WO1985001136A1 - Processor with depictor-linked microcode and logic circuitry - Google Patents

Processor with depictor-linked microcode and logic circuitry

Info

Publication number
WO1985001136A1
WO1985001136A1 PCT/US1984/001387 US8401387W WO8501136A1 WO 1985001136 A1 WO1985001136 A1 WO 1985001136A1 US 8401387 W US8401387 W US 8401387W WO 8501136 A1 WO8501136 A1 WO 8501136A1
Authority
WO
WIPO (PCT)
Prior art keywords
depictor
activity
type
instruction
program
Prior art date
Application number
PCT/US1984/001387
Other languages
French (fr)
Inventor
Howard H. Green
Christopher J. Tomlinson
Original Assignee
Burroughs Corporation
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Burroughs Corporation filed Critical Burroughs Corporation
Publication of WO1985001136A1 publication Critical patent/WO1985001136A1/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44Arrangements for executing specific programs
    • G06F9/448Execution paradigms, e.g. implementations of programming paradigms
    • G06F9/4482Procedural
    • G06F9/4484Executing subprograms
    • G06F9/4486Formation of subprogram jump address

Definitions

  • This invention relates to data processing systems; and in particular, it relates to methods of operating such systems.
  • a data processing system is operated by providing a program which consists of a sequence of instructions of some predetermined language. These instructions are executed one at a time; and the particular instruction that is currently being executed is pointed to by a program counter.
  • the program counter is, incremented to point to the next sequential instruction unless the completed instruction was a JUMP instruction or a CALL instruction.
  • the program counter is loaded with a new address as the JUMP instruction directs; and then instructions are sequentially executed beginning at that new address.
  • a CALL instruction an independent procedure is executed and then control passes back to the instruction which follows the CALL instruction.
  • a procedure is a separate special purpose sequence of instruction from the same predetermined language of which the program that called it is made up. From the point of view of the calling program, a procedure can be regarded as a single new higher level instruction even though it may be quite complicated and made up of hundreds of instructions from the predetermined language.
  • a primary object of the invention is to provide an improved method of operating a data processing system.
  • Another object of the invention is to provide a method of enhancing the speed at which a data processing system operates.
  • Still another object of the invention is to provide a method of operating a data processing system wherein procedure-like functions are performed by low level language microcode routines and hardware logic units.
  • a data processing system in a manner which includes the steps of: storing, in a memory, a software program that contains first and second instructions which respectively call for first and second activities; linking the first instruction via one type of depictor to a software procedure that performs the first activity, while linking the second instruction via a different type of depictor to a non-software mechanism that performs the second activity; executing the software program until either the first or second instruction is encountered; sensing whether the encountered instruction is linked to its activity by the one type or the different type depictor; and executing the procedure if the sensing step detects the one type depictor; but activating the non-software mechanism if the sensing step detects the different type depictor.
  • the activating step of the non-software mechanism can include the substep of executing a microcode program to perform the second activity; or alternatively, it includes the substep of sending parameters to a hardware logic unit to perform the second activity.
  • the depictor in the above-recited method is a partitioned register in which one portion contains the name of the activity that is to be performed; another portion indicates whether the named activity is implemented via a procedure, a microcode routine, or a hardware logic unit; and another portion indicates how the named activity can be accessed.
  • Figure 1 illustrates one embodiment of a data processing system that is constructed according to the invention
  • Figure 2 illustrates the steps that are there taken by the Figure 1 system when a CALL instruction is executed
  • Figure 3 illustrates the details of the hardware components in the computer of the Figure 1 system that are affected by the steps of Figure 2;
  • Figure 4 illustrates another embodiment of a data processing system that is constructed according to the invention.
  • FIG. 1 a data processing system and method of operating the system in accordance with the invention will be described.
  • a digital computer 10 which is coupled to a memory 11 via a bus 12.
  • a software program 13 which computer 10 executes is stored in memory 11.
  • the term "software program” means a sequence of high level language (HLL) instructions with each such instruction being interpreted by a microcode program of low level language (LLL) instructions.
  • the low level language instructions are executed directly by digital logic circuitry in computer 10.
  • Individual high level language instructions in software program 13 are indicated by reference numeral 14; and the low level language microcode programs are indicated by reference numeral 15 in a memory 16.
  • CALL instructions 14a, 14b, and 14c which have been modified to operate in accordance with the present invention.
  • This modified CALL instruction identifies an activity by name, discontinues the execution of the HLL software program of which it is a part, saves sufficient information to resume execution of the discontinued program at a later time at the point of discontinuance, and initiates the execution of the named activity by passing parameters and control to it.
  • CALL instruction 14a identifies an activity A x to which control is to be passed along with parameters j and k.
  • CALL instruction 14b identifies an activity Ay to which control is to be passed along with a parameter 1.
