JPH117730A - Memory apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a capacity of defect management information and shorten a defect search time and an update time, by compressing defective sector information. SOLUTION: A defective cylinder information management part 17 generates and manages a one-bit defective cylinder information 21 indicating the presence/ absence of a defective sector for every cylinder. A pointer information management part 18 forms the one-bit defective cylinder information 21 in groups for every plurality of cylinders, generates and manages a pointer information 22 indicating a start address of a defective sector information. Further, a defective sector information management part 19 generates and manages a defective sector information 23 corresponding to the start address designated by the pointer information 22, e.g. for every four bytes irrespective of kinds of an alternate sector and a split sector. An alternate process part 20 searches the defective cylinder information 21, pointer information 22 and defective sector information 23, thereby carrying out an alternate process, in response to an access requirement from a host apparatus.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for reading defect information for managing a position address of a defective sector on a medium and a replacement sector address of the defective sector from the medium when the power of the apparatus is turned on and storing the defect information in a memory area on the apparatus. The present invention relates to a storage device such as a magnetic disk device that expands and accesses a medium.

[0002]

2. Description of the Related Art Conventionally, defect address information for managing the position address information of a defective sector and the replacement destination sector address of a defective sector is, as shown in FIG.
It is managed by a byte defective sector position information start address. As the start address of this defective sector position information,
When a cylinder has a defective sector, the start address (logical block address LBA) in which the defective sector position of the cylinder is registered is stored. An initial value such as “FFFFFFFFh” is defined for a cylinder having no defective sector.

[0003] Defective sector information managed by the defective sector position information start address of the cylinder defect information shown in FIG. 8A is composed of information in units of 2 bytes or 4 bytes and registered and managed. That is, the defective sector information is information in units of 2 bytes as shown in FIG. 8B for a split sector, and shown in FIG. 8C for an alternate sector.
Is information in units of 4 bytes.

[0004] The defective sector information of the split sector shown in FIG. 8 (B) is set in the upper 4 bits of the first byte 0, with the alternate sector indicating bit ALT, spare sector indicating bit SP, split sector indicating bit SL, and last cylinder indicating bit. DL is assigned, and the lower 4 bits are used as the defective head number. The next byte 1 is assigned a 1-byte long defective sector number.

In the defective sector information of the replacement sector shown in FIG. 8C, the first bytes 0 and 1 are the same as those of the split sector. The display bit SS indicates spare replacement of the same cylinder when set, and spare replacement for a replacement cylinder when cleared. Subsequently, a replacement cylinder number and a replacement logical head number are assigned in 4-bit units. The last byte 3 is a 1-byte long replacement logical sector number.

[0006]

However, in such conventional defect information management, the larger the capacity of the device, the larger the amount of cylinder defect information in FIG. 8A becomes (the number of cylinders). × (2 bytes). Also, in an apparatus having more than 256 sectors per track in a zone where the disk transfer is fast, when the defective sector information is a slip sector, the number of bytes is increased by one byte as shown in FIG. 9
As shown in (B), a total of 6 bytes are required, which is increased by 2 bytes.

[0007] In response to such an increase in apparatus capacity, a high-end apparatus (high-priced apparatus) can sufficiently secure a work buffer memory, so that the defect information management shown in FIG. However, a low-end device (a low-cost device) having a limited work buffer memory cannot continue to use the defect information management of FIG. For this reason, the defective sector position start address is stopped for each cylinder as shown in FIG. 8A, and FIGS. 9A, 9B, and 9C.
Of the defect management information is used to reduce the capacity of the defect management information.

That is, the defect zone information shown in FIG.
The start address (logical block address LBA) of the address pointer is registered and managed for each zone and head number in byte units. The zone pointer information in FIG. Register and manage addresses. The defective sector information shown in FIG. 9C is 6-byte information. Byte 0 is a defective sector number, byte 1 is a split sector display bit SL, a replacement sector display bit ALT, a 3-bit replacement logical head number,
2-bit replacement cylinder number, bytes 2 and 3 are 2
Byte 4 is a defective cylinder number, byte 4 is a 1-byte replacement logical sector number, and byte 5 is reserved for devices having 256 or more sectors.

