WO2010143364A1

WO2010143364A1 - Partition management device, partition management method, and program

Info

Publication number: WO2010143364A1
Application number: PCT/JP2010/003456
Authority: WO
Inventors: 吉岡誠; 國分光裕; 新庄敏男
Original assignee: 株式会社エスグランツ
Priority date: 2009-06-11
Filing date: 2010-05-21
Publication date: 2010-12-16
Also published as: JP2010287066A; JP4846001B2

Abstract

Provided is a technology by which an area can be flexibly partitioned and managed regardless of the capacity of a storage device, and an area can effectively utilized by an effective method when a file is allocated to a partition. A size of a partition of a storage device is expressed by a sum of mutually different power of twos, and when an area having a size of each of the power of twos constituting the sum defines a main section, each main section is divided into sections having sizes obtained by sequentially halving the main section, and an allocation table which stores allocation information indicating the file allocation status with respect to each of the sections having respective sizes including the main section is generated. The partitions are managed on the basis of the allocation information stored in the allocation table.

Description

Partition management apparatus, partition management method and program

The present invention relates to a partition management device, a partition management method, and a program for managing partitions of a storage device.

Conventionally, as a method for managing data stored in a storage device such as a hard disk, a file system such as FAT (File Allocation Tables) or NTFS (NT File System) has been adopted.

FIG. 1A is a diagram for explaining partition assignment in an external storage device such as a hard disk conventionally employed.
As shown in FIG. 1A, for example, the storage area of the external storage device 306 is assigned four partitions 1 to 4 (191, 192, 193, 194) by a utility program such as a disk allocator (not shown), and each of them is assigned to a certain file system. Assigned. A master boot record 195 having a partition management table 195a storing partition management data is arranged at the head of the partition. As illustrated in FIG. 1A with the partition information 196 for the partition number 3, the partition management table 195a includes entries for a partition number 196a, a start position 196b, and a partition size 196c. The partition number 196a stores a partition number for identifying the partition, and is 3 in the illustrated example. Information on the top address of the section is stored at the start position, and the contents thereof are omitted in the illustrated example. The partition size 196c stores the number of allocation unit areas of the partition, which is 52 in the illustrated example.

Disk allocation management for files in each partition is complicated because it depends on the allocation method specific to the file system to which each partition is allocated. For example, according to FAT, a link list indicating a block to which allocation is performed is held for each file. Further, according to NTFS, information about the start position of the block assigned for each file and the number of continuous blocks is held.
In order to solve the problem of allocation management in these file systems, for example, a technique related to a buddy system disclosed in Patent Document 1 below has been proposed. According to the buddy system, the allocation is managed by dividing the area into 2 ^k sizes. A basic buddy system is used for allocating memory areas in a memory management system that does not use the virtual storage method.

FIG. 1B is a diagram illustrating the allocation principle of the buddy system. Shown in FIG. 1B is a bitmap 400 that indicates an allocation area 490 of size 2 ³ = 8, a corresponding tree structure 580, and an area corresponding to a node of the tree structure 580 is in use. . The allocation area 490 is, for example, the sections 1 to 4 illustrated in FIG. 1A.
The tree structure 580 is a hierarchical model of the relationship between the areas obtained by sequentially dividing the allocation area 490 into half the size of the allocation unit. In the example shown in the figure, the level of the root node is 4 when the hierarchy of the area of the allocation unit size is level 1.

The root node 480 of the tree structure 580 corresponds to the entire area (level 4 area) of the allocation area 490 before division, as indicated by the dotted arrow, and “8” described in the node corresponds to This corresponds to the size of the allocation area 490. The numbers in parentheses are intra-level numbers that identify areas within the same level.

A node 440 connected to the root node 480 via a link 540 and a node 441 connected to the root node 480 via a link 541 correspond to an area (level 3 area) obtained by dividing the allocation area 490 into two. The size of each area is 4 as shown in the figure.
Below the node 440 are a size 2 node 420 connected by a link 520 and a size 2 node 421 connected by a link 521. Similarly, a size 2 node 422 connected by a link 522 and a size 2 node 423 connected by a link 523 exist below the node 441. These four nodes correspond to the level 2 region.

Below the node 420 are a size 1 node 410 connected by a link 510 and a size 1 node 411 connected by a link 511. Similarly, a size 1 node 412 connected by a link 512 and a size 1 node 413 connected by a link 513 exist below the node 421, and a node 413 connected by a link 514 below the node 422. The size 1 node 414 connected by the link 515 and the size 1 node 415 connected by the link 515 exist, and the size 1 node 416 connected by the link 516 and the size connected by the link 517 are below the node 423. There is one node 417. These eight nodes correspond to the level 1 region.
Nodes other than the root node 480 correspond to one of the areas obtained by dividing the area corresponding to the parent node into two. Therefore, the sum of the sizes of the areas at each level matches the size of the allocation area 490.

1B has a bit value corresponding to each node of the tree structure 580, as indicated by a dotted arrow. The bit value of the bit position indicated by the in-level number 409 of level 4 indicated by reference numeral 408 in the bitmap 400 indicates whether the level 4 area is in use, and is within the level 3 level indicated by reference numeral 404. The bit values of

numbers

0 and 1 indicate whether the corresponding area is in use. Similarly, level 2 in-level numbers 0 to 3 indicated by reference numeral 402 indicate whether or not each level 2 area is in use, and level 1 in-level numbers 0 to 7 indicated by reference numeral 401 indicate level 1 areas. Indicates whether it is in use.

In the example shown in FIG. 1B, bit “1” is set in the intra-level number 0 of level 4, the intra-level number 0 of level 3, the intra-level number 0 of level 2, and the intra-level number 0 of level 1. This shows that the entire area corresponding to the position cannot be newly used. In other words, since the area of size 1 corresponding to the node 410 is in use as indicated by the level number 0 in level 1, the larger area including that area cannot be used as it is. Is shown.

According to the above-mentioned buddy system, when allocating files and memory areas of a certain size, that even when the assignment searching for free space from the region of the size made of 2 ^k to or exceeds the size and equal, releasing Since it is sufficient to return to the vacant area group of the size, allocation and release of the file and the memory area and management of the area are easy.

JP 7-28893 A

However, according to the above buddy system, since the area is allocated in units of powers of 2, for example, when a 2.1 GB file is allocated, a 4 GB area is required, and data is written to the block after the file allocation. In spite of the occurrence of an unoccupied area, a file cannot be added to the area in the conventional file system, and the area cannot be used effectively.
For example, even if the storage capacity of the entire storage device is 127 GB, a partition that is a physical or logical area allocated to a requester requesting a storage area must be managed as a 64 GB partition at the maximum. I didn't get it.
In other words, according to the buddy system, since the area is allocated in units of powers of 2, it is easy to allocate and release files and memory areas and manage the areas, but from areas of power-of-two sizes to areas that are not power-of-two sizes Since there is no management of the remaining area when acquiring the area, and management of the remaining area when acquiring the area of the power of 2 from the area that is not the power of 2 size, there is a problem that the area cannot be effectively used. is there.

Therefore, the present invention manages the remaining areas even when an area that is not a power of 2 size is acquired from an area of a power of 2 size or an area that is a power of 2 size is acquired from an area that is not a power of 2 size. An object of the present invention is to provide a technique capable of effectively using an area by an efficient method when the area management means allocates a file and a memory area to the area.

According to one aspect of the present invention, when the size of a partition is acquired and the size of the partition is represented by the product of the sum of powers of two different from each other and the size of the allocation unit of the partition, each power of two constituting the sum The area of the size is divided into sections by sequentially assigning the area as the main area to the sections in the order of the size, and the main area is divided into one-by-two main areas corresponding to the area and the main area of each size size Then, generate a multi-layer partition allocation table that stores allocation information indicating the file allocation status of each size area and main area, and allocate the area to the file based on the allocation information stored in the multi-layer partition allocation table Manage.

Further, according to another aspect of the present invention, when the number of powers of 2 that defines the size of the area is a division level of the area, the multi-layer division table shows the allocation information of the area in the order of the division level of the area. In the same division level, store in the order of arrangement within the division of the area, assign a division number that is an identification number for identifying the area to which the allocation information corresponds according to the storage order of the allocation information, and use the division number to Manage the allocation of

Further, according to another aspect of the present invention, when the size of the requested file is the sum of the sizes of the areas having different partition levels, the empty area of the partition level that is one greater than the largest of the partition levels is used. The primary allocation area that is larger than the allocation request size is searched for, and the primary allocation area is divided into the secondary allocation area, which is an area in which areas of different division levels are continuously assigned in the order of the division level, and the remaining areas. The secondary allocation area is divided into adjacent multi-layer areas, which are areas in which areas with different division levels are continuously assigned in the reverse order of the division level in which areas with different division levels of the secondary allocation area are assigned. Use is set for the allocation status of the multi-layer division allocation table corresponding to each of the areas, and the empty candidate status is set for the allocation status of the multi-layer division allocation table corresponding to the area constituting the adjacent multi-layer area That.

According to the present invention, since a partition is managed by a combination of power-of-two sizes, it is possible to efficiently manage a partition of any size efficiently and without waste. As a result, the files are allocated to the consecutive areas without waste, so that the areas can be effectively used.

It is a figure explaining the partition allocation in an external storage device. It is a figure explaining the allocation principle of a buddy system. It is a figure explaining the outline | summary of the division management in one Embodiment of this invention. It is a figure explaining the software and hardware example of an allocation system in one Embodiment of this invention. It is a figure explaining the hardware structural example in one Embodiment of this invention. It is a figure explaining the example of the initialization state of the division corresponding to the division size in one Embodiment of this invention. It is a figure explaining the example of a process flow of the front | former stage which initializes the division in one Embodiment of this invention. It is a figure explaining the process flow example of the back | latter stage which initializes the division in one Embodiment of this invention. It is a figure explaining the example of a processing flow which initializes the multilayer division | segmentation attachment table in one Embodiment of this invention for every division level. It is a figure explaining the outline processing flow of the whole process which allocates an area. It is a figure explaining the example of the processing flow which searches the empty area of an allocation request | requirement division | segmentation level, and acquires the division number of an empty area. It is a figure explaining the example of a processing flow which searches the empty area containing the size of an allocation request | requirement division | segmentation level, and acquires the division number of an empty area. It is a figure explaining the area search when there exists an empty area in the allocation request | requirement division | segmentation level by a specific example. It is a figure explaining the area search in case there is no empty area in the allocation request | requirement division level by a specific example. It is a figure explaining the example of a processing flow which acquires the tail number of the division level in one Embodiment of this invention. It is a figure explaining the example of a processing flow which divides a temporary allocation area into multiple layers and obtains a primary allocation area. It is a figure explaining the process which divides | segments a temporary allocation area into a multilayer and obtains a primary allocation area by a specific example. It is a figure explaining the example of the processing flow which searches the empty candidate area of an allocation request | requirement division | segmentation level, and acquires the division number of an empty candidate area. It is a figure explaining the example of a processing flow which searches the empty candidate area containing the size of an allocation request | requirement division | segmentation level, and acquires the division number of an empty candidate area. It is a figure explaining the example of a processing flow which divides the temporary allocation area whose allocation state is an empty candidate into a multilayer area, and obtains a primary allocation area. It is a figure explaining the example of the process flow of the front | former stage which divides a primary allocation area into multilayer and obtains the secondary allocation area of an allocation request | requirement classification. It is a figure explaining the example of a subsequent process flow which divides a primary allocation area into multiple layers and obtains the secondary allocation area of an allocation demand classification. It is a figure explaining the process which divides a primary allocation area into a multilayer and obtains the secondary allocation area of an allocation request | requirement classification by a specific example. It is a figure explaining the example of a rough processing flow of the whole process which releases the allocation area and tries connection of an empty area. It is a figure explaining the example of a processing flow which searches a multilayer division management table with a division number, and calculates | requires the division level corresponding to a division number. It is a figure explaining the example of a processing flow which tries the release of the area contained in a primary allocation area, and the connection of an empty area. It is a figure explaining the process which tries release of the area contained in a primary allocation area, and connection of an empty area by a specific example. It is a figure explaining the example of a processing flow which calculates | requires the division | segmentation allocation status in an allocation area, and pushes on a division | segmentation allocation status stack. It is a figure explaining the example of a processing flow which tries release of the division pair which a division number points out. It is a figure explaining the example of a processing flow which tries to make the allocation state of a high-order area empty by connecting a primary allocation area and an adjacent empty area. It is a figure explaining the process which tries to make the allocation state of a high-order area empty by connecting a primary allocation area and an adjacent empty area by a specific example. It is a figure explaining the processing flow which tries release of the division pair which a division number points out. It is a figure explaining the functional block structural example of the division management apparatus in one Embodiment of this invention. It is a figure explaining the functional block structural example of the area allocation means in one Embodiment of this invention. It is a figure explaining the functional block structural example of the area release means in one Embodiment of this invention.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described in detail with reference to the drawings.
FIG. 2A is a diagram illustrating an outline of partition management according to an embodiment of the present invention. The outline of the partition management in one embodiment of the present invention will be described with reference to FIG. 2A, and some terms related to the partition management of the present invention will be defined.

First, define the partition again. In the present invention, the “section” is an area allocated in advance as described with reference to FIG. 1A and is an area allocated first. In the present invention, it is assumed that the section has been acquired. In the present invention, a partition is a physical or logical area allocated to a request source requesting a storage area, and a subject that executes the request source and allocation is a file system or a disk allocator. There is no such restriction. In addition, the partition in the present invention is not limited to the storage area of the external storage device, and can be, for example, a pre-allocated area of various storage devices included in the data storage device 308 described later with reference to FIG. 2C.

Furthermore, the memory area defined in the present invention is defined again. The memory area is an area in a partition that is requested by a request source that requests allocation of a partition area and is allocated to the request source by the partition management device. If the request source is a file system, the memory area allocated to the file system is used as a file storage area by the file system. In the following description, a file system will be described as an example of a request source, so expressions such as file allocation and file and memory area allocation are used.

