JP2006127166A - Information processor, task stop method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently debug a computer program. <P>SOLUTION: This information processor includes: a storage part 71 associating and storing task identification information for identifying a task, and stop target task identification information for identifying at least one stop target task stopped when the task comes into a prescribed state; a task state management part 70 detecting that the task comes into the prescribed state; and a stopping task stopping the stop target task stored associatively to the task coming into the prescribed state when a prescribed state detection means detects that it comes into the prescribed state. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information processing apparatus, a task stop method, and a program, and more particularly to a debugging technique when a computer program having a multitask configuration is executed.

  Conventionally, various methods have been proposed for analyzing the cause of a function call under an illegal condition in a computer system. For example, Patent Document 1 describes a technique related to using a test program in order to test a computer program having a multitask configuration.

In the case of multitasking, if there is a function call under an illegal condition in a certain task, even if only that task is stopped, the other tasks are not normally stopped and the computer system remains activated. Then, it is difficult to check the state of the entire computer system including other tasks. Therefore, in this technique, when it is detected that there is a function call with an illegal condition as described above, debugging can be performed by performing a reproduction test using a test program. .
JP-A-9-190365

  However, in the above conventional method, it is necessary to perform a reproduction test by operating a test program. For this reason, the man-hour for specifying the cause becomes very large, and there is a problem that the efficiency of debugging is poor. There is also a method of performing desktop tracing, but in any case, debugging efficiency was poor.

  SUMMARY An advantage of some aspects of the invention is that it provides an information processing apparatus, a task stop method, and a program that can efficiently debug a computer program. is there.

  An information processing apparatus according to the present invention for solving the above problem includes task identification information for identifying a task, and at least one task to be stopped to be stopped when the task is in a predetermined state. Task identification information storage means for storing information in association with each other, predetermined state detection means for detecting that the task has entered the predetermined state, and detection of the predetermined state being detected by the predetermined state detection means And a task stop means for stopping the stop target task stored in association with the task in the predetermined state.

  In this way, when a certain task enters a predetermined state, it is also possible to stop the task to be stopped that is stored in association with the task, so that the state change due to the related task can be suppressed. The program can be debugged efficiently.

  In the information processing apparatus, the task identification information storage unit may store the at least one stop target task identification information in association with each of the plurality of task identification information. In this way, it is possible to suppress state changes due to related tasks for a plurality of tasks.

  Moreover, in the information processing apparatus, the task stop unit identifies stop target task identification information stored in association with task identification information for identifying a task that is the stop target task stopped by the task stop unit. It is also possible to further stop the task to be stopped. In this way, a task group composed of a plurality of related tasks can be stopped at a time.

  In the information processing apparatus, the task stopping unit includes a stopping task starting unit that starts a stopping task when the predetermined state is detected by the predetermined state detecting unit, and the task stopping unit includes: The stop target task may be stopped by the started stop task. In this way, the process for stopping the stop target task can be executed in a stop task that is a task different from the task in the predetermined state and the stop target task, so that the stop target task can be stopped more reliably. it can.

  In the information processing apparatus, when there is communication from the first task to the second task, the task identification information storage means stores the task identification information for identifying the second task in the first task. It is good also as memorizing | storing as stop object task identification information which identifies the stop object task to stop when this task will be in the said predetermined state. In this way, the task to be stopped can be determined dynamically.

  In the information processing apparatus, the stop target task identified by the stop target task identification information stored in the task identification information storage unit in association with the task identification information is a task identified by the task identification information. It may be a task that accesses a common memory address. In this way, it is possible to suppress a change in state due to a related task by suppressing a change in memory.

  Further, in the information processing apparatus, mode storage means for storing one of the first mode and the second mode, and when the stored mode is the first mode, the task stop means And a mode corresponding processing means for performing a predetermined recovery process when the stored mode is the second mode. In this way, the debugging process and the recovery process can be performed according to the mode of the information processing apparatus. Therefore, it is higher priority to restore earlier than to stop the task and perform debugging. It is possible to deal with cases.

  The task stop method according to the present invention includes a predetermined state detection step for detecting that the task has entered the predetermined state, and when it is detected that the predetermined state has been detected in the predetermined state detection step. In association with the task in the predetermined state, task identification information for identifying the task, and stop target task identification information for identifying at least one stop target task to be stopped when the task is in the predetermined state, And a task stop step for stopping the task to be stopped stored by the task identification information storage means for storing in association with each other.

