JPH04280329A - Program abnormality detection 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]

0001

[Field of Industrial Application] The present invention relates to a program operation abnormality detection method for detecting program abnormalities, and in particular, to a program operation abnormality detection method for detecting task operation abnormalities in a time-sliced multitasking OS (operating system). This relates to an anomaly detection method.

[0002] As software processing becomes more complex, multi-task OSes have come into widespread use in place of conventionally used single-task OSs. In the case of a single-tasking OS, one program (task) is executed sequentially, so it is relatively easy to detect program abnormalities, but in the case of a multi-tasking OS, multiple programs (tasks) are executed simultaneously. Therefore, when a program abnormality occurs, it is first necessary to detect in which program (task) the abnormality has occurred.

[0003] There are two types of multitasking OS. One is a method of switching the right to use the central processing unit (CPU) by a task, that is, dispatching, based on the occurrence of events such as system calls and interrupts provided by the OS, and the other is a method using TSS (time sharing system). Similarly, this method dispatches tasks at regular intervals. The major difference between the two is that the dispatch method is different; in the former case, the task itself generates the dispatch event, but in the latter case, the task itself
S causes a dispatch event. O in the latter case
S is called a time slice OS.

[0004] The program malfunction detection method for a multitasking OS is capable of detecting malfunctioning tasks even for tasks under a time slice OS, which dispatches tasks at certain fixed time intervals. is requested.

[0005]

[Prior Art] Fig. 7 explains the operation of a time slice OS, and shows a case where three tasks A, B, and C are operating, and task A, task B, and task C are constant. It is shown that the process is executed sequentially every time T. In the figure, T is the time interval (dispatch interval) at which time slicing is performed, and the solid line represents the execution state.

FIG. 8 shows the operation of the time slice OS when an abnormality occurs, and shows a case where an abnormality such as a deadlock or an infinite loop occurs in task B. P in the figure indicates that during this subsequent time interval T, task B obtains the right to use the CPU by time slicing, but cannot perform any operations.

[0007] In this way, in the case of a time slice OS,
As shown in FIG. 8, even if an abnormality occurs in task B, for example, there is no means to detect the abnormality in the past. The task is executed and the processing shifts to task A and task C. Therefore, the system as a whole continues to operate without any abnormality.

[0008]

[Problem to be solved by the invention] As mentioned above, in the case of a time slice OS, even if a certain task malfunctions,
Dispatch is also performed for tasks in which an abnormality occurs at regular time intervals, and CPU usage time is allocated. Therefore, even if one task malfunctions, other tasks can operate normally, and the system as a whole
The problem is that it continues to operate without becoming abnormal.

[0009] The present invention aims to solve the problems of the prior art as described above, by detecting an abnormal operation of a task in a time-sliced multitasking OS, and thereby preventing the abnormal operation from occurring continuously. The purpose of the present invention is to provide a program abnormality detection method that can prevent such problems.

0010

[Means for Solving the Problems] The present invention provides a kernel section that has a core program, a time slicing interrupt section that generates a time slicing interrupt for the kernel section, and a time slicing interrupt section that generates a time slicing interrupt for the kernel section, and a time slicing interrupt section that generates a time slicing interrupt for the kernel section. In an operating system (OS) that consists of a scheduler that dispatches multiple tasks according to the schedule, there is an interval timer that operates at fixed time intervals, a task operation status table that displays the current operation status of each task, and a task operation status table that displays the current operation status of each task. A pre-task operation state table that displays the operation state, a task operation interval counter that counts the number of operation time intervals of each task, and a task operation interval table that holds the maximum allowable operation interval during normal operation of each task, When the interval timer compares the current operating state and previous operating state of each task and there is no change. When the counted value of the task's operating time interval is compared with the table value of the operating interval and the counted value is greater than the table value. This method is characterized by determining the occurrence of an operational abnormality in a task.