An “area” is an allocated or allocated area. Since a partition is a pre-allocated region, it is one of the zones and is a specific one of the zones.
Dividing the area into a plurality of areas having the size of the product of the power of 2 and the assigned unit area is called “multi-layer division”. A multi-layered area is referred to as a “multi-layer area”. In the following description, reference to the allocation unit area is omitted, and the notation of the area size may be simply a notation such as a size of 2n. For the sake of simplicity, for example, the value 3 stored in the storage area or the set value stored in the partition number 196a illustrated in FIG. 1A is called the partition number 196a. Sometimes called.

Hereinafter, the terms relating to the areas and the terms relating to the corresponding sections will be defined and compared.
“Division” means that an area is assigned or an area is divided and assigned, and is sometimes called “assignment”. A region is an area or section. In particular, when the size of a partition is expressed by a sum of powers of two different from each other, partitioning the partition by sequentially allocating each region of the power of 2 that constitutes the sum in the order of the size. It is called “Main Category”. In addition, the area assigned by the main division is called “main area”. In the present invention, the main division is an initial allocation for a partition. In the following description, it is assumed that the main area is assigned in order from the largest size from the top of the area, but it is possible to implement the present invention even in the reverse order, in order from the smallest size. It will be apparent to those skilled in the art from the following description.

In the illustration of FIG. 2A, the partition 690 has an overall size of 11, a main area 698 of size 2 ³ = 8, a main area 692 of size 2 ¹ = 2 and a main area 691 of size 2 ⁰ = 1. It is divided into. In other words,

main areas

698, 692, 691 are initially assigned to section 690.
The number of powers of 2 that is the area or the allocation size of the main area by the section or the main section is referred to as “section level”. Each division level of the

main areas

698, 692, 691 shown in FIG. 2A is 3,1,0.
Explaining in terms of the above defined terms, the partition management method, which is the principle of the present invention, divides a partition continuously into sections in which the allocation unit areas of the sections are assigned to different powers of two, and the sections are continuously Manage as assigned.

FIG. 2A further describes a tree structure group 790 corresponding to the multilayer section of the partition 690, and a bitmap 600 indicating the allocation state of the area in which the allocation state corresponding to the node of the tree structure group 790 is reduced to a bitmap. Has been.

The tree structure group 790 includes a tree having a node 680 corresponding to the main section to which the main area 698 is allocated as a root node, a tree having a node 624 corresponding to the main section to which the main area 692 is allocated as a root node, and a main area 691. It is composed of a tree having a node 6110 corresponding to the main section to be assigned as a root node. That is, the root node 680 corresponds to a main section (hereinafter, also referred to as a main section 788) that assigns the entire main section 698 of the section level 3 as indicated by a dotted arrow 788, and the root node 624 is a dotted line. As shown by an arrow 782, the root node 6110 corresponds to a main section (hereinafter, also referred to as a main section 782) that allocates the entire main area 692 of the section level 1, This corresponds to a main division (hereinafter, also referred to as a main division 781) in which the entire main area 691 of the division level 0 is allocated. The structure of each tree is similar to that of the tree structure 580 described in FIG. 1B in that it is a binary tree.

“8” described in the root node 680 corresponds to the size of the corresponding main area 698. The numbers in parentheses are division numbers for identifying individual divisions (sometimes referred to as division units) for dividing the division 690 into multiple divisions for each division level. The section number of the main section 788 corresponding to the largest size main section 698 is 0. In the following, the division number may be referred to as a division number of an area divided by the division identified by the division number. That is, instead of the expression of the section number of the main section 788, it may be referred to as the section number of the main section 698. In addition, an area divided by a division unit having a certain division number may be referred to as an area having the division number.

The node 640 connected to the root node 680 via the link 740 and the node 641 connected to the root node 680 via the link 741 correspond to the division in which the main area 698 is divided into two and the area of division level 2 and size 4 is allocated. To do. Each partition number is 1 and 2 following 0 which is the end of the partition number corresponding to the node of partition level 3 that is one level higher.

Below the node 640 are a size 2 node 620 connected by a link 720 and a size 2 node 621 connected by a link 721. Similarly, a size 2 node 622 connected by a link 722 and a size 2 node 623 connected by a link 723 exist below the node 641. These four nodes correspond to the divisions of the division level 1 area. Also, the division numbers corresponding to these four nodes are 3 to 6 following 2 which is the last number of the division number corresponding to the node of the division level 2 which is the next higher division level.

Below the node 620, a size 1 node 610 connected by a link 710 and a size 1 node 611 connected by a link 711 exist. Similarly, a size 1 node 612 connected by a link 712 and a size 1 node 613 connected by a link 713 exist below the node 621, and a node 622 connects by a link 714. A size 1 node 614 connected by a link 715 and a size 1 node 616 connected by a link 716 and a size connected by a link 717. There is one node 617. These eight nodes correspond to the division of the division level 0 area. The partition numbers corresponding to these eight nodes are the main partition 782 following the number at the end of the partition number corresponding to the node of the partition level 1 that is the partition level one higher in the tree having the root node as the node 680. 8 to 15 further following the number at the end of the partition number corresponding to the node of the tree corresponding to the root node of the partition level corresponding to 3.

Similarly, the node 618 connected to the root node 624 via the link 718 and the node 619 connected to the root node 624 via the link 719 divide the main area 692 into two parts and assign an area of division level 0 and size 1. Corresponding to
The partition number of the root node 624 is 7 following the end 6 of the partition number corresponding to the node at the partition level 1 of the tree corresponding to the main partition 788. The partition number of node 618 is 16 following the end 15 of the partition number corresponding to the node at partition level 0 of the tree corresponding to main partition 788, and the partition number of node 618 is 17 following it.
Further, the tree corresponding to the main section 781 to which the main area 691 of size 1 is assigned is configured only by the node 6110 that is the root node, as indicated by the dotted arrow 781. The partition number of the node 6110 is 18 following the end 17 of the partition number corresponding to the node at the partition level 0 of the tree corresponding to the main partition 782.

Next, the property of the division number according to the embodiment of the present invention will be described.
First, when the size of a partition subject to storage management is given, main division is performed according to the size, and multilayer division is performed for each division level. In the example of FIG. 2A, when a partition 690 having a size of 11 is given, a tree structure group 790 corresponding to the partition 690 can be generated. Then, from the node at the highest partition level of the tree structure group to the lower node, the left side is a young number within the same partition level, and a partition number can be uniquely assigned as shown in FIG. 2A.

Then, by managing the division number for each division level, given the division number, the division level that is the size of the area divided by the division unit identified by the division number can be obtained. Based on the difference between the first division number at the division level and the division number, the position on the division of the area to be allocated by the division unit can be obtained.

The bitmap 600 shown in FIG. 2A has a bit value obtained by reducing the allocation state corresponding to each node of the tree structure group 780 to a bitmap, as indicated by a dotted arrow. In the example of FIG. 2A, a 2-bit value corresponds to each node. The bit value of the bitmap 600 will be described in detail later. Here, the illustration of FIG. 2A will be briefly described below.

The bit value of the bitmap 600 represents the allocation state at the end of the initial allocation of the area corresponding to the node having each division number. The bit position 00 of the division level 3 indicated by the reference numeral 608 and having the division level 3 division number 609 value of 0 indicates that the area of the division number 0 division is empty. It is. The bit position 10 of the division level 2 indicated by reference numeral 604 and having the

division numbers

1 and 2 indicates that the allocation state of the corresponding area is pending. Similarly, the bit value having the division number 3 to 6 of the division level 1 indicated by the reference numeral 602 as the bit position is the reserved state of the corresponding area, and the bit value having the division number 7 as the bit position is applicable. Indicates that the allocation status of the area is empty. In addition, the bit values having the bit numbers of the division numbers 8 to 17 of the division level 0 indicated by the reference numeral 601 indicate that the allocation state of the corresponding area is pending, and the bit values having the bit number of the division number 18 are the corresponding area. Indicates that the allocation state of is empty.

The distinction between empty and pending in the allocation state is related to the priority when allocating an empty area of the allocation request size. That is, even if the area is substantially empty, the assigned state is used from the empty state. Details of this will be described later.

FIG. 2B is a diagram illustrating an example of software and hardware environment of the allocation system according to the embodiment of the present invention. The allocation system 100 includes an initialization unit 101 and a multi-layer division management unit 102. Initialization and multi-layer division management will be described in detail later.

The initialization unit 101 receives an initialization request including a partition size from an initialization program 201 such as a disk allocator or an initialization unit of a file system, and configures a data storage device 308, for example, a main storage device 305, an external It initializes the multi-layer division management information of the partitions allocated to the storage device 306, the remote storage device 307 accessed via the communication device (hereinafter also simply referred to as storage device).

The multi-layer division management unit 102 executes division division, that is, division division allocation by writing the multi-layer division management information in response to the area allocation request including the allocation request size from the file system 202, and the written multi-layer The division management information is read and an allocation result including the allocation category number is returned to the file system 202.

The file system 202 receives a file operation request from the system 200 that uses a file such as an application program / OS, and if the file operation request requires allocation of an area to the file, the file system 202 has a multilayer structure. An allocation request including the allocation request size is made to the category management unit 102, the allocation result is received, the category number returned from the multi-layer category management unit of the allocation system 100 is designated as the allocation category number, and the file is sent to the file operation system 203 Make an operation request. The file operation system 203 acquires the address of the area allocated from the multilayer partition management unit 102 by the specified partition number, and executes data writing or data reading as a file operation on the file stored in the storage device. The operation result is returned to the file system 202. The file system 202 returns the operation result returned from the file operation system 203 as a file operation response to the system 200 using a file such as an application program / OS.

Even if the file system 202 and the file operation system 203 are existing, the allocation system 100 according to the present invention can be used by adjusting the interface with the allocation system 100 according to the present invention. In addition to the system 200 that uses a file such as an application program / OS, the initialization program 201 can be applied as well. Therefore, descriptions of these programs and systems are omitted.

FIG. 2C is a diagram illustrating an exemplary hardware configuration according to an embodiment of the present invention.
According to one embodiment of the present invention, partition management of a storage device and allocation and release processing of an area to a file are performed by a data processing device 301 including at least a central processing unit 302 and a cache memory 303 using a data storage device 308. . The data storage device 308 includes a multi-layered division management table 309 for managing a multi-layered division level, which will be described later, a multi-layered division table 310 for holding an allocation state for each division, and a partition 311 to be managed. As shown in FIG. 2C, the data storage device 308 can be realized by the main storage device 305 or the external storage device 306, or a combination thereof, or a device disposed at a remote location connected via the communication device 307. It is also possible to use.

That is, the main storage device 305 can be the one in the data processing device 301, the multi-layer partitioning table 310 and the multi-layer partition management table 309 are held in the main storage device 305, and the partition 311 is assigned to the external storage device 306. However, it will be apparent from the following description that the present invention is applicable even when a partition is assigned to the main storage device 305.
In the example of FIG. 2C, the main storage device 305, the external storage device 306, and the communication device 307 are connected to the data processing device 301 by one bus 304, but the connection method is not limited to this.
Although not particularly illustrated, it is natural that a temporary storage area corresponding to each process is used in order to use various values obtained during the process in a later process. In the following description, the name of the temporary storage area may be the name of the data stored or set in the temporary storage area. Conversely, the name of the temporary storage area May be the name of the data to be stored or set.

Next, with reference to FIG. 3 and FIGS. 4A to 4C, initialization of multi-layer division management information in an embodiment of the present invention will be described. Note that initialization of the multi-layer division management information may be referred to as division initialization.

FIG. 3 is an explanatory diagram illustrating an example of an initialization state of a partition corresponding to the partition size.
As described above with reference to FIG. 2B, partition initialization is performed by allocating a partition to the file system and then receiving an initialization request including the partition size of the partition from the initialization program. This is done by the system initialization unit. In the partition size 120 shown in FIG. 3, the partition size “52” allocated to the file system and included in the initialization request for the partition 311 is displayed in binary format, and 52 = 2 ⁵ +2 ⁴ +2 ² Therefore, the second bit, the fourth bit, and the fifth bit are “1”.

The partition configuration main partition table 130 is created according to the partition size received from the initialization program. The partition configuration main partition table 130 is composed of 1-bit entries corresponding to the partition levels as indicated by the subscripts 0 to 7. In the example of FIG. 3, the partition level 0 is changed from the partition level 0 to the highest partition level 7. It has 8 corresponding entries. As indicated by arrows 122, 124, and 125, the bit value of the corresponding partition level entry is set according to the bit value when the partition size is displayed in binary format.

In the partition 311 shown in FIG. 3, as indicated by the arrows 135, 134 and 132 from the partition configuration main partition table 130, the entry whose bit value is “1” in the partition configuration main partition table 130

Main sections

185, 184, and 182 of the partition level size corresponding to are initially allocated.
In addition, the division level corresponding to the main area may be called the main division level. That is, it can be said that the partition configuration main partition table 130 displays the main partition level with the bit value 1.

As shown in FIG. 3, the multi-layer division management table 309 includes a main division number management table 309a, a head number management table 309b, and a total division number management table 309c. As shown by the subscripts 0 to 7, the main partition number management table 309a and the head number management table 309b are configured by entries corresponding to the partition levels as in the partition configuration main partition table 130. An entry in the main division number management table 309a is composed of the main division number 114, and an entry in the top number management table 309b is composed of the top number 116. The value of each entry is set in the initialization process based on the value of the partition configuration main division table 130. The total number of partitions management table 309c is set with a total number of partitions 115 that is the total number of partition units that are partitioned in accordance with the size of the partition to be initialized. In other words, the total number of sections 115 divided into the size of each section level is set for each section level in the total section number 115. In the example of FIG. 2A, 19 corresponding to the division numbers 0 to 18 is set. In the example shown in FIG. 3, 101 corresponding to the division numbers 0 to 100 is set.

