JPH03204048A - Memory control method - Google Patents

Abstract

PURPOSE:To vary the sizes of memory areas and to control these areas with a small number of tables by adding the information to the unused ones of those memory areas of a heap area to show positions of the front and rear unused areas respectively. CONSTITUTION:A heap area 1 includes (n + 1) pieces of busy areas and unused areas respectively. The information called MCB and showing the semantic (type) and the size of each area are stored in each head of the same semantic area. In addition, the information added with the previous and next free states is stored after the MCB in an unused area. This information is called free MCB. Thus the MCB and the free MCB are stored in the head of each busy area and each unused area of the area 1 respectively. Meanwhile a work area 2 stores only the address data showing the positions of the head and last unused areas.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a memory management method in a device using a microcomputer, such as a controller for a laser printer.

[Conventional technology]

For example, the memory (RAM) installed in the controller of a laser printer consists of a work area that holds essential data for the controller to proceed with processing, and a heap area that stores information such as downloaded fonts, form data, and intermediate buffers. It is managed separately.

In the heap area, when memory is required for downloading fonts, etc., the required amount of memory area is secured and data is stored there for use.

Furthermore, when the data is no longer needed, the memory area used for it is released, and this memory area is used again the next time the memory is needed.

The traditional method of managing memory is as follows:
For example, as shown in Figure 9, memory heap area 1
There is a method of dividing the data into a plurality of blocks of fixed size and having a table in the work area 2 that shows the usage status of each block (hereinafter referred to as "prior art example 1").

In this method, when memory is needed, the table in the work area 2 is checked to find an unused block, and that block is used to store the necessary data.

At this time, the contents of the table corresponding to the newly used block are changed to "in use". If the data is no longer needed and is to be released, the contents of the table corresponding to the block to be released are changed to "runused".

Alternatively, as shown in FIG. 10, there is a method in which work area 2 has a table that manages the position and size of unused areas in heap area 1 (hereinafter referred to as conventional example 2).
).

In this method, when memory is needed, the table in the work area 2 is checked to find an unused area large enough to store the required amount of data, and when it is found, a memory area is secured from there.

At this time, since the size of the unused area changes, the contents of the table are also rewritten. Furthermore, if the unused area is secured from the beginning, the position of the unused area changes, so the position information in the table is also rewritten.

[Problem to be solved by the invention]

However, according to the memory management method of Conventional Example 1, the heap area is managed in blocks, so even if the data to be stored is smaller than the block size, one block is used entirely, resulting in poor memory usage efficiency. There was a drawback.

If you reduce the block size to avoid this drawback, the number of blocks to be managed will increase, and the number of tables that must be kept in the work area will increase.
A problem arises in that the size of the heap area itself, which can be used flexibly, is reduced.

Furthermore, according to the memory management method of Conventional Example 2, the heap area is not managed in blocks of fixed size as in the case described above, but the size of the memory area used can be made variable, resulting in wasted memory. There aren't many.

However, with this method, if memory is allocated and released repeatedly and unused areas are divided, it is necessary to have as many tables as there are unused areas. In addition, it is often difficult to predict how many unused areas will be divided, and in order to manage them without problems, it is necessary to have as many tables in the work area as in Conventional Example 1. There was a problem.

This invention was made in view of the above points, and it is possible to make the size of the memory area used in the memory heap area variable, and to manage it with a small number of tables even if the unused area is divided into many parts. The purpose is to

[Means to solve the problem]

In order to achieve the above object, the present invention uses the on-board memory to create a work area for storing predetermined contents in a predetermined location, and a memory area to secure and store data when necessary. In a memory management method in which data is stored in a heap area and a heap area is released when the data is no longer needed and prepared for the next use, each area in use in the heap area is marked with its purpose. Information indicating the size of the area is added, and information indicating that it is unused is added to each unused area, information indicating where the previous and following unused areas are, and the first unused area is The memory is managed by storing the location and the location of the last unused area in the work area.

[For production]

According to the memory management method according to the present invention, each unused area in the memory area of the heap area is attached with information indicating where the previous and following unused areas are located, so each unused area is linked. can be used to construct a free memory chain.

