JP2012098931A - Information processor, information processing method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To effectively utilize a cache buffer as a limited resource by preventing data that has been stored in the cache buffer from being expelled by data having low reusability.SOLUTION: An information processor includes: distribution means for performing distribution such that when it is determined that such possibility that used data may be reused within a predetermined time is equal to or more than a predetermined value, a cache buffer including the used data is managed by first management means, and that when it is determined that such possibility is less than the predetermined value, the cache buffer including the used data is managed by second management means; and selection means for, when a cache buffer for newly reading data is not sufficient, selecting the cache buffer managed by the second management means as a cache buffer for abandoning data.

Description

  The present invention relates to an information processing apparatus, an information processing method, and a program.

The file system normally uses a storage area called a cache buffer. When there is a file read request from the application, the CPU performs the following processing. In other words, the CPU transfers the data of the corresponding file from the secondary storage device such as a hard disk or a memory card to the cache buffer in units of sectors, and the next requested data is transferred from the cache buffer to the memory area specified by the application. Forward. The contents of the cache buffer are retained as much as possible, and when the application requests to read the same data again, the CPU skips the transfer from the secondary storage device and transfers the data from the cache buffer to the memory area specified by the application. . As a result, access to the secondary storage device that operates slowly can be reduced, which is effective in improving the speed.
When there is a write request to the file from the application, the CPU transfers the data designated by the application to the cache buffer, and then transfers it to the secondary storage device. The actual data transfer from the cache buffer to the secondary storage device can be delayed. If there is an overwrite request for the same data from the application before transferring to the secondary storage device, the CPU changes only the data in the cache buffer and transfers only the last change result to the secondary storage device. Access to the next storage device can be reduced.
Further, the file system has metadata such as directory entry, FAT, and inode necessary for accessing the file. When reading or writing metadata, the CPU once transfers the metadata to the cache buffer and accesses the metadata on the cache buffer.
In this way, the CPU retains the data once transferred to the cache buffer as much as possible, and discards old data only when the cache buffer is completely used up. The CPU efficiently implements file access by using the empty cache buffer area for new data.
A commonly used algorithm for selecting a cache buffer to be replaced at this time is LRU. The LRU selects the cache buffer that has the longest time since it was last used.

JP 2000-122926 A

When measuring the free space of the entire disk, a process that counts the number of free areas by inspecting all the metadata indicating the allocation status of the disk area, such as space bitmap, allocation bitmap, and file allocation table, from the beginning to the end. Is called. In order to perform this processing, the CPU reads a large amount of metadata indicating the allocation status of the disk area, but it is unlikely that the same metadata will be used in the near future.
In addition, even when a new disk area is allocated for writing a file, the CPU reads metadata indicating the allocation status of the disk area in order to search for a free area. However, if a free area is not easily found, a large amount of metadata is read until a free area is found. Also in this case, it is unlikely that metadata that has no free space will be used again in the near future.
When searching for a free area in a directory to create a file, the CPU searches for free areas by sequentially referring to metadata representing the directory from the top. However, a large amount of metadata is read when a free area cannot be found easily. Also in this case, it is unlikely that metadata that has no free space will be used again in the near future.
When a large amount of metadata with low reusability is read in this way, data with high reusability that had been in the cache buffer until then is expelled from the cache buffer, which reduces the efficiency of use of the cache buffer. It was.

  The present invention has been made in view of such a problem, and prevents data that has been stored in the cache buffer by data with low reusability from being expelled, and effectively uses the cache buffer that is a limited resource. The purpose is to do.

  Therefore, the information processing apparatus of the present invention uses a first management unit that manages a cache buffer that is likely to be reused, and a second management unit that manages a cache buffer that is less likely to be reused. Judgment means for judging whether or not the possibility that the finished data is reused within a predetermined time is greater than or equal to a predetermined value, and the determination means determines that the possibility is greater than or equal to a predetermined value. If it is determined that the cache buffer containing the data that has been used is managed by the first management means and is determined to be less than a predetermined value, the data that has been used is contained. A distribution unit for distributing the cache buffer to be managed by the second management unit and a key for discarding the data when there are not enough cache buffers for reading data. As Sshubaffu, having a selection means for selecting a cache buffer that is managed by the second management unit.