The present invention also provides a kernel section having a core program, a time slicing interrupt section that generates a time slicing interrupt for the kernel section, and a plurality of tasks activated by the kernel section based on the interrupt. In an operating system (OS) consisting of a scheduler that dispatches tasks, a task operation status table that displays the current operation status of each task;
A pre-task operation state table that displays the previous operation state of each task, a task operation interval counter that counts the number of operation time intervals for each task, and a task operation interval that maintains the maximum allowable operation interval during normal operation of each task. The time slicing interrupt section operates at fixed time intervals, and when there is no change in the current operating state of each task and the previous operating state, the count value of the task operating time interval and the operating interval are calculated. This feature is characterized in that by comparing the count value with a table value, it is determined that an operational abnormality has occurred in the task when the count value is larger than the table value.

Further, the present invention provides a kernel section having a core program, a time slicing interrupt section generating a time slicing interrupt for the kernel section,
In an operating system (OS) consisting of a scheduler that dispatches multiple tasks in response to startup from the kernel section based on this interrupt, there is a task operation status table that displays the current operation status of each task, and a task operation status table that displays the current operation status of each task. A pre-task operation state table that displays the operation state, a task operation interval counter that counts the number of operation time intervals of each task, and a task operation interval table that holds the maximum allowable operation interval during normal operation of each task, When the scheduler operates at fixed time intervals and there is no change in the current operating state of each task and the previous operating state, compare the counted value of the task operating time interval with the table value of the operating interval and calculate the counted value. This method is characterized in that it is determined that an operational abnormality has occurred in a task when is larger than a table value.

[0013]

[Operation] FIG. 1 shows the basic structure of the present invention. The operating system (OS) that is the subject of this invention
) consists of a kernel part 1 that has a core program, a time slice interrupt part 2 that generates interrupts to this kernel part 1 at regular time intervals, and a plurality of time slice interrupt parts based on the interrupts from this time slice interrupt part 2. and a scheduler 4 that sequentially allocates the right to use the processing device to the tasks 3 of . In the present invention,
In response to this, an interval timer 5 that operates at fixed time intervals is provided, a task operation status table 6 is provided to display the current operation status of each task, and a pre-task operation status table 7 is provided to display each task. A task operation interval counter 8 is provided to count the number of operation time intervals for each task, and a task operation interval table 9 is provided to calculate the maximum allowable operation interval during normal operation of each task. Hold. Then, when the interval timer 5 compares the current operating state and the previous operating state of each task and there is no change, the counted value of the operating time interval of this task is compared with the table value of the operating interval, and the counted value is determined. When it is larger than the table value, it is determined that an operational abnormality has occurred in this task.

Further, in the present invention, in the operating system (OS) as described above, a task operation state table 6 is provided to display the current operation state of each task, and a pre-task operation state table 7 is provided. Displays the previous operation status of each task, and displays the task operation interval counter 8.
is provided to count the number of operation time intervals of each task, and a task operation interval table 9 is provided to hold the maximum allowable operation interval during normal operation of each task. The time slicing interrupt unit 2 operates at fixed time intervals and compares the current operating state and previous operating state of each task, and if there is no change, a table of the counted value of the operating time interval of this task and the operating interval When the counted value is larger than the table value, it is determined that an operational abnormality has occurred in this task.

Further, in the present invention, in the operating system (OS) of the processing device as described above,
A task operation status table 6 is provided to display the current operation status of each task, a pre-task operation status table 7 is provided to display the previous operation status of each task, a task operation interval counter 8 is provided, The number of operation time intervals of each task is counted, and a task operation interval table 9 is provided to hold the maximum allowable operation interval during normal operation of each task. Then, the scheduler 4 operates at fixed time intervals, compares the current operating state and previous operating state of each task, and when there is no change, compares the counted value of the operating time interval of this task with the table value of the operating interval. When the count value is larger than the table value, it is determined that an operational abnormality has occurred in this task.

Therefore, according to the present invention, the time slice
In a multitasking OS, abnormalities occurring in tasks under a time slice OS can be detected by checking changes in the operating states of tasks at regular time intervals.

