JP2007133535A - File system inside storage device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a file system capable of certainly holding management information even if power shutoff during writing occurs, and holding a state at least just before a writing start about a data body. <P>SOLUTION: This file system has: a data area storing a first file and the data body of a second file copied with the first file; a management area managing disposal of the data body inside a storage device; and an area storing flag information showing the file performed with the latest writing of the first file and the second file. When newly producing the file, information related to the disposal of the first file and the second file is produced in the management area, the first file and the second file are produced in the data area, and the file performed with the latest writing is stored in the flag information. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a file system in a storage device, and more particularly to a file system in a storage device that can prevent data destruction and inconsistency even when a power supply is cut off during a writing process in an embedded system.

  There is a flash memory device as a semiconductor memory device having features such as a small size, light weight, low power consumption, and impact resistance. In recent years, with the progress of higher integration and lower cost of this flash memory device, it has come to be used as a storage device even in an embedded system.

In flash memory devices, it is common to manipulate data in a format called a file system. In the file system, the contents to be accessed are manipulated in units of files. The file is divided into a data body and management information.
The data body is stored in a single or a plurality of fixed length areas.
On the other hand, the management information stores the contents of which fixed-length area the data body starts to be stored and in what order when a plurality of fixed-length areas are used. The management information also stores contents such as the size of the data body, the file name, and the file creation date and time.

A typical example of the file system is FAT (File Allocation Table). Since this file system can be used by many computers, FAT has been particularly widely adopted in embedded systems for the purpose of improving convenience.

However, if the power is turned off while a file is being written to the memory, the data body may be destroyed or the consistency of management information may be lost, making subsequent file access impossible.
The cause of the power being cut off while writing a file is not only a physical factor due to lightning or a power failure, but also a human factor such as an operator turning off the power while writing a file.

  A method is known in which an uninterruptible power supply (UPS) is provided and power is supplied until writing is completed in response to such power interruption during the writing process. However, since the uninterruptible power supply uses a battery as a standby power supply, it is not easy to reduce the size and price. There are also problems with environmental resistance, life, and maintainability. Furthermore, this method cannot cope with the case where the system hangs up while writing a file due to a software defect or the like.

  As another solution, there is known a method in which the occurrence of a failure is unavoidable in a personal computer and the entire inspection and repair of file system information is executed after the failure has occurred. In this method, there is a problem that the range of failures that can be repaired is limited such that the entire inspection time is long and the restoration of all files cannot be guaranteed in the first place.

  For this reason, in the conventional technique, the write data is merged with the existing data, additionally described in the unused area, and sequentially written in the two sets of management areas (see, for example, Patent Document 1). Further, the restoration can be performed by sequentially writing to the three sets of management areas (for example, see Patent Document 2).

JP 2005-149291 A Japanese Patent Laid-Open No. 2005-235028

  However, the method of Patent Document 1 consumes an unused area every time data is written, and thus cannot withstand long-term operation. There is also a problem that it is difficult to determine which of the two management areas is correct.

  Furthermore, the method of Patent Document 2 has a problem that the data body of the file being written cannot be recovered.

  The present invention has been made in view of such problems, and management information is always retained even when a power interruption occurs during writing, and at least the state immediately before the start of writing is retained for the data body. The object is to provide a file system that can be used.

  In order to solve the above problem, the present invention is as follows.

  The invention according to claim 1 is a file system constructed in a storage device, wherein a data area storing a data body of a first file and a second file obtained by copying the first file, A management area for managing the arrangement of the data body in the storage device; an area for storing flag information indicating which of the first file and the second file has been most recently written; When creating a new file, information on the arrangement of the first file and the second file is created in the management area, and the first file and the second file are created in the data area. In the flag information, it is stored that the latest writing was performed for the first file, and writing to the file is performed. When performing is characterized by that write only one of the data area of the first file or the second file in accordance with the flag information.

  The invention described in claim 2 is characterized in that an area for storing the flag information is secured in the data area of the first file and the second file.

  According to a third aspect of the present invention, the area for storing the flag information is secured on a volatile memory backed up by a battery.