  According to the present invention, it is possible to prevent data with high reusability that has been in the cache buffer from being expelled by data with low reusability, and to effectively use the cache buffer that is a limited resource. .

2 is a diagram illustrating an example of a hardware configuration of an information processing apparatus (computer) 1. FIG. It is the schematic showing a cache buffer and two LRU. It is a flowchart which shows an example of a cache buffer selection process. It is a flowchart which shows the procedure which searches the vacant area of a directory in a FAT file system. It is a flowchart which shows the procedure which counts up all the free clusters on a disk in a FAT16 file system.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<Embodiment 1>
FIG. 1 is a diagram illustrating an example of a hardware configuration of an information processing apparatus (computer) 1. As shown in FIG. 1, the information processing apparatus 1 includes a CPU 11 as a hardware configuration. When the CPU 11 performs processing based on a program stored in the storage device 13, the file system, the function of the LRU, or processing according to a flowchart described later is realized.
An input device 12, a storage device 13, and a display device 14 are connected to the CPU 11 via the bus 10. The storage device 13 includes, for example, a ROM, a RAM, a hard disk device, and the like, and stores data used in processing based on the program in addition to the above-described programs. It is assumed that a cache buffer and the like described later are mounted on the RAM. The display device 14 is, for example, a display that displays information. The input device 12 is, for example, an operation key for inputting information.
The input device 12 and the display device 14 are not essential components of the information processing device 1.
In the following description, it is assumed that the CPU 11 executes processing unless otherwise specified.

FIG. 2 is a schematic diagram illustrating a cache buffer and two LRUs.
The first LRU 101 manages a cache buffer that is highly likely to be reused in the near future, that is, a cache buffer that is likely to be reused within a predetermined time and that is equal to or greater than a predetermined value. The first LRU 101 is an example of a first management unit. The second LRU 102 manages a cache buffer that is unlikely to be reused in the near future, that is, a cache buffer that is less likely to be reused within a predetermined time. The second LRU 102 is an example of a second management unit. The cache buffer set 103 is a set of cache buffers, and there are eight cache buffers for one sector from B1 to B8. The LRU manages each cache buffer by connecting them with a linked list. In FIG. 2, among the cache buffers connected to the LRU, the first one is the cache buffer that has been used recently, and the last one that has not been used for the longest time. Only links that are likely to be reused in the near future are linked to the link list traced from the first LRU, and only those that are less likely to be reused in the near future are linked to the linked list traced from the second LRU. Yes.
A case where it is necessary to access new data that does not exist in the cache buffer in a state where some data is stored in all the cache buffers B1 to B8 will be described. In order to store new data in the cache buffer, it is necessary to select one of the existing cache buffers and replace the contents with the new data.
In this embodiment, the CPU 11 selects the cache buffer at the end of the first LRU when there is no cache buffer connected to the second LRU, but otherwise selects the cache buffer at the end of the second LRU. . The previous contents of the selected buffer are discarded and overwritten with new data. The cache buffer is then removed from the previously existing LRU. In the example of FIG. 2A, the cache buffer B8 at the end of the second LRU is selected.