[0017]

[Embodiment] FIG. 2 shows the configuration of an embodiment of the present invention, in which numeral 11 indicates a time slice OS section, and in the time slice OS section 11, numeral 12 indicates a process for generating a time slice interrupt. 13 is a kernel section that is the core of the OS, and 14 is a scheduler that performs task switching processing. 15 indicates a task group, and 151 to 15n are tasks, respectively. 16 is an interval timer that measures fixed time intervals;
7 is a task operation state table showing the operation state of each task, 18 is a pre-task operation state table showing the previous operation state of each task, 19 is a task operation interval table showing the operation interval of each task, and 20 is a table of each task. This is a task operation interval counter that counts operation intervals.

FIG. 3 explains the execution of tasks by time slices in the embodiment of FIG.
An example of operation in a system with three tasks B and C is shown. In the figure, 17 is a task operation state table, and T is a time interval at which time slicing is performed. The task operation state table 17 needs to have elements for the number of tasks,
In this example, there are three elements corresponding to tasks A, B, and C.

Task A is processed in a certain time interval T, and when it is finished, the area for task A in the task operation status table 17 is incremented, thereby writing the operation status. That is, the task operation state table 17 is incremented every time a task performs one process. In the next time interval T, task B is processed, and at the end, the area of task B in the task operation status table 17 is incremented, thereby writing the operation status. Furthermore, in the next time interval T, task C is processed, and the area of task C in the task operation state table 17 is incremented, thereby writing the operation state. Thereafter, such processing is sequentially repeated at every time interval T.

FIG. 4 is a flowchart showing the operation of the interval timer in the embodiment of FIG. 2, showing a case where abnormality detection is performed using the interval timer 16. , B, and C.

First, the operating state of task A is obtained from the task operating state table 17 (S1). Next, the interval timer 16 acquires the previous operation state of task A from the pre-task operation state table 18 (S2). Then, check whether there is a change between the previous state and the current state (S3),
If there is no change, the task A area of the task operation interval counter 20 is counted up (S4), and then the time interval count value of task A in the operation interval counter 20 is preset in the task operation interval table 19. Check whether it is larger than the operation interval value of task A (S5).
, when it is large, it is determined that an abnormal operation has occurred in task A. The count value of task A in the operation interval counter 20 is
If the operation interval value is not greater than the operation interval value of task A preset in the task operation interval table 19, the process of task A is finished, and changes in the operation states of tasks B and C are checked in the same way to see if they are normal. If the operation is normal, it is determined whether an abnormality has occurred, and if the operation is normal, the process is terminated.

Returning to step S3, if there is a change between the previous state and the current state, the task operation state table 17
The value of is set in the pre-task operation state table 18 (S6
), the count value in the area of task A of the task operation interval counter 20 is cleared (S7), and the process moves to the next task B and task C.

1 As described above, the interval timer 16 operates by interrupts at fixed time intervals T, and compares the pre-task operation state table 18 and task operation state table 17 of each task during operation, and changes the operation state. Find out if there is. If there is a change, the task has changed from its previous state and is therefore operating normally. If there is no change, the task operation interval table 20
The interval timer interrupt counts up for the number of minutes specified in . If there is no change in the operating state after the specified time has elapsed, it is assumed that an abnormality has occurred in that task and processing is performed. Task operation interval table 2
0 is set to a value larger than the maximum value of the processing time interval for each task. In other words, the fact that the operational state of a task does not change for more than the time listed in the task operation interval table 20 indicates that some abnormality has occurred in the task and the processing time is longer than normal. , This makes it possible to detect abnormalities for each task.

FIG. 5 shows the structure of another embodiment of the invention, in which the same parts as in FIG. 2 are designated by the same numbers, but the interval timer 16 is omitted. In the embodiment of FIG. 5, the time slice interrupt unit 12 performs the same process as the task abnormality detection by the interval timer 16 shown in the embodiment of FIG. 2, and the processing method in this case is shown in FIG. This can be realized in the same manner as in the embodiment.