  The invention according to claim 4 indicates the second file when the flag information indicates the first file after writing to the first file or the second file. The flag information is changed, and when the flag information indicates the second file, the flag information is changed so as to indicate the first file.

According to the present invention, while maintaining compatibility with an existing file system, even if an unexpected power interruption occurs, the entire file system will not be inconsistent. It becomes possible to read the file that has been written last.
That is, there is an effect that a system capable of accessing correct data even when an unexpected power interruption occurs can be realized at a low cost without impairing performance.

  Hereinafter, specific examples of the method of the present invention will be described with reference to the drawings.

  FIG. 1 is a schematic diagram showing the configuration of a storage device to which the present invention is applied. This apparatus includes a CPU 101, ROM 102, RAM 103, I / O (Input / Output) 105, secondary storage device I / F (Interface) 106, and secondary storage device 107.

  The CPU 101 performs input / output of information with other components via the bus 104 and performs numerical operations and logical operations based on programs to control the entire apparatus, and is a typical semiconductor device.

  The ROM 102 is a semiconductor memory device that stores programs. The contents are not lost even when the power is turned off for read-only purposes, and mainly stores the programs necessary for the startup operation of the device. The RAM 103 is a semiconductor memory device that can be freely read and written. During the operation of the apparatus, the program in the secondary storage device 107 is copied and used, or data is temporarily stored.

  The I / O 105 is various inputs / outputs other than the secondary storage device 107. Connect communication devices, sensors, and actuators. Since there is no particular relevance in the present invention, detailed description is omitted. The secondary storage device I / F 106 is used to perform protocol conversion for connecting the secondary storage device 107 and the bus 104.

  The secondary storage device 107 is a flexible disk, a compact flash (registered trademark), a removable media such as a flash memory card, or a hard disk. Also, a flash disk that has recently appeared on the market may be used. A flash disk is not a magnetic disk, but a flash memory device is built in to constitute a storage device, and can be connected with the same protocol as a hard disk.

FIG. 2 shows a schematic software configuration diagram of an apparatus to which the file system of the present invention is applied. The software of this apparatus is a group of programs executed on the CPU 101 and is stored in the ROM 102.
The program group includes an application 201, an OS (operating system) 202, and an I / O library 203.
In a normal configuration, there is one OS 202 and one I / O library 203, but there are both cases where there is only one application 201 and a plurality of applications 201.
In any case, the minimum necessary program group may be stored in the ROM 102 and the rest may be stored in the secondary storage device 107.

  The OS 202 is highly versatile basic software that makes it easy to access the I / O 105 and the secondary storage device 107 from the application 201. The file system 204 is a component for handling data on the secondary storage device 107 as a file. The lowermost device driver 205 directly performs input / output operations with the secondary storage device 107, absorbs differences in protocols and parameters peculiar to individual storage devices, and the upper file system 204 makes the secondary storage device It is abstracted so that it is not necessary every time.

  The I / O library 203 is a means for accessing the I / O 105 and the secondary storage device 107 without using the OS 202. This is a software interface for performing input / output operations with the secondary storage device 107 as in the case of the OS 202. Optimization (high speed, compact) can be expected rather than using the OS 202. On the other hand, the versatility of the I / O library 203 may become poor due to optimization.

  FIG. 3 is an explanatory diagram of recording information required for a standard file system. In the storage area 301 of the secondary storage device 107, file system information is arranged in three areas. The three areas are a cluster number table area (1) 302 and (2) 303, a directory area 304, and a data area 305. Each is arranged at a fixed position on the storage area 301.

An access unit of the secondary storage device 107 has a fixed size called a sector. The three areas are physically accessed in units called sectors. However, only a logical access to the data area 305 uses a unit called a cluster in which a plurality of sectors are collected. It is common to use 1 cluster = 4 sectors or 1 class = 8 sectors.

  The cluster number table area (1) 302 stores information indicating in which cluster the recorded data body is arranged in what order. The cluster number table area (2) 303 is a copy for backing up the cluster number table area (1), and the main body arrangement information can be secured even if any of the cluster number tables is damaged.

