JPH067376B2 - Information Module Loading Area Selection Method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a loading area selection method when loading an information module such as modularized program information from an external storage device to a main storage device, in particular, The present invention relates to a loading area selection method when a plurality of information modules to be loaded have a hierarchical structure.

[Conventional technology]

Conventionally, as a method of loading information such as a modularized program from an external storage device such as a disk device into a predetermined area of the main storage device as needed, for example, an LRU
(Least Recently Used) A method based on the logic is known.

This is, for example, as shown in FIG. 4 (a), four information modules A, B, C, D are divided into two areas (hereinafter referred to as cells) C1,
Assuming that the information module A is used, first, when the information module A is used, the information module A is loaded in the unused cell C1, and then, when the information module B is used, the information module B is not used. It is designed to load the unused cell C2. Then, when the information module C is used in a state where the information modules A and B are already loaded in both cells C1 and C2 as described above, when the information module C is used, one of the already loaded information modules A and B is used. The cell in which the oldest information module existed, that is, cell C1
Then, the new information module C is loaded in the same manner, and thereafter, when loading the new information module, similarly, when loading the new information module, the cell in which the information module that has been used the oldest is present is selected. The cells to be newly loaded are decided.

[Problems to be solved by the invention]

However, the above-mentioned method of selecting the loading area of the information module simply according to the LRU logic has a drawback that the number of loading processes is generally large, and a plurality of information modules to be loaded are not shown in FIG.
In the hierarchical structure as shown in (b), if the lower level information module has meaning in the presence of the upper level information module, for example, main program information and its subroutine program information, etc. In that case,
In order to effectively use such information, it is necessary to secure a relatively large number of cells to be loaded, which makes efficient loading difficult.

[Means for solving problems]

The present invention has been made in view of the above problems, and as a loading area selection method for a plurality of information modules having a hierarchical structure, the plurality of information modularized is modularized in units of loading into areas. In the loading area selecting method of the information module having the relationship of 1) and the information of the corresponding hierarchical level is given from the hierarchical structure, a new 1 is added to one of the areas where the information module is already loaded. When loading one information module, based on the information of the hierarchy level of the information module, the information modules belonging to the lowest hierarchy level and the information belonging to the higher hierarchy level are targeted for a plurality of already loaded information modules. The module is searched and the same level level is searched during the search. When a hierarchy level in which a plurality of information modules belonging to a hierarchy exist is found, the area in which the information module that has been used the oldest among the plurality of information modules belonging to the hierarchy level is newly loaded should be newly loaded. The feature is that it is defined as a region.

Example of Invention

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

FIG. 2 is an explanatory diagram showing a conceptual configuration example of a system to which the method for selecting the loading area of the information module according to the present invention is applied. In this example, one of a plurality of information modules having a hierarchical structure recorded in the floppy disk device FDP is loaded by the loader L into the cell C1 of the main memory MM in the host device.
It is a system to load to any of
The loader L refers to the cell management module to perform the loading operation.

Here, the cell management module has the following tables 1 to 3.

Table 1: Table showing used / unused state of cells C1 to C4 Table 2: Table showing hierarchy (hereinafter referred to as level) number of information module loaded in cells C1 to C4 Table 3: Used in host device A table showing the history of used cells corresponding to the history of the information module, in which new cells are sequentially chained at the beginning.

Then, the loader L is specifically the above tables 1 to 3
Please refer to the following steps and follow the procedure below.
The information module in the FOP is loaded in any of the cells C1 to C4.

Procedure 1: Referring to Table 1, if there is an unused cell, load it.

Step 2-1: If there is no unused cell by referring to Table 1, further refer to Table 2 and there are a plurality of information modules loaded in cells C1 to C4 as the same level, and Find the lowest level.

Step 2-2: Referring to Table 3, the cell in which the information module of the earliest used time exists among the information modules of the levels searched as described above is determined and loaded into this cell.