In the main division number 114, when the multi-layer division table 310 described later has an entry corresponding to a main division of a certain division level, the division number corresponding to the main division is stored in the entry of the division level. At the division level, when the multi-layer division table 310 does not have an entry corresponding to the main division, “−1” is stored as an unexpected division number in the entry at the division level. In the example of FIG. 3, the value of each entry of the main division number 114 is -1, -1, 0, 3, -1, 22, -1, -1, from the higher division level.

The head number 116 is a section number associated with the first entry of the section division table at each section level among the section numbers uniquely associated with each entry of the multi-layer section table 310, and is exemplified in FIG. Further, the value of each entry of the leading number 116 is -1, -1, 0, 1, 4, 10, 23, 49 from the higher division level. The setting of these values will be described in detail later with reference to FIGS. 4A to 4C.

The multi-layer division table 310 is for managing the allocation state of the area in the partition 311 and is the same as the allocation bitmap 600 shown in FIG. 2A. However, as shown in FIG. 3, the in-level number 172 is associated with the multi-layer division table 310 in addition to the division number 171. The multilayer division table 310 is generated and initialized based on the value of the multilayer division management table 309 in the initialization process.

As shown by dotted arrows 140 to 145 in the figure, the multi-layered division table 310 is divided into the division levels from the lowest division level to the highest division level where the bit value of the division configuration main division table 130 is 1. It consists of the corresponding division table 160-165.
An entry in each partition allocation table is composed of a 2-bit allocation state 170 of the section of the partition number corresponding to the entry. The bit values “00”, “01”, “10”, “11” in the allocation state correspond to the “empty”, “empty candidate”, “hold”, and “in use” states of the areas of the respective division numbers. To do. Although the meaning of these states will be described in detail later, by using 2 bits for the allocation state 170, “empty candidates” can be identified, and the storage area can be used more efficiently.

In addition, each entry in the division table 160 to 165 has two main divisions, that is, one corresponding to the main division, and two division levels one level higher, as indicated by arrows 75, 74 and 72 in FIG. There is a division to be divided, that is, a division corresponding to an area obtained by dividing an area of an upper division level into two.

Each entry in the multi-layer partitioning table 310 includes a partition number that is a serial number from 0 from the top entry in the partitioning table 165 at the highest partition level to the last entry in the partitioning table 160 at the lowest partition level. 171 can be associated. Further, for each division table, an in-level number 172 that is a serial number starting from 0 can be associated from the first entry to the last entry. The start number 116 of the start number management table 309b is the start number of the partition number at the corresponding partition level.

As described above with reference to FIG. 2A, when a partition size, that is, a partition configuration main partition table is given, the configuration of the multi-layer partition table is uniquely determined and is partitioned by the partition unit identified by the partition number. The location and size of the area to be determined is also uniquely determined.
In the initial state of the multi-layered division table 310, as indicated by arrows 75, 74, and 72, "00" indicating empty is displayed in the allocation status 170 of the entry corresponding to the main zone, and the allocation status of the other entries. Is initially set to “10” for displaying the hold. Therefore, in the case of an area where the allocation state is empty, the

main area

185, 184, 182 is divided and assigned to the area 311, and an empty area is set among the areas managed in the multi-layered area division table 310. In this sense, the partition 311 is initially allocated. The initial setting of the multi-layer division table 310 will be described in detail later with reference to FIGS. 4A to 4C.

Note that the above-described method of assigning the division number 171 and the in-level number 172 is merely an example, and if the file allocation management described later is possible, for example, the first number is set to 1 instead of 0. It will be apparent to those skilled in the art that various modifications are possible, such as reversing the order of application.
As described above, according to the present invention, the partition 311 assigned to the file system is managed in a multi-layered manner by using the multi-partition partitioning table 310 and the allocation state 170 corresponding to the partition level for the same area. .

According to the allocation of the area using the multi-layer partitioning table 310, according to the size of the allocation request, the division number of an empty area of a certain division level or a continuous empty area of a different division level is obtained by searching the multi-layer partitioning table, This is done by setting the division number in use. If there is no empty area, an empty candidate area is searched.
As shown in FIG. 2B, the file system that has requested the file allocation returns the segment number that is in use as the allocation segment number from the allocation system. When the file operation system receives a file operation request designating this partition number, the file operation system receives address information of the section of this partition number from the multi-layer partition management unit. The multi-layer division management unit searches the multi-layer division management table 309 by the division number, obtains the division level and the intra-level number of the area corresponding to the division number, and thereby determines the position and size in the division of the area assigned to the file. I can know. These processes will be described in detail later.

Next, a process for initializing a partition will be described with reference to FIGS. 4A, 4B, and 4C. Here, the processing for initializing the partitions is specifically processing for initializing the values of the multilayer partition management table 309 and the multilayer partition partitioning table 310 shown in FIG. 3, for example. In the following, description will be made with reference to the multi-layer division management table 309 and the multi-layer division table 310 as an example.

FIG. 4A is a diagram illustrating an example of a processing flow in the previous stage for initializing a partition.
As shown in FIG. 4A, first, in step S401, from the partition size received from the program requesting partition initialization, the bit of the partition level entry according to the bit value when the partition size is displayed in binary format. Create a partition main partition table with values set.

Next, in step S402, the highest partition level of the partition configuration main partition table is set as the main partition level, and in step S403, the total number of sections in the total partition number management table is set to 0 as an initial value. The process proceeds to S404. In the example shown in FIG. 3, a partition configuration main partition table 130 is created, and 7 is set as the main partition level. The main partition level in which the highest partition level of the partition configuration main partition table in step S402 is set is an example of the temporary storage area (not shown) described above.

In step S404, the bit value of the partition configuration main partition table pointed to by the main partition level is extracted. In step S405, it is determined whether or not the extracted bit value is significant, that is, a value of 1.
If the extracted bit value is not 1, the process branches to step S406, the value “−1” is set in the main partition number management table pointed to by the main partition level, and then in step S407, the head number management pointed to by the main partition level is set. A value “−1” is set in the table, and the main division level is decreased by 1 in step S408, and the process returns to step S404.
The loop processing from step S404 to step S408 is repeated for the first time in step S405 until it is determined that the bit value of the partition configuration main division table is significant. In the example shown in FIG. 3, since the bit value of the partition configuration main partition table is 0 until the main partition level is set to 5, this loop process causes the main partition number management table 309a and the head number management table 309b to be The value “−1” is set for both the division level 7 and the division level 6.

On the other hand, if the extracted bit value is 1, the process proceeds to step S409, the value “0” is set as the head number, the value “1” is set as the number of sections in step S410, and step S411 shown in FIG. Proceed to The head number in step S409 and the number of sections in step S410 are also examples of the temporary storage area (not shown) described above. In each case, the name of the data is the name of the temporary storage area.

FIG. 4B is a diagram illustrating an example of a subsequent processing flow for initializing a partition.
In step S411, a value obtained by adding 1 to the main number and subtracting 1 is set to the main division number. In the first process of step S411, 1 is set to the number of sections by the process of step S410, and in the subsequent processes, the value is set by the process of step S423 described later.

Next, in step S412, the main division number is set in the main division number management table pointed to by the main division level, and in step S413, the head number is set in the head number management table pointed to by the main division level. In the first processing of step S412 and step S413 in the illustration of FIG. 3, 0 is set to the division level 5 of the main division number management table 309a and the head number management table 309b, respectively.

Next, proceeding to step S414, the allocation state of the multi-level partitioning table at the partition level indicated by the main partition level is initialized. Details of the processing in step S414 will be described later with reference to FIG. 4C.

In step S415, the number of sections obtained in step S414 is added to the total number of sections in the total section number management table, and the process proceeds to step S416.
In step S416, it is determined whether the main division level is the lowest division level. If it is the lowest division level, the process ends. If it is not the lowest division level, the process branches to step S417.
In step S417, the main division level is decreased by one, the number of divisions is added to the top number in step S418, the number of divisions is doubled in step S419, and the flow proceeds to step S420.

In step S420, the bit value of the partition configuration main partition table at the partition level indicated by the main partition level is extracted. In step S421, it is determined whether the extracted bit value is significant, that is, the value 1. If the bit value of the partition configuration main partition table pointed to by the main partition level extracted in step S420 is not significant, the process branches to step S422, sets the value “−1” in the main partition number, and returns to step S412. For example, it branches to step S423, adds 1 to the number of divisions, and returns to step S411.
The loop processing from step S411 to step S423 is repeated until it is determined in step S416 that the main division level is the lowest division level. At this time, if the bit of the partition configuration main partition table of the partition level pointed to by the main partition level is a bitless bit, that is, the bit value is “0”, the main partition number of the partition level has the value “−” as described above. 1 ”is set.

In the example shown in FIG. 3, 1 is set in the division level 4 of the head number management table 309b, and the division level 4 of the main division number management table 309a is set to the head number 1 and the division in the second processing in step S411. 3 is set by subtracting 1 from the sum of Equation 3. In addition, the division level 3 of the head number management table 309b is set to 4 which is obtained by adding the number of divisions 3 to the head number 1 of the division level 4, and the division level 4 of the main division number management table 309a is set in the process of step S422. −1 is set. In the same manner, 22, -1, and -1 are set in the

division levels

2, 1, and 0 of the main division number management table 309a, respectively. Further, 10, 23, and 49 are set in the

division levels

2, 1, and 0 of the head number management table 309b, respectively, and 101 is set in the total division number management table 309c.

FIG. 4C is a diagram illustrating an example of a processing flow for initializing the multi-layer partitioning table for each partition level in the embodiment of the present invention. By the processing flow illustrated in FIG. 4C, the partitioning table corresponding to each partition level from the lowest partition level constituting the multi-layer partitioning table to the highest partition level having a bit value of 1 in the partition configuration main partition table Is initialized. In the example of FIG. 3, this is a process of setting the values of the partitioning tables 160 to 165 to the values shown. Each time the loop processing from step S411 to step S423 shown in FIG. 4B is executed, the partitioning table from the partitioning table 165 to the partitioning table 160 is initialized, and the initialization of the multi-layer partitioning table 310 is completed. . *

As shown in FIG. 4C, in step S431, the head number is set to the section number, and in step S432, a value obtained by adding the number of sections to the head number and subtracting 1 is set. The head numbers in step S431 and step S432 are set in step S409 shown in FIG. 4A or step S418 shown in FIG. 4B. Further, the number of sections in step S432 is set in step S410 shown in FIG. 4A or step S419 or step S423 shown in FIG. 4B.

The processing of step S431 and step S432 is processing for setting the first division number and the last division number of the division table attached to each division level.
Next, in step S434, it is determined whether the segment number and the end number are equal. If the section number and the end number are not equal, in step S435, the allocation state of the multi-layer partition table attached by the section number is set to hold, and in step S436, the section number is incremented by one and the process returns to step S434. Repeat the determination of whether the end numbers are equal. The loop processing from step S434 to step S436 is performed in the allocation state of the division table attached by the division number from the first division number of the division division attachment table corresponding to the division level being processed to the division number immediately before the end number. This is a process for setting a hold.

On the other hand, if it is determined in step S434 that the section number and the end number are equal, the process proceeds to step S437, where it is determined whether the main section number is the value “−1”. The main division number here is the one set in step S411 or step S422 shown in FIG. 4B.

If it is determined in step S437 that the main division number is the value “−1”, in step S438 the suspension is set in the allocation state of the multi-layer division table attached by the division number, and the process is terminated. In step S437, If it is determined that the main division number is not the value “−1”, the allocation state of the multi-layer division table attached to the division number is set to empty in step S439, and the process is terminated.
The processing in step S438 and step S439 is processing for setting the allocation state of the segment number at the end of the segmentation table corresponding to the segment level being processed. As illustrated in FIG. 3, the division number of the division unit corresponding to the main area is the end of the division number corresponding to the division level of the main area, and the allocation state of the division table of the division number at the end is As seen in the allocation states of the

division numbers

0, 3, and 22, “00”, that is, empty. All other assignment states of the multi-layered division table are “10”, that is, hold.

The reason for introducing “hold” in addition to “empty” in the area allocation state in the section initialization described in detail above is to facilitate securing of large continuous empty areas. For example, in the example of FIG. 3, when the file system requests allocation of an area of size 16, the main area 184, that is, the area of the partition number 3 can be selected from the main areas that are empty. . If there is no allocation state of holding, the areas of the

division numbers

1 and 2 are also empty, and if the allocation request for an area having a size larger than 16 is generated in the future, the division number 1 is assigned so that a continuous area can be assigned. A means for selecting the area of the division number 3 from 2, 3 is required to use the area effectively. However, that means is more complicated than introducing “hold” in the allocation state.

In this embodiment, a state of “pending” is introduced and represented by a 2-bit value “10”. As described above, the allocation state of a zone is set to “pending”. Is excluded from the allocation target. Therefore, in the above-described step S438, it is possible to set a non-empty state indicating that allocation to the area of the division number is not performed in the allocation state of the multi-layered division allocation table indicated by the division number.

Next, with reference to FIGS. 5 to 11C, allocation of areas using the multi-layer division management information in one embodiment of the present invention will be described. As described above with reference to FIG. 2B, area allocation is performed by the multi-layer division management unit of the allocation system by accepting an allocation request including an allocation request size from the file system after partition initialization. .

FIG. 5 is a diagram for explaining an example of a schematic processing flow of the entire processing for assigning areas.
First, in step S501, the allocation request size is set, and in step S502, the power of the minimum power-of-two size including the allocation request size is set as the allocation request classification level. For example, if the allocation request size is 11 (binary number “01011”), the allocation request classification level is set to 4, which is a power of 16 (binary number “10000”) = 2 ⁴ . If the allocation request size is 8, which is a power of 2, then 3 which is the power is set.

Next, in step S503, referring to the multi-layered division table, an empty area in which the assigned state of the area is “empty” is searched from the areas of the allocation requesting division level, and the division number of the empty area is acquired. Details of the processing in step S503 will be described later with reference to FIG.