FIG. 3 is a flowchart illustrating an example of the cache buffer selection process. When the file system needs a cache buffer, a cache buffer selection procedure is started from S401. In S402, the CPU 11 determines whether or not a cache buffer for the designated sector already exists. When a cache buffer of the same sector exists (Yes in S402), the CPU 11 proceeds to S406 and selects the buffer. If there is no cache buffer for the designated sector (No in S402), the CPU 11 proceeds to S403 and determines whether the second LRU is empty. If not empty (No in S403), the CPU 11 selects the last buffer of the second LRU in S404, and proceeds to S407. If the second LRU is empty (Yes in S403), the CPU 11 selects the last buffer of the first LRU in S405, and proceeds to S407. In S407, the CPU 11 discards the contents stored in the selected cache buffer, reads a new sector and stores it in the cache buffer, and proceeds to S408. In S408, the CPU 11 removes the selected cache buffer from the LRU.
Next, the operation when the use of the cache buffer B8 is completed will be described with reference to FIG. 2 (B) and FIG. 2 (C). When it is determined that there is a high possibility that this data will be reused in the near future when B8 is used, B8 is connected to the head of the first LRU (FIG. 2C). On the other hand, if it is determined that there is a low possibility that this data will be reused in the near future, B8 is connected to the head of the second LRU (FIG. 2B).

A method of determining the possibility that this data will be reused in the near future when the cache buffer is used will be described with reference to FIG. FIG. 4 is a flowchart showing a procedure for searching for an empty area of a directory in the FAT file system. When creating a file or directory, a process for searching for a free area in the directory is started from S501. In S502, the CPU 11 stores the head sector number of the target directory in a variable “sect” that stores the sector number. In S503, the CPU 11 obtains a cache buffer of the sector number stored in the variable “sett”. In S504, the CPU 11 stores the start address of the cache buffer in the variable p. The variable p is a pointer variable for storing the head address of a directory entry that is an area for one file or directory.
In step S513, the CPU 11 checks whether the area pointed to by p is an empty area. If the area pointed to by p is an empty area (Yes in S513), the CPU 11 moves the process to S510 and inserts this cache buffer into the first LRU. The reason for inserting into the first LRU is that a file or directory should be created in this free area in the near future, and this cache buffer is reused at that time. After S510, the CPU 11 moves the process to S511 and ends the process.

If the area pointed to by p in S513 is not an empty area (No in S513), the CPU 11 increases p by 32 bytes in S505. In step S506, the CPU 11 checks whether p is the end of the sector. If it is not the end of the sector (No in S506), the CPU 11 shifts the processing to S504. If p is the end of the sector in S506 (Yes in S506), the CPU 11 inserts the cache buffer into the second LRU in S507. The reason for inserting into the second LRU is that there is no possibility that this cache buffer will be reused in the near future because there is no free space in this sector.
In step S508, the CPU 11 checks whether the sector indicated by the variable “sett” is the end of the directory. If it is the end (Yes in S508), the CPU 11 means that there is no free space up to the end of the directory, and thus ends this processing in S512. If it is not the end of the directory in S508 (No in S508), the CPU 11 stores the logical next sector number of the sector pointed to by sect in the directory in sect, and returns the process to S503.
Through the above processing, in the process of searching for a free area in the directory, a cache buffer that has no free area is inserted into the second LRU, and a cache buffer that has found a free area is inserted into the first LRU. This makes it possible to effectively use the cache buffer.

  Further, although this embodiment is an example of searching for a free area in a directory, as another embodiment, a process of searching for a free cluster in order to write new data on a disk can be similarly realized. The difference from the process of searching for a free area in the directory is that a file allocation table (FAT) is checked instead of a directory to search for a free cluster. If one FAT entry indicates a vacancy, the corresponding cluster is vacant. When a free space is found, the CPU 11 inserts a cache buffer including the FAT into the first LRU. The reason is that data should be applied to the free cluster, and the FAT indicating the free space is rewritten. Further, the CPU 11 inserts the cache buffer including the FAT that does not include the empty space into the second LRU. The reason is that the FAT is unlikely to be rewritten in the near future.

