KR940007827B1 - Method of matching between user program and real time os - Google Patents

Abstract

The method for matching a real time operating system and a user program in an electronic switching system includes the steps of: storing register data in a task control table and increasing a call count; checking the user program, if a call has an error, generating error data, and if the call is normal, changing an address space; measuring a central processing unit occupancy of the user program and the operating system; calling the system with reference to a kernel primitive control block; collecting an execution time of a corresponding call; and returning to the user program, thereby raising the stability of the system.

Description

How to Match Operating System with User Program in Real Time Operating System

1 is a block diagram of a hardware system to which the present invention is applied.

2 is an overall flowchart to which the present invention is applied.

3 is a detailed flowchart of the present invention.

* Explanation of symbols for main parts of the drawings

1. MPH: Upper Processor Hardware 2. MPMA: Central Processing Board

3. PCCA: Communication control 4. DCCA: Redundant operation control

5. CPU: Central Processing Unit 6. MSIA: Auxiliary Memory Connection Board

7. IOIA: I / O device connection board 8. C, D, S-ch: Duplex communication channel

9.MPSbus: system main bus 12.VT: vector table

13. TCB: task control table 16. KPCB: kernel command control block

The present invention relates to a method of matching an operating system and a user program in a real-time operating system for a large-scale exchange.

The exchange software requires system call processing software to protect the system and to efficiently manage the kernel function provided by the operating system.

The conventional exchange system does not distinguish between the boundary between the exchange software (hereinafter referred to as a user program) and an operating system, and there is a risk that the system may be infringed depending on the user program.

Accordingly, an object of the present invention is to provide a method of matching an operating system and a user program by dividing a user program space and a system space to increase system safety, to protect against error usage of a system kernel function, and to reduce the burden of redundant design.

In order to achieve the above object, the present invention provides a central processing unit, a memory device having a vector table (VT), a task control table (TCB), and a kernel primitive control block (KPCB), and a multifunction peripheral control chip (MFC). Applied to the central control board (MPMA) of the electronic switching device provided therein, a method of matching the operating system and the user program, wherein when the user program is called, it stores the values of various registers in the TCB and increases the call count. Step 1 and, after performing the first step, a second step of checking the accuracy of the user program to output an error if there is an error call, and to change the address space according to the system mode change if it is normal, and the second step After performing the third step of setting the CPU share measurement start of the user program and the operating system, and after performing the third step, the system A fourth step of actually making a system call with reference to the call management table KPCB; and a fifth step of collecting the execution time of the call when the call is completed after performing the fourth step; and after performing the fifth step, The control is performed by the sixth step of transferring control to the user program and ending.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a configuration diagram of an upper processor (MPH) of an electronic switching system, which is hardware to which the present invention is applied, a is a configuration diagram of redundancy of all hardware, and b is a configuration diagram of an MPMA.

In the drawings, 1 is a central processing unit board (MPMA), 2 is a memory error control board (MECA), 3 is a redundant operation control board (DCCA), 4 is a mass storage board (MSIA), 5 is a processor communication control board (PCCA) 6, I / O Interface Board (IOIA), 7 is Central Processing Unit (CPU), 8 is Memory, 9 is Multifunction Peripheral Control Chip (MFP), 10 is Task Control Table (TCB), 11 is Vector Table ( VT) and 12 represent Kernel Primitive Control Block (KPCB), respectively.

As shown in the figure, the upper processor hardware (MPH) to which the present invention is applied has a central processing unit board (1) and a memory board (4), and the central processing unit board (1) on which the software of the present invention is mounted and operated. : MPMA), a processor communication control board (PCCA) responsible for communication with other processors under the control of the MPMA (1), and redundant operation control in charge of D-channel and C-channel control in connection with redundancy. Board (4: DCCA) is connected to the system main bus.

The central processing unit board (MPMA) 1 is basically a central processing unit 7 connected to the system bus, a memory 8, and a multifunction peripheral control chip 9 (hereinafter referred to as MFP) for performance measurement. The storage device 8 holds a vector table 11 (VT), a task control table 10 (TCB), and a KPCB 12. Such a configuration is a configuration of a general distributed system, but in order to be implemented with agitation software and an operating system, such a configuration should be provided.

The system call processing function is performed by the central processing unit 7 with the vector table 11: VT and the task control table 10: TCB in the storage device 8 as one of the junctions of the operating system.

2 is a process display diagram according to the present invention.

According to the present invention, when a user program requests a function of an operating system and prepares parameters for information transfer and enters the operating system kernel, the matching software supports the kernel primitive control block (hereinafter referred to as KPCB) by reference software. In addition, it analyzes any request among various managers in the kernel and executes the corresponding function by calling the kernel routine, generating the return value for the execution result, and saving and returning to the call. In the figure, PM is a process manager, TM is a time manager, MM is a memory manager, EM is an exception handling manager, and IOM is an input / output manager.