However, in the zone management of the defect information shown in FIG. 9, one zone has several hundred or more cylinders, and the search time of the defect information when an access request is received from a host device is as shown in FIG. Compared with the defect management information of FIG. 8, an extra 2-byte defective cylinder number is loaded from the defective sector information of FIG.
This has to be compared with the cylinder number, so that there is an overhead. Further, when updating the defect management table, since the defective sector information must be sorted in ascending order, the sorting time also becomes an overhead.

Further, in an apparatus having 256 sectors or more per track in a zone having a high data transfer speed, FIG.
Reserved byte 5 of the defective sector information (C)
The upper two bits of the defective sector number and the replacement logical sector number must be secured using ed), and it is expected that the overhead in the defect search will be further added.

The present invention guarantees a defect search time equivalent to a high-end device for a large-capacity low-end device to be developed in the future and compresses defective sector information, thereby reducing the capacity of the defect management information and reducing the update time. It is an object of the present invention to provide a storage device such as a magnetic disk device that can be shortened.

[0012]

FIG. 1 is a diagram illustrating the principle of the present invention. The present invention reads out defect information for managing a position address of a defective sector on a medium and a replacement sector address of the defective sector from the medium at the time of turning on the power of the apparatus and expands the defect information in a memory area on the apparatus to access the medium. It is intended for a storage device such as a magnetic disk device.

According to the present invention, in such a storage device, the defective cylinder information management unit 17 generates and manages 1-bit defective cylinder information 21 indicating the presence or absence of a defective sector for each cylinder. The pointer information management unit 18 groups the one-bit defective cylinder information 21 in units of a plurality of cylinders and generates and manages pointer information 22 indicating a start address of the defective sector information.

Further, the defective sector information management section 19 generates defective sector information 23 corresponding to the start address specified by the pointer information 22 as information of a predetermined byte unit regardless of the types of the replacement sector and the split sector. to manage. Then, in response to an access request from the higher-level device, the replacement processing unit 20 searches the defective cylinder information 21, the pointer information 22, and the defective sector information 23 to perform the replacement processing.

The defective sector information management section 19 generates and manages the defective sector information 23 as information of a 4-byte unit regardless of the types of the replacement sector and the split sector. For this reason, the loading of the defective sector information 23 from the memory at the time of the defect search is all fixed in units of 4 bytes, and the search time can be shortened because the byte length switching is not required in the sector search.

The defective cylinder information management unit 17 shares the start address specified by the pointer information 22 as the defective cylinder information 21 for a plurality of cylinders.
The capacity of the pointer information 22 can be reduced. The defective sector information 23 managed by the defective sector information management unit 19 includes defective cylinder information (for example, lower 4 bits of the defective cylinder information) for distinguishing a specific cylinder from a plurality of defective sector information managed by the same pointer information 22. Each of a defective head number, a replacement position number indicating a replacement destination sector address, a control bit ALT indicating a split sector or a replacement sector, and a control bit LAST indicating a last defect of a track is provided. By arranging on the bit side, acquisition of defective cylinder information and defective head information is accelerated.

According to such a storage device of the present invention, the capacity when the defect management information composed of the pointer information of the defective cylinder information and the defective sector information is compressed and expanded in the work buffer is reduced. The time required for searching and updating the defect management information can be shortened, the defect search time is guaranteed for an increase in the medium capacity of a low-end device having a limited work buffer capacity, and performance degradation due to overhead is prevented.

For a specific cylinder address area of 1 byte or less used for a bench test of a medium, etc.
The defective cylinder information management unit 17 generates a 1-byte length defective sector start address indicating the start address of the defective sector position for each cylinder and manages it as the defective cylinder information 21. It does not manage the pointer information 22 indicating the start address of the information.

