operating a logic test and configuration map control system in a data processing system having processing means and memory means, comprising the steps of:
operating a first test program to determine the functionality of a first section of said memory means;
inhibiting further operation of said system if said first memory section is determined to function erroneously;

operating a second test program to determine the funtionality of a second section of said memory means; and storing in said first memory section the results obtained during the operation of said second test program.

. .