KR20030077696A - Method for task level interrupt handling in realtime operating system - Google Patents

Abstract

PURPOSE: A method for processing a task level interrupt on an RTOS(Real-Time OS(Operating System)) is provided to minimize an interrupt processing time, to reduce interrupt loss, and to enhance system performance by executing an ISR(Interrupt Service Routine) on a task level and managing a delay of the interrupt executing time on an ISR task level. CONSTITUTION: The interrupt process is started if an interrupt is generated(ST12). An interrupt level is calculated after disabling the interrupt(ST14). In case of the calculated interrupt under a preset level, the calculated interrupt is masked(ST15), and a table value of an exception handler registered to a matched level is calculated(ST16). The interrupt is enabled(ST17) and the registered exception handler is executed(ST18). After executing the routine registered to an interrupt vector table, the interrupt is disabled(ST20). The interrupt under the corresponding level is unmasked(ST21). If the interrupt process is completed, the ISR is executed on the ISR task(ST23).

Description

Method for task level interrupt handling in realtime operating system}

The present invention relates to a task level interrupt processing method in which an interrupt handling routine is implemented at a task level in a real time operating system (RTOS). In particular, the interrupt handling routine is registered in an interrupt task to register an ISR at the task level. (Interrupt Service Routine) to minimize the interrupt processing time, reduce the occurrence of interrupt loss, etc., and to give flexibility to the implementation of the ISR task level in the real-time operating system to improve the performance of the system An interrupt processing method.

In general, a real-time operating system generally refers to a multiple programming operating system using a disk, real-time multiple programming at a high priority, and batch processing at a low priority background. Program input from and output to the disk is performed automatically and uses dynamic main memory allocation.

Here, the real time operating system (RTOS) can be classified into two types in terms of structure. One is the Multi Thread model, and the other is the Multi Process model.

The multi-threaded model is a structure in which an operating system (OS) kernel and an application are combined to operate as one large program without being distinguished from each other, and can freely access a common work area (Memory). Although the operating system is small in size and relatively small in size, it is easy and quick to implement. However, since a kernel and an application operate as one program, a minor bug destroys the entire system.

On the other hand, the multi-process model is designed such that the operating system kernel or each application operates as an independent program. Since each application is protected with each other's memory, it is easy to develop a large-scale system because it is easy to develop a module-level application, add or change a module (function), and develop a stable system. However, since the size of the real-time operating system (RTOS) is larger than the multi-threaded model, there is a disadvantage in that a small system is developed.

Considering the advantages and disadvantages of the two models of the real-time operating system, a multi-threaded model is used for developing a relatively small and uncomplicated system, and a multi-process model for a large-scale complex system such as a medical device. It is known to use.

On the other hand, interrupt processing in the real-time operating system as described above has a big impact on system performance. Therefore, the performance of the system should be improved by minimizing the interrupt execution delay by reducing the time of interrupt disable and interrupt mask that occur during interrupt processing time.

To this end, in the related art, interrupts and process masks are minimized by minimizing an interrupt service routine as shown in FIG. 1 and a method of increasing the effect of overall scheduling by associating an interrupt with a processor scheduler. The method of shortening the period was used.

However, in the above method, the execution of the routine registered in the vector table is completed, and then the processing proceeds to the next step, where the interrupt processing ends.

Therefore, when dealing with interrupts of complex hardware, it is difficult to see a particular effect, and in addition, the interrupt execution period must be taken into consideration in implementing an Interrupt Service Routine (ISR).

Accordingly, the present invention is proposed to solve the above problems of the prior art,

An object of the present invention is to register an interrupt processing routine in an interrupt task in a real-time operating system (RTOS) to perform Interrupt Service Routine (ISR) at the task level, and due to the execution of the ISR By managing the delay of the delayed interrupt execution time at the ISR task level, it is possible to minimize the interrupt processing time, reduce the occurrence of interrupt loss, and improve the performance of the system. To provide.

The present invention for achieving the above object,

Starting interrupt processing when an interrupt occurs;

Calculating an interrupt level after the interrupt is disabled when the interrupt processing is started;

Masking the calculated interrupt level when the interrupt level is equal to or lower than a predetermined level, and calculating a table value of an exception handler registered corresponding to the level;

After calculating the table value, enabling an interrupt and executing a registered exception handler;

Disabling the interrupt after performing the routine registered in the interrupt vector table when the exception handler is finished;

After disabling the interrupt, unmasking an interrupt below a corresponding level;

After the interrupt unmasking, it is checked whether or not the interrupt processing is completed, and as a result, if the interrupt processing is completed, the method comprises the step of performing an ISR in an Interrupt Service Routine (ISR) task.

1 is a diagram illustrating a task level interrupt processing process in a real-time operating system according to the prior art,

2 is a diagram illustrating a task level interrupt processing process in a real-time operating system according to the present invention,

3 is a flowchart illustrating a task level interrupt processing method in a real-time operating system according to the present invention.

<Explanation of symbols for the main parts of the drawings>

100 ..... Physical Memory

200 ..... Queue

300 ..... ISR Processing Task

400 ..... Interrupt Vector Table

Hereinafter, a preferred embodiment of the "task level interrupt processing method in a real-time operating system" of the present invention according to the above technical concept will be described in detail with reference to the accompanying drawings.

