JPH0447336A - Task synchronization controlling method for multi-task system - Google Patents

Abstract

PURPOSE:To secure synchronism between tasks even if an event flag is set two times in succession by resetting the event flag while a system is being kept in an interruption state at the same time as the task in an event waiting state has the event flag set and starts start-up. CONSTITUTION:When a condition is concluded by setting the event flag of an interruption handler (or other high-level task) 5, the task 1 in the event waiting processing (state) is made to withdraw from the event waiting processing 3 after turning OF (1 0) the bit of the event flag to meet this condition while the system is left in the interruption inhibition state. Thus, even if the interruption handler (or other task) 5 sets the event flag in succession, the synchronism between the tasks can be secured.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to a method for controlling task synchronization in a multitasking system in which one CPU executes multiple tasks that may occur in parallel at the same time. Note that in the following figures, the same reference numerals indicate the same or corresponding parts.

[Conventional technology]

In a multitasking system in which one CPU executes a plurality of tasks that may occur in parallel at the same time, event flags are used to perform synchronization control between the plurality of tasks. FIG. 2 is a pad chart showing an example of the above-mentioned synchronous control method in a conventional multitasking system. In the figure, the self-task 1 waits for an event 31, resets the event flag 32, processes the self-task 1, as shown in FIG.
It is assumed that each process is executed in this order, and this sequential process is executed in a perpetual cycle (loop). Also in the same figure [Bl
indicates processing related to the self-task 1 of the interrupt handler (or another task at a higher level than the self-task 1) 5. That is, when the invoking task 1 is in the event waiting process (condition B) 31,
When the interrupt handler (or other task) 5 performs the event flag setting process 6, task 1 starts execution if the event flag set as described above satisfies the conditions, and in the next 32 processes, the event flag is set. After resetting the flag and executing the own task 1 in the process 4, the process returns to the event waiting process (state) 31 again. At this time, since the event flag has already been reset, the self-task 1 remains in the event waiting state.

[Problem to be solved by the invention]

However, it is assumed that the interrupt handler (or other task) 5 in FIG. 2 is executed twice in succession, and the event flag is set 6. Then, task 1 sets the event flag for the first time and performs event waiting processing (state) 31.
Since this is canceled, this process 31 is exited. If the event flag is set for the second time before executing the event flag reset in the next process 32, the event flag that was set for the second time will also be reset in the process 32. Therefore, when the task 1 reaches the event waiting process 31 after the process 4 of its own task 1, since the event flag has been reset, the task 1 remains in the event waiting state. In this way, although the side that is activated (interrupt handler (or other task) 5) is executed twice, the side that is activated (self-task 1) is executed only once, and there is no synchronization between the two. There was a problem that I couldn't get it. Therefore, it is an object of the present invention to provide a task synchronization control method that can solve this problem.

[Means to solve the problem]

In order to solve the above problems, the method of the present invention is based on the setting of an event flag by an r interrupt handler or other high-level task (such as 5) indicating that the self-task (such as 1) should be activated. In a multitasking system that exits from an event waiting state (such as 3) and processes its own task (such as 4), when exiting from the event waiting state, the event flag is immediately reset while interrupts are disabled. It shall be.

[For use]

Based on the event flag set based on the interrupt of a high-level task by the interrupt handler or O3, when attempting to start processing of the task from the state of waiting for an event, first reset the event flag while the interrupt is disabled and wait for the corresponding task. By entering task processing, even if the event flag is set twice in succession, the event flag will be set to 1 until the second event flag.
This prevents the event flag from being reset together with the second event flag.

【Example】

FIG. 1 is a pad chart as an embodiment of the present invention, and corresponds to FIG. 2. In FIG. 1, processes 31 and 32 in FIG. 2 are replaced with process 3 of waiting for an event with event flag reset. In other words, when a condition is met by setting the event flag of the interrupt handler (or other high-level task) 5, task 1 in the event wait processing (state) will remain in the interrupt disabled state and select the bit of the event flag that meets the condition. off(
1→O), and then exits from this event waiting process 3. That way, even if interrupt handler (or other task) 5 successively sets the event flag in process 6, task 1 will exit event waiting process 3 by setting the event flag the first time. Even if the event flag is reset at the same time and then set for the second time, it will not be reset because there is no reset processing for the event flag. After executing process 4 of self-task 1, task 1 returns to event waiting process 3, but at this time, the second event flag has already been set, and the abnormal behavior start condition for self-task 1 is satisfied. So,
Immediately exits the event waiting process 3 and at the same time resets the event flag while maintaining the interrupt disabled state. When the task 1 returns to the event wait process 3 after the second process 4 of its own task 1, the event flag has been reset, so the task 1 remains in the event wait state. The side that starts like this (interrupt handler (or task)
5) is executed twice, while the activated side (task 1) is also executed twice, so the two can be synchronized.

【Effect of the invention】

According to the present invention, when a task in the event waiting state starts activation with the event flag set, the event flag is reset while the interrupt is disabled, so that the event flag is set twice in succession. Synchronization between tasks can be achieved even if the tasks are performed multiple times.

[Brief explanation of the drawing]

FIG. 1 is a pad chart as an embodiment of the present invention, and FIG. 2 is a conventional pad chart corresponding to FIG. 1. 1: Invoking task, 2: Eternal loop processing, 3: Event wait processing with event flag reset, 4: Invoking task processing, 5: Interrupt handler or other task, 6: Event flag setting processing. Diagram

Claims (1)

[Scope of Claims] 1) A multitasking system in which an interrupt handler or other high-level task sets an event flag indicating that the task should be started, and exits from a state of waiting for an event and processes the task. A task synchronization control method for a multitasking system, characterized in that when exiting from the event waiting state, the event flag is immediately reset while interrupts are disabled.