JPH02302831A - Data transfer control system - Google Patents

Abstract

PURPOSE:To rapidly and surely transfer data to another task by writing data while referring a memory control table and transferring all or a part of the memory control table. CONSTITUTION:In the case of transferring data from a 1st task A to a 2nd task B, the 1st task A writes data while referring its own memory control table 51 and transfers the table 51 to the 2nd task B. The 2nd task B refers the transferred memory control table 51 as its own memory control table to access the data. Consequently, data transfer between the tasks A, B can be quickly ended and respective tasks can be smoothly processed.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a data transfer method between tasks in a computer system that can execute a plurality of tasks in parallel.

(Prior Art) In recent years, some computer systems are attempting to improve their processing capacity by processing multiple tasks in parallel. Additionally, virtual memory systems are also employed in many computer systems in order to allow each task to be executed without being aware of the storage capacity limitations of the main memory. This virtual memory system not only has great effects when executing relatively large-scale tasks, but also plays an important role when performing parallel processing of tasks. That is, in a virtual memory system, the address space used by each task is unique to the task, and each task is prevented from referencing the address space of another task. Thereby, the address space of each task can be reliably protected, and confidentiality and reliability in the computer system can be maintained.

FIG. 2 shows a configuration diagram of a general computer system.

In the figure, this computer system is comprised of a system console 1, a microprocessor 2, a main storage 3, a secondary storage 4, a floppy disk drive 5, and a printer 6. Further, a plurality of terminal devices 8 are connected to the input/output terminal 7.

The system console 1 is composed of a keyboard, a display, etc., and is a device used by an operator to control the system. The microprocessor 2 controls and manages this system. The main storage device 3 is an RA for loading calculation programs and data.
It is composed of M, etc. The secondary storage device 4 is composed of a hard disk drive or the like that stores and saves various types of information. The floppy disk drive 5 is an external storage device using a floppy disk. The printer 6 is a device that prints various information.

Now, in the computer system as described above, when a task is executed, the task program is first read from the secondary storage device 4 and loaded into the main storage device 3. However, main memory 3
Since there is a limit to the storage capacity of the main storage device 3, it is not possible to load all the programs necessary for a task into the main storage device 3. For this reason, there is a method in which the secondary storage device 4 compensates for the amount that cannot be loaded.

This method loads the minimum necessary programs into the main storage device 3, and loads programs and data from the secondary storage device 4 onto the main storage device 3 each time the need arises. Programs and data that are not stored are transferred from the main storage device 3 to the secondary storage device 4.

In this method, a program is created on a logical storage device that is given almost unlimitedly, and a virtual address is assigned to this logical storage space, while a real address is assigned to the main storage device 3 and the virtual memory configuring the system.

(Problem to be Solved by the Invention) Incidentally, in the computer system as described above, each task may be set to perform a predetermined process exclusively. For example, the first task specializes in receiving data input via a communication line, and the second
It is assumed that this task specializes in processing based on the data received by the first task. In such a case, it becomes necessary to transfer data from the first task to the second task. However, since each task executes processing without being related to other tasks, for example, the first task cannot directly transfer data to the task in clause 2. As explained earlier, this is
This is because the address space of each task is protected.

Therefore, a function related to data transfer between tasks is required.

FIG. 3 shows a conceptual diagram of a conventional data transfer function.

In the figure, it is assumed that a first task A transmits data and a second task B receives data. The first task A secures and occupies an address space 11 on the main storage device 3, and similarly, the second task B secures and occupies an address space 12 on the main storage device 3. shall be. In the address space 11, a data storage area 13 for storing data for transmission is prepared in advance, and similarly, the address space 12
A data storage area 14 for storing received data is prepared in advance. Furthermore, a data transfer table 15 is provided on the main storage device 3.

