JPH04162127A - Magnetic disk storage device - Google Patents

Abstract

PURPOSE:To decrease the number of defective tracks and to reduce the capacity of a substitute memory by assigning only the defective sectors to the substitute sectors when the number of defective sectors of a single track is smaller than a prescribed number. CONSTITUTION:The contents of the substitute sector and track are loaded in a memory 8 prior to an access given to a disk device end right after a power ON state or after the reception of a command given from 8 host controller. When the data are written into the disk device, a data writing command is given to a drive access control circuit 2 from the host controller via an interface circuit 1. Then the circuit 2 analyzes the writing command end checks whether the writing destination includes a substitute sector or not. If not, the data are successively written into the prescribed sectors of the disk 15. If so, the data are written into the memory 8 instead of the disk 15.

Description

[Detailed description of the invention]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic disk storage device that speeds up access to alternative sectors and tracks provided for the purpose of relieving bad sectors and bad tracks. [Prior Art] In a magnetic disk device that divides a magnetic disk into concentric tracks and further divides the tracks into sectors of equal size for management, bad sectors and bad tracks caused by defects in the medium are replaced with alternative sectors. We are providing relief by providing alternative trucks. Therefore, each time an alternative sector or track is accessed, a new seek or rotation wait occurs as shown in FIG. 5(C), resulting in a problem in that the read/write speed to the magnetic disk decreases. This problem becomes even more serious when one file is divided and stored on a large number of magnetic disks, since the end of file access is determined by the end of the magnetic disk that is the slowest to complete access. In order to alleviate this, a spare sector is provided for each track as shown in FIG. 2(a), and if a defective sector exists, a spare sector is provided as shown in FIG. 2(b). By using spare sectors, bad sectors are skipped and sectors are allocated. In this case, a new seek does not occur, and the rotation wait is at most the number of bad sectors that exist on the trunk, so
There is no significant decrease in read/write speed. However, if there are more bad sectors than the number of spare sectors, it is no longer possible to skip all bad sectors and allocate sectors; such tracks are determined to be bad and a replacement track is allocated. , In this case, a new seek and rotation wait will occur. As a method for avoiding new seeks and rotational waits that occur when accessing alternative sectors and tracks as described above, Japanese Patent Laid-Open No. 1-166378 discloses a high-speed memory that stores copies of alternative sectors and tracks, For example, by providing a semiconductor memory, storing a copy of the contents of an alternative sector or track in advance in the memory, and accessing the memory instead of accessing the alternative sector or track, a new seek or rotation wait can be performed. We are trying to prevent this from occurring.

[Problem to be solved by the invention]