  Further, the program according to the present invention is configured such that the predetermined state detection unit that detects that the task has entered the predetermined state, and the predetermined state detection unit detects that the task has entered the predetermined state. Corresponding task identification information for identifying a task in association with a task in a predetermined state and stop target task identification information for identifying at least one stop target task to be stopped when the task enters a predetermined state The computer is caused to function as task stop means for stopping the task to be stopped stored by the task identification information storage means attached and stored.

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

  The information processing apparatus 1 according to the present embodiment is configured as a computer having the structure shown in FIG. That is, the information processing apparatus 1 includes the target CPU 10, the memory 15, and the peripheral device 20 as hardware, and includes a general-purpose OS program 30, an application 40, and a driver 50 as software that uses these hardware. The application 40 includes tasks 41 to 47, and the driver 50 includes tasks 51 to 53. Note that the task here means an execution unit of processing as viewed from the general-purpose OS program 30. More specifically, a task is generated by executing the program, and the task is processed by the target CPU 10 under the management of the general-purpose OS program 30.

  Each task is managed in a preemptive multitasking manner. That is, the processing of each task is managed by the general-purpose OS program 30, and time for using the target CPU 10 is allocated to each task by time division. At this time, the general-purpose OS program 30 stores a task table in the memory 15 and stores the processing state of each task in the table. The task table includes identification information for identifying a task and its status (running / executable status, short pause status, long pause status, end status, pause status, swap-out status, disk wait status, page wait status. Status, etc.) are stored.

  In the present embodiment, the information processing apparatus 1 is a computer that can perform a plurality of processes by using a plurality of tasks. The plurality of processes includes a basic function part and an additional function part, and among the tasks 41 to 47, a task group 60 including the tasks 41 to 44 is a task for processing the basic function part. A task group 61 composed of tasks 45 to 47 is a task for processing the additional function portion. As a specific example, when the information processing apparatus is an exchange, the basic function portion is basic call processing, and the additional function portion is processing for call waiting, for example.

  Processing for stopping these tasks will be described with reference to FIG. As illustrated in FIG. 2, the information processing apparatus 1 includes a task state management unit 70 and a storage unit 71. These are both functional blocks realized by the general-purpose OS program 30. The hardware is realized by the target CPU 10 and the memory 15. The task state management unit 70 performs processing for managing the state of each task (tasks 72-1 to 72-n), and the storage unit 71 stores the above-described task table and a stop target task table described later.

  In the task stop processing that has been performed conventionally, the task 51 has broken the memory 15 used by the task 52 (the task 51 writes information that is not supposed to be written in a certain part of the memory 15). Therefore, a case where an illegal function call has occurred in the task 52 will be described as an example. FIG. 7 is a process flow of a task stop process when a failure occurs according to the conventional technique. First, the task 51 and the task 52 are time-division processed as a multitask. Then, in the task 52, the state in which the task 51 has caused a failure (the memory has been destroyed) is detected (S200). Then, the task 52 requests the task state management unit 70 to stop the task 51 (S202). The task state management unit 70 that has received the request performs a task 51 stop process. On the other hand, the task 52 whose memory has been destroyed by the task 51 is also stopped (S204). In this way, the task 51 and the task 52 are stopped so that the memory destruction by the task 51 does not adversely affect the subsequent processing.

  By the way, according to the process described above, the process progresses for other tasks and the memory is rewritten, and therefore it is difficult to identify the cause by checking the state of the information processing apparatus 1 after the task is stopped. There is. For this reason, in order to specify the cause, a test program is used or a desktop trace is performed, but the debugging efficiency is not good as described above. Therefore, in the present embodiment, a stop target task table is stored in the storage unit 71, and the task is stopped according to the table, thereby enabling debugging without using a test program or performing desktop tracing. .

  FIG. 3 is an example of a stop target task table. In this table, the relationship with other tasks (including own task) is stored for each task. More specifically, a column stored in association with each task (detection task) of the detection task group described at the top of the table (part of the region 80 in the figure) has a predetermined state in the detection task. When detected, for each task (related task) included in the related task group (portion 81 in the same figure), a stop target flag is stored as information indicating whether it is a stop target task to be stopped. . In the figure, a black circle indicates a state indicating that the related task is a stop target task by a stop target flag, and a white circle indicates a state indicating that the corresponding related task is not a stop target task by a stop target flag. . Note that the task stop means that the task is processed as not being a target task for which the CPU execution right is acquired.

