JP2002251299A - Program tracing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a program tracing device with which when which task processing or interruption processing destroys a queue is immediately recognized. SOLUTION: When going through the interruption processing or at the beginning of the task switching processing of a real time monitor, by checking whether or not the interruption processing or the task processing in a RUN state immediately before accesses resources in a possible range, the error of the way of using the resources of a memory and a timer, etc., is detected by the performance of only the task processing and the interruption processing.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a program tracing device for a program which is controlled by a real-time monitor which determines and executes a processing order for each priority when a large number of events occur.

[0002]

2. Description of the Related Art An example of a conventional program trace apparatus is described in a fifth embodiment of Japanese Patent Application Laid-Open No. Hei 10-301805. As shown in FIG. 3, the conventional program tracing apparatus includes a queue 10 for managing memory resources and timer resources, a trace buffer 20 for storing a change history of a chain structure of the queue 10, and a resource for using resources. That is, the queue registration check unit 30 checks whether the change history of the queue chain structure is abnormal when it is desired to acquire or delete the queue.

The conventional program tracing device having such a configuration operates as follows. That is, when the user wants to use the memory resource, the user specifies the desired number of bytes, and for the timer resource, the user specifies that one timer is to be used, and the memory and timer acquisition processing is called.

In the called side acquisition processing, the queue registration check unit 30 is executed to check the change history of the chain structure of the queue 10 from the saved trace buffer 20, and to check the history and the current state of the queue 10. Then, the presence or absence of the destruction of the queue 10 is checked by comparing with.

[0005] If there is an abnormality, the queue registration check unit 30 executes abnormality detection processing such as stopping the program execution by a HALT instruction of the CPU, setting an infinite loop, or resetting the program.

[0006]

However, a real-time monitor (O) which determines a processing order for each priority and executes it for each occurrence of a large number of events, which is a frequently used program structure.
In S), as described below, queue destruction can occur due to all task processing and interrupt processing. Therefore, the above-described conventional program trace device detects queue destruction when it is desired to use resources. However, there is a problem that it is not immediately known which task process or interrupt process has destroyed the queue, and it takes time to correct.

Next, the cause of the queue destruction occurring in the real-time monitor (OS) will be described.

A. Forgetting to release resources If any task process or interrupt process creates a route that does not return any resources acquired such as memory or timer due to a program bug, resources are secured every time the route passes And end up running out of resources. Thereafter, if it is determined that there is no resource, an abnormal process is executed. However, if a process with no resources is not prepared, a route for securing resources is executed even if there are no resources, and data is written to address 0, or data when resources were secured in the past. Is used as it is, so that other task processing or interrupt processing overwrites the used memory or timer. Then, after this, it will be a runaway.

B. Resource not acquired If there is a route that does not acquire any resource acquisition process at any point, or if more memory is used than acquired, the task processing and interrupt processing data that correctly acquired the resource are also overwritten Or runaway to destroy nearby data.

An object of the present invention is to provide a program tracing apparatus which can immediately determine which task processing or interrupt processing has destroyed a queue and when.

[0011]

SUMMARY OF THE INVENTION The present invention relates to a program structure frequently used, which is a program which is controlled by a real-time monitor which executes a processing order in a priority order for each occurrence of a large number of events, and executes any task processing. Errors in the use of dynamically managed resources that may occur due to interrupt processing are detected up to the execution of only the task processing and interrupt processing, and the errors are detected in other task processing and interrupt processing. Characterized in that it is not affected.

[0012]

Next, an embodiment of the present invention will be described with reference to the drawings.

The program tracing device of the present invention is used for a computer capable of operating a plurality of task processes and interrupt processes. FIG. 1 is a block diagram showing an embodiment of the program trace device of the present invention. The program tracing apparatus shown in FIG. 1 manages queues 1 for managing memory resources and timer resources, and dynamically manages the queues 1 (resources such as memory and timers that are free only during use and return when used). A trace buffer 2 for storing access information to resources such as a memory and a timer that are being acquired, written, and released when resources are acquired; A) a resource check unit 3 for checking whether or not the current interrupt processing or the task processing in the RUN state immediately before has accessed resources such as a memory and a timer in a possible range. Have been.

Here, the queue will be described. When the user wants to use the memory resource, the user specifies the desired number of bytes. When the user uses the timer resource, the user specifies that he wants to use one timer. Within the called party's acquisition process, a search is made for a vacant place from the start of the resource, and if free, the called party is notified of the acquired place.

In the acquisition process, resources are often managed in a chain structure using queues. This method is as follows.

