JPH012138A - Underflow processing method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an underflow processing method for processing underflow that occurs during arithmetic processing within a computer.

[Conventional technology]

FIG. 2 is a flowchart for explaining the operation of the conventional underflow processing method. Underflow refers to overflow of the lower digits of an exponent, and is a condition that occurs when a certain amount smaller than the minimum digit that can be represented (memorized) is obtained as a result of calculation, and in the case of floating point numbers. A large negative exponent that cannot be expressed will cause this lower digit confusion. In the figure, 1 indicates a main memory that stores information necessary for arithmetic processing, etc. This main memory 1 contains a Tameno user program 1a that performs processing in response to user requests, as well as program execution control, input/output control, etc. An operating system 1d is provided. lb and lc in the user program la indicate consecutive executable statements. Further, 2 indicates a central processing unit (CPU mechanism) that performs predetermined calculations, etc., and 2a indicates an arithmetic processing mechanism of this central processing unit 2.

Next, the operation will be explained. It is assumed that each executable statement 1b, lc of the user program 1a is sequentially processed by the central processing unit 2. Contents of executable statement 1b r c = a /
b When processing J by the central processing unit 2, first ra
/bJ processing is performed by the arithmetic processing mechanism 2a inside the central processing unit 2.
Executed by If the execution result is an underflow, that is, a small value that cannot be represented by the data representation in the computer, the process moves to a program check process that checks the user program 1a, passes control to the operating system 1d on the main memory la side, and performs a program check. A processing routine is activated and a program stop process is performed to abnormally terminate the user program 1a.

[Problem that the invention seeks to solve]

However, in such conventional underflow processing methods, when an underflow occurs and the program stops as described above, the program cannot be re-executed after the stop.
No processing results were obtained after undefflow. For this reason, the programmer must create a program while paying attention to checking the final calculation data before the calculation to prevent underflow from occurring, which places a heavy burden on program development. There was a problem in that the productivity of the program was reduced by including such check logic when creating the program.

This invention was made to solve the above-mentioned problems, and even if an underflow occurs as a result of arithmetic processing, it is possible to continue processing without performing program check processing.
The purpose of this invention is to provide an underflow processing method that can reduce the burden on programmers during program development and improve program productivity.

[Means for solving problems]

In the underflow processing method according to the present invention, when an underflow occurs in the arithmetic processing process in the computer, the program 1a checks whether the specified mode from the computer is the first mode that avoids the underflow and continues processing. The process S2) of the underflow processing mechanism 2b determines whether the mode is the second mode in which the stop processing is performed, and if it is determined that the mode is the first mode, the process of the underflow processing mechanism 2b sets a predetermined minimum value to the calculation result. S3), avoiding the underflow and continuing the processing immediately after the underflow occurs;
If it is determined that the second mode is selected, a program check processing routine is started and the program 1a is stopped.

[Effect]

The program 1a is executed sequentially, arithmetic processing is performed,
When an underflow occurs, it is determined whether the designated mode is the first mode or the second mode, and when the designated mode is the first mode, the minimum value is set as the calculation result, thereby avoiding the underflow.

Therefore, the processing immediately after the underflow occurs is continued. Further, in the second mode, a program check processing routine is activated and the program 1a is stopped.

[Embodiments of the invention]

Hereinafter, one embodiment of the present invention will be described based on the drawings.

FIG. 1 is a flowchart for explaining the second underflow processing method according to the embodiment of the present invention. In FIG. 1, components corresponding to those shown in FIG. 2 are given the same reference numerals, and their explanations will be omitted. In Figure 1, 2
b indicates an underflow processing mechanism provided in the central processing unit 2, processing S1 of this processing mechanism 2b is a process for determining whether or not there is an underflow, and process S2 is a process for determining what the designated mode is. , Process S3 indicates a process of setting a minimum value to the calculation result.

Next, the operation will be explained. First, the user specifies in advance the first mode or the second mode when an underflow occurs when a computer processing system is generated. These modes can be specified arbitrarily at the time of system generation. The first mode refers to a mode in which underflow is avoided and continuation processing is performed, and the second mode refers to a mode in which stop processing is performed to check the program. The first of these.
The second mode information is stored in the central processing unit 2 and can be referenced for each calculation. With the mode specified in this way, when the user program 1a is sequentially processed in the central processing unit 2, the content r c = a / b J of the executable statement 1b of the user program 1a is stored in the central processing bag W.
2, and the result is given to the underflow processing mechanism 2b (step Nl).
). The underflow processing mechanism 2b determines whether an underflow has occurred in the above result in process S1, and if no underflow has occurred, the content of the executable statement IC of the user program 1a rWRITE (returns to CI J and continues the process)
Step N2).

If an underflow occurs, from process S1 to process S
2 (Step N3), it is determined whether the mode specified by the system is the first mode indicating continued processing or the second mode indicating stop processing, and if it is determined to be the first mode (Step N4 ), a predetermined minimum value (c = O) is added to the calculation result in step S3, and the underflow is avoided and the processing immediately after the underflow occurs, that is, the processing of the executable statement IC of the user program la is continued. (
Step N5). Further, if it is determined that the second mode is selected in the process S2 (step N6), the process moves from process S2 to the control of the operating system 1d, a program check process routine is started, a program check process is performed in the same way as in the conventional system, and the user program 1a is executed (step N7).
).

Note that although the above embodiment shows the case of underflow in division, other functional operations such as multiplication and exponents may be used, and the same effect can be achieved in this case as well. Further, in the above embodiment, the underflow processing is performed by the central processing unit, but it may be performed by the program check processing routine of the operating system.

〔Effect of the invention〕

As described above, according to the present invention, when an underflow occurs in the arithmetic processing process, it is determined whether the mode is the first mode for continuing processing or the second mode for stopping processing, and it is determined that the mode is the first mode. Then, a predetermined minimum value is set for the calculation result to avoid underflow and continue processing immediately after the underflow occurs, and when it is determined that the second mode is selected, a program check processing routine is started and the program is stopped. As a result, if an underflow occurs during program execution, you can select ¥a continuation processing or stop processing as necessary, and this allows you to select continuation processing or stop processing without being aware of the underflow. It is possible to create a program, thereby reducing the burden during program development and improving the productivity of the program.

[Brief explanation of drawings]

FIG. 1 is a flowchart for explaining the operation of an underflow processing method according to an embodiment of the present invention, and FIG. 2 is a flowchart for explaining the operation of a conventional underflow processing method. 1a...User program, 2b...
Underflow processing mechanism. Agent Masuo Ookimi (and 2 others) 1st year high school student

Claims (1)

[Claims] When an underflow occurs in the arithmetic processing process within the computer, the mode specified by the computer is the first mode, which avoids the underflow and continues processing, or the second mode, which halts the program to check the program. If it is determined that the mode is the first mode, a predetermined minimum value is set in the calculation result, the underflow is avoided and the process immediately after the underflow is continued, and the mode is determined to be the second mode. Then, the underflow processing method starts a program check processing routine and stops the program.