JPS63137314A - Timer control system - Google Patents

Abstract

PURPOSE:To improve processing efficiency by comparing the value of a timer table with the value of a self-timer table at the time of processing a time-out event, when the result is dissident, inhibiting said processing. CONSTITUTION:A computer system 2 provided with a timer 1 has a timer control program 3, a body program 4 and each is provided with timer tables 5, 6, respectively. At the time of processing an event requiring time monitoring, the body program 3 adds '1' to the value of the timer table 6, then generates a time monitoring request to the timer control program 3. The program 3 adds '1' to the value of the timer table 5 on the basis of said request, and after the N times of interruption generated every fixed time, adds the value of the timer table 5 to N and forms time-out to the body program 4. The value of the timer table 5 is compared with the value of the timer table 6 and whether both the values are the same or not is checked to decide whether time-out event is generated or not.

Description

[Detailed Description of the Invention] [Summary] Since the timer control program that monitors time using interrupts from the timer and the main program that issues time monitoring requests to this timer control program operate asynchronously,
A timer table is provided for each to enable checking whether the time of the latest event processing is being monitored or not, so that the main program does not need to perform extra processing so that there is no inconvenience in processing when a timeout occurs. .

[Industrial application field]

The present invention relates to a computer system that monitors time using interrupts from a timer, and in particular, a timer control program that controls the timer and a main body program that is notified of the occurrence of a timeout from the timer control program operate asynchronously. This invention relates to a timer management method that eliminates the need for extra processing.

In computer systems, a timer is used to manage the time for processing an event, and if the processing of the event is not completed within a predetermined time, timeout processing is performed by an interrupt from the timer. For example, a timer monitors the time until normal termination of the input/output device, and if the input/output device does not terminate within a predetermined time, it is determined to be abnormal and timeout processing is performed.

By the way, the timer control program that controls this timer depends on the number of times the timer interrupts each unit time.
It manages time, and if there are N interruptions, it is determined that a timeout has occurred, and the main program is notified as a timeout occurrence event.

Since the main program processes various events that occur in the order in which they are notified, there is a time difference between the time when a timeout occurs and the time when the event due to this timeout is processed. processing occurs, and it is necessary that there be no such processing.

[Conventional technology]

FIG. 3 is a block diagram illustrating a conventional timer management system.

The main program 4 that controls the operation of the computer system 2 defines a state and a factor that prompts the transition of the state as an event,
By processing this, the state will be transitioned. When the main program 4 processes an event requiring time monitoring, it issues a time monitoring request to the timer control program 3.

The timer 1 issues an interrupt to the timer control program 3 at regular intervals, and the timer control program 3 notifies the main program 4 of a timeout when this interrupt occurs N times.

FIG. 4 is a time chart explaining the operation of FIG. 3.

If there is a processing queue of events to be processed, such as event A, event B, and event C1, as shown in the event processing queue, the main program 4 processes the events sequentially from the head of this queue. When processing event A, if time monitoring is required,
During the processing of event A, main program 4 sends a time monitoring request to timer control program program 3, as shown by ■.

The timer control program 3 operates asynchronously with the main body program 4, and due to this time monitoring request, timer 1 is activated at time T1.
Starts accumulating the number of interrupts that occur. Then, at time T2 when the event has been accumulated N times, a timeout event is created as shown in (3) and connected as event E to the event processing queue.

[Problem that the invention seeks to solve]

As mentioned above, the timeout as a result of requesting time monitoring in event A is connected to the event processing queue as event E, but main program 4 processes it as event B.

Since event C1 is processed sequentially and event E is processed thereafter, time T4 at which event E is processed is time T4.
There is a time difference later than 2.

When the main program 4 processes event C, if time monitoring is required, the timer control program 3 will have the following information as shown by ■.
A time monitoring request is sent during the processing of event C. Assuming that this time is T, this time is earlier than T4, so it cannot be determined whether the timeout processed by event E is the result of time monitoring requested by event A or time monitoring requested by event C. It disappears. Therefore, the main program 4 performs the following processing.

FIG. 5 is a flowchart illustrating processing when time monitoring is requested.

When the main program 4 requests time monitoring to the timer control program 3, it checks whether the timer is currently operating or not, and if the timer is operating, requests the timer control program 3 to stop the timer. If the timer is not running, the event processing queue is searched for a timeout event, such as event E, and if a timeout event exists, it is deleted.

After that, the main program 4 becomes the timer control program 3.
Issue a time monitoring request to.

Each time the main program 4 issues a time monitoring request to the timer control program 3, it repeats the process shown in the flowchart of FIG. 5, which requires extra processing and reduces the processing efficiency of the computer system.

[Means for solving problems]

FIG. 1 is a block diagram illustrating one embodiment of the present invention.

1 to 4 are the same as in FIG. 5 is a timer table provided in the timer control program 3, and 6 is a timer table provided in the main program 4.