At the starting point of the queue for managing resources, the address of one used resource is stored, and after the indicated address, the address of the next used resource is stored. After the address, the address of one resource used next is stored, and the address of the next resource is stored up to the end point indicating that the used resource ends here.

Next, the overall operation will be described in detail with reference to FIG.

First, when each task process / interrupt process acquires, writes, and releases resources such as a memory and a timer managed by the queue 1, access information to the resources is stored in the trace buffer 2. Keep it. The following values are stored.

(A) At the time of acquisition a. The state of the resource is changed to a state in which only the resource is acquired. b. Stores the address and range of the acquired resource. c. The acquired process number, the process number that can be written, and the process number that can be released are stored.

(B) At the time of writing a. Put the resource information in the written state. b. Stores the written process number. c. The written value is stored (only the amount that can be practically stored by the resource is stored).

(C) Release time a. Releases the resource status. b. Store the released process number.

Next, when exiting the interrupt processing or after executing the processing of the OS involving task switching such as notification between tasks and waiting for event occurrence in the task processing, the resource check unit 3
Then, a process for checking whether the current interrupt process or the task process in the RUN state has accessed a resource within a possible range is executed, and it is detected whether an abnormal process has been performed on the resource.

Whether or not the interrupt processing and the task processing have accessed resources within a possible range is confirmed by the following method.

The resources of the memory and the timer have two main states. The first state is that the task processing / interrupt processing has been acquired because it is desired to use it. At this time, the content of the resource is in one of the following two states.

(1) Acquired state If acquired and not used, the content value of the resource is:
It should not be different from before executing the task processing or the interrupt processing.

At the time of exiting the interrupt processing or at the first timing of the task switching processing of the OS, the content value of the acquired resource is read, and the interrupt processing executed before that, or the task which was in the RUN state immediately before is read out. If the value has changed from the value at the time of processing, the current interrupt processing
This means that the task process in the UN state has performed an abnormal process on the resource.

At this time, it is also possible to determine in advance the amount of resources to be used for judging the abnormality and to check the amount of resources used, for example, when the resource has used 80% of the designed amount, and to judge the abnormality. .

(2) A state in which the acquired and correct processing as designed rewrites the content value of the resource and the resource is being used. At this time, the content value of the resource is the same value as the value rewritten by the correct processing as designed. It should be.

Further, it is necessary to confirm that the correct processing as designed is rewritten in the obtained address in the obtained address, to confirm the information at the time of obtaining only, and to check the processing number determined in advance in the design. You can check with

At the time of exiting the interrupt processing or at the first timing of the task switching processing of the OS, it is checked whether the values are the same, or the address, range, and write processing number are checked to detect an abnormality.

The second state is an initial state after the power of the apparatus is turned on, or a state in which resources acquired by task processing / interrupt processing have been used up and released. At this time, the content of the resource is in one of the following two states.

(3) Initial state after power-on The contents of resources should be initial values. Especially for resources for which no initial value is set, there should be no change from the value at power-on.

When exiting the interrupt processing or at the first timing of the task switching processing of the OS, it is checked whether or not the values are the same, and an abnormality is detected.

(4) Released state The content value of the resource should not change from the value at the time of release.

When exiting the interrupt processing or at the first timing of the task switching processing of the OS, it is checked whether or not the values are the same, and an abnormality is detected.

In this way, it is possible to detect an error in the use of resources such as a memory and a timer, which are being dynamically managed, up to the execution of only the current task processing / interrupt processing.

Next, referring to FIG. 2, the operation of the resource abnormality check performed at the time of exiting the interrupt processing or at the beginning of the task switching processing will be described. FIG. 2 is a flowchart illustrating an operation of performing a resource abnormality check after executing an OS process that involves task switching at the end of an interrupt process or within a task, such as notification between tasks and event occurrence wait.

First, in the case of the task switching process, the exclusive control is performed by inhibiting the interrupt (step 101). next,
Whether or not the memory resource, the timer resource, and other resources have been accessed is confirmed by the above-described checks (1) to (4) (steps 102 to 104). Next, in order of priority,
The presence or absence of an DY task is confirmed (step 105). Here, if there is no READY task, the interruption is permitted and the exclusive control is terminated (step 106). Next, H
The ALT / STOP instruction waits for the occurrence of the next event (step 107). If there is a task switching process, the process returns to the first step 101.

In step 105, if there is a READY task, the state of the READY task is set to RUN,
The stack pointer and register of the UN task are restored (step 108).

Next, the interrupt is permitted, the exclusive control is terminated, and the program counter value of the RUN task is entered in the program counter (step 109).