Therefore, for a specific cylinder address area of 1 byte or less frequently accessed for a bench test or the like, the start address of the defective sector information 23 is immediately obtained from the defective cylinder information 21. In this case,
The defect search time is reduced at the expense of lowering the capacity.

[0020]

FIG. 2 is a block diagram of a magnetic disk drive to which the defect information management of the present invention is applied. A magnetic disk device known as a hard disk drive (HDD) includes a disk enclosure 1 and a control circuit board 2. The disk enclosure 1 has an R / W preamplifier circuit 3 mounted as a head IC, a head assembly 4 supported at the tip of a head actuator and positioned in a direction crossing a track of a magnetic disk medium, and a voice for driving the head actuator. A coil motor (hereinafter referred to as “VCM”) 5 and a spindle motor 6 for rotating a magnetic disk medium are provided.

R / W provided in disk enclosure 1
In the preamplifier circuit 3, in addition to the usual functions of head selection and read / write switching, multi-functionalization such as switching of a write current or a read sense current for an MR head, and further, mode setting such as power saving have been attempted in recent years. Thus, circuit parameters and mode settings can be controlled by register settings from the control circuit board 2 using serial transfer lines.

The head assemblies 4 are provided in a number corresponding to the number of recording surfaces of the magnetic disk medium provided in the disk enclosure 1. Each head assembly 4 uses a write head using an inductive head and, for example, an MR head. Uses a composite head assembly that integrates the read head. On the control circuit board 2 side, an MCU (microcontroller unit) 7 for controlling the entire hard disk drive, a read channel circuit 8, a hard disk controller 9, a nonvolatile flash PEROM 12, a data buffer 13 using a DRAM, a VCM 5, and a spindle motor 6
And an interface connector 14 for a host device.

The management of defect information according to the present invention is performed by MC
The defect management information, which is realized by program control by the MPU provided in the U7 and is stored in the system area of the magnetic disk medium on the disk enclosure 1 side, is stored in the built-in work buffer memory 16a when the power of the apparatus is turned on. The MCU 7 reads out and expands the data, and thereafter executes the defect processing for the access from the host device for the defect management information in the work buffer memory 16a.

FIG. 3 is a functional block diagram of the defect information management of the present invention realized by the MCU 7 of FIG. The defect information management function includes a defect information management unit 15 and a defect information storage unit 16. The defect information management unit 15 includes a defective cylinder information management unit 17, a pointer information management unit 18, a defective sector information management unit 19, and a replacement processing unit 20 for searching for defect information.

The defect information storage section 16 stores the MC of FIG.
This is information developed on the work buffer memory 16a provided in U7, and is composed of defective cylinder information 21, pointer information 22, and defective sector information 23. FIG.
4 is an explanatory diagram of a memory map in a work buffer memory 16a of the defect information storage unit 16 in FIG. In this example, the defective cylinder information 21 has 2 bytes as 1
Defective cylinder information area 21a starting from address X of work buffer memory 16a.
Are stored multiple times.

A plurality of pieces of the next pointer information 22 are stored in a pointer information area 22a starting from the address Y of the work buffer memory 16a, which is 2-byte information. Further, the defective sector information 23 is information in units of 4 bytes, and has an address Z in the work buffer memory 16a.
Are stored in the defective sector information area 23a having the first address as the start address.

FIG. 5 shows details of the defective cylinder information 21, the pointer information 22, and the defective sector information 23 stored in the defect information storage section 16 shown in FIG. FIG.
(A) is defective cylinder information 21 used in the defect information management of the present invention. One bit is assigned to one cylinder of the magnetic disk medium, and the presence or absence of a defective sector on one cylinder is managed by one bit. ing. That is, if there is at least one defective sector on one cylinder, the bit of the corresponding cylinder number is set to 1; if there is no defective sector, the bit of the corresponding cylinder number is cleared to 0. Be composed.

