JPH0254327A - Control system for disk data - Google Patents

Abstract

PURPOSE:To set a postscript type disk under the same control as a reversible medium of a general-purpose operating system by updating a file allocation table and all areas of a directory for each sector and adding together the final information of each configuration sector to refer the whole area. CONSTITUTION:The file allocation tables(FAT) 10 and the directory areas 11 are added and updated every time a file is read and written. Then the logical sector numbers is read out of the pointers for each sector when the FAT part 10 is read. If this pointer is equal to phi, i.e., no data is written, the initialization information is sent to a system. When the FAT part 10 is written, the logical sector addresses are written out of the pointers for each sector. Then an address is newly set by reference to a pointer in the case the pointer is equal to phi. The areas 11 are updated in the same way as the FAT 10. Then the information to be newly stored and added are recorded to the new sector following the recorded sector for each sector. Then the final information of each configuration sector are added together for reference to the entire area.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to a disk data management system, and in particular, in recent years, computer data such as program software has become increasingly popular due to its large capacity, access speed, and data transfer speed. In addition, write-once type (Write 10) is attracting attention as a recording medium for various data such as image data and music data.
This invention relates to a disk data management method suitable for disks (hereinafter referred to as W10).

(Prior Art) Conventionally, disks such as floppy disks and hard disks have been widely used as recording media for recording digital data such as computer programs. The recording format of these disks is determined by the operating system (O8) installed in the computer, and different O8s are usually not compatible, but any O8 can support external storage devices. It is assumed that data can be erased and rewritten, and the storage area of the disk is clearly divided into a file area for recording data and a directory area for recording information between that data. formatted.

FIG. 10 schematically shows a recording area on a flexible disk such as a floppy disk, for example, the upper part of the figure corresponds to the outer circumference of the flexible disk, and the lower part corresponds to the inner circumference. Reserved area 1 is an area for recording data necessary for the system and cannot be used by the user. Area 2 is a FAT (File A11ocatio)
n Table), and this 2 is an area for recording data regarding the file arrangement. 3 is the directory (
DIR) area, in which the date, file name, etc. of the data file recorded in the file area 4 are recorded.

Specific examples will be described below with reference to FIGS. 10 and 11.

First, for convenience, the file area 4 in which data is actually stored is divided into a plurality of blocks, and the blocks are sequentially numbered from 2 onwards.

Next, as the FAT area 2, if one block is represented by 12 bits, for example, a data capacity of (number of blocks + 2) x 12 bits is secured, and as the DIR area 3, a fixed area of 32 bytes per file is secured. Here, the capacity of each area is a multiple of the sector size.

The procedure for creating one file is to first record the file name and the first block number in the file area 4 where data is stored in the fixed length DIR area 3, and then store the data in a predetermined block in the file area 4. , When data spans multiple blocks, the chain information is recorded in the FAT area 2.

In this way, general-purpose O8 uses FAT, directory (DIR)
) on the disk and main RAM,
Data read/write (14ritθ
) Sometimes files are managed by adding and updating their areas. Note that data on the disk is read/written in fixed length units called sectors. Since this management method is performed assuming a freely readable/writable medium, it cannot be used as is for write-once optical discs that can be written only once (FAT and directory areas are frequently rewritten).

Figure 11 shows the FAT inside a reversible disk.
It is clear from this figure that the FAT2a2b and the directory area 3a each consist of a plurality of sectors. For example, in this figure, the FAT area has three sectors (
The directory area consists of four sectors (first sector to fourth sector). In this case, the FAT area handles important file chain information, so to increase reliability, the same content is written twice (
There is a first FAT (2a) and a second FAT (2b).

) In the following description, each sector of the FAT or directory may be referred to as a FAT first sector or a FAT second sector.

By the way, the range of applications of write-once optical discs is expanding due to their large capacity and ease of operation.These write-once optical discs store data in units of predetermined sectors and blocks, for example, from the inner circumference to the outer circumference. Record.

(Problem to be solved by the invention) However, because write-once optical discs have been developed for a short period of time, the data recording format is not necessarily standardized, and some manufacturers have their own file management methods. The current situation is that .

If you try to manage this write-once optical disk using the same management method as the flexible disk, the area other than the file area (for example, the directory area) will be
This is predetermined by the following rules, and files cannot be managed on non-erasable write-once optical discs because they cannot be rewritten.

For example, the directory area 3.3a becomes full once. That is, the conventional O8 cannot handle this, and therefore, even if there is an unrecorded area in the file area 4, unnecessary directory areas 3 and 3a cannot be erased, making it impossible to write any more files in the file area 4.