In order to obtain sufficient effects from the method disclosed in Japanese Patent Application Laid-Open No. 1-166378, it is necessary to have a copy storage memory capable of storing the entire capacity of the alternative sector and track. If the memory capacity is smaller than the total capacity of the alternate sector and alternate track, as stated in Japanese Patent Laid-Open No. 1-166378, the memory capacity is determined in advance before a series of accesses. This is because it is necessary to check the alternate sectors and tracks that need to be accessed, load them into memory, and then return the contents of the memory to the alternate sectors and tracks after a series of accesses has been completed, which results in overhead. . Assuming that the total capacity of alternative sectors and tracks is about 0.3% of the magnetic disk device, it will be 3 MB for a magnetic disk with a capacity of IGB, and if the copy memory is implemented with a semiconductor memory, a large capacity memory will be required. On the other hand, the alternate sector and alternate track must have a capacity sufficient to accommodate the maximum allowable medium defect. For this reason, in actual magnetic disks, the ratio of alternate sectors and tracks used is often less than half, and even if copy memory is implemented to accommodate the full capacity of alternate sectors and tracks, most of the alternate sectors and tracks are not used. There are many cases where it is useless. An object of the present invention is to minimize the capacity of the copy memory of one alternative sector or track, and to eliminate portions that are not actually used and are wasted. [Means for Solving the Problems] In order to achieve the above object, in the present invention, even if there are more bad sectors on one track than the number of spare sectors provided on that track, the number of bad sectors can be determined in advance. If the number is less than the set number K (K≧number of spare sectors per track), it is not considered a bad track, and as shown in Figure 2 (c) and (d), the bad sectors that exist in excess of the number of spare sectors are replaced. Allocate to an alternative sector on the track. When the number of defective sectors exceeds K, the track is regarded as a defective track and an alternative track is allocated to it, and a memory is provided to store copies of the contents of the alternative track consisting of a set of alternative sectors and the alternative track of the defective trunk. Loading data from the alternate sector or track to the memory is performed immediately after power-on or according to instructions from a higher-level controller, and data is stored from the memory to the alternate sector or track immediately before power-off or when the higher-level controller This is done by commands from the control device. Furthermore, in a magnetic disk device having a cache memory, the memory and the cache memory are allocated to successive address spaces, and the boundary between the memory and the cache memory is made variable. [Operation] As explained above, even if there are more bad sectors on one track than the number of spare sectors provided on that track, the number of bad sectors will be reduced to the preset number K (
In the following cases (K≧number of spare sectors per track), this is not treated as a bad track, and only bad sectors that cannot be repaired by spare sectors are allocated to alternative sectors on alternative tracks, so the number of bad tracks can be reduced. I can do it. As a result, fewer alternate sectors and tracks are required, and less memory is required to store copies of the alternate sectors and tracks. In addition, in a magnetic disk storage device with a built-in cache memory, by allocating a continuous address space to the cache memory and the memory for storing copies of alternative sectors and tracks, and making the boundary between the cache memory and the memory variable, 1 replacement sector with fewer media defects,
When the capacity of the alternative track is small, the capacity of the cache memory can be increased. Conversely, if there are many media defects, reduce the cache memory capacity and use one alternative sector. Such devices can be salvaged by increasing the memory capacity for storing copies of alternate tracks, yet read/write speeds are hardly reduced even with many media defects. Embodiment FIG. 1 is a diagram showing the configuration of a magnetic disk device according to a first embodiment of the present invention. In FIG. 1, 15 is a disk on which data is recorded;
16 is a magnetic head for reading/writing data on the disk 15; 14 is an actuator for positioning the head 16 in a desired cylinder; 10 is a read/write amplifier, VF◯. Read/write control circuit consisting of data discrimination circuit, etc.
5 is a selector for selecting data to be written to the disk, 6
1 is a selector for selecting data to be read to the interface circuit; 8 is a memory for storing a copy of an alternative sector and an alternative track; 9 is a battery backup circuit for backing up the memory 8 when the power is cut off; 1 is a host controller; 2 is a drive control unit which has a microprocessor as its main component and controls the magnetic disk device according to instructions from a host controller; 3 is an access request detection circuit that detects the timing to access the memory 8; 4 is a memory control circuit for issuing addresses and control pulses to the memory 8, and its operation will be explained below with reference to the figure. In FIG. 1, the track format of the disk 15 includes at least one spare sector for repairing a defective sector, as shown in FIG. 2(a). If there are bad tracks on the track with less than the number of spare sectors per track, the spare sectors are used and the bad sectors are skipped and sectors are allocated as shown in Figure 2(b). rescue. However, if there are more bad sectors on the track than the number of spare sectors per track, if the number of bad sectors is less than or equal to a predetermined number K (K>number of spare sectors per track), As shown in Figures 2 (c) and (d), bad sectors that exist in excess of the number of spare sectors per track are allocated to alternative sectors on alternative tracks, and the track is not designated as a bad track, but the number of bad sectors is When the track exceeds K, the track is considered a defective track and a replacement track is allocated. The functions described above are realized as functions of the drive access control circuit 2. Memory 8 in FIG. 1 is an alternative trunk provided on disk 15. This is a high-speed memory that stores a copy of an alternative sector, and is made up of, for example, a semiconductor memory.As shown in Figure 3, the contents of an alternative track or sector are loaded immediately after the disk device is powered on or by a command from a host controller. Then, the contents of the memory 8 are written back to the disk 15 immediately before power-off or by a command from a higher-level control device. These operations are executed under the control of the drive access control circuit 2. Next, the specific operation of the apparatus shown in FIG. 1 will be explained. First, prior to accessing the disk device, the contents of the alternative sector and track of the disk 15 are loaded into the memory 8 immediately after power-on or by a command from a higher-level control device. When writing data to a disk device, in FIG. 1, when a data write command is given from the host control device to the drive access control circuit 2 via the interface circuit 1, the drive access control circuit 2 analyzes the write command, Check whether the write destination includes an alternative sector. If the write destination does not include an alternative sector, the drive access control circuit 2 selects the write data 102 given from the host controller via the interface circuit 1 as the output 104 of the selector 5, and transfers the data via the read/write control circuit 1o. and sequentially write to predetermined sectors of the disk 15. On the other hand, if the write destination includes an alternative sector, the drive access control circuit 2 controls the access request detection circuit 3.
sets the defective sector number to be assigned to the alternative sector and the starting address of the alternative sector in the memory 8;
