JPS62271147A - Task control method - Google Patents

Abstract

PURPOSE:To prevent the overhead of a monitor from being increased even if total number of tasks to be processed by the time division system is increased by dividing tasks into high priority tasks and lower priority tasks and regarding the lower priority tasks as one unified task thereby decreasing the overhead of a monitor for the high priority tasks. CONSTITUTION:Let A21 and B22 be the high priority tasks, C1, 23, C2,24 and C3,25 be lower priority tasks, then all the lower priority tasks are regarded as one task C26 with respect to the high priority tasks, and the priority of the C26 is processed as the 3rd rank. Thus, the tasks requiring the processing by the high priority tasks are three, A21, B22 and C26. Since the lower priority tasks are checked only when no execution standby task exists in the high priority task, the task number of the high priority tasks has only to be decreased to reduce the overhead of the monitor with respect to the high priority tasks.

Description

[Detailed Description of the Invention] 3. Detailed Description of the Invention [Field of Industrial Application] The present invention relates to a task control method when a plurality of tasks are processed in real time in a time-sharing manner.

[Conventional method]

Conventional task control methods can be broadly divided into two types: (1) Fixed priority method, (2) Round-tropin method, (3) Combination method of fixed-priority method and Round-tropin method.

In the fixed priority method, the importance of processing (=therefore, priority 1@1) is determined for all tasks among a plurality of tasks to be processed, and tasks are activated according to their priority.

In the round-tropin method, the priority of tasks is not fixed, and tasks are activated in the order in which activation requests are received.

Since the combination of the fixed priority method and the round-tropin method is most widely used, this method will be described below.

FIG. 3 is a block diagram showing the relationship between a plurality of tasks and monitors that control these tasks in a conventional method. C1-C4, D
i to D4 represent each task group, and a priority order is determined for each task group, with the Al-A4 task group in first place, the B1-84 task group in second place, and C1-
It is assumed that the task group C4 is ranked third, the task groups D1 to D4 are ranked fourth, and the priorities of the tasks in the same task group are the same. (20) is a monitor that controls these tasks.

When some event causes control to be transferred to the dispatcher (a program that selects and starts tasks) in the monitor (20), the dispatcher selects tasks that are suspended in execution or tasks that are waiting to be started in each task group according to the priority order. Find out if there is. When a task whose execution is suspended is detected, control is returned to that task, and if there is no task whose execution is suspended in the task group, but there is a task waiting to be started, one of the tasks waiting to be started is returned. The earliest (
=Start the task that was requested to be started.

[Problem that the invention seeks to solve]

The conventional method is performed as described above, so if the number of task groups and the number of tasks in the same task group is large, monitor (
20) becomes the next task (= the amount of processing required to determine the task to which control should be transferred increases, and therefore the overhead of the monitor (20) becomes too large, reducing the throughput of the program. There was a problem in that real-time processing was not possible for tasks that required

This invention was made to solve the above-mentioned problems, and its purpose is to provide a task control method that does not increase monitor overhead even if the total number of tasks to be processed in a time-sharing manner increases. do.

[Means for solving problems]

In this invention, the priority tasks are divided into high-priority tasks from 1st to mth priority, and lower-priority tasks with the (rn+1)th priority, and the lower-priority tasks are the sum of all tasks. Regarding it as one task C: Therefore, the total number of tasks above from the perspective of the higher priority task has been reduced, and the monitoring overhead for the higher priority task has been reduced.

[Effect]

In this invention, lower priority tasks are checked only when there are no tasks waiting to be executed among the higher priority tasks, so if the number of higher priority tasks is kept small, the overhead of monitoring the higher priority tasks can be reduced. .

〔Example〕

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

FIG. 1 is a flowchart showing an embodiment of the present invention, and in the figure (1) to (11) are each step. FIG. 2 is a block diagram showing the priority order in this invention. In the 8g2 diagram, (20) is the monitor, (21) is the task A with the first priority, and (22) is the task A with the first priority. Task B is in second place, (23) is task CI in third place in priority order, and task C2 is in fourth place in priority order (24).
, (25) represents task C3, which is ranked fifth in priority.

(26) c1 (23), C2 (24).

This is task C, which considers C3 (25) as one task.

In this invention, A(21) and B(22) are set as high priority tasks, and CI(23), C2(24), C3(
25) as a lower priority task, and all lower priority tasks CI C23) , C2(24) , C3(25)
is the higher priority task (one task C for two (26)
Therefore, task C (26) is processed as having the third priority.

Therefore, from the perspective of higher priority tasks, the tasks that require processing are A (21), B (22), C (, 26
).

Next, FIG. 1 will be explained. Activation of a task is performed by some event, but when an event requests activation of a task, the request is made separately for a higher priority task and a lower priority task.

At steps +11 and +21, tasks A(21) and B(
22) is waiting for execution, and if none is waiting for execution, the process moves to step (3). In this case, if there is a start request for a lower priority task, the process for that start request:: is performed. .

If the determination in step (1) is Y, the process moves to step (4) to check the task being executed, but if it is A, it means that the task activation has already been completed.

If the task being executed is B (22) or C (26), the contents of the registers used by the task being executed are saved in step (5) or (6), and the next (: these tasks are restarted. After making it available for use when doing so, proceed to step (7) to start the transfer task person.

The judgment in step (1) is N and the judgment in step (2) is Y.
If , go to step (8), check the running task, and if this is A, task A (21) is better than task B.
Since the priority is higher than (22), step (
9) Task B is placed on standby. If the executing task is B in step (8), it means that the task activation has been completed. If the task being executed is C, step (
10) and (11) are performed. Also, step (3)
In the processing of (a), (b), and (b) explained in the conventional technology,
Any method of i/field may be used.

According to the above processing, the lower priority task C1 (23
), C2 (24), and C3 (25) are double-checked, so their processing is slightly delayed, but from the perspective of the higher priority tasks, the total number of tasks has decreased (=,
Monitor overhead is small.

In the above embodiment, the number of high priority tasks is 2 and the number of low priority tasks is 3, but in the present invention, it is only necessary to limit the number of high priority tasks to a relatively small number, and there is no other limitation to these numbers of tasks. In general, tasks with priority rankings 1 to m are considered high priority tasks, and the (m+1)
All tasks that have been demoted are made lower priority tasks.

〔Effect of the invention〕

As described above, according to this invention (:), the selection of higher priority tasks that need to be processed is made faster, so the overhead of monitoring the higher priority tasks can be reduced.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing an embodiment of the present invention, FIG. 2 is a block diagram showing priority order in this invention, and FIG. 3 is a diagram showing a plurality of tasks and a monitor for controlling these tasks in a conventional method. Block diagram showing relationships. (20) Monitor, (21) is task A with the first priority
, (22) is the second priority task B, and (26) is the lower priority task c0. In each figure, the same reference numerals indicate the same or corresponding parts.

Claims (1)

[Claims] In a task control method for performing real-time processing of a plurality of tasks in a time-sharing manner, a processing priority is determined for each task of the plurality of tasks, and a processing priority is determined from the first to the first in the processing priority. A step in which tasks up to the mth rank (m is an arbitrary positive integer and is selected to be a relatively small value) are set as high priority tasks, and all tasks below the (m+1)th rank of processing priority are set as low priority tasks. The processing priority of the lower priority task is (m+1)
A stage in which tasks are controlled according to the processing priority within a task group composed of this one task and all higher priority tasks, and tasks currently being executed among the higher priority tasks are also tasks waiting to be executed. A task control method characterized by comprising the step of performing task control according to processing priority among all lower priority tasks in a state where no lower priority tasks exist.