When the main program 4 sends a time monitoring request to the timer control program 3, it increments the value in the timer table 6 by 1, and the timer control program 3 increments the value in the timer table 5 by 1 in response to this time monitoring request. Then, when a timeout occurs, the timer control program 3 notifies the value set in the timer table 5 when creating a timeout event for the main program 4.

When processing this timeout event, the main program 4 compares the value of this timer table 5 with the value of its own timer table 6, and if they are not the same, it is configured not to process this timeout event.

[Effect]

With the above configuration, when processing a timeout event, the main programmer 4 uses the value of the timer table 6,
By comparing the values in the timer table 5 added to timeout events and checking whether they are the same, it is possible to determine whether the timeout event corresponds to an event that requires timeout processing. The process shown in the flowchart of FIG. 5 for erasing missing timeout events can be made unnecessary.

〔Example〕

In FIG. 1, when the main program t processes an event requiring time monitoring, it increments the value of the timer table 6 by 1 and then issues a time monitoring request to the timer control program 3. The timer control program 3 increments the value of the timer table 5 by 1 in response to this time monitoring request.

The timer issues an interrupt to the timer control program 3 at regular intervals, and when the timer control program 3 receives this interrupt eight times, it adds the value of the timer table 5 and notifies the main program 4 of the timeout.

FIG. 2 is a time chart explaining the operation of FIG. 1.

As shown in the event processing queue, if there is a processing queue of events to be processed, such as event A, event B, and event C1, the main program 4 processes the events sequentially from the head of the queue. When processing event A, if time monitoring is required,
The main program 4 sets the value of the timer table 6 to "1" during processing for an event, and sends a time monitoring request to the timer control program 3 as shown by ■.

The timer control program 3 operates asynchronously with the main program 4, and after setting the value of the timer table 5 to "1" in response to this time monitoring request, it starts accumulating the number of interrupts received from the timer 1 at time T1.

Then, at time T2, which has been accumulated N times, a timeout event is created as shown in ■, the value rlJ of the timer table 5 is added, and the event is connected as event E to the event processing queue.

The main program 4 processes each event sequentially, such as event B and event C1, but if time monitoring is required when processing event C, the value of the timer table 6 is changed to "
2'', a time monitoring request is sent to the timer control program 3 as shown in ■.

The timer control program 3 sets the value of the timer table 5 to "2" in response to this time monitoring request, and then starts accumulating the number of interrupts received from the timer 1 at time T3.

When the main program 4 processes event E, it compares the value "2" in the timer table 6 with the value "1" in the timer table 5 added to the event value.

In this case, since the comparison values do not match, it is determined that event E has already timed out and does not need to be processed, and event E is not processed.

〔Effect of the invention〕

As explained above, in the present invention, by comparing the values in the timer table, the main program can recognize whether or not the event requires timeout processing, so each time a time monitoring request is issued to the timer control program. There is no need to check whether the timer is currently running, and if it is not running, check whether a timeout event is connected to the event processing queue, and if so, delete that event. Therefore, the processing efficiency of the computer system can be improved.

[Brief explanation of the drawing]

FIG. 1 is a block diagram explaining one embodiment of the present invention, and FIG.
The figures are a time chart explaining the operation of Fig. 1, Fig. 3 is a block diagram explaining the conventional timer management method, Fig. 4 is a time chart explaining the operation of Fig. 3, and Fig. 5 is a time chart explaining the operation of Fig. 3.
The figure is a flowchart illustrating processing when a time monitoring request is made. In the figure, 1 is a timer, 2 is a computer system, 3 is a timer control program, 4 is a main program, and 5.6 is a timer table. Sorrow 2 Mouth Spider Timer Contains Method 2 1 Late familiar Blown 20 茅 3 Guo 300) 2 I Otoko Yard 4
− B

Claims (1)

[Claims] A timer control program (3) that monitors a predetermined time using an interrupt from the timer (1), and a time monitoring request that is sent to the timer control program (3) so that the predetermined time elapses. When a timeout occurs, the timer control program (3
), in a computer system (2) controlled by a main program (4) that connects a timeout event to an event processing queue and performs timeout processing, the main program (4) has a first timer table (6).
A second timer table (5) is provided in the timer control program (3), and when the main program (4) sends a time monitoring request to the timer control program (3), the first timer table (5) is
After receiving this time monitoring request, the timer control program (3) also increases the value of the second timer table (5) by 1, and then changes the time by an interrupt from the timer (1). When monitoring is started and the elapse of the predetermined time is detected, the timeout event to which the value of the second timer table (5) has been added is connected to the event processing queue of the main body program (4). Program (4) uses the value of this added second timer table (5) and the first timer table.
A timer management method characterized in that a timeout event to be processed is selected by comparing values of a timer table (6).