3 is a flow diagram of software that performs a system call function.

Automatic linking is achieved by running TRAP exception handling using the code TRAP instruction (MC68030 instruction), which calls a system primitive in the user program. When this TRAP exception processing occurs and is executed, the system call processing is executed with reference to the back table 12.

When execution starts, register values in various CPUs are stored in the TCB 10 (20), and the call count counter is incremented (21). It checks the correctness of the contents called by the user program (22), if it does not match, outputs a fault (23) and sends it back to the user program, and changes the address space so that the user system call can be executed in system mode when used correctly. (24).

If performance measurement is needed prior to making the call, set the performance measurement initiation (25), and refer to KPCB (12) for communication processing functions, multiprocessing functions, main memory device management functions, and kernel functions of the called operating system. Invoke device management, exception handling, file management, and database management system kernel functions (26), then perform performance measurement collection loops (27), and reduce various corruption information stored in the TCB (10). (28) After, the user program is returned.

4 is a process flow diagram of the performance measurement start step (25 in FIG. 3).

Investigate whether there was a request for performance measurement (30), and return immediately if there was no request, and if so, set the kernel to occupy the CPU since it is now executed in the operating system kernel and a specific kernel, and read the performance measurement timer value. (31). Investigate whether the detailed performance measurement is required (32), and if there is a request, examine whether or not the user program is already loaded in the storage device (33). If loading, the detailed performance measurement cannot be performed. The start time is read from the timer (34) to measure the time until return from the system call request.

5 is a process flow diagram of a performance measurement step.

Investigate whether there was a performance measurement request (40), return immediately if there was no request, and if so, set the previous program mode to occupy the CPU as it is now performed in the program before the call (41), and set the performance measurement timer value It reads (42) and changes the CPU occupancy time of the operating system (43). In addition, if the request is made (44), if there is a request, it checks whether the user program is already loaded into the storage device (45), and if the loading is performed, the detailed performance measurement is impossible, and if it is not loaded, the system returns immediately. The time from the call to return is read from the timer and changed (46).

Therefore, the present invention carried out by the above processing procedure can increase the stability of the system, and the effect of reducing the burden of protection and enormous redundancy design for the error use of the system kernel function.

Claims (3)

Storage unit 8 having a central processing unit 7, a vector table 11 (VT), a task control table (hereinafter referred to as TCB) and a kernel primitive control block (hereinafter referred to as KPCB). In the method of matching the operating system and the user program applied to the central processing unit board (1; MPMA) of the electronic switchboard having a multi-function peripheral control chip (9; MFC), various registers are called when called from the user program. Is stored in the TCB 10 and the call count is increased, and after performing the first steps 20 and 21, and after performing the first steps 20 and 21, the accuracy of the user program is examined to determine whether there is an error call. If there is an error, if it is normal, if it is normal, the second multi-system 22 to 24 that changes the address space according to the system mode change, and after performing the second stage 22 to 24, start the CPU occupancy measurement of the user program and the operating system. 3rd stage to set (25), after performing the third step (25), the fourth step (26) of actually performing the system call with reference to the system call management table KPCB (12), and after performing the fourth step (26) When the call is finished, the fifth step (27) of collecting the execution time of the call, and after performing the fifth step (27), the sixth step (28) to transfer the control to the user program and terminate with reference to the TCB (10) Operating system and user program, characterized in that performed by The method of claim 1, wherein the third step (25) checks whether there is a performance measurement request and returns immediately if there is no request, and if so, sets the kernel to occupy the CPU and reads the performance measurement timer value. Putting steps 30 and 31, and after performing the steps 30 and 31, checking whether a detailed performance measurement is required and checking if a user program is already loaded into the storage device if there is a request, or returning to step 32 , 33), and after performing the steps 32 and 33, detailed performance measurement is impossible if loading, and returns immediately. If not, the timer is timed to measure the time until returning from the system call request. Operating system and user program, characterized in that performed by step (34). 5. The method of claim 1, wherein the fifth step 27 examines whether there is a performance measurement request and returns immediately if there is no request, and sets the previous program mode to occupy the CPU if there is a request (40, 41). And after performing the steps 40 and 41, changing the CPU occupancy time of the operating system by reading the performance measurement timer value, and performing the detailed performance measurement after performing the steps 42 and 43. Check if the user program has already been loaded into the storage device, and if so, return to steps 44 and 45, and after performing the steps 44 and 45, which user program is stored in the storage device. If it is loading at, detailed performance measurement is impossible, so it returns immediately, and is not loading, and by changing the time from the timer to returning from the system call, change and return by step 46. How to match the operating system and the user program, characterized in that line.