  • CALL instruction 14c identifies another activity A 2 to which control is to be passed along with parameters m, n, and o.
  • Each of the activities A x , Ay, and A z is either a special purpose software program (i.e., a procedure), a special purpose microcode program, or a special purpose hardware logic unit.
  • the activity is adapted to receive parameters from the CALL instruction, operate on the received parameters in a predetermined fashion, and indicate to the CALL instruction that the operation is complete. For example, one activity may receive the names of several records as parameters from the CALL instruction, read the identified records from various storage units, and sort them in a predetermined order.
  • activity A x is indicated by reference numeral 20 as being a special purpose procedure of HLL instructions in memory 11. Those high level language instructions are from the same language as the instructions which make up program 13.
  • activity A x To execute activity A x , the address of the beginning high level language instruction in activity A x is loaded into a program counter in computer 10; and then instructions from activity A x are read via the high level language bus 12 into computer 10 where they are interpreted by the microcode programs 15.
  • activity A y is indicated by reference numeral 21 in Figure 1 as being a special purpose microcode program in memory 16.
  • Microcode program 21 is made up of a sequence of low level language instructions which are read over a low level language bus 22 into computer 10 where they are executed directly by the computer's digital logic circuitry.
  • activity A z is comprised of a special purpose logic unit which is coupled via an I/O bus 24 to computer 10.
  • computer 10 sends the parameters of the CALL instruction to unit 23 via bus 24 and then merely waits for unit 23 to return the results of the operation which the activity performs.
  • a depictor array 25 is also coupled to the I/O bus
  • FIG. 24 contains respective depictors for each activity that is called in program 13.
  • FIG. 24 contains respective depictors for each activity that is called in program 13.
  • FIG. 25-3 contains three depictors 25-1, 25-2, and 25-3.
  • a depictor is herein meant a mechanism which links a named activity to other information which describes how the named activity can be accessed.
  • each depictor in array 27 is a content addressable hardware register that is partitioned into three parts, "a", "b”. and "c".
  • Part "a” contains the name of the activity.
  • Part "b” identifies whether the named activity is a software program in memory 11 or a microcode program in memory 16 or a hardware logic unit on I/O bus 24.
  • part "c" specifies the memory address or bus address at which the activity can be accessed.
  • the name of an activity is sent over bus 24 to array 25; that name is compared with part "a” of the depictors until a match occurs; and parts "b” and "c" of the matching depictor are sent to computer 10.
  • FIG 1 part “a” of depictor 25-1 shows that it is a depictor for activity A x ; part “b” shows that activity A x is implemented by HLL instructions in memory 11 as a procedure; and part “c” shows that the instructions begin at address 5000.
  • depictor 25-2 shows that activity Ay is implemented by low level language instructions which begin at microcode address 200; and .depictor 25-3 shows that activity A z is implemented by a hardware logic unit which is at I/O bus address 100.
  • Figure 2 illustrates a flow chart for the portion 15a of the microcode programs 15 which is executed when a CALL instruction in program 13 is encountered.
  • the microcode 15a directs computer 10 to fetch the depictor that is linked to the called activity from depictor array 25. Thereafter, as indicated by boxes 31, 32, and 33, the microcode 15a directs computer 10 to examine field "b" of the fetched depictor to sense how the called activity is linked to an implementation mechanism. If the called activity is implemented by high level language instructions in memory 11, then the microcode 15a directs computer 10 to save the present program counter and load the memory address of the called activity into the program counter. This address is contained in portion "c" of the depictor of the called activity. Thereafter, the high level language instructions of the called activity are executed until the activity goes to completion; whereupon control is returned back to the calling program by reloading the program counter with the previously stored address. All of this is indicated in the flow chart at 34a and 34b.
  • the microcode 15a directs computer 10 to send the parameters of the CALL instruction to the hardware logic unit which implements the called activity.
  • processor 10 sends parameters m, n, and o to the unit 25 at address 100 on the I/O bus 26. Thereafter, computer 10 waits for the hardware unit to return the results of the operations that it performs. This is indicated in the flow chart at 35a and 35b.
  • microcode 15a detects that the called activity is performed by a special purpose microcode program
  • the present microcode address counter is saved and the beginning address of the called activity in microcode memory 16 is loaded into the microcode address register.
  • microcode in the called activity is executed until the called activity goes to completion; whereupon the saved microcode address is reloaded back into the microcode address register.
  • FIG. 3 One preferred architecture for the logic circuitry in computer 10 which is adapted to achieve the above steps is illustrated in Figure 3. That architecture includes a HLL bus interface 12, a LLL bus interface 23, and an I/O bus interface 26 which respectively correspond to the buses which were previously described in conjunction with Figure 1.
  • a program counter register 40 is provided for addressing HLL instructions on bus 12; and an instruction register 41 is provided for receiving HLL instructions from bus 12. Also, a stack 42 is provided for saving the present contents of the program counter 40 and for reloading those saved contents back into the program counter.