In step S504, it is determined whether or not the division number of the empty area has been acquired. If it can be acquired, the process proceeds to step S507, and if it cannot be acquired, the process proceeds to step S505.
In step S505, an empty candidate area whose allocation state is “empty candidate” is searched from the areas at the allocation request classification level with reference to the multi-layered division table, and a classification number of the empty candidate area is acquired. Details of the processing in step S505 will be described later with reference to FIG.
In step S506, it is determined whether or not an empty candidate state area can be acquired. If it can be acquired, the process proceeds to step S507. If it cannot be acquired, the area allocation process ends as an acquisition failure.

In step S507, it is determined whether the size of the area corresponding to the classification level of the classification number acquired in step S503 is larger than the allocation request size set in step S501. The allocation request size is not equal to a power of 2, that is, when it is displayed in binary format, there are multiple significant bit positions (in this case, the allocation request is called a multi-bit request, while the allocation request is equal to a power of 2; If there is a single bit request), the size of the acquired area is larger than the allocation request size. Therefore, the determination in step S507 is to determine whether the allocation request is a multibit request or a single bit request.

If the determination in step S507 is a single bit request, the process proceeds to step S509. If the determination in step S507 is a multibit request, the process branches to step S508. In step S508, the primary allocation area corresponding to the acquired classification number is divided into multilayer areas, the secondary allocation area of the allocation request classification is obtained, and the process proceeds to step S509. In addition, although the area | region obtained by secondary allocation is a multilayer area which consists of a several area, the area | region is called a secondary allocation area below.
Details of the processing in step S508 will be described later with reference to FIGS. 11A to 11C. The primary allocation and the secondary allocation will be described later in detail.

In step S509, as the allocation result, the allocation area category number that is primarily allocated to the allocation category number is set, and the process ends. The allocation division number set here is returned to the file system as the allocation result, as shown in FIG. 2A. The multi-layer partition management unit of the allocation system uses the allocation section number corresponding to the file to calculate the relative address from the start position of the partition allocated to the file system to the start position of the area allocated to the file requested for allocation. It can be obtained from the multi-layer division management table.

Next, details of the processing in steps S503, S505, and S508 in FIG. 5 will be described.
FIG. 6 explains the details of the processing in step S503 of FIG. 5, and describes an example of a processing flow for searching for an empty area at the allocation requesting division level from the multi-layered division table and acquiring the division number of the empty area. It is a figure to do.

As shown in the figure, in step S601, an empty area including the size of the allocation requesting division level is searched from the multi-layer division allocation table, and an empty area division number is acquired. According to the multi-layer partition management according to an embodiment of the present invention, even when an empty area at the allocation required partition level cannot be found, if an empty area at a partition level higher than the allocation partition level exists, it can be found. For example, in the example shown in FIG. 3, when the allocation request division level is 3, as shown in the division division table 163 of the division level 3, there is no space in the division level 3 area. However, an empty area with a division number 3 exists in the division level 2, and this division number 3 is acquired. In this way, allocating an empty area at a division level larger than the allocation division level may be referred to as provisional allocation. Details of step S601 will be described later with reference to FIGS. 7A to 7D.

Next, in step S602, it is determined whether or not the division number has been acquired in the process of step S601. If it has not been acquired, an acquisition failure is returned and the process ends. If it has been acquired, the process proceeds to step S603.
In step S603, it is determined whether the classification level (acquired classification level) related to the classification number acquired in the process of step S601 is equal to the allocation request classification level. This determination corresponds to determination of whether or not provisional allocation has been performed. If the acquired category level and the allocation request category level are equal, acquisition success is returned and the process is terminated. If the acquired category level and the allocation request category level are not equal, that is, if provisional allocation is performed, the process branches to step S604. .

In step S604, the temporarily allocated area that has been temporarily allocated is divided into multi-layer areas, and the primary allocation area of the division number of the allocation request classification level is obtained. That is, an area having the size of the allocation request classification level is acquired as the primary allocation, the classification number is acquired, acquisition success is returned, and the process is terminated. Details of step S604 will be described later with reference to FIGS. 8A and 8B.

FIG. 7A explains the details of the processing in step S601 in FIG. 6, and a processing flow for searching the empty area including the size of the allocation requesting division level from the multi-layer division table and obtaining the division number of the empty region It is a figure explaining an example.

As shown in the figure, in step S701, the allocation request category level is set as the category level. Here, the value of the allocation request classification level is set in step S502 of FIG.
Next, in step S702, the head number indicated by the division level is extracted from the head number management table. In step S703a, the extracted head number is set as the section number, and in step S703b, the last number of the section number in the section level partition table currently being processed is set as the end number. Details of the processing in step S703b will be described later with reference to FIG. 7D.

Next, in step S704, the allocation state of the multi-layer partitioning table indicated by the value set in the segment number is read. In step S705, it is determined whether the read allocation state is empty. If it is determined in step S705 that the allocation state is empty, the process proceeds to step S710.

On the other hand, if it is determined in step S705 that the allocation state is not empty, the process branches to step S706. In step S706, it is determined whether the division number is equal to the end number set in step S703b. If the section number is not equal to the end number, the process branches to step S707, and the section number is incremented by 1, and the process returns to step S704. Thereafter, the division number is incremented by 1 within the same division level to search for an empty area.

If it is determined in step S706 that the segment number and the end number are equal, the process proceeds to step S708, and it is determined whether or not the segment level is the highest segment level. If the classification level is the highest classification level, an acquisition failure is returned and the process is terminated.
When it is determined in step S708 that the division level is not the highest division level (when the division level is a lower division level than the highest division level), the process proceeds to step S709, and 1 is added to the division level. The process returns to step S702.

Returning to step S702, the above processing is repeated, and the allocation state of the multi-layer division table for the higher division level is searched one by one. If an empty area is obtained as a result of the search, the process proceeds to step S710.

In step S710, the allocation state of the multi-layer partitioning table indicated by the value set in the segment number is set to be in use, and the process ends. As the processing result in FIG. 7A, the value set for the partition number and the partition level, which are temporary storage areas, and data indicating acquisition success or acquisition failure are output as search results.

7B and 7C are diagrams for explaining the empty area search processing shown in FIG. 7A by a specific example with reference to the multi-layer division management table 310 shown in FIG.
The example shown in FIG. 7B is an example in which the allocation request is a multi-bit request and an empty area is searched at the allocation request classification level. As shown in the figure, the bit value 1 is set in the first bit (partition level 1) and the third bit (partition level 3) of the allocation request size 220. Therefore, the allocation request category level 234 is set with a category level 4 higher than the category level 3 as indicated by the dotted arrow 224. The setting up to this point is performed by the processing in steps S501 and S502 in FIG.

Next, according to the partition level 4 allocation request 240 indicated by the solid line arrow in the figure, the empty search 244 of the partition level 4 partition allocation table before allocation of the multi-layer partition partition table 310 with the reference numeral 164a is performed. In the example shown in the figure, the assignment state of the division number 171 of 1 and 2 is “11”, and the assignment state of the division number 3 is “00” and is empty. Therefore, the empty area is acquired 274a as indicated by the solid line arrow in the figure, and the division number 171 is set to the allocation state of 3 as shown in the division level 4 division division table after the assignment with the reference numeral 164b. 11 "is set and it is displayed that it is in use. As a result, as indicated by the arrow 274 in the figure, the section 311 is assigned with the section level 4 and the section number 3 as the primary allocation section 280.
The above-described allocation of the primary allocation area 280 is executed by the loop processing in steps S704 to S707 and the processing in step S710 shown in FIG. 7A.

Next, a case where the allocation request is a single bit request and an empty area cannot be searched at the allocation request classification level will be described with reference to the example shown in FIG. 7C. Although the allocation request size 220 is not described in FIG. 7C, the allocation request 241a indicated by the dotted arrow is of the allocation request category level 1, and therefore the allocation request is a single bit request in which only the first bit is 1. .

In the flow shown in FIG. 7A, if the empty area cannot be searched for at the allocation request division level, the process branches to step S708 in step S706, and the loop processing of step S702 via step S709 is performed at the upper division level. Will repeat until you find.

Corresponding to this repetitive processing, the partitioning table of the upper partition level is sequentially referred to from the lower partition level (partition level 1 in FIG. 7C) shown in FIG. 7C, and empty in the searched partition level. This is a process of searching for a higher division level when there is no state area. In order from the division level 1, the empty state is searched in order from the top number for each division level area.

First, a search for an empty area of division level 1 that is the allocation requesting division level indicated by the dotted arrow 241a in the figure is requested, and among the multi-layer division table 310, the allocation state of the division level table 161 of the division level 1 is assigned to the division number. In the ascending order, the empty state is searched from the division number 23 which is the head number to the division number 48 which is the tail number (see arrow 241b in FIG. 7C). In the example shown in the figure, there is no empty area even when searching up to the 48 area whose division number is the end number, and therefore, the division level which is the upper division obtained by adding 1 to the division level indicated by the dotted arrow 242a in the figure. A search for two empty areas is required. Then, an empty state is searched for the allocation state in the partition level attached table 162 at the partition level 2 from the partition number 10 as the top number to the partition number 22 as the end number in the ascending order of the partition numbers (see the arrow 242b in FIG. 7C). . Since there is no empty area even if the search is performed up to the area having the end number of the section number 22, the search for the empty state of the section level 3 which is the upper section is requested by adding one to the section level (FIG. 7C). (See arrow 243a). Similarly, with respect to the division level table 163 at the division level 3, the allocation state from the division number 4 area which is the first number to the division number 9 area which is the last number is sequentially searched (see arrow 243b in FIG. 7C). Since an empty area cannot be obtained, an empty area search of an upper level division level 4 is requested by adding 1 to the division level (see 244a in FIG. 7C).
In the example of FIG. 7C, as a result of searching for an empty area (see 244b in FIG. 7C) of the division level 4 division division table before the assignment with the reference numeral 164a, the division number 2 is assigned to the allocation state “00”. As shown in FIG. 7C, the allocation state of the division number 2 in the division level table of the division level 4 after the allocation with the reference numeral 164b is in use. It is changed to “11” to be displayed. That is, as indicated by the arrow 244d, the area of division number 2 is provisionally allocated to obtain a provisionally allocated area 280a.

In addition, about the search process of an empty area, it is not limited to the method of searching said division number in an ascending order, You may employ | adopt arbitrary search algorithms.
Next, a process of acquiring the last number of the division number of the division level attached table at the division level currently being processed will be described.

FIG. 7D is a diagram illustrating an example of a processing flow for acquiring the end partition number of each partition level in an embodiment of the present invention.
As shown in FIG. 7D, in step S721, it is determined whether the currently processed partition level is the lowest partition level. If it is not the lowest partition level, the process proceeds to step S722, and if it is the lowest partition level, step S724 is performed. Proceed to

In step S722, the head number indicated by the value obtained by subtracting the classification level by one is extracted from the head number management table, and in step S723, the value subtracted by one from the extracted head number is set as the tail number, and the process ends. To do.
On the other hand, in step S724, the total number of sections is extracted from the total number of sections management table, and in step S725, a value obtained by subtracting one from the extracted total number of sections is set as the end number, and the process is terminated.
Through the processing described above with reference to FIGS. 7A to 7D, the primary allocation area or the temporary allocation area is obtained.

FIG. 8A explains the details of the process of step S604 in FIG. 6, and the process of dividing the temporary allocation area obtained by the process of step S601 into multiple sections and obtaining the primary allocation area of the section number of the allocation request section level. It is a figure explaining the example of a flow.

First, in step S801, the head number indicated by the division level acquired in step S601 of FIG. 6 is extracted from the head number management table. Since the processing flow in FIG. 8A is based on the assumption that provisional allocation has been performed, the acquired classification level is larger than the allocation request classification. Next, in step S803, a value obtained by subtracting the head number extracted in step S801 from the acquired division number is set to the in-level number. In step S804, a value obtained by subtracting one from the acquired category level is set in the category level, and the process proceeds to step S805.

In step S805, the head number indicated by the division level is extracted from the head number management table. In step S807, a value obtained by doubling the value set in the in-level number is set in the in-level number. This process corresponds to the fact that there are two areas at the lower level of the partition corresponding to a certain area.

Next, in step S808, a value obtained by adding the in-level number to the head number is set to the division number. In step S809, a value obtained by adding 1 to the section number is set to the pair section number. For example, the in-level numbers of the areas in the one sub-division level that is obtained by dividing the area in which the in-level number is “10” at a certain division level are “20” and “21”.

Next, in step S810, “in use” is set in the allocation state of the multi-layer partitioning table indicated by the section number, and in step S811, empty is set in the allocation state of the multi-layer partitioning table pointed to by the pair of partition numbers. The process proceeds to S812.

In step S812, it is determined whether the partition level is higher than the allocation request partition level. If so, the process branches to step S813, and the value set in the partition level is decreased by one, and the process returns to step S805. If the determination result is not large, that is, if the classification level is equal to the allocation request classification level, the process is terminated. As explained above, the acquired classification level is larger than the allocation request classification. Then, since the loop processing from step S805 to step S813 is repeated while decreasing the partition level by one, it is determined in step S812 that the partition level is not larger than the allocation request partition level. Is equal to.

By the above processing, the temporarily assigned area is divided into multiple layers, and the primary assigned area is obtained. The process of step S804 and the loop process of steps S805 to S813 start from the provisionally assigned area, divide the area into two pairs of areas at the next lower division level, and In use in the assigned state, empty is set in the assigned state in the area on the old number side.

According to one embodiment of the present invention, without assigning the entire acquired temporary allocation area in use, an area at the allocation request classification level is primarily allocated to the younger number side of the provisional allocation area classification number, and the remaining continuous Since the area is set as empty, the area can be used efficiently. It is obvious to those skilled in the art that the primary assignment is not limited to the younger number side of the division number, but can be the older number side.