[0041]

As described above, according to the present invention, when exiting the interrupt processing, at the beginning of the task switching processing of the real-time monitor, at the interrupt processing, or immediately before
By checking whether the task process in the state has accessed resources within a possible range, an error in usage of resources such as a memory and a timer is detected up to the execution of only the task process and the interrupt process, and the error is detected. Other task processing,
It is possible not to affect the interrupt processing.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing an embodiment of a program trace device of the present invention.

FIG. 2 is a flowchart illustrating an operation of checking a resource abnormality at the beginning of a task switching process.

FIG. 3 is a diagram showing a conventional program trace method.

[Explanation of symbols]

 1,10 queue 2,20 trace buffer 3 resource check unit 30 queue registration check unit

Claims (10)

[Claims] In a program which is controlled by a real-time monitor, which is a frequently used program structure and which determines a processing order for each priority when a large number of events occur and which may be executed by any task processing and interrupt processing, There is a need to detect an error in the use of resources for which dynamic management is being performed up to the execution of only the task processing and interrupt processing, and to prevent the error from affecting other task processing and interrupt processing. Characterized program trace device. 2. When exiting the interrupt processing or at the beginning of the task switching processing of the real-time monitor, the interrupt processing or the task processing which was in the RUN state immediately before is executed.
2. The program tracing apparatus according to claim 1, wherein the error is detected by checking whether the resource has been accessed within a possible range. 3. The method according to claim 1, wherein the resource is accessed within the possible range when the resource is only acquired, when the interrupt processing is exited, or at the first timing of the task switching processing of the real-time monitor. Read, the content value of the resource is the previously executed interrupt processing,
Or, check whether the value is the same as the value of the task that was in the RUN state immediately before, or when the resource usage is reduced at the time of exiting the interrupt processing or at the first timing of the task switching processing of the real-time monitor. 3. The program tracing device according to claim 2, wherein whether or not the usage amount exceeds a predetermined amount is checked. 4. Whether the resource has been accessed within the possible range depends on whether the resource has been acquired and correct processing as designed rewrites the content value of the resource, and exits the interrupt processing when the resource is in use. Or at the first timing of the task switching process of the real-time monitor, read the content value of the resource, and check whether the content value of the resource is the same as the value rewritten by the correct process as designed, or In the acquired address, the correct processing as designed is rewritten within the acquired range by checking the information at the time of acquisition only and the processing number predetermined in the design and checking it. 3. The program trace apparatus according to claim 2, wherein 5. The method according to claim 5, wherein the resource is accessed within the possible range, when the power is turned on, an initial state in which no initial value is set to the resource and a state in which the resource is released, when interrupt processing is exited, or task switching of the real-time monitor is performed. 3. The program tracing apparatus according to claim 2, wherein at the first timing of the processing, whether or not the values are the same is checked. 6. A program which is controlled by a real-time monitor, which is a program structure frequently used and which determines and executes a processing order for each priority when a large number of events occur, may be generated by any task processing and interrupt processing. There is a need to detect an error in the use of resources for which dynamic management is being performed up to the execution of only the task processing and interrupt processing, and to prevent the error from affecting other task processing and interrupt processing. Characterized program tracing method. 7. When exiting the interrupt processing or at the beginning of a task switching processing of the real-time monitor, the interrupt processing or the task processing which was in the RUN state immediately before is executed.
7. The program tracing method according to claim 6, wherein the error is detected by checking whether the resource has been accessed within a possible range. 8. A method for determining whether a resource has been accessed within the possible range, when the resource is only acquired, when exiting an interrupt process, or at the first timing of a task switching process of a real-time monitor. Read, the content value of the resource is the previously executed interrupt processing,
Or, check whether the value is the same as the value of the task that was in the RUN state immediately before, or when the resource usage is reduced at the time of exiting the interrupt processing or at the first timing of the task switching processing of the real-time monitor. 8. The program tracing method according to claim 7, wherein a check is made as to whether or not a predetermined usage amount has been exceeded. 9. Whether the resource has been accessed within the possible range depends on whether the resource has been acquired and correct processing as designed rewrites the content value of the resource, and exits the interrupt processing when the resource is in use. Or at the first timing of the task switching process of the real-time monitor, read the content value of the resource, and check whether the content value of the resource is the same as the value rewritten by the correct process as designed, or In the acquired address, it is checked that the correct processing as designed within the acquired range is rewritten by confirming the information obtained only when acquiring and the processing number determined in advance in the design. 8. The program trace method according to claim 7, wherein: 10. A method for determining whether a resource has been accessed within the possible range, when an initial state in which an initial value is not set in a resource after power-on and a state in which the resource is released, when exiting an interrupt process, or switching a task of a real-time monitor. 8. The program trace method according to claim 7, wherein at the first timing of the processing, whether or not the values are the same is checked.