  • a microcode address register 43 is provided in the Figure 3 architecture for addressing LLL instructions on bus 23; and a microcode instruction register 44 is provided for receiving LLL instructions from bus 23. Also, another stack 45 is provided for saving the present content of microcode address register 43 and for restoring addresses back into the microcode address register 43.
  • An I/O bus address register 46 is provided in the Figure 3 architecture on bus 26 to transmit addresses on the bus and thereby select one of the units that are to the bus. Also, an I/O bus data register 47 is provided for sending parameters to the selected unit and for receiving data therefrom.
  • One feature of the above-described system of Figures 1-3 is its versatility. To have the option of implementing an activity via a high level language procedure, a low level language microcode program, or a hardware logic unit is very desirable since various speed-cost trade-offs can thereby be made.
  • microcode and logic circuits can perform more primitive operations than high level language instructions. Thus, they can be arranged to perform a given task more efficiently. Also, microcode and logic circuitry operates more quickly than HLL instructions. On the other hand, if low cost is the primary design goal, then most of the activities should be implemented by high level language procedures. This is because the cost of relatively slow HLL memory space is less than the cost of relatively fast LLL microcode memory space or high speed logic circuitry.
  • Another feature of the above-described system is that the means by which a particular activity is implemented is very easy to change with time. For example, an activity that had previously been implemented by a high level language procedure can be implemented at a later date as a low level language program or a hardware logic unit.
  • Such a change is achieved by merely providing the low level language program or hardware logic unit along with a new depictor in depictor array 25.
  • the speed of a particular system can be enhanced from time to time without changing the system's overall architecture.
  • the present invention may be incorporated into a multiprocessor system to improve that system's versatility and performance.
  • One suitable multiprocessor system for example, is described in a co-pending patent application entitled "Method of Performing a Sequence of Related Activities in Multiple Independent Digital Processors" by Christopher J. Tomlinson and Howard H. Green which was filed July 21, 1983 and assigned Serial No. ________. All of the teachings of that application are hereby incorporated by reference.
  • FIG. 4 of the present case illustrates the above-referenced Tomlinson and Green multiprocessor system with modifications that incorporate the present invention.
  • This Figure 4 system includes a pair of digital processors P i and P j which are coupled together via bus 50.
  • Processor P i is also coupled to a memory 51 via a high level language bus 52; and processor P j is coupled to a memory 53 via a high level language bus 54.
  • Included in memory 51 is a program 55 which contains a sequence of high level language instructions 56.
  • One of those instructions 56a is a CALL instruction which calls an activity A x P i that processor P i performs; while another of those instructions 56b is a CALL instruction which calls an activity A y P j that processor P j performs.
  • processor P i executes the instructions 56 until it encounters one of the CALL instructions 56a or 56b. When that occurs, processor P i interrogates an ACTIVITY RECORD 60 for the activity that is being called. That
  • ACTIVITY RECORD is in a shared memory SM; and it is accessed via bus 50 through an intelligent memory controller IMC.
  • each ACTIVITY RECORD 60 includes various FLAGS, a CALLER entry, an ACTIVITY QUEUE HEAD (AQH) entry, a NEXT IN QUEUE (NIQ) entry, and an ACTIVITY QUEUE TAIL (AQT) entry.
  • each activity record also includes an activity depictor 61.
  • each depictor 61 is a pre-written thirty-two bit word in a random access memory or a thirty-two bit register in a register array.
  • processor P j selects an activity to perform, it then examines depictor 61 of the selected activity to determine how the activity is implemented and how it can be accessed.
  • depictor 61 contains a portion "b" which indicates whether the corresponding activity is implemented via a high level language program, a low level language microcode program, or a hardware logic unit. Also preferably, depictor 61 contains a portion "c" which indicates how that activity can be accessed.
  • processor P j performs all of the previously described steps of Figure 2. Thus it examines portion "b" to determine whether the selected activity is implemented by a software program 71, a firmware program 72, or a hardware logic unit 73. Then it utilizes the information in portion "c" to activate the activity as indicated at 34a-36b. In like manner, when processor P i calls an activity which it performs, the processor then examines the depictor for that activity to determine whether it is implemented by a software program 74, a microcode program 75, or a hardware logic unit 76. Then, depending on that determination, it performs steps 34a-34b, 35a-35b, or 36a-36b of Figure 2.
  • the multiprocessor system of Figure 4 has both the versatility and changeability features of the single processor system of Figure 1. If high speed is the primary design goal of the Figure 4 system, then the frequently used activities should be implemented by low level language microcode programs or hardware logic units. And the performance of the multiprocessor system can be enhanced with time by substituting low level language programs or hardware logic units for various activities that previously had been implemented by high level language softv/are programs, and by providing nev/ activity depictors 61 in the activity's record 60 to reflect this change.
  • Various preferred embodiments of the invention have now been described in detail. In addition, hov/ever, many changes and modifications can be made to these details without departing from the nat ⁇ ure and spirit of the invention. Accordingly, the invention is not to be limited to said details but is defined by the appended claims.