FIG. 8B is a diagram illustrating a process for obtaining the primary allocation area by dividing the provisional allocation area shown in FIG. 8A into a multilayer by referring to the multilayer division management table 310 shown in FIG. 3 by a specific example.
In the example shown in FIG. 8B, the allocation request is a single-bit request similar to that illustrated in FIG. 7C, and the primary allocation area is acquired from the temporary allocation area 280a shown in FIG. 7C.

FIG. 8B shows a temporary allocation area (section number 2) which is an allocation state before the primary allocation by reference numeral 280a. A division level 4 division table attached with a reference numeral 164b according to a multi-layer division request that requires division of each related area from division level 4 to division level 1 of the temporary allocation area, indicated by an arrow 241 from the temporary allocation area 280a. The process starts from The division assignment table 164b stores the same assignment state as shown in FIG. 7C.

By the classification request indicated by the arrow 274a from the entry in which the allocation status is set to be in use by provisional allocation and the segment number 171 value of the segment level assignment table 164b is 2, and the value in the level number 172 is 1, “11” in use is set in the allocation state of the area by the division of the division unit with level number 2 in the division level 3 division table 163 (hereinafter referred to as “area with level number 2”). , An area based on the division of the division unit of level number 2 and the division unit of level number 3 constituting the division pair 293a (hereinafter referred to as "the area of level number 2 and the division of level number 3 constituting the division pair 293a") "00" is set in the allocation state. This setting process is executed in the processes of steps S805 to S811 in FIG. Since 4 is extracted as the head number and 1 is set to the in-level number by the initial processing in steps S801 to S803 in FIG. 8A, 2 is set to the in-level number in step S807. Is done. Then, 6 is obtained by adding the in-level number 2 to the head number 4 as the division number. Further, 7 is obtained as the pair division number.

Similarly, in response to the classification request indicated by the arrow 273a corresponding to the entry of classification number 6 being used, the allocation status of the area of level number 4 (division number 14) in the division level allocation table 162 is being used. “11” is set, and an empty “00” is set in the allocation state of the intra-level number 4 and the intra-level number 5 (section number 15) constituting the segment pair 292a.
On the other hand, since the allocation state of the division number 7 is set to empty, as shown by the arrow 273b, the area 283b of the empty division number 7 is added to the temporary allocation area after the primary allocation indicated by the reference numeral 280b. Assigned.

Next, in response to a partition request indicated by an arrow 272a corresponding to the entry of the partition number 14 in use, the allocation state of the area of the level number 8 (section number 31) in the partition level allocation table 161 of the partition level 1 is in use “11 "Is set, and an empty" 00 "is set in the allocation state of the in-level number 9 (section number 32) that forms the section pair 291a with the section in the in-level number 8. Since the allocation state of the division number 15 is set to empty, the area 282b of the empty division number 15 is allocated to the temporary allocation area 280b after the primary allocation, as shown by the arrow 272b.

Since the multi-layer division request is up to division level 1, as indicated by the

arrows

271a and 271b, the temporary allocation area 280b after the primary allocation is primarily allocated as the area 281a of the division number 31 is in use. Area 281b with number 32 is allocated as empty.

Due to the above-mentioned multilayer division, the temporary allocation area 280b after the primary allocation is an adjacent multilayer adjacent to the primary allocation area 281a, which is composed of a primary allocation area 281a in use and

empty areas

281b, 282b, 283b as shown in the figure. Divided into areas 290b.

FIG. 9 explains the details of the processing in step S505 of FIG. 5, and is an example of a processing flow for searching for empty candidate areas at the allocation requesting division level from the multi-layer division table and acquiring the division number of the empty candidate area. FIG. The empty candidates are set as the allocation state of the areas constituting the remaining adjacent multi-layer areas when the secondary allocation is performed by dividing the primary allocation area in the multi-bit request, which will be described later with reference to FIGS. 11A to 11C Is done.

The search for the empty candidate area is performed in the same manner as the search for the empty area, and the flow shown in FIG. 9 of the empty candidate area search corresponds to the flow of the empty area search shown in FIG. Since step S904 is obtained by replacing the empty area with the empty candidate area in steps S601 to S604 shown in FIG. 6, description thereof will be omitted.

FIG. 10A illustrating the details of the processing in step S901 in FIG. 9 is a processing flow for searching for empty candidate areas that are larger than the allocation request classification level size from the multi-layered division table and acquiring the division number of the empty candidate area. 7A corresponds to the diagram illustrating a processing flow for searching for an empty area having a size equal to or larger than the allocation requesting division level from the multi-layer division allocation table shown in FIG. 7A and acquiring the division number of the empty area. Steps S1001 to S1009 in FIG. 10A are different from those in which it is determined in step 705 in FIG. 7A that the allocation state is empty, but only where the allocation state is determined to be an empty candidate in step S1005. A description of FIG. 10A is also omitted.

Similarly, FIG. 10B for explaining the details of the process of step S904 in FIG. 9 is a process of dividing the temporary allocation area obtained by the process of step S901 into multiple sections and obtaining the primary allocation area of the section number of the allocation request section level. FIG. 8 is a diagram for explaining the flow, and illustrates the processing flow for obtaining the primary allocation area of the allocation request classification level by dividing the provisional allocation area obtained in the process of step S601 of FIG. Corresponds to the figure. Steps S1021 to S1033 in FIG. 10B show that the allocation state of the multi-layer partitioning table indicated by the pair partition number is set to empty in step 811 in FIG. 8A, whereas the pair partition number is set in step S1031. Since only the empty candidate is set in the allocation state of the multi-layered division table to be pointed, the description of FIG. 10B is also omitted.

Next, the processing in step S508 in FIG. 5 will be described in detail with reference to FIGS. 11A to 11C.
FIG. 11A is a diagram for explaining the processing flow of the previous stage to obtain the secondary allocation area of the allocation request classification by dividing the primary allocation area of the acquired classification number into a multilayer area.

First, in step S1101, the leading number pointed to by the classification level (acquisition classification level) of the acquired primary allocation area is extracted from the leading number management table. A value obtained by doubling the result obtained by subtracting the head number is set.

Next, in step S1104, in order to divide the primary allocation area, an allocation configuration classification table composed of bit values in which the allocation request size is represented in binary is created from the allocation request size. In step S1105, a value obtained by subtracting 1 from the acquired division level is set as the division division level.

Next, in step S1106, as the minimum division division level, the bit position having a bit value of 1 in the allocation configuration division table and the lowest bit position when viewed from the lower 0 bit are set, and in step S1107 of FIG. move on. For example, when the bit value of the allocation configuration division table is “1010”, bit position 1 is set to the minimum division division level.

FIG. 11B is a diagram illustrating a subsequent process flow in which the primary allocation area of the acquired classification number is divided into multi-layered areas to obtain the secondary allocation area of the allocation request classification.
In step S1107, the head number indicated by the division level set in step S1105 is extracted from the head number management table.
In step S1109, a value obtained by adding the value set in the in-level number to the extracted top number is set in the division number. In step S1110, a value obtained by adding 1 to the division number is set in the paired division number. To do.

Next, in step S1111, a bit value is extracted from the allocation configuration classification table indicated by the division classification level, and in step S1112, it is determined whether the extracted bit value is 1.
If it is determined in step S1112 that the extracted bit value is not 1 (0), the process branches to step S1113. If it is determined that the extracted bit value is 1, the process proceeds to step S1115.

In step S1113, “used” is set in the allocation state of the multi-layer partitioning table indicated by the section number, and in step S1114, an empty candidate is set in the allocation state of the multi-layer partitioning table specified by the pair of partition numbers, and the process proceeds to step S1119. .
On the other hand, in step S1115, it is determined whether the division division level is the minimum division division level set in step S1106. If the division level is not equal to the minimum division level, the process branches to step S1116. If the division level is equal, the process proceeds to step S1121.

In step S1116, “used” is set in the allocation state of the multi-layer partitioning table indicated by the section number, and in step S1117, “used” is set in the allocation state of the multi-layer partitioning table indicated by the paired section number. In step S1118, the in-level number is incremented by 1, and the process proceeds to step S1119.
In step S1119, the in-level number is doubled, the process proceeds to step S1120, the division division level is decreased by 1, and the process returns to step S1107.

When it is determined in step S1115 that the division division level is equal to the minimum division division level and the process proceeds to step S1121, “in use” is set in the allocation state of the multi-layer division attachment table indicated by the division number. In step S1122, the pair division number is set. An empty candidate is set in the allocation state of the multi-layered division table to be pointed, and the process is terminated.

FIG. 11C illustrates the process of obtaining the secondary allocation area of the allocation request classification by dividing the primary allocation area shown in FIG. 11A and FIG. 11B with reference to the multilayer allocation table 310 shown in FIG. It is a figure to do. In the example illustrated in FIG. 11C, since the value set in the allocation request size 220 and the primary allocation area are the primary allocation area 280 of the division number 3, secondary allocation is performed following the primary allocation illustrated in FIG. 7B. It is.

Similarly to the one shown in FIG. 7B, the bit values of the first and third bits of the allocation request size 220 that represents the allocation request size in binary notation are 1. The allocation configuration division table 230 is composed of 1-bit entries corresponding to the division level as indicated by 0 to 5 below it. In the example of FIG. 11C, the division level 0 is changed to the highest division level 5. It has 6 corresponding entries. As indicated by the dotted

arrows

223 and 221 in the figure, the bit value of the corresponding partition level entry is set according to the bit value when the allocation request size is displayed in binary format. The setting up to this point is performed by the processing of step S501 in FIG. 5 and step S1104 in FIG. 11A.

The secondary allocation is executed based on the bit values in the allocation configuration table 230 described above. In response to a multi-layer division request up to the division level 1 which is the minimum division division level of the primary allocation area indicated by the arrow 244 from the primary allocation area 280, the multi-layer division from the division division table of the division level 4 which is the acquisition division level denoted by reference numeral 164b. Is started. The division assignment table 164b stores the same assignment state as shown in FIG. 7B.

11B, since the bit value corresponding to the division level 3 in the allocation configuration division table 230 is 1, as indicated by the dotted arrow 233, the division request 274c for the division number 3 of the division level 4 is shown. By the processing of step S1116 and step S1117, in the partition level attached table 163, which is the partition pair 293b corresponding to one set of regions that occupy the same region as the region of the partition number 3 in the partition level 3 region “11” indicating in-use is set in the allocation state of both the in-level number 4 (section number 8) and the in-level number 5 (section number 9).

Since the intra-level number is increased by 1 in the process of step S1118 following step S1117 of FIG. 11B, the area of the intra-level number 5 (partition number 9) of the segment level 3 becomes the target of the segment request 273d. As indicated by the dotted arrow 232, the bit value corresponding to the division level 2 of the allocation configuration division table 230 is 0. Therefore, the processing of steps S1113 and S1114 in FIG. In the allocation state of the level number 10 (section number 20) in the section level division table 162, which is the section pair 292b corresponding to a set of sections occupying the same area as the section of section number 9, “11” is set, and “01” indicating an empty candidate is set in the allocation state of the in-level number 11 (section number 21).

Next, the section of the division level 2 with the level number 10 (section number 20) becomes the target of the section request 272c. As indicated by the dotted arrow 231, since the bit value corresponding to the division level 1 of the allocation configuration division table 230 is 1 and the division level 1 is the minimum division division level, the step S 1121 and the step of FIG. By the processing of S1122, the intra-level number 20 in the division level 1 division division table 161, which is the division pair 291b corresponding to a set of areas that occupy the same area as the division number 20 area, which is the division level 1 area. “11” indicating in use is set in the allocation state of (section number 43), and “01” indicating an empty candidate is set in the allocation state of in-level number 21 (section number 44).

This completes the update of the multi-layer division table 310 associated with the secondary allocation. By updating the multi-layer division table 310, the primary allocation area 280 is divided into multiple layers as shown in the primary allocation area 280c after the secondary allocation. As shown by the arrow 273c, the section 283c of the section number 8 of the section level 3 is allocated as in use, and subsequently, the section 281c of the section number 43 of the section level 1 is allocated as shown by the arrow 271c. Allocated as in use and secondary allocated area 290c is allocated as a multilayer area. On the other hand, as shown by the arrow 272d, the section 282d of the section number 21 of the section level 2 is assigned as a blank candidate, and subsequently, the section of the section number 44 of the section level 1 as shown by the arrow 271d. 281d is assigned as an empty candidate, and is also assigned as an adjacent multilayer area 290d. In the secondary allocation area 290c, the areas are allocated in descending order of the partition level, and in the adjacent multi-layer area 290d, the areas are allocated in the ascending order of the partition level which is reverse to the secondary allocation area.

Even when secondary allocation is performed, the classification number acquired in the primary allocation is returned to the file system as the allocation classification number. As is clear from the above description, the area of the classification number acquired in the primary allocation (3 in the above example) Since the start position of the second assignment area is the same as the start position of the second assignment area, it is possible to respond to the address inquiry of the assignment area specifying the section number from the file operation system.
In the above description, secondary allocation is performed from the younger number side of the division number, but as described above for primary allocation after provisional allocation, the fact that it can be performed from the old number side of the segment number is, It will be apparent to those skilled in the art.

Next, with reference to FIGS. 12 to 17C, the release of the allocated area in the embodiment of the present invention will be described. Similar to the allocation of areas, the process of releasing areas by deleting files is also performed by the multi-layer division management unit of the allocation system.

FIG. 12 is a diagram for explaining an example of a schematic process flow of the entire process for releasing the allocated area and attempting to connect the empty areas. By allocating the allocated area and the free area shown in FIG. 12, it is possible to reset the allocation state of the area including the allocated allocation area and reallocate the area.

As shown in the figure, in step S1201, an allocation area number is set as the division number. This allocated area number is returned to the file system when the area to be released is allocated to the file, and can be included in the allocation area release request from the file system. In addition, the allocation request size can be included at that time.

