JPH06324914A - Runaway detecting method for computer - Google Patents

Abstract

PURPOSE:To detect the runaway of a computer without using any watchdog timer. CONSTITUTION:A check program, namely, numerical values of a RAM 15 are read out, deciding whether the read numerical values are arranged in a certain fixed order or not and the check program for writing new numerical values is written at the plural positions of a program stored in a program memory 4. When a CPU 3 normally executes the program of the program memory 4, according to the check program, the numerical values are written in the RAM 15 successively from 1, 2 and 3. The CPU 3 confirms that the numerical values of the RAM 15 are regularly changed in the ascending order by the decision processing of the check program. When it is detected that this regularity is broken, the CPU 3 regards the generation of runaway in the program executed by itself and starts a program for dealing with the runaway.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to computers, and in particular to
The present invention relates to a computer runaway detection method suitable for a microcomputer.

[0002]

2. Description of the Related Art Conventionally, a device using a microcomputer, for example, a programmable controller is equipped with a function of detecting a runaway of a microcomputer (an uncertain order of execution of programs), and this runaway is detected to detect the entire device. Control to stop. The cause of the runaway of the microcomputer is caused by a pattern short circuit, release of the printed board, defective soldering of parts, external noise, or the like. A method using a watchdog timer is well known as a method for detecting runaway in general computers, not limited to microcomputers. That is, the watchdog timer is reset (count value is initialized) from the computer side at a substantially constant cycle. Therefore, the watchdog timer does not expire while the computer is operating normally. However, if the microcomputer runs out of control for some reason, the reset period of the watchdog timer changes, so the watchdog timer times out and the runaway of the computer is detected. Normal,
In order to reset the watchdog timer on the computer side, in a loop program as shown in FIG. 5, a program for generating a reset signal (denoted as timer refresh) is inserted in advance at a plurality of points in the execution program.

[0003]

However, the interval for installing the program for timer refresh must be set so that the execution interval time of this program is within the time-up time of the watchdog timer,
There was a problem that the program contents on the installed side were restricted. Further, when a runaway occurs after the completion of the third program in FIG. 5 and the execution program is transferred to the sixth program, the watchdog timer does not time up and the runaway cannot be detected by the conventional method.

Therefore, in view of the above points, a first object of the present invention is to provide a computer runaway detection method capable of asynchronously detecting a computer runaway without using a watchdog timer.

A second object of the present invention is to provide a runaway detecting method for a computer which further improves the runaway detecting accuracy of the runaway detecting method using a watchdog timer.

[0006]

In order to achieve such an object, the invention of claim 1 is a computer runaway detecting method for detecting that the execution sequence of a computer program has run out of control. Corresponding to a plurality of specific positions in the position and the execution order of the computer at those positions, when different identification information and the connection order of the identification information are determined in advance, and when the computer starts the execution of the program. , The first identification information in the connection order is stored in the memory as initial information, and each time the execution position of the program of the computer reaches any one specific position of the specific plurality of positions, The computer matches the identification information stored in the memory with the identification information corresponding to the specific position. If a match is obtained, it is regarded as normal operation, and the computer writes the identification information of the specific position corresponding to the next connection sequence of the current specific position in the memory to determine the mismatch. Is obtained, the computer considers that a runaway has occurred.

The invention of claim 2 is characterized in that, when the inconsistency determination is obtained, the computer activates a predetermined program for coping with the occurrence of runaway.

[0008]

According to the first aspect of the invention, when the execution order of the programs is normal, the order of the identification information stored in the memory is in a predetermined order, but when a runaway occurs,
The order is not predetermined and this is detected by the comparison judgment of the identification information of the computer.

According to the second aspect of the present invention, when a runaway is detected, a program for coping with the runaway is activated, that is, by setting the read start address of the program, the runaway abnormality can be automatically canceled even in a computer that is running out of control.

[0010]

Embodiments of the present invention will now be described in detail with reference to the drawings.

<First Embodiment> FIG. 1 shows the system configuration of a programmable controller system to which the present invention is applied.