Abstract

A method of operating a data processing system (Fig. 4) including the steps of: executing one high level language software program until an instruction is encountered which calls an activity; sensing whether said encountered instruction is linked to the activity which it calls by a first type or a second type depictor (Fig. 2; executing another high level language software program (53, Fig. 4) for performing the called activity if the sensing step detects the first type depictor; and activating a low level language microcode program on hardware logic circuit (74, 75, Fig. 4) for performing the called activity if the sensing step detects the second depictor (Fig. 2).

Description

Processor With Depictor-LInked Microcode and Logic Circuitry
BACKGROUND OF THE INVENTION This invention relates to data processing systems; and in particular, it relates to methods of operating such systems.
Conventionally, a data processing system is operated by providing a program which consists of a sequence of instructions of some predetermined language. These instructions are executed one at a time; and the particular instruction that is currently being executed is pointed to by a program counter.
Each time the execution of an instruction is completed, the program counter is, incremented to point to the next sequential instruction unless the completed instruction was a JUMP instruction or a CALL instruction. When a JUMP instruction is encountered, the program counter is loaded with a new address as the JUMP instruction directs; and then instructions are sequentially executed beginning at that new address. By comparison, when a CALL instruction is encountered, an independent procedure is executed and then control passes back to the instruction which follows the CALL instruction.
A procedure is a separate special purpose sequence of instruction from the same predetermined language of which the program that called it is made up. From the point of view of the calling program, a procedure can be regarded as a single new higher level instruction even though it may be quite complicated and made up of hundreds of instructions from the predetermined language.
By writing a collection of procedures, a programmer can define a new level instruction set. Then programs in the predetermined language can be written which use this new level instruction set by referring to them through the CALL instruction. Additional details on this prior art method of structuring or partitioning a program are found in the Prentice-Hall textbook entitled Structured Computer Organization by Andrew S. Tanenbaum, 1976, at pages 120-130.
One problem, however, with having a data processing system operate to execute a program which includes several
CALL instructions and their corresponding procedures is that the system operates too slowly. This fact and the manner in which the execution speed can be greatly enhanced in accordance with the present invention is explained in detail in the following Detailed Description.
Still another problem with the prior art method of operating a data processing system by means of CALL instructions and corresponding procedures is that the operation is too inflexible. In particular, the operation is too confined to include special purpose microcode routines or special purpose hardware logic units that perform procedure-like functions. BRIEF SUMMARY OF THE INVENTION
Accordingly, a primary object of the invention is to provide an improved method of operating a data processing system. Another object of the invention is to provide a method of enhancing the speed at which a data processing system operates.
Still another object of the invention is to provide a method of operating a data processing system wherein procedure-like functions are performed by low level language microcode routines and hardware logic units.
In the present invention, the above objects and others are achieved by operating a data processing system in a manner which includes the steps of: storing, in a memory, a software program that contains first and second instructions which respectively call for first and second activities; linking the first instruction via one type of depictor to a software procedure that performs the first activity, while linking the second instruction via a different type of depictor to a non-software mechanism that performs the second activity; executing the software program until either the first or second instruction is encountered; sensing whether the encountered instruction is linked to its activity by the one type or the different type depictor; and executing the procedure if the sensing step detects the one type depictor; but activating the non-software mechanism if the sensing step detects the different type depictor.
Preferably, the activating step of the non-software mechanism can include the substep of executing a microcode program to perform the second activity; or alternatively, it includes the substep of sending parameters to a hardware logic unit to perform the second activity. Also preferably, the depictor in the above-recited method is a partitioned register in which one portion contains the name of the activity that is to be performed; another portion indicates whether the named activity is implemented via a procedure, a microcode routine, or a hardware logic unit; and another portion indicates how the named activity can be accessed.
BRIEF DESCRIPTION OF THE DRAWINGS
Various features and advantages of the invention are described in the following Detailed Description in conjunction with the accompanying drawings wherein: Figure 1 illustrates one embodiment of a data processing system that is constructed according to the invention;
Figure 2 illustrates the steps that are there taken by the Figure 1 system when a CALL instruction is executed; Figure 3 illustrates the details of the hardware components in the computer of the Figure 1 system that are affected by the steps of Figure 2;
Figure 4 illustrates another embodiment of a data processing system that is constructed according to the invention.
DETAILED DESCRIPTION OF THE INVENTION
Referring now to Figure 1, a data processing system and method of operating the system in accordance with the invention will be described. Included in the Figure 1 system is a digital computer 10 which is coupled to a memory 11 via a bus 12. A software program 13 which computer 10 executes is stored in memory 11. As used herein, the term "software program" means a sequence of high level language (HLL) instructions with each such instruction being interpreted by a microcode program of low level language (LLL) instructions. In turn, the low level language instructions are executed directly by digital logic circuitry in computer 10. Individual high level language instructions in software program 13 are indicated by reference numeral 14; and the low level language microcode programs are indicated by reference numeral 15 in a memory 16. Three of the illustrated high level language instructions 14 are CALL instructions 14a, 14b, and 14c which have been modified to operate in accordance with the present invention. This modified CALL instruction identifies an activity by name, discontinues the execution of the HLL software program of which it is a part, saves sufficient information to resume execution of the discontinued program at a later time at the point of discontinuance, and initiates the execution of the named activity by passing parameters and control to it. For example, CALL instruction 14a identifies an activity Ax to which control is to be passed along with parameters j and k. Similarly, CALL instruction 14b identifies an activity Ay to which control is to be passed along with a parameter 1. And CALL instruction 14c identifies another activity A2 to which control is to be passed along with parameters m, n, and o.