Now, when the first task A transfers data to the second task B, the first task A stores the data to be transmitted in the data storage area 13 in advance. Furthermore, the first task A is
A request is made to the microprocessor 2 to send data. When microprocessor 2 receives this request,
A free space 20 on the main storage device 3 is secured, the address of this free space 20 is stored in the data transfer table 15, and finally permission is given to the first task A to transfer data. The first task A refers to the data transfer table 15, recognizes a free area 20, and places a data storage area 1 in this free area 20.
Transfer (data transmission) the contents of step 3 (■ in the figure). On the other hand, the second task B regularly updates the data transfer table 15.
Read the contents of. As a result of this reading, if the address of the main storage device 3 has been written, the data in the area indicated by this address is transferred (data reception) to the data storage area 14 (■ in the figure). In this case, the second task B reads the address of the free area 20, immediately reads the contents of the free area 20, and transfers it to the data storage area 14. Note that if the second task B reads the data transfer table 15 and cannot recognize the address on the main memory 3, that is, if the first task A has not yet transmitted data, this transmission will be waiting for it to be done.

As described above, the first task A sends data, and the second
Data transfer until data is received by task B cannot be performed unless the data is once moved to the free area 20. Therefore, data transmission by the first task A, that is, data transfer from the data storage area 13 to the free area 20, and data reception by the second task B, that is, data transfer from the free area 20 to the data storage area 14. Since the transfer is performed twice, a problem arises in that the more data there is, the longer the transfer time becomes. When this transfer time becomes longer, it becomes difficult for each task to smoothly execute its processing, resulting in a problem that the processing time becomes longer.

The present invention has been made with attention to the above points, and an object of the present invention is to provide a data transfer control method that can promptly terminate data transfer between tasks and smoothly process each task. That is.

(Means for Solving the Problems) The data transfer control method of the present invention provides
When transferring data to a task, the first task refers to its own memory management table and writes the data, and the first task delegates the memory management table to the second task. However, the second task refers to the delegated memory management table and accesses the data on the memory.

(Operation) In the system with the above configuration, the first task refers to its own memory management table and stores data at an address on the memory obtained as a result of this reference. Next, the first task is
This memory management table is delegated to the second task. This allows the second task to refer to the delegated memory management table as its own memory management table and access data.

(Embodiment) The data transfer control method of the present invention can be realized in the general computer system shown in FIG. However, the microprocessor 2 requires some innovation.

Therefore, regarding the computer system, only the configuration of the microprocessor 2 will be described here.

FIG. 1 is a block diagram of a microprocessor according to the present invention.

In the figure, the microprocessor 2 includes a main memory management means 31, a main memory allocation means 32, a data transfer means 33,
Transfer notification means 34 and main memory release means 35 are provided. The main memory management means 31 includes the main memory device 3 (FIG. 2)
It consists of a management table and the like for managing the free area of , that is, the area that is not occupied by any task. The main memory allocation means 32 allocates the free space managed by the main memory management means 31 to each task. The data transfer means 33 performs control such as delegation of a memory management table, which will be explained later. The transfer notification means 34 interrupts the task to which data is to be transferred, and notifies the task of data transfer completion. The main memory release means 35 performs processing to place an area of the main memory device 3 occupied by a task into a free area under the management of the main memory management means 31.

Here, the operation of the data transfer control system of the present invention will be explained with reference to FIGS. 4 to 6.

4 to 6 are conceptual diagrams of data transfer according to the present invention.

First, as shown in FIG. 4, it is assumed that two tasks, ie, a first task A and a second task B, have been generated.

These tasks are divided into units called pages, and one page is generally set to a constant value of about 1 to 4 kilobytes.

The first task A is performed on page A1. It consists of two pages, A2, and the second task B is page B+.

B2. It is assumed that the page consists of three B3 pages.

The first task A1 and the second task B are provided with page tables 51 and 52, respectively. The page tables 51 and 52 indicate where the pages constituting the task exist, whether they exist in the main storage device 3 (FIG. 2) or the secondary storage device 4 (FIG. 2). .

Here, it is assumed that both the first task A and the second task B exist on the main storage device 3.

