JPH01120635A - Data control method for inter-process communication - Google Patents

Abstract

PURPOSE:To cut down the program developing cost by using a double layer data chain which includes a data chain block and a frame chain block. CONSTITUTION:The inter-process communication chain saving areas 101 and 102 are provided together with generation of a process. Then the transmission data given to the process from another one are expressed by connecting the data chain blocks 103-105 and 108 to the chain starting at the areas 101 and 102 and also connecting the frame chain blocks 106 and 107 to the chain starting at a frame chain saving area formed in a data chain block. Thus the transmission data can be transferred between processes by releasing the chains of both blocks 106 and 107 as well as the chains of blocks 103-105 and 108 respectively.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a data management method, particularly to a data management method in interprocess communication.

[Technological environment]

In recent distributed combinator systems, a large number of processes operate in parallel, and it is common to perform so-called N:N interprocess communication with each other. With the recent enhancement of microprocessor functions and the spread of personal computers, distributed computers using multi-microprocessors and
As systems are rapidly becoming popular, inter-process communication is also changing from the traditional method of simply synchronizing to advanced communication that organically connects processes. Along with this, the number of man-hours required for software design/manufacturing continues to increase, posing a major problem.

[Conventional technology]

The conventional data management method for inter-process communication is that in a small and simple system, a chain is created for each frame in the order in which the 7 frames arrive, and frames are organized within the process by source process and data ID. It is carried out. In large general-purpose systems, a method is used in which data is passed to a process using a buffer consisting of continuous areas for the source process and the ID* of the data.

[Problem that the invention seeks to solve]

The conventional data management method for inter-process communication described above imposes a high processing load within a process in a small and simple system, and requires advanced software development techniques. On the other hand, in large general-purpose systems, it is necessary to always secure a contiguous area in memory in order to simplify processing within a process.

The internal characteristics of processes using personal computers, etc., which have been increasing noticeably in recent years, are more complex than before.
Although the capacity is increasing, it is difficult to secure a large contiguous area due to the memory capacity, and even with the above technology, it is not possible to suppress the increase in software development costs and the decrease in execution efficiency. was there.

[Means for solving problems]

In the data management method for inter-process communication of the present invention, an inter-process communication chain save area is provided as soon as a process is generated, data sent to the process from other processes is saved, and data chain blocks are saved in the inter-process communication chain save area. This is expressed by connecting the frame chain block to a chain starting from the frame chain save area provided in the data chain block, and transmitting data sent from another process to the process to the frame chain block. The above problem is solved by removing the chain of the data chain block and removing the chain of the data chain block.

〔Example〕

Next, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is an explanatory diagram showing one embodiment of the present invention.

In FIG. 1, an interprocess communication chain save area is provided when a target process is generated.

When the transmission data from the conversion process arrives divided into frames, first allocate one data chain block and one frame chain block, and write the 7 frames x 7 a to the frame buffer address 7 area of the frame chain block. The address of the 7 frame chain block is stored in the head frame safe area 104 and the tail frame 7 area 105 of the data chain block. If there are multiple frames, one frame chain block is secured for frame 11i1, the frame buffer address is stored in the frame buffer address save area in the newly secured frame chain block, and the data chain block text is ] The address of the newly secured frame chain block is stored in the next frame save area and the tail frame safe area 105 in the frame chain block indicated by the gate frame save area 1o5. If this is repeated in sequence until the data is completed, the frames are chained in the order of arrival, forming one data chain managed by the head frame save area 104 and tail frame save area 105 of the data chain block.

On the other hand, the priority of data sent from other processes to the target process is not the same. In this embodiment, the priority of data is digitized and stored inside the first frame, and the source process Set this value. This priority is stored in the data priority save area when a data chain block is first secured, and is set between data line processes that are sent from other processes and are waiting to be processed. Starting from the head save area 101 of the communication chain save area, they are chained in descending order of priority. When inheriting a newly allocated data chain block, the priorities are compared in order from the first data chain block, and the data chain block with a lower priority than the newly allocated data chain block is inherited. When you encounter one, connect the newly secured data chain block to that position. The connection method is to first store the address of the data chain block with a lower priority in the next save area of the newly secured data chain block, and then store one of the data chain blocks with a lower priority. It was chained before. The address of the newly secured data chain block is stored in the next save area of the data chain block that has the same or higher priority than the newly secured data chain block.

The target process extracts and processes data while removing the chain according to the data chain structure of this embodiment.

〔Effect of the invention〕

The data management method in inter-process communication of the present invention simplifies frame management within a process, by source process, and by priority, by using a two-layer data chain called a data chain block and a frame chain block. Compared to the past, programming techniques such as advanced table management are not required, so a reduction in program development costs can be expected.

In addition, there is no need to secure a large contiguous area in memory as is the case with methods used on large machines, and the management unit can be managed in relatively small units called frames, and processes can be generated/disappeared and data reception can be released. In addition, by dynamically allocating and releasing memory, an efficient system can be constructed even on a computer with a relatively small memory capacity, such as a personal combinator.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram showing one embodiment of the present invention. ...Head frame save area, 105...Tail frame save area, 106...Next frame safe area, 107...Frame buffer address save area, 108...Data priority save area. Agent Patent Attorney Susumu Uchihara

Claims (1)

[Claims] 1. An inter-process communication chain save area is provided when a process is generated, and the data chain block for transmitting data to the process from another process is connected to a chain starting from the inter-process communication chain save area. And, the frame chain block is expressed by continuing the chain starting from the frame chain save area provided in the data chain block, and the data sent to the process from another process is removed from the chain of the frame chain block and transferred to the data chain. A data management method in inter-process communication characterized by solving the problem by unchaining blocks. 2. The inter-process communication chain save area includes a head save area and a tail save area, and the data chain block includes a next save area, a head frame save area, and a tail frame save area. 2. The data management method in inter-process communication according to claim 1, wherein the frame chain block includes a next frame save area and a frame buffer address save area. 3. A data management method in inter-process communication according to claim 1, in which, when a data chain block is to be joined to a chain starting from an inter-process chain save area, it is joined to an arbitrary position in the chain. 4. When continuing a frame chain block to a chain starting from a frame chain save area provided in a data chain block, enter the address of the buffer in which the contents of the corresponding frame are stored in the frame buffer address area in the frame chain block. Claim 1 in which the frames are stored and form a chain in the order of frame occurrence.
Data management method in interprocess communication described in Section 2.