2 is a diagram illustrating a task level interrupt processing process in a real-time operating system according to the present invention.

As shown in the drawing, the routines registered in the interrupt vector table are the same as the conventional method, and all the routines registered in the interrupt vector table are assigned to the Interrupt Service Routine (ISR) processing task (Vector Number) and parameters. (Parameter) Record only as a simple routine to pass the value.

In this case, since the handler only needs to send the minimum information, the handler quickly returns to the state to receive the next interrupt, thereby reducing the loss of the interrupt.

In addition, since the ISR is performed at the task level, the routine can be implemented with less restrictions, and the ISR processing task sets the priority level as the highest priority, so that the routine for all interrupts can be set. In case of interrupt, ISR is executed by scheduling with corresponding vector and parameter value.

Therefore, ISR is managed at the task level to manage the delay of interrupt execution time delayed by the execution of Interrupt Service Routine (ISR), which minimizes interrupt execution time and provides flexibility in the implementation of ISR. It can be applied in relation to interrupt handling routines, and in particular, it can be utilized for device driver development and interrupt processing on an operating system (OS).

3 is a flowchart illustrating a task level interrupt processing method in a real-time operating system according to the present invention.

As shown therein, when an interrupt occurs (ST11), interrupt processing starts (ST12), disables the interrupt (ST13), and after disabling the interrupt, sets the corresponding interrupt level (Interrupt Level). Calculating (ST14), and after calculating the interrupt level, masking the calculated interrupt level if it is an interrupt below a predetermined level (ST15), and corresponding to the level after the interrupt mask. Calculating a table value of a registered exception handler (ST16), enabling the interrupt after calculating the table value (ST17), and registering the exception handler after the interrupt is enabled. Performing the routine (ST18), and when the exception handler is finished, executing the routine registered in the interrupt vector table (ST19), and performing the interrupt after performing the routine. Disabling (ST20), after disabling the interrupt, unmasking an interrupt below the level (ST21), and checking the completion of interrupt processing after the interrupt unmasking (ST22), As a result, when the interrupt processing is completed, a step (ST23) of performing an ISR in an Interrupt Service Routine (ISR) task is performed.

The present invention having the above configuration will be described in more detail as follows.

That is, when an interrupt is generated in step ST11, in step ST12, the generated interrupt processing is started (ST12). When interrupt processing is started, the interrupt is disabled as in step ST13.

After the interrupt is disabled, the corresponding interrupt level is calculated in step ST14, and the interrupt generated below the corresponding interrupt level calculated in step ST15 is checked to perform a mask.

Here, the mask is to suppress the occurrence of the interrupt in the operating system (OS), which is called the interrupt mask (Interrupt Mask), even if an interrupt is generated by the interrupt mask until the interrupt is generated until it is released Is ignored.

On the other hand, after the interrupt mask, in step ST16, a table value of an exception handler registered at the corresponding level is calculated, wherein the exception causes a temporary interruption of normal execution processing during program execution. It is a general term for the event which causes a cause, the unexpected data coming in, etc., for example, the overflow when an arithmetic operation is performed, the fault which arises in a peripheral device, etc. are included. In general, for the occurrence of such an exception, a program has a built-in routine for each exception, executes the routine, takes appropriate action, and returns to normal processing. Such processing is called exception handling, and the routine that handles it is called an exception handler.

On the other hand, after calculating the table value of the exception handler, in step ST17, the interrupt is enabled, and in step ST18, the exception handler registered at the corresponding level is performed.

After performing the exception handler, in step ST19, the routine registered in the interrupt vector table is executed, and in step ST20, the interrupt is disabled.

After disabling the interrupt, as in step ST21, the interrupt below the level is unmasked, and in step ST22, it is checked whether the interrupt processing is completed. When the interrupt processing is completed as a result of the check, the ISR is performed in the ISR task (Interrupt Service Routine Task) in step ST23.

According to the present invention described above, the "task level interrupt processing method in a real-time operating system", an interrupt processing routine is registered in an interrupt task in a real-time operating system (RTOS) to perform an Interrupt Service Routine (ISR) at a task level. In addition, by managing the delay of the interrupt execution time delayed by the ISR execution at the ISR task level, it is possible to minimize the interrupt processing process, thereby improving the performance of the system.

In addition, developers can considerably reduce problems such as interrupts caused by ISR implementation, and have the advantage of making the ISR more flexible.

Claims (1)

In the method for handling task level interrupt in a real-time operating system, Starting interrupt processing when an interrupt occurs; Calculating an interrupt level after the interrupt is disabled when the interrupt processing is started; Masking the calculated interrupt level when the interrupt level is equal to or lower than a predetermined level, and calculating a table value of an exception handler registered corresponding to the level; Enabling the interrupt and calculating the registered exception handler after calculating the table value; Disabling the interrupt after performing the routine registered in the interrupt vector table when the exception handler finishes performing; Unmasking an interrupt below the level after the interrupt is disabled; After the interrupt unmasking, checking whether the interrupt processing is completed, and if the interrupt processing is completed as a result, performing the ISR in the Interrupt Service Routine (ISR) task. Way.