The present invention has been made in view of the above-mentioned points, and has the following features: (1) to efficiently utilize the file area of a disk-shaped recording medium other than a flexible disk including a write-once optical disk, and (2) to make a write-once optical disk transparent for general-purpose O8. The purpose is to provide a disk data management method for data management.

(Means for Solving the Problems) The present invention uses write-once media to allow a general-purpose operating system to manage disk data in the same manner as reversible media. This provides a disk data management method in which the entire area is referenced by updating and adding the final information of each constituent sector.

(Example) Fig. 1 is a schematic diagram showing the area allocation of a write-once disk obtained by the method of the present invention, and Fig. 2 is a schematic diagram showing the area allocation of a write-once disk obtained by the method of the present invention. This is a schematic diagram showing the correspondence of disk areas after allocation. This is a schematic diagram showing the correspondence of disk areas after allocation. This is a diagram that shows how many write-once areas are allocated to write-once media. This was achieved by securing the above.

Note that, in the present invention, reading/writing of data on a disk is performed in units of a fixed length called a sector, as is conventional.

In FIGS. 1 and 2, the FATIO and directory areas 11 are areas that are additionally updated each time a file is read/written, as described above.

In order to accommodate write-once media, the present invention continuously and fixedly secures sufficient capacity to rewrite these areas in adjacent areas of the disk.

Each region will be explained below using FIGS. 1 and 2. First, we will start with the directory area 11 in the allocation of all areas of the media.

The directory area 11 is an area for storing file names, and an area is reserved so that data can be added and updated at a fixed length of 32 bytes per registration of one file, and can be rewritten for the number of files managed. Therefore, the number of rewrites until the first sector of the directory becomes full is (1 sector length/32 bytes). Therefore, an area for (number of rewrites of one directory sector x number of directory sectors) sectors is taken on the disk, and this is made to correspond to each sector in the conventional method. however,
When actually registering a file from the general-purpose O8, the directory area is written twice, once when the file is opened and when the file is closed. If this writing is performed sequentially, a rewriting area of two sectors is required for each file creation, and a huge amount of spare area is required.

(See FIG. 3) Therefore, write control of the DIR area is performed as follows, that is, when registering one file from the general-purpose O8, the process is executed in the following manner.

Read the FAT area - Read the DIR area → Write the DIR area → Write the file area - + Write the FATFAT - Write the last sector of the file area → Write the DIR area DIR area when the first file is opened When writing, the file name, file creation date and time information is stored in the area on the disk, and when writing to the DIRfJl area when closing, the total capacity of the file is added and updated.

Therefore, the final data to be left on the disk is the IR.
Only the contents are required, and since the DIR area is not referenced between the first and second times, it is not necessary to record the first information (see Figure 4). Fifth to write information to disk
Control was performed using flags as shown in the figure. In other words, FAT■, DIR■, ■, FIL are used as logical sector determination units.
Set four areas of E■, and as area ■, FAT-FILE flag = 1 area as ■, DIR and FILE flag = φ-Write.
=O (write omitted) → FILE flag = φ (FILE flag reset) Area ■ is DIR and FILE flag = 1 → Correction +
te=1<write) →FILE flag=φ (FILE flag reset) FILE flag=1 is set as area (■). As a result, the area reserved as a DIR area on the disk is halved, making it possible to effectively utilize disk space.

Note that a blank sector portion 13 corresponding to, for example, five tracks of the disk is provided between the directory area 11 and the file area 12. This is provided to shorten the time required to detect the boundary between the directory area 11 and file area 12.

Next, the FAT area 10 will be explained with reference to FIGS. 1 and 2. FAT is an area for storing chain information of data storage blocks, and the amount of information increases in variable length according to the size (capacity) of write data. Therefore,
The maximum number of times the first sector of FAT can be rewritten is based on the assumption that the data capacity of all files is within 1 program.
By setting the data to 12 bits (1 sector length/
12 bits) times. On the disk, a blank sector portion 13 corresponding to the maximum number of rewrites plus five tracks is required for each sector in the conventional method. The purpose of the blank sector section 13 is the same as that of the directory area 11.

On the other hand, if the FAT is updated once every time one file is registered, the number of times the entire FAT area is rewritten will be equal to the number of files that the O8 can manage.

Therefore, the minimum number of sectors required for the entire FATIJi area is (
Number of registrable files)+(blank sectors)Xn.

According to the method of the present invention, in order to minimize the capacity of the entire FAT area, the starting address of each of the second to nth FAT sectors is not fixed, but the following method is adopted. That is, F.A.
Since data is added to the variable length T with one update, how many times should it be updated before the amount of information in one sector exceeds the FAT second
Therefore, when the amount of information in the FAT first sector exceeds the length of one sector and subsequent data is recorded in the FAT second sector and beyond, the FAT first sector must be recorded. The value obtained by adding the number of sectors for 5 tracks to the final address of the sector is registered as the start address of the FAT second sector.