Each of the activities Ax, Ay, and Az is either a special purpose software program (i.e., a procedure), a special purpose microcode program, or a special purpose hardware logic unit. In each case, the activity is adapted to receive parameters from the CALL instruction, operate on the received parameters in a predetermined fashion, and indicate to the CALL instruction that the operation is complete. For example, one activity may receive the names of several records as parameters from the CALL instruction, read the identified records from various storage units, and sort them in a predetermined order. In Figure 1, activity Ax is indicated by reference numeral 20 as being a special purpose procedure of HLL instructions in memory 11. Those high level language instructions are from the same language as the instructions which make up program 13. To execute activity Ax, the address of the beginning high level language instruction in activity Ax is loaded into a program counter in computer 10; and then instructions from activity Ax are read via the high level language bus 12 into computer 10 where they are interpreted by the microcode programs 15. By comparison, activity Ay is indicated by reference numeral 21 in Figure 1 as being a special purpose microcode program in memory 16. Microcode program 21 is made up of a sequence of low level language instructions which are read over a low level language bus 22 into computer 10 where they are executed directly by the computer's digital logic circuitry.
Also as is indicated by reference numeral 23, activity Az is comprised of a special purpose logic unit which is coupled via an I/O bus 24 to computer 10. When a CALL instruction in program 13 is encountered which calls activity Az, computer 10 sends the parameters of the CALL instruction to unit 23 via bus 24 and then merely waits for unit 23 to return the results of the operation which the activity performs. A depictor array 25 is also coupled to the I/O bus
24; and it contains respective depictors for each activity that is called in program 13. Thus the Figure 1 array contains three depictors 25-1, 25-2, and 25-3. By a depictor is herein meant a mechanism which links a named activity to other information which describes how the named activity can be accessed.
In one preferred embodiment, each depictor in array 27 is a content addressable hardware register that is partitioned into three parts, "a", "b". and "c". Part "a" contains the name of the activity. Part "b" identifies whether the named activity is a software program in memory 11 or a microcode program in memory 16 or a hardware logic unit on I/O bus 24. And part "c" specifies the memory address or bus address at which the activity can be accessed. To access a particular depictor, the name of an activity is sent over bus 24 to array 25; that name is compared with part "a" of the depictors until a match occurs; and parts "b" and "c" of the matching depictor are sent to computer 10.
In Figure 1, part "a" of depictor 25-1 shows that it is a depictor for activity Ax; part "b" shows that activity Ax is implemented by HLL instructions in memory 11 as a procedure; and part "c" shows that the instructions begin at address 5000. Similarly, depictor 25-2 shows that activity Ay is implemented by low level language instructions which begin at microcode address 200; and .depictor 25-3 shows that activity Az is implemented by a hardware logic unit which is at I/O bus address 100. Reference should now be made to Figure 2 which illustrates a flow chart for the portion 15a of the microcode programs 15 which is executed when a CALL instruction in program 13 is encountered. To begin, as indicated by box 30, the microcode 15a directs computer 10 to fetch the depictor that is linked to the called activity from depictor array 25. Thereafter, as indicated by boxes 31, 32, and 33, the microcode 15a directs computer 10 to examine field "b" of the fetched depictor to sense how the called activity is linked to an implementation mechanism. If the called activity is implemented by high level language instructions in memory 11, then the microcode 15a directs computer 10 to save the present program counter and load the memory address of the called activity into the program counter. This address is contained in portion "c" of the depictor of the called activity. Thereafter, the high level language instructions of the called activity are executed until the activity goes to completion; whereupon control is returned back to the calling program by reloading the program counter with the previously stored address. All of this is indicated in the flow chart at 34a and 34b.
By comparison, if the called activity is implemented by a hardware logic unit, then the microcode 15a directs computer 10 to send the parameters of the CALL instruction to the hardware logic unit which implements the called activity. Thus when activity Az is called, processor 10 sends parameters m, n, and o to the unit 25 at address 100 on the I/O bus 26. Thereafter, computer 10 waits for the hardware unit to return the results of the operations that it performs. This is indicated in the flow chart at 35a and 35b.
Finally, if the microcode 15a detects that the called activity is performed by a special purpose microcode program, then the present microcode address counter is saved and the beginning address of the called activity in microcode memory 16 is loaded into the microcode address register. Thereafter, microcode in the called activity is executed until the called activity goes to completion; whereupon the saved microcode address is reloaded back into the microcode address register. This is indicated in the flow chart at 36a and 36b. One preferred architecture for the logic circuitry in computer 10 which is adapted to achieve the above steps is illustrated in Figure 3. That architecture includes a HLL bus interface 12, a LLL bus interface 23, and an I/O bus interface 26 which respectively correspond to the buses which were previously described in conjunction with Figure 1. A program counter register 40 is provided for addressing HLL instructions on bus 12; and an instruction register 41 is provided for receiving HLL instructions from bus 12. Also, a stack 42 is provided for saving the present contents of the program counter 40 and for reloading those saved contents back into the program counter.
A microcode address register 43 is provided in the Figure 3 architecture for addressing LLL instructions on bus 23; and a microcode instruction register 44 is provided for receiving LLL instructions from bus 23. Also, another stack 45 is provided for saving the present content of microcode address register 43 and for restoring addresses back into the microcode address register 43.