In FIG. 5 (A), two bytes, byte 0 and byte 1, are taken out. On the byte 0 side, bits 0 to 7 are assigned for cylinder numbers 0 to 7, and on the byte 0 side, Indicates that cylinder numbers 8 to 15 are bits 0 to 7
Have been assigned about. For example, if the number of cylinders of the magnetic disk medium is 8,192, the defective cylinder information 21 of 1 byte is obtained for 8 cylinders, so that the information amount is 1024 bytes for all 8192 cylinders.

FIG. 5B shows pointer information 22 functioning as an address pointer. This pointer information 22
Is a memory area that stores defective cylinder information 21 of each cylinder as two groups of defective cylinder information 21 of bytes 0 and 1 shown in FIG. 5A, that is, defective cylinder information 21 of 16 cylinders. , Ie, defective cylinder information, and defective information start address of byte 0/1.

As a result, the pointer information 22 as an address pointer functions as a shared address pointer for specifying defective sector information for 16 cylinders specified in units of 2 bytes in FIG. The pointer information 22 in FIG. 5B stores the defect information start address of bytes 0 and 1 corresponding to the defective byte information 21 of 2 bytes (16 cylinders) of bytes 0 and 1 in FIG. 5A. Since the processing is performed in units of 4 bytes, the defect information start address corresponding to the defective cylinder information of the next bytes 2 and 3 is also shown.

More specifically, when the total number of cylinders is 8192, the information amount of the pointer information 22 is 1024 bytes for the defective cylinder information 21 in FIG. Since the pointer information 22 also gives information in units of 2 bytes, the pointer information 22 also has an information amount of 1024 bytes. FIG. 5C shows the defective sector information 23. The defective sector information 23 is stored as information in units of 4 bytes in this embodiment, with the address Z of the work buffer memory 16a as a start address, as in the defective sector information area 23a in FIG.

The start address of the defective sector information 23a is
The defect sector information 23 of up to 16 cylinders is stored in order starting from the defect information start address of the defect cylinder information byte 0/1 specified by the 2-byte information of the pointer information 22 in FIG. 5B. . In the present invention, regarding the defective sector information 23 shown in FIG. 5C, regardless of whether the defective sector information is a split sector or a replacement sector, all four bytes of bytes 0, 1, 2, and 3 are used. It consists of information. For this reason, the load and store at the time of registration and search for the defective sector information 23 by the defect information management 15 in FIG. 3 can all be processed with a fixed byte length of 4 bytes, and the bytes in the split sector and the alternate sector can be processed. Since it is not necessary to divide the length, the registration, update and search of the defective sector information 23 can be speeded up accordingly.

The structure of the defective sector information 23 is as follows. First, the lower 4 bits of the replacement position number are assigned to the upper 4 bits of byte 0. The lower 4 bits of the defective cylinder information are assigned to the lower 4 bits of the next byte 0. Here, between the defect information start address specified by the pointer information 22 and the next defect information start address, the defective sector information 23 for 16 cylinders specified by the 2-byte information in FIG. In order to specify a specific defective cylinder in the defective sector information of up to 16 cylinders, the lower 4 bits of the defective cylinder information are stored.

That is, when it is determined that the 8-bit access cylinder information from the higher-level device is a defective cylinder, the defective cylinder information 21 and the pointer information 22 shown in FIGS. Since the shared defect information start address is specified by the upper 4 bits, specific defective cylinder information in the start address can be distinguished by using the lower 4 bits of the access cylinder information.

Byte 1 of the defective sector information 23 is assigned the middle 4 bits of the replacement position number using the upper 4 bits (however, the upper 4 bits in a zone exceeding 256 sectors per track). A defective head number is assigned to the lower 4 bits. Here, the defective cylinder information and the defective head number are assigned to the lower 4 bits for bytes 0 and 1 of the defective sector information 23, whereby the defective cylinder information and the defective cylinder information at the time of loading the defective sector information 23 and searching for the defective sector are obtained. The acquisition of the defective head number can be processed at high speed.