In this way, the undefined number of rewrites for each FAT sector is absorbed by freely moving the start address of the next FAT sector within the FAT area 10.

Next, use Figures 6 and 7 to process when the system starts up, update the entire area of the FAT and directory in sector units, and add the final information of each constituent sector to refer to the entire area. An example will be explained below. Here, FIG. 6 is a schematic diagram showing a method for updating a FAT area, and FIG. 7 is a schematic diagram showing a method for updating a directory area.

As mentioned above, R/W (read/write) of data on the disk
(Write) is performed as a unit of fixed length called a sector, and the FAT and DIR areas each consist of a plurality of sectors.

Updating FAT and DIR information is done by recording the combination of previous information and newly added information in sector units to the new sector next to the already recorded area of FAT and DIR.The amount of information is 1 sector. If the number exceeds the number, a new fixed address is created with the excess amount as the second sector, and the contents of the latest updated sector are written sequentially below that address. The method of storing the latest sector address in the pointer section from the contents of the disk is implemented as follows. First, the FAT area will be explained with reference to FIG. 6. F
The start address of the first sector of the AT is set in advance by the system as a unique value.

When the system wakes up, it checks the number of sectors continuously recorded from this unique address, and stores the address of the last sector in which writing has been performed in the pointer section as the latest contents of the first sector.

The start address of the next second sector is the sum of the number of sectors for five tracks from the final address of the first sector. This is because blank sectors for five tracks are provided at the boundaries of each sector in order to shorten the detection time when examining continuous recording sectors.

The number of consecutive recording sectors after the second sector start address is checked, and the address of the sector a after that is stored in the pointer section as the latest content of the second sector. The same applies to the following sectors.

After the second sector, check the number of consecutively recorded sectors and find out that it is φ
If so, φ is assigned to the pointer section to indicate that the sector is not recorded.

Since writing is performed in the FAT section sequentially from the first sector to the nth sector, if the first unrecorded sector is detected, the subsequent sectors will be unrecorded, and there is no need to check the number of consecutively recorded sectors. The time it takes to start up the system can be significantly reduced.

The system reads each sector of the FAT when starting up, but in order to shorten the time during startup and to make effective use of disk space, the starting address of each sector is fixed. There was no way to initialize the first sector.

Therefore, when the system makes a read request for an unrecorded sector (the contents of the pointer is φ), the initialization information of the FAT sector is transferred to the system side without actually reading the contents of the disk.

In addition to this, it is also possible to scan the entire area and collect data whose sectors are full, but this would require as many scans as the number of writes.

However, by equipping the device with a mode that can detect the boundary between the recorded area and the unrecorded area, the necessary directory or FAT sector can be detected in a much shorter time than with the above method.

Next, the DIR area will be explained with reference to FIG. D
Also for the IRI area, the start address of the first sector is set as a system fixed value.

Unlike the FAT area, the DIR area is updated once with a fixed byte length (32 bytes). Therefore, the number of times one sector is rewritten is multiplied by the sector length (/32) times, so that the starting address up to the m-th sector can be fixed. Therefore, in order to find the final address of each sector, instead of detecting the number of consecutively recorded sectors for each sector as in FAT, the number of consecutively recorded sectors is calculated in the entire DIR area and one position before the m-th sector start address. The address is defined as the final address of the (m-1)th sector.

In this way, by reducing the number of times the number of continuously recorded sectors is checked, the speed at system start-up is increased. Note that FIG. 8 is a block diagram when performing start-up processing, and the pointer control unit checks the number of continuous recording sectors and stores the results in the FAT and DIR unit pointer registers. After storing them, they are handed over to the FAT management section or DIR management section.

Addresses containing the latest information of each sector constituting the FAT etc. are managed by the pointer section shown in FIG.

Pointers are prepared for the number of sectors that make up the FAT. The first fixed address to be set is the first FAT area 10a.
, and the second FAT area 10b, only the first sector start address and the remaining pointers are initialized to φ. Light (Wri)
te) At times, information is recorded in the new sector following the existing edge sector, and the contents of this pointer (sector address of the latest information)
Increase by one. If the information to be written exceeds the FAT first sector and is additionally recorded in the FAT second sector, the FAT
The value obtained by adding the number of sectors for 5 tracks to the latest address of the first sector is registered in the pointer section as the start address of the FAT second sector. Here, blank sectors for five tracks are inserted at sector boundaries in order to shorten the time for detecting sector breaks. Furthermore, when reading data, the actual address of the latest sector is read by referring to the pointer of each sector.