Next, in step S1202, the multi-layered division management table is searched with the set division number, and the division level corresponding to the division number is obtained. Details of the processing in step S1202 will be described later with reference to FIG.
Next, proceeding to step S1203, it is determined whether or not the classification level has been obtained in the process of step S1202, and if not, the process is terminated. If so, the process proceeds to step S1204.

In step S1204, the partition level obtained in the process of step S1202 is set as the release partition level, and in step S1205, a temporary storage area named as a connection request is initially set as a flag for attempting to connect empty areas. .

Next, in step S1206, an attempt is made to release an area included in the primary allocation area indicated by the division number and connect an empty area. If the original allocation request is a single bit request, the processing in step S1206 ends if the allocation state of the division number set in step S1201 is set to empty. If the original allocation request is a multi-bit request, secondary allocation has been performed and not only the secondary allocation area is released, but also the empty candidate area in the adjacent multi-layer area is linked to the released area. An attempt is made to empty the allocation status of the upper division level area. Details of the processing in step S1206 will be described later with reference to FIGS. 14A and 14B.

Next, in step S1207, it is determined whether there is a connection request. If not, the process ends. If there is, an attempt is made in step S1208 to connect the primary allocation area indicated by the section number and the adjacent area adjacent to each other to make the allocation state of the upper area empty. As will be apparent from the following description, there is a connection request and the process proceeds to step S1208 when the primary allocation area is released in the process of step S1206. Details of the processing in step S1208 will be described later with reference to FIGS. 17A, 17B, and 17C.

When the process of step S1208 ends, the process shown in FIG. 12 ends.
FIG. 13 explains the details of the processing in step S1202 of FIG. 12, and is a diagram for explaining an example of a processing flow for searching a multi-layered division management table by division number and obtaining a division level corresponding to the division number. .

First, in step S1301, the total number of divisions is extracted from the total division number management table. In step S1302, it is determined whether the section number is significant and smaller than the total number of sections. Here, as the division number, the allocation division number given from the allocation area release request source in step S1201 of FIG. 12 is set. The processing in step S1302 checks whether the value of the assigned division number is within a valid range, and guarantees the existence of a division level division unit corresponding to the division number for the following processing. .

If the determination result in step S1302 is negative, the corresponding classification level is not returned, and the process ends. If the determination result is affirmative, the process proceeds to step S1303.
In step S1303, the highest partition level of the multi-layer partition management table is set as the initial value for the partition level. In the example shown in FIG. 3, 7 is set as the division level.

Next, in step S1304, the start number indicated by the division level is extracted from the start number management table, and in step S1305, it is determined whether the extracted start number is significant. If the head number is not significant, the process proceeds to step S1306, the division level is decreased by 1, and the process returns to step S1304. The loop processing from step S1304 to step S1306 described above skips the division level determination processing from the highest division level to the division level one level higher than the division level of the actually allocated area.

On the other hand, if the extracted leading number is significant, the process advances to step S1307 to determine whether the division level is the lowest division level. If the division level is the lowest division level, the information of the lowest division level 0 is returned as the division level is present, and the process is terminated. If the partition level is not the lowest partition level, the process proceeds to step S1308, and the start number pointed to by the value obtained by subtracting the partition level from the start number management table is extracted as the next start number. In step S1309, the partition number Determines whether it is less than the next head number. If it is determined that the division number is smaller than the next head number, the division level is returned as the division level is present, and the process is terminated. If it is determined in step S1309 that the segment number is not smaller than the next head number, the process proceeds to step S1310, the segment level is decreased by 1, and the process returns to step S1307.

The above loop processing from step S1307 to step S1310 is repeated while decreasing the partition level by one, and before the partition level becomes the lowest partition level, in the determination of step S1309 of a certain partition level, The partition level when it becomes smaller than the head number of the partition level is the partition level related to the partition unit indicated by the partition number. If the partition level becomes the lowest partition level before the partition number becomes smaller than the head number of the next partition level, the lowest partition level, that is, the partition level 0 becomes the corresponding partition level.

Next, details of the processing in step S1206 in FIG. 12 will be described with reference to FIGS. 14A and 14B.
FIG. 14A is a diagram for explaining an example of a processing flow for attempting to release an area included in the primary allocation area indicated by the division number and connect an empty area.

First, in step S1401, an allocation configuration division table is created from the allocation request size. The allocation request size can be included in the release request of the file system that requested the area release as described above. The allocation configuration division table is created in the same manner as the processing in step S1104 shown in FIG. 11A. Although FIG. 11A shows an allocation request for a multi-bit request, it is clear that an allocation configuration table with a bit value of 1 is formed in the case of a single bit request. .

Next, in step S1402, the bit position where the first bit 1 viewed from the low-order 0th bit of the allocation configuration section table is set at the minimum section level.
In step S1405, the division allocation status in the allocation area is obtained from the allocation configuration classification table and pushed onto the division allocation status stack, and the process proceeds to step S1407. Here, the division allocation status is a division level, a division number, and a release display of an area where the allocation area is divided and assigned. When the bit value of the allocation configuration division table indicated by the division level is 1, the release indication is given, and when the bit value is 0, the release indication is given. Details of the processing in step S1405 will be described later with reference to FIG.

In step S1407, the division allocation status stack is popped, and the division level, division number, and release display are read out.
In step S1409, it is determined whether the read partition level is the release partition level. If the partition level is the release partition level, the process branches to step S1416. If the partition level is not the release partition level, the process proceeds to step S1411. If the original allocation request is a single bit request, the process branches to step S1416 in the determination process of the first step S1409.

In step S1411, it is determined whether the release display read in step S1407 indicates “Yes”. If there is a release display, the process proceeds to step S1413. If there is no release display, the process proceeds to step S1412.
In step S1412, it is determined whether there is a connection request. If there is, the process proceeds to step S1413. If not, the process returns to step S1407.

In step S1413, an attempt is made to release the segment pair indicated by the segment number, and the process returns to step S1407. Details of the processing in step S1413 will be described later with reference to FIG.
The processing after step S1416 in which it is determined that the partition level is equal to the release partition level in the determination of step 1409 described above is a release processing of the area of the release partition level.

In step S1416, it is determined whether there is a connection request. If there is no connection request, the process is terminated. That is, even in the case of a multi-bit request, if there is a connection request, the allocation state for the primary allocation area composed of the secondary allocation area and the adjacent multi-layer area is set to be empty by the process of step S1417.

FIG. 14B refers to the multi-layer partition allocation table 310 shown in FIG. 3 and, according to a specific example, a process of attempting to release an area included in the primary allocation area shown in FIG. 14A and FIG. It is a figure explaining.
The example shown in FIG. 14B releases the secondary allocation area secondary allocated by the multi-bit request similar to that illustrated in FIG. 11C, and concatenates with the adjacent area to empty the allocation status of the upper division level area. It is to set. Therefore, the values set in the allocation request size 220 and the allocation configuration classification table 230 are the same as those shown in FIG. 11C.
The primary allocation area 280c after the secondary allocation shown in FIG. 14B is the allocation state before the secondary allocation area is released and connected, and is an empty candidate for the primary allocation area 280c after the secondary allocation shown in FIG. 11C. The area 282d of the section number 21 that is present is in use. The allocation state of the area 281d of the division number 44 remains an empty candidate.

The release and concatenation processes are sequentially performed from the partition level 1 that is the minimum partition level to the higher partition level while referring to the bit values in the allocation configuration partition table 230.
First, in the division level 1 which is the minimum division level, the corresponding bit value of the allocation configuration division table 230 is 1 (corresponding is indicated by a dotted arrow 231a). That is, since there is a release display, step S1413 in FIG. 14A, that is, the process shown in FIG. 16 is executed. As indicated by the partition level 1 release request arrow 251a, the release of the zone 281c is guaranteed, and the value of the partition number 171 in the partition level 1 partitioning table before the release indicated by the reference numeral 161c is 43. Since the assignment state of the area of the division number 44 that is paired with the area of the level number 172 of 20 is an empty candidate, the determination in step S1605 in FIG. It is assumed that there is a connection request (indicated by an arrow 241c) for setting the allocation state of the division level 1 division pair 291c consisting of the division of the division number 43 and the division number 44 division to be empty, indicated by an arrow 241c. As shown in the division level attached table of division level 1 after release, the allocation state of the division number 43 and the paired division number 44 is set to hold. In the case of the initial setting of the multi-layered division table 310 shown in FIG. 3, when the sky of the upper division level area is set, setting the hold to the allocation state of the lower division level area included in the area It is the same.

In the division level 2, the corresponding bit value of the allocation configuration division table 230 is 0 (indicating the correspondence by the dotted arrow 232a), and since there is a connection request 241c, it is the upper division area as shown by the arrow 241d. , A request for releasing the area of the division number 20 in the division level attached table of the division level 2 before the release with the reference numeral 162c is made. Since the area of the division number 20 and the area of the division number 21 constituting the division pair 292c are in use, the determination in step S1605 of FIG. 16 is negative, and only the area of the division number 20 is shown as indicated by the arrow 242c. The allocation state of the partition pair 292c becomes empty and in use, as shown in the partition level 2 partition allocation table after release with reference numeral 162d. Also, there is no connection request. Therefore, it is the area above the division pair 292c as indicated by the dotted arrow 242d, and the area of the division number 9 in which the allocation state is in use as indicated in the division level 3 divisional table before release, which is denoted by reference numeral 163c. Will not be released.

Next, at the division level 3, the corresponding bit value of the allocation configuration division table 230 is 1 (indicated by the dotted arrow 233a), and the area 283c is indicated by the release level arrow 253a of the division level 3 as shown. To be released. However, as described above, the area of division number 9 whose allocation state is in use is not released and is in use. Therefore, as indicated by the arrow 243c, only the area of the division number 9 and the area of the division number 8 constituting the division pair 293c are released, and the allocation state of the division pair 293c is the level of the division level 3 after the release with the reference numeral 163d. It becomes empty and in use as shown in the Warranty Schedule. In addition, since there is no connection request, it is an upper area of the division pair 293c as shown by the dotted arrow 244d, and the allocation status is used as shown in the division division table of division level 4 before release with reference numeral 164. The area of partition number 3 in the inside is not released, and the allocation state remains in use.

The primary allocation area 280d after the secondary allocation area is released by the above-described release and connection processing corresponds to the allocation state of the multi-layer division allocation table 310 by

arrows

273e and 272e and dotted arrow 272f. As shown, the section number 8 is an empty area 283c, the section number 20 is an empty area 282c, and the section number 21 is an in-use area 282d.

FIG. 15 is a diagram for explaining the details of the processing in step S1405 in FIG. 14A. FIG. 15 is a diagram for explaining an example of a processing flow for obtaining the split allocation status in the allocation area and pushing it to the split allocation status stack using the allocation configuration classification table. It is.

First, in step S1501, the release partition level set in step S1204 shown in FIG. 12 is set as the partition level. Next, in step S1501a, the leading number indicated by the partition level is extracted from the leading number management table, and in step S1501b. Then, a value obtained by subtracting the head number of the read management information from the division number is set in the in-level number, and the process proceeds to step S1502.

In step S1502, the bit value pointed to by the division level is extracted from the allocation configuration division table as a release display. If the bit value is 1, the display is released, and if the bit value is 0, the release is not displayed.
In step S1502a, the division level, the division number, and the release display are pushed onto the division allocation status stack as the division allocation status in the allocation area.

In step S1504, it is determined whether the partition level is equal to the minimum partition level. If it is determined that the segment level is higher than the minimum segment level, the process branches to step S1504a to determine whether there is a release display. If there is no release display, the process proceeds to step S1506. In step S1505, 1 is added to the in-level number, and the flow advances to step S1506.
In step S1506, the in-level number is doubled, and in step S1507, the division level is decreased by one. In step S1507a, the head number indicated by the section level is extracted from the head number management table. In step S1507b, a value obtained by adding the in-level number to the head number is set in the section number, and the process returns to step S1502.

The above-described loop processing from step S1502 to step S1507b is repeated until it is determined in step S1504 that the division level is equal to the minimum division level. If it is determined in step S1504 that the segment level is equal to the minimum segment level, the process ends.

In the example of FIG. 14B, the division number is 3, the division level is 4, and the intra-level number is 2 and the bit value indicated by the division level 4 in the allocation configuration division table is 0. The partition level, the partition number, and the release display pushed to the allocation status stack are 4, 3, and 0 (no release display), respectively.

In the processing of the division level 4, since the release display is not performed, the in-level number is updated to 2 × 2 = 4. The division level is updated to 3, and the division number is updated to “8” which is the sum of the start number “4” in the start number management table indicated by the division level 3 and the in-level number “4”. The bit value indicated by the division level 3 in the allocation configuration division table is 1. Therefore, 3, 8, and 1 (with release indication) are pushed as the division level, the division number, and the release indication on the division allocation status stack, respectively.

In the processing of the division level 3, since there is a release display, the in-level number is updated to (4 + 1) × 2 = 10. The division level is updated to 2, and the division number is updated to “20” which is the sum of the start number “10” of the start number management table indicated by the division level 2 and the in-level number “10”. In addition, the bit value indicated by the division level 2 in the allocation configuration division table is 0. Therefore, 2, 20, and 0 (no release indication) are pushed as the division level, the division number, and the release indication on the division allocation status stack, respectively.

In the processing of the division level 2, there is no release display, so the in-level number is updated to 10 × 2 = 20. The division level is updated to 1 which is the minimum division level, and the division number is updated to “43” which is the sum of the start number “23” of the start number management table indicated by the division level 1 and the in-level number “20”. The In addition, the bit value indicated by the division level 1 in the allocation configuration division table is 1. Therefore, 1, 43 and 1 (with release indication) are pushed as the division level, the division number, and the release indication, respectively, on the split allocation status stack.

In the example shown in FIG. 14B, the area allocated by the multi-bit request is released. If the area allocated by a single bit request is to be released, the release partition level and the minimum partition level are the same, and only one set of partition level, partition number, and release display is pushed to the partition allocation status stack, and the release display There is.