An I/O bus address register 46 is provided in the Figure 3 architecture on bus 26 to transmit addresses on the bus and thereby select one of the units that are to the bus. Also, an I/O bus data register 47 is provided for sending parameters to the selected unit and for receiving data therefrom.
One feature of the above-described system of Figures 1-3 is its versatility. To have the option of implementing an activity via a high level language procedure, a low level language microcode program, or a hardware logic unit is very desirable since various speed-cost trade-offs can thereby be made.
If high speed is the primary design goal of the system, then the frequently used activities should be implemented by a microcode or hardware logic units. This is because microcode and logic circuits can perform more primitive operations than high level language instructions. Thus, they can be arranged to perform a given task more efficiently. Also, microcode and logic circuitry operates more quickly than HLL instructions. On the other hand, if low cost is the primary design goal, then most of the activities should be implemented by high level language procedures. This is because the cost of relatively slow HLL memory space is less than the cost of relatively fast LLL microcode memory space or high speed logic circuitry.
Another feature of the above-described system is that the means by which a particular activity is implemented is very easy to change with time. For example, an activity that had previously been implemented by a high level language procedure can be implemented at a later date as a low level language program or a hardware logic unit.
Such a change is achieved by merely providing the low level language program or hardware logic unit along with a new depictor in depictor array 25. By this means, the speed of a particular system can be enhanced from time to time without changing the system's overall architecture.
The present invention may be incorporated into a multiprocessor system to improve that system's versatility and performance. One suitable multiprocessor system, for example, is described in a co-pending patent application entitled "Method of Performing a Sequence of Related Activities in Multiple Independent Digital Processors" by Christopher J. Tomlinson and Howard H. Green which was filed July 21, 1983 and assigned Serial No. ________. All of the teachings of that application are hereby incorporated by reference.
Figure 4 of the present case illustrates the above-referenced Tomlinson and Green multiprocessor system with modifications that incorporate the present invention. This Figure 4 system includes a pair of digital processors Pi and Pj which are coupled together via bus 50. Processor Pi is also coupled to a memory 51 via a high level language bus 52; and processor Pj is coupled to a memory 53 via a high level language bus 54. Included in memory 51 is a program 55 which contains a sequence of high level language instructions 56. One of those instructions 56a is a CALL instruction which calls an activity AxPi that processor Pi performs; while another of those instructions 56b is a CALL instruction which calls an activity AyPj that processor Pj performs.
In operation, processor Pi executes the instructions 56 until it encounters one of the CALL instructions 56a or 56b. When that occurs, processor P i interrogates an ACTIVITY RECORD 60 for the activity that is being called. That
ACTIVITY RECORD is in a shared memory SM; and it is accessed via bus 50 through an intelligent memory controller IMC.
Some of the entries in each ACTIVITY RECORD 60 are illustrated in Figure 4. They include various FLAGS, a CALLER entry, an ACTIVITY QUEUE HEAD (AQH) entry, a NEXT IN QUEUE (NIQ) entry, and an ACTIVITY QUEUE TAIL (AQT) entry.
If the flags indicate that the called activity is DORMANT, then a pointer to the calling program (i.e., program 55) is placed in the caller entry; whereas if the flags indicate that the activity is ACTIVE, then a pointer to the calling activity is placed in the activity's queue. All of this is explained in greater detail in conjunction with TABLE 1 of the above-referenced co-pending application by Tomlinson and Green. Later, when processor Pj completes the execution of the activity that it was performing, it interrogates the PROCESSOR RECORDS and ACTIVITY RECORDS to determine which activity to execute next. All of the details of the actions that occur to determine which activity processor Pj performs next is described in the above-referenced Tomlinson and Green application in conjunction with TABLE 2.
Now, in accordance with the present invention, each activity record also includes an activity depictor 61. Physically, each depictor 61 is a pre-written thirty-two bit word in a random access memory or a thirty-two bit register in a register array. And after processor Pj selects an activity to perform, it then examines depictor 61 of the selected activity to determine how the activity is implemented and how it can be accessed.
Preferably, depictor 61 contains a portion "b" which indicates whether the corresponding activity is implemented via a high level language program, a low level language microcode program, or a hardware logic unit. Also preferably, depictor 61 contains a portion "c" which indicates how that activity can be accessed.
During its examination of depictor 61, processor Pj performs all of the previously described steps of Figure 2. Thus it examines portion "b" to determine whether the selected activity is implemented by a software program 71, a firmware program 72, or a hardware logic unit 73. Then it utilizes the information in portion "c" to activate the activity as indicated at 34a-36b. In like manner, when processor Pi calls an activity which it performs, the processor then examines the depictor for that activity to determine whether it is implemented by a software program 74, a microcode program 75, or a hardware logic unit 76. Then, depending on that determination, it performs steps 34a-34b, 35a-35b, or 36a-36b of Figure 2.
Thus, the multiprocessor system of Figure 4 has both the versatility and changeability features of the single processor system of Figure 1. If high speed is the primary design goal of the Figure 4 system, then the frequently used activities should be implemented by low level language microcode programs or hardware logic units. And the performance of the multiprocessor system can be enhanced with time by substituting low level language programs or hardware logic units for various activities that previously had been implemented by high level language softv/are programs, and by providing nev/ activity depictors 61 in the activity's record 60 to reflect this change. Various preferred embodiments of the invention have now been described in detail. In addition, hov/ever, many changes and modifications can be made to these details without departing from the natϊure and spirit of the invention. Accordingly, the invention is not to be limited to said details but is defined by the appended claims.