Therefore, since it is only necessary to store the positions of the first and last unused areas in the work area, the work area can be reduced and the heap area can be increased, and the entire memory can be used effectively. Furthermore, the unused area can be managed no matter how many parts it is divided into.

Furthermore, since information indicating the purpose of use and the size of the area is added to the usage area, it is possible to judge whether each usage area is movable or not, and if it is movable, it can be moved and It becomes possible to secure a large memory area by combining the unused areas of .

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1A shows an example of the management state of the work area and heap area in the memory when the present invention is implemented.

Each heap area 1 has n+1 used areas and unused areas, which have the same meaning (areas that are unused or used for the same purpose if they are in use).
At the beginning of each field, the meaning (Type) and size (
The data block is stored as MCB (Me+++ory Contr
ol Block).

Furthermore, in the case of an unused area, following this MOH, Previous-free and Next-free, which are information indicating the position of the nearest unused area before and after this area, are added. This data block, which stores information, is called a free MCB.

FIG. 1B shows the structure of this MCB and free MCB.

"It Type" of MCB indicates the meaning of the area with 8 bits, and in the case of II O71, it is unused. If it is in use, a value predetermined according to the purpose of use is entered. "Type" is also referenced during garbage collection, which will be described later.

Further, EI S x ze ” is 24 bits and indicates the size of the area.

"Next-free" of the free MCB is 32 bits and is link information indicating where the next unused area is, and in the case of the last unused area, this value is "It_O".

"Previous-free" is also 32 bits and is link information indicating where the previous unused area is, and in the case of the first unused area, this value is 0''.

In this way, at the beginning of each used area of heap area 1, M
A free MCB is stored at the beginning of an unused CB area, and only address data indicating the positions of the beginning and end unused areas are stored in the work area 2, as shown in FIG. 1A.

By doing this, the unused area in heap area 1 can be used for link information (previous-free and ~
ext-free) to form a free memory chain.

Therefore, it is only necessary to store the positions of the first and last unused areas in the work area 2, and no matter how many parts the unused area is divided into, it will not become unmanageable.

By the way, as mentioned above, font data, form data, intermediate buffer data, etc. are stored in the heap area of the memory provided in the controller of the laser printer.

Of these, font data and form data are used over multiple pages, and intermediate buffers are created for each page, so they are released when printing for that page is completed.

If such a relatively fixed area is mixed with an area that is frequently allocated and released, the unused area will be divided, which will not only cause management problems but also cause large areas to be allocated. The problem arises that it cannot be done.

Therefore, in order to reduce this problem, data that is used permanently and data that is frequently secured and released are
It is desirable to secure from different directions of the heap area.

For example, a method is used in which font data and form data are secured from areas with smaller addresses, and intermediate buffers are allocated from areas with larger addresses.

Therefore, if memory is required according to this embodiment, an unused area where memory can be secured is searched for by tracing link information from the beginning or the last unused area of the heap area depending on the purpose of use. Allocate memory in the area.

When searching from the beginning, memory areas are secured from the beginning of the unused area, and conversely, when searching from the end, the memory area is secured from the end of the unused area.

When a memory area is secured, M is placed at the beginning of the memory area.
A CB is created, and the purpose of use and the size of the reserved area are written in II Type e'' and '5ize, respectively.

Also, at this time, the free MCB in the unused area is also rewritten. Note that if the beginning or the beginning of the last unused area is changed, the data in the work area 2 is also rewritten.

Figure 2 shows the state before securing the memory area of the heap area, Figure 3 shows the state in which the memory area to be used is secured from the beginning, and Figure 4 shows the state in which the memory area to be used is secured from the end. .

Next, the case of releasing the used memory area will be explained.

When releasing a memory area in use, the free memory chain is traced while comparing the address of the area to be released with the data of the free MCB, and it is determined in which position in the free memory chain the memory to be released should be placed. Once this position is determined, the links between the free MCBs in the used areas before and after are once disconnected, a free MCB is created at the beginning of the freed memory area, and the links with the free MCBs before and after are re-created.

In this way, memory is freed by incorporating the used memory area into the free memory chain,
An unused area is created that can be used the next time memory is needed.