Next, a specific operation of loading the information modules A to H having the hierarchical structure as shown in FIG. 3 (a) into the cells C1 to C4 will be described. The operation flow is the first
As shown in the figure, in FIG. 1, the flow F1 corresponds to the above procedure 1, and the flow F2 corresponds to the above procedure 2-1 and procedure 2-2.
It corresponds to.

First, as shown in step 1 of FIG. 3 (b), when the instruction to use the information module D is issued while the information modules A, B and C are already loaded in the cells C1 to C3, respectively, Since the information module D is not loaded in the main memory MM, a new information module D
Will be loaded. The operation is performed by setting the initial value "1" of n in the flow F1 of FIG.
It is determined whether n (Cn) is unused. N = of the process
When it becomes 4, it is determined that the cell 4 (C4) is an unused cell, and this "4" is determined as the cell number N (= 4) to be loaded. Therefore, in step 1, information module D is loaded into cell C4.

As shown in step 7 of FIG. 3 (b), it is assumed that a new information module C is loaded while the information modules A, B, E, D are already loaded in each of the cells C1 to C4. Then, since the cells C1 to C4 are all in use, the flow F1 of FIG. 1 does not determine the cells to be loaded, and the flow shifts to the flow F2. In the flow F2, the maximum level number, that is, “2” is set to (corresponding to the lowest level), and while decrementing this sequentially, the table 2 is referenced to determine whether there are multiple levels. Then, since the information modules D and E are level 2, a plurality of them are discriminated when = 2 (initial value), and the number "2" of the loaded level 2 information modules is set to m. At the same time, referring to Table 3, the start address of the cell chain is set to a. At this time, as is clear from the column of "cell chain" in FIG. 3 (b), the head address a of the cell chain corresponds to the cell C2. After that,
While sequentially updating the address a, it is determined whether or not the cell corresponding to the address a is loading the level 2 information module. Specifically, the address a
Is updated from the one corresponding to the cell C2 to the one corresponding to the cells C1 and C3, respectively (see FIG. 3 (b), "Cell chain" column). It is determined that the module E is being loaded. Then, when the above number m (= 2, the number of information modules of level 2) is decremented (m = 1) and the address a is updated again from the one corresponding to the cell C3 to the one corresponding to the cell C4, this cell C4 also becomes , It is determined that the level 2 information module D is loaded, and further m
Decrement (= 1). Then, when m = 0 by the decrement processing of the numerical value m, the cell number "4" (C4) corresponding to the address a at that time is determined as the cell number N to be loaded. Therefore, in step 7, the information module C is loaded into cell C4.

Furthermore, as shown in step 10 of FIG. 3 (b), the cell C
Assuming that the information module G is newly loaded in a state where the information modules A, B, F, and C are already loaded in each of 1 to C4, all the cells C1 to C4 are in use, Similar to the above, the flow shifts from the flow F1 to the flow F2. In this flow F2, the maximum level number "2" is set to, and while decrementing this sequentially, referring to Table 2, it is determined whether or not there are a plurality of levels. Then, since the information modules B and F are at level 1, a plurality of them are discriminated when = 1 and the number "2" of the information modules of level 2 which have been loaded are set to m and the table 3 is set. Is set to the start address of the cell chain. At this time, the head address a of this cell chain corresponds to the cell C3 (see the "cell chain" column in FIG. 3 (b)).
Then, since the information module F of level 1 is loaded in the cell C3 corresponding to the head address a, first, the numerical value m (= 2) is decremented (m =
1) is performed, and thereafter, while sequentially updating the address a,
It is determined whether the cell corresponding to the address a is loading the level 1 information module. Specifically, the address a is sequentially updated from the address corresponding to the cell C3 to the address corresponding to the cells C1, C4, and C2 (third address).
Figure (b) "Cell chain" column), in the process, cell C
2 is determined as loading the level 1 information module B. Then, the above numerical value m (= 1)
Is decremented and m =
When it becomes 0, the cell number "2" (C2) corresponding to the address a at that time is determined as the cell number n to be loaded. Therefore, in step 10, the information module G is loaded in the cell C2.

Incidentally, also in the other steps 2, 9, 11, and 14 in FIG. 3 (b), the cells to be loaded are determined in accordance with the flow F2, as in the above.