<Embodiment 2>
Next, an embodiment different from the above-described embodiment will be described with reference to FIG. FIG. 5 is a flowchart showing a procedure for counting all free clusters on the disk in the FAT16 file system. In the FAT16 file system, the status of each cluster is represented by a 2-byte FAT entry, and in order to count the number of free clusters, all FAT entries are inspected every 2 bytes from the beginning to the end, and free space is checked. Count the number of entries shown. The feature of this processing is that all sectors in the FAT area need to be read and accessed sequentially, but it is unlikely that any cache buffer read for this processing will be reused in subsequent processing. is there. Therefore, in this embodiment, the CPU 11 inserts all the read cache buffers into the second LRU. This increases the possibility that the cache buffer contained in the first LRU will be used more effectively before this process starts.

A more specific procedure will be described with reference to FIG. When the file system needs to count free clusters, the process is started from S601. In S602, the CPU 11 stores the leading sector number of the FAT area in the variable “sect” indicating the sector number, and stores “0” in the variable “cnt” indicating the result of counting the free clusters. In step S <b> 603, the CPU 11 acquires a cache buffer for the sector indicated by “sett”. S604, S605, S606, S607, and S608 are the portions that count the empty clusters indicated by the FAT for one sector in the cache buffer, but are not the essence of this embodiment, and thus the description thereof is omitted. When the processing for one sector is completed, the process reaches S609, and the CPU 11 inserts the cache buffer that has been processed here into the second LRU. Thereafter, in S610, the CPU 11 checks whether or not “sett” is the end of the FAT area. If it is the end (Yes in S610), the CPU 11 ends the process in S612. At this time, cnt contains the total number of free clusters on the disk. If sect is not the end of the FAT area in S610 (No in S610), the CPU 11 stores the next sector number in sect and returns the process to S603.
With the above processing, the cache buffer can be used effectively.

<Other embodiments>
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, etc.) of the system or apparatus reads the program. It is a process to be executed.

  As described above, according to each of the above-described embodiments, it is possible to prevent data that has been stored in the cache buffer from being expelled by data with low reusability, and to effectively use the cache buffer that is a limited resource. .

  The preferred embodiments of the present invention have been described in detail above, but the present invention is not limited to such specific embodiments, and various modifications can be made within the scope of the gist of the present invention described in the claims.・ Change is possible.

11 CPU
13 Storage device

Claims (3)

A first management means for managing cache buffers that are likely to be reused;
A second management means for managing cache buffers that are unlikely to be reused;
A determination means for determining whether or not the possibility that the data that has been used is reused within a predetermined time is greater than or equal to a predetermined value;
When it is determined by the determination means that it is equal to or greater than a predetermined value, the first management means manages the cache buffer containing the data that has been used and is less than a predetermined value. If it is determined, a distribution unit that distributes the cache buffer containing the used data to be managed by the second management unit;
Selection means for selecting a cache buffer managed by the second management means as a cache buffer for discarding data when there is not enough cache buffer for reading data;
An information processing apparatus. An information processing method executed by an information processing apparatus,
A first management step for managing cache buffers that are likely to be reused;
A second management step for managing cache buffers that are unlikely to be reused;
A determination step of determining whether or not the possibility that the data that has been used is reused within a predetermined time is greater than or equal to a predetermined value;
If it is determined in the determination step that the value is greater than or equal to a predetermined value, the cache buffer containing the data that has been used is managed in the first management step, and is less than a predetermined value If determined, a distribution step for distributing the cache buffer containing the used data to be managed in the second management step;
A selection step of selecting the cache buffer managed in the second management step as a cache buffer for discarding data when there are not enough cache buffers for reading data;
An information processing method comprising: On the computer,
A first management step for managing cache buffers that are likely to be reused;
A second management step for managing cache buffers that are unlikely to be reused;
A determination step of determining whether or not the possibility that the data that has been used is reused within a predetermined time is greater than or equal to a predetermined value;
If it is determined in the determination step that the value is greater than or equal to a predetermined value, the cache buffer containing the data that has been used is managed in the first management step, and is less than a predetermined value If determined, a distribution step for distributing the cache buffer containing the used data to be managed in the second management step;
A selection step of selecting the cache buffer managed in the second management step as a cache buffer for discarding data when there are not enough cache buffers for reading data;
A program that executes

Images