  In other words, in the stop target task table, task identification information for identifying a detection task and task identification information for identifying a related task are stored in association with each other, and when the detection task is in a predetermined state For at least one related task to be stopped, a stop target flag for identifying that it is a stop target task is stored. When the stop target flag for a combination of a certain detection task and a related task is in a state indicating that it is a stop target task, the related task is a stop target task in relation to the detection task, and the detection task It can be said that the task to be stopped is associated and stored.

  Note that the predetermined state refers to a state where there is an illegal function process, for example. As a specific example of the process for detecting the predetermined state, a process for determining whether or not the return of the function call is an error can be used. By including this processing in each task, a predetermined state can be detected in each task.

  As shown in FIG. 3, in the area A, the detection tasks are tasks 51 to 53 included in the driver 50 or tasks 41 to 44 that are tasks for processing the basic function portion, and related tasks are also tasks 51 to 53 or Tasks 41 to 44. When a predetermined state described later is detected in a task related to the basic function, all the tasks related to the other basic functions need to be stopped. When an illegal function call occurs in a task for processing of the driver 50 or the basic function part, the task forms a basic part of processing performed in the information processing apparatus 1, so that This is because it is appropriate to recover the task by stopping and then restarting the task. Therefore, a stop target flag indicating that a task related to another basic function that is a related task is a stop target task is stored in the area A in association with a task related to the basic function that is each detected task. .

  In area B, the detected tasks are tasks 51 to 53 or tasks 41 to 44, and tasks 45 to 47, which are tasks for processing the related task addition function part. When a predetermined state described later is detected in the task related to the basic function, it is necessary to stop all the tasks related to the additional function. When an illegal function call occurs in a task for processing of the driver 50 or the basic function part, the task forms a basic part of processing performed in the information processing apparatus 1, so that This is because it is appropriate to recover the task by stopping and then restarting the task. Therefore, a stop target flag indicating that the task related to the additional function that is a related task is a stop target task is also stored in the area B in association with the task related to the basic function that is each detected task.

  On the other hand, in area C, the detected tasks are tasks 45 to 47 that are tasks for processing the additional function portion, and the related tasks are also tasks 45 to 47. When a predetermined state is detected in the tasks 45 to 47, the task for processing the other additional function portion is stopped as necessary. Therefore, a stop target flag indicating that the task is a stop target task to be stopped if necessary is stored in the area C. If the detection task is a task 45 to 47 that is a task for processing the additional function part, the task for processing the driver 50 or the basic function part is a stop indicating that the task is to be stopped. The target flag is not stored. This is because when an illegal function call occurs in the task for processing the additional function part, it is sufficient to prevent the basic function part and the driver part from being affected.

  Next, a method for determining a stop target task to be stopped when a predetermined state is detected in a certain task will be described. In other words, as described above, the relationship between the detection task and the related task in the region C is not a relationship in which all other tasks must be stopped when a predetermined state is detected in the detection task. For this reason, there is a need for a method for selecting only related tasks that must be stopped when a predetermined state is detected in the detection task, and storing a stop target flag. One of the methods is, for example, a method of storing in a stop target task table when a program is created. That is, in this method, a stop target flag indicating that the task is to be stopped when a predetermined state is detected in a certain task is stored in advance in the stop target task table. At this time, a task that may access a memory address common to a certain task is preferably set as a stop target task in relation to the task. This is because the state of the memory table when the task is in a predetermined state can be held after the task is stopped.

  In another method, a stop target flag is stored in the stop target task table in accordance with task processing. Specifically, by monitoring inter-task communication, the task received in the communication is stored as a stop target task in relation to the task transmitted in the communication.