The division level, division number, and release display as the division allocation status pushed to the above-described division allocation status stack, and the connection that is initially set to “Yes” in step S1205 of FIG. 12 and updated by the process shown in FIG. According to the request, the allocation area illustrated in FIG. 14B is released.

FIG. 16 is a diagram for explaining details of the processing in step S1413 in FIG. 14A, and is a diagram for explaining an example of a processing flow for attempting to release a pair of segments indicated by a segment number. As shown in the processing flow of FIG. 14A, the processing steps shown in FIG. 16 are executed for each partition level from the minimum partition level to the partition level one level lower than the release partition level by sequentially popping the split allocation status stack. The

As shown in the figure, first, in step S1602, a value obtained by adding the value 1 to the partition number is set in the pair partition number, and the process proceeds to step S1604, where the allocation state of the multi-layer partition table attached by the pair partition number is read, In step S1605, it is determined whether the read allocation state is empty or empty candidate. If the read assignment state is empty or empty candidate, the process proceeds to step S1606. If the read assignment state is neither empty nor empty candidate, the process branches to step S1609.

In step S1606, a hold is set in the allocation state of the multi-layer partition table attached by the category number, and in step S1607, a hold is set in the allocation state of the multi-layer partition table attached by the pair category number. In step S1608, the presence of the connection request is set, and the process ends. The processes in steps S1606 to S1608 correspond to the process with a connection request indicated by the arrow 241c in the example of FIG. 14B.

On the other hand, when branching to step S1609, empty is set in the allocation state of the multi-layer partitioning table indicated by the partition number, and in step S1610, nothing is set in the connection request, and the process ends. In the example of FIG. 14B, this process corresponds to the process of releasing only the own area indicated by the arrows 242c and 243c.

Next, details of the processing in step S1208 in FIG. 12 will be described with reference to FIGS. 17A, 17B, and 17C.
FIG. 17A is a diagram illustrating an example of a processing flow in which the primary allocation area and an adjacent empty area are connected to try to make the allocation state of the upper area empty.

As shown in the figure, in step S1701, the release partition level is set as the partition level, and the flow advances to step S1702.
In step S1702, the start number indicated by the division level is extracted from the start number management table. In step S1703a, the value obtained by adding the in-level number to the extracted start number is set in step S1703a, and the process proceeds to step S1703b. In the first process of step S1703a, that is, the process at the release section level, the in-level number is set by step S1501b of the process flow of FIG. 15 showing details of the process of step S1405 shown in FIG. 14A.

In step S1703b, the main division number indicated by the division level is extracted from the main division number management table, and the process proceeds to step S1705.
In step S1705, it is determined whether the division number is equal to the main division number. If equal, the process branches to step S1710, and if not equal, the process proceeds to step S1706.
In step S1706, an attempt is made to release the pair of segments indicated by the segment number. Details of the processing in step S1706 will be described later with reference to FIG. 17C.

In step S1707, it is determined whether there is a connection request. If there is no connection request, it means that it is determined in step S1605 in FIG. 16 that the assignment state of the paired segment number is neither empty nor empty candidate. If there is, the process branches to the processing of step S1708 and the subsequent steps, and the connection processing of the upper division level area is attempted.
In step S1708, the division level is set one level higher, the process proceeds to step S1709, the quotient obtained by dividing the in-level number by the value 2 is set in the in-level number, and the process returns to step S1702.

In step S1710, which branches when it is determined in step S1705 that the division number and the main division number are equal, the allocation state of the multi-layer division table attached by the division number is set to be empty, and the process ends. When the section number is equal to the main section number, there is no section of the section number that is paired with the section number as is clear from the illustration of FIG. The allocation state is set to empty and the process ends.

FIG. 17B refers to the multi-layer division table 310 shown in FIG. 3 and connects the primary allocation area shown in FIG. 17A and FIG. 17C and the adjacent empty area to make the allocation state of the upper area empty. It is a figure explaining the process which tries this by a specific example. In the following description, connecting the primary allocation area and the adjacent empty area to make the allocation state of the upper area empty may be simply referred to as connection.

In the example shown in FIG. 17B, since the area allocated primarily by the single bit request similar to that illustrated in FIG. 8B is released, the allocation state of the upper division level area is set empty by concatenating with the adjacent area. It is.
In FIG. 17B, the area of the division number 31 that was in use of the temporary allocation area 280b after the primary allocation shown in FIG. 8B is released as the multilayer area in the allocated state before the released primary allocation area is connected, A multi-layered area 280e is illustrated in which the area 283b of the division number 7 that was empty is in use. The allocation state of the area 281b with the division number 32 and the area 282b with the division number 15 remains empty.

The concatenation process is sequentially performed from the segment level 1 that is the release segment level to the higher segment level where the concatenation process is no longer possible. That is, the areas that are adjacent to the area 281a of the section number 31 that has been released and that are not in use are connected.

First, since the determination in step S1207 shown in FIG. 12 is affirmative, a connection request indicated by an arrow 240e is made to the released area 281a, and the processing of the division level 1 that is the release division level is performed. Is called. In the division level 1 division table attached with the reference numeral 161e, the assignment state of the division number 32 area that is paired with the division number 171 value 31 and the in-level number 172 value 8 is empty. , There is a connection request (indicated by arrow 241e) for setting the allocation state of the division level area higher than the division pair 291d consisting of the division number 31 area and the division number 32 area, and is denoted by reference numeral 161f. As shown in the division level attached table of division level 1, the allocation state of division number 31 and paired division number 32 is set to hold. As described in the description of FIG. 14B, when the upper section level area is set to be empty, setting the holding state to the allocation state of the lower section level area included in the area is shown in FIG. This is the same as the case of the initial setting of the multi-layer division table 310 shown.

In division level 2, since there is a connection request, a release request for the area of division number 14 in the division level table of division level 2 before connection, which is the upper division area as indicated by arrow 241f, and is attached with reference numeral 162e. Is done. Since the area of the division number 14 and the area of the division number 15 constituting the division pair 292d are empty, there is a connection request as shown by the arrow 242e. As shown, the allocation is set to the allocation state of the segment number 14 and the pair segment number 15.

Next, at division level 3, since there is a connection request, as shown by the arrow 242f, the area of the upper division, that is, the area of division number 6 in the division level attachment table of division level 2 before connection, which is denoted by reference numeral 163e, is released. A request is made. Since the area of the division number 6 and the area of the division number 7 constituting the division pair 293d are in use, only the allocation state of the area of the division number 6 is set to empty as shown by the arrow 243e, and the allocation state of the division pair 293d is , Empty and in use as shown in the division level attached table of division level 3 after concatenation with reference numeral 163f.
Since there is no connection request, the area of the division number 2 that is an upper area as shown by the dotted arrow 243f and the allocation state is in use as shown in the division division table 164 of the division level 4 is released. Instead, the assigned state remains in use.

The multi-layer area 280f in the allocated state after the primary allocation area is connected to the adjacent empty area by the above processing of releasing and connecting the area is changed to the allocation state of the multi-layer division table 310 by the arrow 273e and the dotted arrow 272f. As shown in the correspondence, the section number is 6 and the area is empty 283e, and the section number is 7 and the area is used 283f.

FIG. 17C explains the details of the processing in step S1706 of FIG. 17A and is a diagram for explaining an example of a processing flow for attempting to release the segment pair indicated by the segment number. Here, the category pair indicated by the category number means the category pair to which the category unit identified by the category number belongs, and as will be described later, the category number is not necessarily the younger number. In other words, the in-level number corresponding to the division number is not always an even number.
As shown in the processing flow of FIG. 17A, the processing step shown in FIG. 17C is executed for each partition level in the direction from the release partition level to the higher partition level, and attempts to release the upper partition level region by concatenating empty regions. It is done.

As shown in the figure, first, in step S1711, it is determined whether the in-level number is an even number. The process flow shown in FIG. 17C is similar to that shown in FIG. 16, but as described above, in the process of FIG. 17A, that is, step S1208 of FIG. 12, the in-level number is not always an even number. Absent. This is because the in-level number corresponding to the primary allocation area that has been released and becomes empty may be an even number or an odd number. In the example shown in FIG. 17B, 8 is set, but when the area 283f with the assigned allocation division number 7 shown in FIG. 17B becomes empty, the in-level number is 3, which is an odd number.

If the in-level number is an even number, in step S1712, the value obtained by adding the value 1 to the division number is set in the pair division number, and the process proceeds to step S1714. If the number in the level is an odd number, the pair division number is entered in step S1713. In step S1714, a value obtained by subtracting 1 from the division number is set.

In step S1714, the allocation state of the multi-layer partition allocation table indicated by the paired segment number is read out, and in step S1715, it is determined whether the read allocation state is empty or empty candidate. If the read assignment state is empty or empty candidate, the process proceeds to step S1716. If the read assignment state is neither empty nor empty candidate, the process branches to step S1719.

In step S1716, the allocation state of the multi-layer partitioning table indicated by the division number is set to hold, and in step S1717, holding is set as the allocation state of the multi-layer partitioning table indicated by the pair of division numbers. In step S1718, the group pair connection request is set to “Yes”, and the process ends. The processes in steps S1716 to S1718 correspond to the connection request processes indicated by arrows 241e and 242e in the illustration of FIG. 17B.

On the other hand, if the process branches to step S1719, the allocation state of the multi-layer partitioning table indicated by the partition number is set to be empty, and in step S1720, none is set in the connection request, and the process ends. This process corresponds to the process of releasing only the own area indicated by the arrow 243e in the example of FIG. 17B.

It is obvious that the partition management device according to the present invention can be constructed on a computer by a program that causes a computer such as the data processing device 301 illustrated in FIG. 2C to execute the partition management method of the present invention.
A functional block configuration example relating to the partition management apparatus of the present invention will be described below.

FIG. 18A is a diagram illustrating a functional block configuration example of a partition management device according to an embodiment of the present invention.
As shown in the figure, the partition management device 800 is roughly composed of an initialization unit 810 and a multilayer section management unit 840. The initialization unit 810 and the multilayer partition management unit 840 correspond to the allocation system (initialization unit) 101 and the allocation system (multilayer partition management unit) 102 illustrated in FIG. 2B.

The initialization unit 810 includes a section size acquisition unit 820 that acquires a section size, and a multi-layer partitioning table generation unit 830. The multi-layer partitioning table generation means 830 represents the size of each partition as the product of the sum of powers of two different from each other and the size of the allocation unit of the partition, and the size of each of the powers of 2 constituting the sum and the allocation unit of the partition The main area of the product size is divided into sections in the order of the size, and the sections are divided, and each main area is divided by half, and each size is sequentially divided into the allocation unit size. A multi-layer division table that stores allocation information indicating the allocation status of each area corresponding to each area including the main area is generated and initialized. The function of the multi-layer partitioning table generation unit 830 can be realized by the processing flow example described with reference to FIGS. 4A to 4C.

The multi-layer division management unit 840 includes an area allocation unit 850 that allocates a file or a memory area to an empty area, and an area release unit 860 that releases the allocation area from the allocation of the file or memory area.

FIG. 18B is a diagram illustrating an example of a functional block configuration of the area allocating unit according to the embodiment of the present invention. As shown in the figure, the area allocation unit 850 includes an allocation request reception unit 851 that receives an allocation request, an empty area search unit 852, a secondary allocation unit 853, and an allocation category number output unit. Allocating means 857 is included.

The function of the area allocating unit 850 can be realized by the processing flow example described with reference to FIG.
In the empty area search means 852, the allocation request size, which is the size included in the allocation request, is represented by the product of the sum of powers of two different from each other and the size of the allocation unit of the partition, and is the sum of the sizes of the areas of different partition levels. In some cases, a search is made for a primary allocation area that is an empty area of a partition level that is one larger than the maximum of the partition level and is larger than the allocation request size, and the allocation request size that is the size included in the allocation request is 2 When it is represented by the product of the power and the size of the allocation unit of the section, an empty area of the allocation request size is searched as a primary allocation area. The function of the empty area searching means 852 corresponds to the processing flow illustrated in FIGS.

The temporary allocation means 857 of the empty area search means 852 refers to the allocation status corresponding to the upper division level area of the multi-layer division division table when the empty area cannot be searched at the division level of the primary allocation area. An empty area at a higher division level is searched, and the allocation state of the multi-layer division allocation table is set to be in use with the area as a temporary allocation area, and the temporary allocation area is defined as the primary allocation area and the remaining areas at the division level. Divide the areas with different division levels in order from the smallest one into adjacent multi-layer areas that are continuously assigned, set the assignment status of the multi-layer division table corresponding to the areas in the primary allocation area to be in use, An empty state is set in the allocation state of the multi-layered division table corresponding to the area constituting the adjacent multi-layered area. The function of the temporary allocation unit 857 corresponds to the processing flow illustrated in FIGS. 7A and 8A, and FIGS. 10A and 10B.

When the size of the primary allocation area is larger than the required allocation size, the secondary allocation means 853 is a secondary allocation in which the primary allocation area is an area in which areas having different division levels are sequentially assigned in descending order of the division level. Divide into areas and adjacent multi-layer areas where areas with different division levels are assigned sequentially from the remaining areas, starting with the lowest division level, and correspond to each of the secondary allocation areas In-use is set in the allocation state of the multi-layer partitioning table, and an empty candidate state is set in the allocation state of the multi-layer partitioning table corresponding to the area constituting the adjacent multi-layer area. The function of the secondary allocation unit 853 corresponds to the processing flow illustrated in FIGS. 11A and 11B.

FIG. 18C is a diagram illustrating an example of a functional block configuration of the area releasing unit according to the embodiment of the present invention. As shown in the figure, the area release means 860 includes a release request reception means 851 that receives a release request, a primary allocation area release means 862 that attempts to release an area in the primary allocation area, and a partition level that is higher than the partition level of the primary allocation area. Higher-level area release means 863 that attempts to release the area.