As described above, according to this embodiment, when all the cells C1 to C4 are in use, the information modules loaded in the cells C1 to C4 are always processed at all levels 0 to 2 because the cells are processed according to the flow F2. Therefore, when using the information module having the hierarchical structure, it is possible to avoid unnecessary loading as much as possible. For example, Figure 3
When using the 15-step information module shown in (b),
Whereas the conventional LRU logic simply requires 10 loadings, the method according to the present invention requires only 7 loadings.

It should be noted that the flow F3 in FIG. 1 is for the case where the number of levels of the hierarchical structure is equal to or greater than the number of cells to be loaded, and when there are no cells loaded with information modules of the same level (all levels). In this case, the operation flow for determining a cell to be newly loaded is shown. Specifically, if there are no cells loaded with information modules of the same level, the cell loaded with the information module with the highest level number (lower level) is determined as the cell to be newly loaded. I am trying.

[Effects of the Invention] As described above, according to the present invention, when one information module is newly loaded in any of the areas in which the information module is already loaded, the hierarchy of the information modules is added. Based on the level information, a search is performed from the information modules belonging to the lowest hierarchical level to the information modules belonging to the higher hierarchical level, targeting a plurality of already loaded information modules, and the same in the process of searching. When a hierarchy level having a plurality of information modules belonging to the hierarchy level is found, an area in which the information module that has been used the oldest among the plurality of information modules belonging to the hierarchy level is newly loaded is newly loaded. The area should be

For example, in the case of an information module having a hierarchical structure as shown in FIG. 3 (a) (if B is used, then C,
When D and E are used and B is used again, and the area to be loaded is only three areas in which B, C and D are loaded, the conventional loading method (conventional method) is used. In the LRU), B, C, D, and E are used in order, so when using the information module E, since B is the oldest information at that time, E is loaded in the area where B is loaded. The information module B to be used next is erased from the already usable area and cannot be used unless reloaded. However, in the loading method of the present invention, the hierarchical level is used. Since the usage time is managed separately, when using the information module E, the information module C with the oldest usage time at the same hierarchical level is loaded. Information area have the module E is loaded, then when the information module B is used, so that can be used without re-loading the information module B.

As a result, the higher the hierarchy level, the more the area can be retained, the LRU management can be performed for each hierarchy level, the finer loading control than the conventional method can be performed, and the number of times of loading is smaller than the conventional method. Thus, a more efficient information module loading area selection method can be realized.

[Brief description of drawings]

FIG. 1 is a flow chart showing an example of a processing flow of an information module loading method according to the present invention, and FIG. 2 is an explanatory diagram showing a conceptual structure of an example system to which the information module loading method according to the present invention is applied. FIG. 3 (a) is an explanatory diagram showing an example of the hierarchical structure of the information module, and FIG. 3 (b).
Is an explanatory view showing the loading state of the information module having the hierarchical structure shown in FIG. 3 (a) by comparing the method according to the conventional LRU logic with the method according to the present invention. FIG. 4 (a) shows the conventional method. FIG. 4 is an explanatory diagram showing an example of a loading state in accordance with the LRU logic of FIG. 4, and FIG. 4 (b) is an explanatory diagram showing an example of a hierarchical structure of information modules.

Claims (1)

[Claims] 1. An information module which is modularized in a unit of loading into an area, and the plurality of modularized information have a hierarchical structure relationship, and information of a corresponding hierarchical level is given from the hierarchical structure. In the loading area selection method, when newly loading one information module into any of the areas where the information module is already loaded, the information module is already loaded based on the information of the hierarchical level of the information module. A hierarchy in which a plurality of information modules belonging to the same hierarchy level are searched in the process of searching from the information modules belonging to the lowest hierarchy level to the information modules belonging to the higher hierarchy level. When a level is found, multiple information belonging to the hierarchical level A method for selecting a loading area for an information module, characterized in that, in the information module, an area in which an information module that has been used for the oldest time is loaded is set as an area to be newly loaded.