  The process for monitoring the communication between tasks will be described with reference to FIGS. Inter-task communication means that, for example, a task 72-1 writes communication contents to the inter-task communication storage area 16 of the memory 15, and another task 72-2 reads the communication contents from the inter-task communication storage area 16. It is communication realized by. Therefore, the inter-task communication storage area 16 is allocated for each inter-task communication. Although not shown in FIG. 8, a plurality of inter-task communication areas 16 are secured in the memory 15, and each inter-task communication area 16 is identified by communication resource information. That is, the communication resource information is information for identifying each inter-task communication area 16. The memory 15 also stores communication resource management information 90. The communication resource management information 90 is information obtained by associating the communication resource information with the program counter of the task performing the inter-task communication using the inter-task communication area 16 identified by the communication resource information. is there. Since there are two tasks on the transmitting side and the receiving side that perform inter-task communication using one inter-task communication area 16, there are two communication resource management information 90 including the same communication resource information. (Communication resource management information 90-1 and 90-2 shown in FIG. 9).

  The memory 15 further stores a task management table. This task management table is a table in which a task ID for identifying each task is associated with a program counter of the task.

  In the present embodiment, the communication resource management information 90 stored in the memory 15 and the communication resource management information 90 as described above are used to share tasks between the tasks using a common resource (here, the intertask communication area 16). Task pair information 92 indicating two tasks performing communication is generated. The task pair information 92 is configured to include task IDs of two tasks that perform communication between tasks using a common resource.

  FIG. 9 is a more specific example of the process of generating the task pair information 92 using the communication resource information and the communication resource management information 90 described above. First, the task state management unit 70 extracts communication resource management information 90 having the same communication resource information. Then, program counter pair information 91 in which two program counters using the inter-task communication area 16 identified by the same communication resource information is created is created. Then, the task ID stored in the task management table in association with the program counter included in the program counter pair information 91 is read. Then, by replacing the task ID read in this way with the program counter included in the program counter pair information 91, two tasks that use the inter-task communication area 16 identified by the same communication resource information, that is, tasks Task pair information 92 for associating two tasks in communication can be created.

  By creating the task pair information 92 in this way, a task performing inter-task communication can be monitored and stored as a stop target task in the stop target task table in association with one task. In other words, the stop target flag can be dynamically stored in the stop target task table, and a task that shares data with a certain task (that is, a task that references a common memory address) is defined as a stop target task. can do.

  Task stop processing for stopping a task with reference to the stop target flag stored as described above will be described with reference to the processing flowcharts shown in FIGS. 4 and 5, only the tasks 45 to 47 are taken up and described for the sake of simplicity. That is, FIG. 4 shows the region C portion of FIG.

  When a predetermined state (in this case, a failure state, for example, an illegal function call) is detected by the task 47 by the task state management unit 70 (S100), the task state management unit 70 first activates a stop task (S102). The stop task activated at this time is assigned time for using the target CPU 10 with the highest priority. The stop task is a task 46 that is stored as a stop target task in the column of task 47 (a stop target flag (stop target flags indicated by reference symbols C1 and C2 in the figure) is stored) The task 47 is stopped (S104, S106). Further, the stop task is a task other than the task that has already been stopped among the tasks stored as the stop target task in the column of the stopped task 46 (stop target flag (stop target flag indicated by C3 in the figure)) Is stored) is stopped (S108). In this way, when a certain task enters a predetermined state, the task associated with the stop target task can be stopped.

  By stopping the task when a certain task enters a predetermined state as described above, it is possible to efficiently debug the computer program. More specifically, since the state of the memory or transaction can be maintained in the state when the task is in a predetermined state, it can be confirmed after the state of the memory or transaction is in the predetermined state. Therefore, it is possible to debug the computer program efficiently. Of course, it is possible to efficiently detect hardware-derived failures.

  Note that when the information processing apparatus 1 is in a commercial operation state, there is a case where the priority is higher when the recovery is performed earlier than when the task is stopped and debugging is performed. In order to cope with such a case, the task state management unit 70 stores the operation mode and the failure analysis mode in the information processing apparatus 1 and stores the mode in the storage unit 71 so that the task state management unit 70 can select the mode. It is good also as performing the process according to. This processing will be described with reference to the processing flowchart shown in FIG.