When writing to an alternative sector, data is written to the memory 8 instead of being written to the disk 15. FIG. 4 is a diagram showing an example of the circuit configuration of the access request detection circuit 3 in FIG. 1. In FIG. 4, 20 is a storage means for storing one or more defective sector numbers to be assigned to an alternative sector, 22 is a storage means for the start address in the memory 8 of a copy of the alternative sector, and 21 is the sector number in FIG. A comparator for comparing the sector number 116 detected by the detection circuit 11 with the sector number stored in the defective sector number storage means 20 and detecting the timing at which a copy of the alternative sector stored in the memory 8 should be accessed. Yes, 20.22 is set in advance by the drive access control circuit 2. That is, the access request detection circuit 3 detects the timing of writing to the defective sector allocated to the alternative sector, and starts the memory control circuit 4 by giving it the starting address 113 of the copy of the alternative sector in the memory 8. The drive access control circuit 2 selects the write data 102 from the host control device as the output 107 of the selector 7 based on the alternative sector access detection signal 112 of the access request detection circuit 3, and sets it as the write data to the memory 8. When writing to , the data is written to the memory 8 instead of being written to the disk 15 . In the data read operation, data is read from the memory 8 instead of from the disk 15 when reading data from the defective sector allocated to the alternative sector. That is, when reading data from the disk 15, the drive access control circuit 2 selects the read data 103 from the memory 8 as the output 106 of the selector 6. At this time, the operation of the access request detection circuit 3 is the same as when writing data. Note that by writing to the memory 8, the alternative sectors and tracks on the disk 15 no longer match the copies of the alternative sectors and tracks in the memory 8. If the contents of the memory 8 are lost due to a power cut in this state, the data cannot be restored, so the power supply of the memory 8 must be backed up. 9 is a battery bank up control circuit for this purpose. FIG. 5 is an explanatory diagram of the operation of the apparatus shown in FIG. 1. FIGS. 5(a) and 5(b) are timing charts when accessing copies of an alternative sector and an alternative track on the memory 8, respectively, and show that there is no delay even when accessing an alternative sector and an alternative track. There is. Figure 5 (
c) is a case where the memory 8 is not provided, and shows that a significant delay occurs in access because an alternative sector and an alternative track on the disk 15 are accessed. The above describes a disk device in which the drive access control circuit 2 collectively manages access to alternative sectors when accessing the disk device. However, if there are a large number of alternative sectors, if they are collectively managed by the drive access control circuit 2, the load on the microprocessor built into the drive access control circuit 2 becomes heavy, and the processing time of the microprocessor that accounts for the access time cannot be ignored. occurs. Therefore, in the device shown in FIG. 1, information is recorded in the ID field of the defective sector for allocation to an alternative sector, and information is detected by the sector number detection circuit 11 for allocation of the alternative sector in the sector ID field. If the request detection circuit 3 performs hardware processing, this problem can be solved. FIG. 6 is a diagram showing an example of a sector ID format when the sector ID field has information on alternative sector allocation. In FIG. 6, 302 is a flag indicating that the sector is a defective sector and an alternative sector has been allocated, and 303 is an alternative sector number when an alternative sector is allocated. FIG. 7 is a diagram showing an example of the configuration of the access request detection circuit 3 in this case. In FIG. 7, the flag detection circuit 26 checks the flag 302 from the sector number detection circuit 11,
When it is detected that the sector is a bad sector and has been allocated to an alternative sector, the alternative sector number 30 is set.
3, the start address 304 of the alternative sector in the memory 8 is obtained from the start address table 27, and is supplied to the memory control circuit 4 and activated. That is, since access to the copy of the alternative sector in the memory 8 is controlled by hardware, the drive access control circuit 2
This eliminates the need to check whether the access destination includes an alternative sector each time the disk device is accessed from the host controller, and even when there are a large number of alternative sectors, the microprocessor built in the drive access control unit 2 Since the processing time of the access point does not increase, the access time does not increase. FIG. 8 is a diagram showing the configuration of a magnetic disk device according to a second embodiment of the present invention.6 In this embodiment, in a magnetic disk device incorporating a cache memory, copies of the cache memory, one alternative sector, and an alternative track are provided. It is an object of the present invention to use memory without waste by allocating a storage memory to a continuous address space and making the boundary between a cache memory and the memory variable. The structure and operation of the apparatus shown in FIG. 8 will be explained below in comparison with the first embodiment shown in FIG. In FIG. 8, 34 is a memory for storing a cache memory, an alternative sector, and an alternative track;
As shown in the figure, the drive access control circuit 30 is allocated to a continuous space, and the boundaries between the cache area, alternative sectors, and alternative track areas are managed by a drive access control circuit 30 whose main component is a microprocessor. In other words, when there are few media defects, the capacity of the required alternative sectors and tracks may be small, so a large cache area can be secured, and the opposite is true when there are many media defects. 31 is memory 3
A cache memory control circuit 3.4 for controlling the operation in the cache area 4 is the same as 3.4 in FIG. be done. In this embodiment, data is read from the alternative sector and track area to the cache area, and conversely, data is written from the cache area to the alternative sector and track area. This is done by controlling the selector 32. Also, as in the first embodiment, it is necessary to back up the power supply of the memory 34 by the battery backup control circuit 9. Effects of the Invention 1 As explained above, according to the present invention, even if there are bad sectors on a track exceeding the number of spare sectors provided on each track, the number of bad sectors existing in excess of the number of spare sectors can be fixed. is assigned to an alternative sector, and if the number of bad sectors on the track is less than a preset number, it is not considered a bad track, so the capacity of the alternative sector and track is small, so it is used for storing copies of the alternative sector and track. This has the effect of requiring less memory capacity. In addition, in disk devices with built-in cache memory,
By allocating the cache memory, alternative sectors, and memories for storing copies of alternative tracks in continuous address spaces and making the boundaries variable, the memory can be used efficiently regardless of the number of media defects.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of a magnetic disk device according to the first embodiment of the present invention, FIG. 2 is an explanatory diagram showing the data storage method on the disk in FIG. 1, and FIG. FIG. 4 is a block diagram showing an example of the configuration of the access request detection circuit in FIG. 1. FIG. 5 is an explanatory diagram of the effect of the device shown in FIG. 1. FIG. 7 is a block diagram showing another configuration example of the access request detection circuit in FIG. 1;
FIG. 8 is a block diagram showing the configuration of a magnetic disk device according to a second embodiment of the present invention, and FIG. 9 is an explanatory diagram showing a data storage method in the memory in FIG. 8. Explanation of symbols 1...Interface circuit, 2...Drive access control, part A, 3...Access request detection circuit, 4...
・Memory control unit, path 8...Memory, 9...Battery bank up control circuit, 10...Read/write control circuit. DESCRIPTION OF SYMBOLS 11... Sector number detection, circuit, 15... Disk, 20... Defective sector number storage means, stage, 21... Comparator, 22... Memory start address, storage means, 2
6...Flag detection circuit, 27...Memory start address table, 30...Drive, access control unit B, ¥
J2 eyes (/, 77 tollers adhering to the sled are also 105) 4th cause VJ5 figure S7th 7th mouth q figure