  In the example of FIG. 3, a numerical value “21” is written in No. 20 of the cluster number table. This indicates that valid data exists in No. 20 of the cluster, and subsequent data exists in No. 21 of the cluster. The number “22” is written in No. 21 of the cluster number table. This indicates that valid data exists in the cluster No. 21 and the following data exists in the cluster No. 22. A code of “E” (meaning END) is written in No. 22 of the cluster number table, but there is valid data in No. 22 of the cluster, there is no following data, and the last data is stored. Indicates that

The directory area 304 stores the information of the head cluster number in addition to the file name, update date and time, file size (size), etc. for each file. The head cluster number indicates in what number cluster the head of the data body exists.
The example of FIG. 3 shows that a certain file contains the first data in the 20th cluster. When referring to the data body of this file, the directory area is first searched by the file name, and it is known from the result that the first cluster number is 20. Next, if the cluster number table area is traced sequentially from No. 20, it can be seen that it is sufficient to refer to No. 21 and No. 22 of the cluster number table.
The cluster number table area and the directory area 304 are collectively referred to as a management area.

  The data area 305 stores the file data body. As described above, data is divided and recorded in clusters. The example in the figure shows that data is stored for each part of “A”, “B”, and “C” in the order of clusters 20, 21, and 22.

  The order of writing to the cluster number table area 302/303, the directory area 304, and the data area 305 is not specified. In particular, the file system does not write each time when the file contents are changed, but writes when a certain amount of changes are accumulated using the internal buffer. For this reason, it is impossible for the application 201 to know which area is being written to when.

The problem here is that if a power interruption occurs during a series of processes for writing a file, recovery is very difficult. If the power is turned off during writing to the data area 305, the contents of the data area 305 are not guaranteed. Also, if the power is shut off during writing to the cluster number table areas 302 and 303 and the directory area 304, the consistency as a file is not maintained.
In addition, when the secondary storage device 107 is composed of a flash memory device, not only the file but also all information in the same sector becomes unreliable. This is because, as a characteristic of the flash memory device, simple overwriting cannot be performed, and a procedure of writing after erasing all predetermined areas is necessary.

  FIG. 4 is a flowchart showing a procedure for writing a file according to an embodiment of the present invention. In response to a file write request from the application 201, the cluster number table areas 302 and 303, the directory area 304, and the data area 305 described in FIG.

First, it is determined whether or not the writing process is the first time (step S01). Whether it is the first time or not can be determined by whether or not a file with the name to be written exists. If it is the first time, the file size is defined (step S02). If the file has a possibility of changing its size, the file is defined with a predetermined maximum size. If the file size is specified, the file is written (step S03). If the size of the data to be written is smaller than the maximum size, the shortage is filled with dummy data. This file is set to “1”.
Subsequently, the file “2” copied from the file “1” is written (step S04).
In practice, a naming rule is set in advance so that the OS 202 can recognize that the file “1” and the file “2” have the same contents.
For example, when a file named “DATA” is generated in response to a request from the application 201, an extension is automatically added, a file “DATA.1” is generated as the file “1”, and the file A file “DATA.2” is generated as “2”. Since the names of the parts before the extension are the same, the OS 202 can determine that the file “DATA.1” and the file “DATA.2” should be handled as a set.

Thereafter, “1” is recorded in the flag information indicating which file is the most recently written file (step S05).
This flag information can be held in the data area of each file. For example, the last byte of the data body is used, the last byte is set to “0” for the first time, and a value obtained by adding 1 each time it is written is recorded. In this method, the last bytes of the file “1” and the file “2” are compared, and it is determined that the file with the larger value is the last file written. In this method, it takes time and effort to compare the contents of the last bytes of both the file “1” and the file “2”. In order to save the time and effort, the flag information may be held not in the secondary storage device 107 but in a battery-backed volatile memory (not shown).
In this case, the numerical value “1” is written when writing to the file “1”, and the numerical value “2” is written when writing to the file “2”. In this method, it is determined that the written numerical value is the file that was last written as it is.
Even if the power is cut off during writing, the contents of the volatile memory are retained by the battery.

However, since the power supply cannot be cut off during the first writing, the processes in steps S01 to S05 must be performed before the system starts actual operation.
The structure of the file system constructed by the above procedure is shown in FIG.