The lower 8 bits of the defective sector number are assigned to 8 bits of byte 2 of the defective sector information 23. To the upper 4 bits of byte 3 of the defective sector information 23, the lower 4 bits of bytes 0 and 1 and the upper 4 bits corresponding to the middle 4 bits are assigned an alternate position number.
Further, a control bit LAST and a control bit ALT are assigned to the third and second bits on the byte 3 side.

The control bit LAST indicates that setting this bit indicates that the defective sector information 23 is the last defect of the track. Control bit AL
T indicates whether the defective sector information 23 is a split sector or a replacement sector. That is, when the control bit ALT is set to 1, it indicates that the sector is a replacement sector, and when it is cleared to 0, it indicates that it is a split sector.

The lower 2 bits of the third byte in the defective sector information 23 are assigned to store the upper 2 bits of the defective sector number with respect to the defective sector number specified by the 8 bits of the second byte. That is, in a zone where the number of sectors per track is 256 or less, 8
If the number of sectors per track exceeds 256 sectors and has a 10-bit configuration of, for example, 1024 sectors, the area of the defective sector number specified by the lower 2 bits of the third byte is sufficient. Use

When the control bit ALT is set to 1 in the defective sector information 23, bytes 0, 1 and 3
Becomes effective, but conversely, the control bit AL
In the case of a split sector in which T is cleared to 0, since the replacement position number is not used, bytes 0, 1, or 3
Are used as control bit A
It can also be used for LT.

FIG. 6 is a flowchart of a process of searching for defect management information when access information is received from a higher-level device by the replacement processing unit 20 of FIG. First, an access cylinder number CC, an access head number HH, an access start sector number SS, and an access end sector number SE are provided with a read or write access from a higher-level device.

Upon receiving such access information, the replacement processing unit 20 first refers to the defective cylinder information 21. That is, in step S1, the access cylinder number CC is set in FIG.
The quotient A and the remainder B are obtained by dividing by 8 which indicates the number of cylinders that can be stored in one byte length of the defective cylinder information 21 in FIG. Subsequently, in step S2, for example, the start address X of the defective cylinder information area 21a as shown in FIG.
The 2-byte word information about the address to which the quotient A obtained in 1 is added is loaded as defective cylinder information C. Regarding the defective cylinder information C loaded in this way,
In step S3, it is checked whether the bit B corresponding to the remainder B obtained in step S1 is "1".

If the bit B is 1, since there is a defective sector in the sector of the access cylinder number CC, the process proceeds to step S4, and the process proceeds to step S1 in the start address Y of the pointer information area 22a in FIG. The quotient A found in
Is loaded with 2-byte word information at an address obtained by adding half (A / 2) of the word. That is, 16 cylinders including one bit of the defective cylinder information of the pointer information 22 of FIG.
Load the defect information start address of (i + 1). Next, the defective sector information of the defect information start address, which is the pointer information loaded in step S4, is loaded in 4 bytes in step S5. In the next step S6, it is checked whether or not the defective cylinder information in the 4-byte defective sector information loaded in the lower 4 bits of the defective cylinder information specified by the remainder B obtained in step S1 matches.

When the remainder B matches the defective cylinder information (lower 4 bits) in the defective sector information, the flow advances to step S7 to check whether the defective head number matches the access head number. If the head number matches the access head number, it is checked in step S8 whether the defective sector number in the defective sector information exceeds the access start sector SS.

If the defective sector number is the access start sector SS
If not, it is checked in step S9 whether the defective sector number is smaller than the access last sector number SE. If the defective sector number is smaller than the access last sector number SE, this defective sector information falls within the sector range to be accessed, so that step S1 is executed.
At 0, the control bit ALT in the defective sector information is checked.