FIG. 5 shows a state in which the area 3 in use in FIG. 2 is released.

During this release process, if the memory area to be released is adjacent to the previous, subsequent, or previous and subsequent unused areas, the unused areas are combined and incorporated into the free memory chain as one unused area. .

FIG. 6 shows a state in which the usage area 2 is further released from the state shown in FIG.

Needless to say, even in such a release operation, if the start address of the first or last unused area changes, the data in the work area 2 will be rewritten.

According to this embodiment, the fixed table that must be kept in the work area 2 is extremely small, and there is no wastage of memory and no memory becomes unmanageable. Area memory management can be performed.

Next, a garbage collection method when this management method is implemented will be explained.

Garbage collection is used to combine unused areas and secure a larger memory area by moving the memory area in use when the unused area is divided into smaller pieces and it becomes impossible to allocate a large memory area. It is to make it possible.

By the way, for example, some laser printer controllers use direct addresses in an intermediate buffer to increase processing speed.

When used in this way, a used area that cannot be moved may occur in the heap area, making complete garbage collection impossible. Even in such a case, there is a possibility that unused areas may be combined due to garbage collection, so garbage collection must be performed while avoiding used areas that cannot be moved.

When garbage collection is performed in the prior art example 1 described above, it is performed on a block-by-block basis, and it is not possible to eliminate wasted areas within a block.

In addition, in the case of Conventional Example 2, since the memory area in use is not managed, it is impossible to move the memory area in use.
Since the purpose of use of the area in use is written in pe II, it is possible to use this to determine whether or not the area in use can be moved.

First, M of the used area after the first unused area
Check CB. As a result, if the area in use can be moved, it is moved and swapped with the previous unused area.

This is repeated in sequence, and when the area after the unused area also becomes an unused area, the two areas are combined into one area. If this process is repeated until the last unused area is reached, the unused areas should be combined into one.

"T y p e" is referenced to know the data that must be changed by swapping. For example, if the area is used to store font data, the data storing where the font data is located is rewritten.

Figures 7(a) to 7(d) show how garbage collection is progressing.

During this process, if an area after the unused area becomes an area that cannot be moved, the process for that unused area is abandoned, and the same process is performed for the next unused area.

FIG. 8 shows the result of garbage collection performed when the currently used area 4 cannot be moved in the state shown in FIG. 7(C).

In this way, by using "Type" of the MCB, garbage collection can be performed while avoiding memory areas that cannot be moved.

〔Effect of the invention〕

As explained above, according to the present invention, in a memory management method that manages memory using a heap area and a work area, the size of the memory area of the heap area is made variable, and the unused area is divided into small pieces. However, it can be managed with a table with very little work area.

Furthermore, garbage collection can be easily performed without waste.

Therefore, the entire memory can be used effectively.

[Brief explanation of drawings]

FIG. 1A is an explanatory diagram showing an example of a management state of a work area and a heap area in a memory when the present invention is implemented. Figures 2 to 4 are explanatory diagrams of the procedure for securing a memory area to be used in the heap area, Figures 5 and 6 are explanatory diagrams of the procedure for releasing the memory area in use, and Figure 7 and FIG. 8 are explanatory diagrams of the garbage collection procedure, and FIGS. 9 and 10 are explanatory diagrams showing different examples of conventional memory management methods, respectively. 1 Heap area 2... Work area (α) Before processing starts Figure 9 (b) Combine unused 1 and 2 Combine unused 1, 2 and 3 Combine all unused Figure 10 Procedure correction band (Moving ceremony) April 26, 1990

Claims (1)

[Scope of Claims] 1. The installed memory has a work area for storing predetermined contents at a predetermined position, and a memory area is secured and data is stored there when needed. In a memory management method that separates the memory area into a heap area and a heap area to prepare for the next use by releasing the memory area when data is no longer needed, each area in use in the heap area is marked with its purpose of use and the size of the area. For each unused area, information indicating that it is unused and information indicating where the preceding and succeeding unused areas are are added, as well as the position of the first unused area and the end. A memory management method, comprising: storing a position of an unused area in the work area.