In FIG. 1, a one-chip microcomputer 2 used in the programmable controller 1
Has a central processing unit (CPU) 3, a program memory 4, a random access memory (RAM) 5, an input / output port 8 and a bus expansion buffer 10. The monitoring timer 6 is used in the second embodiment described later.
These circuits are commonly connected to the internal bus 9. CP
U3 controls the system according to the system program in the program memory 4. In the program memory 4, in addition to the system program executed by the CPU 3, fixed information used for calculation is stored.

Input / output information for the CPU 3 is temporarily stored in the RAM 5. The input / output port 8 transfers information input / output from an electronic device to be controlled by the programmable controller system 1 to the CPU 3 via the output circuit 11 and the input circuit 12. The bus expansion buffer 10 transfers information between the installed expansion program memory 13 and the CPU 3. The expansion program memory 13 is connected to the bus expansion buffer 10 by the external bus 14. The output circuit 11 converts an information signal sent to the electronic device to be controlled from a control signal processable by the CPU 3 into a signal form suitable for signal transmission.
The input circuit 12 converts the transmission signal into a control signal. CPU
3 uses the information signal input from the input circuit 12 to determine the control contents of the electronic device, and sends an operation control signal for instructing the operation to the electronic device via the output circuit 11.
RAM 15 is an external RAM connected to the external bus 14.
In addition to temporarily storing the input / output data with respect to the CPU 3, it also stores identification information in a numerical form used for runaway monitoring according to the present invention.

The runaway detection processing in such a configuration will be described with reference to the flowchart of FIG. The processing procedure of the system program shown by the flowchart of FIG. 2 is actually described in a program language that can be executed by the CPU 3, and therefore is represented functionally in FIG. 2 for convenience of description. This system program is composed of a first program that is executed when the power is turned on, and thereafter, second to eleventh programs that are sequentially and repeatedly executed by loop processing. Five check programs for runaway detection according to the present invention are provided at a plurality of specific positions in any program of the second to eleventh programs. Numerical value 1 as identification information corresponding to a specific position and its connection order
The connection order of → 2 → 3 → 4 → 5 → 1 is predetermined. The position described in the check program may be set to a position desired by the user, and is not limited in any way. The first to tenth programs are the program memory 4
And the eleventh program is stored in the extended program memory 13.

The contents of the nth check program are shown in FIG. The processing contents of this program are as follows. That is, a read address for the monitoring data in the RAM 15 is set and the monitoring data is read (S1
0). It is determined whether or not the numerical value indicated by the read monitoring data and the numerical value n defined on the check program as the numerical value indicating the above connection order match (S2).
0). If an affirmative (YES) determination is obtained, the value obtained by adding 1 to the numerical value n is newly monitored data (corresponding to the identification information of the specific position corresponding to the connection sequence next to the current specific position of the invention of claim 1). Set the write address as
5 is written (S30). Negative in the determination process of S20 (N
O) When the determination is obtained, it is determined that a runaway has occurred, and a program for handling the runaway is activated by jumping to a process for handling the runaway, for example, a process for stopping or resetting a circuit in the system.

When the CPU 3 normally executes such a control procedure, the first check program is executed when the CPU 3 executes the second program.
At this time, the CPU 3 sends the CP from the RAM 15 when the power is turned on.
The initialization value 1 is read by U3 and compared with the reference value 1 defined on the program (S10 to S in FIG. 3).
20). Since a coincidence judgment is obtained as a result of the comparison judgment, the CPU 3 judges that the operation is normal, and stores the numerical value 2 used for the next judgment in the RAM 15 (S30 in FIG. 3). In this way, the CPU 3 updates the numerical value stored in the RAM 15 by 1 each time a normal judgment is obtained in the check program. The other reference value used in the determination of S20 is also set to change by one. In the final position check program, the numerical value used for the next determination is set to 1.

Therefore, while the CPU 3 is operating normally, the order of the numerical values renewedly stored in the RAM 15 is a predetermined numerical order, that is, 1 → 2 → 3 → 4 ...
・ It will change.