However, when the content of the pointer is φ, no writing has actually been performed on that sector, so the initialization information for that sector in the FAT is transferred to the system side. This is done so that unnecessary initialization information does not need to be recorded on the disk. In addition, since the start address of each sector is not fixed and can be made flexible, FATil
This will lead to significant labor savings in the overall capacity of the area.

This point will be explained in detail below using FIG. 9.

Figure 9 shows read/write
FIG. 2 is a block diagram showing time pointer management.

Here, the process when the FAT unit 10 is started up has been described above, so its explanation will be omitted.

(1) First, reading from the FAT section 10 will be explained.

In the FAT section, the logical sector address is read from the pointer for each sector.Next, the pointer is referenced, and if the pointer is φ (the logical sector address is φ), that is, no data has been written, the system The initialization information is transferred to

In this case, data that has not actually been written is transferred to the system side (data requesting side) as if it were read from an optical disk.

(Secondly, in the above, if the pointer is other than φ, data on the optical disk is read. That is,
When the content of the pointer is other than φ, data is stored on the optical disk, and the pointer points to its logical address.

Then, this logical address is converted into a physical address before data is read.

(Thirdly, the write operation of the FAT section 10 will be explained.

In the FAT section, a logical sector address is written using a pointer for each sector.Next, the pointer is referenced, and if the pointer is φ, a new write address is set. In other words, the pointer at the time of writing is φ,
This is the first time to write to that sector,
In that case, calculate the first address of the previous sector as the reference logical sector address (for example, if it is the FAT second sector, refer to the final address of the FAT first sector). The data is written to the pointer section using the address obtained by adding the number as the start address.

(4) Next, in the above case, if the pointer is other than φ, the contents of the pointer are updated. In other words, if the contents of the pointer are other than φ, it means that writing has been done so far, so the new area is Update the contents of the pointer at the same time as writing the data.

Next, a method for updating the directory area 11 will be explained with reference to FIG.

The method of updating the directory area is basically the same as that of FAT, in which new information to be stored is added to the existing information and is recorded in a sector unit in a new sector next to an already recorded sector. Similarly, the amount of information exceeding one sector is recorded as the second sector, and a pointer is provided for each sector, and the latest information address of each sector is managed by the pointer shown in FIG. 8 described above. However, unlike FAT, the directory is 1
Since the information updated each time is fixed at 32 bytes, the number of times each sector is updated is determined by the sector length. That is,
(1 sector length/32 bytes) times. Therefore, in the directory area, if the start address of the first sector of the directory is set, the start addresses up to the m-th address can naturally be determined. Therefore, unlike the FAT area, there is no need to insert a blank sector for detecting the next directory sector start address at each sector boundary, and the disk space can be used effectively. However, a blank sector section 13 is inserted at the boundary between the directory area and the file area for boundary detection.

When writing data, the contents of the pointer are incremented by one and new information is written to the next sector.

When reading, the contents of the latest directory information sector are read by referring to the contents of each pointer. When the contents of the pointer are φ, no writing has actually been done to that sector.
The directory initialization information is transferred to the system side.

This eliminates the need to record extra initialization information on the disk and increases the efficiency of detecting the start address of each sector.

(Effects of the Invention) As described above, according to the present invention, write-once disks can be placed under the control of the general-purpose O8 in the same way as conventional media such as floppy disks and hard disks.
You can use the commands and system calls (software interrupts) provided by O8, leave the management of files in the device to O8, and run various software that runs on general-purpose O8 on a write-once disk. Furthermore, even a device newly created by a user (manufacturer) can be placed under the management of O8 to provide versatility.

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram showing the area allocation of the write-once disk obtained by the method of the present invention, and Fig. 2 is a schematic diagram showing the area allocation of the conventional disk and the write-once disk obtained by the method of the present invention. Fig. 3 is a schematic diagram of the DIR area when write control is not performed, Fig. 4 is a schematic diagram of the DIR area when write control is performed, and Fig. 5 is a schematic diagram of the DIR area when write control is performed. 6 is a schematic diagram showing a method for updating the FAT area on a write-once disk obtained by the method of the present invention, FIG. 7 is a schematic diagram showing a method for updating the DIR area, and FIG. 8 is a schematic diagram showing a method for updating the DIR area. FIG. 9 is a block diagram showing pointer management during reading and writing; FIG. 10 is a schematic diagram showing the internal structure of a conventional block device; FIG. 11 is a diagram showing the internal structure of a conventional disk. It is a schematic diagram. 10... FAT area, 12... File area,

Claims (1)

[Claims] In order to use write-once media and have a general-purpose operating system manage disk data in the same way as reversible media, the entire area of the FAT and directory is updated sector by sector, and the final information of each constituent sector is added. A disk data management method characterized by referencing the entire area.