Therefore, the addresses of the main storage device 3 are stored in the page tables 51 and 52.

On the other hand, the main storage device 3 has an empty area that is not occupied by any task. This free space is also usually divided into pages and managed by the main memory management means 31.

Now, when the first task A transfers data to the second task B, the first task A must first store the data to be transferred on the main storage device 3. In other words, a new area on the main storage device 3 that it occupies is secured. In this case, the first task A issues a request to the main memory allocation means 32 to allocate an area of the main storage device 3 . Upon receiving this request, the main memory allocation means 32 accesses the main memory management means 31, recognizes and secures free space. In this case, it is assumed that the free area C1 is secured among the free areas 01 to Cn managed by the main memory management means 31. The main memory allocation means 32 sets the address of the free area C1 in the third item of the page table 51 in order to make this free area C1 the final page of the first task A. When this third item is set in the page table 51, the free area C1 becomes the area occupied by the first task as page A3.

When this page A3 is secured, the first task A stores the data to be transferred to this page A. When the first task A completes this storage, it issues a transfer request to the data transfer means 33 to transfer data to the second task B.

As shown in FIG. 5, upon receiving this transfer request, the data transfer means 33 transfers the contents of the third item of the page table 51 to the page table 52. That is, when the data transfer means 33 recognizes the third item of the page table 51, the data transfer means 33 transfers the content to the last item (fourth item) of the page table 52.
item). In this case, the fourth item of the page table 52 has contents indicating the area C0 of the main storage device 3.

As a result, the second task B will newly occupy an area on the main storage device 3 as page B4. In addition,
The third item of the page table 51 is deleted at the same time as the delegation.

Now, the second task B cannot recognize the data on page B4 just by setting the fourth item in the page table 52. Therefore, when the transfer notification means 34 erases the third item of the page table 51 of the first task A, it interrupts the second task B and notifies the completion of data transfer. The second task B, which received this notification,
Predetermined data processing will be performed with reference to page B4.

Thereafter, when page B4 is no longer needed, second task B releases the area occupied by page B4, as shown in FIG.

That is, the second task B issues a release request to the main memory release means 35 to release page B4. Main memory release means 35
When it accepts this request, it recognizes the fourth item in the page table 52 and deletes this item. At the same time, the contents of the fourth item are notified to the main memory management means 31, and the area previously occupied as page B4 is converted into a free area Cn*1.
It is placed under the control of the main memory management means 31.

As explained above, when data is transferred from the first task A to the second task B, the page table of the first task A is transferred to the second task B, so processing such as data copying is performed. is not necessary. This is because by delegating the page table, the area on the main storage device 3 that was occupied by another task is taken over. As a result, data is transferred from the first task A to the second task B at high speed, and the processing of each task is not interrupted for a long time.

(Effects of the Invention) The data transfer control method of the present invention described above does not require processing such as copying the data, that is, moving the data to another area on the main storage device, when transferring data to another task. Since there is no such thing, data can be passed to other tasks quickly and reliably without being affected by the amount of data. This results in
It is possible to avoid a situation where the processing of each task is interrupted for a long time due to data transfer, and the task processing can be executed smoothly.

[Brief explanation of the drawing]

□ Figure 1 is a block diagram of a microprocessor according to the present invention, Figure 2 is a configuration diagram of a general computer system, Figure 3 is a conceptual diagram of a conventional data transfer control system, and Figures 4 to 6 are FIG. 2 is a conceptual diagram of a data transfer control method according to the present invention. 3゛1: Main memory management means, 32... Main memory allocation means, 33... Data transfer means, 34... Transfer notification means,
35...Main memory release means. Patent applicant: Oki Electric Industry Co., Ltd.

Claims (1)

[Claims] When transferring data from a first task to a second task, the data is written by referring to a memory management table, and the first task transfers the data to the second task. Data transfer control characterized in that the front part or part of the memory management table is delegated, and the second task refers to the delegated memory management table and accesses the data on the memory. method.