When the CPU 3 runs out of control in FIG. 2 as in FIG. 5 and shifts from the third program to the sixth program, the stored value 2 in the RAM 15 in the third check program (the value set by the first check program).
And the reference value 3 assigned to the third check program are compared (S20 in FIG. 3). Since this comparison judgment does not match, the CPU 3 judges (considers) that a runaway has occurred,
Move to processing to deal with the abnormality.

<Second Embodiment> In the first embodiment, a runaway check is performed by comparing a reference value assigned to a position where a check program is described with a determination numerical value set by a previous check program. This is an example in which a check program for that purpose can be placed at an arbitrary position on the main program. In this example, when the execution of the program is stopped, the abnormality cannot be detected. Therefore, a second embodiment will be described below, which is used in combination with the conventional watchdog timer system so that a stop abnormality can be detected. In the second embodiment, the monitoring timer 6 of FIG. 1 is provided, and the count-up signal 7 is the CPU.
3 is input to the interrupt input terminal (highest priority interrupt input terminal that cannot be masked). A check program executed by the CPU 3 is shown in FIG. The execution position of this check program is determined so that its description position is shorter than the count-up time of the timer 6. The check program of FIG. 4 is an example in which timer refresh processing (S40 of FIG. 4) is added to the check program of FIG. In this example, if normal is confirmed by numerical comparison,
The timer 6 is refreshed, that is, the count value is initialized. When the timer 6 counts up, the abnormality processing program is activated by the interrupt input of the count-up signal. In the second embodiment, the timer refresh process is described in all of the check programs, but if the count-up time of the timer 6 is set to be long, the check program for performing the timer refresh process is appropriately selected according to this count-up time. .
When such a check program is used, when the CPU 3 stops abnormally, for example, in a state where loop processing occurs, the timer 6 detects the abnormality.

In addition to the first and second embodiments described above, the following examples can be implemented.

1) In addition to numerical values, characters, for example, a string of English characters such as A, B, and C, can be used as identification information indicating the execution order of programs used to confirm the execution order of programs. . The execution address of the program in the main program can also be used as the identification information.

2) The reference value assigned to the check program in the first embodiment may be stored in advance in the RAM 5 of FIG. This example is suitable when the length of the main program changes and the number of check programs is increased or decreased.

3) Even if the memory (RAM 15) in which the check program writes numerical values becomes abnormal and the stored data becomes indeterminate, in this example, the numerical value comparison processing of the check program can provide the mismatch determination, so the memory abnormality is detected. it can.

[0024]

As described above, according to the present invention, since the program execution sequence itself is monitored, it is not necessary to use a watchdog timer. Therefore, the runaway detection program can be set at any position, and the device can be downsized. Also,
By using the watchdog timer method together, runaway detection accuracy is further improved.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing a circuit configuration of an embodiment of the present invention.

FIG. 2 is a flowchart showing a processing procedure of the first embodiment.

FIG. 3 is a flowchart showing a runaway detection processing procedure of the first embodiment.

FIG. 4 is a flowchart showing a runaway detection processing procedure of the second embodiment.

FIG. 5 is a flowchart showing a runaway procedure.

[Explanation of symbols]

 3 CPU 4 Program memory 5, 15 RAM 6 (for monitoring) timer

Claims (2)

[Claims] 1. A computer runaway detection method for detecting that the execution sequence of a computer program has runaway, wherein a plurality of specific positions in the program and the execution sequence of the computer at those positions are associated with each other. Different identification information and a connection order of the identification information are determined in advance, and when the computer starts executing the program, the first identification information in the connection order is stored in the memory as initial information. Whenever the execution position of the program of the computer reaches any one specific position of the specific plurality of positions, the computer stores identification information stored in the memory and identification information corresponding to the specific position. If a match is obtained, a normal operation is performed. The computer writes the identification information of the specific position corresponding to the connection sequence next to the current specific position in the memory, and if a mismatch determination is obtained, the computer considers that a runaway has occurred. Runaway detection method. 2. The computer runaway detecting method according to claim 1, wherein when the mismatch determination is obtained, the computer executes a predetermined program for coping with the occurrence of the runaway.