JPH02129724A - Program executing system - Google Patents

Abstract

PURPOSE:To execute the process without a waiting time for loading the process by dividing one function program into several independent processes and storing load information of each process, placing the head process on a main memory, and also, placing other processes in an auxiliary storage device. CONSTITUTION:A DMA device 22 for transferring a process to a main memory 23 from an auxiliary storage device 24 without giving a burden to a CPU 21 is provided. Also, this system is constituted so that this program is divided into several independent processes in order to execute one function program, a head process in which load information of its process is stored is placed on the main memory 23, and loading of the process can be executed in parallel to execution of the process. In such a way, a large program constituted of several independent processes is executed without a waiting time for loading the process, and the whole processing is executed quickly.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to a program execution method used when a large program is executed in a computer, and particularly to a program execution method for speeding up the execution of the entire program.

[Conventional technology]

Traditionally, in systems with large programs,
Due to finite physical resources, it was impossible to store all programs in main memory. For this reason, when processing a large program, the program is divided into several independent processes and the divided processes are sequentially processed to execute one large functional program. Here, each divided process is executed after being loaded onto the main memory from an auxiliary storage device such as an HD (hard disk) or FD (floppy disk) by a process generation system call. FIG. 3 is a diagram for explaining the loading of this divided process, where ■ is the currently executing process, and 2 is the process that is executed after being loaded into the main memory by the system call issued by this process l. It is a process of Next, the operation of the program execution method shown in FIG. 3 will be explained using the flowchart shown in FIG. 4. When started, first, in step STI, process 1 issues a system call. Next, step ST2
, the system software O3 executes memory area securing processing for loading process 2 stored in the auxiliary storage device into the main memory. Next, in step ST3, the file in which process 2 is stored is read from the auxiliary storage device onto the main memory in which the memory area has been previously secured. In this way, once process 2 is stored in main memory, step S
Process 2 is executed at T4. In this way, each process of the program composed of a plurality of processes is executed in sequence, thereby completing the processing of the program. (Problems to be Solved by the Invention) Since the conventional program execution method is configured as described above, when starting one process to the next, it becomes necessary to load the process file into the main memory. As a result, it takes time to start a process, and there is a problem that a process that executes a large program cannot be started quickly.As an approximation technique,
There is ``Analysis Multitasking'' (written by Shigeru Sasaki, published in January 1960, CQ Publishing Co., Ltd.). This invention was made to solve the above-mentioned problems, and when processing a large program made up of multiple processes, it speeds up the transition from one process to the next and improves the overall program performance. The purpose is to obtain a program execution method that speeds up execution.

[Means to solve the problem]

The program loading method according to the present invention uses a direct memory access device (hereinafter referred to as a DMA device) that transfers a program from an auxiliary storage device to a main memory without placing a burden on a central processing unit (hereinafter referred to as a CPU). ), one functional program is divided into several independent processes, the load information of each process is stored, and the first process with the first logic group of the program is placed in the main memory, and the other processes are placed in the auxiliary storage, and while the first process is executed, other processes are loaded from the auxiliary storage onto the main memory.

[Effect]

In this invention, when a program for realizing one function is activated, the first process is executed. Since the load order of the processes is stored in the first process, the next process is loaded onto the main memory by DMA transfer without placing a burden on the CPU. The CPU executes processes regardless of process loading processing, and when there is a system call to start a process, the process already in the main memory is executed without waiting time for loading.