Claims

WHAT IS CLAIMED IS:
1. A method of operating a data processing system including the steps of: storing, in a memory means, one software program that contains first and second instructions v/hich respectively call first and second activities; linking said first instruction via one type of depictor to another software program that performs said first activity, and linking said second instruction via a different type of depictor to a non-software means that performs said second activity; executing said one software program until either said first or second instruction is encountered; sensing whether said encountered instruction is linked to its activity by said one type or said different type depictor; and executing said another software program if said sensing step detects said one type depictor; and activating said non-software means if said sensing step detects said different type depictor.
2. A method according to Claim 1 wherein said activating step includes the substep of executing a microcode program in response to the sensing of said different type depictor to perform the activity called by said encountered instruction.
3. A method according to Claim 1 wherein said activating step includes the substep of sending parameters to a hardware logic means in response to the sensing of said different type depictor to perform the activity called by said encountered instruction.
4. A method according to Claim 1 wherein said linking step includes the substep of v/riting an address of said another software program into said one type depictor and writing an address of a microcode program into said different type depictor.
5. A method according to Claim 1 wherein said linking step includes the substep of writing an address of said another softv/are program into said one type depictor and writing an address of a hardware logic means into said different type depictor.
6. A method of operating a data processing system including the steps of: executing one software program until an instruction is encountered which calls an activity; sensing whether said encountered instruction is linked to the activity which it calls by a first type or a second type depictor; executing another software program for performing the called activity if said sensing step detects said first type depictor; and activating a non-software means for performing said called activity if said sensing step detects said second type depictor.
7. A method according to Claim 6 wherein said activating step includes the substep of executing a microcode program in response to the sensing of said second type depictor to perform the activity called by said encountered instruction.
8. A method according to Claim 6 wherein said activating step includes the substep of sending parameters to a hardware logic means in response to the sensing of said second type depictor to perform the activity called by said encountered instruction.
9. A method according to Claim 6 v/herein said sensing step is performed by executing a set of microcode commands.
10. A method according to Claim 6 and further including the step of repeating all of the steps of Claim 6 until the execution of said one software program is complete.
11. A data processing system comprising: a memory means containing a software program for a computer means to execute; said program including first and second instructions which respectively call first and second activities to perform predetermined tasks; first and second depictor means which respectively correspond to said first and second activities; said first depictor means including a first code and a pointer to another software program for performing said first activity; and said second depictor means including a second code and a pointer to a non-software means for performing said second activity.
PCT/US1984/001387 1983-09-01 1984-08-30 Processor with depictor-linked microcode and logic circuitry WO1985001136A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US528,461 1983-09-01
US06/528,461 US4862351A (en) 1983-09-01 1983-09-01 Method of executing called activities via depictor-linked low level language microcode, hardware logic, and high level language commands; and apparatus for same

Publications (1)

Publication Number Publication Date
WO1985001136A1 true WO1985001136A1 (en) 1985-03-14

Family

ID=24105774

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1984/001387 WO1985001136A1 (en) 1983-09-01 1984-08-30 Processor with depictor-linked microcode and logic circuitry

Country Status (5)

Country Link
US (1) US4862351A (en)
EP (1) EP0138352B1 (en)
CA (1) CA1214565A (en)
DE (1) DE3484725D1 (en)
WO (1) WO1985001136A1 (en)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU597980B2 (en) * 1986-05-30 1990-06-14 Honeywell Bull Inc. Apparatus and method for interprocessor communication
JPS63278145A (en) * 1987-05-11 1988-11-15 Nec Corp Parameter correspondence check system
US4961133A (en) * 1987-11-06 1990-10-02 Visystems, Inc. Method for providing a virtual execution environment on a target computer using a virtual software machine
WO1990004827A1 (en) * 1988-10-21 1990-05-03 A.C. Nielsen Company Space management system incorporating software operating environment
JPH02190930A (en) * 1988-12-29 1990-07-26 Internatl Business Mach Corp <Ibm> Software instruction executing apparatus
US5146593A (en) * 1989-03-06 1992-09-08 International Business Machines Corporation Procedure call interface
US5068821A (en) * 1989-03-27 1991-11-26 Ge Fanuc Automation North America, Inc. Bit processor with powers flow register switches control a function block processor for execution of the current command
GB2232514B (en) * 1989-04-24 1993-09-01 Yokogawa Electric Corp Programmable controller
EP0414624B1 (en) * 1989-08-24 1996-12-18 International Business Machines Corporation System for calling procedures on a remote network node
US5218699A (en) * 1989-08-24 1993-06-08 International Business Machines Corporation Remote procedure calls in heterogeneous systems
GB2242293A (en) * 1990-01-05 1991-09-25 Apple Computer Apparatus and method for dynamic linking of computer software components
JPH0831041B2 (en) * 1991-09-06 1996-03-27 インターナショナル・ビジネス・マシーンズ・コーポレイション Program condition processing method and computer system
US5319784A (en) * 1991-12-18 1994-06-07 International Business Machines Corp. System for automatic and selective compile-time installation of fastpath into program for calculation of function/procedure without executing the function/procedure
US5409200A (en) * 1992-03-05 1995-04-25 Zingher; Arthur R. Printed-circuit-like array of springs with non-linear force vs deflection
US5504920A (en) * 1994-05-16 1996-04-02 Compaq Computer Corporation Video driver system for communicating device specific primitive commands to multiple video controller types
US6542854B2 (en) * 1999-04-30 2003-04-01 Oracle Corporation Method and mechanism for profiling a system
GB2412192B (en) * 2004-03-18 2007-08-29 Advanced Risc Mach Ltd Function calling mechanism

