JPH05165703A - Saving memory allocation method - Google Patents

Abstract

PURPOSE:To obtain a saving memory allocation method for reducing an error processing. CONSTITUTION:In a step S1, a memory security request is given, and the request on one block or more than two blocks in the memory security request is judged in S2. When the number of the request blocks is a piece of YES, an idle block for upper direction is searched in S3. When the number of the request blocks is more than two NO, the idle block in a lower direction is searched in S4. Then, the existence of the idle blocks continuing for the number of the request blocks is judged in S5. In the case of YES, in which the continuous idle blocks for the number of the request blocks exist, the processing of management TBL registration is executed in S6. In S7, the error processing is executed in the case of No, in which the continuous idle blocks for the number of the request blocks does not exist.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a save memory allocation method which is used for window processing and / or multigraph processing of a computer or a word processor and allocates a screen image on a memory and saves it.

[0002]

2. Description of the Related Art Conventionally, in window processing or multigraph processing in information processing of computers, word processors, etc., processing for saving a screen image (bit image) of a program on a memory and restoring it from the memory to a CRT is performed. , The screen is being switched.

FIG. 3 shows this screen switching state.

In FIG. 3, in the save processing [13], the screen image of the first program is saved in the block 1 on the save memory. Next, in the restoration process [14], the screen image of the second program is restored from the block 2 on the save screen to the CRT, and the screen switching process of the program is performed. In this case, it is necessary to manage the allocation and allocation of the save memory to save the screen images of a plurality of programs. In this save memory, one block having a certain fixed length and a required number of blocks in which each program is continuous perform a free block search from a single direction to perform save memory allocation processing.

In FIG. 3, the screen image of the first program is assigned to block 1, the screen image of the second program is assigned to block 2, and the screen image of the third program is assigned to blocks 3 and 4.

FIG. 4 shows a processing procedure for allocating the save memory.

In FIG. 4, a memory reservation request is made in accordance with the required number of consecutive blocks (step (described as S in the following figure) 15). Then, a free block search from a single direction is performed to check whether the requested number of consecutive free blocks exists (step 1).
6). Further, it is determined whether or not there is an empty block that satisfies the request (step 17). Here, Ye with an empty block
In the case of s, the management table (TB
L) A registration process is performed (step 18). If there is no empty block that satisfies the request in step 17, error processing is performed (step 19).

[0008]

By the way, in the conventional save memory allocation method, as shown in FIG. 4, the unidirectional search for empty blocks is performed in step 16, so that the existing empty blocks are There is a drawback that error processing (step 19) frequently occurs without being continuous. For example, as shown in FIG. 5, when the size of the save memory is four blocks, first, the first program of 1 block size, the second program of 2 block size, and the third program of 1 block size are saved in order. When the memory is allocated, the state of processing [20] is entered. Next, when the first program and the third program are released, the state of processing [21] is reached.

If a memory securing request for the fourth program having a 2-block size in the process [22] is issued following this state, the error process [23] is necessary because the empty blocks are not continuous.

An object of the present invention is to solve the above problem and to provide a save memory allocation method capable of reducing the error processing that has been frequently generated in the conventional free block search from a single direction.

[0011]

The save memory allocation method of the present invention is characterized in that the number of blocks required for a memory reservation request is determined, and the direction of block search for memory allocation is changed according to the number of requested blocks. It is a thing.

Further, in the above method, the direction change of the block search for allocating the memory is a bidirectional search in which the upward search and the downward search are performed depending on the number of requested blocks. It is a thing.

Further, each of the above methods is characterized by being applied to window processing and / or multigraph processing in a computer or a word processor.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the save memory allocation method of the present invention will be described below with reference to the drawings.

FIG. 1 shows a procedure for saving memory allocation by bidirectional free block search.

In FIG. 1, first, memory reservation is requested (step 1). Next, it is judged whether or not the number of blocks of this memory reservation request is one or two or more (step 2). If the number of requested blocks is one, Yes, an empty block search is performed by an upward search (step 3). If the number of requested blocks is 2 or more, the empty block search is performed by the downward search (step 4). Next, as a result of performing the upward search in step 3 and the empty block search by the downward search in step 4, it is determined that there are empty blocks in which the number of requested blocks is continuous (step 5). If the number of requested blocks is a continuous empty block, Yes, the start address of the empty block allocated as the position for saving the screen image of the program that made the request, the number of reserved blocks (memory size), etc. Information of the above is registered in the management table (TBL) (step 6). If the number of requested blocks is no empty block, the error processing is performed (step 7).

Next, a concrete operation example will be described in comparison with the conventional FIG.

As shown in FIG. 2, the size of the save memory is 4 blocks as in FIG.

First, the save memory is allocated in the order of the 1-block size first program, the 2-block size second program, and the 1-block size third program.

The first program and the third program each have a required number of blocks, and in FIG. 1, the empty block on the save memory is allocated by the empty block search by the upward search in step 3. Further, the second program has two requested blocks, and in FIG. 1, the empty block on the save memory is allocated by the empty block search by the downward search in step 4.

As a result, the process shown in FIG.

The memory allocation state of [9] is entered. Next, when the first program and the third program are released, the state of processing [10] is reached.

In this state, when a memory reservation request for the fourth program of 2 block size in process [11] occurs, the number of requested blocks is 2, and the process is performed by the downward search in step 4 in FIG. As shown in [12], it is assigned to one block and two blocks on the save memory.

By the way, when such an operation is performed in the same manner as the save memory allocation process described above, the memory reservation request process [22] for the fourth program of 2 blocks size in FIG. 5 is generated. At this time, the save memory state of the process [21] is entered. Therefore, the empty blocks are not continuous and the error processing [23] is frequently required, but in the present embodiment, it is significantly reduced.

[0024]

As is apparent from the above description, the save memory allocation method of the present invention determines the number of reserved request blocks to be saved, and changes the direction of searching for an empty block on the memory according to this request block number. Therefore, it is possible to reduce the error processing that frequently occurs in the conventional free block search from a single direction.

[Brief description of drawings]

FIG. 1 is a flowchart showing a processing procedure by bidirectional search in an embodiment of a save memory allocation method of the present invention.

FIG. 2 is a diagram showing a processing state when the size of a save memory according to the embodiment is four blocks.

FIG. 3 is a diagram showing a screen switching state in a conventional save memory allocation method.

FIG. 4 is a flowchart showing a processing procedure of conventional save memory allocation.

FIG. 5 is a diagram showing a processing state when the size of a conventional save memory is 4 blocks.

[Explanation of symbols]

 S1 Memory reservation request S2 Block number judgment S3 Upward search S4 Downward search S5 Empty block judgment S6 Management TBL registration S7 Error processing

Claims (3)

[Claims] 1. A save memory allocation method, characterized in that the required number of blocks for a memory reservation request is determined and the direction of a block search for memory allocation is changed according to the requested number of blocks. 2. The save memory according to claim 1, wherein the direction change of the block search for allocating the memory is a bidirectional search in which an upward search and a downward search are performed depending on the number of requested blocks. Assignment method. 3. The save memory allocation method according to claim 1, which is applied to window processing and / or multigraph processing in a computer or a word processor.