If the ALT bit is 1, this is a replacement sector, and the process proceeds to step S11, where the replacement position number obtained from the defective sector information is divided by a known number of sectors per track to obtain a quotient T and a quotient T. Find the remainder E. At this time, the quotient T becomes the head number of the replacement sector position, and the remainder E becomes the sector number of the replacement position. In step S12, the replacement sector is accessed. If the ALT bit is 0 in step S10, since this is a split sector, processing is performed on the split sector based on the defective sector information.

On the other hand, if the defective cylinder number of the loaded defective sector information does not match the remainder B in step S6, if the defective head number does not match the access head number in step S7, or if the defective sector number starts accessing in step S8. If smaller than sector number SS,
Proceeding to step S13, the control bit L of the defective sector information
The AST is checked, and if the LAST bit is not 1, the process returns to step S5, loads the next defective sector information by 4 bytes, and repeats the processing from step S6.

Thus, the defective sector information for every 16 cylinders in which the head address of the pointer information is specified is read out in order, and the replacement process is performed by obtaining the head number and the sector number of the replacement sector included in the access information. become. In addition, during the search, LAST in step S13.
If the bit is 1, since there is no defect in the cylinder, the process proceeds to step S14,
The search is terminated assuming that no defect exists in the cylinder.

Of course, if the bit B is not 1 for the load content of the defective cylinder information in step S3, there is no defect in that cylinder, so in step S14 the search is terminated with no defect, and the access is terminated. Perform access by the information itself. Further, in step S11, the quotient T obtained by dividing the replacement position number by the number of sectors per track for a device having one replacement cylinder is the head number, the remainder E
Is used as a sector number. In a device having two or more replacement cylinders, the quotient T is further divided by the number of devices (D /
The quotient obtained by dividing the number of devices by the maximum number of heads is the cylinder number, and the remainder is the head number.

FIG. 7 shows another embodiment of the defect information management according to the present invention. In this embodiment, for example, a defective cylinder having a cylinder number CC = 000 to 256, which is allocated as a cylinder area for bench test of an optical disk medium. As for the information, whether the defective sector exists in one cylinder as shown in FIG.
The one-byte defective sector start address is directly managed.

That is, the work buffer memory 16a shown in FIG.
In the defective cylinder information 25 of the cylinder for the bench test developed in the above, the cylinder number CC = 00 shown on the left side
For 256 cylinders 0 to 255, if a defective sector exists in the cylinder, the 1-byte defective sector start address indicating the start position on the work buffer memory 16a of the defective sector information 23 in FIG. Stored.

As a result, the defective cylinder information 25 of the bench test cylinder does not need the pointer information 22 as shown in FIG. The start address of the defective sector information 23 can be obtained, and the corresponding defective sector information can be searched. Therefore, the search for the defect information can be speeded up for the 256 cylinder area of the cylinder number CC = 000 to 255 frequently accessed for the bench test because the pointer information 22 is not provided.

This bench test cylinder defect information 25
For cylinder numbers other than 256 other than those shown in FIG.
(A) Stores defective cylinder information 21 in which 1 bit indicating the presence or absence of a defective sector is assigned to each cylinder. In the above embodiment, the defective cylinder information 21 in FIG. 5A is used as the pointer information 22 in FIG.
16 are grouped and the defect information start address of the defective cylinder information byte (i / i + 1) is registered as an address pointer. As another embodiment, FIG.
The pointer information 22 shown in FIG. 5B is grouped into eight cylinders (one byte) of the defective cylinder information 21 shown in FIG. The start address may be used.

Of course, also in this case, the defective cylinder information of the defective sector information 23 contains the lower 4
A bit is stored to distinguish one of the eight cylinders for a particular cylinder. The present invention is not limited by the numerical values of the above embodiments.

