JP2965075B2 - Program execution status monitoring method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a program execution state monitoring method for monitoring a program execution state using a watchdog timer.

[Conventional technology]

FIG. 5 is a program configuration diagram showing a conventional program execution state monitoring method, and FIG. 6 is a block diagram showing a hardware environment to which the conventional program execution state monitoring method is applied. In the figure, 1a is a multi-task operating system (hereinafter simply referred to as OS) for managing execution and switching of a plurality of tasks, and 21 to 2n are first to n-th tasks obtained by dividing a processing program into appropriate processing units. 3 sets the count value in the watchdog timer 5
This is an interrupt task for setting WDT. Reference numeral 4 denotes an OS 1a, a central processing unit (hereinafter referred to as a CPU) for executing processing according to a first task 21 to an n-th task 2n, and 5 a built-in counter. A watchdog timer that generates an interrupt operation request signal 7
The WDT5 counter is provided to detect the state of a program that has fallen into an infinite loop because the count value counts down regardless of the program execution state. . 6 is CP
Count value flow when setting the count value from U4 to WDT5, 7
Is a timer provided for generating a timing at which the OS 1a switches the first task 21 to the n-th task 2n, and 9 is a timer provided for the interrupt operation request signal (hereinafter referred to as a watchdog error interrupt). An interrupt operation request signal (hereinafter referred to as a timer interrupt) output from the timer 8 to the CPU 4;
Is a memory in which data related to the program such as the OS 1a, the first task 21 to the n-th 2n is stored, and 11 is a peripheral device to be processed by the program.

Next, the operation will be described. First, the CPU 4 operates according to the OS 1a. That is, the OS 1a is executed. OS1a is the first
It is a program that switches the CPU 4 to operate according to the tasks 21 to 2n of the task 2n.
Operates according to any one of the first task 21 to the n-th task 2n, for example, the task 21. That is, the CPU 4 executes the first task 21. When the execution of the first task 21 ends after an appropriate time has elapsed, the CPU 4 switches to execute the OS 1a, and then, according to the program of the OS 1a, the CPU 4 executes one of the second task 22 to the n-th task 2n. For example, the second task 22 is executed. The operation of switching the tasks executed by the CPU 4 according to the OS 1a in this manner is called scheduling. In addition, during the above operation, in order to monitor the execution state of the program, the WDT setting
Set the count value to 5. For example, a fixed period timer interrupt 9
Occurs, and when the OS 1a recognizes this, the WDT setting interrupt task 3 is started, and the WDT 5 is set to a predetermined count value. When the execution state of the program is normal, a predetermined count value is periodically reset to WDT5, so that the count value of the counter in WDT5 does not become zero. However, when any one of the first task 21 to the n-th task 2n falls into an infinite loop, the periodic setting of the count value in the WDT5 is interrupted, and the counter in the WDT5 is interrupted. The count value becomes 0 and a watchdog error interrupt 7 is generated. At this time, the CPU 4 executes the OS1a and executes this watchdog error interrupt 7
, One of the first to n-th tasks 21 to 2n enters an infinite loop or the like, and is not operating normally.

The second use of WDT5 is as follows:
There is also a method in which the task 2n periodically sets a predetermined count value to the WDT5. Also, the two methods may be used in combination.

[Problems to be solved by the invention]

Since the conventional program execution state monitoring method is configured as described above, for example, as shown in FIG. 7, even if the first task 21 falls into an infinite loop, it can be switched to the OS1a. Then, the second task 22 ~
Schedule one of the nth tasks 2n and call WDT5
Is set, the abnormality of the first task 21 cannot be detected, or all of the first task 21 to the n-th task 2n fall into an infinite loop,
Even if a deadlock occurs or all the tasks are stopped, if at least the WDT setting interrupt task 3 has been activated, the watchdog error interrupt 7 does not occur and it is abnormal. Nevertheless, there has been a problem that the detection cannot be performed.

The present invention has been made in order to solve the above-described problem. In the state where the count value is set to the WDT 5 periodically, any one of the first task 21 to the n-th task 2n is performed. It is an object of the present invention to provide a program execution state monitoring method capable of reliably detecting a crab abnormality even if it occurs.

[Means for solving the problem]

In the program execution state monitoring method according to the present invention, an area (WDT count area) for storing a WDT count value dedicated to each task is provided in a task control block (hereinafter, referred to as TCB) for storing control information of each task. When the OS switches tasks, the count value read from the WDT is stored in the WDT count value area of the TCB corresponding to the task being executed immediately before, and the WDT count value of the TCB corresponding to the task to be executed. The count value is taken out from the area and reset to the WDT.

[Action]

In the present invention, the WDT count value area in the TCB is provided for each task, and a new set value is set when a task is switched by the OS.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 1, 4 is a CPU, and 5 is a watchdog timer (W
DT), 6 is the flow of the count value to this WDT5, 7 is a watchdog error interrupt, 8 is a timer, 9 is a timer interrupt, 10 is a memory, 11 is a peripheral device, and the above are shown in FIG. It is the same as the conventional one shown. Reference numeral 12 denotes a flow of the count value when the CPU 4 reads the count value of the counter in the WDT 5. FIG. 2 is a program configuration diagram showing a program execution state monitoring method, which is almost the same as the conventional one shown in FIG. 5, but includes first tasks 21 to n-th as shown in FIG. TC provided in the memory 10 corresponding to each of the tasks 2n
The difference is that the WDT count value area 15 is added to the content of B14 and that the OS1 has a function of managing the WDT count value in the TCB 14. 2 and 3, 1
Is an OS having a function of managing the WDT count value, 21 is a first task controlled under OS1, 22 is a second task, 2n is an nth task, and 14 is a first task 21 to a first task. n TCBs provided on the memory 10 and managed by the OS 1 corresponding to the respective tasks 2n. The TCB 14 includes the status of the corresponding task (executable ready status, I / O processing In addition to the area for storing information that has been conventionally used, such as the wait state for waiting for completion, the register values when switching tasks, and the task priority, a task-specific WDT
It has a WDT count value area 15 for storing count values.

