JPH02272651A - Area control system - Google Patents

Abstract

PURPOSE:To effectively utilize a main storage device by optionally releasing the allocation of an area in each page and the allocation of an area with a size smaller than a page size. CONSTITUTION:An operating system using a paging type area control device for executing the allocation/release of the areas of the main storage device has a mechanism 101 for executing area control with an optional size of each page as unit in the main storage device, a mechanism 102 for dynamically forming a table for dividedly controlling the allocated pages and a mechanism 103 for controlling the formed table as a page dividing area. An optional page in the main storage device is dynamically allocated as a divided area and areas more than the page size and areas less than the page size are optionally allocated/released. Consequently, areas to be allocated/released in each page can be increased and the using efficiency of the main storage device can be improved.

Description

[Detailed description of the invention] [Industrial application field] The present invention relates to a virtual storage area management system.

[Conventional technology]

Traditionally, operating systems have implemented virtual memory using the paging method, which allocates physical memory to virtual space in page units. It's here.

As a main example of area management, for an area in main memory that can be allocated/released freely, the start address of each unused contiguous area and the size of that area in pages are set as a set of data. There is management of an area in which an array (hereinafter referred to as a management table) is prepared.

FIG. 4 is an explanatory diagram showing an area management mechanism using the above management table. In the same figure (a), the main storage device 4
01 includes a management table 402 prepared in advance and unused areas 403 and 404 in the main storage device. The management table 402 has registered the start addresses and sizes of the unused areas 403 and 404, respectively ( Same figure (
b)).

When allocating a new area, the management table 402 is searched according to the required size to obtain an allocatable area and the management table 402 is updated. When the area is to be released, the start address and size of the area are stored in the management table. 402.

Furthermore, if it is necessary to allocate/release an area in units of a size smaller than the page size, it is necessary to reserve a dedicated area for this purpose. In other words, in addition to the area that is managed in page units, the area that is managed in units of size smaller than the page size is secured as necessary, and a management table for managing that area is created in advance. Allocate an area of the required size from the dedicated area using a table0. For example, a table (hereinafter referred to as a page table) that has information that associates the physical address and virtual address of each page, etc. may have an area that is sufficiently smaller than the page, and can be managed easily. Since they are allocated/released at the same time as the target page, the same number of pages as the target page is required. Therefore, the page table is not allocated in page units, but the necessary size is allocated from the above-mentioned dedicated area.

[Problem to be solved by the invention]

If the main storage device is managed only by area management in units of pages as described above, the usage efficiency of the main storage device will be poor because one page will be allocated even when an area smaller than the page size is required. In addition, unused areas are divided into pages, resulting in area fragmentation.Also, in the method described above, where areas are managed in page units and areas that are managed in units of smaller size than a page, If the area is managed in small size units, area fragmentation will not occur, but since a dedicated area and management table must be reserved in advance, the area that can be allocated/released in page units will be reduced and the main memory This reduces the efficiency of use.

[Means to solve the problem]

The area management method of the present invention provides an operating system that uses a paging-type area management device that allocates/releases areas in the main storage device in page units. It has a mechanism that performs area management, and a mechanism that dynamically creates a table that divides and manages allocated pages and manages the pages as divided areas, and dynamically divides any page in the main storage device. It is configured so that areas larger than the allocated page size and areas smaller than the page size can be freely allocated/released as areas.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

FIG. 1 is a block diagram showing each mechanism of an embodiment of the present invention. In the same figure, the area management method consists of a mechanism (hereinafter referred to as page management mechanism) 101 that manages the area of the main storage device that can be freely handled by the system in units of pages; A mechanism (hereinafter referred to as a division management table creation mechanism) 102 that creates a management table to be managed within the page, and a mechanism (hereinafter referred to as division management mechanism) 103 that divides and manages the inside of a page using the above division management table created by 102. have

FIGS. 2 and 3 are explanatory diagrams showing the state of the main storage device.

In FIG. 2, an area 201 of the main storage device that can be freely handled by the system is divided into nine areas: a management table 202 that manages the entire area 201 in page units, and used pages 203, 205, 208 within the area 201. It has a page 204 allocated as an area and unused pages 206, 207, and 209 in the area 201, and the page 204 allocated as a divided area has its management table 21.
Contains 0.

When allocating a new area in this state, the page management mechanism 101 will allocate/release the area in page units.
carried out by. That is, when allocating pages, the management table 202 is searched to find consecutive unused pages of a required size or more, and the management table 202 is updated. When releasing the page, the released page is registered in the management table 202.

When allocating an area in units smaller than a page, the division management mechanism 103 searches the management table 210 to find and allocate an unused area consecutively equal to or larger than the required size in the division management page 204. If there is no continuous unused area equal to or larger than the required size, the page management mechanism 101 allocates one of the unused pages as a divided management page. The situation in that case is explained in the third section.
As shown in the figure. That is, the page management mechanism 101 searches the management table 202 and sets the unused page 206 as the divided use page 306, and the division management table creation mechanism 10
2, the in-page management table 311 is created within that page.
is created, and divided use page 3 is created by the division management mechanism 103.
Allocate the required size within 06.

When the divided area is to be released, the divided area within the page 306 is released by the partition management mechanism 103 and registered in the management table 311. If all the divided areas within the page 306 are released by releasing the area, the page 306 is also released by the page management mechanism 101 and registered in the management table 202. In other words, the state shown in Figure 2 will occur, and the page 206 will become an unused page again, and the next time the management table 202 is searched and a hit is found, even if it is a page-based area allocation, it will be treated as a divided area. is assigned even if the assignment is

〔Effect of the invention〕

As explained above, according to the present invention, it is possible to freely allocate and release areas in page units and to release areas smaller than the page size. Storage devices can be used effectively. In particular, in operating systems that implement virtual memory, a page table is allocated/released along with page allocation/release, and since this page table requires an area much smaller than a page, you can freely allocate an area of the required size. If this is possible, the main memory can be used more efficiently than when the area can only be allocated/released in page units.

In addition, it is possible to dynamically allocate an area for area management in units of size smaller than the page size.
There is no need to reserve an area in advance as an area for area management in units of size smaller than the page size. Furthermore, since the ratio of areas managed in page units to areas managed in sizes smaller than the page size can be dynamically determined, flexible use of the main storage device can be expected. That is, since an area management table is dynamically created to perform area management in units of size smaller than the page size of the allocated page, there is no need to reserve an area for the management table in advance as in the past. Therefore, effective use of the main storage device can be expected.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIGS. 2 and 3 are explanatory diagrams showing operating states, and FIG. 4 is an explanatory diagram showing a conventional example. 101... Page management mechanism, 102... Division management table creation mechanism, 103... Division management mechanism.

Claims (1)

[Claims] In an operating system that uses a paging-type area management device that allocates/releases areas in a main storage device in page units, a mechanism for managing an area of any size in units of pages in the main storage device, and allocation and a mechanism for dynamically creating a table for dividing and managing pages within the page and managing the pages as divided areas, and dynamically allocating any page in the main memory as a divided area and an area larger than the page size. An area management method characterized by freely allocating/releasing areas smaller than the page size.