Next, the case of file update processing will be described. If there is a file with the specified name, it is determined that it is not the first time (step S01). Next, flag information is determined (step S06). If the flag information indicates “1”, it can be determined that the previous writing was for the file “1”, so the file “2” is written (step S07).
What is important here is that the cluster number table areas 302 and 303 and the directory area 304 remain as they are, and only the data area 305 is written.
Since the file has been generated with a predetermined maximum size, there is no need to update the cluster number table areas 302 and 303 and the directory area 304. When the writing to the data area 305 is normally completed, “2” is recorded in the flag information (step S08).

  If the flag information indicates “2”, it can be determined that the previous writing was for the file “2”, so only the data area 305 is written to the file “1” (step S09). When the writing to the data area 305 is normally completed, “1” is recorded in the flag information.

  FIG. 6 is a flowchart showing a control procedure for reading a file according to an embodiment of the present invention. In response to a file read request from the application 201, the file data body is read using the cluster number table areas 302 and 303, the directory area 304, and the data area 305 described with reference to FIG.

  First, flag information is determined (step S20). If the flag information indicates “1”, the data of the file “1” is read (step S21). If the flag information indicates “2”, the data of the file “2” is read (step S22).

  In either case, since the flag information indicates a file for which writing has been normally completed, the power supply is interrupted during writing, and a file containing a contradiction is not read.

  As described above, according to this embodiment, compatibility with the existing file system is maintained, and no conflict is caused in the entire file system even if an unexpected power interruption occurs, and the file system is duplicated. Of these files, the file that has been successfully written last can be read.

According to the present invention, it is possible to realize a system capable of accessing correct data even if an unexpected power interruption occurs, at a low cost without impairing performance.
For example, the present invention can be applied to a control device for an industrial robot as follows.
Also in the control device for the industrial robot, the part related to the storage device has the same configuration as in FIG. In general, an operator operates a robot with a teaching device or the like connected to the control device to create an operation pattern desired to be performed by the robot, and records the operation pattern in a secondary storage device in the control device. Also, other information such as parameters related to robot control and setting contents can be recorded in the secondary storage device.
However, if the control device is turned off due to an operator's operation error while the taught motion pattern is being written to the secondary storage device, the robot motion pattern taught over a long time may be lost. is there.
By applying the file system of the present invention to an industrial robot controller, such a problem can be solved and at least the contents of the operation pattern immediately before the start of writing can be retained.

Configuration diagram of a storage device to which the present invention is applied Schematic configuration diagram of software showing an application form example of the present invention Explanatory diagram showing the structure of a standard file system The flowchart which shows an example of the write-in procedure of the file by this invention Explanatory drawing which shows the structure of the file system of this invention The flowchart which shows an example of the reading procedure of the file by this invention

Explanation of symbols

101 CPU
102 ROM
103 RAM
104 Bus 105 I / O
106 I / F for secondary storage
107 Secondary storage device 201 Application 202 Operating system (OS)
203 I / O library 204 File system 205 Device driver 301 Storage area 302 of secondary storage device 107 Cluster number table area (1)
303 Cluster number table area (2)
304 Directory area 305 Data area 310 Cluster number table (data recording information)
311 Directory information 312 Data body

Claims (4)

A file system constructed in a storage device,
A data area for storing a first file and a data body of a second file obtained by copying the first file;
A management area for managing the arrangement of the data body in the storage device;
An area for storing flag information indicating which of the first file and the second file has been most recently written;
When a new file is created, information on the arrangement of the first file and the second file is created in the management area, and the first file and the second file are created in the data area. At the same time, in the flag information, it is stored that the latest file was written in the first file,
A file system in a storage device, wherein when writing to a file, writing is performed only to the data area of either the first file or the second file according to the flag information.   2. The file system in the storage device according to claim 1, wherein an area for storing the flag information is secured in the data area of the first file and the second file.   2. The file system in the storage device according to claim 1, wherein an area for storing the flag information is secured on a volatile memory backed up by a battery.   If the flag information indicates the first file after writing to the first file or the second file, the flag information is changed to indicate the second file; 2. The file system in the storage device according to claim 1, wherein when the flag information indicates the second file, the flag information is changed so as to indicate the first file.

Images