Next, the operation will be described. First, when a certain task is generated, the OS1 sets the WDT count value of this task in each area of the corresponding TCB 14, together with the initial value of each register at the start of processing, the priority, and the like. The value of the WDT count value is
Since there is a difference in the processing content of each task, an independent value is set for each task. Next, how the first task 21 and the second task 22 shown in FIG. 2 are executed will be described with reference to the flowchart shown in FIG. The flowchart shown in FIG. 4 shows the first task, OS
And the second task are shown separately. First,
As shown in step ST100, since the first task 21 is a program for setting the WDT count value, for example, 100 at the beginning of the program, the CPU 4 sets 100 to WDT5 (step ST100). The first task is executed (step ST101). During this time, the WDT counter counts down irrespective of the execution state of the program, that is, H / W. Thereafter, the CPU 4 is switched from a state in which the CPU 4 operates according to the first task 21 to a state in which the CPU 4 operates according to OS1 (control is switched to OS1) due to some factor such as a timer interrupt or an external key input, and the OS1 is executed. Be started. OS1 first reads the count value of WDT5, and saves it in the WDT count value area 15 of the TCB 14 corresponding to the first task 21 (step ST102). Next, the task is scheduled (step ST103), and when the control is to be switched to the second task 22, the OS1 sets the T corresponding to the second task 22.
The count value is taken out from the WDT count value area 15 of CB14 and
Set to 5 (step ST104). Assume that this value is 70. Next, the control is switched to the second task 22 and the second task 22 is executed.
Of the task 22 is executed (step ST105). After that, if some switching factor occurs, the control is switched to OS1, and OS1 reads the count value from the same process, that is, WDT5, and saves it in the WDT count value area 15 of the TCB 14 corresponding to the second task 22, Scheduling is performed, and a process of extracting the WDT count value from the TCB 14 corresponding to the first task 21 and resetting it to WDT5 is executed (step ST10).
6). Here, the count value of WDT5 stored in step ST102 is 20
Therefore, WDT5 is set to 20. As described above, the processing of the first task 21 is restarted.
When the count value cannot be set for T5, the WDT5 counter is counted down to 0 and the WD
T5 generates a watchdog error interrupt 7. By this interrupt, control is transferred to OS1, and OS1 checks which task is being executed, and in this case, recognizes that an abnormality has occurred while the first task 21 is being executed. So, the first task
It is possible to execute an abnormality recovery process such as extinguishing 21.

In the above embodiment, two tasks, the first task 21 and the second task 22, are targeted. However, it goes without saying that the present invention is also effective for more tasks.

In the above embodiment, the first task 21 and the second task
It was explained that monitoring by WDT5 was performed for all 22.However, for tasks that do not need to be monitored, the TCB14
Set a specific value in the WDT count value area 15 within, and after OS1 schedules, check the value of the WDT count value area 15 and temporarily stop the operation of WDT5 if it is a specific value. Is also good.

〔The invention's effect〕

As described above, according to the present invention, the program execution state monitoring method stores the WDT count value in the TCB for each task managed by the OS, and monitors the WDT for each task based on the stored count value. Thus, when an error occurs in any task, a task that can surely detect the error is obtained. In particular, since the count value set in the task control block at the time of task switching is stored until the next corresponding task is started, task monitoring can be performed as if there is an independent watchdog timer for each task. In addition, even if the execution time of each task is indefinite, task abnormalities can be reliably monitored.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a hardware environment to which a program execution state monitoring method according to an embodiment of the present invention is applied, FIG. 2 is a program configuration diagram showing the program execution state monitoring method, and FIG. FIG. 4 is a flowchart showing the operation of the program execution status monitoring method, FIG. 5 is a program configuration diagram showing a conventional program execution status monitoring method, and FIG. 6 is a diagram illustrating the conventional program execution status monitoring method. FIG. 7 is a block diagram showing a hardware environment, and FIG. 7 is a flowchart showing an operation by a conventional program execution state monitoring method. 1 is a multitasking operating system (OS), 21
Is the first task, 22 is the second task, 2n is the nth task, 4 is the CPU, 5 is the watchdog timer (WDT), 14 is
TCB, 15 is the WDT count value area. In the drawings, the same reference numerals indicate the same or corresponding parts.

Claims (1)

(57) [Claims] A multi-task operating system that manages execution and switching of a plurality of tasks by task control blocks corresponding to the plurality of tasks, respectively, determines that a count value of a watchdog timer has reached a predetermined value. In the program execution state monitoring method for recognizing and detecting an abnormal state of a program, when the multitasking operating system switches tasks, the count value read from the watchdog timer corresponds to the task executed immediately before. Along with setting the task control block, a count value is read from the task control block corresponding to the task to be executed immediately afterward, and the count value is set in the watchdog timer. The value Constant, and the program execution state monitoring method characterized by storing a count value set in the task control block at the time of switching the task until the next corresponding task startup.