Claims (1)

[Claims] 1. A disk-shaped magnetic recording medium is divided into concentric tracks, and the tracks are divided into a plurality of sectors of equal size, and each track is provided with relief for bad sectors due to media defects. At least one spare sector is provided for the recording medium, and in addition to the spare sector, a bad track due to a media defect, a substitute track for relieving the bad sector, a substitute sector, and a copy of the substitute track and the substitute sector are stored. In a magnetic disk device having a semiconductor memory for use in a magnetic disk drive, if the number of defective sectors in the track is less than or equal to a predetermined number K (K≧the number of spare sectors per track), the track is not considered a bad track and is used as a spare sector. 1. A magnetic disk storage device comprising: control means which allocates only sectors that cannot be salvaged to alternative sectors, and when the number of defective sectors in the track exceeds K, allocates the track as a defective track to the alternative track. 2. The magnetic disk storage device according to claim 1, further comprising a cache memory for write data or read data in addition to the semiconductor memory for storing copies of the alternate tracks and alternate sectors. What is claimed is: 1. A magnetic disk storage device, characterized in that: and are allocated to a continuous address space. 3. The magnetic disk storage device according to claim 2, wherein the boundary between the address spaces of the semiconductor memory and the cache memory is variable. 4. The magnetic disk storage device according to claim 1, 2 or 3, further comprising means for battery-backing up the semiconductor memory for storing copies of the alternative sectors and tracks.