[Embodiments of the invention]

An embodiment of the present invention will be described below with reference to the drawings. 1st
In the figure, 21 is a CPU that executes the process, and 22 is a DMA device that executes DMA transfer between the main memory 23 and the auxiliary storage device 24 in which the process is stored. The main memory 23 contains programs for realizing various functions, but each program is divided into a number of independent processes due to the main memory capacity. processes are located in main memory. Furthermore, the load order of the processes and the logic of the program are described in the first process. And other processes are auxiliary storage device 2
It is stored in 4. In FIG. 1, program A1 stored in the main memory 23 represents the first process located at the beginning of program A, and program Bl represents the first process of program B. Further, program Ai represents the first process of program AO, and program An represents the first process of program AO.
It represents the nth process. Next, the operation will be explained. FIG. 2 is a flowchart showing the operations of the CPU 21 and DMA device 22 when program A is executed. First, when started, in step 5TII,
A command to start program A is given to the CPU 21. Next, in step 5T12, process 1 in the program is executed, and in order to load processes other than process 1 of program A onto the main memory 23, a DMA transfer is generated. On the other hand, the DMA device 22 that received this DMA transfer command,
In step 5T13, the next process 2 to the program is loaded onto main memory. During this loading, as shown in step 5T14, the CPU 21 works independently of the DMA device 22 to execute program processing. On the other hand, when the DMA device 22 finishes loading the process 2 into the main memory 23, in step 5T15, the DMA device 22
The A device 22 notifies the CPU 21 of the completion of the transfer process using interrupt processing. CP that received an interrupt
U21 checks in step 5T16 whether there is a process that must be loaded next. Here, if the determination in step 5T16 is YES, the process moves to step 5T17, and a transfer command is sent to the DMA device 22. The DMA device 22 receives the instruction from the CPU 21 and loads the process to the program in step 1. On the other hand, even while the DMA device 22 is loading the process of program A, the CPU 21 performs step 5T.
As shown in 19, the process of the program is being executed. When the process loading to the main memory 23 is completed, the DMA device 22 uses interrupt processing to notify the CPU 21 that the transfer processing has been completed, as shown in step ST20. When the CPtJ21 completes the process in step 5T19, it executes the process to proceed to step 5T16. In step 5T16, the CPU 21 performs eight rounds to determine whether there are any processes that still need to be loaded.
If not, the process moves to step 5T21 and continues execution of program A. When a process in the program requests execution by the next process, processing for the next process that is already on the main memory 23 is executed without waiting time for loading. Therefore, in such a state, when a program for realizing one function is activated, the first process is executed. Since this first process remembers the load order of the processes, the next process is
The A transfer allows the data to be loaded onto the main memory without placing any burden on the CPU. The CPU executes processes regardless of process loading, and when there is a system call to start a process, it is possible to execute processes already in main memory without waiting for loading. Become. In the above embodiment, the case where the interrupt processing is performed by hardware has been described, but the same effect can be obtained even if the interrupt processing is performed by software.

【Effect of the invention】

As described above, according to the present invention, D
In addition to providing an MA device, in order to execute one functional program, this program is divided into several independent processes, and the first process that stores the load information of the process is placed in the main memory, so that the load of the process can be reduced. Because it is configured so that it can be executed in parallel with the execution of processes, it is possible to execute large programs consisting of several independent processes without waiting for processes to load. This has the excellent effect of speeding up the overall processing.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a program execution method according to an embodiment of the present invention, FIG. 2 is a flow chart showing the operation of the program execution method shown in FIG. 1, and FIG. 3 is a block diagram showing a conventional program execution method. , FIG. 4 is a flow chart showing the operation of the program execution method shown in FIG. 21 is a central processing unit (CPU), 22 is a direct memory access device (DMA device), 23 is a main memory, 2
4 is an auxiliary storage device. In addition, in the figures, the same reference numerals indicate the same or equivalent parts.

Claims (1)

[Claims] In addition to providing a direct memory access device that transfers processes from auxiliary storage to main memory without placing a burden on the central processing unit, a large program for executing a single functional program can be divided into several independent processes. In a program management system that stores the load information of the process and places the first process having the first logic group of the program in the main memory and the other processes in the auxiliary storage, when the program is started, The direct memory access device loads the process located in the auxiliary storage device onto the main memory in parallel with the execution of the first process located in the main memory by the central processing unit. Program execution method.