  FIG. 6 is a flowchart of processing of the task state management unit 70 for performing processing according to the mode. First, mode setting is performed (S150). In the mode setting, a human may set the operation mode / failure analysis mode by an input device (not shown). The result of the mode setting is stored in the storage unit 71. That is, the storage unit 71 stores the operation mode / failure analysis mode. When the task state management unit 70 detects that a failure has occurred (S152), the task mode management unit 70 obtains whether the stored mode is the operation mode / failure analysis mode, thereby operating the current mode. Whether the mode is selected or not is determined (S154), and if it is the operation mode, the recovery process is performed (S160). In the restoration process, a transaction resetting process, a task resuming process, or a resuming process for the entire information processing apparatus 1 is performed according to the failure state. On the other hand, if the failure analysis mode is set, task stop processing is performed (S156). Then, the task state management unit 70 performs a predetermined task information acquisition process (S158). This may be automatic dump processing, for example. Moreover, the process which accepts confirmation of the memory state by a human may be sufficient. When the process ends, a recovery process is performed (S160). Then, the normal operation state is restored (S162). By doing so, the information processing apparatus 1 can perform debugging processing and recovery processing according to the mode.

It is a block diagram of the information processing apparatus which concerns on embodiment of this invention. It is a functional block diagram of an information processor concerning an embodiment of the invention. It is a stop object task table concerning an embodiment of the invention. It is a part of the stop target task table according to the embodiment of the present invention. It is a processing flow figure concerning an embodiment of the invention. It is a processing flow figure concerning an embodiment of the invention. It is a processing flow figure concerning an embodiment of the invention. It is explanatory drawing of the communication between tasks which concerns on embodiment of this invention. It is explanatory drawing of the communication monitoring process between tasks which concerns on embodiment of this invention.

Explanation of symbols

  1 Information processing device, 10 target CPU, 15 memory, 16 task communication area, 20 peripheral device, 30 program, 40 application, 50 driver, 60, 61 task group, 70 task state management unit, 71 storage unit, 72 Task, 90 communication resource management information, 91 program counter pair information, 92 task pair information.

Claims (9)

Task identification information storage means for storing task identification information for identifying a task and stop target task identification information for identifying at least one stop target task to be stopped when the task is in a predetermined state;
Predetermined state detecting means for detecting that the task is in the predetermined state;
Task stopping means for stopping the task to be stopped stored in association with the task in the predetermined state when the predetermined state detecting means detects that the predetermined state has been reached;
An information processing apparatus comprising: The information processing apparatus according to claim 1,
The task identification information storage means includes
Storing the at least one stop target task identification information in association with each of the plurality of task identification information;
An information processing apparatus characterized by that. The information processing apparatus according to claim 1 or 2,
The task stop means further stops the stop target task identified by the stop target task identification information stored in association with the task identification information for identifying the task that is the stop target task stopped by the task stop means.
An information processing apparatus characterized by that. The information processing apparatus according to claim 1,
The task stopping means is
Stop task starting means for starting a stop task when the predetermined state is detected by the predetermined state detecting means;
Including
The task stop means stops the task to be stopped by the started stop task.
An information processing apparatus characterized by that. The information processing apparatus according to claim 1,
When there is communication from the first task to the second task, the task identification information storage means displays the task identification information for identifying the second task, and the first task is in the predetermined state. Stored as stop target task identification information for identifying a stop target task to be stopped when
An information processing apparatus characterized by that. The information processing apparatus according to any one of claims 1 to 5,
The stop target task identified by the stop target task identification information stored by the task identification information storage means in association with the task identification information is a task that accesses a memory address common to the task identified by the task identification information Is,
An information processing apparatus characterized by that. The information processing apparatus according to claim 6,
Mode storage means for storing any one of the first mode and the second mode;
When the stored mode is the first mode, the task stop means is started, and when the stored mode is the second mode, mode corresponding processing means for performing a predetermined recovery process;
An information processing apparatus comprising: A predetermined state detecting step for detecting that the task has entered the predetermined state;
When it is detected in the predetermined state detecting step that the predetermined state is detected, task identification information for identifying the task in association with the task in the predetermined state, and the task is in the predetermined state A task stop step for stopping the stop target task stored by the task identification information storage means for storing the stop target task identification information for identifying at least one stop target task to be stopped,
A method for stopping a task. Predetermined state detecting means for detecting that the task has entered the predetermined state; and
When the predetermined state detecting means detects that the predetermined state is detected, the task identification information for identifying the task in association with the task in the predetermined state, and the task is in the predetermined state Task stop means for stopping the stop target task stored by the task identification information storage means for storing in association with stop target task identification information for identifying at least one stop target task to be stopped
A program characterized by causing a computer to function.

Images