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3983539A (en) * 1969-05-19 1976-09-28 Burroughs Corporation Polymorphic programmable units employing plural levels of sub-instruction sets
US4253145A (en) * 1978-12-26 1981-02-24 Honeywell Information Systems Inc. Hardware virtualizer for supporting recursive virtual computer systems on a host computer system
US4285036A (en) * 1978-02-02 1981-08-18 Tokyo Shibaura Denki Kabushiki Kaisha Data processing device using a subroutine call instruction
US4325121A (en) * 1978-11-17 1982-04-13 Motorola, Inc. Two-level control store for microprogrammed data processor
US4454579A (en) * 1981-09-11 1984-06-12 Data General Corporation System for performing call and return operations
US4454580A (en) * 1981-09-16 1984-06-12 International Business Machines Corporation Program call method and call instruction execution apparatus

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3478322A (en) * 1967-05-23 1969-11-11 Ibm Data processor employing electronically changeable control storage
US3614745A (en) * 1969-09-15 1971-10-19 Ibm Apparatus and method in a multiple operand stream computing system for identifying the specification of multitasks situations and controlling the execution thereof
US3631405A (en) * 1969-11-12 1971-12-28 Honeywell Inc Sharing of microprograms between processors
BE758813A (en) * 1969-11-28 1971-04-16 Burroughs Corp PROGRAM STRUCTURES FOR THE IMPLEMENTATION OF INFORMATION PROCESSING SYSTEMS COMMON TO HIGHER LEVEL PROGRAM LANGUAGES
US3735363A (en) * 1971-04-07 1973-05-22 Burroughs Corp Information processing system employing stored microprogrammed processors and access free field memories
BE795789A (en) * 1972-03-08 1973-06-18 Burroughs Corp MICROPROGRAM CONTAINING A MICRO-RECOVERY INSTRUCTION
US4131941A (en) * 1977-08-10 1978-12-26 Itek Corporation Linked microprogrammed plural processor system
US4149243A (en) * 1977-10-20 1979-04-10 International Business Machines Corporation Distributed control architecture with post and wait logic
US4257096A (en) * 1978-10-23 1981-03-17 International Business Machines Corporation Synchronous and conditional inter-program control apparatus for a computer system
US4463423A (en) * 1982-07-14 1984-07-31 Burroughs Corporation Method of transforming high level language statements into multiple lower level language instruction sets
US4484272A (en) * 1982-07-14 1984-11-20 Burroughs Corporation Digital computer for executing multiple instruction sets in a simultaneous-interleaved fashion

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3983539A (en) * 1969-05-19 1976-09-28 Burroughs Corporation Polymorphic programmable units employing plural levels of sub-instruction sets
US4285036A (en) * 1978-02-02 1981-08-18 Tokyo Shibaura Denki Kabushiki Kaisha Data processing device using a subroutine call instruction
US4325121A (en) * 1978-11-17 1982-04-13 Motorola, Inc. Two-level control store for microprogrammed data processor
US4253145A (en) * 1978-12-26 1981-02-24 Honeywell Information Systems Inc. Hardware virtualizer for supporting recursive virtual computer systems on a host computer system
US4454579A (en) * 1981-09-11 1984-06-12 Data General Corporation System for performing call and return operations
US4454580A (en) * 1981-09-16 1984-06-12 International Business Machines Corporation Program call method and call instruction execution apparatus

Also Published As

Publication number Publication date
DE3484725D1 (en) 1991-07-25
EP0138352B1 (en) 1991-06-19
EP0138352A3 (en) 1988-03-16
US4862351A (en) 1989-08-29
EP0138352A2 (en) 1985-04-24
CA1214565A (en) 1986-11-25

Similar Documents

Publication Publication Date Title
US4862351A (en) Method of executing called activities via depictor-linked low level language microcode, hardware logic, and high level language commands; and apparatus for same
US4323963A (en) Hardware interpretive mode microprocessor
JP3649470B2 (en) Data processing device
KR920001101B1 (en) Microprocessor and internal register device
JPH09231093A (en) Method and system for transfer of program control between two architectures
JPS6081650A (en) Method of synthesizing sequence and data processing system
US4279016A (en) Instruction pre-fetch microprocessor interrupt system
US3959774A (en) Processor which sequences externally of a central processor
US3698007A (en) Central processor unit having simulative interpretation capability
US3940744A (en) Self contained program loading apparatus
CA1289670C (en) Method and device to execute two instruction sequences in an order determined in advance
EP0660229B1 (en) Method and apparatus for modifying the contents of a register
US6502181B1 (en) Method and apparatus for an enhanced processor
US5893928A (en) Data movement apparatus and method
JPS61184643A (en) Starting control system for virtual computer
JP2003505758A (en) Processor architecture
JP2000029690A (en) Method and device for data processing
JP2727023B2 (en) Information processing device
JPS62120542A (en) Information processor
JP2591818B2 (en) Auxiliary function selection control method
RU2340933C2 (en) Method of computation process context switching during program interrupt or subroutine jump
JPS5835648A (en) Program execution controlling system
Hunt Interrupts
JPH05242050A (en) Processor device
JP2002544619A (en) Object-oriented processing using dedicated pointer memory

Legal Events

Date Code Title Description
AK Designated states

Designated state(s): JP