JPH0528053A - Method for memory check of program during system operation - Google Patents

Abstract

(57) [Summary] [Purpose] In a system that does not have a hard memory check function or memory protect function, detect memory damage early, prevent unnecessary malfunction, save damage information, and analyze the cause of abnormality. The purpose is to facilitate. According to the memory check method of the present invention, an FD 1 in which a load module program is stored, a disk 2 having a library, etc., and a load module
Library-3 and A program 4,9 to be checked
, B program 5, 10 etc. and check program 7,
8, loader programs 6 and 11, main memory 12,
It is composed of the CPU 13.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a program memory check method for early detection of destruction of a code portion of a load module program (hereinafter referred to as a program) during operation of software. .. Note that the system applied here is a system that does not have a hard memory check function or memory protect function during system operation, and the target program has a separate code section and data section. Be present.

[0002]

2. Description of the Related Art In a conventional system which does not have a hard memory protection function, if the code part of a program is destroyed during operation, it will not be detected at that time and the information in the main memory or disk will be lost after the system goes down. Was analyzed and detected.

[0003]

However, in the prior art, when the code part of the program is destroyed, an abnormal signal is not detected and the detection of malfunction is delayed, which hinders the operation of the system. In addition, the program in which the code portion is destroyed may be executed, and other memory may be destroyed. In this case, there is a problem that it is impossible to elucidate the cause of the first abnormality even if the information in the main memory or the disk is analyzed after the system goes down.

The present invention has been made in view of the above points, and detects an abnormality early so as to minimize the influence of the above problems, prevents an unnecessary malfunction after being destroyed, and prevents a system from being downed. An object of the present invention is to provide a memory check method for a program during system operation for facilitating the analysis of the cause of an abnormality.

[0005]

In order to solve the above problems, the present invention provides a check program shown in FIG. First, the A program to be checked is read from the floppy disk FD1 into the load module library-3 through the main memory by the loader. At this time
The datum calculates the total sum of the code part of the program and stores it as a check code in the rear part of the code part. The operation is shown in the flow chart in FIG. Other programs are similarly loaded into the load module library-3. The load module library may be in the main memory 12.

When the program is executed as a task, the program to be checked is called by the main memory 12 and executed. When the program to be executed enters the waiting state, when there is no other task to be executed by the CPU,
Execute the check program in the main memory as a calling task. FIG. 3 shows a flowchart of the operation. At this time, the check code of the program to be checked is calculated and compared with the stored check code. If the values are not equal, it is judged that the program to be checked is destroyed and the system is stopped. Then, the abnormal information such as the destroyed address and the code is saved. Perform the above work for all programs to be checked.
The check program itself will also be checked.

[0007]

According to the present invention, the program to be checked is
When first loaded into the library, the check code is calculated and saved as the correct check code. When this program is placed in a waiting state while being executed as a task, the check code is calculated again and both check codes are compared, so that it is possible to know whether or not the program was destroyed during this period.

Since this check program is executed as a task having the lowest priority, it does not adversely affect the system performance. As a cause of destruction, there are cases where it is destroyed due to a hard failure or software-related runaway of its own task or runaway of another task. For example, the A program 9 in FIG. 1 may destroy itself, and the B program 10
Although it may be destroyed by the check, even if it is destroyed by any cause, it can be detected by the check code.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a diagram showing the state of the system at the time of checking according to the present invention, FIG. 2 is a flowchart for calculating and adding a check code when reading a program to be checked into a library, and FIG. 3 is a flowchart of the check program. −
It's a chart.

The program is read from the floppy disk FD1 through the main memory 12 into the load module library-3 of the disk 2. The floppy disk FD1 may be on the disk 2 or another external memory, and the load module library 3 may be in the main memory 12. The load module library-3 stores many programs to be checked, such as A program 4, B program 5, loader program 6 and check program 7.

These programs have a structure in which a code part and a data part are separated and a check code is stored in the rear part of the code part. When these programs are executed, the areas 8, 9, 1 required on the main memory 12 are required.
0, 11 are secured, the load module of the disk 2
It is called from the library-3 to the main memory 12 by the loader program 11, the necessary environmental conditions are prepared, and the CPU 13 executes it as a task. Each task is executed in descending order of priority. The task that executes the check program has the lowest priority. This makes it possible to check the execution of other tasks without disturbing them.

The memory check according to the present invention is performed by the check program 7 in the state described in detail above, and the addition of the first check code is performed by the loader program 6. The operation of adding a check code will be described with reference to FIG. When the program is read into the load module library, the head address of the code section is taken out (step ST1). Next, the final address of the code portion is taken out (step ST2). From the beginning to the end of the code section is added (step ST3) and stored in a predetermined place as a check code (step ST4).

Next, the operation of the check program will be described with reference to FIG. The program index is initialized (step ST11). The leading address and the ending address of the code portion of the program to be checked designated by the index are taken out (step ST12). The check code is calculated in the same manner as in FIG. 2 (step ST1
3) Compare with the stored check code (step ST14). If the same, check whether the current program is the final program (step ST1).
5) If it is not the final program, the index is advanced to the next (step 16), and if it is the final program, the process is repeated from step ST1. If the check codes are not equal in step ST4, it is considered to be abnormal, and abnormal information such as the destroyed address and code is saved and the system is stopped (step ST17).

[0014]

As described above in detail, according to the present invention, the following effects can be expected. The destruction of the code part of the program can be detected early. It is possible to prevent malfunction due to execution of a destroyed program. Damaged information such as destroyed addresses and codes is saved, and the cause analysis after the system is stopped is facilitated in order to minimize unnecessary malfunctions.

[Brief description of drawings]

FIG. 1 is a diagram showing a state of a system during a memory check according to the present invention.

FIG. 2 is a flow chart of check code calculation used in the present invention.
Chart.

FIG. 3 is a flowchart of a memory check program according to the present invention.

[Explanation of symbols]

 1 Floppy Disk 2 Disk 3 Load Module Library-4 A Program 5 B Program 6 Loader Program 7 Check Program 8 Check Program 9 A Program 10 B Program 11 Loader 12 Main Memory 13 CPU

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification number Office reference number FI technical display location G06F 11/30 320 C 8725-5B

Claims (1)

Claim: What is claimed is: 1. A method for checking a memory of a computer system which does not have a hard memory check function or a memory protect function while the system is operating, wherein a load module program is stored in a library. At the time of reading, the check code of the code part of the load module program is calculated and saved, and the target task is used by utilizing the free time during which the CPU processes the task. Load module program code
The check code of the code section is calculated, the result is compared with the stored check code, and if the value is different, it is determined that the load module program is destroyed,
A memory check method characterized by stopping the system, preventing unnecessary malfunctions, storing abnormal information, and facilitating failure analysis.