The function of the area releasing means 860 corresponds to the processing flow example illustrated in FIG.
When the release request size, which is the size included in the release request, is the sum of the areas having different partition levels, the primary allocation area release means 862 obtains the partition number of the smallest area in the adjacent multi-layer area and determines its partition number. If the allocation state is read from the multi-layer division allocation table pointed to by and the read allocation state is empty or empty candidate state, the multi-layer division corresponding to the minimum area in the secondary allocation area and the minimum area in the adjacent multi-layer area Attempts to release the area at the next higher partition level by setting the allocation state in the contingency table, and if the allocation status of the smallest area in the adjacent multi-layer area is in use, the lowest in the secondary allocation area The area assignment state is set to empty. The function of the primary allocation area release means 862 corresponds to the processing flow illustrated in FIG. 14A.

The upper area release means 863 has the same division level as the primary allocation area when the primary allocation area is released and the allocation state of the multi-layered division allocation table indicated by the division number is empty, and is one level higher. Read the allocation status of the other area from the multi-level allocation table when one area is divided into two areas as the primary allocation area, and the allocation status of the other area is empty or empty candidate If so, it sets a hold in the allocation state of the multi-layer partitioning table corresponding to the one area and the other area, tries to release the area of the higher division level, and the allocation state of the other area is in use If there is, the allocation state of one area is made empty. The function of the upper zone release means 863 corresponds to the processing flow illustrated in FIG. 17A.

Although the embodiment for carrying out the present invention has been described in detail above, it is obvious to those skilled in the art that the embodiment of the present invention is not limited to the above and can be variously modified. As described above, according to the present invention, it is possible to effectively and efficiently manage the storage device regardless of the storage capacity. In addition, it is possible to perform file allocation to continuous areas by managing each division level area in a multi-layer manner using a multi-layer division table and a multi-layer division management table.

100 Allocation system 114 Main section number 115 Number of sections 116 First number 120 Section size 130 Section configuration main section table 170 Allocation state 171 Section number 172 Level number 200 Application program / OS
DESCRIPTION OF SYMBOLS 201 Initialization program 202 File system 203 File operation system 220 Allocation request size 230 Allocation configuration division table 301 Data processing unit 302 Central processing unit 303 Cache memory 304 Bus 305 Main storage unit 306 External storage unit 307 Communication unit 308 Data storage unit 309 Multi-layer Division management table 309a Main division number management table 309b First number management table 310 Multi-layer division partitioning table 311 Partition 600 Allocation bitmap 690 Partition 790 Tree structure group 800 Partition management device 810 Initialization unit 820 Partition size acquisition means 830 Multi-layer partition partitioning table Generation means 840 Multi-layer division management unit 850 Area assignment means 851 Assignment request reception means 852 Empty area search means 853 Secondary assignment means 854 Assignment division number output means 8 57 Temporary allocation means 860 Area release means 861 Release request reception means 862 Primary allocation area release means 863 Higher area release means

Claims

In the partition management device of the storage device,
A section size acquisition means for acquiring a section size;
When the size of the partition is expressed by the product of the sum of powers of two different from each other and the size of the allocation unit of the partition, the size of the product of each power of 2 constituting the sum and the size of the allocation unit of the partition The main area is divided by dividing the sections in the order of their sizes, and each main area is divided by half and sequentially divided into areas of each size up to the size of the allocation unit of the section. A multi-layer partitioning table generating means for generating and initializing a multi-layered partitioning table for storing allocation information indicating the allocation status of each of the areas corresponding to each of the areas including the main area;
An initialization unit including
A multi-layer division management unit that manages the allocation of each area based on the allocation information stored in the multi-layer division table.
A partition management device comprising:
The partition management device according to claim 1,
When the power of 2 that defines the size of the area is a division level of the area,
In the multi-layer division table, the allocation information of the areas is stored in the order of the division levels of the areas, in the same division level, in the arrangement order of the areas in the divisions,
The multi-layer division management unit allocates a division number that is an identification number for identifying an area to which the allocation information corresponds according to a storage order of the allocation information, and manages the allocation of the area using the division number.
A partition management device.
The partition management device according to claim 2,
The multi-layer division division table generating means sets an empty state indicating that it is empty as an initial value of the allocation information of the main area, and sets the area as an initial value of allocation information of the area obtained by dividing the main area A partition management device that sets a non-empty state indicating that no assignment is made.
The partition management device according to claim 3, wherein
The multi-layer division management unit
An allocation request receiving means for receiving an allocation request for a file or memory area including a size;
With reference to the multi-layer partition management table, the allocation request size, which is the size included in the allocation request, is represented by the product of the sum of different powers of 2 and the size of the allocation unit of the partition, and the partition level is different If it is the sum of the sizes of the areas, a search is made for a primary allocation area that is an empty area of a partition level that is one larger than the maximum of the partition level and is larger than the allocation request size, and corresponds to the primary allocation area If the allocation request number, which is the size included in the allocation request, is represented by the product of a power of 2 and the size of the allocation unit of the partition, an empty area of the allocation request size is obtained. An empty area search means for searching as a primary allocation area, finding a division number corresponding to the primary allocation area and making it an allocation division number;
When the size of the primary allocation area is larger than the allocation request size, the primary allocation area is divided into a secondary allocation area that is an area in which areas of different partition levels are sequentially allocated in the order of the partition level, and the remaining area. The second allocation area is divided into adjacent multi-layer areas, which are areas in which areas having different division levels are sequentially assigned in the reverse order of the division level in which the areas having different division levels are assigned. The division number corresponding to each area of the next allocation area is obtained, the in-use state is set in the allocation state of the multilayer division allocation table corresponding to the division number, and the division number corresponding to the area constituting the adjacent multilayer area is set. Secondary allocation means for setting an empty candidate state to the allocation state of the multilayer partition allocation table corresponding to the classification number
An allocation category number output means for outputting the allocation category number as one of the allocation results of the allocation request for the file or memory area;
With
The division level of the smallest area in the secondary allocation area and the smallest area in the adjacent multi-layer area are equal, and the division level area one level higher than the division level in the multi-layer division allocation table is divided into two. A partition management device.
The partition management device according to claim 4, wherein
The vacant area searching means refers to an allocation state corresponding to an area of a higher division level in the multi-layer division division table when an empty area cannot be searched at the division level of the primary allocation area. An empty area is searched for, and a division number of the area is obtained, and the temporary allocation area that is the area of the division number is continuously arranged from the primary allocation area and the remaining areas in the order of the division level. Is divided into adjacent multi-layer areas that are allocated areas, and a division number corresponding to the area in the primary allocation area is obtained, and the in-use state is set in the multi-layer division allocation table corresponding to the division number And a partition management means for determining a partition number corresponding to the section constituting the adjacent multilayer section, and setting a temporary allocation means for setting an empty state in the allocation state of the multilayer partition allocation table corresponding to the section number apparatus
The partition management device according to claim 5, wherein
The empty area searching means searches for an empty candidate state area from the multi-layer area partitioning table when an empty area in which the file or memory area requested for allocation cannot be searched. Management device.
The partition management device according to claim 6, wherein
The multi-layer division management unit
A release request receiving means for receiving a release request for a file or memory area including a size and an allocation classification number;
When the release request size, which is the size included in the release request, is the sum of the different areas of the division level, the division number of the smallest area in the adjacent multilayer area is obtained and the multilayer division indicated by the division number is obtained. If the allocation state is read from the attached table, and the read allocation state is empty or empty candidate state, the multilayer partition allocation table corresponding to the minimum area in the secondary allocation area and the minimum area in the adjacent multilayer area If the allocation state is set to a state other than empty or an empty candidate, and an attempt is made to release the area of the next higher partition level, and if the allocation state of the smallest area in the adjacent multi-layer area is in use, the second The primary allocation area release means that sets the allocation status of the smallest area in the next allocation area to an empty state and the primary allocation area are released, and the allocation status of the multi-layer division allocation table pointed to by the division number becomes empty became The allocation state of the other area when the first allocation area is the same division level as the primary allocation area, and one area when the one higher division level area is divided into two. If it is read from the multi-layer partitioning table and the allocation state of the other area is empty or empty candidate, the empty or empty candidate is assigned to the allocation state of the multi-layer partitioning table corresponding to the one area and the other area A higher-level area release means that attempts to release the higher-level division level area and sets the allocation status of the one area to an empty state if the allocation status of the other area is in use. When,
A partition management device comprising:
In the storage device partition management method,
A section size acquisition step for acquiring a section size;
When the size of the partition is expressed by the product of the sum of powers of two different from each other and the size of the allocation unit of the partition, the size of the product of each power of 2 constituting the sum and the size of the allocation unit of the partition A main partitioning step of partitioning the partition by dividing the partition in order of size by a main region;
Each main area is divided by half and sequentially divided into areas of each size up to the size of the allocation unit of the partition, and each allocation state of each area is assigned to each area including the main area. A multi-layer partitioning table generating step for generating a multi-layer partitioning table that stores allocation information to be indicated;
A multi-layer division management step for managing the allocation of each area based on the allocation information stored in the multi-layer division table.
And a partition management method.
The partition management method according to claim 8, wherein
When the power of 2 that defines the size of the area is a division level of the area,
The multi-layered division table stores the allocation information of the areas in the order of division levels of the areas, in the same division level, in the arrangement order of the areas in the divisions,
The multi-layer division management step allocates a division number that is an identification number for identifying an area to which the assignment information corresponds according to a storage order of the assignment information, and manages the assignment of the area using the division number.
A partition management method characterized by the above.
The partition management method according to claim 9, wherein
In the multi-layer division division table generation step, an empty state indicating that the main area is empty is set as an initial value of the allocation information of the main area, and the initial value of the allocation information of the area obtained by dividing the main area is set to the area Set a non-empty state to indicate that no assignments will be made,
A partition management method characterized by the above.
The partition management method according to claim 10, wherein
The multilayer section management step includes:
An allocation request reception step for receiving an allocation request for a file or memory area including a size;
With reference to the multi-layer partition management table, the allocation request size, which is the size included in the allocation request, is represented by the product of the sum of different powers of 2 and the size of the allocation unit of the partition, and the partition level is different If it is the sum of the sizes of the areas, a search is made for a primary allocation area that is an empty area of a partition level that is one larger than the maximum of the partition level and is larger than the allocation request size, and corresponds to the primary allocation area If the allocation request number, which is the size included in the allocation request, is represented by the product of a power of 2 and the size of the allocation unit of the partition, an empty area of the allocation request size is obtained. An empty area search step for searching as a primary allocation area, obtaining a division number corresponding to the primary allocation area, and setting it as an allocation division number;
When the size of the primary allocation area is larger than the allocation request size, the primary allocation area is divided into a secondary allocation area that is an area in which areas of different partition levels are sequentially allocated in the order of the partition level, and the remaining area. The second allocation area is divided into adjacent multi-layer areas, which are areas in which areas having different division levels are sequentially assigned in the reverse order of the division level in which the areas having different division levels are assigned. The division number corresponding to each area of the next allocation area is obtained, the in-use state is set in the allocation state of the multilayer division allocation table corresponding to the division number, and the division number corresponding to the area constituting the adjacent multilayer area is set. A secondary allocation step for obtaining an empty candidate state in the allocation state of the multilayer partition allocation table corresponding to the segment number,
An allocation segment number output step for outputting the allocation segment number as one of the allocation results of the allocation request for the file or memory area;
With
The division level of the smallest area in the secondary allocation area and the smallest area in the adjacent multi-layer area are equal, and the division level area one level higher than the division level in the multi-layer division allocation table is divided into two. A partition management method characterized by the above.
The partition management method according to claim 11,
In the empty area searching step, when an empty area cannot be searched at the division level of the primary allocation area, the higher division level is referred to by referring to the allocation state corresponding to the upper division level area of the multi-layer division allocation table. An empty area is searched for, and a division number of the area is obtained, and the temporary allocation area that is the area of the division number is continuously arranged from the primary allocation area and the remaining areas in the order of the division level. Is divided into adjacent multi-layer areas that are allocated areas, and a division number corresponding to the area in the primary allocation area is obtained, and the in-use state is set in the multi-layer division allocation table corresponding to the division number And a provisional assignment step of obtaining a partition number corresponding to the section constituting the adjacent multilayer section and setting an empty state in the layout state of the multilayer partition allocation table corresponding to the section number. Management method.
The partition management method according to claim 12, wherein
The empty area searching step searches for an empty candidate state area from the multi-layer area partitioning table when an empty area in which a file or a memory area requested to be allocated cannot be searched. Management method.
The partition management method according to claim 13,
The multilayer section management step includes:
A release request receiving step for receiving a release request for a file or memory area including a size and an allocation classification number;
When the release request size, which is the size included in the release request, is the sum of the different areas of the division level, the division number of the smallest area in the adjacent multilayer area is obtained and the multilayer division indicated by the division number is obtained. If the allocation state is read from the attached table, and the read allocation state is empty or empty candidate state, the multilayer partition allocation table corresponding to the minimum area in the secondary allocation area and the minimum area in the adjacent multilayer area If the allocation state is set to a state other than empty or an empty candidate, and an attempt is made to release the area of the next higher partition level, and if the allocation state of the smallest area in the adjacent multi-layer area is in use, the second The primary allocation area release step that sets the allocation status of the smallest area in the next allocation area to an empty state and the primary allocation area are released, and the allocation status of the multi-layer division allocation table pointed to by the division number becomes empty Na The allocation state of the other area when the first allocation area is the same division level as the primary allocation area, and one area when the area of the upper division level is divided into two Is read from the multi-layer partitioning table, and if the allocation state of the other area is empty or empty, the allocation state of the multi-layer partitioning table corresponding to the one area and the other area is empty or empty. Set a state other than the candidate, try to release the upper division level area, and if the other area allocation state is in use, release the upper area to make the allocation state of the one area empty Steps,
A partition management method comprising:
A program for causing a computer to execute the method according to any one of claims 8 to 14.
A computer-readable recording medium having recorded thereon a program that causes a computer to execute the method according to any one of claims 8 to 14.