[0054]

As described above, according to the present invention, defective cylinder information is managed by bit information for each cylinder, and the defective cylinder information is used as an address pointer for every 16 cylinders. Address, and the defective sector information managed by the defective sector start address is all 4-byte defective sector information regardless of the type of the replacement sector or the split sector, so that the total amount of defect management information can be obtained. Can be reduced, and the processing time of replacement processing by searching for defect management information for table registration, updating, and access requests can be reduced by reducing the amount of defect management information when expanded in the work buffer memory on the device. Limited to work buffer area for deploying management information Even a low-end device,
A defect search time equivalent to a high-end device can be guaranteed.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a block diagram of a magnetic disk drive to which the present invention is applied;

FIG. 3 is a block diagram of a defect management function according to the present invention;

FIG. 4 is an explanatory diagram of a memory map in a work buffer development state of defect management information according to the present invention.

FIG. 5 is an explanatory diagram of defective cylinder information, pointer information, and defective sector information used in defect management according to the present invention.

FIG. 6 is a flowchart of a search process for defect management information according to the present invention.

FIG. 7 is an explanatory diagram of a memory map of defective cylinder information used in another embodiment of the present invention.

FIG. 8 is an explanatory diagram of conventional defect management information.

FIG. 9 is an explanatory diagram of conventional defective sector information corresponding to an increase in the number of medium sectors.

FIG. 10 is an explanatory diagram of conventional defect management information managed in zones.

[Explanation of symbols]

 1: Disk enclosure 2: Control circuit board 3: R / W preamplifier circuit 4: Head assembly 5: Voice coil motor (VCM) 6: Spindle motor 7: MCU 8: Read channel circuit 9: Hard disk controller 10: Servo controller 11a , 11b: FPC 12: Flash PEROM 13: Data buffer 14: Interface connector 15: Defect information management unit 16: Defect information storage unit 16a: Work buffer memory 17: Defective cylinder information management unit 18: Pointer information management unit 19: Defective sector Information management unit 20: replacement processing unit 21: defective cylinder information 22: pointer information 23: defective sector information 25: defective cylinder information for bench test

Claims (4)

[Claims] 1. Defect information for managing a position address of a defective sector on a medium and a replacement sector address of the defective sector is read out from the medium when the apparatus is turned on, and is expanded in a memory area on the apparatus to be developed. A defective cylinder information management unit that generates and manages 1-bit defective cylinder information indicating presence / absence of a defective sector for each cylinder in a storage device to be accessed; and groups the 1-bit defective cylinder information into a plurality of cylinder units. A pointer information management unit for generating and managing pointer information indicating a start address of defective sector information; and a start address specified by the pointer information as information in a predetermined byte unit regardless of the type of the alternate sector and the split sector. A defective sector information management unit for generating and managing defective sector information in response to the access to a higher-level device; The defective cylinder information in response to the
A storage device, comprising: a replacement processing unit that searches for pointer information and defective sector information and performs replacement processing. 2. The storage device according to claim 1, wherein the defective cylinder information management unit generates and manages the defective sector information as information of a 4-byte unit regardless of a type of a replacement sector and a type of a split sector. Characteristic storage device. 3. The storage device according to claim 1, wherein said defective cylinder information management unit includes:
Defective cylinder information for discriminating a specific cylinder among a plurality of cylinders sharing the start address specified by the pointer information, a defective head number, a replacement position number indicating a replacement sector address, and indicating whether the sector is a split sector or a replacement sector. A storage device comprising a control bit and a control bit indicating a last defect of a track, and wherein the defective cylinder information and the defective head information are arranged on a lower bit side. 4. The storage device according to claim 1, wherein, for a cylinder address area of 1 byte or less used for a bench test or the like of the medium, the defective cylinder information management unit is configured to have a 1 byte length for each cylinder. A storage device for generating and managing the defective sector information start address, and not managing the pointer information indicating the defective sector information start address by the pointer information management unit.