FIG. 6 shows the configuration of yet another embodiment of the present invention, in which the same components as in FIG. 2 are designated by the same numbers, but the interval timer 16 is omitted. In the embodiment of FIG. 6, the scheduler 14 uses the interval timer 16 shown in the embodiment of FIG.
It performs the same processing as task abnormality detection by
The processing method in this case can be implemented in the same manner as in the embodiment of FIG.

[0026]

Effects of the Invention As explained above, according to the program abnormality detection method of the present invention, in a time-sliced multitasking OS, interrupt operations at fixed time intervals can be performed.
By monitoring changes in the operating state of each task, it is possible to detect an abnormal operation occurring in a task under the time slice OS.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the basic configuration of the present invention.

FIG. 2 is a diagram showing the configuration of an embodiment of the present invention.

FIG. 3 is a diagram illustrating task execution using time slices in the embodiment of FIG. 2;

FIG. 4 is a flowchart showing the operation of an interval timer in the embodiment of FIG. 2;

FIG. 5 is a diagram showing the configuration of another embodiment of the present invention.

FIG. 6 is a diagram showing the configuration of still another embodiment of the present invention.

FIG. 7 is a diagram illustrating the operation of a time slice OS.

FIG. 8 is a diagram showing operations when an abnormality occurs in the time slice OS.

[Explanation of symbols]

1 Kernel part 2 Interrupt section for time slice 3. Task 4 Scheduler 5 Interval timer 6 Task operation status table 7 Pre-task operation status table 8 Task operation interval counter 9 Task operation interval table

Claims (3)

[Claims] 1. A kernel section (1) having a core program, and a time slicing interrupt section (1) that generates a time slicing interrupt for the kernel section (1).
2) and a scheduler (4) that dispatches a plurality of tasks (3) in response to activation from the kernel unit (1) based on the interrupt, an interval operating system that operates at regular time intervals. A timer (5), a task operation status table (6) that displays the current operation status of each task, and a pre-task operation status table (7) that displays the previous operation status of each task.
, a task operation interval counter (8) that counts the number of operation time intervals of each task, and a task operation interval table (9) that holds the maximum allowable operation interval during normal operation of each task, and the interval timer ( 5) Compares the current operation state and previous operation state of each task and when there is no change, compares the count value of the operation time interval of the task with the table value of the operation interval, and determines that the count value is the table value. A method for detecting a program abnormality, characterized in that when the value is larger than that, it is determined that an operational abnormality has occurred in the task. 2. A kernel section (1) having a core program, and a time slicing interrupt section (1) that generates a time slicing interrupt for the kernel section (1).
2) and a scheduler (4) that dispatches a plurality of tasks (3) in response to activation from the kernel unit (1) based on the interrupt, the current operating state of each task. a task operation status table (6) that displays the previous operation status of each task, and a pre-task operation status table (7) that displays the previous operation status of each task.
A task operation interval counter (8) for counting the number of operation time intervals of each task, and a task operation interval table (9) for holding the maximum allowable operation interval during normal operation of each task are provided. The interrupt unit (2) operates at fixed time intervals, compares the current operating state of each task with the previous operating state, and when there is no change, calculates the counted value of the operating time interval of the task and the table value of the operating interval. A program abnormality detection method, characterized in that when the counted value is larger than the table value, it is determined that an operational abnormality has occurred in the task. 3. A kernel section (1) having a core program, and a time slicing interrupt section (1) that generates a time slicing interrupt for the kernel section (1).
2) and a scheduler (4) that dispatches a plurality of (3) tasks in response to activation from the kernel unit (1) based on the interrupt, the current operating state of each task. a task operation status table (6) that displays the previous operation status of each task, and a pre-task operation status table (7) that displays the previous operation status of each task.
, a task operation interval counter (8) that counts the number of operation time intervals of each task, and a task operation interval table (9) that holds the maximum allowable operation interval during normal operation of each task, and the scheduler (4) ) operates at fixed time intervals, and when there is no change when comparing the current operating state and previous operating state of each task, compare the counted value of the operating time interval of the task with the table value of the operating interval. A program abnormality detection method characterized in that it is determined that an operational abnormality has occurred in the task when a count value is larger than the table value.