JPH01169641A - Soft error detecting system - Google Patents

Abstract

PURPOSE:To facilitate the discovery of the error of input data and a program, and to improve the development and maintenance efficiency of the program by providing a means to compare the value of the data, a means to detect the error from the comparing result and a means to output the error information. CONSTITUTION:When a system investigates the relation between the data and the relation is not satisfied, the value of the data, the relation between the data and the position in the user program are outputted and the checking result is returned to a user program. Namely, the user describes the range obtained by the value of the variable and the size relation of the variable and the variable in a program 11, and at the time of executing the program, the system checks these relations by a checking instruction 12. Thus, the user knows the cause and place of the error of the logic of the input data and the program, facilitates the error countermeasure and can achieve the efficiency improvement of the program development and maintenance. By checking at the time of the execution, the reliability of the calculation result of the program can be improved when these checking results are all normal.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to software development and maintenance systems, and in particular, to software error detection suitable for developing rough 1 to software, reducing maintenance man-hours, shortening period, and improving reliability. This invention relates to a soft error detection method that automatically detects and clearly indicates the location of occurrence.

[Conventional technology]

In the conventional method, as described in "Introduction to rPAscAL Programming" by Masayuki Okamoto, Kogaku Tosho Co., Ltd., p. 77, the hardware and operating system (O8) perform a check when the value of a variable exceeds a specified range. Also, HI TAC manual full 8080-3-25
7-40. As described in Pi 34, there was a means to display traceback information such as the location where an error occurred. The same applies to the range of array subscripts.

[Problem that the invention seeks to solve]

The above conventional technology does not allow the user to write detailed check items anywhere in the program. , it was necessary to explicitly describe the traceback information output section.

The purpose of the present invention is to allow the user to write check items in the program to display data values and traceback information at the time of an error, and use this to improve user program development, maintenance efficiency, and reliability. With the goal.

[Means for solving problems]

The above purpose is for a user to send and receive data to be checked and relationships between data to a system consisting of hardware and an operating system (O8), and for the system to use this information to examine relationships between data. This is accomplished by outputting the data value, the relationship between the data, and the position in the user program if the relationship is not satisfied, and returning the check result to the user program.

[Effect]

The user writes in the program the range of values that variables can take, the magnitude relationships between variables, etc., and the system checks the above relationships when the program is executed. As a result, the user can learn the causes and locations of errors in input data and program logic, easily take countermeasures against errors, and improve the efficiency of program development and maintenance. Furthermore, by performing checks during execution, if all the check results are normal, the reliability of the calculation results of the program is improved.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is an example of a user program. 11 is the subroutine name, 12 is the check information to be passed to the system 13.1
4 and variables 1-5 that receive the check results are shown below.

Reference numeral 16 indicates a branch instruction based on the check result, and 17 indicates an example of the instruction number of the user program. 14 indicates that the first data to be checked is ■ and the second data is J. Reference numeral 13 indicates a relationship between data, and indicates a relationship in which the value of the first data J is greater than the value of the second data J. When a user program 11 is executed and it is its turn to execute an instruction 12, relationships 13 and data 14 are passed to the system and checked.

Figure 2 shows the check flowchart 1 in the system.
shows. Control is transferred to user program 11, and instruction 12
When executed, the data relationship is first checked in step 21, and if the result is normal, that is, in FIG.
In the example, if I>J, in 23, CHECK
The value of is set to O. On the other hand, if I>J, control is transferred to 24, and the system prints relation 13 and data 14 from 1 to 25. At 25, the position in the user program, that is, the program name 5UBI of 11 and the instruction number in the program. 2 to pudding 1-. In addition, program 5U
The location of the instruction in the main program that references the BI is also printed. Then, after setting the value of CHECK to 1, return to the user program 11, and return to the left side E in instruction 12.
Assign the CHECK value O or 1 to RR. In the user program, the value of the variable ERR is checked as in the example of instruction 16, and the flow of the program is changed using the result of the check.

FIG. 3 31 is a simplified example of the example shown in FIG. 1 12 .

Checks required in the user program include a≦X.
The form ≦b frequently appears.

In the example of FIG. 3, relationship 13 in FIG. 12 is omitted,
By performing a check exclusively for the form a≦X≦b, the effort of analyzing the relationship is omitted, and in this case, by using dedicated hardware or software, the speed of the check can be increased.

FIG. 4 41 shows an example of a check on a set rather than on a relation of values such as magnitude, equality, unequal, etc. In this example, it is checked whether a value X is included in a set A having N elements. If X is included, it is considered normal and the ICHECK value becomes O. Otherwise, the value of X, the value of A, and the position in the program are printed and the value of ICHECK becomes 1. Become.

According to this embodiment, it is possible to clearly know the validity of the values between variables that take scalar quantities, check the validity of a set, and clearly know the value at the time of an error and the position in the program. Furthermore, it is possible to speed up checking of the type a≦X≦b, which frequently appears.

By writing these check commands in a program, it becomes possible to detect errors that exist in input data or program logic during program execution. This makes it possible to improve efficiency during program development and maintenance, and improve the reliability of completed programs.

In particular, when a single program is developed by multiple programmers, by thoroughly checking data created by others, the cause of an error becomes clear, and development efficiency is significantly improved.

〔Effect of the invention〕

According to the present invention, the programmer describes the contents he or she wants to check, that is, the relationship between data and the data, and the system executes those checks, and in the event of an error, prints the contents of the data and its position in the program. By using this method, a programmer can write the relationships between data everywhere, and errors in input data or programs can be easily discovered, which has the effect of improving the efficiency of program development and maintenance. Furthermore, it has the effect of increasing the reliability of calculation results obtained as a result of performing strict runtime checks.

[Brief explanation of the drawing]

Figure 1 is a description display diagram of a user program which is an embodiment of the present invention, Figure 2 is a flowchart of a system that executes the check command in Figure 1, and Figure 3 frequently appears in the check in Figure 1. FIG. 4 is an explanatory diagram that briefly describes an example, and FIG. 4 is an explanatory diagram that describes set operations in the check contents of FIG. 1. 11... User program, 12... Check command. 24.25 Output content when error is detected.

Claims (1)

[Claims] 1. A computer system consisting of an instruction storage device and an instruction execution device, characterized by being provided with means for comparing data values, means for detecting errors from the comparison results, and means for outputting error information. Soft error detection method.